Capitulo 4

A Training and Dissemination Strategy for HDM-4
Version 1.0
by C C Parkman, J Lowe, A B Sterling, J Rolt and H R Kerali
Unpublished Project Report

PR/INT/653/99
Customer Reference: RETA 5819 CONTRACT NO. COCS/99-400
TRANSPORT RESEARCH LABORATORY
PROJECT REPORT PR/INT/653/99
A TRAINING AND DISSEMINATION STRATEGY FOR HDM-4

VERSION 1.0
by C C Parkman, J Lowe, A B Sterling, J Rolt and H R Kerali
Prepared for: Asian Development Bank

Customer Reference: RETA 5819 Contract No. COCS/99-400
Copyright Transport Research Laboratory December 1999. All rights reserved.
This is an unpublished report prepared for the Asian Development Bank and must not be referred to in any
publication without the permission of the Asian Development Bank. The views expressed are those of TRL and not
necessarily those of the Asian Development Bank.
Approvals
Project Manager
Quality Reviewed
Transport Research Foundation Group of Companies

Transport Research Foundation (a company limited by guarantee) trading as Transport Research Laboratory. Registered in England, Number 3011746.
TRL Limited. Registered in England, Number 3142272. Registered Offices: Old Wokingham Road, Crowthorne, Berkshire, RG45 6AU.
This report was prepared at the Transport Research Laboratory under project reference
11103543. The customer is the Asian Development Bank (Mr C Melhuish).
Project Manager: Mr C C Parkman
Quality Audit Review Officer: Dr J Rolt
The circulation list for this report is as follows:
1-6. C Melhuish ADB

7-16. Steering Committee Via N Robertson at PIARC
17. J Lowe CRDT
18. H Kerali UoB
19. J Mat IKRAM
20. C R Bennett Independent
21. M Riley Independent
22. J Rolt TRL
23. T Toole TRL
24. G Morosiuk TRL
25. J Hine TRL
26. C Parkman TRL
The information contained herein is the property of the Transport Research Laboratory and does not reflect
the views or policies of the customer for whom the report was prepared. Whilst every effort has been made to
ensure that the matter presented in this report is relevant, accurate and up-to-date at the time of publication,
the Transport Research Laboratory cannot accept liability for any error or omission.
DRAFT
Draft Report March 1999 TRL 11.3.99

TABLE OF CONTENTS
1. INTRODUCTION.................................................................................................... 1
2. OVERVIEW AND BACKGROUND...................................................................... 1

2.1 Overview ........................................................................................................... 1
2.2 Project workshop in Kuala Lumpur .................................................................. 2
2.3 Developments since the project workshop in Kuala Lumpur ........................... 3
3. OVERALL GOAL AND PURPOSE....................................................................... 3

3.1 The strategy....................................................................................................... 3
3.2 Sustainability..................................................................................................... 4
3.2.1 Acceptance of HDM-4 .............................................................................. 4
3.2.2 Motivation ................................................................................................. 4
4. DISSEMINATION STRATEGY............................................................................. 5
4.1 Justification for dissemination .......................................................................... 5
4.2 Existing modes of dissemination....................................................................... 6
4.2.1 Newsletter.................................................................................................. 6
4.2.2 Brochure .................................................................................................... 6
4.2.3 Conference papers and presentations and published articles .................... 7
4.2.4 Website at University of Birmingham, PIARC and roadsource.com ....... 7
4.2.5 Training courses ........................................................................................ 8
4.2.6 Support from University of Birmingham and other institutions ............... 8
4.3 Options for dissemination ................................................................................. 8
4.3.1 Long term HDM-4 ‘hub’ overseeing dissemination ................................. 8
4.3.2 Regional road management working groups............................................. 9
4.3.3 Workshop sessions .................................................................................. 10
4.3.4 Targeted conference sessions .................................................................. 10
4.3.5 Published papers and articles .................................................................. 11
4.3.6 Website.................................................................................................... 11
4.3.7 Audio-visual and electronic media.......................................................... 12
4.3.8 Update information with new releases of program ................................. 12
4.3.9 Newsletter and brochure.......................................................................... 12
5. TRAINING STRATEGY....................................................................................... 13
5.1 Identified learning needs ................................................................................. 13
5.2 Access to training ............................................................................................ 14
5.2.1 Modes of training .................................................................................... 14
5.2.2 Methods of training ................................................................................. 15
5.2.3 Developing the strategy during the ADB project.................................... 15
5.3 Modular structure ............................................................................................ 16
5.4 Module content................................................................................................ 16
5.5 Assessment ...................................................................................................... 17
5.6 Training the trainers ........................................................................................ 17
5.7 Major risk / assumptions ................................................................................. 18
5.7.1 Sustainability........................................................................................... 18
5.7.2 Champions and Key Trainers.................................................................. 18
5.7.3 Demand for training ................................................................................ 19
5.7.4 Priority given to HDM-4 ......................................................................... 19
5.7.5 Modes of delivery.................................................................................... 19
5.7.6 Calibre of Trainers .................................................................................. 19
5.7.7 Road management practice versus HDM-4 capability............................ 19
6. THIS PROJECT ..................................................................................................... 19
ANNEX 1.................................................................................................................... 21
LOGICAL FRAMEWORK FOR TRAINING AND DISSEMINATION OF HDM-4 .
ANNEX 2.................................................................................................................... 22
LOGICAL FRAMEWORK FOR ADB FUNDED TRAINING AND
DISSEMINATION PROJECT
ANNEX 3.................................................................................................................... 23
TRAINING PROGRAMME DEVELOPMENT
ANNEX 4.................................................................................................................... 24
NEEDS ASSESSMENT AND PROPOSED MODULAR STRUCTURE FOR HDM-
4 TRAIN THE TRAINER PROGRAMME
HDM-4 TRAINING AND DISSEMINATION STRATEGY
1. INTRODUCTION
This draft strategy is to meet the training and dissemination needs that arise from the
introduction of HDM-4, a replacement for HDM-III. An overview and background to
HDM-4 and its technical characteristics are covered in the “Overview of the New
HDM-4 System” by H R Kerali, C J Hoban and W D O Paterson (1999) given at the
PIARC XXIst World Congress held in Kuala Lumpur.
HDM-III has been in use for over 10 years and the essential differences are that
HDM-4 extends the scope of the HDM-III program by incorporating:
• accumulated knowledge and experience developed during the lifetime of

HDM-III;
• new capabilities needed to model and analyse a wider range of conditions and
environments; and
• network level strategy and programme analyses, as well as the more detailed
project analysis of HDM-III.
HDM-4 is utilised in the “Planning”, “Programming” and “Preparation” processes

within the highway management cycle. The “Operations” element of the Highway
Management Cycle is not addressed by HDM-4. This suggests that the main target
groups for training and dissemination activities are:
• Senior management and policy level groups (Planning)

• Middle-level professionals (Programming) and
• Junior professionals (Preparation)
2. OVERVIEW AND BACKGROUND
2.1 Overview
The Terms of Reference for the project require the production of a strategy for
sustainable training and dissemination of HDM-4. It requires the strategy to be
produced in particular forms as follows:
i) identification of target audiences and characterisation of their needs;

ii) a 5 year dissemination strategy for the Asia Pacific region;
iii) a Logical Framework for the structure of HDM-4 dissemination;
iv) a Logical Framework for training; and
v) a 3 year training plan for the Asia Pacific region.
All the above items go beyond the project timeframe and, as such, aim to create
sustainable training and dissemination processes and structures.
1 HDM-4 Training and Dissemination Strategy

Version 1.0 16 December 1999
The project team designed three draft Logical Frameworks (logframes) to achieve the
above items, for discussion with the key stakeholders in HDM-4 at the PIARC World
Congress held in Kuala Lumpur (see below). These logframes were:
i) An overall training and dissemination strategy for this project in the Asia
Pacific region, which will take place over the next two years;
ii) Dissemination: a 5 year global dissemination programme; and
iii) Training: a 3 year global training programme.
This report summarises the conclusions from those discussions and presents revised
logframes based on the results of the workshop supplemented by further details
developed since that time.
2.2 Project workshop in Kuala Lumpur

The Core Project Team prepared draft Logframes for consideration by the HDM
Project Management Team and a several senior representatives of ADB countries.
The participants at the workshop (held 6th and 9th October) were asked to review,
revise and, having done so, to take ownership of the logframes.
The workshop confirmed most of the broad details of the three logframes, while
making the following changes:
i) The Goal was broadened to extend its scope and to remove the geographical
limitation of reference to the Asia-Pacific region. The draft final version
reads:
‘To improve the development and performance of the road sector and its use
of resources, within the context of national objectives, in order to minimise
total transport costs.’
ii) The Purpose was altered to emphasise not only increasing capability but also
actual application of HDM-4. The draft final version reads:
‘To generate and sustain increasing application of HDM in road management

and sector development in the Asia-Pacific region.’
iii) Outputs were enlarged to
(a) ensure that dissemination to executives and policy makers was

adequately addressed.
(b) emphasise the full range of media through which the training and
dissemination materials will be available.
iv) Verifiable Indicators were expanded to describe some details of the scope of
training and dissemination intended.

2.3 Developments since the project workshop in Kuala Lumpur
Based on the discussions amongst various stakeholders at the PIARC World Congress
in Kuala Lumpur, and subsequent discussions amongst the project team since then,
the strategy has been developed further and changes have been made to the approach
adopted above. In particular, the team considers that
i) the existing logframes did not develop the strategy to the required extent
beyond the timeframe of the existing project; and
ii) the three existing logframes demonstrated considerable overlap, and were
repetitious in content.
The team have therefore decided that two logframes would meet the objectives of the
Terms of Reference and provide a clearer and more focused strategy.
i) Logframe for overall strategy (Annex 1). This key logframe represents the
overall strategy for the training and dissemination component of HDM-4, and
identifies the key outputs which we believe the HDM-4 stakeholders are
aiming to achieve in the long term.
ii) Logframe for ADB funded project (Annex 2). A more detailed logframe for
the two year project has also been developed from the original project
logframe presented at Kuala Lumpur.
The contents of the logframes are described in more detail in the following sections.
First, the overall strategy is discussed (Sections 4 and 5) and then aspects of the
current project which contribute to this overall strategy are described (Section 6).
The Goal and Purpose of each logframe is the same, and so these aspects are
discussed first in a separate section.
It should be noted that, at present, there are no specific provisions for project support
to extend beyond the first two years. This is an issue that will need to be reviewed
during the project as it needs to be commensurate with the overall strategy for
managing HDM-4 in the future. This, in turn, depends on the evolving Business Plan
prepared by PIARC and future institutional arrangements which are yet to be
finalised.
3. OVERALL GOAL AND PURPOSE
3.1 The strategy

The overall strategy aims to:
• develop a sustainable dissemination and communication system, using

channels open to HDM-4 stakeholders that will facilitate dialogue and
exchange of information, knowledge and experience on operational and
topical aspects of HDM-4 use (5 years).

• develop a sustainable training system so that target groups will have the skills
and knowledge needed to optimise the use of HDM-4 in their area of work and
influence (3 years).
• secure extensive adoption of HDM 4 and high levels of technical and
professional performance in the road transport management and policy-
making sectors.
These aims can be encapsulated in the logframes at the Goal and Purpose level, as
agreed at Kuala Lumpur and described in Section 2.2. Whilst achieving the Goal and
Purpose is outside the control of the project, it is important to clarify those risks which
have been identified so that stakeholders can take steps to mitigate against such
possibilities.
Four risks, or assumptions, have been noted in the logframes. Three of these (suitable
data for and calibration of HDM-4, HDM-4 producing improved analyses, application
of results by policymakers without undue distortion) are self explanatory. The final
assumption that ‘HDM-4 gains credibility and acceptance by government agencies’
warrants further comment, based on discussions in Kuala Lumpur.
3.2 Sustainability
3.2.1 Acceptance of HDM-4
From discussions at the PIARC World Congress in Kuala Lumpur it is evident that
there are a number of key stakeholders who have concerns about the performance
parameters of HDM-4 and the strength of support/desire to adopt it. Difficulties with
calibration due to the complexity of the model, and the possible negative reactions of
some potential users to lower levels of projected benefits arising from the use of the
new version of HDM, have also been raised. These are some of the issues the project
team will have to resolve with stakeholders if HDM-4 development and adoption is to
be successfully prosecuted.
Failure to acknowledge and resolve these issues will seriously impair the willingness
of individuals, groups, institutions and governments to take ownership for the long-
term development and maintenance of HDM-4 in their region. Stakeholder
partnerships and participation at all levels will be fundamental to securing a dynamic
and broadly based post-project training and dissemination programme.
3.2.2 Motivation
In developing the project and post-project strategy it must be remembered that the
adoption of HDM 4 is not compulsory, so its benefits and advantages may need to be
made clear and, in some cases, incentives created.
The ADB intend that the project inputs should provide the training and dissemination
programmes with a direction and internal momentum to take the training programme
through to the end of year three and the dissemination programme to the end of year
five. How this post project phase will be supported and funded has yet to be resolved
but it is essential that the project identifies and pursues strategies that lead to early
local commitment and ownership of the HDM-4 programme so that post-project
activity can be considered and planned during the current project phase.

There are two areas which need resolution:
• The motivation of groups and individuals on the ground to adopt HDM-4

stewardship, training and dissemination needs.
• The financial/cost recovery issues of involvement and sustainability.
Motivation is a complex issue but two possible lines to follow in the project are:
• Individuals and groups perceiving HDM-4 as a vehicle for technical and

professional improvement, advancement and influence.
• Individuals and groups seeing HDM-4 as an opportunity to engage in training
and dissemination activities that can be financially and professionally
beneficial.
In truth such a division is simplistic and it is likely that various combinations of these
will exist in the environment in which the project will have to “sell” HDM-4. The
project team must carefully assess and exploit such factors when promoting the
formation of local and regional HDM-4 steering/support institutions/groups.
4. DISSEMINATION STRATEGY
4.1 Justification for dissemination

The starting point to ensure that the use of HDM-4 becomes sustainable worldwide is
the adequate dissemination of information concerning its purpose and use.
Dissemination is required to:
• Alert international donors and financiers to the role of HDM-4 so that they
design projects which have the potential to incorporate the use of HDM-4;
• Alert national policymakers to the role of HDM-4, so that they facilitate its
correct use within road and transport organisations;
• Alert users (including those who train others in the use of HDM-4) to the
scope and limitations of HDM-4; and
• Update the above groups progressively as HDM-4 is developed.
Hence we believe a suitable dissemination strategy comes before training, in so far as

developing the building blocks for sustainable use of HDM-4 is concerned.
The need to generate genuine support for the use of HDM-4 and belief in its value,
before simply informing target groups of its capabilities, has been identified and
emphasised in Section 3. This includes all parties who might be involved with HDM-
4, from policy makers, whose support will be required, through to users, whose
genuine use of the system will be needed in order to provide meaningful outputs. A
sustainable strategy must address this issue.
In order to achieve the above and also to make best use of the resources available for
dissemination and training, it is recommended that the content of dissemination
materials is the prime initial focus. Addressing the many different and more

innovative means of dissemination will follow afterwards. It is also suggested that the
means of prioritising the different options for dissemination should be in terms of the
greatest potential impact for a given cost. Therefore the highest priority should be
given to those methods which:
• target the most strategic audiences, in order to gain long term commitment
• target the greatest size of audience for given cost (benefit-cost)
• require the least amount of on-going input after initial set-up.
4.2 Existing modes of dissemination

Various forms of dissemination are already in place and operating. These are outlined
below, with brief remarks on their assessed impact, and the resources required in
setting up and maintaining their operation.
4.2.1 Newsletter
The ISOHDM Technical Secretariat at the University of Birmingham has been
publishing biannual Newsletters since 1996. This is a four-page document printed on
A3 size paper and folded in half to form the four pages of A4 size. The current print
format is black on white, although a multi-colour print format is under consideration.
The Newsletter is printed at the University of Birmingham at a cost of approximately
GBP 1,500 per print run thus totalling around GBP 3,000 per year.
The Newsletter is widely circulated in printed form to over 3,000 registered readers.
The vast majority of these are from the ISOHDM database which records details of all
individuals and organisations who have enquired about HDM-4 or are current users of
HDM-III/95. Electronic copies of the Newsletters are also posted on the ISOHDM
Web site hosted by the University of Birmingham.
Topics covered in the Newsletter have been wide ranging including the following:
• Overview of the HDM-4 system
• Descriptions of technical models
• Proposed plans for world-wide training programmes
• Pilot implementations of HDM-4
• Plans for release of HDM-4, etc.
Feedback from readers over the past few years indicates that the Newsletter has been
a successful and effective dissemination medium. Some readers have sent comments
on various articles in the Newsletter. However, only few readers have so far
contributed articles to the Newsletter.
4.2.2 Brochure
A full colour brochure was designed and published by the ISOHDM technical
secretariat at the University of Birmingham in 1998. Since then, over 5,000 copies of
the brochure have been distributed at conferences, by mail and to individuals visiting
PIARC and the University of Birmingham. The brochure is printed on hard-glossy A3
sheet and folded to form four-pages of A4. It includes several screen shots of HDM-4,
photographs of several roads with different types of traffic, pavement condition and
terrain.

The brochure is seen as an advertisement medium and it provides a summary of the
key features of HDM-4, as well as the typical areas of applications, information on the
accompanying set of documentation, etc. The total cost of production was of the order
of GBP 6,000.
4.2.3 Conference papers and presentations and published articles

An initial literature search of papers and proceedings which include HDM in the title
or abstract reveals that of the order of 120 papers have been published in the period
from 1987 to present. These are predominantly English, and it is expected that many
more have been published in French, Spanish and German, although no figures are
known. The search is not totally comprehensive since not all conferences are
included (only major international conferences are usually referenced), and also some
articles might concern HDM but not reference it in the title or abstract.
There have been several papers published at international and at regional conferences
covering topics directly related to HDM-4. Recent conferences include:
• Managing Pavements, Durban, May 1998

• PIARC World Road Congress, Kuala Lumpur, October 1999
• Transportation Research Board, Washington DC, January 1999
• ASCE Proceedings of the International Computing Congress, Boston,
Massachusetts, October 1998.
• Joint ARRB/Transit New Zealand Conference, Sydney, December 1998.
In addition to conferences, there have been a few published articles in learned journals
and other periodicals.
The annual circulation of the conference proceedings and journals in which articles
and papers on HDM-4 have been published is not immediately available. However,
the effectiveness of these publications is assessed to be high given that they are
targeted at the more technical audience and potential HDM-4 users. These
publications are seen to be essential in encouraging proper technical use of HDM-4
covering topics such as; calibration, adaptation, implementation, data requirements,
integration with other pavement management systems, etc.
It is therefore proposed that the ISOHDM secretariats at PIARC and at the University
of Birmingham should continue to encourage HDM-4 users to publish the results of
selected applications of HDM-4 which are judged to be of interest to other users. Such
articles and papers must, of necessity, meet the editorial requirements of the
conference and journal publishers.
4.2.4 Website at University of Birmingham, PIARC and roadsource.com

Currently there are three websites which address HDM-4.
The longest established is the site under the University of Birmingham’s site. This
contains details of the earlier studies, copies of newsletters, product information,
contacts and sponsors. There is an opportunity to submit feedback, and a page on
seminars (under preparation). The newsletters and product information appear up to
date; at present, the remaining pages are less comprehensive.

PIARC has also developed a site, although only one page of information on HDM-4 is
currently complete. Pages currently under construction are information on how to
obtain products, information on training opportunities, discussion and feedback
facilities, publications, and regional calibration and default data sets.
The third site, the roadsource site, is more extensive than either of the above, and
focuses on the technical aspects of the HDM-4 system. It is not clear that it is official,
although there is a registration page which seems to be sponsored by PIARC. Other
pages include a discussion site (not clear that this is operational), a contacts page, and
downloads of various technical papers. The papers and documents included are not
comprehensive and suffer from the fact that updates to various technical issues are
continually being made, so that some of the documents are obsolete.
The success of the sites is unknown although the number of visitors to each site is
recorded. This is of the order of one to two thousand per annum.
4.2.5 Training courses

Courses in HDM-III have been run in the past at the University of Birmingham, and
also at locations in France and Spain. In the past, a course was also run by PTRC in
the UK. In addition to these specific courses, there have been numerous more
informal training courses run as part of larger road management projects in different
countries around the world.
Other courses also exist which include sessions on HDM within a broader context.
These have included the annual Management and Financing of Roads Course at
University of Birmingham, the annual Roads and Transport Course at TRL and other
similar events.
4.2.6 Support from University of Birmingham and other institutions

Various institutions, such as University of Birmingham and TRL, receive a number of
ad hoc inquiries for general information and advice concerning HDM. It is expected
that these inquiries will increase dramatically once the new software is released.
However, an estimate of annual current time spent answering such inquiries is of the
order of days, rather than hours or weeks, but this is to be confirmed.
A formalised system will need to be set up once the software is released, as it is

expected that local user support groups would be inadequate for this purpose. At
present, responsibility for this help system lies with PIARC, but clearly there is a role
for local users which will need to be linked with any more formal help system.
4.3 Options for dissemination

4.3.1 Long term HDM-4 ‘hub’ overseeing dissemination
Whilst this is not a true dissemination activity in itself, all dissemination activities will
need to be coordinated in the long-term. Currently, PIARC is providing the role of
coordinator for such aspects but it is understood that this arrangement only extends to
???June 2000. If the following dissemination activities are to be properly
coordinated, then it is recommended that long term support beyond this date is
confirmed and put in place.

Potential impact High. Whilst not a dissemination activity in itself, it provides a coordinating
function for supporting and managing all other dissemination activities
(particularly 5.3.2, 5.3.3, 5.3.4 and 5.3.8).
Potential costs To be estimated
Major risks Low. If implemented, the risk is only that the role would be ineffectual
were there to be lack of funding to implement activities.
Actions required Sponsors should confirm role so that future dissemination plans can be
sustained.
4.3.2 Regional road management working groups

In order to generate stakeholder participation, it is proposed that working groups
should be set up. It is recommended that these are on a regional basis (for example,
Asia-Pacific, Africa, Latin America etc.) and it would be possible to explore this
option during the timeframe of the current ADB project (although this activity is not
funded).
The intention would be that Country Champions (one per country), trainers and users
might all contribute to such a group, which might meet on an annual basis at, for
example, a regional road management conference, with the purpose of exchanging
experiences of using HDM-4 within the road management context.
The use of Champions has been discussed at Kuala Lumpur and it is felt that this is
the best way to proceed with the current ADB funded project. Country champions
will be high profile, high calibre personnel with considerable experience within their
country. As such they will have experience of operating at the policy level, and
understand the political context of road management. It is expected that one per
country will be appointed and their area of contribution would ideally include, but not
be limited to:
• selection of the most suitable ‘institution(s)’ to organise in-country training;

• identification of Government departments, consulting and contracting firms
who use or should use HDM;
• criteria for selection of trainee trainers;
• approaches to advertising and dissemination; and
• approaches most likely to support sustainable in-country development of
HDM-4 in a post project environment.
Country champions will be identified for the six countries during the ADB funded
project. On-going support for them, and future country champions elsewhere, is
beyond the scope of this project. The setting up of any working groups would also be
beyond the scope of the current ADB funded project.
Potential impact High. The target audience would combine users and stakeholders, and
generate a sense of local ownership.
Potential costs (1) Initial set up costs would be administrative time of a coordinating body
to promote the idea (PIARC).
(2) Some on-going support would be needed to fund members to attend
working groups, if they were unable to obtain funding elsewhere. The cost
of use of a suitable venue / facilities / publishing ‘proceedings’ etc. would
also need to be covered.

Major risks (1) Suitable Champions might not be identified or motivated to give long
term support to the objectives of the strategy.
(2) Users might find difficulty attracting sponsorship or finding time to
attend working group. A minimum number of members (perhaps 10 – 15)
would be needed to make exchanges worthwhile.
Actions required (1) The current ADB project is investigating the appointment of Champions.
The working group concept will be explored with them, and at the Train the
Trainer course, to assess viability.
(2) After such investigations, assuming the view is positive, funding will be
required and a co-ordinator would need to be appointed to manage the
implementation.
4.3.3 Workshop sessions

A key concern identified by the project team is that many national policymakers
might not be fully aware of the role that HDM-4 might play in road management. Of
greater concern is the fact that some might actually see little reason to adopt the
model, with a view either that their own systems are currently sufficient, or that the
HDM-4 model offers no benefits or is unsuitable for their institutions. Such people
therefore represent a key target group since, without their support from high levels
within road administrations and government ministries, there is no chance of
achieving sustainability. In addition, international donor staff and financiers who
network with such national policymakers may also hold a similar view.
It is therefore recommended that the project sponsors are proactive in targeting this
group by arranging workshop sessions in which HDM-4 is promoted. These
workshop sessions might be held on a national basis, at an in-country road conference
or similar occasion, or a regional event. The former is preferred, in order to focus on
the needs of each country.
Potential impact High. The target group is the key to long-term sustainability.
Potential costs (1) Coordinating contacts etc to arrange suitable meetings
(2) Contributors will need some preparation time to provide well prepared
contributions.
(3) Costs of attendance by donors and financiers (national policymakers
attendance assumed). Costs of a workshop coordinator.
Major risks Low. Assuming HDM-4 is appropriate, the target group should be
persuaded.
Actions required Identification of suitable forums, and coordination of set-up.
4.3.4 Targeted conference sessions

An existing form of dissemination is the use of special sessions at major world and
regional conferences to bring attendees up to date with HDM-4 developments. This is
merely advertised as part of the conference and, as such, specific groups are not
targeted.

Potential impact Medium. Good general dissemination but no guarantee as to calibre of
attendees.
Potential costs (1) Coordinating role to ensure conferences targeted and sessions set up.
(2) Contributors, if to provide well prepared contributions, will need some
preparation time.
(3) Possibly some costs to assist with attendance.
Major risks None identified.
Actions required Conference targeting and contacts.
4.3.5 Published papers and articles

Conference papers and articles in journals are a traditional form of dissemination and
occur ‘naturally’, stimulated by the professional interest and self marketing of
contributors. They are useful for potential users to learn about different possible
approaches to their own work, but probably are of little impact to higher level
policymakers. However, there is clearly no cost to HDM-4 of such activities, and the
process is self-sustaining whilst there is an on-going use worldwide of the model.
It might be considered worthwhile monitoring the quantity (quality is assumed to be

monitored by the journals and conferences own vetting procedures) and maybe this is
an activity which should be taken on by the coordination body (4.3.1). In addition, in
order to stimulate contributions, a prize (with small financial award) could be
awarded on an annual basis to the best paper on the subject.
Potential impact Medium to low. Such articles and papers only tend to be read by those
proactively interested.
Potential costs (1) A resource guide, identifying all published material, will be developed
under the ADB project in electronic and hardcopy format (subject to ADB
approval).
(2) On-going updates of the guide, and an annual sum for a prize.
Major risks None identified.
Actions required (1) Resource guide under ADB project. Cost estimate to be developed by
project team.
(2) Funding for on-going updates and prize put in place.
4.3.6 Website
As noted in 4.2.3, there are existing official websites under the University of
Birmingham and PIARC, and also a site sponsored by HTC Infrastructure
Management Ltd (www.roadsource.com). Whilst the user who is unaware of where
the websites are might enter the internet in a variety of locations (and so references on
key sites are recommended), it is recommended that a rational strategy is adopted for
the websites that hold information – this will avoid repetition, possible conflict, and
minimise costs (since if a website is to be maintained up to date, funding will be
required for its establishment and on-going management).
Potential impact Difficult to assess but will become more important with time
Potential costs Management fees for web maintenance
Major risks Without a working website, those wishing to attempt independent study or
upgrading of their current HDM-4 knowledge could be severely
compromised.
Actions required (1) Project sponsors rationalise ideas on how many websites are required

4.3.7 Audio-visual and electronic media
Promotional videos can be useful for some dissemination purposes, on topics which
lend themselves to this form of presentation (e.g. “virtual” site visits, visual
comparisons of HDM-4 impact). However, at this stage it is not possible to prescribe
where the application of this communication tools would be best employed in the
process of HDM-4 training and dissemination. In addition, the cost of developing
such videos and computer accessed graphic materials is high and no budget has yet
been identified to support this. Stability of the subject matter is also an issue that will
have to be considered while HDM-4 is still evolving. It is too early to be prescriptive
here, but the situation will be kept under review.
The use of electronic (CD-ROM) media, either as an addition to the HDM-4 program,
or stand-alone, would also be expensive. This aspect is discussed under training, as it
is seen more as a training activity.
Potential impact Medium.

Potential costs High, and on-going updates would be required
Major risks Frequent updates may be necessary in years 1 and 2
Actions required Review in 3 years time, once other forms of dissemination have been
implemented.
4.3.8 Update information with new releases of program

A new release of an updated version of HDM-4 will provide a major focus for
dissemination. This is because users (and to a lesser extent stakeholders) will have an
automatic self-interest to find out about the nature of the updates. It is therefore
recommended that this time be used for a key dissemination output. This would
probably be in the form of a brochure, outlining the updates as well as including
examples of recent best practice and key issues. Such a brochure could be issued with
the software, but also sent out stand-alone to all contacts maintained on the contact
database, and also put on the website and launched at a suitable conference.
The assumption above is that a database of contacts has been maintained up to date.
This supporting activity will enable targeted dissemination for this, and other
purposes such as the newsletter (see 4.3.9). The on-going maintenance of such a
database will require administration and management.
Potential impact High.

Potential costs (1) Maintenance of database of contacts
(2) Development of brochure, publishing costs etc.
Major risks None
Actions required (1) Design of typical brochure and more detailed cost estimates
(2) Suitable conference launches identified
4.3.9 Newsletter and brochure

As noted above, it is considered that the newsletter remains one of the most effective
means of dissemination at present. The content of the existing version seems to be
well received, although its formatting style could be improved. In order to improve
its effectiveness, it is proposed that a higher quality print medium be used, such as
glossy paper with colour prints and photographs. This would update the format to the

standard of similar Newsletters published by others. It is currently available on the
website and this should be continued.
It is also proposed that the brochure (see Section 4.2.2) should now be updated in
order to incorporate the latest features in HDM-4 and the distribution arrangements.
Potential impact High

Potential costs (1) Maintenance and updating of list of contacts.
(2) Publishing costs, and probably some small amount to cover contributions
to ensure a suitable quality is maintained.
Major risks None
Actions required (1) Strategy for publishing once contract with UoB is completed.
(2) Agree budget for improved newsletter style and new brochure
5. TRAINING STRATEGY
5.1 Identified learning needs

The strategy acknowledges that the majority of those being targeted in the early
period will already be professionally qualified and experienced people who have
previously demonstrated their ability to utilise training and learning opportunities.
They are also likely to be familiar with HDM-III applications. However the learning
material and information outputs of the programme will also have to cater for a wider
interest group including:
• Undergraduate civil engineering students and others in similar positions on the

professional ladder
• Users (engineers, economists, managers)
• People in senior positions in companies and government who will use HDM-4
outputs to shape policy and strategic planning.
• Others with no previous experience with HDM-III or 4.
All these groups have differing training and dissemination needs and the project will
have to define and prioritise these needs so that appropriate strategies can be
developed to meet them within the Project and subsequently in the post-project period
(Annex 3 – Training programme development).
The programme will be developed to meet the identified learning needs. The
identification process will be carried out through close collaboration with the project
technical advisory team, Country Champions and a range of other stakeholders. It is
recognised that the likely numbers of people who require training will not be large
and therefore it will not always be possible to assemble groups who will be
homogeneous in their requirements. The training material and methods of delivery
will also need to be tailored to cater for this diversity.
A first assessment of one of the target audiences - the Key Trainers who are to be
trained under the ADB project (see Section 5.6) – in terms of their required
characteristics and needs is provided as Annex 4. The training needs assessments for
future trainees in the countries where the trained Key Trainers will operate are

developed at the Train the Trainers (T3) course by the relevant Trainers. In this way,
the Key Trainers will develop their own capability in assessing training needs in-
country.
5.2 Access to training

The provision of training falls into two components. These are the mode of training
and the method of delivery.
5.2.1 Modes of training

Emphasis will be placed on ensuring easy access to training and learning opportunity.
This will be achieved by developing training support materials in structured learning
packages that can be delivered in a number of modes. The primary modes available
to the project are described in Tables 5.1 to 5.3.
Table 5.1
Advantages and disadvantages of Trainer led courses and workshops
Advantages The trainer is able to tailor the training to meet the precise needs of
the trainees.
The trainees can get instant feedback and guidance from the trainer.
Trainees have a wider learning experience through interaction with
peers. (Group exercises and discussion).
Trainer can enhance the training content with topical and local
examples.
Training material can be produced in most simple print form and
subsequently enhanced by individual trainer to suit target groups.
Disadvantages Requires skilled trainers.
Requires trainees to come together at same time.
Needs appropriate training venue and equipment.
Trainees have to progress at pace of the group.
Not all trainers can accommodate a wide range of learning ability
and experience.
May be difficult to identify good trainers.
Table 5.2
Advantages and disadvantages of Supported Independent Individual/Group
Learner(s)
(For this method, a local tutor or coach is available to provide intermittent support,
which might be face-to-face by appointment, or telephone, or e-mail.)
Advantages Trainee works at own pace and in chosen place of study.

Learners must have reasonable level of study skills and self-
motivation.
Does not need intensive trainer input but allows learner to receive
guidance when necessary.
Can be delivered through printed or computer based material
resources.
Groups can support each other by having regular meetings (works
well if all are working in same department of company).

Disadvantages Requires learning material to be developed in detail and enhance by
many examples and exercises.(expensive and needs large numbers to
achieve acceptable unit costs).
Considerable learner guidance on getting best from the programme
needed. The latter requires writers with considerable experience in
this type of learning.
Learning material more expensive to produce.
Feedback period can be long.
Can be difficult to identify and manage tutor support.
Table 5.3
Advantages and disadvantages of Independent Individual/Group Learner(s)
Advantages Similar to “supported” above.

Disadvantages As “supported” above but:-
Requires high level of study skills and self motivation.
High drop out rate.
Difficult and expensive to provide learner support.
Learning material has to be of very high quality.
The three tables above set out the main features but in practice it is likely that the
methods summarised in Tables 5.1 and 5.2 would be the most effective modes for
learners who are encountering HDM-4 for the first time. Table 5.3 could provide the
mode for training which is intended to enhance the learners ability to use HDM-4 in
specific circumstance or to add a specific process to their current experience i.e.
upgrading.
5.2.2 Methods of training

In delivering the training, a wide range of methods are available, e.g. Face to face
training with supporting texts, trainer managed, audio visual materials, inter-active
computer-based materials, printed manuals and texts, CD ROMs and video, tailor
made commercial computer based platforms to support distance learning programmes,
internet conferencing and video conferencing. However many of these will exclude
themselves on a cost basis and current available bandwidth for transmission.
While the use of distance and independent study approaches will be developed it is
wise to recognise that the inputs and support needed to create such a sustainable
training system is well outside the current project budget. The project will carry out
research to establish what realistic provision might be feasible and viable within the
parameters of need and numbers. It will be part of the project strategy to establish a
network of local trainers/experts who can provide support and advice to those wishing
to adopt independent modes of study. The essence of considering a variety of
delivery modes and methods is so that various emphases and mixes can be achieved to
suit different needs.
5.2.3 Developing the strategy during the ADB project

The project strategy will be to establish the Trainer-led model through the Train the
Trainer programme. Training material will first be developed in print form and

contained within a modular structure. Significant experience building and material
development will be achieved through the T3 course and the subsequent initial in-
country courses. Lessons learned from these initial pilot programmes will be used to
guide the on-going development of module materials.
If the training programme is to be sustainable it will be essential that the project team
continue to provide the Key Trainers with support and guidance throughout the
project. It is also clear that the Key Trainers have a major role to play in the long-
term development and sustainability of the programme in their home countries. The
project will endeavour to ensure that training programme ownership is established at
country, regional and international levels. This issue will be included in the wider
strategy development discussed in Sections 3 and 4 of this document.
5.3 Modular structure

The training programmes will be developed as a series of free-standing modules/units
so that maximum flexibility is available for individuals and groups to build
programmes which are tailored to meet their specific training and knowledge needs.
The full module menu will cover all the identified learning needs of HDM-4 and will
sit within a three dimensional matrix covering;
• Different target groups (people),

• Subject breadth,
• Subject depth.
Account will then need to be taken of different possible training methods and modes
of delivery (See 5.2 above and Annex 4 – Proposed training module structure).
The training programme will provide introductory programmes, user programmes at

various levels and aspects of management, economics, planning and decision making
at senior professional level. The aim is one of confirming HDM-4 as a management
tool.
Module development will begin as soon as possible and an early priority must be on
providing training to those concerned with the use of HDM-4 in policy and planning
roles e.g. senior civil servant, planners and sector leaders. A commitment to using
HDM 4 in planning and management at senior level will ensure a greater likelihood of
commitment to, and bottom up demand for, training. A general consensus amongst
key stakeholders is that this is a vital to ensuring the success of the programme.
It is uncertain how many of those nominated to become key trainers in the project will
have the experience or standing to deliver this “policy” level of training. The project
may have separately to identify and train suitable candidates to carry out this work.
5.4 Module content

Each module will have a technical content which meets the competencies and
knowledge base required. It is emphasised that the competencies and intended
knowledge requirements are seen as minimum standards of achievement and module

writers must ensure they provide opportunities for those with greater capability and
potential to develop to a higher level.
The structuring of technical content will conform to established learning models so

that the learner is lead through a progressive hierarchy beginning with basic
knowledge and comprehension and progressing through to more complex examples
and exercises in application, analysis and synthesis. This will be achieved by
extensive use of problem solving exercises and case studies. Some case studies will
be extensive and require the learner to undertake analysis, problem solving and
decision making work within a real work environment. A tentative list of module
titles is given in Annex 4.
5.5 Assessment
The practicalities and desirability of assessment and possible certification of HDM-4
training programmes is not yet clear. However it is desirable that the effectiveness of
training should be tested and feedback used to improve module content, structures and
delivery. To assist this process the development of the training modules will include
the specification of expected competencies and levels of performance to be achieved
at the end of each module. The political and practical issues of evaluating and
certifying individual learners will be explored and resolved locally as the project
develops.
The required standards for accreditation for the different levels of HDM-4 may be, for
example
Level 1 Ability to train others Trainers

Level 2 Operational competence Users (engineers, economists etc) More
Level 3 Ability to Interpret Managers understanding
Level 4 Awareness Policy makers / students etc.
The three-year plan must not restrict the assessment policy by becoming fixed once it
is documented. It will need to be responsive to the evolving HDM-4 application
environment.
5.6 Training the trainers

The core strategy initiative in training will be to identify and train a group of “Key
Trainers” who can translate the training programme into a sustainable range of
training provisions in their own country. They will have to be trained to a high level
of technical competency and have sufficient training skills and knowledge to enable
them to have the confidence to run in-country “multiplier” courses. They will also
have to understand, promote and “manage” a range of modules and learning modes by
establishing a network of support experts to provide expert guidance and tutorials to
those attempting to study through individual study programmes. However, during the
ADB funded project, the initial training in training methods will focus on
participatory face-to-face training. It is taken as read that a good working relationship
between the in-country “Champion” and “Key Trainer” will be a critical element in
HDM-4 promotion and adoption.

At this stage it is impossible to prescribe good practice as it will only emerge from
pragmatic initiatives during the current ADB funded project. It is probable that a
national institute concerned with civil engineering education and research may be a
possible host for HDM-4 dissemination and development. Several of the target
countries (e.g. Malaysia and Thailand) have well-developed technical and higher
education institutions that may be capable of satisfying many of the required criteria.
The selection of key trainers will be critical to the success of the training programme.
Selection should take note of the nominees interpersonal and communication skills,
previous training experience and capacity to promote training and market HDM-4 as
well as their technical expertise and experience.
It is intended that the initial T3 programme will aim to achieve the strategy objectives
by ensuring that trainees will:
• demonstrate satisfactory knowledge of and skills in the use and application of

HDM-4 in planning and management decisions;
• plan and present a participatory training session on a selected aspect of HDM-
4 to peers;
• prepare a training programme proposal from available modules and modes of
study for a real client training need profile; and
• prepare an action plan for promotion and dissemination of HDM-4 on return to
country.
Given the developmental nature of the Project it will be essential to support “Key
Trainers” in launching new modules and modes of study as they become available.
Following the T3 course the project team will review the original train the trainer
thinking with the aim of strengthening the longer-term in-country support of trainers
and present ADB with an update. This may need some reflection and discussions
with “Key Trainers” and Champions before a formal proposal can be formulated.
5.7 Major risk / assumptions

The key risks associated with the training strategy are summarised below.
5.7.1 Sustainability
Risk There is a concern that sustainability of the strategy is not assured within the
region, particularly after the timeframe of the Project.
Actions required This risk has been discussed in Section 3.2.
5.7.2 Champions and Key Trainers

Risk There is a concern that suitable Champions and Key Trainers might not be
identified or motivated to give long term support.
Actions required This risk has been discussed in Section 4.3.2 and 5.6

5.7.3 Demand for training
Risk Basic assumptions have been made in developing the train the trainer, and in
country strategies, that there is, and will continue to be, a demand for in
country training units and trainers. If there is not, then trainers will lack
motivation and there will be no ‘momentum’ to ensure on-going training
initiatives.
Actions required Project team to discuss and assess with Champions and during first T3
course, and report after T3 course on feedback.
5.7.4 Priority given to HDM-4

Risk Uptake of HDM-4 might not be seen as a priority by countries who have
their preferred methods by which they achieve similar objectives.
Actions required This risk is addressed in the dissemination strategy under Section 3.2 and
Section 4.1.
5.7.5 Modes of delivery

Risk There is a danger that within the budget available, a disproportionate amount
of resources could be spent on developing different media for training and
dissemination, rather than focusing on the quality of the core content.
Actions required The intention is to begin with written materials, and then to expand into
adaptation of these for different modes of delivery on confirmation of the
training and dissemination strategy (See Section 5.2.3).
5.7.6 Calibre of Trainers

Risk Trainers of a suitable calibre to ensure long term sustainability of the training
will not be able to be identified in the target countries, or will require longer
term support than currently provided for in project.
Actions required (1) Champions must be fully involved in identifying and supporting trainers.
(2) Strategy to propose how longer term support might be managed. This
risk can only be properly assessed during and after the T3 course. However,
it is an important sustainability issues, since HDM-4 is continually evolving
and the likelihood is that updates and hence additional training modules will
be required well beyond the timeframe of the existing ADB project.
5.7.7 Road management practice versus HDM-4 capability

Risk Unsustainability of training programmes in countries where road
management practices remain fairly undeveloped.
Actions required (1) Stakeholders will need to accept a longer term view in such countries
where use of HDM-4 might not become institutionalised until well beyond
the timeframe of the existing strategy.
(2) Country training will need to account for such background, and the
expected level of competency expected of participants will need to be
considered against this background.
6. THIS PROJECT
The ADB funded HDM-4 Training and Dissemination Project will be implemented
over the next 18 months to two years. The detailed logframe for this project is
provided in Annex 2. The project will build a project training capacity by
• Carrying out a training needs analysis of stake holder groups

• Producing and distributing training guides and training support materials

• Identifying and training 24 key trainers from 6 countries in the region
(between 2 and 4 from each) – Indonesia, Thailand, Philippines, Pakistan,
India and Bangladesh. (Train the Trainers)
• Providing in-country support to key trainers when establishing training courses
in the six countries identified, aiming to produce 10 national end users per
country.
The project will build a project dissemination and communications capacity by
• Establishing a (or further developing an existing) web site, (interactive; focus

group interaction; bulletin board; training material down-load; calibration
baselines; programming up-grades;– news; to be assessed etc)
• Identifying and supporting HDM-4 in-country/regional “champions” who can
influence and guide the promotion and adoption of HDM-4. (influencing
policy makers, supporting training, promoting local activities and publications,
maintaining regional and international dialogue)
• Developing communications between country/regional groups and
international bodies providing HDM-4 representation/profile at international
conferences and events.
The Training and Dissemination process and its subsequent development and
sustainability will have to be owned and operated by local, regional and international
stakeholders, as stated previously. The early and full involvement of stakeholders is a
critical factor in developing such ownership. Stakeholder commitment is also critical
to Project purpose and goal attainment. In this regard, the early identification and
engagement of country Champions will be crucial to Project progress.
The project team will develop and introduce a selection of training and dissemination
models so that options are available for stakeholders in different countries to utilise
according to their needs and resources. The on-going development of training and
learning materials and modes of delivery will be informed by the feed back from
stakeholders and field experience. However, the principle will be to start from a
simple basis, and develop different media etc from that point. In this regard, the first
task of development of materials will be to ensure there are a core set of documents
from which the training modules can be developed.

ANNEX 1
LOGICAL FRAMEWORK FOR TRAINING AND

DISSEMINATION OF HDM-4
ANNEX 2
LOGICAL FRAMEWORK FOR ADB FUNDED TRAINING AND

DISSEMINATION PROJECT
ANNEX 3
TRAINING PROGRAMME DEVELOPMENT

ANNEX 4
NEEDS ASSESSMENT AND PROPOSED MODULAR

STRUCTURE FOR HDM-4 TRAIN THE TRAINER
PROGRAMME
LOGICAL FRAMEWORK HDM-4 Training and Dissemination Project
Narrative Summary Verifiable Indicator Means of Verification Important Assumptions

Goal 1 Road planning, development and Annual performance reports and
To improve the development and maintenance are more economically National statistics.
performance of the road sector and its and efficiently implemented. N/A
use of resources, within the context of 2 Country programs in region show
national objectives, in order to increasing NPV/C indications.
minimise total transport costs. 3 Other performance indicators
(Country programme performance is
measured increasingly using
performance indicators).
Purpose Purpose to Goal
To generate and sustain increasing 1 Government road projects, 1 Government project, programme 1 Input data available for analysis are
application of HDM in road programmes and strategies and strategy reports. reliable & HDM is calibrated.
management and sector development increasingly developed using HDM-4 2 Consultants’ study reports. 2 Extent and quality of HDM-4 usage
in the Asia-pacific region. from 2001. 3 Research reports results in significantly improved
2 Individuals in government, and 4 Wider scope of criteria used in analysis.
academic/research institutions reports, e.g. congestion, NMT etc 3 Results of analyses are applied by
increasingly using HDM-4 from 2001. 5 Correspondence between user policymakers without undue distortion.
3 Formal/informal interaction taking groups. 4 HDM-4 gains credibility &
place between project countries/staff 6 WEB access hits. acceptance by government agencies.
and other Asia/Pacific countries. 7 Donor agency & client audits.
4 HDM analyses are of high quality
Outputs Output to Purpose
1. Training & dissemination strategy 1.1 Draft TDS Version 1 covering: 1.1 Workshop held at Kuala Lumpur 1 Adequate champions identified.
published and established. a) General dissemination and feedback integrated.
b) User levels 2 Number of trainers sufficient to
c) Training delivery methods sustain in-country programmes.
d) Feedback
produced for PIARC conference (Wk 3 Trained personnel will have
11) adequate opportunities to gain
1.2 ADB approve final Version 1 (Wk 1.2 Project records. practical experience in the use of
21) HDM-4.
1.3 Strategy document in regular use 1.3 Project records.
by project staff & stakeholders (Wk 21) 4 HDM-4 technology can be
and final Version published on project adequately interfaced with existing
completion (updated by project road management systems.
experience).
5 HDM-4 is perceived to be a
2. Primary training and dissemination 2.1 Baseline materials from HDM-4 2.1 Project records. viable/usable tool for road
materials produced and distributed and road management resources management and so there will be
and feedback system operating. produced (Wk 28) adequate demand for HDM-4 trained
2.2 Draft training materials and guides 2.2 Project records. national personnel.
covering technical content, training
structure and approach in hardcopy 6 Feedback system continues to
and computer based form available for respond to user needs.
T3 course
2.3 Final materials and 2.3 Project records.
recommendations for different modes
of delivery (Wk 46)
2.4 Dissemination materials on UoB 2.4 UoB website.
website (Wk 50)
2.5 Project responds to feedback and 2.5 Project records.
updates materials on an on-going
basis
3. 24 T3 trainers trained. 3.1 TT Candidates qualifications 3.1 Candidates CVs and project
satisfy selection criteria (week 26). records.
3.2 At least 50% of TT Candidates 3.2 Project records and course
satisfy performance criteria (week 41) evaluation.
3.3 Action plans produced by country 3.3 Course evaluation.
groups (week 41).
4. 10 National end-users (and others) 4.1 One in-person training course run 4.1 Project records.
in each pilot country trained to at selected institutes in each of 6
appropriate level. selected countries in Asia-Pacific
region (Wk 68)
4.2 End of course evaluations and 4.2 Future courses incorporate
employer feedback. lessons learned.
Activities Activity to Output

1.1 Draft log frames for TDS, and 1. Agreement will be reached on the
Training and Dissemination project TDS at the KL Workshop
and draft TDS report prepared (Wk
11)
1.2 Draft TDS report and log frames 2. Suitable champions are forthcoming
reviewed / revised at workshop
(Wk 11)
1.3 TDS Version 1 produced, approved 3. Sufficient numbers of suitable and
and circulated (Wk 21) well motivated trainee trainers are
1.4 Investigations into 1) dissemination available.
modes, 2) dissemination materials
including translation and
publication, 3) financing, 4) 4. There is sufficient support, both in-
feedback systems, 5) training country and from this project, to make
modes, 6) training levels and 7) running of courses feasible.
support; on-going and updates in
TDS Version 2 (Wk 46)
1.5 Final TDS Version 3 and Final 5. Sufficient numbers of suitable and
Report (Week 69) well motivated trainees are available.
2.1 Identify levels of competence /

scope required at end of training
(Wk 21)
2.2 Identify / review existing training
material for HDM III/4 (Wk 21)
2.3 Produce training materials (Wk 28
then on going updates)
2.4 Supply / advertise / distribute
training materials (Wk 28 onwards)
2.5 Identify range of dissemination
materials required to meet needs
of dissemination strategy and
those to be produced under this
project (Wk 28)
2.6 Produce dissemination materials
(Wk 28 onwards)
2.7 Supply / advertise / distribute
dissemination materials (Wk 46
onwards)
2.8 User feedback system (Wk 21
onwards)
3.1 Appoint country champions (Wk

21)
3.2 Selection criteria specified and
trainees selected (Wk 26)
3.3 T3 course announced (Wk 26)
3.4 Course planned (Wk 30)
3.5 Course implemented (Wk 38)
3.6 Trainees examined/accredited (Wk
42)
3.7 Course reviewed/evaluated (Wk
43)
3.8 Write T3 completion report (Wk 43)
4.1 Mobilisation of country team.

4.2 Course planning – target group /
content / duration / venue /
organisation / materials (Wk 41 –
55)
4.3 Training courses (Wk 55 – 65)
4.4 Courses reviewed/evaluated (Wk
68)
4.5 Reports on courses produced (3
during Wks 57-68)
LOGICAL FRAMEWORK HDM-4 Five year training and dissemination strategy
Narrative Summary Verifiable Indicator Means of Verification Important Assumptions
Goal 1 Road planning, development and 1.1 Annual performance reports.
To improve the development and maintenance are more economically 1.2 National statistics.
performance of the road sector and its and efficiently implemented. N/A
use of resources, within the context of 2 Country programs in region show 2 Annual reports.
national objectives, in order to increasing NPV/C indications.
minimise total transport costs. 3 Country programme performance 3 Annual reports.
are measured increasingly using
performance indicators.
Purpose Purpose to Goal
1 Government road projects, 1.1 Government project, programme 1 Input data available for analysis is
To generate and sustain increasing programmes and strategies and strategy reports reliable & HDM is calibrated before
application of HDM in road increasingly developed using HDM-4 1.2 HDM-4 user feedback system use.
management and sector development from 2001. (PIARC/ISOHDM) 2 Extent and quality of HDM-4 usage
in the Asia-pacific region. 2 Individuals in government, and 2.1 Consultants’ study reports, results in significantly improved
academic/research institutions research reports and academic papers analysis.
increasingly using HDM-4 from 2001. 2.2 Wider scope of criteria used in 3 Results of analyses are applied by
reports, e.g. congestion, NMT etc policymakers without undue distortion.
3 Formal/informal interaction taking 3.1 Correspondence between user 4 HDM-4 gains credibility &
place between project countries/staff groups. acceptance by government agencies.
and other Asia/Pacific countries. 3.2 WEB access hits.
4 HDM analyses are of good quality 4.1 Donor agency & client audits.
Outputs Output to Purpose
1. International structure empowered 1.1 Adequate funding is provided for 1.1 PIARC Annual Report and Annual 1 ADB makes HDM-4 their preferred
and funded for global support of PIARC to continue and extend Accounts method of highway investment
HDM-4 on a sustainable basis role, or for a suitable alternative by appraisal
(4.3.1) Jan 2001
2 Funding secured for achieving the
2. International donors and financiers 2.1 Attendance of 80% target groups 2.1 PIARC reports proposed outputs
and national policymakers aware of at joint workshop sessions by Jan
potential and role of HDM-4 in 2001 3 Financial support secured for group
highway management and project 2.2 ADB staff, project team and others 2.2 Visit reports and anecdotal activities and affiliation to
design (4.3.3) report in-country discussions on evidence international groups
HDM-4
4 Adequate expert support is secured
3. Regional Road Management 3.1 Consists of champions, trainers 3.1 RRMWG reports and minutes for achieving outputs
Working Group established (4.3.2) and key users from countries
within region 5 Training demand will continue
3.2 Establish 6 countries membership 3.2 RRMWG reports, minutes and
by end of ADB funded project (not ADB project final report 6 Other non project countries are
funded yet but within timeframe of recruited to HDM-4
existing project)
4. Increased volume and scope of 4.1 HDM-4 represented at targeted 4.1 Reports
communication and dissemination conferences (4.3.4)
system (4.3.4 – 4.3.9) 4.2 Published papers and articles 4.2 Conference proceedings and
supporting HDM-4 use (4.3.5) journals
4.3 Website has increased range of 4.3 Site visits and user feedback
information and services (4.3.6)
4.4 Increased use of media systems in 4.4 Budget spend, availability of
promotion and training (4.3.7) materials, and training and
4.5 Program updates disseminated dissemination reports
(4.3.8) 4.5 HDM-4 updates with new releases
4.6 Newsletters published biannually 4.6 Mailing list and recipient feedback
and new brochure produced
(4.3.9)
4.7 Help network established and 4.7 System reports and user feedback
operating at local and international
levels by early 2001
5. Sustainable system of maintaining 5.1 Training materials review team 5.1 Reports and feedback from field
and upgrading training modes and formed
materials established (5.1-5.4) 5.2 Upgrades available and distributed 5.2 Reports from trainers and
champions
6. Sustainable training programme 6.1 Training provision meeting 6.1 Country surveys
extended throughout region (5.5) demand with low level of exclusion
Activities to Output
Activities 1. There is increasing demand for
1.1 ToR drawn up for empowering competent users of HDM-4
body to include training and 2. There are candidates who want to
dissemination coordinating function be trained.
1.2 Body (currently PIARC) appointed 3. There is funding available to pay for
for long term candidates to be trained.
4. Fees from courses (and other
2.1 Suitable forums identified and sources) are sufficient to make
sessions targeted and coordinated training financially sustainable.
5. Universities & other institutions
3.1 Concept explored and developed perceive HDM-4 training to be
with champions and trainers at first important.
T3 course. 6. Trainers / institutions continue to
3.2 Proposal for logistics and funding produce training materials for all
requirements in TDS Version 2 media.
3.3 HDM-4 sponsors implement 7. The standards established in the
funding and activities to ensure pilot project are at least maintained.
future developments 8. There is sufficient interest to lead to
WEB sites being used and
4.1 Current dissemination volume and maintained.
scope reported, and conceptual
ideas for increased levels
proposed, in TDS Version 1
4.2 Costed proposals included for
approved ideas, in TDS Version 2
4.3 HDM-4 sponsors implement
projects accordingly
5.1 Current training mechanisms

reported and conceptual ideas for
increase in scope proposed, in TDS
Version 1
5.2 Costed proposals included for
approved ideas, in TDS Version 2
5.3 HDM-4 sponsors implement
projects accordingly
6.1 TDS Version 1 (Dec 99), Version 2

(May 00) and Version 3 (Jan 01)
produced by ADB funded project
6.2 HDM-4 sponsors fund on-going
management and development of
strategy
ANNEX 4
Needs Assessment and Proposed Modular Structure for HDM-4

Train the Trainer Programme
4.1 Intention of Train the Trainer course

The intention of the Train the Trainer (T3) programme has to be two fold:
1. To provide the trainers with appropriate technical knowledge, skills and

resource materials to enable them to bring their trainees up to an appropriate
technical standard (Level 1 – Ability to train others); and
2. To provide the trainers with set of training methods and skills which are
appropriate for the target groups they will be training and compatible with the
technical content of the course.
A vital aspect of the T3 programme must be that the training methods used in the
technical inputs should be models of good practice for the trainees to adopt. All the
evidence is that trainers will teach as they are taught. We must ensure that this is not
a case of “do as we say - not as we do”. This will need inputs to the course on
training methods and approaches.
e.g.
The learning process
Target group profiling and training need assessment
Using training objectives to define training methods and content
Planning training sessions
Participatory learning methods
Communication skills, learning aids
The reflective practitioner - Programme feedback, evaluation and improvement.
4.2 Trainer Characteristics

Considerable practical experience of HDM-III and /or similar applications
Good communication skills
Evidence of good network / linkages in country
Strong support from employer / institution etc
Aware of potential training opportunities in country
Previous training experience desirable
Enjoys professional respect of peers
4.3 Modules and expected training needs

The following tables give a brief description of the proposed technical modules
(draft), with key contents, the expected existing competencies of the participants in
these subjects. and proposed ability of the participants on completing the modules.
A4-1
Note that these include only the technical modules, and do not include details on the
modules for training the participants to train. Such training modules will be detailed
in due course; however they will clearly only be applicable to the T3 course. The
following modules will subsequently be developed, however, to be applicable for
future courses held by the trainers in their own countries, and for other HDM training
purposes.
In addition, note that whilst the participants at the T3 course will already have
experience in the use of HDM-III, the following modules will be prepared with the
assumption of little knowledge of HDM (although a general background in highway
management and engineering). This is because these materials will then be used by
the trainers and pared back for the less competent trainee. Hence at the T3 course the
training materials for each module, as well as being used to teach the participants
themselves, will also provide a means of introducing the participants to the materials
from which they will draw for their own courses in the future.
Module Title Road management

Key contents Two submodules:
(1) Institutional requirements for efficient road management – scope of
road management, hierarchy of factors affecting sustainability, institutional
analyses, exercise on preliminary assessment of situation in their own
countries;
(2) Management cycles and the contribution of HDM – defining the
management cycles, identifying and giving examples of different analyses
required in planning, programming and preparation.
Expected existing Trainers will have considerable experience of road management practice
competencies before T3 and will have a knowledge of the breadth of issues influencing decision
course making. However, they might not have had the opportunity to structure
their understanding of this area, or to be clear as to the fundamental aims of
the various processes involved in road management.
Proposed ability on (1) A clear understanding of the institutional issues, and a framework to
completion of T3 course consider how sustainable road management practices might be
implemented over time.
(2) A clear understanding of the role HDM can play in road management,
the purpose of various analyses and how such practices might be
implemented over time.
Module Title Background to HDM

(1) Overview of history of HDM and principles – the original studies and
subsequent developments, types of models, the fundamental parameters –
roughness, traffic etc., the need for calibration;
(2) Logical framework of HDM analyses – the modular structure of HDM-
4, defining analyses (pavement types etc.), the order of modelling
parameters.
Expected existing Trainers will have experience using HDM and will therefore be aware of
competencies before T3 the background, principles and general approach of the system. However,
course their knowledge will need to be structured and a thorough revision will be
required.
Proposed ability on Understanding of subject, so that they are able to discuss strengths and
completion of T3 course limitations of the HDM approach.
A4-2
Module Title Road investment appraisal
Key contents Three submodules:
(1) Economic analysis – definitions, components of total transport cost,
types of analyses, difference between financial and economic, different
indicators (NPV, IRR etc);
(2) Risk and sensitivity analysis – definitions and the need, examples and
experience of key variables;
(3) Advanced analyses including different optimisation techniques –
different techniques for ranking competing projects, their role in the
context of broader multi-criteria analyses
Expected existing As users of HDM, trainers will have an understanding of economic
competencies before T3 analyses. However, they might not be aware of the more advanced
course application of risk and sensitivity analyses, or of the various optimisation
techniques.
Proposed ability on Full understanding of the subject. Experience will be gained in application
completion of T3 course during the case studies and worked examples (project, programme and
strategy analyses modules below).
Module Title Data requirements

Key contents Four submodules:
(1) Types of data required and information quality level principles – the
different data types, banding data for coarser analyses, matching data
complexity to accuracy expected of outputs;
(2) Dealing with other databases and systems (introduction) –
incorporating HDM modules with other systems, exporting/importing data,
use of models within other systems;
(3) Dealing with other databases and systems (advanced);
(4) Case studies.
Expected existing Local practice and judgement appear to drive data collection techniques
competencies before T3 more than rigorous analysis of cost-benefits. It is expected therefore that
course most participants will not have had the opportunity to consider these
aspects. Many participants might also have experience in using existing
systems, and it is expected that there will be quite a demand to learn how
the HDM-4 models and techniques might integrate with those existing
systems.
Proposed ability on A knowledge of the recommended principles for assessing which data are
completion of T3 course the most important to collect (assuming budget constraints on data
collection) and the required accuracy and sampling rates etc. It will not be
possible to fully cover the aspects of integration with other databases, but a
clear understanding of the key issues and experience from case studies will
be expected.
Module Title Adapting HDM to the locality

Key contents Three submodules:
(1) Configuring HDM – different parameters which need to be configured
for analysis – climate, types of vehicle etc.;
(2) Calibration of HDM – coarse through to long term studies;
(3) Case studies
Expected existing Most participants will be aware of the various aspects required to configure
competencies before T3 the system. However, less will have been involved in calibration – whilst
course they should be aware of the principles.
Proposed ability on Ability to calibrate HDM to a particular environment. Understanding of
completion of T3 course the approach to calibration and ability to develop longer term studies to
carry out more detailed calibrations.
A4-3
Module Title Traffic
(1) Types of traffic and their effects – different types of traffic, volume,
different vehicle types (complexities of different parameters related to each
vehicle for use in RUE models discussed in RUE module), axle loads,
fourth power law, traffic growth;
(2) Case studies – on the transformation of data from traffic studies into
HDM
Expected existing Participants should be well aware of these concepts from HDM-III
competencies before T3
course
Proposed ability on Full knowledge of all aspects, and ability to manipulate data for use in
completion of T3 course HDM-4.
Module Title Modelling road deterioration and works effects

Key contents Four submodules:
(1) Bituminous roads;
(2) Gravel and earth roads;
(3) Concrete roads;
(4) Works effects concepts: hierarchies, strategies and policies.
Expected existing Participants will probably have experience on the predominant road types
competencies before T3 in their own countries. It is unlikely that they will know of the concrete
course roads models, since these are new to HDM-4. Similarly, they will have
experience of using the HDM-III works effects, but the concepts behind
these will need to be clearly explained.
Proposed ability on Knowledge of the various models, and the basis from which they have been
completion of T3 course derived.
Module Title Modelling road user effects

Key contents To be confirmed
Expected existing
course
Proposed ability on
completion of T3 course
Module Title Modelling environmental and safety affects

Key contents To be confirmed
Expected existing
course
Proposed ability on
completion of T3 course
Module Title Project analyses

Key contents A minimum of four submodules to cover the different sorts of project
analysis.
Expected existing Participants will have some experience of particular type of analyses, but
competencies before T3 not necessarily a knowledge of the breadth of analyses of which HDM-4 is
course capable.
Proposed ability on Ability to design and develop analyses for the breadth of possibilities
completion of T3 course within their own countries.
A4-4
Module Title Programme analyses
Key contents A minimum of four submodules to cover the different sorts of programme
analysis.
course capable.
Module Title Strategy analyses

Key contents A minimum of four submodules to cover the different sorts of strategy
analysis.
course capable.
A4-5
TITLE: Priorities in the collection of information for use in UNIT:
road management.
SHORT TITLE: Information strategies
Primary Module: Road and Transport Information
Duration: Presentation 1 hour
Target Group: Engineers/Planners
Trainer Notes Hints and Tips

AIM
To teach the a structured, priority based approach to the

selection of key data and models to support cost effective
decision making
OBJECTIVES
The trainees will:

1. Define what data is essential to support road management
tasks.
2. Explain the concept of information quality levels and its
application in the focusing data collection and calibration
priorities.
3. Describe what is meant by impact elasticity and its relation
to the prediction of outcomes of importance to road
management decision making and performance.
4. Identify what constraints may exist in data collection
assessing needs, and how might you address them?
5. Develop and defend a strategy for the collection of key
data required for the different stages in road management.
RELATED UNITS/MODULES
i) Information types.
ii) Calibration of RDWE and RUE.
iii) Modelling RDWE and RUE.
iv) Configuration and implementation of HDM-4.
INTRODUCTION
This session is best taught by linking the aim to two main issues,
namely the difficulties of sustaining data hungry decision
support tools and the overall sensitivity of decisions to key
parameters.
The presentation should begin by emphasising the aim, and how

a structured, cost effective approach is essential.
The practical experience of the trainees should be drawn out by

inviting group participation to provide feedback to specific
questions, including:
• Should the same data be collected at all stages in the

planning process?
• If not, give examples from local practice?
Version 1.0 1 HDM-4 Technical Training Resource

The relationship with sessions on HDM-4 configuration,
information types and the modelling and calibration of RDWE
and RUE should be emphasised.
DEVELOPMENT
A suggested scope and the order and duration of the various

parts are given below. Indicative parts and timings are as
follows:
i) Introduction to information quality levels and their

relevance to road management (30 mins with
participation).
ii) Explanation of impact elasticity and the outcomes
impacted by relative changes in the value of particular
parameters (30 mins with participation).
iii) Presentation of case study examples (30 mins with
participation).
iv) A short practical to demonstrate the sensitivity of key
dependent parameters, or decisions, to changes in input
data (1 – 2 hours, if models incorporated in a simple
spreadsheet or when using an existing case study in
HDM-4).
v) Task work and trainee presentations on ‘a sustainable
data collection strategy’ (2 hours).
The level of complexity could be increased should time permit

and could build from simple sensitivity tests which examine the
impact of changes in input variables on perhaps quantity
estimates for a given distress type to examining changes in total
transport costs or economic ranking criteria.
Having completed a series of sensitivity tests, the next step

could involve a review of current data practices by comparing
them with the recommendation given in the HDM-4
documentation (Vol. 5). This would then prepare the trainees
for their presentation ‘to develop and defend a sustainable data
collection strategy’.
Should computer techniques be employed the trainees should be

provided with prepared data either within HDM-4 or perhaps in
a spreadsheet. Operational skills should exist prior to this
session.
This session is strongly related to the modelling of RUE and

RDWE and should preferably follow these and the unit on
calibration.
SUMMARY
The initial presentations (i) to iii) above) should be summarised

in 3 or 4 points, giving emphasis to the main points and
stressing how these might be investigated further by the trainees
in practical sessions.

The practicals should include reporting back from the trainees
and summing up by the trainer.
The trainer should be prepared to test the outcome of the

trainees’ exercises in comparison with HDM-4 documentation.
In so doing he/she should also be aware that different results
could arise, but should also be prepared to check or otherwise
challenge the results bearing in mind the wide variety of
outcomes found as a result of reputable case studies.
RESOURCE MATERIALS
Hardcopy (including Manuals):

1. HDM-4 documentation, Volume 5. A guide to the
calibration of HDM-4.
2. Selected research papers.
Visual Aids:
1. Powerpoint presentation (15 – 20 slides).
Case studies/Practicals:
1. As required.
TRAINING FACILITIES
Powerpoint projector, white board/flip chart and overhead

projector/slide projector (optional).
MISCELLANEOUS

TITLE: The role and importance of investment appraisal in UNIT:
Highway management
SHORT TITLE: Investment appraisal for Highway management
Primary Module: Transport Planning and Economics
Duration: 1 hour plus 1 hour practical
Target Group: Engineers / Planners

AIM:
To teach the definition and components of economic analysis
relating to road investment appraisal so that trainees can
interpret the results of HDM-4 analyses effectively.
OBJECTIVES:
The trainee will:
1. Identify decisions which have to be made in the road sector
2. Explain the principle of total transport costs, including
costs and benefits
3. Describe the concept of life cycle analysis, including
discounting
4. Explain the difference between economic and financial
costs
5. Identify primary and secondary benefits
6. Explain the concept of base case and do-minimum
comparisons
7. Describe the differences between estimation of benefits
from Normal, Diverted and Generated traffic
8. Describe the key economic indicators (NPV, IRR, NPV/C)
and their use in investment appraisal
i) Operation of HDM-4 / Modelling framework
ii) HDM-4 Analyses
iii) Overview of Highway management
INTRODUCTION
Appraisal of the funds being spent on the Highway network is
essential to ensure the most effective use of resources. As such,
an understanding of the appraisal methodology and economic
analyses carried out in HDM-4 is key to effective investment
appraisal. However, phrases such as “investment appraisal” and
economic analyses” are not naturally of interest to many
engineers. Hence the relevance to their work needs to be made
to obtain their interest.
It is important to draw out the experience of the trainees and it
may be useful to ascertain the background of the trainees
(engineering / economics/planning). This could assist is
selecting the right words to motivate the trainees.
This session links to sessions on the modelling framework of

HDM-4 and highway management.
The presentation can start with slides and conclude with a

practical hands-on experience with a case study to examine and

interpret the results through the economic indicators. This will
reinforce the learning and provide ability to perform HDM-4
analyses.
A question and answer technique can be used to draw out some

of the frequently posed concerns of highway network managers.
• Why should the Highway Authority carry out appraisal

of road projects? This is aimed to draw out ideas of fund
adequacy, value for money, taxpayers money.
Prompts could be:
• Does the Highway authority have sufficient budget for
all their needs? If not, how do they decide which
projects to do?
• Are you happy, as a taxpayer if decisions on allocation of
funds are made on a random or ill-informed basis?
• What happens if roads are not maintained to an
adequate standard? This is aimed to draw the response of
decreasing roadworthiness and hence increased running
costs and road user costs, which they also pay for – hence
their money.
Hence make the conclusion that appraisal of projects on an

economic basis by a standard method (such as HDM-4) can be a
beneficial tool and that this session will aim to show them how
to do this with HDM-4.
DEVELOPMENT
The purpose and scope of the Unit should be described together
with the order and duration of various parts and practical
sessions. Indicative parts and timings are as follows:
i) Overview (10 minutes)

ii) Analysis logic (life cycle, base case) (10 minutes)
iii) Prices, costs and benefits (Primary & Secondary) (15
minutes)
iv) Benefits from traffic (10 minutes)
v) Economic indicators (15 minutes)
vi) Case study practicals (60 minutes)
The session could start with a number of questions posed by

engineer/planners in the Highway sector and decisions which
they have to take. This could be followed by confirming the key
elements of an economic analysis such as costs and benefits.
The concept of total transport costs here will reinforce the aim
to seek an optimum standard and of including all relevant
costs/benefits.
Additionally the concept of life-cycle analysis should be
introduced or reinforced. This will introduce the concept of
discounting costs to a base year for economic comparison.
Economic principles of prices (economic and financial) should

be introduced. A reference to the location of such data as input
values in HDM-4 (Road User and works activities) should be

made. The main project costs and benefits can then be identified
(possibly by asking questions from trainees) and a distinction
made between primary and secondary benefits, making clear the
potential dangers of double counting. This will lead into the idea
of with and without project and do minimum / do something for
economic comparison.
Traffic is a key parameter which affects benefits and the way in

which increases in traffic resulting form a project investment
must be correctly dealt with. The 3 types dealt with in HDM-4
(Normal, Diverted and Generated) should be explained along
with the different methodologies for assigning benefits to these
traffic.
The different criteria calculated and used to make economic

decisions should be explained. An example here could be shown
of a project (base case and alternative) one with regular costs
throughout the project and the other with high initial costs but
lower subsequent maintenance costs, showing the NPV value
for the project. The choice of economic indicator will depend on
the type of analysis being carried out and this should be shown
with reference to the criteria given in the HDM-4 Analyses.
A practical session could be used to demonstrate the economic

indicators calculated in HDM-4 and their interpretation.
A case study of a project analysis examining a number of

maintenance or improvement alternatives could be run and the
results examined. The trainees could
a) vary the discount rate to see how the discounted costs
change
b) amend one of the maintenance/improvement options to
delay a major expenditure by several years. Reports in
Cost Streams
SUMMARY
Separate summaries and reporting back should be invited for at
the end of each presentation / practical session.
The summary could be done by a series of questions inviting

key points from the trainees to recapping on the main objectives
of the session. This would provide confirmation of the training
objectives.
The importance of the economic analyses in the appraisal should

be reiterated in relation to its impact on the road network
management.
Trainees should be advised of any follow up or related sessions

on HDM-4 analyses.

RESOURCE MATERIALS
1. HDM-4 documentation, Volume 4, Part G
2. HDM-4 documentation, Volume 5, Appendix G
3. PowerPoint presentations as additional handouts.
Visual Aids:
5. PowerPoint presentation (15-20) slides including photos)
1. HDM-4 Case Study
TRAINING FACILITIES
PowerPoint projector, white board/flip chart, overhead
projector/slide projector (optional) and computers.
MISCELLANEOUS

TITLE: The role and effect of road user costs and benefits UNIT:
in HDM-4
SHORT TITLE: Road User effects in HDM-4
Duration: 1 - 2 hours plus practical of 2 hours

AIM:
To teach the modelling of road user effects in HDM-4 so that
trainees will know the contribution of road user effects to
investment appraisal within HDM-4 analysis and be able to use
and interpret the results of the model effectively.
OBJECTIVES:
The trainee will
1. Explain the contribution of road user effects to investment
appraisal, including costs and benefits.
2. Describe the main modelling approach to RUE in HDM-4
and the use of unit costs
3. Define the key components of RUE and their relative
effects on total RUE
4. Recall the location of input parameters for RUE in the
vehicle fleet
5. Describe the speed input requirements for the speed model
6. Recall impact of traffic congestion / traffic interaction on
RUE (fuel, tyres)
7. Describe the impact on tyre consumption, parts and labour
costs
8. Explain the options for calculation of utilisation and service
life
9. Describe the way in which HDM-4 deals with vehicle
depreciation and interest
10. Describe the effect of travel time components and their
contribution to RUE
11. List the effects of non-motorised traffic.
ii) HDM-4 Analyses
iv) Traffic effects
v) Vehicle classification

INTRODUCTION
This session aims to describe the contribution of different Road
User effects to HDM-4 analyses and the way that HDM-4
models these effects. This unit links in with sessions on vehicle
classification, traffic data effects and investment appraisal. The
importance of RUE in terms of generating the main benefits to
traffic from a road investment or maintenance strategy needs
reinforcing strongly.
A variety of learning aids, including pictures and graphics, can

be used to attract the attention of the trainee and convey the
effects being described. The presentation can start with slides
and conclude with a practical hands-on experience with a case
study to examine the effects of different road user parameters.
This will reinforce the learning and provide ability to perform
HDM-4 analyses.
The modelling of RUE is quite complex and the trainer will

need to decide the level of detail that he wishes to pursue. If it is
required to teach the modelling detail of each element, the
session could be separated into a number of sub-sessions
A question and answer technique can be used to draw out the

relevance of RUE by a series of questions such as:
What do we mean by road user effects? Who are the Road

users? What costs do they incur by using the road? This could
draw out people’s perception of different costs and could draw
attention to costs which may not be considered relevant because
they are not “apparent”.
Are costs likely to be the same for all vehicle types? This is
aimed to draw out ideas on different costs (and hence) benefits
to transporters or buses compared to car owners and the idea of
commercial use and value of time.
DEVELOPMENT
sessions. Indicative parts and timings are for short or long
sessions are:
i) Overview including relative contribution (15minutes)

ii) speed and congestion effects (10/20 minutes)
iii) vehicle running costs (10/20 minutes)
iv) vehicle standing costs (10/20 minutes)
v) non-motorised traffic (10/20 minutes)
vi) travel time (10/20 minutes)
vii) practical – case study (2 hours)
The session could start by reiterating the key contribution of

vehicle operating costs and road user benefits in an investment
appraisal tool such as HDM-4. The main factors influencing

RUE should be listed with an indication of their relative
contribution to the User benefits in order to indicate the most
important areas.
An outline should be given of the computational logic within the

RUE model to provide a framework for the following discussion
on the different parameters. The HDM-4 architecture could be
shown here and flow chart to put the RUE model in HDM-4 into
context. The concept of life cycle analysis and costs and benefits
(comparison with the do-minimum) should be revised. The
derivation of relationships based on quantities and not costs and
the use of Unit costs in the estimation of RUE should be
explained.
For an audience who have experience of earlier models, an

explanation of the key changes and new features can be made.
This should be referenced to the necessity for these changes.
Some are improved models of previously modelled factors,
other are new, such as models for non-motorised traffic, effects
of roadworks and accident costs, which will be dealt in another
session.
Speed & Congestion

Speed is an important parameter which is estimated by HDM-4.
The trainees could be invited to describe key elements of RUE
which could be affected by speed - travel time, fuel and tyres.
The modelling approach to speed estimation could be described
explaining the concept of desired speed and constraining speeds.
The detailed effects on speed of geometry, surface condition,
vehicle characteristics, traffic interactions can be described in
broad terms or in details depending on the audience, together
with calibration of the speed model to local conditions. The
effect of vehicle interactions on speeds and acceleration-
volume effects can be described. The concept of PCSE should
have been introduced in another session as well as the use of
speed flow types and traffic flow distribution.
vehicle running costs (fuel, tyres, oil, maintenance)

The interaction between the RUE effects can be described, an
example of which is how speed and speed flow have an effect
on fuel consumption and tyres. Key points to make here are the
ability of HDM-4 to deal with retreads and studded tyres and the
importance of vehicle maintenance (parts & labour) in terms of
their contribution to overall RUE and the need for calibration.
vehicle standing costs (capital costs: depreciation, interest,

overheads)
Vehicle utilisation and service life can be introduced and the
choice of methods for estimating service life (constant life and
optimal life) described. It should be clear that interest is that
related to the purchase of a vehicle. Overheads are given as a
fixed cost (insurance, garaging, etc)
Time costs
The importance of travel time in total RUE should be stressed.

This includes passenger working & non-working time and crew
time. It is dependent on vehicle speed and proportion of
commercial use. These latter are input values and are estimated
for NMT as well as MT.
Effects of NMT
These can be described. In addition to having an impact on MT
in terms of congestion, RUE are estimated for NMT in a similar
manner to those for MT. However, many of the models are
conceptual and need further validation.
Other effects such as safety are the subject of a separate unit.
The role of calibrating vehicle data should be reinforced here.

Volume 5 table 6.1 shows the recommended priority for
calibration of RUE parameters. The importance of this will be
highlighted by the case study practical.
Case study practical

A practical session could be used to demonstrate the relative
effects of
a) change of RUE against changing road condition (eg
increasing roughness)
b) difference in RUE for main vehicle types (eg car, truck, bus)
c) breakdown of costs for one or two vehicle types to show
relative importance
A gravel road section in good condition with low maintenance

would show an increasing change in roughness over a period of
years. A maintenance strategy of improvement to a paved road
after 10 years would show the different RUE on the two types of
road surfaces (roughness the key). Trainees should be
encouraged to plot the results to show the trends. Reports in
Road User Cost (summary by vehicle) and Deterioration
(Annual Road Condition - for roughness)

SUMMARY
The session on Road User Costs should be summarised by
recapping on the main objectives of the session. This would be
best achieved by asking a series of questions, which would
provide confirmation of the training objectives. If a detailed
description of the relationships is given, then it may be
appropriate to invite separate summaries for each section.
The summary could be done by a series of questions inviting

key points from the trainees or be pre-prepared. The former is
preferred as it provides confirmation of the training objectives.
The importance of the road user effects data should be reiterated

in relation to its impact on the investment appraisal / road
network management. The areas which have the greatest impact
on overall costs should be clear.
Trainees should be advised of further related sessions on road

user costs, if appropriate.
RESOURCE MATERIALS
1. HDM-4 documentation, Volume 4, Part E
2. HDM-4 documentation, Volume 5, Appendix E
3. HDM-4 documentation, Volume 7 (Draft).
Visual Aids:
1. Short PowerPoint presentation (10 slides including photos
per section)
2. Long PowerPoint resource (As required)
1. HDM-4 Case Study
TRAINING FACILITIES
MISCELLANEOUS

TITLE: A framework for sustainable road management UNIT:
SHORT TITLE: Road management
Primary Module: Road Management
Duration: Presentation 1-1.5 hrs Practical (if necessary) 1 – 1.5 hrs
Target Group: Senior Engineers and Planners (Could be adapted for PolicyAdvisors)

AIM
OBJECTIVES
1. etc
i) etc
INTRODUCTION
DEVELOPMENT
SUMMARY
RESOURCE MATERIALS

1. etc
Visual Aids:
1. etc.
1. etc.
TRAINING FACILITIES
MISCELLANEOUS

SHORT TITLE: Configuring HDM-4 to suit local conditions UNIT:
TITLE: Adapting HDM-4 to model local conditions by specifying default data for climate,
traffic and pavement characteristics
Primary Module: Role and Implementation of HDM-4 in Road Management
Duration: 1 - 2 hrs Presentation, 4 - 6 hrs Practical

AIM
To demonstrate best practice in adapting HDM-4 to local
conditions so that participants can implement appropriate
default data for use in their countries.
OBJECTIVES
The trainees will:
1. Define the key characteristics of their country’s road
network such as functional classification, pavement types,
geometric design features, capacity, etc.
2. Define the key characteristics of the traffic using the road
network such as the variation in daily flow, axle loading,
etc.
3. Describe the climatic regions within the country that affect
pavement performance.
4. Define the default pavement condition attributes that are
commonly found in their country.
5. Define the currency name and symbol used in their country.
6. Adapt HDM-4 default data sets that suitably represent the
road and traffic characteristics in their country.
1. Information Strategies
2. Modelling Framework
3. Modelling RDWE & RUE
4. Calibration of RDWE and RUE
5. Software Structure, Installation and Operation
6. Road Networks and Vehicle Fleets
7. Implementation of HDM-4
INTRODUCTION
The Configuration module in HDM-4 is essentially a tool for

managing the default data set. The data will be used repeatedly
often under the assumption that the assigned values have been
verified and are correct for the country where HDM-4 will be
used. Consequently, the default data should be carefully selected
and trainees need to be made aware of the vital importance of
doing so.
The unit is best taught through a combination of presentations

and hands-on exploration of the Configuration folder in HDM-4.
The key concepts that need to be highlighted are the relationship
between aggregate (or class) data (e.g. good, fair, poor) and the
underlying default data that are assigned to these. For example,
Good pavement condition translates into several HDM-4

parameters (i.e., cracking, ravelling, potholes, edge-break, and
rutting). This concept applies to all user defined aggregate
variables (riding quality, traffic bands, structural adequacy,
etc.).
The PowerPoint slides should be used to introduce the concept

and describe the required parameters. This should run
concurrently with a hands-on practical on creating and editing
the aggregate variables together with the required default data.
The following topics should be described in detail:
• Climate Zones
• Traffic Flow patterns
• Speed Flow types
• Aggregate variables and the corresponding default data
• Currency definition
DEVELOPMENT
The trainer should start by giving an overview of the

Configuration module using the PowerPoint slides and show the
location of the Configuration Folder in the HDM-4 software.
Trainees should open the HDM-4 Workspace on their
computers and open the Configuration folder.
Thereafter, each topic should follow the same pattern:

PowerPoint presentation by the Trainer(2 ~ 4 slides each), then
hands-on data entry by the trainees for the following:
• Climate Zones
• Traffic Flow patterns
• Speed Flow types
• Aggregate variables and the corresponding default data
• Currency definition
It is estimated that each topic will require 3 minutes of

presentation followed by 12 minutes of hands-on data entry.
More time should be allowed for describing the Aggregate

variables and the associated default data tables.
Total time required 1.5 ~ 2 hours.
It is important that the Trainer checks that each participant has

correctly completed the full set of default data required for the
aggregate variables. The HDM-4 software will not operate
properly if there are missing default data. For example, if a new
Road Class is introduced, it is essential to ensure that all
required parameters have been defined (e.g. traffic flows,
pavement strength, etc.).
The values of default data assigned to each aggregate variable

should be logical. For example, if riding quality is defined as
Good, Fair or Poor, then the IRI values assigned to these should
follow a logical sequence.

SUMMARY
At the end of the Unit session, participants should be able to

locate the default data associated with any aggregate variable.
It is important to conclude the session by advising participants

to regularly review the default data values and update these
periodically. For example, traffic band limits defined for each
road class may need to be reviewed annually.
RESOURCE MATERIALS

1. Handout notes on HDM-4 Configuration
2. HDM-4 Volume 4: Analytical Framework and Model
Descriptions
3. HDM-4 Volume 2: Applications Guide
4. HDM-4 Volume 3: Software User Guide
Visual Aids:
1. PowerPoint Slides (HDM-4 Configuration.PPT)
1. HDM-4 Case Studies Workspace
TRAINING FACILITIES
1. PowerPoint projector or overhead projector/slide projector,

2. White board or flip chart,
3. Computers (preferably for each participant).
MISCELLANEOUS
Filename: T3 Configuration Unit Notes.doc

SHORT TITLE: Implementing HDM-4 within Road UNIT:
Management Systems
TITLE: Implementation of the HDM-4 software within road management systems to
achieve seamless integration
Primary Module: Role and Implementation of HDM-4 in Road Management

AIM
To demonstrate methods of implementing HDM-4 as part of a
road management system so that a seamless integration can be
achieved.
OBJECTIVES
The trainees will:
1. Describe the methods of data collection used in their
countries.
2. Describe the database management system used to store
road network inventory and condition attributes.
3. Describe the methods of data exchange between HDM-4
and road information databases.
4. Define the decision support systems required by road
managers.
5. Identify the role of HDM-4 as a decision support system for
road managers.
6. Define the management information that can be produced
using HDM-4.

2.
INTRODUCTION
This unit puts into practical context the role that HDM-4 plays
within a road management system. HDM-4 is essentially a
decision support tool that can be used to process data collected
about the road network, the environment, the traffic, and other
road users in order to produce pertinent management
information. HDM-4 does not constitute a road management
system on its own. The above points need to be stressed during
the delivery of this unit.
The implementation is a four stage process comprising:

• Data collection: It should be noted that HDM-4 can be
adapted to work with any method of data collection used.
Whilst the various HDM-4 models contain several hundred
parameters, most of these will inherit default values (as
defined in the Configuration module). Data transformation
rules can be applied to convert the data collected to the
parameters required in HDM-4.
• Database management system: HDM-4 can be linked to

any database that can produce a data exchange file to a
defined format. Trainers should point out that the internal
database within HDM-4 has not been designed to perform
this function. A separate database management system is
envisaged for this.
• Decision support system: The role of HDM-4 is to process
data imported from the database and produce work
programmes, long term strategies and project level
economic indicators. Note that other decision support
systems may already exist in some countries, consequently
HDM-4 could be used in such cases as a control or
reference tool against which other systems can be
compared.
• Management information: The key management
information produced by HDM-4 are economic indicators
for project level economic indicators, optimised work
programmes, and network wide performance trends and
budget requirements for strategy analyses.
Each stage of the above process should be described giving

practical examples from some countries, e.g. the implementation
in Bangladesh or other countries.
At the end of the presentation, participants should be asked to

describe the road management systems in their countries and
identify whether HDM-4 has a role to play in this.
DEVELOPMENT
This unit should be presented using the PowerPoint slides with a

short demonstration of the HDM-4 software to show the data
exchange import/export functions. The PowerPoint slides
present the four stages of the process in a logical manner with
detailed information on each of the stages:
• Data collection
• Database management system
• Decision support system
• Management information
It is estimated that each stage will require approximately 10

minutes of presentation (including questions from trainees)
followed by a demonstration with trainees using the HDM-4
software.
The demonstration should take the trainees through the HDM-4

menu / Export / Import functions. Note that the Export function
can also be started using the button in the Workspace. All data
items in the Workspace (i.e. Road Networks, Vehicle Fleets,
Work Standards, Projects, Programmes, Strategies and
Configuration) can be individually exported.
Total time required 1 ~ 1.5 hours.

SUMMARY
The session could conclude with the participants being asked to

describe the road management systems used in their countries.
Alternatively, where the training is being conducted for one
country, this could focus on the methods of data collection,
characteristics of the database management system, the decision
support tools and the management information used by the road
agencies.
RESOURCE MATERIALS

1. Handout Notes on HDM-4 Implementation.
Visual Aids:
1. PowerPoint Slides (HDM-4 Implementation.PPT)
1. HDM-4 Default Case Studies.
TRAINING FACILITIES
1. PowerPoint projector or overhead projector/slide projector

MISCELLANEOUS
Filename: T3 Implementation Unit Notes.doc

SHORT TITLE: HDM-4 Modelling Framework UNIT:
TITLE: Introduction to the HDM-4 concepts and modelling framework
Primary Module: Role and Implementation of HDM-4 in Highway Management
Duration: 1 ~ 1.5 hour Presentation

AIM
1. To introduce the modelling framework used in HDM-4 so that the
HDM-4 applications can be described accurately by trainers.
OBJECTIVES
The trainees will:

1. State the main applications of HDM-4 and their relevant functions
in road management.
2. Explain the analytical framework in terms of the modelling units,
e.g. road deterioration, road work effects, road user effects, etc.
3. Describe the concept of homogeneous sections, road links, sub-
networks and networks.
4. Describe the pavement and vehicle classification system (i.e.
pavement and vehicle types).
5. Describe the hierarchy embodied in the analysis loop.
6. Illustrate the key concept of section alternatives through a flow
diagram.
1. Overview of HDM-4
2. HDM-4 Software Structure and Installation
3. HDM-4 Configuration
5. Road Work Standards
6. HDM-4 Project Analysis Application
INTRODUCTION
It is important that the modelling framework is described to Trainees

in such a way that they appreciate the complexity of trying to model
the future performance of roads and the impact on road users. HDM-4
essentially comprises a set of mathematical models that have been put
together in a coherent analytical framework in order to provide useful
road management information. The mathematical models are for
predicting road deterioration and the effects of road works, for
calculating the effects of road characteristics on road user costs, and
for calculating other socio-economic effects (energy, emissions, noise,
etc).
The HDM-4 models are applied to homogeneous road sections. This

assumes that pavement deterioration will be the same on the entire
road section. Consequently, road user effects will be constant for that
section. Characteristics of the road section are defined in terms of the
pavement type that is user defined and the type of vehicles using the
road. The modelling framework applies the concept of life cycle
analysis to calculate over a number of years (typically, 15 ~ 40 years)
the road deterioration, works effects and road user effects for each
section-alternative.

The above concept is the key to understanding HDM-4. The
applications (Project, Programme and Strategy) are simply multiples
of the analytical framework, i.e. a project analysis comprises one of
more homogeneous sections-alternatives, programme analysis
comprises several projects and strategy analysis is simply a model of
the entire road network broken down into representative sections.
DEVELOPMENT
This session provides a background to the modelling framework

and concepts adopted in HDM-4. It should be presented using
the PowerPoint slides prepared for this unit. There will be little
use of the HDM-4 software during the session. Trainees should
be given the opportunity to ask questions, hence it is suggested
that the presentation should last approximately 40 minutes with
an additional 20 ~ 30 minutes of questions from trainees.
SUMMARY
This session is designed to present the fundamental concepts

and modelling logic adopted in HDM-4. It is therefore essential
that the material is carefully presented. Trainers should spend
sufficient time beforehand preparing for this session.
RESOURCE MATERIALS

1. HDM-4 Volume 4: Analytical Framework and Model
Descriptions.
2. Handout (T3 Modelling Framework Handout Notes.DOC)
3. PowerPoint Slides (T3 Modelling Framework.PPT)
Visual Aids:
1. HDM-4 default Case Studies
TRAINING FACILITIES

MISCELLANEOUS
Filename: T3-UN Modelling Framework.doc

SHORT TITLE: Derivation of optimum maintenance UNIT:
standards
TITLE: Use of HDM-4 to develop optimum maintenance standards that minimise total
transport costs
Primary Module: Cross-Cutting Case Studies & Applications
Duration: 1 hour Presentation and 1.5 ~ 2 hours Practical

AIM
To describe the specification and assignment of HDM-4 maintenance

standards so that Trainers are able to apply a range of maintenance
works.
OBJECTIVES
The trainees will:

1. Prepare input data required for investigating optimum road
maintenance standards for paved roads using HDM-4.
2. Assign a range of maintenance standards to road sections and run
HDM-4.
3. Interpret the results and recommend optimum maintenance
standards for paved roads.
6. HDM-4 Programme Analysis Application
7. HDM-4 Strategy Analysis Application
INTRODUCTION
This case study presents the economic analysis of alternative

maintenance standards for two paved road sections that are 36.5 km
and 10.6 km long, respectively. The existing roads have surface
dressing and asphalt concrete pavement surfacing in good condition,
and carry approximately 6 000 and 16 000 vehicles per day (AADT),
respectively. The objective of this case study is to determine the
optimum maintenance intervals for resealing (surface dressing) and for
overlays on the two roads, respectively.
The case study data is provided in the T3 Case studies

workspace/database installed with the HDM-4 software. The data for
this case study are located in the Projects folder in the Case Studies
Workspace. The name of the case study is Case 2: Optimum
maintenance intervals. To locate the data for this case study:
! Select the T3 Case Studies Workspace
The case study data includes the road network data set (CASE2
Project Sections) and the vehicle fleet (CASE2 Vehicle Fleet)
required for the analyses.
The first step in the procedure would be to calibrate the road sections

using the calibration factors derived from the previous case study. The
second step in the procedure should be to create additional
maintenance standards for the two roads. The data provided includes
one example each for:
! Surface dressing at 5 year intervals (scheduled), and
! Overlay 50mm at 6 IRI (responsive)
Create additional maintenance standards by copying the above to

create separate standards for:
! Surface dressing intervals (scheduled) at 3, 7, 9 and 11 years
! Overlay intervals (responsive) at 4, 8, 10 and 12 IRI
The optimum economic maintenance standards are defined as the

maintenance interval (scheduled or responsive) with the maximum
NPV. This can be illustrated by exporting the Benefit Cost Ratios
report to an Excel spreadsheet and plotting the NPV against the
maintenance interval (years) or intervention criterion (IRI).
Another useful plot would be of the NPV against the increase in
agency cost. The resulting curve from this is defined as the efficiency
frontier. This will be described in detail in the Programme Analysis
case study.
Those Trainers who complete the above exercise within a short time
could attempt to derive the optimum grading and regravelling
frequencies for the Gravel road that is included in the Road Network
database.
DEVELOPMENT
SUMMARY
RESOURCE MATERIALS

1. Project Case Study 2 Handout
2. Project Case Study 2 HDM-4 Workspace
3. HDM-4 Volume 2: Applications Guide (Appendix A)
4. HDM-4 Volume 3: Software User Guide (Chapter 7)
5. HDM-4 Volume 5: A Guide to Calibration and Adaptation
(Chapters 6 and 7)
6. T3 Case Study 1 Handouts
Visual Aids:
1. PowerPoint slides
TRAINING FACILITIES

MISCELLANEOUS

Filename: T3 OptimumStandards Unit Notes.doc

SHORT TITLE: Overview of the role of HDM-4 in road UNIT:
management
TITLE: The role of HDM-4 in performing management functions within a road management
system

AIM
1. To introduce the concept of HDM-4 road investment appraisal and

its applications in road management so that the HDM-4 system
can be properly applied within road management.
OBJECTIVES
The trainees will:

1. Describe the concept of life cycle analysis that is used in road
investment appraisal.
2. Describe the HDM-4 applications (Strategy, Programme and
Project Analysis).
3. Explain the role of the HDM-4 applications within the road
management cycle.
4. Illustrate the HDM-4 applications by using sample output.
1. Modelling Framework
2. Software Structure and Installation
6. HDM-4 Project Analysis Application
INTRODUCTION
This is a key unit that introduces the HDM-4 life cycle analysis
concept and its applications within road management. Life cycle
analysis involves modelling the future performance of roads together
with the impact of road characteristics on user costs. Various titles are
often used to describe the same concept; whole life cycle, total
transport costs, road investment appraisal, etc. It is important that
trainers should stress that these all refer to a common modelling
framework.
The road management process is covered in another Unit. Trainers

should refer to that unit in order to be able to link HDM-4 applications
with the management functions of planning, programming and
preparation. The operations management function is not explicitly
covered by HDM-4.
With regard to the HDM-4 applications (i.e. use of HDM-4 for specific
purposes), the three applications are; Strategy, Programme and Project
analyses. The three applications differ in terms of scope and data
detail. Strategy analysis deals with an entire road network with input
data that can be at an aggregate level (e.g. good/fair/poor pavement

condition, etc.). The objective is to model the performance of the road
network in response to different budget scenarios or maintenance
standards. Programme analysis deals with a selection of road sections
(i.e. a long list of candidate projects). The objective is to prioritise
these and select short list of projects that maximise benefits under
specified budget constraints. Project analysis deals with individual
road sections and models mutually exclusive alternatives in order to
estimate the economic benefits that would accrue from each
alternative. The three applications can also be used to conduct various
research and policy studies, for example to verify pavement design
standards, to model impacts of changing axle load limit regulations,
etc.
DEVELOPMENT
This Unit should be mainly presented using the PowerPoint

slides. It is estimated that the presentation takes 45 minutes with
an additional 15~30 minutes of questions.
SUMMARY
This unit provides an overview of the HDM-4 system and gives

the background to the HDM-4 applications. The unit should be
seen as the scene setter for HDM-4 functions within the overall
context of road management.
RESOURCE MATERIALS

1. Handout notes (T3 Role of HDM-4 Handout.DOC)
2. PowerPoint slides (T3 Role of HDM-4.PPT)
3. Road management cycle
4. HDM-4 Volume 1: Overview
Visual Aids:
1.
TRAINING FACILITIES

MISCELLANEOUS
Filename: T3-UN Role of HDM-4.doc

SHORT TITLE: HDM-4 Programme Analysis Applications UNIT:
TITLE: Applications of HDM-4 Programme analysis module for the preparation of annual
and multi-year work programmes
Module: Cross-Cutting Case Studies & Applications

AIM
1. To introduce the HDM-4 Programme Analysis module concepts

so that Trainees are able to prepare annual or multi-year work
programmes.
2. To introduce the concept of incremental benefit/cost ranking
method so that Trainers appreciate optmisation techniques.
OBJECTIVES
The trainees will:

1. Describe the concepts of the HDM-4 Programme analysis
application.
2. Apply the incremental benefit cost optimisation techniques
through simple manual calculation.
3. Use the HDM-4 model to generate a work programme.
4. Alter maintenance standards and unit costs to study the impact on
budget requirements.
6. Investment Appraisal
INTRODUCTION
This Unit should be taught as a two separate sessions. The first

session should focus on the theoretical aspects of programme
analysis, optimisation and multi-year work programmes. The
second session should be a practical case study illustrating the
concepts through a hands-on exercise.
Programme Analysis is concerned with the preparation of single

or multi-year road work and expenditure programmes under
specified budget constraints. The three factors which define a
work programme are:
• A selection of road sections which are candidate projects

• A specified time period
• A specified budget limit
The selection of candidate projects can be done using a defined

set of criteria, e.g. overlay when IRI > 6. Each candidate project

can be assigned one or more mutually exclusive alternatives.
Note that the alternatives are based on user defined maintenance
or improvement standards. Consequently, the optimised list is
essentially a prioritised list of pre-defined project alternatives. It
should be pointed out that HDM-4 does not internally optimise
within alternatives, e.g. it does not design the optimum overlay
thickness, but selects the best alternative from a amongst a set of
specified alternatives.
The optimisation or prioritisation applies the basic principles of

Capital Budgeting rules. This applies the Efficiency Frontier
concept to select project alternatives that maximise the
NPV/Cost ratio or it’s incremental. Other methods of
optimisation have not been included in HDM-4.
The main output is a prioritised multi-year list of road works

with the following information:
• Road section ID and name
• Length
• Recommended periodic maintenance or improvement works
• Total costs
• Cumulative cost of road works on all higher ranked sections
DEVELOPMENT
Two sessions are envisaged to cover the material. The first

session of about 1 hour is a based on a PowerPoint presentation
of the Programme Analysis and optimisation concepts. The
handout on capital budgeting should be used to illustrate the
optimisation technique used in HDM-4.
The second session should be a hands-on practical to

demonstrate use of the HDM-4 Programme Analysis application
to prepare a work programme using pre-defined Case Study
data. Trainers should review the default Programme Analysis
case study included in the HDM-4 installation package in order
to adapt these to suit local conditions.
SUMMARY
The two sessions should reinforce the Programme Analysis

concept. The session should conclude with a review of the
Programme analysis data produced after running the software.
RESOURCE MATERIALS

1. Handout Notes on Capital Budgeting rules
2. Handout Notes on Progamme Analsysis Case Study

Visual Aids:
1. PowerPoint Slides (Programme.PPT)
TRAINING FACILITIES

MISCELLANEOUS
Filename: T3 Programme Unit Notes.doc

SHORT TITLE: Derivation of optimum maintenance UNIT:
standards
TITLE: Use of HDM-4 to develop optimum maintenance standards that minimise total
transport costs

AIM
To describe the specification and assignment of HDM-4 maintenance

standards so that Trainers are able to apply a range of maintenance
works.
OBJECTIVES
The trainees will:

1. Prepare input data required for investigating optimum road
maintenance standards for paved roads using HDM-4.
2. Assign a range of maintenance standards to road sections and run
HDM-4.
3. Interpret the results and recommend optimum maintenance
INTRODUCTION
This case study presents the economic analysis of alternative

maintenance standards for two paved road sections that are 36.5 km
and 10.6 km long, respectively. The existing roads have surface
dressing and asphalt concrete pavement surfacing in good condition,
and carry approximately 6 000 and 16 000 vehicles per day (AADT),
respectively. The objective of this case study is to determine the
optimum maintenance intervals for resealing (surface dressing) and for
overlays on the two roads, respectively.
The case study data is provided in the T3 Case studies

workspace/database installed with the HDM-4 software. The data for
this case study are located in the Projects folder in the Case Studies
Workspace. The name of the case study is Case 2: Optimum
maintenance intervals. To locate the data for this case study:
The case study data includes the road network data set (CASE2
Project Sections) and the vehicle fleet (CASE2 Vehicle Fleet)
required for the analyses.
The first step in the procedure would be to calibrate the road sections

using the calibration factors derived from the previous case study. The
second step in the procedure should be to create additional
maintenance standards for the two roads. The data provided includes
one example each for:
Create additional maintenance standards by copying the above to

create separate standards for:
The optimum economic maintenance standards are defined as the

maintenance interval (scheduled or responsive) with the maximum
NPV. This can be illustrated by exporting the Benefit Cost Ratios
report to an Excel spreadsheet and plotting the NPV against the
maintenance interval (years) or intervention criterion (IRI).
Another useful plot would be of the NPV against the increase in
agency cost. The resulting curve from this is defined as the efficiency
frontier. This will be described in detail in the Programme Analysis
case study.
Those Trainers who complete the above exercise within a short time
could attempt to derive the optimum grading and regravelling
frequencies for the Gravel road that is included in the Road Network
database.
DEVELOPMENT
SUMMARY
RESOURCE MATERIALS

5. HDM-4 Volume 5: A Guide to Calibration and Adaptation
(Chapters 6 and 7)
Visual Aids:
1. PowerPoint slides
TRAINING FACILITIES

MISCELLANEOUS

Filename: T3 Project Unit Notes.doc

SHORT TITLE: HDM-4 Road Networks and Vehicle Fleet UNIT:
TITLE: Introduction to the HDM-4 Road Networks and Vehicle Fleet data managers
Duration: 30 minutes Presentation, 1 ~ 2 hours Practical

AIM
To introduce the concept of Road Networks and Vehicle Fleets

so that they can be correctly used in all HDM-4 applications.
OBJECTIVES
The trainees will:

1. Create one HDM-4 Road Network and several road
sections together with the required data sets.
2. Create one HDM-4 Vehicle Fleet and several vehicle types
together with the required data sets.
3. Create a sample HDM-4 Project and run the project.
4. Review typical HDM-4 output reports.
INTRODUCTION
This is a key unit that should be scheduled early in the training

programme, preferably immediately after the Software
Installation Unit. The Configuration Unit should be scheduled
after this Unit.
Road Networks and Vehicle fleets are the key data items
required in all HDM-4 applications. Road Networks can be
defined as a group of road sections with similar functional or
design characteristics, or are in the same geographic region.
Examples of road networks include; Trunk (or Main) roads,
Feeder roads, Paved roads, Unsealed roads, Expressways,
Northern Province roads, etc. In the context of a country’s road
network, these could be sub-networks. HDM-4 does not impose
any rules on the definition of a road network – it is up to the
user to decide which road sections should be included in the
network.
The Vehicle Fleet in HDM-4 is a group of vehicle types defined

by the user. Normally, a vehicle fleet will comprise several
vehicle types representing the different types of vehicles that use
a road network. Each vehicle type represents one class of
vehicles commonly found in the country (e.g. Toyota Corrola
representing medium cars). A more detailed discussion of the

factors to be considered when selecting vehicle types is included
in the Unit on Calibrating RUE.
After completing the process of creating road sections and

vehicle types, trainees should be shown how to create a sample
project using the Road Network and Vehicle fleet created in this
session. Note that it is not necessary to define maintenance
standards prior to running the project.
The trainees should run the sample project in order to review

some typical output reports generated by HDM-4.
The above procedure is seen to be essential in cementing the

concept of Road Networks, Vehicle Fleets and their subsequent
use in HDM-4 Project, Programme or Strategy Analyses.
DEVELOPMENT
This Unit should be presented as a practical session starting with

the PowerPoint slides to set the scene (30 minutes) followed by
hands-on data entry using HDM-4 (1 ~ 1.5 hours). It is
recommended that the Trainer should demonstrate the HDM-4
modules as follows:
• Overview of the HDM-4 Workspace and the Road Network

and Vehicle Fleet modules using PowerPoint slides (30
minutes) including discussions
• Demonstration of the Road Network module by the Trainer
(15 minutes)
• Hands-on session by trainees working individually to create
a Road Network and 3 or 4 road sections to illustrate the
different options available for creating new sections (30
minutes)
• Demonstration of the Vehicle Fleet module by the Trainer
(15 minutes)
a Vehicle Fleet and 3 or 4 vehicle types to illustrate the
different options available for creating vehicle types (30
minutes)
• Demonstration of the Projects module by the Trainer to
illustrate the process of creating a new Project (15 minutes)
and run a Project followed by a review of typical output
reports, such as Deterioration and RUE reports (30 minutes)
By the end of the session, trainees should be familiar with the

three HDM-4 modules; Road Network, Vehicle Fleet and
Project Analysis.
SUMMARY
Trainers should be aware that there will be trainees who are

much faster at completing hands-on exercises than others.

Consequently, it would be prudent to encourage some trainees
to work ahead of others, for example by creating several new
road sections or vehicle fleets whilst others catch up.
RESOURCE MATERIALS

1. Handout Notes on Creating Road Networks and Vehicle
Fleets
Visual Aids:
1. PowerPoint Slides (RoadNet & VehFleets.PPT)
TRAINING FACILITIES

MISCELLANEOUS
Filename: T3 RoadNet & VehFleet Unit Notes.doc

SHORT TITLE: HDM-4 Software Structure and Operation UNIT:
TITLE: Description of the HDM-4 software structure, installation procedure and initial
operation

AIM
1. To introduce the HDM-4 software components and the installation

so that Trainees are able to correctly install and use HDM-4 for
the first time.
OBJECTIVES
The trainees will:

1. Describe the HDM-4 installation procedure
2. Describe the main components of the HDM-4 software (i.e. Road
networks, Vehicle fleet, Road work standards, Projects,
Programmes, Strategies, and Configuration).
3. Demonstrate use of the HDM-4 software by running one of the
pre-defined case studies.
INTRODUCTION
It is important that Trainees are able to properly install and operate the
HDM-4 software. They should also appreciate the licensing conditions
under which the HDM-4 software is supplied.
Software installation is achieved through a wizard that guides users

through the installation process. There are several choices that the user
must select, for example the installation language, the default HDM-4
software language, the personalised Software Licence Key (SLK), the
HDM-4 components to be installed (e.g. Reports, Documentation,
etc.). The wizard provides the opportunity to step back through the
process and ensure that the correct choices have been made.
After successful installation, the Trainees should be taken through a

guided demonstration of the HDM-4 software. This should aim to
illustrate the procedure for creating and running one of the Project Cast
Studies (e.g. Case Study 2a) by following the steps below:
• Review the configuration data (climate zones, speed-flow types,
traffic flow patterns, default data)
• Select a Road Network and review the data for one or two selected
road sections
• Select a vehicle fleet and review the data for one or two selected
vehicle types

• Open the Road Works standards folder and review the
maintenance standards used in one of the Project Case Studies
• Open the Project folder and select one Case Study.
• Review the Case Study data and finally run HDM-4 and review
the output Reports.
DEVELOPMENT
It is advisable that the Trainer should first demonstrate the

whole procedure before asking trainees to do the same. This
should be done in stages as follows:
• PowerPoint presentation of software installation and
operation basics (15 minutes)
• Demonstration of HDM-4 installation procedure (15
minutes)
• Hands-on installation by each trainee (30 minutes)
• Guided demonstration of the software operation with
trainees repeating each step (1 hour)
SUMMARY
The session should be the first hands-on use of HDM-4.

Consequently, trainees will gain more if they are guided through
the basics of the HDM-4 software operation. For those trainees
who complete the tasks ahead of others, additional hands-on
experience could be gained by them going through the default
case studies individually.
RESOURCE MATERIALS

Visual Aids:
1. PowerPoint Slides (Software Installation and Operation.PPT)
TRAINING FACILITIES

MISCELLANEOUS
Filename: T3-UN Software Structure & Installation.doc

SHORT TITLE: HDM-4 Strategy Analysis Applications UNIT:
TITLE: Applications of HDM-4 Strategy analysis module for forecasting long term budget
requirements and road network performance trends

AIM
1. To introduce the HDM-4 Strategy Analysis module concepts so

that Trainees are able to forecast long term road budget
requirements and road network performance trends.
2. To highlight the differences between Strategy and Programme
Analysis Applications of HDM-4.
3. To familiarise Trainers with the use of HDM-4 Strategy Analysis
data input and output modules.
OBJECTIVES
The trainees will:

1. Describe the concepts of the HDM-4 Strategy analysis
application.
2. Use the HDM-4 model to calculate a long term budget
requirement
3. Alter maintenance standards and unit costs and study the impact
on budget requirements.
4. Export the Strategy Analysis results and plot performance
indicators using a spreadsheet.
INTRODUCTION
This Unit should be taught as a two separate sessions. The first

session should focus on the theoretical aspects of strategy
analysis, drawing on the material from the presentation on
Programme Analysis. The second session should be a practical
case study illustrating the concepts through a hands-on exercise.
Strategy analysis is concerned with the analysis of entire road

networks to determine funding needs and/or to predict future
performance under budget constraints. The main factors which
define a strategy are:
• Definition of the road network to be analysed. This could

comprise individual road sections within the network, or
representative sections derived by aggregating the length of
all road sections with similar characteristics.

• A specified time period
• A specified budget limit
The analysis conducted by HDM-4 is similar in all respects with

the Programme analysis application.
The main output from a strategy analysis are a prioritised multi-

year list of road works together with charts showing:
• Predicted average performance of the road network over the
analysis period
• Required budget (financial) over the same period
• Graphs/charts of individual pavement defects and summary
road user effects.
DEVELOPMENT
Two sessions are envisaged to cover the material. The first

session of about 1 hour is a based on a PowerPoint presentation
of the Strategy Analysis concepts.
The second session should be a hands-on practical to

demonstrate use of the HDM-4 Strategy Analysis application to
prepare a work programme using pre-defined Case Study data.
Trainers should review the default Strategy Analysis case study
included in the HDM-4 installation package in order to adapt
these to suit local conditions.
SUMMARY
The two sessions should reinforce the Strategy Analysis

concept. The session should conclude with a review of the
Strategy analysis data produced after running the software.
RESOURCE MATERIALS

1. Handout Notes on Strategy Analsysis Case Study
Visual Aids:
1. PowerPoint Slides (Strategy.PPT)
TRAINING FACILITIES


MISCELLANEOUS
Filename: T3 Programme Unit Notes.doc

SHORT TITLE: HDM-4 Road Work Standards UNIT:
TITLE: Creating HDM-4 Maintenance and Improvement Road Works Standards
Duration: 30 minutes Presentation, 1 ~ 2 hours Practical

AIM
To introduce the concept of Road Work Standards so that they

can be correctly used in all HDM-4 applications.
OBJECTIVES
The trainees will:

1. Create several maintenance standards.
2. Create one or tow road improvement standards.
3. Assign one or more maintenance standards to road sections.
4. Create HDM-4 Project alternatives.
5. Review typical HDM-4 output reports illustrating road work
effects.
4. Creating Road Networks and Vehicle Fleets
INTRODUCTION
This Unit should be scheduled immediately after the Road Work

Effects Unit. It is anticipated that this unit will be mainly a
hands-on session that will provide the opportunity to trainees to
practice creating road work standards and the assignment of
maintenance/improvement standards to project, programme or
strategy alternatives.
A road work standard can be either a maintenance or

improvement standard. Maintenance standards comprise one or
more road work items together with the corresponding design,
intervention (scheduled or responsive), unit costs and predicted
effects on pavement condition. A typical maintenance standard
should include both routine and periodic activities. A SIMPLE
standard normally includes only one periodic maintenance
activity, whilst a COMPLEX standard includes several periodic
maintenance activities.
Trainees should be taught the difference between scheduled and

responsive intervention levels. If the latter are used, several
parameters can be specified using boolean AND combinations
to trigger selected types of road works.

It is important to note that the boolean OR equivalent can be
achieved in HDM-4 version 1 by specifying more than one
instance of a type of road work within the same standard (e.g.
overlay when (IRI > 6) or (cracking>5% AND rutting > 10mm).
The above would be implemented by specifying two overlays,
one for (IRI > 6) and the second for (cracking>5% AND rutting
> 10mm).
Note also that COMPLEX maintenance standards provide the

flexibility to model a wide range of maintenance intervention
levels used in various countries.
The SIMPLE maintenance standard would normally be applied

at the strategy analysis level where a decision is being sought
between different road work items (e.g. overlays versus
resealing). A SIMPLE standard would permit road agencies to
compare the effectiveness of different types of road works.
Unlike maintenance standards, improvement standards can only

contain one improvement. For example an upgrading
improvement cannot be specified together with widening. In
addition, responsive improvement standard will only be applied
once during a project in order to avoid repeated application of
the maintenance standard. As an example, consider the case
where road widening would be applied when traffic AADT
exceeds 2500 vehicles per hour. Given that the AADT will
normally grow each year, the improvement would in theory be
required in each year that AADT > 2500 vehicles.
The design of a maintenance or improvement standard

determines the effectiveness of the standard. Note that this does
not have any impact on the unit costs, e.g. the unit cost for a 50
mm overlay will need to be calculated externally. If the overlay
thickness is changed to 100 mm, the unit costs would not be
adjusted accordingly.
DEVELOPMENT
The Unit should be presented mainly as a hands-on session. This

could start with a demonstration by the Trainer of how to create
road work standards. Each trainee should then be asked to repeat
the same procedure.
Specification of Improvement standards should follow after

several maintenance standards have been created by the
participants.
SUMMARY
Trainers should be aware that there will be trainees who are

much faster at completing hands-on exercises than others.
Consequently, it would be prudent to encourage some trainees
to work ahead of others, for example by creating several new
road maintenance or improvement standards whilst others catch
up.

RESOURCE MATERIALS

1. Handout Notes on HDM-4 Maintenance and Improvement
Standards
Visual Aids:
1. PowerPoint Slides (Work Standards.PPT)
TRAINING FACILITIES

MISCELLANEOUS
Filename: T3 Work Standards Unit Notes.doc

TITLE: UNIT:
SHORT TITLE:
Primary Module:
Duration:
Target Group:

AIM
OBJECTIVES
1. etc
i) etc
INTRODUCTION
DEVELOPMENT
SUMMARY
RESOURCE MATERIALS

1. etc
Visual Aids:
1. etc.
1. etc.
TRAINING FACILITIES
MISCELLANEOUS

TITLE: The role and effect of traffic data in HDM-4 UNIT:
SHORT TITLE: Traffic effects in HDM-4
Duration: 1 hour plus 1hour 30 minutes practical

AIM:
To teach the types and effects of different traffic data in HDM-4
so that trainees can know the data requirements of HDM-4 and
understand the effects on HDM-4 analyses.
OBJECTIVES:
The trainee will:
1. Describe the main areas where traffic affects road
investment analysis
2. Describe “Normal”, “Diverted” and “Generated” traffic
3. Describe the input requirements (type and location) for
traffic volume and composition
4. Describe the concept of speed flow types to model capacity
in HDM-4
5. Explain the concept of PCSE as used in HDM-4 and how
existing data may be converted to PCSE values.
6. Describe the use and input of traffic flow patterns in HDM-
4. Show how to calculate these.
7. Demonstrate transfer of current data formats to those
required by HDM-4
ii) HDM-4 Analyses
iv) Investment Appraisal
v) Vehicle classification

INTRODUCTION
This session aims to show the effects of traffic data in HDM-4
analyses. It links with other sessions (vehicle types, Road User
Effects, investment appraisal) and aims to introduce specific
parameters related to all vehicles. It will cover volume,
composition, axle loading and speed but not other Road User
effects. Traffic data needs to be placed in context not only
within HDM-4 but also within road investment appraisal. It is a
key element of the analyses and the importance of this data
needs reinforcing strongly early in the session to gain the
trainee’s interest.

practical hands-on experience of navigating around the traffic
data input areas of the model. This will reinforce the learning
and provide ability to perform HDM-4 analyses.
A question and answer technique can be used to draw out the

relevance of traffic data by a series of questions such as:
• What do we mean by traffic? How is it measured? This

could draw out knowledge of traffic data collection (volume
and composition). The trainees should be encouraged to
consider the full range of vehicles as being “traffic”, not just
cars or trucks.
• Do they consider non-motorised traffic as “traffic”? Is it

measured in their country?
• In what way could traffic affect the road and hence road
management? Ask what would happen to the road of there
were no vehicles. This should aim to draw out ideas on the
aspects of vehicles which could affect road deterioration
such as weight and axle loading, as well as volume.
• How can the traffic be affected by the condition of the

road? This is aimed to draw out ideas on effects on vehicle
speed.
DEVELOPMENT

sessions. Indicative parts and timings are:
i) Overview including key uses for traffic data in HDM-4

(5 minutes)
ii) Traffic data input parameters and their location –
categories, volume, growth, axle loads (20 minutes)
iii) Traffic interaction and speeds (30 minutes)
iv) Case study & practicals (1.5 hours)
The session could start by reiterating the key uses of traffic data
in an investment appraisal tool such as HDM-4. This aims to put

the purpose of collecting and inputting the data into context.
This could be done by asking questions (see above) before
showing the slide.
The location of the data in HDM-4 needs to be identified,

especially as it is located in a number of different places. At this
stage it should be indicated only in outline as the individual
parameters will be identified through the presentation. The
session would then develop through identification of the
different data parameters and their input requirements.
Further participation could be sought by inviting trainees to

describe “Normal”, “Diverted” and “Generated” traffic. It is
important that the distinction is understood as they contribute
differently to the economic analysis calculation and
misunderstanding and misuse can affect the analysis results. It
would be covered in more detail in the session on economic
analyses.
The concept of a representative vehicle together with the vehicle

hierarchy for traffic (categories, classes and types) should have
been introduced in the session on vehicle classification. The key
aspect for the trainees to recall is the purpose of a representative
vehicle which is to simplify input of vehicle classification. This
concept is easy to understand in relation to traffic data collection
as broad vehicle types are used there.
The key parameters for traffic related to the road sections are
volume, composition and growth and these should be explained
in the context of their location in the road network files. It
should be explained that traffic volume is specified as AADT
and is specified separately for MT and NMT. The definition of
AADT should be clear, as it affects many aspects of the
analysis. The input of traffic growth is very flexible with ability
to specify values for different time periods and by different
calculation methods, such that all conventions should be
satisfied. The input of composition is represented simply as a
percentage of the volume.
To introduce parameters of weight and axle loading the trainees

could be asked to recall what affects road deterioration.
Information on vehicle weight and axle loading are located in
the vehicle fleet section – the input screen could be shown here.
The modelling of traffic interactions and congestion could be

introduced by a picture of a busy road. A key point to make here
is that traffic interactions have not been fully included in
previous versions of HDM-4 but can be an important cost to
road users. This session aims to introduce the parameters which
are required as input values for calculating traffic interactions.
The impact of these on Road User costs will be covered in
another session.
The concept of Passenger Space Car Equivalent used for

modelling traffic interactions and congestion in HDM-4 must be

introduced and trainees must be clear how this differs from PCU
or PCE. A question could be asked as to what system is used or
understood by the trainees. Information on how values for PCSE
may be derived should be given. The PCSE parameter is a level
1 input value in the basic vehicle data section in the vehicle
fleet. HDM-4 is an improvement on earlier models where traffic
interaction effects have not been fully included but can be an
important cost to road users. The impact of these on Road User
costs will be covered in another session.
The PCSE parameter is used in the 3-zone speed-flow model

which should be described. Information should be given on the
location of the various input parameters such as nominal and
ultimate capacity, which are related to a Speed Flow Type.
Each road section must nominate a Speed Flow type and Traffic
flow pattern, both of which are specified in the Configuration
section. These can be user defined, although there are some
defaults available. The traffic flow pattern is derived from an
hourly flow distribution and the calculations of this can be
described.
Depending on the level of the trainees, the modelling of speed

can be described further and other impacts on speed (such as
roadside friction and NMT’s) can be described and their
location identified. However, input parameters such as desired
speed are level 2 data in the calibration level of the fleet data
and these values together with other level 2 parameters should
only be used where calibration studies have been carried out.
The existence of default values should be explained.
A practical session could be introduced as a hands-on navigation

around the general, section, vehicle and traffic tabs to locate the
different input parameters. Trainees could be encouraged to add
periods of traffic growth rates. They should also look in the
configuration section at the speed flow and traffic flow patterns.
To indicate the significance of the traffic volume, a case study

could be repeated with a 20% increase in AADT to highlight the
impact to the estimation of both total benefits and surface
condition (roughness).
To indicate the effects of the traffic congestion, a case study

with AADT close to the nominal capacity and a high growth
rate could be run to examine changes in vehicle speeds.

SUMMARY
Separate summaries or reporting back should be invited at he
end of each section or practical session. The summary could be
done by inviting key points from the trainees or be pre-prepared.
The former is preferred as it provides a confirmation of the
training objectives.
The importance of the data should be reiterated in relation to its

impact on the investment appraisal / road network management.
Trainees should be advised of further related sessions on road

user costs, vehicle details or investment appraisal.
RESOURCE MATERIALS
1. HDM-4 documentation, Volume 3, Chapter 11
2. HDM-4 documentation, Volume 4, Part B
3. HDM-4 documentation, Volume 5, Appendix E
4. HDM-4 documentation, Volume 7 (Draft).
Visual Aids:
1. Short PowerPoint presentation (15-20 slides including
photos)
2. Long PowerPoint resource (As required)
1. HDM-4 Case Study
TRAINING FACILITIES

MISCELLANEOUS

TITLE: The importance, types and management of unsealed UNIT:
roads, and the modelling framework used in HDM-4
SHORT TITLE: Management of Unsealed Roads
Primary Module: Highway Management

AIM
To teach the contribution of unsealed roads to the national
network and their classification and management so that
participants can implement improved management programmes
and methods with the aid of HDM-4
OBJECTIVES
The trainees will:
1. Explain the contribution of unsealed roads in the national
network.
2. Describe the main classifications and characteristics of each
recognised type.
3. Describe the deterioration mechanisms and modes of
distress which affect unsealed roads and how they are
modelled within HDM-4
4. Define the types of road works and intervention criteria
commonly applied in the management of unpaved roads.
5. Compare the effect of different construction and
maintenance operations on unsealed road performance and
overall cost effectiveness.
6. Explain the relationship between policy objectives and
maintenance actions.
7. Develop and defend a suitable programme for managing a
network of unsealed roads.
i) Pre-optimisation of standards.
ii) Strategic analysis.
iii) Overview of highway management
iv) Assessing maintenance needs
v) Investment appraisal
vi) Operation of HDM-4
vii) HDM-4 Project and Programme Analysis
APPROACH
This is best taught by employing a variety of learning aids to
capture the interest of the trainees and convey pictorially and
graphically the context, variety of physical conditions and
influence of key factors on unsealed road performance.
The presentation should commence by identifying the target

group, describing the aims and objectives of the session and
build towards a strategic overview of the management of
unsealed roads and how key variables are modelled directly or
prepared and input as exogenous costs into HDM-4. The
impact, on programme effectiveness or costs, could be
illustrated to reinforce points and prepare for the practical
session (Develop and defend a management programme).

It is important to draw out the practical experience of the
trainees, for the purpose of identifying their understanding of
the overall importance of unsealed roads, the modes of distress
and the technical classification and treatments that they
recognise and use.
Questions should be focused on topics such as:
• How important are unsealed roads? This should aim to

identify the proportion or length of network, importance to
the community, the economy, employment, etc.
• How do you classify unsealed roads? This should aim to
develop a classification, the behaviour of the main types and
perhaps their knowledge of the HDM III/4 classification.
• What deterioration mechanisms exist? Climatic
conditions and the adequacy of the surfacing/roadbed
materials are key factors.
• What is the effect of surface maintenance? Ask them to
name the types and effects of different treatments and the
possible role for labour and intermediate technology.
Treatment definitions are notoriously poor, yet the
difference in effect and cost between any two treatments can
be very significant.
• How can we accommodate non-modelled parameters,
and what is their importance? Many works activities,
which are of high priority, are not modelled and must be
entered by the user. This needs to be clearly understood by
the trainees.
The relationship with sessions on Pre-Optimisation of Standards

and Strategy Analysis, emphasising the adoption of affordable
and appropriate standards, should be clearly stated and
reinforced throughout. The link to sessions on investment
appraisal, etc should be highlighted.
DEVELOPMENT
The overall scope of the Unit and the order and duration of the
various parts and practical sessions explained. Indicative parts
and timings are as follows:
i) Overview, including importance, classification,

deterioration mechanisms and distresses (45 mins with
participation).
ii) Unsealed road deterioration and effects of road works,
including introduction to computer entry (45 minutes
with participation).
iii) Approach to maintenance standards and policy
objectives (30 minutes with participation).
iv) Case studies and practicals (4 – 6 hours).
The Trainer should decide which sessions to cover and the

extent of participation, since it is possible to reduce perhaps by
30% the duration of the sessions.

The session could begin by developing a broad overview with
the participation of the trainees. This might involve inviting
answers to two or three relatively simple questions (see above).
The answers to these will give an immediate indication of the
breadth and depth of basic knowledge amongst the trainees.
Simple concepts should be introduced first and might focus on:
• Recognising the role of traffic induced deterioration and that

of the environment, giving particular emphasis to drainage,
rainfall and terrain.
• Illustrating (perhaps by use of drawings on the white board)
the effect of light and heavy maintenance.
• High impact graphical plots, simple numerical data or
photographs the effect of good/poor practice.
Complex concepts and illustrations could be omitted, or perhaps

conveyed through a simplified format that is available for
practical application. Examples include:
• Single or multiple deterioration cycles (SIMPLE)

• Effect of maintenance on single/multiple cycles (SIMPLE)
• Annual average roughness (COMPLEX)
• Steady state roughness (COMPLEX)
• Typical annual grading frequency (by type) for different
traffic levels (SIMPLE)
The trainees should be reminded of whether major factors are

modelled or not modelled by HDM-4, and the importance of
exogenous inputs. A separate session may be required on the
identification and derivation of exogenous costs.
Guided demonstration through use of the computer (with HDM-

4 loaded) to specify alternative maintenance standards, their
design, intervention criterion, costs and effects, should follow.
At this stage the key input data that require direct entry or
specification of calibration factors etc should be identified.
As a preparatory stage to the PRACTICAL, the trainees should

be required to specify alternative maintenance standards and
collate or compute the related inputs for HDM-4. This will help
demonstrate their understanding of the basics prior to the
execution of a full run.
SUMMARY
Reporting back should be invited or else a summary provided by

the Trainer at the end of each presentation/practical session.
The former is preferred, since this will provide confirmation of
the training objectives.
Any intended follow up, perhaps giving examples of a wider

range of case studies, more detailed inputs on costs and

assessing needs or field sessions, should be notified to the
trainees.
RESOURCE MATERIALS

1. Effective planning, implementation and management
strategies to improve local resource utilisation in rural roads
(WB/DFID Rural Transport Knowledge Base).
2. HDM-4documentation,Volume 4, Analytical framework
and model descriptions.
3. HDM-4 documentation, Volume 6, Part D (Draft), Road
deterioration and works effects.
4. Powerpoint presentations as additional handouts.
Visual Aids:
Short Powerpoint presentation (10-15 slides including photos)
Long Powerpoint resource (As required)
1. HDM-4 Project Case Study No 1 – Upgrading an unsealed
road.
2. HDM-4 T3 Extended Case Study No 1 – Optimum timing
of upgrading and effect of variation in grading frequency on
performance.
TRAINING FACILITIES
Powerpoint projector, white board/flip chart, overhead

MISCELLANEOUS

TITLE: The role of traffic classification and vehicle fleet UNIT:
specification in HDM-4
SHORT TITLE: Vehicle classification in HDM-4
Duration: 45 minutes + 45 minutes practical

AIM:
To teach the role and specification of the vehicle fleet data in
HDM-4 so that trainees can know the information required to
populate the vehicle fleet section of HDM-4 and the
methodology of vehicle classification selection for individual
analyses.
OBJECTIVES:
The trainee will:
1. Recall the hierarchy for motorised and non-motorised
traffic in HDM-4
2. Explain the concept of a representative vehicle in HDM-4
3. Explain the use of vehicle fleets for storing vehicle data for
use in HDM-4 analyses
4. Recall the main types of data required for vehicle types,
both motorised (MT) and non-motorised(NMT)
5. Show how to create a new representative vehicle
6. Demonstrate how to select vehicles for a project section
from a defined vehicle fleet
ii) HDM-4 Analyses
iii) Traffic Effects
iv) Road User Costs

INTRODUCTION
The key aspect of this unit is for the trainees to demonstrate
their understanding of the vehicle classification used in HDM-4,
the data requirements for vehicle types and the specification and
use of vehicles within a vehicle fleet. The trainees need to
understand the need for relevant vehicle fleets to be derived for
their networks and how individual vehicle types can then be
selected for individual analyses. Use can be made of pictorial
images of different vehicle types (motorised, non-motorised,
buses, trucks, etc).
This session links in with other sessions on HDM-4 or

investment appraisal, such as traffic effects and Road User
Costs as vehicles and traffic are a key element of the analyses.
This session aims to introduce specific parameters relating to
vehicle characteristics required by HDM-4 but will not cover the
modelling of Road User effects.

practical hands-on experience of navigating around the vehicle
fleet area of the model. This will reinforce the learning and
provide ability to perform HDM-4 analyses.
Questions could be on:

• What do we mean by vehicles? This should aim to draw
out ideas on what the trainees see as different types of
vehicles and encourage them to think across the spectrum of
vehicles, including non-motorised.
• What do you understand by non-motorised vehicles?
• What aspects of vehicles could affect road deterioration?

This is aimed to draw out ideas on vehicle parameters which
would be required to model deterioration such as vehicle
weight and axle loads. This will relate the session to HDM-
4 and investment appraisal / road management.
• If the traffic can affect deterioration of the road surface,

can the road condition affect the vehicles? This is aimed
to draw out ideas on the cost to road users, in the form of
operating costs and that these can vary for different types of
vehicles.
DEVELOPMENT
The overall scope of the Unit and the order and duration of the
various parts and practical sessions should be explained.
Indicative parts and timings are:
i) Overview, including key uses of traffic and vehicle data

(10 minutes)
ii) Vehicle hierarchy, representative vehicle, vehicle fleet
concepts (15 minutes)
iii) Specification of vehicle data in vehicle fleets (20
minutes)
iv) Practical: Navigation around vehicle fleet in HDM-4

(15 minutes)
v) Practical: Input of data for a new representative vehicle
(20 minutes)
vi) Practical: selection of vehicles for HDM-4 analysis (10
minutes)
The session could start by reiterating the key uses of traffic and
vehicle data in an investment appraisal tool such as HDM-4.
This could be done by asking some questions (see above) and
aims to put the value of vehicle information into context.
The vehicle hierarchy used in HDM-4 for traffic (categories,

classes and types) can be described using charts indicating the
different types modelled. The concept of a representative
vehicle should be introduced. It is important that trainees
recognise the purpose of the representative vehicle and that the
user is free to identify his own vehicle types.
The concept of vehicle data being held on a range of

representative vehicles within a vehicle fleet should be
explained. The diagram of HDM-4 architecture could be shown
here to demonstrate the position of the fleet data in the
modelling framework. Additionally, the use of an input screen
could assist here to show that while a fleet is composed of many
vehicles, not all vehicles need to be selected for each analysis. It
should also be explained that the usual mode of operation for a
country would be to create a national or several regional fleets
containing all representative vehicle types likely to be required
for analyses. These can be linked to sub networks.
The trainee should then be guided through the type of vehicle

data required by HDM-4 and its location in the vehicle fleet
section. This could be done with the aid of input screens or
better by hands on navigation through an existing vehicle fleet.
Differentiation should be made between level 1 data and level 2
data and the existence of default data, but level 2 data should not
be described at this stage. This should also include non-
motorised vehicles. The input data for these are fewer than for
MT because of the less well developed relationships available
for estimating Road User costs for these vehicles.
Brief descriptions of the use of the data can be given but the
trainer should avoid detail as this will be covered in the Road
User costs session. Broad categories such as definition, physical
data, tyre data, utilisation, loading, and economic and financial
costs, including time values should be used.
The concept of the use of unit costs in the calculation of Road

User costs can be introduced, requiring unit costs values related
to vehicles to be input. These will be covered in later sessions as
would the difference between economic and financial costs
The input value, PCSE, (a level 1 input value in the basic

vehicle data section) may be noticed during the navigation. This
relates to congestion and will be described in the Unit on Traffic

Effects.
A practical session should be included to:

a) Provide a hands-on navigation around the vehicle fleet
section – definition, basic data, economic costs & financial
costs tabs to locate the different input parameters. This
should include both motorised and non-motorised vehicles.
b) Provide the trainees with practice at creating one or more
new representative vehicles. It is recommended that the data
for this is provided on data sheets in advance. The data sheet
should contain all level 1 (required data) but not level 2
(calibration data).
c) Show how to select vehicles from a vehicle fleet to assign to
an analysis. This could be by adding vehicles to an existing
case study or creating a new study.
SUMMARY
Separate summaries should be provided at the end of the
presentation and each practical session.
The summary could be done by a series of questions aimed to

recapping on the main objectives of the session. This would
provide confirmation of the training objectives
The importance of the data should be reiterated in relation to its

impact on the investment appraisal / road network management.
Trainees should be advised of further related sessions on traffic

data effects, road user costs and investment appraisal.
RESOURCE MATERIALS

2. HDM-4 documentation, Volume 4, Part E
Visual Aids:
1. Short PowerPoint presentation (15-20 slides including
photos)
1. HDM-4 Case Study for navigation exercise
2. Vehicle type input data sheets
TRAINING FACILITIES
MISCELLANEOUS

CAPITAL BUDGETING
Background
There are often situations when the budget available for projects (e.g. in the road sub-
sector) will not be sufficient to undertake all projects shown to have a positive return (i.e.
projects with positive NPV or IRR > MARR). In such situations, a formal method of
selecting projects to be included within the budget can be applied.
One such method is to apply the NPV decision rules in capital budgeting. The NPV rule
states that projects with the highest NPV should be given highest priority as they provide
the largest benefit to society.
Capital budgeting or rationing can be applied to a group of projects which meet either of
the following conditions:
(i) projects that are independent of each other (e.g. list of road projects from different
parts of the country),
(ii) mutually exclusive projects (i.e. projects that are alternatives to each other). Only
one alternative can be selected.
NPV Capital Budgeting Rules
The NPV capital budgeting rules can be applied in both situations where sufficient funds
are available, and also when there is a budget constraint. The rules are summarised below:
A. When sufficient funds are available to undertake all projects;
(i) Select all independent projects with NPV > 0
(ii) Select mutually exclusive project alternatives with the highest NPV
B. When capital rationing is to be applied due to shortage of funds;
(i) Select independent projects with the highest NPV/Cost ratio
(ii) Select mutually exclusive projects using the incremental NPV/Cost method
described below.
Example 1
A total budget of $1.5 million is given for the following projects. Use the Capital
Rationing methods to select the group of projects which will maximise the NPV.
Project Cost NPV NPV/Cost Rank
A 150,000 35,000 0.233 2
B 150,000 45,000 0.300 1
C 200,000 40,000 0.200 3
D 250,000 37,500 0.150 6
E 500,000 70,000 0.140 7
F 500,000 77,000 0.154 5
G 500,000 78,000 0.156 4
Solution 1
The Budget limit of $1.5 million could be exhausted by selecting projects with the highest
NPV, i.e. G, F & E. The total NPV for these would be $ 225,000.
If the NPV/Cost ranking is used instead, projects B, A, C, G & F would be selected giving
a total of $ 275,000.
Incremental Analysis
The incremental analysis is used to test whether the ratio of the increase in NPV to the
increase in costs between alternative projects (mutually exclusive) is greater than a
specified marginal ratio. The formula is defined below:
Incremental Benefit/Cost = [ NPV2 - NPV1 ] / [ C2 - C1 ]
If the above ratio is greater than a specified value, then the project alternative is included
among those to be funded.
Example 2
The table below summarises the capital costs (C) and Net Present Values (NPV) of 8
independent road projects, one of which has 2 mutually exclusive alternatives. If the total
budget available is $ 180,000, select the group of projects will maximise the total NPV.
Summary of NPV and Costs (in $)
A 20,000 4,869 0.2435 2
B 30,000 240 0.0080 8
C 20,000 5,184 0.2592 1
D 20,000 210 0.0105 7
E 40,000 6,401 0.1600 4
F 35,000 4,497 0.1285 5
G 20,000 1,200 0.0600 6
H1 80,000 15,000 0.1875 3
H2 100,000 16,829 0.1683

Solution
In a capital rationing situation, the choice between independent projects is made on the
basis of the NPV/Cost ratio, and the choice between mutually exclusive alternatives is
made by comparing the incremental NPV/Cost ratio against the 'marginal' NPV/Cost ratio
of other competing independent project.
Working down the ranking order until the budget of $180,000 is exhausted, we would
first select projects C, A, H and E. For project H we need to decide between alternatives 1
or 2; Although H2 has the higher NPV, H1 has a higher NPV/Cost ratio. The incremental
NPV/Cost ratio for choosing H2 instead of H1 would be calculated as below:
Incremental NPV/Cost = [ NPVH2 - NPVH1 ] / [ CH2 - CH1 ]
= [ 16829 - 15,000 ] / [ 100,000 - 80,000 ]
= 0.0915
The above is then compared against the 'marginal' NPV/Cost ratio given by the next
projects in the NPV/Cost ranking order which can be carried out with the remaining
budget. We should therefore select project H2 because it has a higher incremental
NPV/Cost ratio when compared to that of projects D or G. Note that projects D and G are
the only ones which could utilise the $20,000 balance from the budget if projects C, A,
H1 and E are selected (totalling $160,000). Project G in this case provides the 'marginal'
NPV/Cost ratio of 0.0600 which is less than the incremental NPV/Cost ratio calculated
for selecting H2 instead of H1.
The selection of projects C, A, H2 and E gives the highest total NPV than any other
possible combination within the specified budget of $180,000.
ECONOMIC ANALYSIS
Questions & Decisions 1.
Is the p
project
j justified
j ?- Are benefits
greater than costs?
Which is the best investment if we have a
set of mutually exclusive alternatives?
If funds
f d are limited,
li it d how
h should
h ld different
diff t
schemes be ranked?
When should the road be built?
Questions & Decisions 2.
What standard of construction should be

used?
What standard and frequency of road
maintenance is required?
Sh ld stage
Should t construction
t ti be
b used?
d?
Are complementary investments required?
Components of Economic Analysis 1
Costs and Benefits are measured in money

terms
Road construction and maintenance costs
are compared with
ith estimates of the direct
Primary Benefits going to road users
Secondary Benefits are usually ignored
Economic p
prices are used using
g constant
prices
Components of Economic Analysis 2
Costs and Benefits are forecast over the

planning time horizon (usually between 10
and 20 years)
Future Benefits are valued less as time
progresses using the planning discount
rate
Costs and Benefits are compared
p usingg
decision criteria such as NPV, IRR, etc.
Economic Prices
Adjust for Taxes, Duties and Subsidies

Use the planning discount rate not the financial
market rate
If exchange rate is overvalued, then value
p
Imports and Exports
p more highly
g y
If there is under employment of labour, then
Unskilled labour is valued less than market rate
If there is a shortage then Skilled labour may be
valued
l d att more than
th marketk t rate
t
Project Costs
Management
g
Labour
Equipment
Materials
Land
Primary Effects 1
Reduced vehicle operating

p g costs (VOC)
( )
z fuel and lubricants
z vehicle maintenance
z depreciation and interest
z overheads
Reduced journey time

z drivers, passengers and goods
Primary Effects 2
Changes
g in road maintenance costs
Changes in accident rates
I
Increased
d travel
t l
Environmental effects
Change in value of goods moved
Secondary Effects
Due to changes
g in:
z Agricultural output
z Services
z Industrial output
z Consumer behaviour
z Land
a d values
a ues
Coverage & Double Counting
Economic analysis should be designed to

give maximum coverage of benefits
Avoid double counting:g do not add
primary and secondary benefits (e.g.
increases in land values added to changes
in transport costs)
Consumer surplus
p approach
pp should be
adequate (in a perfectly competitive
economy)
Economic Comparisons
Economic analysis involves a comparison of

“With”
With and “Without”
Without cases
Forecasts are made of traffic, road condition,
VOC and road maintenance for BOTH scenarios
An unrealistic “Without” case (i.e. with little
maintenance)) can ggive a false result
A range of “With investment” cases should be
analysed
y to find the best solution
A minimum investment approach often gives the
best economic results and should be tested
Traffic Categories
Normal traffic: Existing traffic and growth

that would occur on the same road, with
and without the investment
Diverted traffic: Traffic diverted from
another road to the project road as a result
of the investment
Generated traffic: New traffic induced by
y
the investment
Consumer Surplus Approach
Transport cost savings

Total Benefits = for Normal traffic
Cost +
Additional Benefits from
C1 Generated Traffic
C2
Demand Curve
T1 T2 Traffic
Estimating Benefits
Normal traffic benefits: tripsN * d1 * (VOC1- VOC2)
Diverted traffic benefits: tripsD * ((d1 * VOC1)-(d2*VOC2))
Generated traffic benefits: tripsG * d2 * (VOC1- VOC2)/2
d1 = existing road length d2 new road length

VOC1 = vehicle operating costs per km “without”investment
VOC2 = vehicle operating costs per km “with” investment
VOC data relates to each road section and its condition at the time
Economic Decision Criteria
Net Present Value

NPV = (B1- C1)/(1 + r) + (B2- C2)/ (1 + r)2 + …+ (Bn- Cn)/(1 + r )n
Internal Rate of Return

To calculate IRR, solve for r, such that NPV = 0
B1, B2…Bn = Benefits in years 1, 2 … n

C1, C2…Cn = Costs in years 1, 2 …. n
r = Planning discount rate
n = Planning time horizon
Economic Decision Criteria
Net Present Value/ Investment Cost

NPV/ C = NPV/Ci
First Year Rate of Return

FYRR = (B1- C1) / Ci
B1 , C1 = Benefits and Costs in year 1
after
ft construction
t ti
Ci = Road investment costs
Project Alternative I
Year Constr. & VOC Total Discounted

Maint. Cost Cost Total (15%)
( )
2000 0.8 1.5 2.3 2.3
2001 0.9 1.6 2.5 2.2
2002 10
1.0 18
1.8 28
2.8 21
2.1
2003 1.1 2.1 3.2 2.1
2004 1.2 2.3 3.5 2.0
2005 1.3 2.7 4.0 2.0
2006 1.4 2.9 4.3 1.9
2007 1.5 3.2 4.7 1.8
2008 1.6 3.5 5.1 1.7
2009 1.7 3.8 5.5 1.6
2010 1.8 4.0 5.8 1.4
TOTAL 21.1
Project Alternative II
Year Constr. & VOC Total Discounted

Maint. Cost Cost Total (15%)
2000 5.8 + 0.8 1.5 8.1 8.1
2001 0.4 1.0 1.4 1.2
2002 0.5 1.1 1.6 1.2
2003 06
0.6 13
1.3 19
1.9 12
1.2
2004 0.7 1.4 2.1 1.2
2005 0.7 1.6 2.3 1.1
2006 0.8 1.8 2.6 1.1
2007 0.9 2.0 2.9 1.1
2008 0.9 2.2 3.1 1.1
2009 1.0 2.3 3.3 0.9
2010 1.0 2.5 3.5 0.9
TOTAL 19.1
Economic Decision Criteria: Summary
NPV IRR NPV/C FYRR

Project economic validity V.Good V.Good V.Good Poor
Mutually exclusive projects V.Good Poor Good Poor
Project timing Fair Poor Poor Good
Project screening * Poor V.Good Good Poor
Under budget constraint ** Fair Poor V.Good Poor
* check for robustness to changes in key variables

** with incremental analysis
OVERVIEW OF ROAD USER
EFFECTS MODELS
Overview of Presentation
Introduction Fuel Consumption

RUE components in Tyre Consumption
HDM-4 Maintenance and
Mechanistic modelling Repairs
Speed Prediction Capital Costs
Speed Flow and Oil Consumption
Traffic Interactions Travel Time
Safety/Accidents
Components of Road User Effects
Road User
Effects
Vehicle Operating Motorised Travel Non-Motorised Environmental

Accident Costs Other Effects
Costs Time Costs Traffic Impact
Fuel Uncongested
g Vehicle
T
Travel
l Ti
Time F t lit
Fatality
Consumption Travel Time Emissions
Tyre Delay Due To Operating

Injury Noise
Consumption Congestion Costs
Oil and Impact on

Delay Due To
Lubricant Motorised Damage Only Energy Use
Road Works
Consumption Traffic
Parts
Consumption
Labour Hours
Depreciation
Interest
Overheads
Key Changes from HDM-III
Unlimited number of representative vehicles

Reduced maintenance and repair costs
Changes to utilisation and service life modelling
Changes to capital, overhead and crew costs
New fuel consumption model
New oil consumption model
Changes to speed prediction model
Use of mechanistic tyre model for all vehicles
New Features in HDM-4
Effects of traffic volume on speed, fuel, tyres and

maintenance
i costs
Non-motorised transport modelling
Effects of road works on users
Traffic safety impact
Vehicle emissions impact
Factors Influencing RUE
RUE Component HDM-III

HDM III HDM-4
HDM 4
Geometry Condition Capacity Geometry Condition Capacity
Fuel • • • • •
Tyres • • • • •
Parts • • •
L b
Labour • • •
Depreciation and Interest • • • • •
Crew and Overheads • • • • •
Passenger Time • • • • •
Oil • • • •
Road Works • •
Traffic Safety • • •
Vehicle Noise • • •
Vehicle Emissions • • •
Most important factors contributing > 70% of RUC

RUE Components
z MT Vehicle
V hi l operating
ti costs
t (VOC)
z MT Travel time costs (TTC)
z NMT Time and operating costs (NMTOC)
z Accident costs (AC)
RUE = RUC + Emissions + Noise

RUC = VOC + TTC + NMTOC + AC
Computational Logic
Calculate Free Speed

p for each vehicle type
yp
Calculate for each traffic flow period:
Ï Traffic flow in PCSE/hr
Ï Vehicle operating speed (Speed-flow curve)
Ï Speed
p change
g cycle
y (Acceleration
( noise))
Ï Vehicle operating costs
Ï Travel time costs
Ï Accident costs
Mechanistic Models
Predict that the RUE are proportional to

the forces acting on the vehicle
By
yqquantifying
y g the magnitude
g of the forces
opposing motion one can establish fuel
and tyre consumption
Mechanistic models allow for changes in
vehicle characteristics
Flexible when trying to apply the models
to different conditions.
Forces Opposing Motion
Calculates:
z aerodynamic resistance
vel and
(Fa) D ire c tio n of Tra
Accele
ration
z rolling resistance (Fr) Fg

Fi
Fa
z gradient resistance (Fg)

θ
z curvature resistance
(F )
(Fcr) Fr
z inertial resistance (Fi) Ft
Uses more detailed

equations than HDM-III
Speed Prediction
Road Conditions Curvature
Driver Gradient
Vehicle Roughness
Vehicle Speed
Traffic Conditions Sight Distance
Road Environment Width
Other Factors
Speed Model
Based on HDM-III probabilistic speed

model
Free speed predicted to be a function of
constraining speeds
z Desired speed - f(driver behaviour)
z Drive speed - f(power-to-weight ratio)
z Roughness - f(roughness and suspension)
z C
Curves - f(radius
f( di off curvature))
z Braking - f(downgrade length, brakes)
Vehicle Speed Models
z Desired speeds
Ï User specified for 2
2-lane
lane road
Ï Adjusted for Width, Friction, NMT, Speed Limit
z Free speeds
Ï Uphill (steady state)
Ï Downhill (steady state)
Ï Constrained by limiting factors
z Operating Speeds
Ï Calculated for traffic flow periods
Ï Adjusted for Speed
Speed-Flow
Flow effects
Desired Speeds
VDES = f (VDES2, CW)
VDESIR0 = VDES*XFRI*XNMT
VDESIR = min(VDESIR0,PLIMIT*ENFAC/3.6)
z VDES2 = Specified desired speed on a 2- lane road
z VDES = Calculated desired speed
z VDESIR0 = Adjusted for NMT & side-friction
z VDESIR = Adjusted for speed limit enforcement
z PLIMIT = Posted speed limit in km/h
z ENFAC = Enforcement factor (1.10)
z XFRI = Roadside friction factor
z XNMT = Non-motorised transport factor
Effect of Width on Speed
VDESIR' in m/s
a3
1m
a2
1m
a1
a0
Hoban et al
Hoban, al. (1984) HDM-Q Pan (1995) PhD Thesis
CW1 CW2 Road Width in m

Free Speeds
Constrained by:
y
Ï Desired Speed (VDESIR)
Ï Drive Power (VDRIVE)
Ï Braking Power (VBRAKE)
Ï Road Curvature (VCURVE)
Ï Riding Quality (VROUGH)
Speed Model Form
Model form:
⎛ σ2 ⎞
exp⎜ ⎟
⎝ 2⎠
VSS = β
⎡ 1 1 1 1 1
⎤
⎛
⎢⎜ 1 ⎞ β ⎛ 1 ⎞ β ⎛ 1 ⎞ β ⎛ 1 ⎞ β ⎛ 1 ⎞ β
⎟ +⎜ ⎟ +⎜ ⎟ +⎜ ⎟ +⎜ ⎟ ⎥
⎢⎝ VDRIVE ⎠ ⎝ VBRAKE ⎠ ⎝ VCURVE ⎠ ⎝ VROUGH⎠ ⎝ VDESIR ⎠ ⎥
⎣ ⎦
Examples of Speed Predictions
80
VB RAK E - B raking Speed
70 V CURV E - Curve Speed
60
V ROUGH - Roughnes s S peed
50
S pe e d in m /s
40 V DRIV E - Driving Speed
30
V DES IR - Des ired Speed
20
10
0
-10 -8 -6 -4 -2 0 2 4 6 8 10
Gra die nt in pe r ce nt
Traffic Interactions
HDM-III did not consider traffic

interactions
HDM-95 considered effects of traffic
interactions on speeds but not on VOC
(fuel but only through speed reduction)
HDM 4 h
HDM-4 has expanded
d d the
th HDM
HDM-9595
approach to consider other VOC
components
Acceleration-Volume Effects
Increased VOC consumption

due to traffic congestion
Model Basis
3-Zone speed-flow model predicts that, as

flows increase so do traffic interactions
As interactions increase so do
accelerations and decelerations
HDM-4 adopted concept of ‘acceleration
noise’
i ’ - the
th standard
t d d deviation
d i ti off
acceleration
Acceleration Noise
Uncongested
Congested
0
Acceleration in m/s/s
Acceleration Noise
Modelled using two components: traffic

induced and ‘natural’
natural noise
Traffic noise function of flow
Natural noise function of:
z driver’s natural variations
z road alignment
z roadside friction
z non-motorised transport
z roughness
Fuel Consumption
TRACTIVE FORCES ACCESSORIES INTERNAL

Replaced HDM
HDM-III
III Rolling, air, inertia, grade
and cornering resistance
C li ffan,
Cooling
power steering,
air conditioner,
ENGINE
FRICTION
Brazil model with alternator, etc.
one based on DRIVE - TRAIN

INEFFICIENCIES
ARRB ARFCOM
model TOTAL POWER
Predicts fuel use ENGINE FUEL EFFICIENCY FACTOR
as function of ESTIMATED FUEL CONSUMPTION
power usage:
IFC = max(MinFuel, ξ Ptot)
Model Parameters
Two basic model parameters used:

z idle fuel rate
z fuel conversion efficiency factor
Parameters can be readily derived from
other fuel models
A range of values provided for different
vehicle types from various published
sources
Effect of Speed on Fuel Consumption
200
180
160 India-1
India-2
140 India-3
onsumption in L/1 000km
Caribbean
K eny a
120
100
80
Fuel C o
60
40
20
0
0 20 40 60 80 100 120
Speed in km/h
Implications of New Model
Lower rates of fuel consumption than

HDM-III for many vehicles
Effect of speed on fuel significantly lower
for passenger cars
Considers other factors; e.g. surface
t t
texture andd type
t
Model can be used for congestion
analyses
l
Effects of Traffic Interactions on Fuel
Simulation model part of HDM-Tools

Run as calibration routine once unless vehicle
characteristics changed
U
Uses M
Montet C
Carlo
l simulation
i l ti off a vehicle
hi l
travelling down a road with different levels of
acceleration noise
Determines additional fuel as function of noise
Results in matrix of values of dFUEL vs Mean
Speed vs Acceleration Noise
ects Traffic Interactions on Fuel Consumption
1.2
1.0
0.8
dFUEL 0.6
0.4
0.90
0.2
0 75
0.75
0.60
Acce le ra tion Noise
0.45 in m /s/s
0.0
0.30
10
20
0.15
30
40
50
60
70
0.00
80
0
7
90
S pe e d in km /h
100
Tyre Consumption
Tread wear
z amount of the tread worn due the
mechanism of the tyre coming into contact
with
ith the
th pavementt surface
f
Carcass wear
z combination
bi ti off fatigue
f ti and
d mechanical
h i l
damage to the tyre carcass - affects
number of retreads
Factors Influencing Tyre Consumption
Vehicle
Rubber
Tyre Load
Composition
Applied Force
Properties
Suspension Type
Tyre
Pavement Texture
Type/Construction
yp
Microtexture Tread Pattern
Macrotexture Aspect Ratio
Inflation Pressure
Tread Wear
Energy Per
Rate Per Unit
Unit Distance
of Energy
T
Temperature
t T
Temperature
t
Air Air
Road Road
Tyre Tyre
Interface Operating Conditions

Contaminants
Water Road Alignment
Dust, Mud Tyre Consumption Pavement Condition
Ice, Snow Driver Behaviour
Types of Tyre Models
MECHANISTIC EMPIRICAL
Detailed models Usually aggregate
Relate tyre models
consumption to Based on fleet survey
fundamental data
equations of motion
Developed from
controlled
experiments
Retreads
If tyre carcass is serviceable tyres will

often be retreaded (recapped)
Common with commercial vehicles
The likelihood of surviving for retread
depends on tyre technology and operating
conditions
diti
Decrease in tyre life with increasing
numberb off retreads
t d
Tyre Life and Survival vs No. of Retreads
age of Tyre India Brazil New

Zealand
Life Average Travel as Per Cent Average Travel as Per Cent Per Cent
km Percentage Surviving km Percentage Surviving Surviving
Travelled of New Travelled of New
by Tyre by Tyre
Survivors Survivors
New Tyres 32.,916 100 100 26,939 100 100 100
rst Retreadd 21,515 65 67 18,628 69 54 100
ond Retread 20.853 63 27 18,707 69 27 84
ird Retread 18,191 55 7 18,495 69 13 72
urth Retread 10,364 31 1 18,581 69 5 40
fth Retread - - - 19,165 71 2 11
xth
th Retread
R t d - - - - - 04
0.4 2
Effect of Congestion on Tyre Consumption
9.0
8.0
7.0
6.0
5.0
dTYRE
4.0
3.0
2.0
0.90
0.75
1.0
0.60
0.45 Acce le ra tion Noise
0.0 in m /s/s
0.30
10
20
0.15
30
40
50
60
70
0.00
80
0
7
90
S pe e d in km /h
100
Parts and Labour Costs
Vehicle maintenance and repair costs:

Usually largest single component of VOC
In HDM-III user’s had choice of Kenya, Caribbean,
India and Brazil models
All g
gave significantly
g y different predictions
p
Most commonly used Brazil model had complex
formulation
Few studies were found to have calibrated model
Brazil Parts - Roughness
1.8
1.6
m .
hicle Price/1000 km
1.4
1.2
PC and LDV
Parts Consumpttion as % New Veh
1.0
0.8
MT
0.6
HT
AT
0.4
0.2 HB
0.0
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0
Roughness in IRI m/km

Adjusted Roughness
4.0
3.8
3.6 IRI0 = 4.0

Adjusted Roughness in IRI m/km .
3.4
IRI0 = 3.5
3.2
3.0
IRI0 = 3.25
2.8
IRI0 = 3.1
2.6
24
2.4
2.2
2.0
2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0
Actual Roughness in IRI m/km
Parts Model Parameters
Estimated from HDM-III Brazil model

Exponential models converted to linear models
which gave similar predictions from 3 - 10 IRI
Roughness effects reduced 25% for trucks
For cars, roughness effects same as for trucks
For heavy buses, roughness effects reduced further
25%
Implications of Changes
Vehicle Increase in Parts Consumption as Fraction of New Vehicle Price (x 10-3)

From 3-10 IRI
CKM = 100,000 CKM = 300,000
HDM III
HDM-III HDM 4
HDM-4 HDM III
HDM-III HDM 4
HDM-4
Passenger Car 4.70 1.49 6.59 2.08
Light Truck 2.45 1.49 3.67 2.23
Heavy Truck 1.98 1.49 2.58 2.23
Articulated Truck 1.42 1.49 2.14 2.23
Heavy Bus 0.20 0.99 0.34 1.68
Congestion Effects
Parts consumption is assumed to increase

under congested conditions
Use equation:
z PARTS = PARTS (1 + CPCON dFUEL)
Default value for CPCON is 0.10 indicating
that a 100% increase in fuel results in a
10% increase in parts
Utilisation and Service Life
HDM-III
Contained three utilisation methods:
z Constant Kilometreageg
z Constant Hours
z Adjusted
j Utilisation
Contained two service life methods:
z Constant Service Life
z de Weille’s Varying Service Life
Utilisation and Service Life
HDM-4 has either constant or ‘Optimal

Life’ service life
Utilisation function of hours worked for
work vehicles; lifetime kilometreage for
private vehicles
Optimal Life Method
Proposed by Chesher and Harrison (1987)

based upon work by Nash (1974)
Underlying philosophy is that the service
life is influenced by operating conditions,
particularly roughness
R l t lif
Relates life -- and
d capital
it l costs
t -- to
t
operating conditions
OL Method
Discounted Area = New Vehicle Price Running Costs
RUN(OL)
Costs per year
C
OL
Vehicle Age in years
Application of OL Method
HDM Tools contains a calibration routine

for the OL method
User defines the replacement vehicle
value, an estimated lifetime utilisation, and
the roughness where this lifetime
utilisation applies
Software establishes the effect of
roughness on lifetime utilisation
xample of Roughness on Lifetime Utilisation
100
90
80
Optimal Life as Percentage of Baseliine Utilisation .
70
60
50
40
30
20
10
0
0 2 4 6 8 10 12 14 16 18 20
Capital Costs
HDM-III
Used a simple linear model for depreciation
Affected
Aff t d by
b operating
ti conditions
diti through
th h
the effects of speed on utilisation and
speed on service life (de Weille’s method)
Depreciation in HDM-4
Depreciation calculated multiplying the

replacement vehicle price by the following
equation: (1 - 0.01 RVPLTPCT )
DEP = 1000
LIFEKM
The replacement vehicle price is reduced
b a residual
by id l value
l which
hi h can be
b a
function of roughness
Th denominator
The d i t is
i the
th lifetime
lif ti utilisation
tili ti
which may be constant or predicted with
the OL method to be a function of
Constant Service Life
Equations depend on the percentage of

private use:
z LIFEKM = LIFE x AKM < 50%
z LIFEKM = S x HRWK x LIFE > 50%
Example of OL Roughness on Depreciation
7
PC LT MT
HT AT LB
5
ht/km .
MB HB MC
Deprreciation Cost in Bah
0
0 5 10 15 20 25
Interest Costs
Interest costs are the replacement vehicle

price multiplied by the following equation:
AINV 1000
INT = 0.5
100 S HRWK 0
Function of speed and hours worked as

well as the interest rate
Oil Consumption
HDM-III only function of roughness

Model contains two components
z Fuel use due to contamination
z Fuel use due to operation which is
proportional to fuel use
OIL = OILCONT + OILOPER SFC

Travel Time Components
Passenger Working Hours

1000 PAX PCTWK
PWH =
100 S
Passenger Non-Working Hours
1000 PAX (100 - PCTWK )
PNH =
100 S
Crew Hours
1000 (100 − PP )
CH =
100 S
Accident Rates and Exposure
The accident rate is the number of

accidents divided by the exposure
ACCYR
ACCRATE =
EXPOSURE
The exposure is calculated differently for

intersections and sections
AADT 365
EXPOSINT =
10 6
AADT 365 SECTLEN
EXPOSSEC =
10 6
Modelling Approach
HDM-4 does not contain predictive models

Uses a series of look-up tables
Accidents function of data in HDM-4
database
VOC Calibration Procedure (Level 1)
Vehicle mass & ESAL

R dC
Road Capacity
it & SSpeed-Flow
d Fl factors
f t
Vehicle service life
V hi l Utilisation
Vehicle Utili ti (Annual
(A l KM & Hours)
H )
Desired speed
Vehicle engine power, speed (rpm) & braking
Tyre characteristics (type, rubber volume, etc)
Vehicle depreciation
Aerodynamic factors
THE END
TRAFFIC
Linda Parsley
y
Overview of Presentation
Traffic data input

Traffic categories
Representative vehicles
Traffic composition
Traffic volumes and growth rates
Axle loading
Speed-flow type
Traffic flow pattern
Uses of traffic data in HDM-4
pavement deterioration
p
vehicle operating costs
t
travel
l time
ti
quantities of vehicle exhaust emissions
calculation of energy use
economic analysis
Location of input traffic data
Definition of project details:

Road sections
Vehicle Fleet
Normal Traffic
Specification of Alternatives:
Improvement
Impro ement standard
Configuration
Traffic Categories
Normal traffic
z Existing traffic and expected growth on the
road both with and without the investment
Diverted traffic
z Traffic diverted to the road as a result of
th investment
the i t t
Generated traffic
z N
New ttraffic
ffi iinduced
d d by
b the
th investment
i t t
Representative Vehicles
Purpose:
z to group vehicles with similar
characteristics to simplify input
Method of grouping:
z composite vehicle - average characteristics
of vehicles within the group
z actual
t l vehicle
hi l - actual
t l characteristics
h t i ti off
most common vehicle of the group
Vehicle Hierachy
Three levels:
Categories
z motorised and non-motorised traffic;
Classes e.g.
z passenger cars, trucks, buses, carts….
Types e.g.
z small car, medium car, large car, Toyota
Corolla, Mercedes Benz…...
Motorised traffic
Motorised
Motorcycles Passenger cars Utilities Trucks Buses
Motorcycles Small car Light delivery Light truck Minibus
M di
Medium car Li h goods
Light d H
Heavy truck
k Li h bus
Light b
Large car 4 wheel drive Medium truck Medium bus
Articulated truck Heavy bus
Category Coach
Class
Type
Non-Motorised traffic
Non-Motorised
Pedestrian Bicycle Cycle-Rickshaw Animal Cart Farm Tractor
Pedestrian Bicycle Passenger Animal Cart Farm Tractor

(Commercial)
Freight
(Commercial)
Freight
Category (Private)
Class
Type
Volume, composition and growth rates
Volume - AADT for MT & NMT
Composition - representative vehicles (%)
Growth - for specified periods

z Annual percentage increase %
z Annual increment of AADT
z A t l AADT
Actual
for each section

Axle loading
For each representative

p vehicle
Number of axles
Equivalent standard axles
z affects paved road deterioration
N b off li
Number light
ht vehicles
hi l (< 3
3.5
5 tons)
t )
z affects deterioration of unsealed roads
Traffic Interaction Modelling
HDM-III did not consider traffic

interactions
HDM-95 considered effects of traffic
interactions on speeds but not on other
VOC (fuel but only through speed
reduction)
HDM-4 expanded the HDM-95 approach to
consider other VOC components
Traffic interactions
Passenger car space equivalence
Speed flow types
Traffic flow patterns

Equivalencies
Used to convert heterogeneous

g stream to
equivalent homogeneous stream
Two types
z PCE/PCU: Passenger Car Equivalencies
z PCSE P
PCSE: Passenger C
Car S
Space E
Equivalencies
i l i
HDM-4 uses PCSE
Differences between PCU and PCSE
PCU/PCE
z consider two factors:
z space occupied by vehicle
z speed effects
z used in highway capacity calculations

PCSE (HDM-4)
z considers only space occupied
z speed effects considered separately
through speed model
PCSE
Length Gap
Space (m)
HDM-4 Speed-Flow Model
S3
S2
S1
Snom
Sult
Speed
Qo Flow in PCSE/h Qnom Qult
Km/h
Speed-flow model requirements
Requires speeds
z Free speed (calculated)
z Snom - Speed at nominal capacity (calculated)
z S lt
Sult - Speed
S d att ultimate
lti t capacity
it
Requires capacities
z Q0 - Volume where interactions affect speeds
z Qnom - Nominal capacity
z Qult - Ultimate capacity
specified in Configuration
Types of Capacities
Nominal Capacity
z the capacity where the traffic stream is
travelling at a constant speed
z sustainable over long time periods
Ultimate Capacity
z maximum possible capacity before flow
breaks down
Annual Distribution of Hourly Flows
Flow Periods
Peak
ow Next to Peak
r Medium flow
N t tto L
Next Low
Overnight
N mber of Ho
Number Hours
rs in the Year
Non-Motorised Traffic
Impact of NMT
NMT affect
z Motorised traffic flow and speed
z Motorised traffic operating costs
Impact is specified through a friction factor
NMT also
l affect
ff t
z Safety related costs
z Road maintenance costs
these are not considered in HDM-4
Other effects on speeds
Impact of roadside activities - side friction

Impact of road widths as limiting speeds
Speed limits
Enforcement factor
Key differences from previous models
Unlimited number of representative

p
vehicles
Effects of congestion on speed,
speed fuel,
fuel tyres
and maintenance costs
Non-motorised transport modelling
Definition of project details:

Road sections
z AADT, speed factors, speed-flow type,
t ffi fl
traffic-flow type
t
Vehicle Fleet
z axle loads, PCSE, speed parameters
Normal Traffic
z composition, growth rates
Specification of Alternatives:
Improvement standard
z Diverted traffic / generated traffic
Configuration
z speed flow types, traffic flow patterns
Summary
Aimed to show where traffic data is

required in HDM-4 and its relevance and
impact in the model
Included traffic interaction and speed
estimation which leads to VOC and RUE
Strategy Analysis Case Study
Long Term Budget Forecasts and Performance Trends
1 Background
This case study demonstrates the use of HDM-4 in budget forecasting. The objective of this
study is to determine the required funding levels for user defined network performance
standards, and to demonstrate the effect of budgetary constraints on the long term
performance trends. This involves defining the road network in terms of representative
sections and assigning alternative maintenance standards (investment alternatives) to each
section.
Strategy analysis is normally undertaken at infrequent intervals and is primarily a tool for
resource allocation by, or for, the top managers within a road agency. It is to be used for the
analysis of a chosen road network as a whole to prepare long-range planning estimates of
expenditure needs for road development and preservation under various budget scenarios.
Strategy analysis will be required to produce expenditure estimates for medium to long term
periods of between 5 to 40 years, under various budget constraints, maintenance standards,
and road use scenarios. It will predict key performance indicators of the road network such as
average roughness trends, surface distress, road user costs, safety levels and environmental
effects.
When conducting a strategy analysis, the physical road network may be categorised into a
matrix defined by user-defined parameters such as road class, surface type, pavement
condition or traffic flow. Strategy analysis utilises a model of the road network (or sub-
network) to simulate network-wide performance under specified road use characteristics,
maintenance standards, development standards and resource constraints.
The main difference between strategy and programme analysis is that programme analysis
deals with physical links and sections that can be identified from the road network. In strategy
analysis, the road system essentially loses its individual link and section characteristics as
these are grouped into representative sections to model the road network being analysed.
Thus, whereas in programme analysis the basic unit of the network is an identifiable road
section, the basic unit in strategy analysis is a representative section. Since sections are the
basic units of analysis in HDM-4 for both applications, similar types of economic analysis and
optimisation methods can be utilised for both modules.
2 Case study data

The case study data are included in the default database installed with the HDM-4 software.
The data for this case study are located in the Strategies folder in the Case Studies
Workspace. The name of the case study is 1: Long term budget forecasts and performance
trends.
2.1 Strategy Details

The case study is analysed over a 20 year life cycle analysis period For this case study the
selected optimisation method is maximisation of benefits (NPV).
T3 Strategy Case 1 Handout.doc 1

STRATEGY ANALYSIS CASE STUDY
Road Network
The road network has been modelled using eleven representative sections. The representative
sections for bituminous (Paved) roads have been based on traffic volume (High, Medium or
Low) and road condition (Good, Fair and Poor), giving a total of nine sections.
Unsealed (Gravel) roads have been classified by traffic volume alone, since condition may
change rapidly in any given year. For this case study, no gravel sections are subjected to high
traffic, so the gravel network is represented by two sections only, one for Medium traffic and
one for Low traffic.
Double-click on a Section Description (or selecting the section and clicking the View/Edit
Section button) gives access to the section's details. Some relevant characteristics of the
representative sections are summarised in Table C1.1and Table C1.1. Note that, for both
bituminous and gravel sections, the condition must be defined at the start of the analysis
period. As the analysis period commences in the year 2000, this means that the condition data
must be defined for the year 1999 (that is, end of 1999) or earlier.
Table C1.1
Details of representative sections for gravel roads
Section ID GMT GLT

Length (km) 1385 1760
Traffic level Medium Low
AADT (1998) 175 75
Carriageway width (m) 6 6
Shoulder width (m) 1 1
Condition (1998)
Gravel thickness (mm) 100 70
Roughness IRI (m/km) 9 12
A tabulated data summary is useful in checking consistency of data across sections and also
within a section. This includes checking that the condition classification is compatible with
the condition parameters, and, with bituminous pavements, that the pavement type (defined
under Section/General) is compatible with the pavement details (surfacing material type,
surfacing thicknesses and dates of previous works) defined under Section/Pavement.
Remember that the dates of previous works refer to HDM-4 Works Types. Hence the date of
last resurfacing refers to Works Type resurfacing which includes works activities such as
surface dressing and slurry seal. Works activities such as overlaying, mill and replace, and
inlay are classed as Works Type rehabilitation and would be recorded under date of last
rehabilitation. The bituminous pavement types are reviewed below.

Table C1.1
Details of representative sections for bituminous roads
Section ID PHTGC PHTFC PHTPC PMTGC PMTFC PMTPC PLTGC PLTFC PLTPC
Length (km) 234 392 437 306 483 615 410 670 720
Traffic level High High High Medium Medium Medium Low Low Low
AADT 6200 5240 5180 2500 2300 2060 1400 1150 970
AADT year 1998 1998 1998 1998 1998 1998 1998 1998 1998
Condition (1999) Good Fair Poor Good Fair Poor Good Fair Poor
Roughness (IRI) 2.5 4.0 6.5 2.5 4.4 6.5 3 4.5 6.1
Area of all cracks (%) 1 3 10 1 5 18 1 7 12
Ravelled area (%) 1 5 15 1 8 26 1 10 20
No of (small) potholes per km 0 3 20 0 5 40 0 8 60
Edge break (m2/km) 0 5 10 0 5 15 2 12 30
Mean rut depth (mm) 2 5 10 2 5 12 3 10 15
Texture depth (mm) 3 2 0.5 3 2 0.5 3 2 0.5
Skid resistance (SCRIM 50 kph) 0.7 0.5 0.4 0.7 0.5 0.4 0.7 0.5 0.3
Drainage Excellent Fair Poor Excellent Good Fair Good Fair Poor
Pavement type AMAP AMAP AMAP STGB STGB STGB STGB STGB STGB
Surfacing material type AC AC AC DBSD DBSD AC SBSD SBSD SBSD
New surf thick (mm) 50 50 50 25 25 50 15 15 15
Old surf thick (mm) 50 50 50 50 50 0 0 0 0

Date of last works types:
- reconstruction 1991 1991 1991 1991 1991 1991 1991 1991 1991
- rehabilitation (overlay) 1996 1996 1996 1991 1991 1991 1991 1991 1991
- resurfacing (reseal) 1996 1996 1996 1996 1996 1991 1991 1991 1991
- preventative treatment 1996 1996 1996 1996 1996 1991 1991 1991 1991
SN after last treatment 2.70 2.70 2.70 2.07 2.07 2.26 1.67 1.67 1.67
Key:
Pavement type Traffic class Pavement Condition
P = Paved HT = High Traffic GC = Good Condition
G = Gravel MT = Medium Traffic FC = Fair Condition
LT = Low Traffic PC = Poor Condition

Bituminous pavement types

Table C1.1 indicates Pavement Type AMAP for the bituminous sections carrying high traffic
levels (PHTGC, PHTFC and PHTPC). This is based on an original construction of 50 mm
AC on a granular roadbase (that is, pavement type AMGB) in 1991 (date of last
reconstruction), subsequently overlaid with 50 mm AC in 1996 (date of last rehabilitation).
This results in the updated Pavement Type AMAP, and the dates of last resurfacing and last
preventative treatment are set to 1996 (equal to date of last rehabilitation). Note the most
recent surfacing thickness set at 50 mm (that is, overlay) and previous thickness is also 50 mm
(original construction).
Pavement Type STAP is indicated for the sections identified as PMTGC and PMTFC. This is
based on an original construction of 50 mm AC on a granular roadbase (pavement type
AMGB) in 1991 (date of last reconstruction), subsequently surface dressed in 1996 (date of
last resurfacing). This gives the updated pavement type STAP. The date of the last
rehabilitation is set to 1991 (equal to date of original construction), and the date of last
preventative treatment is set to 1996 (equal to date of last resurfacing). Note the most
recent surfacing thickness is 25 mm (double surface dressing), and previous surfacing
thickness is 50 mm (original construction).
The road section identified as PMTPC has pavement type AMGB that represents the original
construction of 50 mm AC on a granular roadbase in 1991 (date of last reconstruction). Note
that the dates of last rehabilitation, resurfacing and preventative treatment are also set to 1991
(equal to date of last reconstruction). In this case the most recent surfacing thickness is set to
50 mm (original construction), while the previous surfacing thickness is zero.
The three sections with low traffic level (PLTGC, PLTFC and PLTPC) have pavement type
STGB, representing the original construction of double surface dressing on a granular
roadbase in 1991 (date of last reconstruction). The dates of last rehabilitation, resurfacing and
preventative treatment are also set at 1991. The most recent surfacing thickness is set to 25
mm (double surface dressing), while the previous surfacing thickness is zero.
Vehicle Fleet
The vehicles included in the analysis are selected from the pre-defined Western Province
fleet. Individual vehicle attributes may be reviewed by double-clicking on the appropriate
vehicle type description (or by selecting the section and clicking the Edit Vehicle button).
Normal Traffic
The initial traffic composition and growth rates (by vehicle type) can be reviewed by double-
clicking the section description. The initial vehicle compositions are summarised in Table
C1.3. Note that the initial composition is defined for same year as the AADT. The annual
growth rates may be specified for several periods, but must at least be specified for a period
commencing from the analysis period start year (that is, 2000 in this case study). For example
the AADT for section PHTGC is specified for 1998, and the initial vehicle composition
applies to that year. The annual growth rates have been defined only for the period
commencing 2000, and in this case are also applied between 1998 and 2000.
2.2 Assignment of Standards
Alternatives
For this case study, four investment alternatives are considered for each of the representative
sections, ranging from the provision of routine pavement maintenance only (Alternative 1) to

an ideal maintenance case (Alternative 4). For each investment alternative, Road Works
Standards have been pre-defined as shown in Table C1.2 and Table C1.4.
Table C1.2
Assigned Road Works Standards - short codes
Investment alternative High traffic Medium Low traffic

traffic
Bituminous (paved)
1. Patch & Crack seal PHT1 PMT1 PLT1
2. Patch, Reseal & Reconstruct PHT2 PMT2 PLT2
3. Patch, Overlay & Reconstruct PHT3 PMT3 PLT3
4. Patch, Reseal, Overlay & Reconstruct PHT4 PMT4 PLT4
Gravel (unpaved)
1. Grade & Spot regravel G1 G1
2. Grade, Spot regravel & Regravel G2 G2
4. Maintenance & Improvement G1/GU4/PLT3 G1/GW4/G2
With bituminous road sections, maintenance standards have been assigned by alternative and
traffic class. For gravel road sections, a maintenance standard has been defined for each
investment alternative, and an improvement standard has been introduced to Alternative 4
(that is, upgrading gravel roads with medium traffic followed by paved road maintenance, and
widening gravel roads with low traffic). Note that Road Authorities may wish to assign
standards by Road Class, which may be represented in some countries by Surface Class alone.
With condition responsive maintenance intervention, the interval between maintenance
applications will clearly be influenced by the traffic volume.
The definition of investment alternatives for each of the representative sections is indicated in
Table C1.4. This shows the short code identifier associated with each of the pre-defined Road
Works Standards (that is, Maintenance Standards and Improvement Standards).
Maintenance Standards for bituminous sections
The Maintenance Standards for bituminous sections are reviewed first. These are based on
different combinations of the following works activities/operations.
If more than one of these activities is triggered in a given year, the works item that is highest
in the hierarchy of road works (that is, with lowest ranking value) will be applied. Note that
the routine pavement works (patching and crack sealing) have the same ranking, and both can
be applied in the same maintenance year. The hierarchy is defined by the ranking of road
works shown in Chapter D2 of the Analytical Framework and Model Descriptions.

Table C1.3
Details of initial vehicle composition and growth rates
Traffic level
High Medium Low
Initial Annual growth Initial Annual growth Initial Annual growth

Vehicle type composition rate composition rate composition rate
% % % % % %
Leyland bus 10 3 10 3 0 0
Mercedes heavy truck 10 3 25 3 10 3
Nissan pickup 15 3 20 3 20 3
Tata 7.5t lorry 15 3 0 0 10 3
Toyota Corolla car 40 3 30 3 40 3
Toyota Hiace minibus 10 3 15 3 20 3
Total (%) 100 100 100
Note: Traffic details (AADT & associated year, initial composition and associated growth rates) are assigned by section.

Table C1.4
Definition of investment alternatives for representative sections
Represent- Surface Traffic Investment Maintenance Improvem Future

ative class class alternative standard ent maint-
section standard enance
1 PHT1 - -
2 PHT2 - -
PHTGC B H
3 PHT3 - -
4 PHT4 - -
1 PHT1 - -
2 PHT2 - -
PHTFC B H
3 PHT3 - -
4 PHT4 -
1 PHT1 - -
2 PHT2 - -
PHTPC B H
3 PHT3 - -
4 PHT4 - -
1 PMT1 - -
2 PMT2 - -
PMTGC B M
3 PMT3 - -
4 PMT4 - -
1 PMT1 - -
2 PMT2 - -
PMTFC B M
3 PMT3 - -
4 PMT4 - -
1 PMT1 - -
2 PMT2 - -
PMTPC B M
3 PMT3 - -
4 PMT4 - -
1 PLT1 - -
2 PLT2 - -
PLTGC B L
3 PLT3 - -
4 PLT4 - -
1 PLT1 -
2 PLT2 -
PLTFC B L
3 PLT3 - -
4 PLT4 - -


1 PLT1 - -
2 PLT2 - -
PLTPC B L
3 PLT3 - -
4 PLT4 - -
1 G1 - -
2 G2 - -
GMT U M
3 G3 - -
4 G1 GU4 PLT3
1 G1 - -
2 G2 - -
GLT U L
3 G3 - -
4 G1 GW4 G2
Note: Details of Road Works Standards identified by the above short codes are given in Table
C1.8 and Table C1.9 for bituminous and gravel sections respectively.
Table C1.5
Works activities considered for bituminous sections
Works type Works activity Ranking

Routine Pavement Patching 22
Crack sealing 22
Resurfacing Surface dressing double 16
Rehabilitation Overlay dense graded asphalt 9
Reconstruction Pavement reconstruction 6
! Alternative 1
Comprises patching and crack sealing only. This represents the base case (or do minimum)
alternative. In this case study the intervention levels for patching vary from 10 potholes/km
for sections with high traffic, to 50 potholes/km for sections with low traffic (see Table
C1.8).
! Alternative 2
Includes patching, resealing and reconstruction. The objective of this alternative is to adopt
relatively inexpensive treatments which will maintain the existing road in a reasonable
condition for as long as possible until the eventual need for reconstruction.
! Alternative 3

Includes patching, overlaying and reconstruction. This alternative is effectively Alternative 2

with resealing replaced by overlaying. Alternative 3 introduces rehabilitation works at a pre-
defined roughness level which will maintain a higher serviceability level than Alternative 2,
as the structural strength of the pavement will increase at each overlay application.
! Alternative 4
Includes patching, resealing, overlaying and reconstruction. This alternative combines the
benefits of resealing and overlaying within one standard, which should reduce the required
frequency of the relatively expensive overlay works.
The Maintenance Standards for alternatives 2, 3 and 4 include reconstruction as a Works Item.
Details of the pavement layer thicknesses adopted for different road classes are given in Table
C1.8
The pavement details will normally be derived using appropriate design charts, based on a design
subgrade CBR and estimated cumulative traffic loading over the desired design life. The
subgrade CBR value is not requested within the reconstruction works item, as this is already
specified at section level (Section/Pavement). Details of the Maintenance Standards associated
with each alternative are summarised in Table C1.8.
Table C1.6
Pavement reconstruction details for bituminous sections
Traffic class Where specified

Details
Light Medium Heavy in software
Pavement Type STGB STGB AMGB M/Design
Surface material DBSD DBSD AC M/Design
Surface thickness (mm) 25 25 50 M/Design
Structural Number (SN) 1.8 2.1 2.9 M/Design
Subgrade CBR (%) 8 8 8 Section/Pavement
Note: M/Design refers to the Design Tab under the appropriate Maintenance Standard
Maintenance and Improvement Standards for gravel sections
The Maintenance and Improvement Standards for gravel sections are based on combinations of
the works activities listed in Table C1.7.
Table C1.7
Works activities considered for gravel sections
M/I Works type Works activity Ranking

M Routine pavement Grading 7
Spot regravelling 7
M Resurfacing Regravelling 6
I Widening Partial widening 5
Lane addition 4
I Upgrading Upgrading to a new surface class 2

The order of ranking of the Maintenance works items gives regravelling highest priority (that is,
with lowest ranking value). Spot regravelling and Grading have equal ranking, and both of these
can be performed in a given analysis year. Details of precedence rules for road works are given in
Chapter D2 of the Analytical Framework and Model Descriptions.
! Alternative 1
Comprises grading and spot regravelling, both condition responsive. This represents the base
case (do-minimum alternative).
! Alternative 2
Includes three works items, grading, spot regravelling and regravelling (resurfacing), all
condition responsive. Grading is triggered by roughness, while spot regravelling and
regravelling are both triggered by the thickness of gravel surfacing. This alternative aims to
maintain a reasonable thickness of gravel surfacing and so provide continuous protection to
the pavement subgrade.
! Alternative 3
Includes the same works items as Alternative 2, but the intervention levels are set to provide
a higher level of serviceability that should trigger the works more frequently.
! Alternative 4
Introduces upgrading to a paved standard for sections with medium traffic (representative
section GMT), and widening for sections with low traffic (section GLT). This represents the
ideal investment alternative. Suitable maintenance is provided before and after the respective
improvements. Note that the respective Improvement Standards are effective from the year
2003, with works scheduled to start in the year 2004. This means that no maintenance will
be applied during the year 2003, the year prior to improvement.
3 Funding requirements
3.1 No budget constraint

This step in Strategy Analysis produces an unconstrained work programme, based on a life-cycle
analysis that considers the different alternatives for each representative section. The alternative
giving the highest NPV is assigned to each section.
The Run Setup screen for Strategy Analysis is identical to that for life-cycle analysis in
Programme Analysis. The multi-year forward programme option (available under Programme
Analysis) is not applicable to strategy analysis that is concerned with longer term planning.
On completing Run Setup, press the Start button to commence the economic analysis, which
produces a costed work programme containing, for each representative section, the alternative
with the highest NPV. This represents an unconstrained programme, with the total financial cost
given in the cumulative cost column. The road works assigned to each section represent the
optimum maintenance and improvements that should be applied to the road network in
accordance with the specified standards.
3.2 Effect of budget constraints

This step in Strategy Analysis is concerned with budget optimisation, whereby a budget is
defined, and an optimisation procedure is selected which aims to produce a works programme that
satisfies an objective function, that is:

either maximisation of benefits (NPV)

or maximisation of the improvement in network condition
For this case study, two budget scenarios are considered, one at 75% of funding requirements
(from Step 3), and another at 50%. For each of these, optimisation has been based on
maximisation of benefits, and the long term performance trends predicted.
After performing the budget optimisation, revised work programmes are produced. This has a
total cost that falls within the specified budget. The road sections are listed in priority order in
accordance with the incremental benefit/cost ranking procedure.

Table C1.8
Details of investment alternatives for bituminous sections
Investment Maintenance Standard / Condition responsive Units Intervention levels by traffic class
alternative Works item criteria
High traffic Medium traffic Low
traffic
1 Patch & crack seal (PHT1) (PMT1) (PLT1)
Pothole patching Potholing no/km >=10 >=25 >=50
Transverse thermal cracks AND no/km >=15 >=15 >=15

Crack sealing
Wide structural cracking % >=10 >=10 >=10
2 Patch, reseal & reconstruct (PHT2) (PMT2) (PLT2)

Pothole patching Potholing no/km >=10 >=25 >50
Reseal (SBSD) Total damaged area % >=20 >=30 >40
Reconstruct Roughness m/km IRI >=10 >=11 >12
3 Patch, overlay & reconstruct (PHT3) (PMT3) (PLT3)

Roughness AND m/km IRI >=5 >=5.5 >=6

Overlay
Cracking % >=5 >=5 >=5
Reconstruct Roughness m/km IRI >=10 >11 >=12

4 Patch, reseal, overlay & reconstruct (PHT4) (PMT4) (PLT4)

Reseal (SBSD) Total damaged area % >=20 >=30 >=40

Overlay
Cracking % >=5 >=5 >=5
Reconstruct Roughness m/km IRI >=10 >=11 >=12
Notes:
1 Short code identifiers for Maintenance Standards shown in parentheses, for example (PHT1)
2 SBSD = Single Bitumen Surface Dressing

Table C1.9
Details of investment alternatives for gravel sections
Alt Sections M/I Road Works Standard Effective Responsive Criteria [M] Details (design)
from
Works Item year Start Date/Duration [I]
1 GMT & M G1: Grade & Spot regravel 2000

GLT
Grading Roughness >= 8 IRI
Spot regravelling GThk <= 100 mm Replace 10% annual material loss
2 GMT & M G2: Grade , Spot regravel and Regravel 2000

GLT
Regravelling GThk <= 50 mm Final gravel thickness 150 mm
3 GMT & M G3: Grade, Spot regravel and Regravel 2000

GLT

4 GMT M G1 : Grade and Spot regravel (details above) 2000
I GU4: Pave section GMT in 2004 2003 Start year 2004 Upgrade to STGB with 2 m widening
Duration 5 years New Speed-Flow Type is 2-lane wide
M PLT3: Patch, Overlay & Reconstruct 2009
Pothole patching Potholing >= 50/km
Overlay Roughness >= 6 IRI AND
Cracking >= 5%
Reconstruct Roughness >= 12 IRI
4 GLT M G1 : Grade and Spot regravel (details above) 2000
I GW4: Widen section GLT in 2004 2003 Start year 2004 Widen by 3 m
M G2: Grade, Spot regravel and Regravel (details above) 2009
Notes:
Column 1 Alt = Investment Alternative
Column 3 M/I = Maintenance or Improvement Standard
Column 6 GThk = Gravel thickness


Strategy Analysis Case Study
Long Term Budget Forecasts and Performance Trends
1 Unit Aims
1. To introduce the HDM-4 Strategy Analysis module concepts so that Trainees are able to
forecast long term road budget requirements and road network performance trends.
2. To highlight the differences between Strategy and Programme Analysis Applications of
HDM-4.
3. To familiarise Trainers with the use of HDM-4 Strategy Analysis data input and output
modules.
2 Unit Objectives
1. To describe the concepts of the HDM-4 Strategy analysis application.
2. To use the HDM-4 model to calculate a long term budget requirement
3. To alter maintenance standards and unit costs and study the impact on budget
requirements.
4. To export the Strategy Analysis results and plot performance indicators using a
spreadsheet.
3 Background
section.
Strategy analysis is normally undertaken at infrequent intervals and is primarily a tool for
resource allocation by, or for, the top managers within a road agency. It is to be used for the
analysis of a chosen road network as a whole to prepare long-range planning estimates of
expenditure needs for road development and preservation under various budget scenarios.
Strategy analysis will be required to produce expenditure estimates for medium to long term
periods of between 5 to 40 years, under various budget constraints, maintenance standards,
and road use scenarios. It will predict key performance indicators of the road network such as
average roughness trends, surface distress, road user costs, safety levels and environmental
effects.
condition or traffic flow. Strategy analysis utilises a model of the road network (or sub-
network) to simulate network-wide performance under specified road use characteristics,
maintenance standards, development standards and resource constraints.
T3 Strategy Case 4 Notes.doc 1

4 Case study data

trends.
4.1 Strategy Details

The case study is analysed over a 20 year life cycle analysis period For this case study the
selected optimisation method is maximisation of benefits (NPV).
Road Network
The road network has been modelled using eleven representative sections. The representative
sections for bituminous (Paved) roads have been based on traffic volume (High, Medium or
Low) and road condition (Good, Fair and Poor), giving a total of nine sections.
Unsealed (Gravel) roads have been classified by traffic volume alone, since condition may
traffic, so the gravel network is represented by two sections only, one for Medium traffic and
one for Low traffic.
Double-click on a Section Description (or selecting the section and clicking the View/Edit
Section button) gives access to the section's details. Some relevant characteristics of the
representative sections are summarised in Table C1.1and Table C1.2. Note that, for both
bituminous and gravel sections, the condition must be defined at the start of the analysis
period. As the analysis period commences in the year 2000, this means that the condition data
must be defined for the year 1999 (that is, end of 1999) or earlier.

Table C1.1
Length (km) 234 392 437 306 483 615 410 670 720
AADT 6200 5240 5180 2500 2300 2060 1400 1150 970
AADT year 1998 1998 1998 1998 1998 1998 1998 1998 1998
Roughness (IRI) 2.5 4.0 6.5 2.5 4.4 6.5 3 4.5 6.1
Ravelled area (%) 1 5 15 1 8 26 1 10 20
Edge break (m2/km) 0 5 10 0 5 15 2 12 30
Mean rut depth (mm) 2 5 10 2 5 12 3 10 15
Texture depth (mm) 3 2 0.5 3 2 0.5 3 2 0.5
New surf thick (mm) 50 50 50 25 25 50 15 15 15
Old surf thick (mm) 50 50 50 50 50 0 0 0 0

- reconstruction 1991 1991 1991 1991 1991 1991 1991 1991 1991
Key:
Pavement type Traffic class Pavement Condition
P = Paved HT = High Traffic GC = Good Condition
G = Gravel MT = Medium Traffic FC = Fair Condition
LT = Low Traffic PC = Poor Condition

Table C1.2
Section ID GMT GLT

Length (km) 1385 1760
AADT (1998) 175 75
Condition (1998)

Vehicle Fleet
The vehicles included in the analysis are selected from the pre-defined Western Province
fleet. Individual vehicle attributes may be reviewed by double-clicking on the appropriate
vehicle type description (or by selecting the section and clicking the Edit Vehicle button).

Normal Traffic
The initial traffic composition and growth rates (by vehicle type) can be reviewed by double-
clicking the section description. The initial vehicle compositions are summarised in Table
C1.4. Note that the initial composition is defined for same year as the AADT. The annual
growth rates may be specified for several periods, but must at least be specified for a period
4.2 Assignment of Standards
Alternatives
Table C1.3

traffic
Bituminous (paved)
Gravel (unpaved)

Table C1.4
Traffic level
High Medium Low

% % % % % %
Tata 7.5t lorry 15 3 0 0 10 3
Total (%) 100 100 100

investment alternative, and an improvement standard has been introduced to Alternative 4 (that is,
upgrading gravel roads with medium traffic followed by paved road maintenance, and widening
gravel roads with low traffic). Note that Road Authorities may wish to assign standards by Road
Class, which may be represented in some countries by Surface Class alone. With condition
responsive maintenance intervention, the interval between maintenance applications will clearly
be influenced by the traffic volume.
Table C1.5

1 PHT1 - -
2 PHT2 - -
PHTGC B H
3 PHT3 - -
4 PHT4 - -
1 PHT1 - -
2 PHT2 - -
PHTFC B H
3 PHT3 - -
4 PHT4 -
1 PHT1 - -
2 PHT2 - -
PHTPC B H
3 PHT3 - -
4 PHT4 - -
1 PMT1 - -
2 PMT2 - -
PMTGC B M
3 PMT3 - -
4 PMT4 - -
1 PMT1 - -
2 PMT2 - -
PMTFC B M
3 PMT3 - -
4 PMT4 - -
1 PMT1 - -
2 PMT2 - -
PMTPC B M
3 PMT3 - -
4 PMT4 - -


1 PLT1 - -
2 PLT2 - -
PLTGC B L
3 PLT3 - -
4 PLT4 - -
1 PLT1 -
2 PLT2 -
PLTFC B L
3 PLT3 - -
4 PLT4 - -
1 PLT1 - -
2 PLT2 - -
PLTPC B L
3 PLT3 - -
4 PLT4 - -
1 G1 - -
2 G2 - -
GMT U M
3 G3 - -
4 G1 GU4 PLT3
1 G1 - -
2 G2 - -
GLT U L
3 G3 - -
4 G1 GW4 G2
Note: Details of Road Works Standards identified by the above short codes are given in Table
C1.9 and Table C1.10 for bituminous and gravel sections respectively.
If more than one of these activities is triggered in a given year, the works item that is highest in
the hierarchy of road works (that is, with lowest ranking value) will be applied. Note that the
routine pavement works (patching and crack sealing) have the same ranking, and both can be
applied in the same maintenance year. The hierarchy is defined by the ranking of road works
shown in Chapter D2 of the Analytical Framework and Model Descriptions.

Table C1.6

Crack sealing 22
! Alternative 1
Comprises patching and crack sealing only. This represents the base case (or do minimum)
alternative. In this case study the intervention levels for patching vary from 10 potholes/km
for sections with high traffic, to 50 potholes/km for sections with low traffic (see Table
C1.9).
! Alternative 2
Includes patching, resealing and reconstruction. The objective of this alternative is to adopt
relatively inexpensive treatments which will maintain the existing road in a reasonable
condition for as long as possible until the eventual need for reconstruction.
! Alternative 3
Includes patching, overlaying and reconstruction. This alternative is effectively Alternative 2

with resealing replaced by overlaying. Alternative 3 introduces rehabilitation works at a pre-
defined roughness level which will maintain a higher serviceability level than Alternative 2,
as the structural strength of the pavement will increase at each overlay application.
! Alternative 4
Includes patching, resealing, overlaying and reconstruction. This alternative combines the
benefits of resealing and overlaying within one standard, which should reduce the required
frequency of the relatively expensive overlay works.
C1.8
The pavement details will normally be derived using appropriate design charts, based on a design
subgrade CBR and estimated cumulative traffic loading over the desired design life. The
subgrade CBR value is not requested within the reconstruction works item, as this is already
specified at section level (Section/Pavement). Details of the Maintenance Standards associated
with each alternative are summarised in Table C1.9.

Table C1.7

Details
The Maintenance and Improvement Standards for gravel sections are based on combinations of
the works activities listed in Table C1.8.
Table C1.8

Spot regravelling 7
Lane addition 4
The order of ranking of the Maintenance works items gives regravelling highest priority (that is,
with lowest ranking value). Spot regravelling and Grading have equal ranking, and both of these
can be performed in a given analysis year. Details of precedence rules for road works are given in
! Alternative 1
Comprises grading and spot regravelling, both condition responsive. This represents the base
case (do-minimum alternative).
! Alternative 2
regravelling are both triggered by the thickness of gravel surfacing. This alternative aims to
maintain a reasonable thickness of gravel surfacing and so provide continuous protection to
the pavement subgrade.

! Alternative 3
Includes the same works items as Alternative 2, but the intervention levels are set to provide
a higher level of serviceability that should trigger the works more frequently.
! Alternative 4
section GMT), and widening for sections with low traffic (section GLT). This represents the
ideal investment alternative. Suitable maintenance is provided before and after the respective
improvements. Note that the respective Improvement Standards are effective from the year
2003, with works scheduled to start in the year 2004. This means that no maintenance will
be applied during the year 2003, the year prior to improvement.
5 Funding requirements
5.1 No budget constraint

This step in Strategy Analysis produces an unconstrained work programme, based on a life-cycle
analysis that considers the different alternatives for each representative section. The alternative
giving the highest NPV is assigned to each section.
Programme Analysis. The multi-year forward programme option (available under Programme
Analysis) is not applicable to strategy analysis that is concerned with longer term planning.
with the highest NPV. This represents an unconstrained programme, with the total financial cost
given in the cumulative cost column. The road works assigned to each section represent the
optimum maintenance and improvements that should be applied to the road network in
accordance with the specified standards.

defined, and an optimisation procedure is selected which aims to produce a works programme that
satisfies an objective function, that is:
After performing the budget optimisation, revised work programmes are produced. This has a
accordance with the incremental benefit/cost ranking procedure.

Table C1.9
Investment Maintenance Standard / Condition responsive Units Intervention levels by traffic class
alternative Works item criteria
High traffic Medium traffic Low
traffic

Crack sealing
2 Patch, reseal & reconstruct (PHT2) (PMT2) (PLT2)

3 Patch, overlay & reconstruct (PHT3) (PMT3) (PLT3)


Overlay
Cracking % >=5 >=5 >=5

4 Patch, reseal, overlay & reconstruct (PHT4) (PMT4) (PLT4)


Overlay
Cracking % >=5 >=5 >=5
Notes:

Table C1.10
from
Works Item year Start Date/Duration [I]

GLT

GLT

GLT

Cracking >= 5%
Notes:


Project Analysis Case Study
Upgrading a Gravel Road to a Bituminous Pavement
1 Introduction
This case study presents the economic analysis of a project to upgrade an existing gravel road to a
paved standard. The existing road is 50 km long and passes through varying topography. For
analysis purposes, three sections, based on geometry, pavement condition, and traffic volume can
represent the road. Traffic and condition data are available from surveys undertaken in 1998. The
gravel thickness in 1998 was 150 mm.
The objective of the case study is to demonstrate the structure and capabilities of HDM-4, and the
steps needed to undertake the project appraisal.
The purpose of the appraisal is to assess the economic benefits resulting from the proposed
investment. (This differs from a financial appraisal that is concerned with the means of financing
a project and the financial profitability of the project). The economic feasibility of the project is
assessed by comparison against a base-line project alternative (that is, a without project
alternative). The project alternatives are:
! Without Project
Maintain existing gravel road.

! With Project
Maintain existing gravel road before upgrading to a bituminous pavement, followed by

maintenance of the bituminous pavement.
2 Case study data

The project start year has been defined as 2000 with the analysis period (duration) specified as 20
years. The road network to be used for the project has been pre-defined under the name
Northern Province in the Road Network folder. Similarly the vehicle fleet has been pre-defined
in the Vehicle Fleet folder under the same name. The road sections (within the Northern Province
road network) and vehicles types (from the Northern Province vehicle fleet) to be used for the
analysis are selected under the Select Sections and Select Vehicles tabs respectively, as
discussed below:
2.1 Road Sections

The road section is the basic entity for all calculations of pavement deterioration, construction and
maintenance costs, and economic analysis. Each section is considered homogeneous in terms of
its physical attributes (for example, road class, climate, carriageway width, geometry, pavement
condition, traffic flow, and axle loading).
For this case study, the road has been divided into three sections based on those physical
attributes that vary along its length. Table A1.1 contains a summary of the key physical attributes
for each section.
Note that in this case study, no shoulders are specified, and the full width of gravel surfacing is
assigned as carriageway width. In practice it is often difficult to identify shoulders on a gravel
road.
T3 Project Case 4 Description.doc 1

PROJECT ANALYSIS CASE STUDY
Table A1.1
Case study 1 - Summary of section attributes
Section ID B001-01 B001-02 B001-03

General
Section Name Town A to Town B Town B to Town C Town C to Town D
Length (km) 20 10 20
Carriageway width (m) 7 6 8
Shoulder width (m) 0 0 0
Number of lanes 2 2 2
Number of shoulders 0 0 0
AADT (1998) 400 200 600
Flow direction 2 - way 2 - way 2-way
Geometry (before
upgrading)
Rise and Fall (m/km) 10 15 30
*No rises and falls (no/km) 0.1 0.4 1.0
Av horiz curv (deg/km) 50 300 15
*Superelevation % 3.0 3.0 3.0
*σdral 0.1 0.1 0.1
Speed limit (kph) 100 80 100
*Speed limit enforcement 1.1 1.1 1.1
Altitude (m) 120 295 370
Condition 1998
Gravel Thickness 150 150 150
Roughness (IRI, m/km) 8 8 8
Notes:
1 Geometry Data marked with an * (asterisk) is stored under Section/Details/Alignment.

Other geometric data are held under Section/Geometry.
Traffic flow (AADT) used in project analysis is defined by section at project level under and will
often be an update of base traffic data retained at section level.
Each section has been assigned a unique Description and ID, either or both of which may be used
to identify the section on HDM-4 reports. In this case study, the section description identifies
towns at the ends of each section. It is helpful if the Section ID includes reference to the road
class and road number (denoted by B001 in this case study).
2.2 Vehicle Fleet

The vehicles for this study are selected from the pre-defined Northern Province Vehicle Fleet.

2.3 Normal Traffic

For this case study, only one growth period is indicated over the 20-year analysis period. Several
growth periods may be defined within the analysis period. The user is encouraged to define
future growth rates (by vehicle type) as accurately as possible. Note that negative growth rates
are permitted.
3 Alternatives
The two project alternatives considered in this case study are defined below. The first alternative,
Without Project, represents a continuation of current maintenance practice. The second
alternative, With Project, represents the implementation of the project to upgrade the existing
gravel road to paved standard.
! Without Project
This project alternative comprises grading every six months; spot regravelling to replace
30% of material lost each year (if the gravel thickness falls below 100 mm, and gravel
resurfacing (whenever the gravel thickness falls below 50 mm).
Note that in a given year, if gravel resurfacing is triggered, this operation supersedes grading
and spot regravelling.
! With Project
This project alternative includes upgrading the road in the year 2004 (duration of works is
two years for road sections B001-01 and B001-03, and one year for section B001-02).
Before upgrading, the existing gravel road will be maintained, by grading, every six months
and spot regravelling to replace 30% of material lost each year (if the gravel thickness falls
below 100 mm). The latter works will maintain some protection of the subgrade prior to
upgrading.
After upgrading, the road will receive routine maintenance in the form of crack sealing (if
wide structural cracking reaches 5%) and patching (if the severely damaged area reaches
5%).
The analysis period is 20 years, starting in the year 2000. Note that Analysis by project option
has been selected meaning that the economic analysis results will be combined for all three
sections in the Project.
A schematic overview of the Project Alternatives and their constituent Road Works Standards is
shown in Table A1.2. Note that Road Works Standards are sub-divided into Maintenance
Standards denoted by M , and Improvement Standards denoted by I .
The Road Works Standards and associated works assigned to each project alternative are
summarised in Table A1.2. A summary of the intervention limits for maintenance works is given
in Table A1.3. These tables allow the user to check consistency of data across alternatives.
Each Road Works Standard has an associated date, assigned at project level, which represents the
year from which the standard takes effect. For the Without Project alternative, each of the three
road sections has been assigned a Maintenance Standard named Gravel Road Maintenance.
This assignment commences in the year 2000, which is the first year of the 20-year analysis
period (2000-2019).
For the With Project alternative, each section has been assigned a set of Road Works Standards
as follows:
! M Maintenance Standard for gravel road prior to upgrading

! I Improvement Standard representing the upgrading works
! M Maintenance Standard for paved road after upgrading
Note that (for a given feature) only one Maintenance Standard or Improvement Standard will be
effective in any analysis year. However a Maintenance Standard may include more than one
works item which could be implemented in a given year, for example, grading and spot
regravelling on unsealed roads, crack sealing and patching on bituminous roads.
Table A1.2
Details of road works standards for each project alternative
Project Section Road Works Effective Maintenance works/

alternative ID Standards from year Improvement type
M Gravel Road 2000 Grading (GRADE6)
Maintenance
(GRAVEL) Spot regravelling (SPG100)
B001-01
Gravel resurfacing (RESURF)
Without Project M Gravel Road 2000 Grading (GRADE6)
Maintain Maintenance
Gravel (GRAVEL) Spot regravelling (SPG100)
Road B001-02

Maintenance
B001-03
M Maintenance before 2000 Grading (GRADE6)

upgrading (BEFORE)
Spot regravelling (SPG100)
I Pave Section B001-01 2004 Upgrading

B001-01 in 2004 (PAVE01)
M Crack sealing and 2006 Crack sealing (SEAL)

patching paved road
(SEAPAT) Patching (PATCH)

upgrading (BEFORE)
With Project I Pave Section B001-02 2004 Upgrading

B001-02 in 2004 (PAVE02)
Upgrade Gravel
Road M Crack sealing and 2005 Crack sealing (SEAL)
patching paved road

upgrading (BEFORE)
B001-03 I Pave Section B001-03 2004 Upgrading

in 2004 (PAVE03)

patching paved road

Table A1.3
Intervention limits for maintenance works
Road works Effective Maintenance * Last Max Max Interval AADT

standard from year works Year IRI Quantity
S/R Min Max Min Max
M Gravel Road 2000 Grading S 2019 30 n/a 180 days 750 days 0 100,000
Maintenance
Spot regravelling R 2019 30 100 m3/km/yr n/a n/a 0 100,000
Gravel resurfacing R 2017 30 n/a 3 yrs 99 yrs 0 100,000
M Maintenance before 2000 Grading S 2003 30 n/a 180 days 10000 days 0 100,000
upgrading
M Crack sealing and 2006 Crack sealing R 2019 12.5 1500 n/a n/a 0 100,000
patching paved road 2
m /km/yr
Patching R 2019 12.5 1500 n/a n/a 0 100,000

2
m /km/yr
Notes:
n/a not applicable
* S = Scheduled intervention
R = Responsive intervention

4 Project alternatives
The use of the HDM-4 software to set up the Road Works Standards for this case study is
described in more detail below. Each Project Alternative includes one or more Maintenance
Standards and/or Improvement Standards that are assigned to be effective from a specified date.
Each Maintenance Standard may define several Works Items, and each Improvement Standard
defines a specific improvement type.
Project Alternative:
Without Project
M
Gravel Road Maintenance
Works Item: Works Item: Works Item:

Grading Spot regravelling Gravel resurfacing
With Project
M M
I
Maintenance before Crack sealing and
Upgrading works
upgrading patching paved roads
Works Item: Works Item: Works Item: Works Item:

Grading Spot regravelling Crack sealing Patching
Improvement Improvement Improvement

type: type: type:
upgrading upgrading upgrading
section section section
B001-01 B001-02 B001-03
Figure A1.2 Schematic diagram showing details of Project Alternatives

4.1 Without Project Alternative

This Project Alternative includes one Maintenance Standard associated with the existing gravel
road. The Without Project alternative involves grading every six months, spot regravelling to
replace 30% material lost each year (if the gravel thickness falls below 100 mm), and gravel
resurfacing (if the gravel thickness falls below 50 mm). The original gravel thickness was 150
mm (1998).
Note that in a given year, spot regravelling and grading are both permitted as these are
complimentary works. However if gravel resurfacing is triggered, this will supersede the other
works. Hence in a given year, the possible works under the Gravel Road Maintenance
Maintenance Standard are:
! Grading every six months, or
! Spot regravelling AND Grading every six months, or

! Gravel resurfacing
The Works Items and Operations associated with the Maintenance Standard are shown in
Figure A1.1. Full details of the logical rules associated with the implementation of works are
given in Chapter D1 of the Analytical Framework and Model Descriptions.
Name: Gravel Road Maintenance

Maintenance Standard:
Code: GRAVEL
Name: Grading Spot regravelling Gravel resurfacing

Works Items:
Code: GRADE6 SPG100 RESURF
Regravelling/
Operation/Works Activity: Grading Spot regravelling
Resurfacing
Figure A1.1 Hierarchy within a Maintenance Standard
4.2 With Project Alternative

This Project Alternative includes the following Road Works Standards:
Prior to upgrading, the existing gravel road will be given routine maintenance in the form of
grading every six months and spot regravelling (if the gravel thickness falls below 100 mm). The
works are specified on the Maintenance Standard window.
For this case study, the geometric characteristics are significantly different for each road section,
as are the pavement design details and duration of the upgrading works. The different

Improvement Standards for the three sections are reviewed below. A comparison of selected
section details, before and after upgrading, is given in Table A1.4.
Table A1.4
Section details before and after upgrading works
Section ID B001-01 B001-02 B001-03

Section Name Town A - Town B Town B - Town C Town C - Town D
Speed flow type B Two lane road Two lane road Two lane road
A Two lane road Two lane road Two lane road
Traffic flow pattern B Seasonal Seasonal Seasonal
A Seasonal Seasonal Seasonal
Road class B Secondary/Main Secondary/Main Secondary/Main
A Primary/Trunk Primary/Trunk Primary/Trunk
Surface class B Unsealed Unsealed Unsealed
A Bituminous Bituminous Bituminous
Pavement type B Gravel Gravel Gravel
A Surface Treatment on Surface Treatment on Surface Treatment on

Granular Base Granular Base Stabilised Base
Carriageway width (m) B 7 6 8
A 7 6 8
Speed limit (km/h) B 100 80 100
A 120 100 120
Upgraded SN 1.84 1.61 2.62
(excluding subgrade)
Upgrade financial cost 280000 300000 360000
(US dollars/km)
Notes:
B = Before upgrading
A =After upgrading
The Improvement Standard for this section has a different pavement design to that specified for
Section B001-01. The construction period for Section B001-02 is one year.
The Improvement Standard section B001-03 has a different pavement design (detailed under
Pavement below) to the previous sections. The construction period for Section B001-03 is two
years.
For this road section, the proposed pavement design comprises a 150 mm stabilised sub-base
placed on existing formation level with a 200 mm stabilised road base and a double surface
dressing.

4.3 Maintenance Standards

After upgrading, the paved road is to receive routine maintenance, defined under the Maintenance
Standard Crack sealing and patching Paved Road. As summarised in Table A1.2, this includes
two works items:
1 Crack Sealing if wide structural cracking reaches 5%
2 Patching if the severely damaged area reaches 5%
Both works are condition responsive (as defined above) and are effective from the year following
the completion of the upgrading works (that is, effective from 2006 for sections B001-01 and
B001-03, and 2005 for section B001-02.) The effective start date is specified at project level
(under Alternatives).
5 HDM-4 results
The costs and benefits of the Upgrade gravel road alternative will be compared with those for
the Maintain gravel road alternative. Note that the HDM-4 Workspace may hold data for any
number of projects with associated network, fleet and work standards data. The run that has been
set up will consider only those project alternatives defined under 'Alternatives', with the specified
selected sections and vehicles.
For this case study, accident costs, energy balance analysis, emissions calculations, and
acceleration effects are not included in the analysis.
The pre-defined report categories available are:
! Deterioration/Work Effects
! Road User Effects
! Environmental Effects
! Cost Streams
Note that reports generally refer to data presented in tabular form. Certain reports are also
available in graphical form (Charts). For this case study, the Timing of Works (by section) report
should be printed.
Considering the Without Project alternative, the report indicates that, at Section B001-01 (Town
A-Town B), gravel resurfacing would be implemented in the years 2003, 2007, 2010, 2013 and
2016 (with associated preparatory spot regravelling). Spot regravelling would be implemented
every year except the first year and each year immediately following gravel resurfacing. Grading
is implemented every six months as scheduled. These works are specified under the Maintenance
Standard Gravel Road Maintenance described previously.
The With Project alternative for the same road section (B001-01) confirms that the upgrade to
paved standard would be implemented during 2004 and 2005. Before upgrading, the existing
road is graded at six monthly intervals between 2000 and 2003 with spot regravelling. After
upgrading, the paved road has no works specified until 2016, when patching is listed each year
between 2016 and 2019 (end of analysis period). These works are specified under the three works
standards:
! M Maintenance before upgrading
! I Pave Section B001-01 in 2004
! M Crack sealing and patching paved road

The effect of these works on roughness is indicated in the Deterioration report and associated
chart showing roughness at the end of the year.
For this case study, the overall NPV is reported as US$ 0.624 millions. The breakdown by
section indicates that two sections give a positive NPV.
Section ID Section description Discounted NPV
(millions of US Dollars)
B001-01 Town A to Town B 0.117
B001-02 Town B to Town C -0.316
B001-03 Town C to Town D 0.823
All sections 0.624
Note that the traffic levels are lowest for Section B001-02 (negative NPV), also that maximum
benefits are derived for Section B001-03 which has highest traffic flow. A revised project
comprising sections B001-01 and B001-03 only would give an overall NPV of US$ 0.940
millions.

UNIT TITLE: DERIVATION OF OPTIMUM MAINTENANCE STANDARDS
UNIT AIMS
1. To practice the specification and assignment of HDM-4 maintenance standards so that Trainers are able to
apply a range of maintenance works.
UNIT OBJECTIVES
1. To prepare input data required for investigating optimum road maintenance standards for paved roads using
HDM-4.
2. To assign a range of maintenance standards to road sections and run HDM-4.
3. To interpret the results and recommend optimum maintenance standards for paved roads.
GUIDELINES
This case study presents the economic analysis of alternative maintenance standards for two paved road sections
that are 36.5 km and 10.6 km long, respectively. The existing roads have surface dressing and asphalt concrete
pavement surfacing in good condition, and carry approximately 6 000 and 16 000 vehicles per day (AADT),
respectively. The objective of this case study is to determine the optimum maintenance intervals for resealing
(surface dressing) and for overlays on the two roads, respectively.
The case study data is provided in the T3 Case studies workspace/database installed with the HDM-4 software.
The data for this case study are located in the Projects folder in the Case Studies Workspace. The name of
the case study is Case 2: Optimum maintenance intervals. To locate the data for this case study:
The case study data includes the road network data set (CASE3 Project Sections) and the vehicle fleet (CASE2
Vehicle Fleet) required for the analyses.
The first step in the procedure would be to calibrate the road sections using the calibration factors derived from
the previous case study.
The second step in the procedure should be to create additional maintenance standards for the two roads. The
data provided includes one example each for:
Create additional maintenance standards by copying the above to create separate standards for:
After creating the additional maintenance standards, open the Projects Folder and go to the Specify
Alternatives group-tab. Select one of the sections and Edit Alternatives in order to assign the maintenance
standards. Note that the surface dressing standards should be applied to the section with ST surfacing, and the
overlays should be applied to the AC section.
When all of the alternatives have been assigned maintenance standards, go to the Analyse Projects group-tab
and specify the run data directory, and finally run HDM-4.
The results of the analyses can be viewed from the Generate Reports group-tab. The main report containing a
summary of the economic analysis results is the Benefit Cost Ratios within the Cost Streams report group.
HDM-4 Train the Trainers Course Page - 1

KEY POINTS
The optimum economic maintenance standards are defined as the maintenance interval (scheduled or
responsive) with the maximum NPV. This can be illustrated by exporting the Benefit Cost Ratios report to an
Excel spreadsheet and plotting the NPV against the maintenance interval (years) or intervention criterion (IRI).
Another useful plot would be of the NPV against the increase in agency cost. The resulting curve from this is
defined as the efficiency frontier. This will be described in detail in the Programme Analysis case study.
Those Trainers who complete the above exercise within a short time could attempt to derive the optimum
grading and regravelling frequencies for the Gravel road that is included in the Road Network database.
REFERENCES

3. HDM-4 Volume 5: A Guide to Calibration and Adaptation (Chapters 6 and 7)
T3 Project Case 2 Notes.doc
HDM-4 Train the Trainers Course Page - 2

DESCRIPTION
Project Case Study

Optimum Maintenance Standards
1 Description
1.1 Optimum Rehabilitation Standards for a Paved Road
This case study presents the economic analysis of alternative maintenance standards for two
paved road sections, 36.5 km and 10.6 km long, respectively. The existing roads have surface
dressing and asphalt concrete surfacing, and carry approximately 6 000 and 16 000 vehicles
per day (AADT), respectively. The objective of this case study is to determine the optimum
maintenance intervals for resealing (surface dressing) and for overlays on the two roads,
respectively.
1.1 Locate the case study data
The case study data is provided in the T3 Case studies workspace/database installed with the
HDM-4 software. The data for this case study are located in the Projects folder in the Case
Studies Workspace. The name of the case study is Case 2: Optimum maintenance
intervals.
To locate the data for this case study:
! Go to the Case Studies Workspace
! Open the Projects folder
1.1 Case study input data
1.1.1 General
The General Tab page confirms the project description, analysis type, analysis period and the
pre-defined Road Network and Vehicle Fleet.
This case study is presented as a section analysis. The road under study is represented by one
section, and the different rehabilitation proposals represent section alternatives. The road
section will be selected from the Northern Province network, stored in the Road Network
folder. The vehicles using the road will be selected from the Northern Province vehicle fleet,
stored in the Vehicle Fleet folder.
It is also possible to conduct the analysis by Project although we have only one road section.
The analysis by Project option would combine the results of selected sections into one
Project. The analysis period is defined by a base year 2000, and an analysis period of 20
years, (that is, 2000 - 2019).
1.1.2 Select Sections

The sections Tab page indicates that one section will be included in the analysis.
! Definition
The Definition Tab gives details of basic section characteristics including road class,
speed-flow type and traffic flow pattern. The Pavement Type has been specified as
It is other method saying SHIT! 1-1

PROJECT ANALYSIS CASE STUDY 2
Surface Treatment on Asphalt Pavement (STAP). It is important that the correct

Pavement Type is specified, as the deterioration relationships held in HDM-4 are defined
by Pavement Type (see Table A2.1).
Details of the existing pavement are discussed under Pavement below:
! Geometry
Details of the road section's geometry, including horizontal and vertical alignment, are
held under Section/Geometry and Section/Details/Alignment. The data indicates that
the road section under study is in rolling terrain (Rise + Fall = 20 m/km) and subject to a
speed limit of 100 kph.
Alignment data can be assessed from the existing mapping, if available, or measured by a
topographical survey.
! Pavement
The existing pavement construction (Figure A2.1) consists of an asphaltic concrete

surfacing (total thickness 125 mm), over a 200 mm thick granular roadbase and 150 mm
thick granular sub-base. The in situ subgrade CBR is 8%. The pavement details required
by HDM-4 for this pavement type (Surface Treatment on Asphalt Pavement) are
indicated on the Pavement Tab page below:
Note that the specification of Pavement Type (on the Section/Definition page) refers to
the current pavement construction and should be compatible with the Previous Works
details specified on the Pavement Tab page.
The Pavement Type is automatically updated immediately after any maintenance works.
The definition of bituminous Pavement Types based on surface and base types is given in
Table A2.1. (Table A2.2 contains descriptions of surface and base materials.) A
summary of Pavement Type resets after maintenance works is given in Table A2.3.
For the road section under study, the changes in Pavement Type since the date of the last
construction (1988) are summarised in Figure A2.1.
The last reconstruction was 1988, given by pavement type AMGB. After the application
of overlay in 1992, Table A2.3 shows that:
AMGB + Overlay = AMAP ...(1.1)
After surface dressing (reseal) in 1995 (last resurfacing), Table A2.3 gives:
AMAP + Surface Dressing = STAP ...(1.2)
(specified for this case study under Definition)

Note that within HDM-4, the dates of the Previous Works shown under
Section/Pavement are recognised as Age1, Age2, Age3 and Age4 as summarised below:
Age4 = Date of last reconstruction or new construction = 1988
Age3 = Date of last rehabilitation (overlay) = 1992
Age2 = Date of last resurfacing = 1995
Age1 = Date of last preventative treatment = 1995
Note that the dates refer to HDM-4 Works Types. The classification of Works Activities,
by Work Type, is summarised in Table A2.4.
T3 Project Case 2 Handout.doc 1-2

The dates Age2, Age3 and Age4 are used as a base line for the timing of scheduled
treatments (Age2 for surface treatments, Age3 for rehabilitation, Age4 for
reconstruction). For example, if surface dressing was requested every 4 years within a
Maintenance Standard effective from the year 2000, the first possible application would
be done immediately based on:
Date of last resurfacing (1995) + 4 < 2000 ...(1.3)
1995 (STAP) 25mm Surface Dressing New

Surfacing
1992 (AMAP) 50mm AC Overlay
100 mm
Previous Surfacing
1988 (AMGB) 50mm AC Surfacing
200mm Granular Roadbase
150mm Granular Sub-base
Subgrade (CBR 8%)
Figure A2.1 Case Study 2: Details of existing pavements and pavement type
resets

Table A2.1
Generic HDM-4 bituminous pavement types
Surface Surface Base Base Pavement

type material type material type
AC GB CRS AMGB
HRA GM
PMA AB AB AMAB
RAC SB CS AMSB
AM CM LS
PA TNA AMAP
SMA AP FDA
xx
CAPE GB CRS STGB
DBSD GM
SBSD AB AB STAB
ST SL SB CS STSB
PM LS
xx AP TNA STAP
FDA
Note: AM and ST surfacings on concrete pavements, that is, AMCP and STCP, are
modelled in HDM-4 as concrete pavement types in the rigid pavement sub-model.
Table A2.2
Descriptions of surface and base materials
Surface type Surface materials

Abbreviation Description Abbreviation Description
AC Asphaltic Concrete
CM Soft Bitumen Mix (Cold
Mix)
HRA Hot Rolled Asphalt
AM Asphalt Mix PA Porous Asphalt
PMA Polymer Modified
Asphalt
RAC Rubberised Asphalt
Concrete
SMA Stone Mastic
CAPE Cape Seal
DBSD Double Bituminous
Surface Dressing
ST Surface Treatment PM Penetration Macadam
SBSD Single Bituminous
Surface Dressing
SL Slurry Seal
Base types Base materials
AB Asphalt Base CRS Crushed Stone
AP Asphalt Pavement GM Natural Gravel
GB Granular Base CS Cement Stabilisation
SB Stabilised Base LS Lime Stabilisation
TNA Thin Asphalt Surfacing
FDA Full Depth Asphalt

Table A2.3
Pavement type resets after maintenance works
Works Existing pavement type

activity
AMGB AMSB AMAB AMAP STGB STSB STAB STAP

Routine works AMGB AMSB AMAB AMAP STGB STSB STAB STAP
Preventive AMGB AMSB AMAB AMAP STGB STSB STAB STAP

Treatment
Reseal STAP STAP / STAP STAP STGB STSB STAB STAP

STSB1
Overlay AMAP AMAP / AMAP AMAP AMGB AMSB AMAB AMAP

AMSB1
Inlay AMGB AMSB AMAB AMAP STGB STSB STAB STAP
Mill & replace **AP **AP **AP **AP N/A **SB **AB **AP
to intermediate
surface layer
Mill & replace **GB **SB **AB **AP **GB **SB **AB **AP
to base
Source: NDLI (1995)
Notes:
1 The pavement type depends on the critical thickness (Hmin) of the existing
bituminous surfacing that is user-definable in HDM Configuration.
** Indicates that these two characters are dependent on the specific works activity (or
operation) and the surface material.
N/A Not applicable.
The Pavement Tab page shows the current pavement strength, defined by the Adjusted
Structural Number of the pavement (Adjusted SNP). The Adjusted SNP incorporates a
weighting factor that reduces the contribution from the sub-base and subgrade, as
described in the Analytical Framework and Model Descriptions. (Note that the Adjusted
SNP is not the same as the Modified Structural Number used in HDM-III).
The following HDM-4 conventions should be noted:
1 SNP refers to the full structural number including contributions from the sub-base
and subgrade.
2 SN refers to the structural number comprising contributions from the surfacing,
roadbase and sub-base only.
For this case study, the Structural Number (SN) was entered directly together with the
subgrade CBR using Option (1) on the Section/Pavement screen. By clicking Option 1,
the calculated SNP value is displayed at the top of the screen.

Table A2.4
Ranking of road works applicable to the carriageway
Works type Works activity / operation Ranking Unit cost

New section Dualisation of an existing section 1 per km
Upgrading Upgrading to a new surface class 2 per km
Realignment Geometric realignment 3 per km
Lane addition 4 per m2 or per km

Widening
Partial widening 5 per m2 or per km
Reconstruction Pavement reconstruction 6 per m2 or per km
Mill and replace 7 per m2
Overlay rubberised asphalt 8 per m2
Rehabilitation Overlay dense-graded asphalt 9 per m2
Overlay open-graded asphalt 10 per m2
Inlay 11 per m2
Thin overlay 12 per m2
Cape seal with shape correction 13 per m2
Cape seal 14 per m2
Double surface dressing with shape correction 15 per m2

Resurfacing Double surface dressing 16 per m2
(Resealing) Single surface dressing with shape correction 17 per m2
Single surface dressing 18 per m2
Slurry seal 19 per m2
Preventive Fog sealing 20 per m2
Treatment Rejuvenation 21 per m2
Edge-repair1 22 per m2
Routine
Patching1 22 per m2
Pavement
Crack sealing1 22 per m2
Note that Option (2) allows the user to calculate the SNP value with the help of the SNP
Calculation Wizard, which can calculate SNP based on deflections or layer thickness
and coefficients.
! Condition
The condition of the road in 1998 can be reviewed under the Condition Tab page.
For project analysis, this data should preferably be collected as part of a detailed
condition survey. However, users can set up aggregate data specifying detailed values
under HDM Configuration. Default values are shown in Chapter D2.

1.1.3 Select Vehicles

This Tab page confirms the selection of vehicles (from the pre-defined Northern Province
fleet. Individual vehicle attributes can be reviewed by double-clicking on the appropriate
vehicle type description.
1.1.4 Define Normal Traffic

This Tab page confirms the volume of traffic (Motorised and Non-motorised AADT) using
the selected road section in the given year. The initial traffic composition and growth rates
(by vehicle type) can be reviewed by double-clicking the appropriate line.
1.1.5 Alternatives
The Alternatives Tab is displayed after selecting the Specify Alternatives button. The
Alternatives screen is split into two boxes. The upper box shows the names of the six section
alternatives that have been set up for this case study.
The bottom box shows details of the Road Works Standards (assignments) associated with
each section alternative.
The six section alternatives considered for this case study are defined below. The analysis
period is 20 years (from year 2000 - 2019).
Alternative Description
1 This is the do-minimum alternative. Routine pavement maintenance

is undertaken each year, as necessary, based on the pavement
condition. No rehabilitation works are undertaken until the road
condition reaches a poor condition, that is, reconstruction when IRI
>=12 AND total damaged area > = 40% of carriageway area.
2 With this alternative, a 50 mm asphaltic concrete overlay is applied
when the roughness level reaches 6 IRI OR when structural cracking
affects 15% of the carriageway area. As the roughness was at IRI 6
in 1998, this should trigger the first overlay in the year 2000. This
alternative includes routine pavement maintenance (condition
responsive) over the full analysis period.
3 This alternative is similar to Alternative 2, with an extra work item,
surface dressing, undertaken to restore skidding resistance when the
Sideways Force Coefficient (SFC) falls below 0.4. As with
Alternative 2, an overlay should be triggered in the year 2000.
Thereafter this alternative will show what effect the introduction of
regular surface dressing will have on the overlay frequency.
4 With this alternative the upper 75 mm of the bituminous surfacing is
milled out and replaced whenever the roughness value reaches 6 IRI
AND when structural cracking affects 20% of the carriageway area.
Routine pavement maintenance (condition responsive) is applied
throughout the analysis period.

5 This alternative is similar to Alternative 4 except that after 75 mm

surfacing is milled, a thickness of 100 mm asphaltic material is
placed, that is, effectively replacing the existing upper surfacing and
applying a 25 mm overlay. Routine pavement maintenance
(condition responsive) is applied throughout the analysis period.
6 With this alternative, an inlay is recommended, based on the mean rut
depth reaching 20 mm. It is anticipated that the outer wheelpaths
would be inlaid, representing some 25% of the carriageway area.
The Section alternatives can be summarised as follows:
Alternative Proposed maintenance works

1 Do-minimum alternative
Reconstruction if Roughness >= 12 IRI AND
Total Damaged Area > = 40% carriageway area
2 50 mm overlay if Roughness >= 6 IRI OR Cracking >= 15%
3 Surface dressing if SFC <= 0.4
50 mm overlay if Roughness >= 6 IRI OR Cracking >= 15%
4 Mill 75 mm and replace 75 mm if Roughness >= 6 IRI AND Cracking >= 20%
carriageway area
carriageway area
6 Inlay to wheel paths if Mean Rut Depth >= 20 mm
Note that each alternative includes routine pavement maintenance in the form of:
! Crack sealing if area of wide structural cracking > = 5%
! Patching if Severely Damaged Area >= 10%
The Road Works Standards and associated works assigned to each alternative are summarised
in Table A2.5. In this case study, each alternative is defined by one Maintenance Standard.
Note that an identical set of routine pavement works (crack sealing and patching) is common
to each Maintenance Standard. To be considered for implementation during any analysis
year, the routine works and rehabilitation works must be specified within the same
Maintenance Standard. (If routine works and rehabilitation works are specified as two
separate Maintenance Standards with the same effective from date, the software accepts only
one standard, that is, that which is assigned to the alternative first).
The intervention limits for the Works Items are summarised in Table A2.6.

Table A2.5
Case Study 2: Details of road works standards for each project alternative
Project alternative Road Works Standard Effective Maintenance works

from year
Alternative 1 M Routine + Reconstruct (Case Study 2000 Reconstruct at IRI 12 AND 40% Damage (RECON)
Routine + 2) (R&REC) Crack Sealing (CRKSL)
Reconstruct Patching (PATCH)
Alternative 2 M Routine + 50 mm Overlay 2000 Overlay 50 mm at IRI 6 (OVL50)
Routine + 50 mm overlay (R&OV50) Overlay 50 mm at 15% Cracking (OV50)
Crack Sealing (CRKSL)
Patching (PATCH)
Alternative 3 M Routine + SD + 50 mm Overlay 2000 Overlay 50 mm at IRI 6 (OVL50)
Routine + Surface Dressing (RSDOV) Overlay 50 mm at 15% Cracking (OV50)
+ Surface Dressing at SFC < 0.4 (SDRESS)
50 mm Overlay Crack Sealing (CRKSL)
Patching (PATCH)
Alternative 4 M Routine + Mill 75 mm & Replace 2000 Mill 75 mm + Replace 75 mm at 6 IRI (MILREP)
Routine + Mill 75mm 75 mm (MR75) Crack Sealing (CRKSL)
& Replace 75 mm Patching (PATCH)
Routine + Mill 75 mm 100 mm (MR100) Crack Sealing (CRKSL)
Alternative 6 M Routine + Inlay at Rut Depth 20 2000 Inlay at Rut Depth 20 mm (INLAY)
Routine + Inlay mm Crack Sealing (CRKSL)
Patching (PATCH)

Table A2.6
Case Study 2: Intervention limits for maintenance works
Maintenance Effective Maintenance *S/R Last Max Max Interval AADT

Standard from Works Year Quantity
(IRI) 2
(m /km/yr)
Min Max Min Max
year
1 Routine + Reconstruct 2000 Reconstruct R 2099 n/a n/a n/a n/a 0 100,000
Crack Sealing R 2099 12.5 1500 n/a n/a 0 100,000
Patching R 2099 12.5 100 n/a n/a 0 100,000
2 Routine + 2000 Overlay R 2099 12.5 n/a 3 yrs 99 yrs 0 100,000
50 mm Overlay Crack Sealing R 2099 12.5 1500 n/a n/a 0 100,000
Patching R 2099 12.5 100 n/a n/a 0 100,000
3 Routine + SD + 2000 Surface Dressing R 2099 12.5 n/a 1 yrs 99 yrs 0 100,000
50 mm Overlay Overlay R 2099 12.5 n/a 3 yrs 99 yrs 0 100,000
Patching R 2099 12.5 100 n/a n/a 0 100,000
4 Routine + 2000 Mill75&Replace75 R 2099 n/a n/a n/a n/a 0 100,000
Mill 75 mm Crack Sealing R 2099 12.5 1500 n/a n/a 0 100,000
& Replace 75 mm Patching R 2099 12.5 100 n/a n/a 0 100,000
6 Routine + Inlay 2000 Inlay R 2099 12.5 n/a n/a n/a 0 100,000
Patching R 2099 12.5 100 n/a n/a 0 100,000
Notes:
n/a = not applicable; * S = Scheduled intervention, R = Responsive intervention

Note that for certain works items, the unit costs of preparatory works are entered separately on
the Works Item/Costs Tab page. In Case Study 2, this applies to the following works items.
Alternative Maintenance works item Preparatory costs

Patching
2 50 mm overlay
Edge repair
Patching
50 mm overlay
Edge repair
3 Patching
Surface dressing Edge repair
Crack Sealing
Patching
6 Inlay Edge repair
Crack Sealing
Where HDM-4 does not offer a separate facility for entry of preparatory costs, such costs (if
any) should be included in the costs of the Works Item.
Issues relating to the Maintenance Works Standard for each alternative are discussed below:
! Alternative 1 - Maintenance Standard: Routine + Reconstruct
The works included in this Maintenance Standard are the routine pavement works (crack
sealing and patching) and reconstruction. The works are listed in the opening screen for
the Maintenance Standard:
The specification for each of these works can be reviewed by double-clicking the
appropriate works item name.
! Works Item: Crack Sealing and Patching
The specifications for these works, included in all alternatives for this case study, are
similar to those described for Case Study 1 within the Maintenance Standard Crack
Sealing and Patching Paved Road (see Appendix A1).
! Works Item: Reconstruct at IRI 12 and 40% damage
The specification for Reconstruction is held under the tabs:
General/Design/Intervention/Costs/Effects
Note that the Works Activity (Pavement Reconstruction) is specified under General.
The Design Tab shows the proposed new Pavement Type as Asphalt Mix on
Granular Base (AMGB) together with pavement details required by HDM-4.
Note that the details under Design refer to the new pavement construction. The
requested structural number refers to the layers above the subgrade.
The reconstruction Costs to be specified include all costs associated with the works.
The Tab pages specifying the Reconstruction Works Item are shown below:

! Alternative 2 - Maintenance Standard: Routine + 50 mm overlay
The works included in the Maintenance Standard are the routine pavement works (crack
sealing and patching) and a 50 mm thick overlay (condition responsive, based on IRI).
The works are listed in the opening screen for the Maintenance Standard:
The specification for the overlay works can be reviewed by double-clicking the
Note that two works items for overlaying are specified. Both are condition responsive,
one based on roughness, the other on structural cracking. This means that either
roughness OR structural cracking can trigger overlaying. The Works Items are identical
except for the Intervention Tab page. The Works Item based on roughness is reviewed
below:
! Maintenance Works: Overlay 50 mm at IRI 6
The specification for the Overlay works can be reviewed under the tabs:
The Works Activity (overlay dense-graded asphalt) is specified under General. The
overlay material, thickness and dry season strength coefficient are specified under
Design. The Construction defect indicators for the bituminous overlay (CDS) is also
specified.
Note that if overlay is implemented, the Pavement Type will be automatically reset
according to Table A2.3:
STAP + Overlay = AMAP ...(1.4)
The Tab pages specifying the Overlay Works are reviewed below:
! Alternative 3 - Maintenance Standard: Routine + SD + 50 mm overlay
The works included in this Maintenance Standard are the same as for Alternative 2 plus
an extra works item, Surface Dressing (SD).
The specification for the Surface Dressing Works Item can be reviewed by double-
clicking the appropriate works item name:
! Maintenance Works: Surface Dressing at SFC <= 0.4
The specification for Surface Dressing can be reviewed under the tabs:
The Works Activity is specified as Surface Dressing Single (General tab), with
responsive intervention based on the skid resistance (Intervention tab). Note that a
double surface dressing is also available, and shape correction may be applied to
either single or double surface dressings. Shape correction adds to the cost, but has
a more beneficial effect in terms of roughness (details are given in Chapter D2 of the
Analytical Framework and Model Descriptions).
For this case study, intervention is based on the skid resistance. Additional criteria
that may be used with surface dressing are cracking, ravelling, texture depth, and
total damaged area.
Note that the unit costs of surface dressing works and preparatory works are
specified under Costs.

The pavement type (STAP) will be reset automatically according to Table A2.3 if
either the surface dressing or overlay works is implemented.
For example:
STAP + Surface Dressing = STAP ...(1.5)
AMAP + Overlay = AMAP ...(1.8)
The Tab pages specifying the Surface Dressing Works are given below:
! Alternative 4 - Maintenance Standard: Routine + Mill 75 mm & Replace 75 mm
In addition to routine pavement works (crack sealing and patching) this Maintenance
Standard includes Mill and Replace (condition responsive based on IRI).
The specification for Mill and Replace can be reviewed under the tabs:
! Works Activity: Mill 75 mm + Replace 75 mm at 6 IRI

The Works Activity Mill and Replace is specified under General, with details
given under Design. In this alternative, the depth of milling and thickness of new
surfacing are both specified as 75 mm which represents the replacement of the
existing upper layers (25 mm surface dressing over a 50 mm overlay). Note that the
depth of milling will normally not exceed the total thickness of the existing
surfacing (Specified by road section under Section/Pavement). The software does
not check this, as Maintenance Works are not defined at section level.
The intervention criterion for Mill and Replace is based on IRI and cracking. Other
available intervention criteria are:
- Cumulative ESAL
- Mean rut depth
- Potholing
- Ravelling
- Rut depth standard deviation
- Total damaged area
Note that the unit cost of the works should include all associated preparatory works.
If the Mill and Replace works is implemented as specified, the existing Pavement
Type (STAP) will be reset according to Table A2.3.
STAP + Mill & Replace = AMAP ...(1.9)
The Tab pages specifying the Mill and Replace Works for Alternative 4 are
reviewed below:

This Maintenance Standard is similar to that specified for Alternative 4, except in this
case a 100 mm thick surfacing is added after milling out 75 mm.
As with Alternative 4, the Pavement Type after implementation of the specified Mill &
Replace works would be given by:
The Tab pages specifying the Mill and Replace Works for Alternative 5 are reviewed
below:
! Alternative 6 - Maintenance Standard: Routine + Inlay
In addition to the routine pavement works (crack sealing and patching), this Maintenance
Standard includes Inlay, that is condition responsive, based on the mean rut depth.
The specification for the Inlay works can be reviewed under the tabs:
! Maintenance Works: Inlay at Rut Depth 20 mm

The Works Activity Inlay is specified under the General tab. The percentage of the
total carriageway surface area to be inlaid is specified under the Design tab. A
value of 25% has been specified, representing the outer wheel tracks, each taken as 1
metre wide. The carriageway width is 8 m.
For this case study, mean rut depth is adopted as the intervention criterion.
Additional criteria available for use with responsive Inlay are Rut depth standard
deviation and Total damaged area.
The unit costs of the Inlay works and associated preparatory works are specified
under the Costs tab.
If inlay works are implemented the Pavement Type will remain unchanged as
STAP. Table A2.3 indicates that inlay does not affect Pavement Type.
The Tab pages specifying the Inlay Works are reviewed below:
1.2 Run HDM-4 and examine the results
1.2.1 Setup Run

The Setup Run Tab is accessed via the Analyse Projects button.
This screen confirms the base alternative for economic analysis (that is, the first alternative
entered), and that the discount rate is 6%.
For this case study, accident costs, energy balance emissions and acceleration effects are not
included in the analysis.
1.2.2 Run analysis

Starts the analysis and produces the output necessary for report generation.

1.3 Generate reports

The output from the HDM-4 analysis are produced under Generate Reports/Select Reports.
1.3.1 Deterioration/works effects

The Timing of Works report lists the maintenance works items (and associated costs) by
section or by year. The works report by section is useful when identifying the works triggered
by a particular maintenance standard. This report will often be read in conjunction with the
Deterioration Summary report that indicates the progression of certain defects associated
with condition responsive intervention criteria. For example, the progression of roughness
and Total Damaged Area (defined below) can be tracked to check that reconstruction works
are correctly triggered according to the specified intervention criteria.
The works resulting from the maintenance standard specified by section alternative are
summarised below. Observations are made regarding the works triggered, and the need, where
appropriate, to investigate the progression of defects and intervention levels associated with
the condition responsive works.
! Alternative 1: Routine and Reconstruct
2000 - 2007 Routine pavement works

2008 Reconstruction
Observations The absence of any routine pavement works between 2009 - 2014 gives
cause for concern. Intervention levels for routine works would need to
be reviewed in association with progression of defects on the
Deterioration Summary report.
! Alternative 2: Routine and 50 mm overlay
2000 Overlay at 15% cracking including preparatory patching and edge repair
2007 Overlay at 15% cracking
2015 Overlay at IRI 6
Observations The absence of routine pavement works between overlay applications
should be investigated by comparing the intervention criteria with the
progression of defects on the Deterioration Summary report.
! Alternative 3: Routine + SD + 50 mm overlay

2001 - 2007 Surface dressing each year (SFC < 0.4)


Observations The introduction of surface dressing has the effect of delaying the
development of structural cracking, resulting in overlays triggered by
roughness. (With Alternative 2, overlays in 2000 and 2007 were
triggered by structural cracking.) The intervention level for surface
dressing should be reviewed in association with the Summary
Deterioration report.
! Alternative 4: Routine + Mill 75 mm and Replace 75 mm and

Alternative 5: Routine + Mill 75 mm and Replace 75 mm
The works proposed by these alternatives are summarised below. Note the different
thickness of new surfacing (after milling) for alternatives 4 and 5.
2000 Mill and Replace

2008
2010
Routine pavement works
2011
2012
Observations Mill and Replace is condition responsive based on roughness (IRI >= 6)
and structural cracking (>= 20% carriageway area). The levels are
reached in the same year. The intervention levels for routine pavement
works should be reviewed by reference to the Deterioration report.
Alternative 5 shows less benefit than Alternative 4 (see Section 1.3.2).
! Alternative 6: Routine + Inlay

2002 Inlay at rut depth 20 mm

Observations Inlay is triggered five times during the analysis period, based on the
mean rut depth reaching 20 mm. Note that inlay is specified for 25%
carriageway area.
1.3.2 Cost streams

The economic analysis summary (by section) report gives the discounted NPV for each
section alternative when compared against the base alternative (Alternative 1).
Alternative Description Economic Works NPV

Costs discounted
(US$ millions)
1 Routine and Reconstruct 18.8 (base)
2 Routine + 50 mm overlay 18.0 78.1
3 Routine + SD + 50 mm overlay 56.0 50.6
4 Routine + Mill 75 mm & Replace 75 mm 21.2 92.5
5 Routine + Mill 75 mm & Replace 100 24.8 86.5
mm
6 Routine + Inlay 10.2 86.9
1.4 Selected reports

This section contains reports discussed in Section 1.3. The following section contains details
of parameters listed in the Summary report.
The Deterioration Summary Report shows the following condition data by section for each
analysis year:
IRIav = Average roughness m/km IRI

ACAav = Area of All Structural Cracking (expressed as % of carriageway
area)
NPTav = Number of potholes
Note:
Area of Potholes = APOT = 0.1 NPT av
and % Potholes = 100 x 0.1 NPTav / Carriageway area for section
ARVav = Area of Ravelling (expressed as % of carriageway area)
Note that the significant condition parameters available as intervention criteria for Case Study
2 are:
! Roughness (IRI)
May be used as an intervention criteria for Reconstruction, Overlay, and Mill & Replace

! Total (unpatched) damaged area (ADAMR)
May be used as an intervention criterion for Reconstruction and Surface Dressing, and
defined by:
ADAMR = AVEB + APOT + ACRA + ARV ...(1.11)
where:
AVEB = area of edge break

APOT = area of potholes
ACRA* = total area of cracking
ARV = area of ravelling
* Note :
ACRA = ACA + ACT ...(1.12)
where:
ACA = area of all structural cracking

ACT = area of transverse thermal cracking
! Area of Wide Structural Cracking (ACW)
May be used as an intervention criterion for Crack Sealing.

Note that ACW is a component of Area of All Structural Cracking (ACA)
ACA = ACW + ACN ...(1.13)
where:
ACN = area of narrow structural cracking
Note also that transverse thermal cracks (no/km) may be adopted as an intervention
criterion for crack sealing
! Severely Damaged Area (ADAMS)
May be used as an intervention criterion for patching and is defined as:
ADAMS = ACW + ARV + APOT ...(1.14)
where: ACW, ARV, APOT are defined above.

Examination of the terms included under total unpatched damaged area (ADAMR) and
Severely damaged area (ADAMS) indicates that:
ADAMR = ADAMS + AVEB + ACT + ACN ...(1.15)

Programme Analysis Case Study
1 Unit Aims
1. To introduce the HDM-4 Programme Analysis module concepts so that Trainers and Trainees
are able to prepare annual or multi-year work programmes.
2. To introduce the concept of incremental benefit/cost ranking method so that Trainers
appreciate optimisation techniques.
3. To familiarise Trainers with the use of HDM-4 data input and output modules.
2 Unit Objectives
1. To describe the concepts of the HDM-4 Programme analysis application.
2. To apply the incremental benefit cost optimisation techniques through simple manual
calculation.
3. To use the HDM-4 model to generate a work programme.
4. To alter maintenance standards and unit costs and study the impact on budget requirements.
3 Introduction
This Case Study demonstrates the application of Programme Analysis to the production of a
rehabilitation programme for a road network in the Western Province of a country. The study
demonstrates the Life Cycle Analysis method described in Chapter B2.
The objective here is to prepare a prioritised list road projects from a selected group of road
sections that have been identified to be in poor condition. In practice, this would follow some
condition survey of the road network and the application of pavement condition trigger levels or
thresholds to identify those sections in need of some form of rehabilitation.
The Life Cycle analysis method used for this Case Study is the same as that used in Project
Analysis and described in Chapter B1.
To demonstrate this Programme analysis method, the Case Study is presented in the following
steps:
Step 1 : Overview of the case study.
Step 2 : Review of the case study input data.
Step 3 : Run programme analysis with the life-cycle option
Step 4: Examine the results of the work programme
4 Overview of the case study data

The input data for this Case Study is labelled as:
! Western Province Rehabilitation (Life cycle analysis).
When a Programme Analysis study is created, the following data sets must be defined:
T3 Case 3 Programme.doc 1
PROGRAMME ANALYSIS CASE STUDY
! Road network containing the candidate road sections for periodic maintenance and/or
improvement.
! Vehicle fleet defining the characteristics of typical vehicles commonly found on the road
network.
! The average annual daily traffic (AADT) on each road section.
! Maintenance and/or improvement standards to be assigned to each candidate road section.
The above data will usually be pre-defined in HDM-4 before a Programme Analysis study is
created. The following data managers are used for this:
! Road Network manager - to create and edit the road section data
! Vehicle Fleet manager - to create and edit vehicle characteristics

! Maintenance and Improvement Standards manager
To review the case study data, open the Programmes folder and select the Western Province
Rehabilitation (Life Cycle Analysis) study.
5 Input data
The data used for this Case Study has already been created and therefore the procedure described
below does not include the creation of a new study.
The procedure involves defining details of the programme analysis to be carried out, assignment
of work standards, generation of the work programme, optimisation under budget constraints, and
finally generation of reports. This procedure should normally be followed sequentially, although
iteration between the above tasks can be done in order to review previously specified data.
The data is arranged in task-groups and, within these, the data are presented in different Tabs.
6 Define Programme Details
6.1 General
The General Tab can be used to define the study description, analysis type, analysis period (start
year and duration), the currency conversion rates and the discount rate to be used in the economic
analysis. This Tab also shows the selected Road Network and Vehicle Fleet.
This Case Study involves Life Cycle analysis carried out over a fifteen year analysis period
starting from 2000 to 2014.
6.2 Select Sections

It is assumed that the long list of candidate road sections from Western Province was drawn up
following some pavement condition survey. The list would normally contain only those road
sections deemed to require some form of periodic maintenance or rehabilitation during the next
budget period. Consequently, all road sections have been selected for the analysis.
However, if the road network had contained details of all road sections in Western Province, it is
possible to select only those that should be included in the work programme for the next budget
period.
Alternatively, all road sections in the database could be analysed, particularly if this involves a
long budget period.
It is recommended that work programmes should be prepared for short term budget periods, for
example five years or less.
Details of each road section can be edited from within this Tab (press the View/Edit Network
button).
6.3 Select Vehicles

A typical analysis using HDM-4 will involve the calculation of road user costs, primarily in the
form of vehicle operating costs, and travel time costs.
This Tab can be used to select which vehicle types will be included in the analyses, from amongst
those specified in the Vehicle Fleet database.
It is also possible to review or amend details of the vehicle types from within this Tab.
6.4 Define Normal Traffic

The AADT observed on each of the road sections will have been specified together with details of
the road sections. This Tab specifies the traffic composition and annual growth rates that are
applicable on each road section for each vehicle type in this analysis. Note that several traffic
growth periods may be specified, for example, 5% annual growth between 2000 – 2004, followed
by 3.5% annual growth from 2005 onwards. In addition, the traffic details can be copied from one
road section to others, thereby permitting each road section to have a different road set of traffic
growth characteristics, if required.
7 Specify Standards Assignments
7.1 Alternatives
The Life Cycle analysis method requires a comparison to be made between a base case (do
minimum) alternative against a project case (do something) alternative. Details of at least two
alternatives can be specified in this Tab.
For this Case Study, two alternatives have been defined for each road section; a base case
alternative and a rehabilitation alternative. The base case comprises routine pavement
maintenance only (crack sealing and patching potholes only). The rehabilitation alternative
includes routine maintenance, resealing, overlays and reconstruction for bituminous pavements.
Details of the maintenance standards are summarised in Table B1.1.

The P&CS, resealing, overlay and reconstruction standard, adopted for bituminous sections,
includes the five maintenance works items described in Table B1.1. The standard has been
assigned the short code: Brehab for “bituminous rehabilitation”.
Table B1.1
Details of Maintenance Standard (Brehab) assigned to bituminous road sections
Works Type Works Activity Intervention Criteria Ranking *

Routine Patching Potholes >= 10 per km 22
Maintenance
Crack sealing Transverse cracks >= 15 per km 22
Resealing Single Surface Dressing Total damaged area >= 30% of pavement 16
surface area
Overlays Overlay dense graded Roughness >= 5 IRI AND 9

asphalt Cracking >= 5%
Reconstruction Pavement reconstruction Roughness >= 12 IRI 6
Note: * Ranking of road works from Volume 4: Analytical Framework and Model Descriptions,
Chapter D2
If more than one of these activities are triggered in a given year, the works item that is highest in
the hierarchy of road works (that is, with lowest ranking value) will be applied.
Note that routine pavement works (for example, patching and crack sealing) have the same
ranking, and therefore both can be applied in the same year.
8 Generate Programme
8.1 Perform Run
For life-cycle analysis, the Base alternative for economic analysis is specified in the Run Setup
screen. This can be displayed by selecting the Run Setup button within the Perform Run Tab.
8.2 Work Programme
produces an unconstrained work programme. This contains, for each road section, the alternative
with the highest NPV (refer to Chapter B2). This represents an unconstrained programme, with
the total financial cost given in the cumulative cost column.
The ideal budget required for periodic maintenance in year 2000 is approximately US$ 22
million.
Table B1.2
Details of alternatives for life-cycle analysis
Alternative Surface Maintenance Standard Description Code Effective Maintenance Works Code
Class from year
Base Case Bituminous Patching and Crack Sealing Broutine 2000 Pothole Patching PATPOT
Crack Sealing CRKSL
Rehabilitation Bituminous P&CS, Reseal, Overlay & Reconstruct Brehab 2000 Pothole Patching * PATPOT
Crack Sealing * CRKSL
Reseal RSL30
Overlay OV5IRI
Reconstruct REC12
Notes:
1 Details of the intervention criteria for road work activities are given in Table B1.1.
2 The routine pavement works in the Rehabilitation alternative (identified by *) are identical to those in the Base Case.
9 Perform Budget Optimisation
9.1 Define Budget

The budget optimisation provides a facility to select road sections that can be included within
a specified budget in order to maximise the economic benefits. Given that the selected road
sections from Western Province would require US$ 22 million in year 2000 for periodic
maintenance, it is now possible to select the optimal set of road sections if less than 100% of
the required budget is available. Assume that the amount of money available for year 2000
(that is, the next budget period) has been set at US$ 15 million.
Details of the budget optimisation options are specified by pressing the Optimisation Setup
button. For this case study, optimisation is by incremental benefit/cost ranking (see HDM-4
Volume 4). The parameters to be specified are given below:
Minimum incremental value = 0.1
9.2 Optimised Work Programme

After performing the budget optimisation, a revised work programme is produced. This has a
total cost that is within the specified budget. The road sections are listed chronologically (by
year) in accordance with the incremental benefit/cost ranking index.
9.3 Generate Reports

Several reports are included within the Programme Analysis application. For this case study,
the optimised work programme report has been reproduced for this document.
ERROR! NO TEXT OF SPECIFIED STYLE IN DOCUMENT. APPENDIX B1 - PROGRAMME ANALYSIS CASE STUDY 1
HDM-4 Optimised Work Programme
Study Name: 1. Western Province (Life Cycle Analysis)

Run Date: 08-12-1999 All costs are expressed in: US Dollar (millions)
Year Section Length AADT Surface Work Description NPV/Cost Financial Cumulative
(km) Class Costs Costs
2000 MTS 549 km 0 - 1.0 1.0 2521 Bituminous Overlay 50mm 3.38 0.13 0.13
2000 MSC 112 km 35 - 80 21.0 1770 Bituminous Overlay 50mm 1.85 2.65 2.77
2000 MAN 203 km 254 - 262.8 8.8 1781 Bituminous Overlay 50mm 1.83 1.13 3.91
2000 MSC 141 km 28 - 49.5 21.5 2154 Bituminous Resealing SBSD 1.28 0.92 4.83
2000 MSC 131 km 16.5 - 26 9.5 2154 Bituminous Resealing SBSD 1.14 0.42 5.25
2000 MSW 959 km 0.1- 5, 10- 15.4 10.3 1265 Bituminous Overlay 50mm 1.10 1.26 6.76
2000 MSE 203 km 80 - 90 10.0 1191 Bituminous Overlay 50mm 0.76 1.26 8.02
2000 MAN 409 km 105 - 110 5.0 1076 Bituminous Overlay 50mm 0.60 0.63 8.65
2000 MTN 836 km 0 - 13 13.0 926 Bituminous Overlay 50mm 0.60 1.64 10.29
2000 MTN 805 km 145 - 152.6 7.6 873 Bituminous Overlay 50mm 0.41 1.01 13.94
2000 MSC 138 km 5 - 10 & 65 - 70 10.0 1315 Bituminous Resealing SBSD 0.21 0.42 14.36
This report was generated by version 1.0 of HDM-4. Page -1 of 1
Figure B1.1 HDM-4 Optimised Work Programme
It is other method saying SHIT! 9

Version This is only a beginning... wait for the worst to come after Hampehs. It will be much more worst! Visit Hampehs at http://hampehs.cjb.net for disinfection!
PROJECT ANALYSIS CASE STUDIES
Part A Project Analysis Case Studies
Project Analysis Case Study

Optimum Rehabilitation Standards for a Paved Road
1 Background
This case study presents the economic analysis of alternative rehabilitation standards for a 50
km long paved road. The existing road carries 8000 vehicles per day (AADT) and exhibits
significant levels of roughness and surface distress. This case study evaluates several possible
rehabilitation alternatives including overlay, mill and replace, and inlay.
The objective of the case study is to present the definition of section alternatives appropriate
to this type of analysis and demonstrate the HDM-4 deterioration model for paved roads. The
definition and timing of the Maintenance Road Works Standards are discussed, and the results
are examined.
To demonstrate this case study the following steps are followed:
! Locate the case study data (see Section 1.1)
! Review the case study input data (see Section 1.2)
! Run HDM-4 and examine the results (see Section 1.3)

The case study data are included in the default Case studies database installed with the HDM-
4 software. The data for this case study are located in the Projects folder in the Case Studies
Workspace. The name of the case study is Case Study 2 Rehabilitation of paved roads (by
section).
Double-click on the case study named Case Study 2a. Rehabilitation of paved roads (by
section)
1.2 Review the case study input data

The data can be reviewed under the HDM-4 Project tabs listed below:
! Define Project Details
The following Tab pages may be displayed:

! General (see Section 1.2.1)
! Select Sections (see Section 1.2.2)
! Select Vehicles (see Section 1.2.3)
! Define Normal Traffic (see Section 1.2.4)

! Specify Alternatives

! Alternatives (see Section 1.2.5)
! Analyse Projects

! Setup Run (see Section 1.3.1)
! Run Analysis (see Section 1.3.2)
1.2.1 General
The General Tab page confirms the project description, analysis type, analysis period and the
pre-defined Road Network and Vehicle Fleet.
Project.
The analysis period is defined by a start year 2000, and a duration 20 years, (that is, 2000 -
2019).

This Tab page indicates that one section will be included in the analysis. By double-clicking
on the Section Description reveals the Definition/Geometry/Pavement/Condition tabs where
the section details are held.
! Definition
! Geometry
! Pavement

T3 Project Case 1 Handout.doc 2



1995 (STAP) 25mm Surface Dressing New Surfacing

100 mm
Previous Surfacing
Subgrade (CBR 8%)
resets

Table A2.1

AC GB CRS AMGB
HRA GM
PMA AB AB AMAB
RAC SB CS AMSB
AM CM LS
PA TNA AMAP
SMA AP FDA
xx
CAPE GB CRS STGB
DBSD GM
SBSD AB AB STAB
ST SL SB CS STSB
PM LS
xx AP TNA STAP
FDA
The abbreviations in Table A2.1 are described in Table A2.2.

Table A2.2

Mix)
Asphalt
Concrete
SMA Stone Mastic
CAPE Cape Seal
Surface Dressing
Surface Dressing
SL Slurry Seal

Table A2.3

activity


Treatment

STSB1

AMSB1
to intermediate
surface layer
to base
Source: NDLI (1995)
Notes:
N/A Not applicable.
and subgrade.

Table A2.4


Widening
Inlay 11 per m2
Cape seal 14 per m2

Routine
Patching1 22 per m2
Pavement
and coefficients.
! Condition



1.2.5 Alternatives




carriageway area
carriageway area

Table A2.5

from year
Patching (PATCH)
Patching (PATCH)
Patching (PATCH)

Table A2.6

(IRI) 2
(m /km/yr)
Min Max Min Max
year
Patching R 2099 12.5 100 n/a n/a 0 100,000
Patching R 2099 12.5 100 n/a n/a 0 100,000
Patching R 2099 12.5 100 n/a n/a 0 100,000
Patching R 2099 12.5 100 n/a n/a 0 100,000
Notes:


Patching
2 50 mm overlay
Edge repair
Patching
50 mm overlay
Edge repair
3 Patching
Crack Sealing
Patching
6 Inlay Edge repair
Crack Sealing
! Works Item: Crack Sealing and Patching
! Works Item: Reconstruct at IRI 12 and 40% damage
The specification for Reconstruction is held under the tabs:

below:
! Maintenance Works: Overlay 50 mm at IRI 6
The specification for the Overlay works can be reviewed under the tabs:
specified.
! Maintenance Works: Surface Dressing at SFC <= 0.4
The specification for Surface Dressing can be reviewed under the tabs:
The Works Activity is specified as Surface Dressing Single (General tab), with
responsive intervention based on the skid resistance (Intervention tab). Note that a
total damaged area.

For example:
The specification for Mill and Replace can be reviewed under the tabs:
! Works Activity: Mill 75 mm + Replace 75 mm at 6 IRI

- Cumulative ESAL
- Mean rut depth
- Potholing
- Ravelling
reviewed below:

below:
The specification for the Inlay works can be reviewed under the tabs:
! Maintenance Works: Inlay at Rut Depth 20 mm

The Works Activity Inlay is specified under the General tab. The percentage of the
total carriageway surface area to be inlaid is specified under the Design tab. A
under the Costs tab.
1.3.1 Setup Run

1.3.2 Run analysis





2008 Reconstruction





2008
2010
2011
2012


carriageway area.
1.4.2 Cost streams


Costs discounted
(US$ millions)
mm

analysis year:

area)
Note:
2 are:
! Roughness (IRI)

defined by:
where:

* Note :
ACRA = ACA + ACT ...(1.12)
where:


ACA = ACW + ACN ...(1.13)
where:


Programme Analysis Case Study
1 Introduction
condition survey of the road network and the application of pavement condition trigger levels or
thresholds to identify those sections in need of some form of rehabilitation.
steps:
2 Overview of the case study data

improvement.
network.
! Maintenance and/or improvement standards to be assigned to each candidate road section.

Rehabilitation (Life Cycle Analysis) study.
T3 Programme Case 1 Handout.doc 1

3 Input data
The data used for this Case Study has already been created and therefore the procedure described
below does not include the creation of a new study.
The procedure involves defining details of the programme analysis to be carried out, assignment
of work standards, generation of the work programme, optimisation under budget constraints, and
finally generation of reports. This procedure should normally be followed sequentially, although
iteration between the above tasks can be done in order to review previously specified data.
The data is arranged in task-groups and, within these, the data are presented in different Tabs.
4 Define Programme Details
4.1 General
The General Tab can be used to define the study description, analysis type, analysis period (start
year and duration), the currency conversion rates and the discount rate to be used in the economic
analysis. This Tab also shows the selected Road Network and Vehicle Fleet.
4.2 Select Sections

It is assumed that the long list of candidate road sections from Western Province was drawn up
sections deemed to require some form of periodic maintenance or rehabilitation during the next
budget period. Consequently, all road sections have been selected for the analysis.
However, if the road network had contained details of all road sections in Western Province, it is
possible to select only those that should be included in the work programme for the next budget
period.
long budget period.
It is recommended that work programmes should be prepared for short term budget periods, for
example five years or less.
button).
4.3 Select Vehicles

A typical analysis using HDM-4 will involve the calculation of road user costs, primarily in the
form of vehicle operating costs, and travel time costs.
The Vehicles Tab can be used to select which vehicle types will be included in the analyses, from
amongst those specified in the Vehicle Fleet database.

The AADT observed on each of the road sections will have been specified together with details of
the road sections. This Tab specifies the traffic composition and annual growth rates that are
applicable on each road section for each vehicle type in this analysis. Note that several traffic
growth periods may be specified, for example, 5% annual growth between 2000 – 2004, followed
by 3.5% annual growth from 2005 onwards. In addition, the traffic details can be copied from one

road section to others, thereby permitting each road section to have a different road set of traffic
growth characteristics, if required.
5 Specify Standards Assignments
5.1 Alternatives
minimum) alternative against a project case (do something) alternative. Details of at least two
alternatives can be specified in this Tab.
includes routine maintenance, resealing, overlays and reconstruction for bituminous pavements.
Details of the maintenance standards are summarised in Table B1.1.

The P&CS, resealing, overlay and reconstruction standard, adopted for bituminous sections,
includes the five maintenance works items described in Table B1.1. The standard has been
assigned the short code: Brehab for “bituminous rehabilitation”.
Table B1.1
Details of Maintenance Standard (Brehab) assigned to bituminous road sections

Maintenance
surface area

Note: * Ranking of road works from Volume 4: Analytical Framework and Model Descriptions,
Chapter D2
If more than one of these activities are triggered in a given year, the works item that is highest in
the hierarchy of road works (that is, with lowest ranking value) will be applied.
ranking, and therefore both can be applied in the same year.
6 Generate Programme
6.1 Perform Run
For life-cycle analysis, the Base alternative for economic analysis is specified in the Run Setup
screen. This can be displayed by selecting the Run Setup button within the Perform Run Tab.

6.2 Work Programme
produces an unconstrained work programme. This contains, for each road section, the alternative
with the highest NPV (refer to Chapter B2). This represents an unconstrained programme, with
the total financial cost given in the cumulative cost column.
million.

Table B1.2
Class from year
Crack Sealing CRKSL
Reseal RSL30
Overlay OV5IRI
Reconstruct REC12
Notes:

7 Perform Budget Optimisation
7.1 Define Budget

button. For this case study, optimisation is by incremental benefit/cost ranking (see HDM-4
Volume 4). The parameters to be specified are given below:
7.2 Optimised Work Programme

year) in accordance with the incremental benefit/cost ranking index.
7.3 Generate Reports

the optimised work programme report has been reproduced for this document.

ERROR! NO TEXT OF SPECIFIED STYLE IN DOCUMENT. APPENDIX B1 - PROGRAMME ANALYSIS CASE STUDY 1


Version This is only a beginning... wait for the worst to come after Hampehs. It will be much more worst! Visit Hampehs at http://hampehs.cjb.net for disinfection!
PROJECT CASE STUDY 2
Derivation of Optimum
Maintenance Standards
Aims & Objectives
Aim:
z To understand
T d t d theth specification
ifi ti and
d
assignment of HDM-4 maintenance standards
Objectives:
z prepare input data required for investigating
optimum road maintenance standards
z assign a range of maintenance standards to
road sections
z interpret the results and recommend optimum
maintenance standards
Overview
2 Road Sections
z ST, 36
ST 36.5
5 km
km, AADT = 6 000
000.
z AC, 10.6 km, AADT = 16 000.
Maintenance Standards
z Base Case: Patching potholes, Crack sealing,
Drainage maint + Other cyclic works.
z P i di maint
Periodic i t 1 for
f ST road
d section:
ti Reseal
R l every 5
years (scheduled)
z Periodic maint 1 for AC road section: Overlay 50mm
at 6 IRI (responsive)
Procedure
Calibrate the 2 road sections (RD)

Create additional maintenance standards
z For ST pavement - add new standards with
Reseal intervals: 3, 7, 9 & 11 years
z For AC pavement - add new standards with
Overlay 50mm at: 4, 8, 10 & 10 IRI
Create new Al
C Alternatives
i and d assign
i
maintenance standards
R HDM
Run HDM-4
4 & review
i results
lt
Plot summary of the results:
z NPV vs P
Periodic
i di maint
i t iinterval
t l (Years
(Y or IRI)
Optimum Standard
NPV
Maintenance Interval (IRI or Years)

NPV / Agency Cost
NPV
Agency Cost
PROJECT ANALYSIS -
WORKED EXAMPLES
Overview of HDM-4 Cases Studies
z Case studies database + documentation

z 4 Project Analysis
z 2 Programme Analysis
z 1 Strategy Analysis
z Intention is to cover range of features
Project Analysis - Case Studies
z Upgrading a Gravel Road to a Bituminous

Pavement
z Optimum Rehabilitation Standards for a
Paved Road
z Traffic Volume Capacity Improvements
z New Bypass Construction (traffic
diversion effects)
Traffic Volume Capacity Improvements
z Objective - to demonstrate congestion

analysis in HDM-4
z Economic analysis of widening road
z Initially: 7m, AADT 15000 + NMT
z Condition responsive routine maintenance
z 3 schemes:
z Widening by 1m
z Widening by 3m
z Addition of 2 extra lanes
z Volume 2 - Appendix A3
Worked Example Objectives
z Project Analysis concepts

z Setting up road sections
z Defining vehicle fleets
z Defining works standards
z Defining projects alternatives
z Interpreting
p g Reports
p
Review of Objectives
z Project Analysis concepts

z Setting up road sections
z Defining vehicle fleets
z Defining works standards
z Defining projects alternatives
z Interpreting
p g Reports
p
ENERGY BALANCE ANALYSIS
UoB
OBJECTIVES
z Quantify energy used in road transport

z Compare energy impact of transport
policies
z Comparisons avoid price distortions
z Quantify total life-cycle energy use
z Energy efficiency of transport modes
z Quantifyy energy
gy from different sources
APPLICATIONS
z Network Level
z Total energy use by transport mode
z Long term energy requirements
z Renewable energy sources
z Project Level
z Compare energy use for investment
options
z Specific energy use by transport mode
ENERGY COMPONENTS
z Motorised Transport
z Vehicle Operation
z Fuel Production
z Vehicle Manufacture
z Non-Motorised Transport
z Animal drawn carts,
carts cycling,
cycling head-loading
z Road Construction and Maintenance
z Road works
MOTORISED TRANSPORT
z Energy Use Stages

z Vehicle manufacture
z Fuel production
z Vehicle use
z Vehicle maintenance
z Mainly non-renewable (fossil) fuels
VEHICLE OPERATION ENERGY
z Vehicle manufacture
z raw materials,
t i l processing,
i components,t
transportation, assembly, distribution
z Fuel production
z raw materials, stock transport, processing,
distribution
z Vehicle use
z fuel oil & tyre consumption
fuel,
z Vehicle maintenance
z component manufacture,
manufacture distribution
VEHICLE MANUFACTURE
z Motorcycle 20 GJ
z Medium car 100
z Light
g goods
g vehicle 160
z Four wheel drive 180
z Medium truck 600
z Articulated truck 1,500
z Mini-bus
Mini bus 300
z Large bus 1,000
FUEL PRODUCTION CYCLE
Feed stock Processing Distribution End

End--use
Crude Oil Distribution Distribution
and natural pipeline terminal
gas liquids
Petrol, Depot
p
Oil Di l
Diesel
Pipeline storage
refinery road tankers
Gas Petrol,
Separation Crude Oil Service Diesel
plant Tanker stations vehicl
es
FUEL PRODUCTION ENERGY
z Petrol 0.169 MJ/MJ

z Diesel 0.122
z LPG 0.122
z CNG 0.061
z Electric 2.857
z Bio-methanol 0.514
z Bio-ethanol
Bio ethanol 0.510
z Bio-diesel 0.655
FUEL ENERGY CONTENT
z Petrol 34.7 (MJ/Li)

z Diesel 38.7
z LPG 25.5
z CNG 40.0
z Ethanol 23.9
z Methanol 18.1
z Bio-diesel
Bio diesel 32.8
ENERGY CONTENT FACTORS
z Lubricating oil 47.7 (MJ/Li)

z Tyre wear 32 (GJ/Tonne)
z Vehicle repair
p & support
pp 2 ~ 4 % of life
cycle energy
z Replacement parts proportion of
total life cycle
energy
Energy Content Factors
z Lubricating Oil z 47.7 MJ/Li

z Tyre Wear z 32 GJ/Tonne
z Vehicle Repair & z 2~4% of life cycle
support energy
z Replacement parts z proportion of total life-
cycle
y energy
gy
NON-MOTORISED TRANSPORT
z Animal drawn Carts

z animal mass, cart & load mass, speed
z rolling resistance, slope
z Cycling
z total mass
z speed
z Head-loading
z total mass
ROAD CONSTRUCTION AND MAINTENANCE
z Energy consumed per unit of road works

z Excavation (per cu.m)
z Formation level
z Base layers
z Surfacing
z Resurfacing
z Rehabilitation
ENERGY BALANCE RESULTS
z Total energy used

z by transport mode
z Total by renewable/non-renewable
z by fuel type
z Total national/global
z Specific energy consumption
z per km by mode
z per passenger/km
z per tonne/km
ENERGY USE ON GRAVEL ROAD
Poor condition (IRI=15 m/km)

Vehicle Type Av. Speed Energy
(km/h) (MJ/km)
Cars & Utilities 35.6 4.4
Buses 34.6 8.2
Light Truck 35.0 5.4
Medium Truck 34 9
34.9 93
9.3
Heavy Truck 31.6 19.4
GLOBAL AND NATIONAL ENERGY
(Gravel road, IRI=15 m/km)

Vehicle Type Global National
(MJ/km) (MJ/km)
Cars & Utilities 6.0 4.5
Buses 11.8 8.5
Light Truck 8.5 5.7
M di
Medium T
Truck
k 12 8
12.8 96
9.6
ENERGY USE ON PAVED ROAD
(Good condition, IRI = 3 m/km)
Vehicle Type Av. Speed Energy

(km/h) (MJ/km)
Cars & Utilities 108 3.9
B
Buses 87 1
87.1 73
7.3
Light Truck 93.1 4.8
Medium Truck 88.9 7.1
GLOBAL AND NATIONAL ENERGY
(Paved road good condition, IRI = 3 m/km)
Vehicle Type Global National

Cars & Utilities 52
5.2 40
4.0
Buses 9.7 7.5
Light Truck 7.8 5.2
M di
Medium T
Truck
k 94
9.4 74
7.4
NET ENERGY CONSUMPTION
(Comparison of Gravel - Paved Road)

Vehicle Type National Global
Cars & Utilities 0.7 0.8 0.5
B
Buses 09
0.9 21
2.1 10
1.0
Light Trucks 0.6 0.7 0.5
M di
Medium Trucks
T k 22
2.2 34
3.4 22
2.2
Heavy Trucks 3.8 5.1 4.0
(MJ/km)
VEHICLE EMISSIONS
UoB
OBJECTIVES
z To assess the effects, in terms of pollutant

quantities,
titi off changes
h in:
i
z Road characteristics
z Traffic congestion
z V hi l technology
Vehicle t h l
APPLICATIONS
z Network level
z Total emissions by transport mode
z Regional long term impact of pollution
z Project level
z Compare levels of emissions for different
investment options
z Pollution impact on local areas near the
project road
TYPES OF POLLUTANTS
z Hydrocarbon (HC)
z Carbon monoxide (CO)
z Nitrous oxides (NOx)
z Sulphur dioxide (SO2)
z Carbon dioxide (CO2)
z P ti l t (Par)
Particulates (P )
z Lead (Pb)
THE MODELLING
z The models predict vehicle exhaust emissions

as a function of:
z Vehicle speed
z Fuel consumption
z Vehicle service life
z Speeds and fuel consumption depend on both
road and vehicle characteristics
z Thus, enabling the analysis of changes in
emissions arising from different road
investment options, or major changes to the
vehicle fleet using the road network
MODEL PARAMETERS
z Assumes emissions at 0.5* LIFE are suitable

for representing the emissions
z The model parameter a2 represents the
d
degradation
d ti off a vehicle
hi l equipped
i d with
ith
catalytic converter
z F SO2,
For SO2 a0 0 is
i the
th percentage
t content
t t byb
weight, of sulphur in fuels (Petrol: 0.012%)
z For Pb,
Pb a0 is the percentage content by weight
weight,
of lead in fuels (Petrol: 0.15%)
COMPARISON OF INVESTMENT ALTERNATIVES
z The comparison of each pair of

investment alternatives is based on the
changes in the annual difference in the
predicted
di t d quantities
titi off emissions
i i (by
(b
component)
OUTPUTS
z Quantities of emissions
z by component
z by vehicle type
z for each traffic flow period
z Annual quantities of emissions
z by component
z for each investment option
z Annual net quantities of emissions for
each pair of investment alternatives
CONCRETE PAVEMENTS
Pavement Classification
z J i
Jointed
d Plain
Pl i without
ih dowels
d l
z J i t d Plain
Jointed Pl i with
ith dowels
d l
z Jointed Reinforced
z Continuously Reinforced
Jointed Plain without dowels
Joint Spacing
3 - 6 m
Aggregate
Slab Interlock
Base
Jointed Plain with dowels
Joint Spacing
3 - 6 m
Dowels
Jointed Reinforced
10 - 20 m
Slab
Dowels
Base
Welded Wire Fabric

(0,1 - 0,2 %)
Continuously Reinforced
Cracks separation
Slab
Base
Reinforcement Steel
0,6 - 0,8 % Area
Distress Modes Modelled in HDM-4
z Cracking % of slabs cracked JP

number per km JR
z Faulting
g inches JP,, JR
z Spalling % of spalled joints JP, JR
z Failures number per mile CR
z S
Serviceability
i bilit di
dimensionless
i l JR CR
JR,
z Roughness
g inches p
per mile All
Primary Variables
z Structural characteristics
z Traffic
z E i
Environment
t
z Road geometry
z Pavement condition
z Age
z Previous maintenance
Structural Characteristics
z Material type and properties

z Subgrade stiffness
z Slab thickness
z Quantity of reinforcement steel
z Load transfer efficiency (across joints,
and between slabs and shoulder)
z Drainage conditions
z Carriageway width
Material Properties
z Modulus of elasticity of concrete Ec

z M d l off rupture
Modulus t (flexural
(fl l strength)
t th)
of concrete MR28
z Th
Thermal l coefficient
ffi i t off concrete t α
z Drying shrinkage coefficient of
concrete
t γ
z Poisson’s ratio for concrete µ
z M d l off elasticity
Modulus l ti it off dowels
d l Es
z Modulus of elasticity of bases Ebase
z M d l off subgrade
Modulus b d reaction
ti KSTAT
Cracking
z Transverse cracking occurs as a result

of material fatigue due to high stress
levels or defects in the slabs
z The stresses are caused by the

combined effect of thermal curling,
moisture-induced curling and traffic
loading
Transverse Cracking
Distress Distress
width width
A B C D
Longitudinal Joint
C
L
C D
Transv. Transv.
Joint Joint
A B Traffic
Sl b
Slab
Shoulder
Cracking - JP Pavements
Transverse Cracking (% of slabs cracked) is

modelled as a function of cumulative fatigue
damage in the slabs
Key variables are:-
are:
z Cumulative ESAL’s
z Temperature gradient
z Slab thickness
z Material properties
z Joint spacing
z Shoulder
Sh ld supportt
Cracking - JR Pavements
Transverse Cracking (% of slabs cracked) is

modelled as a function of cumulative fatigue
damage in the slabs
Key variables are:-
are:
z Cumulative ESAL’s
z Temperature gradient
z Slab thickness
z Material properties
z Joint spacing
z Shoulder
Sh ld supportt
Faulting
Faultingg represents
p the
elevation difference between
g of a transverse
the edges
joint or crack
Faulting
A
faulting B
Longitudinal Joint
C
L
Transv.
Transv. Joint
Joint
A B
Traffic
Slab
Faulting
Faulting is caused by the loss of fine material

under a slab and the increase in fine material
under nearby slabs
This flow of fine material is called pumping, and

is caused by the presence of high levels of free
moisture under a slab carrying heavy traffic
loading
The effects of thermal and moisture-induced

curling
g and lack of load transfer between slabs
increase pumping
Faulting - JPCP
Faulting is a function of:
• Cumulative ESALs
• Slab thickness
• Joint spacing
• Properties of material
• L d transfer
Load t f efficiency
ffi i
• Climate/environment
• Drainage factor
• Base type
• Width of lane
Spalling
Transverse jjoint spalling

p g is the
cracking or breaking of the edge of
p to a maximum of 600 mm
the slab up
from the joint
Spalling
Distress < 0,6 m

width
A B C D
Crack Joint
Joint
Transv.
T Transv.
T Transv.
Transv
Joint Joint Joint
Low Sev.:
18m
1,8 Low Sev
Sev.:: C D
2m Moder. Sev.:
2,5 m
High Sev.:
15m
1,5 Traffic
A B
Shoulder
Spalling
Spalling can be caused by:
• Presence of incompressible materials

• Disintegration of concrete under high
traffic loading
g
• Improper consolidation of the concrete
in the joint
• Poorly designed or built load-transfer
system
Spalling
Transverse joint spalling is a function of:
z Pavement age
z Joint spacing
z Type of seal
z Dowel corrosion protection
z B
Base ttype
z Climate/environment
Failures in CR Pavements
Localised failures include loosening

g and
breaking of reinforcement steel and
transverse crack spalling
These are caused by high tensile stresses

induced in the concrete and reinforcement
steel by traffic loading and environmental
factors
Failures in CR Pavements
The number
n mber of failures
fail res is a ffunction
nction of
of:
• Slab thickness
• Percentage of reinforcement steel
• Cumulative ESALs
• Base type
Present Serviceability Rating
This is a subjective user rating of the existing

ride
id quality
lit off a pavementt ranging
i from
f 0
(extremely poor) to 5 (extremely good)
For JR pavements, the change in PSR is a

function of cracking,
cracking spalling and faulting
For CR pavements,
F t th
the change
h iin PSR is
i a
function of slab thickness, cumulative ESALs
and pavement age
Roughness
For JP concrete pavements, roughness is

calculated as a function of faulting, spalling
and cracking
For JR and CR concrete pavements,

roughness is calculated as a function of PSR
BITUMINOUS PAVEMENTS
Road Deterioration
AIMS AND OBJECTIVES
z SESSION AIMS
z To introduce the key relationships.
z To make participants aware of the key reference documents for
more detailed explanation.
z SESSION OBJECTIVES
z To outline the principles behind modelling of bituminous pavements.
z T identify
To id tif the
th various
i factors
f t which
hi h affect
ff t pavementt performance.
f
z To explain the basis of the various pavement defects and condition
indicators which are modelled in HDM-4, and to highlight the most
i
important
t t explanatory
l t variables.
i bl
z PRACTICALS
z REFERENCES
Overview of the presentation
z Principles of the deterioration models

z Distress types
z Pavement classification system
z Factors affecting pavement performance
z Environmental factors
z Construction quality
z Pavement strength
z Drainage effects
z Pavement deterioration
z I t
Interactions
ti
PRINCIPLES OF DETERIORATION MODELS
z Models are structured

str ct red empirical
z Individual distresses modelled separately
z R l ti
Relationships
hi are incremental
i t l and
d recursive
i
dY = K a0 f(X1, X2, X3, etc)
z Modelled sequentially through to roughness
z Maintenance intervention at end of each year
PAVEMENT DETERIORATION
z Various defects modelled in the following

order
z Cracking
z Ravelling Damaged Area
z Potholing
z Edge break
z Rutting
z Roughness
z Texture depth
z Skid resistance
i t
PAVEMENT CLASSIFICATION SYSTEM

T
Type M
Material
i l T
Type M
Material
i l T
Type
AC GB CRS AMGB
HRA GM
PMA AB AB AMAB
AM RAC SB CS AMSB
CM LS
PA AP TNA AMAP
SMA FDA
CAPE GB CRS STGB
DBSD GM
SBSD AB AB STAB
ST SL SB CS STSB
PM LS
AP TNA STAP
FDA
FACTORS AFFECTING PAVEMENT
PERFORMANCE
z TRAFFIC - no. of esa’s

z CLIMATE - moisture-temperature, rainfall
z TIME (AGE) - since construction or
maintenance
z STRENGTH - Structural no, DEF, etc
z DRAINAGE - condition/quality
z CONSTRUCTION - type and quality
z MAINTENANCE - type and quality
ENVIRONMENTAL FACTORS
z Moisture environment
z Rainfall
z Drainage from the pavement structure
z Temperature environment
RAINFALL / MOISTURE CLASSIFICATION
Moisture Description Thornthwaite Annual

Classification Moisture Precipitation
Index (mm)
Arid Very
y low rainfall,, -100 to -61 < 300
High evaporation
Semi-arid Low rainfall -60 to -21 300 to 800
Sub-humid Moderate rainfall

rainfall, or -20 to +19 800 to 1600
strongly seasonal rainfall
Humid Moderate warm seasonal +20 to +100 1500 to 3000
rainfall
Per-humid High rainfall, or very > 100 > 2400
many wet-surface days
TEMPERATURE
Temperature Description Temperature range

(C)
Tropical Warm temperatures in small range 20 to 35
Sub-tropical - High day cool night temperatures, - 5 to 45
hot hot-cold seasons
Sub-tropical - Moderate day temperatures
temperatures, cool -10 to 30
cool winters
Temperate - Warm summer, - 20 to 25
cool shallow winter freeze
Temperate - Cool summer, - 40 to 20
freeze deep winter freeze
ENVIRONMENTAL COEFFICIENT
Temperature Classification
Moisture Sub- Sub- Temperate Temperate

Classification Tropical tropical tropical cool freeze
hot cool
Arid 0.005 0.010 0.015 0.025 0.040
Semi-arid 0.010 0.015 0.025 0.035 0.060
Sub humid
Sub-humid 0 020
0.020 0 025
0.025 0 040
0.040 0 060
0.060 0 100
0.100
Humid 0.025 0.030 0.060 0.100 0.200
Per-humid 0.030 0.040 0.070
CONSTRUCTION QUALITY
Poor construction quality or variability in material

properties have a significant effect on the
performance of pavements
3 parameters are used to model these effects
z COMP - relative compaction

p of the pavement
p
layers - affects the rate of deformation (rutting)
z CDS - surface aspects
p - mainly
y driven by
y binder
content - affects cracking, plastic deformation,
ravelling, etc
z CDB - roadbase aspects - affects potholing
PAVEMENT STRENGTH
z Affected by:
z Pavement materials and thicknesses
z Subgrade
z Drainage regime and seasonality
z Adjusted Structural Number (SNP) is used
z Similar to modified structural number of
HDM-III / AASHO
z BUT contribution of pavement layers /
subgrade weighted by depth
z Relationship
p between SNP and DEF
EFFECT OF DRAINAGE
z Depends on seasonal variations

(wet and dry seasons)
z Depends on type and condition of
drains
z Depends on ingress of water into
pavement (cracks and potholes)
VARIATION OF SNP DUE TO DRAINAGE
1.00
DF = 1
0.90
DF = 2
DF = 3
0.80
SNPw / SNPd
0.70
DF = 4
0.60
DF = 5
0.50
MMP = 200 mm/month
0.40
0 0
0 10 20 30 40 50 60 70 80 90 100
ACRA (%)
USE OF DEFLECTION RESULTS
3.5
3.0
n Beam Deflection (mm)
2.5
Granular base
2.0
1.5 Cemented base

Benkelman
1.0
0.5
0.0
1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Adjusted Structural Number
INTERACTIONS BETWEEN PARAMETERS
Environment
Traffic
Loading
Pavement
Strength Pavement
(annual average) Deformation Roughness
Strength
(crack adjusted)
Structural
Cracking
Rutting
Potholing
ROUGHNESS
z IRI = f (structural deformation, surface

condition, age-environment)
z Structural deformation = f(SNPeff, YE4)
z Surface condition = f(ACRA, RDS, NPT)
z Age-environment = f(m, IRI)

ROUGHNESS (2)
6
SNPd = 3
HS = 50 mm
YE4 = 0.1
5 Environmental component
m = 0.023
hness (IRI, m//km)
Rutting
g component
p
4
Cracking component
Rough
Structural component
2
0 5 10 15 20
Pavement Age (years)
ROUGHNESS (3)
8
SNPd = 6
7 HS = 100 mm Environmental component
YE4 = 1.0
m = 0.023
m/km)
6
Rutting component
ghness (IRI, m
5
Cracking component
Roug
4 Structural component
2
0 5 10 15 20
Pavement Age (years)
DISTRESS INITIATION & PROGRESSION
100
90
80
70
ent)
d Area (perce
Initiation Progression
60
Period Period
50
Affected
40
30
20
10
0
Age of Surfacing
CRACKING
z 2 forms of cracking are considered:

z Structural cracking
z All cracking
z Wide cracking
z Transverse thermal cracking
z For each form of cracking, 3 phases:

z Time to initiation
z Subsequent progression
z E ilib i
Equilibrium condition
diti
STRUCTURAL CRACKING INITIATION
14
CDS = 1.25
12
CDS = 1.00
eriod (years)
CDS = 0.75
10
All Crack Initiation Pe
6 SNP = 5
4
SNP = 2
0
0.0 0.2 0.4 0.6 0.8 1.0
Axle Loading (MESAL/year)
STRUCTURAL CRACKING PROGRESSION
100
SNPd = 6 SNPd = 3
80 YE4 = 1.0 YE4 = 0.1
uctural Crackking (%)
60
Area of Stru
40
20
All Cracking
Wide Cracking
0
0 2 4 6 8 10 12 14 16
Time Since Crack Initiation (years)
TRANSVERSE THERMAL CRACKING
120
m)
ermal Cracking (No/km
100
80 sub-tropical hot
arid
id
60
sverse The
40 temperate freeze
humid
Trans
20
0
0 5 10 15 20
Surface Age (Years)
RAVELLING INITIATION
18
16 CDS = 1.25
14
d (years)
12
Ravelling Iniitiation Period
CDS = 1.0
10
8
CDS = 0.75
6
0
0 1000 2000 3000 4000 5000
AADT
RAVELLING PROGRESSION
100
90 CDS = 0.75 CDS = 1.0 CDS = 1.25
80
70
a of Ravelling (%)
60
50
40
Area
30
20
10
0
0 2 4 6 8 10 12 14 16
Time Since Ravelling Initiation (years)
POTHOLING INITIATION
9
Granular base
8
MMP = 100 mm/month
7
Time to Pothole Initiation (years)
HS = 150 mm
6
HS = 100 mm
5
HS = 50 mm
4
HS = 20 mm
3
0
0 1 2 3 4 5 6 7 8 9 10
Traffic (million axles/lane/year)
POTHOLING PROGRESSION
1000
HS = 50 mm
HS = 25 mm
800
HS = 100 mm
year)
oling (no/km/y
600
400
Potho
Granular base
200 ACW = 40%
MMP = 100 mm/month
TLF = 1
0
0 1 2 3 4 5 6 7 8 9 10
Traffic (million axles/lane/year)
EDGE BREAK
90
Edge Step = 50 mm
80
Average Speed = 50 km/h
R i f ll = 200 mm/month
Rainfall / th
70
Width = 3.5 m
m/year)
60
Break (m3/km
50
Width = 4.5 m
40
Edge B
30 Width = 5.5 m
20
10
0
0 500 1000 1500 2000
AADT
RUTTING - DENSIFICATION
Densification distributed Densification within

through pavem ent single pavem ent
structure layer
Surface layer
y
Pavem ent Layer 1
P avem ent Layer 2
Subgrade
RUTTING- PLASTIC FLOW
Depression within
Heaving alongside
wheelpath
wheelpath
Asphalt layer
Pavement layer 1
Pavement layer 2
S b d
Subgrade
Mix deformation
STRUCTURAL DEFORMATION
Structural Deformation
With Cracking
Rut Depth
R
Structural Deformation
No Cracking
Initial Densification
First Year
Age
PLASTIC DEFORMATION
120
YE4 = 1.0
Sh = 60 km/h
m)
ormation (mm
100 HS = 100 mm
CDS = 1.25
80
an Rut Depth - Plastic Defo
60
CDS = 1.0
40
CDS = 0.75
Mea
20
0
0 5 10 15 20
Surfacing Age (years)
SURFACE WEAR
40
35 Salted
Unsalted
30
Speed = 80 km/h
25
m)
Rut Depth (mm
20
Speed = 40 km/h
15
10
0
0 2 4 6 8 10
Vehicle Passes with Studded Tyres (million)
PAVEMENT TEXTURE
SCALE OF TEXTURE
SURFACE MACRO MICRO
COARSE HARSH
A TD~ 2mm SFC ~ 0.6
50
COARSE POLISHED
B TD~ 2mm SFC50~ 0.4
FINE HARSH
C TD~ 0.35mm SFC ~ 0.6

50
FINE POLISHED
D TD~ 0.35mm SFC ~ 0.4

50
CONCLUSIONS
z Models are available which take greater

account of local factors, including
construction quality, drainage and
environment.
z Models are based on ‘average expected’ - you
must calibrate to your environment.
z Some models will not be applicable to your
environment.
z Particular models are sensitive to different
parameters, therefore priority should be given
to identifying and providing these.
these
ROAD MANAGEMENT
A Framework for Sustainable

Road Management
Key Statistics from Emerging Countries
z Transport sector is 5 to 10 per cent GDP
z Paved roads account for 12% of length and 90% non-
urban
rban traffic
z Construction & Maintenance is 1 to 2% GDP
z In some countries
countries, actual expenditure 25% of need
z Every $ not spent on maintenance increases transport
costs by 4$
z Surfacing lives vary from less than 5 to 20 years
z Road crashes account for between 1 - 3 % of GDP
z Transport fuels and machinery account for 40% of
import bill
z Differences in transport tariffs of up to 4 to 5 times
z Rural transport services can vary by more than double
Mix deformation
Bus awaiting repair
Accessibility problem on a feeder road
Culvert design and maintenance problem
Weaknesses in Road Management
z Inadequate and unsecure funding

z Greater priority given to new roads
z Poorly motivated organisations
z Weak or non existent management
systems
z Inappropriate treatments and lack of
quality control
z Poor regulations and/or enforcement
z Equipment failures and lack of spares
What SOLUTIONS can we suggest ?
z Establish a dedicated and secure Road Fund

z Prioritise on a rational and equitable basis
z Adopt affordable standards
z Address deep rooted institutional problems
by…….
z Implement appropriate management and
quality systems
z Tackle accountability and transparency issues
z Address transport management, regulation
and enforcement issues
z S k tto iimprove road
Seek d user b
behaviour
h i and
d
vehicle standards
Country Profile and Road Performance
- What is your country’s profile
COUNTRY PROFILE TREATMENT TREATMENT ANNUAL

TYPE FREQUENCY ROUGHNESS
INCREASE
A Poor construction quality Thin < 5 years > 20 %
High tech management system overlays
Inappropriate treatments
Poor axle load control
Bad mix specifications
B Average construction quality Surface 8 years 2%
p management
Simple g system
y dressings
g
Typical maintenance response
Some overloading
Typical specifications & designs
C High construction quality Surface 12 years Negligible
Simple management system
s stem dressings
Very effective maintenance
response
Good axle load control
Good specs & designs
WHAT IS ROAD MANAGEMENT?
z Purpose: To optimise the overall

performance of the network over time
p
z Road management involves:
z Setting
g priorities
p
z Defining activities
z Planning
z Allocating resources
z Organising and motivating personnel
z Controlling work
z Monitoring and evaluating performance
z Feeding back results to seek improvements
ACTIVITIES ON THE ROAD NETWORK
z Routine works
z Cyclic and reactive works types
z Periodic works
z Preventive, resurfacing, overlay and
reconstruction works types
z Special works
z Emergency and winter maintenance works
types
z Development works
z Widening,
Wid i realignment
li t and
d new
construction works types
THE IMPACTS OF ROAD MANAGEMENT
The effects of road management can be

assessed in terms of the various impacts:
z Level of service (road condition)

z Socio economic impacts
z Road user costs
z Accident levels and costs
z Environmental degradation
z Road administration costs
THE MANAGEMENT FUNCTIONS
z Planning
z Setting standards and policies
z Long term estimates of expenditure
z Programming
z Medium term work programmes
z Preparation
z Detailed project design and work
packaging
z Operations
z Implementation of works in field
DIFFERENCES BETWEEN THE FUNCTIONS
Planning Operations
z Sections/network considerations
z Time horizon
z Staff involved
z Data details
z Computer processing
z U off private
Use i t sector
t
THE MANAGEMENT CYCLE
Aims
A dit
Audit Needs
DATA
Implem. Action
s
Cost/Pr
DIFFERENT COMPUTER SYSTEMS HAVE
DIFFERENT AIMS
Input Data Output

bank
Input Data Output

p
bank
Processing
IMPROVEMENTS NEED TO BE SUSTAINABLE
z User attitudes z Staffing
z C lt
Cultural
l issues
i z Training
z Economics and z Data and

finance computers
Sustainability
IMPROVEMENT PRIORITIES
Technical
Institutional
External
HOW DOES HDM-4 CONTRIBUTE?
z HDM-4 is a decision support system, to

assist in determining impacts of potential
road investments
Management Function HDM-4 Application

Planning Strategy Analysis
Programming Programme Analysis
Preparation Project Analysis

STRATEGY ANALYSIS - Effect of budget levels
Annual
7.0 Budget
50%
6.0
hness
80%
50
5.0
Rough
100%
4.0
Target
3.0
1998 1999 2000 2001 2002 2003 2004
PROGRAMME ANALYSIS - Multi-year works
Rank Road Section km District Work Type Year $m Σ$m

1 N1-2 20.5 2 Resealing 2000 5.4 5.4
2 N4-7 23.5 7 Overlay 40mm 2000 10.9 16.3
3 N2-5 12.5 5 Reconstruct 2000 8.6 24.9
4 R312-1
R312 1 30.0 4 Widen 4 lane 2000 31.4 56.3
5 R458-3 36.2 3 Overlay 60mm 2000 16.3 72.6
: : : : : : : :
1 N4-16 32.1 6 Reconstruct 2001 22.8 22.8
2 R13 23
R13-23 22 4
22.4 4 O
Overlay
l 40mm
40 2001 97
9.7 32 5
32.5
3 N521-5 45.2 2 Widen 4 lane 2001 41.3 73.8
: : : : : : : :
1 N1-6 30.2 4 Resealing
g 2002 8.2 8.2
2 N7-9 17.8 3 Overlay 60mm 2002 9.2 17.4
3 F2140-8 56.1 1 Reconstruct 2002 34.9 52.3
: : : : : : : :
REQUIREMENTS FOR SUSTAINABLE USE OF
HDM-4
1 Obtain commitment
2 Policy and institutional framework
3 Agree objectives
4 Users and outputs
5 Data and models
6 Software and hardware

WHAT ARE THE NEEDS OF YOUR COUNTRY?
z What do you believe are the weaknesses in

road management in your country
z What are the priorities for improvements in
road management?
z How will you assess the range of
possibilities?
z How do we identify the priorities?
z Are your data collection strategies
sustainable
z HDM-4 is a technical improvement - how
might its use be sustained in your country?
Introduction to Treatment Selection
Aims
A dit
Audit Needs
DATA
Implem. Action
s
Cost/Pr
DECISION SUPPORT ACTIVITIES
z Assess Needs
z Road inventory
z Condition and traffic assessment
z Condition and traffic projection
z Determine Actions
z Apply standards
z Identify and select treatments
z Estimate Quantities and Costs
z Assign Priorities under Budget Constraints
z Analyse Past and Future Trends
Treatment selection methods
z NORMS
z Pre-set assumed treatments based on historical data
z SIMPLE RULES
z Responsive to level of distress, perhaps using
up to 2 defects
z Applicable to carriageway and non
carriageway
i defects
d f t
z COMPLEX RULES
z Responsive
R i tot multiple
lti l distresses
di t
z Can be based on optimum intervention levels
z Mainly applicable to carriageway defects
PAVED ROAD - SIMPLE RULES
DEFECT LEVEL TRAFFIC DEFECT EXTENT ACTION WORK

TYPE
RUTTING < 10 mm >1000 WP cracks<5 % Seal RM

ADT >5 % SD PM
10 – 15 >1000 WP Any As above As above

mm ADT Cracks INV ????
> 15 mm >1000 WP Any INV ????

ADT Cracks
Mix deformation
DRAINAGE - LACK OF MITRES
DRAINAGE - SIMPLE RULES
Item Observation Score/ Action
Severity
Vegetation Condition 0 -< 1 m None
1 -1-2 m Cut
2 -> 2m C t
Cut
DRAINAGE Overall condition 0 - VERY GOOD None or programmed
1 - GOOD De-silt
2 - AVERAGE De-silt
3 - POOR De-silt and repair
4 - VERY POOR Re cut or repair
Existence 0 - Exists None
1 - Not required None
2 - Required Cut drain
Scouring 0 - None None
1 - Slight None
2 - Severe LtReshape/Fill/Scour Checks
Blockage 0 - None None
1 - Slight Clear
2 - Severe Clear
SPECIAL DRAINAGE Required
R i d L
Length h off section
i C off
Cut ff drain/Bund
d i /B d
Mitres
Checks
Lining
CALIBRATNG SIMPLE TREATMENT RULES
Works PIL 1 PIL 2 PIL 2.1

Type
Drains 10 1 1
RM Surface 507 151 167
Other RM 1 4 4
PM surface 609 780 269
Other PM 28 52 52
TOTAL 1155 988 493
COST
(MTsh)
Holding 22 29 5
Treats (n)
EXAMPLE - COMPLEX RULES
Roughness Crocodile Cracking (%) Traffic Range – AADT

Range
(IRI m/km)
Narrow Wide < 150 150 - 500 501 -1000 > 1000
<4 < 10 <5 RM RM RM RM

> 10 <5 RM RM RM RM
> 10 >5 Seal Seal Seal Seal
4-5 < 10 <5 RM RM OLAY OLAY

> 10 <5 RM RM OLAY OLAY
> 10 >5 Seal Seal OLAY OLAY
5-6 Any Any SC and Seal OLAY OLAY OLAY
6-8 Any Any SC and Seal OLAY RECON RECON
>8 Any Any RECON RECON RECON RECON

APPLICATION OF DIFFERENT METHODS
Works type
yp Planningg Programming
g g Design
g
Special Norms S/Norms S
Routine off Norms S/Norms S
c/w
R ti
Routine N
Norms/C
/C S/C S/C
surface
Periodic - S/C S/C S/C
reseal
Periodic - C C NA
rehab
PRIORITIES FOR RURAL ROADS
z the basic need is to provide and maintain basic

access
z at higher traffic levels optimum solutions are
desirable
z access difficulties are concentrated at specific
locations, and should be given high priority
z carefully control the quality of all operations and
z maximise use of local resources
z INVOLVE STAKEHOLDERS in identifying and

agreeing ESSENTIAL maintenance and rehabilitation
tasks
UNPAVED ROADS: TROUBLE SPOT
UNPAVED ROADS: GRADING
Desirable Standards for Rural Roads
Road Road Maintenance Indicative Policy objective
Type Class activities traffic level
Unpaved A All urgent Provide efficient
Gravel (including spot access to most
and improvements), > 75 AADT classes of road user
Improved reactive and cyclic and to minimise
Earth routine light and total transport costs
heavyyg grading
g and g
within budget
all periodic works. constraints.
B All urgent Provide year round
(including spot access for most
improvements) and 25 - 75 AADT classes of road user
all drains, culverts, at a reduced
bridges, rivers and standard and to
slopes and ‘cyclic’ minimise loss of
surface work using road assets within
light graders and budget constraints
lengthmen.
C All urgent Provide year round
(including spot access to vital
improvements) and < 25 AADT public services.
services
all drains, culverts,
bridges and rivers
works.
MAINTENANCE COSTS vs STANDARD
ANNUAL MAINTENANCE COSTS (US$/Km)

Maintenance Road Class
Standard A B C
Desirable 1,689 1,123 557
R d
Reduced d 1 132
1,132 634 406
UNSEALED ROADS -
DEVELOPMENTS

access
desirable

tasks
UNPAVED ROADS: TROUBLE SPOT
EFFECTIVE USE OF RESOURCES
z Accessing new areas using simple methods

z Spot improvements to facilitate access
z Material selection and routine maintenance
z Use of water and compaction during
construction and maintenance
z Durable road cross section
z Productive use of intermediate and labour
based technologies
z Designing programmes on the basis of policy
objectives
IMPORTANCE OF EFFECTIVE PLANNING
z Graph of Cost Effectiveness of Upgrading to a

Motorable Track
Headloading
C1
Track
Costs
Improved road
C2
C3
T1 Traffic T2 T3
z Graph of Cost Effectiveness against

Standard of Road
95% of year, access established

oductivity
y
99% of y
year,, access established
Marrginal pro
Maintenance for roughness reduction
0 Maintenance expenditure $/km

heavyyg grading
g and g
within budget
lengthmen.
services
bridges and rivers
works.

Standard A B C
Desirable 1,689 1,123 557
R d
Reduced d 1 132
1,132 634 406
UNPAVED ROADS: DESIGN AND CONSTRUCTION
STANDARDS
Zimbabwean gravel performance guide
1100
Poor performance - Too plastic
1000
Good in dry weather
900 Slippery with potholes when wet
us
Plasticity modulu
800
700
Too fine Too coarse
Good under both wet and dry conditions
600
500
P
400 Good in wet weather

Corrugates in dry weather
300
Lacks cohesion poor
200 performance
0 10 20 30 40 50 60
Coarseness index (IC) (% > 2.36mm)
UNSEALED ROADS: DESIGN AND CONSTRUCTION
STANDARDS
z Specific requirements for calcretes
70
DT
oss per yearr per 100 AD
60 Powder calcretes and

calcified sands
50 Nodular calcretes
40
(mm)
30
Material lo
20
10
0
0 200 400 600 800 1000 1200
Plasticity modulus
z RATE OF MATERIAL LOSS RELATED TO

MAINTENANCE STRATEGY
Material Strategy ML Per Relative
Type Year Per Wear
100 VPD Rates
No maintenance 20 1.0
Plastic
gravel Wet season light grading (1 in 13 0.7
(calcrete) 11,000
, vehicle passes)
p )
Sheeting and light grading (1 in 6 0.3
3,000 vehicle passes)
No maintenance 41 21
2.1
Clayey
sand Wet season light grading (1 in
(calcified 11,000 vehicle passes) 18 0.9
sand)
Sheeting
Sh ti andd light
li ht grading
di (1 in
i 13 07
0.7
Source: Toole (1987) Performance of unpaved calcrete roads in Botswana
Effect of Surface Maintenance ??
z Consider what treatments might apply?
z Consider effect on current roughness

g and
future progression?
z Consider alternative technologies?

EFFECT OF GRADING ON ROUGHNESS
(Kenya: Motor and Towed Graders)
16
14
12
R o u g h n e sss a fte r g ra d i n g m / k m (IR I)
10
2
2 4 6 8 10 12 14 16
Roughness before grading m/km (IRI)
TRACTOR TOWED - GRADER BLADE DRAG
Long term average roughness
35
30 No grading
25
Grading 2 per year

Average IRI (m//km)
20
15 Grading 4 per year
10
Grading 12 per year
0
0 20 40 60 80 100 120 140 160 180 200
AADT
Costs of Tractor-Towed and Motor Graders
Type Availability Output Cost

(km/day) (US$/km)
Motor 30 2 280
grader
Towed 60 1 75
grader
z RATE OF MATERIAL LOSS RELATED TO

MAINTENANCE STRATEGY
Material Strategy ML Per Relative
Type Year Per Wear
100 VPD Rates
No maintenance 20 1.0
Plastic
gravel Wet season light grading (1 in 13 0.7
(calcrete) 11,000
, vehicle passes)
p )
Sheeting and light grading (1 in 6 0.3
No maintenance 41 21
2.1
Clayey
sand Wet season light grading (1 in
(calcified 11,000 vehicle passes) 18 0.9
sand)
Sheeting
Sh ti andd light
li ht grading
di (1 in
i 13 07
0.7
Source: Toole (1987) Performance of unpaved calcrete roads in Botswana
Unpaved
p Analysis
y
z Light
g and Heavy y Grading
g
Deterioration
zOptimum grading cycles
zMaintenance and whole life cost

comparisons
z When should the road be upgraded to
paved standard ?
IRI & Grading Frequency
40 vpd Light Grading
10
9
8
7
6
5
1 2 3 4 5 6 7 8 9 10 11
Transport Costs & Light
Grading Frequency 40 Vpd (Zw $ per yr)
210000
200000
190000
180000
170000
160000
150000
1 3 5 7 9 11
Optimal Heavy Grading (reprocessing)
Vpd Frequency IRI Gravel loss Total Costs

P
Per yr. M /
Mm/yr 000 Zw$/yr
Z $/
50 1 5.96 14.85 238
60 1 6.39 14.86 279
80 1 7.18 14.90 365
100 2 6.28 17.38 460
150 2 7 45
7.45 17 45
17.45 676
200 3 7.09 19.14 890
250 4 7.04 20.42 1112

Optimal Light Grading
Vpd Frequency IRI Gravel loss Total Costs
Mm/yr 000 Zw$/yr
20 0.5 8.56 18.2 123
50 1.5 8.96 22.5 260
60 2 8.5 23.9 306
80 3 85
8.5 26 8
26.8 399
100 3.5 8.5 29.6 490
150 5 8.6 36.8 720
200 7 8.6 43.9 950
250 8.5 8.7 51.0 1180

A Comparison of Net Present Costs over 20 yr life
( Zw $ M).))
Traffic Gravel 1 Gravel 2 Paved (5 yr

maint.cycle)
30 1.843
1 843 1.813
1 813 2 330
2.330
40 2.346 2.299 2.677
50 2.818 2.756 3.024
60 3.284 3.206 3.371
70 3.791 3.693 3.718
80 4 248
4.248 4 131
4.131 4 081
4.081
90 4.707 4.568 4.428
100 5.177 5.024 4.775
UNSEALED ROADS
Aims and Objectives
z Aim - To significantly improve the

management of unsealed roads through
the adoption of suitable management
strategies
t t i
z Objectives a) To explain and improve
basic understanding of key variables,
variables
maintenance effects and treatment
strategies; b) To describe current work
related to future developments
HOW IMPORTANT ARE UNSEALED ROADS?
Key Statistics for Africa. What about Asia?
z Typically > 60% of main roads are unsealed

z Unsealed minor roads perhaps 7 times greater
z 70% of population live in rural areas
z >>> 30%
% off regional G
GDP is agricultural
z >>> 66% of the population work in agriculture
z 40% of regional exports are agricultural
CLASSIFICATION & CHARACTERISTICS ?
z GRAVEL
z EARTH (or SAND)
z BLOCK PAVEMENTS
KENYA - ENGINEERED EARTH ROAD
ETHIOPIA - CINDER GRAVEL ROAD
THAILAND - RURAL ROAD ON INCLINE
ETHIOPIA - WEAK & POORLY DRAINED SOIL CONDITIONS
MONGOLIA - STEPPE TRACK
DETERIORATION MECHANISMS - ??
z Dry weather
z Wet weather - adequate

q pavements
p
z Wet weather - weak surfacing

g
z Wet weather - weak roadbed

MODES OF DISTRESS ??
z Roughness
z Material Loss
z Rutting
z Looseness
z Impassability
z ARE THERE OTHERS ??????

TANZANIA - EROSION OF AN EARTH TRACK
Unsealed Road Work Classification
Preservation Development
z Routine z Improvement
p s
z Off carriageway z Widening
z Grading z Realignment
z Spot regravelling
g g z Off-carriageway
g y works ??
z Other surface z Construction
z Periodic z Upgrading
z Resurfacing z New sections
z Special
z Emergencies
z Wi t maintenance
Winter i t
MAINTENANCE WORKS TYPES
z PROGRAMMED (or Cyclic) activities -

scheduled works carried out at a set
frequency
z CONDITION RESPONSIVE (or Reactive)

activities - works carried out when a
condition
diti state
t t is
i exceeded
d d
UNSEALED ROADS: RDWE MODELS
z Roughness:
z Progression
z Effect of mechanical compaction
z Effect of grading
z Average annual roughness
z Steady
y state roughness
g
z Material Loss
z Passability
y
ROUGHNESS PROGRESSION - WITHOUT COMPACTION
30
25
200 veh/day 100 veh/day 50 veh/day
20 10 veh/day
IRI (m/km)
15
10
0
0 1 2 3 4 5
Surface Age (years)
Roughness progression at different moisture levels
Effect of Surface Maintenance ??
z Consider what treatments might apply?
z Consider effect on current roughness

g and
future progression?
z Consider alternative technologies?

EFFECT OF GRADING ON ROUGHNESS
(Kenya: Motor and Towed Graders)
16
14
12
R o u g h n e sss a fte r g ra d i n g m / k m (IR I)
10
2
2 4 6 8 10 12 14 16
Roughness before grading m/km (IRI)
REPROCESSING
ZIMBABWE - TOWED GRADER
TRACTOR TOWED - GRADER BLADE DRAG
Long term average roughness
35
30 No grading
25
Grading 2 per year

Average IRI (m//km)
20
15 Grading 4 per year
10
Grading 12 per year
0
0 20 40 60 80 100 120 140 160 180 200
AADT
Material loss: Effect of rainfall and traffic
Material loss: Effect of rainfall and terrain
CONCLUSIONS ?
BITUMINOUS PAVEMENTS
Modelling
g of Works Effects
AIMS
z SESSION AIMS
z To introduce the modelling framework.
z To identify the treatments which are modelled
within HDM-4
z To explain the improvements which result
from implementing maintenance or
improvement activities
z To make participants aware of the key
reference documents for more detailed
explanation.
Road Work Classification
Preservation Development
z Routine z Improvement
p s
z Patching, Edge repair z Widening
z Drainage, Crack sealing z Realignment
z Periodic z Off-carriageway
g y works
z Preventive treatments z Construction
z Rehabilitation z Upgrading
z Pavement reconstruction z New sections
z Special
z Emergencies
z Wi t maintenance
Winter i t
Maintenance Works - Preservation
Works Class Works Type Works Activity / Operation

Routine Routine patching, edge-repair, crack sealing,
Maintenance Pavement spot-regravelling, shoulders repair,
etc.
Drainage culvert repairs, clearing side drains
R ti
Routine vegetation
t ti control,
t l markings,
ki signs
i
Miscellaneous
Periodic Preventive fog seal, rejuvenation
Maintenance Treatment
R
Resurfacing
f i surface
f d
dressing,
i slurry
l seal,
l cape
seal, regravelling
Rehabilitation overlay, mill and replace, inlay
Reconstruction partial reconstruction, full pavement
reconstruction
Special Emergency clearing debris, repairing
washout/subsidence, traffic accident
removal, etc.
Winter snow removal,
removal salting
salting, gritting
gritting, etc
Development Works
Works Class Works Type Works Activity /Operation

Widening partial widening, lane addition,
Realignment horizontal and vertical geometric
improvements, junction
Improvement improvement
Off- shoulders addition, shoulders
carriageway upgrading NMT lane addition
upgrading, addition,
side drain improvement, etc.
Upgrading upgrading by changing the
Construction surface class
New section dualisation of an existing section,
new section (link)
Default Budget Heads
z Routine z Recurrent
z Periodic z Capital
z Special z Special
z I
Improvementst z C it l
Capital
z Construction z Capital
Works Standards
Section A
Maintenance Maintenance Improvement Etc.

Standard M1 Standard M2 Standard I1
[[Operation
p 1]+[Intervention
] [ criteria]+[Limits]+[Effects]+[Cost]="Works
] [ ] [ ] [ ] item 1"
[Operation 2]+[Intervention criteria]+[Limits]+[Effects]+[Cost]="Works item 2"

Maintenance Intervention Criteria
Scheduled
• Fixed intervals of time
• Fixed points in time
Responsive
• Pavement condition
• Pavement
P t strength
t th
• Surface age
• Vehicle speeds
• Traffic volumes / loadings
• Accident rates
Limits
• Minimum and maximum intervals between successive

works in order to comply
p y with budget
g or other
constraints
• Maximum applicable roughness beyond which the
works activity must not be done
• Minimum
Mi i and
d maximum
i AADT threshold
th h ld within
ithi
which a standard is applicable on the road section
• Last applicable year for the works
• Maximum annual quantity of the works
Road Works Modelling
z Timing of works over the analysis period

z Calculation of the p
physical
y q
quantities or
amounts of works to be undertaken
z Estimating the costs of works
z Resetting / changing one or more of the
characteristics that define the road
Works Effects and Costs
Note:
Road Cay = Variable at the beginning of year y
variable Cby = Variable at the end of year y
Responsive
R i Cbo = Ca1 Cb1
intervention
Cao Cb4
Cb2 = Ca3
Ca2
0 1 2 3 4 Years
Scheduled intervention
Percent
total
costs
100
0 1 2 3 4 Years
Works Effects and Costs
Note:
Cay = Variable at the beginning of year y
Cby = Variable at the end of year y
Road
variable Cb1 = Ca2 Cb2 = Ca3
Responsive Cbo = Ca1 Cb3
intervention
Cao
Cb4 = Ca5
Ca4
0 1 2 3 4 Y
Years
Percent Scheduled intervention

total
costs
40
30
0 1 2 3 4 Years
Works activity Existing pavement type

treatment

STSB 1

AMSB 1
I l
Mill & replace to **AP **AP **AP **AP N/A **SB **AB **AP
intermediate
surface layer
Mill & replace to **GB **SB **AB **AP **GB **SB **AB **AP
base
Source: NDLI (1995
Routine Maintenance Operations
z Drainage maintenance and rehabilitation

z Sealing of wide structural cracking
z Sealing
g of transverse thermal cracking
g
z Surface patching of ravelled areas
z Deepp patching
p g of potholes
p
z Deep patching of wide structural cracking
z Repair of pavement edge-break
edge break
Drainage Works
z Drainage condition represented by factor DF

z DF ranges from 1-2 (excellent) to 3-5 (bad)
z DF values depend on type of drain
z Without maintenance DF increases each year
y
z With maintenance the deterioration is reduced
z Rehabilitation restores drainage from poor to
good condition
Effect of Drainage Factor
z DF affects
ff t ratio
ti between
b t wett and
d dry
d
season pavement strengths
z Better drainage means higher average
pavement strength
z Higher pavement strength means lower
roughness progression
Effect of Drainage on Roughness
12
AADT = 2,000
SNPd = 3
3.0
0
MMP = 100 mm/month no drainage
10 WET = 0.5 maintenance
hness (IRI, m/kkm)
rehabilitate
when DF = 5
8 annual
maintenance
DF = 2
Rough
2
0 5 10 15 20
Years
Sealing of Wide Structural Cracking
z In HDM-4 the user specifies the

percentage of wide structural cracks to
be sealed each year
z Th “life”
The “lif ” off the
th seall is
i also
l specified
ifi d
Effect of Crack Sealing
z Potholing does not develop from wide

cracks that have been sealed
z Water ingress is inhibited by crack sealing
z Reduced water ingress gives higher
average pavement strength
z Higher strength reduces crack and
roughness progression
z Crack sealing does not inhibit the
development of new cracks
Effect of Crack Sealing on Performance
10
AADT = 2,000
SNPd = 3.0 maintain drainage
MMP = 100 mm/month
/ th no crack sealing
WET = 0.5
8
m/km)
no drainage maintenance
ughness (IRI, m
seal all wide cracks
6
Rou
maintain drainage
4 seal all wide cracks
2
0 5 10 15 20
Years
Annual Areas of Crack Sealing
30 100
Seal 100% Wide Cracks
90
25
Cumulative Arrea of Wide Craccking (%)

80
g (%/y)
CSL = 3 years 70
a of Crack Sealing
CSL = 5 years
20
60
15 50
40
Annual Area
CSL = 10 years
10
30
20
5 Cumulative
Wide Cracking
10
0 0
0 2 4 6 8 10 12 14 16 18 20
Age (years)
Effect of Sealing Percentage
8
AADT = 2,000
SNPd = 3.0
MMP = 100 mm/month
WET = 0.5 seal 25% wide cracking
m/km)
6
ughness (IRI, m
seal 100% wide cracking

Rou
2
0 5 10 15 20
Years
Sealing of Transverse Thermal Cracking
z Similar in principle to sealing of structural

cracking
z Except annual limit can be expressed in
linear rather than square metres
z Seal life is also specified
Effects of Surface Patching of Ravelled Areas
z Replacement of lost surfacing material,

typically
yp y chip
p loss in surface treatments
z Surfacing patching prevents development
of potholes from ravelled areas
z No effect on pavement strength
Patching of Potholes
z An operation that can occur many times

during a year, not just an annual operation
z HDM-4 uses a parameter TLF to represent
frequency of pothole patching operations
z The effect of TLF is inherent in the road
d t i
deterioration
ti models
d l for
f potholing
th li and d
roughness
Effects of Pothole Patching
z Effect of open potholes is removed

z Residual roughness of patches is added
to total roughness
z Pothole repairs (ie. patches) reduce water
ingress and hence increase pavement
strength
t th
Patching of Wide Structural Cracking
z Deep patching of cracks is deemed to be

permanent, unlike sealing
z Can be specified as an annual percentage
of wide cracking to be patched
z Does not prevent the development of new
cracking
ki
Effects of Crack Patching
z Immediate reduction in area of wide

cracking
z Reduction in roughness due to cracking
z Potholes do not develop from patched
cracks
z Loss of structural strength due to
cracking is restored
z Water ingress is reduced
Comparison of Crack Sealing and Patching
8
AADT = 2,000
SNPd = 3.0
MMP = 100 mm/month
/ th
WET = 0.5 seal 100% wide cracking
HS = 50 mm
m/km)
6
ughness (IRI, m
patch 100% wide cracking

Rou
2
0 5 10 15 20
Years
Repair of Pavement Edge Break
z Reduces the cumulative volume of edge

break
z Restores pavement width
z Reduces effective roughness during
partial shoulder use
Periodic Maintenance
z Preventive treatments
z Resurfacing
z Overlays
z Reconstruction
Preventive Treatments
z Proactive operations applied before the

development of surface distress
z Intended to delay the initiation of cracking
and ravelling
z Includes rejuvenation and fog seals
Resurfacings
z Either proactive
Eith ti as preventive
ti treatments
t t t
or reactive to low levels of surface
distress
z Includes surface treatments, slurry seals
or thin
thi overlays
l
z Resurfacings do not give a significant
increase in pavement strength
Overlays
z Normally reactive to high levels of surface

distress, deformation or roughness
z May increase structural capacity of the
pavement
z Includes thick asphaltic overlays and mill
andd replace
l (or
( recycling)
li ) operations
ti
Reconstruction
z Reactive to high levels of surface distress

and/or deformation and/or structural
weakness
z Involves reworking or replacement of one
or more pavement layers
z U
Usually
ll affects
ff t the
th unbound
b d layers
l
Effect on Pavement Classification
z Preventive treatments do not affect

pavement type
z Resurfacings and overlays may change
the surface type
yp
z Reconstruction may affect both surfacing
and base type
1998 (STAP) 25 mm Surface

New Surfacing D
Dressing
i
1994 (AMAP) 50mm AC Overlay 50 mm AC

Overlay
100 mm
50 mm AC
Previous Surfacing Surfacing
150
150mm G
Granular
l S Sub-base
bb
Subgrade (CBR 8%)

Effect on Pavement Structure
z Preventive treatments have no direct

effect on pavement structure
z Resurfacings and overlays increment the
dry season pavement strength depending
on thickness and materials properties
z R
Reconstruction
t ti resets
t the
th structural
t t l
number to a new, specified, value
Pavement Ages
z AGE1 is years since last preventive

treatment
z AGE2 is years since last resurfacing
z AGE3 is years since last structural overlay
z AGE4 is years since original construction
or reconstruction including base layers
Reset of Pavement Ages
Works Type AGE1 AGE2 AGE3 AGE4

Preventive 0 N N N
Resurfacing 0 0 N N
Overlay 0 0 0 N
Mill and Replace 0 0 0 N
Reconstruction 0 0 0 0
0 – reset to zero
ero
N – no reset
Effects on Cracking and Ravelling
z Preventive treatments do not remove cracking

or ravelling
z They retard the initation (if not already initiated)
z They retard the progression to a limited extent
z All other treatments reset cracking and ravelling
to zero
z After resurfacing or overlay, previous cracking
is reset to the amount before treatment
z After reconstruction, previous cracking is reset
to zero
Effects on Rut Depth
z Preventive treatments and seals without

shape correction have no effect
z Seals with shape correction and overlays
reduce rut depth
p to 15%% of its old value
z Reconstruction and mill and replace reset
rut depth to zero
Effects on Roughness
z Preventive treatments have no direct effect

on roughness
z Seals without shape correction can give a
small reduction in roughness
z Seals with shape correction give a larger
reduction in roughness
Roughness Effects - Seals
12
10
Surface treatment
IRI After Seal (m/km)
Slurry seal - 10mm

6
0
0 2 4 6 8 10 12
IRI Before Seal (m/km)
Roughness Effects - Seals with Shape Correction
12
10
Thickness including shape correction 10mm
m/km)
8 30mm
RI After Seal (m
50mm
6
IR
0
0 2 4 6 8 10 12
IRI Before Seal (m/km)
Roughness Effects - Overlays
z Roughness reduction after an overlay is

determined by:
z th thickness
the thi k off the
th overlay
l
z the roughness before overlay
z the quality of workmanship
Roughness Effects - Overlays
9
8
Overlay thickness 20mm
7
40mm
IRI After Overlay (m/km)
6
60mm
5
80mm
4
0
0 2 4 6 8 10 12 14
IRI Before Overlay (m/km)
Roughness Effects - Reconstruction
z The roughness after reconstruction is

user specified
z T i l IRI values
Typical l are:
z 1.0 - 2.0 m/km for asphalt surfacings
z 2.5 m/km for surface treatments
z 6 - 8 m/km for hand-laid
hand laid surfacings
Effects on Surface Texture
z In HDM-4, the values of skid resistance

and texture depth after a periodic
operation must be specified by the user
Improvement Works
z Two operations available in HDM-4

z Widening
z Realignment
Effects of Widening
z Lane addition
z Partial widening - no increase in number
of lanes
z Widening does not change alignment
Effects of Realignment
z Change to horizontal geometry (deg/km)

z Change to vertical geometry (m/km)
z Change in length
z But no change to roadway width
New Construction
z Two options available in HDM-4

z Upgrading
pg g to new pavement
p type
yp
z Construction of a new road section
z All parameters
t are resett to
t user defined
d fi d
values
INTERFACING HDM
HDM-4
4 WITH
EXTERNAL SYSTEMS
Interfacing HDM-4 with External Systems
- Implementation Activities
z NOT just physical linkage to external system!

z Two steps:
p
z Calibration of HDM-4 to local conditions (50%) *
z Data interface development
p (50%)
( )*
* not necessarily in this order. Calibration is an

iterative process of refinement.
What calibration needs to be done for each

implementation of HDM-4?
Calibration/Adaptation Activities
- Initial Data Flow Schematic
Work
Inventory Programs
Projects
Survey
y
Etc.
RDBMS HDM-4
z Vehicle Fleet
z Works Standards
z HDM 4 Configuration Data
HDM-4
z Model Calibration
z R
Reports
t
z Language
- Vehicle Fleet Calibration
z Representative Vehicle Types

z Basic Characteristics
z Unit Costs
z Detailed Calibration
The above data is held in HDM-4 database

(e.g. Bangladesh Vehicle Fleet)
Motorised vehicle types:

z Baby Taxi
z Mishuk
z Tempo
z Small Truck (converted jeep)
z Tata Truck
z Toyota Corolla
z Coach
- Works Standards
z Works types
yp
z maintenance
z improvement
z Unit costs
z Construction quality
q y
z material properties
z construction quality
q y
z compaction, etc.
(e.g. Bangladesh Work Types)
z Works types:
z SBST
z DBST
z Thin Overlay
z Thick Overlay
z Rehabilitation
z Widening
z Intervention levels
- HDM-4 Configuration Data
z Traffic Flow Patterns

z Speed Flow Types
z Climate Zones
z Currencies
z Aggregate Data
Above data is held in HDM-4 database

(not RDBMS)
- HDM-4 Configuration Data
z Aggregate data - country specific:

z traffic volume
z road class
z geometry class
z surface texture
z surface condition
z structural adequacy
z ride quality
- RD Model Calibration
Typical RD relationship:
d C = Kcpw CRP zw [(
dACW [(zw a4 a5 δtW YE4 (S
(SNPa6) + SCW
SC a5 )1/a5
CDS - SCW]
( h
(change in
i area off wide
id structural
t t l cracking)
ki )
dACW = Kcpw CRP zw [(zw a4 a5 δtW YE4 (SNPa6) + SCWa5 )1/a5

CDS - SCW]
Calibration Model
Factor Coefficients
RD calibration factors:
RD model coefficients:
- Other Activities
z Reports
z Special report requirements
z Using user-defined reports facility
z Language
z Is native language version required?
z User interface
z Reports
z Documentation, etc.
- Revised Data Flow Schematic
Configuration Data Vehicle Data

Model Coefficients Work Standards
Work
Inventory Programs
Projects
Survey
y
Special
Calibration Reports
Factors RDBMS HDM 4
HDM-4 Etc.
How do we link HDM-4 to external systems?

How do we develop p the data-exchange
g
interface?
HDM-4 Import / Export Facilities
z Support for all HDM

HDM-4
4 items:
z Road Networks
z Vehicle Fleets
z Projects, Programmes, Strategies
z Configuration,
g , etc.
z For transfer of data between PCs
z Interfacing with external systems
z Uses dBase ‘.dbf’ files
Import / Export Facilities -
Road Network Import Format
Import / Export Facilities -
Sample Road Network Output
And now for the ‘clever’ bit...

- The Data Exchange Interface

Work
Inventory Programs
Projects
Survey
Special
S i l
Calibration Reports
factors HDM 4
HDM-4 Etc
Etc.
RDBMS
S
Generating the Export File
- Outline Procedure
1 0 User selects links from RDBMS

1.0
2.0 Generate sections
2.1 Segmentation
2.2 Selection of defaults ((using
g look-up
p table))
2.3 Aggregation
30
3.0 W it section(s)
Write ti ( ) to
t exportt file
fil (dBase)
(dB )
- Selecting Sections
- Selecting Sections
- Coping with Missing Data
Defaults Total HDM-4 RMS data

for missing
data items
= road data
requirement
compatible
with HDM-4
- Data Interface Schematic
Selected Representative
RDBMS Section Look-up
Links Table
- Export Options
- Outline Procedure

1.0
2.1 Segmentation
g look-up
p table))
2.3 Aggregation
30
3.0 W it section(s)
Write ti ( ) to
t exportt file
fil (dBase)
(dB )
- the Traffic Problem
P bl
Problem
Road networks in HDM-4 version 1.0 only
support TOTAL motorised and NMT traffic,
traffic and
not traffic by vehicle type.
Solution
Don t use a road network export file to transfer
Don’t
data - use a programme analysis export file.
Any Questions?
Institutional Impacts of Implementing HDM-4
z Should we:
z adapt
d t HDM-4
HDM 4 tto suit
it local
l l methods*,
th d * or
z adapt local methods to suit HDM-4?
(* e.g. survey / data collection)

ROAD MANAGEMENT
A Framework for Sustainable

Road Management
THE MANAGEMENT FUNCTIONS
z Planning
z Setting aims and objectives
z Determining standards and policies
z Estimating long term expenditure
z Programming
z Medium term work programmes
z Preparation
z Detailed project design and work packaging
z Operations
z Implementation of works in field
Aims
A dit
Audit Needs
DATA
Implem. Action
s
Cost/Pr
DECISION SUPPORT ACTIVITIES
z Assess Needs
z Road inventory
z Condition and traffic assessment
z Condition and traffic projection
z Determine Actions
z Apply standards
z Identify and select treatments
z Estimate Quantities and Costs
z Assign Priorities under Budget Constraints
z Analyse Past and Future Trends
QUESTIONS
z What is essential to support the TASKS?
z What constraints may

y exist in assessing
g
needs, and how might you address them?
z How might you best deploy your limited

resources?
DATA DESIGN ISSUES
z RELEVANCE
z Only collect items of direct relevance on OUTPUTS and
DECISION MAKING today
z Adopt standard units and dimensions
z APPROPRIATENESS
z Choose a strategy based on intended use, eg. network or
project level planning, design, etc.
z RELIABILITY
z Choose a level of precision depending on application.
Data to be sufficiently
y accurate,, complete
p and up
p to date.
z AFFORDABILITY
z The quantity and quality of all data must be affordable in
resource terms and costs effective in decision making
terms.
Information Quality Levels,
Calibration and Validation
AIMS AND OJECTIVES
z AIMS - To encourage a structured

approach to the selection of key data and
models to support cost effective decision
making
ki
z OBJECTIVES - To increase knowledge
and skills in the identification of key data
items, model calibration and strategies for
collection and application.
SCOPE
z Reliability of results
z Application of results
z Calibration focus
z RUE/RDWE calibration
z Calibration levels
z Reliability concepts
z What to focus on
z Information Quality Levels
Reliability of Results Depend On
z How well data provided represent the real

conditions being analysed as understood
by the model
z How well the predictions of the model fit

th reall behaviour
the b h i and
d the
th interactions
i t ti
between the various factors and
conditions to which it is applied
Application of Model
z Input Data
z Must have a correct interpretation of the
input data requirements
z Have a quality of input data appropriate for
the desired reliability of results
z C lib ti
Calibration
z Adjust model parameters to enhance the
accuracy of its representation of local
conditions
Steps in Analysis
Predictive
Input Data Analysis Results
Models
This Presentation
Calibration Focus
z Road User Effects

z The models must predict the correct
magnitude of costs and relativity between
components
t
z Road Deterioration and Works Effects
z The models
Th d l mustt reflect
fl t local
l l pavementt
deterioration rates and maintenance
practices/effects
HDM Development
z Used structured mechanistic-empirical

approach
z Pavement deterioration validated against four
major overseas studies
z HDM applied in over 100 countries with varied
climate and pavement types
z Found to give reasonable predictions when
calibrated correctly
y
RUE Calibration - Canada
180
2-Axle Truck 5-Axle Truck 7-Axle Truck
160
140
120
Crew
Cost in cents/km
100 Interest
Depreciation
80 Maintenance
Tires
60 Fuel & Oil
40
20
0
Observed HDM Observed HDM Observed HDM
RDWE CALIBRATION IN ZIMBABWE
z Roughness Progression
z Crown height (and earthworks quantities)
z Drainage quality
z Design standard
z Unstable subgrades (Expansive and Dispersive)
z Crack initiation and progression

z Design
g standard
z Original surfacing (single or double)
z Reseals
Sealed Roads: Roughness - Age Relationship
Design Roughness – Age

Standard Description factor
( g )
(Kge)
Low cost/SFRDP design
0.05 MESA with (initial) single
surfacing and low crown 1.25
h i ht (< 0.7m)
height 07 )
DSR standard design with

0.3 MESA (initial) double surfacing
and moderate –highhigh crown 07
0.7
height (> 0.7m)
DSR standard (trunk road)

1.0 MESA design with (initial) double
surfacing and high crown 0.3
height (> 1.0m)
Sealed Roads: Rate of Crack Progression (Kcpa)
Design Original Reseal Original Reseal

Standard Single
g on Double on Double
Single
0.05 MESA 1.5 1.0 1.2 No data
0.3 MESA No data No data 1.1 1.2
1 0 MESA
1.0 No data No data (?) 05
0.5
INDONESIA - ROUGHNESS CALIBRATION
Construction Quality Traffic (MESA) Detailed

Model
Kgm Kgp
Well constructed roads with average to good Heavy > 0.75 1.3 0.9
asphalt surfacings in flat to rolling terrain and
free flowing traffic conditions. Light- < 0.75 2.4 1.2
Medium
Poorly designed/constructed road, exhibiting Heavy > 0.75 1.3 1.1
failures due to poor road widening and
reinstatement prior to overlay and poor mix Light- < 0.75 1.0 1.2
design in flat to rolling terrain and free flow
design, Medium
traffic conditions.
Well constructed roads located in All All n/a n/a
mountainous regions with average to good
p
asphalt surfacings
g
Well constructed roads in mountainous All All 7.0 1.5
regions with poor asphalt surfacings
Typical Road Deterioration Relationship
dACW = KcpwCRP [( w a4 a5 δtW YE4 (SNPa6) + SCWa5 )1/a5

zw [(z
CDS - SCW]
(change in area of wide structural

cracking)
CRP
dACW = Kcpw [(zw a1 a2 δtW YE4 ((SNPa3) + SCWa4 )1/a5 -
zw [(
CDS SCW]
Calibration Model
Factor Coefficients
Calibration Levels
z Level 1: Basic Application

z Addresses most critical parameters
z ‘Desk Study’
z Level 2: Calibration
z Measures key parameters
z Conducts field surveys
y
z Level 3: Adaptation
z Major field surveys to requantify fundamental
relationships
Hierarchy of Effort
Time Required
Experimental
Years
Surveys and
Research
Months
Field Surveys
Weeks
Desk Studies
Resources
Limited Moderate Significant q
Required
General Planning Project Appraisal Research and

Quick Prioritisation Detailed Feasibility Development
Preliminary Screening
Coarse Estimates Reliable Estimates

Level 1 - Application
z Required for ALL HDM analyses

z One off ‘set-up’ investment for the model
z Primarily
y based on secondary y sources
z Assumes bulk of HDM default values
appropriate
Level 2 - Calibration
z Uses local measurements to verify and

adjust predictions
z Requires more data collection and higher
precision
Level 3 - Adaptation
z Comprised of
z Improved data collection
z Fundamental research
z Leads to more accurate data by observing
over long time period
z Often leads to alternative local
relationships
Reliability Concepts
z A model is representation of reality

z How well the model predictions reflect
reality depends on:
z the validity of the underlying relationships
z the accuracy and adequacy of the input
data
z calibration factors used in the analysis
Bias and Precision
z Only way of assessing model’s reliability

is by comparing its predictions to known
data
z Need to take into account two
considerations:
z Bias
Bi
z Precision
Combinations of Bias and Precision
Low Bias Low Bias

A High Precision B Low Precision
Data
Predicted
Predicted
Data
Observed = Predicted Observed = Predicted
Observed Observed
High Bias High Bias

C High Precision D Low Precision

dicted
dicted
Pred
Pred
Data
Data
Observed Observed
Correction Factors
z Used to correct for bias

z Two types of factors
z Rotation (CF = Observed/Predicted)
z Translation (CF = Observed - Predicted)
z Translation factors shift the predictions
vertically; rotation factors adjust the slope
Rotation and Translation Factors
A Rotation B Translation
Translation
Rotation
Predicted
Predicted
Data Data
Observed Observed
Translation
Rotation and
C Translation
Rotation
Observed = Predicted
Prediccted
Data
Observed
Bias and Precision in Input Data
frequency
Bias
Population Sampled
value
frequency Precision
Sampled
P
Population
l ti
value
Important Considerations
z Must calibrate over full range of values

likely to be encountered
z Must have sufficient data to detect the
nature of bias and level of precision
z High correlation (r^2) does not always
mean high
hi h accuracy: there
th can still
till be
b
significant bias
RDWE Calibration - 1
z Simulation of Past
z take sample of roads with historical data
(traffic, design, etc.)
z simulate deterioration from construction to
current age
z compare results
z Average predicted condition should be
similar to current condition
RDWE Calibration - 2
z Controlled Studies
z collects detailed data over time on traffic,
roughness, deflections, condition, rut
d th
depths
z sections must be continually monitored
z long term (5 yr) commitment to quality data
long-term
collection
Road User Effects
z Some data available from field studies;

other from controlled experiments
z Can verify using tariff surveys
7000
Regression to
6000 tariff data
5000
Tariff in Rs/trip
4000
Model prediction for:

Roughness: 6.75 m/km
3000
VCR: 0.21
Width: 7.0 m
2000
1000
0
0 100 200 300 400 500 600 700 800
Trip Length in km
RUE - Fleet Surveys
z In many developed countries data are

available from vehicle fleet management
companies
z Depending on availability and level of
disaggregation may offer scope for
calibration of both level and relative
contributions of RUE components
What to Focus On
z HDM-III had about 80+ data items and

model parameters; HDM-4 over 100
z To assist users, conducted sensitivity
tests and defined ‘impact elasticities’
z Grouped data into ranges
Sensitivity Classes
Impact
p Sensitivity
y Impact
p
Class Elasticity
High S-I > 0.50
Medium S-II 0.20 – 0.50
Low S-III 0.05 – 0.20
Negligible S IV
S-IV <0
0.05
05
Sensitivity Impact Variables or data important Variables or data important
class2 elasticity for total VOC3 for VOC savings4
Replacement vehicle price Replacement vehicle price
S-I > 0.50 Parts model exponent kp Parts model exponent kp
Roughness in IRI Roughness in IRI

S-II 0.20 - 0.50 Parts model a1 Parts model a1
Aerodynamic drag coefficient Aerodynamic drag coefficient

Aerodynamic drag multiplier Cdmult Aerodynamic drag multiplier Cdmult
Annual utilisation Annual utilisation
Base fuel-power factor ξb Base fuel-power factor ξb
Cost of fuel Cost of fuel
Cost of maintenance labour Cost of maintenance labour
Cost of passenger work time Cost of passenger work time
Desired speed Cost of tyre
Driveline efficiency Driveline efficiency
Driving power Driving power
Engine accessories power factor Engine accessories power factor
F lξ
Fuel E i idle
Engine idl speedd
Interest rate Fuel ξ
Labour hours calibration factor Interest rate
Mass density of air Labour hours calibration factor
Maximum average rectified velocity Mass density of air
S-III 0.05 - 0.20
NMT friction factor XNMT Maximum average rectified velocity
Number of passengers Number of passengers
Number of wheels Number of wheels
Parts model a0 Parts model a0
Posted speed limit Posted speed limit
Projected frontal area Projected frontal area
Rolling resistance CR1 Rolling resistance CR1
Rolling resistance CR2 a0 Rolling resistance CR2 a0
Rolling resistance factor CR2 Rolling resistance factor CR2
Roughness-speed a0 Roughness-speed a0
Side friction factor XFRI Speed β
Speed bias correction factor Speed limit enforcement factor
Speed limit enforcement factor Vehicle mass
Vehicle mass Vehicle service life
Vehicle service life Volume of wearable rubber
Wheel diameter Wheel diameter
S-IV <0.05 All others All others

Sensitivity Impact Parameter Outcomes Most Impacted
Class Elasticity Pavement Resurfacing Economic
Performance and Surface Return on
Distress Maintenance
2/
S-I > 0.50 Structural Number
2/
Modified Structural Number
Traffic Volume
3/
Deflection
R
Roughness
h
S-II 0.20 - Annual Loading
0.50 Age
All cracking area
Wide cracking area
Roughness-environment
g factor
Cracking initiation factor
Cracking progression factor
S-III 0.05 - Subgrade CBR (with SN)
0.20 Surface thickness (with SN)
Heavy axles volume
Potholing area
Rut depth mean
Rut depth standard deviation
Rut depth progression factor
Roughness general factor
S-IV < 0.05 Deflection (with SNC)
S b d compaction
Subgrade ti
Rainfall (with Kge)
Ravelling area
Ravelling factor
RUE Model Calibration Priorities
Assume
Required Priority 1 Priority 2
Defaults
Mass
Power
Capacity
Tyre Parameters
Level 1 Unit Costs Speed
Depreciation
All Others
Service Life
Parameters
Utilisation
Priority 1 Priority 2 Priority 3
Fuel
Speed
S d
Power
Level 2 Capacity
Service Life
All Others
PCSE
Utlilisation
Priorities
Depend on
Level 3 Resources and
Objectives
Bituminous Pavement Deterioration Priorities
Deterioration factor Impact class for Impact Typical Net Sensi-
given criteria 1 Elast- values Impact tivity
icity of class
Factor (%)
1 2 3 4 5 6
Roughness-age- D D B C B B 0.20 0.2 - 5.0 10
environment
Cracking initiation A C B B C B 0.25 0.5 - 2.0 6 High
Cracking progression A C C C C C 0.22 0.5 - 2.0 6
Rut depth progression D A B D C C 0.10 0.5 - 2.0 3
Roughness progression D D B D C B 0.09 0.8 - 1.2 1 Low

general
Potholing progression D D C B C D 0.03 0.3 - 3.0 2
Ravelling initiation D D D C D D 0.01 0.2 - 3.0 1

Importance of Data
z Accuracy of data has major impact on

predictions
Condition Confidence
Interval
Existing
Condition Criteria
for
1 Intervention
Condition Condition
2 1 2
Trigger Trigger
Interval Interval
Time
Information Quality Levels - IQL
HIGH LEVEL DATA
System Performance
IQL-5 Performance
Monitoring
Planning and
IQL-4 Structure Condition
Performance Evaluation
Programme Analysis or
IQL-3 Ride Distress Friction
Detailed Planning
Project Level or
IQL-2 Detailed Programme
Project Detail or
IQL-1
Research
LOW LEVEL DATA

IQL Levels
z IQL - 1: Fundamental Research

z many attributes measured/identified
z IQL - 2: Project Level
z detail typical for design
z IQL - 3: Programming Level
z few attributes, network level
z IQL - 4: Planning
z key management attributes
z IQL - 5: Key
y Performance Indicators
IQL’s FOR PAVEMENT DATA
Item 1 2 3 4
Roughness Profile elevations IRI Precise IRI IRI
W b d analysis
Waveband l i approx estimate
ti t
IRI
Cracking Type, Extent, severity, Type, extent Type and Extent
intensity & location and severity extent
Disintegration Extent and severity of Extent only

potholes, etc Any -
Deformation Rut depth (M & SD), Rut depth

transverse pprofile,, ((M & SD))
crossfall, corrugations Mean RD -
Description of IQL Levels
Level Amount of Detail
1 Most comprehensive level of detail, such as would be used as a reference benchmark for
other measurement methods and in fundamental research. Would also be used in detailed
field investigations for an in-depth diagnosis of problems, and for high-class project design.
N
Normallyll usedd att project-level
j tl l iin special
i l cases, and
d unlikely
lik l tto b
be usedd ffor network
t k
monitoring. Requires high level of staff skills and institutional resources to support and utilise
collection methods.
2 A level of detail sufficient for comprehensive programming models and for standard design
methods. For planning, would be used only on a sample coverage. Sufficient to distinguish
the pperformance and economic returns of different technical options p with p practical differences
in dimensions or materials. Standard acquisition methods for project-level data collection.
Would usually require automated acquisition methods for network surveys and use for
network-level programming. Requires reliable institutional support and resources.
3 Sufficient detail for planning models and standard programming models for full network
coverage. For project design, would suit elementary methods such as catalogue-type with
meagre data needs,
needs and low-volume
low volume road/bridge design methods
methods. Able to be collected in
network surveys by semi-automated methods or combined automated and manual methods.
4 The basic summary statistics of inventory, performance and utilisation, of interest to
providers and users. Suitable for the simplest planning and programming models, but for
projects is suitable only for standardised designs of very low-volume roads. The simplest,
most basic collection methods, either entirelyy manual or entirely y semi-automated, provide
direct but approximate measures, and suit small or resource-poor agencies. Alternatively, the
statistics may be computed from more detailed data.
5 A level of detail that would combine pavement quality with other measures such as structural
adequacy, safety aspects, and traffic congestion—that is representing a higher order
information such as “road condition”.
Full Details on RUE Calibration
z Specific details on how to calibrate the

RUE model are given in the HDM-4
Calibration and Adaptation Guide as well
as the
th bbookkMModelling
d lli Road
R d User
U andd
Environmental Effects in HDM-4.
Full Details on RDWE Calibration
z Specific details on how to calibrate the

RDWE model are given in the HDM-4
Calibration and Adaptation Guide
Can We Believe the Output?
z Yes, if calibrated
z HDM has proved suitable in a range of
countries
z As with any model, need to carefully
scrutinise output against judgement
z If unexpected predictions problem with (a)
data (b) calibration (c) the models, or (d)
your judgement
j d t
PRACTICAL: CREATING A ROAD
PRACTICAL
NETWORK & VEHICLE FLEET
Road Network - Key Skills / Points
z C eat g a Road
Creating oad Network
et o
z Creating a Road Section
z Concept of homogeneous section
z Use of aggregate data
z Categories of road section data
z ALL road data is mandatory
z Re use of existing sections
Re-use
z Use of representative section set
z NB importing of road data
NB.
Road Section Data Categories
Road Section
Data
Definition Geometry Pavement Condition
Length Rise + Fall Strength Roughness

Width Curvature Surfacing Rutting
Lanes Speed Limit Works History Potholes
Speed
p Flow Type
yp Altitude Base Crackingg
Traffic Edge Break
Surface Class
Calibration
Road Network - Documentation
4 Road Networks
4.1 Keyy concepts
p 7
4.2 Software overview 7
4.3 Creating new Road Networks with no Sections 9
4.4 Creating Road Networks based on existing networks 10
45
4.5 Viewing / editing Road Networks 10
4.6 Deleting Road Networks 11
4.7 Renaming Road Networks 11
4.8 Exporting Road Networks 12
4.9 Importing Road Networks 12
4.10 Adding Sections based on aggregate data 13
4.11 Adding Sections based on existing Sections 14
4 12
4.12 Editing Section attributes 15
4.13 Specifying which Sections are displayed in the 17
Road Network window
4.14 Editing Section details directly from the Road 18
N
Networkk window
i d
4.15 Deleting Sections 18
Vehicle Fleet - Key Skills / Points
z Creating a Vehicle Fleet

z Creating a Vehicle Type
z HDM-4 default types
z Concept of representative types
z Categories
g of vehicle data
z Most vehicle data are required
z All in volume
o u e 3 section
sect o 5
z Road & Vehicle data entered by
different users
Vehicle Fleet Data Categories
Vehicle
V hi l Type
T
Data
Definition Basic Economic Financial

Data Unit Costs Unit Costs
Calibration
Vehicle Fleet - Documentation
5 Vehicle Fleets
5.1 Key Concepts 20
5.3 Creating Vehicle Fleets 22
5.4 Creating Vehicle Fleets based on existing fleets 23
5.5 Viewing / editing Vehicle Fleets 23
5.6 Deleting Vehicle Fleets 25
5.7 Renaming Vehicle Fleets 25
5.8 Importing Vehicle Fleets 26
5.9 Exporting Vehicle Fleets 26
5 10
5.10 Adding motorised Vehicle Types (MT) to a 27
Vehicle Fleet
5.11 Adding non-motorised Vehicle Types (NMT) 28
to a Vehicle Fleet
5.12
5 12 Editi Vehicle
Editing V hi l Type
T attributes
tt ib t 29
5.13 Specifying which Vehicle Types are displayed in 32
the Vehicle Fleet window
5.14 Editing Vehicle Type attributes directly from the
Vehicle Fleet window 33
5.15 Deleting Vehicle Types 34
RD Inputs and Outputs
z Create our own sections

z Define other road data (e.g.
(e g climate data)
z Setting up a simple project
z Interpreting RD output
RD Sensitivity
z TRAFFIC - no. of esa’s

z CLIMATE - moisture-temperature, rainfall
z TIME ((AGE)) - since construction or
maintenance
z STRENGTH - structural no, DEF, etc.
z DRAINAGE - condition/quality
z CONSTRUCTION - type yp and quality
q y (CDS)
( )
z MAINTENANCE - type and quality
RD Sensitivity
< Hands on Group Session >

HDM-4 CONFIGURATION
PURPOSES
Customising data sets to reflect local

circumstances
Managing aggregate data
Managing HDM-4 default data
COMPONENTS
Climate zones
Road network data
z Road class
z Speed-flow types
z Traffic flow pattern
z Aggregate variables
z Default data
Currency
CLIMATE ZONES
Climate has significant impact on

pavement deterioration
Climatic conditions affect vehicle
operation (e.g., rolling resistance, cold
engine start, etc.)
Definition is based on moisture
classification and temperature regimes
Number of climate zones are defined by
the user
MOISTURE CLASSIFICATION
Arid: very low rainfall, high evaporation,

annual precipitation (AP) < 300 mm
Semi-arid: low rainfall, AP 300 ~ 800 mm
Sub-humid: moderate rainfall or strongly
seasonal rainfall, AP 800 ~ 1600 mm
Humid: moderate warm seasonal rainfall
annual precipitation 1500 ~ 3000 mm
Per-humid: high rainfall or very many wet-
days,
y , AP > 2400 mm
TEMPERATURE CLASSIFICATION
Tropical: warm temperatures within a small

range (20 ~ 35 ºC)
Sub-tropical-hot: high day, cool night
temperatures, hot/cold seasons, (-5 ~ 45 ºC)
Sub-tropical-cool: moderate day temperatures,
cool winters, (-10 ~ 30 ºC)
Temperate-cool: warm summer, shallow freeze
(-20 to 25 ºC)
Temperate-freeze: cool summer, deep freeze
winter (-40 to 20 ºC)
CLIMATE DATA
Moisture-class related data:

z Mean monthly precipitation (mm/year),
Thorntwaithe moisture index
z Duration of dry season (as a proportion of
a year)
Temperature-class related data:
z Mean annual temp, temp range ºC
C
z Number of days with temp exceeding 32 ºC,
freezing index
index, etc.
etc
ROAD NETWORK DATA
Road class
z functional hierarchy
y to facilitate management
g or
analysis of the road network
Speed-flow types
z models the effects of traffic volume on speed and the
economic consequences of road improvements
Traffic flow pattern
z models traffic congestion at different hours of the
day and on different days of the week and year
day,
Aggregate data (defaults)
z pavement condition
condition, strength,
strength road geometry,
geometry etc
etc.
ROAD CLASSES
Managing roads: identify who is responsible

for what class of road
Funding: allocation of funding to roads of
different classes, separate sources of funding
Recognition of different road functions which
could be defined in the policy framework of
road administration
Maintenance and improvement: setting up
diff
different objectives,
bj i standards
d d and d
intervention levels for works for each road
class
SPEED- FLOW MODEL
Speed (km/hr)
S1
S2
S3
S lt
Sult Flow
Qo Qnom Qult PCSE/h
DEFINING A SPEED-FLOW TYPE
Name
Road type
Ultimate capacity (Qult) in PCSE/lane/h
Free flow capacity as a proportion of Qult
Nominal capacity as a proportion of Qult
Jam speed
p at capacity
p y in km/h
TRAFFIC FLOW PATTERN
Flow Periods
Peak
Flow Next to Peak
Medium flow
Next to Low
Low/Overnight
Number of Hours in the Year

DEFINING TRAFFIC FLOW PATTERNS
Name
R d use
Road
Number of flow-frequency periods
Number of hours in each flow-frequency
period (HRYRp)
Percentage of AADT in each flow-
frequency period (PCNADTp), or
Hourly flow in each period (HVp) as a
proportion
p p of AADT
AGGREGATE DATA
User defined
z Traffic levels: e
e.g.,
g low
low, medium
medium, high
high.
z Geometry class: in terms of parameters reflecting
horizontal and vertical alignment
z Pavement characteristics: structure and strength
parameters defined according
p g to pavement
p surface
class
z Road condition: ride quality, surface distress and
surface texture
z Pavement history: construction quality, pavement
age etc.
age, etc
DEFINING TRAFFIC BANDS
Defined by:
y
z Description e.g., high, medium, low
z Pavement surface class (e
(e.g.,
g bituminous
bituminous,
concrete or unsealed)
z Mean AADT
Unlimited number of traffic bands
DEFINING A ROAD GEOMETRY CLASS
Defined by :
z Description e.g., mostly straight and gently
undulating, bendy and winding, etc.
z A
Average rise
i plus
l fall
f ll (m/km)
( /k )
z Average number of rises and falls (no./km)
z Average horizontal curvature (deg/km)
z Average super-elevation (%)
z Other speed related data (e.g., speed limit)
Unlimited number of geometry classes
BITUMINOUS PAVEMENT STRENGTH
Defined in terms of the structural adequacy

to carry traffic loading
Structural adequacy
z defined in qualitative bands e.g., good, fair,
p
poor
z each band is assigned a representative SNP
z default pavement layer thickness for each
band
BITUMINOUS PAVEMENTS - LAYER THICKNESS
Method 1:
z Define several ranges
g of SNP values
z For each range and for each pavement type
define HSNEW, HSOLD, HBASE
z Specify default surface material for each
surface type (i.e., AM or ST), and resilient
modulus for stabilised road base type
Method 2 (HDM-4 Default):
z Layer thickness
L thi k derived
d i d usingi pre-defined
d fi d
range of SNP values, layer strength coefficients
for surfacing
g and base material,, and subgrade
g
CBR of 8%
CONCRETE PAVEMENTS
Structural adequacy is based on the

pavementt characteristics
h t i ti and d traffic
t ffi level
l l
For each structural adequacy and traffic
band, define slab thickness and:
z Modulus of rupture (MPa) for
z Jointed Plain concrete pavement (JPCP)
z Percentage of steel reinforcement for
z Jointed Reinforced (JRCP) and Continuously
Reinforced (CRCP) concrete pavements
UNSEALED ROADS
Requires detailed data on surfacing and

subgrade material:
z Gradation - maximum particle size, and
percent passing 2.0,
20 00.425
425 and 0
0.075
075 mm
sieves
z Plasticity
y index
Structural adequacy and surface condition
are related to the traffic level and are
represented by the thickness of the surfacing
material (gravel)
ROAD CONDITION
Aggregated into 3 measures:

z Ride quality - an indication of the roughness
of the road
z Surface condition - modelled by a number of
distress modes depending on the pavement
type
z Surface texture - gives an indication of texture
depth and skid resistance of the surface
RIDE QUALITY
An important parameter for:

z Indicating road condition and maintenance
needs
z Predicting vehicle operating cost
For each road surface class define the
number of bands e.g., good, fair, poor, bad
For each band define the mean roughness
value, IRI (m/km)
SURFACE CONDITION
Defined by an aggregate measure

representing several distress modes
For each road surface class define the bands
e.g., new, good, fair, poor, bad
Define mean values of the relevant distress
modes e.g., cracking, ravelling, potholing,
etc for bituminous pavements
etc.
SURFACE TEXTURE
Defined by a qualitative measure that gives an

indication of the texture depth and skid
resistance
For bituminous pavements,
pavements define the bands
e.g., good, fair, slippery
For each band
band, and for each of the road
surface types AM or ST, assign mean values
of the sand ppatch texture depth
p and skid
resistance at 50 km/h (SCRIM)
User defined bands such as good, fair, poor

For each band, define
z construction defect indicator for bituminous
surfacing (CDS),
z construction defect indicator for roadbase
(CDB),
z relative compaction (COMP)
For unsealed roads, construction method is
either ‘mechanical’ or ‘manual’
CURRENCY
Defined in terms of:

z Name (e.g., Dollars)
z Symbol
y ((e.g.,
g $))
z Symbol position (i.e., before or after the
figures)
g )
z ‘Thousand’ separator (e.g., comma or point)
Unlimited number of currencies
OVERVIEW
International Study of
Highway Development and Management
z Sponsors:
z World Bank
z Dept for International Development
z Asian Development Bank
z SNRA & Others
z I progress for
In f 5 years comprising
i i
z technical studies
z software development and testing
z PIARC leading the implementation phase
Objective:
Economic basis for selecting investment alternatives

Road standards
Pavement standards
Alignments
Objective:
mproved framework for investigating road investments
Non-motorised
transport
Traffic congestion
Vehicle emissions
Travel times
Transport costs
Road accidents
HDM-4 Applications
z Road sector policy studies

z Strategic planning of road network development,
maintenance and funding requirements
z Optimal allocation of ffunds
nds to competing needs
z Formulation of national road sector funding
programmes
z Economic appraisal of individual road projects
z Research studies
Policy Formulation
Road pricing
p g
z fuel levies
z congestion charges
z weight distance charges
weight-distance
Vehicle regulations
z axle load limits
z energy consumption, vehicle emissions & noise
Engineering Standards
z sustainable road network size
z pavement design and maintenance standards
Analytical Framework
Predicts road network performance as a

function of:
z Traffic volumes and loading
z Road pavement strength
z Maintenance standards
Quantifies benefits to road users from:
z Savings in vehicle operating costs (VOC)
z Reduced road user travel times
z Decrease in number of accidents
z Environmental effects
What has Changed ?
z Types of Analysis & Outputs

z Road Deterioration Models
z Road User Effects Models
z Data
z Di
Dimensions
i
z Road works
z S i l and
Social d Environmental
E i t l Effects
Eff t Models
M d l
z Software
Types of Analysis &Outputs
z Strategy
gy
z Programme
z Project
Network Condition Forecasts
Effect of funding levels on network performance
oor 7.0 Feeder

Roads
$30 /
$30m/yr
6.0
ness
Secondary
Roads
Roughn
5.0 $35m/yr
4.0
0 Trunk
Roads
$20m/yr
Good 3.0
1998 1999 2000 2001 2002 2003 2004
Effect of budget levels
Annual
7.0 Budget
50%
6.0
hness
80%
50
5.0
Rough
100%
4.0
Target
3.0
1998 1999 2000 2001 2002 2003 2004
Optimal budget requirements
250
200
Development
150
Improvement
Periodic
100
Routine
50
0
2000 2001 2002 2003
Multi-year programme
nk Road Section km District Work Type Year $m Σ$m

N1-2 20.5 2 Resealing 2000 5.4 5.4
N4-7 23.5 7 Overlay 40mm 2000 10.9 16.3
N2-5 12.5 5 Reconstruct 2000 8.6 24.9
R312-1
R312 1 30.0 4 Widen 4 lane 2000 31.4 56.3
R458-3 36.2 3 Overlay 60mm 2000 16.3 72.6
: : : : : : :
N4-16 32.1 6 Reconstruct 2001 22.8 22.8
R13 23
R13-23 22 4
22.4 4 O
Overlay
l 40mm
40 2001 97
9.7 32 5
32.5
N521-5 45.2 2 Widen 4 lane 2001 41.3 73.8
: : : : : : :
N1-6 30.2 4 Resealing
g 2002 8.2 8.2
N7-9 17.8 3 Overlay 60mm 2002 9.2 17.4
F2140-8 56.1 1 Reconstruct 2002 34.9 52.3
: : : : : : :
Project Level Outputs
• Road condition indicators

• Road user cost details
• Energy
gy & emissions
• Economic indicators
• Sensitivity
y analysis
y results
• Scenario analysis
Road Deterioration Models
z Bituminous pavements, 8 generic types

z Defects,, including
g surface texture,, skid
resistance, plastic deformation,edge
break.
z Concrete (Rigid) pavements, 3 types
z Unsealed pavements, 3 types
z Wider range of calibration factors
Road User Effects Models
z New/improved VOC models to reflect

changes in vehicle technology
z Acceleration effects
z NMT vehicle operating costs
z Road safety (accident rates and costs)
z Work zone effects
DATA and DIMENSIONS
z Data hierarchy based on Information

Quality Level (IQL) concept
z Import and export of data
z Road sections - unlimited
z Motorised vehicle types - unlimited (based
on 16 basic types)
z NMT vehicle types - unlimited (4)
z Maintenance standards - unlimited
z Study alternatives - unlimited
ROAD WORKS
z Routine maintenance
z Drainage effects
z Periodic maintenance,, including g recycling.
y g
z Road improvements, including widening,
lane separation, diverted traffic.
z Intervention criteria, including road user
effect parameters
Integrating HDM-4 within
Road Management Systems
Long Term
Road
R d St t i
Strategic
Inventory Road Plan
DATABASE HDM-4
Road Multi -Year
Year
Condition Work
HDM4 Program
Traffic
Data
Detailed
Project Level
Bridges
Appraisal
Inventoryy
Bridges Policies
Condition Standards
Research
Summary
HDM 4 is based on a well established economic

HDM-4
analysis framework
Models derived from large scale field experiments
conducted world-wide
Significant
g change
g in scope
p and software technology
gy
from HDM III
International standard tool for road sector planning
p g
and management
y
Provides a common framework for analysis of road
IImplementing
l ti HDM
HDM-4
4 within
ithi
Road Management Systems
Implementation Stages
Data Decision Management

Database
Collection Support
pp Information
Inventory RDBMS HDM-4 Standard

Roughness (RIMS) & Custom
Visual Reports
Deflection
Traffic
Bridges
Integrating HDM-4 within Road Management Systems
Long Term
Road Strategic
Inventory
DATABASE HDM--4
HDM Road Plan
Road Multi -Year

Condition Work
HDM4 Program
Traffic
Data
Detailed
Project Level
Bridges
Appraisal
Inventory
Bridges Policies
P li i
Condition Standards
Research
INPUTS OUTPUTS
Data Collection
Road inventory
z Road
R d links,
li k class,
l pavementt types,
t
z Geometric characteristics (widths,
shoulders etc)
shoulders,
Pavement condition
z Roughness surveys (automated)
z Visual surveys (manual)
z Deflections (sampled once in 5 years)
Traffic data
Bridges
Vehicle Fleet
Unit Costs
Works Standards
HDM-4 Configuration Data
Model Calibration
Reports
L
Language
Linking HDM-4 to External Systems
Configuration Model Vehicle data, Standards,

D t
Data C ffi i t Alternatives,
Coefficients Alt ti etc.
t
Work
Inventory Programs
Projects
Surveys
DXF Special
Calibration R
Reports
t
factors Etc.
RDBMS HDM - 4
Vehicle Fleet Calibration
Representative
p Vehicle Types
yp
Basic Characteristics (Level 1)
z mass, loading, speed, vehicle life, utilisation
Unit Costs
Detailed Calibration (Level 2)
z Tyre wear, depreciation, fuel consumption
The above data are held in HDM-

HDM-4 database
Bangladesh Vehicle Fleet
Motorised vehicle types:

z Baby Taxi
z Mishuk
z Tempo
z Small Truck (converted jeep)
z Tata Truck
z Toyota Corolla
z Coach
Works Standards
Works types
yp
z maintenance
z improvement
Unit costs
Construction quality
q y
z material properties
q y
z compaction, etc.
Bangladesh Work Types
Works types:
z SBST
z DBST
z Thin Overlay
z Thick Overlay
z Rehabilitation
z Widening
Intervention levels
HDM-4 Configuration Data 1
Traffic Flow Patterns

Speed Flow Types
Climate Zones
Currencies
Aggregate Data
Above data is held in HDM-4 database

HDM-4 Configuration Data 2
Aggregate data - country specific:

z traffic volume
z road class
z geometry class
z surface texture
z surface condition
z structural adequacy
z ride quality
Other Activities
Reports
z Special report requirements
z Using user-defined reports facility
Language
z Is native language version required?
z User interface
z Reports
z Documentation, etc.
How do we link HDM-4 to external systems?

How do we develop p the data-exchange
g
interface?
And now for the ‘clever’ bit...

Data Flow Schematic

Work
W k
Inventory Programs
P j t
Projects
Survey
Special
Calibration Reports
Factors RDBMS HDM-4 Etc.

1.0
2.1 Segmentation
g look-up
p table))
2.3 Aggregation
30
3.0 W it section(s)
Write ti ( ) to
t exportt file
fil (dBase)
(dB )
Selecting Sections 1
Coping with Missing Data
Defaults Total HDM-4 RMS data

for missing
data items
= road data
requirement
compatible
with HDM-4
Data Interface Schematic
Selected Representative
RDBMS Section Look-up
Links Table
The End
Questions?
Institutional Impacts of Implementing HDM-4
Should we:
z adapt
d t HDM-4
HDM 4 tto suit
it local
l l methods*,
th d * or
z adapt local methods to suit HDM-4?
(* e.g. survey / data collection)

INTRODUCTION AND INSTALLATION
OF THE HDM-4 SOFTWARE
Objectives of this Session
z software basics
z installation
z user interface basics
z software structure
z i iti l configuration
initial fi ti
z getting help / support
z documentation
Software Basics - Software Description
z Microsoft Windows (95 / 98 / NT *)

z Multiple languages
z Distributed by CD / Internet
z E
Examplel case-studies
t di
z On-line documentation
Software Basics - Hardware Requirements
Minimum recommended requirements:

z Pentium P100 processor (or equivalent)
z 32MB of RAM
z 30MB of hard disk space (for program and
documentation)
z 50MB of hard disk space (for storage of
run data)
run-data)
z Windows 95/98 or NT 4 with Service Pack
6a installed
Software Basics - Licensing
z Owned by Sponsors / PIARC

z Distributors (McTrans / ENPC / PIARC)
z Licence types
z User-seat concept
z Educational licence
z Participant’s copy owned by organisation
Software Basics - Licence Types
Licence Type Price Discount SCC Discount

Price
Demonstration Free 100% Free 100%
Standard 1 200
1,200 0% 780 35%
Four-Pack 3,360 30% 2,184 55%
Ten-Pack 8,400 30% 5,460 55%
Education Std 480 60% 312 74%
Education 4-Pack 1,344 72% 874 82%
Education 10-Pack
10 Pack 3,360
3 360 72% 2 184
2,184 82%
All prices are in US Dollars (US$)

( )
Software Basics - Licensing
z Owned by Sponsors / PIARC

z Distributors (McTrans / ENPC / PIARC)
z Licence types
z User-seat concept
z Educational licence
z Participant’s copy owned by organisation
Software Basics - Licence Terms
"Educational
Educational Purposes
Purposes" means the use of ISOHDM
Controlled Products in formal vocational and
professional teaching courses (for example by
f
faculty
l and d students
d off universities
i i i and
d technical
h i l
training institutes), undergraduate and
postgraduate
p g project,
p j , thesis or research activities,,
and formal HDM-4 training activities. It does not
include consulting activities conducted by
professional researchers and educators
educators, which are
not primarily educational in nature.
Installation - Basics
z Release types:
z Version
z Update
z Upgrade
z Distribution types:
z CD
z Internet download
z Username / Software Licence Key (SLK)
Installation - Instructions
Installation instructions can be found:

z CD insert
z from Internet download page
z Software User Guide ((Volume 3 - Section 2))
z guidance during installation
Installation - Hands On
z Install version 1.0 from CD

z Install upgrade 1.1 from download file
Installation - Hands On
< installation practical >

Installation - Review of Key Points
z Autorun screen
z User registration details
z HDM-4 program group / menu
z Readme file
z Uninstalling
z Documentation
User Interface Basics - Starting HDM-4
z From HDM-4 menu

z Creating a desktop shortcut
z Welcome screen
z Workspace window
z See Volume 3 (section 3 & 12.4)
User Interface Basics - Hands On
< familiarisation session >

UI Basics - Review of Key Concepts
z Road Networks / Road Sections

z Vehicle Fleets / Vehicle Types
z Work Standards - Maintenance /
Improvements
z Studies - Projects,
Projects Programmes,
Programmes Strategies
z Configuration data
z Prompt strings
Software Structure - System Architecture
Data Managers A l i Tools

Analysis T l
Road Vehicle Road HDM Project Program Strategy

N t
Network
k Fl t
Fleet W k
Works C fi .
Config
File Converter Core Data Model libraries

Models
Vehicle Fleets can be used in
Road Networks other systems
transfer data Road Works
to external Projects
Programmes RDWE RUE SEE
systems
Strategies
External RDWE: Road Deterioration and Works Effects

Systems RUE: Road User Effects Road
Databases,
SEE: Safety, Energy and Environmental Effects
PMS, etc.
PMS: Pavement Management Systems
Software Structure - Data / File Handling
z IInputt data
d t
z Using multiple databases (§12.1)
z O t t / run data
Output d t
z Rundata ‘Clean-up’ facility
z Importing / exporting data
z Backing-up (§12.3)
z Reports (§10)
z Log-file
Initial Configuration
z Traffic Flow Patterns

z Speed Flow Types
z Climate Zones
z Currencies
z Aggregate Data
z Volume 3 - Section 11
z NB Needs to be defined for
NB.
each new implementation
Traffic Flow Pattern:

z stores Hourly flow-frequency distribution
d t
data
z one for each road use category
z defined in terms of the number of hours
per year that the traffic volume is at a
certain p
percentage
g of the AADT.
Speed Flow Type:

z defines speed flow characteristics
z one for each road type
z defined in terms of ultimate capacity, jam
speed free-flow
speed, free flow capacity,
capacity nominal
capacity, accident rates
User Support / Getting Help
z On-line Help
z Documentation
z Distributor
z User Groups / Networks
z Web-sites / Knowledge base
z Technical Support
pp (inc.
( UoB))
User Support - Web Sites Key Facilities
z Knowledge Base
z Frequently Asked Questions (FAQs)
z Technical Articles
z Implementation projects
z Discussion / News groups
z Links
User Support / Getting Help
z On-line Help
z Distributor
z User Groups / Networks
z Web-sites / Knowledge base
z Technical Support (inc. UoB)
z Documentation
Documentation - Basics
z Provided on CD - NOT printed

z may be installed or used from CD
z uses Adobe Acrobat (free-ware)
z PDF (Portable Document Format)
Documentation - Volumes
z 1 - Overview
z 2 - Applications Guide
z 3 - Software User Guide
z 4 - Analytical Framework & Model Descriptions
z 5 - A Guide to Calibration & Adaptation
z 6 - Modelling Road Deterioration & Works Effects
z 7 - Modelling Road User & Environmental Effects
Documentation - Hands-on
< hands-on session >

Documentation - Review of Key Points
z Accessed from HDM-4 menu

z Uses Adobe Acrobat Reader
z Navigation
z Hyperlinks
z Searching
z Assistance on Reader Help menu
Review of Session Objectives
z software basics
z installation
z user interface basics
z software structure
z i iti l configuration
initial fi ti
z getting help / support
z documentation
(* All of this is covered in documentation)

Introduction & Installation of HDM-4 Software
Any questions?
Improving Analysis Performance
z Close other applications

z improve specification of PC
z disable log file Effectt on
Eff
accuracy
z disable unnecessary models of analysis
z reduce run
run-data
data export size
z reduce number of flow periods
z reduce
d size
i off study
t d
HDM - 4 MODELLING FRAMEWORK
Linda Parsley
HDM-4 SYSTEM ARCHITECTURE

Analysis T l

N t
Network
k Fl t
Fleet W k
Works C fi .
Config

Models
systems
Strategies
RDWE : Road Deterioration and Works Effects

External
E t l
Systems RUE : Road User Effects
Databases, SEE : Safety, , Energy and Environmental Effects
PMS, etc. PMS : Pavement Management System
HDM-4 ANALYSES
z Strategy analysis
z For
o preparing
p epa g medium
ed u to long
o g term
te
planning estimates of funding needs for
road network development & maintenance
z Programme analysis
z For preparing single or multi-year work
programmes underd budget
b d t constraints
t i t
z Project analysis
z For evaluating
F l ti the
th economic i or
engineering viability of different road
investment projects
HDM-4 ANALYTICAL FRAMEWORK
z The analytical framework is based on the

concept of life cycle analysis
z This is applied to predict the following
over the life cycle of a road pavement,
which is typically 15 to 40 years
z Road
R d deterioration
d t i ti
z Road works effects
z R d user effects
Road ff t
z Socio-economic and environmental effects
TOTAL TRANSPORT COSTS
$ Total
Optimum
Road
User
Construction/
Rehabilitation
Maintenance
R d St
Road Standard
d d
ANNUAL EFFECTS & COSTS
z Input data
z Start loop
z calculate traffic
z estimate road deterioration
z estimate road user effects
z estimate road works effects
z Socio-economic and environmental effects
z exogenous benefits & costs
z Return loop
z Repeat for each analysis year
TRAFFIC ESTIMATES
z Volume AADT
z Composition and growth rates
z Categories
g - Normal,, Diverted,, Generated
z Axle loading
z Road capacity
p y and speed-flow
p relationship
p
z Traffic flow pattern
z Degree of conflict between MT and NMT
z Roadside friction
VEHICLES
z Categories:
z Motorised transport (MT)
( )
z Non-motorised transport (NMT)
z Cl
Classes
z For example: motorcycles, cars, buses,
trucks
z Types
VEHICLE DATA
z General
z Name, category, class, type
z Utilisation
z Performance
z Loading g characteristics
z Unit costs
z RUE
U Ca Calibration
b at o pa
parameters
a ete s
Vehicle fleet data manager

NETWORK ELEMENTS
z Sections
z a length of road that is homogeneous in
terms of its physical attributes (basic unit
off analysis)
l i )
z Network
z S b
Sub-networks
t k
z Nodes
z Links
ROAD SECTION DATA
z General
z ID, inventory, traffic, road use and class
z Geometry
z Pavement characteristics
z Road condition
z Calibration parameters
Road network data manager

HDM-4 PAVEMENT CLASSIFICATION
z Surface category
z Paved
z Unpaved
z Surface class
z bituminous
z concrete
z block
z unsealed
z Pavement type
ROAD WORKS
z MAINTENANCE / PRESERVATION
z Routine - undertaken each year
z Periodic - undertaken at intervals of years
z Special - emergencies, winter maintenance
z DEVELOPMENT
z improvement - widening, realignment
z new construction
WORKS STANDARDS
z Select an operation (e.g., overlay)

z Define Intervention Criteria
z Scheduled or responsive, include limits:
time, roughness, traffic
z Specify design characteristics
z pavement, geometry, road type and class
z Define works effects
z pavement strength, condition, history, road
use, etc.
z D
Duration
ti and
d Costs
C t
ROAD USER EFFECTS MODELS
z Vehicle free speeds

z Speed-flow and acceleration effects
z Vehicle operating
p g costs
z Travel time costs
z Passengers
g and cargo
g
z Accident costs
z Environmental
o e ta e effects
ects
z Emissions and noise
z Energy
gy
EXOGENOUS BENEFITS AND COSTS
z Due to other road related investments

z diversion routes, bridge repair, slope
stabilisation, etc.
z Development benefits
z agricultural, commercial, industrial, etc.
z Accessibility benefits
z Environmental impacts (pollution,
ecological)
z Social & resettlement impacts
SECTION ALTERNATIVES
z A section alternative is the basic unit of

HDM-4 analysis
z It is one of mutually exclusive works
alternatives specified for a road section
z A section alternative is defined by:
z A description
z A set of maintenance and/or improvement
standards,
t d d combinedbi d logically
l i ll
z Associated generated traffic
z Associated exogenous benefits and costs
SECTION ALTERNATIVE
Section A Alternative 2

Model libraries
can be used in
other systems
ANALYSIS BY SECTIONS
Road Section alternatives

sections 1 2 3 4 5
Section A RM Resealing Overlay

Section B RM Overlay Reconstruction Widening
Section C RM Resealing Rehabilitation Lane Realignment
addition
Section D Grading Regravelling Paving
ANALYSIS BY PROJECTS
Road Project alternatives

sections
1 2 3 4
Section A RM Resealing Reconstruction Realignment

Section B RM Overlay Mill & replace Reconstruction
Section C RM Inlay Reconstruction Lane addition
Section D Grading 1 Grading 2 Regravelling Upgrading
ΣProject NPV 0 3.04 4.72 3.85
IMPACT ANALYSIS LOOPS
z Define input data

z Loop A for each option
z Loop B for each road section
z Loop C for each analysis year
z Calculate traffic over the road section
z Model annual effects and costs
z Store result for evaluation and reporting

phase
z End loop C
z End loop B
z End loop A
COMPARISON OF ALTERNATIVES
z Loop A for each pair of investment alternatives

z Loop B for each analysis year
z Loop C for each road section
z Calculate net monetary benefits
z Calculate net emissions and energy use
z End loop C
z Calculate
C l l t nett benefits
b fit over all
ll the
th road
d sections
ti
z End loop B
z Calc late economic indicators (NPV,
Calculate (NPV IRR,
IRR etc
etc.))
z End loop A
z P f
Perform optimisation
ti i ti (Programme
(P & Strategy)
St t )
PROJECT LEVEL OUTPUT
z Economic indicators
z NPV, IRR, FYRR, BCR,..
z Sensitivity analysis results
z Scenario analysis
z Road condition indicators
z Road user cost details
z Energy
e gy & e emissions
ss o s
PROGRAMME ANALYSIS
z Identify candidate road sections

z Determine alternative improvements
z Specify
p y budget
g limits & periods
p
z Optimise using selected objective
z Produce optimal
p list of projects
p j for budget
g
period
PROGRAMME ANALYSIS …
z Method
z Life Cycle Analysis
z Budget Period Analysis
z Objective functions
z maximise NPV
z maximise road network condition
improvement
z Constraints
z budget limits (capital, recurrent, etc.)
PRIORITISATION
z Under budget constraints:

z Independent projects by NPV/Cost
z Mutually exclusive projects by incremental
NPV/Cost
z When there is no budget constraint:
z Independent projects by NPV or IRR
z Mutually exclusive projects by maximum
NPV
STRATEGY ANALYSIS PROCEDURE
z Categorise road network into matrix

z Traffic: High, Medium, Low
z Pavements: Concrete, Bituminous, Gravel
z Condition: Good, Fair, Poor
z Model pavement deterioration
z Apply maintenance standards
z Calculate road user costs
z Estimate long term network funding
requirements and performance trends
HDM-4
HDM 4 ROAD NETWORKS &
VEHICLE FLEETS
Road Network - Key Skills / Points
Creating a Road Network

Creating a Road Section
Concept
p of homogeneous
g section
Use of aggregate data
Categories of road section data
z ALL road section data is mandatory
Re use of existing sections
Re-use
Use of representative section set
Importing road section data
Road Section Data Categories
Road
R d Section
S ti
Data
Definition Geometry Pavement Condition
Calibration
Road Network - Documentation
4 Road Networks
4.1 Keyy concepts
p 7
45
4 12
4.13 Specifying which Sections are displayed in the 17
Road Network window
4.14 Editing Section details directly from the Road 18
N
Networkk window
i d
Vehicle Fleet - Key Skills / Points
Creating a Vehicle Fleet

Creating a Vehicle Type
HDM-4 default types
Concept of representative types
Categories
g of vehicle data
ALL vehicle data is mandatory
All in volume 3 section 5
Road & Vehicle data usually entered by
different users
Vehicle Fleet Data Categories
Vehicle
V hi l Type
T
Data
Definition Basic Economic Financial

Data Unit Costs Unit Costs
Calibration
Vehicle Fleet - Documentation
5 Vehicle Fleets
5.1 Key Concepts 20
5 10
5.10 Adding motorised Vehicle Types (MT) to a 27
Vehicle Fleet
5.11 Adding non-motorised Vehicle Types (NMT) 28
to a Vehicle Fleet
5.12
5 12 Editi Vehicle
Editing V hi l Type
T attributes
tt ib t 29
5.13 Specifying which Vehicle Types are displayed in 32
the Vehicle Fleet window
5.14 Editing Vehicle Type attributes directly from the
Vehicle Fleet window 33
PRACTICAL
Procedure:
Create a Road Network
z Create several road sections and enter data
Create a Vehicle Fleet
z Create several vehicle types and enter data
Create a sample project
z Enter required data and run HDM-4
Review typical outputs
Programme Analysis
Programme Analysis
Programme Analysis is concerned with the preparation

of single or multi-year road work and expenditure programmes
under specified budget constraints.
z Objective: prioritise candidate road projects

in each year within annual budget constraint
z Annual budgets obtained from strategic
maintenance plan
Multi-Year work programme
z Method: use specified standards to screen

network & identify candidate projects, e.g.
z road sections which exceed specified
condition
z roads with inadequate
q capacity
p y
z pavements which need strengthening
z upgrade pavements with high traffic volumes
Procedure
z Select candidate road projects

z Determine maintenance or improvement
options
z Specify budget limits & periods
z Optimise using selected objective
z Produce optimal list of projects for budget
period
Multi-year Programme ...
z Method
z Life Cycle Analysis
z Budget Period Analysis
z Objective functions
z maximise NPV
z maximise road network condition
improvement
z C
Constraints
t i t
z budget limits (capital, recurrent, etc.)
Analysis Procedure
Life Cycle Total

Analysis Enumeration
NPV Minimise Cost

Inputs Results
IRI
Budget Period Incremental

Analysis A l i
Analysis
Capital Budgeting
Independent Projects:
z unrelated road projects e.g. from different
parts of the country,
p y,
Mutually Exclusive Projects:
z alternatives to one project where only one
can be selected.
Capital Budgeting Rules
When sufficient funds are available

z Select all independent projects with NPV > 0
z Select mutually exclusive project alternatives with
the highest NPV
When there is a shortage of funds;
z Select independent projects with the highest
NPV/Cost ratio
z Select mutually exclusive projects using the
incremental NPV/Cost method
method.
Prioritisation
A 150,000 35,000 0.233 2
B 150,000 45,000 0.300 1
C 200,000 40,000 0.200 3
D 250,000 37,500 0.150 6
E 500,000 70,000 0.140 7
F 500,000 77,000 0.154 5
G 500,000
, 78,000
, 0.156 4
A 20,000 4,869 0.2435 2
B 30,000 240 0.0080 8
C 20,000 5,184 0.2592 1
D 20,000 210 0.0105 7
E 40,000 6,401 0.1600 4
F 35,000 4,497 0.1285 5
G 20 000
20,000 1 200
1,200 0 0600
0.0600 6
H1 80,000 15,000 0.1875 3
H2 100 000
100,000 16 829
16,829 0 1683
0.1683
Efficiency Frontier
∆ BENEFITS
Opt
p 4
Opt 3 O t5
Opt
Opt 2
Opt 1
Opt 0
∆COST
Efficiency Frontier
BENEFITS
A4 A5
A3
A2
B3 B4
Priority by
B1 slope:
B2 1. B1
2 A1
2.
3. A2
A1 4. A3
A, B - Independent Projects 5. B3
1,2,3,.. - Mutually Exclusive 6. A4
7. A5
PROJECT COST
Budget Period Analysis
z Life Cycle analysis involves long term

predictions
z Assumes budgets are committed long
term
z Total benefits assume future periodic
maintenance
i t will
ill be
b done
d
z Budget period analysis quantifies relative
b
benefits
fit within
ithi short
h t term
t periods
i d
Budget Period Analysis
BUDGET PERIOD ANALYSIS
Budget Period
IRI Estimated Benefits
TIME
Multi-Year Work Programme
Road
R dNNo. (km))
(k Work
W k Type
T Year
Y ($m))
($ Σ($m))
Σ($
A10031 12.5 Reconstruct 2001 5.4 5.4
A10004 23.4 Resurface 2001 2.9 8.3
B30106 19 8
19.8 Reconstruct 2001 86
8.6 16 9
16.9
A20122 5.5 Patch&Reseal 2001 1.4 18.3
: : : : : :
A11028 34.5 Reconstruct 2002 13.8 13.8
C34567 22.6 Reconstruct 2002 9.7 23.5
: : : : : :
B32109 15.0 Resurface 2003 4.2 4.2
C76543 36.9 Overlay 2003 24.1 28.3
: : : : : :
Strategy Analysis
Strategy Analysis
Strategy analysis is concerned with the analysis of

entire road networks to determine funding needs and/or
to predict future performance under budget constraints
Objectives:
z Determine budget allocations for road
maintenance and improvement
z Prepare for work programmes
z Determine long term network performance
z Assess impact on road users
Procedure
z Categorise road network into matrix cells
z D fi representative
Define t ti traffic
t ffi volume
l and
d loading
l di
z Define maintenance and improvement standards
z Specify budget constraints
z Model pavement deterioration for each matrix
cell
ll
z Apply maintenance and improvement standards
z Calculate road user benefits
z Select maintenance standards which optimise
user benefits
b fit
Road Network Matrix
z Categorise road network into matrix

z Road Class: Primary, Secondary, Tertiary
z Pavements: Rigid, Paved, Gravel
z Relative strength: Good, Fair, Poor
z Condition: Good, Fair, Poor
z Traffic: High, Medium, Low
z Determine road length in each matrix cell
z Record regional distribution
Representative traffic
z Primary roads: Flexible pavements

z High: > 10,000 vpd; (1.0m ESA/lane/year)
z Med.: 5,000 ~ 9,999; (0.4m ESA/lane/year)
z Low: < 5,000; (0.1m ESA/lane/year)
z Primary roads: Gravel pavements
z High: > 1,000
z Med.: 250 ~ 999
z Low: < 250
2 0
Road Network Matrix
Length of Roads (km)

Flexible
Fl ibl Pavement
P t Gravell Pavement
G P t
.
Cond
Road
Traffic Traffic
Class
High Med. Low High Med. Low
G 76 283 307 686 452 703
National F 485 762 561 855 904 1961
P 911 420 321 1905 2432 4321
G
Regional F
P
G
Feeder F
P
Bituminous Pavements
z Routine Maintenance standards

z Patch all potholes
z Edge repairs
z Fixed costs for drainage, safety, etc.
z Periodic Maintenance standards
z Overlay at Roughness > 5.5 IRI
z Resurface at Surface damage > 30%
z Improvement standards
z Widen when V/C ratio > 1.2
Gravel Roads
z Routine Maintenance standards

z Spot regravelling (gravel patching)
z Fixed costs for drainage, safety, etc.
z Periodic Maintenance standards
z Grading scheduled 2/year
z Regravel when gravel loss > 100mm
z Improvement standards
z Upgrade to flexible pavement at ADT > 250
Budget Constraints & Optimisation
z Recurrent budget
z Routine maintenance
z Top sliced from specified budget
z Capital budget
z New construction, upgrading, rehabilitation, etc.
z Multi-year limit
z Optimisation methods
z Maximise NPV
z Maximise network condition improvement
z Minimise cost for target network condition
Pavement Deterioration and Maintenance
Condition
Reconstruct
Overlay
Traffic / Time
Maintenance and improvement costs
z Routine and recurrent maintenance costs

z Budget MUST be sufficient otherwise re-
specify
z Periodic maintenance costs
z For specified alternatives
z Improvement costs
z For specified alternatives
Calculate road user benefits
z vehicle operating cost savings

z fuel, parts, tyres, depreciation, etc
z travel time savings
z reduction in accidents
z pavement preservation
p p
Determine optimum standards
z Select optimum maintenance standards

for each matrix cell under specified
budget constraints
z Calculate total routine, recurrent and
periodic maintenance costs for each cell
z Summarise total maintenance
q
requirements for whole network
Total network needs
Maintenance Costs ($m)
Flexible Pavement Gravel Pavement
.
Cond
Road Traffic Traffic
Class High
g Med. Low High
g Med. Low
G 6 3 0.7 1.6 0.4 0.3
National F 15 13 4 9 4.4 1.1
P 31 22 11 19.6 14.2 9.5
G
Regional F
P
G
Feeder F
P
Budget Summary
Maintenance Budget ($m)

Primary Roads Secondary Roads
Region
Flexible Gravel Flexible Gravel
North 45.9 18.7 34.5 21.8
East 34.6 13.5 24.9 11.8
South 52.1 21.0 45.0 21.8
W t
West 27 2
27.2 20 4
20.4 19 5
19.5 12 3
12.3
Budget Allocation
z Routine maintenance needs based on

road lengths under each category of roads
z Recurrent maintenance needs based on
pavement surface condition
z Periodic maintenance needs based on
economic indicators
z Improvement needs based on economic
indicators
UNIT TITLE: HDM-4 MODELLING FRAMEWORK
UNIT AIMS
1. To introduce the modelling framework used in HDM-4 so that the HDM-4 applications can be described
accurately by trainers.
UNIT OBJECTIVES
1. State the main applications of HDM-4 and their relevant functions in road management.
2. Explain the analytical framework in terms of the modelling units, e.g. road deterioration, road work effects,
road user effects, etc.
3. Describe the concept of homogeneous sections, road links, sub-networks and networks.
4. Describe the pavement and vehicle classification system (i.e. pavement and vehicle types).
5. Describe the hierarchy embodied in the analysis loop.
6. Illustrate the key concept of section alternatives through a flow diagram.
GUIDELINES
It is important that the modelling framework is described to Trainees in such a way that they appreciate the
complexity of trying to model the future performance of roads and the impact on road users. HDM-4 essentially
comprises a set of mathematical models that have been put together in a coherent analytical framework in order
to provide useful road management information. The mathematical models are for predicting road deterioration
and the effects of road works, for calculating the effects of road characteristics on road user costs, and for
calculating other socio-economic effects (energy, emissions, noise, etc).
The HDM-4 models are applied to homogeneous road sections. This assumes that pavement deterioration will
be the same on the entire road section. Consequently, road user effects will be constant for that section.
Characteristics of the road section are defined in terms of the pavement type which is user defined and the type
of vehicles using the road. The modelling framework applies the concept of life cycle analysis to calculate over
a number of years (typically, 15 ~ 40 years) the road deterioration, works effects and road user effects for each
section-alternative.
The above concept is the key to understanding HDM-4. The applications (Project, Programme and Strategy) are
simply multiples of the analytical framework, i.e. a project analysis comprises one of more homogeneous
sections-alternatives, programme analysis comprises several projects and strategy analysis is simply a model of
the entire road network broken down into representative sections.
MATERIALS
2. Handout (T3 Modelling Framework Handout Notes.DOC)
REFERENCES
1. HDM-4 Volume 4: Analytical Framework and Model Description
UNIT TITLE: HDM-4 PROGRAMME ANALYSIS CASE STUDY
UNIT AIMS
1. To introduce the HDM-4 Programme Analysis module concepts so that Trainers and Trainees are able to
prepare annual or multi-year work programmes.
2. To introduce the concept of incremental benefit/cost ranking method so that Trainers appreciate optmisation
techniques.
3. To familiarise Trainers with the use of HDM-4 data input and output modules.
UNIT OBJECTIVES
1. To describe the concepts of the HDM-4 Programme analysis application.
2. To apply the incremental benefit cost optimisation techniques through simple manual calculation.
3. To use the HDM-4 model to generate a work programme.
4. To alter maintenance standards and unit costs and study the impact on budget requirements.
GUIDELINES
KEY POINTS
REFERENCES
UNIT TITLE: DERIVATION OF OPTIMUM MAINTENANCE STANDARDS
UNIT AIMS
1. To describe the specification and assignment of HDM-4 maintenance standards so that Trainers are able to
apply a range of maintenance works.
UNIT OBJECTIVES
1. To prepare input data required for investigating optimum road maintenance standards for paved roads using
HDM-4.
2. To assign a range of maintenance standards to road sections and run HDM-4.
3. To interpret the results and recommend optimum maintenance standards for paved roads.
GUIDELINES
This case study presents the economic analysis of alternative maintenance standards for two paved road sections
that are 36.5 km and 10.6 km long, respectively. The existing roads have surface dressing and asphalt concrete
pavement surfacing in good condition, and carry approximately 6 000 and 16 000 vehicles per day (AADT),
respectively. The objective of this case study is to determine the optimum maintenance intervals for resealing
(surface dressing) and for overlays on the two roads, respectively.
The case study data is provided in the T3 Case studies workspace/database installed with the HDM-4 software.
The data for this case study are located in the Projects folder in the Case Studies Workspace. The name of
the case study is Case 2: Optimum maintenance intervals. To locate the data for this case study:
The case study data includes the road network data set (CASE2 Project Sections) and the vehicle fleet (CASE2
Vehicle Fleet) required for the analyses.
The first step in the procedure would be to calibrate the road sections using the calibration factors derived from
the previous case study. The second step in the procedure should be to create additional maintenance
standards for the two roads. The data provided includes one example each for:
Create additional maintenance standards by copying the above to create separate standards for:
KEY POINTS
The optimum economic maintenance standards are defined as the maintenance interval (scheduled or
responsive) with the maximum NPV. This can be illustrated by exporting the Benefit Cost Ratios report to an
Excel spreadsheet and plotting the NPV against the maintenance interval (years) or intervention criterion (IRI).
Another useful plot would be of the NPV against the increase in agency cost. The resulting curve from this is
defined as the efficiency frontier. This will be described in detail in the Programme Analysis case study.
Those Trainers who complete the above exercise within a short time could attempt to derive the optimum
grading and regravelling frequencies for the Gravel road that is included in the Road Network database.
MATERIALS
REFERENCES
3. HDM-4 Volume 5: A Guide to Calibration and Adaptation (Chapters 6 and 7)
T3 Project Case 2 Unit Notes.doc.doc Page - 1

UNIT TITLE: THE ROLE OF HDM-4 IN ROAD MANAGEMENT
UNIT AIM
1. To introduce the concept of HDM-4 road investment appraisal and its applications in road management so
that the HDM-4 system can be properly applied within road management.
UNIT OBJECTIVES
1. Describe the concept of life cycle analysis that is used in road investment appraisal.
2. Describe the HDM-4 applications (Strategy, Programme and Project Analysis).
3. Explain the role of the HDM-4 applications within the road management cycle.
4. Illustrate the HDM-4 applications by using sample output.
GUIDELINES
This is a key unit that introduces the HDM-4 life cycle analysis concept and its applications within road
management. Life cycle analysis involves modelling the future performance of roads together with the impact of
road characteristics on user costs. Various titles are often used to describe the same concept; whole life cycle,
total transport costs, road investment appraisal, etc. It is important that trainers should stress that these all refer
to a common modelling framework.
The road management process is covered in another Unit. Trainers should refer to that unit in order to be able to
link HDM-4 applications with the management functions of planning, programming and preparation. The
operations management function is not explicitly covered by HDM-4.
With regard to the HDM-4 applications (i.e. use of HDM-4 for specific purposes), the three applications are;
Strategy, Programme and Project analyses. The three applications differ in terms of scope and data detail.
Strategy analysis deals with an entire road network with input data that can be at an aggregate level (e.g.
good/fair/poor pavement condition, etc.). The objective is to model the performance of the road network in
response to different budget scenarios or maintenance standards. Programme analysis deals with a selection of
road sections (i.e. a long list of candidate projects). The objective is to prioritise these and select short list of
projects that maximise benefits under specified budget constraints. Project analysis deals with individual road
sections and models mutually exclusive alternatives in order to estimate the economic benefits that would accrue
from each alternative. The three applications can also be used to conduct various research and policy studies, for
example to verify pavement design standards, to model impacts of changing axle load limit regulations, etc.
MATERIALS
2. Handout notes (T3 Role of HDM-4 Handout.DOC)
REFERENCES
1. UNIT xx: Road management cycle (Tyrone Tool)

2. HDM-4 Volume 1: Overview
UNIT TITLE: HDM-4 SOFTWARE STRUCTURE AND OPERATION
UNIT AIMS
1. To introduce the HDM-4 software components and the installation so that Trainees are able to correctly
install and use HDM-4 for the first time.
UNIT OBJECTIVES
1. Describe the HDM-4 installation procedure
2. Describe the main components of the HDM-4 software (i.e. Road networks, Vehicle fleet, Road work
standards, Projects, Programmes, Strategies, and Configuration).
3. Demonstrate use of the HDM-4 software by running one of the pre-defined case studies.
GUIDELINES
It is important that Trainees are able to properly install and operate the HDM-4 software. They should also
appreciate the licensing conditions under which the HDM-4 software is supplied.
Software installation is achieved through a wizard that guides users through the installation process. There are
several choices that the user must select, for example the installation language, the default HDM-4 software
language, the HDM-4 components to be installed (e.g. Reports, Documentation, etc.). The wizard provides the
opportunity to step back through the process and ensure that the correct choices have been made.
After successful installation, the Trainees should be taken through a guided demonstration of the HDM-4
software. This should aim to illustrate the procedure for creating and running one of the Project Cast Studies
(e.g. Case Study 2a) by following the steps below:
• Review the configuration data (climate zones, speed-flow types, traffic flow patterns, default data)
• Select a Road Network and review the data for one or two selected road sections
• Select a vehicle fleet and review the data for one or two selected vehicle types
• Open the Road Works standards folder and review the maintenance standards used in one of the Project
Case Studies
• Open the Project folder and select one Case Study.
• Review the Case Study data and finally run HDM-4 and review the output Reports.
MATERIALS
2. PowerPoint Slides (Software Installation and Operation.PPT)
REFERENCES
Country Action Plan
For
Implementation of HDM-4 in Highway Planning and Management in
Bangladesh
Aim: Implementation of HDM-4 for Road Maintenance Management System in Bangladesh

so that efficient and effective road infrastructure maintenance management can be achieved.
Objectives:
1. The road management authorities will use HDM-4 as a decision making tool.
2. Prepare optimized maintenance plan within budget constraint
3. Develop maintenance strategies and work standards.
Action Plan:
Action Plan Items Constraints Date of

Implementation
1. Report to the country Not applicable 30-06-00
champion about the training
held at IKRAM
2. Implementation of HDM-4 in Lack of Motivation, lack of technical

highway planning and capabilities.
management.
a. Ensure data collection and data Lack of coordination, motivation, financial 30-06-00
validation. constraints.
b. Creation of HDM interface Time constraints, lack of technical 15-07-00

between database and HDM-4 capabilities.
c. Calibration of user cost model 30-06-00

of HDM-4
d. Calibration of deterioration Lack of technical capabilities, Resource 15-07-00

model. constraints .
e. Development of work Lack of consensus amongst concerned 15-07-00

standards. people.
f. Development of Pavement Lack of consensus amongst concerned 15-07-00

Intervention Level (PIL) people.
matrix.
g. Finalisation of Unit costs of 15-07-00

different items.
h. Strategy analysis for long term Time constraints 15-07-00

budget planning.
Action Plan Items Constraints Date of
Implementation
k. Finalisation of works Lack of consensus amongst concerned 15-07-00
standards under budget people.
constraints.
l. Program analysis to prepare Time constraints. 20-07-00

the yearly maintenance work
program.
m. Project analysis for the Lack of qualified HDM personnel.

selected candidate roads.
3. In-country training and Lack of motivation, lack of knowledge 31-12-00

dissemination of HDM-4 for about decision support system in highway
sustainability. planning, lack of support from policy
makers, lack of resource
In-country training and dissemination of HDM-4

In
Bangladesh
Title: Training and dissemination of HDM-4 in Bangladesh.
Course Target Group Profile:
Group-A: Policy makers ( 15 persons – 1 session)

Group-B: Engineers /Economists (20 X 5 = 100 persons – 5 sessions)
Group-C: Academics, Transport Operators, Officials of different Highway Authorities,
Police and Highway Consultants (20 X 2 = 40 persons – 2 sessions).
NB:- Tentative cost of training = 155 persons @ Tk 100,000 per person = Tk. 15,500,000
Aim of the Course:
Dissemination of highway maintenance and planning using HDM-4 so that the

participants may consider its use in their planning and policy making work.
Objectives:
Group A: The participants will describe how HDM-4 could enhance their decision and
policy-making activities in the road infrastructure and development process.
Group B: The participants will explain and use HDM-4 efficiently in highway planning and
maintenance.
Group C: The participants will describe the role of HDM-4 in determine road user effects.
Course Contents:
See Table B1, B2 and B3.
TABLE C1: BANGLADESH TEAM
Group-A: Policy Makers
Program topics and timing for policy makers
Topic and Subtopics Objectives - The Teaching / Learning Hours Hours

trainee will: Approaches Theory Practical
1. Introduction to Describe the Description with
HDM-4, Evaluation importance of HDM-4 diagram, handout for 1 hr.
of Software and Use in highway planning participants, discussion.
in Bangladesh. and management.
1.1 Basics of Economic Describe the basic Description with 2 hrs. 1 hr.
Analysis and economic criteria and diagram / flowchart
interpretation of interpret HDM outputs support, handout for
HDM output participants, discussion
and demonstration of
HDM run.
TABLE C2: BANGLADESH TEAM

Group-B: Engineers / Economists
Program topics and timing for Engineers / Economists
Topic and Subtopics Objectives - The Teaching / Learning Hours Hours

trainee will: Approaches Theory Practical
1. Introduction to Describe the Description with
HDM-4, Evaluation importance of HDM-4 diagram, handout for 1 hr.
of Software and Use in highway planning participants,
in Bangladesh. and management. discussion.
1.1 Installation of Install HDM-4 Demonstration of 0.5 hrs.

Software. software . software installation
1.2 Introduction to Describe the main Demonstration of 0.5 hrs. 0.5 hrs
HDM-4 Work space components in HDM-4 workspace folders and
workspace. describe the
functionality
2. Introduction to Describe the network Demonstrate the 1 hr.

Network, Creating and explain items network file and
Network included there. explain the items of
different tabs.
2.1 Practical Session: Will create a network Demonstrate creating 1.0 hr
Creating Network from scratch. network.
3. Introduction to Describe the vehicle Description with 0.5 hrs.

Traffic pattern in types and vehicle diagram / flowchart
Bangladesh. characteristics support, handout for
participants
3.1 Modeling Vehicle Explain and describe Description with 0.5 hrs.
Speeds Including the speed model and diagram / flowchart
Congestion. the effects of support, handout for
congestion. participants
3.2 Impact of Road Explain the effect of Description with 1.0 hrs.
Geometry, Road road geometry on diagram / flowchart
Side Use and traffic. support, handout for
Surface Condition. participants
3.3 Vehicle loading, Describe different Description with 0.5 hrs.

determining traffic aspect of vehicle diagram / flowchart
volumes and growth loading, traffic volume support, handout for
rates. and growth rate. participants
4. Project Analysis, Describe the necessity Demonstration of 0.5 hrs.

Worked Example. of project analysis. worked example
4.1 Worked Example Explain and interpret Practice session 0.5 hrs.
without Economic project analysis
Analysis. reports.
4.2 Worked Example Explain economic Practice session. 0.5 hrs.

with Economic returns and interpret
Analysis. reports.
5. Project Analysis, Create a project and Demonstration and 1.0 hr.

Creating and run the project . practice session.
running new project
for paved road and
interpretation of
results.
6. Road deterioration, Describe the modes of Description with 1.0 hr.
Bituminous deterioration, effects of diagram / flowchart
Pavement. distress in VOC. support, handout for
- Modeling Approach participants, case
- Distress Types. study
- Deterioration
Relationship.
- Key Parameters.
- Introduction to
Calibration Parameters.
6.1 Case Study Practice and describe Practice session with 0.5 hr. 1.0 hr.
Examples. the effects of road worked examples
deterioration in HDM
results
1.00 hr.
7. Economic Analysis Describe different Description with
and Road user economic criteria and diagram / flowchart
Costs. parameters of road support, handout for
user costs. participants, case
7.1 Concept of NPV, study, Demonstration,
IRR, B/C ratio. discussion.
7.2 Capital Budgeting.
7.3 Components of
Road User Costs.
Draft Activity Action Plan
Country Training - Pakistan
Time
No of
S.No. Activity Period Action By * Jul Aug Sep Oct Nov Dec Jan Remarks
Participants
(Days)
1 Submission of HDM T3 Course 15 - CG
2 Review meeting for Country Training Prog. 15 - CG-CC-TRL In consultation with TRL
3 Preparation of Final Country Plan 30 - CG-CC-TRL In consultation with TRL
4 Approval of Final Country Plan 15 - CC-LA
5 Coordination and Preparation of Training Material 45 - CG
6 Dove Tailing country pHs with HMM 4 15 - CG-CC-TRL In consultation with TRL
6.1 Calibration of HMM 4 to local Conditions 15 -
6.2 VOICE and Program Analysis of Network 60 - CG-TRL In consultation with TRL
7 Necessary arrangement of resources 45 - CG-CC-TRL Material and Financial
8 Implementation Consensus & Approval 45 - CG-CC-TRL In consultation with TRL
9 Implementation of Training Pro 15 -
9.1 One Day seminar on HMM 4 at Islamabad (for Policy and Decision Makers) 1 30 CG-CC-TRL In consultation with TRL
9.2 Three days training course for Professionals (Engineers & Economists) at Islamabad 3 12 CG-CC-TRL In consultation with TRL
9.3 Three days training course for Professionals (Engineers & Economists) at 4 Provinces 15 112
9.3 One day training course for Students in NED-Krc, UET-Lhr, UET Taxila, UET Pesh 10 15 CG
9.4 Five Days Training course for Technicians at Islamabad 7 12 CG
9.5 Three Days seminar on Unsealed Roads Management by HDM 4 5 15 TRL-CG In consultation with TRL
10 Subsequent Implementation and Training Prog Onwards - CG-CC-TRL In consultation with TRL
Note *
CG : Country Group ( 1.Waqar, 2.Azeem, 3.Saqlain, 4.Atiq)
CC : Country Champion ( Raja Nowsherwan Sultan)
LA : Local Authorities
Draft Activity Action Plan
for
HDM--4 Training by Trainers
HDM
Country Training - Pakistan
Activity Programme for Pakistan
Time
No of
S.No. Activity Period Action By * Jul Aug Sep Oct Nov Dec Jan Remarks
Participants
(Days)
1 Submission of HDM T3 Course 15 - CG
2 Review meeting for Country Training Prog. 15 - CG-CC-TRL In consultation with TRL
3 Preparation of Final Country Plan 30 - CG-CC-TRL In consultation with TRL
4 Approval of Final Country Plan 15 - CC-LA
5 Coordination and Preparation of Training Material 45 - CG
6 Dove Tailing country pHs with HMM 4 15 - CG-CC-TRL In consultation with TRL
6.1 Calibration of HMM 4 to local Conditions 15 -
6.2 VOICE and Program Analysis of Network 60 - CG-TRL In consultation with TRL
7 Necessary arrangement of resources 45 - CG-CC-TRL Material and Financial
8 Implementation Consensus & Approval 45 - CG-CC-TRL In consultation with TRL
9 Implementation of Training Pro 15 -
9.1 One Day seminar on HMM 4 at Islamabad (for Policy and Decision Makers) 1 30 CG-CC-TRL In consultation with TRL
9.2 Three days training course for Professionals (Engineers & Economists) at Islamabad 3 12 CG-CC-TRL In consultation with TRL
9.3 Three days training course for Professionals (Engineers & Economists) at 4 Provinc 15 112
9.3 One day training course for Students in NED-Krc, UET-Lhr, UET Taxila, UET Pesh 10 15 CG
9.4 Five Days Training course for Technicians at Islamabad 7 12 CG
9.5 Three Days seminar on Unsealed Roads Management by HDM 4 5 15 TRL-CG In consultation with TRL
10 Subsequent Implementation and Training Prog Onwards - CG-CC-TRL In consultation with TRL
Note *
CG : Country Group ( 1.Waqar, 2.Azeem, 3.Saqlain, 4.Atiq)
CC : Country Champion ( Raja Nowsherwan Sultan)
LA : Local Authorities
An Overview of Activity Plan
• Role of Country Champion

• Role of Trainers
• Role of TRL & University of Birmingham
Role of Country Champion
• Evaluation of
HDM4 Training & Dissemination
HDM4 Training and Dissemination

Country Action Plan: India
Preliminary Assessment
1.1 Preliminaries
The Trainers will, within a period of ten days after the arrival back to their
respective places of work, submit to the Country Champion a report
containing the essential details of the knowledge and experience gained
on HDM4 Capabilities through the “HDM4: Train the Trainers Course” held
in Kuala Lumpur from 5 June to 23 June 2000. The Trainers would make a
presentation on the major aspects of the model and their relevance in
regard to the strategic planning, program analysis, project analysis,
execution including monitoring and evaluation.
The report will cover the brief details on the adaptation requirements for
the default Road Deterioration and Works Effect (RDWE) and VOC
relationships. The following aspects would also be brought out in the
report:
1) The identification of sources for the availability of pavement performance

data, for earlier studies on existing pavements and new pavements.
2) Requirements for taking up performance studies for newly constructed heavy-
duty pavements for road projects by the MoST, NHAI and State PWDs.
3) Collection and compilation of relevant parameters for VOC.
4) Requirements for Speed-Flow studies for 4-Lane/multi-lane highways.
The Trainers will have to discuss with the Country Champions in regard to
the "Trainers Training at Malayasia","Review Of The Existing Management
System In India" and complementing HDM4 software with the system.
The other relevant essential aspects concerning the “Country Action Plan”
are briefly described in the paragraphs that follow.
1.2 Locations of Institutions for Developing the Training
Region Location Institution

Central Gowalior Regional Engineering College.
East Calcutta Bengal Engineering College, Shibpur
East Kharagpur Indian Institute of Technology, Kharagpur
North Delhi Asian Institute of Transport Development
North Delhi Central Road Research Institute
North Delhi National Institute of Training for Highway
Engineers (NITHE-MoST)
North Delhi RITES (A Government of India Enterprise)
North East Guwahati Indian Institute of Technology, Guwahati
South Bangalore University of Bangalore
South Chennai Highway Research Station
South Chennai Indian Institute of Technology, Chennai
West Mumbai Indian Institute of Technology, Bombay
South Thiruvananthapuram College of Engineering, Thiruvananthapuram
The locations for the training would have to be finalised in consultation

with the Country Champion and the Ministry of Surface Transport.
1.3 Suitable Personnel for Future Trainers of HDM4
The engineers, planners, researchers, academicians from the following

organisations would be suitable to become suitable trainers of HDM4:
1) Ministry of Surface Transport (MoST), Government of India

2) State Public Works Departments
3) Central Road Research Institute
4) National Highways Authority of India (NHAI)
5) Indian Institute of Technology (Delhi, Kharagpur, Guwahati, Chennai and
Mumbai), Universities of (Bangalore, Calcutta), Engineering Colleges
6) Consultants (RITES, CES, Dr. LR Kadiyali & Associates)
1.4 Organisations Benefiting from the HDM4 Training Programme
1) Politicians/Policy Makers
2) Ministry of Surface Transport
3) Public Works Departments
4) Academicians and Researchers
5) Consultants
6) BOT Operators
7) Financial Institutions
1.5 Continuation of Training
The HDM4 training will need resources in terms of funding, equipment and
personnel (trainers). The funding for the training is to come from the
Government of India and International Financial Institutions (World Bank,
Asian Development Bank). It is estimated that a total amount of INR 50
Million will be required for the expenses towards the training of about 550
persons in the first two years of the programme. This cost of INR 50 Million
includes institutional charges but not the cost of transportation and
accomodation and TA/DA of trainees from different parts of the country.
1.6 Action Plan for in-country Training for India
Stage Target Group Operation Need Duration Time

Days
1 Country Champions Finalisation of Draft and To bring to their notice about training 1 Jul 2000
formation of core group received by the trainers, approval of
the proposal and further action
1a Core group Planning of date, duration, To decide and finalise the training 30 Jul 2000
location and target group program
1b Core Group and Other Identification of sponsor and To generate resources, and institutional 30 Aug 2000
officials Locations arrangements
2 Politicians, Top Seminar, Workshop To appraise about capability of HDM4 1X4=4 Sep - Oct 2000
Officials, Financial through Strategy Analysis Example (4 batch)
Institutions
3 PWD Engineers To teach HDM4 run using the To do Project Analysis as supervisor
5 X 5 X 2 = 50 Nov 2000 - Apr
software ( 5 locations, 5 2001.
days, 2 batches in
each location )
4 Consulants and To teach HDM4 run using the To do Project and Program Analysis for 5 X 5 = 25 May - Jul 2001
Contractors software report generation (5 locations,
5 days)
5 Academics and To teach HDM4 run using the To do Project ,Program and Strategy 5 X 10 = 50 Sep - Dec 2001
Researchers software Analysis for further research and (5 locations, 5
development days)
6 Core Group Meeting To Review of progress 10 days Dec 2001
Made
TOTAL PERSONS PROPOSED TO BE TRAINED = 550
TOTAL NUMBER OF TRAINING DAYS = 129
2 Action Plan Schedule
2000 2001
Month JL AU SE OC NO DE JA FE MA AP MY JU JL AU SE OC NO DE
Preparation
Group-Pol
Group-Eng
Group-Con
Group-Aca
Review
ACTION PLAN BY THE MALAYSIAN GROUP
ACTIVITIES PERSON IN-CHARGE TARGET DATE
2000 2001
A. CREATING AWARENESS JULY AUGUST SEPTEMBER OCTOBER NOVEMBER DECEMBER JANUARY FEBRUARY MARCH APRIL MAY JUN JULY AUGUST SEPTEMBER OCTOBER NOVEMBER DECEMBER
1. Seminar For Senior Government Officers,

Policy Makers, and Senior Officers From Private Sectors ZULAKMAL
(30 - 40 PEOPLE), 1 DAY SEMINAR
2. Short course for acedemics DR. JAMALUDIN MAT

(3 DAYS, 20 PEOPLE)
3. HDM-4 course for engineers DR. JAMALUDIN MAT

(2 x 3 DAYS, 20 PEOPLE) Group 1 Group 2 Group 3 Group 4
Session 1 Group 1 Group 2 Group 3
Session 2
B. TACKLING CRITICAL ISSUES FOR IMPLEMENTATION PURPOSERS
1. Develop work standard YAZIP

2. Develop and adopt PIL system ZAMRI
3. Adopt standardized unit costs for economic analysis ZULAKMAL
4. Interface between existing database and HDM-4 ZAMRI
C. Trainers continuous self development

1. Maintain contact with international trainers
2. Frequent discussions between in-country trainers
3. Regular international/regional contact
D. FUNDING FOR TRAINING

1. From the government e.g. JKR, CIDB, MHA etc.
2. From private concession companies
3. From participants fees
COUNTRY ACTION PLAN (INDONESIA)
NO ACTIVITY TIME CONSTRAINT
1 Report to Champion
- Training activities in KL
- Proposed Action Plan
2 Seminar for senior official
3 Study
;-Institutional aspect
- Analyse the position of HDM-4 and existing IRMS
- Training needs analysis
- Calibration level- 1
- Calibration level- 2 and level 3
4 Training
5 Implementation HDM-4
6 Monitoring and Evaluation

NO ACTIVITY OBJECTIVES CONSTRAINT TIME
1 Report to Champions
- Training activities in KL
- Proposed Follow-up Actions
2 Introductory Seminar
- awareness
- commitment
3 Study
- Institutional aspect
- Position of HDM-4 in the existing Systems
- Training needs analysis
- Calibration level- 1
- Calibration level- 2 and level 3
4 Training
5 Implementation HDM-4
6 Monitoring and Evaluation

Draft June 23, 2000
NO ACTIVITY COORDINATOR YEAR 2000 YEAR 2001 YEAR 2002

JN JL AG SP OC NV DC JA FB MR AP MY JN JL AG SP OC NV DC JA FB MR AP MY JN JL AG SP OC NV DC
1 Report to Champions Taufik Widjoyono
- Contact Champion Taufik Widjoyono
- Report on Training activities in KL Salim Mahmud
- Short Present (Overview) on HDM-4 to Champions Djunaedi Kosasih
- Propose Follow-up Actions Erwanto Wahyuwidayat
2 Introductory Seminar Taufik Widjoyono

- Preparation Erwanto Wahyuwidayat
- Formation of Committee & invitation Taufik Widjoyono
- Paper on Overview of HDM-4 Taufik Widjoyono
- Paper on Relationship of HDM-4 to existing RMS's (brief) Erwanto Wahyuwidayat
- Paper on Road Deterioration Model Salim Mahmud
- Paper on Case study (level-1 calibration) Djunaedi Kosasih
- Preparation ( of delivarable trainning materials Erwanto Wahyuwidayat
- Implementation of Seminar (awareness campaign) Taufik Widjoyono
3 Study Erwanto Wahyuwidayat

- Identification of study needs Taufik Widjoyono
- Institutional issue Taufik Widjoyono
- Position of HDM-4 in IRMSs (more in depth study) Erwanto Wahyuwidayat
- Calibration level-1 Erwanto Wahyuwidayat
- Calibration level-2 (tentative) Djunaedi Kosasih
- Calibration level-3 (tentative) Salim Mahmud
4 Trainning and Dissemination Taufik Widjoyono

- Trainning need analysis Erwanto Wahyuwidayat
- Trainning Plan Salim Mahmud
- Trainning Implementation Taufik Widjoyono
- Dissemination (Seminar) Djunaedi Kosasih
5 Implementation of use of HDM-4 By each agency involved
6 Monitoring and Evaluation Taufik Widjoyono
indonesian action plan.xls 7/31/00 - 12:18 PM

PROPOSED ACTION PLAN FOR PAPUA NEW GUINEA
Item No. DESCRIPTION OF ACTIVITY PRESENTED GROUP TYPE TARGET GROUP PERIOD
BY
1 Introduction to Senior Management National Government User - Dept. of Works
Senior Executive Management
Prepare and submitt report ba/pn Secretary and Deputy Secretaries Departmental Head & Champions 1 Week
Give an Overview on HDM-4 ba/pn Secretary and Deputy Secretaries 5 persons 1 hour
Prepare a session plan & other relevant authorities
Finalise action plan and obtain approval ba/pn
2 Pre-Training Requirements and Preparations
Consult other Divisions & Prepare draft

programme ba/pn
Identify Champions ba/pn
Identify Trainers (multipliers) ba/pn
Obtain approval from secretary ba/pn
Secure funding and sponsorship ba/pn
Secure venue and location ba/pn
3 Implementation of In-houseTraining National Government user - Department of Works
Prepare program & session plans ba/pn
Train the Trainers ba/pn A: One senior engineer from each Region Engineers 2 weeks
Introduction & Training B: Planning & Maintenance Branch Engineers & Planners 3 days
10 persons
Conduct Training (Highlands) Trainer - Highlands C: Provincial Works Managers, Provincial Engineers 2 weeks
Civil Engineers - Highlands Region 15 persons
Conduct Training (Southern) Trainer - Southern D: Provincial Works Managers, Provincial Engineers
Civil Engineers - Southern Region 15 persons 2 weeks
Conduct Training (Northern) Trainer - Northern E: Provincial Works Managers, Provincial Engineers
Civil Engineers - Northern Region 15 persons 2 weeks
Conduct Training (Islands) Trainer - Islands F: Provincial Works Managers, Provincial Engineers
Civil Engineers - Islands Region 15 persons 2 weeks
Item No. DESCRIPTION OF ACTIVITY PRESENTED GROUP TYPE TARGET GROUP PERIOD
BY
4 Government Support ba/pn Policy Departments Policy Makers 3 hours

3 - 6 persons
Prepare program & session A : Department of Transport
Assistant Secretary - Infrast., Policy & Planning
Overview presentation of HDM - 4 B: Department of Planning and Monitoring
C: Department of Treasury and Coorporate Affairs
5 Promotion ba/pn Provincial Governments Policy makers 1 day

20 persons
Prepare program & session Works Co-ordinators & Technical Advisors
Overview presentation of HDM - 4
6 Promotion ba/pn Educational Institutions University professors 1 day

10 persons
Prepare program & session Head of School in Universities
eg. School of Civil Engineering, Economics & Planning
Prepared by : Bernard Alois CONTACTS:

Patricia Nagum Secretary - Department of Works Yaip Telue
Deputy Secretary (Technical) - Dept. of Works
Train the trainers course, June 5th -24th, 2000 First Assistant Secretary (Operation) - Dept. of Works
Kumpulan IKRAM Sdn Bhd Assistant Setcretary (Maintenance) - Dept. of Works
KUALA LUMPUR, MALAYSIA
Secretary - Department of Transport Henry Heratau
Completed By
3rd. July
12th. July
28th. July
Completed By
PROPOSED ACTION PLAN FOR PAPUA NEW GUINEA
Item
No. DESCRIPTION OF ACTIVITIES MONTHS MONTHS
Months June July August Septembe October NovemberDecember January February March April May June
1 Introduction to Senior Management
Prepare and submitt report
Give an Overview on HDM-4
Finalise action plan and obtain approval
2 Pre Training Requirements and Preparations

Consultation and prepare draft
programme
Identify Champions
Identify Trainers
Obtain approval from secretary
Secure funding and sponsorship
Secure venue and location
3 Implementation of Training
Prepare program & session plans

Train the Trainers
Training and Utilisation of HDM -4
Conduct Training (Highlands)
Conduct Training (Southern)
Conduct Training (Northern)
Conduct Training (Islands)
4 Government Support
5 Promotion
6 Educational Support and Development

Official Government Request letter
Presentation of HDM - 4
DRAFT PHILIPPINE HDM-4 COUNTRY ACTION PLAN
Training Needs Analysis

1. Planning & Programming • Conduct of Road FS
• Development of Road Maintenance & Improvement Standards
• Creation and Calibration of Road/Transport Databases
2. Institutional • Linkage with related road projects (e.g., RIMMS, Bureau of Maintenance
Projects, Road Network Master Plans for Luzon,Visayas and Mindanao),
BRS Projects
• Linkage/Coordination with Other Government Agencies (DOTC, NEDA,
DBM/DOF, DILG
• Academic Linkage with UP NCTS
Training Plans
• Conduct of Pilot Study on Ad
hoc Road Projects for Country
Case Studies to be used in the Time Frame (mos.)**
Conduct of Training
• Development of Training ! Levels:
Programs/Courses → Appreciation Courses
" Basic 1 for Legislators 1.0
" Basic 2 for Local Government Officials 1.0
" Advanced Course for Technical Senior 1.5
Government Officials involved in
Road Planning and Development
→ Technical Courses
" For Practicing Engineers, Planners and 2.5
Economists
" For University Engineering, Planning and 4.0
Economic Faculty Members
" For University (Undergraduate/Graduate) 5.0
Students
**Arbitrary
Action Plans
Will Commence on*:
• Orientation of key officials regarding HDM-4 framework and relevance of the July 2000
training programs
July 2000
• Establishment of Rapport with Government Agencies related to Road/Transport
Infrastructure Planning and Development and Private and Professional
Organizations (e.g., PICE, REAP, TSSP, Consulting Firms, etc.
• Development of the Training Courses mentioned above August 2000
• Identify Training Requirements (funding, etc.) August 2000
• Commencement of the Conduct of the Training Courses February 2001
• Formulation of Technical and Institutional Measures for the Sustainability of the February 2001
Training Courses
• Institutionalizing HDM-4 Training for Human Resource Development and Capability March 2001
Building
*Suggested
Prepared by: Dr. HS Lidasan & Engr. ED Capucao

DRAFT PHILIPPINE HDM-4 COUNTRY ACTION PLAN
Training Needs Analysis

1. Planning & Programming • Conduct of Road Feasibility Studies (FS)
• Development of Road Maintenance & Improvement Standards
• Creation and Calibration of Road/Transport Databases
2. Institutional • Linkage with related road projects (e.g., RIMMS, Bureau of
Maintenance Projects, Road Network Master Plans for
Luzon, Visayas and Mindanao, BRS Projects)
• Linkage/Coordination with other Government Agencies (DOTC,
NEDA, DBM/DOF, DILG
• Academic Linkage with UP NCTS
Training Plans
• Conduct Calibration and Database
• Conduct of Pilot Study on Ad Hoc
Road Projects for Country Case
Studies to be used in the
Conduct of Training and to test
HDM-4 for local application
• Development of Training Programs • Levels
Courses - Appreciation Courses Participants/Module
- Basic 1 for Legislators 15
- Basic 2 for Local Government Officials 15
- Advance Course for Technical Senior 15
Government Officials involved in Road
Planning and Development
- Technical Courses
- For Practicing Engineers, Planners, and 20
Economist
- For University Engineering, Planning and 20
Economic Faculty Members
- For University (Undergraduate/Graduate) 25
Students
Action Plans
Tentative Dates
• Orientation of key officials regarding HDM-4 framework and relevance of the
training programs July 2000
• Establishment of Rapport with Government Agencies related to Road/Transport July 2000
Infrastructure Planning and Development and Private and Professional
Organizations (e.g., PICE, REAP, TSSP, Consulting Firms, etc.)
• Creation and Calibration of Road/Transport Databases July-Sept. 2000
• Conduct Pilot Study on Ad Hoc Road Projects for Country Case Studies to be Sept.-Dec. 2000
used in the Conduct of Training Programs/Courses and to test HDM-4 for local
application
• Development of the above mentioned Training Courses/Programs Nov. 2000
• Identify Training Sources (funding, etc.) - to be decided August 2000
• Commencement of the Conduct of the Training Courses January 2000
• Formulation of Technical and Institutional Measures for the Sustainability of the February 2001
Training Courses
• Institutionalizing HDM-4 Training for Human Resource Development and Capability March 2001
Building
Coordinating Offices/Persons
• Coordination with DPWH/UP NCTS
• Contact Persons - U/SEC. Teodoro T. Encarnation, DPWH;
"Champions" - A/SEC. Jesus P. Cammayo, DPWH
- Dr. Hussein S. Lidasan, NCTS/UP
Members - to be decided
• Venue of Training May be at UP NCTS
• Budget/Financial Requirements and Sources to be Decided Yet
Prepared by: Dr. HS Lidasan & Engr. ED Capucao
Suggested Working Schedule
2000 2001
Activities
July August Sept. Oct. Nov. Dec. Jan Feb. March
1. Orientation of key officials regarding HDM-4
framework and relevance of the training
program
2. Establishment of Rapport with Government
Agencies related to Road/Transport Infra-
structure Planning and Development and
Private and Professional Organization
(e.g., PICE, REAP, TSSP, Consulting Firms)
3. Creation and Calibration of Road/Transport
Databases
4. Conduct Pilot Study on Ad Hoc Road
Projects for Country Case Studies to be
used in the Conduct of Training Programs
Courses
5. Development of the above mentioned
Training Courses
6. Identify Training Sources (funding, etc.)
- to be decided
7. Commencement of the Conduct of Training
Courses
8. Formulation of Technical and Institutional
Measures for the Sustainability of the
Training Courses
9. Institutionalizing HDM-4 Training for Human
Resource Development and Capability
Building
Tentative Action Plan
Proposed by Siriphan Jitprasithsiri, Thailand
For The HDM-4: Train the Trainer Course at IKRAM, Malaysia
23 June 2000
Time Activities Target Group Main Points

End of 1. A report Chief of Planning • Overview of the HDM-4 software
June submitting to Section • Analytical framework of the HDM-4
2000 my superiors. Chief of Programming • Potential benefit to the highway
Section development plan, maintenance work plan,
Dir.of Planning and highway management.
Bureau • Urgent need: Investigation of the available
Dir.of Training database, Minimum data input requirements,
Division and Method of tranfering data
Dir. Of Maintenance • Proposal on a training course (detail in
Division Action Plan No. 2)
End of 2. A training Junior Economists and Aim: to teach an introduction to the HDM-4
Oct. course on Engineers of the software and its capabilities (see above)
2000 "HDM-4: An Planning Bureau (15) Availability of Equipment:
(3 hrs) Economic Maintenance • Overhead Projector
Appriasal Tool Engineers (5) • LCD Projector
for Highway Senior Executive • Personal Computer
Planning and Engineers (5) Costs (not available):
Management" Consultants (2) • Hand-Outs Materials
Trainer: University prof. (3) • Tea/Coffee Break
Siriphan Venue: Training Division Function Room
Jitprasithsiri Total Participants (30) (subject to availability).
End of 3. A training Economists and Aim: to teach the capabilities of HDM-4 as a
Jan. course on Engineers from DOH beneficial tool for highway maintenance and
2001 "HDM-4: A (200)*** management so that participants will find it
(6 hrs) Tool for Consultants (10) useful to use the HDM-4 software.
Highway University professors Costs*** (not available):
Maintenance (10) • Training fee for Participants from DOH
and Graduate Students • Overhead Projector
Management." (30) • LCD Projector
• Personal Computers (1 for Presentation
Trainers: Total participants(250) and 2 for Hands on experiences)
1. Dr. Siriphan • Trainers fee
Jitprasithsiri Remark: • Hand-Outs Materials
2. Dr. Henry ***: need to seek • Tea/Coffee Break and one lunch
Kelari financial support Venue: A hotel conference room in Bangkok
Note for Support Needs: The training coordinators from TRL shall provide the following
assistances in order to establish the HDM-4 training program and make it sustainable in
Thailand.
• More corporated trainers: I would like the training coordinators from TRL help find
support to invite Dr. Henri Kerali to investigate on the the available database, minimum
data input requirements, method of tranfering data, and deliver on technical issues of the
HDM-4 for the training course no.3 in Thailand.
• Financial Support: I would like the training coordinators from TRL help find financial
support to arrange for the training course no. 3 in Thailand.
A TRAINERS NOTEBOOK
Training Support Notes

for
Train the Trainer Programme
for
Key Trainers in HDM-4 Project
Professor John C Lowe

CRDT
University of Wolverhampton
UK
INTRODUCTION
This "Notebook" has been compiled to provide background texts and supporting information
to a “Train the Trainer” programme aimed at strengthening the training strategies and skills
of Key Country Trainers charged with the development and delivery of training programmes
for the HDM-4 system.
The guide provides Trainers with reference materials that can be used to guide their future
work in planning and delivering HDM4 training. It is strongly emphasised that the Guide has
been designed as a supporting document for the three week “Train the Trainer” course
teaching programme.
The programme has the following Objectives:
By the end of the training programme participants will be able to:
1. Identify their own learning styles and describe how an understanding of these styles and
the Learning Cycle can help them to prepare and deliver better training.
2. Define the main target groups who will need HDM-4 training in their country and carry
out an initial target group profiling exercise for each of these groups.
3. Carry out a training needs analysis of the groups and develop a draft outline training
programme for each of the main groups.
4. Develop a detailed training programme for one of the groups identified.
5. Prepare and successfully deliver to a peer group, an HDM4 training session selected from
the detailed training programme in 4 above.
6. Participate in role playing exercises and practical sessions
7. Demonstrate skills in the use of participatory learning methods.
8. Identify learning difficulties and design and use learning aids to enhance learning in these
areas
9. Utilise the concept of learning hierarchies and learning objectives/competencies in the

structuring and evaluation of training programmes.
10. Develop an action plan for implementing HDM4 training in their individual country/areas
The purpose of the training programme is to introduce participants to the key factors and
techniques in training. Through this knowledge and techniques the key trainers will be better
prepared to pass on their technical and training skills and knowledge more effectively to
others when developing the HDM-4 training and dissemination programme in their respective
countries. This training component of the train the trainers programme lays emphasis on the
importance of preparing and structuring training session content to meet the pre-determined
needs of the trainees and the use of participatory methods suitable for adult learners.
At this initial stage, the project emphasises “face to face” trainer lead training programmes.
However, as the project progresses, a growing range of print and computer-based materials
will be developed which will allow more flexible learning approaches to be supported. These
will be structured to meet the identified information and training needs of those involved with
highway policy, planning and management and may include individual study, self supporting
groups, support networks and web bases information and reference sites
This notebook provides trainers with information and examples from which they can draw
upon in their future work. It is emphasised that the content is not a comprehensive text
covering every aspect of training but simply those aspects that have been identified as
relevant to training activities in this case..
The notes are divided into Units that have common themes. Each Unit is preceded by a brief
introductory text giving the aims and objectives of the UNIT and guidance on how the text
might be used.
The material and approaches proposed are best used within a participatory workshop format.
This is so that principles can be illustrated and applied during individual exercises and
assignments. These exercises and assignments should be based upon realistic working
conditions and technical issues likely to be experience by the trainers being trained.
The text is focused on the practical aspects of the learning process and it should be read in
conjunction with other available texts.
CONTENTS
Page No
UNIT 1
LEARNING STYLES
The Learning Cycle, assessment of learning styles and the implications of
individual styles for training strategies, content and methods.
UNIT 2
TARGET GROUP PROFILES
Gathering and using social, psychological, cultural, economic and technical
information in training planning.
UNIT 3
ADULT LEARNERS and PARTICIPATION
Adult learning and participatory approaches with adult learners.
UNIT 4
TRAINING NEEDS ANALYSIS
Identifying the training needs of individuals and groups.
UNIT 5
AIMS AND OBJECTIVES IN TRAINING
Using aims and objectives to enhance the construction and evaluation of
training programmes and sessions.
UNIT 6
TRAINING SESSION PLANNING
The planning of training sessions and selecting training approaches
UNIT 7
LEARNING AIDS
Designing and using simple learning aids
UNIT 8
TRAINING PROGRAMME PLANNING
Designing and planning in-country training programmes for different target
groups.
UNIT 9
COURSE EVALUATION
Improving feedback and formative evaluation methods in the project.
UNIT TWO
TARGET GROUPS
UNIT AIM
1. To introduce the concept of Target Group profiles so that trainers can prepare
training sessions that more accurately match the trainees needs and are conducted
within a supportive learning environment.
UNIT OBJECTIVES
1. State the main headings which can be used to construct a target group profile.
2. Complete a Profile for a specified target group.
3. Use the profile to explain how the technical content, training methods and learning
environment of a specified session would be influenced by the information
available.
Trainers Notes
This session takes the people focus concept of training developed in learning styles
into a more complex set of characteristics which can also have a major impact on the
training and learning environment. Often the Target Group Profiling and the Training
Needs Analysis which follows in UNIT 3 can be carried out together as they both,
under ideal conditions, involve face to face discussion and information collection. It is
strongly emphasised that both these processes are carried out in the field where the
intended trainees live and work. Desk based analysis is always a very poor second to
field work and desk analysis done from memory and second hand information
invariably results in second rate training.
Version 1.0 HDM-4 Train the Trainer Manual
1
TARGET GROUPS
By now it should be clear that in training, the technical content of a training
programme is only considered after we have clearly defined who the Target Group(s)
are.
Exercise
Consider this situation:
You have been asked to provide a short introductory presentation on
“The advantages of using HDM4 in the design and maintenance of the national road
network”
Your immediate and understandable response might be to start looking out data on the
economic considerations or various permutations on construction options, the scope
of the HDM 4 analytical capacity and the various calibration values needed for
different environments. However before you do all this it might be interesting for you
to stop and ask the question – Who is it that I will be making this presentation to ?
Consider the implications of he two groups below.
1 A group of 15 year old secondary school student doing a project on the national
road network
2 A group of senior government officials with responsibility for the development of
national road policy.
Clearly these two groups will have very different expectations from a session with the
same title
How would you structure presentations for these two groups and what diffence
in content would you make ?.
From this example it will be obvious that the planning of training must begin with the
focus being placed on people and only when they have been considered in detail can
we move to considering the content and the training methods we can adopt.
This Unit on Target Groups provides some suggestions for the main headings you can
use to structure a Target Group Profile. These might be:
Who are you going to train?

Highway Engineers
Technicians
University/College students
Decision Makers
Others?
Number of people to be trained,

Issues
1 Large groups can make participatory approaches difficult
2 Do you need to split large group and run the training a number of times
3 Accommodation and resources need to be matched to numbers involved
2
Age range,
Issues
1 A wide range in ages in the group could suggest a wide range of experience and
different expectations in different age bands
2 More mature people may want more time for discussion and practice in new
methods and concepts They may also appear to be more resistant to change
3 Expectations may vary considerably across age ranges
4 .
Gender
Issues
1 Male or female or a mix in group?
2 Do men and women in the work place have equal job responsibilities?
3 Do men and women have equal access to work place resources?
4 Is training programme scheduled to accommodate domestic commitments?
5 Accommodation and facilities if course is residential
Social/cultural factors relevant to the training programme and approach,

Issues
This can be a major consideration when planning training for trainees from a number
of nationalities. e.g. diet, social custom, etc. However for the HDM 4 programme this
is not likely to be problematic as trainers will be familiar with most issue likely to
arise.
Literacy, numeracy and language.

Issues
Jargon, complex equations. unclear explanations.
HDM 4 is a technical topic and there may be topics, terms and formula and
applications which are not within the scope of the trainee’s current experience or
ability. This is most likely when dealing with those at the ends of the target group
range, e.g Policy makers and field staff concerned with data collection.
Many trainers greatly over-estimate the experience and knowledge of trainees. It is
essential that all trainees understand explanations and demonstrations, and not just a
few who already have a considerable understanding of the topic.
Range of educational ability (general education and technical training)

Issues
1 Linked to previous heading above but analysis may also give indication of
individual ability to manage their own learning.
2 Pace of training programme may be faster with those with greater experience in
the education process.
Level of previous “hands on” experience in the topic to be covered,

Issues
1 Will indicate the starting point for the practical training programme
3
2 Do not assume that a practically experienced trainee also has a full understanding
of the theoretical concepts involved or vice versa ( see learning styles?)
Level of previous knowledge of the topic.

Issues
1 Same as above.
2 A person who can speak knowledgeably about a process or concept may not be
able to apply it in a practical situation
PREPARING A TARGET GROUP PROFILE. – SOME IMPLICATIONS

Using the headings given above the trainer can begin to develop a picture of the
groups to be trained. This information is used in developing the training methods to
be used and the content of the training programme. It also allows the trainer to
develop realistic training Aims and Objectives. Clearly the more detail the profile
contains the more accurately the training programme can be developed and targeted.
To collect really accurate information the trainer should interview a cross section of
the target group and also those who have regular contact with them. e.g supervisors,
clients, subordinates etc.
Numbers to train.
For practical courses it is best to keep groups small - usually practical groups of more
than 10/12 are too big to be able to provide sufficient practice and equipment. Also
on practical training it is not easy to properly supervise larger groups and especially
where safety is an issue. Remember the only way to learn a practical skill is to be
able to have extensive hands on practice. Watching others is not sufficient.
Age of trainees
The age of trainees can influence the training approaches needed in a number of ways.
In general younger people are able to grasp new ideas and concepts more readily than
their older counterparts. However this should not be seen as implying that older
people are slow learners. It simply means that older people have more experience and
often have to compare these new ideas and knowledge against their existing
understandings and practices before considering the acceptance of new ones. This is
why trainers must allow plenty of time for explanation and discussion with older
groups. Remember also that older people do not like to “lose face” so when asking
questions or getting people to do practical tasks try to ensure that you present
questions that they can answer and the tasks in simple steps which they can grasp and
master with the minimum of mistakes. Adult learners are also much more concerned
with the immediate application and utility of the training content being presented. So
keep it relevant and applicable.
Men and Women
4
While, in general, there is no difference between teaching men and women, it is
important to remember that their background experience may well be different. A
good trainer (man or woman) will ensure that an accurate target group profile is
drawn up and that the training programme meets the needs of all participants.
Where there are considerable differences in experience and previous training between
men and women it may be appropriate to run separate courses. The primary guiding
rule is that the trainer must remain sensitive to the needs of all participants, not just
one particular sub-group.
Educational, technical and practical experience

An insight into the abilities of a group will allow a trainer to prepare the training
programme at the right level and to include a good balance of practice and theory.
A major difficulty experienced by many trainers is how to deal with a group which
contains a number of people with wide or specific expertise in the topic. A good
trainer will make use of such people during the course, calling upon them to
contribute to the training sessions and perhaps using them as group leaders in some of
the practical exercises.
It is important for the trainer to find out the qualifications and experience of the group
as early as possible. Some may have had very useful “hands-on” field experience, or
be especially knowledgeable on a particular sector of road construction or design.
Many younger and inexperienced trainers do not like to ask the assistance of someone
in the group as it might reflect badly on their position as trainer. Trainees, however,
rarely think like this. Trainers should take full advantage of any special experience
and knowledge available to them in the group. This can also considerably improve
the involvement and motivation of people who may otherwise feel the course is at too
low a level or not relevant to them.
Motivation
Interest and enthusiasm is infectious and they must start with the trainer. Good
preparation, a sound knowledge of the subject and a good understanding of the target
group, their experience and their problems in the technical area of the training will be
of great value.
The trainer’s own attitude will greatly influence the group. It should be remembered
that it takes a little time for trainer and trainees to get to know each other at the
beginning of a course. This is a time when a trainer can be anxious and nervous.
However, good preparation should make that initial period easier.
Most trainees will be keen to learn, providing the subject matter is relevant to their
situation and needs. There will always be a few who will need encouragement; this
can usually be achieved by actively drawing them into group activities or discussion.
The prospect of having to complete a practical test (e.g. recognition of particular

pavement conditions, carry out estimations during data collection work, data
5
processing with HDM4) can also be a hidden way of motivating the less enthusiastic
trainees.
In any group, there may be great variation and different characteristics and it is
important for the trainer to find out as much as possible about the group to be trained.
In this way, the training can be designed and delivered in a manner suitable for that
particular target group. A great deal of training is less effective than it might be
because those designing and delivering it do not take the trouble to identify their
target groups clearly.
Finally, trainers should remember that even the best trainees can be demotivated if the
domestic details of the course (transport, accommodation, meals, payment of daily
allowances, etc.) are badly organised.
6
Exercise Construct a profile for two different Target groups you will train in the use
of HDM4. Use the format given below
TARGET GROUP PROFILE
Number of people to be trained,
Male or female or a mix,
Age range,
Social/cultural factors relevant to the training programme and approach,
Literacy and numeracy levels, language normally used,
Range of educational ability (general education and technical training)
Level of previous “hands on” experience in the topic to be covered,
Level of previous knowledge of the topic.
Motivational issues
Notes
7
GROUP EXERCISE
This exercise is carried out in small groups of 2 to 5 trainees.
From the list below select a target group.
Government officials responsibly for guiding senior politicians on

Highway Policy..
They are high calibre people but not trained engineers. They need to
know how HDM-4 can improve policy shaping and economic efficiency.
Highway engineers with management responsibility for regional highway

maintenance programmes and budgets.
Engineers and economists responsible for the development and

submission of project proposals concerned with a new Major highway
development.
Private sector consultants who are being considered for appointment to

manage a major section of trunk road and are required to present their
proposals and bid using HDM 4.
University staff responsible for the introduction of HDM-4 application into

a Degree programme
National staff who have been nominated as counter-parts to an

international team from a major Donor who will prepare a Highways
project proposal which requires HDM-4 validation
Task
Using the Target Group profile form provided and your experience draw up a profile
of the group you have chosen
Each group will make a brief presentation of their profile.
8
UNIT THREE
TITLE: ADULT LEARNERS and PARTICIPATION
UNIT AIMS
1. To highlight some of the main issues in adult learning so that trainers can plan more
effective training sessions.
UNIT OBJECTIVES
1. Describe four ways in which Adult learners differ from younger learners.
2. Give three examples of how the trainer can utilise these differences positively in adult
learners
3. Give two examples of participatory teaching methods.
Trainers notes
Your work is concerned mostly with adults and they may have many different backgrounds
and experiences. Most adults already have experience and information related to the subject
being taught. In some cases the training being given may challenge or contradicting what the
trainee already knows and does. To make changes adults often have to UNLEARN what they
already have learned and experienced before they can adopt new methods and ideas. The
notes and tips in the text give direction to the trainer on how more effective participatory
training can be planned
1
ADULT LEARNING
Early learning theory was based on knowledge gained first from studying animals then
children. The term Pedagogy was used to mean the art and science of teaching children. In
early adult education adults were taught as children and because of this much adult education
was unsuccessful. This was often the case with adult literacy programmes.
We have since come to realise that adults learn in a very different way. In the 1960s and
1970s a number of American and European educators were conducting research into the way
adults learn. In 1970 Malcolm Knowles wrote a book entitled “The Modern Practice of Adult
Education: From Pedagogy to Andragogy”. He used the term Andragogy to mean the art
and science of helping adults to learn.
There are many ways in which adult learning differs from child learning. Knowles identified
four major differences.
1 Experience
Adults have a vast reservoir of experience that is a valuable resource for learning. If
these experiences are devalued or ignored then barriers to learning are often created.
2 Self Concept
Adults are independent, self-directing beings who like to exercise control over their
own learning.
3 Immediacy of Application
Adults are generally more motivated if their learning can be applied immediately,
rather than at some time in the distant future.
4 Social Roles
Adults are motivated if their learning is of benefit to them in one of their current roles
in life.
Implications for Trainers:

1 Participatory methods should be used so that the learner can influence and control
their own learning.
2 Adults need a supportive listening environment so that they can contribute to

discussions without fear of being made to feel stupid or devalued.
3 Adults need to be able to diagnose their own needs and to plan their own learning.
They may, however, need help with this.
4 Adults learners need to be given as much information about the programme, methods
and purpose of all aspects of a course in which they are participating. This way they
can take responsibility for their own learning.
5 Negotiation between the trainer and the learner should be an important feature of any
adult training programme.
2
6 Adults need to be able to explore new material and concepts in the light of their own
experiences.
7 Learning should be problem-centred, not subject-centred. Problems are defined and

solutions sought by searching through the information within the subject however the
formal logic and order of a subject may not always fit comfortably with the
complexity and apparent disorder often found in practical problems.
8 Learning may not fall into neat artificially constructed categories: - a more holistic
approach may be needed.
9 Trainers should see themselves as facilitators of learning.
3
PARTICIPATION
What is meant by the participatory approach to training?
Participatory training involves using training methods in which the trainer encourages
trainees to be active in the learning process and give feedback. The trainees share ideas,
experiences and opinions. Participatory training relies on people communicating with each
other.
The big advantage in training adults is that they already have extensive knowledge and
experience. However, because they are so busy they will only go to training if they think it is
relevant. A good way of showing adults that information is useful to them is to relate it to
their own situations and experiences. The best way to do this is to encourage them to
discuss their ideas, to identify their problems, to present their opinions and to share their
experience, i.e. PARTICIPATE! This also requires trainers to have a realistic and preferable
first hand experience of the practical conditions in which the training is to be applied.
Adults are more difficult to convince with just words. They will only adopt technology if
they see it working for themselves. The trainer must spend time showing adults how to do
something and then letting them practice for themselves. This is because they have many
past experiences that they measure new ideas against.
Adults are more sceptical than children, they need more convincing. They can be
conservative and resistant to change. This is particularly true of people in a vulnerable
position where they think new technology or methods may threaten their current position.
They may also have a fear that they may find new methods too difficult to learn. Therefore, it
is particularly important to encourage adults to actively participate during a training session
so that they can put the information into their own words and relate it to the own experiences.
It is not only in the area of knowledge that will need careful handling. Practical processes and
procedures may also be problematic as they are often related to changing ways in which
existing processes are done. Make sure that plenty of time is allowed for guided practice
within the training session. Early failures in practical sessions can demotivate and leave
trainees rejecting a new technique.
Traditionally, adults with little formal education learn by watching others (sitting by Nellie)
who are more experienced, by sharing information and skills, and by Learning by Doing.
They PARTICIPATE.
Advantages of participatory methods.

a) The participants are stimulated and sustain their concentration more easily. Their
involvement leads to greater exposure to the subject which generally raises motivation
and as their knowledge and skills grow they need less stimulation from the trainer and
are able to work independently (intrinsic motivation) This can then lead on to
progressive adoption of new technologies.
b) The trainer can draw on the experiences and abilities of the participants and relate the
subject to the real conditions of the trainees.
e) Training becomes more directly relevant to the trainees needs and experiences.
4
d) Problem solving is more effective using participatory techniques. Many heads are
better than one and the combined experience of the group becomes available to all of
them
e) Participation helps people to remember what they learn in training. For example, in
your own experience do you find it easier to recall ideas you have learnt in a
participatory session such as problem solving in group work and hands on practical
session or ideas presented to you in a lecture?
f) The trainer can get feed back from participating trainees which signals if they are
having difficulties in understanding during the session and so the programme can be
modified to meet the needs of trainees.
g) The trainer can sort out misunderstandings signalled by inappropriate comments or

wrong interpretations etc by trainees.
h) Contributions from trainees enrich the training session and can give trainees greater
self confidence The participants feel valued and on an equal level with the trainer.
This is very important in training adults.
i) Other trainees as well as the trainer can help trainees with difficulties. Trainees
providing help to other trainees also deepen their own learning by having to explain
concepts and procedures.
j) The participants develop a sense of ownership of the knowledge as some of the

information comes from them.
k) Participants and trainers begin to value each other’s opinions and experiences more
and both are learning.
1) Through participation, people learn to accept ambiguity and realise that there may be
many sides to an argument with not just one right answer.
Disadvantages
Of course every method can have disadvantages. Here are some of the problems that can
arise in using participatory techniques?
a) In a participatory approach the trainees take more control of what happens in training.
The trainer must be able to respond to the way in which trainees react in training and
be prepared to adapt. Some trainers find this frightening
b) Adults are status conscious and shy of making fools of themselves; they may be
hesitant to participate at first.
e) Adults may be out of practice at learning and so the trainer will have to carefully plan
how to involve them using participatory methods. This requires considerable skill and
an ability to encourage and motivate people. .
5
d) The trainer must plan ahead and be fully prepared in both the technical and methods
of the session. This often means more extensive planning than that needed for less
participatory teaching styles.
What are the signs that trainees are not participating?

We have emphasised that training and learning should be an interactive process. Try looking
for clues in body language - are they involved and interested or are they removed and silent?
- are they volunteering ideas and discussing or are they passive and letting others do all the
work?. There are many ways to tell if they are involved or confused or not interested. The
trainers job is to try to discover why and change things. A good starting point is to try to
ensure that the topic being covered is relevant to the groups needs.
6
BARRIERS TO GOOD LISTENING
In participation and communications it is just as important to be able to listen effectively so
that as well as giving information you also receive it. You should try to listen with the same
attention and open mind that you hope people will listen to you with. In training it is
essential to develop good listening skills if you are to be a good communicator. Listed below
are some of the things that can happen to stop you hearing the messages.
`On-off' listening
This unfortunate habit in listening arises because most of us think about four times as fast as
the average person speaks. Sometimes we use this extra time to think of our personal
concerns, or interests and troubles, instead of listening and so often miss the point of the other
person’s comments.
'Open ears - closed mind' listening

Here we decide rather quickly that either the subject or the speaker is boring and what is said
makes no sense. Often we jump to the conclusion that we can predict what they know or
what they will say thus, we conclude, there is no reason to listen because we will hear
nothing new.
'Glassy-eyed' listening
Sometimes we look at a person intently and seem to be listening although our minds are on
other things. We get glassy-eyed and a dreamy expression often appears on our faces. We
can tell when people look at us in this way when we are speaking. Similarly, others can see
the same in us.
'Matter over mind' listening

We do not like to have our pet ideas, prejudices, and points of view overturned.
Consequently, when a speaker says something that clashes with what we think, and strongly
believe in, we may unconsciously stop listening or even become defensive and plan a
counter-attack.
Being 'subject-centred' instead of 'speaker-centred'

Sometimes we concentrate on the problem and not the person. Details and facts about an
incident become more important than what people are saying about themselves.
'Fact' listening
Often as we listen to people, we try to remember the facts and repeat them over and over
again to drive them home. Frequently, as we do this, the speaker has gone on to new facts
and we become lost in the process.
'Pencil' listening
Trying to put down on paper everything the speaker says. Eye contact becomes more
difficult and the pencil and paper also act as barriers.
“I've got something to contribute” listening

Sometimes the speaker triggers off an idea in our own mind that we want to contribute. As a
result we eagerly wait to make our contribution and stop listening to the speaker.
7
“HEARING IS NOT ALWAYS LISTENING”
IMPROVING YOUR LISTENING SKILLS
The Basics of good listening skills
1 Responding to the content (subject matter) of what is being said

“The situation in the department is getting increasing difficult”.
!”So the situation is not being resolved”
2 Acknowledging the feelings underlying what is being said
!”It sounds as if you are feeling very uncomfortable about what is going on”
3 Showing you understand and accept what is being said.
!”I can see how much you dislike working in the office when there is so much tension in
the air”
4 Encouraging further detail
!”Can you tell me a little more about what is specifically going on.”
Some more active listening skills to practice
Shut up – you can’t talk and listen at the same time.
Be patient – the time you are prepared to give to listening may be the most valued part of
your contribution.
Concentrate - don’t let you mind wander.
Leave your feelings behind - and don’t use “I know how you feel” because in most cases
you won’t.
Remove distractions – telephone, other people speaking, interruptions, etc.
Look at the speaker – What are their body actions telling you (non-verbal
communications). What is your body saying!?.
Make eye contact – but don’t stare them into submission.
Don’t argue – either mentally or directly. Wait until the full story is heard and then you can
ask questions and make comments. Try to get a feeling about their personality, their likes and
dislikes, what motivates them, what values they hold.
Question your own assumptions before you respond. Avoid making instant judgements.
Most situations are not as straight forward as they first appear.
So listen accurately hear what is said, not what you think is being said.
Try to separate fact from feelings

I have been making a lot of errors at work and I feel that I am hopeless at operating the
HMD-4 software
Fact - I have been making a lot of errors at work
Feelings - I am hopeless at operating the HMD-4 software
8
EFFECTIVE USE OF QUESTIONS
Question and Answer Technique
This is one of the most widely used and most important participatory training methods. It
requires direct communication between the trainer and the trainees. It takes thought, practise
and planning to use questions as an effective teaching technique. In this technique the trainer
uses questions to guide the students thinking and processing of information.
An essential rule in using Questioning in this way is illustrated in a simple Poem by Kipling
“I keep six honest serving men

They taught me all I knew
Their names are What and Why and When
and How and Where and Who.”
The message is – donj’t use questions which only require a “Yes or No” answer. For
example,: Do you understand? – Is that right? As you begin to use questioning techniques
this advice will become clearer.
Use Questions to:

- gain attention;
- increase motivation and interest;
- stimulate group and individual participation;
- encourage communication between the trainer and the trainee;
- give the trainer feedback about the effectiveness of training;
- encourage and develop the participants' confidence;
- draw out the quiet or passive trainee;
- test trainees knowledge;
- develop understanding and reasoning;
- reinforce and clarify ideas and facts;
- open a discussion;
- bring out the participants' own experiences and opinions;
- challenge the trainees and guide their thoughts;
- gather information;
- introduce a problem;
- revise and recall previous training.
Types of Questions
Open - Several possible answers
Closed - One answer (right or wrong)
Subjective - Personal opinions

Objective - Requires a factual answer
Examples
Q1 What is the capital city of France?
Closed - One answer (right or wrong)
Objective - Requires a factual answer.
9
Q2 What is the difference between a diesel engine and a petrol engine?
Objective - Requires a factual answer.
Q3 Why is HDM4 a good choice as a highway management tool.?

Subjective - Personal opinions
Method of Directing Questions

Questions to Answer by Name
To all trainees anyone Overhead
A named trainee a named trainee Directed
To all trainees The trainer Rhetorical
In a participatory session the Directed question is most effective as it allows the trainers to
spread the questions out and also grade the questions so that trainees who are not sure can be
encouraged by giving them easier questions to answer. This method allows the trainer to
nominate a particular person to answer. Remember that trainees will need to understand the
process if they are not to call out during named questions.
The use of questions is usually traditionally thought of as a means of testing or finding out
what the trainee knows. While this is one use’ the real skill of using questions is to lead
trainees to discover and work out how particular concepts are applied or why a particular
process is the best one to follow. Teaching through participatory question and answer
sessions can be very effective but it has to be carefully planned and needs a great deal of
practice to be used well
Questions asked to the trainer

If a trainee asks you a question, what can you do?
- answer it -
- put the question back to the trainee Reverse
- put the question back to another trainee or the whole group Relay
Try to avoid repeating the question as it encourages lazy listening in the group and poor
presentation skills by the trainee asking the question.
Questions to avoid using in a training session

OK?
Do you understand?
Is that clear?
These are non-specific questions and are not very useful or effective questions because they
do not give the trainer much information. (feed back) and are usually answered with “yes”
For example, when you ask the question "Do you understand?” how do you interpret the
meaning of the answer "Yes"?
10
Trainees may answer "Yes" because:
(a) They do understand
(b) They do not want to appear stupid by saying no
(c) They do not really know what the trainer means by the question
(d) They are bored and want the trainer to move on to the next piece of information
Now, imagine that you had just explained the main ways to reduce high rates of deterioration
on a particular road type and you wanted to check whether the trainees had grasped the main
points.
Which one of these questions do you think is most useful for the trainer?
What will the trainer find out from the answers?
1. "Do you understand?" (The trainer will not be sure what the trainees have understood, if
anything).
2. "Can someone tell me one important factor leading to reduced rates of deterioration in
this case.?" (This answer will tell the trainer if the trainees can recall what has been said
and will show which points they think are most important).
3. "In your own experience, what would be the best solution to reducing pavement
deterioration rates in a situation where only hand labour is available.?"(The answer to this
question will show whether the trainees can apply the training matter to their own
situation).
4. "Can you explain to me in your own words how you would explain to a road maintenance
gang how to carry out this maintenance process. (The answer to this question will indicate
to the trainer whether the trainee can correctly recall the subject matter. It will show
whether the most important points have got across. It will also reinforce and clarify the
information for the rest of the group. (“The best way to learn a subject is to have teach
it”)
How to handle answers

"What can happen when you ask a question?"
(a)
No immediate answer
Action
• Manage the silence. - Give eye contact to individuals to encourage a response.
• Wait for someone to answer.
• If there is still no answer, repeat the question in a different way.
(b)
Right answer but not the answer that you wanted
Action
• Be careful to distinguish between a wrong answer and an answer that is right but not the
one you were expecting.
• Be careful to praise the participants and encourage them for giving a correct answer.
11
• If the answer is not the one you expect, you could say: "Yes, that is correct but can
anyone else think of any other answers?" or, "I hadn't thought of that, can you think of
anything else?"
(c)
Wrong answer
Action
• Why may this happen?
• Has the question been misunderstood?
• Has the not understood the information given in the session?
• Rephrase the question For example, at the end of a training session about estimating
pavement condition. The trainer might ask a trainee:
"How do you know when to commence maintenance /repair work?"
The answer you want is, "when the surface falls below X Y standards." However, the trainee
answers, "when you can see pot holes forming." This answer is wrong and also shows that
the trainee has not understood the purpose of preventative maintenance.
How to handle a wrong answer

Be careful not to make the trainee feel foolish or put down. If this happens, they may not
contribute any more to the session so phrase your reply sensitively.
If an answer is wrong, you could handle the situation by

Asking "are you sure, would you like to rethink?"
or
Ask another trainee, "Do you agree with that statement?"
These activities are designed to determine what misunderstanding has occurred to cause the
wrong answer to be given and to determine what subsequent action will then need to be taken
to put things right by:
Supplying more information;
Clarifying the information
Having more discussion
Rephrasing the question;
How to use questions effectively

i. Before a session, plan your questions and think of possible answers and how these
answers will take the session forward.
ii. Think of interesting questions which can lead into or complement your key teaching
points.
iii. At certain points in the session, use questions to find out whether the trainees have
understood the most important points.
iv. Use short clear questions, simply phrased.
When using the directed question, do not pause for too long between asking the question and
nominating the trainee. Trainees may need training in this technique - that is you may have
to explain that when you have asked the question, you do not want trainees to call out the
answer, as you will ask one of them by name.
12
Finally remember that asking questions involves risks so be prepared for surprises.
If you are asked a question which you do not know the answer to, it is best to say ”sorry I
don’t know” BUT then find out as soon as possible and give them the answer.
Some questions can also have different answers for different people
e.g. "Where and when does the sun set?"

Most people in the world would answer, "in the West in the evening."
In the Arctic there are several possible answers, depending upon the season!
"It does not set."
"There is no sun."
"It sets in the West."
"It sets in the East."
"It sets in the morning for 1 hour......"
13
GROUPS
Groups are a normal and important element of most cultures and they can be formed for
many reasons and group work is an important area of activity in participatory learning.
In training it can be helpful to use groups to aid learning by allowing individuals to share
their experiences with others and to learn from each other. Groups allows more complex and
longer term exercises and assignments to be set.
Sometime groups do not work well and the following list may give you some clues why a
particular group does not work.
Common Barriers to good group dynamics - No specific order

Artificial, contrived reasons for formation
Unclear objectives
Ineffective procedures and processes
Unbalanced composition
Inappropriate leadership
Overbearing leadership
Insufficient trust or openness
Unfavourable climate - internal or external
Low creativity
Unfavourable relationships between members
Low achievements on group targets and objectives
14
Summary of issues for working groups
Working Groups?
Purpose Membership and structure Co-operation & Development
Is the purpose clear? Voluntary participation The need to belong (Inclusion) Forming
Whose purpose is it? Group size exert influence (Control)
Group heterogeneity to give and receive approval and support
The need for clear owned goals Common values and principles (Affection)
The need to measure and monitor progress Members individual objectives - Overt and
The need to avoid aimlessness, drifting and covert Pecking order Storming
disintegration Performed leadership or decided by group Muscle flexing
Group processes and procedures clearly Testing of others
defined Shifting loyalties
Mutual accountability
Task and maintenance functions and roles Emergence of Norming
defined Cohesion
Norms
-meetings
-decision making
Leadership Performing
control and limitation mechanisms
Adjourning??)
15
OHP
Stages of group Development
Forming Collection of individuals
Each with own agenda
Little or no shared experience
!
Storming Personal values and principles challenged
Roles and responsibilities accepted or rejected
Conflict as personal agendas are followed
If successfully handled, group objectives and processes
are defined.
!
Norming Rules accepted
Shared goals established
Group and sub-grup settle down and become cohesive
!
Performing Group maturity and energy
Maximum productivity
New insights and solutions appear
!
Adjourning Loss (achievement?) of common goals
Sub groups form with separate agenda
Loss of interest and purpose
Members leave
Group ceases to function
"
16
OHP
Stages of group Development
Forming Lennon and McCartney didn’t know each other
and uncertain of each other’s skills
Members unable to commit to long term
! relationship
The Quarry Men, Johnny and the Moondogs,
The Silver Beatles
Storming Lennon and McCartney recognise each other
talent and begin to practice together.
The group struggle through long periods of
internal strife, drummers, managers and
guitarists quit.
!
Norming Something happens in Hamburg. New
techniques and songs are tried. Tight bonding
and cohesion; camaraderie “were just four parts
! of one” McCartney
Performing The group entered a period of unparalleled
productivity. Concert tours. television and films.
Millions of records sold. They stopped touring
! in 1966 but continued to sell albums until 1970
Adjourning Growing distance amongst members. Lennon
married Ono. McCartney turns more to his wife
Linda. Epstine dies of overdose McCartney
announces he is leaving “All things must pass”
"
17
Participation equals
" Discussion
" Questions - answering and
asking
" Looking Observing
" Explaining Describing
" Problem solving
" Carrying out practical processes
" Sharing ideas, opinions and
experience
" LISTENING
18
UNIT SIX
TRAINING SESSION PLANNING
UNIT AIM
1 To introduce a structured approach to session planning so that trainers can prepare
sessions within a common framework/structure and can share session plans with
others.
UNIT OBJECTIVES
1 State the main parts of the session
2 Explain the purpose of major UNIT and their role in the learning process
3 Produce a session plan to the required format
Trainers Notes
The following text is intended as a framework for the major considerations when
planning a training session It draws upon and itegrates the materials covered in the
preceding Units. While the basic session framework is where the structure and
purpose of the session is developed. The trainer will in addition also need to prepare a
set of personal training notes which set out in detail the content of the session and
the methods and activities planned to achieve participation, group work, use of
learning aids and timing etc. The key to this process is that the session content and
methods are guided by the session aim and objectives. If the objectives have been
careful developed in the training plan then the individual trainer can plan a session to
meet these objectives. The point being that while individual trainers may take
different approaches, the outcomes should be the same. i.e. the objectives are met.
GUIDELINES FOR PREPARING TRAINING SESSIONS
This is a checklist for you to use in the planning process. However keep in mind that
each session while being planned individually should also recognise the training
previously covered and any training sessions following.
The “Proforma Planning Sheet” given below is a suggested layout but individual
trainers may decide to modify it in the light of their own experience and needs.
However the same general principles and headings should still be relevant. A good
guide to use is that if you gave your session plan to another trainer, He/she should be
be able to use it to deliver the same session.
The Proforma Planning sheet is only a summery sheet and must be supplemented with
a set of “Training notes” which set out the content notes needed and show how the
detailed training process is programmed
i.e. It covers detailed timing, when visual aids are to used, what leading question
might be used, when group work might be introduced and guide notes for group task
etc. For a new trainer these notes may be quite extensive and detailed so that they
don’t forget what is being taught and how they intend to implement the training
methods selected. As experience grows these will be reduced to headings and simple
short notes.
CHECKLIST
A Session Title Is it motivating and informative
B Session Aims and Objectives

1 have you written an Aim which states What you will teach and Why the
trainees need it?
2 have you written your Objectives with Action Verbs which clearly state what
the trainee will do?
3 consider what type of activity your objectives are indicating your trainees will
practice during the session:
a) KNOWLEDGE
b) APPLICATION
c) SYNTHESIS
d) EVALUATION
e) ATTITUDES
f) SKILLS
4 Are your aims and objectives closely matched to the training needs of your
TARGET GROUP?
5 What attitudes do you want the participants to develop as a result of the

session'?
Revision example
Title Unsealed Roads –improved to –“The importance, types and management of
unsealed roads in national networks”
Aim
To teach the contribution of unsealed roads to the national network and their
management so that participants can implement improved management programmes
and methods.
Objectives
The trainees will
Explain the contribution of these roads in the national network
Describe the main classifications and characteristics of each
Describe the main factors to be considered in developing an effective management
strategy for these roads.
Develop and defend a suitable management programme from a given HDM4 case
study.
B Introduction to Session
1 Can you use visual aids, anecdotes, real material to motivate your trainees?
2 Will any of your trainees have PRACTICAL EXPERIENCE of the subject

which can be used for the benefit of the class?
3 Have you planned how you will motivate the trainees by INVOLVING them
early in the session?
4 Have you written down the QUESTIONS you will use to involve trainees in
the session?
5 Will you tell trainees the TITLE of the session and give an OVERVIEW of it?
6 Will you tell your trainees how this session fits with other training sessions?
C Development of the Session
1 Does your session develop from SIMPLE ideas to more COMPLEX ones?
2 Have you planned how to INVOLVE trainees in the session?
3 Have you planned what NOTES the trainees will make?
4 Can you use HANDOUTS in your session?

5 Will you use a RANGE of VISUAL aids to increase trainee interest in your
session?
6 Are you trying to relate your session to PRACTICAL APPLICATIONS

wherever possible?
7 Have you planned how you will use the white board. overhead projector and
power-point presentations
8 Do your trainees (REALLY NEED) all the information in your session?
9 Have you estimated how long each part of the session will take?
D Conclusion of Session
1 have you planned how you will summarise the main points covered in the
session and test your objectives?
2 Have you written down the questions you will ask the trainees to test whether
the trainees have learned what you intended?
3 Will you tell trainees how the session will be followed up in future practical or
theory sessions?
E The Room
1 Will you CHECK that all the aids you will use are available and working?
2 Will you check that all your aids can be clearly seen from all parts of the room
If the answer to most of these questions is YES then you are more likely to motivate
your trainees, keep them interested and achieve your objectives, helping trainees to
learn and prepare for their future jobs.
TRAINING SESSION
SELF-EVALUATION CHECKLIST
5 4 3 2 1 0
HOW WELL DID I......? Very good satis- not very poorly not
well factory well applicable
1 link this session to other
sessions
2 introduce this session
3 make the aims clear to the

trainees
4 move clearly through the

structure
5 emphasise key points
6 summarise the session
7 maintain an appropriate pace
8 capture trainees' interest
9 maintain trainees' interest
10 handle problems of
inattention
11 ask questions
12 handle trainees' responses
13 direct trainee tasks
14 cope with the range of ability
15 monitor trainee activity
16 use a range of teaching aids
17 make contact with all class

members
18 cope with individual

difficulties
19 keep the content relevant
20 use body language
21 check learning
22 build trainee confidence
23 convey enthusiasm
24 provide a role model

TRAINING SESSION PLAN Duration
Title
Target group profile
Aim
To teach
Why
Objective(s)
The trainee will:
Session Structure Timing

Introduction
Development
Training points
Summary
Conclusion
Materials, equipment and pre-planning issues

Worksheet
Improving titles
We could have a title “Unsealed Roads”. While this is descriptive it does not really
tell us much. However if we expanded this title to read, “The importance, types and
management of unsealed roads in national networks” we will have a much better idea
of what we re attempting to teach. It will also give the trainee a much better idea of
what to expect
Another important point of having a clear and informative title is it makes it easier to
decide what to cover in the training session. A further advantage of a well thought out
title is it will allow you to use a simple analysis method to highlight the content of the
session
Look at the example below
Road Maintenance and improvement Budgets in Strategic Analysis

R M I B SA
By identifying the important words in the title and applying the six questions to each
the major content to be introduced in the session will become evident.
For example consider the word “improvement” in the title by asking
What improvement?
Why improvements?
Where is the improvement carried out?
When is the improvement done?
Who is responsible for improvement?
How are the improvement implemented
Try the same process for Budgets
The table below indicates with the word code where the question may be relevant.
This not a precise tool but it does make the trainer consider the content of the session
carefully and the answers will begin to reveal what needs to be included and the
possible structure of the session
What Why Where When Who How

R R
M M M M M M
I I I I I I
B B B B
SA SA SA SA SA
Example of an aim and objectives for the title
Aim
To teach the factors influencing budget allocations for road maintenance and
improvement in startegic analysis
So that trainees can identify the key factorsand criteria involved and use them in
HDM-4 strategic planning applications.
Objective
The trainees will
describe the key elements and criteria to be used.in HDM.4
programme and produce an HDM4 report when supplied with a given set of
conditions and support information.
demonstrate their ability to explain and apply the results available in the report.
Example2
Title
Unsealed Roads
Improved title
The importance, types and management of unsealed roads in national networks
Aim
To teach the contribution of unsealed roads to the national network and their
management so that participants can implement improved management programmes
and methods.
Objectives
The trainees will
Explain the contribution of these roads in the national network
Describe the main classifications and characteristics of each
Describe the main factors to be considered in developing an effective management
strategy for these roads.
Develop and defend a suitable management programme from a given HDM4 case
study.
UNIT SEVEN
TITLE: LEARNING AIDS
UNIT AIMS
To make you aware of the potential uses and types of simple learning aids available to
the trainer so that you can make better use of them and enhance the trainees learning
experience.
UNIT OBJECTIVES
Define three ways in which learning aids can improve trainee learning.
Design and use a learning aid in a practice training session to get over a particular
learning point
Define the learning difficulty and how the aid overcomes this.
Trainers notes
There is a wide range of Learning Aids available to the trainer and often the most
effective ones are designed and made by the trainer from locally available resources.
The primary aspect of learning aids is that they must be selected and used to meet a
specific purpose in the training session. The trainer needs to be clear on why they are
being used.
Too often trainers will show a long series of slides/power-point or a video which may
take up to 30 minutes to present. Unless the intention is to present an overview then it
is far better to select only those slides or video clips that are directly related to the
Session topic and Objectives. The most likely learning aids that will be used in the
computer based HDM-4 programme are text in the form of notes, summary lists,
graphs, formulae etc and pictures and diagrams frequently presented through power
point presentations, overhead transparencies or directly on the computer screen.
However this does not mean that that white boards and flip charts, etc are not of any
use here. In fact, the white board, and flipcharts can be extremely useful and flexible
when working with groups where feedback can be collected and presented in this
way. These are also more flexible when working in a participatory way when visual
presentations can be built up and modified in response to the responses and questions
of trainees.
Learning aids are not only a great help to the learner, they are also a useful way to
structure your training plan and session. They provide a useful focus when trying to
build new ideas and concepts with trainees. Careful use of questions centred on the
aid allows you to systematically integrate the trainee’s experience with the new
concepts and knowledge you are introducing. Finally, remember that we learn a great
deal more through our eyes than we do through our ears. Vision is a greatly under-
utilised channel in the learning process
PRODUCTION AND USE OF VISUAL AIDS
Introduction
It is worth remembering that just as people can misunderstand what is said they can
also misinterpret what they see. This means that you must choose and design visual
aids with considerable care
The three examples given at the end of this Unit illustrate how although we assume
that our eyes never deceive us that is not always the case
Example one shows how something that can be drawn cannot always exist in the real
world.
Example two illustrates how we learn to interpret ways in which we can use various
visual “language” (dotted lines in the box in this case) to make something clearer.
Example three shows how we may interpret the drawing (joining up the lines) when
the detail is incomplete. If you see the drawing upside down first you may not realise
it is a chair.
These examples illustrate how careful we should be in developing and using visual
aids. Never assume that everyone sees things in the same way as you do. So when
using visual aids always ask question about them “Can you describe the shape/ colour
of the road surface?” “What is the problem you can see at the edge of the pavement
area in the picture?” “Can you explain what this graph shows?” If you get an answer
you didn’t expect it could indicate that the trainee is not “seeing” the visual aid in the
way you had intended.
Using Learning Aids

In a training session, some points may be hard to get across simply by explanation and
description. It is in these situations that visual aids, also referred to as learning aids,
can help. Specimens and samples of the real thing are often the best learning aids, but
their use can be difficult. In some cases they may be so small that they cannot be seen
without a microscope (fine materials in soils and pavement materials) or they may be
so big, (a road surface or a road section eroded by water) that it is not practical to get
them to place of training.. In these cases, pictures, (photos, slides or computer based
images), drawings. video, or site visits may all be considered. Where the training is
dealing with formulae, graphs, diagrams, matrix etc it is essential to remember that a
close relationship has to be established and maintained between physical conditions
and these forms of representing on-site conditions.
This Unit gives broad guidance on the use of some forms of visual aids. However,
ultimately their effective use depends on the skill of the trainer in diagnosing the areas
of learning difficulty and choosing an aid appropriate for use in that situation. The
illustrations, pictures and visual materials available in the Train the Trainer CD will
assist trainers in developing their training sessions, but they should also try to ensure
that they develop a wider range that are specifically relevant to their own local
conditions and culture.
Handouts
Handouts provide an effective way of providing trainees with information that can be
used for reference in the future. They can also save a great deal of time by cutting
down the amount of time trainees spend making notes. However, the trainer must
design handouts that are not simply a copy from a page in a book. A handout should
present the main points and important information of the session in a logical and
simple way that helps the trainee to recognise the structure of the training session
which will help them to remember and use the information in the future.
Handouts should be used during a training session, not simply given out at the end.
This can be achieved by asking questions on the content or organising small
discussion groups to work on a particular topic in the handout. Spaces can also be left
for key sentences or words to be filled in. This makes important information more
easily remembered.
Handouts covering broader issues involving concepts, explanations, etc. can be

unattractive. When designing handouts, break them into easily digestible paragraphs
and, if possible, include a few diagrams and line drawings, emphasise key words in
“Bold” or in a change of type face or size to make them more visually attractive.
Whiteboards
Boards are frequently used by trainers, but unfortunately, they can be used badly and
become a personal scribbling-pad for the trainer.
If you plan to use the board as your main presentation tool then try to ensure that the
main headings and key points are built up as the training session progresses. This
helps to emphasise the main elements of the session and provides the trainee with a
good structure to follow.
Boards come in many shapes and sizes, from very small portable ones that can be
used in practical field sessions to larger class boards. It is a good idea to split the
larger boards into sections and keeping part free for the sketches, diagrams,
calculations etc which may be needed during a training session.
Use colour to give emphasis and interest to notes and diagrams, but keep diagrams
and sketches very simple and use prepared Acetates and charts for more complicated
topics unless you are good at sketching. Use block capitals for main headings, lower
case underlined for sub-headings and, provided it is easily read, your normal
handwriting for text. Size of writing will depend on room/group size, but as a basic
rule, your letters should be not less than 25 mm high (approximately the distance from
the top of your thumb to the first joint).
Charts
Charts are one of the most flexible and reliable visual aids available to trainers who
conduct field-based training.
Advantages of charts
• Inexpensive, easily and quickly produced;
• Ideas for field-based training
• Easily transported and stored.
Making charts
Some rules on visibility
Lines drawn should be at least 3 mm thick
Lettering should be at least 25 mm high (length of top thumb joint)
Avoid drawing a lot of detail as it is normally unnecessary and often too small to be
seen from the normal viewing distance
The practical test of visibility is to view the chart (or any other visual aid) from the
position of the furthest trainees. Can you see it clearly?
Use of colour
Colour gives emphasis to important parts or shows parts which are related in purpose.
However, using too many colours can cause confusion, so try not to use more than
four colours.
Colour can be added with poster paints, felt-tipped pens or even coloured chalk.
However, filling large block of colour can be time-consuming and tedious,
particularly with felt tipped pens.
Design
Some trainers prefer to develop their own drawings, but this is normally unnecessary.
There are always plenty of examples and ideas to be found in booklets, leaflets and
manuals. As the HDM-4 programme develops it is hoped that more visual materials
will be available in training support materials and on the HDM-4 web site.
There are two basic mistakes that many trainers make:
1 They fail to use the full visual areas available and so produce work which is
too small to be seen properly.
2 They include a lot of writing and labels on the diagrams and pictures . This
often reduces visual impact and adds rather than reduces confusion.
The overhead projector (OHP)

This is a very useful piece of equipment for the training room. The bulb will last a
long time provided that it is allowed to cool down before the OHP is moved.
A good inexpensive screen can be made from a framed half-sheet of hard-board, the
roughened surface of which is given two coats of white emulsion paint. With two
hooks screwed into the top of the frame, it can be hung at an angle from the ceiling in
a corner of the training room. The angle ensures a square picture.
Diagrams can be prepared in advance, on transparent acetate sheets, using either

water-or-spirit based OHP pens of different colours. The latter gives a permanent
drawing which can only be removed using a special eraser. Acetates produced by
photocopying have two major dangers.
They are usually copying something in a book or other publication which often
contains complicated diagrams and too much detail
The size of drawings/diagrams and the print size is invariably too small
As with all visual aids, the lettering and details must be big enough to be visible from
the back of the room and the simplest way of verifying this is to go to the back and
see for yourself. When making an OHP sheet, first draft your diagram and any
lettering on a sheet of A4 lines paper, then place that under an acetate sheet and trace
it. Do not try to put too much information on one sheet.
When an OHP sheet is used for information only, or for presentation of data, it is
often helpful to the trainees if you reveal the contents of one point at a time by
covering the remainder of the slide with a piece of paper.
When using the OHP, you can stand facing the audience using a thin pen (not your
finger) to point out features on the slide. There is no need to turn your back on the
audience.
SOME BASIC RULES

1 Print size on Acetate sheets should be not less than 20 point
2 When presenting points do not have more than 6 or 7 lines on one slide.
3 Do not project large quantities of text, - It is best to put that on a handout.
4 Add some colour to make it more attractive to look at.
5 Make sure that all parts of the diagram, data, graph and text can be read by those
at the back of the room. (Select material in publications carefully as most of it is
too small to be seen when projected.
6 Switch of the projector if you are finished with a slide as it may distract trainees
from a new point you are introducing.
Summary
• Learning aids are intended to be tools for the trainer to use to make the
learning process more easy for the learner.
• Always make sure that the learning aid has a specific purpose in the learning
process.
• Learning aids are for use in the training process so don’t try to turn them into
detailed notes or make them over-complicated. They should complement the
training session an are not intended to stand alone.
• Make sure that everything in the aid is visible to all the trainees
• Use colour to make them more interesting
• Learning aids are useful when trying to link theory to practice.
• The use of learning aids should be carefully planned.
UNIT ONE
UNIT TITLE: - LEARNING STYLES
UNIT AIMS
1 To make you aware of your own learning styles and how your preferences can
influence your training style which in turn may favour some of your students
while disadvantaging others.
2 To introduce the Learning Cycle as a means of providing a framework for
planning training.
3 To emphasise the learner and their learning process as the primary focus in
teaching.
UNIT OBJECTIVES.
1. Complete the learning style inventories and defines your learning style
preferences
2. Give two examples of a learning style profile and state what strategies would be
needed by the teacher to strengthen the learning process in the examples presented
3. Describe a typical trainee behaviour, which would indicate the test profile given
Notes
This UNIT contains a Questionnaire, which allows the trainer to set an exercise,
focusing on individual styles of learning. By completing the questionnaire each
participant can plot his or her own learning style and from comparison with other can
see that everyone is different. It is essential to emphasise that there is no "best" profile
but rather that everyone can improve their learning skills by being aware of their
individual preferences
The results of the questionnaires need to be processed by following through the

process: a, b and c below.
a) complete the questionnaire
b) transfer the results from the questionnaire to the “Learning Style Score Sheet”
c) transfer the totals from the Learning Style Score Sheet to the “Personal Learning
Style Profile Sheet.
The results can then be firstly explained by reference to the “Learning Styles –
General Descriptions” It should then be related to the Learning Cycle diagram and the
Learning Cycle Overhead Transparency (OHT) given at the end of this Unit. This can
be used to demonstrate that the learning cycle (shown as four stages) is one that we
should try to copy. However, we all naturally tend to have preferences for two to
three stages only. The trainer should try to compensate for this by encouraging
students to engage in activities, which are less preferred by them. Also when planning
training sessions, trainers should try to provide activities which take the learner
through all four stages in the Learning Cycle
This questionnaire and diagnosis should not be taken too seriously as it indicates
tendencies rather than providing strict categories. What the exercise can do is to
highlight the importance of the trainer understanding that trainees do have differing
learning styles and the teaching must be able to accommodate these. It is also useful if
students can be encouraged to expand their own learning style to use the full range of

1
learning opportunities available to them. Remember that as a trainer you teach
People a Subject rather than a Subject to People. The subject must always to be
structured to meet the needs of the target group you intend to train.
Finally trainers should be reminded that their own particular learning preferences
could favour students with similar profiles. Therefore trainers must remain conscious
of their own preferred styles when planning and conducting training. This means
taking students through the full learning cycle.
Example
Cycle Student Activity Teacher support action

Experien Experimenting/doing/observing Prepares and structures the activity
ce in a way which maximises student
involvement and exposure to the
important elements.
review Reflecting on the experience and Guides process by asking key
noting actions and results and questions to help student structure
process. Trying to make sense of their experience and observations
the information and experience Asking students to explain the
process
Analyse Ordering and relating and Assisting students to link their
integrating with other previous experience with new
experiences and knowledge. material by asking key questions.
Considering what to do with new Setting group exercise for students
information and experience to present their perceptions and
ideas for next step or application
Plan Considering how to proceed and Key questions
planning the next action. Where Students exercise - present their
do we go from here? How do ideas and plans for comment by
we apply it? teacher and peers
Guided discussion

2
THE LEARNING STYLES QUESTIONNAIRE
This questionnaire is designed to find out your preferred learning style(s). Over the
years you have probably developed learning "habits" that help you benefit more from
some experiences than others. Since you may be unaware of this, this questionnaire
will help you pinpoint your learning preference so that you are in a better position to
select learning experiences that suit your style.
There is no time limit to this questionnaire. It will probably take you 10-15 minutes.
The accuracy of the results depends on how honest you can be. There are no right or
wrong answers. Use the Learning Style Answer sheet to record your answers. If you
agree more than you disagree with a statement put a tick by it (!). If you disagree
more than you agree put a cross by it (").
1 I have strong beliefs about what is right and what is wrong, good and bad.
2 I often act without considering the possible consequences.
3 I tend to solve problems using a step-by-step approach.
4 I believe that formal procedures and policies restrict people.
5 I have a reputation for saying what I think, simply and directly.
6 I often find that actions based on feelings are as sound as those based on careful
thought and analysis.
9 What matters most is whether something works in practice.
10 I actively seek out new experiences.
11 When I hear about a new idea or approach I immediately start working out how
to apply it in practice.
12 I am keen on self-discipline such as watching my diet, taking regular exercise,
sticking to a fixed routine etc.
13 I take pride in doing a thorough job.
14 I get on best with logical, analytical people and less well with spontaneous,
"irrational" people.
15 I take care over the interpretation of data available to me and avoid jumping to
conclusions.
16 I like to reach a decision carefully after weighing up many alternatives.
17 I'm attracted more to novel, unusual ideas than to practical ones.
18 I don't like disorganised things and prefer to fit things into a coherent patter.
19 I accept and stick to established procedures and policies so long as I regard
them as efficient way of getting the job done.
20 I like to relate my actions to a general principle.
21 In discussions I like to get straight to the point.
22 I tend to have distant, rather formal relationships with people at work.
23 I thrive on the challenge of tackling something new and different.
24 I enjoy fun-loving, spontaneous people.
25 I pay meticulous attention to detail before coming to a conclusion.
26 I find it difficult to produce ideas on impulse.
27 I believe in coming to the point immediately.
28 I am careful not to jump to conclusions too quickly.
29 I prefer to have many sources of information as possible - the more data to think

3
over the better.
30 Flippant people who don’t take things seriously enough usually irritate me.
31 I listen to other people’s point of view before putting my own forward.
32 I tend to be open about how I’m feeling.
33 In discussions I enjoy watching the manoeuvrings of the other participants.
34 I prefer to respond to events on a spontaneous flexible basis rather than think
things out in advance.
35 I tend to be attracted to techniques such as network analysis, flow charts,
branching programmes contingency planning etc.
36 It worries me if I have to rush out a piece of work to meet a tight deadline.
37 I tend to judge people’s ideas on their practical merits.
38 Quiet, thoughtful people tend to make me feel uneasy.
39 I often get irritated by people who want to rush things.
40 It is more important to enjoy the present moment than to think about the past?
41 I think that decisions based on a through analysis of all the information are
sounder than those based on intuition.
42 I tend to be a perfectionist.
43 In discussions I usually produce lots of spontaneous ideas.
44 In meetings I put forward practical realistic ideas.
45 More often than not, rules are there to be broken.
46 I prefer to stand back from a situation and consider all the perspectives.
47 I can often see inconsistencies and weaknesses in other people’s arguments.
48 On balance I talk more than I listen.
49 I can often see better, more practical ways to get things done.
50 I think written reports should be short and to the point.
51 I believe that rational, logical thinking should win the day.
52 I tend to discuss specific things with people rather than engaging in social
discussion.
53 I like people who approach things realistically rather than theoretically.
54 In discussions I get impatient with irrelevancies and digressions.
55 If I have a report to write I tend to produce lots of drafts before settling on the
final version.
56 I am keen to try out things out to see if they work in practice.
57 I am keen to reach answers via a logical approach.
58 I enjoy being the one that talks a lot.
59 In discussions I often find that I am the realist, keeping people to the point and
avoiding wild speculations.
60 I like to ponder many alternatives before making up my mind.
61 In discussions with people I often find I am the most dispassionate and
objective.
62 In discussions I’m more likely to adopt a “low profile” than to take the lead and
do most of the talking.
63 I like to be able to relate current actions to a longer-term bigger picture.
64 When things go wrong I am happy to shrug it off and “put it down to

4
experience”.
65 I tend to reject wild spontaneous ideas as being impractical.
66 It’s best to think carefully before taking action.
67 On balance I do the listening rather than the talking.
68 I tend to be tough on people who find it difficult to adopt a logical approach.
69 Most times I believe the end justifies the means.
70 I don't mind hurting people's feelings as long as the job gets done.
71 I find the formality of having specific objectives and plans stifling.
72 I'm usually one of the people who put life into a party.
73 I do whatever is expedient to get the job done.
74 I quickly get bored with methodical detailed work.
75 I am keen on exploring the basic assumptions, principles and theories
underpinning things and events.
76 I'm always interested to find out what people think.
77 I like meetings to be run on methodical lines, sticking to laid down agendas, etc.
78 I steer clear of subjective or ambiguous topics.
79 I enjoy the drama and excitement of a crisis situation.
80 People often find me insensitive to their feelings.

5
LEARNING STYLES SCORE SHEET
Transfer your results from your answer sheet by indicating on the list below which
items were ticked. (!agree)
Add up the total number of questions ticked for each category and enter the column
total in the total box at the bottom of each column.
ACTIVIST REFLECTOR THEORIST PRAGMATIST
2 7 1 5
4 13 3 9
6 15 8 11
10 16 12 19
17 25 14 21
23 28 18 27
24 29 20 35
32 31 22 37
34 33 26 44
38 36 30 49
40 39 42 50
43 41 47 53
45 46 51 54
48 52 57 56
58 55 61 59
64 60 63 65
71 62 68 69
72 66 75 70
74 67 77 73
79 76 78 80
TOTAL: TOTAL: TOTAL: TOTAL

6
Personal Learning Style Profile Sheet
Circle your score from the score sheet for each category
This will show your most and least preferred styles
Activist Reflector Theorist Pragmatist Preference
20 20 20 20
19
18 19
17 19 Very Strong
16 19 18
15 18
14 17
13 18 16 17
12 17 15 16 Strong
16
11 15 14 15
10 14 13 14
9 13 12 13 Moderate
8
7 12 11 12
6 11 10 11
5 10 9 10 Low
4 9 8 9
3 8 7 8
7 6 7
6 5 6
2 5 4 5
4 4 Very Low
1 3 3 3
2 2 2
1 1 1
0 0 0 0
The general descriptions given below explain the tendencies associated with each
Category. No two people score the same but it does help the trainee and the trainer to
consider how they can improve their learning by attempting to strengthen their
performance in the columns showing lower scores. Read through these and consider
where your own scores place you in these descriptions. Remember these scores
simply indicate the strength of your tendency in any one field.

7
LEARNING STYLES - GENERAL DESCRIPTIONS
Activists
Activists involve themselves fully and without bias in new experiences. They enjoy
the 'here and now' and are happy to be dominated by immediate experiences. They
are open-minded, not sceptical, and this tends to make them enthusiastic about
anything new. Their philosophy is, 'I'll try anything once'. They tend to act first and
consider the consequences afterwards. Their days are filled with activity. They
tackle problems by brainstorming. As soon as the excitement from one activity has
died down they are busy looking for the next. They tend to thrive on the challenge of
new experiences but are bored with implementation and longer-term consolidation.
They are gregarious people constantly involving themselves with others but in doing
so; they seek to centre all activities on themselves.
Reflectors
Reflectors like to stand back to ponder experiences and observe them from many
different perspectives. They collect data, both first hand and from others, and prefer
to think about it thoroughly before coming to any conclusion. The thorough
collection and analysis of data about experiences and events is what counts so they
tend to postpone reaching definitive conclusions for as long as possible. Their
philosophy is to be cautious. 'They are thoughtful people who like to consider all
possible angles and implications before making a move. They prefer to take a back
seat in meetings and discussions. They enjoy observing other people in action. They
listen to others and get the drift of the discussion before making their own points.
They tend to adopt a low profile and have a slightly distant, tolerant unruffled air
about them. When they act it is part of a wide picture which includes the past as well
as the present and others' observations as well as their own.
Theorists
Theorists adapt and integrate observations into complex but logically sound theories.
They think problems through in vertical, step by step logical ways. They assimilate
disparate facts into coherent theories. They tend to be perfectionists who won't rest
easy until things are tidy and fit into a rational scheme. They like to analyse and
synthesise. They are keen on basic assumptions, principles, theories, models and
systems thinking. Their philosophy prizes rationality and logic. 'If it's logical it's
good'. Questions they frequently ask are: "Does it make sense?" How does this fit
with that?" "What are the basic assumptions?" They tend to be detached, analytical
and dedicated to rational objectivity rather than anything subjective or ambiguous.
Their approach to problems is consistently logical. This is their mental set and they
rigidly reject anything that doesn't fit with it. They prefer to maximise certainty and
feel uncomfortable with subjective judgements, lateral thinking and anything flippant.
Pragmatists
Pragmatists are keen on trying out ideas, theories and techniques to see if they work in
practice. They positively search out new ideas and take the first opportunity to
experiment with applications. They are the sort of people who return from
management courses brimming with new ideas that they want to try out in practice.
They like to get on with things and act quickly and confidently on ideas that attract
them. They tend to be impatient with ruminating and open-ended discussions. They
are essentially practical, down to earth people who like making practical decisions
and solving problems. They respond to problems and opportunities as a challenge.
Their philosophy is “there is always a better way” and “if it works it's good”.

8
THE LEARNING CYCLE
To enable learning to be complete it is important that the learner is taken through the
full cycle. The learning cycle illustrated below (fig 1) roughly corresponds to the
types of activities described under the headings of Activist, Reflector, Theorist and
Pragmatist.
Fig 1
The Learning Cycle
The Experience
(Exposure)
# $
Plan the Review the

next Step experience
% &
Analyse and Draw conclusions

from the
Experience
The Activist is happiest in the experience

The Reflector is best when reviewing the experience
The Theorist works best when analysing and considering the issues
The Pragmatist likes to plan and get on with the next step
Obviously we all have our strengths and weaknesses and each learner’s style is a
combination of all four characteristics with some characteristics influencing
behaviour more than others.
For example
a) Someone who is a strong Activist and Pragmatist may be quick to pickup the
practical aspects in a learning environment but could experience difficulty if asked
to explain how something works, diagnose a problem or suggest changes in
procedures to improve performance.
b) Someone who is strong on Reflection and Analysis may not be able to pick up the
practical processes of a task quickly but would be more able to grasp and discuss
the underlying principles and their applications.
From these simple examples it will be seen that the trainer can provide a more
effective learning environment if trainees are encouraged to participate more fully in
those parts of the learning cycle which they normally tend to put less emphasis. It is
also important that the Teacher recognises that they also have to plan to give equal
emphasis to all four processes in the cycle irrespective of their own preferences.

9
Exercise
In the tables below two students’ leaning style profiles are given. You are teaching
both people at the same time. Consider these profiles and suggest what practical
approaches you may employ to improve individual learning.
Table 1 Table 2
Activist Reflector Theorist Pragmatist Preference Activist Reflector Theorist Pragmatist Preference
20 20 20 20 Very 20 20 20 20 Very
Strong Strong
19 19
18 19 18 19
17 19 17 19
16 19 18 16 19 18
15 18 15 18
14 17 14 17
13 18 16 17 13 18 16 17
12 17 15 16 Strong 12 17 15 16 Strong
16 16
11 15 14 15 11 15 14 15
10 14 13 14 Moderate 10 14 13 14 Moderate
9 13 12 13 9 13 12 13
8 8
7 12 11 12 7 12 11 12
6 11 10 11 Low 6 11 10 11 Low
5 10 9 10 5 10 9 10
4 9 8 9 4 9 8 9
3 8 7 8 Very Low 3 8 7 8 Very Low
7 6 7 7 6 7
6 5 6 6 5 6
2 5 4 5 2 5 4 5
4 4 4 4
1 3 3 3 1 3 3 3
2 2 2 2 2 2
1 1 1 1 1 1
0 0 0 0 0 0 0 0
Summary
This session has highlighted the importance of viewing each student as an individual.
It has also emphasised the teacher’s own learning style must be consciously
recognised and taken into account when planning and conducting teaching.
The linking of the concept of learning styles with the learning cycle demonstrates the
similarity of thinking in both these concepts. This similarity reinforce the idea that
teaching can and must be planned to ensure that the technical subject content and
structure and the teaching strategies and methods are assembled in such a way that
none of the parts of the learning process missed out or passed over.
This concept of structuring the learning process further reinforces the need for
teachers to structure material to satisfy the ways in which the student will have to
apply the knowledge and skills taught in their future employment and not being
restrained by the traditional structures of the subject being studied.
Conclusion
The principles established in this session must be used to reinforce more practical
sessions on planning, teaching methods and trainee participation.
These are:
The trainee is an individual and has to be seen in that context
Training methods must allow and encourage students to undertake the activities
identified in the learning cycle (telling and showing is not enough)

10
The technical content of learning sessions should be selected to meet the needs of
the trainee and not to satisfy the academic structure of the subject.

11
OHT 1
THE
EXPERIENCE
# $
PLAN THE REVIEW THE
NEXT STEP EXPERIENCE
% ANALYSE and
&
DRAW CONCLUSIONS
from the
EXPERIENCE

12
OHT 2
Personal Learning Style Profile Sheet
Circle your score from the score sheet for each category
This will show you which are your most and least preferred styles
Activist Reflecto Theoris Pragmati Preference
r t st
20 20 20 20
19
18 19
17 19 Very Strong
16 19 18
15 18
14 17
13 18 16 17
12 17 15 16 Strong
16
11 15 14 15
10 14 13 14
9 13 12 13 Moderate
8
7 12 11 12
6 11 10 11
5 10 9 10 Low
4 9 8 9
3 8 7 8
7 6 7
6 5 6
2 5 4 5
4 4 Very Low
1 3 3 3
2 2 2
1 1 1

13
14
Who is What
OHT 3

15
UNIT NINE
TITLE COURSE EVALUATION
UNIT AIM
To introduce a system of course evaluation which is simple to design and use and which can
give trainers useful feedback on which to base improvements in the session/programme.
UNIT OBJECTIVES
1 Identify the major areas in which formative feedback is useful to the trainer
2 Design a simple evaluation document for use in a course being taught by trainers
3 Use the evaluation, process the feedback data and make recommendations for
improvements.
Trainers notes
With any training course it is essential that it should be evaluated. A primary evaluation
criteria will be whether the trainees have achieved the objectives. However it is also essential
to have feedback from the trainees on how they found the course. A simple but useful
method is given in the UNIT text below. What is important is to find out if the trainees have
found the various parts of the course programme useful and where if any there has been parts
which have been less effective. By using the scheme given and particularly the conversion of
the results into a bar graph it is very easy to make comparisons between sessions and equally
importantly between the same course delivered to other groups.
The results of such evaluations are extremely useful in diagnosing where the trainers have to
review and improve units or sessions
Without such evaluation it is impossible to review or improve training. As part of the whole
programme it is important that every effort is made to monitor the training being given by
trainers you may be supervising.
Unit 9 1
COURSE EVALUATION
At this stage in the development of courses in HDM-4 for different groups it is important for
trainers to have early feedback on the groups opinion of the content of the course and how
useful they considered it to be to them in their future work
In the Train the Trainer Course you have just had, such a simple evaluation as given below
might be used. However, because of the duration of the course and large number of training
sessions it contained, many participants would have difficulty remembering all the sessions
and so a different method has been used. In this course the evaluation of each day of the
programme was carried out by the group feed-back sessions held each morning The process
was:
Each group was invited to provide a brief summary of the previous days programme as they
recalled it. This allowed the trainers to assess if the participants had generally grasped the
main concepts and issues relating to these sessions. Secondly groups were asked to highlight
any particular points that were either of particular interest to them or to draw attention to
areas where they felt they were not sure about or that they thought its delivery could be
improved. Tutors then summarised the group feed-back as they had perceived it and
discussed what action was needed.
This process is known as Formative evaluation as it allows participants and trainees to adjust
and change as the programme goes along. Summative evaluation is carried out at the end of a
programme and does not allow change to the existing programme but it can help with the
development of subsequent courses.
For the Summative evaluation on this course we are using two tools.
Forms A1 and A2 given at the end of this Unit allows us to make a comparison between your
entry level of expertise and your exit level as you perceive it. This information and your work
during the many practical exercises and sessions provides an accurate picture of the success
the programme has had in improving your skills and knowledge. However it is not an
examination and in most cases an examination approach is unnecessary and can be a negative
factor in adult courses.
To get a picture of individual trainee opinion on the programme it is possible to put out a
questionnaire that gathers opinion from each participant individually. The simple example
given below is one and for short courses this can be very useful. However as already
mentioned, in a three week programme with a wide range of topics and activities such a
method become less effective and produces large quantities of responses that may or may not
be significant.
In this case we are interested in using a method which will highlight major strengths and
areas in need of improvement and is in a form that is practical to manage. One method that
allows this filtering process to be used is the one outlined in Form C.
There is a wide range of evaluation and assessment tools available to the trainer and trainers
should choose those that best suit their needs. The main issue to be considered is that the
evaluation process in adult courses such as this one is not to test the participant. It is to ensure
that the training programme, its content and the methods used are continuously being
reviewed and adjusted to meet the learning needs of the participants.
Unit 9 2
EXAMPLE
Course evaluation for a short course
This approach may be useful in short course such as those being developed for “HDM-4 for
policy makers” (See Units Eight example programme)
Each Session or Topic received by the course participants is entered and participants provide
responses . This information will allow the training team to see which sessions have been
well received and which ones may need some attention.
From this information the trainer or training team will be able to consider where
improvements can be made to the content (and Methods) in future courses for groups with
similar needs.
Individual participant evaluation sheet

EXAMPLE
Course evaluation Course
We would like to know how useful the programme has HMD-4 for
been. Policy
Please complete the following, by putting a √ in the box makers
that indicates the level of usefulness of the topic to your
future work. Thank you. Date
22/06/00
Topic / Session Very Useful Not
useful very
useful
1. Overview of Highway Management
systems
2. Development of Computer
solutions
3. HDM-4 use and analytical
framework
4. Highway information and its use in
HDM-4 analysis
5. HDM-4 A planner’s tool
6. Adoption of HDM-4
7. Demonstration
Comments
Unit 9 3
SUMMARY OF THE INDIVIDUAL PARTICIPANT RESPONSE SHEETS
Enter the total number of responses taken from the individual participant sheets as below.
(with 20 participants each session line should add up to 20)
Session and Topic Very Useful Not
Useful Very
Useful
(3) (2) (1)
Session 1
Overview of Highway Management systems 7 12 1
Session 2
Development of Computer solutions 14 6
Session 3
HDM-4 use and analytical 17 3
Session 4
Highway information and its use in HDM-4 2 10 8
analysis
Session 5
HDM-4 A planner’s tool 11 9
Session 6
Packaging methods for specific products 16 3 1
Session 7
Adoption of HDM-4 19 1
Session 8
Demonstration 6 9 5
ANALYSING THE RESULTS

A simple analysis can be achieved by comparing the best result (session 7) with the other
sessions. It is clear that sessions 1, 4 and 8 need examination to consider what made them less
successful.
It is possible to make a graphical comparison by giving the score value in ( ) i.e. (3),(2) or (1)
for each entry in each column.
Example Session 6
Very useful =16 x (3) = 48
useful = 3 x (2) = 6
Not useful = 1 x (1) = 1
______
total = 55
If all the participants gave a "very useful" response then the total would be 60
The actual figure recorded in the example was 55 and this converts to 91% (55/60 x 100)
A graph of the results for the programme is below and indicates that session 4 needs attention
and sessions 1 and 8 may also benefit from some considered changes.
Unit 9 4
100
90
80
70
60
50
40
30
20
10
0
session session session session session session session session
1 2 3 4 5 6 7 8
Unit 9 5
FORM A
HDM-4 Train the Trainers Course
Kuala Lumpar. June 2000
Pre-course entry levels assessment sheet

Purpose: To give the Tutors a better understanding of your training needs. Please estimate
your current level of knowledge and experience (using / applying) the HDM 4 system and in
training others.
Name __________________________________ Country _________________
Date 06 June 2000

Level / performance indicator guide
Level 1 Sufficient to train others to all levels
Level 2 Fully competent in the knowledge and skills needed to operate
Level 3 Understand and competent in the basic principles and areas
Level 4 Awareness of systems / process and capabilities
Level 5 No previous Knowledge or Experience
Subject / topic area Knowledge of Experience of

1 2 3 4 5 1 2 3 4 5
Road Management
Earlier Road Planning Models (e.g. HDM-3)
New Models (e.g. HDM-4, DJIMS etc)
Modern Data Collection (including IQL)
Traffic effects and characteristics (including Speed Flow)
Road deterioration models
Road effects models
Social and economical effects models
Road investment and economics project analysis
Project analysis
Programme analysis
Strategy Analysis
Adapting HDM to your locality
Training need identification and course planning
Training methods and process
Conducting Training
Unit 9 6
FORM B
Post-course exit levels assessment sheet

Purpose: To give the Tutors a indication of how much you feel you have added to you entry
levels of Knowledge and Experience (using / applying) the HDM 4 system and training
others.
Name __________________________________ Country _________________
Date 23 June 2000

Level / performance indicator guide
Level 1 Sufficient to train others to all levels
Level 2 Fully competent in the knowledge and skills needed to operate
Level 3 Understand and competent in the basic principles and areas
Level 4 Awareness of systems / process and capabilities
Level 5 No previous Knowledge or Experience
Subject / topic area Knowledge of Experience of

1 2 3 4 5 1 2 3 4 5
Road Management
Earlier Road Planning Models (e.g. HDM-3)
New Models (e.g. HDM-4, DJIMS etc)
Modern Data Collection (including IQL)
Traffic effects and characteristics (including Speed Flow)
Road deterioration models
Road effects models
Social and economical effects models
Road investment and economics project analysis
Project analysis
Programme analysis
Strategy Analysis
Adapting HDM to your locality
Training need identification and course planning
Training methods and process
Conducting Training
General comments and observations
Unit 9 7
FORM C
Course evaluation exercise

Task A
Form into the Feedback Groups you have been in during the course.
As a group, review the programme and experiences you have had.
Try to select THREE areas/experiences that the group can agree were particularly positive
and useful to them for their future work as HDM-4 trainers.
Also select THREE areas/examples of things you would like to change in the course and how
you would suggest these changes could be made for the future.
Record and print out your answers
Task B
Gather together as a whole course group and appoint a chairperson to manage the Task
Consider the responses of the groups and select THREE positive and THREE areas for
change from all the group opinions which everyone can agree with.
Record and print out your answers
As a whole group, compose a short statement (100words) which you feel summarises the
group feelings about the course
Record and print out.
Unit 9 8
UNIT FOUR
TITLE: TRAINING NEEDS ANALYSIS
UNIT AIMS
1. To introduce the concept of Qualitative and Quantitative criteria as a basis for
defining training needs.
UNIT OBJECTIVES
1. State the five criteria to use when carrying out training needs analysis in a work
environment.
2. State why it is important to use a wide range of information sources when
conducting training needs analysis.
3. Demonstrate the ability to complete the tables 3 and 4 successfully within a
practical exercise in the programme
Trainers notes
The text follows through the process in a logical way and is developed from the
Training Cycle diagram provided. The example tables provided models for guidance
and it is recommended that training needs analysis is best largely done as a practical
exercise in the real work situation. The Concept of the Training Gap is a useful place
to begin and also it must be remembered that training needs have to be related to
specific target groups. The quality of a training programme will rely largely upon the
quality of the training needs analysis . An analysis done sitting in the office will
invariable become a work of half remember facts and fiction. The analysis is best
carried out in the work place and with direct observation of the work process and
contact with those doing the job.
Finally the “five rules of training” are a very useful framework to use. If the training
being recommended does not meet these rules then it likely that it will fail and be a
waste of time for trainer and trainee alike.
The OHT on "Training" at the end of the UNIT is a reminder that training is only one
part of the equation and in defining training needs it is always necessary to establish
why a particular process is being followed. The boxes around the training box in the
OHT illustrate other factor that may have to be addressed before training is
considered
1
INTRODUCTION
Training Needs Analysis (TNA) is a a means of establishing the difference between
the present levels of knowledge, skills and attitudes in the individual and what the
work place requires for efficient and effective performance. We refer to this
difference between the present knowledge, skills and attitudes and the desired
knowledge, skills and attitudes as the:
THE TRAINING GAP

(The difference between what is done now and what
will be achieved with better work methods.)
It is worth remembering that training does not take place in a vacuum and there is
little point in providing expensive training if the problem is related to lack of
equipment or money, management structures, or politics etc. However it is usually a
combination of issues in which the provision of training is part of the solution.
Ultimately the criteria for successful training is that the FIVE output criteria for
training must be applied. If training is to be judged successful then it must achieve
one or more of the following improvements
The work is done FASTER
The output is of HIGHER QUALITY
A GREATER QUANTITY is produced
LESS EFFORT achieves the same result
The production process COSTS LESS
Plus
Improved ENVIRONMENT ?
In the training needs analysis it is important to identify the Knowledge, Skills and
Positive Attitudes required to bridge the training gap identified. It is also important to
ensure that the trainer keeps the required balance between the needs of all three areas.
Training needs can be diagnosed by a combination of approaches. Some are:

1 First hand observation and information gathered from the work place.
2 Conversations with those who use or are affected by the output.
3 Conversations with those who work in overlapping or connected areas.
4 Conversations with colleagues.
5 Site reports on application and problems.
6 Research papers, official publications, planning etc.
7 Monitoring reports on implementation of decisions and planning
8 Evidence from people who are affected by the work output/product .
2
The above list illustrates the wide range of sources from which information can be
collected.
Exercise
Review the list above and draw up a list possible sources of information on training
needs for HDM4. Once you have done this rank them in order of descending order of
importance. When preparing your training programme you will have to make use of
these sources. Observing work in progress is just as important as sitting in a
comfortable chair having a discussion with the site and base office staff on their work.
All sources of information are important and useful but the most important ones are
direct observation of, and contact with, the person doing the job in their normal
working environments.
In good training it is important that:
• The information being given and the practical skills and methods introduced
are compatible with the current systems being used;
• That it fits within the financial and physical resources available;
• That it doesn't require equipment and materials that are not available.
PREPARING TO CONDUCT A TRAINING NEED ANALYSIS (TNA)

A preliminary estimate of training needs can be done by carrying out a desk top
analysis which should then be confirmed and refined through site/office visits and
interviews with the staff concerned.
In some cases, identification of training needs will be achieved through the use of job
description analysis. This requires the persons job to be broken down into individual
activities and duties and then each of these are examined individually to discover
where training may achieve improvements
TABLE 1 EXAMPLE
Preliminary Estimate Of Training Needs In Work Areas
Area PROCESS INDICATORS OF

Motorised-Traffic TRAINING NEED
1 Categorisation Placing vehicles in correct Confusion between
categories utilities and light trucks
2 Equivalencies Conversion of Confusion between PCU
heterogeneous to and PCSE use in MDM-4
homogeneous stream
3 calibration Inputs of different values Unbelievable outputs
4
5
3
EXERCISE (GROUPS)
Using Table one layout as an example, consider the HDM4 programme, identify a
main Area and some Processes. and try to describe the likely difficulties that the
target group may have. Select a target group and remember that if the target group
already have some experience with HDM3 or similar programmes, that experience
must be recognised in the training programme design
Task
Using this basic information make a first attempt at defining what difficulties the
target group might experience. Keep it basic and simple
Table 2 gives an example we can use to refine the process. This will require the
various processes associated with HDM4 to be accurately identified.
4
Table 2
POST: e.g. engineer, data collector, on site supervisor, policy maker, university student/
etc (choose only one category at a time)
Q: Do you/would you experience any problems with the following tasks/duties?
Tasks/Duties/ No Some Many Cause of Training

Responsibilities Problems Problems Problems Problems/ Needs
Comments
1 collecting traffic data
2 calibrating HDM-4
3 entering data into HDM-4
4 carrying out maintenance

plan design.
5 Interpreting HDM-4
financial reports
6
10
11
12
13
14
Example of an interview form showing some typical tasks/duties/responsibilities of a

field operator
5
Table 2
Position:
Q: Do you/would you experience any problems with the following tasks/duties?
Tasks/Duties/ No Some Many Cause of Training

Responsibilities Problems Problems Problems Problems/ Needs
Comments
1
10
11
12
13
14
6
OHT
ANALYSIS
OF THE
TRAINING NEED
! "
EVALUATION PLANNING
OF THE THE
TRAINING TRAINING
# CARRYING OUT
$
THE
TRAINING
7
Training does not take place in a vacuum

Consider if other factors are the reason for poor
performance
Policies
Organisation Politics
Motivation TRAINING Money
Management Rewards
Resources
8
Training should equal
FASTER
HIGHER QUALITY
GREATER QUANTITY
LESS EFFORT
COSTS LESS
(ENVIRONMENTAL IMPACT)
9
ASSESSING TRAINING NEEDS
(SOURCES OF INFORMATION)
% First hand observation and information
gathered from the work place.
% Conversations with those who use or are

affected by the output.
% Conversations with those who work in

overlapping or connected areas.
% Conversations with colleagues.
% Site reports on application and problems.
% Research papers, official publications,

planning etc.
% Monitoring reports on implementation of

decisions and planning
% Evidence from people who are affected by the

work output/product .
10
UNIT FIVE
TITLE: AIMS AND OBJECTIVES IN TRAINING
UNIT AIM
1 To introduce the concept of aims and objectives as a planning and evaluation tool
so that planning and teaching can be accurately matched to target groups and
training needs
UNIT OBJECTIVES
1 State the main parts of an aim and objective
2 Explain the relationship between trainer and aim and trainee and objective
3 Illustrate through example how the choice of verbs can indicate the level of
performance required and the possible training strategies to be used.
Trainers notes
This UNIT is a key area for the trainer. As it was once said "If you do not have a
destination any road will do" The purpose of Aims and Objectives is to allow the
trainer and the trainees to have a clear destination. The Aim is to guide the trainer and
the objective is for the trainee to achieve. The text supplied gives guidance on the
writing of aims and objectives. There is also a list of verbs which can be consulted to
indicate the level of performance required by the student As the verbs used in
objectives moves from Knowledge to Evaluation the required student performance
will become more difficult to achieve as each level (1 to 6) builds upon the mastery of
the level below.
It can therefor be helpful in planning training to ensure that the training objectives are
progressively scaled from 1 to 6 as the training programme progresses. A similar
framework exists for the practical skills list of verbs.
It should be remembered that aims and objectives for a complete programme will be
phrased in more general terms and will then become more specific as the planning
process moves to specific training sessions. While objectives are there for students to
achieve the trainer should also remember that testing objectives has to be realist in
terms of time and equipment needed.
Finally the verb in the objective will provide some indication of what teaching
approaches will be needed. For example if the student has to "select" then part of the
training session will be giving information on the criteria needed to be able to select
and to provide a range of what is acceptable and what is not. e.g. ripeness of fruit. . It
will also be necessary to provide a demonstration and student practice in the selection
process. The advantage of using objective is that the teacher knows exactly what to
teach for the student to achieve the performance stated in the objective.
1
AIMS AND OBJECTIVES IN TRAINING AND TEACHING
Aims and objectives are widely used in training and teaching and their use has made a
major contribution to the development of effective training in the last thirty years.
The aim describes your contribution, as a trainer, to the training process.
It states what subject matter you intend to teach and why you feel justified in having
this intention. This justification is as important as the intention and helps ensure that
you:
• select topics which are relevant to your trainees and
• teach appropriate aspects of these topics
• in a way which will enable them to make use of their learning.
As an example we might say that we intend:
To teach the specification and assignment of HDM-4 maintenance standards so

that trainees are able to apply a range of maintenance works.
In themselves the achievement of aims cannot be evaluated because there are external
factor which can influence the long term achievement of the Aim, you cannot
guarantee to achieve your aim. However you can test to ensure that all the component
parts of knowledge and skills have been mastered. The criteria for these tests are
specified in the objectives.
The objective describes what your trainee will be able to do which will demonstrate
that the teaching and learning process has succeeded. An objective based on new
abilities or behaviour is known as a behavioural objective.
A behavioural objective states:
• what the trainee will be able to do to show that learning has taken place;
• the conditions under which the trainee will demonstrate this ability e.g. time
after the training session, the location and limits on facilities or assistance;
• the standard which must be achieved.
In developing training programmes it is more normal to specify only “What the

trainee will do and leave the conditions and criteria until the training session is
planned in detail.
2
An objective for the aim we have already seen could be:
The Trainee will:
From appropriate sources, prepare input data required for investigating optimum
road maintenance standards for paved roads and enter into HDM-4 programme.
Assign a range of maintenance standards to road sections and run HDM-4.
Interpret the results from HDM-4 results and recommend optimum maintenance
Consider the alternative set of objectives below which you could apply to this case
What do you think are the main differences
Through the case study used trainees will
Review and comment upon the input data prior to entry.
Run HMD-4 successfully and achieve an output
Examine and discuss the key issues arising from the output.
Some guidelines in writing aims and objectives

They must be clear and unambiguous so that another trainer could work from them
and achieve the same level of results in the trainees.
The objective must be a statement of observable behaviour. To “understand” is not

acceptable, we must say what the trainee will “do” so that we know that they
understand.
Objectives must be realistic in terms of both what we can expect of a trainee at that
stage of training and what we will actually be able to observe and evaluate.
Some arguments in favour of the use of objectives in training
It is the most widely developed method of planning so it has been seen to work.
Trainers are encouraged to plan in specific detail.
Trainers are encouraged to make sure that their planning is relevant to the needs of
their trainees.
They encourage trainers to make explicit the values underlying their training.
They provide a rational basis for evaluating training and help in the achievement of
success.
Objectives provide a clear basis for assessment.
They assist in the communication of the content of training programmes to trainees

and other trainers.
3
They set goals for the trainees.
Objectives can be used in the development of individual training programmes.
They assist in the development of training which shows maintained or improving

standards.
Arguments against the use of objectives in training

Objectives have been widely and successfully used in vocational training and
education but some objections still exist. E.g.
There is no well-defined prescription for deriving objectives.
Objectives easily become 'set' and accepted as established, valid 'facts''.
Centrally set objectives may not be appropriate or practicable in all situations.
They may slow down the process of change and development in training.
The more superficial objectives are easier to define and assess e.g. recall rather than
evaluation.
Objectives tend to lead to an approach based on teaching discrete ideas (fragmented)

rather than one which recognises that the whole is greater than the sum of the parts.
Writing true behavioural objectives for some worthwhile educational goals is far from
easy e.g. in the affective (attitude) domain.
In a training session the use of objectives may limit the responsiveness of the trainer
to unexpected events and opportunities.
Defining objectives in advance restricts the discovery aspects of discovery teaming

although these may be the most valuable part of the exercise.
Objectives tend to lead to the view that there is one correct sequence for teaching a
subject and one ideal method.
These arguments do not invalidate our use of objectives. We should be aware of

these points and take them into account when we write and use objectives. We must
think about the way we use objectives and not simply follow them mechanically. Try
to consider the standards specified in the objective as the minimum standards which
all students must achieve. It is also possible to greatly exceed them and the trainer
should provide for this.
4
EDUCATIONAL OBJECTIVES IN THE COGNITIVE DOMAIN
This system defines in a hierarchical fashion different general, i.e. not subject-
specific, types of capability that may be developed in a student as a result of learning
experiences. These capabilities are described by reference to the actions that the
student may be able to perform which demonstrate the achievement of the various
levels of objective. The scheme therefore identifies and classifies the products of
learning experiences.
Category Capability and associated action verbs
1 Knowledge Ability to recall or recognise items of information, ideas etc.
define write underline state

recall select list identify
reproduce name label measure
2 Comprehension Ability to show understanding of ideas, e.g. by expressing ideas in

own words or transposing into alternate forms.
identify explain justify represent

judge illustrate select name
contrast indicate formulate classify
3 Application Ability to apply knowledge and ideas in order to meet the

demands of situations or problems which are new to the learner.
predict choose construct select

find compute assess show
use explain perform demonstrate
4 Analysis Ability to deal with situations or problems which demand more

than the straightforward application of ideas, i.e. situations where
elements have to be analysed and restructured before relevant
ideas can be applied.
analyse select justify identify

separate resolve conclude break down
compare criticise contrast differentiate
5 Synthesis Ability to put ideas together in new ways that reach beyond what
has been specifically taught
combine argue select restate

discuss relate summarise organise
precise derive conclude generalise
5
6 Evaluation Ability to judge the quality or value of material by reference to
appropriate criteria
judge support identify evaluate
defend avoid determine attack
select recognise choose criticise
EDUCATIONAL OBJECTIVES IN THE PSYCHOMOTOR DOMAIN
1 Imitation Observes skill and tries to repeat
repeats duplicates imitates copies

replicate reproduce recreate re-do
mimic emulate
2 Manipulation Follows instruction to perform skill
act follow manipulate handle

control use
3 Precision Reproduces a skill with accuracy
accurate exact precise correct

sure proper unerring perfect
4 Articulation Combines different skills smoothly
consistent harmony combines balances

flows continuing seamless
5 Naturalisation Effortless and automatic combination of skills
routine invariable constant automatic

unwavering consistent innate instinctive
spontaneous smooth
6
Learning Style Score Sheet
Activist Reflector Theorist Pragmatist Preference

20 20 20 20
19
18 19
17 19 Very Strong
16 19 18
15 18
14 17
13 18 16 17
12 17 15 16 Strong
16
11 15 14 15
10 14 13 14
9 13 12 13 Moderate
8
7 12 11 12
6 11 10 11
5 10 9 10 Low
4 9 8 9
3 8 7 8
7 6 7
6 5 6
2 5 4 5
4 4 Very Low
1 3 3 3
2 2 2
1 1 1
0 0 0 0
Participation equals
! Discussion
! Questions - answering and
asking
! Looking Observing
! Explaining Describing
! Problem solving
! Carrying out practical processes
! Sharing ideas, opinions and
experience
! LISTENING
3Participation.doc
Training does not take place in a vacuum

Consider if other factors are the reason for poor
performance
Policies
Organisation Politics
Motivation TRAINING Money
Management Rewards
Resources
1
Training should equal
FASTER
HIGHER QUALITY
GREATER QUANTITY
LESS EFFORT
COSTS LESS
(ENVIRONMENTAL IMPACT)
2
ASSESSING TRAINING NEEDS
(SOURCES OF INFORMATION)
! First hand observation and information gathered
from the work place.
! Conversations with those who use or are affected

by the output.
! Conversations with those who work in

overlapping or connected areas.
! Conversations with colleagues.
! Site reports on application and problems.
! Research papers, official publications, planning

etc.
! Monitoring reports on implementation of

decisions and planning
! Evidence from people who are affected by the

work output/product .
3
illusionunit7
What is it?
unit7
Where is the Ball?
Objectives
j - Cognitive
g Hierarchy
y
Evaluation Judge the quality or value of material by reference to

appropriate criteria
Synthesis Put ideas together in new ways that reach beyond what
has been taught
Analysis Deal with situations or problems which have to be

analysed and restructured before relevant ideas can be
applied.
Application Apply knowledge and ideas to new situations or problems
C
Comprehension
h i Expressing ideas in own words or transposing into
E
alternate forms.
Knowledge Recall or recognise

Objectives
j - Physcomotor
y Hierarchy
y
Naturalisation Effortless and automatic combination of skills
Articulation Combines different skills smoothly
Precision Reproduces a skill with accuracy
M i l ti
Manipulation F
Follows instruction to perform skill
Imitation Obser es skill and tries to repeat

Observes
The HDM-4 analytical framwork is based on the

concept of pavement life cycle analysis.
This is applied to predict the following over the life
cycle of the road pavement over, typically
i 15 - 40 yrs
Road deterioration
Roadwork effect
Road user effect
Socio-economic and environmental effects
Speaking
p g
I am happy I am p
puzzled
To be functional we have to simplify the feelings and

emotions we want to communicate. We do this byy
Deletion leaving out a lot of detail
Distortion simplifying otherwise it would become
tedious
Generalisation to avoid spelling out every detail
Target groups
? ?
?
?
Philippines
pp
• Climate : Tropical
Humid
• Roughness Age : 1.09
1 09
• Cracking Initiation : 0.833
• Cracking Progression : 1.2
• Roughness Progression: 11.22
• Factors from Tables 7.3-7.4 : 0.025,
1.2, 1.0
Roughness Comparison
18
16
14
12
10
IRI Before
IRI
IRI After
8
0
1998 2000 2002 2004 2006 2008 2010 2012 2014 2016
Year
Summary
• Calibrating HDM to suit the
environmental conditions in a
tropical-humid region or area
can p
predict the deterioration of
the road section under study.
• Calibrating the HDM, therefore,

will show the deterioration of
the road over time.
Indian factor values
F t
Factors: Roughness
R h Age(Kge)
A (K ) = 11.3 3
Roughness Progression(Kgp) = 0.7
crack initiation (Kci)= 1.5
15
crack progression( Kcp)=1.5
2020
2015
yee a r
2010
2005
2000
1
3
5
7
9
11
13
15
17
19
iri roughness
without vr with adjustment
20.00
15 00
15.00
10.00 iri values
5.00
0 00
0.00
0.00 5.00 10.00 15.00 20.00
Conclusions:
hdm4 default values
- predicts crack initiation early
- predicts faster road deterioration
3 years before the road fails (IRI
(IRI> 16)
Hence the factor values are to be changed appropriately

I
International
i lCCountries
i
• Climate: Tropical Humid
• Roughness age (Kgc) = 11.087
087
• Cracking Initiation (Kci) = 1.2
• C ki Progression
Cracking P i (Kcp)
(K ) = 0 0.83
83
• Roughness Progression (Kgp) = 1
RDWE calibration--Int’l
lib i I ’l Group
G
• Use Papua
p New Guinea Case
• Climate (Table 7.2):

7 2):
Tropical Humid
R
Roughness
h A
Age EEnvironment
i (Kgc)
(K )
• Environmental Coefficient (m) = 0.025
• Km (Table 7.4)
7 4) = 11.0
0
• meff = mKm = 0.025
• Roughness age environment (Kgc) =
meff/0 023 = 11.087
meff/0.023 087
C ki IInitiation
Cracking i i i (Kci)
(K i)
• High Bitumen Quality

• High Construction Quality
• Medium Oxidising Climate
• Cracking
g Initiation (Kci) = 1.2
C ki Progression
Cracking P i (Kcp)
(K )
• Inverse of the Kci
• Kcp = 1/Kci
• Cracking Progression (Kcp) = 0.83

R
Roughness
h Progression
P i (Kgp)
(K )
• No adjustment
j
• Roughness Progression (Kgp) = 1

IRI
18.00
16.00
14.00
12.00
10.00
8.00
International Group
6 00
6.00
4.00 IRI-calibrated
2.00 IRI--Base
0.00
1995 2000 2005 2010 2015 2020
Year
Wide Structural Crack
120.00
100.00
80.00
International Group
60.00
40 00
40.00
Calibrated
20.00 Base
0.00
1995 2000 2005 2010 2015 2020
Year
Summary
• Insignificant difference for IRI results
• Base Case/HDM-4 results are slightly
higher than the Calibrated Case for the
Wide Structural Cracking up until 2007

T3 Knowledge Base: Frequently Asked Questions
Q 1. How can Longitudinal Cracking be modelled?

Answer 1: Longitudinal Cracking is a special defect that normally occurs for specific reasons, e.g.
shrinkage or settlement of embankment or subgrade material, etc. Thus it would not be
considered to be a common pavement defect resulting from normal traffic loading. If
longitudinal cracking is a common occurrence on a particular road, then it could be
modelled as Transverse Cracks, since these also occur as a result of environmental
effects. The resulting area of cracking can be then be calibrated by adjusting the relevant
calibration factors to bring the estimate in line with local experience. An example of this
will be developed prior to the end of the course. In the future, adjustment of the model
coefficients and the guidance in Volume 4 is required. Alternatively, longitudinal
cracking can be taken into account by using a Miscellaneous Maintenance item as a direct
annual cost based on historical information.
Q 2. What is the difference between PCE/PCU and PCSE? Give examples

of the concept.
Answer 2: The concept of Passenger Car Equivalents or Units (PCE or PCU) and Passenger Car
Space Equivalent (PCSE) is to be able to represent the flow of traffic using a standard
unit. This is normally a passenger car. Note that there is essentially no difference between
PCE and PCU. The PCE/PCU represent the impact on traffic flow of different types of
vehicles, when compared to a standard car. For example, a lorry causes more disruption to
traffic flow than a car. The PCE/PCU represent the relative amount of disruption to traffic
flow of each type of vehicle (i.e. buses, lorries, tractors, animal carts, etc). The disruption
results from the size of the vehicles (static component) as well as the differences in travel
speeds (dynamic component). This is estimated from the sum of the effective area
occupied by each vehicle and the headway (or gap) between the vehicle and the next
vehicle. The PCSE concept is used primarily in HDM-III and in HDM-4. In this case,
HDM calculates the differences in speeds between vehicle types, therefore the PCSE only
represents the relative differences in space occupied by the vehicle types (i.e. static
component).
Q 3. How can PCE/PCU be converted to PCSE? Is this reliable?

Answer 3: There is no universal relationship between PCE/PCU and PCSE. The PCE/PCU values
change from one country to another as they include an element of driver behaviour.
However, in a given country, it is possible to conduct experiments to determine PCE/PCU
and also PCSE and thereby derive local conversion table. PCSE factors vary by road type,
and narrow roads have higher PCSE values than wide roads. Refer to Table B1.1 in
HDM-4 Volume 4 for typical values of PCSE by vehicle class and road type. Refer ro
Chapter 6.4 and Appendix E2 in HDM-4 Volume 5 for further information on PCSE and
PCE/PCU.
Q 4.
Answer 4:
HDM-4
IIntroduction
t d ti to t Program
P
Analysis
Obj ti
Objective
• To Prioritize candidate road sections

in each year within the budget
constraints.
Procedure
• Select road sections
• Determine maintenance / improvement
options
p
• Specify the budget limits and budget
p
periods
• Optimize using selected objective
oduce the
• Produce t e final
a list
st of
o projects
p ojects for
o the
t e
budget period
Multi-year Program
• Method:
– Life cycle analysis
– Budget period analysis
• Objective
Obj ti function:
f ti
– Maximize NPV
– Maximize road network condition
Budget allocation
• Routine maintenance costs are based on
road lengths.
• Recurrent costs are based on pavement
p
surface condition.
• Periodic maintenance are based on
economic indicators.
• Improvement needs are based also on
economic indicators.
Budgeting Procedure
• When sufficient budget is available:
– Select all independent projects with NPV>0
– Select
S l t mutually
t ll exclusive
l i projectj t alternative
lt ti
with the highest NPV.
• When
Wh there
th is
i budget
b d t constraints:
t i t
– Select the independent project using the
highest NPV/Cost ratio
– Select mutually exclusive projects using the
i
incremental
t l NPV/Cost
NPV/C t
Prioritization
A 150,000 35,000 0.233 2
B 150 000
150,000 45 000
45,000 0 300
0.300 1
C 200,000 40,000 0.200 3
D 250,000 37,000 0.150 6
E 500 000
500,000 70 000
70,000 0 140
0.140 7
F 500,000 77,000 0.154 5
G 500 000
500,000 78 000
78,000 0 156
0.156 4
Project
j Cost NPV NPV/Cost Rank
A 20,000 4,869 0.2435 2
B 30,000 240 0.0080 8
C 20,000 5,184 0.2592 1
D 20,000 210 0.0105 7
E 40,000 6,401 0.1600 4
F 35,000 4,497 0.1285 5
G 20 000
20,000 11,200
200 0 0600
0.0600 6
H1 80,000 15,000 0.1875 3
H2 100,000 16,829 0.1683
D fi iti
Definition
• Program analysis is concerned
with the preparation of single or
multi-year work program for the
road network by maximizing the
economic return (INPV/Cost)
subject
bj t to
t budgetary
b d t constraints.
t i t
Thank You
D Basic Requirements
D Limitations of HDM-
HDM-4
D Advantages and Benefits of

HDM--4 Use
HDM
Manual and Traditional
Techniques
Computer Based Techniques

z Computation Requirements
z Data
D R
Requirements
i
É Computer
486
486--DX with processing speed 100 MH
Monitor and other accessories
É Environment
Windows (95/98 or advanced)
È Road
R d N
Network
kDData
È Classification of Roads
È Section
S i Number,
N b
È Links Identification,,
È Nodes Number
Vehicle Fleet Data
È Vehicle name,
name Vehicle class,
class Base Type,
Type
Category
È Traffic Data
È Vehicle type, Characteristics, Fleet
Composition Volumes
Composition, Volumes, Growth rates
rates, Axle
loading, Speed flow type and traffic flow
p
pattern
È Pavement
P t Condition
C diti Data
D t
È Cracking, Rutting, Potholes, Edge Break etc.
etc.
È Pavement Ride Quality Data
È Roughness and Skid Resistance
ÈWork Standards
È Maintenance Standards
(items of work defined for maintenance)
È Improvements Standards
È Unit Cost Data
È Economic Unit Price
È Financial Unit Price
È Climatic Data
È R i f ll Temperature,
Rainfall, T t Moisture
M it level,
l l Freeze
F
Index etc.
È Environment
E i t Data
D t
È Emissions, day night temperature etc.
HOWEVER
HDM-4 HAS BULIT IN

HDM-
DEFAULT VALUES FOR THE
IMPORTANT NECESSARY
PERAMETERS
HDM-4 model are based on the research carried out in
HDM-
different countries.
È Models are not directly applicable to all situation
È Can be Made Compatible to Local Conditions
Calibration of Model
È Important Parameters such as
È Road User Effects (RUE) includes
VOC, Travel time, safety, emissions
È Road Deterioration and Works Effects (RDWE)
effect of works,
works maintenance standards,
standards pavement deterioration
Level of Calibration
È Level 1: Low Basic Application
È Level 2: Moderate Calibration
HDM--4 is not based on artificial intelligence system
HDM
È Model output are based on the information provided and
È Calculations of HDM-4 Model
Therefore
È Input data should be carefully collected and grouped
È Output should be carefully examine
È Iterations should be made until the model is reasonably
y responsive
p to
local conditions
Precisely
HDM-4 Model is applicable with suitable
HDM-
calibration and for the set of said
The software is easy to Use and User friendly
Evaluation of many criteria at a time is possible
The Model could be easily modified calibrated
and simulated for local conditions
Standard Reports could be easily generated
Relatively less processing time is required for
analysis
l i
Project, Program and Network Level of analysis
Prioritization of different strategies and works
can be
b carried
i d out
Comprehensive plans for development and
maintenance
i t with
ith optimized
ti i d budgeting
b d ti can be b
done
H l iin decision
Helps d i i making ki withith scientific
i tifi andd
logical reasoning
Th reliability
The li bilit andd confidence
fid level
l l for
f the
th
result is high
Li it d resources andd information
Limited i f ti isi needed
d d
The development is based on the feedback
If applied properly, the software can help in
reducingg the highway
g y budget
g byy allocatingg funds
to the projects which have high economic return
The software is fully compatible with the
existing data base.
The development of HDMHDM-4 4 is based on the
authenticated and documented research work by
leading research organization of world
The software and the support material can
provide the student the basic principals involved
in planning and development of highways and
road
The students may y be having
g advantage
g of usingg
the software earlier and can help in carrier
buildingg
The academic use of HDM-4 can assist the
student in the exposure of practical working
HDM - 4
HIGHWAY DEVELOPMENT & MAINTENANCE
ASSIGNMENT OF
MAINTENANCE AND
IMPROVEMENT OPTIONS IN
HDM-4 ANALYSIS
Presented by:
Md. Amir Azam
Executive Engineer
(Maintenance),
LGED, Bangladesh.
Assignment of Maintenance
and Improvement Options
in HDM-4 Analysis
AIM
To teach how to assign maintenance
p
and improvement options
p in HDM-4
analysis:
4 so that the trainees can select the base option
p
and the other options to compare with, in
HDM-4 analysis.
Assignment of Maintenance and
Improvement Options in HDM-4
Analysis
y
OBJECTIVES
* The trainees will:
* Describe different maintenance and
improvement options available in HDM-
4.
* Demonstrate
D t t their
th i ability
bilit to
t use those
th in
i
HDM-4 analysis answering related
q
questions.
Output Data in HDM-4
Analysis
The three main areas of applications of

HDM-4 are as follows:
* Project analysis - analyses studies.

studies
* Programme analysis - analyses
programmes.
* Strategy analysis - analyses strategies.
Input Data in HDM-4
Analysis
The input data is held in four folders:
* Road Network
- contains data defining the road network.
* Vehicle Fleet
- Contains the characteristics of the vehicle fleet operating on
th road
the d network.
t k
* Road Works Standards
- contains the specifications
p for maintenance and
improvement standards.
* HDM Configuration
- contains default data can be modified to reflect the local
circumstances.
Input Data in HDM-4
Analysis
Road Network
* The Road Network folder provides the basic
facilities for network referencing within HDM-4.
* It is central to the operation of HDM-4 .
* It allows users to define different networks
and sub-networks
sub networks, and section of roads
which are the fundamental unit of analysis in
HDM-4 .
Input Data in HDM-4
A l i
Analysis
Vehicle Fleet
* Representation of vehicles is key data
requirement for HDM-4 .
* Vehicle Fleet folder provides facilities for storage
and retrieval of vehicle fleet details udes for
calculation of:
- vehicle speed
- operating cost
- travel
t a el time cost
- other vehicle effects.
Input Data in HDM-4
Analysis
HDM Configuration
* It is recognised that HDM-4
HDM 4 will be used in wide
range of environment.
* HDM Configuration folder provides the facilities
to costomise system operation to reflect the
local conditions.
Input Data in HDM-4
A l i
Analysis
Road Works Standards

* It refers to the target or level of conditions that
a road administration aims to achieve.
* Road agencies set up different standards that
can be applied to meet specific objectives.
* It provides a list of maintenance and
improvement standards that are followed.
Works Standards
Section A

[[Operation
p 1]+[Intervention
] [ criteria]+[Limits]+[Effects]+[Cost]="Works
] [ ] [ ] [ ] item 1"
[Operation 2]+[Intervention criteria]+[Limits]+[Effects]+[Cost]="Works item 2"

Input Data in HDM-4
A l i
Analysis
HDM-4 Workspace
Input Data in HDM-4
A l i
Analysis
Reference of case study - Proj1.

Input Data in HDM-4
A l i
Analysis
Questions?
Assignment of Maintenance
and Improvement Options
in HDM
HDM-4
4 Analysis
Conclusion:
Selection of alternatives,, and maintenance
and improvement options to the road
pp p
network is central for appropriate
selection of work standards to:
improve the quality of services,such as:
⌧relieving traffic congestion
⌧road safety
⌧road passibility
⌧accommodating increased axle load.
Thank You All
Use of HDM-4 & Its
Analytical Frame Work
T
Target
t Group
G
Students of Final Year Civil

Engineering
Aim
• To give
T i an overview
i off Road
R d
management system
• To create awareness among

students
t d t off ththe use off HDM
HDM-IV
IV
in assisting road development
andd managementt
Why Do we need Roads?
The roads are economic arteries of anyy countryy

Air
Modal Split 1% Rail
14%
of Transport
Roads
85%
Hi t
History off Roads
R d
and
Why to Maintain our Roads

Wh t iis R
What Road
dMManagementt ?
It is a set of activities which leads To
optimise the overall performance of the

network over time
Road management
g activities
involves:
Setting priorities
Defining activities
Planning
Allocating resources
Organising and motivating personnel
Controlling work
Monitoring and evaluating
performance
Feeding back results to seek
improvements
Road Management
g Cycle
y
Planning
Operations Programming
g g
Preparations
ROLE OF HDM-4 IN HIGHWAY
MANAGEMENT
• Planning
• Programming
• Preparations
• Operations
O i
HDM--4 SYSTEM ARCHITECTURE
HDM

Analysis T l

N t
Network
k Fl t
Fleet W k
Works C fi .
Config

Models
systems
Strategies
RDWE : Road Deterioration and Works Effects

External
E t l
Databases, SEE : Safety, , Energy and Environmental Effects
PMS, etc. PMS : Pavement Management System
Implementation Stages
Data Decision Management

Database
Collection Support
pp Information
HDM--4
HDM Standard
& Custom
Reports
Integrating HDM
HDM--4 within Road
Management Systems
Long Term
Road Strategic
Inventory
DATABASE HDM--4
HDM Road Plan
Road Multi -Year

Condition Work
HDM4 Program
Traffic
Data
Detailed
Project Level
Bridges
Appraisal
Inventory
Bridges P li i
Policies
Condition Standards
Research
INPUTS OUTPUTS
NETWORK ELEMENTS
• Links
• Sections
– a length of road that is homogeneous in terms of
its physical attributes (basic unit of analysis)
• Network
• Sub-
S b-networks
Sub t k
PAVEMENT CONDITION
– Roughness surveys
– Road Distress Surveys
– Deflections
TRAFFIC DATA
•VVolume
l AADT (MT andd NMT)
• Traffic Composition
p and ggrowth rates
• Categories - Normal, Diverted,
Generated
• Axle loading
• Traffic flow pattern
• Roadside friction
Calibration To Local Conditions
• Vehicle Fleet
• Unit Costs
• Works Standards
• Model Calibration
• Language
Works Standards
• Works types
– maintenance
– improvement
p
• Unit costs
• Construction
C t ti quality
lit
– material properties
– construction quality
– compaction,
compaction etc.
HDM-4 ANALYSES
z Strategy analysis
z For preparing medium to long term
planning estimates of funding needs for
road network development & maintenance
z Programme analysis
z For preparing single or multi-year work
programmes underd budget
b d t constraints
t i t
z Project analysis
z For evaluating the economic or
engineering viability of different road
investment projects
PRIORITISATION
• Under budget constraints:
– Independent projects by NPV/Cost
– Mutually exclusive projects by incremental
NPV/Cost
• When there is no budget constraint:
– Independent projects by NPV or IRR
– Mutually exclusive projects by maximum NPV
Summary or Conclusion
• The engineers have the responsibility of
playing
l i an important
i t t role
l in
i building
b ildi
infrastructure
s uc u e
Modeling Non
Non--Motorized
Transport (NMT) in HDM
HDM--4
Md. Mahabubul Bari

Roads & Highways Department
Bangladesh
Importance of NMT in Asia and
Pacific Region
• 10 to
t 90 % off Traffic
T ffi ini China,
Chi India,
I di
Indonesia, and Bangladesh is NMT
• In Tokyo
Tokyo, about 30% of the total trip is
cyclist
• Wide range of intermediate modes of transport
between pedestrians and MT in Asia and hence
Transport
p systems
y in Asia are veryy efficient
NMT Vehicle Types in HDM-
HDM-4
• Pedestrians
• Bicycles
• C l Rickshaws
Cycle Ri k h
• Animal Carts
• Farm Tractors (not implemented)
Effects of NMT on Economic
Analysis
• MT Flow and Speed
• p
MT Operatingg Costs
• NMT Speed-Flow ( Partially Implemented)
• NMT Operating
p g Costs
• NMT Energy Consumption
• MT and NMT Safety Related Cost (NI)
• Road Deterioration and Maintenance (NI)
• NMT Travel Demand (NI)
Modeling Impact of NMT
• Impact of NMT on MT » MT Free Speed *

XNMT
• Impact on MT » Acceleration Noise
Operating Cost
• NMT Speed Model » Limiting Speed Model
• NMT Speed-Flow
p » Not Implemented
p
Model
• NMT Operating Cost » Structured
S Regression
g
Model
Computional Logic
Calculate
Average Daily Traffic
For each NMT Type
Calculate
Operating Speed
For each NMT type
Calculate
Time and Operating Costs
Perform
Economic Analysis
and
Energy Balance Analysis
Conclusions
• HDM-4 p provides some conceptual

p
framework for modeling NMT
• Needs further Modification, Calibration and
Validation with respect to Empirical Studies
INTRODUCTION
GOOP PROJECT DEVELOPMENT IS VITAL TO
EFFICIENCY IN STRATEGIC PLANNING AND
EFFECTIVE UTILISATION RESOURCES
Example of a Road Network
WHAT IS PROJECT ANALYSIS and
WHAT IS IT USED FOR ?
• It is an analytical framework based on the
conceptp of ppavement life cycle
y analysis.
y
• Applied to predict: Road deterioration, Road
work effects
effects, Road user effects & Socio -
Economic and Environmental effects.
• Road Pavement Deteriorate due to: Traffic
Loading, Environmental weathering and
Inadequate drainage systems
PROJECT ANALYSIS
•Economic Viability
•Engineering Viability
off R
Road
d IInvestment
t t
projects/alternatives/options
TYPES OF PROJECTS
› Maintenance
› Rehabilitation
› Widening
› G
Geometric
i improvement
i schemes
h
› Pavement upgrading
› New Construction
KEY FACTORS
Structural performance of road pavements
Life cycle predictions of road deterioration,
deterioration
road works effects and costs
Road user costs and benefits
Economic comparison of project alternatives
Structural Performance of Road
Pavements
ƪTYPES CONDITIONS
Basic
B i performance
f under
d ƪTraffic
different conditions over ƪClimate
ƪ
time ƪWeather
ƪDesign standard
ƪConstruction standard
Road User Costs vs Benefits
User Costs Benefits
>Fuel,
F l Oil,
Oil Lubricants
L bi t L
Lower V.O.C
VO C
>Tyres, Spare Parts Shorter travel time
>Travel time
> Vehicle
V hi l accidents
id t
Life Cycle Predictions
• Road deterioration
• Road work deterioration effects
• Road user deterioration effects
• Socio-economic and environmental
deterioration effects
Economic Comparison
Economic analysis for budget

optimisation over entire road
network
Section:
Alternative Net Present Internal Rate
Value of Return
(NPV = B - C) (IRR)
Base 0.000 0.000

RS5 8.28 28.9 (1)
RS3 8.01 22.2 (1)
RS7 7 41
7.41 35 3 (1)
35.3
RS9 7.02 43.5 (1)
RS11 5 95
5.95 52 5 (1)
52.5
HDM-4 Training
Preparing Road Section Data of

Bituminous Roads
for HDM
HDM-4
4
Target Group Profile
z Technicians
z 10 - 15 participants
z Grouped by Experience and

U E
Un-Experience
i
Aim
To teach the Types of Road Section

D t required
Data i d andd the
th Measurement
M t in
i
the format suitable for HDM-4.
So that, the data is accurate and can be
enteredd easily
il into
i the
h database
d b by
b the
h
computer operator in the office.
Objectives
z List the types of Road Section Data

required
i d for
f HDM-4
HDM 4
z Demonstrate how the data are
measured
z Differentiate (only for experienced
technicians) the new from the existing
measurementt techniques
t h i
Introduction
Road Section Data
Surfacing
Road Base
Data Base
Base HDM-4
///\\///\\
Subgrade
Introduction
Effects of Data Inaccuracy on Pavement Distress

14
12
km)
10
IRI (m/k
8
6
4
2 S f
Surface Treatment
Asphalt Concrete
0
2000 2005 2010 2015 2020
Y
Year
Introduction
Effects of Data Inaccuracy on Maintenance Costs
Total Agency Costs (US m$) Total Costs (US m$)
Asphalt Concrete 18.2 1408.6
Surface Treatment 22.1 1559.4

Other Types of Data Required
z Traffic Data
z Vehicle Fleet Data
z Road Condition Data
z Climatic Data
Types of Road Section Data
z Definition
z Road Inventory
z R d Geometry
Road G
z Pavement
z [Configuration Data]
Definition
z Workspace : Database grouped by Road Authorities

(eg National Road,
(eg. Road Provincial Road,
Road
, etc.)
z Network : Roads having the same road function
(eg. Arterial, Collector or Local)
z S ti
Section : Roads
R d in i one corridor,
id individual
i di id l roads,
d
parts of a road.
z Li k
Link : Grouping
G i Sections
S i
Road Inventory
z Pavement Type
z Length (m)
z Carriageway Width (m)
z Shoulder Width (m)
[also No. of Shoulder and Edge Step (mm)]
z Number of Lanes
Road Geometry
z Rise+Fall (m/km) and (no/km)

z Curvature (deg/km)
z Superelevation (%)
z Altitude (m)
z Drainage Type
z σadral
d l (m/s2)
Pavement
z Thickness of each Layer (mm)

z Construction Quality (CDS and CDB)
z Relative Compaction (%)
[ l for
[only f bituminous
bi i layer
l ???)]
z Subgrade CBR (%)
Pavement Classification System
y
1998 (STAP) 25 mm Surface

New Surfacing Dressing
1994 (AMAP) 50mm AC Overlay 50 mm AC

Overlay
100 mm
50 mm AC
Previous Surfacing Surfacing
Subgrade (CBR 8%)

Pavement type resets after
maintenance
i t works
k

treatment
R
Reseal
l STAP STAP / STAP STAP STGB STSB STAB STAP
STSB 1

AMSB 1
intermediate
y
surface layer
base
Source: NDLI (1995
Summary
z Identifying the types of Road Section

D required
Data i d for
f HDM-4
HDM 4
z Practicing the measurement of some
Road Section Data
z Recognizing other types of data
equ ed
required
Conclusion
z Accurate survey data in the format

suitable
i bl for
f HDM-4
HDM 4 will ill help
h l the
h
computer
p operator
p to enter the data into
the data base and will produce sensible
outputs.
outputs
H D M - 4
HIGHWAY DEVELOPMENT & MANAGEMENT
Types of Road Deterioration

Target group:
Senior government officials and politicians
involved in Highway Planning and
Management
AIM:
To introduce to the participants the different
types of road deterioration that will aid in
treatment selection as a basis for establishing
maintenance and improvement standards
through the use of HDM-4
OBJECTIVES:
The pparticipants
p will be able to find the basis of
selecting appropriate maintenance and
p
improvement standards.
The participants will be aware of the importance of
different types of road deterioration as a vital inputs
of HDM-4 in the selection of appropriate
maintenance and improvements standards
standards.
Types
yp of Road Deterioration
Overview of Road Types
• Gravel or earth roads (unsealed roads)

• Paved Roads
• Bituminous
• Cement Concrete
T
Types off Road
R dD Deterioration
t i ti
Assessment of Road Condition
• Road inventory
• Roughness survey
• Deflection survey
T
Types off Road
R dD Deterioration
t i ti

Road inventory
• Ratings (Politicians definitions)
• Good (G)
• Fair (F)
• Bad (B)
• Very Bad (VB)
T
Types off Road
R dD Deterioration
t i ti

Roughness Ratings (HDM-4 definition)
• Good - < 4 m/km
• Fair - 4-6 m/km
• Bad - 6-8 m/km
• Very Bad - > 8m/km
Types
yp of Road Deterioration
Factors Affecting Road Condition

• Traffic - no. of AADT’s or ESA’s
• Climate
Cli - rainfall,
i f ll moisture/temperature
i /
• Strength - structural no., deflection, etc.
• Drainage
D i - condition/quality
diti / lit
• Construction - type and quality
• Maintenance
M i t - type
t and d qualitylit
• Environmental factors
Road Deterioration
Paved Roads
Various distress modelled in the following order
• Cracking
C ki
• Ravelling
• Potholing Damaged Area
• Edge break
• Rutting
• Roughness
p
•Texture depth
• Skid resistance
Road Deterioration
Unsealed Roads
Various distress
• Potholes
P th l
• Materials loss
• Rutting
• Roughness
Course of Action
• Identify and select treatments
• Apply standards
• Estimate quantities and unit cost
• Run all aforementioned data inputs through
HDM-4 for economic analysis.
• Select an optimum standard.
Conclusions
HDM-4 provides the most appropriate standards in

which senior government officials will have the right
judgement in their policy-making decisions and
planning
p g strategies.
g
HIGHWAY INFORMATION & ITS
USE FOR HDM-4 ANALYSIS
presented by
ERWANTO WAHYUWIDAYAT
INDONESIA
for short (25 minutes) course to
International Graduate Student

of Engineering and Economics
Kuala Lumpur
20 June 2000
2/3/2014 Erwanto 1
OUTLINE OF PRESENTATION
• Aims and objectives

• The role of Information in highway
management
• Conceptual framework of HDM-4 Structure
• Type of Data required for HDM-4 Analysis
• Sample of Output of HDM-4
HDM 4 analysis & its
interpretation
• Summary
• Conclusion
• Discussion
2/3/2014 Erwanto 2
AIMS
To teach the student by providing basic
knowledge in making use of highway
information for HDM-4 analysis, so that
the student could interpret the result
appropriately
OBJECTIVES
• The student will be able to describe in
broad manner of the requirement of
highway information for HDM-4 analysis
• The student is expected to be able to
explain
l i the
th potential
t ti l power andd outcome
t off
HDM-4 analysis and its use for decision
making
2/3/2014 Erwanto 3
THE ROLE OF INFORMATION
IN HIGHWAY MANAGEMENT
• It is central in the management process

• It iis a key
k input
i in
i decision
d i i making
ki process,
inaccurate data will lead to wrong decision
• It is costly to manage, therefore it has to be
properly l designed,
d i d for f instance
i t using
i the
th
Information Quality Level Concept (IQL)
2/3/2014 Erwanto 4
CONCEPTUAL FRAMEWORK OF
HDM-4 STRUCTURE
2/3/2014 Erwanto 5
TYPE OF BASIC DATA REQUIRED
FOR HDM-4 ANALYSIS
• Road Data
• Road network
• Road geometry
g y
• Road condition
• Road strength
• Road Traffic and Vehicle Fleet Data

• Vehicle fleet (type of vehicle)
• Volume (including axle loading) of traffic
• T ffi Forecast
Traffic F t (growth)
( th)
• Road Works Data

• Type of works definition,
definition e.g.
e g routine,
routine periodic maint,
maint
reconstruction, improvement, development, etc
• Unit cost for each roadworks item
• Road User Effects Data

• Vehicle operating costs
• Time value
• Environmental,
i l Socio-economic
S i i & Energy impact
i
2/3/2014 Erwanto 6
SAMPLE OF OUTPUT OF HDM-4
ANALYSIS & ITS INTERPRETATION
2/3/2014 Erwanto 7
SUMMARY
• Highway
g y information is one of the
key factor in the decision making
process within road transport sector
management
• HDM-4 is a tool that could be used

at either
i h strategic,
i program andd
project level analysis of road
investments
• HDM-4 is an expert system, it has to

be used by an expert user to ensure it
will yield to a proper output and
recommendation
2/3/2014 Erwanto 8
CONCLUSION
• Road administration has to

build a proper designed
highway information for their
maximum use of its asset
management
• HDM
HDM-44 is
i a tool
t l that
th t could
ld assist
it
road administration and policy
maker in managing their road asset
i the
in h better
b way
2/3/2014 Erwanto 9
THANK YOU
2/3/2014 Erwanto 10
Wh M
Why Manage??
Resource is Limited.
Limited
Maximise Benefits with Limited Resources

Conceiving is Easy
Managing
g g is Difficult
Budget
g Allocation
for
R dM
Road Maintenance
i t
Role of HDM
HDM-4
4
in
R dN
Road Network
k Pl
Planning
i
Process
Role of HDM-4 in
Road Network Planning Process
Outline
• Network Management Process
• The Planning Process
• Performance Indicators
• HDM-4
• Role of HDM-4
Network Management Process
PLANNING
PROGRAMMING
PREPARATIONS
OPERATIONS
Planning
Analysis
y of the road system
y as a whole.
Road Network Development

p and Preservation
under
various budget and economic scenarios.
Pre-Requisites
The Planning Process
Collected Data
Required
What-If Resources?
Strategies
Analysis
Benefits?
The Planningg Process
with HDM-4
Collected Data
What-If
What-
Required
q
Analysis Resources?
Strategies using
HDM--4
HDM Benefits?
Benefits
• Allows modelling and prediction of future
road and traffic conditions.
• Enables testingg of various road
improvements and maintenance strategies.
• Provide objective method to estimate
budget requirements over several years.
• Better estimates of Annual Maintenance
Requirements.
– Planned
Pl d Maintenance
M i t after
ft Conceiving
C i i
Summary
• 4 Stages in the
– Network Management Process
• Planning Stage is Critical
– What if no planning?
• Performance Indicators
– NPV = Benefits
fi - Cost
Conclusion
Role of HDM-4
A Strategic g Planningg Tool

to facilitate testing of various Maintenance
and Improvements works
works.
Finally
Your engineers requested $X billion for next

year.
What would you say?

THANK YOU
INTRODUCTION
ROAD INFORMATION SYSTEM

ROAD MANAGEMENT PLAN
HDM4 - A DECISION SUPPORT TOOL
FETCH DATA FROM ROAD
INFORMATION DATABASE
ADAPTATION TO LOCAL
CONDITION
MANAGEMENT PLAN
HDM4
A DECISION
MAKING TOOL FOR
MANAGEMET
TRAFFIC MAINTENANCE ALTERNATIVES ENVIRONMENT

VEHICLE COST ROAD STRUCTURE
ECONOMIC WORK STANDARD CONDITION
FINANCIAL INTERVENSION GEOMETRY
HDM4 DATA VOLUME
DATA REQUIREMENT IS LARGE
159 FOR ROAD SECTION
160 FOR VEHICLE FLEET
MAINTENANCE DATA
MOST OF THE VALUES ARE DEFAULT
SOME ARE COMPULSORY INPUT
TRAFFIC DATA INFLUENCE VOC,
INTERVENSION
WORK STANDARDS
ROAD SECTION DATA
TRAFFIC FLOW PATTERN
SPEED FLOW TYPE
ROAD TYPE
DIRECTION
SPEED LIMIT
ENFORCEMENT
SPEED REDUCTION FACTOR
FRICTION FACTOR
NMT
TRAFFIC FLOW DATA
TOTAL TRAFFIC VOLUME
VEHICLE COMPOSITION
HOURLY FLOW
DIRECTION OF FLOW
AADT FROM PAST RECORDS
GROWTH RATE
FIELD SURVEY
PAST RECORDS
MANUAL AND OTHER METHODS
VEHICLE FLEET
CATEGORY MOTORISED, NON MOTORISED

BASE TYPE MC, SMALL CAR, MED CAR, LARGE CAR,
( MOTORISED) LIGHT DELEVERY, LIGHT GOODS, FOUR
WHEEL DRIVE, LIGHT TRUCK, MEDIUM
TRUCK, HEAVY TRUCK, ARTICULATED
TRUCK,, MINI BUS,, LIGHT BUS,, MEDIUM BUS,,
COACH
NMT PEDESTRIAN, BICYCLE, RICKSHAW,
ANIMAL CART
CLASS MC, CARS, UTILITIES,TRUCKS,BUSES
VEHICLE CHARACTERISTICS
VEHICLE NAME
VEHICLE CATEGORY
VEHICLE CLASS
BASE TYPE
VEHICLE PCSE
FREE SPEED
FIELD SURVEY
AVARAGE OCCUPANCY
ECONOMIC AND FINANCIAL
COST
VEHICLE
TYRE
TYRE RETREADING
CREW WAGE
MAINTENANCE LABOUR
FREE SPEED
FIELD SURVEY
AVARAGE OCCUPANCY
ECONOMIC AND FINANCIAL
COST
INTEREST RATE
FUEL TYPE AND COST/LIT
LUBRICANT
AVERAGE ANNUAL KM RUN
AVARAGE ANNUAL HOUR RUN
% OF NON HOME WORK TRIPS
VALUE OF PASSENGER WORK AND NON WORK
TIME
VALUE OF GOODS HOLDING TIME
AXLE LOAD
TO TRANSFER VEHICLE LOAD TO
STANDARD AXLE LOAD
NOTE VEHICLE NAME, CATEGORY,
TYPE AND NUMBER OF AXLES
BRING ONE WHEEL OF EACH AXLE
OVER THE WEIGH BRIDGE AND
RECORD
AXLE LOAD (Cont..)
(Cont )
TRANSFER DATA TO A FREQUENCY
TABLE
GET DAMAGE FACTOR FOR EACH
VEHICLE
GET TOTAL AXLE LOAD FOR EACH
AXLE
GET EASL FOR EACH VEHICLE
TYPE
XLE LOAD SURVEY TABLE -1
VEHI AXLE- AXLE- AXLE- AXLE AXLE- AXLE-
CLE 1 2 3 4 5 6
BUS 2 4
(L)
TRUC 2.5 5.4
K(L)
BUS 2.5 6.5
(H)
TRUC 3.0 6.0 5.6
K(H)
SPEED MEASUREMENT
RADAR SPEEDOMETER
SELECTION OF SAMPLE VEHICLES
DIFFERENT HOURS OF SURVEY
AVOID OVERTAKING SITUATION
EXPERIENCED TECHNICIAN
REQUIRED FOR SPEED FLOW CURVE
CONCLUSIONS
COLLECTED DATA ON TRAFFIC AND
VEHICLE IS COMPULSORY INPUT TO
HDM
RELIABILITY OF MANAGEMENT
PLAN DEPENDS UPON ACCURATE
DATA
AADT, GROWTH RATE, EASL
INFLUENCE VOC, INTERVESION
STANDARDS FOR UNSEALED
ROADS IN MONGOLIA
Aims and Objectives
z Aim - To teach the importance of unsealed roads in

Mongolia, their classification and performance and introduce a
common approach to standards for central planners and
regional directors.
z Objectives
1.Explain the contribution of unsealed roads in the national
network.
network
2.Describe the main classifications and characteristics of each
recognised type.
3 Explain the relationship between policy objectives and
3.
maintenance actions
HOW IMPORTANT ARE UNSEALED ROADS?
Key Statistics for Mongolia
z >75% of main roads are unsealed

z 77% of population live in rural areas
z 65% of national GDP is agricultural
z 70%
% off the population work in agriculture
z 85% of national exports are agricultural
CLASSIFICATION & CHARACTERISTICS ?
z GRAVEL
z IMPROVED EARTH
z UNIMPROVED EARTH
MONGOLIA - STEPPE TRACK

access
desirable

tasks
heavyyg grading
g and g
within budget
lengthmen.
services
bridges and rivers
works.

Standard A B C
Desirable 1,689 1,123 557
R d
Reduced d 1 132
1,132 634 406
Target Gro
Group
p
AIM:
To teach the approach towards strategy

analysis
y of a road network
So that engineers can carry out strategy

analysis of a network and interpret the
results
Objectives
j
The Engineers will:

explain the difference between strategy
and
d program analysis
l i applications
li ti off
HDM - 4
describe the criterion to be used and

Interpret results from sample output
Introduction
Strategy Analysis deals with the concept of
strategy planning management, whereby road
network
t k expenditures
dit are off medium
di to
t long
l
term.
This requires that a road organization should

consider the requirements
q of its entire road
network.
Thus, strategy analysis deals with entire

networks or sub-networks managed by one road
organization.
Introd ction
Introduction
Which staff are responsible in the strategy
planning stage?
Technicians,, senior management
g and policy
p y
makers or junior and middle level
professionals
Of course the senior management, however
engineers will be the ones contributing to the
selection
l i off standards
d d and d the
h iimputation
i off
appropriate parameters.
I t d
Introduce th
the D
Development
l tP
Points.
i t
Development Points
1.Differences in Strategy and Program Analysis
applications
2M i P
2.Main Procedure
d iin St
Strategy
t A
Analysis
l i
3 Th A
3.The Application
li ti off St
Strategy
t A
Analysis
l i
4.Interpretation of Results from Sample Outputs

Development Point 1
Sections are a basic unit of analysis in HDM-4.
Both strategy and program analysis have this

however, program has an identifiable road
section which has a physical link.eg Man 203
185-190. In strategy, it is a representative
section
ti which
hi h can be
b determine
d t i from
f a
prepared matrix.eg PHTGC - Paved road with
High Traffic in Good Condition
Refer to next slide for a matrix example.
Development Point 2
Create representative road network matrix
Define representative traffic volume and

loading
g
Model Pavement Deterioration for each
matrix cell
Assign the maintenance and improvement
standard
Run it to determine road user benefits
Select standard which optimizes user benefits

Development Point 3
Medium to long term forecast of funding
requirements for a specified target road
maint.
i t Standards.
St d d Refer R f tot Graph
G h1
Long term network performance with different
levels of funding
funding. Refer to Graph 2
Optimal allocation of funds according to
budget heads
heads,eg.eg Routine
Routine, period etc
etc.
Optimum allocation of funds to sub-network,
eg National Route
eg. Route, National Main etc
etc..
Policy Studies such as impact of changes to
axle load limit
limit, maintenance standards
energy balance and analysis, etc..
Effect of budget levels
Annual
7.0 Budget
50%
6.0
ghness
80%
5.0
R oug
100%
4.0
Target
g
3.0
1998 1999 2000 2001 2002 2003 2004
Summary
W have
We h just
j covered:
d
1.The differences in strategy and program

analysis applications,
2.The main procedures in strategy analysis,
3.The application of strategy analysis,and
4.The interpretation of results from sample

Concl sion
Conclusion
Strategy analysis is important to engineers as
they will the ones to present results policy
makers. Budget for program and project
analysis will very much depend on this.
We have just covered strategy analysis which
is one of the three types of analysis carried
out by HDM-4. Program analysis was briefly
mentioned when I was comparing it to
strategy analysis. In my next session I will
cover in detail program analysis plus program
An Overview of Highway
Management System & Its Related
P bl
Problems
June 2000
DIFFERENT TYPE OF ROADS
• Flexible Pavement
• Composite
p Pavement
• Rigid Pavement
TYPICAL X-
X-SECTION OF ROAD PROJECT
13.300 M
3.000 M 7.3 M 3.000 M

SHOULDER CARRIAGEWAY SHOULDER
C
L
4% 2% 2% 4%
AC SURFACING
AC BASE
AGGREGATE BASE
SUB BASE
What is Road Management
• Purpose: To optimise the overall

pperformance of the network over time
• Goal: To improve the development and

performance of the road sector and its use
of resources, within the context of national
objectives, in order to minimise transport
costs.
Road Management
Road Management involves:

– Settingg priorities
p
– Defining activities
– Planning
– Allocating resources
– Organising and motivating personnel
– Controlling work
– Monitoring and evaluating performance
– Feeding back results to seek improvements
THE IMPACTS OF ROAD MANAGEMENT
The effects of road management can be

assessedd iin terms
t off the
th various
i impacts:
i t
• Level of service (road condition)
• S i economic
Socio i impacts
i
• Road user costs
• A id t levels
Accident l l andd costst
• Road administration costs
Components of Total Road Transport
Costs in the Economy y
15%
50%
Maintenance
- Fuel Consumption
- Oil Consumption
- Tyres Wear & Tear
- Parts Replacement 10% 25%
- Vehicle Depreciation
- Travel Time
- Accidents
Key Statistics from Emerging Countries
• Transport sector is 5 to 10 per cent GDP
• Paved roads account for 12% of length and 90% non non-
urban traffic
• In some countries,
countries actual expenditure is 25% of need
• Every $ not spent on maintenance increases transport
costs
t by
b 4$
• Surfacing lives vary from less than 5 to 10 years
• Transport fuels and machinery account for 40% of
p bill
import
Weaknesses in Road Management
• Inadequate and un-secure funding

• G t priority
Greater i it given
i to
t new roadsd
• Poorly motivated organisations
• Weak or non existent management systems
• Inappropriate treatments and lack of quality
control
• Poor regulations and/or enforcement
• Equipment failures and lack of spares
Pavement Performance
Pavement performance depends on:

– Traffic volumes and loading
– Road pavement strength
– Maintenance
i standards
d d
LONGITUDNAL CRACKS
BLOCK CRACKING
SEVERE FATIGUE CRACKS WITH PERMANENT
DEFORMATIONS
Over Loaded Truck
Over Loaded Truck
Over Loaded Truck
Over Loaded Truck
Over Loaded Truck
Over Loaded Truck
Over Loaded Truck
Over Loaded Truck
Truck Damage Factors in Pakistan & USA
Truck Axle Configuration Truck Truck
Type Factor Factor
P ki t
Pakistan USA
2-Axle Both Single Axles 4.67 0.21
3-Axle One Single & One Tandem Axle 8.84 1.59
4-Axle All Single Axles 12.99 1.32
T
Two Single
Si l & One
O Tandem
T d
4-Axle 10.35 1.32
Axles
One Single
g & Two Tandem
5
5-axle
l 14 73
14.73 1 39
1.39
Axles
One Single, One Tandem & One
6-Axle 10.90 1.39
Tridem Axle
Network Roughness Condition
Good (<3.5 IRI) *
28%
Very Poor
(IRI>7.0)
47%
8%
Fair (3.5<IRI<5.0)
17%
7365 km
two-lane
equivalent Poor
network
t k (5 0<IRI<7 0)
(5.0<IRI<7.0)
Network Condition
RIDE QUALITY
NETWORK
Fair (8.0%)
Good (28.0%)
Poor (17.0%)
Very Poor (47

(47.0%)
0%)
Source: World Bank Study 1998

1998 NETWORK ANALYSIS - REAMINING SERVICE LIFE
4 53% of Network has Remaining Service Life of Zero Years.
4 29% of Network has Remaining Service Life < 3 Years.
RSL > 5
(13%)
RSL 3 - 5
651
(5%) 229
Kms
K
Kms
2706
1483 Kms
Kms RSL <=
< 0
(53%)
RSL 0 - 3
(29%)
RIDE QUALITY & ROAD USER COST
30
25
20
15
10
0
1000 2000 3000 4000 5000 6000 7000
Cars Buses Art. Trucks

ROAD ROUGHNESS (MM/KM)
Funding Status
Demand Allocation and Releases (Development Budget)
Demand,
(1991-92 to 1999-2000)
40
35
30
on
25
5
Rs. Billio
20
15
10
0
Year 1991-92 1992-93 1993-94 1994-95 1995-96 1996-97 1997-98 1998-99 1999-
2000
Demand
Allocation Years
Releases/Utilization
Funding Status
Demand, Allocation & Release
(Maintenance Budget)
7000
6000
5000
Rs. Million
n
4000
3000
R
2000
1000
0
8
9
8
00
-8
-8
-9
-9
-9
-9
-9
-9
-9
-9
-9
-9
87
88
89
90
91
92
93
94
95
96
98
-2
97
99
19
19
19
19
19
19
19
19
19
19
19
19
19
Demand
Allocation
Received Year
Minimum Budgetary Requirement
SCENARIO TO MAINTAIN 1998 CONDITION; MIN. 2-3 BILLION RS PER YEAR *
100%
85
8.5
90%
80%
7.5
70%
ughness (IRI))
ondition (%)
60% 6.5
Poor
50% Fair
Average Rou
Network Co
5.5 G d
Good
Avg.. IRI
40%
30% 4.5
A
20%
3.5
10%
0% 2.5
1998 1999 2000 2001 2002 2003 2004
Year
ORK ANALYSIS - ECONOMIC LOSS OF NON MAINTENANCE
4 Total Economic Loss : Rs. 38 Billion
4 Economic Loss due to N-5 Rs. 26 Billion
Economic VOC Loss Per Year

(Rs. in Billion)
N-70
N-65
N-55 1.91 1.56

0.37 4.25
N-50
0 59
0.59 N-40
N-35
0.88 2.21
26.43
N-25
N-5
What SOLUTIONS can we suggest ?
Establish a dedicated and secure Road Fund
Prioritise on a rational and equitable basis
Adopt affordable standards
Implement appropriate management and
quality systems
Address transport management, regulation and
enforcement issues
Seek to improve road user behaviour and
vehicle
hi l standards
t d d
Overview of Development
of Computer-Based HDM
M d l and
Models d
Development of
HDM-4
HDM 4 Software
Siriphan 1
Manual Calculation
? + - * / % $
$ NPV ?
!!! IRR ?
NPV/C
/C ?
Siriphan 2
EXCEL-- Spread Sheet
EXCEL
Siriphan 3
User Friendly -- HDM
HDM-4
4
Siriphan 4
Objectives
z To Provide Overview of Past

Development of HDM Models
z T Enhance
To E h th
the K
Knowledge
l d
Base of HDM Capabilities:
Capabilities:-
• Basic Capabilities: previous HDM
• Advanced Capabilities: HDM-4
Siriphan 5
Development of HDM
Models -- on Mainframe
z HDM (1979): Highway Design

and Maintenance Standards
Model
z HDM-III (1987): HDM version III
Siriphan 6
Development of HDM
Models -- on PC
z HDM-PC (1989): A Micro Computer
Version of HDM-III
z HDM-Q ((1994):
) Incorp.
p Traffic
Congestion Effect
z HDM Manager (1994): Providing a
Menu-Driven Front End
Siriphan 7
HDM 4
HDM-4
Highway Development
and Management
Tool
(since 1995)
Siriphan 8
Basic Capabilities of
HDM Models
z Road Transport Appriasal/

I
Investment
t t Model,
M d l iie., NPV
NPV,
IRR,, B/C
/C
z Road Work Costs
z Vehicle Operating Costs
z Road Deterioration Model
Siriphan 9
Additional Capabilities
to HDM-4
z Traffic Congestion Effects

z Cold Climate Effects
z A Wider Ranger of Pavement
Types and Structures
z Road Safety
z E i
Environment
t Effects
Eff t
Siriphan 10
Summary
z Why do we need to develop the

computer based HDM models?
computer-based
z What can previous HDM
Appraisal Models perform?
z What can HDM-4
HDM 4 Management
Tool carry out?
z What is the HDM-4?
Siriphan 11
Conclusion
HDM-4 Application Software:

z better capabilities on modellings
techniques
z better
b tt representedt d highway
hi h costs
t
and deterioration models.
Next Session:
z more p
powerful analytical
y
framework
Siriphan 12
THE ROLE OF HDM-4
HDM 4 TO JUSTIFY
THE ROAD NETWORK INVESTMENT
POLICY
TAUFIK WIDJOYONO
KUALA LUMPUR
What is the real problem ???
Road deterioration
Budget short coming
Investment priority
p y
Regional budget
allocation
Budget accountability
Introduction
The Role of HDM
HDM-4
4 as a tool of Decision Support System
Identify key factor affecting road network performance

Economic Parameter relate to user cost
Economic parameter related to investment
Budget
g priority
p y based on benefit cost ratio
Determine Budget Resource Distribution based on economic
judgement and regional development
Recommendation to the Improvement of Existing Policy
Factor Affecting Road Performance
Roughness of the surface

Traffic volume
Level of service
Environment factor
Economic Factor related to user
cost
Vehicle operating cost

(vehicle, maintenance and spare part)
Time value
Economic parameter to determine
road investment
Road network performance

p
Treatment selection
(periodic maintenance, minor improvement, major
improvement include upgrading)
Treatment unit cost
Selection criteria
Discounted
Di t dBBenefit
fit andd costt ratio
ti
NPV
NPV/Cost
EIRR
Prioritization regarding regional
development
Economic consideration
(generated traffic )
Social Consideration
(increasing accessibility)
Towards Better Policy
Classify road networks based on the degree

of utilization
Determine suitable IRR
Determine suitable level of treatment
(bituminous,
bituminous gravel)
THANK YOU
PAVEMENT STRUCTURAL
CONDITION DATA AS
INPUT TO HDM-4
FLEXIBLE PAVEMENT
COMPONENTS
SUBGRADE
SUB-BASE
BASE COURSE
SURFACE COURSE
INTRODUCTION
PAVEMENT TYPES
FLEXIBLE PAVEMENT
SEMI-RIGID PAVEMENT
RIGID PAVEMENT
PAVEMENT STRUCTURAL
CONDITION EVALUATION
TYPES OF EQUIPMENT
COMPONENTS OF THE BENKELMAN
BEAM
PRINCIPLE OF REBOUND DEFLECTION
METHOD OF STRUCTURAL
CONDITION EVALUATION
EQUIPMENT FOR DATA
COLLECTION
• LOADED TRUCK
Wheel Load – 8160 kg.
Tyre
T P
Pressure – 5.6
5 6 kg/sq.cm
k /
Benkelman beam with dial gauge
Thermometer
Glycerol
Pavement Cutting Tools
Rapid Moisture Meter
Tyre Pressure Gauge
SELECTION OF HOMOGENEOUS
PAVEMENT SECTIONS
PAVEMENT SURFACE CONDITION DATA
CRACK AREA
PATCH AREA
RUT AREA
IRC PROCEDURE FOR
DEFLECTION
C O MEASUREMENT
AS
MARKING ON PAVEMENT SECTIONS

PROCEDURE OF DEFLECTION STUDY
NUMBER OF DEFLECTION OBSERVATION
POINTS
DATA FORMAT
• Chainage: DATE:
• Km: from ……….. to ………
• Enumerator:
• Wh l Path:
Wheel P h
• Time of Start:
• Time of Completion:
• Pavement Temperature:
DATA FORMAT (Contd.)
Chainage Do Di Df D
41.400
41 420
41.420
41.440
MARKING OF PAVEMENT SECTION
SECONDARY DATA
SUBGRADE MOISTURE
PAVEMENT TEMPERATURE
PAVEMENT COMPONENT LAYERS
SUBGRADE SOIL TYPE
ROAD DETERIORATION MODELS.
KEY VARIABLES AFFECTING

DETERIORATION OF
BITUMINOUS PAVEMENT.
AIMS AND OBJECTIVES
z SESSION AIMS
To teach the key variables affecting road deterioration modeling
in HDM-4.
To make the participant aware of the key reference document
z SESSION OBJECTIVES
To outline the principles behind modeling of bituminous pavement
deterioration.
To identify the various factors which affects bituminous pavement
performance.
To explain the basis of various pavement defects and condition
indicators which are modelled in HDM-4, and to highlight the most
important explanatory variables.
INTRODUCTION
z Road deterioration is broadly a function of the original

design,material types,construction quality,traffic
volume,axle load characteristics,road
geometry,environmental conditions,age of
pavement,and maintenance policy pursued.
z HDM-4
HDM 4 includes
i l d relationships
l i hi forf modeling
d li Road
R d
Deterioration (RD) and Road Works Effects (WE).These
are used for the p
purpose
p of predicting
p g annual road
condition and for evaluating road works strategies.
PAVEMENT DETERIORATION
z Various defects modelled in the following order

Cracking
R lli
Ravelling
Potholing
Edge break
Rutting
g
Roughness
Texture depth
Skid resistance
PAVEMENT CLASSIFICATION SYSTEM

Type Material Type Material Type
AC GB CRS AMGB
HRA GM
PMA AB AB AMAB
AM RAC SB CS AMSB
CM LS
PA AP TNA AMAP
SMA FDA
CAPE GB CRS STGB
DBSD GM
SBSD AB B STAB
ST SL SB CS STSB
LS
PM AP TNA STAP
FDA
FACTORS AFFECTING PAVEMENT
PERFORMANCE
TRAFFIC - no. of esa’s

CLIMATE - moisture-temperature
TIME (AGE) - since construction or maintenance
STRENGTH - Structural no, DEF, etc
DRAINAGE - condition/quality
CONSTRUCTION - type and quality
MAINTENANCE - type and quality
TRAFFIC
zThe primary traffic-related variables that

effect road deterioration include
number and type of vehicles using the road.
Axle loading characteristics of the different
vehicles types.
CLIMATE AND ENVIRONMENT
zThe climate in a road is situated has a

significant impact on the rate at which the
road deteriorates.
zImportant climatic factors are related to
temperature
precipitation
winter condition
TEMPERATURE
Temperature Description Temperature range

(C°)
Tropical Warm temperatures in small 20 to 35
range
Sub-tropical- High day cool night -5 to 45
hot temperatures,hot cold season
Sub-tropical- Moderate day -10 to 30
cool temperatures,cool winters
Temperate-
p Warm summer, shallow -20 to 25
cool winter freeze
Temperate- Cool summer, deep winter -40 to 20
freeze
ee e freeze
RAINFALL/MOISTURE
CLASSIFICATION

Classification Moisture Precipitation
Index (mm)
A id
Arid Very low
V l rainfall,
i f ll -100
100 to
t -61
61 <300
300
High evaporation
Semi-Arid Low rainfall -61 to -21 300 to 800
Sub-Humid Moderate rainfall, or -20 to +19 800 to 1600

strongly seasonal
rainfall
Humid Moderate warm +20 to +100 1500 to 3000
seasonal rainfall
Per-humid High rainfall,
rainfall or very >100 >2400
ENVIRONMENTAL COEFFICIENT
Temperature
Classification
oisture
assification Tropical
p Sub-tropical
p Sub-tropical
p Temperate
p Temperate
p
hot cool cool freeze
id 0.005 0.010 0.015 0.025 0.040
emi- arid 0.010 0.015 0.025 0.035 0.060
ub - humid 0.020 0.025 0.040 0.060 0.100
umid 0.025 0.030 0.060 0.100 0.200
er - humid 0.030 0.040 0.070

PAVEMENT STRENGTH
z Affected by:
Pavement materials and thickness
S b d
Subgrade
Drainage regime and seasonality
z Adj
AdjusteddS Structurall N
Number(SNP)
b (SNP) isi usedd
Similar to modified structural number of HDM-
III/AASHO
But contribution of pavement layers/subgrade
weighted by depth.
EFFECT OF DRAINAGE
z Depends on seasonal variations (wet and dry

seasons)
z Depend on type and condition of drains
z Depends on ingress of water into pavement (crack
and potholes)
Poor construction quality or variability in

material properties have a significant effect on
the performance of pavement
3 parameters are used to model these effects

COMP - relative compaction of the pavement layers -
affects the rate of deformation(rutting)
CDS - surface
f aspects
t - mainly
i l driven
di by
b binder
bi d
content - affects cracking and plastic deformation
CDB - roadbase aspects - affects potholing
CONCLUSIONS
aPavement deterioration manifests itself in

various kind of distresses, each of which
should be modelled separately.
separately
aAs each mode of distress develops and
progresses att diff
differentt rates
t iin diff
differentt
environments, it is important that RD
relationships
l ti hi should
h ld beb calibrated
lib t d to
t locals
l l
conditions before using them for road
i
investment
t t analyses.
l
AN OVERVIEW OF
STRATEGIC
S G C BUDGET
U G PLANNING
NN NG
USING HDM-4
ZULAKMAL HJ. SUFIAN

PWD, MALAYSIA
T3, JUNE 2000

AIM
To teach engineers an overview of the procedures
of strategic budget planning using HDM-4
WHY
so that the engineers will understand the various
considerations in preparing a strategic budget
planning
OBJECTIVES
The trainee will know the various input
parameters for a strategic
p g budget
g forecasting g
using HDM-4
T3, JUNE 2000
CONTENTS
• INTRODUCTION
• HDM - 4 AND ITS APPLICATION
• STRATEGIC ANALYSIS
– PROCEDURES
– OUTPUTS
• CONCLUSION
T3, JUNE 2000

INTRODUCTION
T3, JUNE 2000

THE ROAD NETWORK IN MALAYSIA
65,000
COMPETING
KILOMETRES
SECTORS
80% PAVED, •BUILDING

20% UNPAVED •WATER
•SLOPE
TYPICAL ANNUAL •BRIDGES
MAINTENANCE
ALLOCATION •NEW ROADS
~RM300 MILLION
T3, JUNE 2000

NETWORK
PERFORMANCE!!!!
BUDGET
CONSTRAINTS??
PUBLIC DEMANDS!!!
T3, JUNE 2000

HDM - 4
AND ITS APPLICATION
T3, JUNE 2000

WHAT IS HDM-4?
Is a decision support system, to assist in
determining impacts of potential road
investments
T3, JUNE 2000

HDM-4 Applications
•Road
Road sector policy studies
•Strategic planning or road network
development maintenance and funding
development,
requirements
•Optimal
O ti l allocation
ll ti off funds
f d tot competing
ti
needs
T3, JUNE 2000

MANAGEMENT HDM-4
FUNCTION APPLICATION
Planning Strategy Analaysis
Programming Programme Analysis
Preparation Project Analysis
T3, JUNE 2000

STRATEGIC ANALYSIS
-Procedures and Outputs-
T3, JUNE 2000

What is a strategy analysis?
Strategy analysis is concerned
with analysis of entire
enti e road
oad
networks to determine funding
needs and/o
and/or to ppredict
edict futu
futuree
performance under budget
constraints
T3, JUNE 2000

Objectives:
• determine budget allocations for road
maintenance and improvements
• Prepare for work programmes
• Determine long term network
performance
• Assess impact on road users
T3, JUNE 2000

PROCEDURES:
• Categories road network into matrix cells
• Define representative traffic volume and
loading
• Define maintenance and improvement
s d ds
standards
• Specify budget constraints
• Mode; pavement deterioration for each matrix
cell
ll
• Apply maintenance and improvement
standards
• Calculate road user benefits
• Select maintenance standards which optimise
user benefits
T3, JUNE 2000
Road Network Matrix
• Categories road network into matrix
•Road Class: Primary, Secondary, Tertiary
•Pavements: Rigid, Paved, Gravel
•Relative strength: Good, Fair, Poor
•Condition: Good, Fair, Poor
•Traffic: High, Medium, Low
• Determine road length in each matrix cell
• Record regional distribution
T3, JUNE 2000

Representative Traffic
• Primary roads: Flexible pavements
•High: >20,000 vpd (2.0m ESA/lane/year)
•Medium: 10,000 - 19,,999 (1.5m
ESA/lane/year)
•Low: <10,000 (1.0m ESA/lane/year)
• Primary
P i roads:
d Gravel
G l pavements
t
•High: >1,000
•Medium: 250 - 999
• Low: <250
T3, JUNE 2000

A Typical Network Matrix
ROAD
CONDITION
TRAFFIC
HIGH MEDIUM LOW
GOOD 76 283 307
FAIR 485 762 561
POOR 911 420 321
T3, JUNE 2000

Specify Maintenance Standards
(Flexible Pavements)
•Routine Maintenance
•Patch all potholes
•Edge repairs
•Fixed costs for drainage, safety, etc.
•Periodic
P i di M Maintenance
i t standards
t d d
•Overlay at roughness > 5.5 IRI
•Resurface
R f at roughness
h damage
d > 30%
• Improvement standards
•Widen when V/C ratio > 1.2
T3, JUNE 2000
Specify Maintenance Standards
(Gravel Roads)
•Spot regravelling
•Fixed costs for drainage
drainage, safety
safety, etc.
etc
•Periodic Maintenance standards
•Grading
G di scheduled
h d l d 2/year
2/
•Regravel when gravel loss>100mm
• Improvement
I standards
d d
•Upgrade to metal when ADT>250
T3, JUNE 2000

Budget Constraints & Optimisation
•Recurrent budget
•Capital budget
•New construction,
construction upgrading,
upgrading rehabilitation
rehabilitation,
etc.
y limit
•Multi-year
• Optimisation methods
Maximise NPV
•Maximise
•Maximise network condition improvement
Minimise cost for target network condition
•Minimise
T3, JUNE 2000
Maintenance And Improvement Costs
•Routine and recurrent maintenance costs
•Budget MUST be sufficient otherwise costs
•Periodic maintenance costs
•For specified alternatives
• Improvement costs
•For
F specified
ifi d alternatives
lt ti
T3, JUNE 2000

Calculate Road User Benefits
•vehicle operating cost savings
•fuel parts
•fuel, parts, tyres
tyres, depreciation
depreciation, etc.
etc
•Travel time savings
•Reduction in accidents
T3, JUNE 2000

Determine Optimum Standards
•Select optimum maintenance standards for each
matrix cell under specified budget constraints
•Calculate total routine, recurrent and periodic
maintenance costs for each cell
•Summarise
S i totall maintenance
i requirements
i for
f
whole network
T3, JUNE 2000

NETWORK IMPROVEMENT AND
MAINTENANCE COSTS
(RM)
Road Traffic Level
g
Condition High Medium Low
Good 6 3 1
Fair 15 13 4
Poor 31 22 11
T3, JUNE 2000

BUDGET SUMMARY
(RM)
Region ROAD CLASS

Primaryy Secondary
y Gravel
North 45.6 33 11
South 75 40 24
East/Central 63 50 12
T3, JUNE 2000

EFFECT OF BUDGET LEVELS
10
8
R O U G H N E SS
6 50%
80%
4 100%
2
0
1997 1998 1999 2000 2001 2002 2003 2004 2005
YEAR
T3, JUNE 2000

WHAT’S
WHAT S MISSING
CALIBRATION
•Vehicle Fleet
•Unit Costs
•Road Deterioration and Works Effects
•Road User Effects
T3, JUNE 2000

CONCLUSION
T3, JUNE 2000

CONCLUSIONS
HDM - 4
•is a useful tool to help managers/engineers in
preparing their strategic budget planning
p
•provides estimates for routine maintenance needs
based on road lengths under each category of roads
•provides recurrent maintenance needs based on
pavement surface condition
•provides periodic maintenance needs based on
economic indicators
•provides improvement needs based on economic
indicators
T3, JUNE 2000

FUTURE SESSIONS ON HDM-4
•practical session on strategic analysis
•Overview and practical sessions on programme
analysis
•Overview and ppractical sessions on project
p j
analysis
T3, JUNE 2000

T3, JUNE 2000
Association
mondiale
de la Route
Wo r l d R o a d
Association
The Highway Development
and Management Series
volumes one - five
H D M - 4
ENGLISH
volume five
1 2 3 4 5
A Guide to Calibration and Adaptation
THE HIGHWAY DEVELOPMENT AND MANAGEMENT SERIES
H D M - 4
ENGLISH
volume four
Analytical Framework and Model Descriptions
H D M - 4
ENGLISH
volume three
Software User Guide
H D M - 4
ENGLISH
volume two
Applications Guide
PYCCKий
FRANÇAIS
H D M - 4
ESPAÑOL
volume one
ENGLISH
Overview of HDM-4
volume one
Overview of HDM-4
Henry G. R. Kerali
Association
mondiale
de la Route
Wo r l d R o a d
Association
1
About This Manual
This Version 1.0 edition of Overview of HDM-4 provides a short executive summary
describing the HDM-4 system. Where necessary it highlights the differences between the
HDM-III and the HDM-4 models.
It is intended to be used by all readers new to HDM-4, particularly high-level management
within a road organisation.
The Overview of HDM-4 is one of five manuals comprising the suite of HDM-4
documentation (see Figure 1).
Overview of HDM-4
Volume 1
Applications
Applications Guide
Guide Software
Software User
User Guide
Guide
Volume
Volume 22 Volume
Volume 33
Analytical
Framework andand
Model Descriptions
Model Descriptions
Volume
Volume 44
A
A Guide
Guide to
to Calibration
Calibration and
and Adaptation
Adaptation
Volume
Volume 55
Figure 1 HDM-4 documentation suite
The suite of documents comprise:

! Overview of HDM-4 (Volume 1)
A short executive summary describing the HDM-4 system. It is intended to be used by all
readers new to HDM-4, particularly high level management within a road organisation.
! Applications Guide (Volume 2)
A task oriented guide describing typical examples of different types of analyses. It is to

be used by the frequent user who wishes to know how to perform a task or create a study.
Overview of HDM-4 i
Version 1.0
About This Manual
! Software User Guide (Volume 3)
Describes the HDM-4 software. It is a general purpose document which provides an

understanding of the software user interface.
! Analytical Framework and Model Descriptions (Volume 4)
Describes the analytical framework and the technical relationships of objects within the
HDM-4 model. It contains very comprehensive reference material describing, in detail,
the characteristics of the modelling and strategy incorporated in HDM-4. It is to be used
by specialists or experts whose task is to carry out a detailed study for a road
management organisation.
! A Guide to Calibration and Adaptation (Volume 5)
Suggests methods for calibrating and adapting HDM models (as used in HDM-III and
HDM-4), to allow for local conditions existing in different countries. It discusses how to
calibrate HDM-4 through its various calibration factors. It is intended to be used by
experienced practitioners who wish to understand the detailed framework and models
built into the HDM-4 system.
Notes:
1 Volumes 1, 2 and 3 are designed for the general user.
2 Volumes 4 and 5 will be of greatest relevance to experts who wish to obtain low level
technical detail. However, Volume 5, in particular, presents very important concepts,
which will be of interest to all users.
Structure of ‘Overview of HDM-4’

The information in this Overview of HDM-4 document is structured as follows:
! Section 1 - Introduction
Provides a general description of HDM-4 and its scope.

! Section 2 - Background
Provides a historical perspective to the design of HDM-4.

! Section 3 - The Role of HDM-4 in Highway Management
Describes the application of HDM-4 in terms of the following highway management

functions:
❏ Planning
❏ Programming
❏ Preparations
❏ Operations
! Section 4 - Analytical Framework
Describes the fundamental analytical framework applied in HDM-4 to model road

deterioration, road user effects, works effects, and social and environmental effects,
followed by the economic analysis framework.
Overview of HDM-4 ii
Version 1.0
About This Manual
! Section 5 - HDM-4 Applications
Describes the three analysis tools used to cater for different types of highway studies,
namely:
❏ Strategy analysis
❏ Programme analysis
❏ Project analysis
! Section 6 - HDM-4 Modules
Describes the modular structure and the main functions of the various modules:
❏ Data managers
❏ Models
! Section 7 - Data Requirements
Describes the data management facilities:

❏ HDM-4 Configuration
Used to customise the characteristics of road sections, vehicles, and the environment
under which the road system will be analysed.
❏ Road Network Manager
Defines the road sections in the network or sub-network to be analysed.
❏ Vehicle Fleet
Defines the characteristics of vehicles in the fleet that operate on the road network
being analysed.
❏ Road Works
Defines maintenance and improvement standards that are applied to different road
sections being analysed.
❏ Importing and Exporting Data
Indicates how HDM-4 can exchange data with other systems.
! Section 8 - User interface
Describes the user interface and compares its improvement to previous models.
! Section 9 - Nomenclature
This provides a reference to the Nomenclature and Glossary of terms used in the HDM-4
suite of documents.
! Section 10 - References
Gives a list of references to relevant documentation sources.
Overview of HDM-4 iii

Version 1.0
About This Manual
ISOHDM Products
The products of the International Study of Highway Development and Management Tools
(ISOHDM) consist of the HDM-4 suite of software, associated example case study databases,
and the Highway Development and Management Series collection of guides and reference
manuals. This Volume is a member of that document collection.
Customer contact
Should you have any difficulties with the information provided in this suite of documentation
please do not hesitate to report details of the problem you are experiencing. You may send an
E-mail or an annotated copy of the manual page by fax to the number provided below.
The ISOHDM Technical Secretariat welcomes any comments or suggestions from users of
HDM-4. Comments on the Overview of HDM-4 should be sent to the following address:
E-mail: isohdm@bham.ac.uk
Telephone: +44 - 121 - 414 6717 (5049)
Fax: +44 - 121 - 414 3675 (5060)
Post: ISOHDM Technical Secretariat.
School of Civil Engineering
The University of Birmingham
Edgbaston
Birmingham B15 2TT
United Kingdom
Change details
This is the first formal edition (Version 1.0) of the HDM-4 documentation.
Related documentation
HDM-4 documents
The Highway Development and Management Series Collection is ISBN: 2-84060-058-7, and
comprises:
Volume 1 - Overview of HDM-4, ISBN: 2-84060-059-5
Volume 2 - Applications Guide, ISBN: 2-84060-060-9
Volume 3 - Software User Guide, ISBN: 2-84060-061-7
Volume 4 - Analytical Framework and Model Descriptions, ISBN: 2-84060-062-5
Volume 5 - A Guide to Calibration and Adaptation, ISBN: 2-84060-063-3
Overview of HDM-4 iv
Version 1.0
About This Manual
Future documentation
The following documents will be issued at a later release:
Volume 6 - Modelling Road Deterioration and Works Effects, ISBN: 2-84060-102-8
Volume 7 - Modelling Road User and Environmental Effects, ISBN: 2-84060-103-6
Terminology handbooks
PIARC Lexicon of Road and Traffic Engineering - First edition. Permanent International
Association of Road Congresses (PIARC), Paris 1991. ISBN: 2-84060-000-5
Technical Dictionary of Road Terms - Seventh edition, English - French. PIARC Commission
on Terminology, Paris 1997. ISBN: 2-84060-053-6
General reference information

Further details on HDM-4 may be obtained from the following:
! ISOHDM Technical Secretariat
Edgbaston
Birmingham
B15 2TT
United Kingdom
Tel: +44 - 121 - 414 6717 (or 5049)
Fax: +44 - 121 - 414 3675 (or 5060)
E-mail: ISOHDM@bham.ac.uk
Web: http://www.bham.ac.uk/isohdm
! ISOHDM Project Co-ordinator
The World Road Association (PIARC)
La Grande Arche
Paroi Nord, niveau 8
92055 La Defénse Cedex
France
Tel: +33 1 41 02 05 84
Fax: +33 1 49 00 02 02
E-mail: piarc.hdm4@attglobal.net
Web: http://hdm4.piarc.org
Acknowledgements
The development of HDM-4 has been sponsored by several agencies, primarily:
! Asian Development Bank (ADB)
! Department for International Development (DFID) in the United Kingdom
! Swedish National Road Administration (SNRA)
! The World Bank
Overview of HDM-4 v
Version 1.0
About This Manual
In addition to these, significant contributions were made by:

! Finnish National Road Administration (Finnra)
! Inter-American Federation of Cement Producers (FICEM)
Many other organisations and individuals in a number of countries have also contributed in
terms of providing information, or undertaking technical review of products being produced.
The study has been co-ordinated by the ISOHDM Technical Secretariat at the University of
Birmingham in the United Kingdom. A number of organisations participated in the research
including:
! Finnra
Specification of the strategic and programme analysis applications.

! FICEM
Development of deterioration and maintenance relationships for Portland cement

concrete roads.
! The Highway Research Group, School of Civil Engineering, The University of
Birmingham
Responsible for system design and software development.
! Road Research Institute (IKRAM) in Malaysia supported by N.D.Lea International
(NDLI)
Responsible for providing updated relationships for road deterioration and road user
costs.
! Transport Research Laboratory (TRL) in the United Kingdom
Responsible for review and update of flexible pavement deterioration relationships.

! SNRA
Responsible for developing deterioration relationships for cold climates, road safety,
environmental effects, and supporting HRG with system design.
All research organisations received support from local and regional staff, visiting experts and
external advisers, to ensure that a high standard of quality and international consensus was
achieved. A number of other countries and individuals have supported this work through
supplying expert advice and reviewing the products.
Copyright statement
These HDM-4 products have been produced by the International Study of Highway
Development and Management Tools (ISOHDM), sponsored by The World Bank, the Asian
Development Bank, the Department for International Development (UK), the Swedish
National Road Administration, and other sponsors. The HDM-4 products are jointly published
by The World Road Association (PIARC), Paris and The World Bank, Washington, DC.
Copyright © 2000 The World Road Association (PIARC) on behalf of the ISOHDM sponsors.
All rights reserved.
This copyright covers all documents and document components, computer software and data
delivered as components of the HDM-4 product, in any physical or electronic forms.
Overview of HDM-4 vi
Version 1.0
Part A Overview
Contents
Volume 1 - Overview of HDM-4

1 Introduction 1
2 Background 3
2.1 Past Developments 3
2.2 Objectives of the HDM-4 Development 4
3 The Role of HDM-4 in Highway Management 5
3.1 Highway management 5
3.2 The management cycle 7
3.3 Cycles within management functions 8
4 Analytical Framework 10
5 HDM-4 Applications 13
5.1 Strategy analysis 13
5.2 Programme analysis 17
5.3 Project analysis 19
6 HDM-4 Modules 20
7 Data requirements 22
7.1 Overview 22
7.2 HDM-4 Configuration 22
7.3 Road Networks 22
7.4 Vehicle Fleets 23
7.5 Road Works 23
7.6 Importing and Exporting Data 23
8 User interface 24
9 Nomenclature 26
10 References 27
Appendix A Example of HDM-4 Strategy Analysis Application 29
A.1 Background 29
A.2 Standards 29
A.3 HDM-4 Application 30
A.4 Summary of Results 30
Appendix B Example of HDM-4 Programme Analysis Application 32
B.1 Background 32
B.2 Standards 32
Overview of HDM-4 vii

Version 1.0
Contents
B.3 HDM-4 Application 32

B.4 Summary of Results 33
Appendix C HDM-4 Project Analysis Application 38
C.1 Introduction 38
C.2 HDM-4 Application 38
C.3 Output 39
C.4 Example Project 1 - Upgrading a Gravel Road 39
C.4.1 Project Description 39
C.4.2 Project Results 40
C.5 Example Project 2 - Widening a Paved Road 40
C.5.2 Analysis Results 41
C.6 Example Project 3 - Construction of a Bypass 41
C.6.2 Project Alternatives 42
C.6.3 Traffic Diversion 42
C.6.4 Results 43
Overview of HDM-4 viii

Version 1.0
Part A Overview
Overview of HDM-4
1 Introduction
The Highway Design and Maintenance Standards Model (HDM-III), developed by the World
Bank, has been used for over two decades to combine technical and economic appraisals of
road projects, to prepare road investment programmes and to analyse road network strategies.
The International Study of Highway Development and Management (ISOHDM) has been
carried out to extend the scope of the HDM-III model, and to provide a harmonised systems
approach to road management, with adaptable and user-friendly software tools. This has
produced the Highway Development and Management Tool (HDM-4).
The scope of HDM-4 has been broadened considerably beyond traditional project appraisals,
to provide a powerful system for the analysis of road management and investment
alternatives. Emphasis was placed on collating and applying existing knowledge, rather than
undertaking extensive new empirical studies, although some limited data collection was
undertaken. Wherever possible, creative new approaches were developed for applying up-to-
date knowledge to the technical problems and management needs of different countries.
Figure 1.1 shows a view of the HDM-4 documentation suite comprising a series of five
volumes. This Overview document is Volume 1 of the series. It contains a short executive
summary describing the HDM-4 system. All readers new to HDM-4, particularly high level
management within a road organisation, should use this document.
Volumes 2 and 3 are considered as guides for users of the HDM-4 software, where user tasks
are documented and an understanding of the product can soon be learnt. Volume 2 is a task-
oriented guide describing typical examples of different types of analyses. Volume 3 describes
the HDM-4 software and is a general-purpose document that provides an understanding of the
software user interface.
Volumes 4 and 5 contain more detailed reference material that is not vital to getting started
using HDM-4, but is designed to provide detailed guidance to the more advanced users.
Volume 4 describes the analytical framework and the technical relationships incorporated in
HDM-4. It is to be used by specialists or experts whose task is to carry out a detailed study for
a road management organisation. Volume 5 describes methods for adapting and calibrating
HDM-4 in different countries. It is intended to be used by experienced practitioners who wish
to understand the detailed framework and models incorporated into the HDM-4 system.
Overview of HDM-4 1
Version 1.0
OVERVIEW
Overview of HDM-4
Volume 1
Applications
Applications Guide
Guide Software
Software User
User Guide
Guide
Volume
Volume 22 Volume
Volume 33
Analytical
Framework andand
Model Descriptions
Model Descriptions
Volume
Volume 44
A
A Guide
Guide to
to Calibration
Calibration and
and Adaptation
Adaptation
Volume
Volume 5 5
Figure 1.1 Documentation suite
Overview of HDM-4 2
Version 1.0
OVERVIEW
2 Background
2.1 Past Developments

The first move towards producing a road project appraisal model was made in 1968 by the
World Bank. The first model was produced in response to terms of reference for a highway
design study produced by the World Bank in conjunction with the Transport and Road
Research Laboratory (TRRL) and the Laboratoire Central des Ponts et Chaussées
(LCPC). Thereafter, the World Bank commissioned the Massachusetts Institute of
Technology (MIT) to carry out a literature survey and to construct a model based on
information available. The resulting Highway Cost Model (HCM) produced by MIT
(Moavenzadeh 1971, 1972) was a considerable advance over other models used for examining
the interactions between the following:
! Road work costs
! Vehicle operating costs
The HCM model highlighted areas where more research was needed to provide a model that
was more appropriate to developing country environments with additional relationships
specific to that environment.
Following this, TRRL, in collaboration with the World Bank, undertook a major field study in
Kenya to investigate the deterioration of paved and unpaved roads as well as the factors
affecting vehicle-operating costs in a developing country. The results of this study were used
by TRRL to produce the first prototype version of the Road Transport Investment Model
(RTIM) for developing countries (Abaynayaka, 1977). In 1976, the World Bank funded
further developments of the HCM at MIT that produced the first version of the Highway
Design and Maintenance Standards model (HDM) (Harral, 1979).
Further work was undertaken in a number of countries to extend the geographic scope of the
RTIM and HDM models:
! The Caribbean Study (by TRRL)
Investigated the effects of road geometry on vehicle operating costs (Morosiuk and
Abaynayaka, 1982; Hide, 1982)
! India Study (by the Central Road Research Institute - CRRI)
Studied particular operational problems of Indian roads in terms of narrow pavements

and large proportions of non-motorised transport (CRRI, 1982)
! Brazil Study (funded by UNDP)
Extended the validity of all of the model relationships (GEIPOT, 1982)

The results of the TRRL studies were used to develop the RTIM2 model (Parsley and
Robinson, 1982), whilst the World Bank developed a more comprehensive model
incorporating the findings from all previous studies and this led to HDM-III (Watanatada et
al., 1987). Both models were originally designed to operate on mainframe computers and, as
computer technology advanced, the University of Birmingham (Kerali et al., 1985) produced
a microcomputer version of RTIM2 for TRRL. Later, the World Bank produced HDM-PC, a
microcomputer version of HDM-III (Archondo-Callao and Purohit, 1989).
Overview of HDM-4 3
Version 1.0
OVERVIEW
Further developments of both models continued with the TRRL producing RTIM3 in 1993 to
provide a user-friendly version of the software running as a spreadsheet (Cundill and
Withnall, 1995), and in 1994, the World Bank produced two further versions of HDM:
! HDM-Q
Incorporating the effects of traffic congestion into the HDM-III program (Hoban, 1987).
! HDM Manager
Providing a menu-driven front end to HDM-III (Archondo-Callao, 1994).
2.2 Objectives of the HDM-4 Development

The various versions of the models have been widely used in a number of countries, and have
been instrumental in justifying increased road maintenance and rehabilitation budgets in many
countries. The models have been used to investigate the economic viability of road projects in
over 100 countries and to optimise economic benefits to road users under different levels of
expenditures. As such, they provide advanced road investment analysis tools with broad-based
applicability in diverse climates and conditions. However, it was recognised that there was a
need for a fundamental redevelopment of the various models to incorporate a wider range of
pavements and conditions of use, and to reflect modern computing practice and expectations.
The technical relationships contained in the RTIM3 and HDM-III models were in excess of 10
years old by 1995. Although much of the road deterioration models were still relevant, there
was a need to incorporate the results of the extensive research that has been undertaken
around the world in the intervening period. In the case of vehicle operating costs, it was
recognised that vehicle technology has improved dramatically since 1980 with the result that
typical vehicle operating costs could be significantly less than those predicted by RTIM3 and
HDM-III models. It was therefore necessary to update the technical relationships to reflect the
state-of-the-art. Whilst most applications of the various models have been utilised in
developing countries, in recent years many industrialised countries have begun to make use of
the model. This has resulted in the need for additional capabilities to be included, such as
models for:
! Traffic congestion effects
! Cold climate effects
! A wider range of pavement types and structures
! Road safety
! Environmental effects (energy consumption, traffic noise and vehicle emissions)
It is against this background that the development of HDM-4 was undertaken.
Overview of HDM-4 4
Version 1.0
OVERVIEW
3 The Role of HDM-4 in Highway Management
3.1 Highway management

When considering the applications of HDM-4, it is necessary to look at the highway
management process in terms of the following functions:
! Planning
! Programming
! Preparation
! Operations
Each of these functions is carried out as a sequence of activities known as the management
cycle (Robinson et al., 1998) described in Section 3.2.
Planning
Planning involves the analysis of the road system as a whole, typically requiring the
preparation of medium to long term, or strategic, estimates of expenditure for road
development and preservation under various budget and economic scenarios. Predictions may
be made of road network conditions under a variety of funding levels in terms of key
indicators together with forecasts of required expenditure under defined budget heads. The
physical highway system is usually characterised at the planning stage by:
! Characteristics of the road network
Grouped in various categories and defined by parameters such as:

❏ road class or hierarchy
❏ traffic flow/loading/congestion
❏ pavement types
❏ pavement condition
! Length of road in each category
! Characteristics of the vehicle fleet which use the road network
The results of the planning exercise are of most interest to senior policy makers in the roads
sector, both political and professional. A planning unit will often undertake this work.
Programming
Programming involves the preparation, under budget constraints, of multi-year road work and
expenditure programmes in which sections of the network likely to require maintenance,
improvement or new construction, are selected and analysed. It is a tactical planning exercise.
Ideally, cost-benefit analysis should be undertaken to determine the economic feasibility of
each set of works. The physical road network is considered at the programming stage on a
link-by-link basis, with each link characterised by homogeneous pavement sections defined in
terms of physical attributes. The programming activity produces estimates of expenditure in
each year, under defined budget heads, for different types of roadwork and for each road
section. Budgets are typically constrained, and a key aspect of programming is to prioritise the
road works in order to find the best use of the constrained budget. Typical applications are the
Overview of HDM-4 5
Version 1.0
OVERVIEW
preparation of a budget for an annual or a rolling multi-year work programme for a road
network, or sub-network. Managerial-level professionals within a road organisation normally
undertake programming activities, perhaps within a planning or a maintenance department.
Preparation
This is the short-term planning stage where road schemes are packaged for implementation.
At this stage, designs are refined and prepared in more detail; bills of quantities and detailed
costing are made, together with work instructions and contracts. Detailed specifications and
costing are likely to be drawn up, and detailed cost-benefit analysis may be carried out to
confirm the feasibility of the final scheme. Works on adjacent road sections may be combined
into packages of a size that is cost-effective for execution. Typical preparation activities are
the detailed design of:
! An overlay scheme
! Road improvement works
For example, construction along a new alignment, road widening, pavement

reconstruction, etc.
For these activities, budgets will normally already have been approved. Preparation activities
are normally undertaken by middle to junior professional staff and technicians within a design
or implementation department of a road organisation, and by contracts and procurement staff.
Operations
These activities cover the on-going operation of an organisation. Decisions about the
management of operations are made typically on a daily or weekly basis, including the
scheduling of work to be carried out, monitoring in terms of labour, equipment and materials,
the recording of work completed, and use of this information for monitoring and control.
Activities are normally focused on individual sections or sub-sections of a road, with
measurements often being made at a relatively detailed level. Operations are normally
managed by sub-professional staff, including works supervisors, technicians, clerks of works,
and others.
As the management process moves from Planning through to operations, it will be seen that
changes occur to the data required. The data detail starts as a coarse summary but progressively
moves towards a detailed level (see Table 3.1).
Overview of HDM-4 6
Version 1.0
OVERVIEW
Table 3.1 Change in management processes
Activity Time horizon Staff Spatial Data detail Mode of

responsible coverage computer
operation
Planning Long term Senior management Network-wide Coarse/ Automatic
(strategic) and policy level summary
Programming Medium term Middle-level Network or sub-

(tactical) professionals network
Preparation Budget year Junior professionals Scheme level/
sections
Operations Immediate/ Technicians/ Scheme level/

very short term sub-professionals sub-sections Fine/detailed Interactive
Source: Paterson and Scullion (1990); Paterson and Robinson (1991)
3.2 The management cycle

Traditionally, in many road organisations, budgets and programmes for road works have been
prepared on a historical basis, in which each year's budget is based upon that for the previous
year, with an adjustment for inflation. Under such a regime, there is no way of telling whether
funding levels, or the detailed allocation, are either adequate or fair. Clearly, there is a
requirement for an objective needs-based approach, using knowledge of the content, structure
and condition of the roads being managed. It will be seen that the functions of Planning,
Programming, Preparation and Operations provide a suitable framework within which a
needs-based approach can operate (Robinson et al., 1998).
In order to undertake each of these four management functions, an integrated system is
recommended. An appropriate approach is to use the management cycle concept that is
illustrated in Figure 3.1 (Robinson et al., 1998). The cycle provides a series of well-defined
steps helping the management process through their decision-making activities. The
management cycle is typically completed once in each year or in one budgeting period.
Overview of HDM-4 7
Version 1.0
OVERVIEW
Policies
Classification
Performance
and
monitoring
standards
Highway Management
Information:
•Inventory
•Condition
•Resources
•Treatments
•Productivity
•Unit Costs
•Economic parameters
Implementation Needs assessment
Finance and
Resources
Source: Robinson et al. (1998)
Figure 3.1 Highway management cycle
3.3 Cycles within management functions

The highway management process as a whole can, therefore, be considered as a cycle of
activities that are undertaken within each of the management functions of Planning,
Programming, Preparation and Operations. Table 3.2 outlines this concept and provides the
framework within which HDM-4 meets the needs of a road management organisation.
Overview of HDM-4 8
Version 1.0
OVERVIEW
Table 3.2
Management functions and the corresponding HDM-4 applications
Management Common descriptions HDM-4 applications

function
Planning Strategic analysis system HDM-4: Strategy Analysis
Network planning system
Pavement management system
Programme analysis system
Programming Pavement management system HDM-4: Programme Analysis
Budgeting system
Project analysis system
Pavement management system
Preparation Bridge management system HDM-4: Project Analysis
Pavement/overlay design system
Contract procurement system
Project management system
Operations Maintenance management system (Not addressed by HDM-4)
Equipment management system
Financial management/accounting system
Source: Kerali, Paterson and Robinson (1998)
Overview of HDM-4 9
Version 1.0
OVERVIEW
4 Analytical Framework
The HDM-4 analytical framework is based on the concept of pavement life cycle analysis.
This is applied to predict the following over the life cycle of a road pavement, which is
typically 15 to 40 years:
! Road deterioration
! Road work effects
! Road user effects
! Socio - Economic and Environmental effects
Once constructed, road pavements deteriorate as a consequence of several factors, most

notably:
! Traffic loading
! Environmental weathering
! Effect of inadequate drainage systems
The rate of pavement deterioration is directly affected by the standards of maintenance

applied to repair defects on the pavement surface such as cracking, ravelling, potholes, etc., or
to preserve the structural integrity of the pavement (for example, surface treatments, overlays,
etc.), thereby permitting the road to carry traffic in accordance with its design function. The
overall long-term condition of road pavements directly depends on the maintenance or
improvement standards applied to the road. Figure 4.1 illustrates the predicted trend in
pavement performance represented by the riding quality that is often measured in terms of the
international roughness index (IRI). When a maintenance standard is defined, it imposes a
limit to the level of deterioration that a pavement is permitted to attain. Consequently, in
addition to the capital costs of road construction, the total costs that are incurred by road
agencies will depend on the standards of maintenance and improvement applied to road
networks.
It is essential to note that the accuracy of the predicted pavement performance depends on the
extent of calibration applied to adapt the default HDM-4 models to local conditions. For
further details, refer to A Guide to Calibration and Adaptation - Volume 5.
Overview of HDM-4 10
Version 1.0
OVERVIEW
Poor
Maintenance Standard
Roughness
Pavement
Performance
Curve
Rehabilitation
Good
Time (years) or Traffic Loading
Figure 4.1 Concept of Life-cycle analysis in HDM-4
The impacts of the road condition, as well the road design standards, on road users are
measured in terms of road user costs, and other social and environmental effects. Road user
costs comprise:
! Vehicle operation costs (fuel, tyres, oil, spare parts consumption; vehicle depreciation
and utilisation, etc.),
! Costs of travel time - for both passengers and cargo, and
! Costs to the economy of road accidents (that is, loss of life, injury to road users,
damage to vehicles and other roadside objects).
The social and environmental effects comprise vehicle emissions, energy consumption, traffic
noise and other welfare benefits to the population served by the roads. Although the social and
environmental effects are often difficult to quantify in monetary terms, they can be
incorporated within the HDM-4 economic analyses if quantified exogenously.
It should be noted that in HDM-4, road user effects can be calculated for both motorised
transport (motorcycles, cars, buses, trucks, etc.) and non-motorised transport (bicycles, human
powered tricycles, animal pulled carts, etc.).
Figure 4.2 illustrates the impact of road condition (represented in terms of the IRI) on the
costs of different modes of transport.
Version 1.0
OVERVIEW
Bus
Road User Costs ($/veh-km)
Heavy Truck
Pickup/utility
Car
Rickshaw
Good Road Condition (IRI) Poor
Figure 4.2 Effect of Road Condition on Vehicle Operating Costs for Rolling
Terrain
Road User Costs in HDM-4 are calculated by predicting physical quantities of resource
consumption and then multiplying these quantities by the corresponding user specified unit
costs. It is necessary to ensure that the vehicle resource quantities predicted are in keeping
with the range of values observed in the area of application. For further details, refer to A
Guide to Calibration and Adaptation - Volume 5.
Economic benefits from road investments are then determined by comparing the total cost
streams for various road works and construction alternatives against a base case (without
project or do minimum) alternative, usually representing the minimum standard of routine
maintenance. HDM-4 is designed to make comparative cost estimates and economic analyses
of different investment options. It estimates the costs for a large number of alternatives year-
by-year for a user-defined analysis period. All future costs are discounted to the specified base
year. In order to make these comparisons, detailed specifications of investment programmes,
design standards, and maintenance alternatives are needed, together with unit costs, projected
traffic volumes, and environmental conditions.
Version 1.0
OVERVIEW
5 HDM-4 Applications
5.1 Strategy analysis

The concept of strategic planning of medium to long term road network expenditures requires
that a road organisation should consider the requirements of its entire road network asset.
Thus, strategy analysis deals with entire networks or sub-networks managed by one road
organisation. Examples of road networks include; the main (or trunk) road network, the rural
(or feeder) road network, urban (or municipal) road network, etc. Examples of sub-networks
include; all motorways (or expressways), all paved (or unpaved roads), different road classes,
etc.
In order to predict the medium to long term requirements of an entire road network or sub-
network, HDM-4 applies the concept of a road network matrix comprising categories of the
road network defined according to the key attributes that most influence pavement
performance and road user costs. Although it is possible to model individual road sections in
the strategy analysis application, most road administrations will often be responsible for
several thousand kilometres of roads, thereby making it cumbersome to individually model
each road segment. The road network matrix can be defined by users to represent the most
important factors affecting transport costs in the country. A typical road network matrix could
be categorised according to the following:
! Traffic volume or loading
! Pavement types
! Pavement condition
! Environment or climatic zones
! Functional classification (if required)
For example, a road network matrix could be modelled using; three traffic categories (high,
medium, low), two pavement types (asphalt concrete, surface treatments), and three pavement
condition levels (good, fair, poor). In this case, it is assumed that the environment throughout
the study area is similar and that the road administration is responsible for one road class (for
example, main roads). The resulting road network matrix for this would therefore comprise (3
x 2 x 3 =) 18 representative pavement sections. There is no limit to the number of
representative pavement sections that can be used in a strategy analysis. The trade-off is
usually between a simple representative road network matrix that would give rather coarse
results compared against a detailed road network matrix with several representative sections
that could potentially provide more accurate results.
Strategy analysis may be used to analyse a chosen network as a whole, to prepare medium to
long range planning estimates of expenditure needs for road development and conservation
under different budget scenarios. Estimates are produced of expenditure requirements for
medium to long term periods of usually 5-40 years. Typical applications of strategy analysis
by road administrations would include:
! Medium to long term forecasts of funding requirements for specified target road
maintenance standards (see Figure 5.1a).
! Forecasts of long term road network performance under varying levels of funding (see
Figure 5.1b).
Version 1.0
OVERVIEW
! Optimal allocation of funds according to defined budget heads; for example routine
maintenance, periodic maintenance and development (capital) budgets (see
Figure 5.1c).
! Optimal allocations of funds to sub-networks; for example by functional road class
(main, feeder and urban roads, etc.) or by administrative region (see Figure 5.1d).
! Policy studies such as impact of changes to the axle load limit, pavement maintenance
standards, energy balance analysis, provision of NMT facilities, sustainable road network
size, evaluation of pavement design standards, etc.
A typical sample application of the HDM-4 Strategy Analysis is given in
Appendix A.
Version 1.0
OVERVIEW
Main Roads Annual

Average Roughness (IRI) 7.0 Budget
$10m
6.0
$15m
5.0
4.0 $20m
Target
3.0
2000 2001 2002 2003 2004 2005 2006
Figure 5.1a Effect of funding levels on road network performance
Budget Allocations
7.0 Feeder
Average Roughness (IRI)
Roads
$30m/yr
6.0
Secondary
Roads
5.0 $35m/yr
4.0 Primary
Roads
$20m/yr
3.0
2000 2001 2002 2003 2004 2005 2006
Figure 5.1b Effect of budget allocations on sub-network performance
Version 1.0
OVERVIEW
250
200
D e v e lo p m e n t
U S $ m /y e a r
150
Im p ro v e m e n t
P e r io d i c
100
R o u tin e
50
0
2000 2001 2002 2003
Figure 5.1c Optimal budget allocations to sub-heads
200
150
U S $ m /y e a r
P rim a ry
100 S e c o n d a ry
F eed er
50
0
2000 2001 2002 2003
Figure 5.1d Optimal budget allocations to sub-networks
Version 1.0
OVERVIEW
5.2 Programme analysis

This deals primarily with the prioritisation of a defined long list of candidate road projects
into a one-year or multi-year work programme under defined budget constraints. It is essential
to note that here, we are dealing with a long list of candidate road projects selected as discrete
segments of a road network. The selection criteria will normally depend on the maintenance,
improvement or development standards that a road administration may have defined (for
example from the output produced by the strategy analysis application). Examples of selection
criteria that may be used to identify candidate projects include:
! Periodic maintenance thresholds (for example, reseal pavement surface at 20% damage).
! Improvement thresholds (for example, widen roads with volume/capacity ratio greater
than 0.8).
! Development standards (for example, upgrade gravel roads to sealed pavements when the
annual average daily traffic exceeds 200 vehicles per day).
The above examples do not imply firm recommendations to be used by road authorities.
When all candidate projects have been identified, the HDM-4 programme analysis application
can then be used to compare the life cycle costs predicted under the existing regimen of
pavement management (that is, the without project case) against the life cycle costs predicted
for the periodic maintenance, road improvement or development alternative (that is, with
project case). This provides the basis for estimating the economic benefits that would be
derived by including each candidate project within the budget timeframe.
It should be noted that the main difference between strategy analysis and programme analysis is
the way in which road links and sections are physically identified. Programme analysis deals
with individual links and sections that are unique physical units identifiable from the road
network throughout the analysis. In strategy analysis, the road system essentially loses its
individual link and section characteristics by grouping all road segments with similar
characteristics into the road network matrix categories.
For both strategy and programme analysis, the problem can be posed as one of seeking that
combination of treatment alternatives across a number of sections in the network that
optimises an objective function under budget constraint. If, for example, the objective
function is to maximise the Net Present Value (NPV), the problem can be defined as:
Select that combination of treatment options for sections that maximises NPV
for the whole network subject to the sum of the treatment costs being less
than the budget available.
The HDM-4 programme analysis application may be used to prepare a multi-year rolling
programme, subject to resource constraints (see Figure 5.2a and Figure 5.2b). The
prioritisation method employs the incremental NPV/cost ratio as the ranking index, described
in more detail in Analytical Framework and Model Descriptions - Volume 4. This provides an
efficient and robust index for prioritisation purposes. Indices such as the NPV, economic rate
of return (ERR), or predicted pavement condition attributes (for example, road roughness) are
not recommended as ranking criteria. The incremental NPV/cost ratio satisfies the objective of
maximising economic benefits for each additional unit of expenditure (that is, maximise net
benefits for each additional $1 of the available budget invested).
A typical sample application of the HDM-4 programme analysis application is
given in Appendix B.
Version 1.0
OVERVIEW
Priority Road Length Province Type of Scheduled Cost Cumulative

Rank Section (km) or District Road Work Year $m S$m
1 N1-2 20.5 2 Resealing 2000 5.4 5.4
2 N4-7 23.5 7 Overlay 40mm 2000 10.9 16.3
3 N2-5 12.5 5 Reconstruct 2000 8.6 24.9
4 R312-1 30 4 Widen 4 lane 2000 31.4 56.3
5 R458-3 36.2 3 Overlay 60mm 2000 16.3 72.6
: : : : : : : :
1 N4-16 32.1 6 Reconstruct 2001 22.8 22.8
2 R13-23 22.4 4 Overlay 40mm 2001 9.7 32.5
3 N521-5 45.2 2 Widen 4 lane 2001 41.3 73.8
: : : : : : : :
1 N1-6 30.2 4 Resealing 2002 8.2 8.2
2 N7-9 17.8 3 Overlay 60mm 2002 9.2 17.4
3 F2140-8 56.1 1 Reconstruct 2002 34.9 52.3
: : : : : : : :
Figure 5.2a Sample output from Programme analysis (Format 1)
2000 2001 2002 2003

Priority Road Length Province Road Cost Road Cost Road Cost Road Cost
Rank Section (km) or District Work $m Work $m Work $m Work $m
1 N1-2 20.5 2 RESEAL 5.4 R.M. 0.185 R.M. 0.185 R.M. 0.185
2 N4-7 23.5 7 OVL40MM 10.9 R.M. 0.212 R.M. 0.212 R.M. 0.212
3 N2-5 12.5 5 RECON 8.6 R.M. 0.113 R.M. 0.113 R.M. 0.113
4 R312-1 30 4 WIDEN-4 31.4 R.M. 0.180 R.M. 0.180 R.M. 0.180
5 R458-3 36.2 3 OVL60MM 16.3 R.M. 0.217 R.M. 0.217 R.M. 0.217
: : : : : : : : : : : :
16 N4-16 32.1 6 R.M. 0.289 RECON 22.8 R.M. 0.289 R.M. 0.289
17 R13-23 22.4 4 R.M. 0.134 OVL40MM 9.7 R.M. 0.134 R.M. 0.134
18 N521-5 45.2 2 R.M. 0.407 WIDEN-4 41.3 R.M. 0.407 R.M. 0.407
: : : : : : : : : : : :
28 N1-6 30.2 4 R.M. 0.272 R.M. 0.272 RESEAL 8.2 R.M. 0.272
29 N7-9 17.8 3 INLAY 0.240 R.M. 0.200 OVL60MM 9.2 R.M. 0.160
30 F2140-8 56.1 1 PATCH 0.202 R.M. 0.202 RECON 34.9 R.M. 0.168
Note: RM = Routine Maintenance
Figure 5.2b Sample output from Programme analysis (Format 2)
Version 1.0
OVERVIEW
5.3 Project analysis

Project analysis is concerned with the following:
Evaluation of one or more road projects or investment options. The
application analyses a road link or section with user-selected treatments, with
associated costs and benefits, projected annually over the analysis period.
Economic indicators are determined for the different investment options.
Project analysis may be used to estimate the economic or engineering viability of road
investment projects by considering the following issues:
! The structural performance of road pavements
! Life-cycle predictions of road deterioration, road works effects and costs
! Road user costs and benefits
! Economic comparisons of project alternatives
Typical appraisal projects would include the maintenance and rehabilitation of existing roads,
widening or geometric improvement schemes, pavement upgrading and new construction.
There are no fundamental changes to the philosophy of the system in this area, but improved
road deterioration relationships have been extended to cover a wider range of pavements and
the performance of materials in temperate and cold climates. Road user cost relationships
include impacts on road safety.
Typical examples of project analysis applications using HDM-4 are given in
Appendix C.
In terms of data requirements, the key difference between the strategy and programme
analyses, with that for project analysis, is in the detail at which data is defined. Use is made of
the concept of information quality levels (IQL) recommended by the World Bank (Paterson
and Scullion, 1990). Project level analysis data is specified in terms of measured defects
(IQL-II), whereas the specification for strategy and programme analyses can be more generic
(IQL-III). For example; for project level analysis, road roughness would be specified in terms
of the IRI value (m/km); but for strategy and programme analyses, roughness could be
specified as good, fair or poor. The relationship between IQL-II and IQL-III level data is
user-defined in the HDM Configuration depending on road class, pavement surface type and
traffic class.
Version 1.0
OVERVIEW
6 HDM-4 Modules
The overall structure of HDM-4 is illustrated in Figure 6.1. The three analysis tools (Strategy,
Programme and Project) operate on data defined in one of four data managers:
! Road Network
Defines the physical characteristics of road sections in a network or sub-network to be

analysed.
! Vehicle Fleet
Defines the characteristics of the vehicle fleet that operate on the road network to be
analysed.
! Road Works
Defines maintenance and improvement standards, together with their unit costs, which
will be applied to the different road sections to be analysed.
! HDM Configuration
Defines the default data to be used in the applications. A set of default data is provided
when HDM-4 is first installed, but users should modify these to reflect local
environments and circumstances.
Technical analysis within the HDM-4 is undertaken using four sets of models:
! RD (Road Deterioration)
Predicts pavement deterioration for bituminous, concrete and unsealed roads.

! WE (Works Effects)
Simulates the effects of road works on pavement condition and determines the
corresponding costs.
! RUE (Road User Effects)
Determines costs of vehicle operation, road accidents and travel time.

! SEE (Social and Environment Effects)
Determines the effects of vehicle emissions and energy consumption.

The model simulates, for each road section, year-by-year, the road condition and resources
used for maintenance under each strategy, as well as the vehicle speeds and physical resources
consumed by vehicle operation. After physical quantities involved in construction, road works
and vehicle operation are estimated, user-specified prices and unit costs are applied to
determine financial and economic costs. Relative benefits are then calculated for different
alternatives, followed by present value and rate of return computations.
These models are described in detail in Analytical Framework and Model Descriptions -
Volume 4, and were largely derived from the equivalent models used in HDM-III.
The HDM-4 system is designed to interface with external systems such as:
! Databases
Road network information systems, pavement management systems, etc., through

intermediate Import/Export files.
Version 1.0
OVERVIEW
! Technical models
Accessed directly by external systems for research applications or other studies.

The system design is modular in structure to enable users to implement the HDM-4 modules
independently within their road management systems. The technical relationships can easily
be calibrated to match local conditions by using HDM-4 Configuration in addition to country
specific default data.
Data Managers Analysis Tools
Road Vehicle Road HDM Project ProgramStrategy

Network Fleet Works Config.
File Core
CoreData
Data Model libraries
FileConverter
Converter Vehicle Models
VehicleFleets
Fleets can be used in
Road
RoadNetworks
Networks other systems
transfer Road
RoadWorks
transferdata
data Works
Projects
totoexternal
external Projects
systems Programmes
systems Strategies
Strategies
External RDWE
RDWE:Road
:RoadDeteriorationand
DeteriorationandWorks
WorksEffects
Effects
RUE : Road User Effects
Databases, SEE
SEE: :Safety,
Safety,Energy
Energyand
andEnvironmental
EnvironmentalEffects
Effects
PMS, etc.
PMS
PMS: :Pavement
PavementManagement
Manageme System
ntSystem
Figure 6.1 HDM-4 System Architecture
Version 1.0
OVERVIEW
7 Data requirements
7.1 Overview
The HDM-4 applications have been designed to work with a wide range of data type and
quality. For example, pavement condition data collected by visual inspection according to
condition classes (for example, Very good, good, fair, poor condition) can be converted to the
HDM-4 model requirements prior to running any of the applications (see section below on
Importing and Exporting Data). Similarly, HDM-4 can work with very detailed measurements
of pavement condition if the data is available. This flexibility in data requirements should
permit all potential users with a variety of data to integrate HDM-4 into their road
management functions. For further details, refer to Applications Guide - Volume 2.
7.2 HDM-4 Configuration

Since HDM-4 will be used in a wide range of environments, HDM Configuration provides the
facility to customise system operation to reflect the norms that are customary in the
environment under study. Default data and calibration coefficients can be defined in a flexible
manner to minimise the amount of data that must be changed for each application of HDM-4.
Default values are supplied with HDM-4, but these are all user-definable and facilities are
provided to enable this data to be modified. The HDM-4 set of tools may be used as additional
modules to current pavement management systems. Import and Export functions, built into
the modules, provide a mechanism for data transfer between existing databases and HDM-4
modules. The data exchange format uses standard data file formats to encourage its wide
adaptation by road organisations.
7.3 Road Networks

Road Networks provides the basic facilities for storing characteristics of one or more road
sections. It allows users to define different networks and sub-networks, and to define road
sections, which is the fundamental unit of analysis. The data entities supported within the
road network are:
! Sections
Lengths of road over which physical characteristics are reasonably constant.

! Links
Comprise one or more sections over which traffic is reasonably constant. This is provided
for purposes of compatibility of the network referencing system with existing pavement
management systems.
! Nodes
Intersections which connect links or other points at which there is a significant change in
traffic, carriageway characteristics, or administrative boundaries.
All network data is entered using the Road Network folder, and facilities are also available for
editing, deleting and maintaining this data. The approach to network referencing is
considerably more flexible than that used in HDM-III, and is designed to handle a wide range
of external referencing conventions as might be used by other systems with which HDM-4
may need to interface.
Version 1.0
OVERVIEW
7.4 Vehicle Fleets

Vehicle Fleets provide facilities for the storage and retrieval of vehicle characteristics required
for calculating vehicle speeds, operating costs, travel time costs and other vehicle effects. The
method used to represent a vehicle fleet is considerably more adaptable than that used in
HDM-III with no limit on the numbers or types of vehicles that can be specified. Motorcycles
and non-motorised vehicles are included. Multiple vehicle fleet data sets can be set up for use
in different analyses, with a wide range of default data provided.
7.5 Road Works

Road Works Standards refer to the targets or levels of conditions and response that a road
management organisation aims to achieve. Road organisations normally set up different
standards that can be applied in practical situations in order to meet specific objectives which
are related to functional characteristics of the road network system.
The Road Works folder provides facilities, within a flexible framework, to define a list of
maintenance and improvement standards that are followed by road organisations in their
network management and development activities. The standards defined in the Road Works
Standards folder can be used in any of the three analysis tools:
! Project analysis
! Programme analysis
! Strategy analysis
7.6 Importing and Exporting Data

The data required for HDM-4 analyses can be imported from existing data sources such as
pavement management systems (PMS), highway information systems, etc. The data import
into HDM-4 (as well as the export from HDM-4) is organised according to the data objects
described above (that is, road networks, vehicle fleets, maintenance and improvement
standards, HDM Configuration). The physical attributes of the selected data objects must be
exported to a data exchange file format defined for HDM-4. This permits all data required by
HDM-4 to be imported directly from any database. Data transformation rules may need to be
implemented for converting the data held in the external database to the format used by HDM-
4. For example, pothole data recorded in the external database in terms of the percentage area
of the pavement surface would need to be converted to the equivalent number of standard
pothole units (10 litres by volume) required in HDM-4. Similarly, other data required by
HDM-4, such as pavement deterioration calibration factors, should be inserted as pre-defined
default values according to the type of pavement, road class, and other defined factors. Other
data required for the HDM-4 analyses can be directly stored within the HDM-4 internal
database. These include data on vehicle fleet characteristics, road maintenance and
improvement standards, unit costs and economic analysis parameters (for example, discount
rate, analysis period, etc.). For further details on exchanging data with HDM-4, refer to
Software User Guide - Volume 3.
Version 1.0
OVERVIEW
8 User interface
A key objective for the development of HDM-4 is to provide a system that is user-friendlier
than the original HDM-III. This has been achieved by addressing the user interface design and
data requirements. The user interface has been improved by developing the system to run
under a standard Microsoft Windows environment. Many computer users are already familiar
with the Windows system and this should make learning HDM-4 more intuitive. A modular
system design has been adopted to enable different modules to be used relatively easily and to
facilitate future system operation on other non-Windows platforms. More details of system
issues are described in the Software User Guide - Volume 3.
Considerable attention has also been paid to the data that must be entered by users,
particularly because the total data requirement is greater than that in HDM-III in order to
allow for the extended facilities included in HDM-4. The concept of data hierarchy is used,
whereby default data is defined for many items in HDM Configuration, and users can choose
the extent to which defaults are used. The system can be used with a level of data entry detail
that is appropriate for particular applications. A comprehensive Help system is also provided.
Figure 8.1 illustrates the HDM-4 Workspace. Detailed instructions for the operation of the
individual modules are given in the Software User Guide - Volume 3.
Version 1.0
OVERVIEW
Figure 8.1 HDM-4 Workspace
Version 1.0
OVERVIEW
9 Nomenclature
A nomenclature explaining any special numbering, markings, etc., that are associated with the
HDM-4 suite of documents is included in Part H of the Analytical Framework and Model
Descriptions - Volume 4.
A glossary of terms is included in Part I of the Analytical Framework and Model Descriptions
- Volume 4 describing the use of phrases found in the HDM-4 documentation.
Version 1.0
OVERVIEW
10 References
Abaynayaka, S.W., Morosiuk G., and Hide H., (1977)

Prediction of road construction and vehicle operating costs in developing countries.
Proceedings of the Institution of Civil Engineers. 62 (Part 1), 419-446.
Archondo-Callao, R. and Purohit R., (1989)
HDM-PC: user's guide. The Highway Design and Maintenance Standards Series.
Washington DC: The World Bank.
Archondo-Callao, R., (1994)
HDM Manager Version 3.0. Transportation Division, Transportation, Water & Urban
Development Department. Washington DC: The World Bank.
CRRI, (1982)
Road user cost study in India. Final Report, New Delhi: Central Road Research Institute.
Cundill, M.A., and Withnall S.J., (1995)
Road transport investment model RTIM3. In: TRB. Sixth International Conference on
Low-Volume Roads, Conference Proceedings 6, Volume 1. Washington DC: National
Academy Press, 187-192.
FHWA, (1994)
Highway economics requirements system. Washington DC: Federal Highway
Administration.
Geipot, (1982)
Research on the interrelationships between costs of highway construction, maintenance
and utilisation (PICR). Final report, 12 volumes, Brasilia, Brazil.
Harral, C.G., and others, (1979)
The highway design and maintenance standards model (HDM): model structure,
empirical foundations and applications. PTRC Summer Annual Meeting, University of
Warwick, 13-16 July 1979. London: PTRC Education and Research Services.
Hide, H., (1982)
Vehicle operating costs in the Caribbean: results of a survey of vehicle operators. TRRL
Laboratory Report 1031. Crowthorne: Transport and Road Research Laboratory.
Hoban, C.J., (1987)
Evaluating traffic capacity and improvements to geometry. Technical Paper Number 74.
Kerali, H.R., Parsley L.L., Robinson R., and Snaith M.S., (1985)
Development of a microcomputer based model for road investment in developing
countries. In: CIVILCOMP. Proceedings of the Second International Conference on
Civil and Structural Engineering Computing, London, December 1985. London:
Institution of Civil Engineers.
Version 1.0
OVERVIEW
Kerali, H.R., Robinson R., and Paterson W.D.O., (1998)

The Role of New HDM-4 in Road Management. Fourth International Conference on
Managing Pavements, Durban, May 1998. South Africa: University of Pretoria.
Moavenzadeh, F., Stafford J.H., Suhbrier J., and Alexander J. (1971)
Highway design study phase I: the model. IBRD Economics Department Working Paper
No 96. Washington DC: International Bank for Reconstruction and Development.
Moavenzadeh, F., Berger F., Brademeyer B., and Wyatt R. (1975)
The Highway Cost Model: General Framework. MIT Department of Civil Engineering
Research Report No 75-4. Cambridge Mass: Massachusetts Institute of Technology.
Morosiuk, G., and Abaynayaka S.W., (1982)
Vehicle operating costs in the Caribbean: an experimental study of vehicle performance.
TRRL Laboratory Report 1056. Crowthorne: Transport and Road Research Laboratory.
Parsley, L.L., and Robinson R., (1982)
The TRRL road investment model for developing countries (RTIM2). TRRL Laboratory
Report 1057. Crowthorne: Transport and Road Research Laboratory.
Paterson, W.D.O., and Scullion T., (1990)
Information systems for road management: draft guidelines on system design and data
issues. Infrastructure and Urban Development Department Report INU77. Washington
DC: The World Bank.
Phillips, S., (1994)
Development of the United Kingdom pavement management system. In: TRB. Third
International Conference on Managing Pavements, San Antonio, Texas, 22-26 May
1994, Volume 2. Washington DC: National Academy Press, 227-236.
Robinson, R. Danielsson U., and Snaith M.S., (1998)
Road Management: Concepts and Systems. Basingstoke and London: Macmillan Press.
Watanatada T., Harral C.G., Paterson W.D.O., Dhareshwar A.M., Bhandari A., and
Tsunokawa K., (1987)
The highway design and maintenance standards model volume 1: description of the
HDM-III model. The Highway Design and Maintenance Standards Series. Baltimore:
Johns Hopkins for the World Bank.
Version 1.0
STRATEGY ANALYSIS APPLICATION
Appendix A Example of HDM-4 Strategy Analysis

Application
A.1 Background
The national road network in a country comprises 4,267 km of paved roads and 3,145 km of
unpaved (gravel) roads. A recent pavement condition survey has shown that 22.3% of the
paved road network is in good condition, 36.2% in fair condition and 41.5% in poor condition.
Traffic surveys were also carried out in the previous year. Following an analysis of the data
available together with the characteristics of the national road network, it has been decided to
categorise the paved road network into three traffic classes (high, medium, low), and three
pavement condition classes (good, fair, poor), and the unpaved road network has been
categorised into two traffic categories only (medium and low). The resulting road network
matrix is summarised in Table A.1:
Table A.1
Road Network Length in each Matrix Category
Traffic categories
Paved roads: High Medium Low
(AADT > 4000) (4000>AADT>1000) (AADT < 1,000)
Good condition 234 km 306 km 410 km
Fair condition 392 km 483 km 670 km
Poor condition 437 km 615 km 720 km
Unpaved roads: Medium Low

(AADT > 100) (AADT < 100)
Gravel surface 1,760 km 1,385 km
A.2 Standards
The national road authority has a defined set of standards for road maintenance and road
network improvement in accordance with its own policies and those set by the government.
! The maintenance and improvement standards for paved roads include
❏ Widening all paved roads with volume to capacity ratio greater than 0.8.
❏ Rehabilitation (structural overlay) of all paved roads in poor condition.
❏ Resealing (surface dressing) paved roads when surface deterioration exceeds 30%.
❏ Reactive routine maintenance comprising patching potholes immediately, sealing
cracks, edge repairs, etc., as required.
❏ Routine maintenance to shoulders, drainage ditches, road markings and all roadside
furniture.
Version 1.0
! The maintenance and improvement standards for unpaved roads include
❏ Upgrading to paved surface, all gravel roads carrying more than 250 vehicles per
day.
❏ Regravelling when the remaining gravel thickness falls below 50 mm.
❏ Grading gravel roads with medium traffic twice a year, and grading once a year for
gravel roads carrying less than 100 vehicles per day.
❏ Routine maintenance to shoulders, drainage ditches and all roadside furniture.
A.3 HDM-4 Application

The objective of this study is to determine the required funding levels for the defined
maintenance and improvement standards, and to monitor the effect of budget constraints on
the long-term network performance trends.
The HDM-4 procedure required to analyse the national road network comprises the following:
! Create the representative road network matrix using the Strategy application;
! Define the characteristics of the vehicles which use the road network;
! Specify traffic growth rates;
! Assign the maintenance and improvement standards to the road network matrix together
with their unit costs;
! Run the HDM-4 Strategy application to determine the total budget requirement;
! Carry out constrained budget analyses;
! Review reports and graphs of the analyses conducted.
A.4 Summary of Results

The results of the analyses can be summarised in chart form as illustrated in Figure A.1.
The analyses indicate that the ideal maintenance and improvement standards specified by the
policy would require approximately US$ 56.2 million per year for the paved road network,
and US$ 21.2 per year for the unpaved road network (based on the unit costs of the various
road works). If only 50% of the required funding were available (represented by the minimum
periodic maintenance option), this would result in a 54% loss in road user benefits (that is,
compared against road user costs for the routine & recurrent option).
Version 1.0
Predicted Performance of Paved Roads Predicted Performance of Gravel Roads

Routine & Recurrent Min Periodic Maint. Ideal Periodic Maint. Routine & Recurrent Min. Periodic Maint. Ideal Periodic Maint.
12.0 25.0

10.0
20.0
8.0
15.0
6.0
4.0
10.0
2.0 5.0
0.0
0.0
2000 2005 2010 2015 2020
2000 2005 2010 2015 2020
Paved Road User Costs (Undiscounted)

Unpaved Road User Costs (Undiscounted)
Routine & recurrent Min Period Maint. Ideal Periodic Maint.
Routine & Recurrent Min Periodic Maint. Ideal Periodic Maint.
200
3,000
2,500 150
(millions)
(millions)
2,000 100
1,500 50
1,000
0
2000 2005 2010 2015 2020
2000 2005 2010 2015 2020
Figure A.1 Summary Output from HDM-4 Strategy Analysis
Version 1.0
PROGRAMME ANALYSIS APPLICATION
Appendix B Example of HDM-4 Programme

Analysis Application
B.1 Background
The national road authority has drawn up a long list of candidate road sections for periodic
maintenance and improvement over the next three years in Western Province. The long list of
candidate road sections follows a review of pavement condition surveys carried out by
consultants. The national road authority has a policy to prioritise the candidate projects and
select those that will be included in the periodic maintenance programme for the three-year
budget period. Given that the candidate projects are from the main road network only, the
objective is to prioritise according to the economic benefits that would be derived from each
candidate road project.
B.2 Standards
The standards for periodic maintenance and road improvement defined by the national road
authority require the following road works to be carried out:
! Road improvement standards
❏ Pave gravel roads with AADT greater than 150 vehicles per day.
❏ Widen roads with peak volume to capacity ratio greater than 0.85.
! Periodic maintenance standards
❏ Reconstruct failed pavements with roughness greater than 9.5 IRI.

❏ Strengthen pavements in critical condition with roughness greater than 5.0 IRI.
❏ Reseal pavements with observed distress on more than 30% of the pavement surface
area (that is, cracking, ravelling, potholes, edge break, etc.). This includes
preparatory works such as crack sealing, pothole patching and edge repairs prior to
the resealing.
! Reactive and cyclic routine maintenance
❏ Patching potholes, crack sealing and edge repairs as required.

❏ Drainage maintenance, shoulder repairs, vegetation control, etc., specified as fixed
costs per km per year.
B.3 HDM-4 Application

The objective of this study is to select a short list of projects for Western Province that can be
carried out within the funding to be made available for periodic maintenance and road
improvement over the next three years.
The HDM-4 procedure required to prioritise the candidate projects comprises the following:
! Import data from the Pavement Management System or use the HDM-4 Road Network
manager to create the candidate road sections.
! Define the characteristics of the vehicles that use the road network.
Version 1.0
! Specify traffic growth rates.

! Assign the maintenance and improvement standards to the candidate road sections
together with the unit costs.
! Run the HDM-4 Programme Analysis application to determine the road works required.
! The unconstrained work programme results give the total funding required for the long
list of candidate road sections.
! Carry out budget optimisation to prioritise and select the short list of projects that can be
carried out within the available budget.
! Review reports of the analyses conducted.
B.4 Summary of Results

The results of the analyses are summarised in Table B.1a and Table B.1b.
The analyses indicate that the selected road sections for periodic maintenance and road
improvement would require approximately US$ 11.345 million over the three year budget
period (based on the unit costs of the various road works).
If only 70% of the required funding were available, Table B.1b shows the short list of
candidate sections that would be included in the three-year program.
Version 1.0
Table B.1a Unconstrained Work Programme
HDM-4 Unconstrained Work Programme

Study Name: Western Province 3 Year Road Investment Programme Run Date: 23-11-1999
All costs are expressed in: Local Currency (millions).
Year Road No. Section Length AADT Work Description NPV/C Financial Cumulative Equivalent
(km) Costs Costs US$(m)
2000 MSC 112 km 35 - 80 21.0 1711 Overlay 20mm 4.91 23.15 23.15 0.609
2000 MSE 203 km 80 - 90 10.0 1152 Overlay 20mm 2.51 11.03 34.18 0.899
2000 MSC 138 km 5 - 10 & 65 - 70 10.0 1271 Single Seal 2.41 3.80 37.98 0.999
2000 MAN 446 km 53.7 - 57.0 3.3 983 Overlay 20mm 2.40 3.31 41.29 1.086
2000 MSE 932 km 36.2 - 52.2 16.0 809 Overlay 20mm 2.19 17.64 58.93 1.551
2000 MSE 334 km 3.8 - 12.8 9.0 932 Overlay 20mm 1.98 9.92 68.85 1.812
2000 MAN 203 km 185 - 190 5.0 1248 Single Seal 1.72 1.90 70.75 1.862
2000 MAN 243 km 0 - 21 21.0 1009 Overlay 20mm 1.55 23.15 93.90 2.471
2000 MTS 549 km 0 - 1.0 1.0 2438 Overlay 20mm 1.37 1.10 95.00 2.500
2000 MSC 142 km 12.5 - 20 7.5 942 Single Seal 0.98 3.04 98.04 2.580
2000 MSW 131 km 10 - 16.5 6.5 428 Overlay 20mm 0.92 7.72 105.76 2.783
2000 MSW 905 km 128 - 170 42.0 717 Single Seal 0.92 15.96 121.73 3.203
2000 MSV 449 km 45 - 59.6 14.6 671 Single Seal 0.46 5.70 127.43 3.353
2000 MTN 748 km 30 - 60 30.0 589 Single Seal 0.21 11.40 138.83 3.653
…Continued
Version 1.0

All costs are expressed in: Local Currency (millions).
2000 MTN 831 km 0 - 3.7 3.7 205 Overlay 20mm 0.17 4.41 143.24 3.770
2000 MTN 830 km 8.6 - 20 11.4 680 Pavement Reconstruction 0.10 51.63 194.87 5.128
2000 MTN 805 km 145 - 152.6 7.6 844 Pavement Reconstruction 0.05 37.80 232.67 6.123
2001 MSW 901 km 180 - 330 60.0 1623 Overlay 20mm 3.16 66.15 298.82 7.864
2001 MSV 537 km 43 - 45 2.0 937 Overlay 20mm 1.42 2.21 301.03 7.922
2001 MSW 935 km 16 - 30 14.0 532 Overlay 20mm 0.33 15.44 316.46 8.328
2002 MSC 131 km 16.5 - 26 9.5 2253 Overlay 20mm 4.91 11.03 327.49 8.618
2002 MSC 141 km 28 - 49.5 21.5 2253 Overlay 20mm 4.88 24.26 351.74 9.256
2002 MSW 956 km 10 - 75 30.0 1733 Overlay 20mm 3.31 33.08 384.82 10.127
2002 MSC 155 km 50 - 55 5.0 1353 Overlay 20mm 2.61 5.51 390.33 10.272
2002 MSW 938 km 3.2 - 10 6.8 913 Overlay 20mm 1.30 7.72 398.05 10.475
2002 MSW 150 km 0.0 - 16.5 16.5 800 Overlay 20mm 0.99 18.74 416.79 10.968
2002 MAN 214 km 90 - 97.7 7.7 759 Overlay 20mm 0.86 8.82 425.61 11.200
2002 MAN 409 km 110 - 115 5.0 512 Overlay 20mm 0.28 5.51 431.12 11.345
This report was generated by Version 1.0 of HDM-4. Neither this report nor the application may be distributed without prior consent.
Version 1.0
Table B.1b Constrained Work Programme
HDM-4 Constrained Work Programme

All costs are expressed in: Local Currency (millions). Available Budget: 300 million (2000 – 2002)
2000 MSC 112 km 35 - 80 21.0 1711 Overlay 20mm 4.91 23.15 23.15 0.609
2000 MSE 203 km 80 - 90 10.0 1152 Overlay 20mm 2.51 11.03 34.18 0.899
2000 MSC 138 km 5 - 10 & 65 - 70 10.0 1271 Single Seal 2.41 3.80 37.98 0.999
2000 MAN 446 km 53.7 - 57.0 3.3 983 Overlay 20mm 2.40 3.31 41.29 1.086
2000 MSE 932 km 36.2 - 52.2 16.0 809 Overlay 20mm 2.19 17.64 58.93 1.551
2000 MSE 334 km 3.8 - 12.8 9.0 932 Overlay 20mm 1.98 9.92 68.85 1.812
2000 MAN 203 km 185 - 190 5.0 1248 Single Seal 1.72 1.90 70.75 1.862
2000 MAN 243 km 0 - 21 21.0 1009 Overlay 20mm 1.55 23.15 93.90 2.471
2000 MTS 549 km 0 - 1.0 1.0 2438 Overlay 20mm 1.37 1.10 95.00 2.500
2000 MSC 142 km 12.5 - 20 7.5 942 Single Seal 0.98 3.04 98.04 2.580
2000 MSW 131 km 10 - 16.5 6.5 428 Overlay 20mm 0.92 7.72 105.76 2.783
2000 MSW 905 km 128 - 170 42.0 717 Single Seal 0.92 15.96 121.73 3.203
2001 MSW 901 km 180 - 330 60.0 1623 Overlay 20mm 3.16 66.15 187.88 4.944
2001 MSV 537 km 43 - 45 2.0 937 Overlay 20mm 1.42 2.21 190.08 5.002
… Continued
Version 1.0
HDM-4 Constrained Work Programme

All costs are expressed in: Local Currency (millions). Available Budget: 300 million (2000 – 2002)
2002 MSC 131 km 16.5 - 26 9.5 2253 Overlay 20mm 4.91 11.03 201.11 5.292
2002 MSC 141 km 28 - 49.5 21.5 2253 Overlay 20mm 4.88 24.26 225.36 5.931
2002 MSW 956 km 10 - 75 30.0 1733 Overlay 20mm 3.31 33.08 258.44 6.801
2002 MSC 155 km 50 - 55 5.0 1353 Overlay 20mm 2.61 5.51 263.95 6.946
2002 MSW 938 km 3.2 - 10 6.8 913 Overlay 20mm 1.30 7.72 271.67 7.149
2002 MSW 150 km 0.0 - 16.5 16.5 800 Overlay 20mm 0.99 18.74 290.41 7.642
2002 MAN 214 km 90 - 97.7 7.7 759 Overlay 20mm 0.86 8.82 299.23 7.874
This report was generated by Version 1.0 of HDM-4. Neither this report nor the application may be distributed without prior consent.
Version 1.0
PROJECT ANALYSIS APPLICATION
Appendix C HDM-4 Project Analysis Application
C.1 Introduction
The project analysis application in HDM-4 can be used to carry out economic appraisal for a
wide range of project types. These include:
! Periodic maintenance and rehabilitation
❏ Overlays
❏ Resealing
❏ Pavement reconstruction
! Upgrading
❏ Paving unsealed roads

❏ Concrete pavements
❏ Full depth asphalt pavements
! Staged construction
❏ Construction of road sections in sequence

! New road construction
❏ Road by-pass schemes

❏ Traffic diversion schemes
! Road widening
❏ Dual carriageways
❏ Lane addition
❏ Carriageway widening
! Non-motorised traffic facilities
❏ Bicycle lanes
❏ Carriageway separation
Note that the above list is not exhaustive.
C.2 HDM-4 Application

The objective of a project level application is to determine the best engineering and economic
alternatives for individual road sections.
The HDM-4 procedure required to carry out a project application comprises the following:
! Specify characteristics of the road sections using the Road Network manager.
! Define the characteristics of the vehicles that use the road sections.
! Specify traffic growth rates.
Version 1.0
! Specify the maintenance and road improvements to be carried out together with the unit
costs.
! Run the HDM-4 Project Analysis application to determine the economic benefits.
! Review reports of the analyses conducted.
C.3 Output
There are several detailed summary reports produced by HDM-4. These include:
! Pavement deterioration and road works
❏ Summary of annual pavement condition

❏ Quantities of road works
❏ Details of annual road work costs
❏ Schedule of road works
! Road user effects
❏ Summary of road user costs (vehicle operation, travel time and accidents)
❏ Traffic flow details
❏ Average travel speeds
❏ Traffic volume to capacity ratios
! Environmental effects
❏ Vehicle emissions
❏ Energy consumption
! Economic analysis results
❏ Annual cost streams

❏ Discounted cash flows
❏ Net Present Values (NPV)
❏ Economic Internal Rate of Return (EIRR)
❏ Benefit Cost Ratio (BCR)
❏ First Year Benefits (FYB)
C.4 Example Project 1 - Upgrading a Gravel Road
C.4.1 Project Description

This example presents the economic analysis of a project to upgrade an existing gravel road to
a paved standard. The existing road is 50 km long and passes through varying topography.
For analysis purposes, three sections, based on geometry, pavement condition, and traffic
volume can represent the road. Traffic and condition data are available from surveys
undertaken in 1999. The gravel thickness in 1999 was 150 mm.
investment. This differs from a financial appraisal that is concerned with the means of
Version 1.0
financing a project and the financial profitability of the project. The economic feasibility of
the project is assessed by comparison against a base-line project alternative (that is, a without
project alternative). The project alternatives are:
! Without Project
Maintain the existing gravel road.

! With Project
Maintain the existing gravel road before upgrading to a bituminous pavement, followed
by maintenance of the bituminous pavement.
C.4.2 Project Results

The Economic Analysis Summary (by Project) report in HDM-4 gives a summary of costs,
discounted Net Present Value (NPV) and Internal Rate of Return (IRR) for the project
alternative. Cost and NPV details are presented by road section in the Economic Analysis
Summary (by Section) report. For this project, the overall NPV is reported as US$ 16.77
(millions). The breakdown by section indicates that all three sections give a positive NPV.
The IRR for the road project is calculated as 14.2%.
C.5 Example Project 2 - Widening a Paved Road

This project presents the economic analysis of widening a paved road. The existing road is
7m wide, with an AADT of 15,000 in 1998. Non-motorised transport contributed an extra
400 vehicles in 1998, comprising animal carts and bicycles. The analysis assumes that
routine pavement maintenance is undertaken on a condition responsive basis for all
alternatives. Three widening alternatives are considered, widening by 1m, widening by 3m,
and adding two extra lanes. The road under study is represented by one section, 10 km long,
and the different widening proposals represent project alternatives. The analysis period is
defined by the start year 2000 and duration 20 years (that is, 2000 - 2019). The project
alternatives are summarised in the table below:
Version 1.0

condition. In addition, a 50 mm overlay is applied when the
roughness level reaches 6 IRI OR when structural cracking affects
15% of the carriageway area.
2 With this alternative, the existing road is widened by 1 m during the
period (2000-2001). The maintenance regime of Alternative 1
(Routine + 50 mm overlay), which is condition responsive, is effective
from year 3 (2002).
period (2000-2001). The condition responsive maintenance regime of
Alternative 1 is effective from year 3 (2002).
4 With this alternative, the existing road is widened by adding two lanes
during the period (2000-2002). The condition responsive maintenance
regime of Alternative 1 (Routine + 50 mm Overlay) is effective from
year 4 (2003).
C.5.2 Analysis Results

The impact of the widening alternatives can be assessed by examination of the Volume-
Capacity Ratio report which tabulates the volume-capacity ratio (VCR) by time period and
calendar year for each project alternative and road section. The effect of widening on vehicle
speeds is demonstrated by the Vehicle Speed report.
The Economic Analysis Summary indicates that Alternatives 3 and 4 give a positive NPV,
whereas Alternative 2 (widening by 1m only) gives a negative value of NPV.
Alternative Widening NPV

(US$ millions)
2 1m -3.31
3 3m +6.69
4 2 lanes +12.18
C.6 Example Project 3 - Construction of a Bypass

This example demonstrates the economic analysis of a project to construct a bypass around a
town centre. The objective is to demonstrate the construction of a new road section, and to
examine the resulting traffic diversion effects.
The road sections included in the project are shown schematically in Figure C.1. Road
sections A, B, C and D represent the existing main road network within a town centre. The
Version 1.0
proposed project is the construction of a bypass, represented by Section E, which is 10 km

long.
B C D
E
Key: Existing road sections
Proposed bypass (Section E)
Figure C.1 Construction of a new bypass
C.6.2 Project Alternatives

The four project alternatives considered are defined below and summarised in the table below.
Alternative 1 represents the existing road sections without the bypass. Alternatives 2, 3 and 4
include the bypass (represented by section E), with carriageway width and pavement
construction as described below. The analysis period is 20 years (from year 2000 to 2019).
1 This alternative represents the base case situation without the bypass.
2 Construct Section E: A two-lane AMGB (Asphalt Mix on Granular Base) road, with
a two-year construction period (2000-2001), and opening in year 2002.
3 Construct Section E: A wide two-lane AMGB road, with a three-year construction

period (2000-2002), and opening in year 2003.
4 Construct Section E: A four-lane AMGB road, with a four-year construction period

(2000-2003), and opening year 2004.
C.6.3 Traffic Diversion

The construction of the bypass (Section E) will cause a significant redistribution of traffic
between the existing roads and the new road. The table below summarises the expected
change in normal traffic after completion of the new section.
Version 1.0
Section AADT (1998) before bypass AADT bypass opening year

A 10,000 10,000
B 4,000 4,000
C 6,000 1,000
D 8,000 3,000
E n/a 5,000
C.6.4 Results
The Economic Analysis Summary indicates that construction of the bypass (Section E) would
be viable in economic terms. The most cost-effective alternative would be a 2-lane bypass as
summarised in the table below.
Alternative New Section Option NPV (US$ millions) EIRR (%)

2 Standard 2 lane 66.360 66.71
3 Wide 2 lane 60.355 52.25
4 Standard 4 lane 49.820 35.69
Version 1.0
A4 back covers 10/12/99 10:25 am Page 1
Volume one: Overview of HDM-4
volume one
A short executive summary describing the HDM-4 system. It is intended to be used by all readers new to HDM-4,
particularly high-level management within a road organisation.
ISBN: 2-84060-059-5
Volume two: Applications Guide

A task oriented guide describing typical examples of different types of analyses. It is to be used by the frequent user
who wishes to know how to perform a task or create a study.
ISBN: 2-84060-060-9
Volume three: Software User Guide

Describes the HDM-4 software. It is a general purpose document which provides an understanding of the software
user interface.
ISBN: 2-84060-061-7
Volume four: Analytical Framework and Model Descriptions

Describes the analytical framework and the technical relationships of objects within the HDM-4 model. It contains Association
very comprehensive reference material describing, in detail, the characteristics of the modelling and strategy mondiale
incorporated in HDM-4. It is to be used by specialists or experts whose task is to carry out a detailed study for a de la Route
road management organisation.
ISBN: 2-84060-062-5
Volume five: A Guide to Calibration and Adaptation
Suggests methods for calibrating and adapting HDM models (as used in HDM-III and HDM-4), to allow for local
conditions existing in different countries. It discusses how to calibrate HDM-4 through its various calibration factors.
It is intended to be used by experienced practitioners who wish to understand the detailed framework and models
built into the HDM-4 system. Wo r l d R o a d
ISBN: 2-84060-063-3 Association
Volume six: Modelling Road Deterioration and Works Effects

This volume describes the technical background of theory, observation, knowledge and international practice in
which the HDM-4 models for road deterioration and works effects have been developed.
ISBN-2-84060-102-8
Volume seven: Modelling Road User and Environmental Effects

which the HDM-4 models for road user and environmental effects have been developed.
ISBN-2-84060-103-6
Highway Development and Management Series

The Highway Development and Management system (HDM-4) provides a harmonised systems approach
to road management, with adaptable and user-friendly software tools. It is a powerful tool for conducting
project appraisals and analyses of road management and investment alternatives.
PIARC, Registered in France These HDM-4 products have been produced by the International Study of Highway
Development and Management Tools (ISOHDM), sponsored by The World Bank, the
Registered Office Asian Development Bank, the Department for International Development (UK), the
The World Road Association (PIARC) Swedish National Road Administration, and other sponsors. HDM-4 is jointly published
La Grande Arche by The World Road Association (PIARC), Paris, and The World Bank, Washington, D.C.
Paroi Nord, Niveau 8
92055 La Defénse Cedex PIARC endeavours to ensure that the information in this document is correct and fairly
France stated but does not accept liability for any error or omission.
Tel: +33 1 41 02 05 84 The development of PIARC products and services is continuous and published
Fax: +33 1 49 00 02 02 information may not be up to date. It is important to check the current position with
Email: piarc.hdm4@attglobal.net PIARC. This document is not part of a contract or licence save insofar as may be
Web: http://hdm4.piarc.org expressly agreed.

Series Collection ISBN: 2-84060-058-7
Copyright © 2000 The World Road Association (PIARC) on behalf of the ISOHDM sponsors All rights reserved English edition
volume two
Applications Guide
Henry G. R. Kerali
Derek McMullen
J.B. Odoki
Association
mondiale
de la Route
Wo r l d R o a d
Association
2
Part A
About This Manual

This Version 1.0 edition of the Applications Guide is a task-oriented guide describing typical
examples of different types of analyses.
It is to be used by the frequent user who wishes to know how to perform a task or create a
study.
The Applications Guide is one of five manuals comprising the suite of HDM-4 documentation
(see Figure 1).
Overview
Overview of
of HDM-4
HDM-4
Volume
Volume 11
Applications
Applications Guide
Guide Software
Software User
User Guide
Guide
Volume
Volume 22 Volume
Volume 33
Analytical
Framework andand
Model Descriptions
Model Descriptions
Volume
Volume 44
A
A Guide
Guide to
to Calibration
Calibration and
and Adaptation
Adaptation
Volume
Volume 55


Applications Guide i
Version 1.0
About This Manual

Describes the analytical framework and the technical relationships used within the HDM-
4 model. It contains very comprehensive reference material describing, in detail, the
characteristics of the modelling and strategy incorporated in HDM-4. It is to be used by
specialists or experts whose task is to carry out a detailed study for a road management
organisation.
Notes:
Structure of the ‘Applications Guide’

The aim of this Applications Guide is to demonstrate the different applications of HDM-4
through selected case studies. The case studies are described in detail and may be reviewed
using the data files included on the CD-ROM.
The information in the Applications Guide is structured in four parts (see Figure 2).
Part A contains an overview of HDM-4, with a short description of the economic analysis
framework used, and the HDM-4 Technical Models are briefly described. Part B describes
the concepts of analysis for projects, programmes and strategies. Part C describes typical
methods of using HDM-4 for Research, Policy and Regulation analyses. Details of the data
requirements for HDM-4 are given Part D. Examples of case studies are given in the
Appendices.
Applications Guide ii
Version 1.0
About This Manual
Applications
ApplicationsGuide
Guide
Introduction
Introduction
Part
PartAA
Concepts
Concepts
ofofAnalysis
Analysis
Part
PartBB
Project
Project Programme
Programme Strategy
Strategy
Analysis
Analysis Analysis
Analysis Analysis
Analysis
Chapter
ChapterB1B1 Chapter
ChapterB2
B2 Chapter
ChapterB3
B3
Research,
Research,
Policy
Policyand
and
Regulation
Regulation
Analyses
Analyses
Part
PartCC
Data
DataManagement
Management
Part
PartDD
Customising Data Road Vehicle Traffic

TrafficData
Customising Data Road VehicleFleet
Fleet Data Road
RoadWorks
Works
HDM-4
HDM-4 Hierarchy
Hierarchy Network
Network Data Standards
Data Chapter Standards
Data ChapterD5D5
Chapter Data
ChapterD1
D1 Chapter
ChapterD2
D2 Chapter
ChapterD4
D4 Chapter
ChapterD6
D6
Chapter
ChapterD3
D3
Case
CaseStudies
Studies
Appendices
Appendices
Figure 2 Schematic layout of the Applications Guide documentation
Applications Guide iii

Version 1.0
About This Manual
A brief description of the four parts follows:

! Part A - Introduction
Provides an overview of HDM-4, with a short description of the economic analysis

framework used.
! Part B - Concepts of analysis
Describes the concepts used for the following analyses:

❏ Project Analysis
Describes the overall procedure for project analysis in HDM-4. This allows users to
assess the physical, functional and economic feasibility of specified project
alternatives by comparison against a base case (without-project) alternative.
❏ Programme Analysis
Describes the concept of Programme Analysis. This deals with the preparation of a
multi-year rolling programme for a road network in which candidate investment
options are identified and selected, subject to resource constraints. Road networks
are analysed section by section and estimates are produced of road works and
expenditure requirements for each road section for each year of the funding period.
❏ Strategy Analysis
Describes the procedure for Strategy Analysis. This is concerned with the analysis of
a chosen network as a whole, for preparing long range planning estimates of
expenditure needs for road development and conservation under different budget
scenarios. The road network is characterised by lengths of road in different
categories defined by parameters such as road class, surface type, and pavement
condition or traffic flow. Estimates are produced of expenditure requirements for
medium to long term periods of 5-40 years.
! Part C - Research, Policy and Regulation Analyses
Describes typical applications of HDM-4 for Research, Policy and Regulation analyses.
! Part D - Data Management
Data Management comprises the following chapters:

Chapter D1 - Customising HDM-4
Chapter D2 - Data Hierarchy
Chapter D3 - Road Transport
Chapter D4 - Vehicle Fleet Data
Chapter D5 - Traffic Data
Chapter D6 - Road Works Standards
! Appendices - Case studies
Case studies for Project, Programme and Strategy analyses are supplied as appendices.
The Applications Guide will be updated with new case studies as future versions of
HDM-4 are released.
Applications Guide iv
Version 1.0
About This Manual
ISOHDM Products
Customer contact
HDM-4. Comments on the Applications Guide should be sent to the following address:
Telephone: +44 - 121 - 414 6717 (5049)
Fax: +44 - 121 - 414 3675 (5060)
Edgbaston
Birmingham B15 2TT
United Kingdom
Change details
HDM-4 documents
comprises:
Applications Guide v
Version 1.0
About This Manual

Edgbaston
Birmingham
B15 2TT
United Kingdom
Tel: +44 - 121 - 414 6717 (or 5049)
Fax: +44 - 121 - 414 3675 (or 5060)
La Grande Arche
France
Tel: +33 1 41 02 05 84
Fax: +33 1 49 00 02 02
Acknowledgements
! The World Bank
Applications Guide vi
Version 1.0
About This Manual

including:
! Finnra

! FICEM

concrete roads.
! The Highway Research Group (HRG), School of Civil Engineering, The University of
Birmingham
! Road Research Institute (IKRAM) in Malaysia supported by N.D. Lea International
(NDLI)
costs.

! SNRA
Copyright statement
Applications Guide vii

Version 1.0
Part A Contents
Contents
Volume 2 - Applications Guide

Part A Introduction
A1 Introduction A1-1
Part B Concepts of Analysis

B1 Concepts of Project Analysis B1-1
B2 Concepts of Programme Analysis B2-1
B3 Concepts of Strategy Analysis B3-1
Part C RPRA
C1 Concepts of Research, Policy and Regulation Analyses C1-1
Part D Data Management

D1 Customising HDM-4 D1-1
D2 Data Hierarchy D2-1
D3 Road Network Data D3-1
D4 Vehicle Fleet Data D4-1
D5 Traffic Data D5-1
D6 Road Works Standards D6-1
Appendix A - Project Analysis Case Studies

Appendix A1 - Project Analysis Case Study 1 1
Appendix B - Programme Analysis Case Studies

Appendix B1 - Programme Analysis Case Study 1 1
Appendix B2 - Programme Analysis Case Study 2 1
Appendix C - Startegy Analysis Case Studies

Appendix C1 - Strategic Analysis Case Study 1 1
Applications Guide viii

Version 1.0
A4 back covers 10/12/99 11:50 am Page 2
volume two
ISBN: 2-84060-059-5

ISBN: 2-84060-060-9

user interface.
ISBN: 2-84060-061-7

ISBN: 2-84060-062-5

ISBN-2-84060-102-8

ISBN-2-84060-103-6


Part C Research, Policy and Regulation Analyses
Contents
Volume 2 - Applications Guide: Part A
A1 Introduction
1 Introduction to HDM-4 A1-1
1.1 Scope of HDM-4 A1-1
1.2 Background A1-1
1.3 Objectives of model development A1-3
1.4 Overview of HDM-4 A1-3
1.5 Input data A1-7
1.6 Applications A1-8
1.7 User interface A1-10
2 References A1-11
Version 1.0
PART A INTRODUCTION
Part A Introduction
A1 Introduction
1 Introduction to HDM-4
1.1 Scope of HDM-4

The Highway Design and Maintenance Standards Model (HDM-III), developed by the
World Bank, has been used for over two decades to combine technical and economic
appraisals of road investment projects, and to analyse strategies and standards. An
international study has now been carried out to extend the scope of the HDM-III model, and
to provide a harmonised systems approach to road management, with adaptable and user-
friendly software tools. This has resulted in the development of the Highway Development
and Management Tool (HDM-4).
The scope of HDM-4 has been broadened considerably beyond traditional project appraisals,
to provide a powerful system for the analysis of road management and investment
alternatives. Emphasis has been on collating and applying existing knowledge, rather than
undertaking new major empirical studies, although some limited data collection has been
undertaken. Wherever possible, creative new approaches have been developed for applying
technical knowledge to the technical problems and management needs of different countries.
1.2 Background
1.2.1 Past developments

The first move to produce a model of this type was made in 1968 by the World Bank. Terms
of reference for a highway design study were produced and considered by the World Bank
in conjunction with Britain's Transport and Road Research Laboratory (TRRL) and the
French Laboratoire Centrale des Ponts et Chaussées (LCPC). The World Bank then
commissioned a group from the Massachusetts Institute of Technology (MIT) to carry out a
literature survey and to construct a model based on information already available. The
Highway Cost Model produced by MIT (Moavenzadeh, 1971 and 1972) was a considerable
advance over any other methods existing at the time for examining the interactions between
road costs, maintenance costs, and vehicle operating costs. However, the model highlighted
areas where more research was needed to replace relationships that were inappropriate to
developing country environments, and to provide additional relationships.
Following on from this, TRRL, in collaboration with the World Bank, undertook a major
field study in Kenya to investigate the deterioration of paved and unpaved roads and factors
affecting vehicle-operating costs in a developing country. The results of this study were
used to calibrate a prototype computer model (RTIM) for evaluating the costs of
construction, maintenance and vehicle operation for a road investment project in a
developing country (Abaynayaka, 1977). In 1976, the World Bank instigated further
developments of the computer model. They awarded a further research contract to MIT to
produce an extended version of the model capable of carrying out economic analysis
directly, separating a road link into homogeneous sections, and carrying out automatic
sensitivity analysis of key variables such as discount rate and traffic growth. The work
resulted in the production of the Highway Design and Maintenance Standards Model
(HDM) (Harral, 1979).
Applications Guide A1-1

Version 1.0
PART A INTRODUCTION A1 INTRODUCTION
Both the RTIM and HDM models were tested and used extensively in the field. Further
fieldwork was undertaken, as listed below, with the general aim of extending the geographic
spread of the results:
! Caribbean - by TRRL
Investigation of the effects of road geometry on vehicle operating costs (Morosiuk and
Abaynayaka, 1982; Hide, 1982).
! India - by the Central Road Research Institute
Study of the particular operational problems of Indian roads in terms of narrow

carriageways and large proportions of animal-drawn transport (CRRI, 1982).
! Brazil - funded by UNDP
Study of the validity of all of the model relationships (GEIPOT, 1982).

! Parallel model developments
TRRL's experience of model use suggested that there was a need for a simpler model
than those, which already existed. The requirement was to obtain a model that was easier
to use by consultants working in the field and by users in developing countries. This
resulted in the development of RTIM2 (Parsley and Robinson, 1982). The World Bank,
on the other hand, saw the need for a more comprehensive model, and this led to the
development of HDM-III (Watanatada, 1987).
Both of these models were originally designed to operate on mainframe computers and,
as computer technology advanced, the University of Birmingham (Kerali, 1985)
produced a microcomputer version of RTIM2 for TRRL. Later, the World Bank
(Archondo-Callao and Purohit, 1989) produced a microcomputer version of HDM-III.
Further development of both models has since taken place. RTIM3 was produced in
1993 to provide an extremely user-friendly version of the software running as a
spreadsheet (Cundill and Withnall, 1995). In 1994, the World Bank produced two further
developments:
❏ HDM-Q
Produced to incorporate the effects of traffic congestion into the HDM program
(Hoban, 1987).
❏ HDM Manager
Produced to provide a menu-driven version of HDM-III (Archondo-Callao, 1994).
1.2.2 Considerations for the present development

The various versions of the models have been used widely in a number of countries, and
have been instrumental in justifying increased road maintenance and road rehabilitation
budgets, and for optimising economic benefits to road users under different levels of
expenditures. As such, they have provided advanced road investment analysis tools with
broad-based applicability in diverse climates and conditions. However, it was recognised in
many quarters that there was a need for a fundamental redevelopment of the various models
to reflect changed conditions of use, and to reflect modern computing practice and
expectations.
The relationships contained in the models are now all in excess of ten years old. In the case
of road deterioration, the existing models used are still relevant, but there is a need to
incorporate the results of the extensive research that has been undertaken around the world in
the intervening period. In the case of vehicle operating costs, the situation is rather different:
vehicle technology has improved dramatically in the last ten years with the result that typical

Version 1.0
operating costs are now considerably less than those predicted by the models. Therefore the
relationships were updated to reflect the present state-of-the-art.
Whilst most applications of the various models have been in developing countries, in recent
years many industrialised countries have begun to make use of the model. This has resulted
in the need for additional facilities to be included, such as the need to deal effectively with
congestion, with temperate and cold climates, with a wider range of pavement types and
structures, and with environmental issues.
It is against this background that the development work has been undertaken.
1.3 Objectives of model development

The international study leading up to the development of HDM-4 had the fundamental
objective of producing a new standard model to be used in the assessment of the technical,
economic, social and environmental impacts of road investments. The new model has been
based on the results of research work conducted over the past ten years, and extends the
scope of the existing HDM-III model in four key areas:
! Project appraisal
For the economic appraisal of road maintenance, rehabilitation, upgrading and new
construction through life cycle analysis of proposed road investments (see Part B).
! Works programming
For the preparation of multi-year and rolling programmes for road network maintenance
and development to facilitate the preparation of medium term budgets for road network
maintenance and extension (see Part B).
! Strategic planning
For policy development, long term resource allocation plans and road network planning
(see Part B).
! Software environment
A user-friendly system, built around a set of modules with the capacity to cope with a
wide range of data requirements and user skill levels.
In order to meet wider system objectives, technical relationships were updated and calibrated
to best current knowledge; additional technical capabilities have been provided to deal with
traffic congestion, non-motorised vehicles, concrete pavements, drainage, environmental and
safety effects.
1.4 Overview of HDM-4

An overview of the HDM-4 data flow is given in Figure A1.1. This identifies the folders
containing the input data and shows the main outputs from each application. Also shown are
the HDM-4 models that undertake the technical analyses.
The analysis procedure is shown in Figure A1.2.

Version 1.0
Data Managers Analysis Tools
Road Vehicle Road HDM Project ProgramStrategy

Network Fleet Works Config.
File Core
CoreData
Data Model libraries
FileConverter
Converter Vehicle Models
VehicleFleets
Fleets can be used in
Road
RoadNetworks
Networks other systems
transfer Road
RoadWorks
transferdata
data Works
Projects
totoexternal
external Projects
systems Programmes
systems Strategies
Strategies
External RDWE
RDWE:Road
:RoadDeteriorationand
DeteriorationandWorks
WorksEffects
Effects
RUE : Road User Effects
Databases, SEE
SEE: :Safety,
Safety,Energy
Energyand
andEnvironmental
Effects
PMS, etc.
PMS : Pavement Management
PMS : Pavement Management SSystem
ystem
Figure A1.1 HDM-4 Flow diagram
1.4.1 Input data

The input data (discussed in Section 1.5) is held in four folders:
! Road Network
Contains data defining the road network or sub-network that will be the basis of the
analysis.
! Vehicle Fleet
Contains the characteristics of the vehicle fleet that will be operating on the road network
being analysed.
! Road Works
Contains the specifications for maintenance and improvement standards that will be
applied to different road sections to be analysed.
! HDM Configuration
Contains all of the default data to be used with the data analysis; a set of default data will
be provided with the system, but users will be able to modify it to reflect local
circumstances.
1.4.2 Life cycle analysis

The underlying operation of HDM-4 is similar for each of project, programme or strategy
analysis. In each case, HDM-4 simulates total life cycle conditions and costs for an analysis

Version 1.0
period under a user-specified scenario of circumstances. The primary cost set for the life
cycle analysis includes the costs of capital investment, maintenance and vehicle operation, to
which travel time costs can be added as an option. The costs of accidents, and environmental
pollution may also be included in the analysis.
The broad concept of the life cycle analysis is illustrated in Figure A1.2. Interacting sets of
costs, related to those incurred by the road administration and those incurred by the road
user, are added together over time in discounted present values. Costs are determined by
first predicting physical quantities of resource consumption and then multiplying these
quantities by their unit costs or prices. Economic benefits are then determined by comparing
the total cost streams for various maintenance and construction alternatives with a base case
(do nothing or do minimum alternative), usually representing minimal routine maintenance.
The following models are contained within HDM-4:
! Road Deterioration (RD)
Predicts road deterioration for paved and unsealed roads.
! Works Effects (WE)
Implements road works programmes and determines works costs.

! Road User Effects (RUE)
Determines costs of vehicle operation, road accidents and travel time.

! Social and Environmental Effects (SEE)
Determines effects of emissions and noise, and predicts numbers of road accidents and
amounts of energy consumption.
More comprehensive details of the HDM-4 models are provided in the Analytical
Framework and Model Descriptions.
HDM-4 is designed to make comparative cost estimates and economic analyses of different
investment options. It estimates the costs for a large number of alternatives year-by-year for
a user-defined analysis period, discounting the future costs. Rates of return, net present
values, and first year benefits can also be determined by HDM-4. In order to make these
comparisons, detailed specifications of investment programmes, design standards, and
maintenance alternatives are needed, together with unit costs, projected traffic volumes, and
environmental conditions.
The model simulates, for each road section, year-by-year, the road condition and resources
used for maintenance under each alternative, as well as the vehicle speeds and physical
resources consumed by vehicle operation. Physical quantities involved in construction,
maintenance and vehicle operation are estimated, and user-specified prices and unit costs
applied to determine financial and economic costs. Relative benefits are then calculated for
different alternatives, followed by present value and rate of return computations.
Details of the benefits and costs considered in HDM-4 are included in Part G of the
Analytical Framework and Model Descriptions.

Version 1.0
INPUTS MODEL OUTPUTS
Vehicle type, volume, growth,

loading, physical parameters,
terrain, precipitation, road Start of analysis loop
geometry, pavement
characteristics, unit costs
Pavement type, strength, age, Cracking, ravelling, pot-holes, rut

condition, and ESAL Road Deterioration depth, faulting (paved); gravel
thickness (unpaved); roughness
Road geometry and roughness; Fuel, lubricant, tyres, maintenance,

vehicle speed, type; congestion Road User Effects fixed costs, speed, travel time, road
parameters; unit costs user costs
Road works standards and Reset cracking, ravelling, pot-

strategies holes, rut depth (paved); gravel
Works Effects thickness (unpaved); roughness,
works quantities and agency costs
Road geometry and surface Social and Levels of emissions and energy
texture, vehicle characteristics Environmental Effects used, and number of accidents
Developmental, accident, Costs and benefits, including

environmental, and other Economic Analysis exogenous benefits
exogenous costs and benefits
Total costs by component; net

Return to start of analysis loop present values and rates of return
by section
Figure A1.2 Life-cycle analysis using HDM-4

Version 1.0
1.4.3 Outputs
The three main types of outputs produced by HDM-4 are as follows (see Section 1.6 for
details):
! Economic efficiency indicators – from analyses of individual road projects
! Multi-year work programmes – produced after prioritisation of several candidate road

projects
! Strategic road maintenance and development plans – produced from long term
predictions of road network performance.
1.4.4 Interface to external systems

HDM-4 also has the ability to interface with external systems:
! Between databases, road network information systems, pavement management systems,
and the like, through a file converter.
! Technical models can be accessed directly by external systems for research applications
or other studies.
The system design is modular in nature, with interfaces between modules being defined
through an open architecture. This enables users to substitute their own technical models for
those that are provided with the system. Existing technical relationships are all implemented
in such a way that they can easily be calibrated and adapted to local conditions. Default data
sets can be defined prior to using HDM-4 thereby reducing the need to specify all data items
during subsequent use; this is done through the HDM Configuration.
1.5 Input data
1.5.1 Road Network

An open road network referencing system provides the basic definition of the way that road
sections in HDM-4 can be integrated with a highway management system. This defines the
way that data pertaining to the road network is located relative to the road, and defines the
technical rules and conventions for doing this. The referencing system is central to the
operation of HDM-4.
The Road Network folder provides the basic facilities for network referencing within HDM-
4. It allows users to define different networks and sub-networks, and to define sections of
road, which are the fundamental unit of analysis within HDM-4. Whereas sections are
lengths of road over which physical characteristics are reasonably constant, the network
referencing system also supports the concept of nodes that connect sections. All network
data are entered to the Road Network folder, and facilities are also available for modifying,
deleting and maintaining this data. The approach to network referencing is considerably
more flexible than that used in HDM-III, and is designed to handle a wide range of external
referencing conventions as might be used by other systems with which HDM-4 may need to
interface.
1.5.2 Vehicle Fleet

Roads are provided to enable their use by vehicles, and the representation of the vehicles is a
key data requirement for HDM-4. The Vehicle Fleet folder provides facilities for the storage
and retrieval of vehicle fleet details used for calculating vehicle speeds, operating costs,
travel time costs and other vehicle effects.

Version 1.0
The method used to represent vehicle fleets is considerably more flexible than that used in
HDM-III. Motorcycles and non-motorised vehicles can be included, and there is no limit on
the numbers or types of vehicles that can be specified. Multiple vehicle fleet data sets can be
set up for use in different analyses, and a wide range of default data are provided.
1.5.3 Road Works Standards

Road Works Standards refer to the targets or levels of conditions and response that a road
administration aims to achieve. Road agencies set up different standards that can be applied in
practical situations in order to meet specific objectives which are related to functional
characteristics of the road network system.
The Road Works Standards folder provides facilities in a flexible framework to define a list of
maintenance and improvement standards that are followed by road administrations in their
network management and development activities. The standards defined in the Road Works
folder can be used with any of the analysis methods.
1.5.4 HDM Configuration

It is recognised that HDM-4 will be used in a wide range of environments. The HDM
Configuration folder provides facilities to customise the default data set to reflect the
conditions encountered in the environment under study. Basic data and calibration
coefficients can be defined in a flexible manner to minimise the amount of data that must be
changed for each application. Default values are supplied with HDM-4, but these are all
user-definable and facilities are provided to permit data modification.
1.6 Applications
HDM-4 can assist with:
! Project analysis for detailed economic appraisal (see Section 1.6.1).
! Programme analysis for annual or rolling work program preparation (see Section 1.6.2).
! Strategic analysis for long term planning (see Section 1.6.2).
! Research and policy studies (see Section 1.6.3).
1.6.1 Project analysis

Project analysis is concerned with:
Evaluation of one or more road projects or investment options. The
application analyses a road section with user-selected treatments, with
associated costs and benefits, projected annually over the analysis period.
Economic indicators are determined for the different investment options.
This is essentially the use for which the earlier generations of HDM systems were developed.
Project analysis is used to estimate the economic or engineering viability of road investment
projects by considering the following issues:
! Pavement structural performance.
! Life cycle predictions of road deterioration, maintenance effects and costs.
! Road user costs and benefits.
! Economic comparisons of project alternatives.

Version 1.0
Typical appraisal projects would include the maintenance and rehabilitation of existing roads,
widening or geometric improvement schemes, pavement upgrading and new construction.
Case studies demonstrating the application of Project analysis are presented in Part B.
The philosophy of this analysis has not fundamentally changed from the HDM-III version, but
improved road deterioration relationships have been extended to cover a wider range of
pavements and the performance of materials in temperate and cold climates. Road user cost
relationships have been updated and extended to include the impacts of road accidents.
Environmental impacts assessment and energy balance analysis is also included.
1.6.2 Programme and Strategy analyses

It is in the areas of programme analysis and strategy analysis that HDM-4 offers significant
improvements over HDM-III.
Programme analysis is concerned with:
Preparation of a multi-year rolling programme for a road network in which
candidate investment options are identified and selected, subject to resource
constraints. Road networks are analysed section by section and estimates
are produced of road works and expenditure requirements for each section for
each year of the funding period.
Strategy analysis is concerned with:

Analysis of a chosen network as a whole, for preparing long range planning
estimates of expenditure needs for road development and conservation under
different budget scenarios. The road network is characterised by lengths of
road in different categories defined by parameters such as road class, surface
type, and pavement condition or traffic flow. Estimates are produced of
expenditure requirements for medium to long term periods of 5-40 years.
The main difference between strategy and programme analysis is the way in which sections
are physically identified:
Deals with sections that are unique physical units identifiable from the road network
throughout the analysis.
! Strategy analysis
Considered as grouped representative sections typical of the road network to be analysed.

The road system essentially loses its individual section characteristics.
For both types of analysis, the problem can be posed as one of seeking that combination of
treatment alternatives across a number of sections in the network that optimises an objective
function under budget constraint. If, for example, the objective function is the maximising of
NPV (Net Present Value), the problem can be defined as:
Select that combination of treatment options for sections that maximises NPV
An important difference between strategic and programme analysis, and that for project
analysis, is in the detail at which data is defined. Use is made of the concept of Information
Quality Levels (IQL) recommended by the World Bank (Paterson and Scullion, 1990).
Project level analysis data is specified in terms of measured defects (IQL-II), whereas the
specification of data for strategic and programme analyses can be more generic (IQL-III).
For example:

Version 1.0
! Project level analysis
Roughness would be specified in terms of m/km IRI.

! Strategic and programme analysis
Roughness could be specified as good, fair or poor.

The relationship between IQL-II and IQL-III level data depends on road class, pavement
surface type and traffic class, and is user-defined. The relationships are defined in the HDM
Configuration folder, or defaults are available.
Case studies demonstrating the application of Programme and Strategy analyses are included
in Appendices B and C respectively.
1.6.3 Research and Policy studies

HDM-4 can be used to conduct a number of road sector policy studies including:
! Funding policies for competing needs; for example, feeder roads versus main roads
! Road user charges for setting up road funds
! Impacts of road transport policy changes on energy consumption
! Impact of axle load limits
! Pavement maintenance and rehabilitation standards
Relevant case studies are included in Part E.
1.7 User interface

A key objective for the new system is that it should be more user-friendly than HDM-III.
This has been achieved by addressing the issues of user interface and data requirements.
! User interface
This has been improved by developing the system from the outset to run on Personal
Computers operating under the Microsoft Windows environment. System design is
modular to enable different technical modules to be used relatively easily and to facilitate
future system operation on other non-Windows platforms, such as UNIX, Linux, etc.
More details of system issues are described in the Software User Guide.
! Data requirements
The total data input to HDM-4 is greater than in HDM-III to allow for the extended
facilities that are available. The concept of a data hierarchy is used, whereby defaults
are set for many data items in the HDM Configuration folder. Users may choose the
extent to which defaults are used, and run the system using a level of data entry detail
that is appropriate to their particular application.
A comprehensive Help system is being designed and will be available in a later release of the
HDM-4 system.

Version 1.0
2 References
Abaynayaka, S.W., Morosiuk G., and Hide H., (1977)
Prediction of road construction and vehicle operating costs in developing countries.
Proceedings of the Institution of Civil Engineers. 62 (Part 1), 419-446.
Archondo-Callao, R. and Purohit R., (1989)
HDM-PC: user's guide. The Highway Design and Maintenance Standards Series.
Archondo-Callao, R., (1994)
HDM Manager Version 3.0. Transportation Division, Transportation, Water & Urban
Development Department. Washington DC: The World Bank.
CRRI, (1982)
Road user cost study in India. Final Report, New Delhi: Central Road Research Institute.
Cundill, M.A., and Withnall S.J., (1995)
Road transport investment model RTIM3. In: TRB. Sixth International Conference on
Low-Volume Roads, Conference Proceedings 6, Volume 1. Washington DC: National
Academy Press, 187-192.
FHWA, (1994)
Highway economics requirements system. Washington DC: Federal Highway
Administration.
Geipot, (1982)
Research on the interrelationships between costs of highway construction, maintenance
and utilisation (PICR). Final report, 12 volumes, Brasilia, Brazil.
Harral, C.G., and others, (1979)
Warwick, 13-16 July 1979. London: PTRC Education and Research Services.
Hide, H., (1982)
Vehicle operating costs in the Caribbean: results of a survey of vehicle operators. TRRL
Laboratory Report 1031. Crowthorne: Transport and Road Research Laboratory.
Hoban, C.J., (1987)
Evaluating traffic capacity and improvements to geometry. Technical Paper Number 74.
Kerali, H.R., Parsley L.L., Robinson R., and Snaith M.S., (1985)
Development of a microcomputer based model for road investment in developing
countries. In: CIVILCOMP. Proceedings of the Second International Conference on
Civil and Structural Engineering Computing, London, December 1985. London:
Institution of Civil Engineers.

Version 1.0
Moavenzadeh, F., Stafford J.H., Suhbrier J., and Alexander J. (1971)

Highway design study phase I: the model. IBRD Economics Department Working Paper
No 96. Washington DC: International Bank for Reconstruction and Development.
Moavenzadeh, F., Berger F., Brademeyer B., and Wyatt R. (1975)
The Highway Cost Model: General Framework. MIT Department of Civil Engineering
Research Report No 75-4. Cambridge Mass: Massachusetts Institute of Technology.
Morosiuk, G., and Abaynayaka S.W., (1982)
Vehicle operating costs in the Caribbean: an experimental study of vehicle performance.
TRRL Laboratory Report 1056. Crowthorne: Transport and Road Research Laboratory.
Parsley, L.L., and Robinson R., (1982)
The TRRL road investment model for developing countries (RTIM2). TRRL Laboratory
Report 1057. Crowthorne: Transport and Road Research Laboratory.
Paterson, W.D.O., and Scullion T., (1990)
Information systems for road management: draft guidelines on system design and data
issues. Infrastructure and Urban Development Department Report INU77. Washington
DC: The World Bank.
Phillips, S., (1994)
Development of the United Kingdom pavement management system. In: TRB. Third
International Conference on Managing Pavements, San Antonio, Texas, 22-26 May
1994, Volume 2. Washington DC: National Academy Press, 227-236.
Robinson, R. Danielsson U., and Snaith M.S., (1998)
Road Management: Concepts and Systems. Basingstoke and London: Macmillan Press.
The highway design and maintenance standards model volume 1: description of the
HDM-III model. The Highway Design and Maintenance Standards Series. Baltimore:
John Hopkins for the World Bank.

Version 1.0
Contents
Volume 2 - Applications Guide: Part B

B1 Concepts of Project Analysis
1 Introduction B1-1
2 Procedure for Project Analysis B1-3
2.1 Creating a project B1-4
2.2 General information B1-4
2.3 Network B1-5
2.4 Vehicle types B1-6
2.5 Traffic characteristics B1-6
2.6 Alternatives to be analysed B1-7
2.7 Running the analysis B1-10
2.8 Print outputs B1-11
2.9 Sensitivity and scenario analyses B1-11
2.10 Description of HDM-4 outputs B1-11
2.11 Input Data Definitions: Project Appraisal B1-16
B2 Concepts of Programme Analysis

1 Introduction B2-1
1.1 Applications B2-1
1.2 Analysis Methods B2-2
1.3 Life cycle analysis B2-2
1.4 Multi-year forward programme B2-4
2 Procedure for Programme Analysis B2-7
2.1 Basic data set B2-7
2.2 Selection of sections B2-7
2.3 Selection of vehicle types B2-9
2.4 Specification of traffic characteristics B2-9
2.5 Specification of maintenance and improvement standards B2-9
2.6 Specification of budget constraints B2-10
2.7 Output of results B2-10
3 Work programming B2-11
Version 1.0
PART C RESEARCH, POLICY AND REGULATION ANALYSES CONTENTS
3.1 Definition of investment alternatives B2-11

3.2 Assignment of standards B2-11
3.3 User selection of project alternatives B2-13
4 Optimisation B2-14
4.1 Total enumeration B2-14
4.2 Incremental NPV/cost ranking B2-16
B3 Concepts of Strategy Analysis

1 Introduction B3-1
1.1 Concept B3-1
1.2 Background B3-2
1.3 Purpose and approach to the analysis B3-5
2 Procedure for strategy analysis B3-7
2.1 Basic data set B3-7
2.2 Road network definition and categorisation parameters B3-7
2.3 Selection of vehicle fleet B3-8
2.4 Specification of traffic characteristics B3-9
2.5 Specification of maintenance and improvement standards B3-9
2.6 Specification of budget constraints B3-9
2.7 Output of results B3-9
3 Strategy analysis B3-10
3.1 Selection of investment options B3-10
3.2 Life-cycle analysis B3-10
3.3 Optimisation B3-11
4 References B3-13
Version 1.0
PART B CONCEPTS OF ANALYSIS
B1 Concepts of Project Analysis
1 Introduction
This chapter describes the overall procedure for project analysis in HDM-4. This allows users
to assess the physical, functional and economic feasibility of specified project alternatives by
comparison against a base case, or a without-project alternative. The key processes of the
analysis are the:
! Life cycle costing of pavement structural performance.
! Prediction of road deterioration.
! Estimation of road user costs (vehicle operating costs, travel time and accidents).
! Modelling of road works effects and the costs of these to the road administration.
! Calculation of economic or financial benefits from comparisons of the project
alternatives.
The aim is to determine which project alternative is most cost-effective.
Four case studies to demonstrate its application are described in Appendices A1, A2, A3, and
A4.
Project analysis is associated with the following types of road projects:
! Maintenance of existing roads
These works cover a wide range of maintenance techniques for different pavement types.
This type of works also includes those that arise when a road pavement has received
insufficient maintenance over its design life, or because the pavement was not built to the
standards of quality required by the original design (sometimes known as rehabilitation).
! Improvement of existing roads
These projects aim to provide additional capacity when a road is nearing the end of its
design life or because there has been an unforeseen change in use of the road. Works
include measures to improve the quality of service, such as relieving traffic congestion,
road safety, road passibility, the need to accommodate increased vehicle axle loads, etc.
Typical examples of road improvement projects are:
❏ Pavement reconstruction
❏ Pavement upgrading
The paving of gravel roads, the provision of concrete pavements in places where the
subgrade soils may be very poor etc.
❏ Road widening and geometric improvements
Includes widening of existing carriageways, the provision of additional lanes, dual-
carriageways, overtaking lanes, climbing lanes, and the betterment of shoulders.
❏ Realignment and pavement upgrading of the existing roadway
Combines the activities described in the above two items (Pavement upgrading and
road widening).
Applications Guide B1-1

Version 1.0
PART B CONCEPTS OF ANALYSIS B1 CONCEPTS OF PROJECT ANALYSIS
! New construction
Involves the construction of a new pavement in an entirely new location, although in

many cases they follow existing alignments. New alignments are required, for example,
for bypasses, or in difficult terrain, or to connect other new infrastructure developments.
! Stage construction
Consists of planned improvements to the pavement standards of a road at fixed stages

throughout the project life. Often, the road alignment needed at the final stages of the
project is provided from the outset. A typical strategy might be initially to construct a
gravel road that will be paved when traffic flows have reached a given level.
! Evaluation of past projects
Assesses the performance of a completed project to see if objectives set out during
appraisal have been met. Project evaluation requires project data that have been collected
and recorded in a systematic way throughout all stages of the project cycle. The
evaluation should result in specific recommendations about improving aspects of the
project design that can be used to improve ongoing and future planning.

Version 1.0
2 Procedure for Project Analysis

A project is defined as the set of road works to be carried out on one or more road sections
that can be grouped together conveniently to be undertaken as one contract or work
instruction. Several project alternatives can be analysed to determine, for example, which is
the most cost-effective to implement. A project alternative could consist of different works
options applied to various sections making up the project. A section option (or section
alternative) consists of one or more works standards combined logically. For example; the
resurfacing of a road section constitutes a section option; similarly the continuation with
existing maintenance practices would constitute another option for the same section.
Figure B1.1 illustrates the relationships between project alternatives, section alternatives,
standards and the constituent works for a particular project. It also shows that certain road
works may generate additional traffic on a road section, or cause traffic to divert from other
road links and modes of transport to a road section. The generated and/or diverted traffic
characteristics would need to be specified for such works.
Project
Project
Diverted
Divertedtraffic
traffic Project
Project Project
Project Project
Project etc
etc
alternative
alternative11 alternative
alternative33
Generated traffic
Exogenous Benefits
Section
Section11 Section
Section22 Section
Section33 etc
etc
and Costs alternative
alternative22
Maintenance
Maintenance Maintenance
Maintenance Improvement
Improvement etc
etc
Standard
StandardAA Standard
StandardBB Standard
StandardYY
[Operation
[Operation1]1]++[Intervention
[Interventioncriteria]
criteria]++[Cost]
[Cost]==“Works
“Works1”
1”
[Operation
[Operation1]1]++[Intervention
[Interventioncriteria]
criteria]++[Cost]
[Cost]==“Works
“Works2”
2”
etc
etc
Figure B1.1 Relationships between project alternatives, section alternatives,

standards and operations
The procedure for project analysis is summarised as follows:

1 Create the project to be studied by defining the road project to be analysed by giving it a
title, specifying the road network to be analysed, and the vehicle fleet that uses the road
network (see Section 2.1).

Version 1.0
2 Define the project by specifying the following:

(a) general information about the project (see Section 2.2).
(b) road sections to be analysed (see Section 2.3).
(c) vehicle types using the selected road sections (see Section 2.4).
(d) traffic characteristics for the selected sections (see Section 2.5).
(e) alternatives to be analysed, in terms of section alternatives and project
alternatives (see Section 2.6).
3 Set up the analysis run, including specifying the comparisons to be undertaken, the
discount rate, choices about which vehicle life model to use, what road safety analysis to
be undertaken, and whether or not to perform emissions analysis and energy balance
analysis (see Section 2.7.1).
4 Run the analysis (see Section 2.7.2).
5 Select the required outputs (see Section 2.10).
6 Print required outputs (see Section 2.8).
7 Optionally, it will be possible for users to carry out sensitivity analyses by varying the
values of selected input parameters (see Section 2.9).
Each of these tasks is described in turn in the following sections of this chapter.
2.1 Creating a project

Specifying the following information creates a project:
! Project title
! Road network
Name of the network that includes the sections to be analysed.

! Vehicle fleet
Name of the vehicle fleet that includes the vehicle types that use the road network.
The components of the project are then specified in detail.
2.2 General information

For the project being studied, the following information is specified:
! Description
A brief description of the works involved in the project.

! Start year of the analysis period
! Duration of the analysis period
! Output currency
The units of currency in which the project outputs are to be presented.

Version 1.0
! Conversion rate
The conversion factors between the output and vehicle fleet currency units, and that
between output and road works currency units used in the analysis.
2.3 Network
2.3.1 Methods of selection

The sections to be analysed are selected and retrieved from the Road Network folder using
one of the following two methods:
1 Selecting the sections individually (see Section 2.3.2)
2 Selection based on a set of criteria (see Section 2.3.3)
One or more sections may be selected for a study.
2.3.2 Individual section selection

Individual sections can be added or removed from the list of those available in the particular
road network identified when the project was created, as above. The resulting list of sections
is that used for subsequent analysis.
2.3.3 Selection criteria

An alternative approach to selecting sections is to define the criteria that should be used by
HDM-4 to select sections from those in the network that were defined when the project was
created. The following criteria are available:
! Pavement
❏ Surface class
All types of surfacing, or one of: bituminous, concrete or unsealed surfacings.
❏ Structural adequacy
Pavements of all strength, or those of: poor, fair or good strength.
! Speed-flow type
All speed-flow types or one of the speed-flow types pre-defined in the Road Network
folder.
! Traffic volume
Road sections carrying any traffic volume, or those carrying one of: low, medium or
high traffic.
! Geometry
Those road sections with minimum or maximum values of:

❏ Rise plus fall (m/km)
❏ Horizontal curvature (deg/km)
❏ Width (m)

Version 1.0
The selection criteria can be used to add new sections to those already selected, or to replace
those selected. At any time during the selection process, a list of those sections currently
selected can be displayed.
2.4 Vehicle types

The vehicle types that use the selected road sections are retrieved from a pre-defined vehicle
fleet comprising vehicle classes or types defined in the Vehicle Fleet folder. HDM-4 base
vehicle types can also be displayed for selection. Hence vehicle types, to be included in the
analysis, can be selected from either of these lists.
2.5 Traffic characteristics

Traffic characteristics are specified for each of the selected road sections. The following
information is required:
! Network element type - section or node
! Annual Average Daily Traffic (AADT) - on section
! Year in which this traffic level applies
Traffic details can be specified for each network element, as follows:

! Total AADT in the current year
! Current year
! Initial composition
Specified in terms of a percentage of the total flow, by vehicle type.

! Annual growth rate
Specified for each vehicle type, for each specified growth period.
A traffic growth period can be added (or amended) by specifying the following:
! Start year of the traffic growth period
! Growth type for normal traffic
❏ Annual percentage increase

❏ Annual increment in AADT
❏ Actual increase in AADT
See Chapter D5 for further information.
Negative traffic growth rates and zero traffic volumes are permitted to allow for changing
vehicle fleet characteristics.
Other required traffic related data (for example, speed-flow characteristics and hourly traffic
flow distributions) are discussed in Chapter D5.

Version 1.0
2.6 Alternatives to be analysed
2.6.1 Section alternatives

The assignment of one or more standards, combined logically, to a road section results in the
formation of a section alternative. A section alternative is one of a set of mutually exclusive
works alternatives specified as options to be analysed for an applicable road section within a
project. The basic unit of analysis in HDM-4 is the section alternative.
A section alternative is defined by the following:
! Description
Typically do something or do nothing or base case.

! Maintenance and/or improvement standards
Apply standards to the section alternative, giving for each the date from which the
standard is effective.
! Generated traffic
See Chapter D5, if applicable.

! Exogenous benefits and costs
These are benefits and costs that are not calculated within HDM-4. These are specified
using the following data:
❏ Description
❏ Starting year relative to the beginning of the road works
❏ Amounts of the costs or benefits in economic and/or financial terms
❏ Annual growth rate in terms of percentage increase, incremental amounts, or actual
amounts
The specified alternative can be edited to alter the description of the alternative, or to add or
change the standards that apply.
2.6.2 Project alternatives

A project alternative may comprise one or more section alternatives. Each of the road sections
within the project is assigned a set of standards. For example, each of the columns numbered 1
to 4 in Table B1.2 constitutes a project alternative. A project alternative is therefore defined
by the following parameters:
! Description of the project alternative
! List of constituent section alternatives
! Diverted traffic
Measured in terms of new AADT, traffic composition and growth rates on all the
sections affected (see Part B of the Analytical Framework and Model Descriptions).
2.6.3 Analysis methods

Two methods for analysing investment options are provided in HDM-4 as follows:

Version 1.0
! Analysis by section
Analyses, individually, each of the road sections that make up the project. Several
alternatives can be defined for each section as shown in Table B1.1 (for example, three
alternatives for section A, four alternatives for section B, etc.), with one alternative
designated as the base case against which all the other alternatives will be compared.
Economic indicators (for example, NPV, IRR, BCR and FYB) are calculated for each
section alternative.
Table B1.1
Analysis by sections
Road Section alternatives

sections
1 2 3 4 5
Section A RM Resealing Overlay
Section B RM Overlay Reconstruction Widening
Section C RM Resealing Rehabilitation Lane addition Realignment
Section D Grading Regravelling Paving
Note: RM = Routine maintenance, and includes: patching, crack sealing, and drainage
works.
! Analysis by project
Analyses road sections together as a package by considering project alternatives as the

basic unit for performing economic analysis. First, the annual costs and benefits are
summed over all the section alternatives within each project alternative to give yearly
totals. Economic indicators are then calculated for each project alternative by comparison
against a base case alternative. Details of these are described in Part G of the Analytical
Framework and Model Descriptions. Analyses involving new sections and diverted
traffic can only be performed using this method.
Table B1.2
Analysis by projects

sections
1 2 3 4
Section A RM Resealing Reconstruction Realignment
Section B RM Overlay Mill and replace Widening
Section C RM Inlay Reconstruction Lane addition
Section D Grading Regravelling Widening Upgrading
ΣProject NPV 0 3.04 4.72 3.85
works.

Version 1.0
2.6.4 New sections

A new road section should be specified as a section alternative within a selected project
alternative. Different specifications of design and construction details can be defined as
separate section alternatives under different project alternatives.
Table B1.3 shows the definition of a project involving the construction of a new section E (for
example, a by-pass), and the maintenance of four existing road sections (A, B, C and D)
which are affected by the introduction of the new link mainly in terms of traffic diversion
effects. Four project alternatives are defined as follows:
! Project alternative 1
Describes the base case situation; that is, without a new road section.
Includes a new section, E, a 2-lane AMSB (asphalt mix on stabilised base) road.
Includes new section, E, a 4-lane AMSB road.

Includes new section, E, a 4-lane STGB (surface treatment on granular base) road.
Table B1.3
Analysis involving new sections

sections
1 2 3 4
(base case)
Section A RM RM RM RM
Section B RM RM RM RM
Section C RM RM RM RM
Section D RM RM RM RM
New Section AMSB AMSB STGB
2-lane 4-lane 4-lane
ΣProject NPV 0 3.04 4.72 3.85
works.
This project can be analysed to determine whether or not the construction of the new section is
cost-effective and what design/construction specifications to follow.
A new section can only be scheduled and not triggered by responsive intervention criteria.
A new section can be added by specifying the following:
! General
❏ Description of the road section

❏ Start year of new construction

Version 1.0
! Construction costs
Costs data consisting of total costs, duration of works, salvage value, and percentage of
total expenditure over each year of the works duration.
! Road works standards
Maintenance and/or improvement standards to apply to the section alternative, giving for
each the date from which the standard is effective.
! Normal traffic
! Section details
Input data as specified in the Road Network folder.

! Generated traffic
Includes any generated traffic resulting from the works option, in terms of the following:
❏ Description
❏ Starting year relative to the beginning of the improvement works
❏ Traffic volumes by vehicle type for the remainder of the analysis period
Further details on specifying generated traffic for a section alternative are discussed in
Part B of the Analytical Framework and Model Descriptions.
Benefits and costs associated with the works option that are significant; they are specified
using the following data:
❏ Description
❏ Starting year relative to the beginning of the improvement works
❏ Monetary value of the cost or benefit in financial and economic terms
Standards can also be added or deleted.
2.7 Running the analysis
2.7.1 Run set-up

In the run set-up, the user can specify the technical models to be used. These include:
! Economic analysis
If economic analysis is to be performed, the following parameters should be specified to

define a study:
❏ Base case project alternative (project analysis)
The base case against which all other project alternatives are compared. Note that
for section analysis, the first section alternative is automatically assigned as the base
case.
❏ Discount rate (%)

Version 1.0
! Road accident analysis
Including whether or not costs are to be included and, if so, the unit costs for accidents in
terms of fatal, injury only, damage only, or all accidents.
! Energy balance analysis
! Vehicle emissions
! Acceleration effects
2.7.2 Run analysis

Following set-up, selecting the Start button under Analyse Projects/Run Analysis runs the
project analysis. During the run, the status is displayed in terms of the stage of analysis
reached and the percentage completion. The analysis can also be stopped at any time by
clicking the Abort button.
2.8 Print outputs

Users are able to select the types of output required from a list of standard outputs produced
by HDM-4. These are listed in Section 2.10. In addition, users may specify customised
reports; details of these are given in the Software User Guide.
2.9 Sensitivity and scenario analyses

These analyses have not been implemented in this software release. Sensitivity and scenario
analyses can be carried out by varrying the magnitude (within appropriate ranges) of the
important variables to determine the net effect on calculated economic indicators. The analyses
are performed typically by specifying different discount rates and multipliers for the following:
! Project costs
! Costs of the maintenance or improvement operations
! Travel time costs
! Accident costs
2.10 Description of HDM-4 outputs
2.10.1 Introduction
Standard outputs that can be generated from HDM-4 are described. The outputs have been
categorised as follows:
! Tabulated input data
! Results of analyses
Input data
Printout of tabulated input data, with appropriate headings, is useful for checking, reviewing
and updating the information that will be used in a study. All key input data, which affect the
results of analyses significantly, should be printed to accompany the reports. Careful checking
of all input data, preferably verified on hard copy, is strongly advised.

Version 1.0
The input data used in a project analysis are grouped as follows:

! Analysis parameters
! Road network characteristics
! Vehicle fleet data
! Traffic data
! Investment standards and alternatives
Analysis results
The results of analyses to be performed within HDM-4 have been grouped as follows:
! Deterioration/works effects
! Road user effects
! Environmental effects
! Energy use
! Cost streams
! Others
! User-definable reports
2.10.2 Details of analysis results

These are presented in tabular format, but where appropriate, a graphical format (chart) is also
available.
Deterioration/works effects
This group includes the following types of reports:
! Timing of works
! Pavement condition versus time
! Pavement strength
Structural Number of Pavement (SNP) versus time.

! Changes in road surface types during the analysis period
! Traffic data
❏ Annual Average Daily Traffic (AADT)

❏ Vehicle axles (YAX)
❏ Equivalent Standard Axle Loads (ESAL)
! Maintenance works
Quantities involved in each section alternative.

! Summary deterioration reports
Includes works quantities, average condition and traffic data (by alternative).

Version 1.0
Road user effects

The reports involving road user effects include the following parameters:
! Vehicle Speeds
❏ Free speeds
❏ Average operating speeds
❏ Operating speeds by traffic flow periods
! Traffic flows
❏ Hourly traffic flows (PCSE/hr)

❏ Volume-capacity ratios
❏ Normal and diverted traffic (vehicles per day)
❏ Generated traffic (vehicles per day)
❏ Total traffic (vehicles per day)
! Vehicle resource consumption
❏ Fuel consumption
❏ Lubricating oil consumption
❏ Tyre consumption
❏ Parts consumption
❏ Maintenance labour hours
❏ Crew hours
❏ Depreciation
❏ Interest
! Time resource consumption
❏ Passenger working hours

❏ Passenger non-working hours
❏ Cargo transit hours
! Work zone effects (not included in this software release)
❏ Delay hours
❏ Additional vehicle operating costs
! Road safety
❏ Number of fatal accidents

❏ Number of injury accidents
❏ Number of damage-only accidents
❏ Total number of accidents

Version 1.0
! Summary reports
Road user effects summary reports by investment alternatives.
Environmental effects
The reports on environmental effects include the following:
! Vehicle emissions
Time streams of quantities for the following components:

❏ Hydrocarbons (HC)
❏ Carbon monoxide (CO)
❏ Nitrogen oxide (NOx)
❏ Sulphur dioxide (SO2)
❏ Carbon dioxide (CO2)
❏ Particulates (Par)
❏ Lead (Pb)
! Net change in vehicle emissions
For each comparison of investment alternatives.

! Traffic noise levels
Measured at selected sites (worst performance level).

Note: The modelling of traffic noise is not included in this software release.
Energy use
The reports on energy use are available as follows:
! National energy use and global energy use
For each investment alternative.

! Renewable and non-renewable energy use
For each investment alternative.
Cost streams
The reports of cost streams are divided into the following subgroups:
! Road agency costs
These costs will be reported under user-defined budget categories for each of the
following cost categories:
❏ Economic
❏ Financial
! Road user costs
Measured in economic and financial terms.

Version 1.0
❏ Vehicle operating costs

❏ Travel time costs
❏ Accident costs
❏ Total road user costs
! Net exogenous costs
! Total transport costs
! Cost streams summary reports by investment alternatives
Economic comparison summaries/indicators

The reports on economic analysis include the following:
! Economic indicators
❏ Net Present Value (NPV)

❏ Internal Rate of Return (IRR)
❏ Benefit Cost Ratio (BCR)
(or Net benefit per investment cost, NPV/C)
❏ First-Year Benefits (FYB)
! Present values of costs
❏ Road agency costs

❏ Road user costs
❏ Net exogenous costs
❏ Total transport costs
! Present values of benefits
❏ Decrease in road agency costs

❏ Savings in road user costs
❏ Net exogenous benefits
❏ Savings in total transport costs
User-definable reports
Users may specify their own desired types of reports, which are different from those presented
above. The user-definable reports are generated using proprietary report writer software
(Crystal Reports). For example users may wish to adopt the following criteria:
! Aggregation of analysis results
For example, according to vehicle classes and road links.

! Export reports
Selection of tables or different types of reports to be exported to spreadsheets and

amalgamated into single reports.
Guidance on the production of user-definable reports is given in the Software User Guide.

Version 1.0
2.11 Input Data Definitions: Project Appraisal

To be included in a subsequent edition of this document

Version 1.0
B2 Concepts of Programme Analysis
1 Introduction
1.1 Applications
Programme Analysis deals primarily with the prioritisation of a defined long list of candidate
road projects into a one-year or multi-year work programme under defined budget constraints.
It is essential to note that here, we are dealing with a long list of candidate road projects
selected as discrete segments of a road network. The selection criteria will normally depend
on the maintenance, improvement or development standards that a road administration may
have defined (for example from the output produced by the strategy analysis application).
Examples of selection criteria that may be used to identify candidate projects include:
! Periodic maintenance thresholds (for example, reseal pavement surface at 20% damage).
! Improvement thresholds (for example, widen roads when volume/capacity ratio exceeds
0.8).
! Development standards (for example, upgrade gravel roads to sealed pavements when the
annual average daily traffic exceeds 200 vehicles per day).
The above examples do not imply firm recommendations to be used by road authorities.
When all candidate projects have been identified, the HDM-4 programme analysis application
can be used to compare the life cycle costs predicted under the existing regimen of pavement
management (that is, the without project case) against the life cycle costs predicted for the
periodic maintenance, road improvement or development alternative (that is, with project
case). This provides the basis for estimating the economic benefits that would be derived by
including each candidate project within the budget timeframe.
It should be noted that the main difference between strategy analysis and programme analysis is
the way in which road links and sections are physically identified. Programme analysis deals
with individual links and sections that are unique physical units identifiable from the road
network throughout the analysis. In strategy analysis, the road system essentially loses its
individual link and section characteristics by grouping all road segments with similar
characteristics into the road network matrix categories.
For programme analysis, the problem can be posed as one of seeking that combination of road
work alternatives across a number of sections in the network that optimises an objective
function under budget constraint. If, for example, the objective function is to maximise the
Net Present Value (NPV), the problem can be defined as:
Select that combination of road work options for sections that maximises NPV
The HDM-4 programme analysis application may be used to prepare a multi-year rolling
programme, subject to resource constraints. The prioritisation method employs the
incremental NPV/cost ratio as the ranking index, described in more detail in Analytical
Framework and Model Descriptions. This provides an efficient and robust index for
prioritisation purposes. Indices such as the NPV, economic rate of return (ERR), or predicted
pavement condition attributes (for example, road roughness) are not recommended as ranking
criteria. The incremental NPV/cost ratio satisfies the objective of maximising economic

Version 1.0
PART B CONCEPTS OF ANALYSIS B2 CONCEPTS OF PROGRAMME ANALYSIS
benefits for each additional unit of expenditure (that is, maximise net benefits for each
additional $1 of the available budget invested).
1.2 Analysis Methods

The purpose of programme analysis is to evaluate maintenance or improvement options, and
to try and select the set of investments to be made on a number of road sections in a road
network, which will optimise an objective function. Programme analysis is concerned with
short to medium term planning and preparation where budget levels are known with
reasonable certainty. Thus, the problem can be posed as one of searching for the combination
of investment alternatives that optimises the objective function, under a budget constraint.
Two methods of analysis are provided in programme analysis:
1 Life cycle analysis
2 Multi-year forward programme.
The two methods are described in detail later in Section 3.
In both cases, optimisation is done using the incremental NPV/Cost ratio where the problem
can be defined as the selection of that combination of investment options on sections that
maximises NPV/cost for the selected sections in the road network subject to the sum of the
investment costs being less than the budget available.
Note that the set of investment options to be optimised is user-defined and is not the set of all
possible options for the particular network; hence the problem is not true optimisation since
all possible solutions are not normally considered. Note also that the investment options on
any one section are mutually exclusive.
1.3 Life cycle analysis

The underlying life cycle analysis concept implemented in HDM-4 is the same for project,
programme and strategy analysis. In each case, HDM-4 predicts the life cycle pavement
conditions and costs over a specified analysis period under a user-specified scenario of
circumstances. The primary cost set for the life cycle analysis includes the costs of capital
investment, maintenance and vehicle operating costs, to which travel time costs can be added
as an option. The cost of accidents, energy and/or environmental pollution may also be
The broad concept of the life cycle analysis is illustrated in Figure B2.1. Interacting sets of
costs, related to those incurred by the road administration and those incurred by the road user,
are added together over time in discounted present values. Costs are determined by first
predicting physical quantities of resource consumption and then multiplying these quantities
by their unit costs or prices. Economic benefits are then determined by comparing the total
cost streams for various maintenance and construction alternatives with a base case (do
nothing or do minimum alternative), usually representing minimal routine maintenance.

Version 1.0

geometry, pavement characteristics,
unit costs



Road geometry and surface texture, Social and Levels of emissions and energy
vehicle characteristics Environmental Effects used, and number of accidents


by section
Figure B2.1 HDM-4 Life Cycle Analysis Concept

Version 1.0
1.4 Multi-year forward programme

For most road agencies, the short term planning cycle is normally based on a one year budget-
period. However, some road agencies adopt a multi-year tactical planning period. A common
example of a multi-year forward programme is a three-year work programme, which could be
considered as three calendar year budget periods.
The multi-year forward programme option in HDM-4 is a simplification of the life-cycle
analysis method based on comparison of the following alternatives:
! Apply the assigned road works within the budget period, or
! Postpone road works until the first year after the budget period.
This simplification is normal practice in work programming and it can be used to keep the
analysis simple and reduce run times because there is usually a large number of network
elements to be analysed. In this method, the work programme is also prepared separately for
each year in the budget period, and the results are summed up for the budget period.
Economic calculations are done by comparing investments made within the budget period
against deferring the action to the first year after the budget period as illustrated in Figure
B2.2. For example, an overlay applied in the first year of the budget period year would be
compared against the road work required after the budget period. The road works are not
necessarily the same, because the road section could be in worse condition after postponement
of the overlay, and this might no longer be applicable on the road section.
IRI
2 years
Deferred investment
Investment now
Benefits due to action
Years
Figure B2.2 Calculation of benefits for a multi-year work programme
The discounted costs and benefits of road works are evaluated only one year beyond the
budget period. Thus, there are two issues to be considered under this approach:
! Determination of the residual value of the investment
! Estimation of future benefits accruing due to the investment
A simplifying assumption is made that the deferred action will be less effective in improving
pavement condition as shown in Figure B2.2. It is also assumed that the pavement
performance after the next year will be similar for both alternatives. Therefore, there is no
need to calculate pavement deterioration and road user costs beyond the first year after the

Version 1.0
budget period. This approach is a modification of that used in the United States Highway
Economics Requirements System (HERS) and in the UK Highway Agencies PMS (HAPMS).
The following two examples in Table B2.1 and Table B2.2 illustrate the above process.
Table B2.1
Example of one-year forward programme
Study title: [Capital budget analysis]

Section Year 1 Road Work required
after budget period
S1 OL50 OL50
S2 OL50 REC
S3 OL50 OL50
Budget Needs 180
Budget Constraint 100
In Table B2.1, a one-year work programme for three sections is analysed. An overlay is
triggered for all sections, however, the budget would be exceeded. Two alternatives are
automatically analysed for each road section in order to calculate economic indicators for the
optimisation:
Alternative 1 Alternative 2
S1 Overlay at year 1 vs. Do nothing at year 1 and overlay at year 2
S2 Overlay at year 1 vs. Do nothing at year 1 and reconstruction at year 2
S3 Overlay at year 1 vs. Do nothing at year 1 and overlay at year 2
The second example in Table B2.2 illustrates a multi-year work programme in which an
overlay is triggered in the second year of the Budget Period. Technically the overlay could be
postponed beyond the budget period, but after that reconstruction would be needed.
The base case in this example is Alternative 2, and the economic benefits are calculated by
comparing Alternative 1 for each of the sections against alternative 2. The results are saved
for the optimisation process described in Section 4.
Table B2.2
Example of multi-year forward programme
Alternative 1 Alternative 2
(Road works applied within (Road works required
Sections the budget period) after the budget period)
Year 1 Year 2 Year 3 Year 4

S1 Reseal Overlay Reconstruction
S2 Overlay Reconstruction
S3 Resealing Overlay
S4 Overlay Reconstruction

Version 1.0
To analyse the above problem, HDM-4 will automatically analyse the equivalent of two
project alternatives for each of the road sections. The with project alternative for each road
section is as follows:
1 Section 1
Resealing in the first year, then overlay in the third year of the Budget Period.
2 Section 2
Overlay in the second year of the Budget Period.

3 Section 3
Resealing in the first year of the Budget Period.

4 Section 4
Overlay in the first year of the Budget Period.

The do minimum or without project alternative for each of the road sections is defined
automatically as the road works which would be required if no periodic maintenance (that is,
capital works) is applied within the budget period. In this case the analysis period applied in
the above example is 4 years for all road sections.

Version 1.0
2 Procedure for Programme Analysis

The procedure for programme analysis is summarised as follows:
1 Define the basic data set for the analysis (see Section 2.1)
2 Select sections for analysis (see Section 2.2)
3 Select the vehicle fleet for the road network (see Section 2.3)
4 Define traffic characteristics for selected sections (see Section 2.4
5 Define maintenance and improvement standards for the selected sections (see Section
2.5)
6 Define budget constraints (see Section 2.6)
7 Execute work programming (see Sections 3 and 4)
8 Finalise and output results (see Section 2.7)
2.1 Basic data set

The programme analysis makes use of the following data:
! Road characteristics - held within the Road Network folder
! Vehicle fleet characteristics - stored in the Vehicle Fleet folder
A programme analysis is characterised by the following information:

! Study title and name label
! Brief description of the study to be undertaken
! Objective function (maximise benefits or improvement in roughness)
! Start and end years of the analysis and budget periods
! Method of analysis (budget period or life-cycle analysis)
! Input and output currencies, and conversion rate between the given currencies
! Discount rate
2.2 Selection of sections

Each study will be carried out on a pre-defined network, accessed via the Road Network
folder. The elements within the network (sections) can be defined in two ways:
1 Entire network
2 Selected set of sections
In the first case, the entire network is selected. In the second case, sections can be selected
through user-defined selection criteria. The parameters available for the selection criteria are
given below:
! Surface class
! Structural adequacy

Version 1.0
! Road type
! Road geometry
! Traffic
! Capacity
! Sub-network (representing geographical or administrative region)
! Link (representing sections to be analysed simultaneously)
The following are examples of the types of selections that might be made:
! All unpaved sections having AADT > 100
! All paved sections with width > 5.0 metres
! All paved sections having AADT > 5000
! Sub-network - North
The group option may be used to combine several sections into a group, which is treated as a
composite unit for analysis. A typical example is a group of consecutive road sections that
form a road link. This kind of analysis might also be used, for example, when a mobile
asphalt or a crushing plant is situated in a different location each year, and the sections near to
this are all to be treated at the same time. For analysis purposes, the results for individual
sections are combined to give a weighted average, which is then compared with other sections
(or groups).
The sub-network option may be used to combine several sections that could be treated as a
composite unit in the reporting of results. Sub-networks differ from links in such a way that
while links are analysed together, road sections in a sub-network are analysed separately and
only summarised for reporting.
Irrespective of the method used, a data set is created as a result of the selection process of the
form shown in Table B2.3.
The user is also able to add or remove one or more sections, but sections in a link or group
cannot be selected individually.
This process is similar to the selection of sections for Project Analysis. Only definitions of
links and sub-networks have been added.

Version 1.0
Table B2.3
Example of selected sections
Scope: [Maintenance]
Section Group or link (optional)
S1 P1
S2 P2
S3 P2
S4 P2
S5 L1
S6 L1
S7 L1
2.3 Selection of vehicle types

The vehicle characteristics to be used in the analyses will have been previously defined in the
Vehicle Fleet folder. Users may select the set of vehicle types to be included in the analysis
from the vehicle fleet.
This process is similar to the selection of vehicle types for Project Analysis.
2.4 Specification of traffic characteristics

For each of the sections to be analysed, it is necessary to specify the average annual daily
traffic (AADT) together with the composition of different vehicle types. The growth in traffic
may be specified for discreet periods defined in terms of the starting year and the type of
growth that may be one of the following:
1 Annual percent growth rate
2 Additional AADT per year
3 New AADT values
If option (3) above is selected, either option (1) or (2) must follow it, unless it is the final year
of the analysis period.
This process is similar to the specification of traffic characteristics for Project Analysis.
2.5 Specification of maintenance and improvement standards

Several maintenance or improvement standards may be specified for the sections selected for
the programme analysis. In the case of multi-year forward programme, analysis will be carried
out by applying only one maintenance or improvement standard for each section, and
selecting one works item (highest in the hierarchy) from the given standard for each year of
the analysis period. In the case of life-cycle analysis, several maintenance or improvement
standards can be analysed for each section (similar to the project analysis). A works item may
comprise maintenance, rehabilitation, reconstruction or improvements (for example,

Version 1.0
upgrading a gravel road). For each of these, the intervention criteria, unit costs, applicable
limits and the effects of the works item must be specified.
This process is generally similar to the specification of Road Works Standards for Project
Analysis.
2.6 Specification of budget constraints

The budget constraints are given for each budget period. Only the total budget is given.
without any division to budget heads.
It is recommended, that these budget constraints do not include the costs of annual routine
maintenance. This assumption will greatly simplify work programming.
2.7 Output of results

The main output expected from a programme analysis module is a schedule of investments
assigned to a selection of road sections to be implemented over a defined period of time. A
sample of a typical output is shown in Figure B2.3. Further outputs are included in the case
studies in Appendices B1 and B2.

1 N1-2 20.5 2 Resealing 2000 5.4 5.4
2 N4-7 23.5 7 Overlay 40mm 2000 10.9 16.3
3 N2-5 12.5 5 Reconstruct 2000 8.6 24.9
4 R312-1 30 4 Widen 4 lane 2000 31.4 56.3
5 R458-3 36.2 3 Overlay 60mm 2000 16.3 72.6
: : : : : : : :
1 N4-16 32.1 6 Reconstruct 2001 22.8 22.8
2 R13-23 22.4 4 Overlay 40mm 2001 9.7 32.5
3 N521-5 45.2 2 Widen 4 lane 2001 41.3 73.8
: : : : : : : :
1 N1-6 30.2 4 Resealing 2002 8.2 8.2
2 N7-9 17.8 3 Overlay 60mm 2002 9.2 17.4
3 F2140-8 56.1 1 Reconstruct 2002 34.9 52.3
: : : : : : : :
Figure B2.3 Typical output of a work programme

Version 1.0
3 Work programming
The Work Programming process is divided into six phases:
1 Selection of maintenance or improvement standards to be used for different
sections (see Section 3.1)
2 Selection of analysis type (multi-year forward or life-cycle) (see Section 3.2)
3 Automatic assignment of works for sections (unconstrained) (see Section 3.2)
4 Identification of pre-defined actions
5 Economic analysis and optimisation (see Section 4)
6 Finalisation of the work programme
3.1 Definition of investment alternatives

The first task in work programming is to assign desired sets of maintenance or improvement
standards to road sections. These constitute the investment alternatives which will be
compared for each road section (or group of sections) in order to calculate the economic
benefits described in Section 1.
Several road sections can be assigned the same set of standards that are applicable to the
pavement surface class. However, users are able to assign any set of standards to any road
section. For example, there might be different sets of maintenance standards for main roads
and for secondary roads, with some improvement standards e.g. upgrading of unsealed road
sections. If an improvement standard is assigned to a road section, this must be followed by a
future maintenance standard.
In addition, there might also be committed road works that have to be included within the
budget. The user is able to pre-select section alternatives prior to the optimisation process (see
Section 3.3 for details).
3.2 Assignment of standards

The assignment of road works to each road section is generally similar to the procedure used
in project analysis if the life-cycle analysis option is selected. However, the assignment is
different for multi-year forward programme.
3.2.1 Multi-year Forward Programme

For multi-year forward programme, only one maintenance standard is allowed per road
section except where an improvement standard has been specified, in which case it must be
followed by a maintenance standard.
The result of this definition process is illustrated in Table B2.4. In this example, an
improvement standard is assigned to sections S2 and S4, which must therefore be followed by
maintenance standards. This process is similar to that of Project Analysis. The only difference
is that the user is able to assign investment options to several sections simultaneously.

Version 1.0
Table B2.4
Example of assignment of standards
Type of analysis: [Multi-Year Forward Programme]

Section Surface Maintenance Improvement Future
class maintenance
S1 B Ideal (B) - -
S2 U Basic (U) Upgrade to bituminous Ideal (B)
S3 U Ideal (U) - -
S4 B Basic (B) Widen to 3 lanes Ideal (B)
S5 B Ideal (B) - -
S6 B Ideal (B) - -
Notes: U = Unpaved B = Bituminous
3.2.2 Life-cycle analysis

The assignment of works in life-cycle analysis is identical to that in the Project Analysis
module (see Part B). The life cycle analysis is performed for each section, and each
maintenance and/or improvement standard selected in Section 3.1. As in project analysis, the
user has to define the base policy (usually do minimum option) which is used in the
calculation of economic indicators. This is done under the Generate Programme/Run Set up
option within of Programme Analysis. For life-cycle analysis, there can be one or more
standards assigned to each road section.
The results of all sections and all alternatives are saved for economic analysis and
optimisation.
3.2.3 Results of the assignment

After either multi-year forward programme or life-cycle analysis, the options with the
highest NPV (net present value) are assigned for each section. The result of this process is a
tentative, unconstrained work program for each budget period, determined by the maintenance
or improvement standards defined by the user. This result reflects the overall technical needs
for road works and it is illustrated in Table B2.5.

Version 1.0
Table B2.5
Example of a tentative work programme

Road Section Surface Length Initial Initial IRI Road Works NPV/ Year
Code Type (km) AADT (mm/km) assigned Cost
P-BE-FE (S01) AC 22.1 2436 6.0 Overlay 50mm 3.46 2001
P-HA-MU (S02) AC 24.3 3892 4.6 Overlay 50mm 3.01 2001
S-HE-MA ST 7.1 1137 5.8 Overlay 40mm 2.95 2001
P-FE-CH ST 38.4 734 6.6 Reconstruct 1.79 2001
S-KA-BI AC 43.2 340 4.2 Overlay 50mm 1.18 2001
S-NY-ZA ST 21.6 612 6.2 Reseal 0.68 2001
S-JO-MA ST 7.0 369 6.2 Reseal 0.57 2001
T-MA-CH GR 22.3 201 10.8 Regravel 0.23 2001
S-RU-NY ** ST 26.5 234 5.6 Reseal 0.01 2001
S-BU-KE ST 21.7 1807 5.5 Reseal 4.56 2002
P-HA-BE AC 54.1 3552 3.8 Overlay 50mm 3.29 2002
P-HA-MU (S01) AC 20.7 4064 3.2 Overlay 50mm 2.58 2002
P-HA-CH (S01) AC 35.3 1927 4.1 Overlay 40mm 1.21 2002
P-BE-FE (S02) AC 21.6 2774 3.2 Overlay 40mm 1.99 2003
S-MU-NY GR 45.4 123 7.5 Regravel 1.26 2003
Codes: P = Primary; S = Secondary; AC = Asphalt Concrete; ST = Surface Treatment;

GR = Gravel
Note: ** S-RU-NY was pre-selected before the optimisation
3.3 User selection of project alternatives

One of the key aspects of road management is the ability to intervene and select projects that
should be carried out for reasons other than economic priority alone. During a high level
workshop held at the University of Birmingham in May 1995 (Kerali et al. 1995),
practitioners stressed the need for user intervention to select discrete projects to be assigned
high priorities for maintenance as a consequence of a multi-criteria analysis. It was agreed that
such projects should meet the minimum criterion of a positive economic return in order to
qualify for the user selected list.
To cater for the above, the results of the analyses provides an unconstrained preview list of
projects (see Table B2.5), from which users may select projects to be committed (that is, given
high priority for funding). The pre-defined works will then be submitted to the optimisation
routines to schedule projects that can be undertaken within the budget constraints taking into
account the pre-defined projects.

Version 1.0
4 Optimisation
If the needed budget for each budget period is below the given budget constraints, no further
economic analysis is necessarily needed. However, if the budget needs are higher than the
available budgets, the user has two choices:
1 Revise maintenance and improvement standards and run the automatic assignment again
2 Execute an economic analysis and optimisation
Programme analysis provides two methods for budget optimisation, as follows:
1 Total enumeration (see Section 4.1)
2 Incremental NPV/cost ranking (see Section 4.2)
Further methods may be added in later versions of HDM-4.
If the number of sections to be analysed is less than 100, and there are no more than five
budget periods and 16 alternatives per section, total enumeration can be used. This will be
externally done in the EBM-HS model of HDM-III. If the above constraints are exceeded,
incremental benefit/cost ranking will used.
4.1 Total enumeration

This is the method used by the EBM-HS model of HDM-III. It requires the user to specify the
following parameters:
! Name of data set
For example, CAPROG97

! Length of analysis period
For example, 20 years

! Budget periods
For example, 1, 2, 3, 4-20 years

! Objective function
Either: maximise NPV or maximise the improvement in roughness; for example, NPV
! Constraints on resources for each budget period
For example, 10, 10, 10, 200

The analysis period is given in terms of the number of years over which the overall analysis
should be performed, together with the initial calendar year. Budget periods are shorter time
periods for which the budget constraints are given. The objective function defines which
parameter is to be optimised. The default is the maximisation of NPV over the analysis period,
but the user can also choose the maximisation of the improvement in roughness.
The program is run for all the sections defined in the preview list (see Table B2.7) with
positive economic return, and for all budget and investment options. The budget requirements
from any committed projects will be deducted from the available budget and the balance is
used for optimisation.

Version 1.0
The optimisation problem is then defined as an integer programming problem of maximising

the total objective function (TOBJ) for the network (extracts from EBM documentation):
K Mk
Maximise TOBJ[ Xkm] =
∑ ∑ OBJ
k =1 m=1
km Xkm ...(4.1)
where:
k a road section
m a investment alternative on a section
K the number of sections in the network analysed
km subscript denoting alternative m for section k
Mk the number of alternatives for section k
Xkm the zero-one decision variable,
Xkm = 1 if alternative m of investment unit k is chosen,
Xkm = 0 otherwise
m = 1,...,Mk
and
k = 1,...,K
OBJkm the objective function to be maximised which may be the discounted net
present value economic benefits, or the average reduction in roughness due to
the investment alternative
The above is subject to resource constraints as follows:

K Mk
∑∑ R
K =1 m =1
kmqt X km ≤ TR qt , q = 1,...., Q; t = 1,...., T ...(4.2)
where:
Rkmqt non-discounted amount of resource of type q incurred by the sector's agency

within a budget period t
Q the total number of resource types
T the total number of budget periods (the duration of t may be one or more years
and need not be equal for different budget periods)
TRqt maximum amount of resource type q available for budget period t
The above is subject to the constraint of mutual exclusivity:

Version 1.0
mk
∑X
m =1
km ≤ 1, k = 1,...., K ...(4.3)
that is, for each section k, no more than one alternative can be implemented.
If M is the average number of alternatives for the sections, the problem then has KM (= K x
M) zero-one variables, QT (= Q x T) resource constraints and K interdependency constraints.
The parameters that define the problem size are K, M and QT. Depending on the solution
method used, different problem-size parameters determine whether the method is suitable for
the problem in terms of the computational effort needed.
The total enumeration method provides the user with an unconditionally optimal solution. It
computes the total net present values of all feasible programme selections, and chooses the
one with the highest value. The computational effort required for this may be considerable, so
the method is only feasible when the number of alternatives per investment unit is relatively
small.
Total enumeration is done externally in the EBM-HS software of HDM-III. The procedure is
as follows:
1 An input file for the EBM-HS is created in the Programme Analysis
(The format of this file is defined in the HDM-III EBM documentation)
2 The user performs the following:
(a) starts EBM-HS
(b) imports the file to EBM-HS
(c) runs EBM-HS
(d) exports the results to an output file
3 The output file is imported to the Programme Analysis for reporting
4.2 Incremental NPV/cost ranking

If the problem to be solved is too large for total enumeration, the incremental benefit/cost
method offers an alternative. It involves searching through investment options on the basis of
the incremental NPV/cost ratio of one alternative compared against another. The aim is to
select options successively with the largest incremental NPV/cost ratio, since this attempts to
maximise the NPV for any given budget constraint. The incremental NPV/cost ratio is defined
as:
(NPV j - NPVi )
E ji =
(cost j - cost i ) ...(4.4)
where:
NPVk the net present value of investment alternative k
costk the investment cost of alternative k
i the index of the cheaper alternative
j the index of the more expensive alternative

Version 1.0
The aim is to select sections successively starting with the largest NPV/cost ratio (Eji). Where
there is more than one investment option on any individual road section, it is assumed that the
user will have specified the alternative with the lowest discounted economic costs as the do
minimum alternative. It is important to note that if the user selects a do minimum alternative
which does not have the lowest discounted total economic cost, then all other options with
lower discounted total economic cost will be ignored.
The incremental method considers all options with higher discounted total economic cost, and
compares these incrementally against the ‘do minimum’ option. The procedure selects the
maximum NPV/Cost ratio for options that would be within budget limit. Table B2.6 shows an
example of the calculation of the incremental NPV/cost ratio.
Table B2.6
Example of Cost-effective Options List
Rank Section Standard j Standard i Eij ∆Financial

Cost ($m)
1 TH-PB-CP Reconstruct @ 5.5 IRI Do minimum 3.29 18.28
2 TH-PB-CF * Reseal @ 4 years Do minimum 2.89 4.67
3 TM-PB-CF Overlay, 60 mm @ 5 IRI Do minimum 2.81 6.84
4 TH-PB-CF * Overlay, 60 mm @ 5 IRI Reseal @ 4 years 2.76 7.26
5 TH-PG-CF Upgrade to paved @ 180 Do minimum 2.76 17.80

:
Note: * - A section with two alternative standards with high incremental benefits
An incremental search technique is used to select the options starting with the highest
incremental NPV/cost ratios, ensuring that at any time there is no more than one option
selected per road section. The process continues until the budget is exhausted for each budget
period. The method is often referred to as the efficiency frontier, which is a line that joins
investments with the highest NPV along the cost axis in a plot of NPV against investment
cost. In essence, the method seeks out those options that are close to the boundary of the
frontier. The algorithm is illustrated in Figure B2.4, and is defined in the following steps:
1 Determine the pre-defined investment options for pre-selected sections and deduct the
financial costs of these options from the available budget in corresponding years.
Exclude these sections from any further optimisation.
2 Determine possible investment options for the remaining sections. If the life cycle
analysis option is being used, set the user-defined base alternatives as the do minimum
for each road section. For the multi-year forward programme, the do minimum option is
that with the delayed capital works.
3 If the total financial cost of the do minimum investment alternatives on each section is
greater than the available budget for any period, then the investment options or budget
constraints must be redefined.
4 Deduct the financial cost of the do minimum investments from the available budget to
determine the remaining budget for each period. Set the do minimum as the first Base
option for each section.

Version 1.0
5 Calculate the incremental NPV/Cost ratio for all remaining section-options compared
against the Base option, and all other option pairs with higher economic cost. For
example, consider the following investment options for a particular section arranged in
the ascending order of discounted total economic costs:
options: A, B, C, D, E
The incremental NPV/Cost ratios for these are given by:
Eba Eca Eda Eea ; Ecb Edb Eeb ; Edc Eec ; Eed
6 Delete incremental NPV/cost ratios that are less than the user specified minimum
incremental value (MIV).
7 List the remaining incremental NPV/cost ratios in decreasing order (with the associated
section-option pair codes) and, within each incremental NPV/cost, in the order of
decreasing economic cost. For example, if Eeb = Edb then Eeb is ranked higher.
8 Select the next incremental NPV/cost ratio from the top of the list. If the lower cost
section-option is not the current Base Option for that section, continue selection until one
is found.
9 If the remaining budget is insufficient in any of the periods for the financial costs of
works required for the section-option selected in Step 8 above, then the selected option
should be rejected, and continue searching by repeating Step 8.
10 If the section-option can fit within the remaining budgets for all periods, deduct the net
increase in financial cost of capital works from all corresponding budget periods. Set the
Base option for this section to be that corresponding to the lower cost option for the
incremental NPV/Cost ratio chosen in Step 8. Providing that the remaining list is not
empty, return to Step 8.
The process described above continues until the budget is exhausted or there are no more
section-options remaining in the list. The resulting list of selected section-alternatives
constitutes the optimal work programme.
NPV
E
D
B
C A, B, C, D, E : Section-Alternatives
Priority for funds:

1. Eba
2. Edb
3. Eed
A
ECONOMIC COST
(relative to Base Option)
Figure B2.4 Efficiency frontier concept

Version 1.0
Table B2.8 illustrates the output of this optimisation. In this example, the Resealing on section
N1-2 was pre-defined before optimisation and is therefore first on the list.
Table B2.7
Comparison of work programme alternatives
Rank Section Standard j Standard i Eij ∆Cost ($m)

1 TH-PB-CP Reconstruct @ 5.5 IRI Do minimum 3.29 18.28
2 TH-PB-CF * Reseal @ 4 years Do minimum 2.89 4.67
3 TM-PB-CF Overlay, 60 mm @ 5 IRI Do minimum 2.81 6.84
4 TH-PB-CF * Overlay, 60 mm @ 5 IRI Reseal @ 4 years 2.76 7.26
5 TH-PG-CF Upgrade to paved @ 180 Do minimum 2.76 17.80
Note: * denotes a section with two alternative standards with high incremental benefits.
Table B2.8
Example of an optimised work program

1 N1-2 20.5 2 Resealing 2000 5.4 5.4
2 N4-7 23.5 7 Overlay 40mm 2000 10.9 16.3
3 N2-5 12.5 5 Reconstruct 2000 8.6 24.9
4 R312-1 30 4 Widen 4 lane 2000 31.4 56.3
5 R458-3 36.2 3 Overlay 60mm 2000 16.3 72.6
: : : : : : : :
1 N4-16 32.1 6 Reconstruct 2001 22.8 22.8
2 R13-23 22.4 4 Overlay 40mm 2001 9.7 32.5
3 N521-5 45.2 2 Widen 4 lane 2001 41.3 73.8
: : : : : : : :
1 N1-6 30.2 4 Resealing 2002 8.2 8.2
2 N7-9 17.8 3 Overlay 60mm 2002 9.2 17.4
3 F2140-8 56.1 1 Reconstruct 2002 34.9 52.3
: : : : : : : :
Detailed examples of Programme Analysis applications are given in Appendix B that

describes typical Case Studies.

Version 1.0
B3 Concepts of Strategy Analysis
1 Introduction
1.1 Concept
The concept of strategic planning of medium to long term road network expenditures requires
that a road organisation should consider the requirements of its entire road network asset.
Thus, strategy analysis deals with entire networks or sub-networks managed by one road
organisation. Examples of road networks include; the main (or trunk) road network, the rural
(or feeder) road network, urban (or municipal) road network, etc. Examples of sub-networks
include; all motorways (or expressways), all paved (or unpaved roads), different road classes,
etc.
In order to predict the medium to long term requirements of an entire road network or sub-
network, HDM-4 applies the concept of a road network matrix comprising categories of the
road network defined according to the key attributes that most influence pavement
performance and road user costs. Although it is possible to model individual road sections in
the strategy analysis application, most road administrations will often be responsible for
several thousand kilometres of roads, thereby making it cumbersome to individually model
each road segment. The road network matrix can be defined by users to represent the most
important factors affecting transport costs in the country. A typical road network matrix could
be categorised according to the following:
! Traffic volume or loading
! Pavement types
! Environment or climatic zones
! Functional classification (if required)
For example, a road network matrix could be modelled using; three traffic categories (high,
medium, low), two pavement types (asphalt concrete, surface treatments), and three pavement
condition levels (good, fair, poor). In this case, it is assumed that the environment throughout
the study area is similar and that the road administration is responsible for one road class (for
example, main roads). The resulting road network matrix for this would therefore comprise (3
x 2 x 3 =) 18 representative pavement sections. There is no limit to the number of
representative pavement sections that can be used in a strategy analysis. The trade-off is
usually between a simple representative road network matrix that would give rather coarse
results compared against a detailed road network matrix with several representative sections
that could potentially provide more accurate results.
Strategy analysis may be used to analyse a chosen network as a whole, to prepare medium to
long range planning estimates of expenditure needs for road development and conservation
under different budget scenarios. Estimates are produced of expenditure requirements for
medium to long term periods of usually 5-40 years. Typical applications of strategy analysis
by road administrations would include:
! Medium to long term forecasts of funding requirements for specified target road
maintenance standards (see Figure B3.5a).

Version 1.0
PART B CONCEPTS OF ANALYSIS B3 CONCEPTS OF STRATEGY ANALYSIS
! Forecasts of long term road network performance under varying levels of funding (see
Figure B3.5b).
! Optimal allocation of funds according to defined budget heads; for example routine
maintenance, periodic maintenance and development (capital) budgets (see
Figure B3.5c).
! Optimal allocations of funds to sub-networks; for example, by functional road class
(main, feeder and urban roads, etc.) or by administrative region (see Figure B3.5d).
! Policy studies such as impact of changes to the axle load limit, pavement maintenance
standards, energy balance analysis, provision of NMT facilities, sustainable road network
size, evaluation of pavement design standards, etc.
A typical sample application of the HDM-4 Strategy Analysis is given in
Appendix C.
1.2 Background
Strategy analysis is undertaken at infrequent intervals and is primarily a tool for resource
allocation by, or for, the top managers within a road agency. It is to be used for the analysis of
a chosen road network as a whole to prepare long-range planning estimates of expenditure
needs for road development and preservation under various budget scenarios. Strategy
analysis will be required to produce expenditure estimates for medium to long term periods of
between 5 to 40 years, under various budget constraints, maintenance standards, and road use
scenarios. It will predict key performance indicators of the road network such as average
roughness trends, surface distress, road user costs, safety levels and environmental effects.
condition or traffic flow (Paterson, 1995). Strategy analysis utilises a model of the road
network (or sub-network) to simulate network-wide performance under specified road use
characteristics, maintenance standards, development standards and resource constraints.

Version 1.0
Main Roads Annual

Average Roughness (IRI) 7.0 Budget
$10m
6.0
$15m
5.0
4.0 $20m
Target
3.0
2000 2001 2002 2003 2004 2005 2006
Figure B3.5a Effect of funding levels on road network performance
Budget Allocations
7.0 Feeder
Roads
$30m/yr
6.0
Secondary
Roads
5.0 $35m/yr
4.0 Primary
Roads
$20m/yr
3.0
2000 2001 2002 2003 2004 2005 2006
Figure B3.5b Effect of budget allocations on sub-network performance

Version 1.0
250
200
D e v e lo p m e n t
U S $ m /y e a r
150
Im p ro v e m e n t
P e r io d i c
100
R o u tin e
50
0
2000 2001 2002 2003
Figure B3.5c Optimal budget allocations to sub-heads
200
150
U S $ m /y e a r
P rim a ry
100 S e c o n d a ry
F eed er
50
0
2000 2001 2002 2003
Figure B3.5d Optimal budget allocations to sub-networks

Version 1.0
1.3 Purpose and approach to the analysis

The primary purpose of conducting a road network strategy analysis is to determine network-
wide resource requirements and the impacts of various development and preservation options
on the road network performance. Strategy analysis is concerned with medium to long term
planning for an entire road network or parts of the network defined by administrative, regional
or functional classification.
The objective of the analysis can be defined as either:
1 Determining funding levels - for a given set of target network performance standards
2 Determining long-term network performance trends - for expected long term funding
levels
The first objective can be solved by defining appropriate maintenance and improvement
standards, whilst the second objective is essentially an optimisation problem, with one or
several long-term budget constraints. In the second objective, one of two objective functions
may be used to select the best network development and/or preservation options that will
satisfy the constraints:
! Maximisation of benefits (NPV)
! Maximisation of the improvement in network condition (reduction in roughness)
The choice of objective function depends on the particular decision situation. Benefit
maximisation is the standard approach and maximisation of the improvement in network
condition may be appropriate where the reduction in roughness can be used as a proxy of
reduction in vehicle operating costs.
It should be noted that the network development or preservation standards to be optimised are
user-defined and are not therefore the set of all possible standards for the network being
analysed.
As with programme analysis, a key difference between strategy and project analysis is in the
detail with which data is defined. Project level analysis data is specified in terms of measured
defects, often at information quality level II (IQL-II) (Paterson and Scullion, 1990), whereas
the specification for strategy analysis is more generic at IQL-III or IQL-IV. As an example,
roughness would be specified in terms of the international roughness index for project level
analysis, whereas for strategy analysis, the riding quality could be specified as good, fair or
poor. The relationship between IQL-II and IQL-III data is user defined depending on road
class, pavement class and traffic class as illustrated in Table B3.1. The default relationships
will be defined in the HDM system configuration (Wightman et al., 1994).

Version 1.0
Table B3.1
Example of relationships between IQL-II and IQL-III data
Strategy Analysis Project Analysis

Road Pavement Traffic (IQL-III) (IQL-II)
class class class
Defect Level Defect Level
Good 3.0
Roughness Fair m/km IRI 6.0
Poor 8.0
Good 5.0
Fair SNC 3.5
Poor 2.0
Good zero
Main Paved High Structural Fair Wide 10
adequacy cracks
Poor 30
Good 10
Fair Deflection 50
Poor 100
Good zero
Surface distress Fair Narrow cracks 10
Poor 30
etc.

Version 1.0
2 Procedure for strategy analysis

The procedure for strategy analysis is summarised as follows:
1 Define the basic data set for the analysis (see Section 2.1)
2 Define the network and categorisation parameters (see Section 2.2)
3 Select the vehicle fleet for the road network (see Section 2.3)
4 Define traffic characteristics for selected sections (see Section 2.4)
5 Define maintenance and improvement standards for selected sections (see Section 2.5)
6 Specify budget constraints (see Section 2.6)
7 Run project analysis (life-cycle analysis) for each representative section with given
maintenance and improvement standards (see Section 3.2)
8 Run economic analyses and output the results (see Sections 3.3 and 2.7)
2.1 Basic data set

Strategy analysis makes use of:
! Road characteristics - held within the Road Network folder
! Vehicle fleet characteristics - stored in the Vehicle Fleet folder
! HDM Configuration - information for the categorisation of roads into a matrix
A strategy analysis is characterised by the following basic information:

! Study title, name label and a brief description of the study
! Objective function - maximisation of benefits or reduction in roughness
! Start and end years of the analysis and budget periods
! Input and output currencies, and conversion rate between the given currencies
! Discount rate
2.2 Road network definition and categorisation parameters

Strategy analysis is undertaken for a road network whose characteristics are derived using one
of two methods:
1 From a previously defined road network held in the Road Network folder
2 By creating a new road network in the Road Network folder
In both cases, the aim is to derive a set of representative sections for the chosen network, with
an indication of the proportion of the network that is represented by each section. A matrix
template is required to define the characteristics of the road network in terms of representative
sections. The matrix is defined from a user-specified combination of selected parameter levels
from the following list to give a series of representative sections, Ci, each with unique
characteristics.
! Road class

Version 1.0
! Road geometry class

! Traffic class
! Pavement surface class
! Pavement structure class
! Pavement condition index
An example of a matrix template is given in Table B3.2.
Table B3.2
Sample matrix template for strategy analysis
Section Road class Traffic class Pavement Condition Length

C1 High Bitumen Good L1
C2 High Bitumen Fair L2
C3 National High Bitumen Poor L3
C4 Medium Bitumen Good L4
C5 Medium Bitumen Fair L5
C6 Medium Bitumen Poor L6
C7 High Bitumen Good L7
C8 Secondary High Bitumen Fair L8
C9 High Bitumen Poor L9
etc. : : : :
The relationships between class names (for example, High, Medium, Low; or Good, Fair,
Poor) are defined in the HDM-4 Configuration.
2.2.1 Previously defined road network

The objective of this method is to populate the specific road network matrix illustrated in
Table B3.2 by using the road network data held within the Road Network folder. The road
sections will be distributed to the matrix template above, according to the physical
characteristics of data held.
This process produces a list of representative sections similar to Table B3.2, which may be
reviewed and manually adjusted before being used for strategy analysis.
2.2.2 Creation of a new road network

This method caters for users who do not have data available on individual road sections in the
Road Network folder. In this case, the representative sections can be created directly by
entering the lengths of each representative section into the matrix template (see Table B3.2).
2.3 Selection of vehicle fleet

The vehicle characteristics to be used in the analyses will have been previously defined using
the Vehicle Fleet folder. Users may select the set of vehicle types to be included in the
analysis from the vehicle fleet. It is expected that the Vehicle Fleet folder will have been used

Version 1.0
to define a comprehensive set of vehicle types from which a few may be selected for the
analysis.
This process is similar to the selection of vehicle types in the Project Analysis module.
2.4 Specification of traffic characteristics

For each of the representative sections to be analysed, the average annual daily traffic
(AADT) is specified in the Road Network folder, and the composition of different vehicle
types is defined for strategy analysis. The growth in traffic may be specified for discreet
periods defined in terms of the starting year and the type of growth that may be one of the
following:
1 Annual percent growth rate
2 Additional AADT per year
3 New AADT values
If option (3) is selected, either option (1) or (2) must follow it, unless it is the final year of the
analysis period.
This process is similar to the specification of traffic characteristics for Project Analysis.
2.5 Specification of maintenance and improvement standards

Several maintenance and/or improvement standards may be specified for the network being
analysed. Improvement standards may comprise, reconstruction or construction works.
For each of these improvement standards, the intervention criteria, unit costs, applicable
limits and the effects of the investment must be specified. In addition, any generated traffic,
diverted traffic or exogenous costs and benefits may also be specified for new construction
options. Optimisation will be carried out by selecting the set of standards for each
representative section that will maximise the objective function within the specified
constraints.
This process is generally similar to the specification of standards for Project Analysis.
2.6 Specification of budget constraints

The budget constraints are given for each budget period. Only the total budget is given.
without any division to budget heads.
It is recommended that these budget constraints do not include the costs of annual routine
maintenance, thereby greatly simplifying strategy analysis.
2.7 Output of results

The two main types of output from strategy analysis are:
1 Long term budget forecasts
2 Long term network condition trends
In addition users can select the types of output required from a list of outputs produced by
HDM-4. (These include all of those currently produced by HDM-III and EBM.) In addition,
graphs can be created from selected types of output. Samples of typical outputs from strategy
analysis are included in Appendix C1 (case studies).

Version 1.0
3 Strategy analysis
The strategy analysis process is divided into four phases:
1 Definition of investment options to be used (see Section 3.1)
2 Life-cycle analysis for representative sections (see Section 3.2)
3 Economic analysis and optimisation (see Section 3.3)
4 Finalisation of the strategy analysis
3.1 Selection of investment options

The first task in strategy analysis is to assign desired sets of maintenance or improvement
standards for representative sections. In the basic situation, the same set of standards is
assigned to all representative sections, but the user may assign any set of standards for any
section. For example, there might be different sets of maintenance standards for the main road
and secondary road networks, road with high or low traffic, and some improvement standards;
for example, unsealed road sections. If an improvement standard is assigned, a future
maintenance standard has also to be given.
The result of this definition process is illustrated in Table B3.3. In this example, two different
standards are applied for each representative section.
Table B3.3
Example of definition of investment options
Study title: [Basic strategy analysis]
Description: [Analysis of current standards]
Label: [STR97]
Representative section Surf. Maintenance Improvement Future
class maintenance
S1 B Basic (B) - -
B Ideal (B) - -
S2 U Ideal (U) Upgrade Basic (B)
U Basic (U) - -
S3 B Ideal (B) Widen Ideal (B)
B Ideal (B) - -
S4 B Ideal (B) - -
B Basic (B)
3.2 Life-cycle analysis

The life-cycle analysis is identical to that for project analysis. The project analysis is run for
each representative section and each maintenance or improvement standard selected in

Version 1.0
Section 3.1. As in the project analysis, the user has to define the base policy that is used in the
calculation of economic indicators. This is done in the Run Setup of the Strategy Analysis.
The results are saved for economic analysis and optimisation (see Analytical Framework and
Model Descriptions - Part G).
After life-cycle analysis, the options with the highest NPV (Net Present Value) are assigned
for each representative section. The result of this process is a tentative unconstrained strategy
for the road network, determined by the maintenance or improvement standards defined by the
user. This result reflects the overall needs for road works and it is illustrated in Table B3.4.
Table B3.4
Example of a Tentative Strategy (Unconstrained)
Description: [Basic strategy analysis]
Label: [STR97]
Representative section Strategy NPV/C
S1 OL50 at 4.0 1,1
S2 OL50 at 4.0 1,2
S3 Rs at 4 years 2,1
S4 Do nothing ---
S5 Do nothing ---
S6 OL50 at 4.5 1,8
---
Annual
B. NEEDS 3000
B. CONSTRAINT 2300
If the needed budget for each budget period is below the given budget constraints, no further
economic analysis is necessarily needed, and the unconstrained solution can be used as an
optimal strategy. However, if the budget needs are higher than the available budgets, the user
has two choices:
1 Revise maintenance and improvement standards and run the life-cycle analysis again
2 Execute optimisation
3.3 Optimisation
Within strategy analysis, three objective functions are available (see Analytical
Framework and Model Descriptions - Part G):
1 Total enumeration
2 Incremental benefit/cost ranking
3 Minimisation of Road Agency costs (for a target road network condition)
(Further methods may be added in later versions of the HDM-4 software.)

Version 1.0
If the number of representative sections to be analysed is less than 100, and there are no more
than five budget periods and 16 alternatives per representative section, total enumeration can
be adopted. This will be externally done using the EBM-HS model of HDM-III. If the above
constraints are exceeded, incremental benefit/cost ranking must be used.
Both total enumeration and incremental benefit/cost ranking are similar to those in the
Programme Analysis (Kerali and Mannisto, 1999).
An alternative objective function is to minimise road agency costs in order achieve specified
target road network condition. In this case, HDM-4 selects from amongst the specified road
work standards, the combination which will minimise road agency costs in achieving the
target road network condition.
The results of the analysis are stored in the database. The structure of this is as shown in Table
B3.5.
Table B3.5
Structure of output from total enumeration optimisation
Representative Length Year 1 Year 2 Year 3

section (km)
S1 1000 Overlay @ 4.0 - -
S2 2202 - Overlay @ 4.5 -
etc.

Version 1.0
4 References
Federal Highway Administration (1994)
Highway Economic Requirements System, (HERS), FHWA, Washington D.C.
Kerali, H.R., and Mannisto V. (1999)
Optimisation methods for Programme and Strategy Analysis in HDM-4. Journal of the
Transportation Research Board, Record 1558, National Academy Press, Washington
DC.
Paterson, W.D.O. (1995)
A Concept Outline of the Application Analysis Levels and the Strategic Analysis
Application for the New Highway Development and Management Model (HDM-4).
(Published in Reference No. 5 below)
Paterson, W.D.O. and Scullion, T. (1990)
Information Systems For Road Management: Draft Guidelines On System Design And
Data Issues. Infrastructure and Urban Development Department Report INU77. The
World Bank, Washington DC.
Robinson, R. and Kerali, H.R. (1995)
Proceedings; Strategic and Programme Analysis Workshop, May 1995, School of Civil
Engineering, The University of Birmingham.
Watanatada T., Harral C. G., Paterson W.D.O., Dhareshwar A.M., Bhandari A., and
The Highway Design and Maintenance Standards Model - Volume 1: Description of the
HDM-III Model.
The Highway Design and Maintenance Standards Series Baltimore:
Johns Hopkins University Press for the World Bank
Wightman, D.C., et al, (1994)
Design Overview and Development Strategy, August 1994, School of Civil Engineering,
The University of Birmingham.

Version 1.0
Contents
Volume 2 - Applications Guide: Part C
C1 Concepts of Research, Policy and Regulation Analyses

1 Introduction C1-1
1.1 Background C1-1
Version 1.0
PART C RPRA
Part C RPRA
C1 Concepts of Research, Policy and

Regulation Analyses
1 Introduction
This chapter describes the procedure for Research, Policy and Regulation analyses within
HDM-4, and includes details of case studies.
1.1 Background
Applications Guide C1-1

Version 1.0
PART C RPRA C1 CONCEPTS OF RESEARCH, POLICY AND REGULATION ANALYSES
Case Study 10: Variable costs of maintenance

Version 1.0
Case Study 11: Pavement design evaluation

Version 1.0
Case Study 12: Energy balance analysis

Version 1.0
Contents
Volume 2 - Applications Guide: Part D

D1 Customising HDM-4
1 Introduction D1-1
2 Overall structure D1-2
2.1 Main data groups D1-2
3 Road network data D1-4
3.1 Road class D1-4
3.2 Speed-flow type D1-4
3.3 Traffic flow pattern D1-5
3.4 Intersection type D1-7
3.5 Climate zone D1-8
3.6 Traffic volume D1-10
3.7 Geometry class D1-11
3.8 Pavement characteristics D1-14
3.9 Road condition D1-20
3.10 Pavement history D1-24
3.11 Miscellaneous D1-25
4 Currency and Budget Categories D1-27
4.1 Currency D1-27
4.2 Budget category D1-27
5 References D1-28
D2 Data Hierarchy
1 Introduction D2-1
2 Data hierarchy D2-2
2.1 Aggregate and detailed data D2-2
Version 1.0
2.2 Information quality levels D2-2

2.3 Key data D2-2
2.4 Calibration data D2-3
3 Data entry to HDM-4 D2-4
3.1 Network data D2-4
3.2 Vehicle fleet data D2-4
3.3 Traffic data D2-4
3.4 Road works standards D2-4
4 References D2-5
D3 Road Network Data

1 Introduction D3-1
2 Basic concepts D3-2
2.1 General D3-2
2.2 Network elements D3-2
2.3 Basis of sectioning D3-3
3 Data levels D3-4
4 Overall structure of data specification D3-5
4.2 Road network data D3-5
4.3 Network element details D3-5
5 Network element details D3-6
5.1 Use of new or existing data D3-6
5.2 Sections D3-6
5.3 Nodes D3-10
6 Special considerations D3-12
6.1 Dual carriageways D3-12
7 Applications issues D3-13
7.1 New road network files D3-13
7.2 Existing road network files D3-13
7.3 Transfer of data from existing road systems D3-13
7.4 Database capabilities D3-13
8 References D3-14
D4 Vehicle Fleet Data

1 Introduction D4-1
2 Basic concepts D4-2
2.1 General D4-2
Version 1.0
2.2 Vehicle fleet D4-2

3 Data levels D4-5
4 Overall structure of data specification D4-6
4.2 Fleet data D4-6
4.3 Vehicle data D4-6
5 Vehicle data D4-7
5.1 Standard and user-defined vehicles D4-7
5.2 Vehicle specific data D4-7
5.3 Costs D4-8
5.4 Calibration data D4-9
6 Special considerations D4-11
6.1 Changes in vehicle technology D4-11
7 Applications issues D4-12
7.1 New vehicle fleet files D4-12
7.2 Existing vehicle fleet files D4-12
7.3 Input Data Definitions: Vehicle Fleet Data D4-12
D5 Traffic Data
1 Introduction D5-1
2 Representation of traffic D5-2
2.1 Purposes of traffic data D5-2
2.2 Data types D5-3
3 Traffic categories D5-4
4 Traffic composition, volumes and growth rates D5-5
4.1 Basic features D5-5
4.2 Composition, volumes and growth rates D5-6
4.3 Input Data Definitions: Traffic Data D5-8
D6 Road Works Standards

1 Introduction D6-1
2 Definition and concepts of standards D6-2
2.1 Definition D6-2
2.2 Types of standards D6-2
2.3 Unit costs and energy use D6-2
3 Maintenance standards D6-3
3.1 Specification of a standard D6-3
3.2 Works on bituminous carriageways D6-3
Applications Guide iii

Version 1.0
3.3 Works on rigid concrete carriageways D6-7

3.4 Works on unsealed roads D6-8
3.5 Off-carriageway and other works D6-10
4 Improvement standards D6-12
4.1 Specifying improvement standards D6-12
4.2 Bituminous roads improvements D6-12
4.3 Unsealed roads improvements D6-19
Applications Guide iv
Version 1.0
PART D DATA MANAGEMENT
D1 Customising HDM-4
1 Introduction
This chapter describes the data sets that can be customised under HDM Configuration. The
primary role of the data configuration system, within HDM-4, is the management of all of the
default data to be used within the analysis. A set of default data are provided with the HDM-4
software, but users may modify this to reflect local circumstances.
Applications Guide D1-1

Version 1.0
PART D DATA MANAGEMENT D1 CUSTOMISING HDM-4
2 Overall structure
2.1 Main data groups

Data in HDM-4 Configuration is structured into three main groups:
! Road network (see Section 2.1.1)
! Technical models (see Section 2.1.2)
! Currency and budget categories (see Section 2.1.3)
2.1.1 Road network

The required sets of data describing a road network are as follows:
! Road class (see Section 3.1)
! Speed-flow characteristics (speed-flow type) (see Section 3.2)
! Hourly traffic flow distribution (traffic flow pattern) (see Section 3.3)
! Intersection type (see Section 3.4)
! Climate zone data (see Section 3.5)
! Traffic level (see Section 3.6)
! Road geometry (see Section 3.7)
! Pavement structure and strength (see Section 3.8)
! Road condition (see Section 3.9)
! Pavement history (see Section 3.10)
! Miscellaneous (see Section 3.11)
Detailed discussions of these data sets are presented in Chapter D3.
2.1.2 Technical models

The following technical models are used in HDM-4:
! Road Deterioration (RD)
! Road User Effects (RUE)
! Social and Environmental Effects (SEE)
! Works Effects (WE)
An overview of the modelling logic is given in Part A1. A more detailed description of the
models is provided in the Analytical Framework and Model Descriptions and the A Guide to
Calibration and Adaptation.

Version 1.0
2.1.3 Currency and budget categories

The following are defined under these categories:
! A list of currencies to be used for defining unit costs of road works and vehicle resources,
and included in HDM-4 reports
! A set of budget heads to which each of the different road works types will be assigned
(not yet implemented)

Version 1.0
3 Road network data
3.1 Road class

It is usually convenient to classify sections in a network by road class according to a
functional hierarchy. Such a classification facilitates management or analysis of the road
network in a number of areas.
A road class is defined by its name, and the number of road classes to be defined is at the
discretion of the user. The following road classes are provided as defaults:
! Primary or trunk
! Secondary or main
! Tertiary or local
3.2 Speed-flow type

The capability to model the effects of traffic volume on speeds is provided to enable the
economic consequences of road capacity improvements to be determined. The factors that
determine speed-flow relationships are described below:
! Capacity
The maximum number of vehicles that can pass a point, or traverse a road section, in one
hour (total both directions). Capacity values determine the shape of speed-flow curves
by establishing the ultimate capacity value.
! Free speed
The speed of each vehicle at zero (or very low) flow. It is unaffected by other traffic but
is affected by the physical characteristics of the road and other non-traffic factors. The
average free speed is calculated for each vehicle type.
! Speed at capacity
As traffic flows increase, average speeds for all vehicles converge towards the speeds of
the slowest vehicles in the stream, as passing becomes more and more restricted. As flow
approaches capacity, average speeds may fall even lower than slow vehicle free speeds,
due to small disturbances in the traffic stream. An estimate of average speed at capacity,
also known as jam speed at capacity, is needed to describe the speed-flow-capacity
relationship.
To model the effects of traffic congestion, the mixed traffic flows are converted into
equivalent standard vehicles. The conversion is based on the concept of Passenger Car Space
Equivalents (PCSE). This accounts only for the relative space taken up by the vehicle on the
road, and reflects the fact that the speed-flow model takes account explicitly of speed
differences of the various vehicles in the traffic stream. PCSE factors vary by road type, and
narrow roads have higher PCSE values than wide roads. Hoban et al. (1994) gives the concept
of PCSEs.
The following data are specified for each speed-flow type:
! Ultimate capacity, Qult (PCSE/lane/hr)
! Free flow capacity as a proportion of the ultimate capacity, XQ1

Version 1.0
Where:
XQ1 = Q0/Qult
Q0 = Free flow capacity
Qult = Ultimate capacity
! Nominal capacity as a proportion of the ultimate capacity, XQ2
Where:
XQ2 = Qnom/Qult
Qnom = Nominal capacity
Qult = Ultimate capacity (=QLult * NLANES)
! Jam speed at capacity, Sult (km/hr)
The number of speed-flow types to be defined is at the discretion of the user. Table D1.1
gives typical examples of capacity and speed-flow data for different road widths. Since this
data applies to individual sections, it is important to ensure that the data relates to a single and
not a dual carriageway. This ensures consistency with the definition of a road section.
Table D1.1
Examples of capacity and speed-flow model parameters for different road types
Road type Width XQ1 XQ2 QL

ult Sult σmaxr
(m) (PCSE/lane/h) (km/h) 2

(m/s )
Single lane road <4 0.0 0.70 600 10 0.75
Intermediate road 4 to 5.5 0.0 0.70 900 20 0.70
Two lane road 5.5 to 9 0.1 0.90 1400 25 0.65
Wide two lane road 9 to 12 0.2 0.90 1600 30 0.60
Four lane road >12 0.4 0.95 2000 40 0.60
Key:
Sult = Jam speed at capacity (km/h)
σmaxr = Maximum acceleration noise (m/s2)
A more detailed discussion on speed-flow relationships is given in Chapter E2 of the

3.3 Traffic flow pattern

There is a need to take account of the differing levels of traffic congestion at different hours of
the day, and on different days of the week and year. Therefore the number of hours of the
year for which different ranges of hourly flows are applicable needs to be considered. By
defining the distribution of hourly flows over the 8760 (365 days x 24 hours per day) hours of
the year, the AADT data can be converted to hourly flows. Congestion analysis can then be
undertaken for a number of hourly traffic flow levels, and the results combined to represent

Version 1.0
the full year. Because congestion delays and costs are greatest during the highest-flow hours,
particular attention should be paid to these hours. These highest-flow hours should be divided
into periods of shorter duration.
The data that describes the hourly traffic flows and the volume-capacity ratios are required for
modelling congestion effects on vehicle speeds and vehicle operating costs. The key
parameters are as follows:
! Qp
The traffic flow, in PCSE per hour, during each traffic flow period p.
! VCRp
The volume-capacity ratio for each traffic flow period p.

Hourly flow-frequency distribution data are specified for each road use category. This reflects
the fact that the predominant use of different roads requires different shapes of flow-frequency
distribution curves. The curves are defined in terms of the number of hours per year that the
traffic volume is at a certain percentage of the AADT. Each specified flow-frequency
distribution can then be assigned to a group of road sections. Distributions are specified as
follows:
! Name
! Number of flow-frequency periods
The number of flow periods into which the total number of hours in a year (8760) are
divided. Only one flow period needs to be specified for aggregate or network level
analyses. A uniform traffic composition and value of travel time across all flow-
frequency periods is assumed.
! The number of hours in each flow-frequency period, p (HRYRp)
! The percentage of AADT in each flow-frequency period, p (PCNADTp)
When specifying a flow-frequency distribution, the following conditions must be satisfied:

! The sum of the number of hours in each flow-frequency period (HRYRp) should be 8760
! The sum of PCNADTp over all the flow periods should be 10
Table D1.2 shows typical values of hourly flow-frequency distribution data.
Table D1.2
Examples of hourly traffic flow distribution data
Flow Period (p) HRYR PCNADT (%)
(hours) Seasonal Commuter Inter-city
1 87.6 4.3 3.1 2.2
2 350.4 13.4 11.5 7.7
3 613.2 16.8 16.8 11.8
4 2978.4 40.8 57.1 40.8
5 4730.4 25.9 13.0 38.9
Source: Hoban et al. (1994)

Version 1.0
The hourly traffic flow for each flow-frequency period is expressed as a proportion of AADT,
and is given by:
365 * PCNADTp
HVp = . . .(3.1)
100 * HRYR p
where:
HVp hourly traffic flow in period p, as a proportion of AADT

PCNADTp percentage of AADT in period p
HRYRp number of hours per year in period p
The traffic flow during each flow period is calculated as follows:

K
Qp =
∑ HV
k =1
p * PCSE k * AADTk . . .(3.2)
where:
Qp hourly traffic flow in period p (PCSE per hour)

AADTk annual average daily traffic of vehicle type k
PCSEk passenger car space equivalent of vehicle type k
The volume-capacity ratio during each traffic flow period is expressed as follows:
Qp
VCRp = . . .(3.3)
Q ult
3.4 Intersection type

A significant contributory factor to the number and severity of road accidents is the type of
interchanges along the road. Depending on the nature and flow of traffic, different types of
interchanges are likely to result in different numbers of fatal, injury or damage-only accidents.
An intersection type is defined by the following information:
! Name
! Accident rates
Default values of accident rates at intersections will not be provided in HDM-4 since this
would be unrealistic due to the wide variations in accident rates from country to country.
The number of intersection types to be defined is at the discretion of the user.
Note: Intersections are not implemented in the current HDM-4 software release.

Version 1.0
3.5 Climate zone

The climate in which a road is situated has a significant impact on the rate at which the road
deteriorates, and on some aspects of road user costs. Important climatic factors are related to
temperature, precipitation, and winter conditions. The impact of these factors on road user
deterioration and user costs depends on the value of the factors, and also on the length of the
season during which they apply.
The following information defines a climate zone:
! Name
! Moisture classification
That is, one of the following:

❏ arid
❏ semi-arid
❏ sub-humid
❏ humid
❏ per-humid*
! Temperature classification
That is, one of the following:

❏ tropical
❏ subtropical-hot
❏ subtropical-cool
❏ temperate-cool*
❏ temperate-freezing*
! Parameters
The parameters are given in Table D1.3 and Table D1.4.

Note:
* The combination of per-humid moisture classification and temperate-cool or
temperate-freezing temperature classification to form a climate zone is not permitted
There are 23 possible default climate zones, each of which is based on the combination of a
moisture classification and a temperature classification. Typical detailed data are given in
Table D1.3 and Table D1.4.

Version 1.0
Table D1.3
Detailed data related to moisture classification
Data item Moisture classification

Arid Semi-arid Sub-humid Humid Per-humid
Mean monthly 15 50 100 175 210
precipitation (mm)
Thornthwaite -80 -40 0 +60 +100

Moisture Index
Duration of dry 0.90 0.75 0.50 0.25 0.10

season as a
proportion of a year
Table D1.4
Detailed data related to temperature classification
Data item Temperature classification

Tropical Sub-tropical Sub-tropical Temperate Temperate
hot cool cool freezing
Mean annual 27 22 18 12 8
temperature (°C)
Temperature range 15 50 40 45 60
(°C)
Number of days per 90 60 30 15 10

year temperature
exceeds 32 °C
Freezing index * 0 0 0 55 220

o
( C-days)
Use of salt on the No No No Yes Yes

roads
Percent of vehicles 0 0 0 10 20
with studded tyres
Percent of driving 0 0 0 10 20
times on snow
covered roads
Percent of driving 20 10 15 5 10
times on water
covered roads
Air density 1.2 1.2 1.2 1.2 1.2

(kg/m3)
Note:
* The Freezing index is measured in OC-days.

Version 1.0
The Freezing index is calculated as follows:

For each day of the year, the mean air temperature (TEMP) is recorded, and the corresponding
number of Degrees Below Freezing (DBF) calculated as:
DBF = 0 - TEMP
If the value of TEMP is zero or above, then DBF defaults to zero.
For example:
if TEMP = 10 oC, DBF = 0
if TEMP = 0 oC, DBF = 0
if TEMP = -3 oC, DBF = 3
A frequency distribution can then be prepared which defines, over 365 days, the number of
days for which each value of DBF applies.
The Freezing index is then calculated by summing, for each DBF value, the product of
(number of days) and (DBF).
A typical calculation of Freezing index is shown below, giving a value of 66 (oC - days)
TEMP (oC) DBF No Days No Days * DBF
o
( C - days)
-5 5 1 5
-4 4 2 8
-3 3 4 12
-2 2 13 26
-1 1 15 15
>=0 0 330 0
Totals 365 66
The number of climate zones to be defined is at the discretion of the user. Each climate zone
will have a unique name, and will have been assigned detailed values (default or otherwise) to
its moisture classification and temperature classification.
3.6 Traffic volume

For each road section, traffic level is specified in terms of annual average daily traffic
(AADT) flow. At aggregate data level, traffic volume will be defined in bands or levels.
Detailed data values are associated with these in terms of the mean AADT.
The following defines a traffic band
! Description
! Road surface class
The road surface class to which the traffic band applies (that is, bituminous, concrete or
unsealed).

Version 1.0
! Mean AADT
The number of traffic bands to be defined for each road surface class is at the discretion of the
user. Table D1.5 shows the detailed traffic levels assumed related to different road surface
classes, with the following default levels:
! Low
! Medium
! High
Table D1.5
Definition of traffic bands by road surface class
Traffic band Traffic volume (mean AADT) by road surface class

Bituminous Unsealed Concrete
Low 750 75 3000
Medium 3000 175 7500
High 7500 800 15000
3.7 Geometry class

At the aggregate level, road geometry will be defined in terms of various parameters reflecting
horizontal and vertical curvature (see Figure D1.1). These represent geometry classes and
should apply to a group of roads.
The following detailed data (see Table D1.6) defines a geometry class:
! Description
! Average rise plus fall (m/km)

o
! Number of rises and falls per kilometre (n /km)
! Average horizontal curvature (deg per km)
! Superelevation (at bends)
Represented as e (%). If not user-defined, values of superelevation can be derived from

the average horizontal curvature, C, as follows:
❏ paved roads (e = 0.012*C)
❏ unpaved roads (e = 0.017*C)
The following data should also be defined together with a road geometry class:
! Speed limit (km/h)
! Speed limit enforcement factor (default = 1.1)
The ratio of mean speed to posted speed limit.

! Speed reduction factors
❏ Non-motorised transport (NMT) speed reduction factor

1 = no reduction, 0.6 = significant reduction

Version 1.0
❏ Motorised transport (MT) speed reduction factor

1 = no reduction, 0.4 = significant reduction
❏ Road side friction factor
1 = no side friction, 0.6 = significant friction

Version 1.0
Plan View of road section

C3
B
C1
C4
LA B
A
C2
Average horizontal curvature (deg/km) = C 1 + C 2 + C 3 + C4 (deg)

L AB (km)
L A B = length of road section AB (km)
Plan View of road section
R1
A F1
R2 F2
B
R3
Average rise plus fall (m/km) = R 1 + F 1 + R 2 + F 2 + R3

LA B (km)
Average number of rises and falls per km = 5 (m)

L A B (km)
Road cross-section at bends (exaggerated vertical scale)
Super-elevation, e (%) = 100 x

y
Figure D1.1 Illustration of geometric parameters

Version 1.0
The number of geometry classes to be defined is at the discretion of the user. Table D1.6.
presents the detailed data assumptions relating to the default geometry classes for HDM-4.
The values given can be amended as required.
Table D1.6
Definition of geometry classes
No. Geometry Rise + Fall Number of Horizontal Super- Speed limit

Class rises and curvature elevation
(m/km) falls per (deg/km) (%) (km/h)
km
1 Straight and 1 1 3 2 110
level
2 Mostly straight 10 2 15 2.5 100

and gently
undulating
3 Bendy and 3 2 50 2.5 100

generally level
4 Bendy and 15 2 75 3 80
gently
undulating
5 Bendy and 25 3 150 5 70

severely
undulating
6 Winding and 20 3 300 5 60

gently
undulating
7 Winding and 40 4 500 7 50

severely
undulating
3.8 Pavement characteristics

The parameters that are used in HDM-4 to describe pavement characteristics vary according
to road surface classes; they are as given below:
3.8.1 Bituminous pavements
Structural adequacy
The strength of bituminous pavements is defined by their structural adequacy to carry traffic
loading. At aggregate data level, structural adequacy is defined in terms of qualitative
descriptors/measures such as good, fair, poor, etc. The detailed data values relating to these
are in terms of the Adjusted Structural Number of the Pavement (SNP). The detailed data
have been assigned by traffic level.
In HDM Configuration, the user should first define the number of qualitative measures of
structural adequacy that should be used, and their names. Then for each qualitative measure,
and for each of the pre-defined traffic bands, the user defines the mean value of the dry season
SNP.

Version 1.0
The default data assumptions for structural adequacy of bituminous pavements in terms of
three qualitative descriptors (good, fair, and poor), for different traffic bands (that is, low,
medium, and high), are shown in Table D1.7.
Table D1.7
Definition of structural adequacy using mean SNP values
Structural adequacy Traffic bands

Low Medium High
Good 2.5 3.5 5.0
Fair 2.0 2.5 3.5
Poor 1.5 2.0 2.5
The HDM-4 definitions relating to Adjusted Structural Number of the Pavement (SNP) are
given in the Analytical Framework and Model Descriptions. Note that these parameters
represent further modification of the Modified Structural Number adopted by HDM-III.
Note also that Benkelman Beam and FWD deflection data can be used to derive SNP values
automatically by HDM-4.
Layer thicknesses and material types

The SNP values defined as measures for structural adequacy may also be used to represent
pavement layer thickness. This can be done through one of the following two ways:
! Method 1 - Defined by the user
The user should define several ranges (for example, 5) of SNP values by entering the
minimum and maximum values for each range. For each SNP range, and for each of the
eight HDM-4 pavement types (plus the option of all pavement types), the user should
define the following:
❏ most recent surfacing thickness (mm)
❏ previous/old surfacing thickness (mm)
❏ road base thickness - if base type is SB-Stabilised Base (mm)
In addition, the user should provide the following information to be used as defaults in all
analyses:
❏ surface material for AM (Asphaltic Mix) surface type
❏ surface material for ST (Surface Treated) surface type
❏ resilient modulus for stabilised roadbase type (GPa)
! Method 2 - HDM-4 default
Alternatively, SNP can be converted into pavement layer thicknesses as shown in Table
D1.8.

Version 1.0
Table D1.8
Pavement Layer Thickness
SNP range Surface thickness (mm) Base thickness Sub-base

(mm) thickness (mm)
New Old
<3 20 0 150 =250 (SNP - 2.2)
3-4 25 25 200 =250 (SNP - 2.9)
4-5 50 50 200 =250 (SNP - 3.6)
>5 50 100 250 =250 (SNP - 4.6)
Note:
Default strength coefficients for the asphalt and base materials are 0.35 and 0.15 respectively.
This method will be provided as default, together with the following information:
❏ surface material for AM will be asphalt concrete (AC)
❏ surface material for ST will be double bituminous surface dressing (DBSD)
❏ resilient modulus for stabilised bases, CMOD = 12 GPa
3.8.2 Concrete pavements
Structural adequacy
The definition of structural adequacy for concrete pavements is based upon the slab thickness,
traffic level, and either the modulus of rupture (MR28) or the percentage of steel, depending
on the pavement type.
In HDM-4 Configuration, first the user should define the number of qualitative measures of
structural adequacy that should be used, and their names. Then for each qualitative measure,
and for each of the pre-defined traffic bands, the user defines the detailed data as described
below:
! Jointed Plain Concrete Pavement (JP concrete pavement)
❏ slab thickness (mm)

❏ modulus of rupture (MPa)
The default data assumptions for structural adequacy of JP concrete pavement is in terms
of three qualitative measures (good, fair, and poor), for different traffic bands (low,
medium, and high), are shown in Table D1.9.

Version 1.0
Table D1.9
Definition of structural adequacy of JP concrete pavement using slab thickness
in mm
Structural adequacy Modulus of Traffic bands

rupture, (MPa)
Low Medium High
Good 5.0 190 200 210
Fair 4.5 170 180 190
Poor 4.0 160 170 180
! Jointed Reinforced Concrete Pavement (JR concrete pavement) and Continuously

Reinforced Concrete Pavement (CR concrete pavement)
❏ slab thickness (mm)

❏ percentage of steel reinforcement
The default data assumptions for structural adequacy of JR concrete pavements and CR
concrete pavements in terms of three qualitative measures (good, fair, and poor), for different
traffic bands (low, medium, and high), are shown in Table D1.10 and Table D1.11,
respectively.
Table D1.10
Definition of structural adequacy of JR concrete pavements using slab
thickness in mm
Structural adequacy Percentage of Traffic bands

steel (%)
Low Medium High
Good 0.25 190 200 210
Fair 0.15 170 180 190
Poor 0.10 160 170 180
Table D1.11
Definition of structural adequacy of CR concrete pavements using slab
thickness in mm
Structural adequacy Percentage of Traffic bands

steel (%)
Low Medium High
Good 2.0 170 180 190
Fair 1.0 155 165 175
Poor 0.5 150 155 160

Version 1.0
Default values for other variables

The user can review and edit the default values of the other variables used for modelling
concrete pavements given in Table D1.12.
Table D1.12
Default values of other variables
Variable Default Unit

Elasticity modulus, Ec 2.9E+07 kN/m2
Joint spacing for JP concrete pavements 4.0 m
Joint spacing for JR concrete pavements 20 m
Load transfer at joints, LT 45 %
Load transfer at slab shoulder edge, LTE 0 %
Poisson's coefficient (ratio) 0.15

Drying shrinkage coefficient, GAMMA 4.0E-04
Coefficient of thermal expansion, ALPHA 1.0E-05
Seal type Liquid
Dowel diameter 30 mm
Chairs were utilised to place dowels? Yes
Dowels were protected against corrosion? No
Shoulder type Monolithic
Widened lane None
Base type Granular
Subgrade type Granular
Modulus of subgrade reaction, KSTAT 54 MPa/m
Drainage coefficient, Cd 1.0
Longitudinal drains None
3.8.3 Unsealed roads
General
The modelling of the deterioration of unsealed roads requires detailed data on layer material
gradation and plasticity index. There is a wide variation in materials from place to place,
making the provision of sensible default materials in HDM-4 difficult. Users must therefore
define specific detailed data for layer materials that can be retrieved when aggregate data are
specified.
For unsealed roads, structural adequacy and surface condition are both represented by the
thickness of the surfacing material, gravel (see Section 3.9.2).

Version 1.0
Surface material
The user should define a surface material by giving its name, and defining the following
detailed data:
! Maximum particle size (mm)
! Percent passing 2.0 mm sieve
! Plasticity index (%)
The default surface materials for unsealed roads that are included in HDM-4 are shown in
Table D1.13.
Table D1.13
Definition of surface materials for unsealed roads
No. Surface Detailed data

material
Maximum Percent Percent Percent Plasticity
particle passing passing passing index in
size (mm) 2.0mm sieve 0.425 mm 0.075 mm (%)
sieve sieve
1 Lateritic 21.9 51.1 41.6 25.5 10.1
gravels
2 Quartzitic 23.8 57.5 44.6 24.2 9.1

gravels
3 Volcanic 25 49.3 38.0 23.5 17

gravels
angular
4 Coral 21 64.3 49 25 13
gravels
angular
5 Earth 4.8 90.5 84.9 70.2 15.8
Subgrade material
The user should define a subgrade material by giving its name, and defining the following
detailed data:
! Maximum particle size (mm)
! Plasticity index (%)
Table D1.14 shows the subgrade materials that have been provided as defaults in HDM-4.
These are based on Casagrande Soil Classification.

Version 1.0
Table D1.14
Definition of subgrade materials for unsealed roads
No. Subgrade Detailed Data

Material
Maximum Percent Percent Percent Plasticity
particle passing passing passing Index (%)
size (mm) 2.0 mm 0.425 mm 0.075 mm
sieve sieve sieve
1 Well-graded 13 60 40 18 15
gravel-sands
with small clay
content, GC
2 Gravel-sand 17 46 34 23 17
mixtures with
excess of fines,
GF
3 Sands with 12 88 68.5 27 8.3

excess fines,
SF
4 Clayey silts 10 88 77 54 17.5

(inorganic),
CL
5 Clays 8 83.5 77 59 18.8

(inorganic) of
medium
plasticity, CI
6 Clays 4 86.3 81.5 74 34.3

(inorganic) of
high plasticity,
CH
3.9 Road condition

Road condition data are grouped as follows:
! Ride quality (see Section 3.9.1)
! Surface distress (see Section 3.9.2)
! Surface texture (see Section 3.9.3)
3.9.1 Ride quality

Ride quality is an indication of the roughness of the road. It is an important parameter for
indicating road condition and maintenance needs, and for predicting vehicle operating costs.
At the aggregate level, ride quality is defined in terms of qualitative measures such as good,
fair, poor, bad, etc. The detailed data values related to these are in terms of roughness IRI
(m/km), and are assigned by road class.
For each road surface class, the user should define the number of qualitative measures of ride
quality that should be used, and their names. Then for each qualitative measure, and for each
of the pre-defined road classes, the user should define the mean roughness value, IRI, in
m/km.

Version 1.0
The default data assumptions for ride quality of bituminous pavements, concrete pavements,
and unsealed roads in terms of four qualitative measures (good, fair, poor and bad), for
different road classes (that is, primary, secondary, and tertiary), are shown in Table D1.15,
Table D1.16 and Table D1.17, respectively.
Table D1.15
Default values for ride quality - bituminous roads
Road class Ride quality (m/km IRI)

Good Fair Poor Bad
Primary or Trunk 2 4 6 8
Secondary or Main 3 5 7 9
Tertiary or Local 4 6 8 10
Table D1.16
Default values for ride quality - concrete roads

Good Fair Poor Bad
Table D1.17
Default values for ride quality - unsealed roads

Good Fair Poor Bad
3.9.2 Surface condition

Surface condition is modelled by a number of distress modes. At the aggregate level surface
condition can be defined by a qualitative measure (for example, new, good, fair, poor, bad,
etc.) that represents several distress modes. These distress modes differ depending on whether
the surface class is bituminous, concrete or unsealed.
For each of the paved road surface classes, the user may define the number of qualitative
measures of surface condition that may be used, and their names. Then for each qualitative
measure, and for each of the pre-defined road classes, the user may define the mean values of
the relevant distress modes. One set of detailed data may be defined for all road classes, or a
set can be specified for each individual road class.
The default modes used at the detailed data level are given in:

Version 1.0
! Table D1.18 for bituminous pavements

! Table D1.19 for JP concrete pavements
! Table D1.20 for JR concrete pavements
! Table D1.21 for CR concrete pavements
Table D1.18
Default distress values for surface condition - bituminous surface
Distress mode Surface condition

New Good Fair Poor Bad
Cracking (%) 0 0 5 15 25
Ravelling (%) 0 1 10 20 30
Number of pot-holes per km 0 0 0 5 50

2
Edge break (m per km) 0 0 10 100 300
Mean rut depth (mm) 0 2 5 15 25
Table D1.19
Default distress values for surface condition - JP concrete pavements

Transverse cracking (%) 0 5 20 30 50
Joint spalling (%) 0 0 10 20 20
Faulting (mm) 0 1 2 4 8
Table D1.20
Default distress values for surface condition - JR concrete pavements

Transverse cracking (%) 0 5 20 30 50
Joint spalling (%) 0 0 10 15 20
Faulting (mm) 0 1 4 7 10
Table D1.21
Default distress values for surface condition - CR concrete pavements

Failures (no/km) 0 2 4 16 30

Version 1.0
For unsealed roads, surface condition and structural adequacy are both related to the traffic
level and are represented by the thickness of the gravel surfacing. The user can define the
number of qualitative measures of surface condition that should be used, and their names.
Then for each qualitative measure, and for each of the pre-defined traffic bands, the user can
define the mean values of the gravel thickness.
The default values of gravel thickness for unsealed roads are given in Table D1.22.
Table D1.22
Default gravel thickness (mm) for surface condition of unsealed roads
Traffic band Surface condition

Good Fair Poor Bad
High 200 150 100 50
Medium 150 100 50 25
Low 100 50 25 0
3.9.3 Surface texture

At the aggregate data level, surface texture can be defined by a qualitative measure (for
example, good, fair, slippery, etc.) that gives an indication of the texture depth and skid
resistance of the surface. In the current version of HDM-4, these parameters are modelled only
for bituminous pavements.
For bituminous pavements, the user defines the number of qualitative measures of surface
texture that should be used, and their names. Then for each qualitative measure, and for each
of the surface types (AM or ST), the user assigns mean values of the sand patch texture depth
and skid resistance at 50 km/h (SCRIM).
The default values used at the detailed data level are given in:
! Table D1.23 for ST pavements
! Table D1.24 for AM pavements
Table D1.23
Default distress values for surface texture - ST pavements
Distress mode Surface texture

Good Fair Slippery
Texture depth sand patch (mm) 1.50 0.70 0.30
Skid resistance SCRIM at 50 km/h 0.60 0.45 0.30

Version 1.0
Table D1.24
Default distress values for surface texture - AM pavements
Distress mode Surface texture

Good Fair Slippery
Texture depth sand patch, (mm) 0.70 0.50 0.30
Skid resistance SCRIM at 50 km/h 0.50 0.40 0.30
For concrete pavements and unsealed roads, the user should define the default values of
texture depth. The default value used is 0.
3.10 Pavement history
3.10.1 Construction quality

The construction quality for bituminous pavements is described at the aggregate data level by
values such as good, fair, poor, etc. The actual data details to be specified relate to
construction defect indicators.
The user defines the number of construction quality classes to be used, and for each class the
following information:
! Description
! Construction defect indicator for bituminous surfacing, CDS
! Construction defect indicator for road base, CDB
! Relative compaction, (%)
The default values for construction quality relating aggregate to detailed data are shown in
Table D1.25.
Table D1.25
Definition of construction quality defaults
Parameter Construction quality

Good Fair- Fair-soft Poor- Poor-
brittle brittle soft
Construction defect indicator for 1.0 0.75 1.25 0.50 1.50
bituminous surfacing (CDS)
Construction defect indicator for 0 0.8 0.8 1.5 1.5

road base (CDB)
Relative compaction 97 91 91 85 85
For unsealed roads, the default construction method is mechanical. The user may change this
in HDM Configuration.

Version 1.0
3.10.2 Age
The three surfacing ages (AGE1, AGE2, AGE3 and AGE4) required for modelling the
deterioration of bituminous pavements can be derived from the input last surfacing year (at
section level) as follows:
If the surface type is ST, then the above input is taken as AGE2; and:
AGE3 = AGE2 + 5
AGE3 = AGE4
AGE1 = AGE2
If the surface type is AM, then the above input (last surfacing year) is taken as AGE3, and:
AGE4 = AGE3 + 5
AGE1 = AGE2 = AGE3
3.10.3 Previous condition

Previous condition data at the detailed level also include the amount of cracking prior to the
last surfacing. The user should define the following:
! Area of previous 'all structural cracking' (PCRA)
Defined as a percentage of total carriageway area; default value = 25% for ST surface
type, and 40% for AM surface type.
! Area of previous ‘wide structural cracking’ (PCRW)
Defined as a percentage of previous all structural cracking; default value = 65 %.

! Number of transverse thermal cracking per km (PNCT)
Default value = 100 for temperate-freezing temperature classification, sub-tropical hot

arid, and sub-tropical hot semi-arid climate zones; and zero for all the other 17 valid
climate zones (see Section 3.5 for definition of climate zones).
3.11 Miscellaneous
3.11.1 Shoulders
Shoulder width is defined by section, with the two unsealed shoulders for each road section
being the default, and with a condition rating which is the same as that of the carriageway.
The default value of the difference in elevation between the carriageway and the shoulder,
ESTEP, is 15 mm.
The user may define the following default information in the configuration file:
! Number of shoulders
! Shoulder type
! ESTEP (mm)

Version 1.0
3.11.2 Drainage
At the aggregate level, drainage condition can be defined by a qualitative measure (for
example, excellent, good, fair, poor, very poor, etc.) that represents the Drainage Factor
(DF). The values of DF are used in the modelling of bituminous pavement performance.
The user should define the number of drainage qualitative measures to be used, and their
descriptions. For each qualitative measure and for each drain type the user should specify the
value of DF.
The default data assumptions for drainage condition for bituminous pavements, in terms of
five qualitative measures (excellent, good, fair, poor and very poor), for different drain types
are shown in Table D1.26.
Table D1.26
Default DF values for drainage condition
Drain type Drainage condition

Excellent Good Fair Poor Very
poor
Fully lined and linked 1.0 1.5 2.0 2.5 3.0
Surface lined 1.0 1.5 2.0 2.5 3.0
V-shaped - hard 1.0 2.0 2.7 3.5 4.0
V-shaped - soft 1.5 2.5 3.2 4.3 5.0

Shallow - hard 2.0 3.0 3.8 4.5 5.0
Shallow - soft 2.0 3.0 3.8 4.5 5.0
No invert, overgrown 3.0 3.5 4.0 4.6 5.0
No drainage effects 1.0 1.0 1.0 1.0 1.0
The user can change the values of Annual Deterioration of Drainage Factor (ADDF) by
temperature type for each drain type. The default values of ADDF are given in Chapter C2 of
the Analytical Framework and Model Descriptions.

Version 1.0
4 Currency and Budget Categories
4.1 Currency
The currency to be used in HDM-4 is specified in HDM Configuration using the following
information:
! Name
! Symbol
! Symbol position
! 'Thousand' separator
4.2 Budget category

Each of the works types considered in HDM-4 is assigned to one of the default budget heads.
The default budget heads for the different road works considered in HDM-4 are given in

Version 1.0
5 References
Hoban C., Reilly W., and Archondo-Callao R., (1994)
Economic Analysis of Road Projects with Congested Traffic
Methods for Economic Evaluation of Highways Investments and Maintenance
Transport Division, Transportation, Water & Urban Development Department
The World Bank, Washington DC, USA

Version 1.0
D2 Data Hierarchy
1 Introduction
This chapter considers the data requirements for the different HDM-4 analyses. The concepts
of key data, aggregate data and detailed data are discussed. The data hierarchy complies with
the concept of information quality levels whereby data is specified at different levels of detail
(tiers) for different purposes (Paterson and Scullion, 1990).

Version 1.0
PART D DATA MANAGEMENT D2 DATA HIERARCHY
2 Data hierarchy
2.1 Aggregate and detailed data

Data used by HDM-4 can be considered at two basic levels:
! Aggregate (class) data
! Detailed data
At aggregate data level, each parameter is specified in terms of user-defined aggregate

descriptions. For example, traffic flow on a road section may be specified in three bands
(low, medium, and high). Each of these bands represents a mean value of annual average
daily traffic (AADT), as shown in Table D2.1. In this example, the AADT values represent
the default-detailed data.
Table D2.1
Definition of traffic bands by road surface class
Traffic volume (mean AADT) by road surface class

Traffic band
Bituminous Concrete Unsealed
Low 750 3000 75
Medium 3000 7500 175
High 7500 15000 800
Within HDM-4, the user may define the number of aggregate level bands and the associated
detailed data. Default data sets (specifying aggregate descriptors and associated detailed
values) are given in Chapter D1.
2.2 Information quality levels

The two data tiers (aggregate and detailed) enable different levels of data to be used
depending on which of project, programme or strategy analysis is being undertaken. Only
minimum data detail and volumes need to be specified when undertaking strategy and
programme analysis at network level. For project analysis, where investigations are focused
on a discrete part of the network, more detailed data can be used. However, the approach also
enables project analysis to be carried out with minimum data requirements in situations where
this is necessary or appropriate. The approach to grouping data into tiers can be related to the
concept of information quality levels (Paterson and Scullion, 1990).
When detailed data is specified, these override aggregate values and are used in any analysis.
When data is defined only at mandatory and aggregate levels, assumptions are made about
values at the detailed level.
2.3 Key data

Key data refers to the minimum set of data needed to obtain a meaningful output from HDM-
4. Key data may be aggregate or detailed, and includes cost data.

Version 1.0
Key network data comprises those attributes needed to define the physical and functional
characteristics of the network; for example, section length and width, speed-flow type and
traffic-flow pattern.
Key vehicle data comprises the vehicle physical characteristics, utilisation details and costs
(including vehicle, labour and time costs).
Works costs are key data, and include costs associated with Maintenance Works and
Improvement Works. A default set of costs for both works categories is held under Works
Standards/Default Works Costs. Maintenance Works are assumed to take place within a
specified calendar year. With Improvement Works, the percentage of costs incurred over each
year of the works period (between 1 and 5 years) is specified.
2.4 Calibration data

Calibration data refers to certain parameters, which are used to modify the behaviour of the
HDM-4 models. For example, the road deterioration model incorporates calibration factors
for the relationships defining the rate of deterioration. The deterioration model for bituminous
roads considers the effect of trafficking on parameters such as rut depth, roughness and
structural number, while the model for unsealed roads considers the effect on gravel loss.
Calibration data can be considered as detailed data, whether the values are default or
specified.
Details of the HDM-4 models and associated calibration factors are given in Analytical
Framework and Model Descriptions. Details of the calibration procedures, including
associated data collection, are included in A Guide to Calibration and Adaptation.

Version 1.0
3 Data entry to HDM-4

Data entry to HDM-4 can be effected manually via the keyboard, or by importing data in a
database format. Data sets that may be imported to the HDM-4 database are:
! Network data (see Section 3.1)
! Vehicle fleet data (see Section 3.2)
! Traffic data (see Section 3.3)
! Road Works Standards (see Section 3.4)
3.1 Network data

When a section is added to a specific Road Network, this can be based either on
aggregate data, or on an existing section. If based on aggregate data, the user selects, for
various section attributes, the aggregate descriptor from a predefined list. The aggregate data
sets are set up under HDM Configuration, as described in Chapter D1. Further details of
specific network data requirements are given in Chapter D3.
3.2 Vehicle fleet data

When a new vehicle type is added to a specific Vehicle Fleet, the user selects the base vehicle
type from a list of base types pre-defined under. Further details of specific data requirements
are included in Chapter D4.
3.3 Traffic data

For each road section in the network, the user must specify details of traffic composition,
volumes and growth rates, axle loading (defined under Vehicle Fleet), speed-flow type and
traffic flow pattern. Details of specific data requirements are given in Chapter D5.
3.4 Road works standards

Road Works Standards are used to specify the works to be applied to a given road section
either over a given period (scheduled) or in response to a deteriorating condition (responsive).
The standards include unit costs for a range of works items, which may be specified
individually for each standard or, by default, assigned from a cost database.
Details of data requirements for the specification of Maintenance Standards and
Improvement Standards are given in Chapter D6.

Version 1.0
4 References
Paterson W.D.O., and Scullion T., (1990)
Information Systems for Road Management: Draft Guidelines on System Design and
Data Issues, Technical Paper, Report INU 77. Infrastructure and Urban Development
Department. The World Bank, Washington DC, USA

Version 1.0
D3 Road Network Data
1 Introduction
This chapter discusses the concept of road network referencing used in HDM-4. A flexible
framework is provided to cater for a wide range of road network representations. The Road
Network folder holds the road network data that is used by all of the HDM-4 applications.

Version 1.0
PART D DATA MANAGEMENT D3 ROAD NETWORK DATA
2 Basic concepts
2.1 General
Road network referencing is a system used for the locational positioning of data pertaining
to roads. It defines the technical rules and conventions employed to identify data on a
particular part of a road network. There are two types of locational referencing data:
! Data that describe the position and geometry of a road segment
! Attribute data that describe characteristics of the road or inventory associated with it
Attribute data are assigned to a location or segment of a road according to the segmentation
scheme used. Locational referencing and road segmentation together forms the basis for the
retrieval of road network information.
The establishment of a road network referencing system is principally a modelling process
that maps the real world on to its equivalent computer abstraction. It is essential to make a
distinction between objects and events that exist in the real world (that is, physical) and those
that are stored in the computer (that is, logical). The physical view is both locational and
quantifiable, whilst logical data are merely an abstraction (that is, a representation) of the real
world. Information is also needed about how objects and events are related to each other (that
is, logical relationship). Each entity or data item associated with a road network may have
both logical and physical attributes, together with the logical relationships between them.
2.2 Network elements

A road network may be defined as a logical system of road links, sections and nodes. Each of
these network elements is usually identified uniquely by a numbering or labelling system.
The following definitions of network elements are used in HDM-4.
Sections
A section is a length of road that is homogeneous in terms of its physical attributes. A road
section may be split further into sub-sections, normally for purposes of data collection.
Sub-sections
A sub-section is defined as a part of a road section that has strong consistency in particular
attributes at a higher degree of precision and reliability.
Network
A network is a collection of sections that are grouped together for administrative or analysis
purposes.
Sub-networks
A sub-network is defined as a smaller grouping of sections within a larger road network.
These may be selected according to common logical attributes such as pavement type, road
function, administrative boundary, etc.

Version 1.0
Nodes
Within a road network, nodes mark points at which traffic characteristics change or road
features change significantly. Such points would include changes in traffic volume or
composition, significant changes in geometry, road layout, population centres or
administrative boundaries.
Links
A link is defined as a length of road joining two nodes. A link should have a consistent
geometric layout along which the volume of traffic flow is uniform (that is, traffic does not
vary substantially).
Note: Nodes Links are not implemented in the current release of the HDM-4 software.
Road networks can be represented or modelled in two ways:
! Physical networks
Physical networks comprise the geometric description of the roads within a region.
Physical networks usually contain locational and inventory information, and are
represented graphically using two-dimensional maps.
! Logical or administrative networks
Logical or administrative networks comprise a group of sections within a physical

network that perform a defined function. An example of such a network in a country
would be the trunk or principal roads that serve the purpose of carrying through traffic
between cities or towns.
2.3 Basis of sectioning

Fixed-length sectioning is used as the basis for road network representation in HDM-4. This
may have implications when interchanging data with external systems, where all necessary
attributes must be specified for each section. Responsibility for ensuring the validity of the
section data imported into HDM-4 will be with the external systems. The referencing process
involves both logical and physical activities (Robinson et al., 1998). With fixed length
sectioning, the sections are defined on a logical basis, so that each has homogeneous
characteristics. Within HDM-4, the attributes that are considered uniform along a section
include:
! Road class
! Speed flow type
! Climate zone
! Traffic flow pattern
! Carriageway width
! Pavement construction (that is, surface class and pavement type)
! Traffic volume
Each logical section is identified by a unique name and ID, either of which may be used to
identify the section on the various HDM-4 output reports. The logical definition of the
network within the HDM-4 database must be matched by physical referencing on the ground.
This may take the form of road markings (for example, studs), concrete posts (for example,
kilometre posts) or bar codes, typically mounted on posts. It is important that the physical
referencing is maintained to facilitate the future location of sections.

Version 1.0
3 Data levels

Road Network information can either be specified in detail or it can be aggregated to reduce
the input requirements. Thus, two tiers exist for these data types:
! Aggregate network data
These data items include attributes such as road class, traffic level, geometry, pavement
structure and history. These attributes also include condition data that changes over time.
! Detailed network data
These data items are those used in the models or relationships built into HDM-4. These
include, for example, the properties of layer materials used in the pavement, layer
thicknesses, number of effective lanes, superelevation. Such data will normally be
determined following detailed investigations of the road section characteristics.
Within HDM-4, when detailed data are specified, these override aggregate values and are used
in any analysis. When data are defined only at aggregate level, assumptions are made about
values at the detailed level. Chapter D1 describes the relationship between aggregate and
detailed data, and gives particulars of the detailed data values that are assumed when default
values of aggregate data have been specified. The relationship between aggregate and
detailed data are user-definable in HDM Configuration.

Version 1.0
4 Overall structure of data specification

Data in the Road Network folder is structured into two main groups:
! Road network data
! Network element data
Facilities are provided for adding (new), modifying (edit attribute) and deleting attributes in
these data groups.
The data groups are summarised in sections 4.2 and 4.3. The specification of road network
data (for example, speed-flow type, traffic flow pattern, and climate zone) is described in
Chapter D1. Network element data (aggregate and detailed) are described in Section 5.
4.2 Road network data

Road network data allows the specification and labelling of a specific network to which all
other attributes can be ascribed. There is no limit to the number of networks that can be
defined in this way. In addition to this basic network information, data values can be
specified for:
! Network information (see Section 4.3)
! Speed-flow types (see Chapter D1)
! Traffic flow patterns (see Chapter D1)
! Climate zones (see Chapter D1)
! Intersection types (not implemented in current release of HDM-4)

These data values can be associated with individual network elements.
4.3 Network element details

Network element details enable data to be specified for sections and nodes. Aggregate and
detailed data can be specified for sections.
Note: Nodes are not implemented in the current release of HDM-4.

Version 1.0
5 Network element details
5.1 Use of new or existing data

The network elements of sections or nodes can be entered in one of two ways, either each
element can be:
1 Entered from scratch, or
2 Created by modifying the details of an existing element.
Where elements share many similar characteristics, the second method of specification can
save considerable time with data entry.
5.2 Sections
5.2.1 Description and attributes

A road may comprise several homogeneous pavement sections. A section is the basic entity
for all calculations of pavement deterioration, construction and maintenance costs, and
economic analysis, and are specified by two sets of data:
1 Description and identification
A unique reference for each road section within a network.

2 Attributes
A list of parameters associated with a particular section, defined by their type and value
as may be appropriate in the data field format; examples are geometry, pavement
structure, climate, topography.
Road sections may be pre-defined as in the case where data are imported from an existing
road management system.
5.2.2 Key data

A section is identified uniquely by a textual description and a label or ID. The ID should
consist of a short alphanumeric code that provides a key to the section. It is often helpful if
the ID includes the road number, where used, and reference to the road class. Surface class
and pavement type is also specified according to the codes given in Table D3.1.
Table D3.1
Surface class and pavement type
Surface class Pavement type

Bituminous AMAB, AMAP, AMGB, AMSB, STAB, STAB, STGP, STSB
Concrete CR concrete pavements, JP concrete pavements no dowels, JP concrete pavements

with dowels, JR concrete pavements
Unsealed Earth, gravel

Version 1.0
Note:
Definitions of pavement types are given in the Analytical Framework and Model
Descriptions.
The speed-flowtype, traffic-flow pattern, and climate zone of the section are assigned as
described in Chapter D1.
The geometric characteristics of the section are assigned in terms of the following:
! Length of section (km)
! Width of the carriageway (metres)
That part of the road traversed by traffic

! Width of each shoulder (metres)
! Direction of traffic flow

One-way on one way streets or dual carriageways
Two-way on other roads
! Number of traffic lanes on the carriageway
5.2.3 Aggregate data

Data are specified under the headings of:
! General
❏ road class
❏ traffic
❏ geometry
! Pavement
❏ structural adequacy
! Condition
❏ ride quality
❏ surface condition
❏ surface texture
! History
❏ construction quality
These aggregate data types are described below. Detailed definitions are assigned in HDM
Configuration and are described in Chapter D1.
General
Road Class
It is usually convenient to classify sections in a network by road class according to a
functional hierarchy. Such a classification facilitates management or analysis of the road
network in a number of areas:

Version 1.0
! Managing roads
Different road administrations or agencies may have the responsibility for managing
roads of different classes. Classification can be used to assist in identifying who is
responsible for what.
! Funding
Funding may also be allocated separately to roads of different classes, and the source of
funding may also differ. For example; main or trunk roads may be funded with monies
raised and allocated by central government, and district or village councils may fund
local roads.
! Road functions
Classification enables different road functions to be recognised. Main roads may have
the purpose of carrying long distance traffic whereas local roads may have the purpose of
providing only access to industrial, commercial, agricultural or domestic premises.
These functions should be defined in the policy framework of the road administration,
and different objectives, standards and intervention levels for works should be set up for
each road class.
! Types and numbers of traffic
The different functions of road classes are likely to be reflected in different types and
numbers of traffic being carried by each. The capacity that is provided by each road class
is also likely to differ. However, there may be road sections in the network that have
relatively low traffic but are of key strategic importance because of the places that they
link. The maintenance standards, intervention levels and priorities assigned to such
sections can be related to the road class, rather than the traffic level.
! Maintenance and improvement standards
The division of roads into classes enables users of HDM-4 to set up maintenance and
improvement standards, and priorities that are appropriate to each functional class.
Default values of road class are:
! Primary or trunk
! Secondary or main
! Tertiary or local
Traffic
For each road section, traffic level is specified in terms of annual average daily traffic
(AADT) flow. At aggregate data level, traffic volume can be defined, with the following
default levels:
! Low
! Medium
! High
Geometry
Horizontal and vertical curvature is defined in terms of various combinations of the following
default values:
! Horizontal curvature
Mostly straight, bendy, winding.

Version 1.0
! Vertical curvature
Generally level, gently or severely undulating.
Pavement
Pavement details are defined in terms of the following attributes and default values:
❏ good
❏ fair
❏ poor
Condition
Ride quality
Ride quality is an indication of the roughness of the road. It is an important parameter for
indicating road condition and maintenance needs, and for predicting vehicle operating costs.
It is defined in terms of the following default levels:
! Good
! Fair
! Poor
! Bad
Surface condition
This is an aggregate measure of several deterioration parameters. The following default
values are used:
! New
! Good
! Fair
! Poor
! Bad
Surface texture
This value gives an indication of the surface texture and skid resistance of the surface. The
default values are:
! Good
! Fair
! Slippery
History
Construction quality is specified under this heading with the following defaults:
! Good
! Fair

Version 1.0
! Poor
5.2.4 Detailed data

Detailed data are specified under the following headings:
! Geometry
Horizontal and vertical alignment, speed factors, etc.

! Pavement
Pavement layers, construction materials, strength characteristics, subgrade, etc.

! Condition
Pavement defects.
! History
Previous works records and past conditions.

! Miscellaneous
Other data not included in any of the above groups.

! Calibration
Calibration factors for models and relationships used in HDM-4.

The actual data details to be specified differ depending on whether the surface class is
bituminous, concrete or unsealed. The relationship between aggregate and detailed data
default values is defined in Chapter D1. Definitions of the detailed data parameters are also
given in Chapter D1.
5.3 Nodes
The specification of nodes is an optional attribute of the road network representation in HDM-
4. Nodes are considered as entities (that is, separate data items) in all analyses. Typically
nodes are intersections or junctions where two or more road sections meet. In addition to this
are intermediate nodes that include the start and end of dual carriageways, administrative
boundaries, and the like. The start or end of a road (that is, a dead-end) is considered to be a
separate type of node. Thus, a road network may comprise three types of node:
! Intermediate
Positioned at some other point on the section where a locational position needs to be
referenced for some particular purpose.
! Intersection
Positioned at the junction of this and other sections.

! Road end
Positioned at a dead end of a section, where there are no junctions with other sections.
A node should be specified by its name and a unique reference number through a textual
description and a label or ID. The ID should consist of a short alphanumeric code that
provides a key to the node. It is often helpful if the ID is set to the grid co-ordinates of the
node, since this may assist subsequent spatial location.

Version 1.0
The intersection type can also be defined, with a value being assigned from those specified
earlier. For analytical purposes, it is assumed that only intersections require attributes to be
specified. The results calculated for nodes may be combined with those of the adjoining road
section, or taken individually. Each node may be assigned more than once to different
sections within the same road network.
Note: Nodes are not implemented in the current version of HDM-4.

Version 1.0
6 Special considerations
6.1 Dual carriageways

Intermediate nodes should mark the start and end of dual carriageways. Dual carriageways
are therefore represented in one of the following ways:
! Two parallel sections
Running in opposite directions, with one-directional traffic, if the characteristics of the

carriageways differ substantially, see Link B-C in Figure D3.1. In this case, appropriate
speed-flow types are assigned to each section.
! A single link with two-directional traffic
If both carriageways exhibit uniform characteristics, the dual carriageway may be

represented as a single section with an appropriate speed-flow type. See Link A-B in
Figure D3.1.
Figure D3.1 illustrates the above convention.
Intermediate node
direction
left-hand side carriageway
local dualling
A A’ B C
right-hand side carriageway
single-carriageway double-carriageway
A B C
Figure D3.1 Alternative dual carriageway representation schemes

Version 1.0
7 Applications issues
7.1 New road network files

Each set of network elements (that is, a combination of sections and nodes) within an area, is
contained in a user-specified road network file. Facilities are provided for creating new files,
opening existing files, saving, exporting, importing and printing files. This is intended to
encourage users to maintain a database of their physical (or logical) network for future
analyses. The resources of the host machine (that is, available memory and hard disk space)
only limit the number of sections stored in the Road Network folder.
A new road network is built by adding network elements and then entering the details for each
of these elements. This is done by selecting the type of network element to be created from a
list of the valid options, and then defining it in terms of a unique label and the attributes
associated with the element. The order in which the types of network elements are entered is
not important. Building the road network could start with the entry of either sections or
nodes. The entry of nodes is optional, since they may not be required by many HDM-4
applications.
7.2 Existing road network files

Once a road network file has been created, its contents can be modified subsequently by
adding new network elements, by amending information on existing elements, or by deleting
pre-defined elements as may be required. It is also possible to select and view the information
on each of the network elements by name or by label.
7.3 Transfer of data from existing road systems

Many countries and organisations already have established road management systems that use
different referencing systems, ranging from simple gazetteers to fully computerised formats.
HDM-4 is capable of accepting and utilising data from a wide range of referencing systems
and a mechanism for transferring existing network data has been provided. This has the
benefit of efficiency, but before undertaking the data transfer exercise, users should satisfy
themselves that their system is robust enough for HDM-4 purposes. All imported data must
be specified according to homogeneous sections. All aggregate attributes, at least must be
included, otherwise the data will be invalid. Any optional data not specified will initially
assume user-defined default values held within the data file system.
7.4 Database capabilities

The HDM-4 system is not intended to have the functionality of an extensive database
management system. Whilst the road network data required by HDM-4 is stored internally in
the HDM-4 database, this database is limited to servicing the data manipulation requests of
the applications within HDM-4. There are no facilities provided for applications such as
indexing, table restructuring or other functions normally found in database management
systems.

Version 1.0
8 References
Robinson R., Danielson U., and Snaith M., (1998)
Road Maintenance Management Concepts and Systems. London and Basingstoke:
Macmillan Press.

Version 1.0
D4 Vehicle Fleet Data
1 Introduction
Vehicle Fleet data describes the characteristics of vehicles that use a road network. Vehicle
fleet characteristics are required in HDM-4 for the estimation of traffic flow and capacity,
vehicle operating costs (VOC), travel time costs, accident costs, and the evaluation of
environmental effects from vehicle emissions and traffic noise. Within HDM-4, vehicle fleet
data are held in the Vehicle Fleet folder.
This chapter describes the methodology used for the storage and retrieval of the vehicle
parameters to be used in HDM-4. A vehicle fleet representation framework has been provided
to suit the needs of many countries and will also satisfy the analytical requirements of HDM-
4.

Version 1.0
PART D DATA MANAGEMENT D4 VEHICLE FLEET DATA
2 Basic concepts
2.1 General
Road administrations normally only deal with the vehicle fleet used in their region or country.
However, international consultants often undertake road investment studies in many different
countries, and would therefore deal with several vehicle fleets. It is possible to deal with both
of these requirements for vehicle fleet representation within HDM-4. Several sets of vehicle
types may be defined corresponding to the requirements of the different users and types of
analyses.
A vehicle classification system is primarily a modelling process that attempts to represent the
diverse characteristics of the vehicles in a traffic mix. It uses a set of representative vehicles
for which a number of physical and performance characteristics are defined. The
representative vehicles may or may not exist physically. However, the parameters or
attributes specified for the representative vehicles should reflect the overall physical and
performance characteristics of the range of vehicles that use a road network.
2.2 Vehicle fleet
2.2.1 Grouping of vehicles

A vehicle fleet comprises a mix of several vehicle types that use a road network. The
characteristics of the vehicle fleet are normally represented by grouping the vehicles into
classes defined according to common attributes such as size, utilisation or performance.
The standard terminology used in HDM-4 refers to vehicles according to the following:
! Categories (see Section 2.2.2)
! Classes (see Section 2.2.3)
! Types (see Section 2.2.4)
Figure D4.1 illustrates the hierarchical structure of the relationships between vehicle
categories, vehicle classes and vehicle types.

Version 1.0
Motorised
Motorised Categories
Motor
Motorcycles
cycles Passenger
Passengercars
cars Utilities
Utilities Trucks
Trucks Buses
Buses Classes
Motor
Motorcycles
cycles Small
Smallcar
car Medium
Medium Light
Lighttruck
truck Medium
Medium truck
truck
(1) Types
(1) (2)
(2) car
car (8)
(8) (9)
(9)
(3)
(3)
Large
Largecar
car
(4)
(4)
Heavy
Heavytruck
truck Articulated
Articulated truck
truck
(10)
(10) (11)
(11)
Light Light
Lightgoods Mini-bus Light
Light goods Mini-bus Lightbus
bus
delivery
delivery vehicle
vehicle (12) (13)
(12) (13)
vehicle
vehicle (6)
(6)
(5)
(5)
Four
Fourwheel
wheel Medium
Mediumbusbus Heavy
Heavybus
bus
drive
drive (14)
(14) (15)
(15)
(7)
(7)
Coach
Coach
(16)
(16)
Source: NDLI (1995)
Figure D4.1 Definition of motorised vehicle categories, classes and types

Version 1.0
2.2.2 Vehicle categories

The two vehicle categories included in HDM-4 are motorised and non-motorised vehicles
(NMT).
2.2.3 Vehicle classes

These are essentially sets of vehicles with similar physical, utilisation and performance
characteristics. The vehicle classes included are:
! Motorised vehicles
❏ Motorcycles
❏ Cars
❏ Utilities
❏ Trucks
❏ Buses
! Non-motorised vehicles
❏ Animal Cart
❏ Bicycle
❏ Pedestrian
❏ Rickshaw
2.2.4 Vehicle types

A vehicle type represents a set of vehicles within a class for which a separate set of
relationships exists for the prediction of vehicle speeds and for the estimation of vehicle
operating costs. For example, trucks could be sub-divided into light, medium, heavy, and
articulated types.
There are 16 motorised vehicle types provided as the standard default representative vehicles
within HDM-4. It is also possible to define a number of vehicle types based upon these
standard vehicles.
There are 4 non-motorised vehicle base types. Further details are given in Part E of the
2.2.5 Trailers
Trailers have not been included as a separate class/type of vehicle. Vehicles with semi-trailers
or drawbar trailers should, therefore, be included as a separate vehicle type, with appropriate
values for the gross vehicle mass, number of wheels, and other parameters.

Version 1.0
3 Data levels

Vehicle Fleet data can be specified in one of two ways, either in detail, or it can be
aggregated to reduce input requirements.
Aggregate data includes attributes such as the physical characteristics of vehicles, in terms of:
! The number of passengers and crew carried
! Number of wheels
! Utilisation in terms of working hours
! Vehicle loading
Detailed data items are those used in the models or relationships built into HDM-4. These
include, for example, details of the physical characteristics, utilisation, performance, and other
calibration parameters. Such data will normally be determined following detailed
investigations of the particular vehicle fleet under study (see Section 5.4).
Default values are available for several items of data in both tiers.

Version 1.0
4 Overall structure of data specification

Data in the Vehicle Fleet folder is structured into two main groups:
1 Fleet data (see Section 4.2)
2 Vehicle data (see Section 4.3)
Facilities are provided for adding (new), modifying (edit attribute) and deleting (remove)
attributes.
4.2 Fleet data

Fleet data allows the specification and labelling of a specific vehicle fleet to which all vehicle
categories, classes and types can be ascribed. There is no limit to the number of fleets that can
be defined in this way.
The vehicle classes to be used in the analysis are also assigned in this section of the data
definition.
4.3 Vehicle data

The second main data group enables data to be specified for vehicles in terms of general
parameters and costs, and detailed data (see Section 5.4).

Version 1.0
5 Vehicle data
5.1 Standard and user-defined vehicles

An HDM default fleet is available and defines standard (or base) vehicle classes and types.
Although this fleet can be used directly, it will normally be more appropriate to specify a
number of vehicle types based on the standard vehicles. New vehicle classes and types
specified in this way will be added to the list of vehicles. User-defined vehicles will take on
initially the characteristics of the standard vehicles, but these can then be modified as
appropriate.
The number of vehicle types required to represent a vehicle fleet will depend on the situation
in the particular country or region, and on the type of analysis to be performed. The key
factor is to ensure that the representation of the vehicle fleet is robust and efficient for the
required analysis. For example; a research application involving the impacts of axle loads in a
vehicle weights and dimensions study would require a more comprehensive representation of
the vehicle fleet than that needed for a strategic network level analysis.
The minimum criterion for the specification of vehicle classes is the proportion of heavy
vehicles in the traffic mix that could be obtained from sample spot counts. On average, a
specification of at least three vehicle classes is required, with five classes being recommended
to correspond to the standard vehicle classes (that is, cars, utilities, light trucks, heavy trucks
and buses). Non-motorised vehicles will form an important traffic component in some
countries where such vehicles exert a significant influence on traffic flow characteristics.
Motorcycles can also form a significant proportion of traffic in some regions. These should
be specified as separate classes.
For analyses where more detailed investigations are being carried out, the number of vehicle
types should typically correspond to the number of standard vehicle types, with no upper limit
set for the maximum number.
5.2 Vehicle specific data

Vehicle types are created by selecting the appropriate base vehicle type from a list of valid
options.
5.2.1 Definition
A vehicle type can be defined or modified by specifying the following set of data:
! Name
A description or unique reference for each vehicle type within the vehicle fleet.
! Base vehicle type
The pre-defined representative vehicle to which the vehicle type is related.

! Vehicle class
See Section 2.2.3.

! Category
Either motorised or non-motorised.

Version 1.0
! Description
A textual description that can be included to give more details relating to user-defined
vehicle types.
! Life method
This may be either Constant Life or Optimal Life, used for RUE analysis.
5.2.2 Basic Characteristics

The basic characteristics of each vehicle are specified under the headings:
! Physical
❏ Passenger Car Space Equivalency factor (PCSE)

❏ number of wheels
❏ number of axles
! Tyres
❏ tyre type
❏ base number of recaps per tyre carcass
❏ retread cost as a percentage of the new tyre cost
! Utilisation
❏ annual average distance travelled by the vehicle (km)

❏ working hours - hours per year expended on the essential tasks of complete round
trips
❏ average service life of the vehicle (years)
❏ percentage of vehicle use on private trips
❏ average number of passengers per vehicle
❏ percentage of passenger trips which are work related
! Loading
❏ number of equivalent standard axles per vehicle (ESAL)

❏ average operating weight of the vehicle (tonnes)
5.3 Costs
Costs are specified in terms of economic and financial unit rates. Costs data are specified
under the headings:
! Vehicle resources (economic and financial costs)
❏ Cost of an equivalent new vehicle

❏ Cost of an individual tyre
❏ Cost of fuel used by the vehicle (per litre)
❏ Cost of lubricating oil (per litre)

Version 1.0
❏ average maintenance labour costs (per hour)

❏ total crew wages (per hour)
❏ average annual overhead costs (lump sum per year)
❏ annual bank interest rate (per cent)
! Time value (economic costs)
❏ Passenger working time

A value of time for passengers travelling on business (per hour).
❏ Passenger non-working time
A value of time for passengers travelling for other than business activities (per
hour).
❏ Cargo
A value of time for cargo carried by the vehicle (per hour).
! Bank interest rate (%)
! Overheads associated with operating the vehicle (lump sum per year)
5.4 Calibration data

Road User Effects calibration parameters, in general, provide calibration factors for technical
relationships, and are specified under the following headings:
! Forces
Forces acting on the vehicle, under the headings:

❏ Aerodynamic
❏ Power
❏ Rolling resistance
❏ Unsealed road impassability
! Speed
Speed model calibration parameters.

! Fuel
Fuel and oil consumption and engine speed model parameters.

! Tyres
Parameters related to tyre wear.

! Maintenance
Vehicle maintenance requirements defined in terms of parameters for the parts

consumption model and the labour hour’s model.
! Emissions
Calibration factors relating to vehicle emissions.

Version 1.0
! Optimal life
Parameters for optimal life model.

! Acceleration effects
Acceleration noise from different sources and roughness at which maximum acceleration
noise occurs.
! Energy
Energy use parameters, also tyre weight and unladen vehicle weight.
Further details of calibration data are given in Part E of the Analytical Framework and Model
Descriptions.

Version 1.0
6 Special considerations
6.1 Changes in vehicle technology

Improvements in technology produce new models and designs of vehicles. It is expected that,
over time, there will be many improvements such as:
1 Vehicles will be more economical with fuel efficient engines
2 Vehicles will be environmentally friendly with reduced emissions, possibly through
exhaust gas re-circulation and the use of more effective catalytic converters
3 There will be increased numbers of electrically driven vehicles
4 There will be improved control systems leading to improved safety and efficiency of
highway use, etc.
The vehicle fleet representation system incorporated in HDM-4 permits the vehicle fleet
parameters to be updated when appropriate, and has the ability to modify or replace the
standard set of relationships, as the need arises.

Version 1.0
7 Applications issues
7.1 New vehicle fleet files

Information on the vehicle fleet using a particular road network under study is stored in a
user-specified vehicle fleet file. Facilities are provided for creating new files, opening existing
files, saving, exporting, importing and printing vehicle fleet files. This is intended to
encourage users to maintain a database of their vehicle fleets for future analyses. The
resources of the host machine (that is, available memory and hard disk space) only limit the
number of vehicle fleets stored in the Vehicle Fleet folder.
A new vehicle fleet is built by adding vehicle types and then entering the details of each of
these vehicle types. Selecting the vehicle type to be created from a list of the valid options
does this, and then defining this in terms of a unique name and the attributes associated with
the type. The order in which the vehicle types are entered is not important.
7.2 Existing vehicle fleet files

Once a vehicle fleet file has been created, its contents can be modified subsequently by adding
new vehicle types, by amending information on existing vehicle types, or by deleting pre-
defined vehicle types, as may be required. It is also possible to select and view the
information on each of the vehicle types by name.
7.3 Input Data Definitions: Vehicle Fleet Data


Version 1.0
D5 Traffic Data
1 Introduction
This chapter describes the traffic characteristics used in HDM-4. Section 2 summarises the
traffic data requirements for different HDM-4 applications, and lists the traffic data types.
The three basic traffic categories (normal, diverted and generated) are described in Section 3,
and their specification detailed in Section 4. Further details are given in Part B of the

Version 1.0
PART D DATA MANAGEMENT D5 TRAFFIC DATA
2 Representation of traffic
2.1 Purposes of traffic data

The results of economic analyses are quite sensitive to traffic data, and most benefits that
justify road improvements arise from savings in road user costs. To perform economic
analyses in HDM-4, traffic characteristics of roads therefore need to be described and
represented at an appropriate level of detail.
Traffic characteristics need to be represented for the following analytical purposes:
! Project analysis
This requires a detailed representation of traffic characteristics on the road being

analysed. For each road section, the representation should include data items that
describe the details of changing traffic composition and volumes, axle loading, capacity
and speed-flow relationships, hourly traffic flows, traffic induced by road improvements,
and of demand shifts. These sets of traffic data may need to be specified separately for
each lane of the carriageway.
The required traffic data for this type of analysis is similar to that for project analysis, with
the exception that the data are at a more aggregate level. For example; traffic volumes may
be specified by vehicle classes, and several representative road sections could use the data.
! Strategy analysis
This requires the specification of an aggregate set of traffic data, that is representative of
a group of road sections being analysed. Traffic levels should be expressed in terms of
daily flows, and may be described as low, medium, or high. The composition of traffic
may be expressed as a percentage of daily flow for each vehicle class/type.
For the purposes of representing traffic characteristics both for project and network level
analyses, road links within a network must be categorised according to the following:
! Speed flow type
The speed flow type defines the parameters for capacity, speed-flow relationship shape,
width effects and passenger car space equivalents for different road widths (for example,
single lane road, two lane road, wide two lane road, four lane road).
! Traffic-flow pattern
The Traffic flow pattern describes the patterns of traffic flow along each road category;
commuter routes, for example, tend to have weekday peaks but low weekend traffic,
whereas recreational routes have a more peaked distribution.
! Non-motorised transport factor
This measures the effect of non-motorised transport on motorised traffic speeds.

! Motorised transport factor
This is an assessment of the effect of motorised transport on motorised traffic speeds.

Version 1.0
! Roadside friction
This is a measure of the effect of roadside activity on traffic speeds; and includes the
effects of land use, roadside stalls, bus stops, parking, etc.
2.2 Data types

The traffic data are incorporated in several modules (or applications) rather than being
specified in one place. Traffic data types can be considered under the following headings.
! Traffic categories (see Section 3)
Traffic data types are normal, diverted and generated.

! Traffic composition, volumes and growth rates (see Section 4)
These are specified within the individual applications programs (that is, project analysis,
programme analysis and strategy analysis) as required by the purpose of the particular
study.
! Axle loading
This information is specified for each vehicle type in the Vehicle Fleet folder (see
Overview of HDM-4). Provision is made to derive most of the required parameters (for
example, equivalent standard axle load factors, etc.) from the user-specified data. Details
of axle loading data requirements are given in Part B of the Analytical Framework and
Model Descriptions.
! Road capacity and speed-flow relationships
These parameters are defined by the speed flow type within HDM-4 Configuration.
Further details are given in Chapter D1.
This defines the hourly flow frequency distribution over a year, and is used to estimate
hourly flow ranges from annual average traffic data. Traffic flow patterns are road
dependent and are specified for each road use category within HDM-4 Configuration.
Further details are given in Chapter D1.

Version 1.0
3 Traffic categories
Traffic categories consist of the following traffic data types:
! Normal (see Section 4.2.1)
Normal traffic is defined as traffic that would pass along the project road if no investment
took place. It is specified for each road section within the applications programs.
! Diverted (see Section 4.2.2)
Diverted traffic is defined as traffic that changes from another route (or transport mode)
to the project road. It is specified together with the road investment option that causes it,
and is specified within the relevant applications programs.
! Generated (see Section 4.2.3)
Generated traffic is defined as additional traffic that occurs in response to the road
investment. It is specified together with the road investment option that induces it, and is
specified within the relevant applications programs.

Version 1.0
4 Traffic composition, volumes and growth rates
4.1 Basic features

The traffic composition is defined as the proportions of the different vehicle types that use the
road. Information on traffic composition is required for several analytical purposes, including
predictions relating to:
! Pavement deterioration
! Travel time
! Vehicle emissions (particulates, hydrocarbons, noxious gases)
! Total life cycle energy consumption

! Economic analysis and comparisons
The existing traffic volumes on the road being analysed are specified in terms of vehicle type
or class, depending on the kind of analysis to be performed. The value entered for each
vehicle type is expressed as the annual average daily traffic (AADT):
Total annual traffic in both directions

AADT =
365
This constitutes the baseline flow for the analysis period. It is assumed that seasonal variations
in traffic flows have already been accounted for when estimating the AADT from traffic
counts carried out over shorter periods.
For project analysis, traffic composition data are specified for each section. For network
level analysis, several representative sets of traffic composition data can be specified, and
each is assigned to a group of road sections with similar traffic characteristics.
In most situations, traffic growth has a major effect on the level of benefits obtained. Changes
in traffic composition over time can be modelled by specifying different growth rates for each
vehicle type/class. Uncertainties always exist in estimating initial traffic, but there is even
greater uncertainty in forecasting future growth rates. Thus, it is recommended that sensitivity
analyses are always undertaken to evaluate the effects of different forecast growth rates.
Traffic volumes are derived from the baseline AADT and the composition data. The
computation of annual road user effects requires the AADT for each road section to be broken
down by vehicle type. For normal traffic, the annual average daily traffic for each vehicle type
(AADTkn) is obtained by multiplying the normal traffic AADTn by the composition data
defined for the vehicle type k in the base year. For generated traffic, AADTkg is either
specified directly or calculated, based on AADTkn and the traffic growth type used. Volumes
for normal and generated traffic are required as separate inputs into the Road User Effects
(RUE) model to enable the economic benefits of generated traffic to be determined.
The specification of traffic composition data for each section is carried out for each traffic
category as described in Sections 4.2.1, 4.2.2 and 4.2.3.

Version 1.0
4.2 Composition, volumes and growth rates
4.2.1 Normal traffic

Normal traffic is specified using the following data.
! Annual average daily traffic (AADT)
This is specified in vehicles per day at the traffic start year. It is permissible and
sometimes convenient to enter AADT for some year before the start of the analysis
period. However, it must be ensured that the traffic start year is always less than or
equal to the start year of the analysis period.
The proportion of each representative vehicle that uses the road, as a percentage of the
AADT. For example; medium cars 10%; articulated trucks 5%; etc. These data items are
used to calculate the actual volumes, by vehicle type, for the applicable year.
! Future traffic
This data item is specified as a traffic growth period defined in terms of its start year and
by one of the following growth types:
❏ annual percentage increase.
❏ annual incremental increase in AADT.
❏ actual AADT, which must be followed by another traffic growth period, unless the
year in which it applies is the last year of the analysis period.
Then, depending on the selected traffic growth type, the growth rate is specified for each
representative vehicle. For example, light trucks increasing at 4% per year from 1995 to 2000,
if annual percentage increase was selected.
Several traffic growth periods may be defined, each with a minimum length of one year. It is
important to ensure that the defined traffic growth periods cover consistently each and every
year of the analysis period. Therefore, if only one growth period has been defined, this will be
assumed to apply to all successive analysis years.
4.2.2 Diverted traffic

Forecasting diverted traffic can be difficult, particularly where traffic diverts from other
transport modes, and for analysis over a complete network. Ideally, estimates of diverted
traffic should be made by utilising the results from external traffic demand models. Thus, in a
situation where a road works causes diverted traffic, a new set of traffic data representing the
net effect of traffic diversion over all the sections affected should be defined for each project
option. For each project option, a new set of traffic data are specified as follows:
! Name
! Start year
The calendar year in which traffic diversion starts, typically this should coincide with the
year following the completion of the road works.
! New AADT
The annual average daily traffic in the start year for each road section.

Version 1.0
! New composition
The traffic composition in the start year for each road section.
! Future traffic
Specified by selecting one of the following growth types:

❏ annual percentage increase in AADT.
❏ annual incremental increase in AADT.
Economic comparisons of investment options involving diverted traffic can only be done
meaningfully, at the project analysis level, if the following conditions are met:
1 All the road sections from and to which traffic diverts are analysed together with the
section(s) being considered under the investment analysis; this implies that a study area
be defined to comprise all the sections that are affected significantly by traffic diversion
as a result of carrying out the road works;
2 In any given analysis year, the total traffic volume entering the study area equals the total
traffic volume exiting the area; this implies a fixed trip matrix.
The analysis of a new section in an entirely new location will always involve diverted traffic.
4.2.3 Generated traffic

The main causal factors of generated traffic are reductions in travel cost or time, or the
increased development brought about by a road investment. The amount of generated traffic is
difficult to forecast accurately, so the time period over which traffic is generated should be
limited.
The recommended approach to forecasting generated traffic is to use demand relationships.
The price elasticity of demand for transport measures the responsiveness of traffic to a
change in transport costs following a road investment.
For each road investment standard, generated traffic is specified as follows:
! Name
! Relative start year
A number of years, after the start of the road works, that generated traffic appears. Thus,
relative start year j means the jth year following that year in which the works are started
that cause the generated traffic.
! Future traffic
This is specified by selecting one of the following growth types:

❏ additional annual percentage increase in AADT
❏ annual incremental increase in AADT of generated traffic
❏ actual AADT of generated traffic
❏ percentage of the normal traffic in the current year
4.2.4 Traffic at intersections

Total traffic entering an intersection node is needed for road safety analysis. The total traffic
entering the node is entered as AADT. Traffic composition data are not required for this.
Note: Intersections are not modelled in the current release of HDM-4.

Version 1.0
4.2.5 Changes in vehicle fleet characteristics

If changes in the characteristics of the vehicle fleet are expected to occur in the future, the
following procedure should be followed:
Any vehicles that come into use at a future date should be specified in the
Vehicle Fleet folder, together with those vehicles that are already in use. The
AADT of future vehicles should be set to zero until the year in which they
come into effect. In that year new AADT for the vehicle is defined using
incremental increase in AADT, or the actual AADT, followed by another growth
period. Negative growth rates can be defined for vehicles that are already in
use in such a manner that they are gradually phased out (that is, AADT=0) and
replaced by future vehicles.
4.3 Input Data Definitions: Traffic Data


Version 1.0
D6 Road Works Standards
1 Introduction
This document describes the specification of Road Works Standards used within HDM-4.
The standards include the road maintenance and improvement standards that are followed by
road administrations in their network management and development activities. The standards,
which are pre-defined by users and stored in the Road Works Standards folder, can
subsequently be selected for use with a particular HDM-4 analysis.

Version 1.0
PART D DATA MANAGEMENT D6 ROAD WORKS STANDARDS
2 Definition and concepts of standards
2.1 Definition
Standards refer to the targets or levels of conditions and response that a road
administration aims to achieve. Road agencies set up different standards that
can be applied in practical situations in order to meet specific objectives
which are related to functional characteristics of the road network system.
In HDM-4, a standard is defined by a set of operations or works activities with
definite intervention criteria to determine when to carry them out. In general
terms, intervention levels define the minimum level of service that is allowed.
2.2 Types of standards

A standard is user-defined according to the road surface class to which it is applied, the
characteristics of traffic on the section, and the general operational practice in the study area
based upon engineering, economic and environmental considerations. Standards are grouped
into two types for input purposes:
! Maintenance standards
! Improvement standards
A number of maintenance and improvement standards can be defined within the Road Works
Standards folder, and these can be assigned to physical road sections for project and
programme analysis, or to representative road sections for strategy analysis.
For a given feature type (that is, carriageway, shoulders, non-motorised traffic (NMT) lanes,
miscellaneous and special), only one maintenance standard and/or improvement type will be
effective in any analysis year.
2.3 Unit costs and energy use

The default values of unit costs, in economic and financial terms, for all the operations
considered in HDM-4 can be specified in a separate data input facility (Road Works
Standards/Default Works Costs and Energy Consumption). These default values can be
amended to reflect differences in the costs of different specifications of works of the same
operation type. The facility also allows users to specify default values of unit rates of energy
used in carrying out each of the operations. These unit rates are required if energy balance
analysis is to be performed.

Version 1.0
3 Maintenance standards
3.1 Specification of a standard

A maintenance standard can be set up or amended by defining the following:
! Name
! Short code
! Road surface class to which the standard is applicable
! A list of constituent works items for the standard
A detailed discussion of road maintenance works, the intervention criteria that determine
when they should be applied, their design, effects on the road characteristics, and the costs of
implementing each works activity is given in Part D of the Analytical Framework and Model
Descriptions.
The following sections describe the specification of constituent works for a Maintenance
Standard.
3.2 Works on bituminous carriageways

The specification of the works for the standard differs depending on the type of works:
! Routine pavement works (see Section 3.2.1)
For example, crack sealing, patching, and edge repair.

! Preventive treatment works (see Section 3.2.2)
For example, fog sealing and rejuvenation.

! Resurfacing works (see Section 3.2.3)
For example, surface dressing, slurry sealing, and cape seal.

! Rehabilitation works (see Section 3.2.3)
For example, inlay, overlay, and mill and replace.

! Reconstruction (see Section 3.2.4)
For example, partial reconstruction, and full reconstruction.
3.2.1 Routine pavement works

A routine pavement works item is defined by the following:
! General
❏ name of the Works Item

❏ short code for the Works Item (unique within the Maintenance Standard)
❏ type of operation selected from a pre-defined set (for example, crack sealing,
patching, edge repair)
❏ intervention type - whether scheduled or responsive

Version 1.0
! Intervention
The range of factors required to trigger the operation:

❏ time interval in years for scheduled interventions
❏ intervention criteria, and minimum and/or maximum limits, for responsive
interventions
❏ limits of interventions:
- last year in which intervention should be carried out
- maximum road roughness, IRI, (m/km) at which a maintenance intervention
should be used
- maximum quantity or unit to be applied (in m2/km/year or m3/km/year)
- minimum intervention in terms of traffic flow (AADT)
- maximum intervention in terms of traffic flow (AADT)
! Costs
❏ unit rates (economic & financial) for the works item

! Effects on pavement condition (responsive)
These effects are for responsive interventions only. The effects are specified in terms of
the percentage of the following types of distress repaired, based on the operation
type/works activity. Examples of distress type repairs specified are given below for the
following works activities:
❏ crack sealing
- transverse thermal cracking
- wide structural cracking
❏ patching
- severely damaged area, wide structural cracking, potholing and ravelling
❏ edge repair
- edge break
3.2.2 Preventive treatment

A preventive treatment works is defined by the following:
! General

❏ short code for the Works Item
❏ type of operation selected from a pre-defined set (for example, fog sealing,
rejuvenation)

Version 1.0
! Intervention
The range of factors required to trigger the operation

❏ time interval, in years, for scheduled interventions
interventions
should be used
- minimum intervention interval in terms of time (years) and traffic flow(AADT)
- maximum intervention interval in terms of time (years) and traffic flow
(AADT)
! Costs
Unit rates (economic and financial) for the works item.
3.2.3 Resurfacing and rehabilitation works

A resurfacing or rehabilitation works is defined by the following:
! General

❏ type of operation selected from a pre-defined set (for example, resurfacing works
include slurry seal and surface dressing, while rehabilitation works include overlay,
inlay, and mill and replace)
! Design
❏ surface material type (for new surface layer)

❏ thickness of new surfacing (mm)
❏ AASHTO strength coefficient during the dry season
❏ depth of material to be milled (applicable to mill and replace only)
❏ area of carriageway to inlay (applicable to inlays only)
❏ construction details in terms of defect indicator, CDS
! Intervention
The range of factors required to trigger the operation

interventions
❏ limits of interventions

Version 1.0

should be used
- minimum intervention interval in terms of time (years) and traffic flow (AADT)
(AADT)
! Costs
Unit rates (economic and financial) for the operation specified under General, and
associated preparatory works.
! Effects
Resulting from the works in terms of the following:

❏ Roughness, IRI, (m/km)
❏ mean rut depth (mm)
❏ surface texture (mm)
❏ skid resistance (SCRIM at 50 kph)
3.2.4 Pavement reconstruction

A pavement reconstruction works item is defined by the following:
! General

❏ type of operation - specified as pavement reconstruction
! Design
❏ pavement type
❏ surface material type
❏ thickness of new surfacing (mm)
❏ AASHTO strength coefficient during the dry season
❏ relative compaction
❏ base thickness (for stabilised base only)
❏ resilient modulus (for stabilised base only)
❏ construction details in terms of defect indicators for bituminous surfacing (CDS) and
base (CDB)
! Intervention


Version 1.0
❏ intervention criterion, and minimum and/or maximum limits, for responsive

interventions
- minimum intervention interval, in terms of traffic flow (AADT)
- maximum intervention interval, in terms of traffic flow (AADT)
! Costs
Unit rates (economic and financial) for operation pavement reconstruction.

! Effects
Resulting from the works in terms of the following:

❏ skid resistance (SCRIM at 50 kph)
3.3 Works on rigid concrete carriageways

These works include bonded overlay, diamond grinding, full depth repair, joint sealing, load
transfer restoration, partial depth repair, retrofit edge drains, slab replacement, and tied
concrete shoulders.
Each works item is defined by the following:
! General

❏ type of operation selected from a pre-defined set (for example, bonded overlay and
diamond grinding)
! Intervention

interventions
should be used
- minimum intervention interval in terms of time (years) and traffic flow (AADT)

Version 1.0

(AADT)
! Costs
Unit rates (economic and financial) of the works item.

! Effects
Effects on road characteristics as a result of the works.
3.4 Works on unsealed roads

These include grading, spot-regravelling, and regravelling/resurfacing (see Sections 3.4.1,
3.4.2 and 3.4.3 respectively).
3.4.1 Grading
Grading can be scheduled to take place at a fixed time interval or after a fixed amount of
traffic; alternatively, grading can be triggered by a roughness threshold. The works to which
the standard applies are defined by the following:
! General

❏ type of operation (that is, grading) selected from a pre-defined set
! Design
Compaction method (mechanical or non-mechanical).

! Intervention

❏ time interval, (days) OR traffic interval (AADT) for scheduled interventions
interventions
- maximum road roughness, IRI, (m/km) beyond which a maintenance
intervention should not be used
- minimum intervention interval in terms of time (days) and traffic flow (AADT)
- maximum intervention interval in terms of time (days) and traffic flow (AADT)
! Costs
Costs per kilometre (economic and financial) of the grading works.
3.4.2 Spot regravelling

Spot regravelling can be carried out on a scheduled basis, or in response to a reduction in
gravel thickness. The works to which the standard applies are defined by the following:

Version 1.0
! General

❏ type of operation (that is, spot regravelling) selected from a pre-defined set
! Design
❏ Either:
- volume of gravel to add (m3/km/year)
or
- percentage of annual gravel loss to replace (%)
! Intervention

interventions
intervention should not be used
- maximum annual quantity to be applied (in m3/km/year)
- minimum intervention interval in terms of traffic flow (AADT)
- maximum intervention interval in terms of traffic flow (AADT)
! Costs
Economic and financial costs of spot regravelling (per m3)
3.4.3 Regravelling/resurfacing
Regravelling can be scheduled to take place at a fixed interval based on the age of the gravel,
or can be responsive to roughness and gravel thickness. The works to which the standard
applies are defined by the following:
! General

❏ type of operation (that is, regravelling/resurfacing) selected from a pre-defined set
! Design
❏ material properties (details of particle size distribution and Plasticity Index)

❏ final gravel thickness (mm) or increase in gravel thickness (mm)

Version 1.0
❏ compaction method (mechanical or non-mechanical)

! Intervention

interventions
intervention should not be used minimum intervention interval in terms of time
(years) and traffic flow (AADT)
(AADT)
! Costs
Economic and financial costs of regravelling (per m3 and associated preparatory spot
regravelling (per m3).
! Effects
❏ derived, or
❏ user defined - initial roughness, IRI, (m/km)
3.5 Off-carriageway and other works

These include the following works, by feature:
"
Feature Operation Cost units

Miscellaneous Miscellaneous per km per year
Non-motorised traffic lane NMT lane repair per km
Shoulders Shoulder repair per km
Special Emergency per year
Winter per year
For these works, only the following need to be specified:

! General

❏ type of operation selected from a pre-defined set (see above table)
❏ intervention type - scheduled only
! Intervention limits
Specified for:

Version 1.0
❏ last applicable year

❏ minimum time interval in terms of time (years) and traffic flow (AADT) as
applicable
❏ maximum time interval in terms of time (years) and traffic flow (AADT) as
applicable
! Costs
Economic and financial costs (units defined in table above).

! Effects
The effects of drainage repairs on bituminous pavements are modelled internally. There
are no effects as a result of the other works.

Version 1.0
4 Improvement standards
4.1 Specifying improvement standards

The following types of improvement standard can be applied:
! Bituminous Carriageways
❏ partial widening
❏ lane addition
❏ realignment of bituminous roads
❏ upgrading of bituminous roads
❏ pavement reconstruction
! Unsealed carriageways
❏ lane addition
❏ realignment of unsealed roads
❏ upgrading of unsealed roads
An improvement standard can be set up or amended by defining data under the following
headings:
! General
! Design
! Intervention
! Costs
! Pavement
! Geometry
! Effects
These are each described in turn for each of the works types in the sub-sections associated
with Sections 4.2 and 4.3.
4.2 Bituminous roads improvements
4.2.1 Partial widening
General
The works to which the standard applies are defined by the following:
! Name
Name of the improvement works.

! Short Code

Version 1.0
! Road surface class
The road surface class to which the standard is applicable.

! Improvement type
! Duration of the works (years)
! Intervention type
whether scheduled or responsive.
Design
The design of the works is specified in terms of the following:
! Speed-flow type
! Road class
! Pavement type
For example; asphalt mix on an asphalt base, surface treatment on a granular base, etc.
! Increase in carriageway width (m)
In the range 0-3 metres.

More detail of the design data input can be specified as follows:
! Construction quality indicators
❏ bituminous surfacing CDS (0.5 ≤ CDS ≤ 1.5)

❏ road base CDB (0 ≤ CDB ≤ 1.5)
Details of CDS and CDB are given in Chapter C2 of the Analytical Framework and
Model Descriptions.
CDS indicates the general level of binder content and stiffness relative to the optimal
material design for the specified bituminous mixture. It is used to illustrate whether a
bituminous surfacing is prone to cracking and ravelling (low value of CDS) or prone to
rutting through plastic deformation (high value of CDS). Values are in the range 0.5 to
1.5 with typical values as follows:
Dry (brittle) surfacing, with binder content nominally 10% below optimal: CDS = 0.5
Normal surfacing, with binder content at optimal value: CDS = 1.0
Rich (soft) surfacing, with binder content nominally 10% above optimal: CDS = 1.5
CDB ranges between 0 (no construction defects) and 1.5 (several defects). Defects
considered are: poor gradation of base material, poor aggregate shape and poor
compaction.
! Resurfacing
Check on whether resurfacing the existing carriageway is part of the standard and, in
those cases where it is, then the following is specified:
❏ thickness of the new surfacing (mm)

Version 1.0
❏ AASHTO strength coefficient of the layer

❏ unit cost of the layer per m2
! Unit costs of preparatory works
❏ patching per m2
❏ edge repairs per m2
❏ crack sealing per m2
Intervention
Partial widening can be on a scheduled based upon a calendar year. Alternatively,
intervention can be based on the following criteria:
! Speed
❏ average annual speed (km/h)

❏ minimum speed (km/h)
! Traffic
❏ average annual daily traffic (two-way)

❏ peak period volume-capacity ratio
❏ daily average volume-capacity ratio
The time of the intervention can be limited by specifying the last applicable year in which it
can take place for responsive intervention criteria, but not for scheduled intervention.
Costs
Costs are specified in terms of the following:
! Unit rates (per km)
! Percentage of costs incurred over each year of the works period
Pavement
Details of the pavement to be provided under an improvement standard are specified with the
following data:
! Details of the surfacing layers
Specified in terms of:

❏ surface material (for example; asphalt concrete, cold mix, hot rolled asphalt, etc.)
❏ thickness of layer (mm)
! Adjusted structural number (excluding subgrade contribution) in the dry season
! Average relative compaction of pavement (%)
! Details of road base
(if a stabilised road base)

❏ thickness of roadbase (mm)

Version 1.0
❏ resilient modulus (GPa)
Geometry
The following data under geometry can be specified for partial widening:
2
! Acceleration noise due to driver and road alignment (m/s )
! Speed limit enforcement factor

❏ due to non-motorised transport, with a value between 0.6 (significant reduction) and 1.0
(no reduction)
❏ due to motorised transport, with a value between 0.4 (significant reduction) and 1.0 (no
reduction)
❏ due to roadside friction, with a value between 0.6 (significant reduction) and 1.0 (no
reduction)
Effects
The effects of an improvement intervention define the condition of the road after the
intervention has been made:
! Change in traffic flow pattern
! Pavement condition after works have been carried out
These effects are specified in terms of the percentage of the following types of distress
repaired:
❏ skid resistance, (SCRIM at 50 km/h)
In addition, the pavement deterioration factors for the initiation and progression of defects and
other calibration factors can be specified as optional data.
A detailed discussion of road improvement works, the intervention criteria that determine
when they should be applied, their design, effects on the road characteristics, and the costs of
implementing each works activity is given in Part D of the Analytical Framework and Model
Descriptions.
4.2.2 Lane addition
General
The specification is the same as for partial widening (see Section 4.2.1) in all areas except
for design. (Note however, that with partial widening the increase in width is limited to the
range 0-3 metres.)

Version 1.0
Design
The design of the lane addition works is specified in terms of the following:
! Speed-flow type
! Road class
! Pavement type
! Additional number of lanes (in the range 1-9)
4.2.3 Realignment of bituminous roads
General
The specification of this is similar to that for partial widening (see Section 4.2.1), with the
exception of design and geometry.
Design
The design of realignment works is specified in terms of the following:
! Speed-flow type
! Road class
! Pavement type
! Length adjustment factor
! Proportion of new construction
The ratio of new construction length to the section length after realignment works, with a
value between zero and one.
Geometry
The following data under geometry can be specified for realignment:
! Rise plus fall (m/km)
! Number of rises and falls (number per km)
! Average horizontal curvature (deg/km)
! Superelevation (%)
2
! Acceleration noise (m/s ) - due to driver and road alignment

(no reduction)
reduction)
reduction)

Version 1.0
4.2.4 Upgrading of bituminous roads
General
Upgraded pavements can be constructed in either bituminous or concrete materials, and the
specification of the upgrading standard reflects this. The specification for this is similar to
partial widening (see Section 4.2.1), with the exception of design, pavement, geometry and
effects.
Design
The design of the improvement works is specified in terms of the following:
! Speed-flow type
! Road class
! Pavement type
! Increase in road width - resulting from the improvement (metres)
! Additional number of lanes (between 1 and 9)

For bituminous pavements, design data input can be specified in more detail in the same
manner as for partial widening (see Section 4.2.1). For concrete pavements, the following
data may be specified:
! Shoulders
❏ shoulder type
❏ load transfer efficiency (%)
! Drainage
❏ edge drains
❏ drainage coefficient
Pavement
Depending on the road surface class resulting from the upgrading, the pavement details are
specified as follows:
! Bituminous
As for partial widening of bituminous roads.

! Concrete
Layers thickness, slab length, material strength properties.
Geometry
The following data under geometry can be specified for upgrading:

Version 1.0
2

(no reduction)
reduction)
reduction)
Effects
intervention has been made.
Bituminous roads
The following effects are specified:
These effects are specified after the works have been carried out, in terms of the
following types of distress:
❏ surface texture depth (mm)
other calibration factors can be specified under Section/Details/Surface Distress.
Concrete pavements
The following effects must be specified:
These effects are specified after the works have been carried out, in terms of the
following type of distress:
In addition, detailed calibration factors can be specified under Section/Details/Surface
Distress.

Version 1.0
4.2.5 Pavement Reconstruction

This improvement works is similar to the maintenance works item Pavement Reconstruction
described in Section 3.2.4. The main difference is that, under improvement works, widening
(up to 3 metres) is allowed. The specification of reconstruction under both maintenance and
improvement standards is discussed in Chapter D2 of the Analytical Framework and Model
Descriptions.
The specification of pavement construction as improvement works is given under the Design
and Pavement Tabs. (Note that under maintenance works, pavement details are specified
under the Design Tab).
! Design
❏ Pavement Type
following the improvement (reconstruction)
❏ Increase in carriageway width (up to 3 metres)
! Pavement
❏ Surface material
❏ Dry season Adjusted Structural Number (excluding subgrade contribution)
❏ Surface thickness (mm)
❏ Relative compaction
4.2.6 New construction of paved roads

Specification of new construction is effected by specifying a section alternative within a
selected project alternative, as described in Chapter B1 under Section 2.6.4.
4.3 Unsealed roads improvements
4.3.1 Partial widening
General
The works to which the standard applies are defined by the following:
! Name - of the improvement works
! Short code
! Surface class - to which the standard is applicable
! Improvement type
! Duration of the works (years)
! Intervention type
whether scheduled or responsive.
Design

Version 1.0
! Speed-flow type
! Road class
! Pavement type
of the section following improvement.

! Increase in carriageway width
resulting from the improvement (metres).

! Existing carriageway is to be regravelled
❏ gravel thickness (mm)

❏ unit cost of regravelling (per m3)
! Existing carriageway is 'not' to be regravelled
❏ Unit cost of spot-regravelling (per m3)

❏ unit cost of grading (per km)
Intervention
Partial widening can be on a scheduled based upon a calendar year. Alternatively,
intervention criteria can be based on the following parameters:
! Daily average volume capacity ratio
! Peak period volume capacity ratio
! Mean speed
! Minimum speed
! Two-way AADT
For both scheduled and responsive works, the time of intervention can be limited by
specifying the last applicable year in which the works can take place.
Costs
Costs are specified in terms of the following:
! Unit rates (per km)
! Percentage of costs incurred over each year of the works period
Pavement
The following data can be specified:
! Material properties
! Compaction method (mechanical or non-mechanical)
! Gravel thickness (mm)
In the case of earth roads gravel thickness is set to zero
Geometry
The following data under road geometry can be specified:

Version 1.0
2
! Acceleration noise (m/s ) due to driver and road alignment

(no reduction)
reduction)
reduction)
Effects
intervention has been made. The following effects are specified:
These effects are specified in terms of:

In addition, the following effects can be specified under Effects/Edit Detailed Calibration:
! Gravel loss factor
! Traffic-induced material loss factor
! Maximum roughness, IRI, (m/km)
! Minimum roughness, IRI, (m/km)
A detailed discussion of road works, the intervention criteria that determine when they should
be applied, their design, effects on the road characteristics, and the costs of implementing each
works activity is given in Part D of the Analytical Framework and Model Descriptions.
4.3.2 Lane addition
General
Specification of lane addition to unsealed roads is identical to that for partial widening of
unsealed roads (see Section 4.3.1), with the exception of the specification under design.
Design
The design of the lane addition works is specified in terms of the following:
! Speed-flow type
! Road class
! Pavement type
! Additional number of lanes

Version 1.0
4.3.3 Realignment of unsealed roads
General
Specification of realignment of unsealed roads is identical to that for partial widening of
unsealed roads (see Section 4.3.1), with the exception of the specification under design and
geometry.
Design
The design of realigned unsealed roads is specified in terms of the following:
! Speed-flow type
! Road class
! Pavement type
! length adjustment factor
! Proportion of new construction
The ratio of new construction length to the section length after realignment works, with a
value between zero and one.
Geometry
The following data under geometry can be specified for realignment:
2
! Acceleration noise (m/s ) due to driver and road alignment

(no reduction)
reduction)
reduction)
4.3.4 Upgrading of unsealed roads

Upgrading of unsealed roads involves constructing bituminous or concrete pavements, or the
upgrading of an earth road to a gravel road.

Version 1.0
General
Specification of the upgrading of unsealed roads is similar to that for partial widening of
unsealed roads (see Section 4.3.1), but there are exceptions under design, intervention,
pavement, geometry and effects. These are described below:
Design
! Speed-flow type
! Road class
! Pavement type
! Increment in road width - resulting from the improvement (metres)
Depending on the type of road resulting from the upgrading, the construction details are
! Bituminous roads
As for Full Reconstruction of bituminous roads (see Sections 4.2.5).

! Concrete roads
As for upgrading of bituminous roads (see Section 4.2.4).

! Gravel roads
In this case no additional construction details may be specified.
Intervention
Partial widening can be on a schedule based upon a calendar year. Alternatively, intervention
can be based on the following criteria:
! Speed
❏ mean speed (km/h)

❏ minimum speed (km/h)
! Traffic
❏ two-way AADT
❏ cumulative equivalent standard axle loads (CESAL)
❏ Peak period volume-capacity ratio
❏ daily average volume-capacity ratio
Pavement
Depending on the road surface class resulting from the upgrading, the pavement details are

Version 1.0
! Bituminous
As for partial widening of bituminous roads (see Section 4.2.1).

! Concrete
As for upgrading of bituminous roads (see Section 4.2.4).

! Gravel
Gravel material properties, thickness and compaction method.
Geometry
The following data under geometry can be specified for upgrading:
2

(no reduction)
reduction)
reduction)
Effects
intervention has been made, and depend on the type of road resulting from the upgrading.
Bituminous roads
! Change in Traffic flow pattern
These effects are specified in terms of the following types of distress:

❏ roughness, IRI, (m/km)

Version 1.0
Concrete pavements
! Changes in road use
These effects are specified in terms of:

Gravel pavements
! Change in Traffic flow pattern
Measured in terms of:

other calibration factors can be specified under Section/Details/Surface Distress.
4.3.5 New construction of unsealed roads

Specification of new construction is effected by specifying a section alternative within a
selected project alternative, as described in Chapter B1 under Section 2.6.4.

Version 1.0
Part A Appendices
Contents
Volume 2 - Applications Guide: Appendix A -

Project Analysis Case Studies
A1 – Project Analysis Case Study 1:

1 Upgrading a Gravel Road to a Bituminous Pavement 1
2 Locate the case study data 2
3 Review the case study input data 3
3.1 General 4
3.2 Select Sections 4
3.3 Select Vehicles 9
3.4 Define Normal Traffic 12
3.5 Alternatives 14
3.6 Overview of project alternatives 19
3.7 Project Alternative: Without Project 20
3.8 Project Alternative: With Project 33
4 Run HDM-4 and examine the results 61
4.1 Setup Run 61
4.2 Run Analysis 61
4.3 Deterioration and works effects 62
4.4 Cost streams 64

1 Optimum Rehabilitation Standards for a Paved Road 1
1.1 Locate the case study data 1
1.2 Review the case study input data 1
1.3 Run HDM-4 and examine the results 38
1.4 Generate reports 39
1.5 Selected reports 42
Version 1.0
APPENDICES CONTENTS
1 Traffic Volume Capacity Improvements 1

1.4 Generate reports 21

1 New Bypass Construction (traffic diversion effects) 1
Version 1.0
Appendix A1 - Project Analysis Case

Study 1
1 Upgrading a Gravel Road to a Bituminous

Pavement
This case study presents the economic analysis of a project to upgrade an existing gravel road
to a paved standard. The existing road is 50 km long and passes through varying topography.
For analysis purposes, three sections, based on geometry, pavement condition, and traffic
volume can represent the road. Traffic and condition data are available from surveys
undertaken in 1998. The gravel thickness in 1998 was 150 mm.
The objective of the case study is to demonstrate the structure and capabilities of HDM-4, and
the steps needed to undertake the project appraisal.
investment. (This differs from a financial appraisal that is concerned with the means of
financing a project and the financial profiTability of the project). The economic feasibility of
the project is assessed by comparison against a base-line project alternative (that is, a without
project alternative). The project alternatives are:
! Without Project (see Section 3.7)
Maintain existing gravel road.

! With Project (see Section 3.8)
Maintain existing gravel road before upgrading to a bituminous pavement, followed by

maintenance of the bituminous pavement.
! Locate the case study data (see Section 2)
! Review the case study input data (see Section 3)
! Run HDM-4 and examine the results (see Section 4)
Applications Guide 1
Version 1.0
PROJECT ANALYSIS CASE STUDIES APPENDIX A1 - PROJECT ANALYSIS CASE STUDY 1
2 Locate the case study data

The case study data are included in the default daTabase installed with the HDM-4 software.
The data for this case study are located in the Projects folder in the Case Studies Workspace.
The name of the case study is Case Study 1: Upgrading a gravel road.
! Double-click on the case study named Case Study 1: Upgrading a gravel road
Version 1.0
3 Review the case study input data

The case study input can be reviewed under the HDM-4 Projects work flow buttons and
associated Tabs as follows:

❏ General (see Section 3.1)
❏ Select Sections (see Section 3.2)
❏ Select Vehicles (see Section 3.3)
❏ Define Normal Traffic (see Section 3.4)
The following Tab page may be displayed:

❏ Alternatives (see Section 3.5)
! Analyse Projects

❏ Setup Run (see Section 4.1)
❏ Run Analysis (see Section 4.2)
! Generate Reports
❏ Select Reports
It is recommended that the user keep a record of all input data in hard copy format, noting the
sources of all information. This will assist when input data is checked (for example, on
HDM-4 reports) and outputs reviewed, possibly by different personnel.
Version 1.0
3.1 General
This Tab page confirms the project title and type of analysis required - project analysis in this
case, as the task is to compare, for all three road sections, the upgrade alternative against the
do-minimum alternative.
The project start year has been defined as 2000 with the analysis period (duration) specified as
20 years. The road network to be used for the project has been pre-defined under the name
Northern Province in the Road Network folder. Similarly the vehicle fleet has been pre-
defined in the Vehicle Fleet folder under the same name. The road sections (within the
Northern Province road network) and vehicles types (from the Northern Province vehicle
fleet) to be used for the analysis are selected under the Select Sections and Select Vehicles
Tabs respectively, as discussed below:
Finally the General Tab page confirms that the input and output currencies are both US
dollars.
3.2 Select Sections

The road section is the basic entity for all calculations of pavement deterioration, construction
and maintenance costs, and economic analysis. Each section is considered homogeneous in
terms of its physical attributes (for example, road class, climate, carriageway width, geometry,
pavement condition, traffic flow, and axle loading).
For this case study, the road has been divided into three sections based on those physical
attributes that vary along its length. Table A1.1 contains a summary of the key physical
attributes for each section.
Note that in this case study, no shoulders are specified, and the full width of gravel surfacing
is assigned as carriageway width. In practice it is often difficult to identify shoulders on a
gravel road.
Version 1.0
Table A1.1
Case study 1 - Summary of section attributes
Section ID B001-01 B001-02 B001-03

General
Section Name Town A to Town B Town B to Town C Town C to Town D
Length (km) 20 10 20
Shoulder width (m) 0 0 0
Number of shoulders 0 0 0
AADT (1998) 400 200 600
Flow direction 2 - way 2 - way 2-way
Geometry (before
upgrading)
Rise and Fall (m/km) 10 15 30
*No rises and falls (no/km) 0.1 0.4 1.0
Av horiz curv (deg/km) 50 300 15
*Superelevation % 3.0 3.0 3.0
*σdral 0.1 0.1 0.1
*Speed limit enforcement 1.1 1.1 1.1
Altitude (m) 120 295 370
Condition 1998
Gravel Thickness 150 150 150
Roughness (IRI, m/km) 8 8 8
Notes:
1 Geometry Data marked with an * (asterisk) is stored under Section/Details/Alignment.

Other geometric data are held under Section/Geometry.
Traffic flow (AADT) used in project analysis is defined by section at project level under
Define Project Details/Define Normal Traffic, and will often be an update of base traffic
data retained at section level under Network/Section/Definition.
Version 1.0
The Select Sections Tab page displays those sections of the Northern Province road network
that will be included in the analysis. For this case study, the three sections listed should have
a tick in the Include column to confirm their selection.
Each section has been assigned a unique Description and ID, either or both of which may be
used to identify the section on HDM-4 reports. In this case study, the section description
identifies towns at the ends of each section. It is helpful if the Section ID includes reference
to the road class and road number (denoted by B001 in this case study).
By double-clicking on one section in the Select Section Tab page, the characteristics of that
section can be reviewed under four Tabs (Definition, Geometry, Pavement, and Condition).
The corresponding Tab pages are reviewed below for one of the sections from this case study
(Section Description: Town A to Town B).
! Section/Definition
Lists basic characteristics such as speed-flow type and traffic-flow pattern, also
carriageway length and width. The pavement type is defined as Gravel (HDM-4
pavement type GRUP, Granular Unsealed Pavement).
Section pavement details are stored under the Pavement Tab.
Version 1.0
! Section/Geometry
Summarises existing geometric parameters (before upgrading) including those for

horizontal and vertical alignment. Note that certain geometric parameters (marked with
* in Table A1.1) are held under Section/Details/Alignment.
❏ Section/Geometry screen
Version 1.0
❏ Section/Geometry/Alignment screen
! Pavement
❏ Section/Pavement screen
Gives pavement layer descriptions and year of last regravelling. The material properties
can be reviewed under Section/Details/Material Gradation.
Version 1.0
❏ Section/Details/Material Gradation screen
! Condition
Confirms gravel thickness 150 mm and roughness 8 m/km (IRI) in1998. Note that, to
enable road condition to be modelled through the analysis period, condition data must be
specified for a date prior to the start of the analysis period. In this case study the
condition data refer to the end of 1998, which is prior to the analysis start year 2000.
❏ Section/Condition screen
3.3 Select Vehicles

This Tab page displays those vehicles that were selected (from the pre-defined Northern
Province Vehicle Fleet) for this case study analysis.
Version 1.0
The tick in the Include column confirms the selection. This list defines the range of vehicle
types that can be assigned to each road section, although the full range need not be used. The
assignment of traffic by section is effected under the Define Normal Traffic Tab, discussed
later.
Vehicle attributes for a particular vehicle type can be reviewed by double-clicking on the
appropriate vehicle type descriptions. The vehicle attributes are held under four Tabs:
Definition, Basic Characteristics, Economic Unit Costs, and Financial Unit Costs.
An example of each Tab page for one vehicle type is shown below:
! Vehicle Attributes/Definition screen
Version 1.0
! Vehicle Attributes/Basic Characteristics screen
! Vehicle Attributes/Economic Unit Costs screen
Version 1.0
! Vehicle Attributes/Financial Unit Costs screen
Note that the vehicle attributes are defined for a representative vehicle within each vehicle
class. For example, under Basic Characteristics, the loading details (No of equivalent
standard axles and operating weight) are based on the average weight of vehicles in the
vehicle class.

This Tab page gives the traffic volume (AADT) on each road section in the specified year.
When setting up a specific case study, the base AADT held by section is automatically
assigned to the Define Normal Traffic Tab page. However the user may update the
AADT/year information for the purposes of the case study. In this event the AADT data held
at section level (under Network/Section/Definition) remains unchanged.
Version 1.0
By double-clicking on the row representing a specific section, the traffic composition in the
specified year (by % of each vehicle type) can be reviewed, together with vehicle growth
rates. Thus for section A-B, the normal traffic details are as given below:
! Define Normal Traffic/Motorised screen
For this case study, only one growth period is indicated over the 20-year analysis period.
Several growth periods may be defined within the analysis period. The user is encouraged to
define future growth rates (by vehicle type) as accurately as possible. Note that negative
growth rates are permitted.
Version 1.0
3.5 Alternatives
The Alternatives Tab is displayed after selecting the Specify Alternatives button.
The two project alternatives considered in this case study are defined below. The first
alternative, Without Project, represents a continuation of current maintenance practice. The
second alternative, With Project, represents the implementation of the project to upgrade the
existing gravel road to paved standard.
! Without Project (see Section 3.7)
This project alternative comprises grading every six months; spot regravelling to replace
30% of material lost each year (if the gravel thickness falls below 100 mm, and gravel
resurfacing (whenever the gravel thickness falls below 50 mm).
Note that in a given year, if gravel resurfacing is triggered, this operation supersedes
grading and spot regravelling.
! With Project (see Section 3.8)
This project alternative includes upgrading the road in the year 2004 (duration of works
is two years for road sections B001-01 and B001-03, and one year for section B001-02).
Before upgrading, the existing gravel road will be maintained, by grading, every six
months and spot regravelling to replace 30% of material lost each year (if the gravel
thickness falls below 100 mm). The latter works will maintain some protection of the
subgrade prior to upgrading.
After upgrading, the road will receive routine maintenance in the form of crack sealing (if
wide structural cracking reaches 5%) and patching (if the severely damaged area reaches
5%).
The analysis period is 20 years, starting in the year 2000, and analysis is by project as
specified under the General Tab.
The Alternatives Tab page is split into two boxes. The upper box shows the names of the two
project alternatives that have been set up for this case study.
Version 1.0
The bottom box shows details, by section, of the Road Works Standards (assignments)
associated with each project alternative. Note that the assignments box shows the Road
Works Standards corresponding to the selected project alternative and section. For the
selected alternative, the Road Works Standards assigned to a different section can be reviewed
by selecting that section.
A schematic overview of the Project Alternatives and their constituent Road Works Standards
is shown in Figure A1.1. Note that Road Works Standards are sub-divided into Maintenance
Standards denoted by M , and Improvement Standards denoted by I .
The Road Works Standards and associated works assigned to each project alternative are
summarised in Table A1.2. A summary of the intervention limits for maintenance works is
given in Table A1.3. These Tables allow the user to check consistency of data across
alternatives.
Each Road Works Standard has an associated date, assigned at project level, which represents
the year from which the standard takes effect. For the Without Project alternative, each of
the three road sections has been assigned a Maintenance Standard named Gravel Road
Maintenance. This assignment commences in the year 2000, which is the first year of the 20-
year analysis period (2000-2019).
For the With Project alternative, each section has been assigned a set of Road Works
Standards as follows:

Note that (for a given feature) only one Maintenance Standard or Improvement Standard will
be effective in any analysis year. However a Maintenance Standard may include more than
one works item which could be implemented in a given year, for example, grading and spot
regravelling on unsealed roads, crack sealing and patching on bituminous roads. Details of
the logical rules associated with implementation of works are given in Chapter D1 of the
Version 1.0
Without project With project
Calendar year All sections Section Section Section

B001-01 B001-02 B001-03
1998 Traffic and Condition
data
1999
2000 (1) (2) (2) (2)
2001 (1) (2) (2) (2)
2002 (1) (2) (2) (2)
2003 (1) (2) (2) (2)
2004 (1) (3) (4) (5)
2005 (1) (3) (6) (5)
2006 (1) (6) (6) (6)
2007 (1) (6) (6) (6)
2008 (1) (6) (6) (6)
2009 (1) (6) (6) (6)
2010 (1) (6) (6) (6)
2011 (1) (6) (6) (6)
2012 (1) (6) (6) (6)
2013 (1) (6) (6) (6)
2014 (1) (6) (6) (6)
2015 (1) (6) (6) (6)
2016 (1) (6) (6) (6)
2017 (1) (6) (6) (6)
2018 (1) (6) (6) (6)
2019 (1) (6) (6) (6)
(1) M Maintenance Standard: Gravel road maintenance
(2) M Maintenance Standard: Maintenance before upgrading
(3) I Improvement Standard: Pave section B001-01
(6) M Maintenance Standard: Crack sealing and patching paved road
Figure A1.1 Case Study 1: Overview of project alternatives and road work
standards
Version 1.0
Table A1.2
Project Section Road Works Effective Maintenance works/

alternative ID Standards from year Improvement type
Maintenance
B001-01
Without Project M Gravel Road 2000 Grading (GRADE6)
Maintain Maintenance
Gravel (GRAVEL) Spot regravelling (SPG100)
Road B001-02

Maintenance
B001-03

upgrading (BEFORE)
I Pave Section B001-01 2004 Upgrading

B001-01 in 2004 (PAVE01)

patching paved road

upgrading (BEFORE)
With Project I Pave Section B001-02 2004 Upgrading

B001-02 in 2004 (PAVE02)
Upgrade Gravel
Road M Crack sealing and 2005 Crack sealing (SEAL)
patching paved road

upgrading (BEFORE)
B001-03 I Pave Section B001-03 2004 Upgrading

in 2004 (PAVE03)

patching paved road
Version 1.0
Table A1.3
Case study 1: Intervention limits for maintenance works
Road works Effective Maintenance * Last Max Max Interval AADT

standard from year works Year IRI Quantity
S/R Min Max Min Max
M Gravel Road 2000 Grading S 2019 30 n/a 180 days 750 days 0 100,000
Maintenance
Gravel resurfacing R 2017 30 n/a 3 yrs 99 yrs 0 100,000
M Maintenance before 2000 Grading S 2003 30 n/a 180 days 10000 days 0 100,000
upgrading
M Crack sealing and 2006 Crack sealing R 2019 12.5 1500 n/a n/a 0 100,000
patching paved road
m2/km/yr
Patching R 2019 12.5 1500 n/a n/a 0 100,000

2
m /km/yr
Notes:
n/a not applicable
* S = Scheduled intervention
R = Responsive intervention
Version 1.0
3.6 Overview of project alternatives

The use of the HDM-4 software to set up the Road Works Standards for this case study is
described in more detail below. Each Project Alternative includes one or more Maintenance
Standards and/or Improvement Standards that are assigned to be effective from a specified
date. Each Maintenance Standard may define several Works Items, and each Improvement
Standard defines a specific improvement type.
Without Project
M
Gravel Road Maintenance
Works Item: Works Item: Works Item:

Grading Spot regravelling Gravel resurfacing
With Project
M M
I
Maintenance before Crack sealing and
Upgrading works
upgrading patching paved roads
Works Item: Works Item: Works Item: Works Item:

Grading Spot regravelling Crack sealing Patching
Improvement Improvement Improvement

type: type: type:
upgrading upgrading upgrading
section section section
B001-01 B001-02 B001-03
Figure A1.2 Schematic diagram showing details of Project Alternatives for Case
Study 1
Version 1.0
3.7 Project Alternative: Without Project

This Project Alternative includes one Maintenance Standard associated with the existing
gravel road.
3.7.1 Maintenance Standard: Gravel Road Maintenance

The specification for this Maintenance Standard can be reviewed via the Edit Standards
button (or alternatively via HDM-4 Workspace). This displays a list of the names of all
Maintenance Standards that have been pre-defined within the Road Works Standards folder.
By double-clicking on Gravel Road Maintenance, the window for the specification of this
maintenance standard is displayed.
Version 1.0
This window shows the names and codes assigned to the Gravel Road Maintenance
Maintenance Standard and constituent Works Items. These names and codes appear on
certain HDM-4 reports so it is essential that they are unique within a Maintenance Standard.
The software checks for unique names and codes when Maintenance Standards are set up.
The window lists the three works items that are associated with the Gravel Road
Maintenance Maintenance Standard. As previously described, the Without Project
alternative applies grading every six months, spot regravelling to replace 30% material lost
each year (if the gravel thickness falls below 100 mm), and gravel resurfacing (if the gravel
thickness falls below 50 mm). The original gravel thickness was 150 mm (1998).
Note that in a given year, spot regravelling and grading are both permitted as these are
complimentary works. However if gravel resurfacing is triggered, this will supersede the
other works. Hence in a given year, the possible works under the Gravel Road Maintenance
Maintenance Standard are:
! Grading every six months, or
! Spot regravelling AND Grading every six months, or
! Gravel resurfacing
The Works Items and Operations associated with the Maintenance Standard are shown in
Figure A1.3. Full details of the logical rules associated with the implementation of works are
given in Chapter D1 of the Analytical Framework and Model Descriptions.
Name: Gravel Road Maintenance

Maintenance Standard:
Code: GRAVEL
Name: Grading Spot regravelling Gravel resurfacing

Works Items:
Code: GRADE6 SPG100 RESURF
Regravelling/
Operation/Works Activity: Grading Spot regravelling
Resurfacing
Figure A1.3 Hierarchy within a Maintenance Standard
Note that, when setting up a Maintenance Standard, the user defines the name and code for the
Maintenance Standard and Maintenance Works. The name of the Operation/Works Activity
is selected from a pre-defined list. It is useful if the name/code assigned to the Works Item
also includes information on intervention (for example, grading every six months). It may
also be useful if the name given to the Maintenance Standard refers to the surface type (for
example, Gravel Road Maintenance).
Works Item: Grading

By double-clicking on a particular works item, the Maintenance Works Item specification
window is displayed. The specification for Grading is contained under four Tabs, General,
Design, Intervention, and Costs:
Version 1.0
! Grading/General
The General Tab page for Works Item Grading is shown below:
This confirms the Name and Short Code of the Works Item. The operation type/works
activity (specified from a pre-defined list) is confirmed as grading, and intervention type
is scheduled.
! Grading/Design
The Design Tab page for grading specifies the compaction method as mechanical.
Version 1.0
! Grading/Intervention
The Intervention Tab page confirms the frequency at which the grading works will be
undertaken (that is, 180 days).
In general, scheduled grading works are timed from the last regravel year, specified
under Section/Pavement (1998 in this case study). However the works can only be
implemented after the date when the Maintenance Standard becomes effective (specified
at project level, 2000 for this case study).
Note that for scheduled works, the frequency of works can also be specified in terms of
traffic interval (that is, number of vehicles between successive gradings).
The Intervention screen also shows limits outside which the grading works would not
be triggered:
❏ Last year
The last year in which grading should be considered. For this Maintenance Standard
the last year for grading is set at 2019 (the last year in the analysis period).
This parameter could be used to ensure that a given works is not triggered too close
to the end of the analysis period.
❏ Maximum roughness
Sets the maximum roughness at which grading would be undertaken (30 m/km IRI).
❏ Interval
The minimum interval between successive grading operations has been set at 180
days (six months).
The maximum interval between successive grading operations, above which grading
will not be implemented, has been specified as 750 days (just over two years) to
allow for the fact that grading will not be triggered in a year when gravel resurfacing
Version 1.0
is implemented. If the specified interval is exceeded, grading will not be triggered,

but other Works Items under this Maintenance Standard would be considered.
❏ AADT
Allows the user to specify the AADT range over which the works are considered
appropriate. For this case study, the range is specified as 0-100,000 to avoid
elimination of grading based on this parameter.
! Grading/Costs
The unit costs for grading works (expressed in US dollars per km) can be reviewed under
the Costs Tab. Note that default costs for all works operations are specified under the
HDM Workspace option Road Works Standards/Default Works Costs & Energy
Consumption.
By clicking the OK button (on the Costs screen), the user is returned to the entry screen
for the Maintenance Standard Gravel Road Maintenance.
Works Item: Spot regravelling

The specification for Spot regravelling is contained under four Tabs, General, Design,
Intervention, and Costs. The Tabs are revealed by double-clicking the Spot regravelling
Works Item in the Maintenance Standard window.
Version 1.0
! Spot regravelling/General
The General Tab page for the Works Item Spot regravelling is shown below:
This confirms the Name and Short Code of the Works Item. The operation type/works
activity is spot regravelling, with intervention type responsive.
! Spot regravelling/Design
Two design options for spot regravelling are given under the Design Tab.
Version 1.0
Either a specified volume of gravel can be added each year (specified as m3/km/yr), or a
specified percentage of the annual gravel loss is replaced (for example, 100% would
replace all gravel lost). For this case study the latter design option is adopted, replacing
30% gravel lost each year. Note that the amount of gravel actually applied to the road
section will be subject to user defined limits specified under Intervention.
! Spot regravelling/Intervention
The Intervention Tab page gives the responsive criterion that has been specified for spot
regravelling (that is, whenever gravel thickness falls below 100 mm). Note that with
some works activities, more than one intervention criterion can be specified; however
with spot regravelling, gravel thickness is the only parameter that is available.
The Intervention Tab page also gives limits affecting the spot regravelling works:
❏ Last year
The last year in which spot regravelling should be considered. For this case study,
the last year is set at 2019 (last year of analysis period).
Sets the maximum roughness at which spot regravelling would be undertaken (30
m/km IRI).
❏ Maximum quantity
Sets an upper limit on the quantity of material to be used each year for spot
regravelling, specified as m3/km/year. This could reflect the upper limit considered
practicable before complete resurfacing should be considered, but should also take
account of availability of material and work capacity of the Roads Authority. For
this case study, the quantity specified is 100 m3/km/year.
Version 1.0
❏ AADT
This allows the user to specify the AADT range over which the works are
considered appropriate. For this case study, the range is specified as 0-100,000 to
avoid elimination of grading based on this parameter.
! Spot regravelling/Costs
The unit assigned to spot regravelling (expressed in US Dollars per cubic metre) can be
reviewed via the Costs Tab.
Works Item: Gravel Resurfacing

The third Works Item considered under the Maintenance Standard Gravel Road
Maintenance is Gravel Resurfacing. The specification for this can be reviewed under the
Tabs General, Design, Intervention, Costs, and Effects:
Version 1.0
! Gravel Resurfacing/General
The General Tab page, shown below, confirms that the Operation Type/Works Activity
is Regravelling/Resurfacing and that the Intervention Type is responsive (the intervention
criterion is defined under the Intervention Tab).
Version 1.0
! Gravel Resurfacing/Design
The Design Tab page for Gravel Resurfacing gives material properties (particle size
distribution and Plasticity Index) of the proposed surfacing material. In addition, the
final gravel thickness (or increase in gravel thickness) and compaction method are
specified. For this case study, a final gravel thickness of 150 mm is specified, using
mechanical compaction.
Version 1.0
! Gravel Resurfacing/Intervention
For this case study, the intention is to resurface only if the gravel thickness falls below 50
mm. This is reflected in the intervention criterion adopted. The implementation of
resurfacing is subject to the user-defined limits indicated under the Intervention Tab,
shown below:
Version 1.0
Note that, for gravel resurfacing, roughness could also be used as the intervention
criterion (for example, if the roughness is 15 m/km IRI or greater). Both parameters
(layer thickness and roughness) could be used together., If more than one criterion is
used with a given works item, a logical AND is assumed, and displayed on-screen (for
example, if gravel thickness falls below 50 mm AND roughness is 15 m/km IRI or
greater).
If the user wishes to trigger gravel resurfacing based on either layer thickness OR
roughness (logical OR), then two separate Works Items for gravel resurfacing would
be set up under the same Maintenance Standard with the appropriate intervention
criteria.
To summarise, examples of intervention options for responsive gravel resurfacing
are:
Intervention is based on:
! Gravel Thickness only
Works Item (gravel resurfacing) with one intervention criterion:

❏ Gravel thickness <= 50 mm
! Roughness only
Works Item (gravel resurfacing) with one intervention criterion:

❏ Roughness >= 15 IRI
! Gravel Thickness AND Roughness
Works Item (gravel resurfacing)) with two intervention criteria:

❏ Gravel Thickness <= 50 mm
AND
! Either Gravel Thickness OR Roughness
Works Item (gravel resurfacing) with thickness criterion:

❏ Gravel Thickness<=50 mm
Works Item (gravel resurfacing) with roughness criterion:
Version 1.0
The Intervention Tab page also shows the limits on the use of gravel resurfacing:
❏ Last year
The last year in which gravel resurfacing would be considered, set at 2017, year 18
of the 20 year analysis period.
Sets the maximum roughness at which gravel resurfacing would be undertaken (30
m/km IRI).
❏ Interval
The minimum interval between successive resurfacings has been specified as three
years for this case study.
The user must also specify the maximum interval between successive resurfacing
operations, above which resurfacing will not be considered. As resurfacing is the
highest ranking maintenance operation available under this Maintenance Standard, a
value of 99 years is used to avoid elimination by this limit.
❏ AADT
For this case study, no limits on AADT are imposed (that is, minimum and
maximum values set at 0 and 100,000 respectively). In this case the minimum
interval specified above will override.
! Gravel Resurfacing/Costs
The unit cost assigned to gravel resurfacing (expressed as US Dollars per cubic metre)
can be reviewed under the Costs Tab.
Version 1.0
! Gravel Resurfacing/Effects
Note the Effects Tab with this Works Item, which gives the initial roughness value
immediately after the gravel resurfacing works are completed.
A user defined value of 3 m/km IRI is used for this case study.
The model calibration factors relating to gravel loss can be reviewed under
Section/Details. Details of the modelling logic for unsealed roads are given in Chapter
D4 of the Analytical Framework and Model Descriptions.
3.8 Project Alternative: With Project

This Project Alternative includes the following Road Works Standards:
3.8.1 Maintenance Standard: Maintenance Before Upgrading

Prior to upgrading, the existing gravel road will be given routine maintenance in the form of
grading every six months and spot regravelling (if the gravel thickness falls below 100 mm).
The works are specified on the Maintenance Standard window.
Version 1.0
Works Item: Grading

The specification for Grading within the Maintenance Standard Maintenance Before
Upgrading is similar to that described under the Standard Gravel Road Maintenance. The
Tab pages under General, Design, Intervention and Costs are shown below:
! Grading/General
Version 1.0
! Grading/Design
! Grading/Design
Version 1.0
! Grading/Costs
The intervention limits associated with the grading works are given on the Intervention
screen, and summarised in Table A1.3. Note that the Last Year is specified as 2003, the year
before the start of the upgrading works.
Works Item: Spot regravelling before upgrading

The specification for spot regravelling under the Maintenance Standard Maintenance Before
Upgrading is similar to that described under the Maintenance Standard Gravel Road
Maintenance. Details of the Tab pages (General, Design, Intervention, and Costs) are
shown below:
Version 1.0
! Spot regravelling/General
! Spot regravelling/Design
Version 1.0
! Spot regravelling/Intervention
! Spot regravelling/Costs
The intervention limits associated with the spot regravelling works are given on the
Intervention Tab page, and summarised in Table A1.3.
Version 1.0
3.8.2 Improvement Standards

The specification for an Improvement Standard includes details of the improvement type
(upgrading in this case study), and geometric characteristics of the road section after
improvement. The geometry data set may be assigned by selecting a geometry class (for
example, mostly straight and gently undulating), as defined under HDM-4 Configuration
(see Chapter D1).
For this case study, the geometric characteristics are significantly different for each road
section, as are the pavement design details and duration of the upgrading works. The different
Improvement Standards for the three sections are reviewed below. A comparison of selected
section details, before and after upgrading, is given in Table A1.4.
Table A1.4
Section details before and after upgrading works
Section ID B001-01 B001-02 B001-03

Section Name Town A - Town B Town B - Town C Town C - Town D
Speed flow type B Two lane road Two lane road Two lane road
A Two lane road Two lane road Two lane road
Traffic flow pattern B Seasonal Seasonal Seasonal
A Seasonal Seasonal Seasonal
Road class B Secondary/Main Secondary/Main Secondary/Main
A Primary/Trunk Primary/Trunk Primary/Trunk
Surface class B Unsealed Unsealed Unsealed
A Bituminous Bituminous Bituminous
Pavement type B Gravel Gravel Gravel
A Surface Treatment on Surface Treatment on Surface Treatment on

Granular Base Granular Base STabilised Base
Carriageway width (m) B 7 6 8
A 7 6 8
Speed limit (km/h) B 100 80 100
A 120 100 120
Upgraded SN 1.84 1.61 2.62
(excluding subgrade)
Upgrade financial cost 280000 300000 360000
(US dollars/km)
Notes:
B = Before upgrading
A =After upgrading
Version 1.0
3.8.3 Improvement Standard: Pave Section B001-01 in 2004

The specification for this improvement can be reviewed under the Tabs: General, Design,
Intervention, Costs, Pavement, Geometry, and Effects. The information needed for each
Tab page is reviewed below.
! General
Confirms the improvement type upgrading, with duration of two years, and scheduled
intervention.
! Design
Version 1.0
Confirms Pavement Type as Surface Treatment on a Granular base (STGB), as shown in

Figure A1.4. The intention is to replace the existing gravel surfacing with imported
granular material (150 mm) as sub-base, add 200 mm granular roadbase, with a double
surface dressing on top. The design parameters required by HDM-4 are entered under the
Pavement Tab below.
The length adjustment factor is specified as 1.0 as no change in length is planned (that is,
no realignment). Similarly no increase in width is proposed.
Version 1.0
H1 = 25mm Surface Treatment
H2 = 200mm Imported granular roadbase (CBR 100%)
Gravel Surfacing 150mm max H3 = 150mm Imported granular sub-base (CBR 60%)
Subgrade Subgrade
Before Upgrading After Upgrading
Design: Pavement Type = STGB (Surface Treatment on Granular Base)
Pavement: Surface Material = Double Bitumen Surface Dressing
Dry Season Structural Number = SN = 1.84
Surface Thickness = 25mm
H2 and H3 not used directly by HDM-4 but included in costs
and contribute to SN
Figure A1.4 Case Study 1: Pavement design for Section B001-01 (Town A to Town B)
Version 1.0
The Construction Quality Indicators (CDS and CDB) are specified under Design/Edit
Construction Details. For this Improvement Standard, CDS = 1 and CDB = 0. Further
details of these parameters are given in Chapters C2 and D2 of the Analytical Framework and
Model Descriptions.
! Intervention
This confirms the scheduled implementation date as the year 2004. The Effective from
year assigned to the Improvement Standard (at project level) is also specified as 2004.
Note that the Effective from year could be specified differently to the implementation
year. For example, if no maintenance was needed for the year prior to the start of
improvement works, then the assignment (effective from) dates for standards would be:
M Maintenance before upgrading 2000 (start of analysis period)
I Pave Section B001-01 in 2004 2003
(Note: Improvement Standard effective from 2003, works implementation is 2004. The
Improvement Standard takes precedence over the previous Maintenance Standard, so no
works would be triggered in 2003).
M Maintenance after upgrading 2006 (condition responsive)
Version 1.0
! Costs
The costs (expressed in US Dollars per km) are spread across the two-year construction
period specified under the General Tab. 50% costs have been assigned to the year 2004,
and 50% to 2005.
! Pavement
This Tab page is used to specify details of the new pavement after upgrading.
❏ The proposed pavement design is included in Figure A1.4.

❏ Surface Material confirms the use of a double bitumen surface dressing. Note that
the options given in the Surface Material drop-down list depend on the Pavement
Type selected under the Design Tab (STGB for this road section).
Version 1.0
❏ Dry Season Structural Number, SN, is derived for the pavement layers indicated in
Figure A1.4 (that is, surface, granular base and sTabilised sub-base). Note that SN
does not include a contribution from the subgrade.
For road section B001-01, the value of SN has been derived from layer thicknesses
and coefficients as indicated below:
Layer Layer type Thickness Layer coefficient
(mm)
Surfacing ST 25 0.2
Base GB 200 0.14
Granular Sub-base - 150 0.11
SN s = SNBASU s + SNSUBA s = 1.84
where:
SNBASUs = contribution of surfacing and base layers for season, s

= 0.0394 * [(25 * 0.2) + (200 * 0.14)] = 1.30
SNSUBAs = contribution of sub-base for season, s

= 0.54 from Equation 3.3 in Chapter C2 of the Analytical
Framework and Model Descriptions
Recommended values for layer coefficients are given in Chapters C2 and D2 of the
Analytical Framework and Model Descriptions, which also gives full details of the
HDM-4 definitions relating to Structural Number.
Note that roadbase details are not requested as the proposed roadbase is not
sTabilised.
❏ The surface thickness refers to the thickness of the new surfacing that is 25 mm for
this road section (Double bitumen surface dressing).
❏ Relative compaction of pavement layers, taken as 97% on average.
Version 1.0
! Geometry
In this case study, the effect of the upgrading works on Geometry is summarised in
Table A1.4. There are no changes to the horizontal or vertical alignment. Note that the
posted speed limit (after upgrading) is 120 km/h, compared with 100 km/h before
upgrading.
These data were based on the data held for the gravel road prior to upgrading. Note that
the geometry data before upgrading is held in the Road Network folder under
Section/Geometry and Section/Details/Alignment.
! Effects
The screen confirms that the road condition after the improvement works will be derived
by the HDM-4 Works Effects model. The calibration factors can be reviewed under the
Version 1.0
Edit Detailed Calibration button. Further details of calibration of Road Works Effects
are given in Part D of the Analytical Framework and Model Descriptions.

The Improvement Standard for this section has a different pavement design (detailed under
Pavement below) to that specified for Section B001-01. The construction period for Section
B001-02 is one year. The specification for this improvement can be reviewed under the Tabs:
General, Design, Intervention, Costs, Pavement, Geometry, and Effects. The information
needed for each Tab page is reviewed below.
! General
Confirms the improvement type upgrading, with duration of one year, and scheduled
intervention.
Version 1.0
! Design
Confirms the Pavement Type after improvement as Surface Treatment on Granular Base
(STGB). The pavement design is described under Pavement below. The factors CDS
and CDB can be reviewed under Design/Edit Construction Details.
! Intervention
The improvement has been scheduled to start in 2004.

! Costs
The construction period for this section is one year, and so all works costs are allocated
to the year 2004.
Version 1.0
! Pavement
This screen confirms details of the pavement after upgrading.
The pavement design for the upgrading of Section B001-02 is given in Figure A1.5. For
this road section, the intention is to remove the existing gravel surfacing to formation
level, then sTabilise the existing subgrade material (to a depth of 150 mm) below
formation. This improved subgrade will be considered (for HDM-4 purposes) as a
sTabilised sub-base on top of which a granular roadbase layer (150 mm thick) will be
placed. A double bitumen surface dressing (thickness 25 mm) will be applied as
surfacing.
This design may be appropriate where the existing subgrade material is generally weak
but is suiTable (in terms of plasticity and particle size distribution) for strengthening by
the addition of a lime or cement sTabiliser. Note that sTabilisation would normally be
considered only if the cost of the process is less than the cost of replacing the existing
weak material with imported suiTable material.
The structural number, SN is derived for the pavement layers indicated in Figure A1.5
(that is, surface, granular base and sTabilised sub-base). Note that SN does not include a
contribution from the subgrade.
For road section B001-02, the value of SN has been derived from layer thicknesses and
coefficients as indicated below:
(mm)
Surfacing ST 25 0.2
Base GB 150 0.14
STabilised sub-base - 150 0.11
Version 1.0
where:

= 0.0394 * [(25 * 0.2) + (150 * 0.14)] = 1.02

! Geometry
The geometric characteristics of road section B001-02 after upgrading are shown on the
Geometry Tab page. Note that the posted speed limit is changed from 80 km/h (before
upgrading) to 100 km/h.
! Effects
The condition after the upgrading works is derived by the HDM-4 Works Effects models.
Version 1.0
Gravel Surfacing 150mm max H2 = 150mm Imported granular roadbase (CBR 60%)
Subgrade
H3 = 150mm Cement sTabilised sub-base (CBR 60%)
Subgrade

Design: Pavement Type = STGB (Surface Treatment on Granular Base)
H2 and H3 not used directly by HDM-4 but included in
costs and contribute to SN
Figure A1.5 Case Study 1: Pavement Design for Section B001-02 (Town B to Town C)
Version 1.0

The Improvement Standard for this road section has a different pavement design (detailed
under Pavement below) to the previous sections. The construction period for Section B001-
03 is two years. The specification for this improvement can be reviewed under the Tabs:
General, Design, Intervention, Costs, Pavement, Geometry, and Effects. The information
needed for each Tab page is reviewed below.
! General
Name and Short Code refer to Section B001-03. The construction period for the
proposed improvement works is two years.
! Design
Confirms the pavement type after improvement as Surface Treatment on STabilised Base
(STSB). The pavement design is described under Pavement below.
! Intervention
Scheduled for 2004, as with the other road sections.

! Costs
As specified with 50% works costs incurred in each of the construction years 2004 and
2005.
Version 1.0
! Pavement
For this road section, the proposed pavement design comprises a 150 mm sTabilised sub-
base placed on existing formation level with a 200 mm sTabilised road base and a double
surface dressing (Figure A1.6).
For road section B001-03, the value of SN has been derived from layer thicknesses and
coefficients as indicated below:
(mm)
Surfacing ST 25 0.2
STabilised base SB 200 0.22
STabilised sub-base - 150 0.14
where:

= 0.0394 * [(25 * 0.2) + (200 * 0.22)] = 1.93

Version 1.0
The layer coefficients are representative of the as new condition of the pavement. It is
recognised that the sTabilised layers will develop cracking. Details of the deterioration
models for sTabilised layers are given in the Analytical Framework and Model
Descriptions.
The proposed design includes a sTabilised sub-base and sTabilised base. As the roadbase
is sTabilised, its thickness (200 mm) and Resilient Modulus (15 GPa) are requested.
HDM-4 does not require these parameters for the sTabilised sub-base.
Note that use of a sTabilised roadbase gives a significant risk of reflection cracking
through the surfacing. The risk is effectively reduced if a granular layer is placed above
the sTabilised layer (for example, as for section B001-02 using a sTabilised sub-base and
granular roadbase).
! Geometry
The geometric characteristics of road section B001-03 after upgrading are shown on the
Geometry Tab page. Note that the posted speed limit is changed from 100 km/h (before
upgrading) to 120 km/h.
! Effects
The condition after the upgrading works is derived by the HDM-4 Works Effects models.
Version 1.0
H2 = 200mm Cement sTabilised roadbase
Gravel Surfacing 150mm max H3 = 150mm Cement sTabilised sub-base (CBR 60%)
Subgrade
Design: Pavement Type = STSB (Surface Treatment on STabilised Base)

STabilised Roadbase: Base thickness = 200mm
Resilient Modulus = 15GPa
Figure A1.6 Case Study 1: Pavement design for Section B001-03 (Town C to Town D)
Version 1.0
3.8.6 Maintenance Standard: Crack Sealing and Patching Paved Road

After upgrading, the paved road is to receive routine maintenance, defined under the
Maintenance Standard Crack sealing and patching Paved Road. As summarised in Table
A1.2, this includes two works items:
1 Crack Sealing if wide structural cracking reaches 5%
2 Patching if the severely damaged area reaches 5%
Both works are condition responsive (as defined above) and are effective from the year
following the completion of the upgrading works (that is, effective from 2006 for sections
B001-01 and B001-03, and 2005 for section B001-02.) The effective start date is specified at
project level (under Alternatives).
The opening screen of this Maintenance Standard confirms the two work items Crack Sealing
and Patching.
Works Item: Crack Sealing

The specification for Crack Sealing is held under four Tabs: General, Intervention, Costs,
and Effects. The details required for each Tab page are reviewed below:
Version 1.0
! General
This Tab page confirms the Name and Code assigned to this Works Item. The works
activity is specified as crack sealing and intervention type is responsive.
! Intervention
The Intervention Tab page confirms the responsive criterion (Wide structural cracking
affecting 5% or more of carriageway area over the section). The other criterion available
is based on the number of transverse thermal cracks per km).
Version 1.0
The user defined limits are:
Last Year 2019, the last year of the analysis period.

Maximum Roughness 12.5 m/km IRI as the road is now bituminous.
Maximum Quantity The upper limit on crack sealing works, expressed in
m2/km/year. For this Works Item, the upper limit is set
at 1500 m2/km/year. This reflects the capacity of the road
authority to undertake crack sealing.
AADT Applicable range set to 0-100,000 to avoid elimination
based on traffic volume.
! Costs
The Costs Tab confirms the unit cost of the crack sealing works, (expressed in US
Dollars per square metre).
! Effects
The Effects Tab confirms the percentage of distress to be repaired in terms of

Transverse thermal cracking and Wide Structural Cracking (that is, set at 100% for
each). This is subject to the limits defined under Intervention above.
Works Item: Patching

The specification for Patching can be reviewed under the Tabs: General, Intervention,
Costs and Effects. The details required for each Tab page are reviewed below:
! General
Version 1.0
This Tab page confirms the Name and Code assigned to this Works Item. The works
activity is specified as patching and intervention type is responsive.
! Intervention
The Intervention Tab page confirms the responsive criterion (when 5% of section
carriageway area is severely damaged). Note that other criteria could be defined, based
on potholing, ravelling, or wide structural cracking.
The user defined limits are:
Last Year 2019, the last year of the analysis period.

Maximum Roughness 12.5 m/km IRI as the road is now bituminous.
Maximum Quantity The upper limit on patching works, expressed in
m2/km/year. For this works item, the upper limit is set at
1500 m2/km/year. This reflects the capacity of the road
authority to undertake patching.
AADT Applicable range set to 0-100,000 to avoid elimination
based on traffic volume.
! Costs
The Costs Tab confirms the unit cost of the patching works, (expressed in US Dollars per
square metre)
Version 1.0
! Effects
The Effects Tab confirms the percentage of distress to be repaired in terms of Potholing
(set at 100%). This is subject to the limits defined under Intervention above.
Note that the user may select one of the radio buttons shown. For this case study, the
responsive criterion is based on the number of potholes per kilometre, and this is
intended to trigger pothole patching.
Further details of road works effects are given in Chapter D2 of the Analytical
Version 1.0
4 Run HDM-4 and examine the results

The analysis setup is defined under the Setup Run Tab.
4.1 Setup Run

This screen confirms the base project alternative for economic analysis (maintain gravel road)
and the discount rate (6%).
The costs and benefits of the Upgrade gravel road alternative will be compared with those
for the Maintain gravel road alternative, as defined under
Specify Alternatives/Alternatives (Section 3.5). Note that the HDM-4 Workspace may hold
data for any number of projects with associated network, fleet and work standards data. The
run that has been set up will consider only those project alternatives defined under
'Alternatives', with the specified selected sections and vehicles.
For this case study, accident costs, energy balance analysis, emissions calculations, and
acceleration effects are not included in the analysis.
4.2 Run Analysis

When Start is selected, the analysis commences and produces the output necessary for report
generation. Any serious errors or omissions in the input data will stop the analysis, and an
appropriate error message given. Otherwise, missing data (for example, certain costs
information) may be advised as a Warning message on screen.
A status bar at the bottom of the screen indicates progress and the current analysis stage.
When the analysis is finished, the dialogue stops scrolling and reports Analysis completed
successfully. The status message will read:
Analysis Stage: Completed - total analysis time = hh : mm : ss
Version 1.0
The user may then select Generate Reports to display the folders holding pre-defined report
options. The pre-defined report categories available with the current program release are:
! Deterioration/Work Effects
! Road User Effects
! Environmental Effects
! Cost Streams
The full list of standard reports is displayed under Project/Generate Reports.

Note that reports generally refer to data presented in Tabular form. Certain reports are also
available in graphical form (Charts). In addition, users may set up customised reports using a
report writer facility.
The results of the HDM-4 run for Case Study 1 are examined by reference to the reports and
(where available) the associated charts produced under the Generate Reports button:
! Deterioration/Works Effects
! Cost Streams
4.3 Deterioration and works effects

The Timing of Works report allows the user to check the works that would be implemented
in each analysis year for each project alternative. Two variants are available:
1 Timing of Works (by year)
For each project alternative, this report lists, by analysis year, the works description,
quantity and costs for each road section. A summary of total annual economic costs is
provided for each project alternative.
2 Timing of Works (by section)
This report lists similar details by section together with the summary of total annual
economic costs.
For this case study, the Timing of Works (by section) report is included in the Reports section
at the end of this chapter.
Considering the Without Project alternative, the report indicates that, at Section B001-01
(Town A-Town B), gravel resurfacing would be implemented in the years 2003, 2007, 2010,
2013 and 2016 (with associated preparatory spot regravelling). Spot regravelling would be
implemented every year except the first year and each year immediately following gravel
resurfacing. Grading is implemented every six months as scheduled. These works are
specified under the Maintenance Standard Gravel Road Maintenance described previously.
The With Project alternative for the same road section (B001-01) confirms that the upgrade
to paved standard would be implemented during 2004 and 2005. Before upgrading, the
existing road is graded at six monthly intervals between 2000 and 2003 with spot regravelling.
After upgrading, the paved road has no works specified until 2016, when patching is listed
each year between 2016 and 2019 (end of analysis period). These works are specified under
the three works standards:
! M Maintenance before upgrading
! I Pave Section B001-01 in 2004
! M Crack sealing and patching paved road
Version 1.0
The effect of these works on roughness is indicated in the Deterioration report and associated
chart showing roughness at the end of the year.
Case study 1: Plot of Roughness vs Time for 1 section (B001-01)
insert 1 page plot of roughness for each project alternative.
This plot shows the roughness progression during the analysis period for each project
alternative. Comparison with the Timing of Works (by section) report indicates that, before
upgrading, the regravelling works (listed as gravel resurfacing) have a significant impact on
roughness. Regravelling has been triggered when the roughness reaches a level of about 16 or
17 IRI. Note that intervention (regravelling) was requested if the thickness of gravel surfacing
was reduced to 50 mm.
After upgrading section B001-01, the roughness increases less dramatically. Patching is
triggered between 2016 to 2019, towards the end of the analysis period.
Note that the Improvement Standard specification included the option to define the effects of
the upgrading works in terms of roughness, mean rut depth, skid resistance and surface texture
(under the Effects Tab). For this case study, these values are derived by the HDM-4 Works
Effects model. An overview of the HDM-4 modelling logic is given in Chapter A1; full
details are given in Chapter D4 of the Analytical Framework and Model Descriptions.
Version 1.0
4.4 Cost streams

The economic analysis reports are accessed via Generate Reports/Cost Streams.
The Economic Analysis Summary (By Alternative) report gives a summary of costs,
discounted Net Present Value (NPV) and Internal Rate of Return (IRR) by project alternative.
Cost and NPV details are presented by road section in the Economic Analysis Summary (By
Section) report.
For this case study, the overall NPV is reported as 1.51 (millions of US Dollars). The
breakdown by section indicates that two sections give a positive NPV.
Section ID Section description Discounted NPV
(millions of US Dollars)
B001-01 Town A to Town B 0.62
B001-02 Town B to Town C - 0.57
B001-03 Town C to Town D 1.46
All sections 1.51
Note that the traffic levels are lowest for Section B001-02 (negative NPV), also that
maximum benefits are derived for Section B001-03 which has highest traffic flow. A revised
project comprising sections B001-01 and B001-03 only would give an overall NPV of 2.08
(millions of US Dollars).
Version 1.0

Study 2
1 Optimum Rehabilitation Standards for a Paved

Road
This case study presents the economic analysis of alternative rehabilitation standards for a 50
km long paved road. The existing road carries 8000 vehicles per day (AADT) and exhibits
significant levels of roughness and surface distress. This case study evaluates several possible
rehabilitation alternatives including overlay, mill and replace, and inlay.
The objective of the case study is to present the definition of section alternatives appropriate
to this type of analysis and demonstrate the HDM-4 deterioration model for paved roads. The
definition and timing of the Maintenance Road Works Standards are discussed, and the results
are examined.

The name of the case study is Case Study 2 Rehabilitation of paved roads (by section).
Double-click on the case study named Case Study 2a. Rehabilitation of paved roads (by
section)

The data can be reviewed under the HDM-4 Project tabs listed below:

❏ General (see Section 1.2.1)
❏ Select Sections (see Section 1.2.2)
❏ Select Vehicles (see Section 1.2.3)
❏ Define Normal Traffic (see Section 1.2.4)
Version 1.0

❏ Alternatives (see Section 1.2.5)
! Analyse Projects

❏ Setup Run (see Section 1.3.1)
❏ Run Analysis (see Section 1.3.2)
1.2.1 General
This Tab page confirms the project description, analysis type, analysis period and the pre-
defined Road Network and Vehicle Fleet.
Project.
The analysis period is defined by a start year 2000, and a duration 20 years, (that is, 2000 -
2019).
Version 1.0
This Tab page indicates that one section will be included in the analysis. By double-clicking
on the Section Description reveals the Definition/Geometry/Pavement/Condition Tabs
where the section details are held.
! Definition
Version 1.0
! Geometry
Version 1.0
! Pavement


Version 1.0

1995 (STAP) 25mm Surface Dressing New Surfacing

100 mm
Previous Surfacing
Subgrade (CBR 8%)
resets
Version 1.0
Table A2.1

AC GB CRS AMGB
HRA GM
PMA AB AB AMAB
RAC SB CS AMSB
AM CM LS
PA TNA AMAP
SMA AP FDA
xx
CAPE GB CRS STGB
DBSD GM
SBSD AB AB STAB
ST SL SB CS STSB
PM LS
xx AP TNA STAP
FDA
The abbreviations in Table A2.1 are described in Table A2.2.
Version 1.0
Table A2.2

Mix)
Asphalt
Concrete
SMA Stone Mastic
CAPE Cape Seal
Surface Dressing
Surface Dressing
SL Slurry Seal
Version 1.0
Table A2.3

activity


Treatment

STSB1

AMSB1
to intermediate
surface layer
to base
Source: NDLI (1995)
Notes:
N/A Not applicable.
and subgrade.
Version 1.0
Table A2.4


Widening
Inlay 11 per m2
Cape seal 14 per m2

Routine
Patching1 22 per m2
Pavement
and coefficients.
! Condition
Version 1.0


Version 1.0
1.2.5 Alternatives
Version 1.0

Version 1.0

carriageway area
carriageway area
Version 1.0
Table A2.5

from year
Patching (PATCH)
Patching (PATCH)
Patching (PATCH)
Version 1.0
Table A2.6

(IRI) 2
(m /km/yr)
Min Max Min Max
year
Patching R 2099 12.5 100 n/a n/a 0 100,000
Patching R 2099 12.5 100 n/a n/a 0 100,000
Patching R 2099 12.5 100 n/a n/a 0 100,000
Patching R 2099 12.5 100 n/a n/a 0 100,000
Notes:
Version 1.0

Patching
2 50 mm overlay
Edge repair
Patching
50 mm overlay
Edge repair
3 Patching
Crack Sealing
Patching
6 Inlay Edge repair
Crack Sealing
Version 1.0
❏ Works Item: Crack Sealing and Patching
❏ Works Item: Reconstruct at IRI 12 and 40% damage
The specification for Reconstruction is held under the Tabs:
Version 1.0
Version 1.0
Version 1.0
Version 1.0
below:
❏ Maintenance Works: Overlay 50 mm at IRI 6
The specification for the Overlay works can be reviewed under the Tabs:
specified.
Version 1.0
Version 1.0
Version 1.0
❏ Maintenance Works: Surface Dressing at SFC <= 0.4
The specification for Surface Dressing can be reviewed under the Tabs:
The Works Activity is specified as Surface Dressing Single (General Tab), with
responsive intervention based on the skid resistance (Intervention Tab). Note that a
total damaged area.
For example:
Version 1.0
Version 1.0
Version 1.0
The specification for Mill and Replace can be reviewed under the Tabs:
❏ Works Activity: Mill 75 mm + Replace 75 mm at 6 IRI

Version 1.0

- Cumulative ESAL
- Mean rut depth
- Potholing
- Ravelling
reviewed below:
Version 1.0
Version 1.0
Version 1.0
below:
Version 1.0
Version 1.0
Version 1.0
The specification for the Inlay works can be reviewed under the Tabs:
Version 1.0
❏ Maintenance Works: Inlay at Rut Depth 20 mm

The Works Activity Inlay is specified under the General Tab. The percentage of
the total carriageway surface area to be inlaid is specified under the Design Tab. A
under the Costs Tab.
Version 1.0
Version 1.0
1.3.1 Setup Run

Version 1.0
1.3.2 Run analysis



Version 1.0

2008 Reconstruction

Version 1.0


2008
2010
2011
2012

carriageway area.
1.4.2 Cost streams

Version 1.0

Costs discounted
(US$ millions)
mm

analysis year:

area)
Note:
2 are:
! Roughness (IRI)
defined by:
where:
Version 1.0

* Note :
ACRA = ACA + ACT ...(1.12)
where:


ACA = ACW + ACN ...(1.13)
where:

Version 1.0
Part B Project Analysis Case Studies

Study 3
1 Traffic Volume Capacity Improvements

This case study presents the economic analysis of widening a paved road. The existing road is
7 m wide, with an AADT of 15,000 in 1998. Non-motorised transport contributed an extra
400 vehicles in 1998, comprising pedestrians, animal carts, and bicycles. The analysis
assumes that routine pavement maintenance is undertaken on a condition responsive basis for
all alternatives. Three widening alternatives are considered, widening by 1m, widening by
3m, and adding two extra lanes. It is recognised that a 1m widening is an impractical
proposition; this alternative is included to observe the resulting economic indicators.
The objective of this case study is to demonstrate the congestion analysis of HDM-4, with
particular reference to specification of the input data and examination of the results.

The name of the case study is Case Study 3 Capacity improvements.
Double-click on the case study named Case Study 3. Capacity improvements

The case study run data is held in the Projects folder under Case Study 3.
The data can be reviewed under the HDM-4 Project Tabs listed below:

❏ General (see Section 1.2.1)
❏ Select Sections (see Section 1.2.2)
❏ Select Vehicles (see Section 1.2.3)
❏ Define Normal Traffic (see Section 1.2.4)
Version 1.0
The following Tab page may be displayed:

❏ Alternatives (see Section 1.2.5)
! Analyse Projects

1.2.1 General
This screen confirms the project description, analysis type, analysis period and the pre-defined
Road Network and Vehicle Fleet.
This case study is presented as a project analysis. The road under study is represented by one
section, 10 km long, and the different widening proposals represent project alternatives. The
analysis period is defined by the start year 2000 and duration 20 years (that is, 2000 - 2019).
Version 1.0
The screen confirms that only one road section will be included in the analysis. By double-
clicking on the section description reveals the Definition/Geometry/Pavement/Condition
Tabs which gives access to the section details.
The following details have particular relevance to this case study; they are concerned with the
effects of road widening.
! Definition
The speed-flow type of the existing road is specified as a two lane road. The existing
carriageway width is 7 m, with two traffic lanes. If lane addition is implemented
(alternative 4 in this case study), the width of each additional lane will be calculated as
Version 1.0
the existing carriageway width divided by the existing number of traffic lanes (that is, 7/2
= 3.5 m for this case study).
The speed-flow type defines the capacity of the existing road. The capacity may be
increased by Improvement Works such as partial widening (by up to 3 m) or lane
addition (by 1 - 9 lanes). The capacity of the road after improvement is defined by the
speed-flow type specified within the Improvement Standard. Default values of the
ultimate capacity for several speed-flow types are given in Table A3.1. Further details of
speed-flow model parameters are given in Chapter D5, and in Part B of the Analytical
Table A3.1
Ultimate capacity for selected speed-flow types
Speed-flow type Carriageway width Ultimate capacity Qult

(m) (PCSE/lane/hr)
Single lane road <4 600

Intermediate road 4 to 5.5 900
Two lane road 5.5 to 9 1400
Wide two lane road 9 to 12 1600
Four lane road >12 2000
The traffic flow (AADT) on the road section is confirmed as 15,000 in 1998. Details of
AADT, traffic composition and growth rate, as defined for this project case study, will be
reviewed under Define Normal Traffic.
Note also that for this road section there is some speed reduction due to roadside
activities or motorised/non-motorised transport. The speed reduction factors are
specified under Section/Details/Alignment.
Version 1.0
The speed reduction factors are defined as follows:

XNMT Speed reduction: effect of NMT on MT
1 = no reduction
0.6 = significant reduction
XMT Speed reduction: effect of MT on NMT
1 = no reduction
0.6 = significant reduction
Road side Speed reduction due to roadside activities (effect on MT only), where:
friction
1 = no roadside friction
0.6 = significant friction
! Pavement
The pavement type has been defined as Asphalt Mix on Granular Base (AMGB) under
the General Tab. The dates of Previous works listed under the Pavement Tab indicates
that the pavement type is unchanged since the original construction in 1988.
The Structural Number (SN), estimated as 3.3, represents the existing pavement layers
above the subgrade. The SNP value includes the subgrade (8% CBR) contribution. Note
that for the improvement works (partial widening or lane addition), a value of 3.3 has
also been specified for the Structural Number, SN, (under Improvement/Pavement).
Version 1.0
! Condition
This confirms the condition levels defined for 1998. For project appraisal, this data
would normally be collected as part of a detailed condition survey.
The vehicles selected from the pre-defined Northern Province fleet are shown on this screen.
Note that Non-Motorised Transport (NMT) is included in the analysis (NMT box ticked at top
of Select Vehicles Tab page). Individual vehicle attributes can be reviewed by double-
clicking on the appropriate vehicle type description.
Version 1.0

This confirms the volume of traffic using the selected road section in the given year. The
initial traffic composition and growth rates (by vehicle category and type) can be reviewed by
double-clicking the section description. The Normal Traffic details are given on two Tab
pages Motorised and NMT:
Version 1.0
1.2.5 Alternatives
The four different alternatives considered for this case study are defined below. The analysis
period is 20 years (from years 2000 to 2019).
Version 1.0

condition. In addition, a 50 mm overlay is applied when the
roughness level reaches 6 IRI OR when structural cracking affects
15% of the carriageway area.
period (2000-2001). The maintenance regime of Alternative 1
(Routine + 50 mm overlay), which is condition responsive, is effective
from year 3 (2002).
period (2000-2001). The condition responsive maintenance regime of
Alternative 1 is effective from year 3 (2002).
4 With this alternative, the existing road is widened by adding two lanes
during the period (2000-2002). The condition responsive maintenance
regime of Alternative 1 (Routine + 50 mm Overlay) is effective from
year 4 (2003).
The key data relating to the alternative widening scenarios are given in Table A3.2 and Table
A3.3.
Table A3.2
Summary of widening alternatives
Alternative Widening Duration Economic cost After Widening

of widening
(years) Carriageway Speed-flow
(US$/km) width relationship
(m)
1 None - - (7) (Two lane road)
2 +1 m 2 51,000 8 Two lane road
3 +3 m 2 102,000 10 Wide two lane
4 +2 lanes 3 238,000 14 Four lane road
Notes:
1 All data for Section ID A003-01, length 10 km
2 Alternative 1 represents the base case (7 m wide two-lane commuter road)
As noted previously, the speed-flow type determines the capacity of the road section. The
base case (Alternative 1) has been assigned the speed-flow type for a two-lane road. The
same speed-flow type has been assigned to Alternative 2. With Alternatives 3 and 4, the
capacity can be expected to increase due to the revised speed-flow types specified under the
respective Improvement Standards.
The various widening alternatives have an impact on the speed reduction due to NMT, MT,
and roadside friction. For the existing road section, these data are specified under
Version 1.0
Section/Details/Alignment. For the widening alternatives, the data are specified under
Improvement/Geometry. The values assigned for each alternative are given in Table A3.3.
Table A3.3
Speed limit and speed reduction factors
Alternative Widening Speed limit Speed reduction factors
km/h NMT MT Roadside

friction
1 None 100 0.7 0.9 0.7
2 +1 m 100 0.75 0.9 0.7
3 +3 m 100 0.9 1 0.8
4 +2 lanes 120 1 1 1
in Table A3.4. Each widening assignment is scheduled to start in year 1 of the analysis
period (year 2000), with a construction period of two or three years.
Each widening alternative includes a Maintenance Standard (Routine + 50 mm overlay)
providing condition responsive routine pavement works and condition responsive overlay,
effective from the year following completion of the widening works. This Maintenance
Standard includes the following works items.
1 Crack Sealing if Area of Wide Structural Cracking > = 5% Carriageway Area
2 Patching if Severely Damaged Area > = 5% Carriageway Area
3 50 mm overlay if Roughness > 6 IRI OR if Area of Structural Cracking >= 15%
carriageway area
These maintenance works define the base alternative.
Version 1.0
Table A3.4
Project Road Works Effective from Maintenance

alternative Standard year Works/Improvement Type
Alternative 1 M Routine + 50 mm 2000 Crack Sealing (CRKSL)
Overlay
Base case (R&OV50) Patching (PATCH)
Without widening Overlay 50 mm at 6 IRI (OVL50)
road
Overlay 50 mm at 15% cracking
(OV50)
Alternative 2 I Partial Widening 2000 Partial Widening

by 1 m
Widening by 1m (WIDE1m)
M Routine + 2002 Crack Sealing (CRKSL)

50 mm Overlay
(R&OV50) Patching (PATCH)
Overlay 50 mm at 6 IRI (OVL50)

(OV50)
Alternative 3 I Partial Widening 2000 Partial Widening

by 3 m
Widening by 3 m (WIDE3m)

50 mm Overlay

(OV50)
Alternative 4 I Lane addition 2000 Lane addition

(ADD2L)
Widening by 2 lanes
50 mm Overlay

(OV50)
This case study is concerned with the effects of road widening. Note that the base case road
geometry can be reviewed under Section/General, as discussed previously under Select
Sections. Relevant details of each alternative are discussed below:
Alternative 1: Base Case Without Widening

! M Routine + 50 mm overlay (Effective from year 2000)
This project alternative includes one Maintenance Standard Routine + 50 mm overlay which
is effective from the first year of the analysis period. As noted in Table A3.4, this
Maintenance Standard comprises three works activities (Crack Sealing, Patching, and
Overlay), each of which is condition responsive.
Version 1.0
Details of this Maintenance Standard can be reviewed by double-clicking the above

description under Maintenance Standards. The details are similar to those described for
Case Study 2, Alternative 2.
Details of each works activity can be reviewed by double-clicking the appropriate works
name. The details are similar to those described for Case Study 2, Alternative 2.
The intervention criteria can be reviewed via the Intervention Tab associated with each works
activity.
Version 1.0
Version 1.0
Version 1.0
Alternative 2: Widening by 1 m
! I Partial Widening by 1 m (Effective from year 2000)
Details of the Improvement Standard (partial widening) are given under seven Tabs:
General/Design/Intervention/Costs/Pavement/Geometry/Effects
Version 1.0
Note: ‘Re-surface existing carriageway’ is not selected
Version 1.0
Version 1.0
The improvement type (partial widening) and duration are specified under the General Tab.
The increase in width, together with the pavement type (of the partial widening) and speed-
flow type for the widened road is specified under Design. For this alternative the speed-flow
type after widening has been specified as a two-lane road (the same as the base case), as the
widening is only 1m. The pavement type (of the partial widening) is Asphalt Mix on Granular
Base (AMGB), which is the same as the existing pavement (Section/Pavement). Note that
the construction quality indicators (CDS and CDB) of the partial widening can be reviewed
via Design/Edit Construction Details. The main works costs and associated preparatory
costs are held under the Costs Tab, and are spread across the two-year construction period.
Version 1.0
Details of the surfacing used for widening (material and thickness) and the Structural Number
(SN) of the partial widening are specified under Pavement. For improvements, SN will
normally be assessed using design layer thicknesses and as new layer coefficients.
Alternative 3: Widening by 3 m
! I Partial Widening by 3 m (Effective from year 2000)
This alternative is similar to Alternative 2, except that the widening is 3 m, specified under
Improvement Standard/Design. The speed-flow type after widening is a wide two-lane road,
which has increased capacity over the original two-lane road (see Table A3.1). In addition,
speed reduction due to NMT is less significant (that is, higher factor in Table A3.3).
Alternative 4: Lane addition (2 lanes)

! I Lane addition (2 lanes) - Effective from year 2000
! M Routine + 50 mm overlay - Effective from year 2003
This alternative proposes the addition of two lanes, specified as Improvement Type Lane
Addition with a construction period of three years. The additional number of lanes is
specified as 2. As noted earlier, the assumed lane width is 3.5 m (based on an existing two-
lane carriageway width of 7 m), giving a carriageway width of 14 m after the addition of the
two lanes. (Exogenous benefits and costs have been included in this assignment, and may be
reviewed via Specify Alternatives/Edit Alternatives/Edit Assignment).
The General Tab page is shown below:
Version 1.0
Note that on the Design Tab page the speed-flow type after improvement has been specified
as four lane road, which has improved capacity over the two lane road specified for the base
case (see Table A3.1). In addition, there is no speed reduction due to NMT, MT, or roadside
friction.
The cost of the improvement works and associated preparatory works are specified on the
Costs Tab page.
Version 1.0
1.3.1 Setup Run

The Setup Run screen confirms that the base alternative for economic analysis is Alternative
1, and that the discount rate is 6%.
1.3.2 Run Analysis

This starts the analysis and produces the output necessary for report generation.

The output reports from the HDM-4 analysis are produced under Generate Reports/Select
Reports.

The Timing of Works (by section) report lists, by alternative, the works that would be
implemented by the specified works standards in each analysis year. The report confirms the
dates of the improvement works for alternatives 2, 3 and 4.
Alternative 1 2 3 4
Widening None 1m 3m 2 lanes
Widening Works date - 2000 - 2001 2000 - 2001 2000 - 2002
Baseline IRI (year) 6.14 (2000) 5.98 (2002) 5.38 (2002) 4.64 (2003)
Version 1.0
Note that with Alternatives 2-4, the roughness value immediately following the widening
works is based on the full carriageway width, and hence the baseline IRI value decreases as
the widening (with no defects immediately after construction) increases.
The Timing of Works report should be read in conjunction with the Deterioration Summary
report, which indicates the progression of certain defects throughout the analysis period.
1.4.2 Road user effects

The impact of the widening alternatives can be assessed by examination of the Volume-
Capacity Ratio report which tabulates the volume-capacity ratio (VCR) by time period and
calendar year for each project alternative and road section.
The long-term effectiveness of the road widening is indicated by the time taken for the VCR
to reach a value of 1.0 (that is, when volume equals capacity).
The effect of widening on vehicle speeds is demonstrated by the Vehicle Speed report.
1.4.3 Cost streams

The Economic Analysis Summary indicates that, for the range of widening represented by
alternatives 2-4, the NPV increases significantly with width of carriageway.
Carriageway
width after NPV
Alternative Widening widening
(millions of US$)
(m)
2 1m 8
3 3m 10
4 2 lanes 14
Version 1.0
CONCEPTS OF ANALYSIS
Part A Concepts of Analysis

Study 4
1 New Bypass Construction (traffic diversion

effects)
This case study presents the economic analysis of a project to construct a bypass around a
town centre. The objective is to demonstrate the specification of the bypass (as a section
alternative within a project alternative), and to examine the resulting traffic diversion effects.
The roads under study are shown schematically in Figure A4.1. Road sections A, B, C and D
represent the network at a town centre. The network shown, models traffic movements
between Zone 1 and Zone 2. The proposed project is the construction of a bypass, represented
by one section (Section E), 10 km long.
Zone centroid 1
A
C D
B
E
Key: Zone centroid 2
Existing road sections
Proposed bypass (Section E)
Figure A4.1 Case Study 4: Construction of a new bypass
The case study considers four project alternatives as defined below:

Alternative 1: Base case without the bypass
Alternative 2: Bypass width 8m, constructed between 2000-2001, opening in 2002.
The traffic flows (AADT) on each section after bypass construction will normally be derived
using an external traffic demand model (that is, derived outside HDM-4). The AADT data
used with this case study are summarised in Table A4.1. The vehicle composition data (Table
Version 1.0
CONCEPTS OF ANALYSIS APPENDIX A4 - PROJECT ANALYSIS CASE STUDY 4
A4.2) predict a transfer of heavy vehicles from Sections A, C and D to Section E once the
bypass (Section E) is constructed.
Table A4.1
AADT values by project alternative
Alternative 1 Alternative 2 Alternative 3 Alternative 4

Base case Without With Without With Without With
Road Section bypass bypass bypass bypass bypass bypass
Without 2002 2002 2003 2003 2004 2004
bypass 1998
A 5600 6062 2062 6183 2103 6306 2144
B 3200 3464 7464 3533 7613 3604 7766
C 3200 3464 464 3533 473 3604 483

D 8800 9526 2526 9716 2576 9910 2627
E (bypass) n/a n/a 7000 n/a 7140 n/a 7283
Notes:
1 Bypass opening year is 2002 for Alternative 2, 2003 for Alternative 3, and 2004 for
Alternative 4.
2 Data for Sections A-D may be reviewed via Specify Alternatives/Edit
Alternative/Diverted Traffic.
3 Data for Section E may be reviewed via Specify Alternatives/Edit Alternative/Edit New
Section.
4 Shaded data included for information, and is not entered to HDM-4.
5 n/a = not applicable
Version 1.0
Table A4.2
Traffic composition and growth rates
Alternative 1 Alternatives 2-4: With bypass (opening year)
Without bypass:
Sections A,C,D Sections B,E
Vehicle type Sections A-D (1998)
Compo- Annual Compo- Annual Compo- Annual
sition growth sition growth sition growth
(%) rate (%) (%) rate (%) (%) rate (%)
Medium bus 5 2 5 2 5 2
Heavy truck 5 2 3 2 10 2
Medium truck 10 2 7 2 14 2
Light goods vehicle 10 2 10 2 10 2
Medium car 60 2 70 2 50 2
Mini-bus 5 2 5 2 5 2
Artic truck 5 2 0 0 6 2


The name of the case study is Project Case Study 4: Construction of new bypass.
Double-click on the case study named Case Study 4. Construction of a new by-pass

The case study input data can be reviewed under the HDM-4 Projects work flow buttons and
associated Tab pages/screens as follows:
! Define Project Details (see Section 1.2.1)
The following Tab pages are reviewed:

❏ General
❏ Select Sections
Version 1.0
❏ Select Vehicles
❏ Define Normal Traffic
! Specify Alternatives (see Section 1.2.2)
The following screens are reviewed:

❏ Alternatives
Alternative Details
New Construction Section Option
Normal Traffic
Section Details (Definition/Geometry/Pavement/Condition)
Diverted Traffic
Diverted Traffic Details
! Analyse Projects
The following Tab pages are reviewed:

1.2.1 Define Project Details

The project details are held under four Tabs:
General/Select Sections/Select Vehicles/Define Normal Traffic
These screens are shown on the following pages.

! General
The General screen confirms the project description, analysis type, analysis period
(2000-2019) and the pre-defined Road Network and Vehicle Fleet. Note that the
Version 1.0
construction of a new road section can only be modelled as a section alternative within a
project alternative (analysis by project).
! Select Sections
The four road sections (Sections A-D) included in the analysis are shown in the Select
Sections screen. Section details can be accessed by double-clicking the section
description. The section representing the bypass (Section E) is not included. Section E
represents a proposed new section and is defined under specific project alternatives (see
later). Note that, even after the analysis, Section E is not automatically added to the
Road Network folder, as the section does not actually exist; however the user may add it
to the appropriate Network after construction.
! Select Vehicles
The vehicle types that may be assigned to the sections (including the new section) are
given under Select Vehicles.
Version 1.0
! Define Normal Traffic
Traffic data (AADT, vehicle composition and growth rates) for Sections A, B, C and D are
defined under Define Normal Traffic. The vehicle composition and growth rates for a
particular section can be reviewed by double-clicking the appropriate row on the above Tab
page. Note that traffic data for the proposed bypass (Section E) are assigned separately under
the appropriate Project Alternatives.
1.2.2 Specify Alternatives

! Alternatives
The four project alternatives considered for this case study are displayed above.
Alternative 1 represents the base case, that is, existing road sections A-D without the bypass.
Alternatives 2, 3 and 4 include the bypass (Section E), with carriageway width and pavement
Version 1.0
construction as defined in Table A4.3. The screen above indicates the maintenance standard
assigned to Section A under Alternative 1 (base case).
Table A4.3
Details of proposed bypass by project alternative
Section details Project alternative

2 3 4
Length (m) 10 10 10
Pavement type AMGB AMGB AMGB
Construction start year 2000 2000 2000
Works duration (years) 2 3 4
Opening year 2002 2003 2004
Total economic cost (millions US Dollars) 2.735 3.418 4.786
Total financial cost (millions US Dollars) 3.2 4.0 5.6
Speed flow type 2-lane wide 2-lane 4-lane
Traffic flow pattern Commuter Commuter Commuter
The maintenance standards associated with each project alternative are summarised in Table
A4.4. For each section, the same maintenance standard has been assigned to each constituent
section alternative. This allows the effect of different bypass carriageway widths to be
assessed directly.
All alternatives include condition responsive routine maintenance (crack sealing and
patching), as well as a condition responsive overlay.
Version 1.0
Table A4.4
Summary of maintenance standards by project alternative
Road section Project alternatives

1 (base case) 2 3 4
Section A RMA + OVLA RMA + OVLA RMA + OVLA RMA + OVLA
Section B RMB + OVLB RMB + OVLB RMB + OVLB RMB + OVLB
Section C RMC + OVLC RMC + OVLC RMC + OVLC RMC + OVLC
Section D RMD + OVLD RMD + OVLD RMD + OVLD RMD + OVLD
AMGB (2-lane) AMGB (wide 2- AMGB (4-lane)
Section E lane)
(New section) RM + OVL RM + OVL
RM + OVL
Notes:
1 RM = condition responsive Routine Pavement Maintenance: includes patching and crack
sealing (section denoted by subscript).
2 OVL = condition responsive 50mm overlay (section denoted by subscript).
3 New section AMGB = pavement type (Asphalt Mix on Granular Base).
4 ( 2-lane) = speed-flow type given in parentheses.
in Table A4.5. Each alternative includes a Maintenance Standard (Routine + 50 mm overlay)
providing condition responsive routine works and condition responsive overlay. This is
effective from the year 2000 for the base case. The ‘effective from’ date for Section E is the
year of opening of the bypass (applies Alternatives 2, 3 and 4).
Note that, for Section E, the ‘effective from’ date could also have been specified as the year
2000, giving the same resulting works. This is because the maintenance standard would not
be applied during the construction period. (Based on the works ranking hierarchy, new section
construction supersedes works associated with the maintenance standard).
The Maintenance Standard includes the following works:
! Crack Sealing if Area of Wide Structural Cracking > = 5% Carriageway Area.
! Patching if Severely Damaged Area > = 5% Carriageway Area.
! 50 mm overlay if Roughness > 6 IRI.
! 50 mm overlay if ‘All Structural Cracking’ > = 15% Carriageway Area.
Version 1.0
Table A4.5
Project alternative Road Works Standard "Effective Maintenance Works/Improvement Type

from" year
Alternative 1 M Routine + 50mm Overlay (R&OV50) Crack Sealing (CRKSL)
Base Case - without bypass Sections A, B, C and D . . . . . . . . . . . . . 2000 Patching (PATCH)
Overlay 50 mm at 15% cracking (OV50)
Bypass completed in 2001 Sections A,B,C and D . . . . . . . . . . . . . . 2000 Patching (PATCH)

Opening year 2002 Section E (construction 2000-2001) . . . 2002 Overlay 50 mm at 6 IRI (OVL50)


Version 1.0
Specification of bypass details

The bypass is modelled as a section alternative within a project alternative. Before the bypass
is created, the Alternatives screen shows only Sections A-D. The bypass (Section E) was
created by selecting the appropriate Alternative, then Edit Alternative/Add New Section.
! Alternatives
Clicking on the Edit Alternative button gives:

! Alternative Details
This shows the maintenance standard assigned to Section E under Alternative 2. The
assignments for a particular section can be reviewed by selecting that section. The pre-
defined maintenance standards were assigned via the Assign Maintenance button.
Version 1.0
! New Construction Section Option
The following screen is displayed by clicking the Edit New Section button on the
Alternative Details screen (see previous page).
This gives Section E details that have been specified for Alternative 2:
❏ Details of Bypass
General Name of new section (Section E).
Start year of construction of Section E is 2000.
Costs Economic and financial costs specified as shown (US Dollars per km).
Duration of construction is 2 years.
Allocation of construction cost within construction period is 50% each
year.
Version 1.0
! Normal Traffic
The following screen is displayed by clicking the Normal Traffic button on the New
Construction Section Option screen (see previous page).
This screen defines the AADT and vehicle composition using the bypass in its opening year
(2002 for Alternative 2). The traffic data for this case study are summarised in Table A4.1
and Table A4.2.
! Section Details
The following screen is displayed by clicking the Section Details button on the New
Construction Section Option screen.
Version 1.0
The Tab pages Definition/Geometry/Pavement/Condition are used to specify the

bypass characteristics (by Alternative). As with all section data specification, care is
needed to ensure consistency of data on the different Tab pages. For this case study:
❏ Surfacing material type (Pavement Tab) should be consistent with Pavement Type
(Definition Tab).
❏ AADT/year (Definition Tab) refers to bypass opening year.
❏ Dates of previous works (Pavement Tab) refer to last year of bypass construction
period.
❏ Date of condition (Condition Tab) refers to last year of bypass construction period.
! Diverted Traffic
Once Section E has been defined for a specific project alternative via the New
Construction Section Option screen, it is listed in the Alternative Details screen.
The top button on this screen allows the user to specify Diverted Traffic. In the context
of this case study, "Diverted traffic" refers to traffic using sections A,B C and D after the
opening of the bypass.
Version 1.0
The Diverted Traffic screen indicates the initial AADT (1998) on sections A-D, and the
new "diverted" AADT values in 2002 (the bypass opening year under Alternative 2).
The vehicle composition and growth rates for a particular section (in 2002) may be
reviewed by double-clicking the relevant cell under the "New AADT in 2002" column or
by selecting the appropriate cell and clicking the Edit Diversion Details button.
! Diverted Traffic Details
The screen below shows the vehicle composition and growth rates (in 2002) for Section
A under Alternative 2.
Version 1.0
1.3.1 Setup Run

The Setup Run Tab page is accessed via the Analyse Projects button. For this case study the
base alternative for economic analysis is Alternative 1 (without the bypass) and the discount
rate is 6%.
1.3.2 Run Analysis

The HDM-4 analysis commences when ‘start’ is selected, and produces the output necessary
for report generation.
1.3.3 Deterioration/Works Effects

The Timing of Works (by section) report can be used to check the works that would be
implemented under each project alternative. Note that for alternatives 2-4, works are listed for
the bypass (new section: Section E). The progression of defects on the existing roads
(Sections A-D) and the new bypass (Section E) is given on the Deterioration Summary
Report. This can be used to check when condition responsive works should be triggered.
1.3.4 Road User Effects

The reports produced under Road User Effects can be used to check the impact of the new
bypass on vehicle speeds and volume capacity ratio (VCR) on the sub-network under study.
The traffic flows on each section (with and without the bypass), derived externally, have been
defined in Table A4.1. The construction of the bypass is expected to reduce the daily volume
of traffic on sections A, C and D.
1.3.5 Cost Streams

The economic analysis (see Benefit Cost Ratio report) produces the economic indicators for
each project alternative. The results for this case study indicate that Alternative 2 (bypass
with 8m carriageway) is the most cost-effective, having the highest values of NPV,
Benefit/Cost ratio, and IRR.
Version 1.0
Part A Appendices
Contents
Volume 2 - Applications Guide: Appendix B -

Programme Analysis Case Studies
B1 – Programme Analysis Case Study 1:

1: Western Province rehabilitation programme – Life Cycle Analysis Method 1
1.1 Introduction 1
1.2 Overview of the case study data 1
1.3 Input data 2
1.4 Define Programme Details 3
1.5 Specify Standards Assignments 5
1.6 Generate Programme 8
1.7 Perform Budget Optimisation 8
B2 – Programme Analysis Case Study 2:

1 Eastern Province rehabilitation programme – Multi-Year Forward Programme1
1.1 Introduction 1
1.2 Overview of the case study data 2
1.3 Multi-year forward programme data 3
1.4 Generate Programme 8
1.5 Perform Budget Optimisation 9
Version 1.0
Appendix B1 - Programme Analysis

Case Study 1
1 Western Province rehabilitation programme –

Life Cycle Analysis Method
1.1 Introduction
condition survey of the road network and the application of pavement condition trigger levels
or thresholds to identify those sections in need of some form of rehabilitation.
steps:
1.2 Overview of the case study data

improvement.
network.
! Maintenance and/or improvement standards to be assigned to each candidate road
section.
Version 1.0
CONCEPTS OF ANALYSIS APPENDIX B1 - PROGRAMME ANALYSIS CASE STUDY 1
Rehabilitation (Life Cycle Analysis) study as illustrated in Figure B1.1.
Figure B1.1 Workspace screen
1.3 Input data

The data used for this Case Study has already been created and therefore the procedure
described below does not include the creation of a new study.
The procedure involves defining details of the programme analysis to be carried out,
assignment of work standards, generation of the work programme, optimisation under budget
constraints, and finally generation of reports. This procedure should normally be followed
sequentially, although iteration between the above tasks can be done in order to review
previously specified data.
The data is arranged in groups (corresponding to the tasks) and, within these, the data are
presented in different Tabs as summarised below:
Define Programme Details

! General Tab
! Select Sections Tab
Version 1.0
! Select Vehicles Tab
! Define Traffic Tab
Assign Standards
! Specify Alternatives Tab
Generate Programme
! Perform Run Tab
! Unconstrained Programme Tab
Perform Budget Optimisation

! Define Budget Tab
! Optimised Programme Tab
Generate Reports
! Reports Tree
1.4 Define Programme Details
1.4.1 General
The General Tab can be used to define the study description, analysis type, analysis period
(start year and duration), the currency conversion rates and the discount rate to be used in the
economic analysis. This Tab also shows the selected Road Network and Vehicle Fleet.
Version 1.0

It is assumed that the long list of candidate road sections from Western Province was drawn
up following some pavement condition survey. The list would normally contain only those
road sections deemed to require some form of periodic maintenance or rehabilitation during
the next budget period. Consequently, all road sections have been selected for the analysis.
However, if the road network had contained details of all road sections in Western Province, it
is possible to select only those that should be included in the work programme for the next
budget period.
long budget period.
It is recommended that work programmes should be prepared for short term budget periods,
for example five years or less.
button).

A typical analysis using HDM-4 will involve the calculation of road user costs, primarily in
the form of vehicle operating costs, and travel time costs.
This Tab can be used to select which vehicle types will be included in the analyses, from
amongst those specified in the Vehicle Fleet database.
It is also possible to review or amend details of the vehicle types from within this Tab.

The AADT observed on each of the road sections will have been specified together with
details of the road sections. This Tab specified the traffic composition and annual growth rates
that are applicable on each road section for each vehicle type in this analysis. Note that several
traffic growth periods may be specified, for example, 5% annual growth between 2000 –
2004, followed by 3.5% annual growth from 2005 onwards. In addition, the traffic details can
be copied from one road section to others, thereby permitting each road section to have a
different road set of traffic growth characteristics, if required.
Version 1.0
1.5 Specify Standards Assignments
1.5.1 Alternatives
minimum) alternative against a project case (do something) alternative. Details of at least
two alternatives can be specified in this Tab.
includes routine maintenance, resealing, overlays and reconstruction for bituminous
pavements. Details of the maintenance standards are summarised in Table B1.1.
1.5.2 Maintenance Standards

The P&CS, resealing, overlay and reconstruction standard, adopted for bituminous
sections, includes the five maintenance works items described in Table B1.1. The standard has
been assigned the short code: Brehab for “bituminous rehabilitation”.
Version 1.0
Table B1.1
Details of Maintenance Standard (Brehab) assigned to bituminous road
sections

Maintenance
surface area

Note: * Ranking of road works from Chapter D2 of Analytical Framework and Model
Descriptions
If more than one of these activities are triggered in a given year, the works item that is highest
in the hierarchy of road works (that is, with lowest ranking value) will be applied.
ranking, and therefore both can be applied in the same year. (The hierarchy is defined by the
ranking of road works described in Chapter D2 of Analytical Framework and Model
Descriptions).
Version 1.0
Table B1.2
Class from year
Crack Sealing CRKSL
Reseal RSL30
Overlay OV5IRI
Reconstruct REC12
Notes:
Version 1.0
1.6 Generate Programme
1.6.1 Perform Run
For life-cycle analysis, the Base alternative for economic analysis is specified in the Run
Setup screen. This can be displayed by selecting the Run Setup button within the Perform
Run Tab.
1.6.2 Work Programme
produces an unconstrained work programme. This contains, for each road section, the
alternative with the highest NPV (refer to Chapter B2). This represents an unconstrained
programme, with the total financial cost given in the cumulative cost column.
million.
1.7 Perform Budget Optimisation
1.7.1 Define Budget

Version 1.0
button. For this case study, optimisation is by incremental benefit/cost ranking (see Chapter
B2). The parameters to be specified are given below:
Efficiency frontier Zone = 75%
1.7.2 Optimised Work Programme

year) in accordance with the incremental benefit/cost ranking index described in Chapter B2.
Version 1.0
1.7.3 Generate Reports

the optimised work programme report has been reproduced for this document (see Figure
B1.2).
Version 1.0

Version 1.0
Appendix B2 - Programme Analysis

Case Study 2
1 Eastern Province rehabilitation programme –

Multi-Year Forward Programme
1.1 Introduction
rehabilitation programme for a road network in the Eastern Province of a country. The study
demonstrates the Multi-Year Forward Programme method described in Chapter B2.
The objective is again to prepare a prioritised list road projects from a selected group of road
condition survey of the road network and the application of pavement condition trigger levels
or thresholds to identify those sections in need of some form of rehabilitation.
This Case Study demonstrates the application of the Multi-year Forward Programme method
for the production of a rehabilitation programme for a road network. The only difference with
the life cycle analysis data set is the way that the maintenance standards are assigned to road
sections. The multi-year forward programme method requires one maintenance standard to be
assigned to each road section, while life-cycle analysis requires at least two standards per
section to permit economic analysis. The multi-year forward programme method assigns a
default do minimum alternative which corresponds to deferring capital road works (periodic
maintenance and improvements) to the first year after the budget period.
To demonstrate the multi-year forward programme method, the Case Study is presented in the
following steps:
Step 1 : Overview of the case study data.
Step 2 : Review of the case study data.
Step 3 : Run multi-year forward programme analysis
Step 4: Examine the analysis results
Version 1.0
1.2 Overview of the case study data

Eastern Province Rehabilitation (Forward programme)
When a Programme Analysis study is created for the Multi-year forward programme method,
the following data sets must be defined:
improvement.
network.
! Maintenance and/or improvement standards to be assigned to each candidate road
section.
! Maintenance and Improvement standards manager
To review the case study data, open the Programmes folder and select the Eastern Province
Rehabilitation (Forward Programme) study as illustrated in Figure B2.1.
Figure B2.1 Workspace screen
Version 1.0
The data used for this Case Study has already been created and therefore the procedure
described below does not include the creation of a new study.
The procedure is similar to that of the Life Cycle method and involves defining details of the
programme analysis to be carried out, assignment of work standards, generation of the work
programme, optimisation under budget constraints, and finally generation of reports. This
procedure should normally be followed sequentially, although iteration between the above
tasks can be done in order to review previously specified data.
The data is arranged in groups (corresponding to the tasks) and, within these, the data are
presented in different Tabs as summarised below:
1.2.1 Define Programme Details

! General
! Select Sections
! Select Vehicles
! Define Traffic
1.2.2 Assign Standards

! Work Programming
1.2.3 Generate Programme

! Perform Run
! Unconstrained Programme
1.2.4 Perform Budget Optimisation

! Define Budget
! Optimised Programme
1.2.5 Generate Reports

! Reports Tree
1.3 Multi-year forward programme data

The Case Study run data is held in the Programmes folder under:
Eastern Province Rehabilitation (Forward programme)
The data can be reviewed under the HDM-4 Programme Analysis Tabs listed below.
1.3.1 Define Programme Details

General
The General Tab can be used to define the study description, analysis type, analysis duration,
the currency conversion rates and the discount rate to be used in the economic analysis. This
Tab also shows the selected Road Network and Vehicle Fleet.
The analysis will be performed using the Multi-Year Forward Planning option based on a
three-year budget period from 2000-2002.
Version 1.0
Select Sections
The road network under study is represented by 21 sections (one section with unsealed surface
class and all others with bituminous surfacing).
It is assumed that the long list of candidate road sections from Eastern Province was drawn up
sections deemed to require some form of periodic maintenance or rehabilitation during the
next budget period. Consequently, all road sections have been selected for the analysis.
However, if the road network had contained details of all road sections in Eastern Province, it
is possible to select only those that should be included in the work programme for the next
budget period. Alternatively, all road sections in the database could be analysed, particularly if
this involves a long budget period. The selected sections define a sub-network for which a
prioritised programme is required.
It is recommended that work programmes should be prepared for short-term budget periods,
for example less than five years.
button).
Select Vehicles
The vehicles selected from the pre-defined Provincial Vehicle Characteristics fleet are
indicated below. Individual vehicle attributes may be reviewed by double-clicking on the
appropriate vehicle type description (or by selecting the section and clicking the Edit Vehicle
button).
Version 1.0
Define Traffic
This Tab defines the traffic composition and growth rates on the selected road sections. The
initial traffic composition and growth rates (by vehicle type) can be reviewed by double-
clicking the section description (or by selecting the section and clicking the Section Traffic
Details button).
This Tab specified the traffic composition and annual growth rates that are applicable on each
road section for each vehicle type in this analysis. Note that the traffic details can be copied
from one road section to others, thereby permitting each road section to have a different road
set of traffic growth characteristics, if required.
1.3.2 Assign Standards
Work programming
For Multi-Year Forward Programme, one maintenance standard (representing the
rehabilitation alternative) is assigned to each road section, or, if an improvement is proposed,
this is assigned together with associated maintenance standards before and after the
improvement is applied. For Multi-Year Forward Programme, the base alternative is
automatically assigned as the works needed in the first year following the defined budget
period (with no works during the budget period).
In this Case Study, one maintenance policy has been assigned to all paved road sections and
one to unsealed sections based on the Surface Class (that is, bituminous or unsealed) as
follows:
Bituminous sections: M P&CS, Reseal, Overlay & Reconstruct (short code Brehab)
where P&CS = Patching and Crack Sealing
Unsealed section: M Grade, Spot regravel & Resurface (short code Grehab)
Each maintenance standard comprises a set of condition responsive works activities and
establishes a maintenance policy for the road sections to which the standard is assigned. For
this Case Study, one standard has been set up for each surface class. In practice this may be
further refined by assigning different maintenance standards based on, for example, road class
or traffic level, both of which indicate the relative importance of the section..
Details of the works activities and condition responsive intervention criteria associated with
each maintenance standard are summarised in Table B2.1 and Table B2.2. The details may be
reviewed under Maintenance Standards in HDM-4 Workspace shown in Figure B2.1.
Version 1.0
Paved Road Maintenance Standard

This standard (P&CS, Reseal, Overlay & Reconstruct - Brehab), adopted for bituminous
sections, includes five maintenance works items defined in Table B2.1 below:
Version 1.0
Table B2.1
Details of Maintenance Standard assigned to bituminous road sections
short code: Brehab

Pavement
Resurfacing Surface Dressing Single Total damaged area >= 30% 16
Rehabilitation Overlay dense graded asphalt Roughness >= 5 IRI AND 9

Cracking >= 5%
Reconstruction Pavement Reconstruction Roughness >= 12 IRI 6
Note: * Ranking of road works from Chapter D2 Analytical Framework and Model
Descriptions.
If more than one of these activities are triggered in a given year, the works item that is highest
be applied in the same maintenance year. (The hierarchy is defined by the ranking of road
works in Chapter D2 of Analytical Framework and Model Descriptions - Modelling Road
Works Effects).
Unsealed Maintenance Standard

The Works Activities that make up this standard for unsealed road sections (Grade, Spot
Regravel & Resurface - Grehab) are as follows.
Table B2.2
Details of Maintenance Standard assigned to unsealed road sections
short code: Grehab

Grading Roughness >= 8 IRI 7
Routine Pavement
Spot regravelling Gravel thickness <= 100 mm 7
Resurfacing Regravelling Gravel thickness <= 50 mm 6
Note: * Ranking of road works from Chapter D2 of Analytical Framework and Model
Descriptions
The ranking order assigned to these road work items implies that Resurfacing has the highest
priority. Spot regravelling and Grading have equal ranking, and therefore both of these can be
performed in a given analysis year.
Details of the maintenance works items can be reviewed under Maintenance Standards in
HDM-4 Workspace shown in Figure B2.1.
Version 1.0
1.4 Generate Programme
Perform Run
For Multi-Year Forward Programme Analysis, the user can choose to perform the run either
with or without economic analysis. Economic analysis is required before budget optimisation.
This can be selected using the Run Setup button.
The Run Setup button also allows the user to select Energy Balance and Emissions models,
to model acceleration effects, and to include accident costs in the economic analysis.
When Run Setup is completed, the Work Programme (prior to budget optimisation) is
generated by pressing the Start button. Once completed, the user is strongly advised to Save
the generated programme, especially before viewing any reports.
Version 1.0
Unconstrained Programme
The generated Work Programme can be viewed on screen prior to performing a budget
optimisation.
1.5 Perform Budget Optimisation
Define Budget
sections from Eastern Province would require US$ 80 million in over the three-year period
from 2000 to 2002 for periodic maintenance, it is now possible to select the optimal set of
road sections if less than 100% of the required budget is available. Assume that the amount of
money available for the three-year period has been set at US$ 60 million.
Optimised Programme
year) in accordance with the incremental benefit/cost ranking index described in Chapter B2.
Version 1.0
Table B2.3 Details of alternatives for life-cycle analysis
Alternative Surface Class Maintenance Standard Effective Maintenance Works

from year
Pothole Patching (PATPOT)
Bituminous Patching and Crack Sealing (B/rout) 1999
Base
Grading (GRADE)
Unsealed Grade & Spot Regravel (G/rout) 1999
Spot Regravel (SPRG)
Pothole Patching (PATPOT) *
Crack Sealing (CRKSL) *
Bituminous P&CS, Reseal, Overlay & Reconstruct (Brehab) 1999 Reseal (RSL30)
Overlay (OV5IRI)
Rehabilitation
Reconstruct (REC12)
Grading (GRADE) *
Unsealed Grade, Spot Regravel & Resurface (Grehab) 1999 Spot Regravel (SPRG) *
Resurface (RESURF)
Notes:
1 Details of intervention criteria for works activities are given in Table B2.1 and Table B2.2.
2 Routine pavement works in rehabilitation alternative (identified by *) are identical to those in base alternative.
Version 1.0
Study Name: 2. Eastern Province (Forward Programme)

Year Section Road Class Length AADT Surface Class Work Description NPV/C Financial Cumulative
(km) Costs Costs
2000 Gaborone-Lobatse Section 2 Primary or Trunk 34.0 1334 Bituminous Resealing SBSD 128.46 1.37 1.37
2000 Gaborone-Pilane Primary or Trunk 30.0 1516 Bituminous Resealing SBSD 4.06 1.21 2.57
2000 Pilane-Rasesa Secondary or Main 7.0 459 Bituminous Resealing SBSD 3.38 0.28 2.85
2000 Gaborone-Lobatse Section 1 Primary or Trunk 60.0 1814 Bituminous Resealing SBSD 3.34 2.41 5.27
2000 Serule-Francis Town Primary or Trunk 98.0 1119 Bituminous Resealing SBSD 3.08 3.94 9.21
2000 Mosetse-Nata Secondary or Main 68.0 177 Bituminous Resealing SBSD 3.00 2.73 11.94
2000 Gaborone-Tlokweng Primary or Trunk 10.0 2586 Bituminous Overlay 50mm 1.63 1.21 13.15
2000 Sebina - Mosetse Secondary or Main 79.0 202 Bituminous Resealing SBSD 1.53 3.18 16.32
2000 Francis Town - Selebi Phikwe Primary or Trunk 85.0 1226 Bituminous Resealing SBSD 1.33 3.42 19.74
2000 Palapye-Serule Primary or Trunk 70.0 963 Bituminous Overlay 50mm 0.77 8.44 28.18
2000 Lobatse-Jwaneng Section 1 Primary or Trunk 120.0 812 Bituminous Overlay 50mm 0.38 15.55 43.73
2000 Jwaneng-Sekoma Section 1 Secondary or Main 80.0 607 Bituminous Overlay 50mm 0.34 9.65 53.38
2000 Monametsana-Mahalapye Secondary or Main 142.0 446 Bituminous Overlay 50mm 0.33 17.13 70.51
2000 Gravel Road Tertiary or Local 10.0 114 Unsealed Resurface 0.18 0.29 70.80
2000 Sekoma-Morwamosu Secondary or Main 79.0 313 Bituminous Overlay 50mm 0.17 8.53 79.33
2002 Rasesa-Monametsana Secondary or Main 15.0 270 Bituminous Resealing SBSD 0.22 0.60 79.93
Figure B2.2 HDM-4 Unconstrained Work Programme
Version 1.0
Part A Appendices
Contents
Volume 2 - Applications Guide: Appendix C -

Strategy Analysis Case Studies
C1 – Strategy Analysis Case Study 1:

1 Long Term Budget Forecasts and Performance Trends 1
1.3 Funding requirements without budget constraint 22
1.4 Effect of budget constraints 22
Version 1.0
Part C Concepts of Analysis
Appendix C1 - Strategic Analysis Case

Study 1
1 Long Term Budget Forecasts and Performance

Trends
section.
The following steps are followed:
1 Locate the case study data (see Section 1.1)
2 Review the case study data (see Section 1.2)
3 Run strategy analysis (without budget constraint) (see Section 1.3)
4 Perform budget optimisation (see Section 1.4)

trends.
! Open the Strategies folder
Double-click on the case study named:
1: Long term budget forecasts and performance trends.

The data can be reviewed under the HDM-4 Strategies work flow buttons and associated Tabs
listed below:
! Define Strategy Details (see Section 1.2.1)
The following screens may be displayed:

❏ General
❏ Select Sections
❏ Select Vehicles
Applications Guide Version 1.0 1

CONCEPTS OF ANALYSIS APPENDIX C1 - STRATEGY ANALYSIS CASE STUDY 1
❏ Define Normal Traffic

! Specify Standards Assignments (see Section 1.2.2)
The following screen may be displayed:

❏ Alternatives
1.2.1 Define Strategy Details
General
This screen confirms the study description, optimisation method, analysis period (start year
and duration) and the names of the pre-defined Road Network and Vehicle Fleet. For this
case study the selected optimisation method is maximisation of benefits (NPV).
Select Sections
The list of representative sections is displayed under the Select Sections Tab.

The road network under study has been modelled by eleven representative sections. The
representative sections for bituminous (paved) roads have been based on traffic volume
(High, Medium or Low) and road condition (Good, Fair and Poor), giving a total of nine
sections.
The section ID (5 letters) for the bituminous sections is derived as shown by the following
example:
For PHTGC:
P Paved
HT High Traffic
GC Good Condition
Unsealed (gravel) roads have been classified by traffic volume alone, since condition may
traffic, so the gravel network is represented by two sections only, one for medium traffic and
one for low traffic. The section ID (3 letters) is based on pavement type and traffic level as
indicated below:
For GLT:
G Gravel
LT Light traffic

For GMT:
G Gravel
MT Medium traffic
The use of a matrix template to define the characteristics of the road network in terms of
representative sections is discussed in Section 2.2 of Chapter B3. While this is useful in
identifying groups of road sections with specific characteristics, the definition of properties
for each representative section (for example, based on the average or weighted average of
values within each group) is currently undertaken manually.
Double-clicking on a specific Section Description (or selecting the section and clicking the
View/Edit Section button) reveals the General/Geometry/Pavement/Condition Tabs that
give access to the section's details. Some relevant characteristics of the representative sections
are summarised in Table C1.1and Table C1.2. Note that, for both bituminous and gravel
sections, the condition must be defined at the start of the analysis period. As the analysis
period commences in the year 2000, this means that the condition data must be defined for the
year 1999 (that is, end of 1999) or earlier.

Table C1.1
Length (km) 234 392 437 306 483 615 410 670 720
AADT 6200 5240 5180 2500 2300 2060 1400 1150 970
AADT year 1998 1998 1998 1998 1998 1998 1998 1998 1998
Roughness (IRI) 2.5 4.0 6.5 2.5 4.4 6.5 3 4.5 6.1
Ravelled area (%) 1 5 15 1 8 26 1 10 20
Edge break (m2/km) 0 5 10 0 5 15 2 12 30
Mean rut depth (mm) 2 5 10 2 5 12 3 10 15
Texture depth (mm) 3 2 0.5 3 2 0.5 3 2 0.5
New surf thick (mm) 50 50 50 25 25 50 15 15 15
…Continued
Version 1.0
Old surf thick (mm) 50 50 50 50 50 0 0 0 0

- reconstruction 1991 1991 1991 1991 1991 1991 1991 1991 1991
Version 1.0
Table C1.2
Section ID GMT GLT

Length (km) 1385 1760
AADT (1998) 175 75
Condition (1998)
Version 1.0
Note that the Adjusted Structural Number (SN or SNP as appropriate) assigned to each section
represents the condition of the pavement immediately after the most recent treatment (that is,
date of last preventative treatment as defined under the Pavement Tab). For example, the SN
value assigned to section PHTGC above represents Pavement Type AMAP (specified under
the Definition Tab) immediately after the overlay application in 1996. The value of SN
(surfacing, roadbase and sub-base only) can be quickly approximated as indicated below,
based on the original AASHTO relationship:
SN = 0.0394
∑a h i i ...(1.1)
where:
ai layer coefficient
hi layer thickness (mm)
Note the reduced value of layer coefficient (0.2) applied to the old surfacing due to its
condition.
Table C1.3
Layer coefficients and thicknesses
Construction Layer Layer Layer aihi

date thickness (mm) coefficient
hi ai
1996 AC Overlay 50 0.4 20
1991 AC 50 0.2 10
1991 Granular 150 0.14 21
1991 Granular 150 0.11 16.5
Total 67.5
Giving:
SN ≅ 0.0394 * 67.5 = 2.7 ...(1.2)
This value has been assigned to section PHTGC under the Section/Pavement Tab.
The subgrade contribution (added to SN to give SNP) is computed by the HDM-4 software.
Note that this approximation does not take account of the reduced contribution from the sub-
base, but is a much quicker manual calculation. Details of HDM-4 Structural Number
concepts are given in Chapter C1 of the Analytical Framework and Model Descriptions.
Version 1.0
Select Vehicles
The vehicles selected from the pre-defined Western Province fleet are indicated above.
Individual vehicle attributes may be reviewed by double-clicking on the appropriate vehicle
type description (or by selecting the section and clicking the Edit Vehicle button).
Define Normal Traffic
This screen confirms the traffic volume (AADT) on the selected road section in the given
year. The initial traffic composition and growth rates (by vehicle type) can be reviewed by
double-clicking the section description (or by selecting the section and clicking the
Section/Traffic/Details button). The initial vehicle compositions are summarised in Table
C1.4.
Note that the initial composition is defined for same year as the AADT. The annual growth
rates may be specified for several periods, but must at least be specified for a period
Version 1.0
Version 1.0
Table C1.4
Traffic level
High Medium Low

% % % % % %
Tata 7.5t lorry 15 3 0 0 10 3
Total (%) 100 100 100
Version 1.0
1.2.2 Specify Standards Assignments

For Strategy Analysis, the assignment of works is identical to that of Project Analysis. The
assigned standards can be checked under:
Specify Standards Assignments/Alternatives
By clicking the View Details box, the user may review the assignment of Road Works
Standards by Alternative and Section.
Alternatives
Table C1.5

traffic
Bituminous (paved)
Gravel (unpaved)
Version 1.0
investment alternative, and an improvement standard has been introduced to Alternative 4
(that is, upgrading gravel roads with medium traffic followed by paved road maintenance, and
widening gravel roads with low traffic).
(Note that Road Authorities may wish to assign standards by Road Class, which may be
represented in some countries by Surface Class alone. With condition responsive
maintenance intervention, the interval between maintenance applications will clearly be
influenced by the traffic volume.)
For bituminous (paved) sections, the Maintenance Standard short code is derived from the
pavement surfacing, traffic class, and investment alternative, for example:
PHT1 is based on:
P Paved
HT High Traffic
1 Alternative 1
For gravel sections, the short code for Maintenance Standards is derived from the pavement
surfacing material and the investment alternative:
For example:
G1 represents the Maintenance Standard for a gravel section under Alternative 1

GU4 represents the Improvement Standard to upgrade under Alternative 4
GW4 represents the Improvement Standard to widen under Alternative 4
Version 1.0
Table C1.6
Represent- Surface Traffic Investment Maintenance Improve- Future

ative class class alternative standard ment maint-
1 PHT1 - -
2 PHT2 - -
PHTGC B H
3 PHT3 - -
4 PHT4 - -
1 PHT1 - -
2 PHT2 - -
PHTFC B H
3 PHT3 - -
4 PHT4 -
1 PHT1 - -
2 PHT2 - -
PHTPC B H
3 PHT3 - -
4 PHT4 - -
1 PMT1 - -
2 PMT2 - -
PMTGC B M
3 PMT3 - -
4 PMT4 - -
1 PMT1 - -
2 PMT2 - -
PMTFC B M
3 PMT3 - -
4 PMT4 - -
1 PMT1 - -
2 PMT2 - -
PMTPC B M
3 PMT3 - -
4 PMT4 - -
1 PLT1 - -
2 PLT2 - -
PLTGC B L
3 PLT3 - -
4 PLT4 - -
…Continued
Version 1.0
Represent- Surface Traffic Investment Maintenance Improve- Future

ative class class alternative standard ment maint-
1 PLT1 -
2 PLT2 -
PLTFC B L
3 PLT3 - -
4 PLT4 - -
1 PLT1 - -
2 PLT2 - -
PLTPC B L
3 PLT3 - -
4 PLT4 - -
1 G1 - -
2 G2 - -
GMT U M
3 G3 - -
4 G1 GU4 PLT3
1 G1 - -
2 G2 - -
GLT U L
3 G3 - -
4 G1 GW4 G2
Note: Details of Road Works Standards identified by the above short codes are given in
Table C1.9 and Table C1.11 for bituminous and gravel sections respectively.
Table C1.7

Crack sealing 22
If more than one of these activities is triggered in a given year, the works item that is highest
be applied in the same maintenance year. (The hierarchy is defined by the ranking of road
works shown in Chapter D2 of the Analytical Framework and Model Descriptions.
Version 1.0
! Alternative 1
Comprises patching and crack sealing only. This represents the base case (or do
minimum) alternative. In this case study the intervention levels for patching vary from
10 potholes/km for sections with high traffic, to 50 potholes/km for sections with low
traffic (see Table C1.9).
! Alternative 2
Includes patching, resealing and reconstruction. The objective of this alternative is to

adopt relatively inexpensive treatments which will maintain the existing road in a
reasonable condition for as long as possible until the eventual need for reconstruction.
! Alternative 3
Includes patching, overlaying and reconstruction. This alternative is effectively

Alternative 2 with resealing replaced by overlaying. Alternative 3 introduces
rehabilitation works at a pre-defined roughness level which will maintain a higher
serviceability level than Alternative 2, as the structural strength of the pavement will
increase at each overlay application.
! Alternative 4
Includes patching, resealing, overlaying and reconstruction. This alternative combines

the benefits of resealing and overlaying within one standard, which should reduce the
required frequency of the relatively expensive overlay works.
C1.8. The pavement details will normally be derived using appropriate design charts, based
on a design subgrade CBR and estimated cumulative traffic loading over the desired design
life. The subgrade CBR value is not requested within the reconstruction works item, as this is
already specified at section level (Section/Pavement).
Table C1.8

Details
Details of the Maintenance Standards associated with each alternative are summarised in
Table C1.9. The details may be reviewed under Maintenance Standards in HDM-4
Workspace or under Specify Standards Assignments/Alternatives/Edit Standards.
Version 1.0
Table C1.9
Investment Maintenance Standard Condition responsive criteria Units Intervention levels by traffic class
alternative
Works item High Medium traffic Low traffic
traffic

Crack sealing
2 Patch, reseal & (PHT2) (PMT2) (PLT2)

reconstruct
3 Patch, overlay & (PHT3) (PMT3) (PLT3)

reconstruct

Overlay
Cracking % >=5 >=5 >=5
…Continued
Version 1.0
Intervention levels by traffic class

4 Patch, reseal, overlay & (PHT4) (PMT4) (PLT4)
reconstruct

Overlay
Cracking % >=5 >=5 >=5
Notes:
Version 1.0

The Maintenance and Improvement Standards for gravel sections are based on combinations
of the works activities listed in Table C1.10.
Table C1.10

Spot regravelling 7
Lane addition 4
The order of ranking of the Maintenance works items gives regravelling highest priority (that
is, with lowest ranking value). Spot regravelling and Grading have equal ranking, and both of
these can be performed in a given analysis year. Details of precedence rules for road works
are given in Chapter D2 of the Analytical Framework and Model Descriptions.
! Alternative 1
Comprises grading and spot regravelling, both condition responsive. This represents the
base case (do-minimum alternative).
! Alternative 2
regravelling are both triggered by the thickness of gravel surfacing. This alternative aims
to maintain a reasonable thickness of gravel surfacing and so provide continuous
protection to the pavement subgrade.
! Alternative 3
Includes the same works items as Alternative 2, but the intervention levels are set to
provide a higher level of serviceability that should trigger the works more frequently.
! Alternative 4
section GMT), and widening for sections with low traffic (section GLT). This represents
the ideal investment alternative. Suitable maintenance is provided before and after the
respective improvements. Note that the respective Improvement Standards are effective
from the year 2003, with works scheduled to start in the year 2004. This means that no
maintenance will be applied during the year 2003, the year prior to improvement.
Details of the Road Works Standards for gravel road sections are given in Table C1.11. The
details can be reviewed under Maintenance Standards in HDM-4 Workspace or under:
Specify Standards Assignments/Alternatives/Edit Standards
Version 1.0
Table C1.11
from year
Works Item Start Date/Duration [I]

GLT

GLT

GLT
…Continued
Version 1.0
from year
Works Item Start Date/Duration [I]
Cracking >= 5%
Notes:
Version 1.0
1.3 Funding requirements without budget constraint

This step in Strategy Analysis produces an unconstrained work programme, based on a life-
cycle analysis that considers the different alternatives for each representative section. The
alternative giving the highest NPV is assigned to each section.
1.3.1 Generate Strategy
Perform Run / Run Setup

Programme Analysis. The multi-year forward programme option (available under
Programme Analysis) is not applicable to strategy analysis that is concerned with longer term
planning.
Work Programme
with the highest NPV (refer to Section 3.2.3 of Chapter B2). This represents an unconstrained
programme, with the total financial cost given in the cumulative cost column. The road works
assigned to each section represent the optimum maintenance and improvements that should be
applied to the road network in accordance with the specified standards.

defined, and an optimisation procedure is selected which aims to produce a works programme
that satisfies an objective function, that is:
Version 1.0
1.4.1 Perform budget optimisation (budget at 75% funding requirement)
Define Budget
The budget for the defined budget period has been set at 900 million US dollars.
Details of the budget optimisation set-up are specified on this screen. For this case study,
optimisation is by incremental benefit/cost ranking (see Section 4.2 of Chapter B2). The
parameters to be specified are given below:
Minimum incremental value = 0
Efficiency frontier zone = 95
Work Programme
accordance with the incremental benefit/cost ranking procedure described in Section 4.2 of
Chapter B2.
Version 1.0
1.4.2 Perform Budget Optimisation (budget at 50% funding requirement)
Define Budget
The budget for the defined budget period has been set at 450 million US dollars, representing
approximately 50% of the funding requirements from Step 3.
Work Programme
After performing the budget optimisation, the revised work programme is produced which
shows the road sections listed in priority order in accordance with the incremental benefit/cost
ranking procedure.
Version 1.0
Version 1.0
volume three
Software User Guide
David C. Wightman
Eric E. Stannard
Jo M. Dakin
Association
mondiale
de la Route
Wo r l d R o a d
Association
3
ABOUT THIS MANUAL
Part A About This Manual
About This Manual

Part B About This Manual
This HDM-4 V1 edition of the Software User Guide describes the HDM-4 software. It is a
general purpose document which provides an understanding of the software user interface.
The Software User Guide is one of five manuals comprising the suite of HDM-4
documentation (see Figure 1).
Overview
Overview of
of HDM-4
HDM-4
Volume
Volume 11
Applications
Applications Guide
Guide Software
Software User
User Guide
Guide
Volume
Volume 22 Volume 3
Volume 3
Analytical
Framework and
and
Model
Model Descriptions
Descriptions
Volume
Volume 44
A
A Guide
Guide to
to Calibration
Calibration and
and Adaptation
Adaptation
Volume
Volume 55


Software User Guide i

Version 1.0
ABOUT THIS MANUAL

Notes:
Structure of the ‘Software User Guide’

The aim of the Software User Guide is to provide information for using HDM-4. It covers
installing, running and using the software, describes the user interface, and details many of the
tasks related to working in the HDM-4 environment. It is a general-purpose document for all
users.
The information in the Software User Guide is structured in twelve sections:
1 About HDM-4
2 Installing HDM-4
3 Running HDM-4
4 Road Networks
5 Vehicle Fleets
6 Maintenance and Improvement Standards
7 Project analysis
8 Programme analysis
9 Strategy analysis
10 Working with reports
11 Configuring HDM-4
12 Managing your HDM-4 installation
Software User Guide ii

Version 1.0
ABOUT THIS MANUAL
Conventions Used in this Manual

The following symbols are used throughout this manual. The following legend describes the
meaning of each symbol:
Caution: the accompanying text warns of the consequences of not following the
recommended course of action.
Tip: the accompanying text contains details that may help your understanding of the
software, or may describe best practice.
Trouble-shooting: the accompanying text is intended to assist you if you encounter
a problem whilst using the respective part of the software.
ISOHDM Products
(ISOHDM) consist of the HDM-4 software suite, associated example case study databases,
manuals. This volume is a member of that document collection.
Customer contact
please report details of the problem by E-mail, or an annotated copy of the manual page by fax
to the number provided below.
HDM-4. Comments on the Software User Guide should be sent to the following address:
Telephone: +44 - 121 - 414 6717 (5049)
Fax: +44 - 121 - 414 3675 (5060)
Edgbaston
Birmingham B15 2TT
United Kingdom
Change details
Software User Guide iii

Version 1.0
ABOUT THIS MANUAL
HDM-4 documents:
comprises:

Edgbaston
Birmingham
B15 2TT
United Kingdom
Tel: +44 - 121 - 414 6717 (or 5049)
Fax: +44 - 121 - 414 3675 (or 5060)
Software User Guide iv

Version 1.0
ABOUT THIS MANUAL

La Grande Arche
France
Tel: +33 1 41 02 05 84
Fax: +33 1 49 00 02 02
Acknowledgements
! The World Bank

including:
! Finnra

! FICEM

concrete roads.
! The Highway Research Group (HRG), School of Civil Engineering, The University of
Birmingham
(NDLI)
costs.
Software User Guide v

Version 1.0
ABOUT THIS MANUAL
! SNRA
Copyright statement
Software User Guide vi

Version 1.0
Part A
Contents
Volume 3 - Software User Guide
1 About HDM-4
2 Installing HDM-4
2.1 Installing HDM-4 for the first time 2
2.2 Installing an HDM-4 update 3
2.3 Installing an HDM-4 upgrade 4
3 Running HDM-4
3.1 Working with the HDM-4 Workspace window 6
4 Road Networks
4.1 Key concepts 7
4.13 Specifying which Sections are displayed in the Road Network window 17
4.14 Editing Section details directly from the Road Network window 18
5 Vehicle Fleets
5.1 Key Concepts 20
Software User Guide i

Version 1.0
CONTENTS

5.10 Adding motorised Vehicle Types (MT) to a Vehicle Fleet 27
5.11 Adding non-motorised Vehicle Types (NMT) to a Vehicle Fleet 28
5.12 Editing Vehicle Type attributes 29
5.13 Specifying which Vehicle Types are displayed in the Vehicle Fleet window 32
5.14 Editing Vehicle Type attributes directly from the Vehicle Fleet window 33
6 Maintenance & Improvement Standards

6.1 Key concepts 35
6.3 Creating Maintenance Standards 37
6.4 Creating Maintenance Standards based on existing standards 38
6.5 Viewing / editing Maintenance Standards 39
6.6 Deleting Maintenance Standards 39
6.7 Renaming Maintenance Standards 40
6.8 Adding Work Items 40
6.9 Adding Work Items by copying existing Work Items 41
6.10 Editing Work Items 41
6.11 Deleting Work Items 42
6.12 Creating Improvement Standards 43
6.13 Creating Improvement Standards based on existing standards 44
6.14 Viewing / editing Improvement Standards 44
6.15 Deleting Improvement Standards 45
6.16 Renaming Improvement Standards 45
6.17 Updating default works unit costs and energy consumption 46
6.18 Importing Work Standards 47
6.19 Exporting Work Standards 47
7 Project analysis
7.1 Key concepts 49
Software User Guide ii

Version 1.0
CONTENTS
7.3 Creating Projects 50

7.4 Creating Projects based on existing Projects 52
7.5 Viewing / editing Projects 52
7.6 Deleting Projects 53
7.7 Renaming Projects 53
7.8 Exporting Projects 54
7.9 Importing Projects 54
7.10 Selecting Sections for a Project 55
7.11 Selecting Vehicle Types to include in a Project 57
7.12 Defining normal traffic 57
7.13 Analysis by Section 59
7.14 Analysis by Project 63
7.15 Customising the analysis run 69
7.16 Running the analysis 71
7.17 Generating reports 71
8.1 Key concepts 73
8.3 Creating Programmes 75
8.4 Creating Programmes based on existing Programmes 76
8.5 Viewing / editing Programmes 77
8.6 Deleting Programmes 77
8.7 Renaming Programmes 78
8.8 Exporting Programmes 78
8.9 Importing Programmes 79
8.10 Selecting Sections for a Programme 79
8.11 Selecting Vehicle Types to include in a Programme 81
8.13 Life-cycle analysis 83
8.14 Multi-year forward Programme analysis 87
8.15 Generating a work programme 88
8.16 Optimisation using budget constraints 90
9 Strategy analysis
9.1 Key concepts 93
Software User Guide iii

Version 1.0
CONTENTS

9.3 Creating Strategy analyses 94
9.4 Creating Strategy analyses based on existing Strategy analyses 96
9.5 Viewing / editing Strategies 96
9.6 Deleting Strategies 97
9.7 Renaming Strategies 98
9.8 Importing Strategy analyses 98
9.9 Exporting Strategy analyses 99
9.10 Creating a network matrix 99
9.11 Selecting Sections for a Strategy analysis 100
9.12 Selecting Vehicle Types to include in a Strategy analysis 101
9.14 Defining alternatives for a Strategy analysis 103
9.15 Generating a work programme 106
9.16 Optimisation using budget constraints 108

10.1 Key concepts 111
10.2 Printing reports 112
10.3 Exporting report data to other applications 112
10.4 Adding user defined reports 113
11.1 Traffic Flow Patterns 115
11.2 Speed Flow Types 120
11.3 Climate Zones 125
11.4 Currencies 130
11.5 Section aggregate data and tables 133

12.1 Working with multiple databases 136
12.2 Backing up your HDM-4 data 136
12.3 Improving analysis performance 137
12.4 Adding an HDM-4 shortcut to your desktop 138
Software User Guide iv

Version 1.0
1 About HDM-4
The Highway Development and Management (HDM-4) system is a software
tool that is used to appraise the technical and economic aspects of road
investment projects.
It provides facilities for storing characteristics of road networks, vehicle types,
and road works. The available analysis types are:
! Project analysis: the economic evaluation of individual road projects or
investment options.
! Programme analysis: the preparation of prioritised work programmes in
which investment alternatives are defined, and selected subject to
resource constraints.
! Strategy analysis: the analysis of a whole road network.
Software User Guide 1

Version 1.0
2 Installing HDM-4
This section describes how to install the HDM-4 software on your PC. There
are a number of scenarios in which you may want to install HDM-4 software:
! Installing HDM-4 for the first time
! Installing an upgrade (installing a new version over an existing one)
! Installing an update (installing a bug fix for an existing version)
You may have obtained the software on CD-ROM from a distributor, or via
Internet download. The following sections describe the various installation
scenarios, and cover CD and Internet use.
The recommended minimum platform requirements for running HDM-4 are
as follows:
! Pentium P100 processor (or equivalent)
! 32MB of RAM
! 30MB of hard disk space
! Windows 95/98 or NT 4.0 (with service pack 6)
HDM-4 is a sophisticated modelling tool which involves many processor
intensive calculations. As such, execution speed is significantly affected by
the specification of the PC on which it is to be run. The above platform
requirements are the minimum recommended. Obviously, the performance
will improve in relation to the specification of your PC.
2.1 Installing HDM-4 for the first time

The method of installation will differ slightly according to whether you have
obtained the software on CD-ROM or via Internet download. If you have the
HDM-4 CD-ROM, do the following: Insert the HDM-4 CD in your CD-ROM
drive. Microsoft Windows automatically scans the CD and displays the HDM-
4 CD window. First choose your preferred installation language. Then click
Install HDM-4.
Alternatively, if you down-loaded the installation program from the Internet,

do the following: Select the Run option from the Windows Start menu.
Specify the path of the HDM-4 installation program (SETUP.EXE). Click
OK.
Now do the following:

1 In the first window select the I Accept the Agreement option button and
click Next to proceed with the installation.
2 In the next window, enter the registration name and 16-digit licence key
supplied by the distributor. Click Next to proceed.
3 In the next window, select the Typical installation option button and
click Next.

Version 1.0
4 For most users the Typical installation option is recommended. The
Custom option should only be used either when doing an upgrade or
update, or when replacing individual components or files (see Section
2.2).
5 Do one of the following:
❏ Accept the default folder for storing HDM-4
❏ Specify a different folder for storing HDM-4
6 Click Next to begin installing the HDM-4 program files on your PC.
Please note that the installation program may need to restart Windows in order
to register the various user interface controls used by HDM-4. If you are
asked whether you want to restart Windows now, answer Yes. When the
installation is complete, you will notice that the HDM-4 program group has
been added to the Windows Start menu. The program group contains the
HDM-4 application, on-line help, and the README.TXT file which includes
information on late-breaking changes, and other useful details.
If the licence key is not on the back of the CD case, do one of the following:
! Consult the person who purchased the software
! Contact the distributor from whom the software was purchased
If the licence key you entered is rejected, do the following:
! Make sure that you have entered the serial number correctly
2.2 Installing an HDM-4 update

HDM-4 updates will be issued from time-to-time in response to bug reports.
In most cases these will be distributed via the Internet. Having downloaded an
update installation program, do the following:
1 Select the Run option from the Windows Start menu. Specify the
location of the HDM-4 update installation program. Click OK.
2 In the first window, enter the username and 16-digit licence key
originally supplied by the distributor. Click Next to proceed.
3 Click Next to begin installing the HDM-4 update on your PC.
If the serial number is not on the back of the CD case, do one of the
following:
! Contact the dealer from whom the software was purchased
If the serial number you entered is rejected, do the following:
! Make sure that you have entered the serial number correctly
! Contact the dealer from whom the software was purchased

Version 1.0
2.3 Installing an HDM-4 upgrade
The method of installation will differ slightly according to whether you have
obtained the software on CD-ROM or via Internet download. If you have the
HDM-4 CD-ROM, do the following: Insert the HDM-4 CD in your CD-ROM
drive. Microsoft Windows automatically scans the CD and displays the HDM-
4 CD window. Now choose your preferred installation language. Then click
Install HDM-4.
Alternatively, if you downloaded the installation program from the Internet,

do the following: Select the Run option from the Windows Start menu.
Specify the path of the HDM-4 installation program (SETUP.EXE). Click
OK.
Now do the following:

1 In the first window select the I Accept the Agreement option button and
click Next to proceed with the installation.
2 In the next window, enter the username and 16-digit licence key supplied
by the distributor. Click Next to proceed.
3 Now select the Custom installation option button and click Next. In the
next window select all components except Database. In this way, you
will install the upgrade whilst retaining your current ‘default’ database
(i.e. the database in the \hdmdict directory).
4 The installation program should automatically detect the folder in which
the existing version of HDM-4 is stored. If not, you will be prompted to
specify the folder.
5 Click Next to begin installing the HDM-4 upgrade on your PC.
If the serial number is not on the back of the CD case, do one of the
following:
If the serial number you entered is rejected, do the following:
! First make sure that you have entered the serial number correctly

Version 1.0
3 Running HDM-4
To run HDM-4:
1 Click the Windows Start menu.
2 Select the HDM-4 application item from the HDM-4 program group.
3 The HDM-4 splash screen is displayed. Once HDM-4 is loaded the
HDM-4 Welcome screen is displayed.
The HDM-4 Welcome screen

The Welcome screen is designed to introduce you to the concepts and
facilities of HDM-4. As you become familiar with the system, you will find it
easier to use the Workspace window where you can access all the HDM-4
facilities.
The Welcome screen provides access to the following core HDM-4 facilities:
! Create a new study
Select this option to create a new Project, Programme or Strategy
analysis. For more information, see the following sections:
❏ Creating Projects
❏ Creating Programmes
❏ Creating Strategy analyses
! Open an existing study
Select this option to open an existing Project, Programme or Strategy
analysis. For more information, see the following sections:
❏ Viewing / editing Projects
❏ Viewing / editing Programmes
❏ Viewing / editing Strategies
! Go to the HDM-4 Workspace
Select this option to open the HDM-4 Workspace window. For more
information, see Section 3.1 below.

Version 1.0
Make it easier to start-up HDM-4 by adding a shortcut to your Windows
desktop (see Section 12.4).
3.1 Working with the HDM-4 Workspace window

The Workspace window is the central hub of HDM-4. Folders are listed on
the left, and buttons on the right.
To help you navigate through the system, the folders are listed in the sequence
that you would use to create and analyse a new study if you were beginning
with a clean system.
The general procedure for performing an operation is:

1 Expand a folder by clicking the + to the left of the folder name.
2 Select an item in the folder.
3 Click the appropriate button for the operation you wish to perform.
The button functions are also available from the Workspace menu.

Version 1.0
4 Road Networks
4.1 Key concepts

An HDM-4 Road Network stores details of the roads that you wish to analyse.
Each Road Network consists of a number of Sections. A Section typically
corresponds to an identifiable length of road, but may also be a
‘representative’ Section created solely for the purpose of analysis.
HDM-4 uses the concept of ‘homogenous Sections’, where each Section has
uniform strength, geometry, traffic and condition characteristics over its entire
length. When adapting existing Road Network data for use in HDM-4, or
when entering new road data directly into HDM-4, you will have to divide
your network into homogenous Sections.
4.2 Software overview

The Sections in a Road Network are displayed in a spreadsheet in the Road
Network window (an example is shown below). Options on the View menu
determine which Sections and/or attributes are displayed.
D
B
C E
A. Each row displays a separate Section.
B. These buttons allow you to add, delete and edit Sections.
C. The status bar shows information about the currently active control.
D. The title bar shows the name of the Road Network and the category
of data currently on display.
E. You can change the size of the window by dragging the Resize Box.
You can edit most (but not all) Section attributes here in the spreadsheet. To
edit attributes that are not available in the spreadsheet, double-click a Section.
The resulting Section dialog box provides access to all Section attributes.

Version 1.0
F
F. Click this button to view detailed Section parameters.
Road Section data is stored on two levels:

Key Section ! Consists of data that are readily available and/or are
attributes applicable in most environments.
! Attributes differ according to the surface class of the
Section (i.e. bituminous, concrete or unsealed).
! Key attributes can be viewed and edited in the
spreadsheet views.
Detailed ! Access detailed data by clicking Details.
Section
attributes ! Detailed data include those parameters which are either
not readily available, are only required in certain
circumstances, or should only be changed by those
calibrating the various models.
! Attributes differ according to the surface class of the
Section (i.e. bituminous, concrete or unsealed).
! Detailed attributes cannot be viewed or edited in the
spreadsheet views.
HDM-4 places no restriction on the number of Road Networks that can be
created, or the number of Sections in each Road Network. The only constraint
is the amount of free disk space on your PC.
The sections in this chapter describe the tasks of creating, editing and deleting
both Road Networks and Sections.

Version 1.0
4.3 Creating new Road Networks with no Sections
This section describes how to create a new Road Network that initially has no
Sections.
❏ Click New in the Workspace window
❏ With the Workspace window active, select New from the
Workspace menu
The New HDM-4 Item dialog box is displayed.
2 Select Road Network and click OK.

The New Road Network dialog box is displayed.
Type the name of the new Road Network in the text box and click OK.
The Road Network window is displayed.
You can now add Sections to the blank Road Network (see Sections 4.10
and 4.11).
When you have finished working with the Road Network, click Save before
closing the window. If do not click Save, your changes will be lost.
The new Road Network is added to the Workspace window. To view the new
copy, see Section 4.5.

Version 1.0
In step 2 above, if the Road Networks folder, or any existing Road Network
is selected in the Workspace window, the Road Network Item will be
automatically selected in the New HDM-4 Item dialog box, and you will only
need to click OK.
4.4 Creating Road Networks based on existing networks

If you want to create a Road Network that is similar to an existing one, you
can copy the existing network and edit the copy as necessary.
1. In the Workspace window, select the existing network that you wish to
copy.
2. To copy the selected network, do one of the following:
❏ Click Copy
❏ With the Workspace window active, select Copy from the
Workspace menu
The Copy Road Network dialog box is displayed.

3. The name of the Road Network you wish to copy is shown. In the To edit
box, enter the name of the new network. This name must be unique.
4 Click OK.
The new Road Network is added to the Workspace window.
To view the new copy, see Section 4.5 below.
4.5 Viewing / editing Road Networks

1 In the Workspace window, expand the Road Networks folder.
2 To open a Road Network, do one of the following:
❏ Select a Road Network and click Open
❏ Select a Road Network. With the Workspace window active, select
Open from the Workspace menu
❏ Double-click the Road Network

Version 1.0
When you have finished working with the Road Network, click Save before
closing the window. If you do not click Save, your changes will be lost.
To edit the Section details, see Section 4.12.
4.6 Deleting Road Networks

1 In the Workspace window, select the Road Network that you wish to
delete.
2 To delete the selected network, do one of the following:
❏ Click Delete
❏ With the Workspace window active, select Delete from the
Workspace menu
You are prompted to confirm the deletion. Click Yes to delete the Road
Network, or No to cancel the delete action.
If you click Yes, the deleted Road Network is no longer displayed in the
Workspace window.
! When you delete a Road Network, all corresponding Sections are also
deleted.
! Projects, Programmes and Strategies that use the deleted Road Network
will be affected. You will be informed if this is the case.
! You cannot reverse a deletion.
4.7 Renaming Road Networks

1 In the Workspace window, select the Road Network that you wish to
rename.
2 To rename the selected network, do one of the following:
❏ Click Rename
❏ With the Workspace window active, select Rename from the
Workspace menu
The Rename Road Network dialog box is displayed.

3 The current name of the Road Network you wish to rename is shown. In
the To edit box, enter the new name for the Road Network. The new
name must be unique.
4 Click OK.
The renamed Road Network is displayed in the Workspace window. The
name change is reflected throughout the system.
To work with the renamed Road Network, see Section 4.5.

Version 1.0
4.8 Exporting Road Networks
You can use HDM-4 Road Network data in other software applications by
exporting individual Road Networks in a ‘.dbf’ file. See the file HDM-4
RoadNet Export Format.doc for details of the Road Network export file
format.
1 In the Workspace window, expand the Road Networks folder and select
a Road Network.
2 Click Export.
The Export Road Network As dialog box is displayed.
3 Double-click the folder in which the export file is to be stored.

4 Enter a file name for the export file.
5 Click Save.
4.9 Importing Road Networks

You can import Road Network data into HDM-4 from other software
applications. To do this, the data must be in a .dbf file of a specific format.
See the file HDM-4 RoadNet Export Format.doc for details of the Road
Network export file format.
1 With the Workspace window active, select Import from the Workspace
menu.
The Import HDM-4 Export File dialog box is displayed.
2 Double-click the folder that contains the file to be imported.

3 Select the file to be imported and click Open.
The Enter Name for Imported Road Network dialog box is displayed.

Version 1.0
4 Enter a unique name for the Road Network and click OK.
In step 4, above, you are informed if a Road Network of the same name
already exists. You must then decide whether to overwrite the existing Road
Network, or to rename the Road Network that is being imported, or to cancel
the import operation.
4.10 Adding Sections based on aggregate data

1 In the Workspace window, create or open a Road Network.
The Road Network dialog box is displayed.
2 Click Add New Section.
The New Section dialog box is displayed.
3 Select the based on aggregate data option button and click OK.
The New Section From Aggregate Data dialog box is displayed.
4 Enter a name and ID for the Section. These must be unique.

5 Specify values for all the aggregate parameters.
Aggregate parameters correspond to more particular values for Key and
Detailed parameters as defined in the Configuration folder (Section
Aggregate Data and Section Aggregate Tables). When you specify
aggregate parameters, values are derived for all Key and Detailed

Version 1.0
parameters which are then used to populate the edit boxes in the Section
dialog box (see next step).
6 Click OK.
The Section dialog box is displayed.
7 Work through the tab-pages and edit the necessary fields.

8 If necessary, you can click Details to edit detailed Section parameters.
When you have finished working in the Details dialog box, click OK to
save any changes.
9 When you have finished working in the Section dialog box, click OK to
save any changes and return to the Road Network window.
4.11 Adding Sections based on existing Sections

If you want to create a Section that is similar to an existing one, you can copy
the existing Section and edit the copy as necessary.
1 In the Workspace window, create or open a Road Network.
2 Click Add New Section.
3 Select the based on existing Section option button. In the drop-down

list, select the existing Section that you wish to copy and click OK.
The Section dialog box is displayed. The data in the new Section is a
copy of the data in the existing Section.

Version 1.0
4 Enter a description and an ID for the new Section. These must be unique.
Work through the tab-pages and make any necessary changes.
save any changes.
save any changes.
4.12 Editing Section attributes

1 In the Workspace window, open a Road Network.
2 To edit Section attributes do one of the following:
❏ Click anywhere in a Section row and click Edit
❏ Double-click anywhere in a Section row
3 Work through the tab-pages and make any necessary changes.

save any changes.

Version 1.0
save any changes.
4.12.1 Calculating the SNP (Structural Number of Pavement) for a Section

Pavement strength is characterised in HDM-4 by the adjusted structural
number - SNP. This can either be entered directly or calculated using other
pavement characteristics such as Falling Weight Deflections, Benkelman
Beam Deflections, or surface layer thicknesses and coefficients.
2 To edit the Section details, do one of the following:
❏ Click anywhere in a Section row and click Edit
❏ Double-click anywhere in a Section row
3 Click the Pavement tab.
4 In the Strength group, select the Calculated SNP option button.

5 Click Calculate SNP.
The SNP Calculator Wizard is displayed. The Step 1 of 3 tab-page is
active.
6 Select one of the method option buttons to calculate the SNP.

Version 1.0
7 Click Next.
The Step 2 of 3 tab-page is displayed.
The fields in this dialog box vary according to the calculation method
you selected.
8 Complete the fields as required.
9 Click Next.
The Step 3 of 3 tab-page is displayed. It shows the calculated SNP
results for the calculation method you selected.
10 Click Finish to accept the calculated SNP value and automatically

transfer it to the Pavement tab-page of the Section dialog box.
When using the wizard, you may return to the previous step (click Back) or
cancel the procedure (click Cancel) at any time.
4.13 Specifying which Sections are displayed in the Road Network

window
You can specify which Sections are displayed in the Road Network window
on the basis of surface class.
2 To display the View menu, do one of the following:
❏ In the main HDM-4 window, click the View menu

Version 1.0
❏ Right-click anywhere in the Road Network window
The following surface classes are available:
❏ Bituminous Sections
❏ Concrete Sections
❏ Unsealed Sections
❏ All Sections
The active menu item is marked with a ".
Select a surface class from the menu.
Only the Sections with the selected surface class are displayed.
4.14 Editing Section details directly from the Road Network

window
2 To display the View menu, do one of the following:
❏ Right-click anywhere in the Road Network window
Data categories from the Section dialog box are listed in the lower half
of the menu:
❏ Definition
❏ Geometry
❏ Pavement
❏ Condition
❏ History
❏ Miscellaneous
Categories that are relevant to the current surface class are available. The
active category is marked with a ".
3 Select a category from the menu.
The columns in the spreadsheet change according to the category
selected. You can edit the tabular information in the Road Network
window.
In step 3 above, if the category options are greyed out, try selecting a different
surface class, then select the required category.
4.15 Deleting Sections


Version 1.0
2 In the Road Network window, click anywhere in the Section row you
wish to delete.
3 Click Delete.
You are prompted to confirm the deletion. Click Yes to delete the
Section, or No to cancel the delete action.
If you click Yes, the Section is no longer displayed in the Road Network
window.
4 Click Close to return to the Workspace window.
! When you delete a Section, all corresponding data are also deleted.
! Studies that use the deleted Section will be affected. You will be
informed if this is the case.

Version 1.0
5 Vehicle Fleets
5.1 Key Concepts

Vehicle Fleets are used to store details of the Vehicle Types to be included in
HDM-4 analyses. A Vehicle Fleet consists of a number of Vehicle Types.
Each Vehicle Type represents a class of vehicles in the traffic mix being
modelled (e.g. small car, large truck, etc.). When creating a Vehicle Fleet, you
should include one Vehicle Type for each vehicle class in the traffic mix
being modelled.
HDM-4 includes 20 default Vehicle Types. The models for vehicle speed,
road user effects, and social and environmental effects have been developed
for these default types. The default Vehicle Types fall into the following two
categories:
Motorised category
! Motorcycle ! Medium Truck
! Small Passenger Car ! Heavy Truck
! Medium Passenger Car ! Articulated Truck
! Large Passenger Car ! Minibus
! Light Delivery Vehicle ! Light Bus
! Light Goods Vehicle ! Medium Bus
! Four Wheel Drive ! Heavy Bus
! Light Truck ! Coach
Non-motorised category
! Pedestrian
! Bicycle
! Rickshaw
! Animal Cart
You can use these Vehicle Types directly in your analyses, or you can create
your own user-defined types. A user-defined type must be based on a default
type so that HDM-4 can determine which models to use for analysis.

Version 1.0
The Vehicle Types in a Vehicle Fleet are displayed in a spreadsheet in the
Vehicle Fleet window (an example is shown below). Options on the View
menu determine which Vehicle Type category and/or group of vehicle
attributes is displayed.
You can edit most (but not all) attributes here in the spreadsheet. To edit
attributes that are not available in the spreadsheet, you can double-click a
Vehicle Type. The resulting Vehicle Attributes dialog box provides access to
all vehicle attributes.
Vehicle Type data is stored on two levels:

Key vehicle ! Consists of data that are readily available and/or are
attributes applicable in most environments.
! Attributes differ according to the category of the Vehicle
Type (i.e. motorised or non-motorised).
! Key attributes can be viewed and edited in the
spreadsheet views.
Detailed ! Access detailed data by clicking Details.
vehicle
attributes ! Detailed data include those parameters which are either
not readily available, are only required in certain
circumstances, or should only be changed by those
calibrating the various road deterioration models.
! Attributes differ according to the category of the Vehicle

Version 1.0
Type (i.e. motorised or non-motorised).
! Detailed attributes cannot be viewed or edited in the
spreadsheet views.
HDM-4 places no restriction on the number of Vehicle Fleets that can be
created, or the number of Vehicle Types in each fleet. The only constraint is
the amount of free disk space on your PC.
The sections that follow describe the tasks of creating, editing and deleting
both Vehicle Fleets and Vehicle Types.
5.3 Creating Vehicle Fleets

This section describes how to create a new Vehicle Fleet which initially has
no Vehicle Types.
Workspace menu
2 Select Vehicle Fleet and click OK.

The New Fleet dialog box is displayed.
3 Type the name of the new Vehicle Fleet and choose a unit cost Currency
from the drop-down list. This is the Currency in which unit costs should
be specified for all Vehicle Types that are added to the fleet.
4 Click OK.
The Vehicle Fleet window is displayed. You can now add Vehicle Types
to the blank Vehicle Fleet (see Section 5.10 and 5.11).

Version 1.0
When you have finished working with the Vehicle Fleet, click Save before
closing the window. If you do not click Save, your changes will be lost.
The new Vehicle Fleet is added to the Workspace window. To view it, see
Section 5.5.
In step 2 above, if the Vehicle Fleets folder, or any existing Vehicle Fleet is
selected in the Workspace window, the Vehicle Fleet item will be
need to click OK.
5.4 Creating Vehicle Fleets based on existing fleets

If you want to create a Vehicle Fleet that is similar to an existing one, you can
copy the existing fleet and edit the copy as necessary.
1 In the Workspace window, select the existing Vehicle Fleet that you
wish to copy.
2 To copy the selected fleet, do one of the following:
❏ Click Copy in the Workspace window
Workspace menu
The Copy Vehicle Fleet dialog box is displayed.

3 The name of the Vehicle Fleet you wish to copy is shown. In the To edit
box, enter the name of the new fleet. This name must be unique.
4 Click OK.
The new Vehicle Fleet is added to the Workspace window.
5.5 Viewing / editing Vehicle Fleets

1 In the Workspace window, expand the Vehicle Fleets folder.
2 To open a Vehicle Fleet, do one of the following:
❏ Select a Vehicle Fleet and click Open

Version 1.0
❏ Select a Vehicle Fleet. With the Workspace window active, select
❏ Double-click the Vehicle Fleet.

The Vehicle Fleet window is displayed.
When you have finished working with the Vehicle Fleet, click Save before
To edit the Vehicle Types, see Section 5.12.
5.5.1 Changing the fleet Currency

You specify the fleet Currency when you create a Vehicle Fleet. Unit costs for
all Vehicle Types in the Vehicle Fleet are specified in this Currency. If you
wish to change the fleet Currency all unit costs should be updated to reflect
the change. To change the Fleet Currency do the following:
1 In the Workspace window, expand the Vehicle Fleets folder.
2 To open a Vehicle Fleet, do one of the following:
❏ Select a Vehicle Fleet and click Open
❏ Select a Vehicle Fleet. With the Workspace window active, select
❏ Double-click the Vehicle Fleet

3 Click Info.
The Fleet Information dialog box is displayed.
4 Select a Currency from the drop-down list.

Version 1.0
5 Click OK to save the change and return to the Vehicle Fleet window.
! You can define Currencies in the Configuration folder. See Section
11.4.2.
! In the Fleet Information dialog box, there is also an edit box where you
can enter a general description for the Vehicle Fleet.
5.6 Deleting Vehicle Fleets

1 In the Workspace window, select the Vehicle Fleet that you wish to
delete.
2 To delete the selected fleet, do one of the following:
❏ Click Delete
Workspace menu
Vehicle Fleet, or No to cancel the delete action.
If you click Yes, the deleted Vehicle Fleet is no longer displayed in the
Workspace window.
! When you delete a Vehicle Fleet, all corresponding Vehicle Types are
also deleted.
! Studies that use the deleted Vehicle Fleet will be affected. You will be
5.7 Renaming Vehicle Fleets

1 In the Workspace window, select the Vehicle Fleet that you wish to
rename.
2 To rename the selected fleet, do one of the following:
❏ Click Rename
Workspace menu
The Rename Vehicle Fleet dialog box is displayed.

The current name of the Vehicle Fleet you wish to rename is shown. In
the To edit box, enter the new name for the Vehicle Fleet. The new name
must be unique.
Click OK.
The renamed Vehicle Fleet is displayed in the Workspace window.
To work with the renamed Vehicle Fleet, see Section 5.5.

Version 1.0
5.8 Importing Vehicle Fleets
You can import Vehicle Fleet data into HDM-4 from other software
applications. To do this, the data must be in a .dbf file of a specific format.
See the file HDM-4 Vehicle Fleet Export Format.doc for details of the export
file format.
menu.

3 Select the file to be imported and click Open. The import routine detects
the file type. If it is a Vehicle Fleet export file then the Enter Name for
Imported Vehicle Fleet dialog box is displayed.
4 Enter a unique name for the Vehicle Fleet and click OK.
In step 4, above, you are informed if a Vehicle Fleet of the same name already
exists. You must then decide whether to overwrite the existing Vehicle Fleet,
to provide an alternative name for the Vehicle Fleet that is being imported, or
to cancel the import operation.
5.9 Exporting Vehicle Fleets

You can use HDM-4 Vehicle Fleet data in other software applications by
exporting individual Vehicle Fleets as a .dbf file. See the file HDM-4 Vehicle
Fleet Export Format.doc for details of the export file format.
1 In the Workspace window, expand the Vehicle Fleets folder and select
a Vehicle Fleet.
2 Click Export.
The Export Vehicle Fleet As dialog box is displayed.

Version 1.0
5 Click Save.
5.10 Adding motorised Vehicle Types (MT) to a Vehicle Fleet

1 In the Workspace window, create or open a Vehicle Fleet.
2 In the Vehicle Fleet window, click Add New Vehicle.
The New Vehicle Type dialog box is displayed.
3 Select the Motorised option button and select a base Vehicle Type from
the drop-down list.
The drop-down list contains default motorised Vehicle Types and user
defined Vehicle Types for the current Vehicle Fleet. Initial data for the
new Vehicle Type is derived from the base Vehicle Type.
4 Click OK.
The Vehicle Attributes dialog box is displayed.

Version 1.0
5 Work through the tab-pages and complete the necessary fields.
6 If necessary, you can click Calibration to edit detailed Vehicle Type
parameters. When you have finished working in the Calibration dialog
box, click OK to save any changes.
7 When you have finished working in the Vehicle Attributes dialog box,
click OK to save any changes.
5.11 Adding non-motorised Vehicle Types (NMT) to a Vehicle Fleet

1 In the Workspace window, create or open a Vehicle Fleet.
2 In the Vehicle Fleet window, click Add New Vehicle.
The New Vehicle Type dialog box is displayed.
3 Select the Non-motorised option button and select a base Vehicle Type
from the drop-down list.
The drop-down list contains four default non-motorised Vehicle Types
and user defined Vehicle Types for the current Vehicle Fleet. Initial data
for the new Vehicle Type is derived from the base Vehicle Type.
4 Click OK.
The NM Vehicle Attributes dialog box is displayed.
5 Work through the tab-pages and complete the necessary fields.

7 When you have finished working in the NM Vehicle Attributes dialog

Version 1.0
5.12 Editing Vehicle Type attributes
1 In the Workspace window, open a Vehicle Fleet.
2 To open the selected fleet, do one of the following:
❏ Click anywhere in a Vehicle Type row and click Edit
❏ Double-click anywhere in a Vehicle Type row
3 Work through the tab-pages and make any necessary changes.

box, click OK to save any changes and return to the Vehicle Attributes
dialog box.
5.12.1 Calculating Annual km and Hours Worked

The values for the utilisation parameters Annual km and Working Hours can
either be entered directly (if the data is available) or can be calculated by
HDM-4 from age spectrum data. The Vehicle Utilisation Wizard is provided
to simplify the calculation process.
2 To edit the Vehicle Type details, do one of the following:
3 Click the Basic Characteristics tab.

Version 1.0
4 Click Calculate in the Utilisation group.
The Utilisation Parameters Wizard is displayed. The Step 1 of 2 tab-
page is active.
5 Click Add New Year.

A blank row is created in the table.
6 Enter age spectrum data values in the table.
7 Add the necessary number of years.
8 Click Next.
The Step 2 of 2 tab-page is displayed. It shows the values calculated
from the data on the previous tab-page.
9 Click Finish to accept the calculated values and insert them in the Basic
Characteristics tab-page of the Vehicle Attributes dialog box.

Version 1.0
5.12.2 Calculating ESAL

The value for the vehicle loading parameter ESAL may be entered directly (if
the data is available), or calculated by HDM-4 from raw vehicle axle loading
data that you supply. The ESAL Wizard assists with the calculation process.
2 To edit the Vehicle Type attributes, do one of the following:
3 Click the Basic Characteristics tab.
4 Click Calculate in the Loading group.

The Calculate ESAL Wizard is displayed. The Step 1 of 3 tab-page is
active.
5 Enter the axle load equivalency exponent (LE) in the edit box.
6 Click Next.

Version 1.0
7 Click Add New Axle Group.
A blank column is created in the table. Note: If you click Insert New
Axle Group, a blank column will be created at the cursor position,
moving any other columns to the right.
8 Click Add New Vehicle.
A blank row is added to the bottom of the table.
9 Enter loading data by axle group for the Vehicle Type.
10 Add the necessary number of Vehicles Types
11 Click Next.
12 Click Finish to accept the calculated ESAL factor and insert it in the
Basic Characteristics tab-page of the Vehicle Attributes dialog box.
5.13 Specifying which Vehicle Types are displayed in the Vehicle

Fleet window
You can specify which Vehicle Types are displayed in the Vehicle Fleet
window on the basis of category (motorised or non-motorised).

Version 1.0
❏ Right-click anywhere in the Vehicle Fleet window
The following categories are available:
❏ Motorised vehicles
❏ Non-motorised vehicles
❏ All categories
The active item is marked with a ".
3 Select a category from the menu.
Only the Vehicle Types with the selected category are displayed.
5.14 Editing Vehicle Type attributes directly from the Vehicle Fleet
window
You can edit Vehicle Type details in the Vehicle Fleet window spreadsheet.
2 To select the group of attributes, do one of the following:
❏ Right-click anywhere in the Vehicle Fleet window
The following groups of attributes are listed in the lower half of the
menu:
❏ Definition
❏ Basic Characteristics
❏ Economic Unit Costs
❏ Financial Units Costs
The active attribute group is marked with a ".
3 Select an attribute group from the menu.
The columns in the Vehicle Fleet window change according to the
attribute group you selected. You can edit the tabular information in the
Vehicle Fleet window.
In step 3 above, if the attribute group you require is greyed out, select a
different vehicle category (Motorised, Non-motorised) first, then select the
attribute group.

Version 1.0
5.15 Deleting Vehicle Types
2 In the Vehicle Fleet window, click anywhere in the Vehicle Type row
you wish to delete.
3 Click Delete.
Vehicle Type, or No to cancel the delete action.
If you click Yes, the Vehicle Type is no longer displayed in the Vehicle
Fleet window.
4 Click Close to return to the Workspace window.

! When you delete a Vehicle Type, all corresponding data are also deleted.
! Analyses that use the deleted Vehicle Type will be affected. You will be
informed if this is the case. If you continue with the deletion, the Vehicle
Type will be removed from the analysis and any associated traffic data
will be lost.

Version 1.0
6 Maintenance & Improvement Standards
6.1 Key concepts

In HDM-4 Maintenance and Improvement Standards are used to represent the
targets or levels of condition and response that you aim to achieve.
Maintenance Standards define the maintenance works required to maintain the
road network at the target level. Each Maintenance Standard consists of a set
of one or more Works Items. Each Works Item is defined in terms of the road
surface class to which it applies, an intervention level, an operation type (e.g.
reseal, overlay, etc.), and the resultant effect on the pavement.
Improvement Standards define the road improvement works to be carried out
should the state of the road network fall below a certain level. Each
Improvement Standard is defined in terms of the road surface class to which it
applies, an intervention level, an improvement type, the costs & duration of
the works, and the resultant effect on the pavement in terms of its condition,
geometry, strength, etc. Supported improvement types include: lane addition,
partial widening, reconstruction, and upgrading.
Typically you will create a set of Maintenance and Improvement Standards
that can then be assigned to road sections in many subsequent Projects,
Programmes and Strategies that you create.

Below is an example of the Maintenance Standard dialog box. The top part of
the dialog shows the definition of the standard including the surface type. The
bottom half of the dialog shows the works items defined for the standard.
A Maintenance Standard consists of one or more Works Items. Double-

clicking on one of the Works Items in the list box causes the Works Item
dialog box to be displayed – an example of which is shown below:

Version 1.0
The data that defines the Works Item is separated into 5 categories, each
assigned to a separate tab-page as shown in the figure above:
! General – the definition of the Works Item: description, operation type
and intervention type (scheduled or responsive).
! Design – the design and structure of the pavement following the
completion of the works: surface material, thickness, strength coefficient,
etc.
! Intervention – the content of this tab-page will differ according to the
intervention type selected. For scheduled intervention you will be
prompted to specify the time interval at which the selected Works Item
will be triggered. For responsive intervention you will be prompted to
specify one or more condition levels at which the Works Item should be
triggered (e.g. when Area of All Cracking > 10%).
! Costs – the unit cost for the selected operation, appropriate preparatory
works costs, and the unit of work (e.g. per m, per m², per km, etc.).
HDM-4 stores default costs and work units for each of the supported
operation types. Each time you change the operation type for a Works
Item, you will be asked if you want to adopt the current default cost and
units (see Section 6.17).
! Effects – the condition of the pavement following the completion of the
works (e.g. roughness and rutting).
Please note that not all of these tab-pages will be displayed for any given
Works Item. The operation type determines which pages are displayed.
Below is an example of the Improvement Standard dialog box. Unlike
Maintenance Standards, there is no hierarchical Standard / Works Items
relationship. For an Improvement Standard only one works type is involved:

Version 1.0
The data that defines the Improvement Standard is separated onto various tab-
pages, as with Maintenance Standards. Two extra tab-pages are included:
! Pavement – detailed description of the pavement structure following
improvement.
! Geometry – the new geometry / alignment of the road following the
improvement.
A common misconception is that the Surface Class parameter on the General
tab-page refers to the surface of the pavement following the improvement.
This is incorrect. The parameter actually refers to the class of roads to which
this standard applies. For example, if the Bituminous surface class is selected,
then the standard will only be considered for bituminous road sections.
6.3 Creating Maintenance Standards

Workspace menu
2 Select Maintenance Standard and click OK.

The Maintenance Standard dialog box is displayed.

Version 1.0
3 Enter a name and short code for the new Maintenance Standard. These
must be unique.
You can now add Works Items to the empty Maintenance Standard (see
Section 6.8).
4 Click OK to save the Maintenance Standard and close the dialog box.
The new Maintenance Standard is added to the Workspace window. To
view it, see Section 6.5.
In step 2 above, if the Work Standards folder, Maintenance Standards
folder, or any existing Work Item is selected in the Workspace window, the
Maintenance Std item will be automatically selected in the New HDM-4 Item
dialog box, and you will only need to click OK.
6.4 Creating Maintenance Standards based on existing standards

If you want to create a Maintenance Standard that is similar to an existing
one, you can copy the existing standard and edit the copy as necessary.
1 In the Workspace window, expand the Work Standards folder.
2 Expand the Maintenance Standards folder.
3 Select the Maintenance Standard that you wish to copy.
4 To copy the selected Maintenance Standard, do one of the following:
Workspace menu
The Copy Maintenance Standard dialog box is displayed.

5 The name of the Maintenance Standard you wish to copy is shown. In the
To edit box, enter the name of the new standard. This name must be
unique.
6 Click OK.
The new Maintenance Standard is added to the Workspace window.

Version 1.0
6.5 Viewing / editing Maintenance Standards
3 To open a Maintenance Standard, do one of the following:
❏ Select a Maintenance Standard and click Open
❏ Select a Maintenance Standard. With the Workspace window
active, select Open from the Workspace menu
❏ Double-click the Maintenance Standard
4 When you have finished working with the Maintenance Standard, click
OK to save any changes and close the dialog box.
To edit the Work Items, see Section 6.10.
6.6 Deleting Maintenance Standards

3 Select the Maintenance Standard that you wish to delete.
4 To delete the selected Maintenance Standard, do one of the following:
❏ Click Delete
Workspace menu
Maintenance Standard, or No to cancel the delete action.
If you click Yes, the deleted Maintenance Standard is no longer
displayed in the Workspace window.
! When you delete a Maintenance Standard, all corresponding Work Items

are also deleted.

Version 1.0
! Studies that use the deleted Maintenance Standard will be affected. You
will be informed if this is the case.
6.7 Renaming Maintenance Standards

3 Select the Maintenance Standard that you wish to rename.
4 To rename the selected Maintenance Standard, do one of the following:
❏ Click Rename
Workspace menu
The Rename Maintenance Standard dialog box is displayed.

5 The current name of the Maintenance Standard you wish to rename is
shown. In the To edit box, enter the new name for the Maintenance
Standard. The new name must be unique.
6 Click OK.
The renamed Maintenance Standard is displayed in the Workspace
window.
To view the renamed Maintenance Standard, see Section 6.5.
6.8 Adding Work Items

1 In the Workspace window, create or open a Maintenance Standard (in
the Work Standards folder).
2 Click Add New Work Item.
The Work Item dialog box is displayed.

Version 1.0
3 Enter a name and short code for the Work Item. The Work Item name
and code must be unique within a Maintenance Standard.
4 Work through the tab-pages and complete the necessary fields. The
number of tab-pages and their content changes according to the operation
type you select in the General tab-page.
5 When you have finished working in the Maintenance Works dialog box,
6 When you have finished working in the Maintenance Standard dialog
6.9 Adding Work Items by copying existing Work Items

If you want to create a Work Item that is similar to an existing one, you can
copy the existing item and edit the copy as necessary.
2 Select an existing Work Item from the Work Items group list.
3 Click Copy Work Item.
A new Work Item is added to the list. It is automatically assigned a
unique name and short code which you can edit if necessary (see Section
6.10).
6.10 Editing Work Items

2 Select a Work Item in the Work Items group list.
3 Click Edit.
The Maintenance Work Item dialog box is displayed.

Version 1.0
number of tab-pages and their content changes according to the operation
type you choose in the General tab-page.
5 When you have finished working in the Maintenance Work Item dialog

6.11 Deleting Work Items

1 In the Workspace window, open a Maintenance Standard (in the Work
Standards folder).

2 Select a Work Item in the Work Item group list.
3 To delete the selected Work Item, do one of the following:
❏ Click Delete
Workspace menu
You are prompted to confirm the deletion. Click Yes to delete the Work
Item, or No to cancel the delete action.
If you click Yes, the deleted Work Item is no longer displayed in the
Maintenance Standard dialog box.

! When you delete a Work Item, all corresponding data are also deleted.
! Studies that use the deleted Work Item will be affected. You will be

Version 1.0
6.12 Creating Improvement Standards
Workspace menu
2 Select Improvement Standard and click OK.

The Improvement Standard dialog box is displayed.
3 Enter a name and short code for the new Improvement Standard. These
must be unique.
number of tab-pages and their content changes according to the
improvement type you choose in the General tab-page.
5 Click OK to save the Improvement Standard and close the dialog box.
The new Improvement Standard is added to the Workspace window. To
view it, see Section 6.14.
In step 2 above, if the Improvement Standards folder or any existing

Improvement Standard is selected in the Workspace window, the
Improvement Standard item will be automatically selected in the New HDM-4
Item dialog box, and you will only need to click OK.

Version 1.0
6.13 Creating Improvement Standards based on existing standards
If you want to create an Improvement Standard that is similar to an existing
one, you can copy the existing standard and edit the copy as necessary.
2 Expand the Improvement Standards folder.
3 Select the Improvement Standard that you wish to copy.
4 To copy the selected Improvement Standard, do one of the following:
Workspace menu
The Copy Improvement Standard dialog box is displayed.

The name of the Improvement Standard you wish to copy is shown. In the To
edit box, enter the name of the new standard. This name must be unique.
Click OK.
The new Improvement Standard is added to the Workspace window.
6.14 Viewing / editing Improvement Standards

Expand the Improvement Standards folder.
To open an Improvement Standard, do one of the following:
❏ Select a Improvement Standard and click Open
❏ Select a Improvement Standard. With the Workspace window
❏ Double-click the Improvement Standard
The Improvement Standard dialog box is displayed.
Work through the tab-pages and complete the necessary fields.

Version 1.0
The tab-pages change according to the improvement type you choose in
the General tab-page.
When you have finished working with the Improvement Standard, click OK to
save any changes.
6.15 Deleting Improvement Standards

3 Select the Improvement Standard that you wish to delete.
4 To delete the selected Improvement Standard, do one of the following:
❏ Click Delete
Workspace menu
Improvement Standard, or No to cancel the delete action.
If you click Yes, the deleted Improvement Standard is no longer
displayed in the Workspace window.
! When you delete an Improvement Standard, all corresponding data are
also deleted.
! Studies that use the deleted Improvement Standard will be affected. You
will be informed if this is the case.
6.16 Renaming Improvement Standards

3 Select the Improvement Standard that you wish to rename.
4 To rename the selected Improvement Standard, do one of the following:
❏ Click Rename
Workspace menu
The Rename Improvement Standard dialog box is displayed.

5 The current name of the Improvement Standard you wish to rename is
shown. In the To edit box, enter the new name for the Improvement
Standard. The new name must be unique.
6 Click OK.
The renamed Improvement Standard is displayed in the Workspace
window.

Version 1.0
To work with the renamed Improvement Standard, see Section 6.14.
6.17 Updating default works unit costs and energy consumption

HDM-4 stores default unit costs for all works operation types that it supports.
When you create a new Work Item, the relevant default costs are inherited.
Energy consumption rates for the works operation types are also stored. You
can edit both the energy consumption rates and the default works costs.
2 To edit the default works costs and energy consumption, do one of the
following:
❏ Select Default Works Costs & Energy Consumption. Click Open.
❏ Select Default Works Costs & Energy Consumption. With the
Workspace window active, select Open from the Workspace
menu.
❏ Double-click Default Works Costs & Energy Consumption.
The Default Works Costs And Energy Consumption dialog box is
displayed.
3 Click Energy to display the energy consumption columns.
4 Enter values as necessary.

5 When you have finished working in the Default Works Costs And
Energy Consumption dialog box, click OK to save any changes.

Version 1.0
6.18 Importing Work Standards
You can import Work Standards that were originally exported from HDM-4.
This enables data exchange to take place between HDM-4 users.
menu.
2 Go into the folder that contains the file to be imported.

3 Select one of the following files:
❏ IMPROVES.DBF
❏ MAINTSTDS.DBF
❏ RESPONINT.DBF
❏ WORKDEFAULTS.DBF
❏ WORKS.DBF
4 Click Open.
! In step 3, all the Maintenance Standards (including Work Items),
Improvement Standards and default works costs and energy consumption
records are imported, irrespective of the file you select.
! In step 3, above, you are informed if a Work Standard of the same name
already exists. You are prompted to rename the Work Standard that is
being imported.
6.19 Exporting Work Standards

When you export Work Standards from HDM-4, several .dbf and .hdbf files
are created. If you are using the export facility to exchange data with other
HDM-4 users, you must supply all the files.
1 In the Workspace window, select Work Standards or any of the folder
contents.
2 Click Export.
The Browse for Folder dialog box is displayed.

Version 1.0
3 Select the folder in which the export file is to be stored.
4 Click OK.
All the Maintenance Standards (including Work Items), Improvement
Standards and default works costs and energy consumption records are
exported irrespective of the item that you select in the Work Standards
folder.

Version 1.0
7 Project analysis
7.1 Key concepts

An HDM-4 Project contains information about existing Road Sections, new
Road Sections, Vehicle Types, traffic (normal, diverted and generated), and
investment alternatives. An alternative is a combination of maintenance
and/or Improvement Standards that is applied to a Section. You can analyse
the Project using several different alternatives, allowing you to determine
which is the most cost effective.

A key feature of the Project window (shown below) is the four workflow
buttons on the left-hand side.
The following table shows how the workflow buttons guide you through the
process of defining and running a Project analysis.

Version 1.0
Workflow Tab-page Procedure
button
1 Specify the Road Network that contains the Sections to be

analysed.
2 Specify the Vehicle Fleet that contains the Vehicle Types to be

analysed.
3 Specify general Project information such as the start year,

duration, and output Currency.
General 4 Specify the analysis mode. There are two modes of analysis:
! By Project
Define Project
Details Determines the best alternative for the group of Sections
analysed.
! By Section
Determines the best alternative for each Section.
Select Sections 5 Select specific Sections for analysis.
Select Vehicles 6 Select specific Vehicle Types for analysis.
Define Normal 7 Define the traffic composition and expected growth for each
Traffic Section / Vehicle Type.
Specify Alternatives 8 Define the alternatives to be analysed. Some alternatives may

Alternatives include diverted and/or generated traffic, exogenous benefits
and costs, and new Sections.
Setup run 9 Customise the run setup. This includes:
! Specifying a base alternative, with which the other

alternatives should be compared.
! Specifying various sub-models such as energy balance,

Analyse Projects economic analysis and emissions analysis.
Run Analysis 10 Run the analysis. If vital information is missing, you are
informed, and must supply the necessary information so that
HDM-4 can complete the run. The time required to perform
the analysis depends on the complexity of the Project.
Generate Reports Select Reports 11 Generate reports and examine the results of the run.
HDM-4 places no restriction on the number of Projects that can be created.

The only constraint is the amount of free disk space on your PC.
7.3 Creating Projects

Workspace menu

Version 1.0
2 Select Project and click OK.
The New Project dialog box is displayed.
3 Enter a name for the new Project, and select a Road Network and Vehicle
Fleet.
4 Click OK.
The Project dialog box is displayed. The Define Project Details
workflow button is selected and the General tab-page is active.
5 Select the mode of analysis:

Analyse by Section Alternatives are defined for each Section. Each
alternative defines a set of works standards. The
analysis determines the optimum alternative for
each Section. NPV is calculated for each
section alternative. Analysis by Section
supports neither New Sections nor diverted
traffic.
See Section 7.13.
Analyse by Project Each alternative defines a set of work standards
for each section. The analysis determines the
optimum alternative for the sections as a group.
NPV is calculated for each alternative (i.e. for

Version 1.0
the group of sections). Analysis by Project
supports the analysis of New Sections and
diverted traffic.
See Section 7.14.
6 In the Currencies group, select an output Currency from the drop-down
list.
7 Edit the other information on the General tab-page as necessary. You
can now work through the tab-pages to select Sections and Vehicle
Types, and define traffic.
When you have finished working with the Project, click Save before closing
the window. If do not click Save, your changes will be lost.
The new Project is added to the Workspace window. To view it, see Section
7.5.
In step 2 above, if the Projects folder or any existing Project is selected in the
Workspace window, the Project item will be automatically selected in the
New HDM-4 Item dialog box, and you will only need to click OK.
7.4 Creating Projects based on existing Projects

If you want to create a Project that is similar to an existing one, you can copy
the existing Project and edit the copy as necessary.
1 In the Workspace window, select the Project that you wish to copy.
2 To copy the selected Project, do one of the following:
❏ Click Copy
Workspace menu
The Copy Project dialog box is displayed.

3 The name of the Project you wish to copy is shown. In the To edit box,
enter the name of the new Project. This name must be unique.
4 Click OK.
The new Project is added to the Workspace window.
7.5 Viewing / editing Projects

1 In the Workspace window, expand the Projects folder.
2 To open a Project, do one of the following:
❏ Select a Project and click Open
❏ Select a Project. With the Workspace window active, select Open
from the Workspace menu
❏ Double-click a Project

Version 1.0
The Project window is displayed. The Define Project Details workflow
button is selected and the General tab-page is active.
When you have finished working with the Project, click OK to save any
changes.
In steps 1 and 2 above, you can alternatively, double-click the Project in the
Workspace window.
7.6 Deleting Projects

1 In the Workspace window, select the Project that you wish to delete.
2 To delete the selected Project, do one of the following:
❏ Click Delete
Workspace menu
Project, or No to cancel the delete action.
If you click Yes, the deleted Project is no longer displayed in the
Workspace window.
! When you delete a Project, associated alternatives, traffic and run

settings are also deleted.
7.7 Renaming Projects

1 In the Workspace window, select the Project that you wish to rename.
2 To rename the selected Project, do one of the following:
❏ Click Rename
Workspace menu
The Rename Project dialog box is displayed.

Version 1.0
3 The current name of the Project you wish to rename is shown. In the To
edit box, enter the new name for the Project. The new name must be
unique.
4 Click OK.
The renamed Project is displayed in the Workspace window.
To work with the renamed Project, see Section 7.5.
7.8 Exporting Projects

When you export Project data from HDM-4, a number of files are created—a
.dbf file and several .hdbf files. If you are using the export facility to
exchange data with other HDM-4 users, you must supply all the files. When
importing the Project, only the .dbf file is seen by the importer.
1 In the Workspace window, expand the Projects folder and select a
Project.
2 Click Export.
3 Select the folder in which the export files are to be stored.

4 Click OK.
You can only export one Project at a time. If you want to export a number of
Projects one after the other, select a different folder for each Project. Do not
rename any of the files.
7.9 Importing Projects

You can import Project data that was originally exported from HDM-4. This
enables data exchange to take place between HDM-4 users.
menu.

Version 1.0
You are informed if a Project of the same name already exists. You are
prompted to rename the Project that is being imported.
When you import a Project, other items that are used by the Project are also
imported (for example, Vehicle Fleet and Maintenance Standards).
7.10 Selecting Sections for a Project

1 In the Workspace window, create or open a Project.
2 Click the Select Sections tab.
In a new Project, all Sections in the Road Network are initially listed, but
none are selected for the Project.
3 Click the buttons in the Include column to select/de-select Sections.

Selected Sections are marked with a ".
! You can choose to display or hide un-selected Sections by ticking or
clearing the Show unselected Sections check box.
! Click Unselect All to exclude all Sections from the Project. The " is
removed from all the Section lines.
! You can also select Sections by specifying selection criteria (see below).

Version 1.0
! You can edit/view detailed Section information. Click Edit/View Section
to access the Section dialog box.
! You can edit/view detailed Road Network information. Click Edit/View
Network to access the Road Network window.
7.10.1 Specifying selection criteria

You can select Sections for an analysis by specifying search criteria. The
Sections that match the search criteria are selected for analysis. This is
particularly useful if the Road Network you are using contains a large number
of Sections.
1 In the Select Sections tab-page click Select by Criteria.
The Network Section Selection dialog box is displayed.
2 Specify pavement, speed flow, traffic and geometry selection criteria.

3 Click the Add to currently selected option button, or the Replace
currently selected option button.
4 Click Select.
The Selected Sections dialog box is displayed. The Sections that match
the selection criteria are listed.
5 Do one of the following.

Version 1.0
❏ Click OK to include the selected Sections in the analysis. The
Sections are entered in the spreadsheet and marked with a " to show
they are included in the analysis.
❏ If you want to repeat the selection cycle, click Continue Selection.
7.11 Selecting Vehicle Types to include in a Project

2 Click the Select Vehicles tab.
In a new Project, all Vehicle Types in the fleet are listed and marked with
a " to show they are included in the Project.
3 Click the buttons in the Include column to de-select/select Vehicle Types.
! You can show/hide un-selected Vehicle Types by ticking/clearing the

Show unselected vehicles check box.
! The Include NMT in analysis option is only enabled if there are non-
motorised vehicle types in the selected fleet.
! You can exclude non-motorised Vehicle Types from the analysis by
clearing the Include NMT in analysis check box.
! You can edit/view the attributes of the selected Vehicle Type. Click
View/Edit Vehicle to display the Vehicle Attributes dialog box.
! You can edit/view the selected Vehicle Fleet. Click View/Edit Fleet to
display the Vehicle Fleet window.
7.12 Defining normal traffic

Define normal traffic for each Section you have selected for the Project.
Specify the initial traffic mix and the traffic growth for the selected Vehicle
Types.

Version 1.0
You must select Sections and Vehicle Types before you can define
traffic.
2 Click the Define Normal Traffic tab.
The Sections that you selected for the Project are listed. You can now
define traffic details for each Section.
3 Click anywhere in a Section row and click Edit Section Traffic Details.
The Normal Traffic Details dialog box is displayed. The Vehicle Types
that you selected for the Project are listed. Motorised Vehicle Types and
non-motorised Vehicle Types are separated into two tab-pages.
4 Define the normal traffic details:

❏ The sum of the initial compositions must be 100%.
❏ Enter traffic growth values for first year of analysis. Use the buttons
to add, edit and delete additional traffic growth periods as necessary.
N.B. You cannot edit the AADT from this screen. The section AADT
may only be changed from the Section Details dialog.
5 Click OK to save any changes.

Version 1.0
You can copy traffic growth periods and compositions from one Section to
another. Click the Section row with the traffic details you want to copy. Click
Copy. Click the Section row where you want to paste the traffic details. Click
Paste.
7.13 Analysis by Section

The purpose of analysing a Project by Section is to determine the most cost
effective scheme of works for each Section included in the Project. The
scheme of works is called an alternative and will typically consist of a number
of Maintenance and Improvement Standards. Part of the Project definition is,
therefore, to define alternatives for each Section, and assign Maintenance and
Improvement Standards to each Section/alternative combination.
7.13.1 Defining alternatives

2 Click the Analyse by Section option button.
3 Click the Specify Alternatives workflow button.
The Sections that you selected for the Project are listed. You can now
define alternatives for each Section.
Note: The Add New Alternative, Copy Alternative and Delete

Alternative buttons are not available for this type of analysis.
4 Select the Section for which you want to define alternatives.

5 Click Edit Alternatives.
The Section Alternatives Details dialog box is displayed. The
Description field shows the name of the Section that you are defining
alternatives for. A base alternative is automatically created for each
Section that has been added to the Project. No maintenance or
Improvement Standards are initially assigned to it.

Version 1.0
6 Click Add New Alternative.
The New Section Alternative dialog box is displayed.
7 Enter the name of the new alternative and click OK.

The new alternative is added to the list. To assign Maintenance and
Improvement Standards, see Section 7.13.2 and Section 7.13.3.
8 Click OK to save any changes.
! In step 3 above, tick / clear the View Details check box to display / hide
alternatives, and assigned Maintenance and Improvement Standards.
! In the Section Alternative Details dialog box, you can edit an
alternative name by selecting it and clicking Edit Alternative.
! In the Section Alternatives Details dialog box, you can delete an
alternative by selecting it and clicking Delete Alternative. The
alternative no longer appears in the list.
7.13.2 Assigning Maintenance Standards

1 In the Section Alternatives Detail dialog box, select the alternative that
you want to assign Maintenance Standards to.
2 Click Assign Maintenance.
The Assign Maintenance Standard dialog box is displayed.

Version 1.0
3 Enter the year from which the Maintenance Standard will be considered
for implementation, select a Maintenance Standard, and click OK.
The dialog box is closed, returning you to the Section Alternatives
Details dialog box. When you select an alternative, the standards that
have been assigned to it are shown in the Assignments column.
7.13.3 Assigning Improvement Standards

1 In the Section Alternatives Details dialog box, select the alternative that
you want to assign Improvement Standards to.
2 Click Assign Improvement.
The Assign Improvement Standard dialog box is displayed.
3 Enter the year from which the Improvement Standard will be considered
for implementation, and select an Improvement Standard.
Road improvements may have associated benefits, costs, and traffic
effects. To specify exogenous benefits and costs for an Improvement
Standard, see Section 7.13.4. To define generated traffic for an
Improvement Standard, see 7.13.5.
4 Click OK.
The dialog box is closed, returning you to the Section Alternatives
Details dialog box. When you select an alternative, the standards that
have been assigned to it are shown in the Assignments column.
! You can make alterations to the standard assignments you have made.
Select the Section, select the assignment, and click Edit Assignment.
! You can remove a standard assignment. Select the Section, select the
assignment, and click Remove Assignment.
! You can copy standard assignments from one Section to another. Select
the Section that you want to copy from and click Copy Assignment.
Select the Section that you want to copy to, and click Paste
Assignment.
7.13.4 Specifying exogenous benefits and costs

To define benefits and costs for an Improvement Standard, perform the steps
listed below:

Version 1.0
1 In the Assign Improvement Standard dialog box, tick the Ex Bens &
Costs check box.
2 Click Ex Bens & Costs.
The Exogenous Benefits and Costs dialog box is displayed.
3 Enter a description and the number of years from the start of the
improvement work until benefits and costs will start.
4 Click Append Benefit and Append Cost to add benefit and cost rows to
the spreadsheet.
5 Click Add Period to add a growth period column to the spreadsheet.
Add the necessary number of growth periods to reflect the growth rate of
the benefits and costs. You must define a growth period for the second
year of benefits and costs.
6 Enter initial amounts and growth rates for the benefits and costs you have
added.
! You can make changes to the growth period definition by clicking
anywhere in the period column and clicking Edit Period.
! You can delete periods by clicking anywhere in the period column and
clicking Delete Period.
! You can delete benefits and costs by clicking anywhere in the benefit or
cost row and clicking Delete Ben/Cost.
7.13.5 Specifying generated traffic

Generated traffic is used to model the new trips generated by road
improvement.
1 In the Assign Improvement Standard dialog box, tick the Generated
Traffic check box.
2 Click Generated Traffic.
The Generated Traffic dialog box is displayed. The Vehicle Types that
you selected for the Project are listed in a spreadsheet.

Version 1.0
3 Enter a description and the number of years from the start of the
improvement until generated traffic begins.
4 Click Append Period.
The Generated Traffic Growth Period dialog box is displayed. Add the
necessary number of growth periods to reflect the growth of the
generated traffic. You must define a growth period for the first year of
generated traffic.
5 Enter a start year for the period, select a growth type and click OK.
A traffic growth period column is added to the spreadsheet.
6 Enter appropriate traffic growth values for the Vehicle Types and click
OK.
If you do not click OK before closing the dialog box, any changes you have
made will be lost.
! You can make changes to the traffic growth period definition. Click
anywhere in the period column and click Edit Period.
! You can delete traffic growth periods. Click anywhere in the period
column and click Delete Period.
7.14 Analysis by Project

The purpose of analysing by Project is to determine the optimum scheme of
works for the analysis sections as a whole. Each scheme of works is called an
alternative and consists of a number of Maintenance and Improvement
Standards. Part of the Project definition is, therefore, to define alternatives for
each Section, and assign Maintenance and Improvement Standards to each
alternative/Section combination.

Version 1.0
2 Click the Analyse by Project option button.

The New Alternative dialog box is displayed.
5 Enter a name and click OK.

The new alternative is added to the list. To define Maintenance and
Improvement Standards for alternative/Section combinations, see Section
7.14.2 and Section 7.14.3.
! In step 3 above, tick or clear the View Details check box to display or
hide Sections, and assigned Maintenance and Improvement Standards.
! You can copy an alternative by selecting it and clicking Copy
Alternative. A new copy of the alternative with a unique name is added
to the list.
! You can delete an alternative by selecting it and clicking Delete
Alternative. The alternative no longer appears in the list.

1 In the Alternatives tab-page of the Project window, select the alternative
that you want to assign Maintenance Standards to.
2 Click Edit Alternative.

Version 1.0
The Alternative Details dialog box is displayed. The Description field
shows the name of the alternative that you are defining Maintenance
Standards for. The Sections that you selected for the Project are listed.
3 Select a Section and click Assign Maintenance.

The dialog box is closed, returning you to the Alternative Details dialog
box. When you select a Section, the standards that have been assigned to
the Section/alternative combination are shown in the Assignments
column.

1 In the Alternatives tab-page of the Project window, select the alternative
that you want to assign Improvement Standards to.
shows the name of the alternative that you are defining Improvement
Standards for. The Sections that you selected for the Project are listed.

Version 1.0
3 Select a Section and click Assign Improvement.
Improvement Standard, see Section 7.13.5.
5 Click OK.
column.
! You can make alterations to standard assignments you have made. Select
the Section, select the assignment, and click Edit Assignment.
Select the Section that you want to copy to and click Paste Assignment.

Version 1.0
7.14.4 Defining diverted traffic
You can define diverted traffic for each Project alternative when you use the
Analyse by Project analysis mode. The traffic diversion facility is used to
model traffic diversions resulting from a Project alternative.
1 In the Alternative Details dialog box, click Traffic Diversion.
The Traffic Diversion for Alternative dialog box is displayed. The
Sections that you selected for the Project are listed in a spreadsheet. The
spreadsheet also shows the AADT values that you specified when
defining normal traffic for the Project.
2 Click Add Diversion Period.
The Traffic Diversion Period dialog box is displayed.
3 Enter a description and period start year. The start year cannot be before
the second year of analysis.
4 Click OK to return to the Traffic Diversion for Alternative dialog box.
A column is added to the spreadsheet for the new diversion period.
5 Click a Section row in the diversion period column.
6 Click Edit Diversion Details.
The Traffic Diversion Details dialog box is displayed. The Vehicle
Types that you selected for the Project are listed. Motorised and non-
motorised Vehicle Types are separated onto two tab-pages.
7 Select the growth type and enter the initial traffic compositions and
growth.
8 Click OK.
9 Repeat steps 5 to 8 for each Section that you want to define diverted
traffic for.

Version 1.0
10 When you have finished defining diverted traffic, click OK in the Traffic
Diversion for Alternative dialog box to save the changes.
! In step 2 above, you can make alterations to a diversion period. Click in

the diversion period column and click Edit Diversion Period.
! In step 2 above, you can delete a diversion period. Click in the diversion
period column and click Delete Diversion Period.
7.14.5 Defining new Sections

If you are using the Analyse by Project analysis mode, you can define new
Sections for one or more Project alternatives. This allows you to compare
Project alternatives with and without a new Section.
1 In the Alternative Details dialog box, click Add New Section.
The New Construction Section Option dialog box is displayed.
2 Enter a name for the new Section, the start year, and the duration of
construction. Complete the other fields as necessary.
3 Click Normal Traffic.
that you selected for the Project are listed. Motorised Vehicle Types and
non-motorised Vehicle Types are separated into two tab-pages.
4 Define the normal traffic details for the new Section:

Version 1.0
to add, edit and delete traffic growth periods as necessary.
5 Click OK to save the changes and return to the New Construction
Section Option dialog box.
6 Click Section Details.

The Section dialog box is displayed. It is divided into four tab-pages:
Definition, Geometry, Pavement and Condition.
7 Work through the tab-pages and enter the necessary values. See Section
4.12.1 for details of how to use the SNP calculation wizard in the
Pavement tab-page.
8 When you have finished defining the new Section, click OK to save the
changes and return to the New Construction Section Option dialog
box.
9 Click OK to save the changes and return to the Alternative Details
dialog box.
The new Section is added to the list. The icon that is used for a Section
added in this way is slightly different to the icon used for Sections
already included in the Project.
! You can make alterations to a new Section. Select the Section and click
Edit New Section.
! To remove a new Section, select the Section and click Remove New
Section.
7.15 Customising the analysis run

2 Click the Analyse Projects workflow button.
The Setup Run tab-page is active.

Version 1.0
3 The check boxes on this tab-page allow you to include / exclude the
following models from the analysis:
❏ Economic analysis
❏ Accident costs
❏ Energy balance
❏ Emissions
❏ Acceleration effects
Deciding which of these models to include / exclude will depend upon
the requirements of the particular Project. Typically, you should only
include those models that you require, as each additional model will
reduce the speed of the analysis run.
The acceleration effects model is different from the others listed here in
the sense that the others merely affect the types of outputs that are
produced. If a particular model isn’t included then the respective results
will not be calculated. Acceleration effects on the other hand will always
be calculated. The setting of the Acceleration effects control determines
the method of calculation and therefore the actual results. If selected,
then the acceleration effects indicators are calculated using the current
congested speed and acceleration noise; if not, then a fixed default is
used regardless of the values for speed and acceleration noise.
The setting of the Write Log File check-box determines whether a log
file is produced for the analysis run. The log file is a detailed trace of all
the individual calculations performed during the course of the analysis.
The log file is written to the filename hdm4log.txt in the HDM-4
installation folder. The log file is intended for use by modellers and
testers.
The content of the log file is very detailed. As a result, the file can
become very large even for a relatively small Project (e.g. 12MB is not
uncommon). This has a detrimental effect on analysis speed. If analysis
speed is important, and you do not need the detailed calculation results,
then the log file should be disabled. For further discussion of methods of
improving analysis performance please see Section 12.3.

Version 1.0
7.16 Running the analysis
Before attempting to run an analysis, you must define Project details and
specify alternatives.
2 Click the Analyse Projects workflow button.
3 Click the Run Analysis tab.
4 Click Start.
As the Project is analysed, messages are displayed in the list box and the
status of the analysis is shown below the list box.
! If vital information has not been supplied, you are warned and the
analysis run stops. The message tells you which information is missing.
Supply the information and start the run again.
! If non-vital information has not been supplied, you are warned, but the
run continues.
! At the end of the analysis, you can view a complete warning message.
Select the message in the list box and click View.
! You can abort an analysis at any time by clicking Abort.
7.17 Generating reports


Version 1.0
Before attempting to generate reports, you must complete an analysis
run.
2 Click the Generate Reports workflow button.
The Select Reports tab-page is displayed.
3 Expand a folder and select a report.
4 Click Generate Report.

The report is displayed.
For more information about reports, see Section 10.

Version 1.0
8.1 Key concepts

An HDM-4 Programme Analysis allows you to generate a short to medium
term programme of works for a selected set of Sections which maximises the
economic return subject to budgetary constraints. The investment alternatives
defined depend on the mode of analysis selected; Multi-year Forward
Programming or Life Cycle Analysis. Each method allows the user to define a
combination of maintenance and/or improvement standards for a Section. In
the case of Life Cycle Analysis, several investment alternatives can be
defined. Multi-year Forward Programming allows you to enter one set of
standards and analyse the benefit of performing the work when required or
delaying the work until the end of the analysis period.

A key feature of the Programme window (shown below) is the five workflow
process of defining and running a Programme analysis.

Version 1.0
button
1 Specify the Road Network that contains the Sections to be

analysed.
2 Specify the Vehicle Fleet that contains the Vehicle Types to

be analysed.
3 Specify general Project information such as the start year,

duration, and output Currency.
Specify the analysis mode. There are two modes of analysis:
! Life Cycle Analysis

Define Project
General Determines the best investment from a set of
Details
4 alternatives.
! Multi-year Forward Programming
Determines whether the work should be performed

during this budget period, or delayed.
5 Select specific Sections for analysis.
6 Select specific Vehicle Types for analysis.
Define the traffic composition and expected growth for the

7
Section/ Vehicle Type combinations.
Alternatives (for a The steps depend on the analysis mode you are using:
Life Cycle
Analysis) For a Life Cycle Analysis, define the set of alternatives to be
analysed.
Specify Standards
8
Assignments Work Programming For a Multi-year Forward Programme, define the
(for a Multi-year Maintenance Standards and Improvements for each Section.
Forward
Programme)
9 Customise the run setup.
For a Life Cycle Analysis, specify the base alternative.

Generate
Programme For a Multi-year Forward Programme, specify whether an
economic analysis should be performed. An unconstrained
Perform Run Programme is generated if economic analysis is not selected.
10 Run the analysis. If vital information is missing, you are

informed. You must supply the necessary information so
that HDM-4 can complete the run. The time required to
perform the analysis depends on the complexity of the
specified Programme.
Work Programme 11 The generated work programme is displayed. If you want to

perform a budget optimisation, you can manually select
works that you do not want to be considered for
optimisation.
12 Define budget periods and values (Life Cycle Analysis

only).
Perform Budget Define Budget
13 Define the optimisation parameters.
Optimisation
14 Run the budget optimisation.
Work Programme 15 The optimised work programme is displayed.
Generate Reports Select Reports 16 Generate reports and examine the results of the run

Version 1.0
HDM-4 places no restriction on the number of Programmes that can be
created. The only constraint is the amount of free disk space on your PC.
8.3 Creating Programmes

Workspace menu
2 Select Programme and click OK.

The New Programme dialog box is displayed.
3 Enter a name for the new Programme, and select a Road Network and
Vehicle Fleet.
4 Click OK.
The Programme window is displayed. The Define Programme Details
5 Enter a description for the Programme.

Version 1.0
6 Select the analysis mode:
Life-cycle: This option allows you to compare
two or more alternatives comprising
different works for the selected
sections.
See Section 8.13.
Multi-year forward This option is a simplification of the
Programme: lifecycle method. It compares the
benefit of performing capital works
when triggered, and postponing the
capital works to the first year after the
budget period.
See Section 8.14.
list.
can now work through the tab-pages to select Sections and Vehicle
Types, and define traffic.
When you have finished working with the Programme, click Save before
The new Programme is added to the Workspace window. To view it, see
Section 8.5.
In step 2 above, if the Programmes folder or any existing Programme is
selected in the Workspace window, the Programme item will be
need to click OK.
8.4 Creating Programmes based on existing Programmes

If you want to create a Programme that is similar to an existing one, you can
copy the existing Programme and edit the copy as necessary.
1 In the Workspace window, select the Programme that you wish to copy.
2 To copy the selected Programme, do one of the following:
❏ Click Copy
Workspace menu
The Copy Programme dialog box is displayed.

3 The name of the Programme you wish to copy is shown. In the To edit
box, enter the name of the new Programme. This name must be unique.
4 Click OK.
The new Programme is added to the Workspace window.

Version 1.0
8.5 Viewing / editing Programmes
1 In the Workspace window, expand the Programmes folder.
2 To open a Programme, do one of the following:
❏ Select a Programme and click Open
❏ Select a Programme. With the Workspace window active, select
❏ Double-click a Programme
3 When you have finished working with the Programme, click Save to
save any changes.
8.6 Deleting Programmes

1 In the Workspace window, select the Programme that you wish to
delete.
2 To delete the selected Programme, do one of the following:
❏ Click Delete
Workspace menu
Programme, or No to cancel the delete action.
The deleted Programme is no longer displayed in the Workspace
window.
! When you delete a Programme, associated alternatives, traffic and run
settings are also deleted.
! You cannot undo a deletion.

Version 1.0
8.7 Renaming Programmes
1 In the Workspace window, select the Programme that you wish to
rename.
2 To rename the selected Programme, do one of the following:
❏ Click Rename
Workspace menu
The Rename Programme dialog box is displayed.

3 The current name of the Programme you wish to rename is shown. In the
To edit box, enter the new name for the Programme. The new name must
be unique.
4 Click OK.
The renamed Programme is displayed in the Workspace window.
To work with the renamed Programme, see Section 8.4.
8.8 Exporting Programmes

When you export Programme data from HDM-4, a number of files are
created—a .dbf file and several .hdbf files. If you are using the export facility
to exchange data with other HDM-4 users, you must supply all the files.
When importing the Programme, only the .dbf file is seen by the importer.
1 In the Workspace window, expand the Programmes folder and select a
Programme.
2 Click Export.

4 Click OK.
You can only export one Programme at a time. If you want to export a number
of Programmes one after the other, select a different folder for each
Programme. Do not rename any of the files.

Version 1.0
8.9 Importing Programmes
You can import Programme data that was originally exported from HDM-4.
menu.
2 Go into the folder that contains the files to be imported.

3 Select the Programme file to be imported and click Open.
You are informed if a Programme of the same name already exists. You
are prompted to rename the Programme that is being imported.
When you import a Programme, other items that are used by the Programme
are also imported (for example, Vehicle Fleet and Maintenance Standards).
8.10 Selecting Sections for a Programme

1 In the Workspace window, create or open a Programme.
In a new Programme, all Sections in the Road Network are initially
listed, but none are selected for the Programme.
3 Click the buttons in the Include column to select/de-select Sections.


Version 1.0
! You can choose to display or hide unselected Sections by ticking or
! Click Unselect All to exclude all Sections from the Programme. The " is
removed from all the Section rows.
! You can select Sections by specifying selection criteria (see Section
7.10.1).
! You can group Sections (see Section 8.10.1).
! Click View/Edit Section to display the Section dialog box.
! Click View/Edit Network to display the Road Network window.
8.10.1 Grouping Sections

Sections can be assigned to groups which are then considered together when
performing budget optimisation.
1 In the Programme window Select Sections tab-page, click Groups.
The Network Section Groups dialog box is displayed.
2 Click Add New and enter a name and description for the group.

Version 1.0
3 Add Sections to the group using one of the following methods:
❏ Click the buttons in the Include column to select/ de-select Sections.
Selected Sections are marked with a "
❏ Click Select by criteria and specify selection criteria (see Section
7.10.1)
4 Click Finished.
The group is added to the list.
clearing the Show unselected sections check box.
! You can make changes to a group. Select the group and click Edit.
! You can delete a group. Select the group and click Delete.
8.11 Selecting Vehicle Types to include in a Programme

In a new Programme, all Vehicle Types in the Vehicle Fleet are listed
and marked with a " to show they are included in the Programme.

Version 1.0
3 Click the buttons in the Include column to de-select/select Vehicle Types.
! Unselected Vehicle Types are listed only when you tick the Show
unselected vehicles check box.
! The Include NMT in analysis option is only enabled if non-motorised

vehicles are defined in the selected vehicle fleet.
! Click View/Edit Vehicle to display the Vehicle Attributes dialog box.
! Click View/Edit Fleet to display the Vehicle Fleet window.

Define normal traffic for each Section you have selected for the Programme.
Specify the initial traffic mix and the traffic growth for the selected Vehicle
Types.
traffic.
The Sections that you selected for the Programme are listed. You can
now define traffic details for each Section.

Version 1.0
that you selected for the Programme are listed. Motorised Vehicle Types
and non-motorised Vehicle Types are separated into two tab-page pages.

Copy. Select the Section rows where you want to paste the traffic details.
Click Paste.
8.13 Life-cycle analysis

If you want to analyse data using the life-cycle method, you must define a
base alternative and at least one other.

Version 1.0
2 Click the Life-cycle option button.

3 Click the Specify Standards Assignments workflow button.
The Alternatives tab-page is displayed.


8.13.2 and Section 8.13.3.

Version 1.0
to the list.
! You can add, edit or delete Maintenance Standards by clicking Edit

Standards.

1 In the Alternatives tab-page of the Programme window, select the
alternative that you want to assign Maintenance Standards to.
Standards for. The Sections that you selected for the Programme are
listed.

column.

Version 1.0
1 In the Alternatives tab-page of the Programme window, select the
alternative that you want to assign Improvement Standards to.
Standards for. The Sections that you selected for the Programme are
listed.

effects. To specify exogenous benefits and costs for an improvement, see
Section 7.13.4. To define generated traffic for an improvement, see
Section 7.13.5.
5 Click OK.
column.

Version 1.0
! You can make alterations to the standard assignments you have made to a
Section. Select the Section, select the assignment, and click Edit
Assignment.
8.14 Multi-year forward Programme analysis
8.14.1 Assigning Work Standards

2 Click the Multi-year forward Programme option button.

3 Click the Specify Standards Assignments workflow button.
The Work Programming tab-page is displayed. The Sections you chose
for the Programme are listed in the spreadsheet.

Version 1.0
4 Click the Maintenance column for the first Section.
5 Click Assign standard.
The Assign Standard dialog box is displayed.
6 Select a standard and click Select.

The Maintenance Standard is displayed into the spreadsheet.
7 Repeat this procedure as required for the Improvement column, the
Maintenance After Improvement column, and for each of the listed
Sections.
In step 4, you can select more than one Section if you want to assign the same
standard to each.
8.15 Generating a work programme

Before attempting to generate a work programme, you must define
Programme details and assign Work Standards.
2 Click the Generate Programme workflow button.
3 Click Run Setup.
The Run Setup dialog box is displayed.
4 Complete the fields according to the type of analysis you are doing:

Version 1.0
❏ Life-cycle analysis
Choose a base alternative from the drop-down list.
❏ Multi-year forward programming—unconstrained
Select the option button Perform run without economic analysis.
If you choose this option no economic indicators will be calculated,
and you will not be able to perform a budget optimisation after
running the analysis.
❏ Multi-year forward programming—constrained
Select the option button Perform run with economic analysis. If
you choose this option, you will be able to perform a budget
optimisation after running the analysis (see Section 8.16).
5 Tick the check boxes for the additional options you want to run. The
options are:
❏ Accident costs
❏ Energy balance
❏ Emissions
the requirements of the particular Programme. Typically, you should
only include those models that you require, as each additional model will
then the acceleration effects indicators are calculated for the current
testers.
6 Click OK.
7 Click Start.

Version 1.0
As the Programme is analysed, messages are displayed in the list box and
the status of the analysis is shown below the list box.
8 Click the Work Programme tab.
The work programme is displayed. The work programme maximises the
Net Present Value irrespective of budget constraints.
9 If you are performing a life-cycle analysis or a constrained multi-year

forward Programme, you can select the work programme for a Section
and exclude it from budget optimisation. Click the Section row and click
Manual Assignment. The row is shown in italics.
run continues.
! At the end of the Programme analysis, you can view a complete warning
message. Select the message in the list box and click View.
! You can abort a Programme analysis at any time by clicking Abort.
8.16 Optimisation using budget constraints

This facility is only available if you perform a life-cycle analysis or a
constrained multi-year forward Programme.
2 Click the Perform Budget Optimisation workflow button.
The Define Budget tab-page is displayed. A budget period equivalent to
the Programme duration is listed in the spreadsheet.

Version 1.0
3 Define budget periods.
❏ If you are optimising a budget for a life-cycle analysis, you can
insert, append and delete budget periods.
❏ If you are optimising a budget for a constrained multi-year forward
Programme, only one budget period is possible and the Insert
Period, Append Period and Delete Period buttons are not
available.
4 Click Optimisation Setup.
The Optimisation Setup dialog box is displayed.
5 Enter a minimum incremental value and a value for the efficiency

frontier zone.
6 Click OK.
7 Click Perform Budget Optimisation.
As the budget is optimised, messages are displayed in the list box and the
processing status is shown below the list box.
The optimised work programme is displayed for the selected Sections.


Version 1.0
run.

For more information about reports, see Section 10.

Version 1.0
9 Strategy analysis
9.1 Key concepts

An HDM-4 Strategy Analysis allows you to generate a medium to long term
investment strategy for a selected set of Sections. For each Strategy Analysis,
you define optimisation criterion and budget constraints. Three optimisation
methods are available: Maximise NPV, Maximise dIRI, and Minimise Cost
for Target IRI. The investment alternative is a combination of Maintenance
and Improvement Standards that are applied to a Section.
When defining a Strategy Analysis you can either select Sections from a
predefined Road Network, or you can define a Network Matrix within the
Strategy. The Network Matrix option allows you to define a number of
representative Sections using aggregate parameters to represent the Road
Network being analysed.

A key feature of the Strategy window (shown below) is the five workflow
process of defining and running a Strategy analysis.

Version 1.0
button
Define Strategy General 1 Specify the Road Network that contains the Sections to be
Details analysed.
2 Specify the Vehicle Fleet that contains the Vehicle Types to be

analysed.
3 Specify general Strategy information such as the start year,

duration, and output currency.
4 Specify one of three available optimisation methods:
! Maximise NPV
! Maximise dIRI
! Minimise Cost for Target IRI
Select Sections 5 Select specific Sections for analysis.
Select Vehicles 6 Select specific Vehicle Types for analysis.
Define Normal 7 Define the traffic composition and expected growth for the
Traffic Section/Vehicle Type combinations.
Specify Standards Alternatives 8 Define the alternatives to be analysed. Some alternatives may
Assignments include diverted and/or generated traffic, exogenous benefits
and costs.
Generate Perform Run 9 Customise the run setup.

Programme
Specify the base alternative and select models to include in the
analysis.
10 Run the analysis. If vital information is missing, you are

informed, and must supply the necessary information so that
HDM-4 can complete the run. The time required to perform
the analysis depends on the complexity of the Strategy.
Work Programme 11 The generated work programme is displayed. If you want to

perform a budget optimisation, you can manually select works
that you do not want to be considered for optimisation.
Perform Budget Define Budget 12 Define budget periods and values.

Optimisation
13 Define the optimisation parameters.
14 Run the budget optimisation
Work Programme 15 The optimised work programme is displayed.
Generate Reports Select Reports 16 Generate reports and examine the results of the run.
HDM-4 places no restriction on the number of Strategies that can be created.

The only constraint is the amount of free disk space on your PC.
9.3 Creating Strategy analyses


Version 1.0
Workspace menu
2 Select Strategy and click OK.

The New Strategy Analysis dialog box is displayed.
3 Enter a name and select a Vehicle Fleet. Select one of the option buttons:
❏ Create new network matrix—allows you to create representative
Sections for the Strategy analysis.
❏ Use existing network—allows you to choose Sections from an
existing Road Network. Select the Road Network that you want to
use from the drop-down list.
4 Click OK.
The Strategy window is displayed. The Define Strategy Details
If you chose to select Sections from an existing Road Network, the
Select Sections tab-page is displayed (see Section 9.11).
If you chose to create a network matrix, the Network tab-page is

displayed (see Section 9.10).
5 Enter a description for the Strategy analysis.

6 Select an optimisation method:
Maximise NPV: Maximises the Net Present Value
when performing a budget
optimisation.

Version 1.0
Maximise dIRI: Maximises the change in roughness
when performing a budget
optimisation.
Minimise cost for target IRI: Maximises the cost for a target
roughness when performing a budget
optimisation.
list.
can now work through the tab-pages to select or create Sections, select
Vehicle Types, and define traffic.
When you have finished working with the Strategy analysis, click Save
before closing the window. If do not click Save, your changes will be lost.
The new Strategy analysis is added to the Workspace window. To view it,
see Section 9.5.
In step 2 above, if the Strategies folder or any existing Strategy analysis is
selected in the Workspace window, the Strategy item will be automatically
selected in the New HDM-4 Item dialog box, and you will only need to click
OK.
9.4 Creating Strategy analyses based on existing Strategy

analyses
If you want to create a Strategy analysis that is similar to an existing one, you
can copy the existing analysis and edit the copy as necessary.
1 In the Workspace window, select the Strategy analysis that you wish to
copy.
2 To copy the selected Strategy analysis, do one of the following:
❏ Click Copy
Workspace menu
The Copy Strategy dialog box is displayed.

3 The name of the Strategy analysis you wish to copy is shown. In the To
edit box, enter the name of the new Strategy analysis. This name must be
unique.
4 Click OK.
The new Strategy analysis is added to the Workspace window.
To work with the new copy, see Section 9.5 below
9.5 Viewing / editing Strategies

1 In the Workspace window, expand the Strategies folder.

Version 1.0
2 To open a Strategy analysis, do one of the following:
❏ Select a Strategy analysis and click Open
❏ Select a Strategy analysis. With the Workspace window active,
select Open from the Workspace menu.
❏ Double-click a Strategy analysis
3 When you have finished working with the Strategy analysis, click OK to
save any changes.
In steps 1 and 2 above, you can double-click the Strategy analysis in the
Workspace window.
9.6 Deleting Strategies

delete.
2 To delete the selected Strategy analysis, do one of the following:
❏ Click Delete
Workspace menu
Strategy analysis, or No to cancel the delete action.
The deleted Strategy analysis is no longer displayed in the Workspace
window.
! When you delete a Strategy analysis, associated network matrices,

alternatives, traffic and run settings are also deleted.
! You cannot undo a deletion.

Version 1.0
9.7 Renaming Strategies
rename.
2 To rename the selected Strategy analysis, do one of the following:
❏ Click Rename
Workspace menu
The Rename Strategy dialog box is displayed.

3 The current name of the Strategy analysis you wish to rename is shown.
In the To edit box, enter the new name for the Strategy analysis. The new
name must be unique.
4 Click OK.
The renamed Strategy analysis is displayed in the Workspace window.
To work with the renamed Strategy analysis, see Section 9.5.
9.8 Importing Strategy analyses

You can import Strategy analysis data that was originally exported from
HDM-4. This enables data exchange to take place between HDM-4 users.
menu.

You are informed if a Strategy analysis of the same name already exists.
You are prompted to rename the Strategy analysis that is being imported.
When you import a Strategy analysis, other items that are used by the Strategy
analysis are also imported (for example, Vehicle Fleet and Maintenance
Standards).

Version 1.0
9.9 Exporting Strategy analyses
When you export Strategy analysis data from HDM-4, a number of files are
created—a ‘.dbf’ file and several ‘.hdbf’ files. If you are using the export
facility to exchange data with other HDM-4 users, you must supply all the
files. When importing the Strategy analysis, only the ‘.dbf’ file is seen by the
importer.
1 In the Workspace window, expand the Strategies folder and select a
Strategy analysis.
2 Click Export.

4 Click OK.
You can only export one Strategy analysis at a time. If you want to export a
number of Strategy analyses one after the other, select a different folder for
each Strategy analysis. Do not rename any of the files.
9.10 Creating a network matrix

You can create representative Sections in a network matrix only if you
selected the Create new network matrix option button when creating this
Strategy analysis.
1 In the Workspace window, create or open a Strategy analysis.
2 Click the Network tab.

Version 1.0
3 Click Add new Section.
Do one of the following:

❏ to create a new Section select the based on aggregate data option
button and click OK.
❏ to create a Section that is similar to an existing one select the based
on existing Section option button. In the drop-down list, select the
existing Section that you wish to copy and click OK.
The Section dialog box is displayed. The data for the new Section is
a copy of the data in the existing Section.
The New Section from Aggregate Data dialog box is displayed.
9.11 Selecting Sections for a Strategy analysis

You can select Sections only if you selected the Use existing network option
button when creating this Strategy analysis.
In a new Strategy analysis, all Sections in the Road Network are initially
listed, but none are selected for the Strategy analysis.

Version 1.0
3 Click the buttons in the Include column to select / de-select Sections.
! Click Unselect All to exclude all Sections from the Strategy analysis.
The " is removed from all the Section lines.
! You can also select Sections by specifying selection criteria (see Section
7.10.1).
! Click View/Edit Section to display the Section dialog box.
! Click View/Edit Network to display the Road Network window.
9.12 Selecting Vehicle Types to include in a Strategy analysis

In a new Strategy analysis, all Vehicle Types in the Vehicle Fleet are
listed and marked with a " to show they are included in the Strategy
analysis.
3 Click the buttons in the Include column to de-select / select Vehicle

Types.

Version 1.0
! Unselected Vehicle Types are listed only when you tick the Show
unselected vehicles check box.

! Click View/Edit Vehicle to display the Vehicle Attributes dialog box.
! Click View/Edit Fleet to display the Vehicle Fleet window.

Define normal traffic for each Section you have selected for the Strategy
analysis. Specify the initial traffic mix and the traffic growth for the selected
Vehicle Types.
traffic.
The Sections that you specified for the Strategy analysis are listed. You
can now define traffic details for each Section.
that you selected for the Strategy analysis are listed. Motorised Vehicle
Types and non-motorised Vehicle Types are separated into two tab-
pages.

Version 1.0
Copy. Click the Section row where you want to paste the traffic details. Click
Paste.
9.14 Defining alternatives for a Strategy analysis



Version 1.0
9.14.1 and Section 9.14.2.
to the list.

1 In the Alternatives tab-page of the Strategy window, select the
alternative that you want to assign Maintenance Standards to.
Standards for. The Sections that you selected for the Strategy analysis are
listed.


Version 1.0
column.

1 In the Alternatives tab-page of the Strategy window, select the
alternative that you want to assign Improvement Standards to.
Standards for. The Sections that you selected for the Strategy analysis are
listed.


Version 1.0
Improvement Standard, see Section 7.13.5.
5 Click OK.
column.
! You can make alterations to the standards assigned to a Section. Select
the Section, select the assignment, and click Edit Assignment.
9.15 Generating a work programme

1 In the Workspace window, create or open a Strategy.
Before attempting to generate a work programme, you must define
Strategy details and specify alternatives.
2 Click the Generate Strategy workflow button.
3 Click Run Setup.
The Run Setup dialog box is displayed.

Version 1.0
4 Choose a base alternative from the drop-down list.
5 Tick the check boxes for the additional options you want to run. The
options are:
❏ Accident costs
❏ Energy balance
❏ Emissions
the requirements of the particular Strategy. Typically, you should only
include those models that you require, as each additional model will
then the acceleration effects indicators are calculated for the current
testers.
6 Click OK.

Version 1.0
7 Click Start.
As the Strategy is analysed, messages are displayed in the list box and
the status of the analysis is shown below the list box.
The work programme is displayed.
9 You can select the work programme for a Section and exclude it from
budget optimisation. Click the Section row and click Manual
Assignment. The row is shown in italics.
run continues.
! At the end of the Strategy analysis, you can view a complete warning
message. Select the message in the list box and click View.
! You can abort a Strategy analysis at any time by clicking Abort.
9.16 Optimisation using budget constraints

You must generate a work programme before attempting a budget
optimisation.
1 In the Workspace window, create or open a Strategy.
2 Click the Perform Budget Optimisation workflow button.
The Define Budget tab-page is displayed. A budget period equivalent to
the Strategy duration is listed in the spreadsheet.

Version 1.0
3 Define budget periods and values
4 Click Optimisation Setup.
The Optimisation Setup dialog box is displayed.
5 Enter a minimum incremental value and a value for the efficiency

frontier zone.
6 Click OK.
7 Click Perform Budget Optimisation.
As the budget is optimised, messages are displayed in the list box and the
processing status is shown below the list box.
The optimised work programme is displayed for the selected Sections.

run.

Version 1.0
For more information about reports, see Section 10 below.

Version 1.0
10.1 Key concepts

HDM-4 provides sophisticated reporting facilities for the viewing of input and
output data. The reporting facilities were developed using Seagate Crystal
Reports. A number of default report templates are provided. These can be
added to by downloading new templates from the ISOHDM web-site
(www.bham.ac.uk/isohdm). Alternatively, if you have access to the Crystal
Reports Designer you can create your own report templates and add them to
HDM-4 (see Section 10.4).
Previous sections of this guide have dealt with how to display reports:
! Generating reports for Projects - Section 7.17
! Generating reports for Programmes - Section 8.17
! Generating reports for Strategy analyses - Section 9.17
This section deals with making the most of reports once they have been
generated. Every HDM-4 report is displayed in a separate report window, an
example of which is shown below:
Typically, an HDM-4 report will be larger than the window in which it is

displayed. To move about the report you can use the scroll bars. Alternatively
you can resize the window so that more of the report is displayed by dragging
the resize box in the bottom right-hand corner of the window.
In addition, you can use the navigation controls which appear on the toolbar.
These operate as follows:
Go to the first page of the report

Version 1.0
Go to the previous page of the report
Go to the next page of the report
Go to the last page of the report
Stop – some reports may take a long time to generate due to

their complexity and/or the volume of output data resulting
from the last run. Click this button to abort the generation of
the report.
The other controls that appear on the toolbar operate as follows:
Print the report
Export the report to a file
Change the size of the report to the selected setting
The Print and Export controls are explained in the following sections.
10.2 Printing reports

At the end of a Project, Programme or Strategy analysis, you can generate a
report, which is displayed on the screen. To print the report, perform the
following steps:
1 Click to print the report.

The Print dialog box is displayed.
2 Specify whether you want to print all the pages in the report, or a range.
You can print several copies of the report and have them collated.
3 Click OK.
10.3 Exporting report data to other applications

You can save generated reports to disk to print later or to export to another
software application.
1 When you have generated a report, click to export it to a file.

The Export dialog box is displayed.

Version 1.0
2 Select a format and destination for the export file.
If you intend to print the report file later, select one of the following
formats:
❏ Paginated text
❏ Rich text format
❏ Text
❏ Word for Windows document
If you intend to import the report file into another software application,
select one of the following formats:
❏ Character separated values
❏ Comma separated values (CSV)
❏ Excel
❏ Lotus
❏ Tab separated text
❏ Tab separated values
3 You are prompted to enter information, depending on the format that you
selected in step 2. The final step is to specify a name and destination for
the export file.
The default destination folder is c:\Windows\Temp, but you can select

any folder. HDM-4 also enters a default file name, but, again, you can
specify any name.
4 Click Save.
The report file is saved and can be printed or imported into other
software applications.
When you export a report to a file format other than Seagate Crystal Reports
format (.RPT), you may lose some or all of the formatting that appears in your
report. However, the program attempts to preserve as much formatting as the
export format allows.
10.4 Adding user defined reports

You can define additional reports for HDM-4 using Crystal Reports (version 7
and above). Save the report template in the folder c:\Program Files\HDM-
4\Reports. (If you changed the default folder at installation, store the report

Version 1.0
template in the reports subfolder of the HDM-4 installation folder). To make
the report available from within HDM-4, perform the following steps:
1 In the Workspace window, open any Project, Programme or Strategy
analysis.
2 Run an analysis and click the Generate Reports workflow button (You
cannot click the Generate Reports workflow button until you have run
an analysis).
3 Click Add Report.
The Add Report dialog box is displayed.
4 Select the report template that you want to add, and the folder in which
you would like it to appear. Enter a name for the report.
5 Click OK.
To remove a user-defined report, select it and click Remove Report.

Version 1.0
11.1 Traffic Flow Patterns
11.1.1 Key concepts

Traffic Flow Patterns are used to represent the varying traffic intensities that
occur on roads throughout the day. Different road Sections exhibit different
traffic flow patterns according to their use. Examples of Traffic Flow Patterns
are:
! Commuter
! Seasonal
! Inter-city
Traffic Flow Patterns are defined as a set of flow periods. A flow period
represents the hours of the day (over the course of a year) with the same
traffic flow.
For each flow period you should specify:
! Total hours per year (HRYR) that the period occupies, and
! The amount of yearly traffic occurring during the period, as either:
❏ The proportion of AADT (HV) that occurs during the period, or
❏ The percentage of AADT (PCNADT) that occurs during this period
A number of default road uses are defined in HDM-4. Each has an associated
set of flow period data. Changing the selection of road use for a Traffic Flow
Pattern causes the corresponding flow period default data to be adopted. To
reset the flow period data to the default values for the currently selected road
use, click Defaults.

Version 1.0
11.1.2 Creating Traffic Flow Patterns
Workspace menu
2 Select Traffic Flow Pattern and click OK.

The Traffic Flow Pattern dialog box is displayed.
3 Enter a name for the new Traffic Flow Pattern.

4 Add flow periods which define the Traffic Flow Pattern. Click Add New
Period to add flow periods, and enter flow distribution data for each.
5 Click OK to save the Traffic Flow Pattern and close the window.
The new Traffic Flow Pattern is added to the Workspace window. To
view it, see Section 11.1.4.
! In step 2 above, if the Traffic Flow Patterns folder or any existing
Traffic Flow Pattern is selected in the Workspace window, the Traffic
Flow Pattern item will be automatically selected in the New HDM-4 Item
dialog box, and you will only need to click OK.
! Using more flow periods improves accuracy but slows down analysis
time.

Version 1.0
11.1.3 Creating Traffic Flow Patterns based on existing Traffic Flow Patterns
If you want to create a Traffic Flow Pattern that is similar to an existing one,
you can copy the existing pattern and edit the copy as necessary.
1 In the Workspace window, expand the Configuration folder.
2 Expand the Traffic Flow Patterns folder.
3 Select the Traffic Flow Pattern that you wish to copy.
4 To copy the selected Traffic Flow Pattern, do one of the following:
Workspace menu
The Copy Traffic Flow Pattern dialog box is displayed.

5 The name of the Traffic Flow Pattern you wish to copy is shown. In the
To edit box, enter the name of the new pattern. This name must be
unique.
6 Click OK.
The new Traffic Flow Pattern is added to the Workspace window.
To view the new copy, see Section 11.1.4 below.
11.1.4 Viewing / editing Traffic Flow Patterns

3 To open a Traffic Flow Pattern, do one of the following:
❏ Select a Traffic Flow Pattern and click Open
❏ Select a Traffic Flow Pattern. With the Workspace window active,
select Open from the Workspace menu
❏ Double-click a Traffic Flow Pattern

Version 1.0
The Traffic Flow Pattern dialog box is displayed.
4 Edit the flow distribution data as necessary.

5 When you have finished working with the Traffic Flow Pattern, click OK
to save any changes and close the window.
11.1.5 Deleting Traffic Flow Patterns

3 Select the Traffic Flow Pattern that you wish to delete.
4 To delete the selected Traffic Flow Pattern, do one of the following:
❏ Click Delete
Workspace menu
You are prompted to confirm the deletion. Click Yes to delete the Traffic
Flow Pattern, or No to cancel the delete action.
If you click Yes, the deleted Traffic Flow Pattern is no longer displayed
in the Workspace window.
! When you delete a Traffic Flow Pattern, all corresponding data are also
deleted.
! There may be Sections which use the Traffic Flow Pattern to be deleted.
These Sections will be affected by the deletion. You will be informed if
this is the case.
11.1.6 Renaming Traffic Flow Patterns

3 Select the Traffic Flow Pattern that you wish to rename.

Version 1.0
4 To rename the selected Traffic Flow Pattern, do one of the following:
❏ Click Rename
Workspace menu
The Rename Traffic Flow Pattern dialog box is displayed.

5 The current name of the Traffic Flow Pattern you wish to rename is
shown. In the To edit box, enter the new name for the Traffic Flow
Pattern. The new name must be unique.
6 Click OK.
The renamed Traffic Flow Pattern is displayed in the Workspace
window.
To work with the renamed Traffic Flow Pattern, see Section 11.1.4.
11.1.7 Importing Traffic Flow Patterns

You can import Traffic Flow Patterns that were originally exported from
menu.

3 Select one of the following files:
❏ FLOWPERIODS.DBF
❏ TRAFFICFLOWPERIODS.DBF
4 Click Open.
! In step 3, all the Traffic Flow Patterns are imported, irrespective of the
file you select
! In step 3, above, you are informed if a Traffic Flow Pattern of the same
name already exists. You are prompted to rename the Traffic Flow
Pattern that is being imported.

Version 1.0
11.1.8 Exporting Traffic Flow Patterns
When you export Traffic Flow Patterns from HDM-4, several .dbf and .hdbf
files are created. If you are using the export facility to exchange data with
other HDM-4 users, you must supply all the files.
1 In the Workspace window, select Traffic Flow Patterns or any of the
folder contents.
2 Click Export.

4 Click OK.
All the Traffic Flow Patterns are exported irrespective of the item that you
select in the Traffic Flow Patterns folder.
11.2 Speed Flow Types
11.2.1 Key concepts

The primary function of HDM-4 Speed Flow Types is to represent the
capacity characteristics of different road types. Examples of Speed Flow
Types are:
! Single Lane Road
! Intermediate Road
! Two Lane Road
! Four Lane Road
Capacity characteristics are defined in terms of the various parameters (such
as ultimate capacity and free flow capacity) that constitute the HDM-4 speed
flow curve model.

Version 1.0
HDM-4 includes a set of default Road Types. Each Road Type has associated
default values for the speed flow parameters. Selecting a Road Type from the
drop-down list causes the speed flow parameters to be overwritten with the
appropriate defaults. To reset the speed flow parameters to the defaults for the
currently selected Road Type, click Defaults.
Speed Flow Types also define accident rates and speed model calibration
factors for the road types they represent.
11.2.2 Creating Speed Flow Types

Workspace menu
2 Select Speed Flow Type and click OK.

The Speed Flow Type dialog box is displayed.

Version 1.0
3 Enter a name for the new Speed Flow Type.
4 Enter capacity data, accident rates, and speed factors as required.
5 Click OK to save the Speed Flow Type and close the dialog box.
The new Speed Flow Type is added to the Workspace window. To view
it, see Section 11.2.4.
In step 2 above, if the Speed Flow Types folder or any existing Speed Flow
Type is selected in the Workspace window, the Speed Flow Type item will
be automatically selected in the New HDM-4 Item dialog box, and you will
only need to click OK.
11.2.3 Creating Speed Flow Types based on existing Speed Flow Types
If you want to create a Speed Flow Type that is similar to an existing one, you
can copy the existing type and edit the copy as necessary.
2 Expand the Speed Flow Types folder.
3 Select the Speed Flow Type that you wish to copy.
4 To copy the selected Speed Flow Type, do one of the following:
Workspace menu
The Copy Speed Flow Type dialog box is displayed.

5 The name of the Speed Flow Type you wish to copy is shown. In the To
edit box, enter the name of the new Speed Flow Type. This name must be
unique.
6 Click OK.

Version 1.0
The new Speed Flow Type is added to the Workspace window.
11.2.4 Viewing / editing Speed Flow Types

3 To open a Speed Flow Type, do one of the following:
❏ Select a Speed Flow Type and click Open
❏ Select a Speed Flow Type. With the Workspace window active,
select Open from the Workspace menu
❏ Double-click a Speed Flow Type
The Speed Flow Type dialog box is displayed.
4 Edit the fields as necessary.

5 When you have finished working with the Speed Flow Type, click OK to
save any changes and close the dialog box.
11.2.5 Deleting Speed Flow Types

3 Select the Speed Flow Type that you wish to delete.
4 To delete the selected Speed Flow Type, do one of the following:
❏ Click Delete

Version 1.0
Workspace menu
You are prompted to confirm the deletion. Click Yes to delete the Speed
Flow Type, or No to cancel the delete action.
If you click Yes, the deleted Speed Flow Type is no longer displayed in
the Workspace window.
! When you delete a Speed Flow Type, all corresponding data are also
deleted.
! There may be Sections which use the Speed Flow Type to be deleted.
These Sections will be affected by the deletion. You will be informed if
this is the case.
11.2.6 Renaming Speed Flow Types

3 Select the Speed Flow Type that you wish to rename.
4 To rename the selected Speed Flow Type, do one of the following:
❏ Click Rename
Workspace menu
The Rename Speed Flow Type dialog box is displayed.

5 The current name of the Speed Flow Type you wish to rename is shown.
In the To edit box, enter the new name for the Speed Flow Type. The
new name must be unique.
6 Click OK.
The renamed Speed Flow Type is displayed in the Workspace window.
To work with the renamed Speed Flow Type, see Section 11.2.4.
11.2.7 Importing Speed Flow Types

You can import Speed Flow Types that were originally exported from
menu.

Version 1.0
You are informed if a Speed Flow Type of the same name already exists.
You must then decide whether to overwrite the existing Speed Flow
Type, or to rename the Speed Flow Type that is being imported, or to
cancel the import operation.
11.2.8 Exporting Speed Flow Types

You can export Speed Flow Types as a .dbf file which can then be imported
by other HDM-4 users.
1 In the Workspace window, expand the Configuration folder and click
Speed Flow Types.
2 Click Export.
The Export Speed Flow Type As dialog box is displayed.

5 Click Save.
11.3 Climate Zones
11.3.1 Key concepts

Climate Zones are used to represent the climatic conditions found in different
parts of a Road Network. The data items that represent these climatic
conditions affect pavement deterioration. Climate Zone data are divided into
two categories:
! Moisture

Version 1.0
! Temperature
For each of the two categories, aggregate level parameters are provided.
These are Temperature Classification and Moisture Classification. HDM-4
supports a fixed set of Temperature Classifications, each of which has an
associated set of defaults for the temperature parameters. Changing the
Temperature Classification selected for a Climate Zone causes the
corresponding set of temperature data to be adopted.
Similarly, HDM-4 supports a fixed set of Moisture Classifications, each of
which has an associated set of defaults for the moisture parameters. Changing
the Moisture Classification currently selected for a Climate Zone causes the
corresponding set of moisture defaults to be adopted.
To reset the default values for Temperature and Moisture Classifications,
click Defaults.
11.3.2 Creating Climate Zones

Workspace menu
2 Select Climate Zone and click OK.

The Climate Zone dialog box is displayed.

Version 1.0
3 Enter a name for the new Climate Zone.
4 Click OK to save the Climate Zone and close the dialog box.
The new Climate Zone is added to the Workspace window. To view it,
see Section 11.3.4.
In step 2 above, if the Climate Zones folder or any existing Climate Zone is
selected in the Workspace window, the Climate Zone item will be
need to click OK.
11.3.3 Creating Climate Zones based on existing Climate Zones

If you want to create a Climate Zone that is similar to an existing one, you can
copy the existing Climate Zone and edit the copy as necessary.
2 Expand the Climate Zones folder.
3 Select the Climate Zone that you wish to copy.
4 To copy the selected Climate Zone, do one of the following:
Workspace menu
The Copy Climate Zone dialog box is displayed.

5 The name of the Climate Zone you wish to copy is shown. In the To edit
box, enter the name of the new Climate Zone. This name must be unique.
6 Click OK.
The new Climate Zone is added to the Workspace window.

Version 1.0
11.3.4 Viewing / editing Climate Zones
3 To open a Climate Zone, do one of the following:
❏ Select a Climate Zone and click Open
❏ Select a Climate Zone. With the Workspace window active, select
❏ Double-click a Climate Zone

The Climate Zone dialog box is displayed.
4 Edit the fields as necessary.

5 When you have finished working with the Climate Zone, click OK to
save any changes and close the dialog box.
In steps 2 and 3 above, you can alternatively, double-click the Climate Zone
in the Workspace window.
11.3.5 Deleting Climate Zones

3 Select the Climate Zone that you wish to delete.
4 To delete the selected Climate Zone, do one of the following:
❏ Click Delete
Workspace menu
Climate Zone, or No to cancel the delete action.

Version 1.0
If you click Yes, the deleted Climate Zone is no longer displayed in the
Workspace window.
! When you delete a Climate Zone, all corresponding data are also deleted.
! There may be Sections which use the Climate Zone to be deleted. These
Sections will be affected by the deletion. You will be informed if this is
the case.
11.3.6 Renaming Climate Zones

3 Select the Climate Zone that you wish to rename.
4 To rename the selected Climate Zone, do one of the following:
❏ Click Rename
Workspace menu
The Rename Climate Zone dialog box is displayed.

5 The current name of the Climate Zone you wish to rename is shown. In
the To edit box, enter the new name for the Climate Zone. The new name
must be unique.
6 Click OK.
The renamed Climate Zone is displayed in the Workspace window.
To work with the renamed Climate Zone, see Section 11.3.4.
11.3.7 Importing Climate Zones

You can import Climate Zones that were originally exported from HDM-4.
menu.

Version 1.0
You are informed if a Climate Zone of the same name already exists.
You must then decide whether to overwrite the existing Climate Zone, or
to rename the Climate Zone that is being imported, or to cancel the
import operation.
11.3.8 Exporting Climate Zones

You can export Climate Zones as a .dbf file which can then be imported by
other HDM-4 users.
Climate Zones.
2 Click Export.
The Export Climate Zone As dialog box is displayed.

5 Click Save.
11.4 Currencies
11.4.1 Key concepts

HDM-4 maintains a list of Currencies. These are used to specify the
following:
! Vehicle Fleet unit costs
! Default works unit costs
! Project analysis output
! Programme analysis output
! Strategy analysis output
You can add and delete Currencies.
11.4.2 Adding a Currency

2 To access the Currency details, do one of the following:

Version 1.0
❏ Select Currencies and click Open
❏ Select Currencies. With the Workspace window active, select Open
❏ Double-click Currencies
The Currencies dialog box is displayed.
3 Click Add.
A row is added to the table.
4 Enter a description and symbol for the Currency. Select a symbol
position from the drop-down list.
5 Click OK to save the Currency definition and close the dialog box.
11.4.3 Deleting a Currency

2 To access the Currency details, do one of the following:
❏ Select Currencies and click Open
❏ Select Currencies. With the Workspace window active, select Open
❏ Double-click Currencies
The Currencies dialog box is displayed.
3 Select the Currency you wish to delete.
4 Click Delete.
One of the following will happen:
❏ If the Currency has not yet been saved, it is deleted immediately
❏ If the Currency has been saved and is not being used elsewhere in
the system, you are prompted to confirm the deletion. Click Yes to
delete the Currency, or No to cancel the delete action.
❏ If the Currency is being used elsewhere in the system, you are
informed of the places it is being used and the consequences of
deleting it. You are prompted to confirm the deletion. Click Yes to
delete the Currency, or No to cancel the delete action.
5 Click OK to save any changes and close the dialog box.
If you delete a Currency that is in use, the system sets the respective Currency
field(s) to <none> or <undefined>. You must specify another Currency before

Version 1.0
continuing with the analysis. The system informs you that the Currency is
missing if you try to perform an analysis without specifying a Currency.
11.4.4 Importing Currencies

You can import Currencies that were originally exported from HDM-4. This
enables data exchange to take place between HDM-4 users.
menu.

You are informed if a Currency of the same name already exists. You
must then decide whether to overwrite the existing Currency, or to
rename the Currency that is being imported, or to cancel the import
operation.
11.4.5 Exporting Currencies

You can export Currencies as a .dbf file which can then be imported by other
HDM-4 users.
Currencies.
2 Click Export.
The Export Currency As dialog box is displayed.

5 Click Save.

Version 1.0
11.5 Section aggregate data and tables
Values for the following aggregate parameters (from the New Section from
Aggregate Data dialog box) are defined in Section Aggregate Data in the
Configuration folder:
! Traffic volume
! Road class
! Geometry class
! Construction quality
! Ride quality
! Surface condition
! Surface texture
The parameter values that you define (for example good, fair, bad) correspond
to detailed values for related parameters in the Section Aggregate Tables
in the Configuration folder.
Each HDM-4 system can, therefore, have a different definition of what
constitutes good, bad or fair to suit the local conditions.
When you select a parameter value in the New Section from Aggregate Data
dialog box, the corresponding detailed values from the Section Aggregate
Tables are used when you run an analysis.
11.5.1 Adding values for an aggregate parameter

2 To access the Section Aggregate Data, do one of the following:
❏ Select Section Aggregate Data and click Open
❏ Select Section Aggregate Data. With the Workspace window
❏ Double-click Section Aggregate Data
The Road Network Aggregate Parameters dialog box is displayed.
3 Click one of the Aggregate Parameter option buttons.

Version 1.0
The variables for that parameter are displayed in the list box.
4 Click Add.
<New HDM-4 Item> is displayed in the list box.
5 Overwrite <New HDM-4 Item> with the name of the new value.
6 Click Edit Related Tables to access the Section Aggregate Tables,
where you can define detailed values for the selected aggregate
parameter.
To save the changes you have made, click OK in both windows.
! You can also delete a parameter value. Select the value and click Delete.
! You can rename a parameter value. Select the value and click Rename.
! You can also access the Section Aggregate Tables by expanding the
Configuration folder and double-clicking Section Aggregate Tables.
All the tables are displayed in tab-pages in the resulting dialog box. The
Bituminous layers table can only be accessed in this way.
11.5.2 Importing Section Aggregate Data and tables

You can import Section Aggregate Data that were originally exported from
menu.

3 Select the file SECAGGCONFIG.DBF.

Version 1.0
4 Click Open.
In step 3, selecting the file SECAGGCONFIG.DBF imports both Section

Aggregate Data and tables.
11.5.3 Exporting Section Aggregate Data and tables

When you export Section Aggregate Data from HDM-4, several files are
created— one ‘.dbf’ file and several ‘.hdbf’ files. If you are using the export
facility to exchange data with other HDM-4 users, you must supply all the
files.
1 In the Workspace window, select Section Aggregate Data or Section
Aggregate Tables.
2 Click Export.

4 Click OK.
In step 1, both the Section Aggregate Data and the Section Aggregate Tables
are exported irrespective of which you select.

Version 1.0
12.1 Working with multiple databases

By default, all HDM-4 input data items (e.g. Road Networks, Projects, Work
Standards, etc.) are stored in a single database which is set-up during
installation. The database is physically stored as a number of files in the
following sub-directories of the HDM-4 installation directory:
! hdmdbase - contains the files objects.dat and objects.idx in which the
actual data is stored. These files are called repository files.
! hdmdict - contains the files _objects.dat and _objects.idx which describe
the database structure. These files are called dictionary files.
You cannot view the contents of these files unless you have the necessary
software development tools.
HDM-4 allows you to create multiple databases, and to partition your data
accordingly. A typical scenario might involve a consultant who works on a
number of projects and who prefers to separate his data by using separate
databases. HDM-4 supports this functionality by supporting multiple
Workspaces. A Workspace can be regarded as a separate work-area. Each
Workspace is associated with its own separate database. Each separate
database consists of a separate set of repository files (objects.dat and
objects.idx – as described above). The single set of dictionary files in the
\hdmdict sub-directory are shared between all separate databases. Functions
are provided to allow the user to:
! Open an existing Workspace
! Create a new / blank Workspace
! Create a copy of an existing Workspace
! Delete a Workspace
These options are listed on the Workspace menu. Note that only one
Workspace (and thus database) may be active at any time. Therefore when
you open an existing Workspace, the Workspace that is currently active is
closed.
Workspaces are intended to be self-contained. At present it is not possible to
directly move/copy one HDM-4 data object (e.g. a Vehicle Fleet) from one
Workspace to another. For the time being, this can only be achieved by the
use of the import/export facilities (see relevant Sections).
12.2 Backing up your HDM-4 data

It is good practice to regularly back up any data you create in HDM-4. This
will protect you from problems such as file corruption, accidental deletion,
hardware failure, etc. To be sure of retaining all your data there are three sets
of files you should backup:

Version 1.0
! Dictionary files – the HDM-4 dictionary files are stored in the \hdmdict
sub-directory of the installation directory. These two files (_objects.dat
and _objects.idx) store information about the structure of the HDM-4
input data stored in repository files (see below). If you are using a new
version of HDM-4 or an upgrade, you do not need to backup your
dictionary files as you will be able to retrieve them from your CD.
However, if you are currently using an update (for which you may not
have a CD) you should backup the dictionary files following installation.
! Databases – these store the HDM-4 input data. At any one time you
may have several of these in use (see Section 12.1 above). You should
backup all databases that you need to retain. Each database consists of
two repository files objects.dat and objects.idx.
! Run-data files – these are stored at the end of each HDM-4 analysis run
in a user specified directory. Each set of run-data consists of a number of
dBase compatible files (with the extensions ‘.dbf’ and ‘.cdx’). As the
total size of each set of files can range from around 10MB for a Project
analysis to over 100MB for a Programme or Strategy analysis, you will
probably only want to backup those run-data sets which you need to
keep.
From the above it can be seen that in the process of using HDM-4
(particularly if you create several databases and retain several sets of run-
data), you may use a significant amount of hard disk space. To reduce this
overhead, you should regularly review your storage and remove any obsolete
databases and/or run-data.
12.3 Improving analysis performance

When running larger analyses you may find that the run-time increases
significantly. This is understandable if you consider the complexity of the
analysis that is being performed: the level of detail of the various models; the
number of sections, vehicles, alternatives, and years, etc. being analysed.
There are a number of things you can do to improve the analysis speed –
particularly if you are willing to reduce the accuracy of the results. The
following is a list of changes that can be made to the software / analysis
configuration in order to improve analysis speed:
! close other applications – before starting the analysis run, close down
any other Windows applications. Each additional application uses
Windows resources, reducing the resources available to HDM-4, and
therefore increasing the analysis time.
! improve specification of PC – since the problem is one of limited
resources, increase the power of your hardware. Use a PC with more
RAM, a faster processor, and more free disk space.
! disable log file – make sure that the log-file is disabled by de-selecting
the option on the Run Setup tab-page. The log-file is very detailed. It is
primarily intended for modellers and testers, and may not be useful for
your purposes. It is not unusual for the log-file to exceed 12MB, so you
can imagine the time required to generate it. Take a look at the log-file
and you will appreciate the amount of work involved.

Version 1.0
! disable unnecessary models – turn off those models which you don’t
need in your analysis. The controls on the Run Setup tab-page allow you
to disable the calculation of emissions, energy balance, acceleration
effects, and economic indicators.
! reduce run-data export size – a significant part of run-time is spent
saving run-data to the hard-disk. If you are not interested in the very
detailed results, you can instruct HDM-4 not to save them. This detailed
data is the annual road user effects by vehicle, and the annual road user
effects by vehicle and by flow period. These are stored in the files
ANNUALVEHICLE.DBF and ANNUALVEHPERIOD.DBF
respectively. If these aren’t required then select the appropriate options in
the Run Data Export Detail section of the Run Setup tab-page.
! reduce number of flow periods – check the Traffic Flow Patterns that
you are using. The defaults have five flow periods each. Is it possible to
reduce the number of flow periods? Doing so may affect the accuracy of
your results, but will increase analysis speed.
! reduce size of study – this is probably the most drastic way of
improving analysis speed. Re-examine your study. Do you need quite so
many Vehicle Types? Could the fleet mix you are modelling be
represented using fewer Vehicle Types? How about the number of
Sections? Could you replace your two hundred sections with eight or so
representative sections (in the fashion of strategy analysis)? Also
consider the number of years, and the number of options. Obviously
changing these aspects of your study will affect the accuracy of the
results. However, you may be in the early stages of putting together your
study, and initially, detailed results may not be required.
12.4 Adding an HDM-4 shortcut to your desktop

To access HDM-4 more quickly, add a shortcut to your desktop.
1 Right-click the Windows desktop.
2 From the menu select New & Shortcut as shown below:
3 The Create a Shortcut dialog box is displayed.

4 Enter HDM-4.EXE and the full path of where the software is installed.
For example:
c:\Program Files\HDM-4\hdm4.exe

Version 1.0
5 Click Next, and then enter HDM-4 as the name for the shortcut.
6 Click Finish.
The HDM-4 icon is added to your Windows desktop. In future, to run HDM-4
you can simply double-click the HDM-4 desktop icon.

Version 1.0
A4 back covers 10/12/99 4:48 pm Page 3
volume three
ISBN: 2-84060-059-5

ISBN: 2-84060-060-9

user interface.
ISBN: 2-84060-061-7

ISBN: 2-84060-062-5

ISBN-2-84060-102-8

ISBN-2-84060-103-6


volume four
Analytical Framework and
Model Descriptions
J. B. Odoki
Henry G. R. Kerali Association
mondiale
de la Route
Wo r l d R o a d
Association
4
Part A
About This Manual

This Version 1.0 edition of the Analytical Framework and Model Descriptions describes the
analytical framework and the technical relationships of objects within the HDM-4 model. It
contains very comprehensive reference material describing, in detail, the characteristics of the
modelling and strategy incorporated in HDM-4.
It is to be used by specialists or experts whose task is to carry out a detailed study for a road
management organisation. The Analytical Framework and Model Descriptions is one of five
manuals comprising the suite of HDM-4 documentation (see Figure 1).
Overview
Overview of
of HDM-4
HDM-4
Volume
Volume 11
Applications
Applications Guide
Guide Software
Software User
User Guide
Guide
Volume
Volume 22 Volume
Volume 33
Analytical
Framework andand
Model Descriptions
Model Descriptions
Volume
Volume 44
A
A Guide
Guide to
to Calibration
Calibration and
and Adaptation
Adaptation
Volume
Volume 55


Analytical Framework and Model Descriptions i

Version 1.0
ABOUT THIS MANUAL

HDM-4), to allow for local conditions existing in different countries.
Notes:
1 Volumes 1, 2 and 3 are designed for the general user
2 Volumes 4 and 5 are only to be used by experts who wish to obtain low level technical
detail
Structure of the ‘Analytical Framework and Model

Descriptions’
The information in the Analytical Framework and Model Descriptions is structured in nine
parts (see Figure 2).
Analytical Framework and Model Descriptions ii

Version 1.0
ABOUT THIS MANUAL
Analytical
AnalyticalFramework
Framework
and
andModel
Model
Descriptions
Descriptions
Introduction
Introduction
Part
PartAA
Traffic
Traffic
Part
PartBB
RD
RDModel
Model WE
WEModel
Model RUE
RUEModel
Model SEE
SEEModel
Model
Part
PartCC Part
PartDD Part
PartEE Part
PartFF
Economic
EconomicAnalysis
Analysis
RD = Road Deterioration Part
PartGG
Nomenclature
Nomenclature
WE = Road Works Effects Part
PartHH
RUE = Road User Effects Glossary

Glossary
Part
PartI I
SEE = Social and Environmental Effects
Figure 2 The Analytical Framework and Model Descriptions documentation
The nine parts in the Analytical Framework and Model Descriptions contain the following
information:
Provides an overview of the Analytical Framework and Model Descriptions.

! Part B - Traffic
Describes the traffic characteristics used in HDM-4, and provides details of the traffic
data that is required.
! Part C - Road Deterioration Models
❏ C1 Modelling Concepts and Approach

Describes the types of pavements modelled in HDM-4, and the possible combinations
of pavement surface types and base types. It discusses the key variables that affect
deterioration, which include climate and environment effects, traffic, and pavement
history.
❏ C2 Bituminous Pavements
Describes the specifications used in the Bituminous Pavements (Road Deterioration)
model. It provides a detailed discussion of pavement materials, surface behaviour,
surface distress, pavement strength, and construction quality.
Analytical Framework and Model Descriptions iii

Version 1.0
ABOUT THIS MANUAL
❏ C3 Concrete Pavements
Describes the specifications used to model rigid (or Portland) cement concrete
pavements. This includes details of the different types of rigid concrete pavement
construction that are modelled in HDM-4.
❏ C4 Unsealed Roads
Describes the specifications used in the unsealed Road Deterioration models. It
provides an overview of the HDM-4 Road Deterioration modelling logic, and the
relationships and default coefficient values for each of the distresses modelled.
! Part D - Road Works Effects
❏ D1 Types of Works
Describes the types of road works and their effects on road pavements, including the
calculation of physical quantities of road works and the corresponding costs.
The three types of pavements considered are:
1 Bituminous pavements (see Chapter D2)
2 Concrete pavements (see Chapter D3)
3 Unsealed roads (see Chapter D4)
❏ D2 Bituminous Pavements
Describes the detailed modelling of Road Works Effects for bituminous pavements.
This includes methods of defining works activities and intervention criteria, the
calculation of physical quantities of road works and their corresponding costs, and
the effects of road works on pavement characteristics and road users. Road works
are grouped into the following classes:
" Routine maintenance
" Periodic maintenance
" Special works
" Improvement works
" Construction works
❏ D3 Concrete Pavements
Describes the detailed modelling of Road Works Effects for concrete pavements.
This includes methods of defining works activities (maintenance and rehabilitation)
and intervention criteria, and classification of the road works. The following three
types of concrete pavements are analysed:
1 Jointed Plain Concrete Pavements
2 Jointed Reinforced Concrete Pavements
3 Continuously Reinforced Concrete Pavements
The following classes of road works are performed:
" Routine maintenance
" Restorations
Analytical Framework and Model Descriptions iv

Version 1.0
ABOUT THIS MANUAL
" Overlays
" Reconstruction
❏ D4 Unsealed Roads
Describes the detailed modelling of road works effects for unsealed roads, including
the methods of defining works activities and intervention criteria, the calculation of
physical quantities of road works and their corresponding costs, and the modelling
of the effects of road works on pavement characteristics and road users. Unsealed
road works are grouped into the following classes:
" Maintenance
" Improvement works
" Construction works
! Part E - Road User Effects
❏ E1 Introduction
Provides an overview of the implementation of Road User Effects (RUE) models for
calculating motorised vehicle operating costs and travel time, including an overview
of the HDM-4 representative vehicle framework and the default representative
vehicle types.
❏ E2 Vehicle Speeds and Operating Costs
Describes the implementation of Road User Effects (RUE) models for calculating
motorised vehicle speeds, operating costs and travel time. It provides an overview of
the modelling concepts and logic, the relationships used, and the default parameter
values for each of the RUE components.
❏ E3 Non-Motorised Transport
Discusses the use of non-motorised modes of transport and their effect on the
motorised transport. The impact of non-motorised transport on other road users and
road characteristics is included.
❏ E4 Road Safety
Describes the implementation of road safety models through a series of look-up
tables of accident rates with user defined accident costs.
! Part F - Social and Environmental Effects
❏ F1 Introduction
Gives an overview of the modelling logic used to quantify the energy consumed by
both motorised and non-motorised vehicles, together with the emissions produced
by motorised vehicles.
❏ F2 Energy Balance Analysis
Describes the energy implications of alternative transport projects and policies. The
models estimate separately the global and national consumption of non-renewable
energy by motorised vehicles, the consumption of renewable energy by non-
motorised vehicles, and energy use in roadworks.
Analytical Framework and Model Descriptions v

Version 1.0
ABOUT THIS MANUAL
❏ F3 Vehicle Emissions
Describes the models used to quantify vehicle emissions. The document presents the
emissions models together with default parameters for the 16 standard vehicle types.
Details of the input data and the output reports are also given.
! Part G - Economic Analysis
Describes how the annual costs streams calculated by the HDM-4 components are
compared to determine the benefits and costs associated with a road investment. Standard
discounting methods are then applied to calculate key economic indicators such as net
present values, internal rates of return and benefit/cost ratios. Investment optimisation
methods are also described.
! Part H Nomenclature
Provides details of the units and numbering system used within the HDM-4 suite of
documents.
! Part I Glossary
Provides descriptions of terminology used throughout the HDM-4 suite of

documentation.
ISOHDM Products
Customer contact
HDM-4. Comments on the Analytical Framework and Model Descriptions should be sent to
the following address:
Telephone: +44 - 121 - 414 6717 (5049)
Fax: +44 - 121 - 414 3675 (5060)
Edgbaston
Birmingham B15 2TT
United Kingdom
Analytical Framework and Model Descriptions vi

Version 1.0
ABOUT THIS MANUAL
Change details
HDM-4 documents:
comprises:
Volume 6 - Modelling Road Deterioration and Works Effects in HDM-4,
ISBN: 2-84060-102-8
Volume 7 - Modelling Road User and Environmental Effects in HDM-4,
ISBN: 2-84060-103-6

Analytical Framework and Model Descriptions vii

Version 1.0
ABOUT THIS MANUAL

Edgbaston
Birmingham
B15 2TT
England
Tel: +44 - 121 - 414 6717 (or 5049)
Fax: +44 - 121 - 414 3675 (or 5060)

La Grande Arche
France
Tel: +33 1 41 02 05 84
Fax: +33 1 49 00 02 02
E-mail: piarc.hdm4@ibm.net
Web: http://www.piarc.lcpc.fr
Acknowledgements
The authors would like to express their appreciation to all those who have contributed to
developing this manual, in particular, but not limited to, the following people:
Road Deterioration and Works Effects

! William D.O. Paterson of the World Bank
Contributed to bituminous pavements and unsealed roads.
! Greg Morosiuk of Transport Research Laboratory (UK)
Managed the bituminous pavements RD specifications.
! Michael J. Riley (UK)
Contributed to bituminous pavements RD and WE specifications.
! Juan-Pablo Covarrubias of the Inter-American Federation of Cement Producers (FICEM)
Managed the development of RD and WE models for concrete pavements.
! Peter Cenek of Opus International Consultants (NZ)
Contributed to pavement texture modelling.
! Richard Robinson (UK)
Contributed to the classification of roadworks.
! Lennart Djarf of Swedish Road and Transport Research Institute (VTI)
Contributed to RD models for cold climates (rutting due to wear by studded tyres)
Analytical Framework and Model Descriptions viii

Version 1.0
ABOUT THIS MANUAL
Road User Effects

! Christopher Bennett of HTC Infrastructure Management Ltd (NZ)
Contributed extensively to RUE model development.
! Ian Greenwood of Opus International Consultants (NZ)
Contributed to fuel consumption modelling and developed the HDM Tools.
! Chris Hoban of the World Bank
Contributed to speed-flow modelling and non-motorised transport modelling.
! John Hine of Transport Research Laboratory (UK)
Contributed to vehicle utilisation and service life.
! Technical comments were received from a number of reviewers including, but not
limited to:
❏ Rodrigo Archondo-Callao
❏ Clive Daniels
Social and Environmental Effects

! Ulf Hammerstrom of Swedish Road and Transport Research Institute (VTI)
Contributed to vehicle emissions.
! Simon Collings of ETSU, AEA Technology (UK)
Contributed to energy balance analysis.
Economic Analysis
! Vesa Mannisto of Inframan Oy (Finland)
Contributed to investment optimisation methods.
Development of HDM-4
! The World Bank

! Finnish Road Administration (Finnra)
Analytical Framework and Model Descriptions ix

Version 1.0
ABOUT THIS MANUAL
including:
! Finnra

! FICEM

concrete roads.
Birmingham
(NDLI)
costs.

! SNRA
Note: Within each Part/Chapter there will be an Acknowledgements chapter. Each listing
contains references to specific documentation for the topics described in the Part/Chapter.
Copyright statement
Analytical Framework and Model Descriptions x

Version 1.0
Part A
Contents
Volume 4 - Analytical Framework & Model
Descriptions
Part A Overview
A1 Introduction A1-1
Part B Traffic
B1 Traffic characteristics B1-1
Part C Road Deterioration Models

C1 Modelling concepts and approach C1-1
C2 Bituminous Pavements C2-1
C3 Concrete Pavements C3-1
C4 Unsealed Roads C4-1
Part D Road Works Effects

D1 Types of roadworks D1-1
D2 Bituminous Pavements D2-1
D3 Concrete Pavements D3-1
D4 Unsealed Roads D4-1
Part E Road User Effects

E1 Overview E1-1
E2 Vehicle Speeds and Operating Costs E2-1
E3 Non-Motorised Transport E3-1
E4 Road Safety E4-1
Part F Social and Environmental Effects

F1 Overview F1-1
F2 Energy Balance Analysis F2-1
F3 Vehicle Emissions F3-1
Part G Economic Analysis

G1 Economic Analysis G1-1
Part H Nomenclature
H1 Nomenclature H1-1

Version 1.0
CONTENTS
Part I Glossary
I1 Glossary I1-1

Version 1.0
volume four
ISBN: 2-84060-059-5

ISBN: 2-84060-060-9

user interface.
ISBN: 2-84060-061-7

ISBN: 2-84060-062-5

ISBN-2-84060-102-8

ISBN-2-84060-103-6


Part A Overview
Contents

Descriptions: Part A
A1 Introduction
1 Structure of this manual A1-1
2 Economic appraisal of road projects A1-5
2.1 Purpose A1-5
2.2 Transport cost components A1-6
3 Economic appraisal method A1-10
3.1 Project alternatives A1-10
3.2 Discounting A1-10
4 Applications of economic appraisal A1-16
4.1 Financial and economic costs A1-16
4.2 Project types A1-16
4.3 Sensitivity and risk analysis A1-16
4.4 Prioritisation A1-18

Version 1.0
Part A Road Map
Analytical
AnalyticalFramework
Framework
and
andModel
Model
Descriptions
Descriptions
Introduction
Introduction
Part
PartAA
Traffic
Traffic
Part
PartBB
RD
RDModel
Model WE
WEModel
Model RUE
RUEModel
Model SEE
SEEModel
Model
Part
PartCC Part
PartDD Part
PartEE Part
PartFF
Economic
EconomicAnalysis
Analysis
PartGG
Nomenclature
Nomenclature
PartHH

Glossary
Part
PartI I
Figure A Analytical Framework and Model Descriptions Road Map
Analytical Framework and Model Descriptions 1

Version 1.0
PART A OVERVIEW
A1 Introduction
1 Structure of this manual

This manual describes the analytical framework and the technical relationships used within
the HDM-4 model. Before reading this overview it is suggested that you should refer to the
more general Overview of HDM-4 - Volume 1 which outlines the features and functions of
the HDM-4 model.
Figure A1.1 shows a schematic view of the HDM-4 documentation suite. The Overview is
shown as Volume 1 because that needs to be the first document to be read by new users as it
contains introductory material. Volumes 2 and 3 are considered to be guides for users of the
software, where tasks are documented and an understanding of the product can soon be learnt.
Volumes 4 and 5 contain more detailed reference material that is intended for more advanced
users who wish to acquire detailed knowledge of the HDM-4 system.
This manual is split into seven descriptive parts and two appendices. The appendices are
numbered as parts H and I for consistency. Within each part are a number of chapters which
each describe a major topic. Each chapter is then split into sections each describing part of the
topic. The following gives a brief overview of each part:
Provides an overview of the manual, and a summary discussion of road investment

appraisal and prioritisation.
! Part B - Traffic
Describes the traffic characteristics used in HDM-4, and provides details of the traffic
data that are required.
! Part C - Road Deterioration Models
❏ C1 Modelling Concepts and Approach

Describes the pavement types and classification for the following pavements:
" Bituminous
" Concrete
" Unsealed
It includes the modelling approach and a discussion on the key variables that affect
deterioration, which include climate and environment effects, traffic, pavement
history.
❏ C2 Bituminous Pavements
Describes the specifications for modelling the performance of bituminous
pavements. Details of the modelling logic, distress modes, pavement strength,
construction quality, and the relationships are discussed.
Analytical Framework and Model Descriptions A1-1

Version 1.0
PART A OVERVIEW A1 INTRODUCTION
Volume 1
Overview
Volume 2 of HDM-4 Volume 3
Applications Software
Guide User Guide
Volume 4
Analytical
Framework
and Model
Descriptions
Glossary &
Nomenclature
Volume 5
A Guide to
Calibration
and
Adaptation
Figure A1.1 HDM-4 Documentation suite
❏ C3 Concrete Pavements
Describes the implementation of Road Deterioration models for concrete pavements.
It provides an overview of the modelling framework, a brief analysis of the concrete
pavement types and distresses considered, and a description of the models.

Version 1.0
❏ C4 Unsealed Roads
Describes the specifications for modelling unsealed Road Deterioration. It provides
an overview of the modelling logic and the relationships for each of the distresses
modelled.
! Part D - Road Works Effects
❏ D1 Types of Works
Describes the classification of road works and the generic framework for modelling
works quantities, costs and effects for the three road surface classes bituminous,
concrete and unsealed.
❏ D2 Bituminous Pavements
Describes the detailed modelling of Road Works Effects for bituminous pavements.
The methods of defining works activities and intervention criteria, the calculation of
physical quantities of works and the costs to road administration, and works effects
on road characteristics and road use are discussed for different road works classes.
❏ D3 Concrete Pavements
Describes the detailed modelling of Road Works Effects for concrete pavements.
on road characteristics and road use are discussed for different works classes.
❏ D4 Unsealed Roads
Describes the detailed modelling of road works effects for unsealed roads.
on road characteristics and road use are discussed for different works classes.
! Part E - Road User Effects
❏ E1 Introduction
Describes the Road User Effects (RUE) components modelled in HDM-4. It
provides an overview of the HDM-4 representative vehicle framework, which
considers both motorised vehicles and non-motorised transport.
❏ E2 Vehicle Speeds and Operating Costs
Describes the Road User Effects models for calculating motorised vehicle speeds,
operating costs and travel time.
❏ E3 Non-Motorised Transport
Discusses the use of non-motorised modes of transport and its effects on motorised
transport. Since non-motorised modes of transport account for the vast majority of
the movement of people and goods in many developing countries it was essential to
include this model in HDM-4.
❏ E4 Road Safety
Describes the specification for implementing road safety analysis. The model allows
users to define a series of look-up tables of accident rates.

Version 1.0
! Part F - Social and Environmental Effects
❏ F1 Introduction
Introduces the modelling of energy balance and vehicle emissions.
❏ F2 Energy Balance Analysis
The energy used in the road transport sector forms a significant share of the total
energy consumption in most countries. It is therefore important for planners and
decision-makers to be able to understand the energy implications of alternative
transport projects, strategies and policies.
❏ F3 Vehicle Emissions
Describes the implementation of Vehicle Emissions analysis. The document presents
the emissions model together with default parameters for the 16 HDM-4 standard
vehicle types.
! Part G - Economic Analysis
Describes how HDM-4 is used to determine the benefits and costs associated with a road
investment, and how these are used in economic analysis and optimisation procedures.
! Part H - Nomenclature
Lists details about the units and numbering methods used within the HDM-4 suite of
documents.
! Part I - Glossary
Provides descriptions of terminology used throughout the HDM-4 suite of

documentation.

Version 1.0
2 Economic appraisal of road projects
2.1 Purpose
The purpose of road investment appraisal is to select projects with high economic returns.
The decision of whether to invest in roads or in some other infrastructure development is not
the primary objective of road investment appraisal since in most cases such a decision will
already have been made. The purpose of an economic appraisal of road projects therefore is
to determine how much to invest and what economic returns to expect. The size of the
investment is determined by the costs of construction and annual road maintenance. The
economic returns are mainly in the form of savings in road user costs due to the provision of a
better road facility. These three costs constitute what is commonly referred to as the total
(road) transport cost or the whole life cycle cost.
Three primary objectives of road investment appraisal can therefore be identified as follows:
! To determine the appropriate size of investment and the returns expected from this
investment,
! To determine the appropriate geometric and structural design standards for the size of
investment in order to obtain the expected returns,
! To determine the relative priorities for investment among competing road projects when
there is a budget constraint, and
! To assess the economic and socio-economic impact of investments in roads such as the
improvement to the community of industrial, agricultural, educational, and health
services.
The assessment of the socio-economic benefits of road investments however is difficult to
quantify in monetary terms. This is usually done separately after an economic appraisal has
been carried out using a road investment appraisal model.
The primary function of a road investment appraisal model is therefore to calculate the costs
of road construction, road maintenance and road user costs for a specified analysis period.
This is accomplished by modelling the interrelationships between the environment,
construction standards, maintenance standards, geometric standards and road user costs.
A road investment appraisal model may be used to assist with the selection of appropriate
road design and maintenance standards which minimise the total transport cost. The effect of
providing better road standards on the components of the total transport cost is illustrated in
Figure A1.2. If a low standard road is constructed, high maintenance and road user costs can
be expected resulting in a high total transport cost despite the low cost of construction.
Conversely, if a high standard road is constructed, the very high cost of construction will
outweigh the low maintenance and road user costs. A road investment appraisal model can
therefore be used to search for the road design or maintenance standard for which the total
transport cost is a minimum. This alternative is represented by the dotted vertical line in
Figure A1.2. The interaction between the transport cost components however is more
intricate than is suggested in Figure A1.2. For example a high construction cost will not
necessarily lead to lower road user costs, for example, the construction of a very wide but
structurally weak road.
When planning investments in the roads sector, it is necessary to evaluate all costs associated
with the proposed project. These include construction costs, maintenance and rehabilitation
costs, road user costs, and all other external or exogenous costs or benefits that can be directly

Version 1.0
attributed to the road project. It is normal to consider such costs or benefits over an extended
analysis period usually equal to a longer than the expected life of the road, hence the term
'life-cycle' cost analysis. The costs of construction, routine and periodic maintenance are
usually borne by the agency or authority in charge of the road network. However, road user
costs are borne by the community at large in the form of vehicle operating costs (VOC), travel
time costs, accident costs and other indirect costs.
Total Transport Costs

Total Costs
Optimum
Construction Maintenance
& Road User
Low Road Standard High
Figure A1.2 HDM-4 Documentation suite
A road investment model simulates the interaction between pavement construction standards,
maintenance standards and the effects of the environment and traffic loading in order to
predict the annual trend in road condition. This, together with the geometric standards of the
road, has a direct effect on vehicle speeds and on the costs of vehicle operation and accident
rates on the road.
2.2 Transport cost components

The economic evaluation carried out in the road investment models is based on the generated
annual cost streams for road construction, road maintenance and road user costs. The cost
streams will usually begin in a specified base year which may either be the first year of
construction, the first year of trafficking or simply the current calendar year. The number of
years for which costs are calculated depend on the specified analysis period usually chosen to
equal the design life of new roads.
2.2.1 Cost of road construction

The cost of new road construction is calculated from the sum of the costs of site preparation,
earthworks, pavement construction, bridge and drainage structures, and overheads. Site
preparation includes clearing vegetation from the right-of-way, and the stripping and disposal
of topsoil. The cost of earthworks however is greatly influenced by the land terrain and the
geometric standard specified for the road. This includes the costs of excavations, material

Version 1.0
haulage and compaction. The cost of pavement construction is a function of the number,
thickness and type of pavement layers, and includes the costs of shoulders and kerbs.
A sizeable percentage of road construction costs can be attributed to the costs of bridges and
the provision of adequate drainage facilities. Road construction will usually involve the
setting up of temporary site camps and the transportation of equipment, materials and
manpower to the construction site. The costs of these activities together with the profit
margin for the contractor and consultants fees are usually grouped into an overhead cost
which may be specified as a fixed cost per kilometre or as a percentage of the total cost of
construction. At the end of the analysis period, a salvage value may be specified,
representing the percentage of total costs incurred on permanent structures such as
embankments, cuttings, bridges and drainage facilities.
In HDM-4, the costs of construction or any other major road improvements, are user
specified.
2.2.2 Road deterioration

Pavement deterioration is modelled in terms of a number of defects including; road roughness,
pavement cracking, rutting, ravelling, potholes, gravel loss on unsealed roads, and joint
faulting and spalling on concrete pavements. The rate of deterioration is as a function of the
initial pavement design standard, traffic loading, maintenance standards, and the effects of the
environment. The amount of maintenance carried out in a given year depends on user
specified maintenance standards and the predicted road condition.
Pavement performance is largely modelled as a function of several factors (see Part C). Two
of the most important factors are:
! Pavement strength
! Road roughness
Pavement strength is represented by the Structural Number (SNP), an index which represents
pavement strength such that all pavements with the same SNP are assumed to have similar
performance. The concept of pavement structural numbers is described in more detail in Part
C. The structural number of a pavement is defined by an empirical relationship in which the
thickness and strength of each pavement layer are combined using the sum of their product.
The pavement layer strength mainly depends on the type and quality of the constituent
materials. The structural number of a pavement includes the contribution to pavement
strength made by the subgrade. In general, pavements with high structural numbers will have
a low rate of deterioration under the same regime of traffic and environmental loading.
Road roughness is the second important parameter used in modelling pavement performance.
It is the most significant pavement condition which is used in the calculation of road user
costs, particularly VOC. It represents the unevenness of a road surface and is taken to be the
primary cause of wear and tear in vehicles. Roughness is commonly measured using one of
three systems:
! The towed fifth wheel bump integrator (BI) in mm/km,
! The Portland Cement Association (PCA) road meter which simulates the vertical motion
of a standard Quarter-car Index (QI) measured in counts per km,
! Cumulative vertical profile movements, represented in terms of the International
Roughness Index (IRI) measured in m/km.
The following relationships, developed by the World Bank, are used to convert roughness
values between the three units:

Version 1.0
BI = 55 × QI …(2.1)
BI = 630 × IRI1.12 …(2.2)
BI = 900 IRI - 1000 (linear estimate) …(2.3)
A new road has to be assigned a structural number and an initial roughness. The high
dependence of road user costs on roughness, and of roughness progression on the pavement
structural number, suggests that the results of an economic analysis will be affected by the
values assigned to these two variables at the beginning of the analysis.
2.2.3 Cost of road works

The annual cost of road works is calculated from the amounts of reactive routine maintenance,
periodic maintenance, and any road improvements applied in a given year. These depend on
the predicted pavement condition and the specified maintenance or improvement standard
which may include a combination of one or more of different types of road works (see Part
D).
Cyclic routine maintenance works which includes activities that are required regardless of the
condition of the road or the traffic level, for example, grass cutting, road sign painting,
drainage clearance, etc., may be specified as fixed annual costs per kilometre. Emergency
maintenance works such as repairs to flood damage, debris clearance, etc., may also be
specified as an overhead cost since such have to be carried out regardless of pavement
condition.
2.2.4 Road user costs

Road user costs may be defined as the costs incurred by vehicle operators and by the
travelling public at large. The four types of road user costs usually considered are associated
with vehicle operation, travel time, accidents and discomfort (see Part E). The last two costs
are difficult to quantify in monetary terms, although accident costs can be estimated in several
ways in terms of both the resource content (for example, cost of spares and vehicle
replacement), and the injuries and fatalities. However, the lack of acceptable methods of
estimating accident and discomfort costs in developing countries is the main reason why these
two components of road user costs are not included in existing road investment appraisal
models for developing countries.
Vehicle operating costs

These are calculated from the sum of the vehicle resource components, including:
! Fuel and lubricating oil consumption,
! Tyres and spare parts,
! Vehicle maintenance labour costs,
! Vehicle crew wages,
! Vehicle depreciation and interest on capital.
Separate sets of equations are used for the different vehicle types specified by the user. For
each vehicle type, the models predict average travel speeds as a function of road geometry and
road condition. The above VOC components, with the exception of vehicle depreciation and
interest, depend largely on road roughness and the geometric characteristics of the road. The
consumption of the above VOC components are predicted in resource terms. For example,
equations for fuel consumption calculate the quantity of fuel consumed over the travel

Version 1.0
distance. Unit costs for the various resources are specified by the user in order to calculate the
annual total costs of vehicle operation. Vehicle depreciation is considered to be a function of
the predicted travel time and of the level of vehicle utilisation.
Travel time costs

Travel time costs are calculated from average vehicle speeds, travel distances and the unit
costs per hour of road users' time. The average vehicle speeds are a function of road
roughness, road width, and the vertical and horizontal alignment of the road. The values to be
specified for unit time costs for road users in developing countries are not easily justifiable.
Many authors on this subject recommend that projects in developing countries should be
appraised without benefits from time savings. The road user benefits derived from savings in
travel time costs can in this case be considered to be a "consumer surplus" in addition to the
savings in VOC.
2.2.5 Other exogenous costs and benefits

Other costs or benefits that can be directly associated with a road project may be included in
the economic analyses. These usually include independently assessed benefits accruing from
socio-economic developments such as increased agricultural productivity, industrial output,
accessibility benefits, etc. Exogenous costs could include the costs of providing diversion
routes, noise barriers, and other impediments during construction. Such costs or benefits are
not calculated by road investment models and therefore their inclusion in any economic
analyses must be clearly justified as they can easily affect the ranking of alternative projects.

Version 1.0
3 Economic appraisal method
3.1 Project alternatives

Roads are normally constructed to reduce costs and therefore to increase benefits derived from
reduced user costs and from improvements to socio-economic services. The economic
appraisal of road projects is therefore essentially a comparison of transport cost components
calculated for at least two alternatives for road construction, usually referred to as the Do
Minimum or Without Project alternative and one or more Do Something or With Project
alternatives.
3.1.1 Without project alternative

The Do Minimum or Without Project alternative in most cases represents the current
situation from which a reduction in transport costs is sought. It is usually the alternative
which involves minimum input of capital, for example, a continuation with the current road
standard. The annual cost matrix for a Do Minimum alternative will usually have little or no
construction cost component but with generally high maintenance and road user costs.
3.1.2 With project alternatives

The selection of project alternatives to be analysed depends on several factors, in particular,
national road standards, previous road projects, traffic levels, availability of materials as well
as political and socio-economic considerations. A project alternative will usually involve the
provision of a higher standard road. This could be achieved by either new construction,
reconstruction, upgrading, or improvements to pavement or geometric standards, all of which
can be analysed as independent project alternatives. The cost matrices from these project
alternatives will have varying levels of capital and recurrent costs but with generally lower
road user costs.
3.2 Discounting
It is necessary to discount the transport costs in each year of an analysis period to their value
in the base year. This is carried out to reflect the time value of money represented by the
opportunity cost of the capital invested in a road project. Discounting is performed by
multiplying the cost in a given year by the discount factor for that year. Discount factors are
derived from the equation below:
r -N
D.F. = (1 + ) …(3.1)
100
where:
r discount rate in percent (%)

N number of years from the base year
Table A1.1 illustrates the principle of discounted cash flow analysis (DCF) applied to a gravel
road which will be paved after one year. The paved road has a design life of 10 years after
construction. The economic comparisons are based on the totals of the discounted costs. This
is referred to as the Present Value (PV) of costs. For example, in Table A1.1, the PV of the
Without Project alternative is $ 23.8 million, and the PV of the With Project alternative is

Version 1.0
$20.2 million, both at 12% discount rate. If the example given in Table A1.1 represents a real
road project, the alternative chosen for implementation would depend on the economic criteria
used to compare the alternatives. The most commonly used criteria for selecting projects are
the net present value (NPV), the internal rate of return (IRR) and the Benefit Cost Ratio
(BCR).
3.2.1 Net Present Value

The NPV is defined as the difference between the discounted benefits and costs of a project.
In the economic appraisal of road projects, benefits are derived mainly from savings in road
user costs and also in road maintenance costs (where possible). Thus the benefits of paving a
gravel road would be obtained by subtracting the total transport cost for the sealed road from
that of the gravel road. The calculation of NPV is therefore simplified by taking the
difference between the present value of costs for the two alternatives being compared. Thus
from Table A1.1, the NPV of the With Project alternative when compared against the
Without Project alternative would be $3.6 million at a 12% discount rate.
3.2.2 Internal Rate of Return

The NPV depends on the discount rate used in the calculation of present values. When high
discount rates are used, a lower NPV is obtained resulting in negative values. The IRR of a
project is defined as the discount rate at which the present value of costs equals the present
value of benefits, that is, when NPV is zero. In the example given in Table A1.1, the IRR
would be the discount rate for which the two present value of costs are equal; in this case
approximately 15.2%. Projects with high IRR values are generally preferred as this will give
positive NPV at high discount rates. In general, the calculated IRR should be greater than the
test discount rate used to assess projects funded by government.

Version 1.0
Table A1.1 Comparison of With and Without Project Alternatives
Without Project Alternative Costs ($ m) With Project Alternative Costs ($ m)

Year 12% Annual Vehicle Total Discounted Road Annual Vehicle Total Discounted
Discount Maintenance Operation Transport Total Construction Maintenance Operation Transport Total
Factors
0 (Base) 1.0000 0.8 1.5 2.3 2.3 0.8 1.5 2.3 2.3
1 0.8929 0.9 1.6 2.5 2.2 5.8 0.4 1.0 7.2 6.4
2 0.7972 1.0 1.8 2.8 2.2 0.5 1.1 1.6 1.3
3 0.7118 1.1 2.1 3.2 2.3 0.6 1.3 1.9 1.4
4 0.6355 1.2 2.3 3.5 2.2 0.7 1.4 2.1 1.3
5 0.5674 1.3 2.7 4.0 2.3 0.7 1.6 2.3 1.3
6 0.5066 1.4 2.9 4.3 2.2 0.8 1.8 2.6 1.3
7 0.4523 1.5 3.2 4.7 2.1 0.9 2.0 2.9 1.3
8 0.4039 1.6 3.5 5.1 2.1 0.9 2.2 3.1 1.3
9 0.3606 1.7 3.8 5.5 2.0 1.0 2.3 3.3 1.2
10 0.3220 1.8 4.0 5.8 1.9 1.0 2.5 3.5 1.1
TOTALS 29.4 PV = 23.8 18.7 PV = 20.2

Version 1.0
3.2.3 Benefit Cost Ratio

The BCR provides a simple measure of the profitability of a project (that is, amount of
benefits derived for every dollar invested). It represents the dimensionless index obtained by
dividing the calculated benefits of the project by the discounted capital costs of the
investment. This can be estimated from the NPV as follows: (see Part G for details):
NPV
BCR = +1 …(3.2)
C
From the example given in Table A1.1, the estimated BCR would be 1.55.

Version 1.0
4 Applications of economic appraisal
4.1 Financial and economic costs

The financial cost of a project is the sum of the market price of materials, labour, equipment,
and overheads incurred in the construction of a project. The economic appraisal of road
projects should be carried out using economic costs which represents the true cost of a project
to the economy of a country. For example, the market price of fuel in most countries includes
a percentage of government tax. The economic price of fuel would therefore be estimated
from the market price less the tax. In most cases the economic cost of fuel will be less than
the financial cost. This relationship is reflected in VOC since most of the components are
taxed by governments, for example, vehicle tax. It is also possible for economic costs to
exceed financial costs if a government subsidises instead of taxing various cost components.
4.2 Project types

Investment appraisal models can be used to perform a number of economic analyses. One
important type on analysis is the choice of pavement surfacing during the design of a
pavement. A simple example is the choice between a surface dressing or an asphalt concrete
surfacing for a pavement to be constructed in a developing country. In this case the surface
dressing is likely to have a high initial roughness with higher roughness progression rates
resulting in a high VOC component but with a lower cost of construction. The choice
between the two types of surfacing will therefore be governed by the total VOC calculated for
the two alternatives.
In the geometric design of new roads, the investment models can only be used to reject
designs that would be extravagant, for example, the provision of a dual carriageway when a
single carriageway would suffice. It should be noted that the geometric design of roads
should be done to satisfy capacity requirements and to provide safety to road users.
Although investment appraisal models do not carry out mathematical optimisation of, for
example, maintenance options or the timing of construction or maintenance, they can be used
to conduct a sensitivity analysis to study the effects of changes in construction cost,
maintenance cost, VOC, traffic growth, discount rates, and time values.
4.3 Sensitivity and risk analysis

All road projects involve some degree of uncertainty in the outcome of the project. The
decision to proceed with a project therefore includes some element of risk taken by road
authorities. Many road projects will have a significant element of risk attached to them. These
will in general be due to several factors amongst which the following are the main causes:
! Unforeseen events beyond the control of the engineer; for example, improved
technology, political changes.
! National economic changes; for example, future economic growth, traffic growth rates.
! Unpredictability of pavement performance due to environment, traffic, construction.
! Impact on socio-economic factors that can not be evaluated.
It is necessary to assess the impact of uncertainty on the viability of road projects for the
following reasons:

Version 1.0
! Road investments often take up a large proportion of national income, therefore any
failure will be expensive,
! Alterations during implementation can be very expensive and prohibitive and should
therefore be avoided by selecting the most suitable alternative at the outset,
! To determine impact of possible changes (for example, to the environment, socio-
economic) on the overall viability of road projects and plan for these accordingly.
The formal method of economic appraisal is only one step in the process of quantifying risk.
Some of the causes of uncertainty can be assessed by additional analyses such as;
! Sensitivity analysis
❏ This is applied to study the effects of changes in one parameter (for example,
construction cost, or traffic growth rate) on the overall viability of a road project.
! Scenario analysis
❏ This is carried out to determine the broad range of parameters that together would
affect the viability of the road project. For example, a review of government long-
term policies could yield alternative economic growth rates which would affect both
traffic growth rates and the cost of road construction.
! Risk analysis
❏ Most commonly done by assigning probabilities of occurrence to project parameters

and then study combined effects of changes in the parameters. This can be done by
reviewing past trends in project parameters (for example, traffic growth rates, final
construction costs, etc.) on other projects to arrive at a suitable probability
distribution.
Sensitivity analysis is the simplest form of risk analysis. Essentially, it involves repeated
economic evaluation with systematic changes made to one parameter each time. The
procedure can be summarised as follows:
! Identify parameters which affect the viability of a road project, for example, discount rate
used, cost of construction, traffic growth rate, road maintenance standards, etc.
! Systematically change values of these parameters and repeat the economic appraisal.
Generally, changes to the parameters should be made relative to a base value, for
example, +/- 25%, 50%, 100%.
! Study effect of changes and report on parameters to which the project is most sensitive.

Version 1.0
4.4 Prioritisation
There are often situations when the budget available for road projects will not be sufficient to
undertake all projects shown to have a positive return (that is, projects with positive NPV). In
such situations, a formal method of selecting projects to be included within the budget can be
applied. Capital budgeting or rationing can be applied to a group of projects that meet either
of the following conditions:
! Projects that are independent of each other (for example, list of road projects from
different parts of the country),
! Mutually exclusive projects (that is, projects that are alternatives to each other) when
only one alternative can be selected.
The NPV capital budgeting rules can be applied in both situations where sufficient funds are
available, and also when there is a budget constraint. The rules are summarised below:
! When sufficient funds are available to undertake all projects;
❏ Select all independent projects with NPV > 0
❏ Select mutually exclusive project alternatives with the highest NPV
! When capital rationing is to be applied due to shortage of funds;
❏ Select independent projects with the highest NPV/Cost ratio
❏ Select mutually exclusive projects using the incremental NPV/Cost method
described below.
The incremental analysis is used to test whether the ratio of the increase in NPV to the
increase in costs between alternative mutually exclusive projects is greater than a specified
marginal ratio. The formula is defined as:
(NPV2 - NPV1 )
IBCR = …(4.1)
(C 2 - C1 )
where:
IBCR Incremental Benefit/Cost Ratio

NPV2, 1 Net Present Values of two mutually exclusive project alternatives
C2, 1 Investment costs of the two mutually exclusive project alternatives
If the above ratio is greater than a specified marginal value, then the project alternative is
included among those to be funded. The marginal value is usually determined from the BCR
of the road project at the budget boundary.
Further details of the above method are given in the Applications Guide (Volume 2).

Version 1.0
Part B Traffic
Contents

Descriptions: Part B
B1 Traffic characteristics
1 Introduction B1-1
2 Representation of traffic B1-1
2.1 Purposes of traffic data B1-1
2.2 Data types B1-2
3 Traffic categories B1-3
4 Traffic composition, volumes and growth rates B1-4
4.1 Basic features B1-4
4.2 Composition, volumes and growth rates B1-5
5 Axle loading B1-9
5.1 The need for axle loading data B1-9
5.2 Vehicle axles B1-9
5.3 Equivalent standard axle load factors B1-9
5.4 Light and heavy vehicles B1-10
5.5 Cumulative traffic loading B1-11
6 Road capacity and speed-flow relationships B1-13
6.1 Basic concepts B1-13
6.2 Data to be specified B1-15
7 Hourly flow-frequency distribution data B1-17
7.1 Basic concepts B1-17
7.2 Data to be specified B1-17
7.3 Limits on traffic flows B1-20
8 References B1-22

Version 1.0
Part B Road Map
Analytical
AnalyticalFramework
Framework
and
andModel
Model
Descriptions
Descriptions
Introduction
Introduction
Part
PartAA
Traffic
Traffic
Part
PartBB
RD
RDModel
Model WE
WEModel
Model RUE
RUEModel
Model SEE
SEEModel
Model
Part
PartCC Part
PartDD Part
PartEE Part
PartFF
Economic
EconomicAnalysis
Analysis
PartGG
Nomenclature
Nomenclature
PartHH

Glossary
Part
PartI I
Figure B Analytical Framework and Model Description Road Map

Version 1.0
PART B TRAFFIC
Part B Traffic
B1 Traffic characteristics
1 Introduction
This chapter describes the traffic characteristics used in HDM-4, and provides details of the
traffic data that are required. These include the methods used for calculating future traffic and
axle loading together with the approaches for modelling traffic congestion. The traffic data are
used in all four sets of HDM-4 models (Road deterioration, Works effects, Road user effects
and Social and environmental effects, see Figure B).
As travel demand varies both in time and space, traffic data should be representative averages
for a road section. The representation of traffic needs to be at an appropriate level of detail in
accordance with the type of analysis (project, programme or strategy) to be performed.
A list of research documents referenced from this chapter is given in Section 8.
2 Representation of traffic
2.1 Purposes of traffic data

The results of economic analyses are quite sensitive to traffic data, and most benefits that
justify road improvements arise from savings in road user costs (see Part E). To perform
economic analyses in HDM-4, traffic characteristics of roads therefore need to be described
and represented at an appropriate level of detail.
Traffic characteristics need to be represented for the following analytical purposes:
! Project analysis
This requires a detailed representation of traffic characteristics on the road being

analysed. For each road section, the representation should include data items that
describe the details of changing traffic composition and volumes, axle loading, capacity
and speed-flow relationships, hourly traffic flows, traffic induced by road improvements,
and of demand shifts.
The required traffic data for this type of analysis is similar to that for project analysis,
with the exception that the data is at a more aggregate level (see Applications Guide). For
example; traffic volumes may be specified by vehicle classes, and the data could be used
by several representative road sections.
! Strategy analysis
This requires the specification of an aggregate set of traffic data, that is representative of
a group of road sections being analysed. Traffic levels should be expressed in terms of
daily flows, and may be described as low, medium, or high. The composition of traffic
may be expressed as a percentage of daily flow for each vehicle class/type.
For the purposes of representing traffic characteristics both for project and network level
analyses, road sections within a network must be categorised according to the following:
Analytical Framework and Model Descriptions B1-1

Version 1.0
PART B TRAFFIC B1 TRAFFIC CHARACTERISTICS
! Road types
For example, single lane roads, four-lane roads, motorways, etc. This data is used to
determine the parameters for capacity, speed-flow relationship shape, width effects and
passenger car space equivalents for each road type.
For example; commuter, recreational, inter-city, etc. This data is necessary for describing
the patterns of traffic flow along each road category; commuter routes, for example, tend
to have weekday peaks but low weekend traffic, whereas recreational routes have a more
peaked distribution.
! Non-motorised transport factor
Measures the effect of non-motorised transport (for example, bicycles, animal-drawn

carts, pedestrians, etc.) on motorised traffic speeds.
! Roadside friction
Measures the effect of roadside activity on traffic speeds; this includes the effects of land
use, roadside stalls, bus stops, parking, access points, etc.
! Motorised transport factor
Measures the effect of motorised transport on non-motorised transport (NMT) speeds.
2.2 Data types

The traffic data required is incorporated in several modules (or applications) rather than being
specified in one place. Traffic data types can be considered under the following headings.
! Traffic categories (see Section 3)
These are normal, diverted and generated traffic.

! Traffic composition, volumes and growth rates (see Section 4)
Specified within the individual applications programs (that is, project analysis,
programme analysis and strategy analysis) as required by the purpose of the study being
performed.
! Axle loading (see Section 5)
Specified for each vehicle type in the Vehicle Fleet folder (see Overview of HDM-4).
Provision is made to derive most of the required parameters (for example, equivalent
standard axle load factors, etc.) from the user-specified data.
! Road capacity and speed-flow relationships (see Section 6)
Defined according to the road type within the HDM-4 Configuration (see Overview of
HDM-4).
! Hourly flow-frequency distribution data (see Section 7)
Required to estimate hourly flow ranges from annual average traffic data. These traffic
data types are road dependent and are specified for each road use category within HDM-4
Configuration (see Overview of HDM-4).
The specification for each of these data sets is discussed in Sections 3 to 7 (as referenced above).

Version 1.0
3 Traffic categories
Traffic is separated into the following categories in order to assess benefits, TRRL Overseas
Unit (1988):
! Normal (see Section 4.2.1)
Normal traffic is defined as traffic that would pass along the project road if no investment
took place, including normal growth. It is specified for each road section within the
applications programs.
! Diverted (see Section 4.2.2)
Diverted traffic is defined as traffic that changes from another route (or transport mode)
to the project road, but still travels between the same origin and destination (this is
termed reassigned traffic in transport modelling). It is specified together with the road
investment option that causes it, and is specified within the relevant applications
programs.
! Generated (see Section 4.2.3)
Generated traffic is defined as additional traffic that occurs in response to the road
investment (this includes redistributed traffic as defined in transport models). Generated
traffic arises either because a journey becomes more attractive because of a cost or time
reduction, or because of the increased development that is brought about by a road
investment. It is specified together with the road investment option that induces it, and is
specified within the relevant applications programs.
These categories are each treated separately in an economic analysis (see Part G).

Version 1.0
4 Traffic composition, volumes and growth rates
4.1 Basic features

The traffic composition is defined as the proportions of the different vehicle types that use the
road. Information on traffic composition is required for several analytical purposes, including:
! Predicting pavement deterioration
! Estimation of vehicle operating costs
! Estimation of travel time
! Predicting quantities of vehicle exhaust emissions
! Calculation of energy use
! Economic analysis
The existing traffic volumes on the road being analysed are specified in terms of vehicle type
or class, depending on the kind of analysis to be performed. The value entered for each
vehicle type is expressed as the annual average daily traffic (AADT):
Total annual traffic in both directions

AADT =
365
This constitutes the baseline flow for the analysis period. It is assumed that seasonal variations
in traffic flows have already been accounted for when estimating the AADT from traffic
counts carried out over shorter periods.
For project analysis, traffic composition data is specified for each section. For network level
analysis, several representative sets of traffic composition data can be specified, and each is
assigned to a group of road sections with similar traffic characteristics.
In most situations, traffic growth has a major effect on the level of benefits obtained.
Specifying different growth rates for each vehicle type/class can effect changes in traffic
composition over time. Uncertainties always exist in estimating initial traffic, but there is even
greater uncertainty in forecasting future growth rates. Thus, it is recommended that sensitivity
analyses be always undertaken on the effects of different forecast growth rates.
Traffic volumes are derived from the baseline AADT and the composition data. The
computation of annual road user effects requires the AADT for each road section to be broken
down by vehicle types. For normal traffic, the annual average daily traffic for each vehicle
type (AADTkn) is obtained by multiplying the normal traffic AADTn by the composition data
defined for the vehicle type k in the base year. For generated traffic, AADTkg is either
specified directly or calculated, based on AADTkn and the traffic growth type used. Volumes
for normal and generated traffic are required as separate inputs into the Economic Analysis
module (see Part G) to enable the economic benefits of generated traffic to be
determined.
The specification of traffic composition data for each section is carried out for each traffic
category as described in Sections 4.2.1, 4.2.2 and 4.2.3.

Version 1.0
4.2 Composition, volumes and growth rates
4.2.1 Normal traffic

Normal traffic is specified using the following data.
! Annual average daily traffic (AADT)
It is permissible and sometimes convenient to enter AADT, in vehicles per day, for some
year before the start of the analysis period. However, it must be ensured that the traffic
start year is always less than or equal to the start year of the analysis period.
The proportion of each representative vehicle that uses the road, as a percentage of the
AADT. For example; medium cars 10%; articulated trucks 5%; etc. These data items are
used to calculate the actual volumes, by vehicle type, for the applicable year.
! Future traffic
This is specified as a traffic growth period defined in terms of its start year and by one of
the following growth types:
❏ annual percentage increase (p)
( y −1)
 p 
AADTy = AADTstartyear 1 +  . . .(4.1)
 100 
where:
AADTy annual average daily traffic in year y (vehicles per day)

AADTstartyear annual average daily traffic in the start year, (vehicles per day)
p annual percentage increase in AADT (%)
❏ annual incremental increase in AADT (vpd)
AADTy = AADTstartyear [1 + VPD(y − startyear )] ...(4.2)
where:
AADTy annual average daily traffic in year y (vehicles per day)

AADTstartyear annual average daily traffic in the start year, (vehicles per day)
VPD annual incremental increase in AADT (vehicles per day)
❏ actual AADT, which must be followed by another traffic growth period, unless the
year in which it applies is the last year of the analysis period
Then, depending on the selected traffic growth type, the growth rate is specified for each
representative vehicle. For example, light trucks increasing at 4% per year from 2000 to
2005, if annual percentage increase was selected.

Version 1.0
Several traffic growth periods may be defined, each with a minimum length of one year.
It is important to ensure that the defined traffic growth periods cover consistently each
and every year of the analysis period. Therefore, if only one growth period has been
defined, this will be assumed to apply to all successive analysis years. A combination of
different growth types may be used to specify future traffic over the analysis period.
4.2.2 Diverted traffic

Forecasting diverted traffic can be difficult, particularly where traffic diverts from other
transport modes, and for analysis over a complete network. Ideally, estimates of diverted
traffic should be made by utilising the results from external traffic demand models. Thus, in a
situation where a road works causes diverted traffic, a new set of traffic data representing the
net effect of traffic diversion over all the sections affected should be defined for each project
alternative.
For each project alternative, a new set of traffic data is specified as follows:
! Name
! Start year
The calendar year in which traffic diversion starts, typically this should coincide with the
year following the completion of the road works.
! New AADT
The annual average daily traffic in the start year for each road section.
! New composition
The traffic composition in the start year for each road section.
! Future traffic
Specified by selecting one of the following growth types:

❏ annual percentage increase in AADT, calculated using Equation 4.1 above
❏ annual incremental increase in AADT, calculated using Equation 4.2 above
Economic comparisons of investment options involving diverted traffic can only be done
meaningfully, at the project analysis level, if the following conditions are met:
1 All the road sections from and to which traffic diverts are analysed together with the
as a result of carrying out the road works.
2 In any given analysis year, the total traffic volume entering the study area equals the total
The analysis of a new section in an entirely new location will always involve diverted traffic.
4.2.3 Generated traffic

The main casual factors of generated traffic are reductions in travel cost or time, or the
increased development brought about by a road investment. The amount of generated traffic is
difficult to forecast accurately, so the time period over which traffic is generated should be
limited.

Version 1.0
The recommended approach to forecasting generated traffic is to use demand relationships.

The price elasticity of demand for transport measures the responsiveness of traffic to a
change in transport costs following a road investment.
For each road investment standard, generated traffic is specified as follows:
! Name
! Relative start year
The number of years, after the start of the road works, that generated traffic appears.
Thus, relative start year j means the jth year following the start year of the works that
cause the generated traffic.
! Future traffic
This is specified by selecting one of the following growth types:

❏ additional annual percentage increase in AADT (q)
the generated traffic in year y is given by:
AADTgen( y ) = AADTtotal( y ) − AADTnorm( y ) ...(4.3)
 (p + q) 
AADTtotal( y ) = AADTtotal( y −1) 1 +  ...(4.4)
 100 
 p 
AADTnorm( y ) = AADTnorm( y −1) 1 +  ...(4.5)
 100 
where:
AADTgen(y) generated traffic in year y (vehicles per day)

AADTtotal(y) total traffic in year y (vehicles per day)
AADTtotal(y-1) total traffic in year y-1 (vehicles per day)
AADTnorm(y) normal traffic in year y (vehicles per day)
AADTnorm(y-1) normal traffic in year y-1 (vehicles per day)
q additional annual percentage increase in AADT (%)
p annual percentage increase in AADT for normal traffic (%)
❏ percentage of the normal traffic in the current year
p
AADTgen( y ) = AADTnorm( y ) ...(4.6)
100
❏ annual incremental increase in AADT of generated traffic

❏ actual AADT of generated traffic

Version 1.0
4.2.4 Traffic at intersections

Total traffic entering an intersection node is needed for road safety analysis (this is not
available at HDM-4 Version 1.0). The total traffic entering the node is entered as AADT.
Traffic composition data are not required for this.
4.2.5 Changes in vehicle fleet characteristics

If changes in the characteristics of the vehicle fleet are expected to occur in the future, the
following procedure should be followed:
Any vehicles that come into use at a future date should be specified in Vehicle
Fleet folder, together with those that are already in use. The AADT of future
vehicles should be set to zero until the year in which they come into effect. In
that year new AADT for the vehicle is defined using incremental increase in
AADT, or the actual AADT, followed by another growth period. Negative
growth rates can be defined for vehicles that are already in use in such a
manner that they are gradually phased out (that is, AADT=0) and replaced by
future vehicles.

Version 1.0
5 Axle loading
5.1 The need for axle loading data

The following measures of axle load are required to predict the impacts of traffic on pavement
deterioration and maintenance effects:
! Numbers of vehicle axles (YAX)
Defined as the total number of axles of all vehicles traversing a given road section in a
given year.
! Numbers of equivalent standard axle loads (ESA)
Combines the damaging effects of the full spectrum of axle loading in a common
damage-related unit. ESA is considered on each road section, for each year of the
analysis period.
5.2 Vehicle axles

For each vehicle type; the number of vehicle axles, YAXk, traversing a given road section in a
particular year is computed as the volume of traffic multiplied by the number of axles per
vehicle of the type involved. The total number of all axles, YAX, in a given year is obtained
by summing YAXk for all vehicle types.
T * NUM_AXLES k
YAX k = k ...(5.1)
ELANES * 10 6
YAX =
∑ YAX
k =1
k ...(5.2)
where:
YAX annual total number of axles of all vehicle types (millions per lane)
Tk annual traffic volume of vehicle type k, (k = 1, 2, . . , K)
NUM_AXLESk number of axles per vehicle type k
ELANES effective number of lanes for the road section
The effective number of lanes ELANES is used to model the effect of traffic load distribution
across the width of paved roads. This may either be specified by the user or be taken by
default as equal to the number of lanes (NLANES) for the road section.
5.3 Equivalent standard axle load factors

The equivalent standard axle load factor is defined as the number of applications of a standard
80kN dual-wheel single axle load that would cause the same amount of damage to a road as
one application of the axle load being considered. The value of ESALF for each vehicle type

Version 1.0
may be specified directly by the user, or calculated from axle load information defined in the
Vehicle Fleet folder.
For each vehicle type, ESALFk is computed using information on the different damaging
effects of various axle configurations. For each type of axle group j a standard load, SAXLj, is
used to determine the loading ratio. The expression for calculating ESALF is as follows
Watanatada et al. (1987):
LE
Ik Jk
 AXL kij 
ESALFk =
∑ 100 ∑
i =1
Pki
j =1

 SAXL
 j



...(5.3)
where:
ESALFk equivalent standard axle load factor for vehicle type k, in equivalent standard
axle loads
Ik the number of subgroups i (defined in terms of load range) of vehicle type k (i
= 1, 2, . . ., Ik). Note that i may represent each individual vehicle
Pki percentage of vehicles in subgroup i of vehicle type k. If i represents each
individual vehicle then Pki = 100 (%)
LE axle load equivalency exponent (default = 4.0)
Jk the number of single axles per vehicle of type k (j = 1, 2, . . ., Jk)
AXLkij the average load on axle j of load range i in vehicle type k (tonnes)
SAXLj the standard single axle load of axle group type j, for example, 6.60 tonnes for
single-wheel single axle, 8.16 tonnes for dual-wheel single axle, etc.
The factor ESALFk is therefore an average over all vehicles of type k, loaded and unloaded, in
both directions on the given road section.
The annual total number of equivalent standard axles is calculated as:
K
YE4 =
∑ ELANES
T * ESALF
k =1
* 10
k k
6
...(5.4)
where:
YE4 annual total number of equivalent standard axles (millions per lane)
All other parameters are as previously defined in Sections 5.2 and 5.3.
5.4 Light and heavy vehicles

The modelling of some pavement distress modes and the calculations of unsealed road
deterioration requires input of the amounts of light and heavy MT vehicles. Heavy vehicles are
categorised as those with operating weight equal to or greater than 3.5 tonnes; other vehicles
are categorised as light. The Average Daily Light vehicles (ADL) and the Average Daily

Version 1.0
Heavy vehicles (ADH) are specified in terms of vehicles per day for each year of the analysis
period.
The modelling of the changes in pavement skid resistance requires the specification of the
flow of heavy commercial vehicles per lane per day (QCV).
ADH
QCV = ... (5.5)
ELANES
where:
QCV flow of heavy commercial vehicles per lane per day

ADH average daily heavy vehicles (numbers in both directions)
ELANES effective number of lanes for the road section
The modelling of changes in pavement texture depth requires the specification of the annual
number of equivalent light vehicle passes (∆NELV) over the road section. This is calculated
from the following expression:
∆NELV = 365 * (ADL + 10 * ADH) ...(5.6)
where:
∆NELV Number of equivalent light vehicle passes during an analysis year
The number of vehicles with studded tyres is required for modelling pavement rutting during
freezing seasons.
365 * ST * AADT y * 10 −5
PASS = ...(5.7)
NTFD
where:
PASS annual number of vehicle passes with studded tyres in one direction (measured
in 1000s)
AADTy annual average daily traffic (AADT) in year y (veh/day)
ST percentage of annual number of vehicle passes with studded tyres (%)
NTFD number of traffic flow directions
5.5 Cumulative traffic loading

The cumulative traffic loading parameters are used for modelling road deterioration and as
intervention criteria for some road works activities. These parameters are calculated from the
accumulated traffic since the time of the last surfacing or construction works on the road
section (see below).

Version 1.0
The cumulative number of equivalent standard axle loads since the last rehabilitation,
reconstruction or new construction works (NE4) is given by:
∑
AGE 3
NE4 = YE4 y ... (5.8)

y =1
where:
NE4 cumulative number of equivalent standard axle loads since last rehabilitation,
reconstruction or new construction (millions/lane)
YE4y number of equivalent standard axle loads in year y (millions/lane)
AGE3 number of years since last rehabilitation, reconstruction or new construction
(years)

Version 1.0
6 Road capacity and speed-flow relationships
6.1 Basic concepts

The capability to model the effects of traffic volume on speeds is provided to enable the
economic consequences of road capacity improvements to be determined. The factors that
determine speed-flow relationships are described below:
! Capacity
The maximum number of vehicles that can pass a point, or traverse a road section, in one
hour (total both directions). Capacity values determine the shape of speed-flow curves
by establishing the ultimate capacity value.
! Free speed
The speed of each vehicle at zero (or very low) flow. It is unaffected by other traffic but
is affected by the physical characteristics of the road and other non-traffic factors (see
Part E). The average free speed is calculated for each vehicle type.
! Speed at capacity
As traffic flows increase, average speeds for all vehicles converge towards the speeds of
the slowest vehicles in the stream, as passing becomes more and more restricted. As flow
approaches the ultimate capacity, average speeds may fall even lower than slow vehicle
free speeds, and any small disturbances in the traffic stream causes a stop-and-go
situation. An estimate of average speed at ultimate capacity, also known as jam speed,
is needed to describe the speed-flow-capacity relationship.
6.1.1 Speed-flow model

The speed-flow model adopted for motorised transport (MT) is the three-zone model
proposed by Hoban et al. (1994). This model is illustrated in Figure B1.1.

Version 1.0
S1
S2
Speed km/h
S3
Snom
Sult
Qo Qnom Qult
Flow in PCSE/h
Figure B1.1 Speed-flow model
The following notation applies to Figure B1.1:
Qo the flow level below which traffic interactions are negligible in PCSE/h
Qnom nominal capacity of the road (PCSE/h)
Qult the ultimate capacity of the road for stable flow (PCSE/h)
Snom speed at the nominal capacity (km/h)
Sult speed at the ultimate capacity, also referred to as jam speed (km/h)
S1 to S3 free flow speeds of different vehicle types (km/h)
PCSE passenger car space equivalents (see below)
6.1.2 Passenger car space equivalents

To model the effects of traffic congestion, the mixed traffic flows are converted into
equivalent standard vehicles. The conversion is based on the concept of Passenger Car Space
Equivalents (PCSE) Hoban et al. (1994). This accounts only for the relative space taken up by
the vehicle on the road, and reflects the fact that the speed-flow model takes account explicitly
of speed differences of the various vehicles in the traffic stream. PCSE factors vary by road
type, and narrow roads have higher PCSE values than wide roads. Table B1.1 gives the values
of PCSE by vehicle class and road type.

Version 1.0
Table B1.1 PCSE Values
Vehicle Average Space Total Basic Recommended Values

Length Headway Space PCSE 2-Lane Narrow
(m) (m) (m) 1-Lane
4-Lane 2-Lane
Car 4.0 32.0 36.0 1.0 1.0 1.0 1.0
Utilities (Pickup) 4.5 36.0 40.5 1.0 1.0 1.0 1.0
Heavy Bus 14.0 44.0 58.0 1.6 1.8 2.0 2.2
Light Truck 5.0 40.0 45.0 1.3 1.3 1.4 1.5
Medium Truck 7.0 44.0 51.0 1.4 1.5 1.6 1.8
Heavy Truck 9.0 48.0 57.0 1.6 1.8 2.0 2.4
Trailer 11.0 50.0 65.0 1.8 2.2 2.6 3.0
Notes: The basic PCSE values have been used as defaults for the HDM-4 standard vehicle
types (see Part E)
6.2 Data to be specified

The key parameters for use in the speed-flow model vary depending upon the road type and
width. Table B1.2 lists the recommended values for these parameters. The values for Qo and
Qnom are expressed relative to Qult.
Note that in HDM-4 the ultimate capacity for the road section Qult is obtained from the
product of the ultimate capacity per lane (QLult) and the number of lanes for the road section
(NLANES).
The ratio of Qo to Qult is designated by XQ1, and is expressed as follows:
Qo
XQ1= ...(6.1)
Qult
The ratio of Qnom to Qult is designated by XQ2, and is expressed as follows:
Qnom
XQ2 = ...(6.2)
Qult
Table B1.2 Examples of capacity and speed-flow model parameters for different
road types
Road type Width XQ1 XQ2 QLult Sult σmaxr

(m) (PCSE/lane/hr) (km/hr) 2
(m/s )
Single lane road <4 0.0 0.70 600 10 0.75
Intermediate road 4 to 5.5 0.0 0.70 1200 20 0.70
Two lane road 5.5 to 9 0.1 0.90 1400 25 0.65
Wide two lane road 9 to 12 0.2 0.90 1600 30 0.60
Four lane road >12 0.4 0.95 2000 40 0.60

Version 1.0
Since these data apply to individual road sections, it is important to ensure that the data relates
to a single and not a dual carriageway. This ensures consistency with the definition of a road
section as used in HDM-4.
The data in Table B1.1 describing the capacity of road sections are specified for each road
type:
! Ultimate capacity per lane (QLult) (PCSE/lane/hr). The ultimate capacity for the road
section Qult = QLult*NLANES
! Free flow capacity as a proportion of the ultimate capacity (XQ1)
! Nominal capacity as a proportion of the ultimate capacity (XQ2)
! Speed at ultimate capacity (Sult) (km/hr)
Free speed value for each vehicle type is determined internally using the model described in
Part E.
The speed at nominal capacity is estimated to be 85% of the free speed of the slowest vehicle
in the traffic stream.
The maximum acceleration noise (σmaxr) represents the maximum standard deviation of
acceleration for each road type. This is required for modelling the effect of speed change
cycles (that is, speed variations along the road) on vehicle operating costs. In addition to
driver behaviour, speed fluctuations are sensitive to road geometry, road condition, the
presence of NMT, roadside friction, intersections, etc. (See Part E).

Version 1.0
7 Hourly flow-frequency distribution data
7.1 Basic concepts

There is a need to take account of the differing levels of traffic congestion at different hours of
the day, and on different days of the week and year. Therefore the number of hours of the year
for which different ranges of hourly flows are applicable needs to be considered. By defining
the distribution of hourly flows over the 8760 (365 days x 24 hours per day) hours of the year,
the AADT data can be converted to hourly flows. Congestion analysis can then be undertaken
for a number of hourly traffic flow levels, and the results combined to represent the full year.
Because congestion delays and costs are greatest during the highest-flow hours, particular
attention should be paid to these hours. These highest-flow hours should be divided into
periods of shorter duration. Figure B1.2 shows a typical example of a flow-frequency
distribution in which the number of hours in a year are divided into five traffic flow levels or
periods.
7.2 Data to be specified

Hourly flow-frequency distribution data is specified for each road use category. This reflects
the fact that the predominant use (see Section 2.1) of different roads requires different shapes
of flow-frequency distribution curves. The curves are defined in terms of the number of hours
per year that the traffic volume is at a certain percentage of the AADT. Each specified flow-
frequency distribution is referred to as traffic flow pattern, and can be assigned to a number
of road sections with similar use pattern.
F lo w P e rio d s
M ean Peak
H o u rly F lo w -fre q u e n c y N e x t to P e a k
F lo w d istrib u tio n c u rv e
M e d iu m flo w
(% A A D T )
N e x t to L o w
O ve rn ig h t
C u m u la tiv e n u m b e r o f h o u rs in th e Y e a r (8 7 6 0 )
Figure B1.2 Hourly flow-frequency distribution
Traffic flow patterns are specified as follows:

! Name

Version 1.0
! Number of flow-frequency periods
The number of flow periods into which the total number of hours in a year (8760) are
divided. Only one flow period needs to be specified for aggregate or network level
analyses. A uniform traffic composition and value of travel time across all flow-
frequency periods is assumed.
! The number of hours in each flow-frequency period, p (HRYRp) and
either
! Hourly traffic flow in each period as a proportion of AADT (HVp)
or
! The percentage of AADT in each flow-frequency period, p (PCNADTp)
When specifying a flow-frequency distribution, the following conditions must be satisfied:

! The sum of the number of hours in each flow-frequency period (HRYRp) should be 8760
! The sum of PCNADTp over all the flow periods should be 100
! The sum of the products of HVp and HRYRp (over all the traffic flow periods) divided by
365 should be equal to 1.00 ± 0.05
Table B1.3 shows typical values of hourly flow-frequency distribution data for three road use
categories: seasonal, commuter and inter-urban.
Table B1.3 Examples of hourly traffic flow distribution data
Flow Period (p) HRYR HV
(hours) Seasonal Commuter Inter-urban
1 87.6 0.18 0.13 0.09
2 350.4 0.14 0.12 0.08
3 613.2 0.10 0.10 0.07
4 2978.4 0.05 0.07 0.05
5 4730.4 0.02 0.01 0.03
Table B1.4 shows typical values of hourly flow-frequency distribution data for the three road-
use categories with hourly flows expressed as a percentage of AADT.
Table B1.4 Examples of hourly traffic flow distribution data
Flow Period (p) HRYR PCNADT (%)
(hours) Seasonal Commuter Inter-urban
1 87.6 4.25 3.05 2.17
2 350.4 13.24 11.33 7.59
3 613.2 16.60 16.55 11.64
4 2978.4 40.32 56.26 40.24
5 4730.4 25.59 12.81 38.36

Version 1.0
The parameter PCNADT is converted to hourly traffic flow as a proportion of AADT using
the following expression:
365 * PCNADTp
HVp = ...(7.1)
100 * HRYR p
where:
HVp hourly traffic flow in period p, as a proportion of AADT

PCNADTp percentage of AADT in period p
HRYRp number of hours per year in period p
The data that describes the hourly traffic flows and the volume-capacity ratios are required for
modelling congestion effects on vehicle speeds and vehicle operating costs. Therefore the key
parameters are as follows:
! Qp
The traffic flow, in PCSE per hour, during traffic flow period p.
! VCRp
The volume-capacity ratio for traffic flow period p.

The traffic flow during each flow period is calculated as follows:
K
Qp =
∑ HV
k =1
p * PCSE k * AADTk ...(7.2)
where:
Qp hourly traffic flow in period p (PCSE per hour)

AADTk annual average daily traffic of vehicle type k
PCSEk passenger car space equivalent of vehicle type k
The volume-capacity ratio during each traffic flow period is expressed as follows:
Qp
VCR p = ...(7.3)
Qult
where:
VCRp Volume-to-capacity ratio for traffic flow in period p (dimensionless)
All the other parameters are as defined previously.

Version 1.0
7.3 Limits on traffic flows

The annual growth of traffic over the analysis period can lead to high values of AADT and
hourly traffic flows, and a change in the hourly flow-frequency distribution. For realistic
analyses, the traffic flows in terms of AADT and hourly traffic flows in each flow period
should be limited to the capacity of the road. The following flow-capacity limits are therefore
included:
For each year of the analysis period, the annual average daily traffic on a road section is
limited as follows:
AADTy ≤ AADTylim
and:
24 * Qult
AADTylim = K
...(7.4)
∑ (PCSE
k =1
k * MIX k )
where:
AADTy the total AADT in year y (veh/day)

AADTylim the limiting road capacity in year y (veh/day)
Qult the ultimate road capacity in (PCSE/hr)
MIXk the proportion of vehicle type k in the traffic on the road in year y
PCSEk the passenger car space equivalent of vehicle type k
Since the proportion of each vehicle type in the traffic stream (MIXk) may vary each year, and
the ultimate road capacity Qult may also change due to a works intervention, the value of
AADTylim is calculated for each year of the analysis period. A check on the AADT limit is
carried out within the traffic module for every analysis year. Necessary adjustments to the
values of the traffic data are then made before passing on the data for use in the subsequent
modules. If the projected AADTy is higher than the AADT limit for the road, AADTy will be
set equal to AADTylim, and the analysis in year y will be performed using AADTylim. In these
circumstances, a printed warning is included in the HDM-4 outputs.
In reality, changes in the hourly flow-frequency distribution take place on a continuous basis
as traffic volume on a road increases with time. As a result, adjustments to the values of the
flow-frequency distribution parameters HVp and HRYRp are required. However, for the
purposes of analyses, the hourly flow-frequency distribution is only changed under one of the
following conditions:
1 When a spill-over from one high flow period to the next period occurs; that is, when the
hourly traffic flows computed for a high flow period exceeds the ultimate capacity of the
road, the excess flows will spill-over into the next highest flow period, so that some peak
spreading occurs. In this circumstance, new values of HVp are calculated, but HRYRp
values remain unchanged in order to simplify the analysis.
2 An intervention occurs that changes road users' pattern of journey time (for example, an
increase in road capacity that reduces traffic congestion).

Version 1.0
8 References
Methods for Economic Evaluation of Highways Investments and Maintenance
Transport Division, Transportation, Water & Urban Development Department
World Bank, Washington D.C., USA
TRRL Overseas Unit, (1988)
A guide to road project appraisal, Road Note 5
Transport and Road Research Laboratory
Crowthorne, Berkshire, UK
HDM-III Model.

Version 1.0
Contents

Descriptions: Part C
C1 Modelling concepts and approach

1 Introduction C1-1
2 Pavement classification C1-2
3 The modelling approach C1-7
3.1 Classes and types of models C1-7
3.2 Pavement distresses C1-7
3.3 Side-drains deterioration C1-8
3.4 Shoulders deterioration C1-8
4 Key variables affecting deterioration C1-9
4.1 Climate and environment C1-9
4.2 Traffic C1-11
4.3 Pavement history C1-11
4.4 Other road related variables C1-11
5 References C1-12
C2 Bituminous Pavements
1 Introduction C2-1
2 Model framework and logic C2-2
2.1 Classification and concepts C2-2
2.2 Computational logic C2-5
3 Pavement strength C2-11
3.1 Adjusted structural number C2-11
3.2 Seasonal and drainage effects C2-13
3.3 User input options C2-16
4 Construction quality C2-18
5 Crack modelling C2-20
5.1 Structural cracking C2-20
5.2 Transverse thermal cracking C2-27

Version 1.0
PART C ROAD DETERIORATION MODELS CONTENTS
5.3 Total areas of cracking C2-29

6 Ravelling C2-30
6.1 Initiation C2-30
6.2 Progression C2-30
7 Potholing C2-33
7.1 Initiation C2-33
7.2 Progression C2-34
8 Edge-break C2-36
9 Damaged and undamaged surface area C2-38
9.1 The logic C2-38
9.2 Distress values at the end of the year C2-39
9.3 Total damaged surface area C2-45
10 Rut depth C2-47
10.1 Initial densification C2-47
10.2 Structural deformation C2-48
10.3 Plastic deformation C2-49
10.4 Surface wear C2-50
10.5 Total rut depth C2-50
10.6 Standard deviation of rut depth C2-51
11 Roughness C2-52
11.1 Structural C2-52
11.2 Cracking C2-53
11.3 Rutting C2-53
11.4 Potholing C2-54
11.5 Environment C2-54
11.6 Total change in roughness C2-55
12 Pavement surface texture C2-57
12.1 Texture depth C2-57
12.2 Skid resistance C2-58
13 Calibration factors C2-61
14 References C2-62
C3 Concrete Pavements
1 Introduction C3-1
2 Modelling framework and logic C3-2
2.1 Concrete pavement structure C3-3
2.2 Concrete pavement distress modes C3-5
2.3 Primary modelling parameters C3-9
2.4 Computational procedure C3-10

Version 1.0
3 Structural characteristics C3-12

3.1 Introduction C3-12
3.2 Properties of the materials C3-12
3.3 Drainage conditions C3-15
3.4 Percentage of reinforcement steel C3-17
3.5 Load transfer efficiency C3-17
3.6 Widened outside lanes C3-18
4 Cracking C3-19
4.1 Jointed plain concrete pavements C3-19
4.2 Jointed reinforced concrete pavements C3-28
5 Faulting C3-30
5.1 JP concrete pavements without load transfer dowels C3-30
5.2 JP concrete pavements with load transfer dowels C3-31
6 Spalling C3-35
7 Failures C3-38
8 Serviceability loss C3-39
8.2 Continuously reinforced concrete pavements C3-39
9 Roughness C3-41
9.3 Continuously reinforced concrete pavements C3-42
11 References C3-44
C4 Unsealed Roads
1 Introduction C4-1
2 Modelling logic C4-2
2.1 Classification, concepts, and logic C4-2
2.2 Primary model parameters C4-3
2.3 Basic computational procedure C4-5
2.4 Initialisation of variables C4-6
3 Road roughness C4-7
3.1 General C4-7
3.2 Roughness progression C4-7
3.3 Effect of compaction on roughness progression C4-9

Version 1.0
3.4 Effect of grading C4-10

3.5 Average roughness during the analysis year C4-11
3.6 Roughness cycle 'steady state' C4-13
4 Material loss C4-14
5 Passability C4-15
7 References C4-17
Analytical Framework and Model Descriptions iv

Version 1.0
Part C Road Map
Analytical
AnalyticalFramework
Framework
and
andModel
Model
Descriptions
Descriptions
Introduction
Introduction
Part
PartAA
Traffic
Traffic
Part
PartBB
RD
RDModel
Model WE
WEModel
Model RUE
RUEModel
Model SEE
SEEModel
Model
Part
PartCC Part
PartDD Part
PartEE Part
PartFF
Economic
EconomicAnalysis
Analysis
PartGG
Nomenclature
Nomenclature
PartHH

Glossary
Part
PartI I
Figure C Analytical Framework and Model Descriptions Road Map

Version 1.0
PART C ROAD DETERIORATION MODELS
C1 Modelling concepts and approach
1 Introduction
Road deterioration is broadly a function of the original design, material types, construction
quality, traffic volume, axle load characteristics, road geometry, environmental conditions,
age of pavement, and the maintenance policy pursued.
HDM-4 includes relationships for modelling Road Deterioration (RD) and Road Works
Effects (WE). These are used for the purpose of predicting annual road condition and for
evaluating road works strategies. The relationships should link standards and costs for road
construction and maintenance to road user costs through road user cost models. In HDM-III
these relationships were combined into a single module called the Road Deterioration and
Maintenance Effects (RDME), as described by Watanatada et al. (1987). In HDM-4, this
module has been separated in order to address properly the expanded scope of modelling Road
Deterioration and Works Effects. The analysis now includes:
! Physical environments (climatic zones)
Encompassing cold (freeze/thaw) climates, very high temperatures and a very wide range
of temperature variations such as desert conditions, and very high moisture regimes and
arid conditions.
! Rigid/concrete and semi-rigid pavements, and a wide range of flexible pavements.
! Models for the following distresses:
Edge-break, texture depth and skid resistance.

! Road shoulders and the impact on non-motorised transport, and side-drains and
the effects on pavement strength.
! Road capacity improvement and a broader range of maintenance techniques for
different pavement types.
This chapter deals with the pavement classification system used, and describes the RD
modelling approach for the different road surface classes considered in HDM-4. The key
variables that affect road deterioration (in particular those that are associated with climate and
environment) are also discussed. A comprehensive classification of climate in terms of
temperature and moisture is also given. The modelling of Road Works Effects is described in
Part D.
Analytical Framework and Model Descriptions C1-1

Version 1.0
PART C ROAD DETERIORATION MODELS C1 MODELLING CONCEPTS AND APPROACH
2 Pavement classification
A versatile framework of pavement classification is required to cater for the expanded scope
of RD and WE analysis. A system of classifying pavements has therefore been formulated
which uses broad definitions of road surfacing and roadbase types as illustrated in Table C1.1
(NDLI, 1995).

Version 1.0
Table C1.1 Pavement classification system in HDM-4
Surface Surface Pavement Surface Surface Base Base material

category class type type material type
AMGB AC, HRA, GB NG, CRS, WBM, etc.
AMAB RAC, PA, AB AB, EB, etc.
AMSB AM CM, etc. SB CS, LS, etc.
AMAP AP TNA, FDA, etc.

Bituminous AMRB RB JUC, RBC, CUC, etc.
STGB SBSD, PM, GB NG, CRS, WBM, etc.
STAB DBSD, SL, AB AB, EB, etc.
STSB ST CAPE, etc. SB CS, LS, etc.
STAP AP TNA, FDA, etc.
STRB RB JUC, RBC, CUC, etc.
JPGB VC, RC, GB NG, CRS, WBM, etc.
JPAB FC, PC, AB AB, EB, etc.

Paved
JPSB JP etc. SB CS, LS, etc.
JPAP AP TNA, FDA, etc.
JPRB RB JUC, RBC, CUC, etc.
JRGB VC, GB NG, CRS, WBM, etc.
JRAB FC, etc. AB AB, EB, etc.

Concrete
JRSB JR SB CS, LS, etc.
JRAP AP TNA, FDA, etc.
JRRB RB JUC, RBC, CUC, etc.
CRGB VC, GB NG, CRS, WBM, etc.
CRAB FC, etc. AB AB, EB, etc.
CRSB CR SB CS, LS, etc.
CRAP AP TNA, FDA, etc.
CRRB RB JUC, RBC, CUC, etc.
CBSG CB CB SG SA, NG, etc.
Block BRLC BR BR LC LC
SSGB SS SS CG LC, NG, etc.
GRUP GR LT, QZ, etc.
Unpaved Unsealed EAUP EA EA UP
SAUP SA SA

Version 1.0
Key:
Surface type Surface material

AM Asphaltic Mix AC Asphalt Concrete
ST Surface Treatment HRA Hot Rolled Asphalt
JP Jointed Plain RAC Rubberised Asphalt Concrete
JR Jointed Reinforced PA Porous Asphalt
CR Continuously Reinforced CM Cold Mix (Soft Bitumen Mix)
CB* Concrete Block SBSD Single Bituminous Surface Dressing
BR* Brick PM Penetration Macadam
SS* Set Stone DBSD Double Bituminous Surface Dressing
GR Gravel SL Slurry Seal
EA* Earth CAPE Cape Seal
SA* Sand VC Vibrated Concrete
RC Rolled Concrete
FC Fibre Concrete
PC Porous Concrete
LT Lateritic Gravel
QZ Quartzitic Gravel
Note:
Asterik (*) indicates that different types of material or construction pattern may be defined.
Base type Base material

GB Granular Base NG Natural Gravel
SB Stabilised Base WBM Water Bound Macadam
AP Asphalt Pavement EB Emulsified Base
RB Rigid (Concrete) Base CS Cement Stabilised
SG Sand/Gravel LS Lime Stabilised
LC Lean Concrete TNA Thin Asphalt Surfacing
CG Concrete/Gravel FDA Full Depth Asphalt
UP Unpaved – base types not applicable JUC Jointed Unbonded Concrete
RBC Reinforced Bonded Concrete
CUC Continuously Unbonded Concrete

Version 1.0
The definitions are as follows:

! Surface category
Divides all pavements into two groups:

❏ paved
❏ unpaved
These are mainly used for the reporting of network statistics.
! Surface class
Subdivides the paved category into bituminous, concrete and block surfaces; together
with the unsealed class there are thus four classes that are used to define the family of
distress models used for performance modelling.
! Pavement type
Integrates surface and roadbase types. Each type is designated by a four-character code,
combining the surface and roadbase type codes.
! Surface type
Divides bituminous surfacings into two types:

❏ asphaltic mix (AM)
❏ surface treatment (ST)
Divides concrete surfacings into three types:
❏ jointed plain (JP)
❏ jointed reinforced (JR)
❏ continuously reinforced (CR)
Divides three types of block:
❏ concrete (CB)
❏ brick (BR)
❏ set stone (SS)
Divides three types of unsealed surfacings:
❏ gravel (GR)
❏ earth (EA)
❏ sand (SA)
A surface type is designated by a two-character code.
! Base type
There are eight generic types, including those which allow for overlays of asphalt on
concrete and vice versa. Each base type is designated by a two-character code.
! Surface material
Defines more specific surface types (for example, different types of asphalt mixes).
These are user definable.

Version 1.0
! Base material
Allows the user to specify more detailed characteristics of roadbase types.

Note that during an analysis period, the road surface class and pavement type might change
depending on the types of works applied to the pavement (see Part D). For example, the initial
pavement type for a section may be AMGB (asphaltic mix surface on granular base); if an
asphaltic overlay is applied the pavement type will change to AMAP (asphaltic mix surface on
asphalt pavement) and different model parameters will apply. If the same initial pavement is
given a surface treatment it will change to STAP (surface treatment on asphalt pavement).

Version 1.0
3 The modelling approach
3.1 Classes and types of models

The two general classes of models used for Road Deterioration (RD) and Works Effects (WE)
analyses are mechanistic and empirical (NDLI, 1995). Mechanistic models use sound
fundamental theories of pavement behaviour for their development; but they are usually very
data intensive and rely on parameters which are difficult to quantify in the field. Empirical
models are usually based on statistical analyses of locally observed deterioration trends, and
may not be applicable outside the specific conditions upon which they are based.
To minimise these problems Paterson (1987) adopted a structured empirical approach for
developing the HDM-III RDME model. This was based on identifying the functional form and
primary variables from external sources and then used various statistical techniques to
quantify their impacts. This had the advantage that the resulting models combined both the
theoretical and experimental bases of mechanistic models with the behaviour observed in
empirical studies. The RD and WE relationships included, in HDM-4, are therefore mainly
structured empirical models.
There are two types of models that can be used for predictive purposes:
! Absolute models
! Incremental models
Absolute models predict the condition (or distress) at a particular point in time as a function of
the independent variables. Incremental models give the change in condition from an initial
state as a function of the independent variables.
The families of pavement performance models used are based on the road surface classes:
Bituminous incremental models (described in Chapter C2)

Concrete absolute models (described in Chapter C3)
Unsealed incremental models (described in Chapter C4)
Block incremental models (not included in this software release)
3.2 Pavement distresses

Pavement deterioration manifests itself in various kinds of distresses, each of which should be
modelled separately. Table C1.2 gives a summary of the pavement defects that are modelled
in HDM-4. As each mode of distress develops and progresses at different rates in different
environments, it is important that the RD relationships should be calibrated to local conditions
before using them for road investment analyses. To facilitate this, the relationships include a
number of user-definable deterioration factors to change the scale of a particular distress. The
model coefficients should be used to adjust the rates of deterioration for different types of
pavement material.
In order to model road deterioration properly it is required that homogeneous road sections in
terms of physical attributes and condition should be identified so that a particular set of RD
relationships can be applied. The basic unit of analysis is therefore the homogeneous road
section, to which several investment options can be assigned for analysis.

Version 1.0
Table C1.2 Pavement defects modelled in HDM-4
Bituminous Concrete Block* Unsealed

Cracking Cracking Rutting Gravel loss
Ravelling Joint spalling Surface texture Roughness
Potholing Faulting Roughness
Edge repair Failures
Rutting Serviceability loss
Surface texture Roughness
Skid resistance
Roughness
* Not implemented in this HDM-4 release
3.3 Side-drains deterioration

The condition of the drains will deteriorate unless they are maintained adequately through, for
example, routine maintenance. The deterioration of side-drains has the effect of reducing
pavement strength and accelerating its deterioration. Drain life is expressed as a function of
terrain, drain type, climate type and the maintenance policy pursued. A number of drain types
are considered in RD modelling (see Chapters C2 and C3).
3.4 Shoulders deterioration

The modelling of road shoulder deterioration is required in order to assess the effect on the
rate of pavement deterioration; and the impact on non-motorised transport and traffic flow in
terms of Road User Costs. It is proposed to include this feature in a future release of HDM-4.

Version 1.0
4 Key variables affecting deterioration

The key variables that are used in the various deterioration models are associated with the
following:
! Climate and environment
! Traffic
! Pavement history
! Road geometry
! Pavement structural characteristics
4.1 Climate and environment

The climate in which a road is situated has a significant impact on the rate at which the road
deteriorates. Important climatic factors are related to temperature, precipitation and winter
conditions. This section describes the principal climatic data that is used to model the
deterioration of the different categories of roads considered in HDM-4.
4.1.1 Classification
It is necessary for the user to define climatic and environment information as per Table C1.3
and Table C1.4:
Table C1.3 Moisture classification

classification moisture precipitation
index (mm)
Arid Very low rainfall, high evaporation -100 to -61 < 300
Sub-humid Moderate rainfall, or strongly seasonal -20 to +19 800 to 1600

rainfall
Humid Moderate warm seasonal rainfall +20 to +100 1500 to 3000
Per-humid High rainfall, or very many wet-surface days > 100 > 2400
Table C1.4 Temperature classification
Temperature Description Temperature

classification range (oC)
Tropical Warm temperatures in small range 20 to 35
Sub-tropical - hot High day cool night temperatures, hot-cold seasons -5 to 45
Sub-tropical - cool Moderate day temperatures, cool winters -10 to 30
Temperate - cool Warm summer, shallow winter freeze -20 to 25
Temperate - freeze Cool summer, deep winter freeze -40 to 20

Version 1.0
4.1.2 Precipitation
The Mean Monthly Precipitation (MMP) is used in the modelling of bituminous pavement
deterioration and unsealed road deterioration, and is expressed in mm/month. The annual
average precipitation (PRECIP) is used in the modelling of concrete pavement deterioration
and is expressed in inches/year.
4.1.3 Thornthwaite moisture index

The Thornthwaite moisture index (MI) is defined as follows (LAST, 1996):
100 * SWAT − 60 * DWAT

MI = I h - 0.6 * I a = ...(4.1)
NWAT
where:
MI Thornthwaite moisture index

Ih humidity index
Ia aridity index
SWAT excess of water (mm)
DWAT water deficiency (mm)
NWAT necessary water (mm)
It is important to know if a given place is continually wet or dry, or if it is wet in a given

season and dry in another. The Moisture Index is capable of indicating how wet or dry is a
given climate zone, but it is not capable of distinguishing climates with or without seasonal
dampness variations.
Wet climates will have a positive Moisture Index; on the other hand dry climates will have a
negative index. The Thornthwaite Moisture Index indicates the free humidity in a particular
area.
4.1.4 Freezing index

The freezing index (FI) is defined as the difference between the mean ambient temperature
and 0ºC (degrees per day). The freezing index is negative when the ambient temperature is
below 0ºC and positive otherwise.
The freezing index is calculated as:
ndays
FI =
∑ ABS[MIN(TEMP,0)]
i =1
...(4.2)
where:
FI freezing index
TEMP temperature (oC)
Ndays number of days in one freezing season

Version 1.0
Note that FI is only required as input data for the two temperate temperature zones, and are
used in modelling the performance of concrete pavements.
4.1.5 Temperature range

Temperature range (TRANGE) is defined as the mean monthly ambient temperature range. Its
calculation is based on the temperature ranges for each of the 12 months of the year, hence the
difference between the maximum and minimum temperature for each month. The 12 values
obtained are then averaged to obtain the variable TRANGE, which is used for modelling
concrete pavements.
4.1.6 Days with temperatures greater than 90ºF

The number of days, in a year, in which the ambient temperature exceeds 90ºF (32ºC) is
denoted as DAYS90. This variable is required for modelling the performance of concrete
pavements.
4.2 Traffic
The primary traffic-related variables that affect road deterioration include the number and
types of vehicles using the road, and axle loading characteristics of the different vehicle types.
Details of the variables required are described in Part B, and the way in which they are used in
the various deterioration relationships is defined in the appropriate Sections of this document.
4.3 Pavement history

The required variables refer to the age of pavement. These variables are related to the previous
maintenance, rehabilitation and construction works carried out on the pavement, and have
been discussed in the appropriate sections of this document.
4.4 Other road related variables

The other key variables affecting pavement performance are related to road geometry,
pavement structural characteristics and material properties. These are also described in the
appropriate Sections.

Version 1.0
5 References
LAST (1996)
Modelling road design and maintenance effects for pavements in HDM-4
Final Report, FICEM, Latin American Study Team, International Study of Highway
Development and Management Tools, Santiago, Chile
NDLI (1995)
Modelling Road Deterioration and Maintenance Effects in HDM-4
Final Project Report, Asian Development Bank, RETA 5549
N. D. Lea International Limited, Vancouver, Canada
Paterson W. D. O., (1987)
Road Deterioration and Maintenance Effects: Models for Planning and Management
The Highway Design and Maintenance Standards Series
World Bank, Johns Hopkins Press, Baltimore, USA
The Highway Design and Maintenance Standards Model - Volume 1 Description
World Bank, Johns Hopkins Press, Baltimore, USA

Version 1.0
C2 Bituminous Pavements
1 Introduction
This chapter describes the detailed modelling of bituminous pavement deterioration in
HDM-4 (see Figure C2.1).
Road
RoadDeterioration
Deterioration
Models
Models
Pavement Bituminous Concrete Unsealed

UnsealedRoads
Pavementtypes
types Bituminous Concrete Roads
Pavements
Pavements Pavements
Pavements
Chapter C-1
Chapter C-1 Chapter C-4
Chapter C-4
Chapter
ChapterC-2
C-2 Chapter
ChapterC-3
C-3
Figure C2.1 Road Deterioration Modules
The development of the road deterioration models was based on three previous documents:
1 Riley and Bennett (1995 & 1996) - based on Paterson (1987)
2 Watanatada et al. (1987)
3 NDLI (1995)
A series of formal workshops followed. These were held at the University of Birmingham
(UoB), UK, in April 1996, December 1996 and October 1997. Informal meetings were then
held in Washington, the University of Birmingham and the Transport Research Laboratory
(TRL), UK that supplemented these workshops.
As a result of these discussions, major changes were regularly proposed to the road
deterioration models which were presented in various versions of the fourth and fifth draft
specifications (Morosiuk, 1996 & 1998a). Following on from the beta testing of the HDM-4
software in November 1998, a sixth then a seventh draft of specifications were produced. A
further workshop was held at TRL in June 1999 to resolve outstanding issues. The decisions
reached at this workshop have been incorporated in the eighth draft version of the
specifications. The main contributors to the specifications were Paterson (IBRD), Morosiuk
(TRL), Riley (Riley Partnership), Odoki and Kerali (UoB).
Section 2 provides an overview of the Road Deterioration modelling framework. This is
followed by the relationships and default coefficient values for each of the distresses to be
modelled. The model coefficient values are stored in data files instead of having them hard
coded into the model. This facilitates local calibration and adaptation. The HDM-4 model
has more calibration factors than the previous HDM-III model.

Version 1.0
PART C ROAD DETERIORATION MODELS C2 BITUMINOUS PAVEMENTS
2 Model framework and logic
2.1 Classification and concepts

The road deterioration framework developed for HDM-4 is much more flexible than in HDM-
III and is able to handle a wider range of pavement types. This has been accomplished by
providing a single set of generic models whose coefficient values are altered based on the
surface and base type. The pavement classification system that forms the basis for defining
the model framework is shown in Table C2.1.
The formal structure of the framework is comprised of the non-shaded cells in Table C2.1.
The pavement type is defined by the combination of the surface type and base type. This is
given in the right hand column of Table C2.1.
Within a given pavement type, there are various combinations of surface and base materials.
As the performance of the pavement can be anticipated to be a function of these materials, the
user is able to associate model coefficients with each combination of surface and base
materials. The same basic models for the pavement type is used with the different coefficient
values.
Table C2.1 HDM-4 Bituminous pavements classification system

AC CRS
GB AMGB
HRA GM
PMA AB AB AMAB
RAC CS
AM SB AMSB
CM LS
PA TNA
AMAP
SMA AP FDA
Xx
CAPE CRS
GB STGB
DBSD GM
SBSD AB AB STAB
SL CS
ST SB STSB
PM LS
Xx TNA STAP
AP
FDA
Note: The modelling of AM and ST surfacings on concrete pavements, that is, AMRB and STRB, is
not included in this software release.
The abbreviations in Table C2.1 are described in Table C2.2.

Version 1.0
Table C2.2 Descriptions of surface and base materials


Mix)
PA Porous Asphalt
AM Asphalt Mix
Asphalt

Concrete
SMA Stone Mastic
CAPE Cape Seal

Surface Dressing

Surface Dressing
SL Slurry Seal

AP Asphalt Pavement NG Natural Gravel
The available models are constructed from different factors. Many are created from surface
and base types, whilst others are created from surface materials. Accordingly, the modelling
is done in terms of surface material and base type, even though base materials can be
specified. HDM-4 contains default coefficient values for the bituminous pavement types given
in Table C2.3.

Version 1.0
Table C2.3 Generic HDM-4 bituminous pavement types
Pavement Surface Base type Description of pavement types

type type
AMGB AM GB Asphalt Mix on Granular Base
AMAB AM AB Asphalt Mix on Asphalt (Dense Bitumen Macadam) Base
AMSB AM SB Asphalt Mix on Stabilised Base
AMAP AM AP Asphalt Mix on Asphalt Pavement
STGB ST GB Surface Treatment on Granular Base
STAB ST AB Surface Treatment on Asphalt (Dense Bitumen Macadam) Base
STSB ST SB Surface Treatment on Stabilised Base
STAP ST AP Surface Treatment on Asphalt Pavement
Currently, there are no coefficient values to differentiate between the performances of

different base materials, so all materials of a given base type are assigned the same coefficient
values. Each combination of surface and base material results in a set of coefficient values
associated with the pavement.
NDLI (1995) gives definitions of the characteristics used to define different types of
pavements into the above framework and alternative terminology applied to the same
pavement materials.
The resets for pavement type after maintenance works are discussed in detail in the Road
Works Effects (see Part D). These resets are summarised in Table C2.4.

Version 1.0
Table C2.4 Pavement type resets after maintenance works

treatment

STSB1

AMSB1
intermediate
surface layer
base
Source: NDLI (1995)
Notes:
1 The pavement type will depend on the critical thickness (Hmin) of the existing
bituminous surfacing, that is user-definable in HDM-4 Configuration
** Indicates that this, two character variable, is dependent on the specific works activity
(that is, operation
N/A not applicable
2.2 Computational logic
2.2.1 Pavement distress modes

Road deterioration is predicted through eight separate distress modes, namely:
! Cracking (see Section 5)
! Ravelling (see Section 6)
! Potholing (see Section 7)
! Edge-break (see Section 8)
! Rutting (see Section 10)
! Roughness (see Section 11)
! Texture depth (see Section 12.1)
! Skid resistance (see Section 12.2)

Version 1.0
These are defined in Table C2.5, and can be considered under the following three categories:
! Surfacing distress
This category comprises:

❏ Cracking
❏ Ravelling
❏ Potholing
❏ Edge-break
The first three distress modes are characterised by two phases referred to as initiation and
progression. The initiation phase is the period before surfacing distress of a given mode
or severity develops. The progression phase refers to the period during which the area
and severity of distress increases. Edge-break is modelled only through its continuous
progression.
! Deformation distress

❏ Rutting
❏ Roughness
Deformation distress modes are continuous, and represented by only progression
equations. As they are partly dependent upon the surfacing distress, they are computed
after the change of surfacing distress in the analysis year has been calculated.
! Surface texture

❏ Texture depth
❏ Skid resistance
Surface texture distress modes are continuous, and like deformation distress modes they
are modelled only through their progression.
2.2.2 Primary modelling parameters

The primary variables used from one analysis year to the next may be grouped as described
below. The road characteristics at the beginning of the analysis year are initialised either
from input data if it is the first year of the analysis or the first year after construction,
or otherwise from the result of the previous year’s maintenance and improvement
works.
Pavement structural characteristics

These include measures of pavement strength, layer thickness, material types, construction
quality, and subgrade stiffness.
The RD models require as input data the thickness of new and old bituminous surfacing
layers. An original pavement that has not been resurfaced or overlaid since it was
constructed/reconstructed has a new surfacing and no old surfacing. For a pavement that has
been resurfaced or overlaid, the following relationship applies:
HSOLD 2 = HSNEW1 + HSOLD 1 - MLLD ...(2.1)

Version 1.0
HSNEW2 = user specified value ...(2.2)
where:
HSOLD2 thickness of old surfacing after works (mm)

HSNEW1 thickness of the most recent surfacing (mm)
HSOLD1 total thickness of previous underlying surfacing layers (mm)
MLLD mill depth (mm)
HSNEW2 thickness of new surfacing after works (mm). This is the user-specified
thickness when an intervention is to be made

Version 1.0
Table C2.5 Definitions of distress measures
Measure Definition
Area (of distress) Sum of rectangular areas circumscribing manifest distress (line cracks are
assigned a width of 0.5 m), expressed as a percentage of the carriageway area
All cracking Narrow and wide structural cracking inclusive
Narrow cracking Interconnected or line cracks of 1-3 mm crack width (equivalent to AASHTO
Class 2)
Wide cracking Interconnected or line cracks of 3 mm crack width or greater, with spalling
(equivalent to AASHTO Class 4)
Indexed cracking Normalised sum of AASHTO Classes 2 to 4 cracking weighted by class, see
Section 5.3
Transverse thermal cracking Unconnected crack running across the pavement
Ravelling Loss of material from wearing surface
Pothole Open cavity in the road surface with at least 150 mm diameter and at least 25
mm depth
Edge-break Loss of bituminous surface material (and possibly base materials) from the edge
of the pavement
Rutting Permanent or unrecoverable traffic-associated deformation within pavement

layers which, if channelled into wheelpaths, accumulates over time and
becomes manifested as a rut
Rut depth Maximum depth under 2 m straightedge placed transversely across a wheelpath
Roughness Deviations of a surface from a true planar surface with characteristic

dimensions that affect vehicle dynamics, ride quality, dynamic loads and
drainage (ASTM E-867-82A) – typically in the ranges of 0.1 to 100 m
wavelength and 1 to 100 mm amplitude
IRI International Roughness Index, the reference measure expressing roughness as

a dimensionless average rectified slope statistic of the longitudinal profile and
defined in Sayers et al. (1986)
Mean Texture depth The average depth of the surface of a road surfacing expressed as the quotient
of a given volume of standardised material [sand (sand patch test), glass
spheres] and the area of that material spread in a circular patch on the surface
being tested, (PIARC, 1997)
Skid resistance Resistance to skidding expressed by the sideways force coefficient (SFC)
measured using the Sideways Force Coefficient Routine Investigation Machine
(SCRIM)
Source: Watanatada et al. (1987)
Road condition
Pavement and side-drain condition data at the beginning of the first analysis year or the first
year after construction are required. The data items for surfacing and deformation distress
modes and surface texture, are as described in Table C2.5.
The pavement conditions at the end of the year (that is, before road works) are predicted as
follows:

Version 1.0
[CONDITION]b = [CONDITION]a +∆ [CONDITION] ...(2.3)
[CONDITION]av = 0.5 * {[CONDITION]a + [CONDITION]b } ...(2.4)
where:
[CONDITION]b condition at the end of the year

[CONDITION]a condition at the start of the year
∆[CONDITION] change in condition during the year
[CONDITION]av average condition for the year
Pavement history
The required data items refer to pavement ages, and these are related to the previous
maintenance, rehabilitation and construction works carried out on the pavement.
There are four variables defining the age of the pavement used in the models; AGE1, AGE2
AGE3 and AGE4. These variables are defined as follows:
1 AGE1 is referred to as the preventive treatment age. It is defined as the time, in number
of years, since the latest preventive treatment, reseal, overlay (or rehabilitation),
pavement reconstruction or new construction activity.
2 AGE2 is referred to as the surfacing age. It is defined as the time, in number of years,
since the latest reseal, overlay, pavement reconstruction or new construction activity.
3 AGE3 is referred to as the rehabilitation age. It is defined as the time, in number of
years, since the latest overlay, pavement reconstruction or new construction activity.
4 AGE4 is referred to as the base construction age. It is defined as the time, in number of
years, since the latest reconstruction that involves the construction of a new base layer
or new construction activity.
Road geometry and environment

These include carriageway and shoulder widths, vertical alignment and the mean monthly
precipitation.
Traffic
The required traffic data are the flow of all vehicle axles (YAX) and the flow of equivalent
standard axle loads (YE4), both expressed on an annual basis in millions per lane. The data are
calculated for each analysis year based on the user-specified trafficand vehicle characteristics.
The annual average traffic speed and the average speed of heavy vehicles are also required
in some RD relationships
2.2.3 Computational procedure

The overall computational logic for modelling the deterioration of each road section in each
analysis year can be summarised by the following steps:
1 Initialise input data and the conditions at the beginning of the year
2 Compute pavement strength parameters

Version 1.0
3 Calculate the amount of change in each surfacing distress mode during the analysis year
in the following order:
(a) Cracking
(b) Ravelling
(c) Potholing
(d) Edge-break
4 Check that the total damaged and undamaged carriageway surface area equals 100%
based on the limits defined for each distress mode, and determine the amount of each
surfacing distress mode at the end of the year and the average value for the year
5 Compute the change in each deformation distress mode during the year, and determine
the amount of the distress mode at the end of the year and the average value for the year
6 Compute the change in each surface texture distress mode during the year, and determine
the amount of the distress mode at the end of the year and the average value for the year
7 Store results for use in subsequent modules (that is, RUE, WE, SEE) and in the following
analysis year, and for reporting

Version 1.0
3 Pavement strength
3.1 Adjusted structural number

Pavement strength is characterised by the adjusted structural number, SNP, (Parkman and
Rolt, 1997). This has been derived from the modified structural number, which was adopted
as the pavement strength descriptor for HDM-III. The adjusted structural number applies a
weighting factor, which reduces with increasing depth, to the sub-base and subgrade
contributions so that the pavement strength for deep pavements is not over-predicted (a
concern with the use of the modified structural number). It is calculated as:
SNPs = SNBASU s + SNSUBA s + SNSUBG s ...(3.1)
SNBASU s = 0.0394
∑ ais hi
i =1
...(3.2)
 b 0 exp ( −b 3 z j ) b 1exp ( − (b 2 + b 3 )z j )  
m  +  − 
∑  − b3 (b 2 + b 3 )  
SNSUBA s = 0.0394 a js   ...(3.3)
 0
b exp ( −b z
3 j−1 ) b 1exp ( − (b 2 + b 3 )z j−1 )  
j =1 + 
 − b3 (b 2 + b 3 ) 
 
[
SNSUBG s = [b 0 - b 1exp (- b 2 z m )] [exp(- b 3 z m )] 3.51 log10 CBR s - 0.85 (log10 CBR s ) − 1.43
2
]
...(3.4)
where:
SNPs adjusted structural number of the pavement for season s

SNBASUs contribution of surfacing and base layers for season s
SNSUBAs contribution of the sub-base or selected fill layers for season s
SNSUBGs contribution of the subgrade for season s
n number of base and surfacing layers (i = 1, 2,…, n)
ais layer coefficient for base or surfacing layer i for season s
hi thickness of base or surfacing layer i (mm)
m number of sub-base and selected fill layers (j = 1, 2,…, m)
z depth parameter measured from the top of the sub-base (underside of base)
(mm)
zj depth to the underside of the jth layer (z0 = 0) (mm)
CBRs in situ subgrade CBR for season s
ajs layer coefficient for sub-base or selected fill layer j for season s
b0, b1, b2, b3 model coefficients

Version 1.0
The values of the model coefficients b0 to b3 are given in Table C2.6 and the values of the
layer coefficients ai and aj are given in Table C2.7.
Table C2.6 Adjusted structural number model coefficients
Pavement type b0 b1 b2 b3
All pavement types 1.6 0.6 0.008 0.00207
Table C2.7 Pavement layer strength coefficients
Layer Layer Condition Coefficient

type
ST Usually 0.2 ai = 0.20 to 0.40
hi < 30 mm, low stability and cold ai = 0.20

mixes
Surfacing hi > 30 mm, MR30 = 1500 MPa ai = 0.30
AM hi > 30 mm, MR30 = 2500 MPa ai = 0.40
hi > 30 mm, MR30 ≥ 4000 MPa ai = 0.45
Default ai = (29.14 CBR - 0.1977 CBR2 + 0.00045

CBR3) 10-4
GB CBR > 70, cemented sub-base ai = 1.6 (29.14 CBR - 0.1977 CBR2 +
0.00045 CBR3) 10-4
Base CBR < 60, max. axle load > 80kN ai = 0
AB, AP Dense graded with high stiffness ai = 0.32
SB Lime or cement ai = 0.075 + 0.039 UCS - 0.00088(UCS)2
Sub-base Granular aj = -0.075 + 0.184(log10 CBR) - 0.0444(log10

CBR)2
Cemented UCS > 0.7 MPa aj = 0.14
Notes:
1 The table reproduces information from the source, with the exception of the granular
sub-base coefficient
2 If the user quotes a CBR value for a stabilised (lime or cement) layer, the
corresponding granular coefficient should be used
3 Unconfined Compressive Strength (UCS) is quoted in MPa at 14 days
4 MR30 is the resilient modulus by the indirect tensile test at 30 °C
5 CBR is the California Bearing Ratio
Equation 3.4 above predicts negative values for the subgrade contribution below CBR = 3.
This is different to HDM-III where the values were set to 0 and reflects the detrimental impact
of weak subgrades on pavement performance.

Version 1.0
3.2 Seasonal and drainage effects

It has been well recognised that pavement strength changes during the course of a year due to
climatic effects. Hence, both seasonal and drainage effects have been included in the
modelling of road deterioration in HDM-4. The average annual strength of the pavement is
used in the deterioration models. This is estimated from the strength of the pavement during
the dry season and during the wet season, and the duration of each season. The user is
required to input the dry season SNP (or the wet season SNP) and the length of the dry season.
The average annual SNP is derived as follows:
SNP = f s SNPd ...(3.5)
where:
f
fs = ...(3.6)
[(1 - d) + d(f p )]1/p
and:
SNP Average annual adjusted structural number

SNPd dry season SNP
f SNPw / SNPd ratio
d length of dry season as a fraction of the year
p Exponent of SNP specific to the appropriate deterioration model (see Table
C2.8)
Table C2.8 Values of exponent p for calculating SNP
Distress Model p
Cracking Initiation of structural cracking 2.0
Initial densification 0.5

Rut depth
Structural deformation 1.0
Roughness Structural component 5.0
If only one season’s SNP value is available then the following relationship (Riley, 1996a)
should be used to calculate the wet/dry season SNP ratio. This relationship will also be used
to calculate the wet/dry season SNP ratio for each year of the analysis period, taking into
account changes in the drainage factor and the amount of cracking.
 [1 - exp(a 0 MMP )] 
f = K f 1 - (1 − a 2 DFa )(1 + a 3 ACRA a + a 4 APOTa ) ...(3.7)
 a1 
where:
f SNPw / SNPd ratio

Version 1.0
SNPw wet season SNP

SNPd dry season SNP
MMP mean monthly precipitation (mm/month)
DFa Drainage factor at start of analysis year
ACRAa total area of cracking at the start of the analysis year (% of total carriageway
area)
Kf Calibration factor for wet/dry season SNP ratio (range 0.1 to 10)
The default coefficient values a0 to a4 are given in Table C2.9.
Table C2.9 Default coefficient values for the seasonal SNP ratio
Coefficient a0 a1 a2 a3 a4
Default value -0.01 10 0.25 0.02 0.05
The drainage factor, DF, is a continuous variable whose value can range between 1 (excellent)
and 5 (very poor), depending on the type of drain (Paterson, 1998). The user will be required
to input the type of drain (as listed in Table C2.10) and the condition of the drain as excellent,
good, fair, poor or very poor.
Table C2.10 Suggested range of drainage factor values
Drain type Drain condition

Excellent Very poor
DFmin DFmax
Fully lined and linked 1 3
Surface lined 1 3
V-shaped – hard 1 4
V-shaped – soft 1.5 5
Shallow – hard 2 5
Shallow – soft 2 5
No drain - but required 3 5
No drain - not required 1 1
The minimum (excellent) and maximum (very poor) values for DF suggested for various
types of drain are given in Table C2.10. The value of DF for drains in a good, fair or poor
condition is determined by linearly interpolating between these values.
In some instances there will be an absence of drains. In situations where a drain is required the
value of DF ranges between 3 and 5, whereas in situations where a drain is unnecessary a
value of 1 for DF is suggested.
The condition of the drains will deteriorate unless they are well maintained, for example,
through routine maintenance. The incremental annual change in DF due to deterioration is

Version 1.0
given in Equation 3.8 below: (the change in DF due to maintenance, ∆DFw, is detailed in Part
D - Road Works Effects).
∆DFd = MAX {0, MIN [K ddf ADDF, (DFmax - DFa )] } ...(3.8)
and:
(DFmax - DFmin )
ADDF = ...(3.9)
Drain Life
where:
∆DFd annual change in DF due to deterioration

Kddf calibration factor for drainage factor
ADDF annual deterioration of DF
Drain Life life of the drain (years) (see Table C2.8)
Drain life has been expressed as a function of the terrain as given below. The proposed
default coefficient values a0 and a1 are given in Table C2.11 (Morosiuk, 1998b) for the
climatic categories classified by moisture (see Chapter C1 - Section 4).
Drain Life = K drain a 0 (1+ a 1RF ) ...(3.10)
where:
RF rise and fall (m/km)

Kdrain calibration factor for drain life
Table C2.11 Default coefficient values for drain life
Drain type Arid Semi-arid Sub-humid Humid Per-humid
a0 a1 a0 a1 a0 a1 a0 a1 a0 a1
Fully lined 20 -0.0033 20 -0.0033 13 -0.0031 6 -0.0022 5 -0.0027

and linked
Surface lined 20 -0.0033 15 -0.0031 8 -0.0017 5 -0.0027 4 -0.0033
V-shaped - 20 -0.0033 15 -0.0031 10 -0.0027 6 -0.0022 4 -0.0033

hard
V-shaped - 15 -0.0031 8 -0.0033 6 -0.0022 5 -0.0027 4 -0.0033

soft
Shallow - 15 -0.0031 6 -0.0022 5 -0.0027 4 -0.0033 3 -0.0022

hard
Shallow - soft 10 -0.0033 5 -0.0027 4 -0.0033 3 -0.0022 3 -0.0033
No drain - but 3.5 -0.0029 2.5 -0.0027 2 -0.0033 1.5 -0.0044 1.5 -0.0044
required
No drain - not 50 0 50 0 50 0 50 0 50 0
required

Version 1.0
3.3 User input options

Pavement strength may be input in the following forms:
1 SNP
2 Benkelman beam deflections
3 FWD deflections
4 Layer thickness, strength coefficients and subgrade CBR
Option 4 has been described by Equations 3.1 above to 3.4 above. If either of options 2 or 3
are used, the model will convert the input data to SNP as follows:
! Option 2 - Benkelman beam deflections
The relationships used to convert Benkelman beam deflections (DEF) to SNP values are
based on those in HDM-III (Paterson, 1987) and are given below:
Base is not cemented
−0.63
SNPs = 3.2 (DEFs ) + dSNPK ...(3.11)
Base is cemented
−0.63
SNPs = 2.2 (DEFs ) + dSNPK ...(3.12)
and:
dSNPK = 0.0000758 {MIN (63, ACX a ) HSNEW + MAX [MIN (ACX a − PACX,40 ),0] HSOLD}
...(3.13)
where:
DEFs benkelman beam rebound deflection under 80 kN axle load, 520 kPa
tyre pressure and 30°C average asphalt temperature for season s
(mm)
dSNPK reduction in adjusted structural number due to cracking
ACXa area of indexed cracking at the start of the analysis year (% of total
carriageway area)
HSNEW thickness of the most recent surfacing (mm)
PACX area of previous indexed cracking in old surfacing (% of total
carriageway area); that is, 0.62 (PCRA) + 0.39 (PCRW)
HSOLD total thickness of previous underlying surfacing layers (mm)
Some models need Benkelman beam deflection values. Where these are not user input,
DEF values are derived from SNP values using relationships originated from those in
HDM-III, that is:
❏ Base is not cemented
−1.6
DEFs = 6.5 (SNPK s ) ...(3.14)

Version 1.0
❏ Base is cemented
−1.6
DEFs = 3.5 (SNPK s ) ...(3.15)
and:
SNPK s = SNPs - dSNPK ...(3.16)
where:
SNPKs adjusted structural number due to cracking for season s
! Option 3 - FWD deflections
The central FWD deflection at 700 kPa is used as the equivalent Benkelman beam
deflection. The equations in Option 2 are then used to calculate SNP

Version 1.0
4 Construction quality
Poor construction quality results in greater variability in material properties and performance.
HDM-4 does not provide a method of modelling proportions of road that are classified as
good, fair and poor, so only an average level of construction defects is usually included. The
construction defects indicators (CDS and CDB) used in the deterioration models are described
below.
The relative compactions of the base, sub-base and selected subgrade layers (COMP) is
important in predicting the initial densification of rut depth. Paterson (1987) gives an
equation to calculate COMP, but it is proposed that users are also able to estimate it based on
the values in Table C2.12.
Table C2.12 Default values for relative compaction
Compliance Relative Compaction COMP (%)

Full compliance in all layers 100
Full compliance in some layers 95
Reasonable compliance in most layers 90
Poor compliance in most layers 85
The initiation (and in some cases progression) of certain distresses is more accurately
attributed to problems in material handling, preparation, or construction than to structural
weakness in the pavement. In HDM-III, a construction quality code (CQ) was used in the
crack initiation and ravelling models. However, in HDM-4 the construction defects are input
through two indicators:
! CDS
Construction defects indicator for bituminous surfacings

! CDB
Construction defects indicator for the base
CDS is a factor indicating the general level of binder content and stiffness relative to the
optimal material design for the specified bituminous mixture. It is used as an indicator to
illustrate whether a bituminous surfacing is prone to cracking and ravelling (low value of
CDS), or prone to rutting through plastic deformation (high value of CDS).
CDS is a continuous variable, generally ranging in value between 0.5 and 1.5 as shown in
Table C2.13. Intermediate values are chosen by judgement. This may involve back-analysis
to verify that the plastic deformation and cracking predictors are valid (see A Guide to
Calibration and Adaptation).

Version 1.0
Table C2.13 Selection of construction defect indicator for bituminous

surfacings
Surfacing condition CDS

Dry (Brittle) Nominally about 10% below design optimal binder content 0.5
Normal Optimal binder content 1.0
Rich (Soft) Nominally about 10% above design optimal binder content 1.5
For potholing, the base construction defects indicator (CDB) is used. CDB is a continuous
variable ranging between 0 (no construction defects) and 1.5 (several defects). The type of
defects that should be considered in setting a value of CDB is given in Table C2.14.
Table C2.14 Selection of construction defect indicator for base
Construction defect CDB

Poor gradation of material 0.5
Poor aggregate shape 0.5
Poor compaction 0.5

Version 1.0
5 Crack modelling
Cracking is one of the most important distresses in bituminous pavements. Fatigue and ageing
have been identified as the principal factors which contribute to cracking of a bituminous
pavement layer. The propagation of cracking is accelerated through the embrittlement
resulting from ageing and the ingress of water, which can significantly weaken the underlying
pavement layers.
There are two types of cracking considered in HDM-4:
! Structural cracking
This is effectively load and age/environment associated cracking (see Section 5.1).
! Transverse thermal cracking
This is generally caused by large diurnal temperature changes or in freeze/thaw

conditions, and therefore usually occurs only in certain climates (see Section 5.2).
For each type of cracking, separate relationships are given for predicting the time to initiation
and then the rate of progression. These relationships include the construction defects indicator
for bituminous surfacings, CDS, as a variable (see Section 4).
5.1 Structural cracking

Structural cracking is modelled as All and Wide cracking, based on the relationships derived
by Paterson (1987).
5.1.1 Initiation of All structural cracking

Crack initiation is said to occur when 0.5% of the carriageway surface area is cracked.
Initiation of All structural cracking depends on the base:
! Stabilised base
if HSOLD = 0 (that is, original surfacings)
 a 1 HSE + a 2 log e CMOD  

ICA = K cia CDS 2 a 0 exp 
( ) ( ) + CRT  ...(5.1)
 + a 3 log e DEF + a 4 YE4 DEF  
if HSOLD > 0 (that is, overlays or reseals)
 (0.8 KA + 0.2KW ) (1 + 0.1HSE ) + (1 − KA ) (1 − KW )a 0  

 2  + CRT 
ICA = K cia CDS
* exp  1  
a HSE + a 2 log e CMOD

  + a 3 log e DEF + a 4 (YE 4 ) (DEF )  
    
...(5.2)
! Other bases
if HSOLD = 0 (that is, original surfacing)
   YE4  
ICA = K cia CDS 2 a 0 exp a 1 SNP + a 2   + CRT  ...(5.3)
   SNP 2  

Version 1.0
if HSOLD > 0 (that is, overlays or reseals)

For all surface materials except CM, SL and CAPE
     YE4   
   a 0 expa 1 SNP + a 2  2   
      + CRT 
ICA = K cia CDS 2 MAX SNP
 ...(5.4)
   PCRW  
 
   * MAX1 − a ,0 , a 4 HSNEW 

   3  
For surface materials - CM, SL and CAPE
     YE4   
   a 0 exp a 1 SNP + a 2    
 MAX   SNP 2   
ICA = K cia CDS 2   + CRT  ...(5.5)
   PCRA  
  * MAX1 − ,0 , a 4 
    a3   
5.1.2 Initiation of Wide structural cracking

ICW = K ciw MAX[(a 0 + a 1 ICA ), a 2 ICA ] ...(5.6)
where:
ICA time to initiation of All structural cracks (years)

ICW time to initiation of Wide structural cracks (years)
CDS construction defects indicator for bituminous surfacings
YE4 annual number of equivalent standard axles (millions/lane)
SNP average annual adjusted structural number of the pavement
DEF mean Benkelman beam deflection in both wheelpaths (mm)
CMOD resilient modulus of soil cement (GPa) (in the range between 0 and 30 GPa for
most soils)
PCRA area of All cracking before latest reseal or overlay (% of total carriageway
area)
PCRW area of Wide cracking before latest reseal or overlay (% of total carriageway
area)
KW MIN [0.05 MAX (PCRW - 10, 0), 1]
KA MIN [0.05 MAX (PCRA - 10, 0), 1]
HSE MIN [100, HSNEW + (1 - KW) HSOLD]
Kcia calibration factor for initiation of All structural cracking
Kciw calibration factor for initiation of Wide structural cracking
CRT crack retardation time due to maintenance (years) (see Part D)

Version 1.0
The proposed default coefficient values a0 to a4 for the initiation of All cracking are given in
Table C2.15, and those of a0 to a2 for the initiation of Wide cracking are given in Table C2.16.
Table C2.15 Default coefficient values for initiation of All structural cracking
models
Pavement Surface HSOLD Equn a0 a1 a2 a3 a4

type material value
All 0 5.3 4.21 0.14 -17.1
AMGB All except CM >0 5.4 4.21 0.14 -17.1 30 0.025
CM >0 5.5 13.2 0 -20.7 20 1.4
All 0 5.3 4.21 0.14 -17.1

AMAB
>0 5.4 4.21 0.14 -17.1 30 0.025
AMAP All >0 5.4 4.21 0.14 -17.1 30 0.025
All 0 5.1 1.12 0.035 0.371 -0.418 -2.87

AMSB
>0 5.2 1.12 0.035 0.371 -0.418 -2.87
All 0 5.3 13.2 0 -20.7
STGB All except SL, >0 5.4 13.2 0 -20.7 20 0.22

CAPE
SL, CAPE >0 5.5 13.2 0 -20.7 20 1.4
All 0 5.3 13.2 0 -20.7
STAB All except SL, >0 5.4 4.21 0.14 -17.1 20 0.12
CAPE
SL, CAPE >0 5.4 4.21 0.14 -17.1 30 0.025
STAP All >0 5.4 4.21 0.14 -17.1 20 0.12
All 0 5.1 1.12 0.035 0.371 -0.418 -2.87

STSB
>0 5.2 1.12 0.035 0.371 -0.418 -2.87

Version 1.0
Table C2.16 Default coefficient values for initiation of Wide structural cracking
models
Pavement type Surface material HSOLD value a0 a1 a2

All 0 2.46 0.93 0
AMGB All except CM >0 2.04 0.98 0
CM >0 0.70 1.65 0
All 0 2.46 0.93 0

AMAB
>0 2.04 0.98 0
AMAP All >0 2.04 0.98 0
All 0 1.46 0.98 0

AMSB
>0 0 1.78 0
All 0 2.66 0.88 1.16
STGB All except SL, CAPE >0 1.85 1.00 0
SL, CAPE >0 0.70 1.65 0
All 0 2.66 0.88 1.16
STAB All except SL, CAPE >0 1.85 1.00 0
SL, CAPE >0 2.04 0.98 0
STAP All >0 1.85 1.00 0
All 0 1.46 0.98 0

STSB
>0 0 1.78 0
5.1.3 Progression of All structural cracking

The general form of the model for the progression of All structural cracking is given below:
 CRP 
dACA = K cpa 
 CDS 
[
 Z A (Z A a 0 a 1 δt A + SCA )
a1 1/a1 − SCA ] ...(5.7)
Progression of All structural cracking commences when δtA > 0 or ACAa > 0
where:
if ACA a > 0 δtA = 1 otherwise δt A = MAX {0, MIN [(AGE2 − ICA ), 1]}
if ACA a ≥ 50 then: zA = -1 otherwise: zA = 1
ACA a = MAX (ACA a , 0.5 )
SCA = MIN [ACA a , (100 - ACA a )]
Y = [a 0 a1 Z A δt A + SCA a1 ] ...(5.8)

Version 1.0
! if Y < 0
then:
dACA = K cpa 
CRP 
(100 − ACA a ) ...(5.9)
 CDS 
! if Y ≥ 0
then:
dACA = K cpa 
CRP 
 (
Z A Y 1/a1 − SCA ) ...(5.10)
 CDS 
! if ACA a ≤ 50 and ACA a + dACA > 50
then:
dACA = K cpa 
CRP 

 CDS 
(
100 - c 1
1/a1
− ACA a ) ...(5.11)
where:
c 1 = MAX { [2(50 a1 ) − SCA a1 − a 0 a1 δt A ], 0} ...(5.12)
5.1.4 Progression of Wide structural cracking

The general form of the model for the progression of Wide structural cracking is given below:
 CRP 
dACW = K cpw 
 CDS 
[
 Z w (Z w a 0 a 1 δt w + SCW )
a1 1/a1 − SCW ] ...(5.13)
where:
dACW = MIN [ACA a + dACA - ACWa, dACW ] ...(5.14)
Progression of Wide structural cracking commences when δtW > 0 or ACW a > 0
where:
if ACWa > 0 δtW = 1 otherwise δt w = MAX { 0, MIN [(AGE2 − ICW ), 1] }
The initiation of Wide structural cracking is constrained so that it does not commence before
the area of All structural cracking (ACAa) exceeds 5% as follows:
δtW = 0 if ACA a ≤ 5 and ACWa ≤ 0.5 and δtW > 0
If patching of Wide structural cracking was performed in the previous analysis year, reducing
the area of Wide cracking to below 1% but with the area of All structural cracking remaining
at over 11% at the start of the current analysis year (that is, ACWa ≤1 and ACAa > 11), then
the rate of progression of Wide structural cracking is assumed to begin not at the low initial
rate, but at a higher rate similar to the rate before patching.
For this situation a temporary value of Wide structural cracking, ACWtemp is defined to be 5%
less than ACAa; that is:
ACWtemp = ACAa – 5 if ACWa ≤ 1 and ACAa > 11

Version 1.0
This value is then used as the temporary value of ACWa for the computation of dACW in that
analysis year.
dACW is computed each analysis year as follows:
if ACWa ≥ 50 ACWa ≥ 50 then: zw = -1
otherwise: zw = 1
ACWa = MAX (ACWa , 0.5 )
[
SCW = MIN ACWa , (100 - ACWa ) ]
Y = [a 0 a 1 Z w δt w + SCW a1 ] ...(5.15)
! if Y < 0
then:
dACW = K cpw 
CRP 
MIN [(ACA a + dACA − ACWa ), (100 − ACWa )] ...(5.16)
 CDS 
! if Y ≥ 0
then:
 CRP 
MIN [(ACA a + dACA − ACWa ), Z w (Y
dACW = K cpw  1/a1 − SCW )] ...(5.17)
 CDS 
! if ACWa ≤ 50 and ACWa + dACW > 50

then:
 CRP 
dACW = K cpw  [ (
MIN (ACA a + dACA − ACWa ), 100 − c1
1/a1
− ACWa )]
 CDS 
...(5.18)
where:
c 1 = MAX { [2(50 a1 ) − SCW a1 − a 0 a1 δt w ], 0} ...(5.19)
and:
dACA incremental change in area of All structural cracking during the analysis
year (% of total carriageway area)
dACW incremental change in area of Wide structural cracking during the analysis
ACAa area of All structural cracking at the start of the analysis year
ACWa area of Wide structural cracking at the start of the analysis year (% of total
carriageway area)
δtA fraction of analysis year in which All structural cracking progression
applies

Version 1.0
δtW fraction of analysis year in which Wide structural cracking progression

applies
AGE2 pavement surface age since last reseal, overlay, reconstruction or new
construction (years)
ICA time to initiation of All structural cracking (years)
ICW time to initiation of Wide structural cracking (years)
Kcpa calibration factor for progression of All structural cracking
Kcpw calibration factor for progression of Wide structural cracking
CRP retardation of cracking progression due to preventative treatment, given
by CRP = 1 - 0.12 CRT
The proposed default coefficient values a0 and a1 for the progression of All cracking and those
for the progression of Wide cracking are given in Table C2.17.
Table C2.17 Default coefficient values for progression of All and Wide
structural cracking
Pavement Surface HSOLD All cracking Wide cracking

type material value
a0 a1 a0 a1
All 0 1.84 0.45 2.94 0.56
AMGB All except CM >0 1.07 0.28 2.58 0.45
CM >0 2.41 0.34 3.40 0.35
0 1.84 0.45 2.94 0.56

AMAB All
>0 1.07 0.28 2.58 0.45
AMAP All >0 1.07 0.28 2.58 0.45
0 2.13 0.35 3.67 0.38

AMSB All
>0 2.13 0.35 3.67 0.38
0 1.76 0.32 2.50 0.25

STGB All
>0 2.41 0.34 3.40 0.35
All 0 1.76 0.32 2.50 0.25
STAB All except SL, >0 2.41 0.34 3.40 0.35

CAPE
SL, CAPE >0 1.07 0.28 2.58 0.45
STAP All >0 2.41 0.34 3.40 0.35
0 2.13 0.35 3.67 0.38

STSB All
>0 2.41 0.34 3.40 0.35

Version 1.0
5.2 Transverse thermal cracking

Transverse thermal cracking is modelled as cracking intensity expressed as the number of
cracks per kilometre. A coefficient of thermal cracking (CCT) is used as a variable to predict
time to initiation of thermal cracks for the various climate zones described in Chapter C1.
Suggested values of CCT are given in Table C2.18. Table C2.19 gives the proposed values of
the maximum number of thermal cracks (NCTeq) per kilometre of road and the time since
initiation to reach this level of cracking (Teq), for the various climate zones.
Table C2.18 Proposed default values of CCT
Model Tropical Sub- Sub- Temperate Temperate

parameter tropical tropical
Arid 100 5 100 100 2
Semi-arid 100 8 100 100 2
Sub-humid 100 100 100 100 1
Humid 100 100 100 100 1
Per-humid 100 100 100
Table C2.19 Proposed default values of NCTeq and Teq
Model Tropical Sub- Sub- Temperate Temperate

parameter tropical tropical
NCTeq 0 100 0 0 20
Teq 50 7 50 50 7
5.2.1 Initiation of transverse thermal cracking

A distinction is made between the time to initiation of transverse thermal cracking in original
surfacings and in overlays or reseals.
! if HSOLD = 0 (that is, original surfacings)
ICT = K cit MAX[a 0 , (CDS )(CCT )] ...(5.20)
! if HSOLD > 0 (that is, overlays or reseals)
ICT = K cit MAX [a 0 , CDS (CCT + a 1 + a 2 HSNEW )] ...(5.21)
5.2.2 Progression of transverse thermal cracking

Progression of transverse thermal cracking commences when δtT > 0
where:
if ACTa > 0 δtT = 1 otherwise δt T = MAX { 0, MIN [(AGE2 − ICT ), 1 ] }
! if HSOLD = 0 (that is, original surfacings)

Version 1.0
   2 NCTeq (AGE3 - ICT - 0.5 )   

dNCT = K cpt 
1 
 MAX0, MIN(NCTeq − NCTa ),    δt T
 CDS    (Teq )2  
   
...(5.22)
! if HSOLD > 0 (that is, overlays or reseals)
   
   
 MIN(a 0 PNCT, (PNCT - NCTa )), 
 
dNCT = K cpt 
1 
 MIN(NCTeq − NCTa ), MAX  
δt T
 CDS   
  2 NCTeq (AGE3 - ICT - 0.5 )   
  ,0  
  (Teq )2   
    
...(5.23)
A transverse thermal crack is assumed to traverse the full width of the carriageway. Thus the
area of transverse thermal cracking is given by:
dNCT
dACT = ...(5.24)
20
where:
ICT time to initiation of transverse thermal cracks (years)

dNCT incremental change in number of transverse thermal cracks during the analysis
year (no/km)
dACT incremental change in area of transverse thermal cracking during the analysis
CCT coefficient of thermal cracking (see Table C2.18)
PNCT number of transverse thermal cracks before latest overlay or reseal (no/km)
NCTa number of (reflected) transverse thermal cracks at the start of the analysis year
(no/km)
NCTeq maximum number of thermal cracks (no/km) (see Table C2.19)
Teq time since initiation to reach maximum number of thermal cracks (years) (see
Table C2.19)
Kcit calibration factor for initiation of transverse thermal cracking
Kcpt calibration factor for progression of transverse thermal cracking
The default coefficient values a0 to a2 for the initiation, and those of a0 for the progression of
transverse thermal cracks, are given in Table C2.20.

Version 1.0
Table C2.20 Default coefficient values for transverse thermal cracking
Pavement type Initiation Progression
a0 a1 a2 a0
All pavement types except STGB and STSB 1.0 -1.0 0.02 0.25
STGB and STSB 100 -1.0 0.02 0.25
5.3 Total areas of cracking

The above cracking models predict areas of All and Wide structural cracking (ACA and ACW
respectively) and transverse thermal cracking (ACT). In several of the deterioration models,
areas of cracking other than ACA, ACW or ACT are required. These are defined in Sections
5.3.1 and 5.3.2.
5.3.1 Area of indexed cracking

The area of indexed cracking is a weighted average of All and Wide structural cracking,
defined by Paterson (1987) as follows:
ACX = 0.62 ACA + 0.39 ACW ...(5.25)
where:
ACX area of indexed cracking (% of total carriageway area)

ACA area of All structural cracking (% of total carriageway area)
ACW area of Wide structural cracking (% of total carriageway area)
5.3.2 Total area of cracking

The total area of cracking combines the structural and transverse thermal cracking and is
defined as follows:
ACRA = ACA + ACT ...(5.26)
where:
ACRA total area of carriageway cracked (% of total carriageway area)

ACA area of All structural cracking (% of total carriageway area)
ACT area of transverse thermal cracking (% of total carriageway area)

Version 1.0
6 Ravelling
Ravelling is the progressive loss of surface material through weathering and/or traffic
abrasion. The occurrence of ravelling varies considerably among different regions and
countries according to construction methods, specifications, available materials, and local
practice. Ravelling is a common distress in poorly constructed, thin bituminous layers such as
surface treatment, but it is rarely seen in high quality, hot-mix asphalt.
The construction defects indicator for bituminous surfacings, CDS, (see Section 4) is used as a
variable in the ravelling models. The initiation model is basically as proposed by Paterson
(1987), with CDS replacing the original construction quality variable CQ. The progression
model is also based on that proposed by Paterson (1987) but with a traffic variable introduced
as proposed by Riley (1999).
6.1 Initiation
Ravelling is said to occur on a given road section when 0.5% of the carriageway surface area
is classified as ravelled. The initiation is given as:
IRV = K vi CDS 2 a 0 RRF exp (a1 YAX ) ...(6.1)
where:
IRV time to ravelling initiation (years)

YAX annual number of axles of all motorised vehicle types in the analysis year
(millions/lane)
Kvi calibration factor for ravelling initiation
RRF ravelling retardation factor due to maintenance (see Part D)
The proposed default coefficient values a0 to a1 for the ravelling initiation model is given in
Table C2.21.
Table C2.21 Default coefficient values for ravelling initiation model
Surface type Surface material a0 a1

All except CM 100 -0.156
AM
CM 8.0 -0.156
All except SL, CAPE 10.5 -0.156

ST
SL, CAPE 14.1 -0.156
6.2 Progression
The general form of the model for the progression of ravelling is given below:

Version 1.0
 K vp   1 
dARV =   [
 Z (Z( a 0 + a 1 YAX)a 2 δt v + SRV )
a2 1/a2 − SRV ] ...(6.2)
 RRF   CDS 
2
Progression of ravelling commences when δtv > 0 or ARVa > 0

where:
if ARVa > 0 δtv = 1 otherwise δt v = MAX { 0, MIN [(AGE2 − IRV ), 1] }
if ARVa ≥ 50 then: z = -1
otherwise: z = 1
ARVa = MAX (ARVa , 0.5 )
SRV = MIN [ARVa , (100 - ARVa )]
YAX = MAX [MIN (YAX, 1), 0.1]
[
Y = (a 0 + a 1 YAX)a 2 Z δt v + SRV a2 ] ...(6.3)
! if Y < 0
then:
 K vp   1 
dARV =   2 
(100 − ARVa ) ...(6.4)
 RRF   CDS 
! if Y ≥ 0
then:
 K vp   1 
 Z(Y
dARV =  1/a2 − SRV )
 ...(6.5)
 RRF   CDS 2 
! if ARVa ≤ 50 and ARVa + dARV > 50
then:
 K vp   1 
dARV =   
 RRF   CDS 2 
(
100 − c 1
1/a2
− ARVa ) ...(6.6)
and:
c 1 = MAX { [2(50 a2 ) − SRV a2 − (a 0 + a1 YAX)a 2 δt v ], 0} ...(6.7)
where:
dARV change in area of ravelling during the analysis year (% of total carriageway
area)
ARVa area of ravelling at the start of the analysis year (% of total carriageway area)
δtv fraction of analysis year in which ravelling progression applies

Version 1.0
AGE2 pavement surface age since last reseal, overlay, reconstruction or new
Kvp calibration factor for ravelling progression
IRV time to ravelling initiation (years)
Other parameters are as defined for ravelling initiation.

The proposed default coefficient values a0 and a2 for the ravelling progression model is given
in Table C2.22.
Table C2.22 Default coefficient values for ravelling progression model
Pavement type a0 a1 a2
All pavement types 0.6 3.0 0.352

Version 1.0
7 Potholing
Potholes usually develop in a surface that is either cracked, ravelled, or both. The presence of
water accelerates pothole formation both through a general weakening of the pavement
structure and lowering the resistance of the surface and base materials to disintegration.
The potholing models use the construction defects indicator for the base, CDB, as a variable
(see Section 4). In the models, potholing is expressed in terms of the number of pothole units
of area 0.1 m2. The volume of each of these pothole units is assumed to be 10 litres (that is,
100 mm in depth). The relationships for the initiation and progression of potholing have been
modified from those given in the NDLI (1995) and Riley (1996b).
7.1 Initiation
Initiation of potholes due to cracking only arises once the total area of Wide structural
cracking (ACW) exceeds 20%. Ravelling initiated potholes arise when the ravelled area
(ARV) exceeds 30%.
 (1 + a 1HS ) 
IPT = K pi * a 0   ...(7.1)
 (1 + a 2 CDB ) (1 + a 3 YAX ) (1 + a 4 MMP ) 
where:
IPT time between the initiation of Wide structural cracking or ravelling and the
initiation of potholes (years)
HS total thickness of bituminous surfacing (mm)
CDB construction defects indicator for the base
YAX annual number of axles of all motorised vehicle types in the analysis year
(millions/lane)
Kpi calibration factor for pothole initiation
The values for IPT are calculated separately for potholing due to cracking and due to
ravelling. The separation between these two mechanisms of potholing is maintained
throughout the analysis with the progression being modelled differently for potholes due to
cracking, due to ravelling and due to the enlargement of existing potholes.
The proposed default coefficient values a0 to a4 for the pothole initiation model is given in
Table C2.23.

Version 1.0
Table C2.23 Default coefficient values for pothole initiation model
Cause of Pavement a0 a1 a2 a3 a4
pothole initiation type
AMGB, STGB 2.0 0.05 1.0 0.5 0.01
Cracking
All except GB bases 3.0 0.05 1.0 0.5 0.01
AMGB, STGB 2.0 0.05 1.0 0.5 0.01

Ravelling
All except GB bases 3.0 0.05 1.0 0.5 0.01
7.2 Progression
Pothole progression arises from potholes due to cracking, ravelling and the enlargement of
existing potholes. The progression of potholes is affected by the time lapse between the
occurrence and patching of potholes. A time lapse factor (TLF) has been introduced as an
indicator of the response time to patching potholes (Odoki, 1997; and Riley, 1997).
The annual incremental increase in the number of pothole units due to each of these three
distresses is calculated as:
 (1 + a 1CDB )(1 + a 2 YAX )(1 + a 3 MMP)

dNPTi = K pp * a 0 * ADIS i (TLF)  ...(7.2)
 (1 + a 4HS ) 
Pothole progression from wide cracking or ravelling commences as follows:

! If at the start of the first year of the analysis period ACWa = 0, then potholing progression
from wide cracking commences when:
AGE2 > ICW + IPT and ACW a > 20
! If at the start of the first year of the analysis period ARVa = 0, then potholing progression
from ravelling commences when:
AGE2 > IRV + IPT and ARVa > 30
! If at the start of the first year of the analysis period 0 < ACWa ≤ 20, then potholing
progression from wide cracking commences when ACWa > 20
! If at the start of the first year of the analysis period 0 < ARVa ≤ 30, then potholing
progression from ravelling commences when ARVa > 30
! If at the start of the first year of the analysis period ACWa > 20, then potholing
progression from wide cracking commences immediately
! If at the start of the first year of the analysis period ARVa > 30, then potholing
progression from ravelling commences immediately
! Potholing progression from enlargement commences if NPTa > 0 at the start of an
analysis year
The total annual increase in the number of pothole units per kilometre of road length is given
by:

Version 1.0
dNPT =
∑ dNPT
i =1
i ...(7.3)
where:
dNPTi additional number of potholes per km derived from distress type i (Wide
structural cracking, ravelling, enlargement) during the analysis year
ADISi the percentage area of Wide structural cracking at the start of the analysis year,
or the percentage area of ravelling at the start of the analysis year, or number
of existing potholes per km at the start of the analysis year
TLF time lapse factor (see Table C2.25)
dNPT total number of additional potholes per km during the analysis year
Kpp calibration factor for pothole progression
Other parameters are as defined previously.

The proposed default coefficient values a0 to a4 for the pothole progression model are given in
Table C2.24 and the proposed default values for TLF are given in Table C2.25.
Table C2.24 Default coefficient values for pothole progression model
Cause of pothole Pavement type a0 a1 a2 a3 a4

progression
AMGB, STGB 1.0 1.0 10 0.005 0.08
Cracking
All except GB bases 0.5 1.0 10 0.005 0.08
AMGB, STGB 0.2 1.0 10 0.005 0.08

Ravelling
AMGB, STGB 0.07 1.0 10 0.005 0.08

Enlargement
Table C2.25 Default values for TLF
Time lapse between the TLF

occurrence of potholes and
patching
Less than 2 weeks 0.02
1 month 0.06
2 months 0.12
3 months 0.20
4 months 0.28
6 months 0.43
12 months 1.00

Version 1.0
8 Edge-break
Edge-break can be defined as the loss of surface and base materials at the pavement edge,
caused by shear failure and attrition. This commonly arises on narrow roads with unsealed
shoulders, where vehicle wheels pass on or close to a pavement edge.
The measure for edge-break that the user provides as input into the model, and the
corresponding output data, will be in square metres per km, and not in cubic metres per km.
The value in square metres is then multiplied internally by 100 mm (same depth as potholes)
to obtain the volume of edge-break in cubic metres for modelling purposes.
The edge-break model predicts that edge-break occurs on roads with a carriageway width of
up to a user defined maximum width of CWmax. The default value of CWmax is 7.2 metres and
an upper limit of CWmax has been set to 7.5 metres (that is, no edge-break is predicted for
roads with a carriageway width greater than 7.5 metres).
The edge-break model is as follows:
dVEB = K eb a 0 PSH (AADT ) ESTEP (S ) a 2 +

2 a1 MMP  − 6
10 ...(8.1)
 1000 
and:
   CWmax - CW   
PSH = MAXMINMAX a 3 + a 4 CW, , 1, 0 ...(8.2)
   a5   
where:
dVEB annual loss of edge material (m3/km)

PSH proportion of time vehicles use the shoulder due to road width
AADT annual average daily traffic (veh/day)
ESTEP elevation difference from pavement to shoulder (mm)
S average traffic speed (km/h)
CW carriageway width (metres)
CWmax user definable maximum carriageway width for the occurrence of edge-break
(metres) (default = 7.2, maximum = 7.5)
Keb calibration factor for edge-break progression
The proposed default coefficient values a0 to a5 for the edge-break model are given in Table
C2.26.

Version 1.0
Table C2.26 Default coefficient values for edge-break model
Pavement type a0 a1 a2 a3 a4 a5
AMGB 50 -1 0.2 2.65 -0.425 10
AMAB, AMSB, AMAP 25 -1 0.2 2.65 -0.425 10
STGB 75 -1 0.2 2.65 -0.425 10
STAB, STSB, STAP 50 -1 0.2 2.65 -0.425 10

Version 1.0
9 Damaged and undamaged surface area

In modelling pavement deterioration, it is important to ensure that the sum of damaged and
undamaged surface area must be equal to 100%, in any given analysis year. The total road
surface consists of the following:
! Edge-break
! Potholes
! Cracking
! Ravelling
! Undamaged
This area consists of the original road surface which is still in good condition since the
last surfacing and the area which has been patched.
The logic devised for calculating the distress values at the end of an analysis year is described
below (Odoki, 1998).
9.1 The logic

For modelling purposes, the above types of deterioration need to be converted to the
equivalent surface area and these are assumed to be mutually exclusive. Therefore the sum of
the surface area with edge-break, potholes, cracking, ravelling and undamaged must equal
100%.
It is acknowledged that an area of road can be both cracked and ravelled. However, the
hierarchy employed in HDM-4 classifies cracking above ravelling because cracking is
considered to be a more severe distress than ravelling. Once substantial amounts of damaged
area are being modelled, the area of ravelling will therefore be re-classified as area of
cracking. This will result in the reported area of ravelling decreasing, although this re-
classified area could be regarded as both cracked and ravelled.
In devising a logic that satisfies the constraint of 100% total surface area, the following
simplifying assumptions are made:
! Cracking develops first from the undamaged area and then, after the latter is exhausted,
from the ravelled area if any. Furthermore, an area once cracked can develop potholes but
cannot ravel.
! Ravelling can only develop from the undamaged area. After an area is ravelled it can also
crack, at which stage it is reclassified from ravelled to cracked. (Note: this does not mean
that ravelled areas would physically disappear).
! Potholes can only develop from cracked, ravelled and undamaged areas (as reflected in
the formulas for computing the change in the number of potholes), and unless it is
repaired, an area potholed cannot revert to cracking, ravelling or undamaged.
! An upper limit of 10% is imposed on the potholed area. This is because above this level
the pavement surface becomes ill defined and the roughness function becomes invalid.
! Edge-break can only develop from cracked, ravelled and undamaged areas, and unless it is
repaired, an area of edge-break cannot revert to potholes, cracking, ravelling or
undamaged.

Version 1.0
! An upper limit of 18% is imposed on the area of edge-break. The upper value of 18% is
based on the assumption that edge-break will not extend beyond 0.5 metres from either
edge of a pavement of 5.5 metres width.
9.2 Distress values at the end of the year

The assumptions given in Section 9.1 lead to Equations 9.1below - 9.20 below (see Sections
9.2.1 - 9.2.4) for computing the damaged areas at the end of each analysis year and before
road works.
9.2.1 Edge-break
AVEB b = MIN [ 18, (AVEB a + dAVEB ) ] ...(9.1)
where:
AVEBb area of edge-break at the end of the analysis year (% of total carriageway area)
AVEBa area of edge-break at the start of the analysis year (% of total carriageway
area)
dAVEB unadjusted increase in the area of edge-break during the analysis year (% of
total carriageway area)
Equation 9.1 above requires that the volume of edge-break, VEB, be converted into an area of
edge-break measured as a percentage of total carriageway area. The area of edge-break
expressed as a percentage of the total carriageway area, AVEB, is obtained from the following
expression:
VEB
AVEB = ...(9.2)
CW
where:
AVEB area of edge-break (% of total carriageway area)

VEB volume of edge-break per km (m3/km)
CW carriageway width (metres)
Thus, by substituting VEB in Equation 9.2 above with VEBa the value of AVEBa is obtained,
and by substituting VEB with dVEB, the value of dAVEB is obtained;
where:
VEBa volume of edge-break per km at the start of the analysis year (m3/km)
dVEB unadjusted increase in volume of edge-break per km during the analysis year

Version 1.0
(m3/km)
9.2.2 Potholes
APOTb = MIN [10, (APOTa + dAPOT )] ...(9.3)
where:
APOTb area of potholes at the end of the analysis year (% of total carriageway area)
APOTa area of potholes at the start of the analysis year (% of total carriageway area)
dAPOT unadjusted increase in the area of potholes during the analysis year (% of total
carriageway area)
Equation 9.3 above requires that the number of potholes per km be converted into area of
potholes in per cent of total carriageway area. The area of potholes expressed as a percentage
of total carriageway area (APOT) is obtained from the expression:
(NPT ) (STDAPOT )
APOT = ...(9.4)
10(CW )
where:
APOT area of potholes (% of total carriageway area)

NPT number of potholes per km
STDAPOT standard area of one pothole (m2 ) (default = 0.1)
Thus, by substituting NPT in Equation 9.4 above with NPTa the value of APOTa is obtained,
and by substituting NPT with dNPT the value of dAPOT is obtained;
where:
NPTa number of potholes per km at the start of the analysis year (no/km)
dNPT unadjusted increase in the number of potholes per km during the analysis year
9.2.3 Cracking
Total area of cracking
ACRA b = MAX { 0, MIN [(100 - AVPD b ), (ACRA a + dACRA − dAVEBCR − dAPOTCR )] }

...(9.5)
where:
ACRAb total area of cracking at the end of the analysis year (% of total carriageway
area); that is, ACAb + ACTb

Version 1.0
ACRAa total area of cracking at the start of the analysis year (% of total carriageway
area); that is, ACAa + ACTa
dACRA unadjusted increase in total area of cracking during the analysis year (% of
total carriageway area); that is, dACA + dACT
AVPDb AVEBb + APOTb
dAVEBCR increase in area of edge-break arising from cracked area during the analysis
dAPOTCR increase in area of potholes arising from cracked area during the analysis year
(% of total carriageway area)
ACAb area of All structural cracking at the end of the analysis year (% of total
carriageway area)
ACAa area of All structural cracking at the beginning of the analysis year (% of total
carriageway area)
dACA unadjusted increase in area of All structural cracking during the analysis year
ACTb area of transverse thermal cracking at the end of the analysis year (% of total
carriageway area)
ACTa area of transverse thermal cracking at the start of the analysis year (% of total
carriageway area)
dACT unadjusted increase in area of transverse thermal cracking during the analysis
The value of dAVEBCR is obtained as follows:

! if ACRAa > 0
then:
dAVEBCR = 0.01 (VBCR ) (∆AVEB ) ...(9.6)
otherwise:
dAVEBCR = 0
and:
∆AVEB = AVEB b - AVEB a
where:
∆AVEB adjusted increase in the area of edge-break during the analysis year (% of
VBCR user definable percentage of dAVEB arising from cracked areas (default =
20)
The value of dAPOTCR is obtained as follows:

Version 1.0
! if dNPT > 0
∆NPTc (STDAPOT )
dAPOTCR = ...(9.7)
10 (CW )
otherwise:
dAPOTCR = 0
 dNPTc 
∆NPTc =   ∆NPT ...(9.8)
 dNPT 
and:
∆NPT = NPTb − NPTa
where:
∆NPT adjusted total increase in the number of potholes per km during the
analysis year
∆NPTc adjusted increase in the number of potholes per km derived from Wide
structural cracking during the analysis year
dNPTc unadjusted increase in the number of potholes per km derived from Wide
structural cracking during the analysis year (see Section 7)
NPTb total number of potholes per km at the end of the analysis year
NPTa total number of potholes per km at the start of the analysis year
All structural cracking
ACA b = MIN [(ACA a + ∆ACA ), ACRA b ] ...(9.9)
! if ACRAa > 0
 [dACA - q(dAVEBCR ) − dAPOTCR ] 

∆ACA = MAX 0, ∆ACRA  ...(9.10)
 (dACRA − dAVEBCR − dAPOTCR ) 
! if ACRAa = 0 and ACRAb > 0
(dACA - dAPOTCR )
∆ACA = ∆ACRA ...(9.11)
(dACRA − dAPOTCR )
otherwise:
∆ACA = 0
and:

Version 1.0
 ACTa 
q = 1 -  ...(9.12)
 ACRA a 
and:
∆ACRA = ACRA b - ACRA a ...(9.13)
where:
∆ACRA adjusted increase in total area of cracking during the analysis year (% of
∆ACA adjusted increase in area of All structural cracking during the analysis
Wide structural cracking
ACWb = MAX { 0, MIN [ACWa + dACW - dAPOTCR - q (dAVEBCR ), ACA b ] }
...(9.14)
where:
ACWb area of Wide structural cracking at the end of the analysis year (% of total
carriageway area)
ACWa area of Wide structural cracking at the start of the analysis year (% of total
carriageway area)
dACW unadjusted increase in area of Wide structural cracking during the analysis
Transverse thermal cracking

! if ACTa + dACT > 0
then:
ACTb = ACRA b - ACA b ...(9.15)
otherwise:
ACTb = 0
All the parameters are as defined previously.

Version 1.0
9.2.4 Ravelling
 (100 - AVPC b ), 
ARVb = MAX 0, MIN  
 (ARVa + dARV − dAVEBRV − dAPOTRV − dACRARV ) 
...(9.16)
where:
ARVb area of ravelling at the end of the analysis year (% of total carriageway area)
ARVa area of ravelling at the start of the analysis year (% of total carriageway area)
dARV unadjusted increase in area of ravelling during the analysis year (% of total
carriageway area)
AVPCb AVEBb + APOTb + ACRAb
dAVEBRV increase in area of edge-break arising from ravelled area during the analysis
dAPOTRV increase in area of potholes arising from ravelled area during the analysis
dACRARV increase in area of cracking arising from ravelled area during the analysis

The value of dAVEBRV is obtained as follows:
! if ARVa > 0
then:
dAVEBRV = 0.01 (VBRV ) (∆AVEB ) ...(9.17)
otherwise:
dAVEBRV = 0
where:
VBRV user definable percentage of dAVEB arising from ravelled area (default =
20)
The value of dAPOTRV is obtained as follows:

! if dNPT > 0
then:
∆NPTr (STDAPOT )
dAPOTRV = ...(9.18)
10(CW )
otherwise:

Version 1.0
dAPOTRV = 0
and:
 dNPTr 
∆NPTr =   ∆NPT ...(9.19)
 dNPT
where:
∆NPTr adjusted increase in the number of potholes per km derived from ravelling
during the analysis year
dNPTr unadjusted increase in the number of potholes per km derived from
ravelling during the analysis year (see Section 7)

The value of dACRARV is obtained as follows:
! if ARVa > 0
then:
dACRARV = 0.01 (CRV ) (∆ACRA ) ...(9.20)
otherwise:
dACRARV = 0
where:
CRV user definable percentage of dACRA arising from the ravelled area
(default = 10)
9.3 Total damaged surface area

The total non-patched damaged surface area at any time is calculated from the following
expression:
ADAMR b = AVEB b + APOTb + ACRA b + ARVb ...(9.21)
where:
ADAMRb total non-patched damaged surface area at the end of the analysis year (%
of total carriageway area)

Severely damaged surface area that can be patched is given by the expression:
ADAMS b = APOTb + ACWb + ARVb ...(9.22)
where:

Version 1.0
ADAMSb severely damaged surface area at the end of the analysis year (% of total
carriageway area)

Version 1.0
10 Rut depth
Rutting is defined as the permanent or unrecoverable traffic-associated deformation within
pavement layers which, if channelised into wheelpaths, accumulates over time and becomes
manifested as a rut (Paterson, 1987).
Rut depth modelling is performed after the values of all the surface distresses (that is,
cracking, ravelling, potholing and edge-break) at the end of the year have been calculated.
The rut depth model is based on four components of rutting:
! Initial densification (see Section 10.1)
! Structural deformation (see Section 10.2)
! Plastic deformation (see Section 10.3)
! Wear from studded tyres (see Section 10.4)
The rut depth at any time is the sum of the four components.
For HDM-4 the rut depths have been standardised to a 2.0 m straight-edge. Since HDM-III
was based on a 1.2 m straight-edge, the default model coefficients have been changed
accordingly.
10.1 Initial densification

The initial densification depends upon the degree of relative compaction of the base, sub-base
and selected subgrade layers; that is, COMP. Suggested values of COMP have been given in
Section 4.
The initial densification is:

(
RDO = K rid a 0 YE4 10 6 )(a + a
1 2 DEF )
SNP a3 COMP a 4 

...(10.1)
where:
RDO rutting due to initial densification (mm)

DEF average annual Benkelman beam deflection (mm)
COMP relative compaction (%) (see Section 4)
Krid calibration factor for initial densification
The proposed default coefficient values a0 to a4 for the initial densification model is given in
Table C2.27.

Version 1.0
Table C2.27 Default coefficient values for initial densification model
Pavement type a0 a1 a2 a3 a4
AMGB, AMAB, AMSB, STGB, STAB, STSB 51740 0.09 0.0384 -0.502 -2.30
AMAP, STAP 0 0 0 0 0
Initial densification only applies to new construction or reconstruction that involves the
construction of a new base layer (that is, from when AGE4 = 0), for a period of time of one
year. AGE4 is defined as follows:
AGE4 age since reconstruction (including base) or new construction (years)
10.2 Structural deformation

The structural deformation model used in HDM-III has been simplified into a linear form for
inclusion in HDM-4 (Morosiuk, 1998c). Separate terms are proposed for structural
deformation without cracking and structural deformation after cracking as follows:
! Structural deformation without cracking
(
∆RDSTuc = K rst a 0 SNP a1 YE4 a 2 COMP a3 ) ...(10.2)
! Structural deformation after cracking
[
∆RDSTcrk = K rst a 0 SNP 1 YE4 2 MMP 3 ACX a
a a a a4
] ...(10.3)
The total annual incremental increase in structural deformation is as follows:

! if ACRA = 0
then:
∆RDST = ∆RDSTuc ...(10.4)
! if ACRA > 0
then:
∆RDST = ∆RDSTuc + ∆RDSTcrk ...(10.5)
where:
∆RDST total incremental increase in structural deformation in the analysis year

(mm)
∆RDSTuc incremental rutting due to structural deformation without cracking in the
analysis year (mm)
∆RDSTcrk incremental rutting due to structural deformation after cracking in the
analysis year (mm)

Version 1.0
ACXa area of indexed cracking at the beginning of the analysis year (% of total
carriageway area)
Krst calibration factor for structural deformation
The proposed default coefficient values a0 to a4 for the structural deformation models are
given in Table C2.28.
Table C2.28 Default coefficient values for structural deformation model
Without cracking All pavement types 44950 -1.14 0.11 -2.3
After cracking All pavement types 0.0000248 -0.84 0.14 1.07 1.11
10.3 Plastic deformation

The plastic deformation model includes a variable, CDS, which indicates whether the
surfacing is prone to plastic deformation.
A more accurate method of determining the plastic deformation of a bituminous surfacing is
detailed in A Guide to Calibration and Adaptation. The method includes the use of variables
to predict changes in material properties, such as the softening point of the binder and voids in
the mix, to model the incremental increase in plastic deformation.
The general plastic deformation model (that is, without material properties) is given by:
∆RDPD = K rpd CDS 3 a 0 YE4 Sh a1 HS a 2 ...(10.6)
where:
∆RDPD incremental increase in plastic deformation in the analysis year (mm)

Sh speed of heavy vehicles (km/h)
HS total thickness of bituminous surfacing (mm)
Krpd calibration factor for plastic deformation
The proposed default coefficient values for the plastic deformation model are given in Table
C2.29.

Version 1.0
Table C2.29 Default coefficient values for plastic deformation model
Surface type a0 a1 a2
AM 2.46 -0.78 0.71
ST 0 -0.78 0.71
10.4 Surface wear

The surface wear model (Djarf, 1995) is applied to environments where vehicles use studded
tyres during the freezing period.
RDW = K rsw [a 0 PASS a1 W a 2 S a 3 SALT a 4 ] ...(10.7)
where:
∆RDW incremental increase in rut depth due to studded tyres in the analysis year
(mm)
PASS annual number of vehicle passes with studded tyres in one direction (1000s)
SALT variable for salted or unsalted roads (2 = salted; 1 = unsalted)
W road width (m) (carriageway plus total shoulder width)
Krsw calibration factor for surface wear
The proposed default coefficient values a0 to a4 for the surface wear model is given in Table
C2.30.
Table C2.30 Default coefficient values for surface wear model
All pavement types 0.0000248 1.0 -0.46 1.22 0.32
10.5 Total rut depth

The annual incremental increase in total rut depth, ∆RDM, is derived as follows:
! if AGE4 ≤ 1
∆RDM = RDO + ∆RDPD + ∆RDW ...(10.8)
otherwise:
∆RDM = ∆RDST + ∆RDPD + ∆RDW ...(10.9)
where:

Version 1.0
∆RDM incremental increase in total mean rut depth in both wheelpaths in the
analysis year (mm)
RDO rutting due to initial densification in the analysis year (mm)
∆RDST incremental increase in structural deformation in the analysis year (mm)
∆RDPD incremental increase in plastic deformation in the analysis year (mm)
∆RDW incremental increase in wear by studded tyres in the analysis year (mm)
The total rut depth, RDMb, at any given time is given as:
RDMb = MIN [(RDM a + ∆RDM),100 ] ...(10.10)
where:
RDMb total mean rut depth in both wheelpaths at the end of the analysis year (mm)
RDMa total mean rut depth in both wheelpaths at the start of the analysis year (mm)
10.6 Standard deviation of rut depth

The standard deviation of rut depth is used in the roughness model. It is calculated from the
mean total rut depth as:
RDS b = MAX[0.3, (0.9 - 0.04 RDMb )] RDMb ...(10.11)
where:
RDSb rut depth standard deviation at the end of the analysis year (mm)
RDMb mean rut depth at the end of the analysis year (mm)
The standard deviation of rut depth at the beginning of the analysis year (RDSa) is also
calculated from Equation 10.11 above with RDMa replacing RDMb.

Version 1.0
11 Roughness
The roughness model consists of several components of roughness (cracking, disintegration,
deformation and maintenance). The total incremental roughness is the sum of these
components. The surface distress values used in predicting roughness are those that have
been adjusted so that the total damaged surface area plus the undamaged area equals 100%.
11.1 Structural
The structural component of roughness relates to the deformation in the pavement materials
under the shear stresses imposed by traffic loading. It is given by:
−5
∆RI s = a 0 exp (m K gm AGE3 ) (1 + SNPK b ) YE4 ...(11.1)
and:
SNPK b = MAX [(SNPa - dSNPK ), 1.5 ] ...(11.2)
and:
MIN (a 1, ACX a ) HSNEW + 

dSNPK = K snpk a 0   ...(11.3)
MAX [MIN (ACX a − PACX, a 2 ), 0] HSOLD 
where:
∆RIs incremental change in roughness due to structural deterioration during the

analysis year (IRI m/km)
dSNPK reduction in adjusted structural number of pavement due to cracking
SNPKb adjusted structural number of pavement due to cracking at the end of the
analysis year
SNPa adjusted structural number of pavement at the start of the analysis year
ACXa area of indexed cracking at the start of the analysis year (% of total
carriageway area)
PACX area of previous indexed cracking in the old surfacing (% of total carriageway
area); that is, 0.62 (PCRA) + 0.39 (PCRW)
AGE3 pavement age since last overlay (rehabilitation), reconstruction or new
m environmental coefficient (see Table C2.31)
Kgm calibration factor for environmental coefficient
Ksnpk calibration factor for SNPK

Version 1.0
The default values for the environmental coefficient m are given in Table C2.31.
Table C2.31 Roughness environmental coefficient ‘m’ by climate zones
Moisture Temperature classification

classification
Tropical Sub- Sub- Temperate Temperate
tropical tropical

Arid 0.005 0.010 0.015 0.025 0.040
Semi-arid 0.010 0.015 0.025 0.035 0.060
Sub-humid 0.020 0.025 0.040 0.060 0.100
Humid 0.025 0.030 0.060 0.100 0.200
Per-humid 0.030 0.040 0.070
11.2 Cracking
The incremental change in roughness due to cracking is given by:
∆RI c = a0 ∆ACRA ...(11.4)
where:
∆RIc incremental change in roughness due to cracking during the analysis year (IRI
m/km)
∆ACRA incremental change in area of total cracking during the analysis year (% of
11.3 Rutting
The incremental change in roughness due to variation of rut depth is given by:
∆RIr = a 0 ∆RDS ...(11.5)
where:
∆RIr incremental change in roughness due to rutting during the analysis year (m/km
IRI)
∆RDS incremental change in standard deviation of rut depth during the analysis year
(mm) (= RDSb – RDSa)

Version 1.0
11.4 Potholing
The potholing effect depends upon the number of vehicles that actually hit the potholes, which
in turn depends upon the traffic volume and the freedom to manoeuvre. A freedom to
manoeuvre variable (FM), ranging between 0 and 1, is used and is predicted using Equation
11.6 below:
FM = (MAX { MIN [0.25 (CW - 3 ) , 1 ] , 0} ) MAX 1 −

AADT   
, 0  ...(11.6)
  5000   
The change in roughness is calculated as follows:

! if nil patching (TLF = 1) or 100% patching policy options is specified, then
 ∆NPT * TLF 
a2 
∆RI t = a 0 (a 1 − FM) NPTa * TLF +  − NPTa 
a2
...(11.7)
 2  
 
! otherwise (for partial patching policy options)
a2
 ∆NPT 
∆RI t = a 0 (a 1 − FM) * ∆NPT *  NPTa +  ...(11.8)
 2 
where:
FM freedom to manoeuvre
CW carriageway width (m)
∆RIt incremental change in roughness due to potholing during the analysis year (IRI
m/km)
∆NPT incremental change in number of potholes per km during the analysis year
NPTa number of potholes per km at the start of the analysis year
TLF time lapse factor (see Table C2.25)
11.5 Environment
The environmental component of roughness is due to factors which include temperature and
moisture fluctuations, and also foundation movements (for example, subsidence). It is given
by:
∆RIe = m * K gm RIa ...(11.9)
where:
∆RIe incremental change in roughness due to the environment during the analysis
year (IRI m/km)
RIa roughness at the start of the analysis year (IRI m/km)

Version 1.0
m environmental coefficient
Kgm calibration factor for the environmental component
11.6 Total change in roughness

The total change in the roughness of the pavement is derived as:
∆RI = K gp [∆RIs + ∆RIc + ∆RIr + ∆RI t ] + ∆RI e ...(11.10)
where:
∆RI total incremental change in roughness during the analysis year (IRI m/km)
Kgp calibration factor for roughness progression
The default coefficient values for the various roughness components are given in Table C2.32.
The roughness of the pavement at the end of the analysis year is given by:
RIb = MIN[(RIa + ∆RI), a 0 ] ...(11.11)
where:
RIb roughness of the pavement at the end of the analysis year (IRI m/km)
RIa roughness of the pavement at the start of the analysis year (IRI m/km)
a0 upper limit of pavement roughness, specified by the user (default = 16 IRI
m/km)
Table C2.32 Default coefficient values for roughness components
Pavement type Roughness Equation a0 a1 a2

component
Structural 11.1 134
dSNPK 11.3 0.0000758 63.0 40.0

All pavement types
Cracking 11.4 0.0066
Rutting 11.5 0.088
Potholing 11.7 0.00019 2.0 1.5
The annual average roughness of the pavement for a given analysis year is calculated as:
RI av = 0.5 (RI a + RIb ) ...(11.12)
where:

Version 1.0
RIav annual average roughness of the pavement for the analysis year (IRI m/km).
This is the roughness used in the Road User Effects model.

Version 1.0
12 Pavement surface texture

Pavement texture is perhaps the most important variable which determines the magnitude of
longitudinal and lateral forces at the tyre-road interface. A road surface exhibits two types of
texture classified as macrotexture and microtexture. In general, microtexture determines the
maximum skid resistance afforded by a dry pavement, while macrotexture determines the
drainage ability and therefore how effective the microtexture will be when the pavement is
wet. Most skidding related accidents occur on wet pavements. The changes in macrotexture
due to wear and compaction from traffic action have important safety and economic
consequences since rolling resistance is a function of texture.
12.1 Texture depth

This refers to the macrotexture of pavement. Cenek and Griffith-Jones (1997) proposed an
incremental macro-texture model that can be expressed as:
   (ITD − TD a )(a ITD ) 

∆TD = K td ITD − TD a − a 0 ITD log10 10 0
+ ∆NELV  ...(12.1)
  
 
where:
∆TD incremental change in sand patch derived texture depth during analysis year
(mm)
ITD initial texture depth at construction of surfacing (mm)
TDa texture depth at the beginning of the analysis year (mm)
∆NELV number of equivalent light vehicle passes during the analysis year (one heavy
truck or heavy bus is equal to 10 NELV; light vehicles equal 1)
Ktd calibration factor for texture depth
The proposed default coefficient values for a0 for the texture depth model are given in Table
C2.33. This table also includes values for the initial texture depth (ITD) which is used as
default when resetting pavement surface type. These can be replaced by user definable
values.

Version 1.0
Table C2.33 Default parameter values for texture depth model
Surface type Surface Parameter

material
ITD a0
AC 0.7 0.005
HRA 0.7 0.005
PMA 0.7 0.005

AM RAC 0.7 0.005
CM 0.7 0.005
SMA 0.7 0.005
PA 1.5 0.008
SBSD 2.5 0.120
DBSD 2.5 0.120

ST
CAPE 0.7 0.006
SL 0.7 0.006
PM 1.5 0.008
The texture depth at the end of the analysis year is given by the following relationship:
TD b = MAX[(TD a + ∆TD ), 0.1] ...(12.2)
where:
TDb texture depth at the end of the analysis year (mm)

TDa texture depth at the start of the analysis year (mm)
∆TD incremental change in texture depth during the analysis year (mm)
The annual average texture depth for a given analysis year is calculated as follows:
TD av = 0.5 (TD a + TD b ) ...(12.3)
where:
TDav annual average texture depth for the analysis year (mm). This is the texture
depth used in the Road User Effects model.
12.2 Skid resistance

This is strongly influenced by the microtexture, which is a measure of the degree of polishing
of a pavement surface or of the aggregate and the surface. The proposed skid resistance model
is as follows:

Version 1.0
∆SFC 50 = K sfc a 0 MAX[0, ∆QCV ] ...(12.4)
where:
∆SFC50 incremental change in sideway force coefficient during the analysis year,
measured at 50 km/h
∆QCV annual incremental increase in the flow of commercial vehicles (veh/lane/day)
Ksfc calibration factor for skid resistance
The proposed default coefficient values for a0 for the skid resistance model are given in Table
C2.34.
Table C2.34 Default coefficient values for skid resistance model
Surface type Surface Coefficient

material
a0
AC -0.663 x 10-4
HRA -0.663 x 10-4
PMA -0.663 x 10-4

AM RAC -0.663 x 10-4
CM -0.663 x 10-4
SMA -0.663 x 10-4
PA -0.663 x 10-4
SBSD -0.663 x 10-4
DBSD -0.663 x 10-4

ST
CAPE -0.663 x 10-4
SL -0.663 x 10-4
PM -0.663 x 10-4
The skid resistance measured at 50 km/h at the end of the analysis year is given by the
following expression:
SFC 50b = MAX [(SFC 50a + ∆SFC 50 ), 0.35 ] ...(12.5)
where:
SFC50b sideway force coefficient, measured at 50 km/h, at the end of the analysis year
SFC50a sideway force coefficient, measured at 50 km/h, at the start of the analysis year
∆SFC50 incremental change in sideway force coefficient, measured at 50 km/h, during
the analysis year
The annual skid resistance value for a given analysis year is calculated as follows:

Version 1.0
SFC 50av = 0.5 (SFC 50a + SFC 50b ) ...(12.6)
where:
SFC50av annual average sideway force coefficient, measured at 50 km/h, for the
analysis year

The average skid resistance value at a given annual average traffic speed is calculated as
follows:
 SFC 50av {400 − [2 − MIN(TD av , 2)] [MAX(50, S ) − 50]}

SFCs = K sfcs   ...(12.7)
 400 
where:
SFCs sideway force coefficient at an average traffic speed of S km/h

TDav annual average texture depth for the analysis year (mm)
Ksfcs calibration factor for skid resistance speed effects
The user needs to define a value of SFC50 in order for skid resistance modelling to be
performed. This will also need to be supplied after maintenance treatments.

Version 1.0
13 Calibration factors
The deterioration models contain calibration factors to facilitate local calibration. These
factors have default values of 1.0 and are summarised in Table C2.35.
Table C2.35 Calibration factors used in the deterioration models
Deterioration model Calibration factor

Wet/dry season SNP ratio Kf
Drainage deterioration factor Kddf
Drain life factor Kdrain
All structural cracking - initiation Kcia
Wide structural cracking - initiation Kciw
All structural cracking - progression Kcpa
Wide structural cracking - progression Kcpw
Transverse thermal cracking - initiation Kcit
Transverse thermal cracking - progression Kcpt
Rutting - initial densification Krid
Rutting - structural deterioration Krst
Rutting - plastic deformation Krpd
Rutting - surface wear Krsw
Ravelling - initiation Kvi
Ravelling - progression Kvp
Pothole - initiation Kpi
Pothole - progression Kpp
Edge-break Keb
Roughness - environmental coefficient Kgm
Roughness - SNPK Ksnpk
Roughness - progression Kgp
Texture depth - progression Ktd
Skid resistance Ksfc
Skid resistance - speed effects Ksfcs

Version 1.0
14 References
Cenek P.D., and Griffith-Jones G., (1997)
Incremental Forms for Modelling Texture Depth and Friction Parameters in HDM-4
Communication to the ISOHDM
Works Central Laboratories, New Zealand
Djarf L., (1995)
Road deterioration and maintenance effects models in cold climates. Swedish Road
Research Institute, Sweden
Morosiuk G., (1996, 1998a & 1999)
Specifications for the HDM-4 Road Deterioration Model – Fourth, Fifth, Sixth, Seventh
and Eighth Drafts, ISOHDM, UK
Morosiuk G., (1998b)
Notes on the coefficient values for estimating drain life
Morosiuk G., (1998c)
Derivation of a new rut depth model for the structural deformation component in HDM-4
TRL Unpublished Project Report PR/ORC/610/98
Transport Research Laboratory, Crowthorne, UK
NDLI, (1995)
Modelling Road Deterioration and Maintenance Effects in HDM-4. Final Report Asian
Development Bank Project RETA 5549
N.D. Lea International, Vancouver
Odoki J.B., (1997)
Notes on the effects of patching on the progression of potholing. Communication to the
ISOHDM
Odoki J.B., (1998)
Notes on the logic for total damaged and undamaged carriageway surface areas
ISOHDM Working Paper, University of Birmingham, UK
Parkman C.C., and Rolt J., (1997)
Characterisation of pavement strength in HDM-III and possible changes for HDM-4
TRL Unpublished Project Report PR/ORC/587/97
Transport Research Laboratory, Crowthorne, UK
Paterson W.D.O., (1987)
Road Deterioration and Maintenance Effects
World Bank Publications
Washington, D.C., USA
Notes on modelling the deterioration of drainage

Version 1.0
PIARC Commision on Terminology, (1997)

Technical Dictionary of Road Terms, 7th Edition, English - French
PIARC World Road Congress, Paris, France
Riley M.J., (1996a & 1999)
Notes on Seasonal Adjustment of Pavement Strength
Riley M.J., (1996b)
Modifications to Chapter 7 - Potholing, Delamination and Edge-break - of the NDLI
report
Riley M.J., (1997)
Notes on the effects of patching on the progression of potholing. Communication to the
ISOHDM
Riley M.J., (1998)
Proposed amendments to the potholing component of roughness. Communication to the
ISOHDM
Riley M.J., and Bennett C.R., (1995 & 1996)
Specifications for the HDM-4 Road Deterioration Model - Preliminary, Second and
Third Drafts, ISOHDM, UK
Sayers M., Gillespie T.D., and Paterson W.D.O., (1986)
Guidelines for the Conduct and Calibration of Road Roughness Measurements
Technical Report No 46
Watanatada T., Harral C.G., Paterson W.D.O., Dhareshwar A..M., Bhandari A., and
HDM-III Model. World Bank Publications, Washington, D.C.

Version 1.0
C3 Concrete Pavements
1 Introduction
The prediction models for concrete pavement deterioration included in HDM-4 are based on
research carried out by the Latin American Study Team, in Chile (1996). The research
involved a comprehensive bibliographical review of the existing concrete pavement
deterioration models. The models implemented in HDM-4 are mainly based on previous work
carried out by SHRP (1993), Al-Omari and Darter (1994), Lee and Darter (1994) and ERES
Consultants (1995).
The process of model selection considered the following aspects:
! Validity range of the model
Modelled in terms of types of climate, traffic range, and pavement structure.

! Statistical parameters
Number of observations, correlation coefficients, estimated errors, etc.

! Sensitivity analysis
! Year of model development
This chapter describes the Road Deterioration (RD) models for concrete pavements included
in HDM-4. It commences with an overview of the modelling framework, followed by a brief
analysis of the concrete pavement types and distress modes considered, and finally a complete
description of the models.
It should be noted that the RD models for concrete pavements are basically absolute models
(as opposed to incremental models used for bituminous pavements). The models have been
developed in imperial units. However, for consistency within HDM-4 user-interface, the data
required for modelling is initially input in metric. This data is then converted into imperial
units for use in the relationships, and then the results of calculations are converted back into
metric for reporting purposes.

Version 1.0
PART C ROAD DETERIORATION MODELS C3 CONCRETE PAVEMENTS
2 Modelling framework and logic

The framework used for concrete pavement modelling conforms to the general HDM-4
pavement classification system described in Chapter C1. This is a versatile framework that is
able to handle a wide range of pavement types. The formal structure of concrete pavement
classification is shown in Table C3.1.
Table C3.1 Structures of concrete pavements
Surface Base Pavement Description

type type type
JP GB JPGB Jointed Plain Concrete over Granular Base
JP AB JPAB Jointed Plain Concrete over Asphalt Base
JP AP JPAP Jointed Plain Concrete over Asphalt Pavement
JP SB JPSB Jointed Plain Concrete over Stabilised Base
JP RB JPRB Jointed Plain Concrete over Rigid/concrete Base
JR GB JRGB Jointed Reinforced Concrete over Granular Base
JR AB JRAB Jointed Reinforced Concrete over Asphalt Base
JR AP JRAP Jointed Reinforced Concrete over Asphalt Pavement
JR SB JRSB Jointed Reinforced Concrete over Stabilised Base
JR RB JRRB Jointed Reinforced Concrete over Rigid/concrete Base
CR GB CRGB Continuously Reinforced Concrete over Granular Base
CR AB CRAB Continuously Reinforced Concrete over Asphalt Base
CR AP CRAP Continuously Reinforced Concrete over Asphalt Pavement
CR SB CRSB Continuously Reinforced Concrete over Stabilised Base
CR RB CRRB Continuously Reinforced Concrete over Rigid/concrete Base
There are different sets of deterioration models for concrete pavements included in HDM-4
that are based on pavement surface type and construction type (see Table C3.2). Calibration
parameters have also been provided to account for variations in surface material and to
facilitate local adaptation of the models.
Table C3.2 Surface types upon which the concrete RD models are based
Surface type Description

JP Jointed Plain concrete pavement - without load transfer dowels
JP Jointed Plain concrete pavement - with load transfer dowels
JR Jointed Reinforced concrete pavement
CR Continuously Reinforced concrete pavement
The modelling of concrete pavement performance is considered in two separate phases:

Version 1.0
! Phase 1
Refers to the time before any major periodic maintenance or reconstruction.

! Phase 2
Refers to the time after the pavement has received a major maintenance or has been
reconstructed.
Phase 1 models are described in this chapter; Phase 2 models are discussed in Chapter D3.
2.1 Concrete pavement structure

In rigid pavement roads, the thickness of concrete slab often ranges from 15 cm for light
traffic to 30 cm for heavy traffic. Thicker slabs in excess of 28 cm may be constructed without
a base course. A brief description of the concrete pavement structures considered in HDM-4 is
given below.
2.1.1 Jointed plain concrete pavements without load transfer dowels

This type of JP concrete pavement (JPCP n/d) is built using short slabs without reinforcement
steel (see Figure C3.1). Spacing between transverse joints (or slab length) is such that the
strains induced by changes in temperature and/or moisture content do not produce
intermediate cracking between the joints. The maximum spacing between joints is limited to
minimise slab movement and maximise load transfer. Typical values of slab length vary
between 3.0 and 6.0 metres for this type of pavement. Transverse load transfer from one slab
to the next one is accomplished through aggregate interlock.
Joint spacing
3 - 6 m
Slab Aggregate
Interlock
Base
Figure C3.1 Jointed plain concrete pavements without dowels
2.1.2 Jointed plain concrete pavements with load transfer dowels

This type of JP concrete pavement is similar to (JPCPn/d) described above, except that dowel
bars are added in the transverse joints to help load transfer (see Figure C3.2).

Version 1.0
Joint spacing
3 - 6 m
Dowels
Figure C3.2 Jointed plain concrete pavements with dowels
2.1.3 Jointed reinforced concrete pavements

This type of concrete pavement is designed with a quantity of longitudinal reinforcement
steel, which permits longer slab lengths, between 10 and 20 m (see Figure C3.3).
Reinforcement steel control transverse cracking that could occur due to movements of the
foundation subgrade, and/or strains produced by temperature or humidity changes. Load
transfer in transverse joints is accomplished through load transfer dowels.
Joint spacing
10 - 20 m
Slab
Dowels
Base
Welded wire fabric (0.1 – 0.2%)
Figure C3.3 Jointed reinforced concrete pavements
2.1.4 Continuously reinforced concrete pavements

This type of concrete pavement has longitudinal reinforcement throughout its length;
therefore, it has no transverse joints. The objective of the longitudinal reinforcement steel is to
control the cracks that are produced in the pavement due to shrinkage in the concrete (see
Figure C3.4).

Version 1.0
Cracks separation
Slab
Base
Reinforcement steel
0,6 - 0,8 % of area
Figure C3.4 Continuously reinforced concrete pavements
2.2 Concrete pavement distress modes

There are six concrete pavement distress modes modelled in HDM-4. These distresses
together with the pavement surface type to which they apply are presented in Table C3.3.
Table C3.3 Distress modes modelled in HDM-4
No. Distress mode Units of measurement Pavement surface type

1 Cracking Percent of slabs cracked JP
Number per mile JR
2 Faulting inches JP and JR
3 Spalling Percent of spalled joints JP and JR
4 Failures Number per mile CR
5 Serviceability loss Dimensionless JR and CR
6 Roughness Inches per mile (or m/km) JP, JR and CR
Each of these distress modes is described in the following sections (2.2.1 - 2.2.6).
2.2.1 Cracking
There are three types of cracking distress that are commonly identified on concrete
pavements:
1 Transverse cracking
2 Longitudinal cracking
3 Durability cracks
Transverse cracks are predominantly perpendicular to the central axis of the road (see Figure
C3.5). They manifest three severity levels, according to SHRP (1993):
! Low
Cracks with a width of less than 3 mm, without visible spalling or faulting; or well
sealed, with a non-determinable width.

Version 1.0
! Medium
Cracks with a width between 3 and 6 mm, or with spalling less than 75 mm, or faulting
less than 6 mm.
! High
Cracks with a width greater than 6 mm, or spalling greater than 75 mm, or faulting
greater than 6 mm.
Distress Distress
width width
A B C D
Longitudinal Joint
C
L
C D
Transv. Transv.
Joint Joint
A B Traffic
Slab
Shoulder
Figure C3.5 Transverse cracking
Transverse cracking may have significant impact on the riding quality and are therefore
modelled in HDM-4.
Longitudinal cracks are predominantly parallel to the axis of the road. Durability cracks are
fine cracks, slightly spaced, and often occur adjacent to joints, cracks, or free edges. They
begin in the slab corner as a group of obscure cracks just as in the surrounding area.
2.2.2 Faulting of transverse joints and cracks

This distress refers to a joint or crack having a difference in elevation between both sides of
the joint or crack (see Figure C3.6). Faulting causes significant increases in road roughness.
Faulting is measured as the average fall of all transverse joints within the pavement section
under consideration.

Version 1.0
A B
Longitudinal Joint
C
L
Transv.
Transv. Joint
Joint
A B
Traffic
Slab
Shoulder
Figure C3.6 Faulting of transverse joints and cracks
2.2.3 Spalling of transverse joints

These are breaks or cracks of the joint edge, occurring within a maximum distance of 0.6 m
from the transverse joint (see Figure C3.7). Spalling occurs on Jointed Plain and Jointed
Reinforced concrete pavements. They manifest three severity levels, according to SHRP
(1993):
! Low
Spalling of less than 75 mm of distress width, measured from centre of the joint, with or
without loss of material.
! Medium
Spalling of between 75 and 150 mm of distress width, measured from centre of the joint,
with loss of material.
! High
Spalling of greater than 150 mm of distress width, measured from the centre of the joint,
with loss of material.

Version 1.0
Distress < 0,6 m

width
A B C D
Crack Joint
Joint
Transv. Transv. Transv.
Joint Joint Joint
Low Sev.:
1,8 m Low Sev.: C D
2m Moder. Sev.:
2,5 m
High Sev.:
1,5 m Traffic
A B
Shoulder
Figure C3.7 Spalling of transverse joints
2.2.4 Failures
This distress is a major defect that occurs in Continuously Reinforced (CR) concrete
pavements. Located failures include loosening and breaking of reinforcement steel and
transverse crack spalling. Failures are measured in number per mile (or km).
Many of the maintenance activities in CR concrete pavements are directly related to failures.
Often, it is necessary to estimate the quantity of these distresses for the purpose of preventive
designs and rehabilitation planning.
2.2.5 Serviceability loss

Present Serviceability Rating (PSR) is a subjective user rating of the existing ride quality of
pavement condition. The ratings based on key distress types (for example, transverse
distortions, cracking, spalling, faulting and surface deterioration) range from 0 for extremely
poor condition to 5 for extremely good condition, as shown in Table C3.4. This concept for
rating pavement surface condition was developed by engineers at the AASHO Road Test
(1962), and it has since been correlated with various roughness indicators, such as slope
variance and International Roughness Index (IRI).
Table C3.4 Scale of PSR values used
PSR Condition
0-1 Very poor
1-2 Poor
2-3 Fair
3-4 Good
4-5 Very good

Version 1.0
2.2.6 Roughness
This is a measure of the deviations of a surface from a true planar surface with characteristic
dimensions that affect vehicle dynamics, ride quality, dynamic loads and drainage (ASTM E-
867-82A) - typically in the ranges of 0.1 to 100 m wavelength and 1 to 100 mm amplitude.
International Roughness Index (IRI), is the reference measure expressing roughness as a
dimensionless average rectified slope statistic of the longitudinal profile and defined in Sayers
et al. (1986).
2.2.7 Other defects

There are several other defects on concrete pavements that include the following:
! Scaling
! Polishing of aggregate
! Popouts
! Blow-up
! Punchouts
! Water bleeding and pumping
! Patch-deterioration
! Deterioration of transverse joints
2.3 Primary modelling parameters

The primary variables used for modelling the performance of concrete pavements may be
considered under pavement structural characteristics, condition, history, traffic, road geometry
and the environment. The road characteristics at the beginning of the analysis year is
initialised either from input data if it is the first year of the analysis or the first year after
construction, or otherwise from the result of the previous year’s maintenance and
improvement works.
Pavement structural characteristics

These include measures of pavement strength, slab thickness, material types and properties,
amount of reinforcement steel, the presence of tied concrete shoulders and widened outside
lanes, and subgrade stiffness. These parameters are described in detail in Section 3.
Road condition
The pavement and side-drain condition data at the beginning of the first analysis year or the
first year after construction are required as inputs. The data for pavement condition is as
described above in Section 2.2.
The average pavement condition indicators in a given analysis year (that is, before road
works) are predicted using absolute models. Absolute models predict condition (or distress) at
a particular point in time as a function of the independent variables, and can be represented as
follows:
(CONDITION) t = f [ (TIME ), (TRAFF ), (STRENGTH), (ENVIRON), ETC.] ...(2.1)
where:

Version 1.0
(CONDITION)t condition at time t

(TIME) time since the construction year of pavement
(TRAFF) cumulative traffic loading since the construction year of pavement
(STRENGTH) pavement strength parameters
(ENVIRON) environment/climate related parameters
Pavement history
The required data items refer to pavement age and the year of previous major maintenance
and construction works carried out on the pavement.
Road geometry and environment

The key road geometry data required are carriageway and shoulder widths. Several
environment-related parameters are used for concrete pavement deterioration modelling.
These include the mean annual precipitation, freezing index, Thornthwaite moisture index,
temperature range, and number of days with temperature greater than 32°C (90°F). These
parameters are described in detail in Chapter C1.
Traffic
The required traffic data are the annual flow of equivalent standard axle loads (ESAL) and the
cumulative equivalent standard axle loads (NE4), both expressed in millions per lane. These
data are calculated for each analysis year based on the user-specified traffic and vehicle
characteristics.
2.4 Computational procedure

The overall computational logic for modelling the deterioration of each road section in each
analysis year can be summarised by the following steps:
1 Initialise input data and the conditions at the beginning of the year
2 Convert input data from metric to imperial units
3 Compute pavement strength parameters
4 Calculate the amount of each distress mode in the analysis year, in the following order
depending on the pavement surface type:
(a) Cracking
(b) Faulting
(c) Spalling
(d) Failures
5 Calculate present serviceability rating (PSR) if pavement type is JR or CR
6 Calculate average roughness value for the analysis year
7 Store results in imperial units for use in Works Effects (WE) module and in the following
analysis year

Version 1.0
8 Convert the required outputs into metric for use in the RUE and SEE modules and for
reporting

Version 1.0
3 Structural characteristics
3.1 Introduction
This section describes the principal pavement structural data that are necessary to predict the
deterioration of concrete pavements. These include the following:
! Properties of the materials (see Section 3.2)
! Drainage conditions (see Section 3.3)
! Percentage of reinforcement steel (see Section 3.4)
! Load transfer efficiency (see Section 3.5)
! Widened outside lanes (see Section 3.6)
3.2 Properties of the materials

! Modulus of elasticity of concrete (Ec)
The Modulus of elasticity of concrete denoted by Ec can be obtained from an analysis of

deflection measures or from a laboratory testing (for example, according to the procedure
described in ASTM C469). It can also be estimated from a correlation with the
compressive strength of concrete expressed by Equation 3.1 below (Pauw, 1960):
0.5
E c = 57000( f' c ) ...(3.1)
where:
Ec elasticity Modulus of concrete (psi)
fc' compressive strength of concrete, in psi, as determined using

procedures AASHTO (T22-92), AASHTO (T140-92) or ASTM C39
The value of the Modulus of elasticity of concrete used in the pavement deterioration
models is 5,000,000 psi
! Modulus of Rupture of concrete (MR28)
Stresses in concrete pavements are mainly caused by the effects of traffic and
environmental action. The Modulus of Rupture is a measure of the concrete flexural
strength in providing a sustained resistance to the stresses. During the useful life of the
pavement the stress levels may exceed the Modulus of Rupture at certain points, causing
fatigue damage and cracking in the slabs.
The Modulus of Rupture measured after 28 days and denoted by MR28 can be
determined using AASHTO T97 and ASTM C78 procedures, or estimated from the
compressive strength of concrete, as follows:
MR28 = RUP * f c'

0.5
( ) ...(3.2)
where:

Version 1.0
MR28 modulus of Rupture of concrete after 28 days (psi)
fc' compressive strength of concrete, in psi, as determined using procedures

AASHTO (T22-92), AASHTO (T140-92) or ASTM C39
RUP model parameter (varies between 8 and 10, default = 9)
The Modulus of Rupture can also be estimated using the modulus of elasticity of
concrete, which can be obtained from Falling Weight Deflectometer (FWD) test results
or from laboratory testing. The empirical equation (Foxworthy, 1985) for estimating the
Modulus of Rupture is as follows:
E 
MR = 43.5 *  c  + 488.5 ...(3.3)
 10 6 
where:
MR modulus of Rupture of concrete (psi)

Ec modulus of Elasticity of concrete (psi)
The deterioration models for concrete pavements consider the modulus of rupture (MR)
in the long term. The long-term value is estimated by increasing the modulus of rupture
at 28 days (MR28) by 11%.
! Thermal coefficient of concrete (α)
The thermal coefficient of expansion is used to determine the warping (or curling)
stresses produced in a concrete pavement when it is subjected to a temperature difference
between the top and the bottom of the slab. The stresses are greatest at the edges of the
slab, and may result in slab cracking usually near its mid point.
The thermal coefficient of expansion varies with such factors as water to cement ratio,
concrete age, richness of the mix, relative humidity, and the type of aggregate in the mix.
Table C3.5 shows typical values of the Thermal coefficient of concrete according to the
type of aggregate. A value of 5.5 x 10-6 per ºF is commonly used in concrete pavement
analysis.
Table C3.5 Typical values of the thermal coefficient of concrete
Type of aggregate Thermal coefficient of concrete (α)
(10-6 per ºF)

Quartz 6.6
Sandstone 6.5
Gravel 6.0
Granite 5.3
Basalt 4.8
Limestone 3.8
Source: AASHTO (1993)

Version 1.0
! Drying shrinkage coefficient of concrete (γ)
Concrete pavement slabs are subjected to daily variations in temperature and humidity
with associated expansion or contraction effects. A slab with unrestricted movement in
the horizontal direction would not develop stresses under the effects of expansion and
contraction. However, in real site conditions there exists some resistance between the
slab and the underlying base.
Shrinkage in concrete pavements is caused by the loss of water in the drying process. The
drying shrinkage coefficient (γ) is used in the evaluation of the opening and closing of the
joints caused by variations in the mean temperature to which the slabs are subjected.
The shrinkage and strength of concrete are strongly dependent upon the water to cement
ratio. High values of water to cement ratio will reduce the strength of the concrete and
increase the drying shrinkage potential. Shrinkage can therefore be considered to be
inversely proportional to the indirect tensile strength of the concrete. Table C3.6 can be
used to estimate the drying shrinkage coefficient of concrete.
Table C3.6 Approximate relationship between shrinkage coefficient and indirect

tensile strength of Portland cement concrete
Indirect tensile strength (psi) Shrinkage coefficient (in/in)

300 (or less) 0.0008
400 0.0006
500 0.00045
600 0.0003
700 (or greater) 0.0002
! Poisson’s ratio for concrete (µ)
For most cement treated materials, the value of µ normally varies between 0.10 and 0.25,
with 0.15 generally accepted as a representative value.
! Modulus of elasticity of dowel bars (Es)
Dowel bars may be used to transfer (or distribute) load across discontinuities such as
transverse joints. The value of the Modulus of elasticity of load transfer dowels assigned
within the deterioration model is 2.9*107 psi (or 2.0*105 MPa).
! Modulus of elasticity of bases (Ebase)
The stiffness of base influences the overall behaviour of concrete pavements, mainly as a
result of the support provided to the slabs. Drainage effects also have a significant
influence on the behaviour of the base, as discussed below. A more rigid base will
generally provide better support to the slab, and this should reduce the occurrence of
faulting at transverse joints. However, a very rigid base may increase the warping effect
induced by changes in temperature or humidity, and transverse cracking will increase.
Table C3.7 gives typical values of Modulus of elasticity for different base types.

Version 1.0
Table C3.7 Elasticity Modulus by base type
Base type Elasticity Modulus, Ebase
(in psi)
Granular (GB) 25,000
Asphalt treated (AB) 600,000
Cement treated (SB) 400,000
Lean concrete (RB) 1,000,000
The effects of a stabilised base is considered in the cracking model, see Section 4.
! Modulus of subgrade reaction (KSTAT)
The modulus of reaction of a material is an elastic constant that defines the stiffness of
the material or resistance to deformation, under certain loading conditions. The Modulus
of subgrade reaction (KSTAT) is defined by Equation 3.4 below:
RPRESS
KSTAT = ...(3.4)
DEF
where:
KSTAT modulus of subgrade reaction (pci)

RPRESS reactive pressure (psi)
DEF deflection of the plate (inches)
The value of KSTAT can be determined through the plate load test, where the deflection
is the displacement of a circular plate of 30 inches diameter subjected to a static pressure.
KSTAT is expressed in pounds per cubic inch (pci). The assumption associated with the
determination of the value of KSTAT is that the plate is in complete contact with the
subgrade soil and that the soil is elastic.
3.3 Drainage conditions

It is widely recognised that drainage is a major factor that influences the performance of many
concrete pavements. Water which infiltrates through cracks and joints in a concrete slab often
results in the loss of uniform subgrade support and in pavement faulting due to redistribution
of base/sub-base material.
The effect of drainage on concrete pavement performance is incorporated in the HDM-4
deterioration models through the use of a drainage coefficient (Cd). The drainage coefficient
is defined by the quality of drainage and the percentage of time during the year the pavement
structure would normally be exposed to moisture levels approaching saturation (AASHTO,
1986). The quality of drainage is based on the speed at which water is removed from the
pavement structure, and is determined by:
The time that a base needs for draining 50% of the free saturation water.

Version 1.0
This is equivalent to the saturation time (T50) given in Table C3.8, and the associated values
for the drainage coefficient Cd are given in Table C3.9 (AASHTO, 1986).
Table C3.8 Relationship between drainage time and the quality of drainage
Drainage quality Free water removed within, (T50)

Excellent 2 hours
Good 1 day
Fair 7 week
Bad 1 month
Very bad (water will not drain)
Table C3.9 Recommended values of the drainage coefficient (Cd) for concrete
pavements
Drainage Percentage of time that the pavement structure is exposed to

quality humidity levels near to saturation
Less than 1% 1 - 5% 5 - 25% Greater than 25%
Excellent 1.25 - 1.20 1.20 - 1.15 1.15 - 1.10 1.10
Good 1.20 - 1.15 1.15 - 1.10 1.10 - 1.00 1.00
Regular 1.15 - 1.10 1.10 - 1.00 1.00 - 0.90 0.90
Bad 1.10 - 1.00 1.00 - 0.90 0.90 - 0.80 0.80
Very Bad 1.00 - 0.90 0.90 - 0.80 0.80 - 0.70 0.70
Table C3.10 shows a simplified matrix that can be used to estimate the drainage coefficient
(FWHA, 1995).
Table C3.10 Modified AASHTO simplified matrix for drainage coefficient (Cd)
Lateral Precipitation Fine subgrade Coarse subgrade

drains levels
Impermeable Permeable Impermeable Permeable
base base base base
No Wet (humid) 0.85 - 0.95 0.70 - 0.90 0.75 - 0.95 0.90 - 1.00
Dry (arid) 0.95 - 1.05 0.90 - 1.10 0.90 - 1.15 1.00 - 1.10
Yes Wet (humid) 1.00 - 1.10 0.75 - 0.95 0.90 - 1.10 1.05 - 1.15
Dry (arid) 1.10 - 1.20 0.95 - 1.15 1.10 - 1.20 1.15 - 1.20
Source: FHWA (1995)

Version 1.0
Notes:
1 (a) Coarse Subgrade Groups A-1 to A-3

(b) Fine Subgrade Groups A-4 to A-7 , according to AASHTO Soil
Classification System
2 Permeable Base k = 1000 ft/day (305 m/day) or Cu ≤ 6
3 (a) Wet (humid) Precipitation >25 in/year (635 mm/year)
Climate
(b) Dry (arid) Climate Precipitation ≤ 25 in/year (635 mm/year)
3.4 Percentage of reinforcement steel

The purpose of distributed steel reinforcement in reinforced concrete pavement is to resist
cracking due to induced tensile stresses and to reduce the opening of any cracks that may
form, thus maintaining the pavement as an integral unit. The amount of reinforcement
required is expressed as a percentage of concrete cross-sectional area, denoted as PSTEEL.
The requirement of steel reinforcement in concrete pavement varies between JR and CR type
construction.
3.5 Load transfer efficiency
3.5.1 Efficiency of load transfer in the transverse joints

The effective transfer of traffic loads from one slab to another reduces tensile stress levels in
the slabs and the associated deformations of the slabs at the joints. This situation helps to
decrease deterioration by reducing pumping, loss of support and breaking of slab edges. Load
transfer through transverse joints can be effected through dowel bars, aggregate interlock or a
combination of both mechanisms.
Load transfer in the joints can be evaluated with equipment such as the FWD, by registering
the deformations from both the loaded and unloaded sides of the joint. The percentage of load
transferred across a joint, denoted by LT, is expressed as follows:
 DEFunld 
LT =   * 100
 ...(3.5)
 DEFload 
where:
LT percentage of load transferred across a joint

DEFunld deflection in the unloaded side of the joint (inches)
DEFload deflection in the loaded side of the joint (inches)
The efficiency of load transfer is used in the calculation of the maximum bearing stress of the
dowel-concrete system. Theoretically, if a dowel is 100% efficient it is capable of assigning
half of the applied load to each adjacent slab. However, a reduction in load transfer efficiency
would occur over the pavement life, either due to the loss of bond in the zone where the load
transfer device is imbedded in the concrete slab or due to the deterioration of the aggregate

Version 1.0
interlock mechanism. Generally, the reduction in load transfer efficiency increases as traffic
loads increase since aggregate load transfer decreases with load repetitions. The reduction in
the load transfer efficiency can be assumed to be around 5% to 10%, therefore the value of LT
used in the deterioration model is 45%.
3.5.2 Efficiency of load transfer between slab and shoulder

Tied concrete shoulders contribute substantially to improve the overall performance of the
pavement, by providing a reduction in slab stress and an increase in the service life. These
effects are considered in the cracking model through the efficiency of load transfer (LTEsh)
between the slab and shoulder defined in terms of stress. The variable LTEsh is given by the
expression:
 STRESS unld 
LTE sh =   * 100
 ...(3.6)
 STRESS load 
where:
LTEsh efficiency of load transfer between slab and shoulder (%)

STRESSunld stress in the unloaded side of the joint (psi)
STRESSload stress in the loaded side of the joint (psi)
If tied concrete shoulders are provided in the original pavement construction, a value for
LTEsh = 20% should be used. If the shoulders are provided on an existing pavement the value
of LTEsh should be taken as 10%.
3.6 Widened outside lanes

This refers to an original construction that incorporates a wider lane (or standard lane with
hard strip) adjacent to the shoulders. The main benefit associated with the provision of a wider
outer lane is stress reduction at the outer edge of the slab since wheel loads are kept at a
distance from the pavement edge.
The effects of widened outside lanes on concrete pavement performance are considered in
both the cracking and faulting models.

Version 1.0
4 Cracking
The HDM-4 cracking model considers transverse cracking in concrete pavements due to high
stress levels in the slabs or defects originating from material fatigue. The stresses are caused
generally by the combined effect of thermal curling, moisture-induced curling and traffic
loading.
Separate relationships are given for predicting the amounts of transverse cracking over the life
cycle of Jointed Plain concrete pavements and Jointed Reinforced concrete pavements. The
models are deterministic and predict the expected average deterioration based on the input
variables.
4.1 Jointed plain concrete pavements

Transverse cracking is modelled as a function of cumulative fatigue damage in the slabs
(ERES Consultants, 1995).
The percentage of slabs cracked is given by:
100
PCRACK = Kjp c * ...(4.1)
1 + 1.41 * FD -1.66
where:
PCRACK percent of slabs cracked

FD cumulative fatigue damage, dimensionless
Kjpc calibration factor (default = 1.0)
4.1.1 Fatigue damage determination

The cumulative fatigue damage is calculated in terms of Miner’s damage analysis, by
summing the damage index over each slab thermal condition or temperature gradient and axle
load distribution as follows:
G
∑N
n tg
FD = ...(4.2)
tg
tg=1
where:
FD cumulative fatigue damage

tg temperature gradient (tg = 1, …., G)
ntg number of 18 kip equivalent single axle load passes during temperature
gradient tg (ESALs per lane)
Ntg maximum number of 18 kip equivalent standard axle load repetitions during
temperature gradient tg before flexural failure occurs (ESALs per lane)

Version 1.0
According to Miner’s theory, transverse cracking is expected to occur when the cumulative
fatigue damage (FD) approaches 1.0.
! Temperature gradients
The variations in concrete pavement temperature over the year can be represented by a
distribution of temperature gradients. The average temperature gradient is defined as the
difference between the temperature at the top and at the bottom of the slab divided by the
slab thickness. A positive gradient indicates the top of the slab is warmer than the
bottom, which normally occurs during the daytime. Negative gradient condition typically
occurs during the cooler hours of the evening. In all types of climate the positive
temperature gradients occur with greater frequency than the negative temperature
gradients.
Owing to the difficulty that may be experienced in obtaining field data on temperature
gradient distribution, a default data set based on climate zones is provided in HDM-4 as
illustrated in Table C3.11.
Table C3.11 Temperature gradient distribution
Temperature Frequency
difference
(FREQ)
∆T) in °F
(∆ Dry with Dry without Wet with Wet without
freezing freezing freezing freezing
-8 0.086660 0.073237 0.090494 0.086209
-6 0.092003 0.067994 0.094611 0.072691
-4 0.076447 0.057834 0.081522 0.052129
-2 0.058163 0.039585 0.067007 0.039496
0 0.057014 0.031803 0.052426 0.033466
2 0.034749 0.029573 0.036817 0.030790
4 0.036162 0.024472 0.039393 0.031347
6 0.037122 0.019472 0.033196 0.021113
8 0.031273 0.021223 0.033254 0.024858
10 0.036200 0.028565 0.032462 0.032160
12 0.021978 0.027069 0.026291 0.025427
14 0.037272 0.029359 0.034706 0.038571
16 0.026134 0.036464 0.029423 0.037274
18 0.032394 0.030194 0.034758 0.038976
20 0.033724 0.037439 0.032034 0.038803
22 0.023131 0.032684 0.017874 0.037385
24 0.009683 0.036172 0.006422 0.027180
26 0.000047 0.024021 0.000078 0.011631
28 0.000000 0.013717 0.000000 0.001188

Version 1.0
Note: The frequencies do not add up to 1.0 because the data relative to temperature
differences of negative 8 ºF do not provide significant (meaningful) information to the
concrete models.
There are factors other than temperature that cause curling (which may be concave) in
slabs. A correction to the difference in temperature measured in the slab is applied
according to climate zones as follows, Eisenmann and Leykauf (1990):
a1 * (SLABTHK - 2 ) 
∆Ts = ∆T - a0 + ...(4.3)
 SLABTHK 3 
where:
∆Ts adjusted difference in temperature at the top and bottom of the slab (ºF)
∆T difference between the temperature measured at the top and bottom of the
slab (ºF )
(= Ttop - Tbottom)
SLABTHK slab thickness (inches)
a0 and a1 model coefficients based on climate zones
The model coefficient values are given in Table C3.12.
Table C3.12 Model coefficient for temperature correction
Climate type a0 a1
Dry with freezing 6.29 436.36
Dry without freezing 7.68 436.36
Wet with freezing 5.03 327.27
Wet without freezing 6.66 218.18
! Distribution of total traffic loading according to temperature gradients
The total traffic loading since the construction of pavement is distributed over the
temperature gradients as follows:
NE4
n tg = * FREQ tg ...(4.4)
LCR tg
where:
ntg number of 18 kip equivalent standard axle load passes during

temperature gradient tg (ESALs per lane)
NE4 cumulative number of ESALs since construction of pavement, in
millions 18-kip axles per lane
FREQtg frequency of each temperature gradient tg
LCRtg lateral coverage ratio of traffic, for temperature gradient tg

Version 1.0
! Determination of the lateral coverage ratio of traffic
The lateral coverage ratio of traffic (LCR) is simply a measure of the likelihood of the
wheel loading passing through the critical edge location. The edge loading location is
considered critical for jointed plain concrete pavements as this is the location of the
maximum stress and will be the point of crack initiation.
Assuming that the average location of vehicle wheels is 22 inches from the edge of the
slab, with a standard deviation of 8.4 inches, the following regression equation can be
used to calculate LCR:
LCR tg = 418.9 - 1148.6 * SR tg + 1259.9 * SR 2tg - 491.55 * SR 3tg ...(4.5)
SIGMA tg
SR tg = ...(4.6)
MR
where:
LCRtg lateral coverage ratio of traffic, for temperature gradient tg

SRtg ratio between combined stress in slab and the Modulus of Rupture of
concrete, for temperature gradient tg
SIGMAtg combined stress in the slab edge due to loading and curling for
temperature gradient tg (psi)
MR modulus of Rupture of concrete (psi)
! Maximum number of Ntg
The maximum number of load repetitions to the failure of concrete slab (N) applied
during temperature gradient tg depends on the induced stress level, and is calculated
through the law of fatigue as follows:
Log10 (N tg ) = 2.13 * SR -1.2

tg ...(4.7)
where:
Ntg maximum number of 18 kip equivalent standard axle load repetitions

during temperature gradient tg before flexural failure occurs (ESALs per
lane)
SRtg ratio between combined stress in slab and the Modulus of Rupture of
concrete, for temperature gradient tg
4.1.2 Calculation of stresses

The combined stress due to curling and loads, for each temperature gradient tg, is obtained
from Equation 4.8 below:
SIGMA tg = f SB * ( σ load (tg) + R tg * σ curl (tg) ) ...(4.8)
where:

Version 1.0
SIGMAtg combined stress in slab edge for temperature gradient tg (psi)

fSB adjustment factor for stabilised bases
σload(tg) stress in slab edge due to traffic loading (psi)
Rtg regression coefficient

σcurl(tg) stress in slab edge due to curling (psi)
! Calculation of load induced stress
The stress at the slab edge produced by traffic loads, for each temperature gradient, is
expressed as follows:
σ load = fES * f WL * σ e ...(4.9)
where:
σload stress in slab edge due to traffic loading (psi)

fES adjustment factor for edge support (for example, shoulder)
FWL adjustment factor for widened outside lanes

σe edge stress obtained from Westergaard’s equations (psi)
! Calculation of edge stress (σe)
Edge stress in the slab is calculated using Westergaard’s equation (Westergaard, 1948)
for a circular load, in which the load application radius for a simple axle is replaced by
the equivalent radius for a single axle dual wheel, as follows:
3 * (1 + µ ) * P   E c * SLABTHK 3  4µ 1 - µ a eq 
σe = ln  + 1.84 -

+ + 1.18 * (1 + 2µ ) 
π (3 + µ ) * SLABTHK 2 4
  100 * KSTAT * a eq  3 2 l 
...(4.10)
where:
σe edge stress obtained from Westergaard’s equation (psi)
µ Poisson’s ratio
P total load applied by each wheel of a single-axle dual wheel (lb), default =
9000
Ec modulus of elasticity of concrete (psi)
aeq equivalent load application radius for a dual-wheel single axle (inches)
radius of relative stiffness of the slab-foundation system (inches)

Version 1.0
The equivalent load application radius (aeq) is calculated from the following expression:

2 2
 SP  a  SP   SP  a
0.909 + 0.339485 *  a  + 0.103946 *  l  - 0.017881 *  a  - 0.045229 *  a  * l 
a eq  
= 
a  3 3 2 3 
+ 0.000436 *  SP  - 0.301805 * SP *  a  + 0.034664 *  SP  + 0.001 *  SP  * a 
        
 a  a l  l   a  l 
...(4.11)
Limits: 0 < SP/a < 20

0 < a/ l < 0.5
where:
aeq equivalent load application radius for a dual-wheel single axle (inches)
a load application radius for a single-wheel single axle, in inches. This is
given by the square root of (P/π*p)
p tyre pressure (psi)
SP spacing between central wheels of dual wheel single axle (inches)
l radius of relative stiffness of the slab-foundation system (inches)
The radius of relative stiffness of the slab-foundation system is calculated from the
0.25
 E c * SLABTHK 3 
l=  ...(4.12)
 12 * (1 − µ ) * KSTAT 
2
where:

µ Poisson’s ratio
! Calculation of the adjustment factor for shoulders (f ES)
In pavement sections with concrete shoulders or other forms of edge support (such as
kerb), load stress should be multiplied by the edge support adjustment factor, calculated
as follows:

Version 1.0
100
fES = ...(4.13)
100 + LTE sh
where:
fES adjustment factor for the edge support

LTEsh efficiency of load transfer between slab and edge support (for example,
shoulder), (%)
Default:
= 20 if concrete shoulders are placed during initial construction
= 10 if concrete shoulders are placed after initial construction
! Calculation of the adjustment factor for widened outside lanes (f WL)
In sections with widened lanes, the load stress should be multiplied by an adjustment
factor, calculated as follows (Benekohal et al., 1990):
2 3
0.013211 * l
+ 0.386201 * 
a   a  + 0.053891 *  a 
f WL = 0.454147 +  - 0.24565 *    
DW  DW   DW   DW 
...(4.14)
where:
fWL adjustment factor for widened outside lanes

DW average wheels location, given by the average distance of the exterior
wheel to slab edge (inches)
a load application radius for single-wheel single axle (inches)
! Calculation of stresses produced by curling
Curling stress is given by Equation 4.15 below:
COEF * E c * α * ∆Ts
σ curl = ...(4.15)
2
where:
σ curl stress in slab edge due to curling (psi)

COEF curling stress coefficient
α thermal coefficient of concrete (default = 5.5*10- 6), (per ºF)

Version 1.0
The curling stress coefficient (COEF) in Equation 4.15 above is obtained from Equation 4.16
below developed by Westergaard (1926) and Bradbury (1938):
2 * cos(λ ) * cosh(λ )  * tan( λ ) +  sinh( λ ) 

COEF = 1 -    ...(4.16)
 ( sin(2λ ) + 2 * sinh( λ ) * cosh(λ ))    cosh( λ ) 
and:
12 * JTSPACE
λ= ...(4.17)
l* 8
where:
λ intermediate parameter expressed in sexagesimal degrees

JTSPACE average transverse joint spacing (ft)
! Calculation of the regression coefficient (R)
The load induced stresses and curling stresses cannot be added directly since curling
produces a debonding effect between the slab and the base. This effect is taken into
account by the regression coefficient (Rtg) in Equation 4.8 above.
The regression coefficient is calculated for each temperature gradient using the following
equation:

(
86.97 * Y 3 - 1.051 * 10 -9 * E c * dT * KSTAT + 1.7487 * dT * Y 2 ) 

 

(
R = − 1.068 - 0.387317 * dT - 1.84 * 10 -11 2 2
* E c * dT * KSTAT + 8.16396 * dT * Y ) 

 
 

(
+ 1.062 - 1.5757 * 10 -2 * dT - 8.76 * 10 -5 * KSTAT + (1.17 - 0.181 * dT) * 10 -11 * E * dT * KSTAT 
c 
)
...(4.18)
The intermediate parameters Y and dT in Equation 4.18 above are expressed as follows:
12 * JTSPACE
Y= ...(4.19)
100 * l
dT = α * ∆Ts * 10 5 ...(4.20)
where:


Version 1.0
α thermal coefficient of concrete (per ºF)
! Calculation of the adjustment factor for stabilised bases (fSB)
The effect of stabilised bases on the performance of concrete pavement structure is

considered in Equation 4.8 above through the adjustment factor fSB. The adjustment
factor is based on the effective slab thickness, which represents the equivalent thickness
of a plain concrete slab that would give the same structural response of the current
pavement (that is, the slab and the base).
The adjustment factor is calculated as follows:
2 * (SLABTHK - NAXIS)
f SB = ...(4.21)
EFFETHK
where:
fSB adjustment factor for stabilised bases

= 1.0 if EFFETHK is equal to SLABTHK
NAXIS location of the neutral axis
EFFETHK effective slab thickness (inches)
The location of the neutral axis (NAXIS) and the effective slab thickness (EFFETHK)
are calculated from Equations 4.22 below and 4.23 below, respectively:
 E base 
2
 0.5 * SLABTHK + E * BASETHK * (SLABTHK + 0.5 * BASETHK )
NAXIS =  c 
 E base 
 SLABTHK + * BASETHK 
 E c 
...(4.22)
0.5
 E * BASETHK 
EFFETHK = SLABTHK 2 + BASETHK 2 * base 
 E c * SLABTHK 
...(4.23)
where:

Ebase modulus of elasticity of stabilised base (psi)
BASETHK thickness of the stabilised base (inches)

Version 1.0
4.1.3 Key factors

The following factors have significant effects on the propagation of transverse cracking:
! Slab thickness
! Joint spacing
! Concrete flexural strength
! Climate/environment
4.2 Jointed reinforced concrete pavements

Low severity transverse cracks usually occur in JR concrete pavements due to shrinkage,
curling and contractions of concrete caused by variations in the mean temperature. The
reinforcement steel in a JR concrete pavement holds the cracks tightly closed and ensures load
transfer by aggregate interlock, thus reducing distress progression. However, crack
propagation may increase due to repetitive traffic loading and environmental effects (leading
to corrosion of reinforcement). Only medium and high severity transverse cracks in JR
concrete pavements are modelled in HDM-4, since these types may increase road roughness
significantly.
The number of deteriorated transverse cracks per mile is given by the following relationship
(ERES Consultants, 1995):
6.88 * 10 -5 * FI/SLABTHK + NE4 * (0.116 - 0.073 * BASE) 

 
DCRACK = Kjrc * AGE 2.5 * * (1 - exp(-0.032 * MI) 
 
* exp[7.5518 - 66.5 * PSTEEL - (1 - 5 * PSTEEL) * E c * 10 -6 ]
...(4.24)
where:
DCRACK number of deteriorated transverse cracks per mile

AGE number of years since pavement construction
FI freezing Index (ºF-days)
NE4 cumulative ESALs since pavement construction (millions 18-kip axles per
lane)
BASE base type:
0 if non stabilised
1 if stabilised
MI Thornthwaite moisture index
PSTEEL percentage of longitudinal steel reinforcement
Kjrc calibration factor (default = 1.0)

Version 1.0
Note that this model does not use the spacing between joints to predict crack deterioration.
The model highlights the following:
! Crack deterioration increases with pavement age and traffic.
! A significant increase in the quantity of longitudinal steel reinforcement (greater than
0.15%) reduces the number of deteriorated cracks.
! Crack deterioration occurs at a higher rate in cold and wet climates (higher values of MI).
! Stabilised bases gives less cracking than non-stabilised bases.
! Crack deterioration can be reduced by using higher strength concrete or by increasing
slab thickness.

Version 1.0
5 Faulting
Faulting is caused by the loss of fine material under a slab and the increase in fine material
under nearby slabs. This flow of fine material is called pumping, and is caused by the
presence of high levels of free moisture under a slab carrying heavy traffic loading. The
effects of thermal and moisture-induced curling and lack of load transfer between slabs
increase pumping.
The HDM-4 pavement deterioration model considers faulting in jointed plain concrete
pavements (with and without load transfer dowels) and jointed reinforced concrete pavements
as described in Sections 5.1, 5.2 and 5.3.
5.1 JP concrete pavements without load transfer dowels

The relationship for modelling transverse joint faulting in jointed plain concrete pavements
without load transfer dowels is as follows (ERES Consultants, 1995):
0.2347 - 0.1516 * Cd - 0.00025 * (SLABTHK 2 /JTSPACE 0.25 )

 
FAULT = Kjpn f * NE4 0.25 * − (0.0115 * BASE + 7.78 * 10 -8 * FI1.5 * PRECIP 0.25 ) 
 
- (0.002478 * DAYS90 0.5 - 0.0415 * WIDENED ) 
 
...(5.1)
where:
FAULT average transverse joint faulting (inches)

NE4 cumulative ESALs since pavement construction (millions 18-kip axles per
lane)
Cd drainage coefficient, modified AASHTO
BASE base type:
0 if not stabilised
1 if stabilised
PRECIP annual average precipitation (inches)
DAYS90 number of days with mean temperature greater than 90ºF
WIDENED widened lane:
0 if not widened
1 if widened
Kjpnf calibration factor (default = 1.0)

Version 1.0
The following design characteristics can be used to reduce faulting in jointed plain concrete
pavements without load transfer dowels:
! Provision of better drainage conditions
For example, use of longitudinal drains, and more permeable bases.

! Use of stabilised bases (or lean concrete bases)
! Use of widened lanes or concrete shoulders
! Provide shorter joint spacing or use thicker slabs
Since this model predicts the average faulting, it is recommended that the critical level of
intervention should be set fairly low (at around 0.07 in) to provide some safety factor. In
situations where faulting in JP concrete pavements without load transfer dowels is predicted to
be excessive, the use of load transfer dowels has to be considered.
5.2 JP concrete pavements with load transfer dowels

The use of dowels as a load transfer mechanism reduces transverse joint faulting in concrete
pavements. Transverse joint faulting in JP concrete pavements with load transfer dowels is
predicted as described below, (ERES Consultants, 1995):
0.0628 * (1 - Cd) + 3.673 * 10 -9 * BSTRESS 2 

 
FAULT = Kjpd f * NE4 0.25 * + (4.116 * 10 - 6 * JTSPACE 2 + 7.466 * 10 -10 * FI 2 * PRECIP 0.5 ) 
 
- (0.009503 * BASE - 0.01917 * WIDENED + 0.0009217 * AGE )
 
...(5.2)
where:

NE4 cumulative ESALS since pavement construction (millions 18-kip axles
per lane)
BSTRESS maximum concrete bearing stress, in the dowel-concrete system (psi)
BASE base type:
0 if not stabilised
1 if stabilised
0 if not widened
1 if widened or shoulders provided during initial construction
0.5 if concrete shoulders are placed after initial construction

Version 1.0

Kjpdf calibration factor (default = 1.0)
The value of the maximum concrete bearing stress (BSTRESS) has a significant impact on the
predictions of the faulting model, and is calculated as follows (Heinrichs et al., 1989):
DFAC * P * LT * Kd * ( 2 + BETA * OPENING )

BSTRESS = ...(5.3)
4 * E s * INERT * BETA 3
where:

DFAC distribution factor, given by 24/( l + 12)
l radius of relative stiffness of the slab-foundation system (inches). This is

calculated using Equation 4.12 above
P total load applied by each wheel of a single-axle dual wheel (lb) (default =
9,000)
LT percentage of load transfer between joints (default = 45)
Kd modulus of dowel support, in pci (default = 1.5*106 psi/in)
BETA relative stiffness of the dowel-concrete system
OPENING average transverse joint opening (inches)
Es modulus of elasticity of dowel bar (psi)
INERT moment of inertia of the transverse section of the dowel bar (in4)
The relative stiffness of the dowel-concrete system (BETA) is obtained from Equation 5.4
below:
0.25
 Kd * DOWEL 
BETA =   ...(5.4)
 4 * E s * INERT 
where:

Kd modulus of dowel support (pci) (default = 1.5*106 psi/in)
DOWEL dowel diameter (inches)
Es modulus of elasticity of dowel bar (psi)
INERT moment of Inertia of the transverse section of the dowel bar (in4)
The average transverse joint opening (OPENING) is given by:

Version 1.0
α * TRANGE 
OPENING = 12 * CON * JTSPACE *  
+γ  ...(5.5)
 2  
where:

CON adjustment factor due to base/slab frictional restraint:
0.80 if non stabilised base
0.65 if stabilised base
α thermal coefficient of concrete (per °F)
TRANGE temperature range (the mean monthly temperature range obtained from data on
the difference between the maximum and the minimum temperature for each
month) (°F)
γ drying shrinkage coefficient of concrete
The moment of inertia of the dowel bar (INERT) is given by:

4
 DOWEL 
INERT = 0.25 * π *   ...(5.6)
 2 
where:
INERT moment of Inertia of the transverse section of the dowel bar (in4)
DOWEL dowel diameter (inches)
Based on model results, the following design characteristics can reduce faulting:
! Use of load transfer dowels of greater diameter to reduce the stress levels at the
dowel-concrete support system
For example, use of longitudinal drains or permeable bases.

! Use of widened outside lanes
! Use of concrete shoulders
! Shorter spacing between transverse joints
Faulting in JP concrete pavements is more likely to occur in cold and wet climates than in
warm climates.

The relationships for modelling faulting in jointed reinforced concrete pavements are the same
as those used for jointed plain concrete pavements with load transfer dowels with the
exception of the calibration factor Kjrf.

Version 1.0
0.0628 * (1 - Cd) + 3.673 * 10 -9 * BSTRESS 2 

 
FAULT = Kjr f * NE4 0.25 * + (4.116 * 10 - 6 * JTSPACE 2 + 7.466 * 10 -10 * FI 2 * PRECIP 0.5 ) 
 
- (0.009503 * BASE - 0.01917 * WIDENED + 0.0009217 * AGE )
 
...(5.7)
where:

NE4 cumulative ESALS since pavement construction (millions 18-kip axles
per lane)
BASE base type:
0 if not stabilised
1 if stabilised
0 if not widened
1 if widened or shoulders provided during initial construction
0.5 if concrete shoulders are placed after initial construction
Kjrf calibration factor (default = 1.0)
The following design characteristics can reduce faulting in JR concrete pavements and JP
concrete pavements with dowels:
! Use of load transfer dowels of greater diameter to reduce the stress levels at the
dowel-concrete support system
For example, use of longitudinal drains or permeable bases.

! Use of widened outside lanes
! Shorter spacing between transverse joints
! Use of stabilised bases
Note that the model does not include the effects of shoulder types.

Version 1.0
6 Spalling
Transverse joint spalling is the cracking or breaking of the edge of the slab up to a maximum
of 0.6 metres from the joint. Spalling generally does not extend through the whole thickness
of the slab, but intercepts the joint at an angle. Transverse joint spalling can be caused by a
variety of factors including:
! Presence of incompressible materials
The presence of incompressible materials in the joint which produces excessive stress in
the joint. This produces a fracture or detachment of the joint edges when the slab expands
in warm conditions.
! Disintegration of concrete under high traffic loads
! Improper consolidation of the concrete in the joint
! Wrongly designed or built load transfer system
The HDM-4 model considers medium and high severity spalled transverse joints.

Transverse joint spalling in jointed plain concrete pavements is predicted using Equation 6.1
below (ERES Consultants, 1995):
549.9 - 895.7 * (LIQSEAL + PREFSEAL ) 

 
 
+ 1.11 * DAYS90 3 * 10 -3 + 375 * DWLCOR 
 
SPALL = Kjp s * AGE2 * JTSPACE * 10- 6 * 
 + ( 29.01 - 27.6 * LIQSEAL ) * FI 
 
 
 − ( 28.59 * PREFSEAL + 27.09 * SILSEAL ) * FI
 
...(6.1)
where:
SPALL percentage of spalled transverse joints

AGE age since pavement construction (years)
LIQSEAL presence of liquid sealant in joint:
0 if not present
1 if present
PREFSEAL presence of pre-formed sealant in joint:
0 if not present
1 if present
DAYS90 number of days with temperature greater than 90ºF

Version 1.0
DWLCOR dowel corrosion protection:

0 if no dowels exist, or are protected from corrosion
1 if dowels are not protected from corrosion
SILSEAL presence of silicone sealant in joint:
0 if not present
1 if present
Kjps calibration factor (default = 1.0)
The following observations can be made regarding the behaviour of the JP concrete pavement
model:
! The transverse joint spalling increases with the pavement age to the second power
! Pre-formed seals are more effective in reducing joint spalling than other types of
seals considered
! Transverse joints without seals, show a great quantity of spalling
! Liquid sealant gives a better performance than the silicon seals
! An increase in joint spacing increases the percentage of spalled joints
! An appropriate dowel protection against corrosion reduces joint spalling
The model also shows that the ageing of JP concrete pavements has a significant effect on
joint spalling. Effective maintenance strategies that include joint cleaning and resealing at
regular intervals can reduce the ageing effect, and therefore, reduce joint spalling
significantly.

Transverse joint spalling in jointed reinforced concrete pavements is predicted using Equation
6.2 below (ERES Consultants, 1995):
1.94 * DWLCOR + 8.819 * BASE * (1 - PREFSEAL )

SPALL = Kjr s * AGE * JTSPACE * 10
3 -5
*  

+ 7.01 * FI * 10 -3 
...(6.2)
where:
SPALL percentage of spalled transverse joints

JTSPACE average transverse Joint Spacing (ft)
DWLCOR dowel corrosion protection:
0 if no dowels exist, or are protected from corrosion
1 if dowels are not protected from corrosion

Version 1.0
BASE base type:

0 if not stabilised
1 if stabilised
PREFSEAL presence of pre-formed sealant in joint:
0 if not present
1 if present
Kjrs calibration factor (default =1.0)
The following observations can be made regarding the behaviour of the JR concrete pavement
model:
! The percentage of spalled joints increases rapidly with the age of the pavement
! Transverse joint spalling is more likely to occur in cold climates than in warm
climates
! An increase in joint spacing results in an increase of the percentage of spalled
joints
! Protecting dowels against corrosion will reduce spalling
! The use of a stabilised dense base increases joint spalling
! Pre-formed seals reduce spalling in pavements with stabilised bases

Version 1.0
7 Failures
This is the main deterioration mode that occurs in Continuously Reinforced concrete
pavements. Located failures include loosening and breaking of reinforcement steel and
transverse crack spalling. These are caused by high tensile stresses induced in the concrete
and reinforcement steel by traffic loading and changes in environmental factors. Most
maintenance activities carried out on CR concrete pavements are directly related to failures.
Failures in continuously reinforced concrete pavements are predicted using Equation 7.1
below (Lee et al., 1991):
6.8004 - 0.0334 * SLABTHK 2 - 6.5858 * PSTEEL 

 
Log e (FAIL ) = Kcr f * + 1.2875 * log e (NE4 ) - 1.1408 * AB - 0.9367 * SB 
 
- 0.8908 * GB - 0.1258 * CHAIRS 
...(7.1)
where:
FAIL number of failures per mile in the more trafficked lane (number/mile)
SLABTHK slab thickness (in)
PSTEEL percentage of longitudinal reinforcement steel (%)
NE4 cumulative equivalent standard axle load (ESALs) since pavement
construction (millions per lane)
AB 1 if base type is asphaltic
0 in other cases
SB 1 if base type is cement stabilised
0 in other cases
GB 1 if base type is granular
0 in other cases
CHAIRS 1 if chairs are used for installation of the reinforcement
0 if tubes are used
Kcrf calibration factor (default = 1.0)
Slab thickness and percentage of reinforcement steel have a significant effect on the number
of failures in CR concrete pavements. Installation of reinforcement steel with chairs produces
fewer failures than installation using tubes. The use of stabilised or asphaltic bases further
reduces the risk of failures.

Version 1.0
8 Serviceability loss
Present Serviceability Rating (PSR) is a subjective user rating of the existing ride quality of
pavement condition. PSR has been correlated with various roughness indicators, such as slope
variance and IRI. It is a reflection of the user response to pavement condition.

PSR values for JR concrete pavements are predicted in HDM-4 using the following
relationship (ERES Consultants, 1995):
PSR = 4.165 - 0.06694 * TFAULT 0.5 - 0.00003228 * DCRACK 2 - 0.1447 * SPALL0.25

...(8.1)
where:
PSR present serviceability rating

TFAULT total transverse joint faulting per mile (in/mile)
DCRACK number of deteriorated transverse cracks per mile
SPALL percentage of spalled joints
The total joint faulting per mile (TFAULT) is calculated as:
FAULT * 5280
TFAULT = ...(8.2)
JTSPACE
where:
TFAULT total transverse joint faulting per mile (in/mile)

FAULT average transverse joint faulting (in)
It should be noted that cracking has a very significant influence on the predictions of the PSR
model.
8.2 Continuously reinforced concrete pavements

The model of serviceability loss for continuously reinforced concrete pavements predicts loss
of serviceability in the traditional scale of 0 to 5, based on pavement age since construction,
cumulative equivalent standard axle loads and slab thickness.
The model is as follows (Lee et al., 1991):
0.79 - 1.3121 * log10 (SLABTHK ) 

Log 10 (PSR 0 - PSR t ) =   ...(8.3)
 + 0.1849 * log10 (AGE ) + 0.2634 * log10 (NE4 )
where:

Version 1.0
PSR0 initial PSR at the time of pavement construction (default = 4.5)

PSRt predicted PSR value at time t
SLABTHK slab thickness (in)
NE4 cumulative equivalent standard axle load (millions 18-kip axles per lane)

Version 1.0
9 Roughness

Roughness on JP concrete pavements is calculated as a function of faulting, spalling and
transverse cracking (ERES Consultants, 1995):
RI t = Kjp r * (RI0 + 2.6098 * TFAULT + 1.8407 * SPALL + 2.2802 * 10 -6 * TCRACKS 3 )

...(9.1)
where:
RIt roughness at time t (in/mile)

RI0 initial roughness at the time of pavement construction (in/mile), (default =
98.9)
TFAULT total transverse joint faulting per mile (in/mile) (calculated from Equation 8.2
above)
SPALL percentage of spalled joints
TCRACKS total number of cracked slabs per mile
Kjpr calibration factor (default = 1.0)
The total number of cracked slabs per mile (TCRACKS) is calculated as:
PCRACK * 5280
TCRACKS = ...(9.2)
JTSPACE * 100
where:
TCRACKS total number of transverse cracks per mile


Roughness on JR concrete pavements is calculated as a function of PSR (Al-Omari and
Darter, 1994):
RI t = Kjrr * − log e  
0.2 * PSR t
  ...(9.3)
 0.0043 
where:

Version 1.0
PSRt serviceability rating at time t

Kjrr calibration factor (default = 1.0)
9.3 Continuously reinforced concrete pavements

Roughness on CR concrete pavements is calculated as a function of PSR (Al-Omari and
Darter, 1994):
RI t = Kcrr * − log e  
0.2 * PSR t
  ...(9.4)
 0.0043 
where:

PSRt serviceability rating at time t
Kcrr calibration factor (default = 1.0)

Version 1.0
The deterioration models contain calibration factors to facilitate local calibration. The
calibration factors have default values of 1.0 and are summarised in Table C3.13.
Table C3.13 Calibration factors used in the deterioration models of concrete

pavements
Pavement Calibration Deterioration model

surface type factor
Kjpc Transverse cracking calibration factor
Kjpnf Faulting calibration factor in JP concrete pavements without dowels
JP Kjpdf Faulting calibration factor in JP concrete pavements with dowels
Kjps Joint spalling calibration factor
Kjpr Roughness (IRI) progression calibration factor
Kjrc Cracking deterioration calibration factor
Kjrf Faulting calibration factor

JR
Kjrs Joint spalling calibration factor
Kjrr Roughness progression calibration factor
Kcrf Failures calibration factor
CR
Kcrr Roughness progression calibration factor

Version 1.0
11 References
The AASHO Road Test, (1962)
Report 5 - Pavement research, Special Report No. 61E
The Highway Research Board
Washington DC, USA
AASHTO, (T22-92) American Association of State Highway and Transportation Officials
Standard Method of Test for Compressive Strength of Cylindrical Concrete Specimens
AASHTO, (T140-92) American Association of State Highway and Transportation Officials
Standard Method of Test for Compressive Strength of Concrete Using Portions of Beams
Broken Inflexure (ASTM C116)
AASHTO, (T97) American Association of State Highway and Transportation Officials
Standard Method of Test for Time of Setting of Concrete Mixtures by Penetration
Resistance (ASTM C403-90)
AASHTO, (1986) American Association of State Highway and Transportation Officials
Guide for Design of Pavement Structures
Washington DC, USA
Washington DC, USA
ASTM C39, American Society for Testing Materials (1994)
Standard Test Method for Compressive Strength of Cylindrical Concrete Specimens
ASTM C78, American Society for Testing Materials (1994)
Standard Test Method for Flexural Strength of Concrete (Using Simple Beam with
Third-Point Loading)
ASTM C469
ASTM E-867-82A
Al-Omari B., and Darter M.I., (1994)
Relationships Between IRI and PSR
Transportation Research Board, 73rd Annual Meeting
Washington DC, USA
Benekohal R.F., Hall K.T., and Miller H.W., (1990)
Effects of Lane Widening on Lateral Distribution of Truck Wheels
Transportation Research Record 1286
Transportation Research Board
Washington D.C.

Version 1.0
Bradbury R.D., (1938)

Reinforced Concrete Pavements
Wire Reinforcement Institute
Eisenmann J., and Leykauf G., (1990)
Simplified Calculation Method of Slab Curling Caused by Surface Shrinkage,
Proceedings, 2nd Internacional Workshop on Theoretical Design of Concrete Pavements
Madrid, Spain
ERES Consultants, and FHWA (1995)
Performance of Concrete Pavements, Volume III: Improving Concrete Pavement
Performance
U.S. Department of Transportation and Federal Highway Administration
Champaign, Illinois
Foxworthy P.T., (1985)
Concepts for the Development of a Non-destructive Testing and Evaluation System for
Rigid Airfield Pavements
PhD. Thesis, University of Illinois
Illinois, USA
Heinrichs, K.W., et al, (1989)
Rigid Pavements Analysis and Design
Federal Highway Administration, FHWA-RD-88-068
Lee Y.H., and Darter M.Y., (1994)
Development of Pavements Performance Prediction Models
Report Nº529-2, Illinois Department of Transportation
University of Illinois at Urbana-Champaign
Lee Y.H., Mohseni A., and Darter M.Y., (1991)
Pavement Model Enhancements for The Highway Performance Monitoring System
(HPMS)
Final Report, Illinois Department of Transportation and Federal Highway Administration
Pauw, A., (1960). Static Modulus of Elasticity of Concrete as Affected by Density.
ACI Journal, Proceedings Volume 57, No.6, Dic, pp. 679-687.
Sayers, M., Gillespie, T.D., and Paterson, W.D.O. (1986). Guidelines for the Conduct and
Calibration of Road Roughness Measurements. Technical Report no. 46, World Bank,
Washington, D.C.
SHRP, (1993)
Distress Identification Manual for the Long-Term Pavement Performance Project Report
SHRP-P-338, Strategic Highway Research Program
National Research Council
Westergaard H.M., (1926)
Analysis of Stresses in Concrete Pavements Due to Variations of Temperature
Proceedings, Sixth Annual Meeting, Highway Research Board
Washington D.C.

Version 1.0
Westergaard H.M., (1948)

New Formulas for Stresses in Concrete Pavements of Airfields
Transactions, American Society of Civil Engineers, Volume 113

Version 1.0
C4 Unsealed Roads
1 Introduction
This chapter describes the detailed modelling of unsealed road deterioration (see Figure C4.1).
Road
RoadDeterioration
Deterioration
Models
Models
Pavement Bituminous Concrete Unsealed

UnsealedRoads
Pavementtypes
types Bituminous Concrete Roads
Pavements
Pavements Pavements
Pavements
Chapter C1
Chapter C1 Chapter
ChapterC4
C4
Chapter
ChapterC2
C2 Chapter
ChapterC3
C3
Figure C4.1 Road Deterioration Modules
The HDM-4 Road Deterioration model for unsealed roads is based on the specifications given
in the HDM-III documentation by Watanatada et al. (1987) reproduced with the approval of
the World Bank. Minor modifications have been made in the text and to the models by
incorporating calibration factors to facilitate local calibration and adaptation. The background
of the model is given in Paterson (1987).

Version 1.0
PART C ROAD DETERIORATION MODELS C4 UNSEALED ROADS
2 Modelling logic
2.1 Classification, concepts, and logic

Unpaved roads comprise the lower classes of the road network hierarchy, and generally carry
low volumes of traffic ranging from a few vehicles up to several hundred vehicles per day.
The geometric standards vary considerably and it is necessary to make a primary classification
of unpaved roads into engineered roads, which have controlled alignment, formation width,
cross-section profile and drainage; and tracks, which are essentially ways formed by
trafficking along natural contours with or without the removal of topsoil. Unpaved roads
classified in a country's network are usually engineered or partly engineered, and tracks are
usually not classified.
The analysis of unsealed road deterioration is designed primarily for engineered unpaved
roads, of either gravel or earth surfacing, because the empirical models are based on a variety
of such roads. When necessary it is possible to use the relationship also for tracks as a first
estimate, but the user needs to be aware that the environmental effects of drainage and rainfall
may be poorly represented.
The deterioration of unpaved roads is characterised primarily by roughness and by material
loss from the surfacing. The prediction relationships for these are based on analyses of the
Brazil-UNDP study (Visser, 1981; and Paterson, 1987). Wheelpath ruts also develop under
traffic but the ruts are usually not straight forward, often being mixed with water-induced
surface erosion. Thus the concept of rut depth was not used in HDM-III and is subsumed in
the property of roughness; prediction relationships may be found in Visser (1981). The
looseness of surfacing material, which was analysed in the Kenya study (Hodges et al., 1975),
was also observed in the Brazil-UNDP study (GEIPOT, 1982). But as it was found to have no
substantial effect on vehicle speed, no prediction relationships were incorporated to HDM-III.
Finally, road passability is an important criterion for upgrading tracks or earth roads to gravel
roads. HDM-4 allows for an increase in vehicle operating costs by a factor specified by the
user. This is to reflect the economic effects of reduced passability when the gravel thickness
drops below a minimum level (see Part E).
The periodic grading of unpaved roads is usually undertaken on a more-or-less regular basis
for management purposes, either seasonally or frequently enough to keep the roughness
within tolerable limits.
These repeated cycles of roughness deterioration and grading maintenance are treated as
continual by the model. The average roughness during each analysis year is computed as a
function of the roughness at the beginning of the year, of material, traffic, geometry and
rainfall parameters and the specified grading frequency. Over a period of time depending on
the traffic volume and frequency of grading, the annual average roughness tends towards a
long-term average roughness that is also computed.
Maintenance of the gravel surfacing is accounted each analysis year through the surfacing
thickness and the net change from material loss, spot regravelling and gravel resurfacing
maintenance. The material loss from earth roads, although computed, is accounted only for the
purpose of predicting spot regravelling quantities and is otherwise ignored.
The computational logic described above is simplified by considering that an unpaved road
comprises two layers, a gravel surfacing and a subgrade. A gravel road has both layers, but an
earth road has a zero thickness of gravel surfacing and its surface characteristics are those of
the subgrade. When a gravel road loses all of its gravel surfacing, then its classification

Version 1.0
reverts to that of an earth road. Upon gravel resurfacing, all unpaved roads become gravel
roads by definition of the new surfacing layer.
Deterioration is predicted using the properties of the surfacing layer, whether that is gravel or
subgrade, as it is defined for the analysis year. Thus the user must specify the physical
properties of both gravel surfacing and subgrade for unpaved roads.
The pavement classification framework for defining the modelling of unsealed road
deterioration and works effects is shown in Table C4.1.
Table C4.1 Generic HDM-4 unsealed pavement types
Pavement type Surface type Base type Description
GRUP GR n/a Granular Unsealed Pavement (for example, gravel

road)
EAUP EA n/a Earth Unsealed Pavement
SAUP SA n/a Sand Unsealed Pavement
n/a not applicable for unsealed pavements
NDLI (1995) give definitions of the characteristics used to define different types of pavements
into the above framework and alternative terminology applied to the same pavement materials
(see also Chapter C1).
2.2 Primary model parameters
2.2.1 Material properties

Previously, deterioration relationships have been categorised by material type (lateritic,
quartzitic, coral, volcanic, etc.), but from the Brazil-UNDP study it has been possible to
replace these by material properties that should improve the transferability of the
relationships.
The material properties, which were found to affect the rate of deterioration in Brazil, include
the maximum particle size, the particle size distribution and the soil plasticity (Paterson,
1987). The specific soil properties are used subsequently to define various summary metrics
of the particle size distribution, which are parameters in the deterioration prediction equations.
The minimum and maximum levels of roughness (QIMIN and QIMAX) are predicted
endogenously from the soil properties but the user may override those by specifying input
values. The soil properties are defined for both the gravel and subgrade layers, and this is
denoted by the subscript j, where j = g for gravel surfacing layer, and j = s for subgrade (or
earth road surfacing) layer, in Table C4.2.

Version 1.0
Variable Definition
ADH the average daily heavy vehicle traffic (GVW ≥ 3,500 kg) in both directions (veh/day)
ADL the average daily light vehicle traffic (GVW < 3,500 kg) in both directions (veh/day)
AADT the annual average daily traffic in both directions (veh/day)
C the average horizontal curvature of the road (deg/km)
D95j the maximum particle size of the material, defined as the equivalent sieve opening through which
95% of the material passes (mm)
MGDj dust ratio of material gradation, see Section 3.2
MGj slope of mean material gradation, see Section 3.4
PIj the plasticity index of the material (%)
P075j the amount of material passing the 0.075 mm sieve (or ASTM No. 200 sieve) (% by mass)
P425j the amount of material passing the 0.425 mm sieve (or ASTM No. 40 sieve) (% by mass)
P02j the amount of material passing the 2.0 mm sieve (or ASTM No. 10) (% by mass)
QIavg average roughness during analysis year (QI)
QI(ag) roughness after grading (QI)
QI(bg) roughness before grading (QI)
QIMINj the minimum roughness of the material (either estimated in Section 3.4 or specified) (QI)
QIMAXj the maximum roughness of the material (either estimated in Section 3.2 or specified) (QI)
RF the average absolute rise plus fall of the road (m/km)
Note: RF = 10 times average absolute gradient (%)
SW average width of shoulder (m)
2.2.2 Traffic loading measures

The three traffic-loading variables used in predicting unpaved road deterioration are simply
those of two-way motorised traffic (MT) counts:
1 All vehicles (AADT)
Used in the prediction of material loss. It equals ADL plus ADH,

2 Light vehicles (ADL)
Used in the prediction of roughness,

3 Heavy vehicles (ADH)
Used in the prediction of roughness,

as defined in Table C4.2 and also in Part B of this Manual.
2.2.3 Road geometry measures

The geometric characteristics found to influence the deterioration of unpaved roads in the
Brazil-UNDP study were horizontal curvature (C) and longitudinal gradient (here represented

Version 1.0
by the rise plus fall variable, RF). Roughness progression, and in particular the maximum
roughness, is influenced by both characteristics. In material loss prediction the horizontal
curvature affects the rate of traffic-induced material whip-off and the gradient interacts with
rainfall in causing erosion. Cross-sectional geometry, including crown, camber and
superelevation, were not measured in the study and are discussed in the following section. The
average shoulder width (SW) is used to compute the amount of gravel used in spot
regravelling and gravel resurfacing. The variables RF, C and SW are defined in Table C4.2.
2.2.4 Environment: Climate and Drainage

While the climate of the Brazil-UNDP study area is classed as humid to warm- or wet-humid,
the rainfall pattern was seasonal, ranging from:
! Precipitation of less than 20 mm per month and air humidity less than 40% during a
continuous six to eight months of a year,
to
! Precipitation's of 200 to 600 mm per month and air humidity in excess of 60% over four
months of a year.
The effects of the full range of highly seasonal rainfall were analysed in the study, and are
represented by the average monthly rainfall in the deterioration prediction relationship. The
predictions of annual average roughness and material loss transforms this to an annual average
rainfall and thus makes no specific distinction between uniform and seasonal-rainfall climates.
Geometric cross-sectional characteristics, particularly crown, camber, table side-drains and
run-off points, have pronounced effects on drainage and road deterioration during high
rainfall. In the study area, roughness levels on level, tangent sections that were poorly drained
were very high during wet periods due largely to the rapid development of potholes. On
vertical slopes, roughness levels were frequently low despite extensive erosion by surface
water run-off because the longitudinal profile was affected less than the transverse profile.
The study sections generally had moderate drainage facilities and maintenance, and positive
crowns. The prediction relations therefore apply to unpaved roads with moderate to good
cross-sectional geometry and for dry to wet conditions but may not apply to bathtub type
roads with negative crown or lack of surface drainage in high rainfall conditions.
2.3 Basic computational procedure

The model assumes that the grading operations and spot regravelling specified for each year,
both for gravel and earth roads, are distributed uniformly throughout the year. However, the
gravel resurfacing operation, when it occurs, is performed at the end of the year. Like the
periodic paved road maintenance operations, gravel resurfacing is not permitted in an
effective construction completion year. The computational procedure for road deterioration of
the unsealed roads for each analysis year comprises the following steps:
1 Initialise road characteristics and traffic loading variables at the beginning of the analysis
year.
2 If earth road skip to step 3. Otherwise, check whether the gravel thickness is zero (that
is, no gravel remaining) at the beginning of the analysis year. If the thickness is zero,
reset the road type to earth.
3 If grading is specified compute the annual average road roughness as a function of the
grading frequency, traffic volume, environmental conditions, and attributes of the gravel
(if gravel road) or the subgrade (if earth road). Otherwise, if no grading is specified, set
the average roughness equal to the predicted maximum roughness (QIMAXj) (see
Section 3.2).

Version 1.0
4 Compute the depth of material loss during the analysis year as a function of the traffic
volume, monthly rainfall, and road geometry and the attributes of the gravel (if gravel
road) or the subgrade (if earth road) (see Section 4).
5 Store the results for later use in the RUE (see Part E) and WE (see Part D) models and in
the evaluation and reporting phase.
2.4 Initialisation of variables

At the beginning of the analysis year the traffic variables are computed based on the user-
specified traffic data. The values of the environment, road geometry, and material property
variables are provided in one of the following three ways:
1 From the preceding analysis year
If the analysis year is neither the first year of the analysis period nor an
improvement/construction opening year;
2 From the existing section characteristics data
If the analysis year is the first year of the analysis period;

3 From the improvement/construction option data
If the analysis year is a construction opening year.

The only history variable for unpaved roads is the gravel age, denoted by GAGE, which is
relevant only for gravel roads. It is defined as the number of years elapsed since the last gravel
surfacing or resurfacing. It is initialised as follows:
! When the analysis year is not a construction opening year, the value of GAGE is
provided either from:
❏ the preceding year - if the analysis year is the second or a subsequent year of the
analysis period, or
❏ the existing section characteristics data - if the analysis is the first year of the
analysis period and is increased by one year; and
! When the analysis year is a construction-opening year (of a gravel road project), the
value of GAGE is set to 1 and the unpaved road surface type to gravel, irrespective of the
previous surface type.

Version 1.0
3 Road roughness
3.1 General
The roughness of unpaved roads increases through the shear, mechanical disintegration, and
erosion of the surfacing material caused by traffic and surface water runoff. Roughness levels
are usually 4 to 15 m/km IRI (50 to 200 QI) although lower levels sometimes occur with fine
materials. Roughness in excess of 13 m/km IRI (180 QI) is usually related to depressions,
potholes or transverse erosion gullies, and levels above 22 m/km IRI (300 QI), which
correspond to numerous wheel-sized potholes, are very rare and usually apply only on short
sections or unclassified tracks. The roughness modelled for economic analysis is the profile in
the wheel-paths of the traffic, since this generates the vehicle operating costs. The location of
the wheel-paths tends to vary when roughness reaches high levels as vehicles seek to
minimise the dynamic impact. Hence the prediction of roughness progression must take this
self-regulating tendency into account. On account of the high variability of:
! Material properties,
! Drainage,
! Surface erosion, and
! High roughness levels of unpaved roads,
prediction errors tend to be large, in the order of 1.5 to 2.5 m/km IRI (20 to 32 QI) standard
error, or equivalent to 95 percentile confidence intervals of 20% to 40%.
A number of different model forms have been applied to roughness progression and to the
effects of maintenance grading (Hodges et al., 1975; Visser, 1981; Paterson, 1987). As the
objective of policy analyses can be satisfied by computation of the average roughness
resulting from a specified policy, the model selected for predicting roughness was one which
both represented the progression and grading phases of the roughness cycle realistically and
also permitted a closed-form solution. The model form and its derivation are described in
detail elsewhere (Paterson, 1987). The primary principles and parameter estimates are
described in Sections 3.2 - 3.6.
Although the IRI roughness measure, or other compatible measure, could have been used in
the following relationships (see Sections 3.2 - 3.6) because many parameters are
dimensionless, the nomenclature of QI is used for roughness for internal consistency with the
remainder of the model.
3.2 Roughness progression

In previous models, progression followed either cubic (Hodges et al., 1975; Visser, 1981;
Paterson, 1987) concave curves which, unless restrained, led to unrealistically high
predictions of roughness for policies of infrequent grading. The model form adopted here
constrains the roughness to a high upper limit, or maximum roughness (QIMAXj), by a
convex function in which the rate of progression decreases linearly with roughness to zero at
QIMAX conforms well with practical observations. The predictions of both forms differ
significantly only at high levels of roughness; at low levels of roughness the concave curve is
often more realistic in shape, but quantitatively there is little difference between the two. From
the Brazil-UNDP study (GEIPOT, 1982):
! Maximum roughness was found to be a function of material properties and road
geometry.

Version 1.0
! Rate of roughness progression was found to be a function of the roughness, maximum

roughness, time, light and heavy vehicle passes and material properties, as given by
(Paterson, 1987):
QITG 2 = QIMAX j - b * (QIMAX j - QITG1 ) …(3.1)
where:
QITG1 roughness at time TG1 (QI)

QITG2 roughness at time TG2 (QI)
TG1, TG2 time elapsed since latest grading (days)
QIMAXj maximum roughness for material j (QI)
b model parameter (0 < b < 1)
The variable b is computed as follows:
b = exp[c * (TG 2 - TG1 )] …(3.2)
( )
c = - 0.461 + 0.0174 * ADL + 0.0114 * ADH - 2.87 * AADT * MMP * 10 -5 * 10 −3
…(3.3)
where:
ADL average daily light vehicle traffic in both directions (veh/day)

ADH average daily heavy vehicle traffic in both directions (veh/day)
AADT annual average daily traffic in both directions (veh/day)
The maximum roughness is given by:
[
QIMAX j = MAX 279 - 421 (0.5 - MGD j ) + 0.22 * C − 9.930 * RF * MMP * 10 −3 , 150
2
]
…(3.4)
where:
C average horizontal curvature of the road (deg/km)

RF average rise plus fall of the road (m/km)
MGDj material j gradation dust ratio
The material gradation dust ratio is defined as follows:

Version 1.0
if P425j = 0, then:
MGD j = 1 …(3.5)
if P425j > 0, then:
P075 j
MGD j = …(3.6)
P425 j
where:
P425j the amount of material j passing the 0.425 mm sieve (% by mass)

P075j the amount of material j passing the 0.075 mm sieve (% by mass)
Note: The standard error of this prediction on the original database was 1.5 m/km IRI
(19.8 QI).
3.3 Effect of compaction on roughness progression

Observations on gravel and earth roads in the first few grading cycles after construction or
rehabilitation with full mechanical shaping and compaction, indicate rates of roughness
progression that are much slower than given by the model in Section 3.2. These figures were
derived from roads under repeated grading cycles with no special compaction (Paterson,
1987). Thus if mechanical compaction is specified in the model inputs, the coefficient c is
reduced, initially to one quarter of its predicted value and rising to the full-predicted value
after a few grading cycles, but in a period not exceeding 4 years. Therefore mechanical
compaction can be considered as follows:
[ (
c' = c * MIN 1, 0.25 * t * MAX 1, n 0.33 )] …(3.7)
where:
t time in years since regravelling or construction with mechanical compaction

n frequency of grading (cycles/year)
and thus:
 c' 
…(3.8)
b' = exp365 * 
 n
where:
b', c' are the values of b and c (see Section 3.2) when mechanical compaction is in
effect

Version 1.0
3.4 Effect of grading

The effect of grading maintenance on roughness was found to depend on the roughness before
grading, the material properties and the minimum roughness (QIMINj) (Paterson, 1987). The
minimum roughness, below which grading cannot reduce roughness, increases as the
maximum particle size increases and the gradation of the surfacing material worsens. The
prediction of roughness after grading is expressed as a linear function of the roughness before
grading, dust ratio and the minimum roughness, as given by:
QI(ag ) = QIMIN j + a * (QI(bg ) − QIMIN j ) …(3.9)
where:
QI(ag) roughness after grading (QI)

Ql(bg) roughness before grading (QI)
QIMINj minimum roughness for material j (QI)
a model parameter
The model parameter a is calculated as follows:
a = 0.553 + 0.230 * MGD j …(3.10)
The minimum roughness is given by:
QIMIN j = MAX {10, MIN [100, 4.69 * D95 j * (1 − 2.78 * MG j )]} …(3.11)
where:
MGj slope of mean material gradation

D95j the maximum particle size of the material, defined as the equivalent sieve
opening through which 95% of the material passes (mm)
The slope mean material gradation is calculated as follows:
MG j = MIN (MGM j , 1 − MGM j , 0.36 ) …(3.12)
(MG075 j + MG425 j + MG02 j )

MGM j = …(3.13)
3
The value of parameter MG075j is obtained as follows:

if D95j > 0.4
then:
 P075 j 
log e 
 95 
MG075 j = …(3.14)
 
log e  0.075
 D95 j 

Version 1.0
otherwise:
MG075 j = 0.3

if D95j > 1.0
then:
 P425 j 
log e 
 95 
MG425 j = …(3.15)
 
log e  0.425
 D95 j 
otherwise:
MG425 j = 0.3

if D95j > 4.0
then:
 P02 j 
log e 
 95 
MG02 j = …(3.16)
 
log e  2.0
 D95 j 
otherwise:
MG02 j = MG425 j
Note: The standard error of this prediction on the original database was 2.4 m/km IRI (31.6
QI).
3.5 Average roughness during the analysis year

The average roughness during the analysis year is computed by combining the progression
and grading-effect relationships and integrating (see Paterson, 1987). The year's average is
expressed in terms of the roughness at the beginning of the year and the parameters in the
previous expressions (see Sections 3.2 - 3.4) as follows:
Case 1: if (t*n) ≥ 1
The average roughness during year t (QIavg) is computed as follows:
 y * NS 
QIavg = QIMAX j * (1 - y ) +  …(3.17)
 n 
(b - 1) * n
y= …(3.18)
365 * c

Version 1.0

[ ] (
k 1 − (a * b )n
n * k + 1 − (a * b ) * QIa − 
n )  
 
(1 − a * b )
 
NS = …(3.19)
(1 − a * b)
k = (1 - a) * QIMIN j + a * (1 − b ) * QIMAX j …(3.20)
where:
QIavg average roughness during year t (QI)

QIMAXj maximum roughness for material j (QI)
QIMINj minimum roughness for material j (QI)
n frequency of grading (cycles/year)
QIa roughness at beginning of year t (QI)
a as defined above (see Equation 3.10 above)
b as defined above (see Equation 3.2 above), except that b takes the value b’
when mechanical compaction has been specified
c as defined above (see Equation 3.3 above), except that c takes the value c’
when mechanical compaction has been specified
The roughness at the beginning of the year is obtained as follows:

! First year of analysis period
For the first year of analysis period when t = 1, QIa = QIo (the value specified by the
user).
! Subsequent analysis year
For any subsequent analysis year t, QIa = QIb (= roughness at end of the previous year t-
1, as given below in QI).
In any given analysis year t, the roughness at the end of the year (QIb) is given by:
QIb = (a * b ) * QIa +
n [
k * 1 − (a * b )
n
] …(3.21)
(1 − a * b)
where:
Case 2: if (t*n) < 1

The average roughness during year t (OIavg) is given by:
[exp(365 * c ) - 1]
QIavg = QIMAX j - (QIMAX j - QI a ) * …(3.22)
365 * c
The roughness at the end of the year (QIb) is given by:
QIb = QIMAX j - (QIMAX j - QIa ) * exp(365 * c ) …(3.23)

Version 1.0
where:
3.6 Roughness cycle 'steady state'

When grading is performed regularly at constant time intervals, or a fixed roughness level, or
fixed traffic intervals, the process of roughness change described by these relationships
without restriction eventually leads to a steady state, as shown in Paterson (1987). This steady
state is characterised by a saw-toothed pattern of roughness-time profile, in which the highs
and lows represent the roughness immediately before and after grading, respectively. These
highs and lows, denoted by QIH and QIL, are given by:
[QIMAX j * (1 − b ) + QIMIN j * (1 − a ) * b]
QIH = …(3.24)
(1 − a * b)
[QIMIN j * (1 − a ) + QIMAX j * a * (1 − b)]

QIL = …(3.25)
(1 − a * b )
where:
QIH roughness immediately before grading (QI)

QIL roughness immediately after grading (QI)

The long-term average roughness, denoted by QIlta, at this steady state under a maintenance
policy is dependent on the grading frequency (embodied in the variable b above) and is
obtained by integration over the roughness-time profile, so the annual average roughness
tends to the following:
QI avg → QIlta …(3.26)
and:
[QIMAX j - QIMIN j ]
QIlta = QIMAX j + (1 - a ) * (1 − b ) * …(3.27)
[(1 - a * b ) * log e b]
where:
These relationships are illustrated in Watanatada et al. (1987).

Version 1.0
4 Material loss
From the Brazil-UNDP study the following relationship for predicting the annual quantity of
material loss as a function of monthly rainfall, traffic volume, road geometry and
characteristics of the gravel (if a gravel road) and the subgrade (if an earth road) was
obtained (Paterson, 1985):
(
MLA = K gl * 3.65 3.46 + 2.46 * MMP * RF * 10 -4 + KT * AADT ) …(4.1)
where:
MLA the predicted annual material loss (mm/year)

RF average rise plus fall of the road (m/km)
KT the traffic-induced material whip-off coefficient
Kgl gravel material loss calibration factor
The traffic-induced material whip-off coefficient is expressed as a function of rainfall, road

geometry and material characteristics, as follows:
   0.969 * C   
 0.022 +   + 3.42 * MMP * P075 j * 10 - 6  
   57300   
KT = K kt * MAX 0,    …(4.2)
 - 9.2 * MMP * PI j 10 - 6 - 1.01 * MMP * 10 - 4  
   
   
where:

PIj the plasticity index of material j where:
j = g if a gravel road
j = s if an earth road
Kkt traffic-induced material loss calibration factor
These predictions are illustrated in Watanatada et al. (1987).

Version 1.0
5 Passability
Passability is the quality of the road surface that ensures the safe passage of vehicles. In the
vehicle operating cost model, provision has been made to determine the economic impact of a
partial reduction in passability through factors augmenting the operating costs of the various
vehicle types (see Part E). This augmentation comes into effect when the gravel surfacing
thickness drops below a minimum, and relates to the risk of the subgrade material being
impassable.
The user however must determine endogenously whether passability will be a problem in the
subgrade material, because no physical estimation of it is made within the model. The
following criteria from Visser (1981) are adequate for ensuring passability and surface
stability:
1 Passability
is a function of the shear strength of the saturated material, and is satisfactory when:
SFCBR ≥ 8.25 + 3.75 * log10 AADT …(5.1)
2 Surfacing stability
relates to ravelling and looseness, is satisfactory when:
P075 ≥ 14 …(5.2)
where:
SFCBR the (minimum) soaked California Bearing Ratio at standard Proctor

laboratory compaction for ensuring passability
All other parameters are as defined previously.

Version 1.0
The deterioration models contain calibration factors to facilitate local calibration. These
factors have default values of 1.0 and are summarised in Table C4.3.
Table C4.3 Calibration Factors used in the Deterioration Models
Calibration factor Deterioration model

Kgl Gravel loss factor
Kkt Traffic-induced material loss factor

Version 1.0
7 References
GEIPOT, (1982)
Research on the Interrelationships between Costs of Highway Construction, Maintenance
and Utilisation (PICR)
Final Report 12 Volumes
Brasilia, Brazil
Hodges J.W., Rolt J., and Jones T.E., (1975)
The Kenya Road Transport Cost Study: Research on Road Deterioration
Report LR 673. Crowthorne, England
Department of the Environment, Transport and Road Research Laboratory
NDLI, (1995).
Modelling Road Deterioration and Maintenance Effects in HDM-4
Final Report Asian Development Bank Project RETA 5549
World Bank Publications, Washington, D.C., USA
Visser A.T., (1981)
An Evaluation of Unpaved Road Performance and Maintenance
Ph.D. thesis. Austin, Texas
University of Texas at Austin, department of Civil Engineering
HDM-III Model
World Bank Publications, Washington, D.C., USA

Version 1.0
Part C Road Works Effects
Contents

Descriptions: Part D
D1 Types of roadworks
1 Introduction D1-1
2 Road works classification D1-3
2.1 Works categories D1-4
2.2 Works classes D1-4
2.3 Works types D1-5
3 Modelling framework D1-9
3.1 Road feature types D1-9
3.2 Works standards D1-9
3.3 Intervention criteria D1-10
3.4 Works design D1-11
3.5 Works duration D1-11
3.6 Unit costs of works D1-14
3.7 Works effects D1-14
4 Costs to road administration D1-15
4.1 Annual costs and budget categories D1-15
4.2 Cost spread over works duration D1-15
4.3 Salvage value D1-16
5 References D1-17
D2 Bituminous Pavements
1 Introduction D2-1
2 Modelling logic D2-2
2.1 Overall computational procedure D2-2
2.2 Hierarchy of works D2-2
2.3 Pavement types reset D2-4
3 Routine maintenance D2-6
3.1 Patching D2-6

Version 1.0
PART C ROAD WORKS EFFECTS CONTENTS
3.2 Crack sealing D2-14

3.3 Edge-repair D2-17
3.4 Drainage works D2-18
3.5 Other routine works D2-19
4 Periodic maintenance D2-20
4.1 Preventive treatment D2-20
4.2 Resealing works D2-23
4.3 Overlay D2-30
4.4 Mill and replace D2-34
4.5 Inlays D2-38
4.6 Reconstruction D2-42
5 Special works D2-46
6 Improvement works D2-47
6.1 Reconstruction with minor widening D2-47
6.2 Widening D2-49
6.3 Realignment D2-60
6.4 Junction improvement D2-70
6.5 Non-carriageway improvement D2-71
7 Construction D2-74
7.1 Upgrading D2-74
7.2 Dualisation D2-77
7.3 New section D2-77
8 References D2-79
D3 Concrete Pavements
1 Introduction D3-1
2.3 Road characteristics reset D3-5
3 Routine maintenance D3-6
4 Preventive treatment D3-7
4.1 Load transfer dowels retrofit D3-7
4.2 Tied concrete shoulders retrofit D3-8
4.3 Longitudinal edge drains retrofit D3-10
4.4 Joint sealing D3-12
5 Restoration works D3-15
5.1 Slab replacement D3-15
5.2 Full depth repairs D3-19

Version 1.0
PART C ROAD WORKS EFFECTS CONTENTS
5.3 Partial depth repairs D3-25

5.4 Diamond grinding D3-27
6 Rehabilitation D3-30
6.1 Bonded concrete overlays D3-30
6.2 Unbonded concrete overlays D3-38
7 Reconstruction D3-40
8 Special works D3-41
9 New construction D3-42
10 References D3-43
D4 Unsealed Roads
1 Introduction D4-1
3 Maintenance works D4-4
3.1 Periodic grading D4-4
3.2 Spot regravelling D4-6
3.3 Gravel resurfacing D4-7
3.4 Routine-miscellaneous maintenance D4-9
4 Improvement works D4-11
4.1 Widening D4-11
4.2 Realignment D4-16
5 Construction works D4-21
5.1 Upgrading D4-21
5.2 Dualisation D4-22
5.3 New section D4-23
6 References D4-24

Version 1.0
Part D Road Map
Analytical
AnalyticalFramework
Framework
and
andModel
Model
Descriptions
Descriptions
Introduction
Introduction
Part
PartAA
Traffic
Traffic
Part
PartBB
RD
RDModel
Model WE
WEModel
Model RUE
RUEModel
Model SEE
SEEModel
Model
Part
PartCC Part
PartDD Part
PartEE Part
PartFF
Economic
EconomicAnalysis
Analysis
PartGG
Nomenclature
Nomenclature
PartHH

Glossary
Part
PartI I
Figure D Analytical Framework and Descriptions Road Map

Version 1.0
PART D ROAD W ORKS EFFECTS
D1 Types of roadworks
1 Introduction
The HDM-4 system includes relationships for modelling Road Deterioration (RD) (see Part C)
and Road Works Effects (WE). In HDM-III these were combined into a single module called
the RDME, Watanatada et al. (1987). In HDM-4, RD and WE have been separated in order to
meet the requirements to trigger the extended number and types of road works using a wider
range of criteria based on threshold values of road user effects parameters.
Roadworks modelling in the context of the HDM-4 system imply the following:
! Defining roadworks in a logical manner to achieve a target standard
! Timing of works over the analysis period
! Calculation of the physical quantities or the amounts of works to be undertaken
! Estimating the costs of works which is part of the cash flow analysis used for economic
analysis and budget preparation
! Resetting/changing one or more of the characteristics that define the road as a result of
implementing the roadworks
Thus, the WE module is used to estimate road agency resource needs for road preservation
and development. These needs are expressed in terms of the physical quantities and the
monetary costs of works to be undertaken. Road agency costs are used together with road user
costs and social and environmental costs to determine the economic viability of different road
investment options and strategies.
This chapter describes the classification of road works and the generic framework for
modelling road works effects in HDM-4. The detailed modelling logic for the different types
of works are described separately for each of the three road surface classes (see Figure D1.1)
as follows:
1 Bituminous pavements (see Chapter D2)
2 Concrete pavements (see Chapter D3)
3 Unsealed roads (see Chapter D4)
Analytical Framework and Model Descriptions D1-1

Version 1.0
PART D ROAD W ORKS EFFECTS D1 TYPES OF ROADWORKS
Road
RoadWorks
WorksEffects
Effects
Types Bituminous Concrete Unsealed

UnsealedRoads
TypesofofWorks
Works Bituminous Concrete Roads
Pavements
Pavements Pavements
Pavements
Chapter
ChapterD1
D1 Chapter
ChapterD4
D4
Chapter
ChapterD2
D2 Chapter
ChapterD3
D3
Figure D1.1 Road Works Effects modules

Version 1.0
2 Road works classification

In HDM-4, road works are considered in a hierarchical structure of category, class, and type.
Each works type comprises several works activities or operations. This classification of road
works is presented in Table D1.1.
Table D1.1 Classification of road works in HDM-4
Works category Works class Works type Works activity /operation

Routine Pavement patching, edge-repair, crack
sealing, spot-regravelling,
shoulders repair, etc.
Routine
Maintenance Drainage culvert repairs, clearing side
drains, etc.
Routine Miscellaneous vegetation control, line-markings,

signs, etc.
Preventive Treatment fog seal, rejuvenation, load

transfer dowel retrofit, joint
sealing, etc.
Resurfacing surface dressing, slurry seal, cape

Periodic seal, regravelling, slab
Maintenance (or Restoration) replacement, diamond grinding,
Preservation etc.
Rehabilitation thick overlay, mill and replace,

inlay, bonded concrete overlay,
unbonded concrete overlay
Reconstruction partial reconstruction, full

pavement reconstruction
Special Emergency clearing debris, repairing

washout/subsidence, traffic
accident removal, etc.
Winter snow removal, salting/gritting,

etc.
Widening partial widening, lane addition
Realignment horizontal and vertical geometric

improvements, junction
improvement
Improvement
Off-carriageway shoulders addition, shoulders
Development upgrading, NMT lane addition,
side drain improvement, etc.
Upgrading upgrading by changing the road

surface class
Construction
New section dualisation of an existing section,
new section (link)

Version 1.0
2.1 Works categories

Road works are divided under two categories:
1 Preservation
Preservation of the existing pavement involves performing maintenance works required

to offset the deterioration of roads, and to lower road user costs by providing a smooth
running surface, and keeping the road open on a continuous basis.
2 Development
Development works aim to expand road network capacity, provide stronger pavement,
and improve road geometric characteristics in order to minimise the total cost of road
transportation and mitigate environmental impacts.
2.2 Works classes

Within each category, road works are considered in classes. Works classes consider road
works in terms of their frequency of application and the budget head used to fund them,
Robinson (1995).
2.2.1 Maintenance
There are three works classes under maintenance:
1 Routine maintenance (see Section 2.3.1)
Comprises works that may need to be undertaken each year.

2 Periodic maintenance (see Section 2.3.2)
Comprises works that are planned to be undertaken at intervals of several years.

3 Special (see Section 2.3.3)
Comprises works whose frequencies cannot be estimated with certainty in advance.
2.2.2 Development
Development works are divided into two classes as follows:
1 Improvement (see Section 2.3.4)
Comprises works that aim to provide additional capacity when a road is nearing the end
of its design life or because there has been an unforeseen change in use of the road. These
include measures of improving quality of service on existing roads such as relieving
traffic congestion, road safety, road passability, etc.
2 Construction (see Section 2.3.5)
Works to create a new pavement.

Each of the works classes described above is divided further into works types as discussed in
Section 2.3.

Version 1.0
2.3 Works types

Works types consider road works in terms of their impact (or effects) on the road
infrastructure. Under each works type, there are several works activities or operations
considered in terms of the pavement type to which they can be applied, and the technique
used. Each works type has a default budget category (recurrent, capital or special). These
budgets are discussed in Section 4.1. Recurrent and capital budgets may also be referred to as
routine and periodic budgets, respectively.
2.3.1 Routine maintenance

Routine maintenance works are divided into the following works types:
! Routine pavement
Works responding to minor pavement defects caused by a combination of traffic and

environmental effects, for example, crack sealing, patching, edge repair; shoulders repair,
non motorised transport (NMT) lane repair, spot-regravelling and grading.
! Drainage
Clearing side drains, clearing culverts and culvert repairs.

! Routine Miscellaneous
Includes all other works that are not modelled endogenously in HDM-4, for example,
vegetation control, line-marking, road signs repair, guard rail repair, etc.
All routine works are considered under the recurrent budget (by default).
2.3.2 Periodic maintenance

Periodic maintenance works are divided into the following works types:
! Preventive treatments
For bituminous pavements this is the addition of a thin film of surfacing to improve
surface integrity and waterproofing that does not increase the strength of the pavement.
This group includes:
❏ fog sealing
❏ rejuvenation
For concrete pavements preventive treatments include:
❏ joint sealing
❏ load transfer dowels retrofit
❏ longitudinal edge drains retrofit
❏ tied concrete shoulders retrofit
Preventive treatments are considered under capital budget (by default).
! Resurfacing or Restoration
Resurfacing of bituminous pavements involves the addition of a thin surfacing to

improve surface integrity and waterproofing, or to improve skid resistance, that does not
increase the strength of the pavement significantly. For bituminous pavements this
includes:

Version 1.0
❏ single surface dressing

❏ double surface dressing
❏ slurry seal
❏ cape seal
Regravelling of unsealed roads is included under resurfacing.
Restoration works on concrete pavements includes:
❏ slab replacement
❏ full depth repair
❏ partial depth repair
❏ diamond grinding
Resurfacing is considered under capital budget (by default).
! Rehabilitation
The addition of thick surfacings, or the removal of part of the existing pavement and the
addition of layers to restore or improve structural integrity and to increase the strength of
the pavement. For bituminous pavements this group includes:
❏ open-graded asphalt overlay
❏ dense-graded asphalt overlay
❏ rubberised asphalt overlay
❏ mill-and-replace
❏ inlays
Rehabilitation of concrete pavements includes:
❏ bonded concrete overlay
❏ unbonded concrete overlay
Rehabilitation is considered under capital budget (by default).
! Reconstruction
The removal of part (both bound and unbound layers) or all of the existing pavement
layers and the construction of a new pavement. This group includes:
❏ strengthening by multiple-layer overlays thicker than 125 mm
❏ granular overlays
❏ recycling of base
❏ membrane-interlayer overlays
❏ reconstruction of concrete pavements
Reconstruction is considered under capital budget (by default).
2.3.3 Special works

Special works are divided into the following works types:

Version 1.0
! Emergency
Comprises works undertaken to clear a road that has been cut or blocked. This group
includes:
❏ traffic accident removal
❏ clearing debris
❏ repairing washout/subsidence
Emergency works is considered under special budget head (by default).
! Winter
Comprises works undertaken to prevent the formation of ice or to remove snow from the
pavement. This group includes:
❏ salting/gritting
❏ snow removal
Winter works is considered under special budget head (by default).
Note: Winter works are not modelled endogenously in this release.
2.3.4 Improvement works

Improvement works are divided into the following works types:
! Full Reconstruction with minor widening
Full pavement reconstruction with the allowance for minor widening has been included
under improvement works since it is unlikely to be funded from the maintenance budget.
Full reconstruction is considered under capital budget (by default).
! Widening
Comprises works that retain the existing pavement, but increase width throughout the
length of the section. This group includes:
❏ lane addition
Widening is considered under capital budget (by default).
! Realignment
Comprises works that changes the road geometry for part of a section, but that retains
some of the existing pavement structure. This group includes:
❏ local geometric improvements
❏ junction improvements
Realignment works is considered under capital budget (by default).
! Off-Carriageway Works
Comprises improvement works that is carried out off-carriageway on road features like
shoulders, NMT lanes, and drainage structures. This group includes:
❏ shoulders addition or upgrading
❏ side drain improvement

Version 1.0
❏ NMT addition or improvement

Off-carriageway improvement works are considered under capital budget (by default).
2.3.5 Construction works

Construction works are divided into the following works types:
! Upgrading
Involves changing the pavement surface class and improving geometric characteristics of
an existing road section. This group includes upgrading:
❏ An unsealed road to bituminous or rigid concrete pavement
❏ A bituminous pavement to rigid concrete pavement
❏ A low grade bituminous pavement to a high grade bituminous pavement
❏ An earth road to a gravel road
Pavement upgrading is considered under capital budget (by default).
! New section
Comprises works to create a new pavement in an entirely new location. This group
includes:
❏ dualisation of existing road sections
❏ new link construction
New section works are considered under capital budget (by default).

Version 1.0
3 Modelling framework
3.1 Road feature types

A road can be considered as a number of complementary features or elements, to each of
which a variety of roadworks activities can be applied.
The different road works activities considered in HDM-4 can be divided into two groups as
follows:
1 Those whose effects are modelled and their costs are used in economic analyses
2 Those that are not modelled but their costs can be considered in economic analyses
Each works activity that is modelled has effects on one or more of the following road feature
types:
! Carriageway
! Shoulders
! NMT lanes
! Intersection or junction
! Drainage structures
The above distinction provides a logical framework that enables a detailed modelling of road
works effects on each of the feature types to be performed separately.
3.2 Works standards

Standards refer to the targets or levels of conditions and response that a road administration
aims to achieve. Road agencies set up different standards that can be applied in practical
situations in order to meet specific objectives which are related to functional characteristics of
the road network system. A standard is defined by a set of operations or works activities with
definite intervention criteria to determine when to carry them out. In general terms,
intervention levels define the minimum level of service that is allowed.
A standard is user-defined according to the road surface class to which it is applied, the
characteristics of traffic on the section, and the general operational practice in the study area
based upon engineering, economic and environmental considerations.
Standards are grouped into two types for input purposes:
1 Maintenance standards
2 Improvement/construction standards
For a given road feature, only one maintenance standard and/or improvement type will be
effective in any analysis year.
The operation type plus the following information define a works standard:
! Intervention criteria (see Section 3.3)
! Works design (see Section 3.4)
! Works duration (see Section 3.5)

Version 1.0
! Unit cost (and optionally energy use) (see Section 3.6)
! Works effects (see Section 3.7)
A standard may therefore comprise several operations each with definite intervention criteria
to determine when to carry them out. The following sections (Sections 3.3 to 3.7) discuss
briefly the information that is required to define a works standard. A detailed discussion of
standards is given in the Applications Guide.
3.3 Intervention criteria

For each works activity, user-specified intervention criteria are used to determine the timing
and limits on the works to be carried out. The intervention criteria can be defined through one
of the following methods:
! By scheduling at fixed intervals of time (for example, resurface at four years intervals,
overlay when pavement age reaches 7 years) or point(s) in time for maintenance works
(not included in this release); and at a fixed time (for example, widen in the year 2005)
for improvement and construction works
! In response to critical threshold levels specified by the user in terms of any of the
following:
❏ Pavement condition
❏ Pavement structure and strength
❏ Drainage condition
❏ Vehicle Speeds
❏ Traffic volumes, loading and flows
Limits
In addition to the above criteria, user-specified limits can be defined in order to simulate
realistic policies, Watanatada et al. (1987). For example, it is recognised that practical
considerations would normally preclude the situation of periodic maintenance being applied
immediately before a major overlay or reconstruction was planned. Also due to the economic
implications, different maintenance standards would be applied to a road section at different
thresholds of traffic volume. Provision is therefore made to specify:
! Minimum and maximum intervals between successive works in order to comply with
budget or other constraints
! Maximum applicable roughness beyond which the works activity must not be done
! Minimum and maximum AADT threshold within which a standard is applicable on the
road section
! Last applicable year for the works
! Maximum annual quantity of the works
A works activity will be triggered when any one or a combination of the user-specified criteria
has been met. When more than one works activity meets the criteria for being applied in a
given analysis year, the works hierarchy established for each road surface class and feature
type is used to prioritise and select the highest placed activity. The details of these modelling
logic are discussed for each road surface class in Chapters D2, D3 and D4.

Version 1.0
3.4 Works design

Before implementing any contemplated road works, it is necessary to carry out a preliminary
or coarse design of the operation. The preliminary design will be refined through technical
and economic analyses of different alternative options, leading to a final detailed design that
can be realised. Works design refers to the explicit specifications of the works activity by the
user in terms of the following:
! Pavement structure
Includes pavement type, strength, layer thickness and material properties.

! Road geometry
Road length, width and number of lanes.

! Road type and road class
Note: No endogenous design or selection from alternative design options is performed.
3.5 Works duration

The duration of a road works determines when its effects should be considered within the
logic of the analysis process. This is as follows:
! For works of duration not greater than one year (see Figure D1.2)
It is assumed that the works are carried out at the end of the analysis year in which it has
been scheduled or triggered, and its effects become effective at the beginning of the
following analysis year. All maintenance works are considered under this category.

Version 1.0
Note:
Road
Cby = Variable at the end of year y
variable
Cb1
Responsive Cbo = Ca1
intervention
Cao
Cb4
Cb2 = Ca3
Ca2
0 1 2 3 4 Years
Scheduled
intervention
Percent
total
costs
100
0 1 2 3 4 Years
Figure D1.2 Modelling of road condition and costs for road works of one year
duration

Version 1.0
! For works of duration greater than one year (see Figure D1.3)
Road characteristics are assumed to remain the same from the end of the first year of
works until the end of the year of works completion. The effects of works become
effective at the beginning of the analysis year following the year of works completion.
Note:
Road Cby = Variable at the end of year y
variable
Responsive
Cb1 = Ca2 Cb2 = Ca3 Cb3
Cbo = Ca1
intervention
Cao
Cb4 = Ca5
Ca4
0 1 2 3 4 Years
Scheduled
intervention
Percent
total
costs
40
30
0 1 2 3 4 Years
Figure D1.3 Modelling of road condition and costs for road works of duration
greater than one year (for example, 3 years)

Version 1.0
3.6 Unit costs of works

The user must specify the unit cost of each operation in economic and/or financial terms,
using one of the following methods as may be appropriate:
! Cost per square metre (for example, overlay)
! Cost per cubic metre (for example, gravel resurfacing)
! Cost per kilometre (for example, upgrading a gravel road to a bituminous surface road)
! Cost per kilometre per year (for example, drainage works)
! Lump sum cost per year (for example, junction improvements)
Unit costs are used to multiply the physical quantities (or amount) of works in order to obtain
the total cost required to implement the operation. The amount of each road works is
computed for each year of the analysis period in which the works activity is applicable.
3.7 Works effects

When a works activity is performed, the immediate effects on road characteristics and road
use need to be specified in terms of the following:
! Pavement strength
! Pavement history
! Road deterioration calibration factors
! Road use patterns
The change in characteristic can be specified in several ways, summarised as:

1 The parameter is set to zero; for example, after an overlay cracking becomes zero.
2 The parameter is reset to an absolute value, which is defined as part of the operation; for
example, the roughness after an overlay is set to 2 m/km IRI.
3 The parameter is reset using a formula which may include other model parameters; for
example, the roughness after an overlay is reset as a function of the previous roughness
and the thickness of the overlay.
4 The parameter is not reset; for example, the carriageway width is unchanged after an
overlay.
Details of the methods used for resetting the parameters that describe the road characteristics
are given in Chapters D2, D3 and D4.
The long-term effects of a works operation are considered through the relevant models, for
example:
! Rate of road deterioration (see Road Deterioration model in Part C)
! Changes in road user costs (see Road User Effects model in Part E)
! Changes in energy use and environmental impacts (see Social and Environmental Effects
model in Part F)
Thus, both the immediate and long-term effects are combined to determine the benefits of
carrying out different sets of roadworks activities at different times over the analysis period.

Version 1.0
4 Costs to road administration
4.1 Annual costs and budget categories

The annual costs to road administration (or road agency costs) incurred in the implementation
of road works will be calculated in economic and/or financial terms depending on the type of
analysis being performed.
The costs of each works activity will be considered under the corresponding user-specified
budget category, or under the default budget category (that is, capital, recurrent, or special)
assigned to that particular operation, as discussed in Section 2.3.
The annual total costs to road administration for each investment option j will be calculated as
follows:
RAC j = CAPj +REC j + SPEC j ...(4.1)
where:
RACj annual total road agency cost incurred under investment option j (currency)
CAPj annual road agency capital (or periodic) cost incurred under investment option
j (currency)
RECj annual road agency recurrent (or routine) cost incurred under investment
option j (currency)
SPECj annual road agency special cost incurred under investment option j (currency)
For each investment option (or alternative), the annual total agency costs will be reported by
works activities and by budget heads or categories.
4.2 Cost spread over works duration

In economic analysis (or cash flow analysis), road agency costs are considered as follows:
! For works of duration not greater than one year
The total agency costs are incurred in the analysis year in which the works is scheduled
or triggered by the specified responsive criteria. It is assumed that all maintenance works
fall under this category (see Figure D1.2).
! For works of duration greater than one year
The total agency costs are broken-down into annual cost streams according to user-
specified percentages. The first percentage of the total costs is incurred in the analysis
year in which the works is scheduled or triggered. Subsequent annual costs are incurred
in each of the following years of works (see Figure D1.3).
The annual total costs to road administration are used in economic analyses of different
investment options (see Part G).

Version 1.0
4.3 Salvage value

The salvage value of construction works undertaken is the value of benefits remaining to be
realised after the end of the analysis period. This may be specified by the user to represent the
percentage of total costs incurred on permanent structures such as embankments, cuttings,
bridges and drainage facilities.
The salvage value for each investment option is calculated as:
S W
SALVA j =
∑∑ SALVA
s =1 w =1
jsw ...(4.2)
where:
SALVAj salvage value of works performed under investment option j (currency)

SALVAjsw salvage value of works w performed on road section s under investment option
j (currency) (see Chapters D2 and D4)
In the economic analysis of investment options, the salvage value SALVAj is considered as a
benefit in the last year of analysis period (see Part G).

Version 1.0
5 References
Classification of Road Works - ISOHDM Working Paper
Communication to ISOHDM Secretariat, University of Birmingham, UK
Riley M., (1995)

Framework for Defining the Effects of Works on Road parameters in HDM-4 -
ISOHDM Working Paper
Robinson R., (1995)

Road Works Classification, RR002\6_35
University of Birmingham, UK
Watanatada T., Harral C.G, Paterson W.D.O., Dhareshwar A.M., Bhandari A., and
World Bank, John Hopkins University Press
The Highway Design and Maintenance Standards Model - Volume 2 User's Manual

Version 1.0
D2 Bituminous Pavements
1 Introduction
This chapter describes the detailed modelling of Road Works Effects for bituminous
pavements (see Figure D2.1).
Road
RoadWorks
WorksEffects
Effects

UnsealedRoads
TypesofofWorks
Pavements
Pavements Pavements
Pavements
Chapter
ChapterD1
D1 Chapter
ChapterD4
D4
Chapter
ChapterD2
D2 Chapter
ChapterD3
D3
The methods of defining works activities and intervention criteria, the calculation of physical
quantities of works and the costs to road administration, and works effects on road
characteristics and road use are discussed for the following works classes:
! Routine maintenance (see Section 3)
! Periodic maintenance (see Section 4)
! Special works (see Section 5)
! Improvement works (see Section 6)
! Construction works (see Section 7)
The modelling logic described comprises the overall computational procedure, the
hierarchical ranking of works activities and pavement type resets after works. The background
of the modelling logic is given in Watanatada et al. (1987).
Analytical Framework of Model Descriptions D2-1

Version 1.0
PART D ROAD W ORKS EFFECTS D2 BITUMINOUS PAVEMENTS
2 Modelling logic
2.1 Overall computational procedure

The overall computational procedure for modelling road works that is applied in each analysis
year can be summarised by the following steps:
! Determine the works standard(s) that is applicable in the given year. Only one
maintenance standard and/or one improvement standard can be applied to a road section
feature in any analysis year.
! Check the intervention criteria and the limits defined for works in the following order:
❏ first improvement works, then
❏ maintenance works
A scheduled operation takes priority over a responsive operation of the same type.
! Apply drainage works (if specified).
! Identify and apply the works activity at the top of the hierarchy.
! Compute the physical quantities of works.
! Compute works effects and reset modelling parameter values to reflect post-works road
geometry, pavement structure, strength, condition, history, and road use.
! Apply any other works whose effects on pavement performance are not modelled
endogenously, (for example, routine-miscellaneous works).
! Calculate the costs of works by applying unit costs to the physical quantities of works.
! Store results for economic analysis and for use in the following analysis year.
2.2 Hierarchy of works

A works activity (or an operation) is triggered when any one or a combination of the user-
specified criteria has been met. When more than one works activity meets the criteria for
being applied in a given analysis year, the highest placed operation for the particular road
feature is selected.
Table D2.1 shows the hierarchy of works activities that are applicable to the carriageway. The
operation dualisation of an existing road section is placed at the top of the list as number 1,
and takes priority over all the other operations, while routine pavement works (that is,
patching, edge-repair, and crack sealing) placed at the bottom are given the lowest priority.

Version 1.0
Table D2.1 Hierarchy of road works applicable to the carriageway
Works type Works activity / operation Hierarchy Unit cost


Widening
Inlay 11 per m2
Cape seal 14 per m2

Routine
Patching1 22 per m2
Pavement
Note:
1
Routine pavement works (that is, crack sealing, patching, edge-repair) have the same
ranking, and all of them can be performed in the same analysis year.
An improvement, or construction works, of a fixed specification is applied to a given road
section only once during the analysis period. This rule applies particularly to improvement
works that have been defined as responsive to the levels of user-specified intervention criteria
based on road user effects parameters.
Routine pavement works, defined by the user, can be applied as separate operations in each
year, or used to repair some distresses before applying the higher-priority works (for example,
preventive treatment, resealing, or overlays). In the former case, routine pavement works are
performed every year in which no periodic maintenance works are applied. In the latter case,

Version 1.0
routine pavement works are considered to be an integral part of the periodic maintenance
works, and are referred to as preparatory works. Although preparatory works are
automatically triggered and performed together with the periodic maintenance works, the
amount and cost of each of the operations involved are modelled and reported separately.
Drainage works are applied in any given analysis year, if specified by the user, regardless of
the hierarchy for carriageway works activities given in Table D2.1. Improvement of side
drains takes priority over routine drainage maintenance should both works be applicable in an
analysis year.
The operations that apply to shoulders and non-motorised transport (NMT) lanes are also
performed in any analysis year, if specified by the user, regardless of the works hierarchy
described above. Shoulder or NMT lane improvement works takes priority over shoulders
repair or NMT lane repair, respectively.
For all road feature types, if more than one works activity of the same operation type (for
example, different specifications of overlay) are applicable in an analysis year, the one with
the highest cost takes priority over the others.
The works activities whose effects on pavement performance are not modelled endogenously
(for example, emergency works, winter maintenance, and routine- miscellaneous works) are
applied in a given analysis year, if specified by the user, regardless of any works hierarchy.
2.3 Pavement types reset

Maintenance works reset pavement types in accordance with pavement classification (see
Table D2.2 and Table D2.3).
Table D2.2 Pavement type resets after maintenance works

activity


Treatment

STSB1

AMSB1
to intermediate
surface layer
to base
Source: NDLI (1995)

Version 1.0
Notes:
bituminous surfacing that is user-definable in HDM Configuration
operation) and the surface material
N/A Not applicable
Table D2.3 Maintenance effects on resultant surface material
Maintenance works Resultant surface material options

All resealings SBSD, DBSD, CAPE, SL, PM
All overlays, including thin overlay AC, HRA, PMA, RAC, CM, SMA, PA
Mill and replace SBSD, DBSD, CAPE, SL, PM, AC, HRA, PMA, RAC,
CM, SMA, PA
Notes: For the definition and full description of the abbreviated surface materials refer to
Chapter C2.

Version 1.0
3 Routine maintenance
The routine maintenance works on bituminous roads whose effects on pavement performance
are modelled, comprises the following operations:
! Patching (see Section 3.1)
! Crack sealing (see Section 3.1.6)
! Edge-repair (see Section 3.3)
! Drainage works (see Section 3.4)
Other routine maintenance works (for example, vegetation control, repairs to road
appurtenances, etc.) are considered in the analyses only in terms of their costs to the road
administration.
3.1 Patching
Patching is used to repair the following surface distresses:
! Potholing
! Wide structural cracking
! Ravelling
The user may specify patching to repair the individual surface distresses, or a combination of
all the three distresses above, which is referred to as the severely damaged area (ADAMS). If
more than one kind of patching works is applicable in any analysis year then patching works
specified to treat ADAMS will override those specified to treat the individual distresses.
3.1.1 Severely damaged area

The severely damaged area is defined as the sum of the areas with wide structural cracking,
ravelling and potholing; and this can be treated by patching specified in one of the following
ways:
! Option 1: Scheduled
The user may specify a fixed percentage of the severely damaged area (Pdam) to be
patched and a maximum annual quantity of patching (in square metres per kilometre).
The fixed amount might, for example, reflect the maximum resources available from the
road authority when averaged over all roads within the road class. The amount of
patching performed is computed as the lesser of the specified maximum amount and the
severely damaged area to be repaired.
! Option 2: Condition - Responsive
The user may specify the percentage of the severely damaged area (Pdam) which is to be
patched in each year and impose a limit of the maximum annual quantity of patching.
The intervention criteria for this case will be the level of ADAMS.
The total area patched is given by the following expression:
APAT = MIN[APATlim , ASP ] ...(3.1)
ASP = Pdam * ADAMS bw * CW * 10 -1 ...(3.2)

Version 1.0
where:
APAT total patched area (m2/km)

APATlim maximum annual quantity of patching, input by the user (m2/km)
ADAMSbw severely damaged area before patching works (= ADAMSb) (% of total
carriageway area)
ADAMSb severely damaged area at the end of the year (% of total carriageway area)
Pdam percentage of the severely damaged area to be patched, input by the user
(default = 100)
When patching is performed, the unpatched damaged area is reduced by the amount of
patching as described by the pseudo code given in Equation 3.3 below. It is assumed that
potholing, wide structural cracking, and ravelling have priorities in that order, and no
patching is performed to fix these individual distress-areas until those of higher priorities are
completely repaired.
The percentage repaired of each distress is calculated as follows:

Version 1.0
START
If APAT ≥ 10*APOTbw*CW
Ppt = 100
LIMCW = APAT - 10*APOTbw*CW
If LIMCW ≥ 10*ACW bw*CW

Pcw = 100
LIMRV = LIMCW - 10*ACW bw*CW
If LIMRV ≥ 10*ARVbw*CW
Prv = 100
Else
10 * LIMRV
Pr v = (if ARVbw = 0, set Prv = 0)
ARVbw * CW
End if
Else
10 * LIMCW
Pcw = (if ACW bw = 0, set Pcw = 0)
ACWbw * CW
Prv = 0
End if
Else
10 * APAT
Ppt =
APOTbw * CW
Pcw = 0
Prv = 0
End if
END
...(3.3)
where:
APOTbw area of potholing before patching works (= APOTb) (% of total carriageway

area)
APOTb area of potholing at the end of the year (% of total carriageway area)
ACWbw area of wide structural cracking before patching works (= ACWb) (% of total
carriageway area)
ACWb area of wide structural cracking at the end of the year (% of total carriageway
area)

Version 1.0
ARVbw area of ravelling before patching works (= ARVb) (% of total carriageway

area)
ARVb area of ravelling at the end of the year (% of total carriageway area)
Ppt percentage of potholed area to be patched (0 - 100%)
Pcw percentage of the area with wide structural cracking to be patched (0 - 100%)
Prv percentage of ravelled area to be patched (0 - 100%)
3.1.2 Potholing only

Patching may be specified to repair only potholing in one of the following ways:
The user may specify a fixed percentage of potholing area (Ppt) to be repaired and
impose a limit of the maximum annual quantity of patching.
The user may specify the percentage of potholing area to be patched (Ppt) and impose a
limit of the maximum annual quantity of patching. The intervention criteria to use in this
case will be the level of potholing only.
For both options the total area patched is given by Equation 3.1 above, and the parameter ASP
is calculated as follows:
ASP = Ppt * APOTbw * CW * 10 -1 ...(3.4)
where:
Ppt percentage of potholing area to be patched, input by the user (default = 100%)

When patching is performed, the area of potholing is reduced by the amount of patching. The
areas of wide structural cracking and of ravelling is not changed by the patching (that is,
Pcw = 0 and Prv = 0).
The percentage repaired of potholing is calculated as follows:
If ASP > APATlim
then:
 10 * APATlim 
Ppt =   ...(3.5)
 APOTbw * CW 
otherwise:
Ppt equals the user-specified value (or the default value = 100%)

Version 1.0
3.1.3 Wide structural cracking only

Patching may be specified to treat only wide structural cracking in one of the following ways:
The user may specify a fixed percentage of wide structural cracking area (Pcw) to be
patched and impose a limit of the maximum annual quantity of patching.
Under this option, the user may specify the percentage of wide structural cracking area to
be patched (Pcw) and impose a limit of the maximum annual quantity of patching. The
intervention criteria to use in this case will be the level of wide structural cracking only.
For both options; the total area patched is given by Equation 3.1 above, and the parameter
ASP is calculated as follows:
ASP = Pcw * ACWbw * CW * 10 -1 ...(3.6)
where:
Pcw percentage of wide structural cracking area to be patched, input by the user
(default = 100%)
All the other variables are as defined previously.

When patching is performed, the area of wide structural cracking is reduced by the amount of
patching. The areas of potholing and of ravelling will not be changed by the patching, (that is,
Ppt = 0 and Prv = 0).
The percentage repaired of wide structural cracking is calculated as follows:
If ASP > APATlim, then
Pcw =   ...(3.7)
 ACWbw * CW 
otherwise:
Pcw equals the user-specified value (or the default value = 100%)
3.1.4 Ravelling only

Patching may be specified to treat only ravelled areas in one of the following ways:
The user may specify a fixed percentage of ravelled area (Prv) to be patched and impose
a limit of the maximum annual quantity of patching.
Under this option, the user may specify the percentage of ravelled area to be patched
(Prv) and impose a limit of the maximum annual quantity of patching. The intervention
criteria to use in this case will be the level of ravelling only.

Version 1.0
For both options, the total area patched is given by Equation 3.1 above and the parameter ASP
ASP = Prv * ARVbw * CW * 10 -1 ...(3.8)
where:
Prv percentage of ravelled area to be patched, input by the user (default = 100%)
All the other variables are as defined previously.

When patching is performed, the area of ravelling is reduced by the amount of patching. The
areas of potholing and of wide structural cracking will not be changed by the patching, (that
is, Ppt = 0 and Pcw = 0).
The percentage repaired of ravelling is calculated as follows:
If ASP > APATlim, then:
Prv =   ...(3.9)
 ARVbw * CW 
otherwise:
Prv equals the user-specified value (or the default value = 100%)
3.1.5 Total area and cost of patching

The total area patched (TAPAT), in square metres, is given by the product of APAT and the
section length (L), in kilometres. The total cost of patching performed on the entire road
section is obtained from the product of TAPAT and the user-specified unit cost per square
metre.
3.1.6 Effects of patching

In all the options for defining patching works, the effects of patching on pavement condition is
computed as follows:
Potholing
NPTaw = NPTbw * 1 -

Ppt 
 ...(3.10)
 100
where:
NPTaw number of potholes per km after patching works

NPTbw number of potholes per km before patching works (= NPTb)
NPTb number of potholes per km at the end of the year
The change in the number of potholes per km due to works (∆NPTw) is given by:

Version 1.0
∆NPTw = NPTbw - NPTaw ...(3.11)
Cracking
ACWaw = ACWbw * 1 -

Pcw 
...(3.12)
 100 
∆ACW w = ACWbw * ACWaw ...(3.13)
ACA aw = ACA bw * ∆ACW w ...(3.14)
ACX aw = 0.62 * ACA aw + 0.39 * ACWaw ...(3.15)
ACRA aw = ACA aw + ACTbw ...(3.16)
where:
ACWaw area of wide structural cracking after patching works (% of total carriageway
area)
ACAaw area of all structural cracking after patching works (% of total carriageway
area)
ACAbw area of all structural cracking before patching works (= ACAb) (% of total
carriageway area)
ACAb area of all structural cracking at the end of the year (% of total carriageway
area)
ACXaw area of indexed cracking after patching works (% of total carriageway area)
ACRAaw total area of cracking after patching works (% of total carriageway area)
ACTbw area of transverse thermal cracking before patching works (= ACTb) (% of
ACTb area of transverse thermal cracking at the end of the year (% of total
carriageway area)
∆ACWw reduction in area of wide structural cracking due to patching works
(= ∆ACWpat) (% of total carriageway area)
Ravelling
ARVaw = ACVbw * 1 -

Prv 
 ...(3.17)
 100
where:
ARVaw area of ravelling after patching works (% of total carriageway area)
Roughness
Roughness after patching is calculated as follows:

Version 1.0
  APAT  
RI aw = RIbw - MINa0 * ∆CRX w + ∆RI t - a1 * MIN ,10 , (a2 - RIbw )
 (10 * CW ) 
   
...(3.18)
and:
∆RIpat = RIbw - RI aw ...(3.19)
where:
RIaw roughness after patching works (IRI m/km)

RIbw roughness before patching works (= RIb) (IRI m/km)
RIb roughness at the end of the year (IRI m/km)
APAT total patched area (m2/km) (for computing increase in roughness due to
depression or protrusion of patches)
∆CRXw reduction in the area of indexed structural cracking due to patching works (=
ACXbw - ACXaw)
ACXbw area of indexed structural cracking before patching works (= ACXb) (% of
ACXb area of indexed structural cracking at the end of the year (% of total
carriageway area)
∆RIt reduction in roughness due to pothole patching (IRI m/km)
∆RIpat reduction in roughness due to patching works

a0 to a2 user-definable model coefficients (default = 0.0066, 0.01 and 16.0
respectively)
The model coefficient a1 in Equation 3.18 above represents an average depression or

protrusion of about 2 mm for skin patches. When the standard of workmanship observed in
maintenance patching differs significantly from this, the coefficient 0.01 should be adjusted
by the ratio of the observed average depression/protrusion to 2 mm.
The reduction in roughness is a function of the freedom to manoeuvre variable (FM), ranging
between 0 and 1, which is predicted using Equation 3.20 below:
FM = (MAX{MIN[0.25 * (CW - 3 ),1],0}) * MAX 1 −

AADT   
, 0  ...(3.20)
  5000   
The reduction in roughness is calculated as follows:

! if nil patching (TLF = 1) or 100% patching policy options is specified, then
 ∆NPTw * TLF 
a2 
∆RI t = a 0 (a 1 − FM)  NPTa * TLF +
  − NPTa 
a2
...(3.21)
 2  
 

Version 1.0
! otherwise (for partial patching policy options)
a2
 ∆NPTw 
∆RI t = a 0 (a 1 − FM) * ∆NPTw *  NPTa +  ...(3.22)
 2 
where:
∆NPTw reduction in number of potholes per km due to patching

FM freedom to manoeuvre
NPTa number of potholes per km at the start of the analysis year
AADT annual average daily traffic, two-way flow (veh/day)
a0 to a2 user-definable model coefficients (default = 0.00017, 2.0, and 1.5,
respectively)
3.2 Crack sealing

Crack sealing treats transverse thermal cracking and wide structural cracking. However, it is
assumed that crack sealing is not applied to treat wide structural cracking if the area of wide
structural cracking exceeds 20 % (that is, ACWb>20).
The user may specify crack sealing in one of the following manners:
Crack sealing may be scheduled by specifying a fixed percentage of transverse thermal

cracking and/or a fixed percentage of wide structural cracking, and a maximum annual
quantity in square metres per kilometre.
By specifying the percentage of the area with transverse thermal cracking (Pcrt) and/or
the percentage of the area with wide structural cracking (Pcrw) to be sealed in each year,
and imposing a limit of the maximum area. The intervention criteria to use in this case
will be the level of transverse thermal cracking and/or the level of wide structural
cracking.
The carriageway area sealed is computed as follows:
ACSL = MIN (ACSL lim , ASEAL ) ...(3.23)
ASEAL = [(Pcrt * ACTbw ) + (Pcrw * ACWbw )] * CW * 10 -1 ...(3.24)
where:
ACSL area of crack sealing (m2/km)

ACSLlim maximum annual quantity of crack sealing, input by the user (m2/km)
ACTbw area of transverse thermal cracking before crack sealing (= ACTb) (% of total
carriageway area)

Version 1.0
ACWbw area of wide structural cracking before crack sealing (% of total carriageway
area)
Pcrt percentage of transverse thermal cracking area to be sealed, input by the user
(%)
Pcrw percentage of wide structural cracking area to be sealed, input by the user (%)
The values of ACWbw, ACAbw, and RIbw to be used in Equations 3.24 above, 3.30 below, 3.31
below and 3.34 below are obtained as follows:
if both crack sealing and patching are specified to be performed in a given analysis year, it
is assumed that patching takes priority over crack sealing in reducing the area of wide
structural cracking, thus:
ACWbw = ACWb - ∆ACWpat ...(3.25)
ACA bw = ACA b - ∆ACWpat ...(3.26)
RIbw = RIb - ∆RIpat ...(3.27)
where:
∆ACWpat reduction in area of wide structural cracking due to patching works (% of total
carriageway area). This is calculated using Equation 3.13 above, with the
subscript w replaced with pat
ACAbw adjusted area of all structural cracking before crack sealing (after patching) (%
of total carriageway area)
RIbw adjusted roughness before crack sealing works (after patching) (IRI m/km)
∆RIpat roughness after patching works (IRI m/km). This is calculated using Equation
3.19 above
otherwise:
ACWbw = ACWb
ACA bw = ACA b
RIbw = RIb
The total area of crack sealing (TACSL), in square metres, is given by the product of ACSL
and the section length (L), in kilometres. The total cost of crack sealing performed on the
entire road section is computed by multiplying TACSL by the user-specified unit cost per
square metre.
3.2.1 Effects of crack sealing - option 1

The effects of crack sealing on pavement condition are reset as follows:

Version 1.0
Cracking
When crack sealing is performed, it is assumed that the treatment of transverse thermal
cracking takes priority over that of wide structural cracking, and no crack sealing is performed
to fix wide structural cracking until transverse thermal cracking is completely repaired.
The areas of cracking are reduced by the amount of sealing as follows:
 ACSL 
ACTaw = ACTbw - MIN Pcrt * ACTbw * 10 - 2 , ...(3.28)
 (10 * CW ) 
  ACSL  
∆ACW w = MAX0,   − [ACTbw − ACTaw ] ...(3.29)
  (10 * CW )  
ACWaw = ACWbw - ∆ACW w ...(3.30)
ACA aw = ACA bw - ∆ACW w ...(3.31)
ACX aw = 0.62 * ACA aw + 0.39 * ACWaw ...(3.32)
ACRA aw = ACA aw + ACTaw ...(3.33)
where:
ACTaw area of transverse thermal cracking after crack sealing works (% of total
carriageway area)
ACTbw area of transverse thermal cracking before crack sealing works (= ACTb) (% of
∆ACWw reduction in area of wide structural cracking due to crack sealing (% of total
carriageway area)
ACWaw area of wide structural cracking after crack sealing works (% of total
carriageway area)
ACWbw adjusted area of wide structural cracking before crack sealing (% of total
carriageway area)
ACAbw adjusted area of all structural cracking before crack sealing (% of total
carriageway area)
ACAaw area of all structural cracking after crack sealing works (% of total carriageway
area)
ACXaw area of indexed cracking after crack sealing works (% of total carriageway
area)
ACRAaw total area of cracking after crack sealing works (% of total carriageway area)
Roughness
Crack sealing has some little effect on roughness, which is computed as follows:
RI aw = RIbw - a0 * ∆ACRA ...(3.34)

Version 1.0
∆ACRA w = ACRA bw - ACRA aw ...(3.35)
where:
RIaw roughness after crack sealing works (IRI m/km)

RIbw adjusted roughness before crack sealing works (IRI m/km)
∆ACRAw reduction in the total area of cracking due to crack sealing (% of total
carriageway area)
a0 model coefficient (default = 0.0066)
3.3 Edge-repair
The user may specify edge-repair in one of the following manners:
Edge-repair may be scheduled by specifying a fixed percentage area of edge-break to be

repaired and a maximum annual quantity, in square metres per kilometre. The amount
performed is computed as the lesser of the specified maximum annual amount and the
area of edge-break to be repaired.
By specifying the percentage area of edge-break to be repaired in each year (Pver) and
imposing a limit of the maximum annual quantity of edge-repair. The intervention criteria
to use will be the level of edge-break.
The area of edge-repair is computed as follows:
{ [
AVERP = MIN AVERPlim , Pver * AVEB bw * CW * 10 -1 ]} ...(3.36)
where:
AVERP area of edge-repair (m2/km)

AVERPlim maximum annual quantity of edge-repair, input by the user (m2/km)
AVEBbw total area of edge-break before edge-repair works (= AVEBb) (% of total
carriageway area)
AVEBb total area of edge-break at the end of the year (% of total carriageway area)
Pver percentage of edge-break area to be repaired, input by the user (%), default =
100
The product of AVERP and the section length (L), in kilometres, gives the total area of edge-
repair (TAVER) in square metres. The total cost of edge-repair performed on the entire road
section is computed by multiplying TAVER by the user-specified unit cost per square metre.
3.3.1 Effects of edge-repair

When edge-repair is performed, the area of edge-break is reduced by the amount repaired as
follows:

Version 1.0
 AVERP 
AVEB aw =  AVEB bw - ...(3.37)
 (CW * 10 ) 
where:
AVEBaw area of edge-break after edge-repair works (% of total carriageway area)
3.4 Drainage works

Drainage maintenance is an important works activity that prevents accelerated pavement
deterioration. Drainage works is modelled through its effects on pavement strength.
The user may specify drainage works in one of the following ways:
Drainage works should be scheduled at a fixed interval of year(s) between successive

treatments (for example, 1 year), commencing from the start year of the maintenance
standard under which it is included.
Drainage works is performed when the level of drain deterioration measured by drainage
factor (DF) exceeds the user-specified intervention criteria (see Chapter C2).
The unit costs of drainage works is specified in terms of currency per km per year. The annual
costs are obtained by multiplying the length of the road by the unit cost.
3.4.1 Effects of drainage works

When drainage works is performed, the drainage factor after works (DFaw) is reset as follows:
DFaw = MAX [DFdmin , (DFbw - ∆DFw )] ...(3.38)
and:
∆DFw = (DFdmax - DFdmin ) * DMCF ...(3.39)
where:
DFaw drainage factor after maintenance works

DFbw drainage factor before maintenance works
DFdmax maximum drainage factor, denoting very poor drainage condition for drain
type d
DFdmin minimum drainage factor, denoting excellent drainage condition for drain type
d
∆DFw change in DF due to the drainage works performed
DMCF drainage maintenance cost factor, defined as the ratio of the annual cost of
drainage works performed to the annual cost required to maintain the drainage
system in excellent condition

Version 1.0
The drainage factor after works is used to compute the adjusted structural number of
pavement (SNP).
3.5 Other routine works

The effects of the following routine works on pavement performance are not modelled
endogenously, and therefore only their costs are considered in an analysis:
! Shoulders repair
! NMT lanes repair
! Routine miscellaneous works
For example, vegetation control, road sign repairs and replacement, line marking, guard
rail repair, etc.
These works activities are scheduled at a fixed interval of time (a minimum of one year), and
are performed on an annual basis. If specified by the user, the works activities are applied in a
given analysis year regardless of the works hierarchy.
The unit cost should be specified in terms of currency per km per year, and the annual cost of
the operation is obtained from the product of the road section length and the unit cost.

Version 1.0
4 Periodic maintenance
The periodic maintenance works on bituminous roads comprises the following:
! Preventive treatment (see Section 4.1)
! Resealing (see Section 4.2)
! Overlay (see Section 4.3)
! Mill and replace (see Section 4.4)
! Inlays (see Section 4.5)
! Reconstruction (see Section 4.6)
4.1 Preventive treatment

Preventive treatments include fog sealing and rejuvenation. These can be defined in one of the
following ways:
A fixed interval between successive treatments is specified (for example, three years),
and treatment is applied whenever the surfacing preventive treatment age (AGE1)
exceeds this interval.
Preventive treatment is applied at the first signs of cracking or ravelling distress and is
constrained by the user-specified limits of the minimum and maximum allowable
preventive treatment intervals, in years. Preventive treatment is not applied if:
ACRAb ≥ 5,
ARVb ≥ 5, or
NPTa > 0
even if the maximum allowable interval has been exceeded (as may occur in the first
analysis year of an old pavement).
4.1.1 Quantities and costs

If performed, the amount of preventive treatment is given by:
APVT = 1000 * CW ...(4.1)
where:
APVT area of road treated (m2/km)

ACRAb total area of cracking at the end of the year (% of total carriageway area)
ARVb area of ravelling at the end of the year (% of total carriageway area)

Version 1.0
The total area treated (TAPVT), in square metres, is given by the product of APVT and the
section length (L), in kilometres. The total cost of preventive treatment performed is
computed by multiplying TAPVT by the user-specified unit cost per square metre
4.1.2 Effects of preventive treatment

The effects of preventive treatment are reset as follows:
! Surface distresses
When preventive treatment is performed, any surface distress (which is minimal) is reset
to zero.
! Pavement strength
The pavement structure and strength remains unaltered.

! Surfacing age
Preventive treatment age (AGE1) is reset to zero.

! Cracking and ravelling retardation
Preventive treatment has the effects of delaying the initiation of cracking and ravelling.
These effects are considered in the models through changes in the cracking retardation
time (CRT) and ravelling retardation factor (RRF) as described below:
❏ The cracking retardation time after preventive treatment is reset as follows:
CRTaw = MIN CRTbw +

CRM CRTMAX 
, , 8 ...(4.2)
 YXK YXK 
YXK = MAX (0.1, YAX ) ...(4.3)
where:
CRTaw cracking retardation time after preventive treatment (years)

CRTbw cracking retardation time before preventive treatment (years)
CRM change in cracking retardation time due to preventive treatment
CRTMAX maximum limit on the value of cracking retardation time
YAX annual number of axles of all vehicle types (millions per lane)
❏ The ravelling retardation factor after preventive treatment is reset as follows:

For surface type AM:
RRFaw = 1 ...(4.4)
For surface type ST:
RRFaw = MIN[RRFbw * RRM, RRFMAX] ...(4.5)
where:

Version 1.0
RRFaw ravelling retardation factor after preventive treatment

RRFbw ravelling retardation factor before preventive treatment
RRM change in ravelling retardation factor due to preventive treatment
RRFMAX maximum limit on the value of ravelling retardation factor
The default values for CRM, CRTMAX, RRM and RRFMAX for rejuvenation and fog
sealing are given in Table D2.4 and Table D2.5, respectively.
Table D2.4 Model parameters for rejuvenation
Pavement Surface HSOLD CRM CRTMAX RRM RRFMAX

type material value
All 0 1.5 3.0 1.15 2.0
CM >0 0.75 1.5 1.15 2.0
AMAB All 1.5 3.0 1.15 2.0
AMAP All 1.5 3.0 1.15 2.0
AMSB All 1.5 3.0 1.15 2.0
All 0 3.0 6.0 1.15 2.0

STGB
>0 1.5 3.0 1.15 2.0
STAB All 0 1.5 3.0 1.15 2.0
STAP All 1.5 3.0 1.15 2.0
STSB All 1.5 3.0 1.15 2.0

Version 1.0
Table D2.5 Model parameters for fog sealing
Pavement Surface HSOLD CRM CRTMAX RRM RRFMAX

type material value
All 0 0.8 1.6 1.3 3.0
CM >0 0.4 0.8 1.3 3.0
AMAB All 0.8 1.6 1.3 3.0
AMAP All 0.8 1.6 1.3 3.0
AMSB All 0.8 1.6 1.3 3.0
All 0 1.6 3.2 1.3 3.0

STGB
>0 0.8 1.6 1.3 3.0
STAB All 0 0.8 1.6 1.3 3.0
STAP All 0.8 1.6 1.3 3.0
STSB All 0.8 1.6 1.3 3.0
4.2 Resealing works

Resealing without shape correction can repair surface distress but cause little change to
roughness or structural strength of the pavement. However, resealing with shape correction
can achieve some reduction in roughness through the filling of depressions and repair of
damaged areas. The corrective material is assumed to be bituminous, with an average
thickness of less than 50 mm and placed to a quality of less than that of automatic-levelling
paver-finishers, Watanatada et al. (1987).
Resealing works can be defined as described in options 1 and 2:
A fixed interval between successive reseals is specified, and the resealing is applied
whenever the surfacing age (AGE2) exceeds this interval.
Resealing is applied when the levels of pavement defect or roughness exceeds the user-
specified values.
For both options (1 and 2), resealing is not performed if the surfacing age (AGE2) is less than
the user-specified minimum applicable interval. However, resealing is always performed if
AGE2 exceeds the user-specified maximum allowable interval between reseals. In all cases,
resealing is not performed if the user-specified last applicable year or maximum applicable
roughness has been exceeded.
A resealing works is specified using the following:
! New surfacing thickness
! Layer strength coefficient
! Surface material, and

Version 1.0
! Construction defect indicator for bituminous surfacing (CDS).

If performed, the amount of resealing is calculated as follows:
ARSL = 1000 * CW ...(4.6)
where:
ARSL carriageway area resealed (m2/km)
Preparatory works
If the area of wide structural cracking is larger than 20%, or the areas of potholes, transverse
thermal cracking or of edge-break are not zero at the end of the year, preparatory works of
the following amount is assumed to be carried out along with resealing:
! Patching
APAT = 10 * CW * {MAX [0.1 * (ACWb - 20 ) , 0] + APOTb } ...(4.7)
where:
APAT area of preparatory patching (m2/km)

ACWb area of wide structural cracking at the end of the year (% of total
carriageway area)
APOTb area of potholes at the end of the year (% of total carriageway area)
! Crack Sealing
ACSL = 10 * CW * ACTb ...(4.8)
where:
ACSL area of crack sealing (m2/km)

ACTb area of transverse thermal cracking at the end of the year (% of total
carriageway area)
! Edge-repair
AVERP = 10 * CW * AVEB b ...(4.9)
where:
AVERP area of edge-repair (m2/km)

AVEBb area of edge-break at the end of the year (% of total carriageway area)

Version 1.0
The total area resealed (TARSL), in square metres, is obtained from the product of ARSL and
the section length (L), in kilometres. The total cost of resealing is obtained by multiplying
TARSL by the user-specified unit cost per square metre. The additional areas and costs of
preparatory patching, crack sealing and of edge-repair are reported separately under patching,
crack sealing and edge-repair, respectively.
4.2.2 Effects of resealing

Resealing resets the pavement structure as described below:
! Pavement type
Pavement type after resealing works is reset as defined in Table D2.2 and Table D2.3
above.
! Surfacing layer thickness
The total surface thickness after works is given by:
HS aw = HS bw + HSNEWaw ...(4.10)
where:
HSaw total surface thickness after resealing (mm)

HSbw total surface thickness before resealing (mm)
HSNEWaw user-specified thickness of reseal (mm)
The thickness of the old, underlying bituminous surfacing after works is given by:
HSOLD aw = HSNEWbw + HSOLD bw ...(4.11)
where:
HSOLDaw thickness of old surfacing after works (mm)

HSOLDbw total thickness of previous, underlying surfacing layers before works
(mm)
HSNEWbw thickness of the most recent surfacing before works (mm)
! Pavement strength
To take account of the net strengthening of the pavement due to both maintenance and
cracking, the pavement strength parameters are updated through the following steps:
❏ Calculate the dry season SNP after works as follows:
SNPdaw = MAX[1.5, (SNPdbw + 0.0394 * a sw * HSNEWaw - dSNPK )] ...(4.12)
where:
SNPdaw dry season adjusted structural number of pavement after resealing

works

Version 1.0
SNPdbw dry season adjusted structural number of pavement before resealing

works
dSNPK reduction in the adjusted structural number of pavement due to
cracking (existing before resealing)
asw layer strength coefficient of the reseal (see Chapter C2 - Table D2.7)
❏ Calculate the parameter f, defined as the ratio of wet season SNP to dry season
SNP using DFaw and ACRAaw as detailed in Chapter C2
❏ Calculate the parameter fs as specified for the different Road Deterioration models,
see Chapter C2
❏ Calculate the annual average adjusted structural number of pavement (SNP)
The Benkelman beam deflection after works is given by:
−1.6
 SNPaw 
DEFaw = DEFbw *   ...(4.13)
 SNPbw 
where:
SNPaw adjusted structural number of pavement after works

SNPbw adjusted structural number of pavement before works
DEFaw Benkelman beam deflection after works (mm)
DEFbw Benkelman beam deflection before works (mm)
! Construction defects indicators
The construction defect indicator for bituminous surfacing (CDS) is reset to a user
specified value. If this is not specified, a good construction quality is assumed, with the
CDS value set to 1.0.
! Pavement surface distresses
Resealing works resets surface distresses to zero, and thereafter the pavement condition
is considered to be new.
! Rutting
The following resealing operations has no effect on rutting:

❏ single surface dressing without shape correction
❏ double surface dressing without shape correction
❏ cape seal without shape correction
❏ slurry seal
The effect of resealing works with shape correction on rutting is user-specified. If these
are not specified, the mean rut depth is calculated for each works activity from the
following expression, as given by Riley (1995):
RDM aw = a0 * RDMbw ...(4.14)

Version 1.0
where:
RDMaw mean rut depth after works (mm)

RDMbw mean rut depth before works (= RDMb) (mm)
a0 user-definable coefficient (default = 0.15)
! Roughness
The effects of resealing works on roughness are user-specified. If these are not specified,
the roughness is calculated for each works activity in the following manner:
First, the roughness value at the end of the year is adjusted to take account of
preparatory works as follows:
RIap = RIb - MIN {[a0 * (∆CRX p + ∆ACTp ) + ∆RI t ], a1} ...(4.15)
where:
RIap adjusted roughness after preparatory patching (IRI m/km)

∆CRXp reduction in the area of indexed cracking due to preparatory works
(=ACXb - ACXap)
∆ACTp reduction in the area of transverse thermal cracking due to preparatory
works (= ACTb)
a0 and a1 user-definable model coefficients (default = 0.0066 and 4.6 respectively)
The value of ACXap is obtained as follows:
ACXap = 0.62 * ACA ap + 0.39 * ACWap ...(4.16)
ACA ap = ACA b - ∆ACWp ...(4.17)
ACWap = ACWb - ∆ACWp ...(4.18)
∆ACWap = MAX [0.1 * (ACWb - 20 ), 0] ...(4.19)
The value of ∆RIt is computed using Equations 3.20 above to 3.22 above with ∆NPTw
taken as equal to NPTbw (= NPTb), since all existing potholes should be patched.
The adjusted roughness after preparatory works (RIap) is then used to compute the final
roughness after resealing works as follows (NDLI, 1995):
❏ Single and double bituminous surface dressing without shape correction
RI aw = RIap - MAX {0, MIN [a0 * (RIap - a1), a2]} ...(4.20)
where:

Version 1.0
RIaw roughness after resealing works (m/km)

a0 to a2 user-definable parameters (defaults = 0.3, 5.4, 0.5 respectively)
❏ Slurry Seal and Cape Seal without shape correction
RI aw = RIap - MAX {0, MIN [a0 * (RI ap - a1), a2 * Hsl]} ...(4.21)
where:

Hsl thickness of slurry seal or cape seal in mm (that is, HSNEWaw)
a0 to a2 user-definable parameters (defaults = 0.3, 4.6, 0.09 respectively)
❏ Cape Seal, SBSD and DBSD with Shape Correction
RI aw = RIap - MAX { 0, MIN [a0 * Hsc * RIap , a1 * Hsc * MAX (0, (RIap - a2 ))]}
...(4.22)
where:

Hsc the thickness of reseal including shape correction layer (that is,
HSNEWaw) (mm), Hsc = MIN(Hsc, 45)
a0 to a2 user-definable parameters (defaults = 0.0075, 0.0225, 4.0
respectively)
! Texture depth and skid resistance
Resealing resets texture depth and skid resistance to the user-specified values. If these are
not specified, texture depth after works (TDaw) is reset to the default value of initial
texture depth given in Table D2.6; skid resistance after works (SFCaw) is reset to 0.6 for
all types of reseals.

Version 1.0
Table D2.6 Default values for Initial Texture Depth (ITD)
Surface type Surface material Initial Texture Depth

(ITD) in mm
AC 0.7
HRA 0.7
PMA 0.7
AM RAC 0.7
CM 0.7
SMA 0.7
PA 1.5
SBSD 2.5
DBSD 2.5
ST CAPE 0.7
SL 0.7
PM 1.5
! Previous cracking
The area of previous cracking (PCRA and PCRW) is updated to equal the cracking in the
current surfacing before resealing, and a weighting (w) of the cracking in the previous
surfacing, as follows:
❏ if CRAi bw ≥ PCRi bw
PCRi aw ≥ PCRi bw ...(4.23)
❏ if CRAi bw < PCRi bw
PCRi aw = w * CRAi bw + (1 - w ) * PCRi bw ...(4.24)
w = MIN (0.70 + 0.1 * HSNEWaw , 1) ...(4.25)
where:
PCRiaw area of previous cracking type i (i = all structural cracking or wide

structural cracking) after works (% of total carriageway area)
CRAibw area of cracking type i before works (% of total carriageway area)
PCRibw area of previous cracking type i before works (% of total carriageway
area)
w weighting used for averaging the cracking in the old and new
surfacing layers
HSNEWaw thickness of the reseal (mm)

Version 1.0
The number, per km, of previous transverse thermal cracking (PNCT) is also reset in
the same way as for all and wide structural cracking using Equations 4.23 above,
4.24 above and 4.25 above, except for the following definitions:
PCRiaw number of previous transverse thermal cracking after works (per km)
CRAibw number of transverse thermal cracking before works (per km)
PCRibw number of previous transverse thermal cracking before works (per
km)
! Pavement age
The surfacing age (AGE2) and the preventive treatment age (AGE1) is reset to zero after
resealing works.
4.3 Overlay
Overlay works can be defined in one of the following manners:
The overlay of fixed specifications is applied whenever the rehabilitation age (AGE3)
equals or exceeds a fixed time interval specified by the user.
The overlay of fixed specifications is applied when the levels of the user-specified
intervention criteria are met.
For both options (1 and 2), overlay is not performed if the rehabilitation/overlay age (AGE3)
is less than the user-specified minimum applicable overlay-interval. An overlay is also not
performed if either the preventive treatment age (AGE1) or the surfacing age (AGE2) is less
than the respective minimum preventive treatment or resurfacing intervals. If the minimum
intervals have not been defined within any of the standards that have been assigned to the road
section, the following default values (in years) are used: preventive treatment - 2 years,
resealing - 4 years.
An overlay is always performed when AGE3 exceeds the user-specified maximum allowable
overlay interval. In all cases, an overlay is not performed if the last applicable year or
maximum applicable roughness has been exceeded.
Overlay is specified using the following:
! New surfacing thickness.
! Layer strength coefficient.
! Surface material.
! Construction defect indicator for bituminous surfacing (CDS).

If performed, the amount of overlay is given by:
AOVL = 1000 * CW ...(4.26)
where:

Version 1.0
AOVL overlay area (m2/km)

Before overlay is performed, it is often necessary to carry out some preparatory works. It is
assumed that the following amounts of preparatory works is carried out along with overlay:
! Patching
APAT = 10 * CW * APOTb ...(4.27)
where:
APAT area of preparatory patching (m2/km)
! Edge-repair
Calculated using Equation 4.9 above

The total amount of overlay (TAOVL), in square metres, is obtained from the product of
AOVL and the section length (L), in kilometres. The total cost of overlay is obtained by
multiplying TAOVL by the user-specified unit cost per square metre. The additional areas and
costs of preparatory patching and edge-repair are reported separately under patching and edge-
repair, respectively.
4.3.2 Effects of overlay

Overlay resets the pavement structure as described below:
! Pavement type
The pavement type after overlay is reset as detailed in Table D2.2 and Table D2.3. The
surface material after works is user-specified.
! Surface thickness
The total surface thickness after works is obtained in the following manner:
HS aw = HS bw + HSNEWaw ...(4.28)
where:
HSaw total surface thickness after overlay works (mm)

HSbw total surface thickness before overlay works (mm)
HSNEWaw user-specified thickness of overlay (mm)
The thickness of the old, underlying bituminous surfacing after works is calculated using
Equation 4.11 above.
! Pavement strength
The pavement strength parameters are updated to take account of the net change in
pavement strength due to the new overlay and the underlying cracks (if any), as follows:

Version 1.0
❏ The dry season adjusted structural number of pavement is given by:
SNPdaw = MAX [1.5, (SNPdbw + 0.0394 * a sw * HSNEWaw - dSNPK )] ...(4.29)
where:
SNPdaw dry season adjusted structural number of pavement after overlay

works
SNPdbw dry season adjusted structural number of pavement before overlay
works
dSNPK reduction in the adjusted structural number of pavement due to
cracking
HSNEWaw thickness of the overlay (mm)
asw layer strength coefficient of the overlay
❏ The wet season adjusted structural number of pavement (SNPwaw) and the annual
average adjusted structural number of pavement (SNPaw) after works are then
calculated as discussed in Section 4.2.
The Benkelman beam deflection after works is computed using Equation 4.13 above.
The construction defect indicator for bituminous surfacing CDS is reset to a user
specified value. If these are not specified, a good construction quality is assumed, with
the CDS value set to 1.0.
Overlay works resets surface distresses to zero, and thereafter the pavement condition is
considered to be new.
! Rutting
The effects of overlay on rutting are user-specified. If these are not specified, the mean
rut depth is calculated as follows:
RDM aw = a0 * RDMbw ...(4.30)
where:

RDMbw mean rut depth before works (= RDMb) (mm )
! Roughness
The effects of overlay on roughness are user-specified. If these are not specified, the
roughness value after works is calculated in the following manner:
The roughness value at the end of the year is adjusted to take account of preparatory
works as follows:

Version 1.0
RI ap = RIb - MIN (∆RI t , a0 ) ...(4.31)
where:
RIap adjusted roughness after preparatory patching (IRI m/km)

a0 user-definable model coefficients (default = 4.6)
The adjusted roughness after preparatory works (RIap) is then used to compute the final
roughness after overlay as follows (NDLI, 1995):
RI aw = = a0 + a1 * MAX [0, (RIap - a0 )] * MAX [0, (a2 - HSNEWaw )] ...(4.32)
where:
RIaw roughness after overlay (IRI m/km)

RIap adjusted roughness after preparatory works (IRI m/km)
HSNEWaw thickness of overlay (mm)
a0 to a2 user-definable parameters (defaults = 2.0, 0.01, 80 respectively)
After overlay, the texture depth and the skid resistance are reset to the user-specified
values. If these are not specified, texture depth after works (TDaw) is reset to the default
value of initial texture depth given in Table D2.6; skid resistance after works (SFCaw) is
reset to 0.5 for all overlays.
! Previous cracking
The amounts of previous cracking (PCRA, PCRW, and PNCT) are reset using Equations
4.23 above and 4.24 above. The weighting factor (w) of the cracking in the previous
surfacing is computed as follows:
❏ For asphalt roadbase (AB), asphalt pavement (AP) and granular roadbase (GB):
 HSNEWbw 
w = MAX  , 0.6  ...(4.33)
 HSOLD aw 
❏ For stabilised roadbase (SB)
 HSNEWbw 
w = MAX  , 0.6  ...(4.34)
 (HSOLD aw + HBASE ) 
where:

Version 1.0
w weight used for averaging the cracking in the old and new surfacing
layers
HBASE thickness of the base layer in the original pavement (required only for SB
roadbase types) (mm)
! Pavement age
The rehabilitation age (AGE3), the surfacing age (AGE2), and the preventive treatment
age (AGE1) are reset to zero after overlay works.
4.4 Mill and replace

This operation involves the removal of all or part of the existing bituminous surfacing and
replacing it with a new bituminous surfacing. It is usually performed to correct defects that
have occurred mainly due to poor construction quality and the bituminous material being too
rich or brittle, or where the road surface levels need to comply with some requirements related
to drainage facilities, bridge underpasses and other such structures.
Mill and replace works, which is a type of Rehabilitation works, can be defined in one of the
following ways:
The mill and replace of fixed specifications is applied whenever the rehabilitation age
(AGE3) equals or exceeds a fixed time interval specified by the user.
The mill and replace of fixed specifications is applied when the levels of the user-
specified intervention criteria are met.
For both options (1 and 2), a mill and replace works is not performed if the rehabilitation age
(AGE3) is less than the user-specified minimum applicable interval, or if the last applicable
year has been exceeded. A mill and replace works is always performed when AGE3 exceeds
the user-specified maximum allowable interval.
The following information is required to specify mill and replace works:
! New surfacing thickness
! Layer strength coefficient
! Surface material
! Depth of milling
! Construction defect indicator for bituminous surfacing

If performed, the amount of mill and replace is given by:
AMR = 1000 * CW ...(4.35)
where:

Version 1.0
AMR area of carriageway milled and replaced (m2/km)

The total amount of mill and replace works (TAMR), in square metres, is obtained from the
product of AMR and the section length (L), in kilometres. The total cost of mill and replace
works is obtained by multiplying TAMR by the user-specified unit cost per square metre.
4.4.2 Effects of mill and replace

Mill and replace works resets the pavement structure as described below:
! Pavement type
After mill and replace, the pavement type is reset as detailed in Table D2.2 and Table
D2.3. The surface material after works is user-specified.
The following conditions determine which kind of mill and replace activity is being
applied:
if MILLD ≥ HS bw then it is a mill and replace to roadbase
if MILLD < HS bw then it is a mill and replace to intermediate layer
! Surface thickness
The total surface thickness after works is obtained as described below:
HS aw = HS bw + HSNEWaw - MILLD ...(4.36)
where:
HSaw total surface thickness after works (mm)

HSbw total surface thickness before works (mm)
HSNEWaw user-specified thickness of new surfacing (mm)
MILLD depth of milling (mm)
The thickness of the old, underlying bituminous surfacing after works is given by:
HSOLD aw = MAX(HSNEWbw + HSOLD bw - MILLD, 0 ) ...(4.37)
where:

HSOLDbw total thickness, in mm, of previous, underlying surfacing layers before
works

Version 1.0
! Pavement strength
The pavement strength parameters are updated to take account of the net change in
pavement strength due to the new bituminous surfacing and the depth of milling as
follows:
❏ If MILLD ≤ HSNEWbw , then the dry season adjusted structural number of
pavement is given by:
SNPdaw = MAX[1.5, (SNPdbw - 0.0394 * a hsn * MILLD + 0.0394 * a sw * HSNEWaw )]
...(4.38)
❏ If HSNEWbw < MILLD ≤ HS bw , then the dry season adjusted structural number
of pavement is calculated as follows:
 SNPdbw - 0.0394 * a hsn * HSNEWbw + 0.0394 * a sw  

SNPdaw = MAX1.5,  
 * HSNEWaw − 0.0394 * a hso * (MILLD - HSNEWbw ) 
...(4.39)
❏ If MILLD > HS bw , then the dry season adjusted structural number of pavement
is given by:
 SNPdbw - 0.0394 * a hsn * HSNEWbw − 0.0394 * a hso * HSOLD bw  

SNPdaw = MAX1.5,  
 − 0.0394 * a b * (MILLD - HS bw ) + a sw * HSNEWaw 
...(4.40)
where:
SNPdaw dry season adjusted structural number of pavement after works

SNPdbw dry season adjusted structural number of pavement before works
ahsn strength coefficient of the most recent surfacing before works
ahso strength coefficient of the old surfacing before works
asw strength coefficient of the new surfacing after works
ab strength coefficient of the roadbase layer
The wet season adjusted structural number of pavement (SNPwaw) and the annual average
adjusted structural number of pavement (SNPaw) after works are calculated using the
same procedure as discussed in Section 4.2.
The Benkelman beam deflection after works is calculated using Equation 4.13 above.
the CDS value set to 1.0.

Version 1.0
Mill and replace works resets surface distresses to zero, and thereafter it is assumed that
the pavement behaves as if new.
! Rutting
Mill and replace resets rutting to zero by default, unless the user specifies a different
value.
! Roughness
The effects of mill and replace on roughness are user-specified. If these are not specified,
the following default values are used:
For AM surface type, RIaw = 2.0 (IRI m/km)
For ST surface type, RIaw = 2.8 (IRI m/km)
After mill and replace, the texture depth and the skid resistance are reset to the user-
specified values. If these are not specified, texture depth after works (TDaw) is reset to the
default value of initial texture depth given in Table D2.6; skid resistance after works
(SFCaw) is reset to the following default values:
For surface type AM, SFCaw = 0.5
For surface type ST, SFCaw = 0.6
! Previous cracking
The amounts of previous cracking (PCRA, PCRW, and PNCT) are reset as follows:
If MILLD < HSNEWbw , then the amounts of previous cracking are reset as follows:
if CRAi bw ≥ PCRi bw
PCRi aw = wf * CRAi bw ...(4.41)
if CRAi bw < PCRi bw
PCRi aw = wf * CRAi bw + (1 − w ) * PCRi bw ...(4.42)
  MILLD  
wf = w * 1 -   ...(4.43)
  HSNEWbw  
The weighting (w) of the cracking in the previous surfacing is computed as follows:
❏ For asphalt base (AB), asphalt pavement (AP) and granular roadbase (GB)
 HSNEWbw 
w = MAX  , 0.6  ...(4.44)
 HSOLD aw 
❏ For stabilised base (SB)
 HSNEWbw 
w = MAX  , 0.6  ...(4.45)
if MILLD ≥ HSNEWbw

Version 1.0
then:
PCRi aw = wg * PCRi bw ...(4.46)
 HS bw − MILLD 
wg = MAX 0, ...(4.47)
 HSOLD bw 
Equation 4.47 above assumes that if MILLD = HSbw any cracking in the roadbase is
rectified before surfacing.
All the parameters used in Equations 4.41 above to 4.47 above are as defined previously
under resealing and overlay.
! Pavement age
The rehabilitation/overlay age (AGE3), the surfacing age (AGE2), and the preventive
treatment age (AGE1) is reset to zero after a mill and replace works.
4.5 Inlays
Inlay is a special works activity, considered under Rehabilitation, that is normally applied to
treat rutting along wheelpaths (and this involves some milling of the existing pavement
layers). It is therefore not scheduled, but defined as a condition-responsive works, in which
an inlay of fixed specifications is applied when the level of pavement defect exceeds the user-
specified values.
Inlay is not performed if the user-specified last applicable year or the maximum applicable
roughness has been exceeded.
To define an inlay works, the percentage of total carriageway area to be repaired and the
construction quality factors should be specified.

If performed, the amount of inlay is given by:
AINLY = 10 * Pin * CW ...(4.48)
where:
AINLY area of inlay (m2/km)

Pin area to be repaired (% of the total carriageway area) (0 < Pin < 100)
It is likely that patching, crack sealing and edge-repair will be carried out along with inlay
works. The modelling logic assumes therefore that first inlay is performed, and then patching
and crack sealing repairs the remaining areas of potholes and cracking. The amounts of
patching, crack sealing and edge-repair that may be performed are computed as follows:
! Patching
APAT = 10 * CW * ∆ACWpat + APOTb * 1 -

Pin  
 ...(4.49)
  100  

Version 1.0
  Pin  
∆ACWpat = MAX 0.1 * (ACWb − 20 ) * 1 −  , 0 ...(4.50)
  100  
where:
APAT area of patching (m2/km)

∆ACWpat area of wide structural cracking treated by patching (% of the total
carriageway area)
! Crack sealing
It is assumed that crack sealing repairs the entire remaining area of transverse thermal
cracking and wide structural cracking as follows:
  Pin 
ACSL = 10 * CW * Pacw * ACWbcs + Pact * ACTb * 1 -  ...(4.51)
  100 
ACWbcs = ACWb -  ∆ACWpat + ACWb * 1 -

Pin 
 ...(4.52)
  100 
where:
ACSL area sealed (m2/km)

Pacw proportion of wide structural cracking to be sealed (default = 1.0)
Pact proportion of transverse thermal cracking to be sealed (default = 1.0)
! Edge-repair
The amount of edge-repair is calculated using Equation 4.9 above

The total amount of inlay (TAINLY), in square metres, is obtained from the product of
AINLY and the section length (L), in kilometres. The cost of inlay is obtained by multiplying
TAINLY by the user-specified unit cost per square metre. The additional areas and costs of
patching, crack sealing and of edge-repair are reported separately under patching, crack
sealing and edge-repair, respectively.
4.5.2 Effects of inlays

Inlays does not alter the pavement type and surface thickness, since it is assumed that the
same surface material of the existing pavement is used for inlay works.
To take account of the net strengthening of the pavement due to maintenance, the dry season
structural number of pavement is updated as follows:
SNPdaw = MAX(1.5, SNPdbw ) ...(4.53)
where:

Version 1.0
The wet season adjusted structural number of pavement (SNPwaw) and the annual average
adjusted structural number of pavement (SNPaw) after works is then calculated as discussed in
Section 4.2 above for resealing works.
The Benkelman beam deflection after works is computed using Equation 4.13 above.
a CDS value of 1.0.
After inlay works, surface distresses are reset as follows:

❏ Potholes are reset to zero (that is, NPTaw = 0), because of the additional patching that
is assumed to be performed during inlay works.
❏ Cracking is reset as follows:
ACWaw = MAX {[ACWb - (∆ACWinlay + ∆ACWpat + ∆ACWcs )], 0} ...(4.54)
 Pin 
∆ACWinlay = ACWb * 1 -  ...(4.55)
 100 
∆ACWcs = Pacw * [ACWb - (∆ACWinlay + ∆ACWpat )] ...(4.56)
ACA aw = ACA b - (ACWb - ACWaw ) ...(4.57)
ACTaw = Pact * ACTb ...(4.58)
where the parameters Pacw and Pact are as defined in Equation 4.51 above
The amounts of indexed cracking (ACXaw) and total area of cracking (ACRAaw) are
reset after inlay works using Equations 3.32 above and 3.33 above, respectively.
❏ Ravelling is reset as follows:
ARVaw = ARVb * 1 -

Pin 
 ...(4.59)
 100 
where:
ARVaw area of ravelling after works (% of the total carriageway area)

ARVbw area of ravelling before works (= ARVb) (% of the total carriageway
area)
❏ Edge-break is reset to zero (that is, AVEBaw = 0), due to the edge-repair works that
is assumed to be performed during inlay works.

Version 1.0
! Rutting
The effects of inlay on rutting are user-specified. If these are not specified, the mean rut
depth is calculated as follows:
RDM aw = a0 * RDMbw ...(4.60)
where:

RDMbw mean rut depth before works (= RDMb) (mm)
! Roughness
Inlay reduces roughness by treating rutting, all the potholes, transverse thermal cracking
and wide structural cracking. The effect of inlay on roughness is user-specified. If these
are not specified, the values of roughness after works is computed as follows:
RI aw = MAX[a0, RIbw - a1 * ∆RDS w - ∆RI t - a2 * (∆CRX w + ∆ACTw )] ...(4.61)
where:
RIaw roughness after inlay (IRI m/km)

RIbw roughness before works (= RIb) (IRI m/km)
∆RDSw reduction in rut depth standard deviation due to inlay works (mm)
(=DSbw – RDSaw, where RDSaw is calculated using RDMaw)
∆CRXw reduction in the amount of indexed cracking due to inlay works

(= ACXb - ACXaw)
a0 to a2 user-definable parameters (defaults = 2.8, 0.088, and 0.0066 respectively)
Inlay resets texture depth and skid resistance in the same way as described for mill and
replace in Section 4.3.2 above.
! Previous cracking
The amounts of previous cracking (PCRA, PCRW, and PNCT) is reset as follows:
PCRi aw = 1 -
Pin 
 * PCRi bw ...(4.62)
 100 
where:

Version 1.0
PCRiaw amount of previous cracking type i (i = all structural cracking or wide

structural cracking or transverse thermal cracking) after inlay works
PCRibw amount of previous cracking type i before inlay works
! Pavement age
The surfacing age after inlays is reset as follows:
AGEi aw = 1 -
Pin 
 * AGEi bw ...(4.63)
 100 
where:
AGEiaw AGE type i (i = 1, 2 or 3) after inlay works (years). (This value should be
adjusted to an integer)
AGEibw AGE type i before inlay works (years)
4.6 Reconstruction
Pavement reconstruction refers to all works that require the re-specification of the surfacing
and roadbase types.
Reconstruction may be specified either as a maintenance standard or as an improvement
standard if the works involve minor widening of the carriageway.
Note: the latter is permitted only for the sake of convenience in the modelling logic as
discussed in Section 6.1.
Reconstruction may be specified in one of the following ways:
A reconstruction works of fixed specifications is to be performed whenever the

construction age (AGE4) equals or exceeds the user-specified maximum allowable age.
! Option 2: Responsive
A reconstruction works of fixed specifications is applied when the levels of the user-
specified intervention criteria, based on pavement condition and traffic loading, are met.
Reconstruction is not performed if the last applicable year has been exceeded.
For both options (1 and 2), reconstruction is not performed if:
AGE4 is less than the minimum reconstruction interval
AGE1, AGE2 and AGE3 are less than their respective minimum intervals
It is a construction-opening year
Reconstruction is always performed if AGE4 exceeds the maximum allowable reconstruction

interval, if specified by the user. In all cases, reconstruction is not performed if the last
applicable year has been exceeded.
Reconstruction as a maintenance standard is specified using the following:
! New pavement type

Version 1.0
! Surface material
! Surfacing thickness
! Structural number of pavement (SN) of the layers above the subgrade
! Relative compaction
! Construction defect indicators
For stabilised roadbases, the base thickness and the resilient modulus are also required.

If performed, the amount of pavement reconstruction is given by:
ARCON = 1000 * CW ...(4.64)
where:
ARCON area of road reconstructed (m2/km)

The total area reconstructed is given by:
TARCON = ARCON * L ...(4.65)
where:
TARCON total carriageway area reconstructed (m2)

L section length (m)
The total cost of reconstructing the entire road section is obtained from the product of
TARCON and the unit cost per square metre, or from the product of the section length (L) and
the unit cost per kilometre.
4.6.2 Effects of reconstruction

After reconstruction, the pavement type is reset to the new type specified by the user. All
bituminous pavement types are permitted, except STAP and AMAP. The required modelling
parameters of the surfacing and roadbase/sub-base layers are reset as described below:
! Pavement strength
The dry season adjusted structural number of pavement is computed as follows:
SNPdaw = SN new + SNSG s ...(4.66)
SNSG s = 3.51 * log10 CBR - 0.85 * (log10 CBR ) − 1.43

2
...(4.67)
where:

Version 1.0
SNPdaw dry season adjusted structural number of pavement after reconstruction

works
SNnew user-specified new dry season structural number of pavement for the
layers above the subgrade. This value may be obtained from the sum of
SNBASU and SNSUBA as described in Chapter C2
CBR in situ subgrade CBR in dry season. Note also that SNSGs equals
SNSUBGs (see Chapter C2) by assuming a sub-base thickness of about
183 mm
The Benkelman beam deflection after reconstruction works is given by:

−1.6
DEFaw = a0 * (SNPaw ) ...(4.68)
where:
SNPaw adjusted structural number of pavement after reconstruction works

DEFaw Benkelman beam deflection after reconstruction works (mm)
a0 coefficient depending upon the roadbase type (a0 = 6.5 for GB, AB and
AP; a0 = 3.5 for SB roadbase type)
The construction defect indicators for bituminous surfacing (CDS), and the construction
defect indicator for the roadbase, (CDB), are reset to user specified values. If these are
not specified, a good construction quality is assumed, with the CDS value set to 1.0 and
the CDB value set to 0.
Surface distresses (that is, edge-break, potholing, cracking and ravelling) are all reset to
zero.
! Rutting
After reconstruction, the mean rut depth is reset to zero.

! Roughness
Pavement reconstruction resets roughness to a user-specified value. The roughness after

works depends on the construction quality. The following values are used as defaults:
For AM surface type, RIaw = 2.0 (IRI m/km)
For ST surface type, RIaw = 2.8 (IRI m/km)
Reconstruction resets texture depth and skid resistance in the same way as described for
mill and replace in Section 4.4.2.
! Previous cracking
The area of previous cracking (PCRA, PCRW, and PNCT) is reset to zero.

Version 1.0
! Road deterioration factors
The cracking retardation time, ravelling retardation factor and the other road deterioration
factors are also reset to user-specified values.
! Pavement age
The pavement ages: AGE1, AGE2, AGE3, and AGE4 are all reset to zero.

Version 1.0
5 Special works
The effects of the following special works on pavement performance are not modelled
endogenously, and therefore only their costs can be considered in an analysis:
! Emergency works
For example, (repairing washout/subsidence, clearing debris, traffic accident removal,

etc.).
! Winter maintenance
These works types are scheduled at a fixed interval of time (minimum of one-year), and are
performed on an annual basis. If specified, the works activities are applied in a given analysis
year regardless of the works hierarchy given in Table D2.1.
Their unit costs are specified in terms of currency per kilometre per year, and the annual costs
are obtained by multiplying the road section length (L) by the unit cost.

Version 1.0
6 Improvement works
Road improvement works aim to provide additional capacity and comprises the following:
! Partial widening
! Lane addition
! Realignment
! Non-carriageway improvement works
For example, improvement or addition of shoulders, NMT lanes, and side drains.
Although it is not an improvement works, reconstruction that involves minor widening of the
carriageway is modelled using the logic framework for improvement works.
6.1 Reconstruction with minor widening

Pavement reconstruction with minor widening (see also notes given in Section 4.6) may be
specified in one of the following ways:
A reconstruction of fixed specifications is performed at a specified time, defined by a

calendar year.
A reconstruction of fixed specifications is applied when the levels of the user-specified

intervention criteria, based on pavement condition and traffic loading, are met.
Reconstruction is not performed if the last applicable year has been exceeded.
Reconstruction works are specified using the following:
! New pavement type
! Surface material
! Surfacing thickness
! Structural number of pavement (SNnew) of the layers above the subgrade
! Relative compaction
! Increase in carriageway width
! Construction defect indicators
! For stabilised roadbases, the base thickness and the resilient modulus are also required.

ARCON = 1000 * CWaw ...(6.1)
where:

Version 1.0
ARCON area of road reconstructed (m2/km)

CWaw new carriageway width after reconstruction works (m)
The total area reconstructed is given by:
TARCON = ARCON * L ...(6.2)
where:
TARCON total carriageway area reconstructed (m2)

L road section length (km)
The total cost (CSTCON) of reconstructing the entire road section is obtained from the
product of TARCON and the unit cost per square metre, or from the product of the section
length and the unit cost per kilometre.
The salvage value is given by:
SALVA = PCTSAV * CSTCON ...(6.3)
where:
SALVA salvage value of the works (currency)

PCTSAV percent of total cost salvageable (%)
6.1.2 Effects of reconstruction

After reconstruction, the pavement type, strength, condition, age, deterioration factors, and
construction quality indicators as described in Section 4.6.2. In addition, the new carriageway
width after reconstruction with minor widening is given by:
CWaw = CWbw + ∆CW ...(6.4)
where:
CWaw carriageway width after reconstruction works (m)

CWbw carriageway width before reconstruction works (m)
∆CW user-specified increase in carriageway width (m)
The effective number of lanes after works is reset to a new value ELANESaw, which should be
input by the user. If this is not specified ELANESaw will be reset to the value of NLANES
(that is, the number of lanes for the road section) see Part B Section 5.2.

Version 1.0
6.2 Widening
There are two operations included under widening, lane addition and partial widening. The
major distinction is that partial widening does not increase the number of lanes. Both widening
operations do not alter the road alignment.
Widening works can be specified in one of the following ways:
A widening works of fixed specifications is applied at a specified time, defined by a

calendar year.
A widening works of fixed specifications is applied when the levels of the user-specified
intervention criteria, based on road user effects parameters, are met. Widening is not
performed if the last applicable year has been exceeded.
The following information is required to specify a widening works:
! New road type
! Road class
! Increase in carriageway width for partial widening
! Additional number of lanes and increase in carriageway width for lane addition
! Pavement type for the entire section
! Pavement details of the widened area of the carriageway
! Whether or not the existing carriageway is to be provided with a new surfacing

If performed, the amount of widening works is given by:
AWDN = 1000 * ∆CW ...(6.5)
where:
AWDN widened area of carriageway (m2/km)

∆CW increase in carriageway width (m)
The total area of widening over the entire section is given by the expression:
TAWDN = AWDN * L ...(6.6)
where:
TAWDN total widened area of carriageway for the entire section (m2)

Version 1.0
The cost of widening (CSTWDN) is obtained from the product of TAWDN and the unit cost
per square metre, or from the product of the section length and the unit cost per kilometre.
SALVA = PCTSAV * CSTWDN ...(6.7)
where:

Additional works
It is likely that widening works will include re-surfacing or at least repairs of the severely
damaged area of the existing carriageway. The additional works that may be required are
modelled as described below. (Note that re-surfacing or re-surfaced has been used to denote
the provision of a new bituminous surfacing and to distinguish it from resurfacing as a type
of roadwork).
Case1: re-surfacing the existing carriageway
If the existing carriageway is to be re-surfaced either by an overlay or resealing (as
determined from the user-specified new pavement type after works), the amount of re-
surfacing works for the entire road section is given by:
TANSF = 1000 * CWbw * L ...(6.8)
where:
TANSF total carriageway area re-surfaced (m2)

! An overlay is provided over the existing carriageway
TANSF would be denoted as TAOVL, and the total cost of overlay is obtained from the
product of TAOVL and the unit cost of overlay per square metre.
Before overlay is performed, it is often necessary to carry out some preparatory works.
The amount of preparatory works to be performed for the entire road section is
❏ Patching
TAPAT = 10 * CWbw * APOTb * L ...(6.9)
where:
TAPAT total area of preparatory patching (m2)
The total cost of preparatory patching is obtained from the product of TAPAT and
the unit cost of patching per square metre.

Version 1.0
❏ Edge-repair
TAVER = 10 * CWbw * AVEB b * L ...(6.10)
where:
TAVER total area of edge-repair (m2)

AVEBbw area of edge-break before widening works (= AVEBb) (% of total
carriageway area)
The total cost of edge-repair is obtained from the product of TAVER and the unit
cost of edge-repair per square metre.
! The existing carriageway is resealed
ANSF would be denoted by TARSF, and the total cost of resealing is obtained from the
product of TARSF and the unit cost of resealing per square metre.
Before resealing is applied, the amount of preparatory works that may be required for
the entire road section is computed as follows:
❏ Patching
TAPAT = 10 * CWbw * L * {MAX[0.1 * (ACWb - 20 ), 0] + APOTb } ...(6.11)
where:

❏ Crack sealing
It is assumed that 100% of transverse thermal cracking is sealed, and the amount of
crack sealing performed is given by:
TACSL = 10 * CWbw * ACTbw * L ...(6.12)
where:
TACSL total area of crack sealing (m2)

ACTbw area of transverse thermal cracking before widening works (= ACTb)
The total cost of crack sealing is obtained from the product of TACSL and the unit
cost of crack sealing per square metre.

Version 1.0
❏ Edge-repair
The total area of edge-repair is calculated using Equation 6.10 above. The total cost
of edge-repair is obtained from the product of TAVER and the unit cost of edge-
repair per square metre.
Case 2: no re-surfacing of the existing carriageway

If the existing carriageway is not to be re-surfaced, it is assumed that the following amounts of
patching, crack sealing and edge-repair are performed along with the widening works:
! Patching
Assuming that 100% of severely damaged area is patched, the amount of patching is
TAPAT = 10 * CWbw * L * ADAMS bw ...(6.13)
where:
TAPAT total carriageway area patched (m2)

ADAMSbw severely damaged carriageway area before widening works (= ADAMSb)
! Crack sealing
The total amount of crack sealing is calculated using Equation 6.12 above.
! Edge-repair
Assuming that all the edge-break area is repaired, the amount of edge-repair works is
given by Equation 6.10 above.
The total cost of widening works is the sum of carriageway widening cost and the cost of
additional works comprising re-surfacing of the existing carriageway and preparatory
works. The amount and cost of re-surfacing works are reported separately under resealing or
overlay. Similarly, the amounts and costs of patching, of crack sealing and edge-repair are
reported separately under patching, crack sealing and edge-repair, respectively.
In economic analysis, it is assumed that the cost of all these additional works is incurred in the
last construction year.
6.2.2 Effects of widening

After widening, the pavement type is reset to the new type specified by the user. The required
modelling parameters are also reset to user-specified values, or computed endogenously as
described below:
! Carriageway width
The new carriageway width after widening is given as follows:
CWaw = CWbw + ∆CW ...(6.14)
where:

Version 1.0
CWaw carriageway width after widening works (m)

CWbw carriageway width before widening works (m)
For partial widening, the increase in carriageway width (∆CW) is specified directly by
the user. For lane addition, the increase in carriageway width is user-specified, if
this is not specified the increase will be given by:
ADDLN * CWbw
∆CW = ...(6.15)
NLANES bw
where:
ADDLN additional number of lanes, input by the user

NLANESbw number of lanes before widening works
For lane addition, the number of lanes after widening works (NLANES aw) is equal
to the number of lanes before works (NLANESbw) plus the user-specified additional
number of lanes (ADDLN).
The effective number of lanes after works is reset to a new value ELANESaw, which
should be input by the user. If this is not specified, ELANESaw will be reset to the value of
NLANESaw (that is, the number of lanes for the road section following widening) see Part
B Section 5.2.
! Thickness of surfacing layers
❏ Case 1: re-surfacing the existing carriageway
If the existing carriageway is to be re-surfaced, the thickness of new surfacing after

widening works is obtained as follows:
(CWbw * HRESF + ∆CW * HSNEW ww )

HSNEWaw = ...(6.16)
CWaw
where:
HSNEWaw new surfacing thickness after widening works (mm)

HSNEWww surfacing thickness of the widened part of the carriageway (mm)
HRESF user-specified thickness of the re-surfacing layer on the existing
carriageway (mm)
The thickness of old surfacing after widening is given as:
(CWbw * HS bw )
HSOLD aw = ...(6.17)
CWaw
where:

Version 1.0
HSOLDaw thickness of old surfacing after widening works (mm)

HSbw total surfacing thickness of the existing carriageway before widening
works (mm)
❏ Case 2: no re-surfacing of the existing carriageway
If the existing carriageway is not to be re-surfaced, the thickness of new surfacing

after widening works is obtained as follows:
(CWbw * HSNEWbw + ∆CW * HSNEW ww )

HSNEWaw = ...(6.18)
CWaw
where:
HSNEWaw new surfacing thickness after widening works (mm)

HSNEWbw new surfacing thickness before widening works (mm)
HSNEWww surfacing thickness of the widened part of the carriageway (mm)
The thickness of old surfacing after widening is given as:
(CWbw * HSOLD bw )
HSOLD aw = ...(6.19)
CWaw
where:

HSOLDbw thickness of old surfacing before widening works (mm)
! Pavement strength
The dry season adjusted structural number of pavement is reset to the weighted
average of that of the existing carriageway and that of the widened part of carriageway,
as follows:
(CWbw + ∆CW )
0.2
 
SNPdaw = SNPdexcw * SNPdww * 
 (CWbw * [SNPdww ] 5 + ∆CW * [SNPdexcw ] 5 )
...(6.20)
SNPdww = SN dww + SNSG s ...(6.21)
❏ If the existing carriageway is to be re-surfaced
SNPdexcw = MAX[1.5, (SNPdbw + 0.0394 * a resf * HRESF - dSNPK )] ...(6.22)

Version 1.0
❏ If the existing carriageway is not to be re-surfaced
SNPdexcw = MAX[1.5, (SNPdbw - dSNPK )] ...(6.23)
where:
SNPdaw dry season adjusted structural number of pavement after widening works
SNPdbw dry season adjusted structural number of pavement before widening
works
SNPdww dry season adjusted structural number of pavement of the widened part of
the carriageway
SNdww user-specified structural number of pavement (of the layers above the
subgrade) of the widened part of the carriageway. This may be computed
in a similar manner to that described for SNnew in Equation 4.66 above
SNSGs subgrade contribution to the structural number of pavement, calculated
using Equation 4.67 above
dSNPK reduction in the adjusted structural number of pavement due to cracking
aresf strength coefficient of the re-surfacing layer on the existing carriageway

The Benkelman beam deflection after widening works is given by:
−1.6
 SNPaw 
DEFaw = DEFbw *   ...(6.24)
 SNPbw 
where:
DEFaw benkelman beam deflection after works (mm)

DEFbw benkelman beam deflection before works (mm)
! Surface material
If the existing carriageway is to be re-surfaced, the surface material after works is reset to
that specified for the widening works. This is based on the assumption that the same
surfacing material is used for the re-surfacing.
If the existing carriageway is not to be re-surfaced, the surface material after works is
reset as follows:
If CWbw is greater than ∆CW, the surface material after widening works is
reset to that of the existing carriageway.
Otherwise the surface material after widening works is reset to that of the widened
part of the carriageway.

Version 1.0
The construction defect indicators for bituminous surfacing (CDS) and the construction
defect indicator for the roadbase (CDB) are reset to a weighted average computed in the
following manner:
 (CDi bw * CWbw + CDi ww * ∆CW ) 

CDi aw =   ...(6.25)
 CWaw 
where:
CDiaw construction defect indicator i (i=CDS or CDB) after widening

CDibw construction defect indicator i before widening works
CDiww construction defect indicator i specified for the widening works
If the existing carriageway is to be re-surfaced, the amounts of all surface distresses after
widening works are reset to zero.
If the existing carriageway is not to be re-surfaced, the areas of edge-break, potholing,
transverse thermal cracking, wide structural cracking and ravelling after widening works
are all reset to zero. The area of all structural cracking and the total area of cracking are
calculated as follows:
 (ACA b - ACWb ) * CWbw 

ACA aw =   ...(6.26)
 CWaw 
ACRA aw = ACA aw ...(6.27)
where:
ACAaw area of all structural cracking after widening works (% of total

carriageway area)
ACRAaw total area of cracking after widening works (% of total carriageway area)
! Rutting
The mean rut depth is reset to a user-specified value. If this is not specified, the mean rut
 (CWbw * a0 * RDMbw ) 
RDM aw =   ...(6.28)
 CWaw 
where:
RDMaw mean rut depth after widening works (mm)

RDMbw mean rut depth before widening works (mm)
a0 user-definable coefficient (default = 0.15 if the existing carriageway is to

Version 1.0
be re-surfaced, otherwise a0 = 1.0)
! Roughness
After widening works, roughness is reset to a user-specified value. If this is not specified,
the value of roughness is obtained as follows:
❏ If the existing carriageway is to be re-surfaced the following values are used as
defaults:
For AM surface type: RIaw = 2.0 (IRI m/km)
For ST surface type: RIaw = 2.8 (IRI m/km)
❏ If the existing carriageway is not to be re-surfaced, it is assumed that patching and
crack sealing that may be performed on the existing carriageway would affect the
roughness after widening works as follows:
 (RIn * ∆CW + CWbw * RI ap )

RI aw =   ...(6.29)
 CWaw 
RI ap = RIbw - MIN {[a0 * (∆CRX w + ∆ACTw ) + ∆RI t ], a1} ...(6.30)
where:
RIaw roughness after widening works (IRI m/km)
RIbw roughness before widening works (= RIb) (IRI m/km)
RIn user-specified roughness for new construction, (default = 2.0 for AM, and
2.8 for ST)
RIap roughness after patching and crack sealing (IRI m/km)
∆CRXw reduction in the area of indexed structural cracking, (= ACXb - ACXaw)
∆ACTw reduction in the area of transverse thermal cracking (= ACTb)

a0 and a1 user-definable model coefficients (default = 0.0066 and 4.6 respectively)
The value of ∆RIt is computed using Equations 3.20 above to 3.22 above by taking
∆NPTw as equal to NPTbw (= NPTb), since all existing potholes should be patched.
After widening, texture depth and skid resistance is reset to user-specified values. If these
are not specified, the values of texture depth and skid resistance after works are obtained
in the following manner:
❏ The existing carriageway is to be re-surfaced
Texture depth after works (TDaw) is reset to the default value of initial texture depth
given in Table D2.6; and skid resistance after works (SFCaw) is reset to the following
default values:
For surface type AM: SFCaw = 0.5
For surface type ST: SFCaw = 0.6

Version 1.0
❏ The existing carriageway is not to be re-surfaced

The values of texture depth and skid resistance after works are computed as follows:
 (TD bw * CWbw + ITD * ∆CW ) 

TD aw =   ...(6.31)
 CWaw 
 (SFC bw * CWbw + SFC n * ∆CW ) 

SFC aw =   ...(6.32)
 CWaw 
where:
TDaw texture depth after widening works (mm)

TDbw texture depth before widening works (mm)
ITD default initial texture depth as given in Table D2.6
SFCaw skid resistance after widening works, SCRIM
SFCbw skid resistance before widening works, SCRIM
SFCn default skid resistance at 50 km/h (= 0.5 for AM, and 0.6 for ST)
! Previous cracking
❏ If the existing carriageway is to be re-surfaced

The amounts of previous cracking (PCRA, PCRW, and PNCT) are reset as follows:
if CRAib ≥ PCRibw
 (CWbw * CRAi bw ) 
PCRi aw =   ...(6.33)
 CWaw 
if CRAi b < PCRi bw
 CWbw * [w * CRAi b + (1 − w ) * PCRi bw ]

PCRi aw =   ...(6.34)
 CWaw 
The weighting factor (w) is obtained in the following manner:

If the re-surfacing is an overlay (that is, surface type AM) consider roadbase types
as follows:
for roadbase types AB, AP, GB:
 HSNEWbw  
w = MAX  , 0.6  ...(6.35)
 HSOLD aw  
for roadbase type SB:
 HSNEWbw  
w = MAX   , 0.6  ...(6.36)
 (HSOLD aw + HSBASE )  
If the re-surfacing is a reseal (that is, surface type ST)

Version 1.0
w = MIN(0.70 + 0.1 * HSNEWaw , 1) ...(6.37)
where:
PCRiaw amount of previous cracking type i (i = all structural cracking, wide

structural cracking, or transverse thermal cracking) after works
CRAib amount of cracking type i at the end of the year
PCRibw amount of previous cracking type i before widening works
w weighting used for averaging the cracking in the old and new
surfacing layers
HBASE thickness of the roadbase layer in the original pavement (required for
the roadbase type SB only) (mm)
HSNEWaw thickness new surfacing after widening works (mm)
❏ If the existing carriageway is not to be re-surfaced, the area of previous cracking

(PCRA, PCRW, and PNCT) is reset as follows:
 (CWbw * PCRi bw ) 
PCRi aw =   ...(6.38)
 CWaw 

! Pavement age
The pavement ages after widening are reset as follows:

❏ If the existing carriageway is to be re-surfaced by an overlay, AGE1, AGE2 and
AGE3 are reset to zero. AGE4 is calculated from the expression:
 (CWbw * AGE4 bw ) 
AGE4 aw =   (returns an integer value) ...(6.39)
 CWaw 
❏ If the existing carriageway is to be resurfaced by resealing, AGE1 and AGE2 are

reset to zero. AGE3 and AGE4 are given as:
(CWbw * AGEi bw )
AGEi aw = (for i = 3 or 4) ...(6.40)
CWaw
❏ If the existing carriageway is not to be re-surfaced, the pavement ages are calculated
as follows:
(CWbw * AGEi bw )
AGEi aw = ...(6.41)
CWaw
where:
AGEiaw AGE type i (i = 1, 2, 3 or 4) after widening works (years)

Version 1.0
AGEibw AGE type i before widening works (years)
factors (that is, the K factors) are also reset to user-specified values.
! Speed factors
These are the speed limit, speed enforcement factor, roadside friction factor, non-
motorised transport and motorised transport speed reduction factors and acceleration
noise, which depend primarily on the individual road section.
! Traffic flow pattern (road use)
The data that describes the hourly traffic flow distribution are also reset to a user-
specified type.
6.3 Realignment
Refers to local geometric improvements of road sections, which may also result in a reduction
of the road length. It is assumed that the carriageway width remains unaltered when a
realignment works is performed.
Realignment works can be defined as follows:
A realignment works of fixed specifications is applied at a specified time, defined by a

calendar year.
A realignment works of fixed specifications is applied when the levels of the user-
specified intervention criteria, based on road user effects parameters, are met.
Realignment is not performed if the last applicable year has been exceeded.
A realignment works is specified using the following:
! New road type
! Road class
! Proportion of new construction - defined as the ratio of new construction length to the
section length after realignment works
! Road geometry for the whole section
! Pavement details of the new realigned segments
! Whether or not the non-realigned segments (or parts) of the existing carriageway is
provided with a new surfacing (that is, re-surfaced)

If performed, the total amount of realignment works is given by:

Version 1.0
REAL = Pconew * L aw ...(6.42)
L aw = L bw * LF ...(6.43)
where:
REAL length of road realigned (km)

Law section length after realignment works (km)
Lbw section length before realignment works (km)
LF length adjustment factor, (LF > 0)
Pconew proportion of new construction, (0 < Pconew < 1)
The cost of realignment (CSTREAL) is obtained from the product of REAL and the user-
specified unit cost per kilometre.
SALVA = PCTSAV * CSTREAL ...(6.44)
where:

Additional works
It is likely that realignment works will include re-surfacing or at least repair of the severely
damaged area of the existing carriageway segments that are not to be realigned. These
additional works are modelled as described below.
Case 1: re-surfacing the non-realigned segments
If the non-realigned segments of the existing carriageway are to be re-surfaced either by an
overlay or resealing, the amount of re-surfacing works for the entire road section is given by:
TANSF = 1000 * CW * Law * (1 - Pconew ) ...(6.45)
where:
TANSF total carriageway area re-surfaced (m2)
! If an overlay were provided over the non-realigned parts of the existing carriageway,
TANSF would be denoted as TAOVL, and the total cost of overlay is obtained from the
product of TAOVL and the unit cost of overlay per square metre.
Before overlay is performed, the amounts of preparatory works to be performed for the
entire road section are computed as follows:

Version 1.0
❏ Patching
TAPAT = 10 * CW * APOTb * L aw * (1 - Pconew ) ...(6.46)
where:
the unit cost of patching.
❏ Edge-repair
TAVER = 10 * CW * AVEB b * Law * (1 - Pconew ) ...(6.47)
where:
TAVER total area of edge-repair (m2)

AVEBbw area of edge-break before realignment works (= AVEBb) (% of total
carriageway area)
The total cost of edge-repair is obtained from the product of TAVER and the unit
cost of edge-repair.
! If the non-realigned parts of the existing carriageway were resealed, ANSF would be
denoted by TARSF, and the total cost of resealing is obtained from the product of
TARSF and the unit cost of resealing per square metre.
Before resealing is applied, the amounts of preparatory works to be performed for the
entire road section are computed as follows:
❏ Patching
TAPAT = 10 * CW * Law * (1 - Pconew ) * {MAX [0.1 * (ACWb − 20 ), 0] + APOTb }
...(6.48)
where:
❏ Crack sealing
It is assumed that 100% of transverse thermal cracking on the non-realigned parts of

the existing carriageway is sealed, and the amount of crack sealing performed is
given by:
TACSL = 10 * CW * L aw * (1 - Pconew ) * ACTbw ...(6.49)
where:

Version 1.0
TACSL total area of crack sealing (m2)

ACTbw area of transverse thermal cracking before realignment works
(=ACTb) (% of total carriageway area)
The total cost of crack sealing is obtained from the product of TACSL and the unit
cost of crack sealing per square metre.
❏ Edge-repair
The total area of edge-repair is calculated using Equation 6.47 above. The total cost
of edge-repair is obtained from the product of TAVER and the unit cost of edge-
repair per square metre.
Case 2: no re-surfacing of the non-realigned segments
If the non-realigned parts of the existing carriageway are not to be re-surfaced, it is assumed
that the following amounts of patching, crack sealing and edge-repair are performed along
with the realignment works:
! Patching
Assuming that 100% of severely damaged area is patched, the amount of patching is
TAPAT = 10 * CW * L aw * (1 - Pconew ) * ADAMS bw ...(6.50)
where:
TAPAT total carriageway area patched (m2)

ADAMSbw severely damaged carriageway area before realignment works
(=ADAMSb) (% of total carriageway area)
! Crack sealing
The total amount of crack sealing is calculated using Equation 6.49 above.
! Edge-repair
Assuming that all the edge-break area is repaired, the amount of edge-repair works is
The total cost of realignment works is the sum of carriageway realignment cost and the cost of
additional works comprising re-surfacing of the existing carriageway and preparatory
works. The amount and cost of re-surfacing works are reported separately under
resealing or overlay. Similarly, the amounts and costs of patching, of crack sealing and edge-
repair are reported separately under patching, crack sealing and edge-repair,
respectively.
In economic analysis, it is assumed that the costs of all these additional works are incurred in
the last construction year.

Version 1.0
6.3.2 Effects of realignment

After realignment, the pavement type is reset to the new type specified by the user. The
required modelling parameters are also reset to user-specified values, or computed
endogenously as described below:
! New length
The new length of the road section after realignment is given by Equation 6.43 above.
! Thickness of surfacing layers
❏ Case 1: re-surfacing non-realigned segments
If the non-realigned parts of the existing carriageway are to be re-surfaced, the

thickness of new surfacing after realignment works is obtained as follows:
HSNEWaw = [(1 - Pconew ) * HRESF + Pconew * HSNEWrw ] ...(6.51)
where:
HSNEWaw new surfacing thickness after realignment works (mm)

HSNEWrw surfacing thickness of the new construction parts of the carriageway
(mm)
HRESF user-specified thickness of the re-surfacing layer on the existing
carriageway (mm)
The thickness of old surfacing after realignment works is given by:
HSOLD aw = (1 - Pconew ) * HS bw ...(6.52)
where:
HSOLDaw thickness of old surfacing after realignment works (mm)

HSbw total surfacing thickness of the existing carriageway before
realignment (mm)
❏ Case 2: no resurfacing of non-realigned segments
If the non-realigned parts of the existing carriageway are not to be re-surfaced, the
thickness of new surfacing after realignment works is obtained as follows:
HSNEWaw = [(1 - Pconew ) * HSNEWbw + Pconew * HSNEWrw ] ...(6.53)
where:
HSNEWaw new surfacing thickness after realignment works (mm)

HSNEWbw new surfacing thickness before realignment works (mm)
HSNEWrw surfacing thickness of the new construction parts of the carriageway
(mm)

Version 1.0
The thickness of old surfacing after realignment works is given by:
HSOLD aw = (1 - Pconew ) * HSOLD bw ...(6.54)
where:

HSOLDbw thickness of old surfacing before realignment works (mm)
! Pavement strength
The dry season adjusted structural number of pavement is reset to the weighted
average of the structural number of the non-realigned parts of the existing carriageway
and that of the newly constructed segments, as follows:
SNPdaw = [(1 - Pconew ) * SNPdexcw + Pconew * SNPdrw ] ...(6.55)
SNPdrw = SN drw + SNSG s ...(6.56)
If the non-realigned parts of the existing carriageway are to be re-surfaced:
SNPdexcw = MAX [1.5, (SNPdbw + 0.0394 * a resf * HRESF - dSNPK )] ...(6.57)
If the non-realigned parts of the existing carriageway are not to be re-surfaced:
SNPdexcw = MAX [1.5, (SNPdbw + 0.0394 * a resf * HRESF - dSNPK )] ...(6.58)
where:

SNPdrw dry season adjusted structural number of pavement of the newly
constructed parts of carriageway
SNdrw user-specified structural number of pavement (of the layers above the
subgrade) of the realigned parts of the carriageway. This may be
computed in a similar manner to that described for SNnew in Equation 4.66
above
SNSGs subgrade contribution to the structural number of pavement, calculated
using Equation 4.67 above
dSNPK reduction in the adjusted structural number of pavement due to cracking
HRESF thickness of the re-surfacing layer on the non-realigned parts of the
existing carriageway
aresf strength coefficient of the re-surfacing layer on the non-realigned parts of
the existing carriageway
The Benkelman beam deflection after realignment works is given by:

Version 1.0
−1.6
 SNPaw 
DEFaw = DEFbw *   ...(6.59)
 SNPbw 
where:

DEFbw Benkelman beam deflection before works (mm)
! Surface material
If the existing carriageway is to be re-surfaced, the surface material after works is reset to
that specified for the realignment works. This is based on the assumption that the same
surfacing material is used for the re-surfacing.
If the existing carriageway is not to be re-surfaced, the surface material after works is
reset as follows:
if Pconew is less than 0.5, the surface material after works is reset to that of
the existing carriageway.
otherwise the surface material after works is reset to that of the realigned parts of the
carriageway.
The construction defect indicators for bituminous surfacing (CDS) and the construction
defect indicator for the roadbase (CDB) is reset to a weighted average computed as
follows:
CDi aw = [CDi bw * (1 - Pconew ) + CDi rw * Pconew ] ...(6.60)
where:
CDiaw construction defect indicator i (i=CDS or CDB) after works

CDibw construction defect indicator i before works
CDirw construction defect indicator i specified for the realignment works
If the non-realigned parts of the existing carriageway are to be re-surfaced the surface
distresses (that is, edge-break, potholing, cracking and ravelling) are all reset to zero.
If the non-realigned parts of the existing carriageway are not to be re-surfaced, the area
of edge-break, potholing, transverse thermal cracking, wide structural cracking and
ravelling are reset to zero. The area of all structural cracking and total area of cracking
after realignment works are reset as follows:
[(1 - Pconew ) * (ACA b - ACWb )]

ACA aw = ...(6.61)
LF

Version 1.0
ACRA aw = ACA aw ...(6.62)
where:
ACAaw area of all structural cracking after realignment works (% of total

carriageway area)
ACRAaw total area of cracking after realignment works (% of total carriageway
area)

! Rutting
The mean rut depth is reset to a user-specified value. If this is not specified, the mean rut
RDM aw = [(1 - Pconew ) * a0 * RDMbw ] ...(6.63)
where:
RDMaw mean rut depth after realignment works (mm)

RDMbw mean rut depth before realignment works (mm)
a0 user-definable coefficient (default = 0.15 if the non-realigned parts of the
existing carriageway are to be re-surfaced, otherwise a0 = 1.0)
! Roughness
After realignment works, roughness is reset to a user-specified value. If this is not

specified, the value of roughness is obtained as follows:
❏ If the non-realigned parts of the existing carriageway are to be re-surfaced the
following values are used as defaults:
❏ If the non-realigned parts of the existing carriageway are not to be re-surfaced,
roughness after realignment works is reset as follows:
RI aw = [RIn * Pconew + (1 - Pconew ) * RI ap ] ...(6.64)
RI ap = RIbw - MIN {[a0 * (∆CRX w + ∆ACTw ) + ∆RI t ], a1} ...(6.65)
where:
RIaw roughness after realignment works (IRI m/km)
RIbw roughness before realignment works (= RIb) (IRI m/km)
RIn user-specified roughness for realignment new construction,
(default = 2.0 for AM, and 2.8 for ST)
RIap roughness after patching and crack sealing (IRI m/km)

Version 1.0
∆CRXw reduction in the area of indexed structural cracking,

(= ACXb - ACXaw)
∆ACTw reduction in the area of transverse thermal cracking (= ACTb)

a0 and a1 user-definable model coefficients (default = 0.0066 and 4.6
respectively)
The value of ∆RIt is computed using Equations 3.20 above to 3.22 above by taking
∆NPTw as equal to NPTbw (= NPTb), since all existing potholes should be patched.
After realignment, texture depth and skid resistance is reset to user-specified values. If
these are not specified, texture depth and skid resistance after works are obtained in the
following ways:
❏ If the non-realigned segments of the existing carriageway are to be re-surfaced,
texture depth after works (TDaw) is reset to the default value of initial texture depth
given in Table D2.6; and skid resistance after works (SFCaw) is reset to the following
default values:
❏ If the non-realigned segments of the existing carriageway are not to be re-surfaced,
the values of texture depth and skid resistance after works are computed as follows:
TD aw = TD bw * (1 - Pconew ) + ITD * Pconew ...(6.66)
SFC aw = SFC bw * (1 - Pconew ) + SFC n * Pconew ...(6.67)
where:
TDaw texture depth after realignment works (mm)

TDbw texture depth before realignment works (mm)
ITD default initial texture depth as given in Table D2.6
SFCaw skid resistance after realignment works, SCRIM
SFCbw skid resistance before realignment works, SCRIM
SFCn default skid resistance at 50 km/h (= 0.5 for AM, and 0.6 for ST)
! Previous cracking
If the non-realigned parts of the existing carriageway are to be re-surfaced, the amount of
previous cracking (PCRA, PCRW, and PNCT) is reset as follows:
❏ if CRAi b ≥ PCRi bw
[(1 - Pconew ) * CRAi b ]

PCRi aw = ...(6.68)
LF

Version 1.0
❏ if CRAi b < PCRi bw
{(1 - Pconew ) * [w * CRAi b + (1 − w ) * PCRi bw ]}

PCRi aw = ...(6.69)
LF
The weighting factor (w) is obtained in the following manner:

If the re-surfacing is an overlay (that is, surface type AM), the value of w is obtained as
follows:
❏ for roadbase types AB, AP, GB
 HSNEWbw 
w = MAX  , 0.6  ...(6.70)
 HSOLD aw 
❏ for roadbase type SB
 HSNEWbw 
w = MAX  , 0.6  ...(6.71)
If the re-surfacing is a reseal (that is, surface type ST), the value of w is given by:
w = MIN (0.70 + 0.1 * HSNEWaw , 1) ...(6.72)
where:
PCRiaw amount of previous cracking type i (i = all structural cracking, wide

structural cracking or transverse thermal cracking) after works
CRAib amount of cracking type i at the end of the year
PCRibw amount of previous cracking type i before realignment works
LF length adjustment factor
w weighting used for averaging the cracking in the old and new surfacing
layers
HBASE thickness of the roadbase layer in the original pavement (required for SB
roadbase type only) (mm)
HSNEWaw thickness of new surfacing after realignment works (mm)
If the non-realigned parts of the existing carriageway are not to be re-surfaced, the
amount of previous cracking (PCRA, PCRW, and PNCT) is reset as follows:
 [(1 - Pconew ) * PCRi bw ]

PCRi aw =   ...(6.73)
 LF 

! Pavement age
The pavement ages after realignment are reset as follows:

Version 1.0
❏ If the non-realigned parts of the existing carriageway are to be re-surfaced by an

overlay, AGE1, AGE2 and AGE3 are reset to zero. AGE4 is calculated as follows:
AGE4 aw = (1 - Pconew ) * AGE4 bw (returns an integer value) ...(6.74)
❏ If the non-realigned parts of the existing carriageway are to be re-surfaced by

resealing, AGE1 and AGE2 are reset to zero. AGE3 and AGE4 are calculated from
the expression:
AGEi aw = (1 - Pconew ) * AGEi bw (i = 3 or 4) ...(6.75)
❏ If the non-realigned parts of the existing carriageway are not to be re-surfaced, the
pavement ages are calculated as follows:
AGEi aw = (1 - Pconew ) * AGEi bw ...(6.76)
where:
AGEiaw AGE type i (i = 1, 2, 3 or 4) after realignment works (years)

AGEibw AGE type i before realignment works (years)
! Deterioration factors
factors (that is, the K factors) are also reset to user-specified values.
! Speed factors
motorised transport and motorised transport speed reduction factors, and acceleration
noise which depend primarily on the individual road section.
specified type.
6.4 Junction improvement

This operation refers to the improvement of a road intersection from one type to another, for
example, from a cross-priority junction to a roundabout. It is required for the purpose of road
safety effects analysis.
Note: This feature is not included in this software release.
A junction improvement works is defined as follows:
A junction improvement works of fixed specifications is applied at a specified time,

defined by a calendar year.
A junction improvement works of fixed specifications is applied when the predicted

number of accidents reaches a critical user-specified threshold level. Junction
improvement is not performed if the last applicable year has been exceeded.

Version 1.0
A junction improvement works is specified by the new intersection type (to which a different
set of accident rates apply), the duration of works and the cost involved in terms of currency
per year.
6.5 Non-carriageway improvement

The major non-carriageway improvement works considered relate to shoulders, NMT lanes,
and side-drains.
6.5.1 Shoulders
The modelling of shoulders improvement works is not included in this release. It will be
implemented when the specifications are finalised. The improvement works for shoulders
includes the following:
! Shoulders addition
! Shoulders upgrading
6.5.2 NMT lanes

The effects of Non-Motorised Transport (NMT) lane upgrading and NMT lane addition are
modelled as described below:
NMT lane upgrading

This involves upgrading the pavement type of the existing NMT lanes to a new type. The
upgrading of NMT lanes can be defined as follows:
An NMT lane-upgrading works of fixed specifications is performed at a specified time,

defined by a calendar year.
An NMT lane upgrading works of fixed specifications is performed when the user-
specified intervention criteria, based on road user effects parameters, are met. Upgrading
is not performed if the last applicable year has been exceeded.
The following information is required to specify upgrading NMT lanes:
! New road type
! New surface type
It is assumed that the geometry details are the same as that of the existing road section.
If performed, the total amount of works is given in terms of the total length of NMT lanes
constructed as follows:
NMTL aw = L * NMTLN ...(6.77)
where:
NMTLaw total length of NMT lanes upgraded (km)


Version 1.0
NMTLN number of existing NMT lanes
The total cost of NMT lane upgrading is obtained from the product of NMTLaw and the user-
specified unit cost per lane per kilometre.
The effects of upgrading NMT lane is reflected by specifying new values for the speed
reduction factors, and resetting the data that describe the resulting traffic flow pattern.
NMT lane addition

This includes the addition of NMT lanes of all types, and can be defined as follows:
An NMT lane(s) of fixed specifications is constructed at a specified time defined by a

calendar year.
An NMT lane(s) of fixed specifications is constructed when the levels of the user-
specified intervention criteria, based on road user effects parameters, are met.
Construction is not performed if the last applicable year has been exceeded.
The addition of new NMT lanes is specified by the number of lanes, the road type and the
pavement surface type. It is assumed that the geometry details are the same as that of the road
section.
If performed, the total amount of works is given in terms of the total length of NMT lanes
constructed as follows:
NMTL aw = L * NEWLN ...(6.78)
where:
NMTLaw total length of NMT lanes constructed (km)

NEWLN number of new NMT lanes constructed
The total cost of NMT lane construction is obtained from the product of NMTLaw and the
user-specified unit cost per lane per kilometre.
The effects of NMT lane addition should be reflected by specifying new values for the speed
reduction factors, and resetting the data that describe the resulting traffic flow pattern.
6.5.3 Side drain improvement

The effect of drainage on pavement performance depends to some extent on the type of side
drains provided. For example, a fully lined and linked drain type would give a better
performance (if maintained in a good condition) than would a shallow-soft drain.
A side drain improvement works is defined as scheduled, in which the improvement works of
fixed specifications is applied at a specified time, defined by a calendar year.
A side drain improvement works is specified by the new drain type (to which a different set of
drainage factors, and annual drainage deterioration factors applies) and the duration.

Version 1.0
The total cost of side drain improvement is obtained from the product of road section length
and the unit cost given in terms of currency per kilometre per year.

Version 1.0
7 Construction
Construction works comprises the following:
! Upgrading (see Section 7.1)
! Dualisation (see Section 7.2)
! Construction of a new section (or link) (see Section 7.3)
7.1 Upgrading
This operation involves both pavement upgrading and geometric improvements of an existing
road. Normally pavement upgrading would change the existing surface class to another
surface class of a higher performance grade. For example, a bituminous pavement road may
be upgraded to a rigid concrete pavement road. The modelling framework also allows for the
upgrading of a lower-grade bituminous pavement to a higher-grade bituminous pavement.
Upgrading works can be specified in one of the following ways:
An upgrading works of fixed specifications is applied at a specified time, defined by a

calendar year.
An upgrading works of fixed specifications is applied when the levels of the user-
specified intervention criteria, based on road user effects parameters, are met. Upgrading
is not performed if the last applicable year has been exceeded.
An upgrading works is specified by the:
! Road type
! Road class
! Road geometric characteristics
! Pavement details
! Increase in width

If performed, the total amount of upgrading works is given by:
LUPGRD = L aw ...(7.1)
L aw = LF * L bw ...(7.2)
where:

Version 1.0
LUPGRD amount of upgrading works (km)

Law road length after upgrading works (km)
Lbw road length before upgrading works (km)
LF length adjustment factor (LF > 0)
The total cost of upgrading (CSTUPGRD) is obtained from the product of LUPGRD and the
user-specified unit cost per kilometre.
SALVA = PCTSAV * CSTUPGRD ...(7.3)
where:

7.1.2 Effects of upgrading

After upgrading, the required modelling parameters are reset to user-specified values, or
computed endogenously as described below:
! New length
The new length of the road section after upgrading is given by Equation 7.2 above.
! Carriageway width
The new carriageway width after upgrading is given as follows:
CWaw = CWbw + ∆CW ...(7.4)
where:
CWaw carriageway width after upgrading works (m)

CWbw carriageway width before upgrading works (m)
The number of lanes after upgrading (NLANESaw) is equal to the number of lanes before
works (NLANESbw) plus the user-specified additional number of lanes (ADDLN).
The effective number of lanes after works is reset to a new value ELANESaw, which
should be input by the user. If this is not specified, ELANESaw will be reset to the value of
NLANESaw (that is, the number of lanes for the road section following upgrading) see
Part B Section 5.2.
! Pavement characteristics
The new pavement characteristics are reset according to the new road surface class
specified by the user as described in Sections 7.1.3 and 7.1.4.

Version 1.0
7.1.3 Bituminous pavements

If the new road surface class is bituminous, the pavement type, structure, strength and layer
material properties are reset to the user-specified values or calculated endogenously.
Note: In this case the pavement types STAP and AMAP are not permissible options.
! Pavement strength
The dry season adjusted structural number of pavement (SNPdaw) is computed using the
specified structural number of pavement (SNnew) from Equations 4.66 above and 4.67
above. The Benkelman beam deflection after upgrading works will be given as:
−1.6
DEFaw = a0 * (SNPaw ) ...(7.5)
where:

a0 coefficient depending upon the roadbase type (a0 = 6.5 for types GB and
AB; a0 = 3.5 for SB base type)
After upgrading, the pavement condition is new, and all distress values are therefore set
to zero.
! Rutting
After upgrading, the mean rut depth is set to zero.

! Roughness
Roughness after upgrading is set to a user-specified value. If this is not specified, the
following default values are used:
After upgrading, texture depth and skid resistance is reset to the user-specified values. If
these are not specified, texture depth after works (TDaw) is reset to the default value of
initial texture depth given in Table D2.6; and skid resistance after works (SFCaw) is reset
to the following default values:
! Pavement ages
After upgrading, all the pavement age parameters (that is, AGE1, AGE2, AGE3 and
AGE4) are reset to zero.
! Previous cracking
The amounts of previous cracking are all reset to zero.

Version 1.0
The deterioration factors for modelling the performance of the new pavement are reset to
user-specified values.
! Speed factors
These are the speed limit, speed enforcement factor, acceleration noise, roadside friction
factor, non-motorised transport factor, and the speed reduction factor due to motorised
transport which depend primarily on the individual road section.
specified type.
7.1.4 Rigid concrete pavements

If the new surface class is rigid concrete, the required modelling parameters for concrete
pavement are used. The pavement condition is set to new, and the history data is reset to
reflect a completely new construction.
The new section length, carriageway width, number of lanes, speed factors and traffic flow
pattern are reset as described above for bituminous pavements.
7.2 Dualisation
The dualisation of an existing road section can be either scheduled or triggered by responsive
intervention criteria.
Note: The modelling of dualisation has not been included in this release.
7.3 New section

The construction of a new section can only be scheduled and not triggered by responsive
intervention criteria. In a project analysis, a new road section (that is, a new link) can be
specified as a section alternative within a selected project alternative.
The required components for defining a new section are as follows:
! All the data items that are required to define a road section in HDM-4. The user can
specify these items using aggregate data. Note that the pavement types STAP and AMAP
are not valid options.
! Traffic data
❏ diverted traffic - traffic that is diverted from the nearby routes and other transport
modes.
❏ generated traffic – additional traffic that occurs in response to the new investment.
! Construction costs, duration and salvage value.
! Exogenous benefits and costs.
! Maintenance standards, and future improvements to be applied after opening the new
section to traffic.
The amount of new construction can be expressed in terms of the number of kilometres
constructed (NEWCON) and this is equal to the new section length. The total cost of

Version 1.0
construction (CSTNEW) is obtained from the product of NEWCON and the user-specified
unit cost per kilometre.
The salvage value is given as:
SALVA = PCTSAV * CSTNEW ...(7.6)
where:


Version 1.0
8 References
NDLI, (1995)
Modelling Road Deterioration and Maintenance Effects in HDM-4. Final Report Asian
Development Bank Project RETA 5549. N.D. Lea International, Vancouver, Canada
Riley M., (1995)
Framework for Defining the Effects of Works on Road parameters in HDM-4.
Tsunokawa K., (1987a)
The Highway Design and Maintenance Standards Model: Volume-1 Description. The
Tsunokawa K., (1987b)
The Highway Design and Maintenance Standards Model: Volume- 2 User's Manual

Version 1.0
D3 Concrete Pavements
1 Introduction
This chapter describes the detailed modelling of road works effects for rigid concrete roads in
HDM-4. It is based on the specification documents prepared by the Latin American Study
Team in Chile (LAST, 1995; and LAST, 1996).
! Routine maintenance
! Periodic maintenance
! Special works
! Construction works
hierarchical ranking of works activities and pavement type resets after works. The background
of the modelling logic is given in LAST (1996).
Table D3.1 shows the maintenance works activities for the different pavement surface types
considered in HDM-4.
Table D3.1 Maintenance works for concrete pavements
Works Works type Works activities Pavement surface type

class
JP JR CR
Routine Routine Vegetation control, line marking, drain a a a
maintenance cleaning, etc.
Load transfer dowels retrofit a

Preventive Tied concrete shoulders retrofit a a
treatment
Longitudinal edge drains retrofit a a
Joint sealing a a
Slab replacement a
Periodic Full depth repair a a
Restoration
Partial depth repair a
Diamond grinding a a
Bonded concrete overlay a a a
Rehabilitation
Unbonded concrete overlay a a a
Reconstruction Pavement reconstruction a a a
Notes:
a indicates that the works activity is applicable to the pavement surface type

Version 1.0
PART D ROAD W ORKS EFFECTS D3 CONCRETE PAVEMENTS
JP Jointed plain
JR Jointed reinforced
CR Continuously reinforced

Version 1.0
2 Modelling logic
The modelling of concrete pavement performance is considered in two separate phases:
! Phase 1
Refers to the time before any major periodic maintenance or reconstruction.

! Phase 2
Refers to the time after the pavement has received a major maintenance or has been
reconstructed.
This chapter discusses the Phase 2 modelling logic. Phase 1 modelling logic is described in
Chapter C3.

The overall computational procedure for modelling road works that is applied in each analysis
year can be summarised by the following steps:
1 Determine the road works standard(s) that is applicable in the given year. Only one
maintenance standard and/or one improvement standard can be applied to a road feature
in any analysis year.
2 Check the intervention criteria and the limits defined for works included within the
standard. A scheduled operation takes priority over a responsive operation of the same
type.
3 Identify and apply the highest-priority works activity.
4 Compute the physical quantities of works.
5 Compute works effects and reset modelling parameter values to reflect post-works road
6 Check intervention criteria, identify and apply the next highest-priority works activity.
7 Repeat (4), (5) and (6).
8 Apply any other works whose effects on pavement performance are not modelled
endogenously in HDM-4, (for example, routine maintenance works).
9 Calculate the costs of works by applying unit costs to the physical quantities of works.
10 Store results for economic analysis and for use in the following analysis year.

A works activity (or an operation) will be triggered when any one or a combination of the
user-specified criteria has been met. When more than one works activity meet the criteria for
being applied in a given analysis year, the highest-priority operation for the particular road
feature will be applied first, and their effects will be computed to reset road characteristics.
Most interventions on concrete pavements will require a combination of several preventive
and restoration works activities. The activities defined within a standard should therefore
address properly the causes of pavement deterioration and repair them.

Version 1.0
Table D3.2 shows the hierarchy of works activities that are applicable to the carriageway for
JP concrete pavements.
Table D3.2 Maintenance works applicable to JP concrete carriageway
Works type Works activity / operation ID code Hierarchy Unit cost

Reconstruction Pavement reconstruction REC 1 per m2
Unbonded concrete overlay UOL 2 per m2

Rehabilitation
Bonded concrete overlay BOL 3 per m2
Slab replacement SLR 4 per m2
Restoration Partial depth repair PDR 5 per m (joint length)
Diamond grinding* DGR 6 per m2
Load transfer dowels retrofit* DWL 7 per m (joint length)
Preventive Tied concrete shoulders retrofit* TCS 7 per km

treatment
Longitudinal edge drains retrofit* RED 7 per km
Joint sealing* SLJ 7 per m (joint length)
Note:
* Works activity can be applied together with slab replacement or partial depth repair
in the same analysis year
JR concrete pavements.
Table D3.3 Maintenance works applicable to JR concrete carriageway

2
Reconstruction Pavement reconstruction REC 1 per m
Rehabilitation Unbonded concrete overlay UOL 2 per m2
Full depth repair FDR 4 per m2

Restoration
Diamond grinding* DGR 5 per m2
Tied concrete shoulders retrofit* TCS 6 per km
Preventive Longitudinal edge drains retrofit* RED 6 per km

treatment
Joint sealing* SLJ 6 per m (joint length)
Note:
* Works activity can be applied together with full depth repair in the same analysis
year

Version 1.0
CR concrete pavements.
Table D3.4 Maintenance works applicable to CR concrete carriageway

Reconstruction Pavement reconstruction REC 1 per m2
Unbonded concrete overlay UOL 2 per m2

Rehabilitation
Restoration Full depth repair FDR 4 per m2
The works activities whose effects on pavement performance are not modelled endogenously
in HDM-4 (for example, emergency works, winter maintenance, and routine maintenance
works) will be applied in a given analysis year, if specified by the user, regardless of any
works hierarchy.
2.3 Road characteristics reset

When a road works activity is performed, two kinds of effects on the road characteristics are
considered:
1 The immediate effect on the road geometry, pavement type, strength, condition, age (or
history), status, etc.
2 The long-term effect on pavement performance.
The RD model parameters are reset to reflect the effects of the works activity and are
subsequently used to predict the long-term performance of the pavement.
This chapter discusses in detail the immediate effects of different works activities. The long-
term effect on pavement performance is modelled using the Phase 1 models described in
Chapter C3, with some adjustments introduced to shift the performance curve either vertically,
horizontally or both in order to maintain continuity at the point of time the intervention was
applied.

Version 1.0
3 Routine maintenance
The routine maintenance activities on concrete roads include vegetation control, drain
cleaning, road marking, repair to road appurtenances, and sign posts. These are considered in
the analyses only in terms of their costs to the road administration. The effect of drainage on
pavement performance is accounted for through the drainage coefficient (Cd), as defined by
AASHTO (1993).
Routine maintenance works should be scheduled at a fixed interval of time (a minimum of
one-year), and will be performed on an annual basis. If specified by the user, routine
maintenance works will be applied in any analysis year regardless of the works hierarchy.
The unit cost should be specified in terms of currency per kilometre per year, and the annual
cost of routine maintenance is obtained from the product of the road section length and the
unit cost.

Version 1.0
4 Preventive treatment
The preventive maintenance works on concrete pavements comprises the following:
! Load transfer dowels retrofit (see Section 4.1).
! Tied concrete shoulders retrofit (see Section 4.2).
! Longitudinal edge drains retrofit (see Section 4.3).
! Joint sealing (see Section 4.4).
4.1 Load transfer dowels retrofit

This is a technique used to increase the load transfer efficiency of JP concrete pavements by
cutting slots and inserting dowels at undoweled joints (and cracks).
Retrofitting load transfer dowels can be scheduled at a fixed point in time defined by the
calendar year.
Quantities and costs

If performed, the amount of works is given by:
5280 * CW
LDOW = ...(4.1)
1.6093 * JTSPACE
where:
LDOW joint length fitted with load transfer dowels (m/km)

The product of LDOW and the section length (L) in kilometres gives the total length of joints
fitted with load transfer dowels (TLDOW) in metres. The total cost of retrofitting load transfer
dowels is computed by multiplying TLDOW by the user-specified unit cost per metre length.
The effects
Load transfer restoration has the ability to increase the structural capacity of a pavement.
After works, the pavement type will be reset to JP concrete pavement with dowels, and the
progression of cracking, faulting and spalling will be computed using the models for JP
concrete pavements with dowels. If the dowel diameter is less than 20 mm, the model for JP
concrete pavement without dowels will be used. The user should also specify whether or not
the dowels fitted are protected against corrosion.
Figure D3.1 shows the effect of retrofitting load transfer dowels on faulting in JP concrete
pavements without dowels.

Version 1.0
0.40
0.35
0.30
Average Joint Faulting (in)
0.25
0.20
0.15
0.10
Original
0.05
Dowels Placement
0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00
Cumulative ESAL (millions per lane)
Figure D3.1 Retrofitting load transfer dowels effect on transverse joint faulting
in JP concrete pavements without dowels
4.2 Tied concrete shoulders retrofit

This is the addition of tied concrete shoulders to an existing JP or JR concrete pavement.
Retrofitting tied concrete shoulders can be scheduled at a fixed point in time defined by the
calendar year.

If performed, the amount of works is given as follows:
TCS = L ...(4.2)
where:
TCS amount of tied concrete shoulders retrofitted (km)

The total cost of retrofitting tied concrete shoulders is computed by multiplying TCS by the
The effects
Retrofitting tied concrete shoulders can increase the structural capacity of a pavement by
decreasing the critical stresses at the concrete edge and by decreasing the deflections at the
corners.
The value of the parameter LTEsh used in the equations for predicting cracking and faulting
will be updated as described in Table D3.5.

Version 1.0
Table D3.5 Cracking and faulting model parameter values
Pavement construction status Parameter
WIDENED LTEsh (%)

1 No widened lane or tied concrete shoulders 0 0
2 Widened outside lanes provided 1.0 20
3 Concrete shoulders placed during initial pavement construction 1.0 20
4 Concrete shoulders placed after initial pavement construction 0.5 10
Figure D3.2 shows the effect of retrofitting tied concrete shoulders on cracking in JP concrete
100%
90%
Load transfer efficiency (stress), LTEsh = 10 %
80%
Transverse cracking of slabs [%]
70%
60%
50%
40%
30%
Original
20%
Concrete shoulder, LTEsh = 10
10%
0%
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00
Figure D3.2 Retrofitting tied concrete shoulder effect on transverse cracking in

JP concrete pavements
Figure D3.3 shows the effect of retrofitting tied concrete shoulders on faulting in JP concrete

Version 1.0
0.40
0.35
0.30
0.25
0.20
0.15
0.10
..... Original
0.05 Concrete Shoulders
0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00
Figure D3.3 Retrofitting tied concrete shoulders effect on joint faulting in JP

concrete pavements without dowels
Figure D3.4 shows the effect of retrofitting tied concrete shoulders on faulting in JP concrete
pavements with dowels.
0.40
0.35
0.30
Average Joint Faulting
0.25
0.20
0.15
0.10
Original
0.05 Diamond Grinding
0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00
Figure D3.4 Retrofitting tied concrete shoulders effect on joint faulting in JP

concrete pavements with dowels
4.3 Longitudinal edge drains retrofit

This refers to the technique used to add longitudinal drains (or lateral drains) to an existing
pavement system to aid in the rapid removal of water from the system.

Version 1.0
Retrofitting longitudinal edge drains can be scheduled at a fixed point in time defined by the
calendar year.

If performed, the road section length gives the amount of works, as follows:
RED = L ...(4.3)
where:
RED amount of tied concrete shoulders retrofitted (km)

The total cost of retrofitting longitudinal edge drains is computed by multiplying RED by the
The effects
Since many pavement distresses can be attributed to water, removing it decreases the
opportunity for deterioration and increases the pavement’s life.
After retrofitting longitudinal edge drains, the parameter Cd (drainage coefficient) used in the
equation for calculating faulting should be reset to the user-specified value. The maximum
value of Cd is 1.25.
Figure D3.5 shows the effect of retrofitting longitudinal edge drains on faulting in JP concrete
0.40
0.35
0.30
0.25
0.20
0.15
0.10
Original, Cd = 0.5
0.05 Lateral Drains, Cd =1.0
0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00
Figure D3.5 Lateral drains placement on joint faulting in JP concrete

pavements without dowels

Version 1.0
Figure D3.6 shows the effect of retrofitting longitudinal edge drains on faulting in JP concrete
0.40
0.35
0.30
0.25
0.20
0.15
0.10
Original, Cd = 0.5
0.05
Lateral Drains, Cd =1.0
0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00
Figure D3.6 Lateral drains placement effect on joint faulting in JP concrete

pavements with dowels
4.4 Joint sealing

Joint sealing is applied to minimise infiltration of surface water and incompressible material
into the joint system. Minimising the amount of water in and under a pavement reduces the
potential for subgrade softening, pumping, and erosion of the subgrade/sub-base fines.
Minimising the amount of incompressible material reduces the potential for spalling and
blow-ups.
Joint sealing works can be defined in one of the following ways:
Joint sealing is applied at a fixed point in time defined by the calendar year.
A fixed interval between successive seals is specified, and the joint sealing is applied in
an analysis year t defined as follows:
t = interval + tprev ...(4.4)
tprev = max (tstart, tlast ) - tstart ...(4.5)
where:
t analysis year, defined relative to the start year of the analysis period
interval a fixed interval between successive joint sealings (years)
tprev number of years since last joint sealing or start year of analysis period

Version 1.0
tstart start year of analysis period, defined by calendar year

tlast last year in which joint sealing was applied, defined by calendar year

If performed, the amount of joint sealing is calculated as follows:
5280 * CW
LJSL = + (1000 * NLNGJTS ) ...(4.6)
1.6093 * JTSPACE
where:
LJSL amount of joint sealing (m/km)

NLNGJTS number of longitudinal joints
The product of LJSL and the section length (L) in kilometres gives the total length of joints
sealed (TLJSL) in metres. The total cost of joint sealing is computed by multiplying TLJSL
by the user-specified unit cost per metre length.
The effects
Joint sealing has an effect only on the progression of spalling. This effect is modelled through
the parameters AGE and seal type used in the spalling equation. Following joint sealing, the
seal type will be reset to the user-specified type and AGE will be reset to zero.
Figure D3.7 shows the effect of joint sealing on spalling in JP concrete pavements without
dowels.
16.00
14.00
12.00
10.00
Joints Spalling [%]
8.00
6.00
4.00
Original, Liquid Seal

2.00
Joints must be properly clean Liquid Seal
Preformed Seal
to obtain these effects.
0.00
0 5 10 15 20 25 30
Age [years]
Figure D3.7 Joint sealing effect on joint spalling in JP concrete pavements

Version 1.0
Figure D3.8 shows the effect of joint sealing on spalling in JR concrete pavements.
100
90
80
70
60
Joint Spalling [%]
50
40
30
20 Original
80% Joint Sealing
10
0
0 5 10 15 20 25 30
Age [years]
Figure D3.8 Joint sealing effect on joint spalling in JR concrete pavements

Version 1.0
5 Restoration works
The following four generic types of restoration works are modelled:
1 Slab replacement
2 Full depth repair
3 Partial depth repair
4 Diamond grinding
5.1 Slab replacement

Slabs that are extensively and severely damaged need to be replaced. This works activity
normally involves slab stabilisation to restore support to concrete slabs by filling small voids
that develop underneath the slab at joints, cracks, or the pavement edge. These voids (not
much deeper than 3 mm) are usually caused by pumping, consolidation, and subgrade failure
due to high deflections at the joints, cracks, and pavement edge.
Slab replacement works can be defined in one of the following ways:
A slab replacement works is performed at a fixed point in time defined by the calendar
year.
A fixed interval between successive slab replacement works is specified, and the slab
replacement works is applied in an analysis year t defined by Equations 4.4 above and
4.5 above.
Slab replacement is performed when the level of pavement defect (cracking) exceeds the
user-specified value.
In all cases, slab replacement will not be performed if the user-specified last applicable year or
A slab replacement works will be specified using the percentage of damaged slabs to be
replaced.

If performed, the amount of slab replacement works is calculated as follows:
0.3048 * JTSPACE * CW * REPSLBS

SLB = ...(5.1)
1.6093 * (NLNGJTS + 1)
PCRACK * PCTREP * NTSLBS

REPSLBS = ...(5.2)
(100 * 100 )
5280 * (NLNGJTS + 1)
NTSLBS = ...(5.3)
JTSPACE

Version 1.0
where:
SLB slab replacement area (m2/km)

REPSLBS number of replaced slabs per mile
PCTREP percent of cracked slabs to be replaced, input by the user
NTSLBS number of slabs per mile
NLNGJTS number of longitudinal joints
The product of SLB and the section length (L) in kilometres gives the total slab replacement
area (TSLB) in square metres. The total cost of slab replacement is computed by multiplying
TSLB by the user-specified unit cost per square metre.
The effects
The immediate effects of slab replacement on cracking, spalling and faulting are computed as
described below. The long-term effect on pavement distress progression is modelled by
applying weighting factors based on the proportions of the new and old slabs.
Cracking
The amount of cracking remaining after works is given by:
100 * (CRKSLBS - REPSLBS )

PCRACK aw = ...(5.4)
NTSLBS
where:
PCRACKaw percent of cracked slabs remaining after works

CRKSLBS number of cracked slabs per mile
The number of cracked slabs before works is given by:
PCRACK bw * NTSLBS
CRKSLBS = ...(5.5)
100
where:
CRKSLBS number of cracked slabs per mile

PCRACKbw percent of cracked slabs before works

Version 1.0
Figure D3.9 shows the effect of slab replacement on cracking in JP concrete pavements
without dowels.
100%
90%
80%
Transverse cracking of slabs [%]
70%
60%
50%
40%
30%
20% Original
40% Slabs Replacement
10%
0%
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00
Cumulative ESAL [millions per lane]
Figure D3.9 Slab replacement effect on transverse cracking of slabs in JP

concrete pavements
Spalling
The amount of spalling remaining after works is given by:
SPALL bw * (NTSLBS - REPSLBS )

SPALL aw = ...(5.6)
NTSLBS
where:
SPALLaw percentage of spalled transverse joints remaining after works

SPALLbw percentage of spalled transverse joints before works
Faulting
The amount of faulting remaining after works is given by:
FAULTbw * (NTSLBS - REPSLBS )

FAULTaw = ...(5.7)
NTSLBS
where:

Version 1.0
FAULTaw average transverse joint faulting after works (inches)

FAULTbw average transverse joint faulting before works (inches)
Figure D3.10 shows the effect of slab replacement on faulting in JP concrete pavements
without dowels.
0.40
0.35
0.30
0.25
0.20
0.15
0.10
Original
0.05 Slab replacement
0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00
Figure D3.10 Slab replacement effect on joint faulting in JP concrete

Figure D3.11 shows the effect of slab replacement on faulting in JP concrete pavements with
dowels.

Version 1.0
0.40
0.35
0.30
0.25
0.20
0.15
0.10
Original
0.05
Slabs replacement
0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00
Figure D3.11 Slab replacement effect on joint faulting in JP concrete

5.2 Full depth repairs

This technique is most often used on JR concrete pavements to repair joint deterioration that
extends more than one third of the slab depth. It consists of removing and replacing a portion
of the existing slab from the top right up to the bottom. Joint deterioration includes cracking,
breaking and spalling of slab edges on either side of a transverse or longitudinal joint or crack.
Full depth repairs are also applied to repair shattered slabs and punchouts in CR concrete
pavements.
Full depth repair can be defined in one of the following ways:
A full depth repair works is performed at a fixed point in time defined by the calendar
year.
A fixed interval between successive full depth repair works is specified, and the full
depth repair is applied in an analysis year t defined by Equations 4.4 above and 4.5
above.
Full depth repair is performed when the level of pavement defects (cracking, spalling and
failures) exceeds the user-specified values.
In all cases, full depth repair will not be performed if the user-specified last applicable year or
The amounts of works and effects of full depth repair are calculated depending on the
pavement surface type, as described below.

Version 1.0
5.2.1 JR concrete pavements

The amount of full depth repairs on JR concrete pavements is computed as follows:
(CKAREA + SPAREA )
FDR = ...(5.8)
1.6093
where:
FDR area of full depth repair (m2/km)

CRAREA area of deteriorated transverse cracks repaired (m2)
SPAREA area of spalling repaired (m2)
The area of deteriorated transverse cracks repaired is given by:
CW * PCTCKS * DCRACK bw * 2 * CKWDTH

CRAREA = ...(5.9)
100
where:
CRAREA area of deteriorated transverse cracks repaired (m2)

PCTCKS percentage of deteriorated transverse cracks to be repaired, input by the user
DCRACKbw number of deteriorated transverse cracks per mile before works
CKWDTH width considered on each side of a deteriorated transverse crack (m)
The area of spalling repaired is given by:
CW * PCTSPL * SPALL bw * NJTS * 2 * JTWDTH

SPAREA = ...(5.10)
100 * 100
where:
SPAREA area of spalling repaired (m2)

PCTSPL percentage of spalled transverse joints to be repaired, input by the user
NJTS number of transverse joints per mile
JTWDTH width considered on each side of a transverse joint (m)
The number of transverse joints per mile (NJTS) is given by:

Version 1.0
1.5 * 5280
NJTS = ...(5.11)
JTSPACE
where:

JTSPACE average spacing of transverse joints (ft)
The product of FDR and the section length (L) in kilometres gives the total area of full depth
repair (TFDR) in square metres. The total cost of full depth repair is computed by multiplying
TFDR by the user-specified unit cost per square metre.
The effects
Following full depth repair, pavement defects will be reduced as detailed below.
! Cracking
The amount of cracking remaining after works is given by:
DCRACK aw = 1 -
PCTCKS 
 * DCRACK bw ...(5.12)
 100 
where:
DCRACKaw number of deteriorated transverse cracks per mile remaining after

works
PCTCKS percentage of deteriorated transverse cracks to be repaired, input by the
user
DCRACKbw number of deteriorated transverse cracks per mile before works
Figure D3.12 shows the effect of full depth repair on cracking in JR concrete pavements.

Version 1.0
30.00 JRCP - cracking deterioration: full depth repair
25.00
Transverse Cracking [Nº / mi]
20.00
15.00
10.00
5.00
Original
50 % Full Depth Repair
0.00
0.00 10.00 20.00 30.00 40.00 50.00 60.00
Age [years]
Figure D3.12 Full depth repair effect on deteriorated transverse cracking in JR

concrete pavements
! Spalling
The amount of spalling remaining after works is given by:
SPALL aw = 1 -
PCTSPL 
 * SPALL bw ...(5.13)
 100 
where:

Figure D3.13 shows the effect of full depth repair on spalling in JR concrete pavements.

Version 1.0
100
90
80
70
Joint spalling (%)
60
50
40
30
20
Original
10
50% Full Depth Repair
0
0 5 10 15 20 25 30
Age [years]
Figure D3.13 Full depth repair effect on joint spalling in JR concrete pavements
! Faulting
The amount of faulting remaining after works is given by:
FAULTbw * (NJTS - REPJTS )

FAULTaw = ...(5.14)
NJTS
where:
FAULTaw average transverse joint faulting after works (inches)

FAULTbw average transverse joint faulting before works (inches)
REPJTS number of repaired transverse joints per mile
The number of repaired transverse joints is given by:
PCTSPL * SPALL bw * NJTS

REPJTS = ...(5.15)
100
where:
REPJTS number of repaired transverse joints per mile

SPALLbw percentage of spalled transverse joints before works (%)

Version 1.0
NJTS Number of transverse joints per mile
Figure D3.14 shows the effect of full depth repair on faulting in JR concrete pavements.
0.25
0.20
0.15
Average Faulting [in]
0.10
0.05 Original
Full Depth Repair
0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00
Figure D3.14 Full depth repair effect on faulting in JR concrete pavements
5.2.2 CR concrete pavements

The amount of full depth repairs on CR concrete pavements is computed as follows:
PCTFAIL * FAIL bw * FRESA

FDR = ...(5.16)
1.6093 * 100
where:
FDR area of full depth repair (m2/km)

PCTFAIL Percentage of failures to be repaired, input by the user
FAILbw Number of failures per mile before works
FRESA Average restoration area of each failure (m2)
The product of FDR and the section length (L) in kilometres gives the total area of full depth
repair (TFDR) in square metres. The total cost of full depth repair is computed by multiplying
TFDR by the user-specified unit cost per square metre.

Version 1.0
The effects
Following full depth repair, pavement defects will be reduced as follows:
! Failures
The amount of failures remaining after works is given by:
FAIL aw = 1 -
PCTFAIL 
 * FAIL bw ...(5.17)
 100 
where:
FAILaw number of failures per mile remaining after works

PCTFAIL percentage of failures to be repaired, input by the user
FAILbw number of failures per mile before works
Figure D3.15 shows the effect of full depth repair on failures in CR concrete pavements.
35.00
30.00
25.00
Failures [ Nº /mi]
20.00
15.00
10.00
5.00 Original
50% Full depth repair
0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00
Figure D3.15 Full depth repair effect on failures in CR concrete pavements
5.3 Partial depth repairs

This technique is used on JP concrete pavements to repair surface deterioration in the top one
third of the slab. If the deterioration extends deeper than one third of the slab depth, the slab
should be replaced. Generally partial depth repairs are performed at transverse joints, however
they may also be applied anywhere in the slab where surface defects occur.
Partial depth repair can be defined in one of the following ways:
A partial depth repair works is performed at a fixed point in time defined by the calendar
year.

Version 1.0
A fixed interval between successive partial depth repair works is specified, and the partial
depth repair is applied in an analysis year t defined by Equations 4.4 above and 4.5
above.
Partial depth repair is performed when the level of pavement defect (cracking) exceeds
the user-specified value.
In all cases, resealing will not be performed if the user-specified last applicable year or

If performed, the amount of works is calculated as follows:
0.6214 * CW * PREPJTS * SPALL bw * NJTS

PDR = ...(5.18)
100 * 100
where:
PDR amount of partial depth repair (m/km)

PREPJTS percentage of spalled transverse joints to be repaired, input by the user
The product of PDR and the section length (L) in kilometres gives the total length of partial
depth repair (TPDR) in metres. The total cost of partial depth repair is computed by
multiplying TPDR by the user-specified unit cost per metre.
The effects
Following partial depth repair, spalling will be reduced as follows:
SPALL aw = 1 -
PREPJTS 
 * SPALL bw ...(5.19)
 100 
where:

PREPJTS percentage of spalled transverse joints to be repaired, input by the user
Figure D3.16 shows the effect of partial depth repair on spalling in JP concrete pavements
without dowels.

Version 1.0
16.00
14.00
12.00
10.00
Joints Spalling [%]
8.00
6.00
4.00
2.00 Original, liquid

60% Partial Depth
0.00
0 5 10 15 20 25 30
Age [years]
Figure D3.16 Partial depth repair effect on joint spalling in JP concrete

pavements
5.4 Diamond grinding

Diamond grinding is used to restore or improve the rideability of a pavement by providing
smooth level surface. It removes faulting at joints, slab warping, and surface deformations
caused by studded tyres. Diamond grinding can also be used to correct inadequate slope for
drainage and excessive surface polishing, and to increase surface friction by creating a
roughened corduroy-like surface capable of draining water and reducing hydroplaning
potential.
Diamond grinding works can be defined in one of the following ways:
Diamond grinding is performed at a fixed point in time defined by the calendar year.
A fixed interval between successive diamond grinding is specified, and the grinding
works is performed in an analysis year t defined by Equations 4.4 above and 4.5 above.
Diamond grinding is performed when the level of pavement defect (faulting and
roughness) exceeds the user-specified values.
In all cases, diamond grinding will not be performed if the user-specified last applicable year
or maximum applicable roughness has been exceeded.

If performed, the amount of diamond grinding works is calculated as follows:
DGR = 1000 * CW ...(5.20)

Version 1.0
where:
DGR diamond grinding area (m2/km)

The product of DGR and the section length (L) in kilometres gives the total area of diamond
grinding (TDGR) in square metres. The total cost of diamond grinding is computed by
multiplying TDGR by the user-specified unit cost per square metre.
The effects
It is considered that the first application of diamond grinding will not reduce slab thickness.
However, when performed for the second (or more time) diamond grinding will reduce the
slab thickness as follows:
SLABTHK aw = SLABTHK bw - GRIND ...(5.21)
where:
SLABTHKaw slab thickness after diamond grinding (mm)

SLABTHKbw slab thickness before diamond grinding (mm)
GRIND depth of grinding (mm)
Diamond grinding will remove all faulting, that is, faulting is reset to zero and thereafter the
pavement distress modes will be modelled using the reduced slab thickness.
Roughness after diamond grinding will be computed using the Phase 1 model based on the
amounts of distresses remaining.
Figure D3.17 shows the effect of diamond grinding on faulting in JP concrete pavements
without dowels.

Version 1.0
0.40
0.35
0.30
0.25
0.20
0.15
0.10
Original
0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00
Figure D3.17 Diamond grinding effect on joint faulting in JP concrete

Figure D3.18 shows the effect of diamond grinding on faulting in JP concrete pavements with
dowels.
0.40
0.35
0.30
0.25
0.20
0.15
0.10
Original
0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00
Figure D3.18 Diamond grinding effect on joint faulting in JP concrete


Version 1.0
6 Rehabilitation
A concrete pavement may require a new surfacing layer either because the riding quality has
become unacceptable or for structural strengthening. There are two basic types of concrete
overlays on existing concrete pavements that are modelled in HDM-4 under rehabilitation
works:
! Bonded concrete overlays
! Unbonded concrete overlays
6.1 Bonded concrete overlays

The techniques used for constructing bonded concrete overlays ensure that the new concrete
overlay adheres to the existing concrete. Bonded overlays increase the structural capacity of
the existing pavement by creating a thicker monolithic section. The thickness of bonded
overlays depends primarily on the condition of the existing pavement, the traffic level and the
required life. Typically, unbonded concrete overlays are less than 100 mm thick. Due to their
monolithic nature, the underlying concrete is the main load carrying section of the pavement
structure and therefore it must be in good condition to carry the traffic load. Bonded overlays
can also be used to improve the skid resistance of an existing pavement.
A bonded concrete overlay works can be defined in one of the following manners:
A bonded concrete overlay of fixed specifications is applied at a fixed point in time

defined by the calendar year.
A fixed interval between successive bonded concrete overlay is specified, and the overlay
is performed in an analysis year t defined by Equations 4.4 above and 4.5 above.
Bonded concrete overlay is performed when the level of pavement defect (cracking,
failures and/or roughness) exceeds the user-specified values.
In all cases, bonded concrete overlay will not be performed if the user-specified last
applicable year or maximum applicable roughness has been exceeded.
6.1.1 JP concrete pavements

If bonded concrete overlay is performed on a JP concrete pavement, the amount of works is
BOL = 1000 * CW ...(6.1)
where:
BOL bonded concrete overlay area (m2/km)


Version 1.0
Before concrete overlay is applied, it is often necessary to carry out some preparatory
works. For JP concrete pavements, this is equal to the amount of slab replacement to be
performed (SLB), computed using Equations 5.1 above to 5.3 above.
The total amount of bonded concrete overlay (TBOL) in square metres is obtained from the
product of BOL and the section length (L) in kilometres. The total cost of bonded concrete
overlay is obtained by multiplying TBOL by the user-specified unit cost per square metre. The
additional areas and cost of preparatory works will be reported separately under slab
replacement.
The effects
The pavement type after bonded concrete overlay will not change. The seal type for JP
concrete pavements will be changed to the user-specified type.
! Slab thickness
Bonded concrete overlay will reset the concrete slab thickness as follows:
SLABTHK aw = SLBATHK bw + OVLTHK ...(6.2)
where:
SLABTHKaw slab thickness after bonded concrete overlay (mm)

SLABTHKbw slab thickness before bonded concrete overlay (mm)
OVLTHK thickness of bonded concrete overlay (mm)
! Cracking
The amount of cracking remaining after bonded concrete overlay works is computed
using Equation 5.4 above. Thereafter, cracking progression is predicted using the Phase 1
RD model with the new slab thickness (SLABTHKaw) and the new values of the
maximum permissible number of equivalent standard axle load repetitions during each
temperature gradient (Ntg). The cumulative fatigue damage (FD) is maintained and used
in the calculation of the new Ntg values (see Section 4 of Chapter C3).
Figure D3.19 shows the effect of bonded concrete overlay on cracking in JP concrete
pavements.

Version 1.0
90%
80%
Same fatigue
70% Slab thickness changes
Determination of the number of admissible axles N
60%
Transverse Cracking [%]
50%
40%
30%
20%
Original
10% Overlay
0%
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 50.00
Figure D3.19 Bonded concrete overlay effect on cracking in JP concrete

pavements
! Faulting
Faulting after bonded concrete overlay will be reset to zero. Thereafter, the progression
of faulting will be computed using the Phase 1 RD model. The value of NE4
(cumulative traffic loading in equivalent standard axle loads) will be reset to zero.
Figure D3.20 shows the effect of bonded concrete overlay on faulting in JP concrete
0.20
0.18
0.16
0.14
0.12
0.10
0.08
0.06
0.04
Original
0.02 Overlay
0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00
Figure D3.20 Bonded concrete overlay effect on joint faulting in JP concrete

Figure D3.21 shows the effect of bonded concrete overlay on faulting in JP concrete

Version 1.0
0.18
0.16
0.14
0.12
Average Joint Faulting [in]
0.10
0.08
0.06
0.04
Original
0.02 Bonded Overlay
0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00
Figure D3.21 Bonded concrete overlay effect on faulting in JP concrete

! Spalling
Spalling after bonded concrete overlay will be reset to zero. Thereafter, spalling
progression will be computed using the Phase 1 RD model with the parameter pavement
age (AGE) counted since the time of overlay.
Figure D3.22 shows the effect of bonded concrete overlay on spalling in JP concrete
16.00
14.00 Original
Overlay
12.00
10.00
Joint Spalling [Nº /mi]
8.00
6.00
4.00
2.00
0.00
0 5 10 15 20 25 30
Age [years]
Figure D3.22 Bonded concrete overlay effect on joint spalling in JP concrete


Version 1.0
! Roughness
Roughness after bonded concrete overlay will be reset to a new value, and its progression
computed using the Phase 1 RD model based on the amounts of cracking, spalling, and
faulting remaining.
6.1.2 JR concrete pavements

If bonded concrete overlay is performed on a JR concrete pavement, the amount of works
(BOL), in square metres per kilometre, is calculated using Equation 6.1 above.
The amount of preparatory works required will be given by the amount full depth repair to be
performed (FDR), which is computed using Equations 5.8 above and 5.11 above.
additional areas and cost of preparatory works will be reported separately under full depth
repair.
The effects
! Slab thickness
The new slab thickness after works is calculated as follows:
SLABTHK aw = SLABTHK adj + OVLTHK ...(6.3)
where:
SLABTHKaw slab thickness after bonded concrete overlay (mm)

SLABTHKadj adjusted slab thickness to account for the effect of the remaining
cracking (mm)
OVLTHK thickness of bonded concrete overlay (mm)
! Cracking
The adjusted slab thickness to account for the effect of the remaining cracking is given
by:
SLABTHK adj = SLABTHK bw * 1 - 0.5 *  

DCRACK aw
  ...(6.4)
 MAXCKS 
where:
SLABTHKadj adjusted slab thickness to account for the effect of the remaining
cracking (mm)
SLABTHKbw slab thickness before bonded concrete overlay (mm)
DCRACKaw number of deteriorated transverse cracks per mile remaining after
bonded concrete overlay (mm)

Version 1.0
MAXCKS maximum number of deteriorated transverse cracks per mile
The maximum number of deteriorated transverse cracks per mile (MAXCKS) is given
by:
3 * 5280
MAXCKS = ...(6.5)
JTSPACE
where:
MAXCKS maximum number of deteriorated transverse cracks per mile

If DCRACKaw is greater than MAXCKS, then set DCRACKaw to be equal to MAXCKS.

The amount of cracking remaining after works is computed using Equation 5.12 above.
Thereafter, cracking progression is computed using the Phase 1 RD model with the new
slab thickness (SLABTHKaw) and new pavement age (AGE) counted since performing
the concrete overlay.
Figure D3.23 shows the effect of bonded concrete overlay on cracking in JR concrete
pavements.
25.00
/ ]
Transverse Cracking Deterioration [Nº mi
20.00
Original
Bonded Overlay
15.00
10.00
5.00
0.00
0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00
Figure D3.23 Bonded concrete overlay effect on cracking in JR concrete

pavements
! Faulting
Faulting will be reset to zero. Thereafter the progression of faulting will be computed
using the Phase 1 RD model. The value of NE4 (cumulative traffic loading in
equivalent standard axle loads) will be reset to zero.

Version 1.0
Figure D3.24 shows the effect of bonded concrete overlay on faulting in JR concrete
pavements.
0.18
0.16
Original
0.14
Average Faulting [in]
Bonded Overlay
0.12
0.10
0.08
0.06
0.04
0.02
0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00
Figure D3.24 Bonded concrete overlay effect on faulting in JR concrete

pavements
! Spalling
Spalling after bonded concrete overlay will be reset to zero. Thereafter spalling
progression will be computed using the Phase 1 RD model with the parameter pavement
age (AGE) counted since the time of overlay.
Figure D3.25 shows the effect of bonded concrete overlay on spalling in JR concrete
pavements.

Version 1.0
50
45 Original
Overlay
40
35
30
Joint Spalling [%]
25
20
15
10
0
0 5 10 15 20 25 30
Age [years]
Figure D3.25 Bonded concrete overlay effect on joint spalling in JR concrete

pavements
! Roughness
is computed using the Phase 1 RD model based on the PSR value.
6.1.3 CR concrete pavements

If bonded concrete overlay is performed on a CR concrete pavement, the amount of works
(BOL), in square metres per kilometre, is calculated using Equation 6.1 above.
The amount of preparatory works required will be given by the amount full depth repair to be
performed (FDR), which is computed using Equation 5.16 above.
additional areas and cost of preparatory works will be reported separately under full depth
repair.
The effects
Following bonded concrete overlay the slab thickness will be reset as expressed by Equation
6.2 above.
! Failures
The number of failures after bonded concrete overlay will be reset to zero.
Figure D3.26 shows the effect of bonded concrete overlay on failures in CR concrete
pavements.

Version 1.0
160.00
140.00 Original
Bonded Overlay
120.00
Failures [Nº /mi]
100.00
80.00
60.00
40.00
20.00
0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00
Figure D3.26 Bonded concrete overlay effect on failures in CR pavements
! Roughness
is computed using the Phase 1 RD model based on the PSR value.
6.2 Unbonded concrete overlays

The techniques used for constructing unbonded concrete overlays ensure that the new concrete
layer does not adhere to the existing concrete. This involves placing a separation interlayer,
usually of asphaltic concrete, over the existing concrete and then constructing a new concrete
pavement on top of the interlayer. Unbonded concrete overlays on heavily trafficked roads are
typically greater than 150 mm and for lightly trafficked roads not less than 100 mm. Because
the two concrete layers are separated and act independently of each other, the unbonded
concrete overlay behaves as a new concrete pavement on top of a very rigid support. The
separation interlayer helps to delay the onset of reflection cracking.
An unbonded concrete overlay works can be defined in one of the following ways:
An unbonded concrete overlay is performed at a fixed point in time defined by the

calendar year.
A fixed interval between successive unbonded concrete overlays is specified, and the
overlay is performed in an analysis year t defined by Equations 4.4 above and 4.5 above.
Unbonded concrete overlay is performed when the level of pavement defect (cracking,
failures and/or roughness) exceeds the user-specified values.
In all cases, unbonded concrete overlay will not be performed if the user-specified last
applicable year or maximum applicable roughness has been exceeded.

Version 1.0

If performed, the amount of unbonded concrete overlay works is calculated as follows:
UOL = 1000 * CW ...(6.6)
where:
UOL unbonded concrete overlay area (m2/km)

The total amount of unbonded concrete overlay (TUOL) in square metres is obtained from the
product of UOL and the section length (L) in kilometres. The total cost of unbonded concrete
overlay is obtained by multiplying TUOL by the user-specified unit cost per square metre.
The effects
After unbonded concrete overlay, the roadbase type will change to rigid/concrete base (RB).
The seal type for JP and JR concrete pavements will be changed to the user-specified types. It
is considered that the pavement will behave as new and the performance will be modelled
using the Phase 1 RD models. Roughness after unbonded concrete overlay will be reset to the
user-specified value.

Version 1.0
7 Reconstruction
Pavement reconstruction refers to all works that require the re-specification of part or the
entire pavement structure and characteristics, which involves the removal and replacement of
the surfacing, roadbase and sub-bases. The re-specification of concrete pavement
reconstruction also allows for the adjustment to roadside geometry and safety features, and to
improve or add drainage structures.
Reconstruction may be specified in one of the following ways:
A pavement reconstruction works is to be performed at a fixed point in time defined by

the calendar year.
A reconstruction works of fixed specifications is applied when the levels of the user-
specified intervention criteria, based on pavement condition, are met. Reconstruction will
not be performed if the last applicable year has been exceeded.

REC = 1000 * CWaw ...(7.1)
where:
REC area of pavement reconstruction (m2/km)

CWaw carriageway width after works (m)
The total area of reconstruction (TREC) in square metres is obtained from the product of REC
and the road section length (L) in kilometres. The total cost of pavement reconstruction is
obtained by multiplying TREC by the user-specified unit cost per square metre.
The effects
After reconstruction, the pavement type will be reset to the new type specified by the user.
The pavement will behave as new and the performance will be modelled using the Phase 1 RD
models. Roughness will be reset to the user-specified value. The calibration factors for
modelling pavement deterioration will also be reset to user-specified values.

Version 1.0
8 Special works
The effects of the following special works on pavement performance are not modelled
endogenously in HDM-4, and therefore only their costs can be considered in an analysis:
! Emergency works
For example, repairing washout/subsidence, clearing debris, traffic accident removal, etc.
These works will be scheduled at a fixed interval of time (minimum of one-year interval), and
will be performed on an annual basis. If specified, these works activities will be applied in a
given analysis year regardless of the works hierarchy.
Their unit costs will be specified in terms of currency per kilometre per year, and the annual
costs will be obtained by multiplying the road section length (L) by the unit cost.

Version 1.0
9 New construction
intervention criteria. In a project analysis, a new road section (that is, a new link) can be
specified as a section alternative within a selected project alternative.
The required components for defining a new section are as follows:
! Road section data
All the data items that are required to define a road section in HDM-4. The user can
specify these items using aggregate data.
! Traffic data
❏ diverted traffic
traffic that is diverted to the new section (link) from the nearby routes and other
transport modes.
❏ generated traffic
additional traffic that occurs in response to the new investment.
! Construction costs and duration
! Maintenance standards
Applied after opening the new section to traffic.

constructed (NEWCON) and this is equal to the new section length. The total cost of
construction is obtained from the product of NEWCON and the user-specified unit cost per
kilometre.
The new pavement will be modelled using the Phase 1 RD models discussed in Chapter C3.

Version 1.0
10 References
Washington D.C., USA
LAST, (1995)
Concrete pavement performance equations
Latin American Study Team
International Study of Highway Development and Management Tools
Santiago, Chile
LAST, (1996)
Modelling road design and maintenance effects for pavements in HDM-4
Final Report, FICEM, Latin American Study Team
Santiago, Chile

Version 1.0
D4 Unsealed Roads
1 Introduction
This document describes the detailed modelling of Road Works Effects for unsealed roads
(see Figure D4.1).
Road
RoadWorks
WorksEffects
Effects

UnsealedRoads
TypesofofWorks
Pavements
Pavements Pavements
Pavements
Chapter
ChapterD1
D1 Chapter
ChapterD4
D4
Chapter
ChapterD2
D2 Chapter
ChapterD3
D3
! Maintenance (see Section 3)
! Improvement (see Section 4)
! Construction (see Section 5)
hierarchical ranking of works activities and pavement type resets after works.

Version 1.0
PART D ROAD W ORKS EFFECTS D4 UNSEALED ROADS
2 Modelling logic
An unsealed road is considered to comprise two layers, a gravel surfacing and a subgrade. A
gravel road has both layers, but an earth road has a zero thickness of gravel surfacing and its
surface characteristics are those of the subgrade. When a gravel road loses all of its gravel
surfacing, then its classification reverts to that of earth road. Upon gravel resurfacing, all
unpaved roads become gravel roads by definition of the new surfacing layer. The background
of the modelling logic is given in Watanatada et al. (1987) and Paterson (1987).

The overall computational procedure for modelling works on unsealed roads that are applied
in each analysis year can be summarised by the following steps:
! Determine the works standard(s) that is applicable in the given year. Only one
maintenance standard and/or one improvement standard can be applied to a road section
feature in any analysis year.
! Check the intervention criteria and the limits defined for works in the following order:
❏ improvement works then maintenance works
❏ a scheduled operation takes priority over a responsive operation of the same type
! Identify and apply the highest-ranking works activity.
! Compute the physical quantities of works.
! Compute works effects and reset modelling parameter values to reflect post-works road
! Apply any other work activities whose effects on pavement are not modelled
endogenously, (for example, routine-miscellaneous maintenance).
! Calculate the costs of works by applying unit costs to the physical quantities of works.
! Store results for economic analysis and for use in the following analysis year.

A works activity (or an operation) is triggered when any one or a combination of the user-
specified criteria has been met. When more than one works activity meets the criteria for
being applied in a given analysis year, the highest placed operation for the particular road
feature is selected.
Table D4.1 shows the hierarchy of works activities that are applicable to the carriageway. The
operation dualisation of an existing road section is placed at the top of the list as number 1,
and takes priority over all the other operations, while routine pavement works (that is, spot-
regravelling and grading) are given the lowest priority.

Version 1.0
Table D4.1 Hierarchy of road works applicable to the carriageway
Works type Works activity / Operation Hierarchy Unit cost


Widening
Resurfacing Regravelling 6 per m3
Routine Spot regravelling 1 7 per m3
Pavement Grading 1 7 per km
Notes:
1 Spot regravelling and grading have the same ranking, and both of them can be
performed in the same analysis year
An improvement or construction works of fixed specifications is applied to a given road

section only once during the analysis period. This rule applies particularly to improvement
works that have been defined as responsive to the user-specified intervention criteria based on
road user effects parameters.
Spot regravelling and grading can be defined by the user to be applied as separate works
activities in each year, or defined to repair areas of severe depression and reduce roughness
before applying the higher-ranking works (for example, regravelling, widening). In the former
case, spot regravelling or grading is performed every year in which no periodic maintenance
or capital works is applied. In the latter case, spot regravelling and grading are considered to
be an integral part of the periodic maintenance or capitals works, and are referred to as
preparatory works. Although preparatory works are automatically triggered and performed
together with the capital works, the amount and the cost of each of the works activities
involved is modelled and reported separately.
The works activities that apply to shoulders and Non-Motorised Transport (NMT) lanes are
applied in any analysis year, if specified by the user, regardless of the works hierarchy given
in Table D4.1. Shoulder or NMT lane improvement works takes priority over shoulders repair
or NMT lane repair, respectively.
For all road feature types, if more than one works activity of the same operation type (for
example, different specifications of regravelling) are applicable in a given analysis year, the
one with the highest cost takes priority over the others.
The following works activities whose effects are not modelled endogenously:
! Emergency works
! Routine-miscellaneous maintenance
are applied in a given analysis year, if specified by the user, regardless of any works
hierarchy.

Version 1.0
3 Maintenance works
The maintenance of unsealed roads comprises the following operations:
! Periodic grading (see Section 3.1)
! Spot regravelling (see Section 3.2)
! Gravel resurfacing (see Section 3.3)
! Routine-miscellaneous maintenance of drainage and verges (see Section 3.4)
Maintenance of gravel surfacing is accounted each analysis year through the surfacing
thickness and the net change from material loss, spot regravelling and gravel resurfacing
maintenance. The material loss from earth roads, although computed, is accounted only for the
purpose of predicting spot regravelling quantities and is otherwise ignored, Watanatada et al.
(1987).
3.1 Periodic grading

Periodic grading by motorised or towed grader to restore surfacing gravel from the shoulders
to the roadway and to reduce roughness is one of the principal routine maintenance for
unsealed roads. The periodic grading of unpaved roads is usually undertaken on a more-or-
less regular basis for management purposes, either seasonally or frequently enough to keep the
roughness within tolerable limits.
Grading may be specified by the user in one of the following three ways:
Grading works is performed at a fixed time interval in days between successive grading.
! Option 2: Traffic-responsive
Grading is performed at a fixed traffic interval in number of vehicle passes between

successive grading.
! Option 3: Roughness-responsive
Grading is performed after a maximum allowable roughness value has been exceeded.

When performed, the amount of grading works (LGRD) is equal to the road section length in
kilometres. The cost of grading is obtained from the product of LGRD and the unit cost per
kilometre. The annual cost of grading is obtained by multiplying this product with the number
of grading performed in a given year.
3.1.2 Effects of grading

In all options, the average roughness between successive grading (QIavg) is computed as a
function of the number of days between grading (DG) as described in Part C - Chapter C4. In
the scheduled option, DG is specified directly by the user. In the traffic and roughness-
responsive options, DG is determined as follows:
If DGMAX < DG'

DG = DGMAX ...(3.1)
If DGMIN < DG' ≤ DGMAX

Version 1.0
DG = DG' ...(3.2)
If DG' ≤ DGMIN
DG = DGMIN ...(3.3)
where:
DGMAX the maximum allowable time interval between successive grading, in days,
specified by the user as an option or equal to the default value of 10,000 days
DG' the number of days between successive grading determined from the traffic or
roughness parameter
DGMIN the minimum applicable time interval between successive grading, in days,
specified by the user as an option or equal to the default value of 5 days
The parameter DG' is determined as follows:

for the traffic-responsive option:
VEHG
DG' = ...(3.4)
AADT
for the roughness-responsive option:
 (QIMAX j − QIMAX 0 ) 
DG' =   * log e 
1
 ...(3.5)
c  [QIMAX j − (1 − a ) * QIMIN j − a * QIMAX 0 ]
where:
VEHG the traffic interval between successive grading, in vehicles, specified by the
user
QIMAXj the maximum roughness of material j (QI)
QIMINj the minimum roughness of material j (QI)
QIMAXO the maximum allowable roughness specified by the user (QI)
a and c are model parameters which are defined in Chapter C4, Sections 3.2, 3.4 and
3.5
If no grading is specified, the long-term average roughness (QIlta) is equal to the maximum
roughness, as follows:
QIlta = QIMAX j ...(3.6)
Note that if the historic maintenance of the section has been nil-grading over several years,
then the existing roughness is the best estimate of the average roughness and the user can
provide this by specifying QIMAX endogenously with a value equal to the existing
roughness.

Version 1.0
3.2 Spot regravelling

Spot regravelling provides repair to areas of severe depression (gravel loss, rutting, etc.), and
may be specified by the user as follows:
A fixed number of cubic metres per kilometre (VGS) per year is applied.
A percentage of gravel or subgrade material loss in the current analysis year is replaced
subject to a maximum limit per year. In this option VGS is computed as follows:
VGS = Pmla * MLA * (CW + SW ) * 10 −2 ...(3.7)
where:
VGS the in-place volume of material added due to the spot regravelling
(m3/km)
Pmla percentage of annual material loss to be replaced, specified by the user
(%)
MLA annual material loss (mm)
SW shoulder width (m)

When spot regravelling is performed, the added material is assumed to be the same type as the
existing. The total amount of spot regravelling performed is given by:
TVGS = VGS * L ...(3.8)
where:
TVGS Total amount of spot regravelling for the road section (m3)
The cost of spot regravelling is computed as the product of TVGS and the unit cost of material
per cubic metre.
3.2.2 Effects of spot regravelling

! Gravel thickness
For gravel roads, the thickness of the gravel layer is increased to reflect the volume of
material added, according to the following formula (trapezoidal rule):
VGS
∆THGS = ...(3.9)
(CW + SW )

Version 1.0
THG aw = THG bw + ∆THGS ...(3.10)
where:
∆THGS the increase in gravel thickness due to spot regravelling (mm)

THGaw gravel thickness after works (mm)
THGbw gravel thickness before works (mm)

! Roughness
Spot regravelling is predicted to reduce the average roughness on the assumption that the
gravel is applied in the major depressions and potholes that have appeared in the surface
in the upper ranges of roughness. Roughness levels above 190 QI (15 m/km IRI) are
invariably associated with the presence of visible birdbath type depressions or potholes,
which become larger or more frequent as the roughness level increases, and these can be
effectively patched, with high benefits, by spot regravelling. Over the roughness range of
150 to 190 QI (that is, 11 to 15 IRI m/km), such patchable birdbath depressions are
frequently observed but not always present so that, in this range, spot regravelling may
not always be effective. For example, spot regravelling is not effective maintenance on
corrugations or on runoff-induced surface erosion, which conditions commonly induce
roughness levels within this range. At roughness levels below 150 QI (that is, 11 IRI
m/km) spot regravelling is considered to be ineffective on roughness. This logic is
defined in the following algorithm, adopting the roughness to volume of depression ratio
as equal to 2 QI per m3/lane/km), allowing for the spot regravelling to be only 60%
effective (that is, 1.2 QI per m3/lane/km), and adopting an average effective lane width of
3 m:
  [
 QI avg(bw) - 150
QIavg (aw ) = MAX150, QI avg(bw ) − MIN1,
] * 3.6 * VGS 
   ...(3.11)
   40  CW  
where:
QIavg(aw) average roughness after spot regravelling (QI)

QIavg(bw) average roughness before spot regravelling (QI)

The effects of spot regravelling are illustrated in Watanatada et al. (1987). It should be noted
that spot regravelling affords only a temporary repair of depressions, and that the most
effective means is by grading, or in severe cases by scarifying, grading and recompacting.
3.3 Gravel resurfacing

Gravel resurfacing is performed to replace or augment the gravel-surfacing layer in response to
material loss. A gravel resurfacing works is specified as follows:

Version 1.0
Gravel resurfacing is applied when the gravel age (GAGE) equals or exceeds the fixed
time interval specified by the user.
Gravel resurfacing is applied when either:

# The current gravel thickness (THG) falls below the user-specified minimum allowable
thickness, provided that GAGE equals or exceeds the user-specified minimum
applicable resurfacing interval, in years; or
# The gravel age (GAGE) equals or exceeds the user-specified maximum allowable
resurfacing interval, in years.
However, gravel resurfacing is not performed if either of the following is true:
! The specified last applicable year has been exceeded, or
! The analysis year is a construction year, or
! The final thickness specified in a scheduled policy is smaller than the predicted thickness
at the end of the analysis year before the resurfacing decision.

When performed, the amount of regravelling is given by:
TVGR = VGR * L ...(3.12)
where:
TVGR amount of regravelling for the entire road section (m3)

VGR the in-place volume of gravel added due to the regravelling (m3/km)
The volume of gravel added per km is computed according to the following trapezoidal
formula:
VGR = (THG aw - THG bw ) * (CW + SW ) ...(3.13)

The cost of regravelling is obtained from the product of TVGR and the unit cost per cubic
metre.
Preparatory works
If roughness before works (RIbw) is greater than 11.6 IRI m/km, it is assumed that the
following amount of spot regravelling is performed before gravel resurfacing:
(QIbw - 150 ) * CWbw

VGS = ...(3.14)
3.6 * {MIN [1, (QIbw - 150)/40 ]}
where:

Version 1.0
VGS volume of spot regravelling (m3/km)

QIbw roughness before works (= QIb) (QI)
The additional cost of spot regravelling is obtained by multiplying the product of VGS and the
section length (L) by the user-specified unit cost per cubic metre.
3.3.2 Effects of gravel resurfacing

! Pavement type
When gravel resurfacing is performed the pavement type is set to gravel (GRUP)
regardless of the previous surface type.
! Gravel age
After regravelling, the gravel age (GAGE) is reset to zero.

! Gravel thickness
The thickness of the gravel surfacing is increased according to the formula below:
If the final gravel thickness is specified:
THG aw = THG 0 ...(3.15)
If an increase in the gravel thickness is specified:
THG aw = THG bw + ∆THG ...(3.16)
where:
THG0 the final gravel thickness after gravel resurfacing, specified by the user
(mm)
∆THG the increase in the gravel thickness due to gravel resurfacing, specified by
the user (m)
The existing surface material is changed to the material specified by the user (which may
still be of the same attributes as the existing). The surface material attributes (P075,
P425, P02, D95, PI, QIMIN and QIMAX) are replaced either by the new values provided
by the user, or by the default values from the previous gravel attributes.
! Roughness
Roughness after regravelling is reset to a user specified value. If this is not specified, the
roughness after works is reset to the minimum allowable value (QIMIN) converted into
IRI m/km units.
3.4 Routine-miscellaneous maintenance

This includes drainage maintenance, vegetation control, safety installations, and other items
that are not modelled as affecting the riding quality of the pavement. A lump sum cost per km
per year is used as the basis for costing routine maintenance. Because the unpaved road
deterioration relationships employed are based on the assumption of adequate drainage, the

Version 1.0
cost of drainage maintenance should be included, when it is normally done. Otherwise, some
allowance due to the lack of drainage, for example, in the form of frequent road closures,
washouts, etc., should be incorporated in the economic analysis.
When specified by the user, the total annual cost of routine-miscellaneous maintenance is
obtained from the product of the section length (L) and the unit cost per kilometre per year.

Version 1.0
4 Improvement works
Improvement works for unsealed roads comprises the following:
! Widening (see Section 4.1)
! Realignment (see Section 4.2)
4.1 Widening
The operations included under widening are lane addition and partial widening. The
difference between the two is that partial widening does not increase the number of lanes. It is
considered that these operations do not alter the road alignment, hence there is no change in
section length.
Widening works can be defined in one of the following ways:
A widening works of fixed specifications is applied at a specified time, defined by a

calendar year.
Widening of fixed specifications is applied when the user-specified intervention criteria,

based on road user effects parameters, are met. Widening is not performed if the last
applicable year has been exceeded.
The following information is required to specify a widening works:
! Road type.
! Road class.
! Increase in width for partial widening.
! Additional number of lanes and increase in carriageway width for lane addition.
! Pavement type for the entire section.
! Pavement details of the widened area of carriageway.
! Whether or not the existing carriageway is resurfaced.

If performed, the amount of widening works is given by:
AWDN = 1000 * ∆CW ...(4.1)
where:
AWDN widened area of carriageway (m2/km)

The total area of widening over the entire road section is given by:

Version 1.0
TAWDN = AWDN * L ...(4.2)
where:
TAWDN widened area of carriageway (m2)

The cost (CSTWDN) of widening the section is obtained from the product of TAWDN and
the user-specified unit cost per square metre, or from the product of the section length L and
the unit cost per kilometre.
SALVA = PCTSAV * CSTWDN ...(4.3)
where:

Additional works
It is likely that widening works will involve resurfacing, or repair to the severely damaged
area of the existing carriageway. The required additional works are modelled as described
below:
! Case 1: resurfacing the existing carriageway
If the existing carriageway is to be resurfaced, the amount of gravel resurfacing to be

provided is given by:
VGR = (THG aw - THG bw ) * (CWbw + SW ) ...(4.4)
where:
VGR volume of regravelling (m3/km)

THGaw gravel thickness after works (mm)
THGbw gravel thickness before works (mm)
CWbw carriageway width before works (m)
The total amount of regravelling the existing carriageway is given as:
TVGR = VGR * L ...(4.5)
where:

Version 1.0
TVGR total volume of regravelling the existing carriageway (m3)
The cost of regravelling the existing carriageway is obtained from the product of TVGR
and the user-specified unit cost per cubic metre.
! Case 2: no resurfacing of the existing carriageway
If the existing carriageway is not to be resurfaced, it is assumed that spot regravelling

and grading the full length of the existing carriageway is performed along with the
widening works.
If roughness before works, RIbw, is greater than 11.6 IRI m/km, the amount of spot
regravelling performed is computed using Equation 3.14 above. The additional cost of
spot regravelling is obtained by multiplying the product of VGS and the section length
(L) by the user-specified unit cost per cubic metre.
The amount of grading performed (LGRD) is equal to the road section length (L) in
kilometres, and the additional cost of grading is obtained from the product of LGRD and
the user-specified unit cost per kilometre.
The total cost of widening works is the sum of the carriageway widening cost and the cost of
additional works comprising regravelling of the existing carriageway, or grading and any spot
regravelling performed. The costs and the amounts of the additional works are reported
separately under regravelling, grading or spot regravelling.
In economic analysis, it is assumed that these additional costs are incurred in the last
construction year.
4.1.2 Effects of widening

It is considered that widening works do not alter the road surface class. After widening, the
required modelling parameters are reset as described below:
! Carriageway width
The new carriageway width after works is given as follows:
CWaw = CWbw + ∆CW ...(4.6)
where:
CWaw carriageway width after works (m)

CWbw carriageway width before works (m)
For partial widening the increase in carriageway width (∆CW) is specified directly by the
user. For lane addition, the increase in carriageway width is user-specified. If this
is not specified the increase in width is given by:
(ADDLN * CWbw )
∆CW = ...(4.7)
NLANES bw
where:

Version 1.0
ADDLN additional number of lanes, input by the user

NLANESbw number of lanes before works
For lane addition, the number of lanes after widening works (NLANES aw) is equal
to the number of lanes before works (NLANESbw) plus the user-specified additional
number of lanes (ADDLN).
! Gravel thickness
❏ Gravel thickness after widening is calculated as a weighted average as follows:
(CWbw * THG excw + ∆CW * THG ww )

THG aw = ...(4.8)
CWaw
where:
THGaw gravel thickness after widening works (mm)

THGww gravel thickness on the widened part of the carriageway (mm)
THGbw gravel thickness before widening works (mm)
THGexcw gravel thickness over the existing carriageway after widening works
(mm)

❏ The gravel thickness over the existing carriageway after widening (THGexcw) is
obtained as follows:
If the existing carriageway is to be regravelled:
THG excw = THG bw + ∆THG gr ...(4.9)
and the pavement type is set to gravel (GRUP) regardless of the previous pavement
type
If the existing carriageway is not to be regravelled:
THG excw = THG bw + ∆THGS ...(4.10)
where:
∆THGgr increase in gravel thickness over the existing carriageway due to

regravelling (mm)
∆THGS increase in gravel thickness over the existing carriageway due to spot
regravelling (mm)
❏ The increase in gravel thickness over the existing carriageway due to spot
regravelling (∆THGS) is obtained as follows:
If pavement type is gravel (GRUP):

Version 1.0
∆THGS is computed using Equation 3.9 above

If pavement type is earth (EAUP):
∆THGS is set to zero
! Surface material properties
After widening, the surface material properties (SMPi) are reset as follows:
❏ If the existing carriageway is to be regravelled, all the surface material properties are
reset to those of the new gravel material.
❏ If the existing carriageway is not to be regravelled
 (CWbw * SMPi bw + ∆CW * SMPi ww )

SMPi aw =   ...(4.11)
 CWaw 
where:
SMPiaw surface material property i after widening works, (i = P075, P425,

P02, PI, D95)
SMPibw surface material property i before widening works (i = P075, P425,
P02, PI, D95)
SMPiww Surface material property i of the widened part of the carriageway
works (i = P075, P425, P02, PI, D95)
! Roughness
Roughness after widening (RIaw) is reset to a user-specified value. If this is not specified,
reset RIaw to the minimum allowable roughness (QIMIN) converted into IRI m/km units.
! Gravel age
The gravel age (GAGE) after widening works is reset as follows:

❏ If the existing carriageway is to be regravelled, GAGE is reset to zero.
❏ If the existing carriageway is not to be regravelled, GAGE is calculated as follows:
 (CWbw * GAGE bw ) 
GAGE aw =   ...(4.12)
 CWaw 
where:
GAGEaw gravel age after widening works (years) (Note: returns an integer
value)
GAGEbw gravel age before widening works (years)
! Material loss calibration factors
After widening works, the material loss calibration factors (that is, Kgl and Kkt) are reset
to user-specified values.

Version 1.0
! Speed factors
noise which depends primarily on the individual road section.
The data that describes the hourly traffic flow distribution is also reset to a user-specified
type.
4.2 Realignment
Refers to local geometric improvements of existing roads, which may also result in a
reduction of the road length. It is assumed that the carriageway width remains unaltered when
a realignment works is performed.
Realignment works can be defined as follows:
A realignment works of fixed specifications is applied at a specified time, defined by a

calendar year.
A realignment works of fixed specifications is applied when the user-specified

intervention criteria, based on road user effects parameters, are met. Realignment is not
performed if the last applicable year has been exceeded.
A realignment works is specified by the following information:
! Road type.
! Road class.
! Proportion of new construction - defined as the ratio of new construction length to the
section length before works.
! Length adjustment factor.
! Geometry details.
! Pavement type for the entire section.
! Pavement details of the new construction segments.
! Whether or not the non-realigned segments of the existing carriageway is regravelled.

If performed, the total amount of realignment works is given by:
REAL = Pconew * L aw ...(4.13)
L aw = L bw * LF ...(4.14)
where:
REAL length of road realigned (km)

Version 1.0
Pconew proportion of new construction (0 < Pconew < 1)

Law new section length after realignment works (km)
Lbw section length before realignment works (km)
LF length adjustment factor (LF > 0)
The cost of realignment (CSTREAL) is obtained from the product of REAL and the user-
specified realignment unit cost per kilometre.
SALVA = PCTSAV * CSTREAL ...(4.15)
where:

Additional works
It is assumed that the following amounts of additional works would be carried out together
! Case 1: resurfacing the non-realigned segments
If the non-realigned parts of the existing carriageway are to be resurfaced, the amount of
regravelling works is given as follows:
( )
VGR = THG aw - THG bw * (CW + SW ) ...(4.16)
where:
VGR volume of regravelling (m3/km)

THGaw gravel thickness after realignment works (mm)
THGbw gravel thickness before realignment works (mm)
The total amount of regravelling the non-realigned parts of the existing carriageway is
given by:
TVGR = VGR * (1 - Pconew ) * Law ...(4.17)
where:
TVGR total volume of regravelling the non-realigned parts of the existing

carriageway (m3)

Version 1.0
The cost of regravelling the non-realigned parts of the existing carriageway is obtained
from the product of TVGR and the user-specified unit cost per cubic metre.
! Case 2: no resurfacing of the non-realigned segments
If the non-realigned parts of the existing carriageway are not to be resurfaced, it is

assumed that the following amounts of grading and spot regravelling are performed along
❏ Grading
LGRD = (1 - Pconew ) * L aw ...(4.18)
where:
LGRD total length of road graded (km)

The cost of grading the non-realigned parts of the existing carriageway is obtained
from the product of LGRD and the user-specified unit cost per kilometre.
❏ Spot regravelling
If roughness before works (RIbw) is greater than 11.6 IRI m/km, the amount of spot
regravelling performed on the non-realigned parts of the existing carriageway is
TVGS = VGS * (1 - Pconew ) * Law ...(4.19)
where:
TVGS volume of spot regravelling performed on the non-realigned parts of

the existing carriageway (m3)
VGS volume of spot regravelling computed using Equation 3.14 given
above (m3/km)
The cost of spot regravelling the non-realigned parts of the existing carriageway is
calculated from the product of TVGS and the user-specified unit cost per cubic
metre.
The total cost of realignment works is the sum of realignment construction cost and the cost of
additional works comprising regravelling of the non-realigned parts of the existing
carriageway, or grading and any spot regravelling performed. The additional costs and the
amounts of works should be reported separately under regravelling, grading or spot
regravelling.
In economic analysis it is assumed that these additional costs are incurred in the last
construction year.
4.2.2 Effects of realignment

It is considered that realignment works do not alter the road surface class. The required
modelling parameters are reset as described below:

Version 1.0
! New section length
The new length of the road section after realignment is given by Equation 4.14 above.
! Gravel thickness
❏ Gravel thickness after realignment is calculated as follows:
THG aw = [(1 - Pconew ) * THG excw + Pconew * THG rw ] ...(4.20)
where:
THGaw gravel thickness after realignment works (mm)

THGexcw gravel thickness of the non-realigned parts of the existing
carriageway after realignment works (mm)
THGrw gravel thickness of the realigned parts of the carriageway (mm)
THGbw gravel thickness before realignment works (mm)
❏ The gravel thickness over the non-realigned parts of the existing carriageway after a
realignment works is obtained as follows:
If the non-realigned parts of the existing carriageway are to be regravelled
THG excw = THG bw + ∆THG gr ...(4.21)
and the pavement type is set to gravel (GRUP) regardless of the previous pavement
type.
If the non-realigned parts of the existing carriageway are not to be regravelled
THG excw = THG bw + ∆THGS ...(4.22)
where:
∆THGgr increase in gravel thickness over the non-realigned parts of the

existing carriageway due to regravelling (mm)
∆THGS increase in gravel thickness over the non-realigned parts of the
existing carriageway due to spot regravelling (mm)
❏ The increase in gravel thickness over the non-realigned parts of the existing
carriageway due to spot regravelling (∆THGS) is obtained as follows:
If pavement type is gravel (GRUP):
∆THGS is computed using Equation 3.9 above
If pavement type is earth (EAUP):
∆THGS is set to zero
! Surface material properties
After realignment works, the surface material properties (SMPi) are reset as follows:

Version 1.0
❏ If the non-realigned parts of the existing carriageway are to be regravelled, all the
surface material properties are reset to those of the new gravel material
❏ If the non-realigned parts of the existing carriageway are not to be regravelled
SMPi aw = [(1 - Pconew ) * SMPi bw + Pconew * SMPi rw ] ...(4.23)
where:
SMPiaw surface material property i after realignment works, (i = P075, P425,

P02, PI, D95)
SMPibw surface material property i before realignment works, (i = P075,
P425, P02, PI, D95)
SMPirw surface material property i for the realigned road segments after
works, (i = P075, P425, P02, PI, D95)
! Roughness
Roughness after realignment works (RIaw) is reset to a user-specified value. If this is not
specified, reset RIaw to the minimum allowable roughness (QIMIN) converted into IRI
units of m/km.
! Gravel age
The gravel age (GAGE) after realignment works is reset as follows:

❏ If the non-realigned parts of the existing carriageway are to be regravelled, GAGE is
reset to zero.
❏ If the non-realigned parts of the existing carriageway are not to be regravelled,
GAGE is calculated as follows:
GAGE aw = [(1 - Pconew ) * GAGE bw ] ...(4.24)
where:
GAGEaw gravel age after realignment works (returns an integer value, in years)
GAGEbw gravel age before realignment works (years)
! Material loss calibration factors
After realignment, the material loss calibration factors (that is, Kgl and Kkt) are reset to
! Speed factors
type.

Version 1.0
5 Construction works
Construction works for unsealed roads comprises the following:
! Upgrading (see Section 5.1)
! Dualisation (see Section 5.2)
! New section (see Section 5.3
5.1 Upgrading
An unsealed road can be upgraded to a bituminous or concrete pavement. It is also possible to
upgrade an earth road to a gravel road, although both are of the same surface class.
Upgrading works can be defined as follows:
An upgrading works of fixed specifications is applied at a specified time, defined by a

calendar year.
An upgrading works of fixed specifications is applied when the user-specified

intervention criteria are met. Upgrading is not performed if the last applicable year has
been exceeded.
An upgrading works are specified by the following information:
! Road type.
! Road class.
! Length adjustment factor.
! Increase in width.
! Additional number of lanes.
! Geometry details of the entire road section.
! New pavement details (according to the new pavement type).
! Other modelling parameters that depend on the new pavement type (for example,
construction quality indicators for bituminous roads).

If performed, the amount of upgrading works in kilometres of road length is given by:
LUPGRD = L aw ...(5.1)
where:
LUPGRD amount of upgrading works (km)

Law new road length after works (= Lbw*LF) (km)

Version 1.0
The total cost of upgrading (CSTUPGRD) is obtained from the product of LUPGRD and the
SALVA = PCTSAV * CSTUPGRD ...(5.2)
where:

5.1.2 Effects of upgrading

After upgrading, the pavement type is reset to the new type specified by the user. Depending
upon the new pavement type, the required modelling parameters are obtained in the following
ways:
! Pavement structure, strength, layer material properties and construction quality are set to
! Pavement condition after works is reset to as new.
! Pavement history data is reset to reflect new construction.
! Calibration factors are user-specified.
! Carriageway width and number of lanes
The new carriageway width after upgrading is calculated using Equation 4.6 above.
The increase in carriageway width is either specified directly by the user, or calculated
using Equation 4.7 above.
The number of lanes after upgrading works (NLANESaw) is equal to the number of lanes
before works (NLANESbw) plus the user-specified additional number of lanes (ADDLN).
! Speed factors
type.
5.2 Dualisation
The dualisation of an existing road section can be either scheduled or triggered by responsive
intervention criteria.
Note: The modelling of dualisation works is not included in this release.

Version 1.0
5.3 New section

intervention criteria. In a project analysis, a new section (that is, a new link) can be specified
as a section alternative within a selected project alternative, as described in the Applications
Guide.
The required components of the new section to be constructed is defined using the following
information:
! Road section data
All the data items that are required to define a road section in HDM-4. The user is able
to specify these data items in aggregate terms.
! Traffic data
❏ diverted traffic - traffic that is diverted from the nearby routes and other transport
modes
❏ generated traffic – additional traffic that occurs in response to the new investment
! Construction costs, duration and salvage value
! Maintenance and improvement standards
To be applied after opening the new section to traffic.

constructed (NEWCON) and this is equal to the new road section length. The total cost of
construction is obtained from the product of NEWCON and the user-specified unit cost per
kilometre.

Version 1.0
6 References
World Bank Publications, Washington D.C.

Version 1.0
Contents

Descriptions: Part E
E1 Overview
1 Introduction E1-1
2 Vehicle classification system E1-2
3 Vehicle speeds and operating costs E1-5
4 Non-motorised transport E1-6
5 Road safety E1-7
6 Total road user costs E1-8
7 References E1-9
E2 Vehicle Speeds and Operating Costs

1 Introduction E2-1
2 Modelling concepts and logic E2-2
2.1 Representative vehicles E2-2
2.2 Primary modelling parameters E2-2
2.3 Computational procedure E2-3
3 Free speeds E2-7
3.1 Free speed model E2-7
3.2 The constraining speeds E2-10
3.3 One-way traffic sections E2-22
4 Traffic congestion modelling E2-23
4.1 Modelling framework E2-23
4.2 Congested speeds E2-24
4.3 Vehicle operating speed E2-27
4.4 Annual average vehicle operating speed E2-27
4.5 Annual average traffic speed E2-28
4.6 Acceleration effects E2-28
5 Fuel consumption E2-34
5.1 Modelling approach E2-34
Analytical Framework and Model descriptions i

Version 1.0
PART E ROAD USER EFFECTS CONTENTS
5.2 Power requirements E2-35

5.3 Efficiency factor E2-45
5.4 Additional fuel consumption E2-45
5.5 Fuel consumption per 1000 vehicle-km E2-46
5.6 Annual average fuel consumption E2-47
6 Lubricating oil consumption E2-48
7 Tyre consumption E2-50
7.2 Rate of tread wear E2-51
7.3 Tyre consumption per 1000 vehicle-km E2-54
7.4 Modification to the tyre consumption model E2-55
7.5 Annual average tyre consumption E2-56
8 Vehicle utilisation and service life E2-57
8.1 Utilisation E2-57
8.2 Service life E2-58
9 Parts consumption E2-60
9.2 Roughness effects E2-61
9.3 Vehicle age effects E2-62
9.4 Acceleration effects E2-63
9.5 Annual average parts consumption E2-63
10 Labour hours E2-65
10.1 The model E2-65
10.2 Annual average labour hours E2-66
11 Capital costs E2-67
11.2 Depreciation E2-68
11.3 Interest E2-69
11.4 Annual average capital cost E2-70
12 Crew hours E2-71
12.2 Annual average number of crew hours E2-71
13 Overhead costs E2-72
13.2 Annual average overhead costs E2-72
14 Passenger travel time E2-73
14.1 Working passenger-hours E2-73
14.2 Non-working passenger-hours E2-73
14.3 Annual average number of passenger-hours E2-74
Analytical Framework and Model descriptions ii

Version 1.0
15 Cargo holding time E2-75

16 Road impassability costs E2-76
17 Vehicle resource costs E2-77
17.1 Unit costs E2-77
17.2 Vehicle-trip costs over the road section E2-77
17.3 Annual cost streams E2-78
17.4 Annual vehicle-kilometres E2-78
18 References E2-79
E3 Non-Motorised Transport
1 Introduction E3-1
2 Modelling concepts and logic E3-2
2.1 NMT vehicle types E3-2
2.2 Modelling issues E3-3
2.3 Data requirements E3-4
2.4 Computational logic E3-4
3 Impact of NMT on motorised transport E3-6
3.1 Impact on MT speed E3-6
3.2 Impact on MT operating costs E3-6
4 NMT speeds E3-7
4.1 Factors influencing NMT speeds E3-7
4.2 The speed model E3-7
4.3 VROUGH E3-8
4.4 VGRAD E3-9
4.5 Resistance to motion E3-10
5 NMT time and operating costs E3-14
5.1 Travel time cost E3-14
5.2 Operating cost E3-15
5.3 Capital cost E3-15
5.4 Repair and maintenance cost E3-16
5.5 Crew cost E3-17
5.6 Energy cost E3-17
5.7 Overheads E3-18
6 Estimation of economic benefits E3-19
7 References E3-20
E4 Road Safety
1 Introduction E4-1
2 Modelling logic E4-2
Analytical Framework and Model descriptions iii

Version 1.0
2.1 Accident types E4-2

2.2 Accident rates E4-2
2.3 Primary data E4-3
2.4 Computational procedure E4-3
2.5 Number of accidents E4-4
2.6 Accident costs E4-4
3 Economic analysis and comparisons E4-5
3.1 Economic analysis E4-5
3.2 Net change in the number of accidents E4-5
3.3 Outputs E4-5
4 References E4-6
Analytical Framework and Model descriptions iv

Version 1.0
Part E Road Map
Analytical
AnalyticalFramework
Framework
and
andModel
Model
Descriptions
Descriptions
Introduction
Introduction
Part
PartAA
Traffic
Traffic
Part
PartBB
RD
RDModel WE RUE
RUEModel SEE
SEEModel
Model WEModel
Model Model Model
Part
PartCC Part Part
PartEE Part
PartFF
PartDD
Economic
EconomicAnalysis
Analysis
PartGG
Nomenclature
Nomenclature
PartHH

Glossary
Part
PartI I
Figure E Analytical Framework and Descriptions Road Map

Version 1.0
PART E ROAD USER EFFECTS
E1 Overview
1 Introduction
The modelling of Road User Effects (RUE) in HDM-4 (see Figure E1.1) comprises analysis
of the following:
! Motorised vehicle (MT) speed, operating costs and travel time (see Figure E1.2)
! Non-motorised transport (NMT) speed and operating costs (see Figure E1.3)
! Road safety (see Chapter E4)
This chapter gives an overview of the HDM-4 vehicle classification system; and describes the
different RUE components considered in HDM-4.
Road
RoadUser
UserEffects
Effects
Overview Vehicle
VehicleSpeeds
Speeds Non-Motorised
Non-Motorised Road
RoadSafety
Safety
Overview
and
andOperating
Operating Transport
Transport
Chapter
ChapterE1
E1 Chapter
ChapterE4
E4
Costs
Costs Chapter
ChapterE3
E3
Chapter
ChapterE2
E2
Figure E1.1 Road User Effects modules
Analytical Framework and Model Descriptions E1-1

Version 1.0
PART E ROAD USER EFFECTS E1 OVERVIEW
2 Vehicle classification system

The vehicle classification system uses a flexible approach in which vehicles are divided into
motorised and non-motorised categories, and each category is divided into vehicle classes
(Kerali et al., 1994). A class comprises several vehicle types or representative vehicles,
which can be user specified based on one of several standard vehicle types. This approach
suits the needs of many countries and satisfies all the analytical requirements in the system.
Thus, vehicles are defined by a three-level hierarchy:
1 Categories
Differentiates between motorised and non-motorised transport.

2 Classes
Groupings of similar vehicles, for example passenger cars, trucks.

3 Types
Individual representative vehicle types, for which a set of RUE relationships has been
provided.
Figure E1.2 shows the hierarchical representation of motorised vehicles into categories,
classes and types (NDLI, 1995). Figure E1.3 shows a similar hierarchical representation of
non-motorised transport (PADECO, 1996).

Version 1.0
Motorised
Motorised Categories
Motor
Motorcycles
cycles Passenger
Passengercars
cars Utilities
Utilities Trucks
Trucks Buses
Buses Classes
Motor
Motorcycles Small
Smallcar Medium Light
Lighttruck Medium
cycles car Medium truck Medium truck
truck Types
(1)
(1) (2)
(2) car
car (8)
(8) (9)
(9)
(3)
(3)
Large
Largecar
car
(4)
(4)
Heavy
Heavytruck
truck Articulated
Articulated truck
truck
(10)
(10) (11)
(11)
Light Light
Lightgoods Mini-bus Light
Light goods Mini-bus Lightbus
bus
delivery
delivery vehicle
vehicle (12) (13)
(12) (13)
vehicle
vehicle (6)
(6)
(5)
(5)
Four
Fourwheel
wheel Medium
Mediumbusbus Heavy
Heavybus
bus
drive
drive (14)
(14) (15)
(15)
(7)
(7)
Coach
Coach
(16)
(16)
Source: NDLI (1995)
Figure E1.2 Definition of motorised vehicle categories, classes and types

Version 1.0
Fleet
Fleetrepresentation
representation
Categories
Motorised
Motorisedtransport
transport Non-motorised
Non-motorised
(MT)
(MT) transport
transport(NMT)
(NMT)
Pedestrian
Pedestrian Bicycle
Bicycle Cycle Animal
Cycle Animalcart
cart Farm
Farmtractor
tractor Classes
rickshaw
rickshaw
Pull
Pulltype
type Push
Pushtype
type Horse
Horsecart
cart Ox
Oxcart
cart Types
Pedestrian
Pedestrian Bicycle
Bicycle Small
Small Large
Large
Source: PADECO (1996)
Figure E1.3 Definition of non-motorised vehicle categories, classes and types

Version 1.0
3 Vehicle speeds and operating costs

Motorised vehicle speeds and operating resources are determined as functions of the
characteristics of each type of vehicle and the geometry, surface type and current condition of
the road, under both free flow and congested traffic conditions. The operating costs are
obtained by multiplying the various resource quantities by the unit costs or prices, which are
specified by the user in financial or economic terms.
Financial costs represent the actual costs incurred by transport operators in owning and
operating vehicles over the road. Economic costs represent the real costs to the economy of
that ownership and operation, where adjustments are made to allow for market price
distortions such as taxes, subsidies, foreign exchange restrictions, labour wage laws, etc.,
(Watanatada et al., 1987).
The following components of vehicle operating cost (VOC) are considered (see Chapter E2):
! Fuel consumption
! Lubricating oil consumption
! Tyre wear
! Parts consumption
! Maintenance labour hours
! Depreciation
! Interest
! Crew hours
! Overheads
Travel time is considered in terms of passenger-hours during working and non-working time,
and cargo holding hours. Travel time costs are expressed more appropriately only in economic
terms. Additional costs due to impassability of seriously damaged unsealed roads are also
included in the total amount of motorised road user cost.

Version 1.0
4 Non-motorised transport
Non-motorised transport (NMT) modes such as bicycles, cycle rickshaws, animal carts, and
pedestrians play a major role in moving passengers and freight in many countries (see Chapter
E3). The use of NMT is increasing in some regions mainly because of their affordability,
flexibility and cost-effectiveness in providing low cost transportation. Furthermore, the
increasing focus on efficiency in energy use, and the environmental impacts arising from the
ever-increasing use of motorised transport (MT), has highlighted the need for better provision
of NMT facilities. This has led to the recognition that the full range of transport needs in many
countries would not be catered for adequately by MT alone. Therefore, investment policies in
the road transport sector should include NMT issues.
A formal method has been developed for calculating the operating costs incurred by NMT on
roads and thereby for estimating the benefits derived by NMT from road improvements
(Odoki and Kerali, 1999). The presence of NMT can influence the speed of motorised
transport, thereby affecting the operating costs of motorised vehicles. In addition, policies
such as road improvements influence the costs and benefits to both motorised and non-
motorised road users.

Version 1.0
5 Road safety
The HDM-4 system allows users to define a series of look-up tables for accident rates. These
are basically broad, macro descriptions of the expected accident rates defined according to a
particular set of road and traffic attributes (for example, road type, traffic level and flow
pattern, presence of NMT, and geometry class). This tabular approach to implementing road
safety analysis was recommended (ISOHDM, 1995) following a detailed review of various
road safety studies, modelling and analysis methods. For each road type or intersection type,
users are required to specify the accident rate for each severity (that is, fatal, injury or damage
only), in terms of the numbers of accidents per 100 million vehicle-kilometres. When a road is
improved (for example, providing separate NMT lanes, and widening of road shoulders) a
new set of accident rates can be specified based on data observed for roads with similar traffic
flow and geometric characteristics. Thus, it is possible to analyse the change in total numbers
of accidents and the costs resulting from the improvement.

Version 1.0
6 Total road user costs

The total road user cost comprises:
! Motorised transport (MT) vehicle operating cost
! MT travel time cost
! Non-motorised transport (NMT) time and operating cost
! Accident cost
The annual road user cost for each investment option is given by:
RUC j = VOC j + TTC j + NMTOC j + AC j ...(6.1)
where:
RUCj road user cost under investment option j (currency)

VOCj MT vehicle operating cost under investment option j (currency)
TTCj MT travel time cost under investment option j (currency)
NMTOCj NMT time and operating cost under investment option j (currency)
ACj annual accident cost under investment option j (currency)

Version 1.0
7 References
ISOHDM, (1995)
Predicting changes in accident rates in developing countries following modifications in
road design
International Study of Highway Development and Management
School of Civil Engineering, The University of Birmingham
Kerali H.G.R., Odoki J.B., and Wightman D.C., (1994)
Vehicle Fleet Representation, Draft Working Paper
School of Civil Engineering, The University of Birmingham
NDLI, (1995)
Modelling Road User Effects in HDM-4 - Final Report
Asian Development Bank Project RETA 5549. International Study of Highway
Development and Management Tools
Odoki J.B., and Kerali H.G.R., (1999)
Modelling Non-motorised Transport in HDM-4 - TRB, Paper No. 991129
Transportation Research Board, 78th Annual Meeting, Washington D.C., USA
PADECO Co. Ltd., (1996)
Non-Motorised Transport (NMT) Modelling in HDM-4 - Draft Final Report (second
Version)
International Study of Highway Development and Management Tools. Transport
Division, The World Bank, Washington D.C., USA
The World Bank, John Hopkins University Press

Version 1.0
E2 Vehicle Speeds and Operating Costs
1 Introduction
This chapter describes the implementation of Road User Effects (RUE) models for calculating
motorised vehicle speeds, operating costs and travel time (see Figure E2.1). It provides an
overview of the modelling concepts and logic, and a description of the relationships and
default parameter values for each of the RUE components and HDM-4 standard representative
vehicles. For further details and background to the equations refer to NDLI (1995) and
Watanatada et al. (1987a).
Road
RoadUser
UserEffects
Effects
Overview Vehicle
VehicleSpeeds
Non-Motorised Road
RoadSafety
Safety
Overview
and
andOperating
Operating Transport
Transport
Chapter
ChapterE1
E1 Chapter
ChapterE4
E4
Costs
Costs Chapter
ChapterE3
E3
Chapter
ChapterE2
E2
Figure E2.1 Road User Effects module
After describing the modelling concepts and logic the chapter is divided into four modules as
follows:
! Module A (Sections 3 and 4)
Describes the methods of calculating the different vehicle speed components.

! Module B (Sections 5 - 13)
Describes models of the vehicle operating resources.

! Module C (Sections 14 - 16)
Discusses travel time and unsealed road impassability.

! Module D (Section 17)
Costs vehicle resources.


Version 1.0
PART E ROAD USER EFFECTS E2 VEHICLE SPEEDS AND OPERATING COSTS
2 Modelling concepts and logic
2.1 Representative vehicles

HDM-III allowed up to 10 representative vehicles in a single analysis (Watanatada et al.,
1987a). HDM-4 is more flexible with respect to the number of representative vehicles that can
be used in the analysis. The user may define any number of vehicles based on 16 motorised
representative vehicle types. This enables the user to model, for example, several heavy
trucks, each having different loading patterns. The HDM-4 standard representative vehicles
are given in Table E2.1.
For modelling RUE, it is necessary to assign certain key characteristics to the representative
vehicles. These include:
! Vehicle physical attributes
For example, number of axles, number of wheels, etc.

! Performance characteristics such as driving power and braking power
! Vehicle utilisation and service life
The basic data for each of the 16 representative vehicle types are also given in Table E2.1.
These values were estimated from a variety of sources, as reported in NDLI (1995).
2.2 Primary modelling parameters

The primary data which are required together with the key vehicle characteristics for
modelling RUE, may be grouped as follows:
! Road geometry
Includes the road alignment data, speed limit, roadside friction factor, section length,
width and the number of lanes.
! Speed-flow relationship
Includes road capacity and the parameters that determine the vehicle operating speeds
and flow characteristics at different traffic flow levels.
! Traffic flow pattern
Includes parameters that describe the road use in terms of the hourly traffic flow
distribution, and are used to determine the traffic flow in passenger car space equivalents
per hour (PCSE/h) for each traffic flow period.
! Road condition
Comprises data on the annual average surface roughness and texture depth for the road
section. These are obtained from the output data of the Road Deterioration module.
! Traffic
Comprise traffic volumes specified in terms of AADT (annual average daily traffic),
traffic composition and growth for each road section. These data should be obtained from
the traffic models.
! Unit costs
Includes costs of vehicle resources, (for example, fuel cost per litre, crew wages, new
vehicle price, tyre cost, etc.) and value of time. The vehicle resource data should be

Version 1.0
defined in economic and financial terms, but the value of time should be specified only in
economic terms.

The overall computational procedure for modelling motorised vehicle speeds, operating costs
and travel time for each section alternative, for each vehicle type in a given analysis year can
be summarised by the following steps:
1 Calculate vehicle speeds
The following speed components are calculated for a given road section:
(a) Free speed of each vehicle type
(b) Congested speeds by vehicle type - these are the operating speeds at different traffic
flow levels
(c) Annual average operating speed of each vehicle type
(d) Annual average traffic speed – this is the weighted average speed of all the vehicles
in a traffic stream.
2 Compute quantities of vehicle operating resources
in the following order:
(a) Fuel
(b) Lubricating oil
(c) Tyre
(d) Spare parts
(e) Maintenance labour hours
(e) Capital costs (comprises depreciation and interest)
(f) Crew hours
(g) Overhead
3 Calculate travel time
in terms of passenger-hours during working and non-working time, and cargo holding
hours
4 Cost vehicle resources and travel time
by applying unit costs to the predicted quantities of the resources consumed
5 Calculate increased operating cost
due to reduced passability on seriously damaged unsealed roads
6 Summarise and store data
for use in subsequent analysis and for reporting

Version 1.0
Table E2.1 Default representative vehicle classes and basic characteristics
Vehicle Type Description Abbre- Fuel Number Number of Aero- Projected Tare Operating
Number viation type of axles wheels dynamic frontal area weight weight
drag Coeff.
(m2) (t) (t)
1 Motorcycle Motorcycle or scooter MC P 2 2 0.70 0.8 0.1 0.2

2 Small car Small passenger cars PC-S P 2 4 0.40 1.8 0.8 1.0
3 Medium car Medium passenger cars PC-M P 2 4 0.42 1.9 1.0 1.2
4 Large car Large passenger cars PC-L P 2 4 0.45 2.0 1.2 1.4
5 Light delivery Panel van, utility or pickup truck LDV P 2 4 0.50 2.0 1.3 1.5
vehicle
6 Light goods vehicle Very light truck for carrying LGV P 2 4 0.50 2.8 0.9 1.5
goods (4 tyres)
7 Four wheel drive Landrover/Jeep type vehicle 4WD P 2 4 0.50 2.8 1.5 1.8
8 Light truck Small two-axle rigid truck LT D 2 4 0.55 4.0 1.8 2.0
(approx. < 3.5 t)
9 Medium truck Medium two-axle rigid truck MT D 2 6 0.60 5.0 4.5 7.5
(> 3.5 t)
10 Heavy truck Multi-axle rigid truck HT D 3 10 0.70 8.5 9.0 13.0
11 Articulated truck Articulated truck or truck with AT D 5 18 0.80 9.0 11.0 28.0
drawbar trailer
12 Mini-bus Small bus based on panel van MNB P 2 4 0.50 2.9 1.1 1.5
chassis (usually 4 tyres)
13 Light bus Light bus (approx. < 3.5 t) LB D 2 4 0.50 4.0 1.75 2.5
14 Medium bus Medium bus (3.5 - 8.0 t) MB D 2 6 0.55 5.0 4.5 6.0
15 Heavy bus Multi-axle or large two-axle bus HB D 3 10 0.65 6.5 8.0 10.0
16 Coach Large bus designed for long COACH D 3 10 0.65 6.5 10.0 15.0
distance travel
Source: NDLI (1995)
Notes:
1 Fuel type P=Petrol, D=Diesel
2 Classification of vehicle types are shown in Chapter E1

Version 1.0
Module A: Vehicle Speeds

The average speed of each vehicle type is required for calculating vehicle operating costs,
travel time, energy use and emissions. The speeds of MT vehicles are influenced by a number
of factors, which include:
! Vehicle characteristics
! Road severity characteristics, for example, road alignment, pavement condition, etc.
! The presence of non-motorised transport (NMT) (see Section 3.2.5)
! Roadside friction, for example, bus stops, roadside stalls, access points to roadside
development, etc. (see Section 3.2.5)
! Total MT traffic volume (see Section 4)
The methods of calculating the different vehicle speed components are described below:
! Free speed (see Section 3)
! Average operating speeds at different traffic flow levels (see Section 4)

Version 1.0
3 Free speeds
This is defined as the speed of each vehicle travelling on uncongested wide sections of road in
the environment under investigation. Free speeds are required for determining the operating
speeds of each vehicle type on a given road section under different flow characteristics.
The modelling of free speed described in this section is focused on the individual
representative vehicle types.
3.1 Free speed model

The free speeds are calculated using a mechanistic/behavioural model predicting that the
steady-state speed for each vehicle type k is a probabilistic minimum of five constraining
speeds based on driving power, braking capacity, road curvature, surface roughness and the
desired speed. The expression steady-state implies that the effects of speed variations or speed
change cycles along the road section are not considered. The speed constraints, generated by
the interaction of road severity factors and the relevant characteristics of the vehicle, are
described in Section 3.2. A more detailed representation of the methodology and its validation
is given in Watanatada et al. (1987).
Free speed analysis over a road section is carried out separately for each of the possible two
traffic-flow directions, known as the uphill segment and the downhill segment, and the results
averaged for a round trip. Both idealised homogeneous segments have the same
characteristics, except that the uphill is of positive grade and the downhill is of negative grade.
The modelling of the uphill speed, the downhill speed and the average round trip speed is
described in sections 3.1.1, 3.1.2 and 3.1.3, respectively.
3.1.1 Uphill segment speed

The free speed for the uphill segment is given by the expression:
σ 2 
exp  
 2 
VS ku = β
 1 1 1 1 1
 1 β  1 β  1 β  1 β  1 β 
  +   +   +   +   
  VDRIVEu   VBRAKEu   VCURVE   VROUGH   VDESIR  
 
...(3.1)
where:
VSku the predicted steady-state speed for the uphill segment (m/s)
VDRIVEu the speed limited by gradient and used driving power for the uphill segment
(m/s) (see Section 3.2.1)
VBRAKEu the speed limited by gradient and used braking power for the uphill segment
(m/s) (see Section 3.2.2)

Version 1.0
VCURVE the speed limited by curvature (m/s) (see Section 3.2.3)

VROUGH the speed limited by roughness (m/s) (see Section 3.2.4)
VDESIR the desired speed under ideal conditions (m/s) (see Section 3.2.5)
σ SPEED_SIG Weibull model parameter (see Table E2.2)
β SPEED_BETA Weibull model parameter (see Table E2.2)
The model parameter β determines the shape of the assumed Weibull distribution of the
constraining speeds. When β approaches zero, the mean speed would be equal to the
minimum of the five constraining speeds. The greater the value of β the further away the
predicted mean speed will be from the constraining speed.
As described in Watanatada et al. (1987), the model parameter σ is an estimate of the
standard error of residuals in the estimation, which involves a logarithmic transformation. The
numerator of Equation 3.1 above gives the value of the bias correction factor.
3.1.2 Downhill segment speed

The free speed for the downhill segment is given by the expression:
σ2 
exp  
 2 
VS kd = β
 1 1 1 1 1
  1  β  1  β  1  β  1  β  1  β 
  +   +   +   +   
  VDRIVEd   VBRAKEd   VCURVE   VROUGH   VDESIR  
 
...(3.2)
where:
VSkd the predicted steady-state speed for the downhill segment (m/s)
VDRIVEd the speed limited by gradient and used driving power for the downhill segment
(m/s)
VBRAKEd the speed limited by gradient and used braking power for the downhill segment
(m/s)
All other parameters are as defined above in Section 3.1.1.

Version 1.0
Table E2.2 Default steady-state speed model parameters
Vehicle Model Parameters VDRIVE VBRAKE

Number
SPEED_ SPEED_ PDRIVE PBRAKE CGR_a0 CGR_a1 CGR_a2
SIGMA BETA
σ β (kW) (kW)
1 0 0.151 12 5 94.9 0.85 2.80
2 0 0.151 26 20 94.9 0.85 2.80
3 0 0.151 33 20 94.9 0.85 2.80
4 0 0.151 36 20 94.9 0.85 2.80
5 0 0.151 40 25 94.9 0.85 2.80
6 0 0.151 40 20 94.9 0.85 2.80
7 0 0.151 45 25 94.9 0.85 2.80
8 0 0.191 50 45 94.9 0.85 2.80
9 0 0.164 87 70 94.9 0.85 2.80
10 0 0.110 227 255 94.9 0.85 2.80
11 0 0.110 227 255 94.9 0.85 2.80
12 0 0.151 40 26 94.9 0.85 2.80
13 0 0.191 50 45 94.9 0.85 2.80
14 0 0.191 65 70 94.9 0.85 2.80
15 0 0.110 120 120 94.9 0.85 2.80
16 0 0.110 180 180 94.9 0.85 2.80
Source: Bennett and Greenwood (1996)
3.1.3 Round trip speed

The average speed for a round trip is computed to correspond to the space-mean speed over
the two segments, that is, the round trip distance divided by the round trip time. This is given
by the harmonic mean of the uphill and downhill speeds as follows:
7.2
Sk = ...(3.3)
 1   1 
 VS + 
ku   VS kd 
where:
Sk the average steady-state free speed (km/h) for vehicle type k

Version 1.0
3.2 The constraining speeds

The following sub-sections describe how the constraining (or limiting) speeds are calculated.
3.2.1 Limiting speed based on road gradient and engine power (VDRIVE)
The limiting speed due to driving power is related to the used driving power and the road
gradient through the balance of forces in the absence of acceleration. It is calculated by
solving the following cubic equation, which is based on the hypothesis that the vehicle is
driven at steady-state speed on a smooth, straight road:
1000 * PDRIVE = z0 * VDRIVE 3 + z1 * VDRIVE ...(3.4)
where:
PDRIVE used driving power (kW)

z0 and z1 are functions of the forces opposing motion (see below)
The used driving power is generally less than the rated power of the engine. The forces
opposing motion, under the hypothesis given above, are the aerodynamic resistance, gradient
resistance and rolling resistance (see Section 5.2).
The parameters z0 and z1 are calculated as follows:
z0 = 0.5 * RHO * CDmult * CD * AF + b13 * CR1 * CR2 * FCLIM
b11 * CR2 * FCLIM * NUM_WHEELS 

z1 =  
+ b12 * CR1 * CR2 * FCLIM * WGT_OPER + WGT_OPER * g * GR
where:
RHO mass density of air (kg/m3) (default value = 1.20)

Cdmult CD multiplier
CD aerodynamic drag coefficient
AF 2
projected frontal area of the vehicle (m )
CR1 tyre type dependent coefficient of rolling resistance
CR2 pavement dependent coefficient of rolling resistance
FCLIM climatic adjustment factor
NUM_WHEELS number of wheels per vehicle
WGT_OPER vehicle operating weight (kg)
g 2
acceleration due to gravity taken as equal to 9.81 m/s
GR average gradient of the road section (as a fraction )
b11, b12 and b13 rolling resistance parameters

Version 1.0
The mass density of air is required for computing the aerodynamic resistance, and is given by
St. John et al. (1978) as:
(
RHO = 1.225 * 1 - 2.26 * ALT * 10 -5 )4.255 ...(3.5)
where:
ALT road altitude, defined as the elevation of the road section above the mean
sea level (m)
Rolling resistance is calculated as a function of tyre and pavement characteristics, and climatic
factors.
The tyre factor CR1 depends upon the tyre type as follows:
CR1 = 1.0 if TYRE_TYPE = Radial
CR1 = 1.3 if TYRE_TYPE = Bias-ply
The pavement dependent coefficient of rolling resistance CR2 is calculated as:
CR2 = Kcr2 * (CR_CR2_a0 + CR_CR2_a1 * TD av + CR_CR2_a2 * RIav )
...(3.6)
where:
Kcr2 rolling resistance factor

TDav average sand patch texture depth (mm) (TD is set to zero for concrete and
unsealed roads)
RIav average roughness (IRI m/km)
Table E2.3 gives the default values for the various rolling resistance model parameters.

Version 1.0
Table E2.3 Rolling resistance model parameters
WGT_OPER < = 2500 kg WGT_OPER > 2500 kg

Surface Surface type CR_ CR_ CR_ Kcr2 CR_ CR_ CR_ Kcr2
class CR2_ CR2_ CR2_ CR2_ CR2_ CR2_
a0 a1 a2 a0 a1 a2
Bituminous AM or ST 0.90 0.022 0.022 1 0.84 0.03 0.03 1
Concrete JP, JR or CR 0.90 0.022 0.022 1 0.64 0.03 0.03 1
Unsealed GR 1.00 0.00 0.075 1 1.00 0.00 0.075 1
Unsealed EA 0.80 0.00 0.10 1 0.80 0.00 0.10 1
Unsealed SA 7.50 0.00 0.00 1 7.50 0.00 0.00 1
Block CB, BR or SS 2.00 0.00 0.00 1 2.00 0.00 0.00 1
Source: NDLI (1995)
Notes: The following abbreviations are used:

AM = Asphalt Mix,
ST = Surface Treatment,
JP = Jointed Plain,
JR = Jointed Reinforced,
CR = Continuously Reinforced,
CB = Concrete Block,
BR = Brick,
SS = Set Stone.
For the definition of surface types see Chapter C1.

The diameter and number of wheels influence rolling resistance as follows:
b11 = CR_B_a0 * WHEEL_DIA
CR_B_a1
b12 =
WHEEL_DIA
CR_B_a2 * NUM_WHEELS
b13 =
(WHEEL_DIA )2
where:
WHEEL_DIA wheel diameter (m)

NUM_WHEELS number of wheels per vehicle
CR_B_a0 to model coefficients

Version 1.0
CR_B_a2
The default values for tyre parameters CR_B_a0 to CR_B_a2 are given in Table E2.4.
The rolling resistance depends upon the percentage of time travelled on snow covered
(PCTDS) and water covered (PCTDW) roads:
FCLIM = 1 + 0.003 * PCTDS + 0.002 * PCTDW
The average road section gradient GR is estimated from the following expression taken from
Watanatada et al. (1987a):
RF
GR ± ...(3.7)
1000
where:
RF average road rise plus fall (m/km)
Thus, solving the cubic equation with GR = + [RF/1000] would yield the value for VDRIVEu,
and solving it with GR = - [RF/1000] would yield the value for VDRIVEd.
The solution of the cubic equation is through Descartes’ rule of signs:
z1
z2 =
(3 * z0)
1000 * PDRIVE
z3 =
(2 * z0)
DT = z2 3 + z3 2
if DT > 0 :
VDRIVE = 3 DT + z3 − 3 DT - z3
if (√DT - z3) < 0, set (√DT - z3) = 0

if (√DT + z3) < 0, set (√DT + z3) = 0
if DT ≤ 0 :
  2π   4π 
VDRIVE = MAX r * cos(z ), r * cos z + , r * cos z + 
  3   3 
where:
1  − 2 * z3 
z= arccos 
3  z2 * r 
r = 2 * − z2

Version 1.0
The default values for the VDRIVE model parameters are given in Table E2.2.

Version 1.0
Table E2.4 Parameters for calculating aerodynamic, rolling and inertial resistance
Vehicle Aerodynamic resistance Rolling resistance parameters Inertial Resistance Parameters

number parameters
CD Aero. Projected Number Wheel Type of Tyre parameters
multiplier Drag frontal of wheels diameter tyre
Coeff. Area (m2)
(m)
CDMULT CD AF NUM_ WHEEL_ TYRE_ CR_B_a0 CR_B_a1 CR_B_a2 EMRAT_a0 EMRAT_a1 EMRAT_a2
WHEELS DIA TYPE
1 1.10 0.70 0.8 2 0.55 Bias 37 0.064 0.012 1.10 0 0
2 1.10 0.40 1.8 4 0.60 Radial 37 0.064 0.012 1.14 1.010 399.0
3 1.10 0.42 1.9 4 0.60 Radial 37 0.064 0.012 1.05 0.213 1260.7
4 1.10 0.45 2.0 4 0.66 Radial 37 0.064 0.012 1.05 0.213 1260.7
5 1.11 0.50 2.8 4 0.70 Radial 37 0.064 0.012 1.10 0.891 244.2
6 1.11 0.50 2.8 4 0.70 Bias 37 0.064 0.012 1.10 0.891 244.2
7 1.11 0.50 2.8 4 0.70 Bias 37 0.064 0.012 1.10 0.891 244.2
8 1.13 0.55 4.0 4 0.80 Bias 37 0.064 0.012 1.04 0.830 12.4
9 1.13 0.60 5.0 6 1.05 Bias 37 0.064 0.012 1.04 0.830 12.4
10 1.14 0.70 8.5 10 1.05 Bias 37 0.064 0.012 1.07 1.910 10.1
11 1.22 0.80 9.0 18 1.05 Bias 37 0.064 0.012 1.07 1.910 10.1
12 1.11 0.50 2.9 4 0.70 Radial 37 0.064 0.012 1.10 0.891 244.2
13 1.13 0.50 4.0 4 0.80 Bias 37 0.064 0.012 1.10 0.891 244.2
14 1.14 0.55 5.0 6 1.05 Bias 37 0.064 0.012 1.04 0.830 12.4
15 1.14 0.65 6.5 10 1.05 Bias 37 0.064 0.012 1.04 0.830 12.4
16 1.14 0.65 6.5 10 1.05 Bias 37 0.064 0.012 1.04 0.830 12.4

Version 1.0
3.2.2 Limiting speed based on road gradient and braking capacity (VBRAKE)
For uphill segments, the value of VBRAKE is infinite, (that is, the speed on upgrades is not
limited by the braking power).
VBRAKEu = ∞ ...(3.8)
The speed on downgrades is dependent upon the length of gradient. Once the gradient length
(GL) exceeds a critical value, the brakes are used to reduce the speed. Below this critical
gradient there is no effect of downgrade on speed.
The critical gradient length CGL is calculated as follows:
CGL = CGR_a0 * exp(CGR_a1 * GR ) + CGR_a2 ...(3.9)
where:
CGL critical gradient length (km)

GR average gradient of the road section (absolute value as a fraction)
a0 to a2 regression coefficients
Thus, the following two conditions need to be analysed for the downhill segment:
If GL < CGL
VBRAKEd = ∞ ...(3.10)
If GL > CGL,
VBRAKEd is obtained by solving the following cubic equation, which is formulated using the
mechanistic principle of balancing forces:
- 1000 * PBRAKE = z0 * VBRAKEd 3 + z1 * VBRAKEd ...(3.11)
where:
PBRAKE used braking power of the vehicle (kW)
z0 = 0.5 * RHO * Cdmult * CD * AF + b13 * CR1 * CR2 * FCLIM
z1 =  
b11 * CR2 * FCLIM * NUM_WHEELS
 + b12 * CR1 * CR2 * FCLIM * WGT_OPER + WGT_OPER * g * GR 
All other parameters are as defined above in Section 5.2.1.

The average road gradient length, GL, (km) is estimated from the expression given below:
1
GL = ...(3.12)
NUM_RF
where:

Version 1.0
NUM_RF average number of road rise and fall per kilometre (minimum value = 0.1)
The solution of the above cubic equation is through Descartes’ rule of signs:
z1
z2 =
3 * z0
-1000 * PBRAKE
z3 =
2 * z0
DT = z2 3 + z3 2
if DT ≥ 0
VBRAKEd = ∞
else if DT < 0
1  −2 * z3 
z= arccos 
3  z2 * r 
r = 2 * - z2
 4π 
VBRAKEd = r * cos z + 
 3 
The default values for the VBRAKE model parameters are given in Table E2.2 (see Section
5.1.2).
3.2.3 Limiting speed determined by road curvature (VCURVE)

The limiting curve speed is calculated as a function of the radius of curvature. It is based on
the postulate that drivers select their curve speed such that the side friction generated would
not cause the wheels to skid. The limiting curve speed, based on the work carried out by
McLean (1991), is given as:
VCURVE = VCURVE_a0 * R VCURVE_a1 ...(3.13)
where:
R average radius of road curvature (m)

VCURVE_a0
and regression parameters
VCURVE_a1
The average radius of road curvature, R, is estimated from the following expression taken
from Watanatada et al. (1987a):

Version 1.0
180,000
R= ...(3.14)
π * MAX(18 , C)
π
where:
The default values for the limiting curve speed parameters are given in Table E2.5.
Table E2.5 Default model parameters for VCURVE and VROUGH
Vehicle VCURVE VROUGH

Number
VCURVE_a0 VCURVE_a1 ARVMAX VROUGH_a0
(mm/s)
1 3.9 0.34 203 1.15
2 3.9 0.34 203 1.15
3 3.9 0.34 203 1.15
4 3.9 0.34 203 1.15
5 3.9 0.34 203 1.15
6 3.9 0.34 200 1.15
7 3.9 0.34 200 1.15
8 4.8 0.29 200 1.15
9 4.8 0.29 200 1.15
10 4.6 0.28 180 1.15
11 4.2 0.27 160 1.15
12 3.9 0.34 203 1.15
13 4.8 0.29 200 1.15
14 4.8 0.29 200 1.15
15 4.6 0.28 180 1.15
16 4.6 0.28 180 1.15
3.2.4 Limiting speed based on road roughness (VROUGH)

This is the constraining speed corresponding to the maximum allowable suspension motion of
the vehicle, which governs ride severity. The ride suspension motion is measured by the rate
of absolute displacements of the vehicle rear axle relative to the body, which is termed the
average rectified slope (ARS). ARS is usually expressed in units of m/km or mm/m. The
ride suspension motion is related to vehicle speed and road roughness as follows:
ARV = V * ARS ...(3.15)

Version 1.0
where:
ARV the average rectified velocity of suspension motion of the standard Opala-
Maysmeter vehicle in response to roughness (mm/s)
V the vehicle speed (m/s)
ARS the average rectified slope (mm/m)
The limiting speed due to the effect of road roughness is calculated as:
ARVMAX
VROUGH = ...(3.16)
VROUGH a0 * R1av
where:
ARVMAX the maximum allowable average rectified velocity of suspension motion of

the standard Opala-Maysmeter vehicle in response to roughness (mm/s)
VROUGH_a0 regression parameter
RIav average roughness of the road (m/km)
The default values for the roughness effects parameters are given in Table E2.5 (see Section
5.2.3).
3.2.5 Desired speed (VDESIR)

This is the speed at which a vehicle is assumed to be operated in the absence of the constraints
based on the vertical grade, curvature, ride severity, and traffic congestion, that is, the desired
speed on a flat, straight, smooth, uncongested road section. The desired speed is influenced by
driver’s behaviour in response to psychological, safety, cultural, and economic considerations
in addition to factors such as:
! Road width
! Roadside friction
! The presence of non-motorised transport
! Speed limits and enforcement
The desired speed in the absence of posted speed limits is calculated as:
VDESIR0 = VDES * XFRI * XNMT * VDESMUL ...(3.17)
where:
VDESIR0 desired speed in the absence of posted speed limits (m/s)

VDES desired speed adjusted for carriageway width effects (m/s)
XFRI speed reduction factor due to roadside friction for the section
(dimensionless, in the range 0.6 - 1; default = 1.0)

Version 1.0
XNMT speed reduction factor due to non-motorised transport for the section
(dimensionless, in the range 0.6 - 1; default = 1.0)
VDESMUL desired speed multiplication factor (dimensionless, in the range 0.85 - 1.3;
default = 1.0). The basic model is for two lane roads. This factor is used to
adapt the model for single lane roads and multi-lane highways
Since desired speeds are affected by road widths, the speed values (VDES) need to be
adjusted for these effects. The adjustment is based on the work carried out by Hoban et al.
(1994), which assumes that there is critical width (CW1) below which speeds will be
unaffected by carriageway width. Between this minimum speed (VDESMIN) and the desired
speed on two-lane roads (VDES2), there is a linear increase in the speed. On roads wider than
two-lanes, Yuli (1996) showed that there is a continued increase in speed, but at a much lower
rate.
The values of VDES are adjusted as follows:
if CW ≤ CW1 VDES = VDESMIN
if CW1 < CW ≤ CW2 VDES = VDESMIN + VDES_a3 * (CW - CW1)
if CW > CW2 VDES = VDES2 + VDES_a1 * (CW - CW2 )
where:

CW1 minimum carriageway width for a single lane road (m)
VDESMIN the minimum desired speed on a very narrow (single lane) straight road (m/s)
CW2 minimum width for a two lane road (m)
VDES2 desired speed on a two lane road (m/s)
VDES_a3 the rate of increase in desired speed for a single to two lane road (m/s per m
road width)
VDES_a1 the rate of increase in desired speed for a two or more lane road (m/s per m
road width) (see Table E2.6)
The speed/width slope between CW1 and CW2 is calculated from the expression:
( VDES2 - VDESMIN )
VDES_a3 =
(CW2 - CW1)
The relationship between the minimum desired speed and the desired speed on two-lane roads
is given as:
VDESMIN = VDES_a2 * VDES2
where:
VDES_a2 the ratio of the desired speed on a single lane road to the desired speed on a

Version 1.0
two lane road
The default values for the desired speed model parameters are given in Table E2.6.
Table E2.6 Default model parameters for VDESIR
Vehicle Desired speed

Number
(Bituminous surface roads)
VDES2 VDES_a0 VDES_a1 VDES_a2 CW1 CW2
(m/s)
1 40.0 0.0020 2.9 0.75 4 6.8
2 40.1 0.0020 2.9 0.75 4 6.8
3 34.8 0.0020 2.9 0.75 4 6.8
4 34.4 0.0020 2.9 0.75 4 6.8
5 42.0 0.0020 2.9 0.75 4 6.8
6 40.0 0.0020 2.9 0.75 4 6.8
7 39.2 0.0020 2.9 0.75 4 6.8
8 35.6 0.0028 0.7 0.75 4 6.8
9 29.3 0.0028 0.7 0.75 4 6.8
10 24.6 0.0033 0.7 0.75 4 6.8
11 29.1 0.0039 0.7 0.75 4 6.8
12 46.1 0.0020 0.6 0.75 4 6.8
13 34.4 0.0028 0.6 0.75 4 6.8
14 39.4 0.0028 0.6 0.75 4 6.8
15 24.8 0.0033 0.6 0.75 4 6.8
16 24.5 0.0033 0.6 0.75 4 6.8
A complete set of desired speed parameters is required for each of the four road surface
classes: bituminous, block, concrete, and unsealed.
The actual desired speed is the minimum of the desired speed and the enforced speed limit:
VDESIR = MIN VDESIR0,

PLIMIT * ENFAC 
...(3.18)
 3.6 
where:
PLIMIT posted speed limit (km/h)

ENFAC speed enforcement factor (default = 1.10)

Version 1.0
3.3 One-way traffic sections

For the analysis of one-way traffic sections, the user has to define whether the direction of
flow is generally uphill or downhill.
For uphill sections, the steady-state speed VSku is calculated as explained in Section 5.1.1. The
average steady-state free speed (km/h) is given as:
S k = 3.6 * VS ku ...(3.19)
For downhill sections, the steady-state speed VSkd is calculated as explained in Section 5.1.2.
The average steady-state free speed (km/h) is given as:
S k = 3.6 * VS kd ...(3.20)

Version 1.0
4 Traffic congestion modelling
4.1 Modelling framework

The speed-flow model adopted for motorised transport (MT) is the three-zone model
proposed by Hoban et al. (1994). This model is illustrated in Figure E2.2.
S1
S2
Speed km/h
S3
Snom
Sult
Qo Qnom Qult
Flow in PCSE/h
Figure E2.2 Speed-flow model
The following notation applies to Figure E2.2:
Qo the flow level below which traffic interactions are negligible in PCSE/h
Qnom nominal capacity of the road (PCSE/h)
Qult the ultimate capacity of the road for stable flow (PCSE/h)
Sult speed at the ultimate capacity, also referred to as jam speed (km/h)
Snom speed at the nominal capacity (km/h)
S1 to S3 free flow speeds of different vehicle types (km/h)
PCSE passenger car space equivalents (see Part B)
The model predicts that below a certain volume there are no traffic interactions and all
vehicles travel at their free speeds. Once traffic interactions commence, the speeds of the
individual vehicles decrease until the nominal capacity where all vehicles will be travelling at
the same speed, which is estimated as 85% of the free speed of the slowest vehicle type. The

Version 1.0
speeds can then further decrease towards the ultimate capacity beyond which unstable flow
will arise.
The values of the key parameters that define the speed-flow relationship vary depending upon
the road type and width (see Part B), and are specified by the user. The free speeds of different
vehicle types (S1, S2, etc.) and the speed at nominal capacity (Snom) are computed internally
as described in Section 3 and Section 4.2, respectively.
For a given road section and for each analysis year, the modelling of traffic congestion is
carried out for each traffic flow period. This is done separately for the uphill segment and the
downhill segment, and the results are combined to give an average for a round trip over the
road section. Further aggregation of the results is performed to obtain annual average values
for each vehicle type and for all the vehicles using the road section.
The computational procedure can be summarised as follows:
1 Define the speed-flow relationship
by determining the speed at nominal capacity and using the free speeds calculated, for:
(a) Uphill segment
(b) Downhill segment
2 For each vehicle type k and for each traffic flow period p calculate:
(a) Steady-state congested speed for the uphill segment, the downhill segment, and for a
round trip (see Section 4.2)
(b) Steady-state operating speed (see Section 4.3)
(c) Acceleration noise - a measure of the severity of speed changes (see Section 4.6)
3 For each vehicle type k calculate the annual average operating speed
(see Section 4.4)
4 Calculate the annual average traffic speed
(see Section 4.5)
(a) For all vehicles
(b) For heavy commercial vehicles only
4.2 Congested speeds

As traffic flows increase vehicle interactions also increase, and this leads to a reduction in
speed. The resulting reduced speed is modelled as the steady-state congested speed, since it
does not consider the effects of speed variations along the road section. The steady-state
congested speed is modelled by traffic flow period (for the uphill and the downhill segments)
and the values are used for calculating fuel consumption (see Section 5) and tyre consumption
(see Section 7).
4.2.1 Uphill segment speed

The speed at nominal capacity is equal to 85% of the free speed of the slowest vehicle type:
VSnomu = 0.85 * MIN(VS ku ) ...(4.1)
where:

Version 1.0
VSnomu speed at nominal capacity for the uphill segment (m/s)

VSku free speed of vehicle type k for the uphill segment (m/s)
The uphill segment speed, VU, at each traffic flow period p and flow Qp is calculated as
follows:
! for Qp < Qo
VU kp = VS ku ...(4.2)
! for Qo ≤ Qp ≤ Qnom
 (VS ku - VSnomu) * (Q p - Qo) 

VU kp = VS ku −   ...(4.3)
 (Qnom - Qo) 
! for Qnom < Qp ≤ Qult
 (VSnomu - VSult) * (Q p - Qnom) 

VU kp = VSnomu −   ...(4.4)
 (Qult - Qnom) 
The steady-state speed at each traffic flow period is adjusted as follows:
VU kp = MAX (VU kp * CALBFAC, VSult ) ...(4.5)
where:
VUkp steady-state congested speed of vehicle type k during period p for the uphill
segment (m/s)
Qp traffic flow during period p (PCSE/h)
CALBFAC speed calibration factor (default = 1.0, range 0.1 to 10). This is a road type
dependent translation factor for the speed-flow model
Here, the speeds Sult (km/h) are converted into VSult (m/s) by dividing them by 3.6.
The computed uphill segment speeds (VUkp) are used in the calculation of fuel consumption
(see Section 5).
4.2.2 Downhill segment speed

The speed at nominal capacity is calculated as:
VSnomd = 0.85 * MIN(VS kd ) ...(4.6)
where:
VSnomd speed at nominal capacity for the downhill segment (m/s)

VSkd free speed of vehicle type k for the downhill segment (m/s)

Version 1.0
The downhill segment speed, VD, (m/s) at each traffic flow period p and flow Qp (PCSE/h) is
! for Qp < Qo
VD kp = VD kd ...(4.7)
! for Qo ≤ Qp ≤ Qnom
 (VS kd - VSnomd) * (Q p - Qo) 

VD kp = VS kd −   ...(4.8)
 (Qnom - Qo) 
! for Qnom < Qp ≤ Qult
 (VSnomd - VSult) * (Q p - Qnom) 

VD kp = VSnomd −   ...(4.9)
 (Qult - Qnom) 
These speeds (VDkp) are used in the calculation of fuel consumption (see Section 5).
The steady-state speed at each traffic flow period is adjusted as follows:
VD kp = MAX (VD kp * CALBFAC, VSult ) ...(4.10)
where:
VDkp steady-state congested speed of vehicle type k during period p for the downhill
segment (m/s)
Qp traffic flow during period p (PCSE/h)
CALBFAC speed calibration factor (default = 1.0, range 0.1 to 10)
4.2.3 Round-trip congested speed

The average steady-state congested speed for a round trip (km/h) at each traffic flow period p
and flow Qp is calculated as follows:
 
 
 7.2 
S kp =   ...(4.11)
  1   1
+
 

  VU   VD  
  kp   kp  
where:
Skp average steady-state congested speed (km/h) for vehicle type k for traffic flow
period p

Version 1.0
4.3 Vehicle operating speed

To account for the bias introduced to the analysis through the use of the time mean speed
instead of the individual space mean speed, the individual congested speeds calculated above
in Section 4.1 are adjusted by multiplying them with a speed-bias factor. Spot speeds are
speeds measured as vehicles pass a point. Time mean speed is the arithmetic average speed of
all vehicles passing a point on the road over a specified time period. Space mean speed (also
called journey speed) is the average speed of all vehicles occupying a given section of road
over a specified time period.
Thus, the adjusted speed, herein called steady-state vehicle-operating speed, is calculated as
follows:
SS kp = S kp * SPEEDBIAS ...(4.12)
where:
SSkp vehicle operating speed (km/h) in traffic flow period p

Skp steady-state congested speed (km/h) for traffic flow period p
SPEEDBIAS speed adjustment factor to account for the bias introduced through the use
of the time mean speed instead of the space mean speed
The speed adjustment factor SPEEDBIAS is given by the expression:
SPEEDBIAS = 1.0000 + 0.0122 * COV - 0.8736 * COV 2 ...(4.13)
where:
COV coefficient of speed variation within the traffic stream (default = 0.15)
These speed values (SSkp) are used for the calculation of vehicle utilisation, crew hours,
passenger travel time, and cargo transit time and for reporting purposes.
4.4 Annual average vehicle operating speed

The annual average vehicle operating speed is calculated as follows:
 n


 ∑
p =1
HRYR p * HVp * SS kp 

SS kav =   ...(4.14)
 n




∑
p =1
HRYR p * HVp 


where:
SSkav annual average operating speed of vehicle type k (km/h)

SSkp Operating speed of vehicle type k during traffic flow period p (km/h)

Version 1.0
HRYRp the number of hours in traffic flow period p

HVp the hourly traffic flow in period p expressed as a proportion of AADT
These speeds (SSkav) are used for reporting purposes.
4.5 Annual average traffic speed

The annual average traffic speed for a road section option is required for the modelling of
pavement deterioration: edge-break, skid resistance and rutting due to wear by studded tyres,
and are calculated as follows:
K
∑ SS
k =1
kav
S= ...(4.15)
K
where:
S annual average traffic speed (km/h)

SSkav annual average operating speed of vehicle type k (for k = 1, 2, ..., K) (km/h)
The annual average speed of heavy vehicles is required for the modelling of pavement rutting
due to plastic deformation, and is calculated as follows:
KH
∑ SS
kH=1
kav
Sh = ...(4.16)
KH
where:
Sh annual average speed of heavy vehicles (km/h)

kH heavy vehicles (WGT_OPER > 3500 kg) (for kH = 1, 2, ..., KH)
4.6 Acceleration effects
4.6.1 Concepts
The speed-flow model presented in Figure E2.2 shows that as flows increase, there is an
increase in vehicle interactions and a decrease in speeds. The interactions are accompanied by
an increase in the frequency and magnitude of vehicle accelerations and decelerations. Under
ideal conditions drivers would maintain a steady-state speed without any decelerations and
accelerations. However, this is not possible in reality since drivers are forced to adjust their
speeds in response to traffic congestion, road alignment, pavement surface condition, the
presence of NMT and other roadside activities.
The standard deviation of accelerations, referred to as the acceleration noise, gives an
indication of the severity of speed changes. Low values of acceleration noise indicate that

Version 1.0
there are minor speed changes, large values indicate major speed changes. For each road
section, it is considered that the additional vehicle operating costs due to speed change cycles
is proportional to the magnitude of acceleration noise.
The total acceleration noise for a vehicle type k operating on a road section during traffic flow
period p is considered in two components:
1 Natural acceleration noise
2 Traffic induced acceleration noise
These are combined to give the total acceleration noise as:
σa kp = (σan k
2
+ σat kp
2
) ...(4.17)
where:
σakp total acceleration noise (m/s2) for vehicle type k in flow period p
σank natural acceleration noise for vehicle type k (m/s2)
σatkp traffic induced acceleration noise for vehicle type k in flow period p (m/s2)
4.6.2 Natural acceleration noise

The total natural acceleration noise arises due to the following:
1 Driver’s natural behaviour (on uncongested, straight, smooth road section without NMT
and other roadside activities)
2 Road alignment
3 Roadside friction
4 Non-motorised transport
5 Road surface roughness
The total natural acceleration noise (σank) for each vehicle type k is given by the following
expression:
[
σan k = MAX 0.1, MAX(σ adral k , σ asfk , σanmt k , σ airi k
2 2 2 2
)] ...(4.18)
where:
σadralk natural acceleration noise due to driver behaviour and road alignment (m/s2)
(default value = 0.1)
σasfk acceleration noise due to roadside friction
σanmtk acceleration noise due to non-motorised transport
σairik acceleration noise due to road roughness

Version 1.0
The driver behaviour acceleration noise (σadr) and the road alignment acceleration noise
(σaal) are combined into a single value (σdral) as it is difficult to differentiate between these
two components. The other three components of natural noise are modelled as linear
functions. This is done using the following relationships:
σasfk = 2.5 * (1.0 - XFRI) * FRIAMAX k ...(4.19)
σanmt k = 2.5 * (1.0 - XNMT) * NMTAMAX k ...(4.20)
  RIAMAX k 
σairi k = MINRIAMAX k ,  RIav *  ...(4.21)
  AMAXRIk 
where:
XFRI Speed reduction factor due to road side friction for the road section (0.6 to
1)
FRIAMAXk maximum acceleration noise due to side friction (default value = 0.20 m/s2
for all vehicle types)
XNMT Speed reduction factor due to non-motorised transport for the road section
(0.6 to 1)
NMTAMAXk maximum acceleration noise due to non-motorised transport (default value
= 0.40 m/s2 for all vehicle types)
RIAMAXk maximum acceleration noise due to roughness (default value = 0.30 m/s2 for
all vehicle types)
RIav average roughness for a road section (IRI m/km)
AMAXRIk roughness at which maximum acceleration noise RIAMAXk occurs (default
value = 20 for all vehicle types) (IRI m/km)
4.6.3 Traffic acceleration noise

The acceleration noise due to traffic interactions is modelled as a sigmoidal function
depending upon the volume-to-capacity ratio (VCR) and the flow level (Qo) at which traffic
interactions start on the speed-flow model illustrated in Figure E2.2. Thus, when traffic flow
at any flow period (Qp) is less than Qo, there is only natural acceleration noise; and when Qp
exceeds Qo there is a combination of natural and traffic acceleration noise.
The traffic induced acceleration noise for vehicle type k during traffic flow period p is
calculated as:
 1.04 
σat kp = σatmax k   ...(4.22)
 (
1 + exp a0 + a1 * VCR p )


where:
σatmaxk maximum traffic acceleration noise (m/s2) for vehicle type k

Version 1.0
VCRp volume to capacity ratio (or, relative flow) in period p

a0 and a1 regression coefficients
The relative flow in traffic flow period p is given by the expression:
Qp
VCR p = ...(4.23)
Q ult
The regression coefficients a0 and a1 are quantified as follows:
a0 = 4.2 + 23.5 * XQ12 ...(4.24)
a1 = - 7.3 - 24.1 * XQ12 ...(4.25)
where:
XQ1 the ratio of Qo to Qult
The maximum traffic acceleration noise for each vehicle type can be determined by taking
measurements of total acceleration noise under severely congested conditions. This gives the
maximum total acceleration noise on the road section, which includes both natural
acceleration noise and the maximum traffic induced acceleration noise. The maximum traffic
acceleration noise for each vehicle type k is then calculated as:
σatmax k = ( 2
σamax k − σan k
2
) ...(4.26)
where:
σamaxk maximum total acceleration noise for vehicle type k (m/s2)
σank natural acceleration noise for vehicle type k (m/s2)
The maximum total acceleration noise is given by the expression:
σamax k = MAX [0.1, MIN (0.75, σamaxv k , σamaxgr )] ...(4.27)
where:
σamaxvk maximum acceleration noise for a vehicle type k, input by the user in
Vehicle Fleet (default = 0.75 m/s2 for all vehicle types, with a range 0.1 -
0.75)
σamaxgr the gradient adjusted maximum acceleration noise for a road section (m/s2)
To account for the changes in driver behaviour due to gradients effect, the maximum
acceleration noise for a road type (σamaxr) is adjusted using the following expression:

Version 1.0
σamaxgr = σamaxr {1 - 0.05 * ABS[MIN(GRAN, 10 )]} ...(4.28)
where:
GRAN if GR ≤ 0.02 GRAN=0

if GR > 0.02 GRAN=100*GR
σamaxr maximum acceleration noise for a road type (m/s2) (range from 0.1 to 0.75,
see Part B)
GR average gradient of road section (as a fraction)
Section 5.4 describes the application of acceleration noise within the RUE models.

Version 1.0
Module B: Vehicle Operating Resources

Savings in vehicle operating cost are the main benefits that justify road improvements. Road
users perceive these benefits in terms of lower expenditures.
Vehicle operating costs depend on the following:
! Types of vehicles using the road
! Traffic volume on the road section
! Road geometry (particularly, the curvature, gradient and road width)
! Road surface condition (primarily roughness and texture depth)
! Driver behaviour
Changes in any of these parameters as a result of a project will result in a change in vehicle
operating costs.
For a given road section and for each analysis year, vehicle resource consumption is modelled
by considering the operation of each vehicle type under the conditions of each traffic flow
period, and the results are aggregated into annual totals. The prediction of vehicle resource
consumption is done for each vehicle type in the following order:
1 Fuel consumption (see Section 5)
2 Oil consumption (see Section 6)
3 Tyre consumption (see Section 7)
4 Vehicle utilisation (see Section 8)
5 Parts consumption (see Section 9)
6 Labour hours (see Section 10)
7 Capital costs (see Section 11)
8 Crew hours (see Section 12)
9 Overhead (see Section 13)

Version 1.0
5 Fuel consumption
5.1 Modelling approach

Fuel consumption contributes significantly to the total vehicle operating cost. The HDM-4
fuel consumption model is based on the ARFCOM mechanistic fuel model (Biggs, 1988).
The mechanistic model predicts that fuel consumption is proportional to the total power
requirements of the engine. These are made up of three components:
! Tractive power – this is the power required to overcome forces opposing motion
! Engine drag – this is the power required to overcome internal engine drag (or friction)
! Accessory power – this is the power required to run the vehicle accessories such as the
cooling fan, power steering, air conditioner, alternator, etc.)
For each vehicle type, fuel consumption is calculated for each traffic flow period separately
for the uphill segment and for the downhill segment, the results are then averaged for a round-
trip over the road section.
(a) Total power requirements of the engine for the uphill segment and the downhill
segment (see Section 5.2)
(b) Fuel-to-power efficiency factor for the uphill segment and the downhill segment
(see Section 5.3)
(c) Instantaneous fuel consumption for the uphill segment and the downhill segment.
This combines the steady-state fuel consumption with the additional fuel
consumption due to speed change cycles (see Section 5.4).
(d) Specific fuel consumption for a round trip over the road section (see Section 5.5)
2 For each vehicle type k calculate the annual average fuel consumption
(see Section 5.6)
5.1.1 Uphill segment instantaneous fuel consumption

For the uphill segment, the instantaneous fuel consumption for each vehicle type k during
traffic flow period p is given by:
IFC kpu = MAX[IDLE_FUEL k , ZETA kpu * PTOTkpu * (1 + dFUEL kpu )] ...(5.1)
where:
IFCkpu instantaneous fuel consumption of vehicle type k during traffic flow period
p (ml/s)
IDLE_FUELk idle rate of fuel consumption of vehicle type k (ml/s)
ZETAkpu uphill fuel-to-power efficiency factor of vehicle type k (ml/kW/s)

Version 1.0
PTOTkpu uphill total power requirement for steady-state motion (kW)

dFUELkpu additional fuel consumption factor due to vehicle speed-change cycles (that
is, accelerations and decelerations) discussed in Section 5.4
The expression steady-state implies that speed-change cycles are not considered. Thus, the
steady-state instantaneous fuel consumption for the uphill segment is calculated using
Equation 5.1 above with the value of dFUEL set to zero.
5.1.2 Downhill segment instantaneous fuel consumption

For the downhill segment, the instantaneous fuel consumption for each vehicle type k during
traffic flow period p is given by:
IFC kpd = MAX[IDLE_FUEL k , ZETA kpd * PTOTkpd * (1 + dFUEL kpd )] ...(5.2)
where:
IFCkpd instantaneous fuel consumption of vehicle type k during traffic flow period
p (ml/s)
IDLE_FUELk idle rate of fuel consumption of vehicle type k (ml/s)
ZETAkpd downhill fuel-to-power efficiency factor of vehicle type k (ml/kW/s)
PTOTkpd downhill total power requirement for steady-state motion (kW)
dFUELkpd additional fuel consumption factor due to vehicle speed-change cycles
The steady-state instantaneous fuel consumption for the downhill segment is calculated using
Equation 5.2 above with the value of dFUEL set to zero.
5.1.3 One-way traffic flow

For the analysis of one-way traffic flow, the instantaneous fuel consumption is calculated as
follows:
! For uphill segments
IFCkpu is calculated using Equation 5.1 above
! For downhill segments
IFCkpd is calculated using Equation 5.2 above
5.2 Power requirements

The total power requirements of the engine comprise the tractive power needed to overcome
forces opposing motion and the power to overcome engine drag and run vehicle accessories.
These are calculated separately for the uphill segment and the downhill segment.
5.2.1 Tractive power

At any instant of vehicle movement along a road section, the tractive power may be positive,
negative or zero depending upon the road and vehicle characteristics. The tractive power
required for each vehicle type k during traffic flow period p (PTR) is given by the expression:

Version 1.0
 (FTR kp * Vkp )
PTR kp =   ...(5.3)
 1000 
where:
PTRkp tractive power for vehicle type k during traffic flow period p (kW)
FTRkp total resistance to steady-state motion experienced by vehicle type k during
traffic flow period p (N)
Vkp Speed of vehicle type k during traffic flow period p (m/s)
! For calculating the tractive power (PTRkpu) for the uphill segment
Use Vkp = VUkp, and FTRkpu is given by Equation 5.4 below.

! For calculating the tractive power (PTRkpd) for the downhill segment
Use Vkp = VDkp, and FTRkpd is given by Equation 5.5 below.

The total resistance to steady-state motion comprises aerodynamic resistance, gradient
resistance, rolling resistance, and curvature resistance. Inertial resistance is considered only
under the regime of speed change cycles. The total resistance to steady state motion is
! For the uphill segment
FTR kpu = FA u + FG u + FR u + FCVu ...(5.4)
! For the downhill segment
FTR kpd = FA d + FG d + FR d + FCVd ...(5.5)
where:
FA aerodynamic resistance to motion (N)

FG gradient resistance to motion (N)
FR rolling resistance to motion (N)
FCV curvature resistance to motion (N)
The subscripts u and d denote the uphill and downhill segments, respectively.
The components of total resistance to steady state motion are calculated as given below:
! Aerodynamic resistance
The aerodynamic resistance to motion is calculated as:

2
FA = 0.5 * RHO * CDmult * CD * AF * Vkp ...(5.6)

Version 1.0
The default parameter values to calculate aerodynamic resistance for each vehicle type
are given in Table E2.4.
! Gradient resistance
The gradient resistance to motion is calculated separately for the uphill and downhill
segments using the following expression:
FG = WGT_OPER * g * GR ...(5.7)
where:
All the parameters are as previously defined.
For the uphill segment FGu use the positive value of GR, and for the downhill segment
FGd use the negative value of GR.
! Rolling resistance
The rolling resistance to motion is calculated as:
(
FR = FCLIM * CR2 * b11 * NUM_WHEELS + CR1 * b12 * WGT_OPER + CR1 * b13 * Vkp
2
)
...(5.8)

The default parameter values to calculate rolling resistance for each vehicle type are
given in Table E2.4.
! Curvature resistance
The curvature resistance to motion is calculated as follows:
 2 
   WGT_OPER * V 2   
 
  MAX 0,  − WGT_OPER * g * e   
kp
 
  
R   

FCV =   ...(5.9)
 (NUM_WHEELS * CS )  
  1000  
 
 
where:
e superelevation of the road (as a fraction)

CS cornering stiffness of the tyres
The cornering stiffness is calculated as:
 WGT_OPER  
2
+ CS_a2 * 
CS_a1 * WGT_OPER
CS = Kcs * CS_a0 +  
 NUM_WHEELS  NUM_WHEELS  
...(5.10)
where:

Version 1.0
Kcs tyre stiffness factor

CS_a0 to model parameters
CS_a2
All the other parameters are as previously defined.

Table E2.7 gives the parameter values for the cornering stiffness model.
Table E2.7 Cornering stiffness model parameters
Coefficient WGT_OPER < =2500 kg WGT_OPER > 2500 kg

Bias Radial Bias Radial
CS_a0 30 43 8.8 0
CS_a1 0 0 0.088 0.0913
CS_a2 0 0 -0.0000225 -0.0000114
Kcs 1 1 1 1
Source: NDLI (1995)
! Inertial resistance
The inertial resistance is not included in the total tractive power for steady-state motion.
It is considered in the modelling of speed-change cycle effects described in Section 5.4.
The inertial resistance is calculated as follows:
FI = WGT_OPE R * EMRAT * ACC ...(5.11)
The parameter EMRAT is calculated as:
 EMRAT_a2 
EMRAT = EMRAT_a0 + EMRAT_a1 * atan  
 ...(5.12)
 3
Vkp
 
where:
EMRAT_a0 to inertial resistance parameters

EMRAT_a2
ACC 2
vehicle acceleration in m/s
Note that the inertial resistance to motion is considered to be zero under steady-state
conditions.
The default parameter values to calculate inertial resistance for each vehicle type are given in
Table E2.4.

Version 1.0
5.2.2 Engine and accessories power

The total power required for overcoming engine drag and running vehicle accessories
(PENGACCS) by each vehicle type is calculated as a function of the engine speed and vehicle
speed:
 (PACCS_a0 - PACCS_a1) * (RPMkp - RPM_IDLE )

PENGACCS kp = Kpea * PRATk * PACCS_a1 + 
 (RPM100 - RPM_IDLE ) 
...(5.13)
where:
PENGACCSkp total engine and accessories power of vehicle type k during traffic flow period
p (kW)
Kpea calibration factor for total engine and accessories power (default 1.0)
PRATk the maximum rated engine power of vehicle type k (kW)
RPMkp engine speed (rev/min)
RPM_IDLE idle engine speed (rev/min)
RPM100 the engine speed calculated at 100 km/h (rev/min)
PACCS_a0 the ratio of engine and accessory drag to rated engine power when travelling
at 100 km/h
PACCS_a1 a model parameter
The parameter PACCS_a1 is related to the idle fuel consumption rate. It is calculated from the
user-supplied value for IDLE_FUEL as follows:
(100 - PCTPENG )
a = ZETAB * EHP * Kpea 2 * PRAT *
100
b = ZETAB * Kpea * PRAT
c = - IDLE_FUEL
PACCS_a1 =
(- b + b 2 - 4ac ) ...(5.14)
2a
where:
ZETAB base fuel-to-power efficiency factor (ml/kW/s) (see Table E2.8)

EHP decrease in engine efficiency when producing higher power
PCTPENG percentage of the total engine and accessories power produced from the engine
(default = 80)

Version 1.0
The engine speed (RPM) depends upon the vehicle speed, and it is calculated using Equations
5.15 below to 5.17 below with Vkp = VUkp for the uphill segment and Vkp = VDkp for the
downhill segment:
! If Vkp ≤ 5.6 m/s
RPM kp = RPM_a0 + 5.6 * RPM_a1 + 31.36 * RPM_a2 ...(5.15)
! If Vkp ≤ RPM_a3
2
RPM kp = RPM_a0 + RPM_a1 * Vkp + RPM_a2 * Vkp ...(5.16)
! If Vkp > RPM_a3
(RPM_a0 + RPM_a1 * RPM_a3 + RPM_a2 * RPM_a3 2 ) * Vkp

RPM kp = ...(5.17)
RPM_a3
The engine speed RPM100 at 100 km/h is calculated as follows:

If 27.8 ≤ RPM_a3
RPM100 = RPM_a0 + RPM_a1 * 27.8 + RPM_a2 * 27.8 2 ...(5.18)
else
RPM100 =
(RPM_a0 + RPM_a1 * RPM_a3 + RPM_a2 * RPM_a3 2 ) * 27.8 ...(5.19)
RPM_a3
The values for the above model are given in Table E2.8.
5.2.3 Total power requirement

The total power requirement of the engine (PTOT) is calculated depending on whether or not
the total tractive power is negative as follows:
If PTR kp ≥ 0
 PTR kp 
PTOTkp =  + PENGACCS kp  ...(5.20)
 EDT 
else
PTOTkp = (PTR kp * EDT + PENGACCS kp ) ...(5.21)
where:
PTOTkp total power requirement for steady-state motion by vehicle type k during
traffic flow period p (kW)
PTRkp total tractive power of vehicle type k during traffic flow period p (kW)
EDT drivetrain efficiency

Version 1.0
PENGACCSkp total engine and accessories drag power (kW)

Version 1.0
Table E2.8 Default fuel model parameters
Vehicle Engine speed model parameters Idle Idle Base fuel Decrease Rated Efficiency Engine and
number engine fuel efficiency in engine of the Accessories Power
speed rate efficiency power drivetrain
RPM_a0 RPM_a1 RPM_a2 RPM_a3 RPM_ IDLE_ ZETAB EHP PRAT EDT PACCS_ PCTPENG
IDLE FUEL a0
RPM RPM/ RPM/ m/s RPM ml/s ml/kW/s kW

(m/s) (m/s)2
1 2790 94.0 2.83 31 800 0.12 0.067 0.25 15 0.95 0.20 80
2 2280 17 0.83 42 800 0.25 0.067 0.25 60 0.90 0.20 80
3 2280 17 0.83 42 800 0.36 0.067 0.25 70 0.90 0.20 80
4 1709 7.16 0.99 42 800 0.48 0.067 0.25 90 0.90 0.20 80
5 2490 -30.4 2.25 34 800 0.48 0.067 0.25 60 0.90 0.20 80
6 2574 -27.8 2.46 32 800 0.37 0.067 0.25 55 0.90 0.20 80
7 2490 -30.4 2.25 34 800 0.48 0.057 0.10 60 0.90 0.20 80
8 1214 17.6 2.32 22 500 0.37 0.057 0.10 75 0.86 0.20 80
9 1214 17.6 2.32 22 500 0.37 0.057 0.10 100 0.86 0.20 80
10 1167 -24.0 1.76 22 500 1.12 0.056 0.10 280 0.86 0.20 80
11 1167 -24.0 1.76 22 500 1.12 0.055 0.10 300 0.86 0.20 80
12 2490 -30.4 2.25 34 800 0.48 0.067 0.25 60 0.90 0.20 80
13 1214 17.6 2.32 22 500 0.37 0.057 0.10 75 0.86 0.20 80
14 1214 17.6 2.32 22 500 0.37 0.057 0.10 100 0.86 0.20 80
15 1167 -24.0 1.76 22 500 1.12 0.057 0.10 130 0.86 0.20 80
16 1167 -24.0 1.76 22 500 1.12 0.057 0.10 150 0.86 0.20 80

Version 1.0
5.3 Efficiency factor

The fuel-to-power efficiency factor ZETA relates instantaneous fuel consumption to the total
power requirement of the engine as expressed by Equations 5.1 above and 5.2 above. For each
vehicle type and for each traffic flow period ZETA is calculated separately for the uphill
segment and the downhill segment as given below.
5.3.1 Uphill segment efficiency factor

  PTOTkpu - PCTPENG * PENGACCS kpu 
  100  
ZETA kpu = ZETAB * 1 + EHP *  
 PRATk
 
  
...(5.22)
5.3.2 Downhill segment efficiency factor

  PTOTkpd - PCTPENG * PENGACCS kpd 
  100  
ZETA kpd = ZETAB * 1 + EHP *  
 PRATk
 
  
...(5.23)
All the other parameters are as previously defined.

The values for the above model are given in Table E2.8.
5.4 Additional fuel consumption
5.4.1 Fuel-acceleration simulation model

The additional fuel consumption due to vehicle speed-change cycle effects is estimated using
a simulation model called ACCFUEL. This model is described in detail by Bennett (1996c).
The model computes the additional fuel consumption factor dFUEL as a function of the total
acceleration noise and the vehicle mean speed. These factors are then used in the calculation
of fuel consumption, tyre consumption and spare parts consumption as described in this
document. High magnitudes of total acceleration noise lead to high consumption of fuel, tyres
and spare parts.
The method is as follows:
1 The user runs a calibration routine external to HDM-4 (that is, ACCFUEL) which
generates a matrix of values for dFUEL as a function of mean speed and acceleration
noise, for each vehicle type.
2 These matrices are read in HDM-4. A default set of matrices for the standard vehicle
types is contained within HDM-4. Values of dFUEL are linearly interpolated for
intermediate speeds and acceleration noise, and applied in the analyses.

Version 1.0
5.4.2 Mean speeds and acceleration noise

The mean speeds (km/h) that are used for determining the corresponding values of dFUEL for the
uphill and downhill segments are given as follows:
The mean speed for each vehicle type k and traffic flow period p is given by:
SU kp = 3.6 * VU kp ...(5.24)
The mean speed for each vehicle type k and traffic flow period p is given by:
SD kp = 3.6 * VD kp ...(5.25)
The total acceleration noise (σakp) for each vehicle type k and for each traffic flow period p
used for determining dFUEL is calculated as described in Section 4.6.
5.5 Fuel consumption per 1000 vehicle-km

The specific fuel consumption (ml) per vehicle-kilometre on the road section is calculated
from the expression:
 IFC kpu IFC kpd 

SFC kp = 500  +  ...(5.26)
 VU kp VD kp 
where:
SFCkp specific fuel consumption (ml/km)

IFCkpu instantaneous fuel consumption for uphill travel (ml/s)
VUkp uphill speed (m/s) of vehicle type k in traffic flow period p
IFCkpd instantaneous fuel consumption for downhill travel (ml/s)
VDkp downhill speed (m/s) of vehicle type k in flow period p
For one-way traffic, the specific fuel consumption is calculated as follows:

1000 * IFC kpu

SFC kp = ...(5.27)
VU kp
1000 * IFC kpd

SFC kp = ...(5.28)
VD kp
The fuel consumption (litres per 1000 vehicle-kilometres) is thus given by the expression:

Version 1.0
FC kp = SFC kp * FUELBIAS ...(5.29)
where:
FCkp fuel consumption of vehicle k in traffic flow period p (l/1000 veh-km)

FUELBIAS fuel adjustment factor to account for the bias introduced through the use of the
time mean speed instead of the space mean speed
The traffic stream is comprised of vehicles travelling at different speeds and, thus, different
fuel consumption rates. Since fuel consumption is non-linear with speed, the mean fuel
consumption does not correspond to the fuel consumption at the mean speed. The fuel
adjustment factor (FUELBIAS) is therefore applied to correct the bias introduced to the
analysis through the use of the mean speed instead of individual vehicle speeds. It is given by
the following expression:
FUELBIAS = MAX (dFUEL, 1.0000 - 0.0182 * COV + 0.7319 * COV 2 ) ...(5.30)
5.6 Annual average fuel consumption

The annual average fuel consumption in litres per 1000 vehicle-kilometres of each vehicle
type is required for reporting purposes, and it is calculated as follows:
n
∑ HRYR
p =1
p * HVp * FC kp
FC kav = n
...(5.31)
∑ HRYR
p =1
p * HVp
where:
FCkav annual average fuel consumption of vehicle type k (km/h)

FCkp fuel consumption of vehicle type k during traffic flow period p

Version 1.0
6 Lubricating oil consumption

The model used for predicting lubricating oil consumption is based on that developed by
Pienaar (1984), reported in English by du Plessis, editor (1989). This models lubricating oil
consumption in two components: oil loss due to contamination and that due to operation. Oil
loss due to contamination is a function of distance between oil changes. Oil loss due to
operation is calculated as a function of fuel consumption.
Thus, the oil consumption for each vehicle type k, for each traffic flow period p is calculated
from the expression:
OIL kp = OILCONT + OILOPER * FC kp ...(6.1)
where:
OILkp oil consumption (l/1000 km)

OILCONT oil loss due to contamination (l/1000 km)
OILOPER oil loss due to operation (l/1000 km)
FCkp fuel consumption (l/1000 km) in traffic flow period p
The loss due to contamination is determined as follows:
OILCAP
OILCONT = ...(6.2)
DISTCHNG
where:
OILCAP engine oil capacity (litres)

DISTCHNG distance between oil changes (1000s kilometres)
The values in Table E2.9 are the defaults for oil consumption model values.

Version 1.0
Table E2.9 Default oil consumption model values
Vehicle type Distance Engine oil Oil loss due

between oil capacity to operation
changes
(km) (l) OILOPER

Motorcycle 5000 2.0 0.0014
Passenger car 10000 4.0 0.0028
Light goods and delivery vehicle, mini-bus, 4WD 7500 5.0 0.0028
Light and medium truck 9000 14.0 0.0021
Heavy and articulated truck 10000 31.0 0.0021
Light and medium bus 8000 14.0 0.0021
Heavy bus and coach 8000 20.0 0.0021
The annual average oil consumption (litres per 1000 vehicle-kilometres) is given by:
OIL kav = OILCONT + OILOPER * FC kav ...(6.3)
where:
OILkav oil consumption (l/1000 km)

OILCONT oil loss due to contamination (l/1000 km)
OILOPER oil loss due to operation (l/1000 km)
FCkav annual average fuel consumption (l/1000 km)

Version 1.0
7 Tyre consumption

The tyre consumption model is based on slip energy theory used in the HDM-III model. As
described by Watanatada et al. (1987a), the tyre consumption of a vehicle is proportional to
the energy requirements. The energy requirements are calculated as a function of the
circumferential, lateral and normal forces acting on each wheel.
The rate of tyre consumption is expressed in terms of the number of equivalent new tyres
consumed per 1000 vehicle-kilometres for each vehicle wheel. This is calculated separately
for the uphill and the downhill segments. The results are then the averaged to represent the
tyre consumption for a round trip over the road section.
(a) The circumferential, lateral and normal forces acting on a tyre for the idealised
uphill and downhill segments (see Section 7.2.2)
(b) The tyre energy for the uphill segment and the downhill segment (see Section 7.2.1)
(c) Tyre consumption per 1000 vehicle-kilometres for a round trip over the road section
(see Section 7.3)
2 For each vehicle type k calculate the annual average tyre consumption
(see Section 7.5)
7.1.1 Uphill segment tyre consumption

The number of equivalent new tyres consumed per 1000 vehicle-kilometres for each wheel of
vehicle type k during traffic flow period p is calculated as follows:
1 + 0.01 * RREC k * NR k
EQNTkpu = + 0.0027 ...(7.1)
DISTOTk pu
where:
EQNTkpu number of equivalent new tyres consumed per 1000 veh-km for each wheel
during traffic flow period p
RRECk retread cost as a percentage of new tyre cost (default = 15 for all vehicle types)
NRk the number of retreads per tyre carcass
DISTOTkpu total distance travelled in the uphill direction by the tyre (1000s of kilometres)
The number of retreads per tyre carcass is given by the expression:
NR k = MAX[0, NR0 k * exp(-0.03224 * RImod ) − 1] ...(7.2)

Version 1.0
where:
NR0k base number of recaps specified by the user (default = 1.30 for all vehicle
types)
RImod modified value of the average road roughness (m/km), see Section 7.4
The total distance travelled by a tyre carcass during traffic flow period p is given by:
VOL k
DISTOTkpu = (1 + NR k ) ...(7.3)
TWTkpu
where:
TWTkpu the rate of tread wear (dm3/1000 veh-km) during traffic flow period p
VOLk volume of wearable rubber (dm3), see Table E2.10
7.1.2 Downhill segment tyre consumption

The number of equivalent new tyres consumed per 1000 vehicle-kilometres for each wheel of
vehicle type k during traffic flow period p is calculated as follows:
1 + 0.01 * RREC k * NR k
EQNTkpd = + 0.0027 ...(7.4)
DISTOTk pd
where:
EQNTkpd number of equivalent new tyres consumed per 1000 veh-km for each wheel
DISTOTkpd total distance travelled in the downhill direction by the tyre (1000s of
kilometres) during traffic flow period p
The total distance travelled by a tyre carcass during traffic flow period p is given by:
VOL k
DISTOTkpd = (1 + NR k ) ...(7.5)
TWTkpd
7.2 Rate of tread wear

The rate of tread wear during each traffic flow period is calculated as a function of tangential
energy. This is done separately for the uphill and downhill segments, as follows:
TWTkp = C0tc + Ctcte * TE kp ...(7.6)
where:

Version 1.0
TEkp the tangential energy of each tyre (J-m)

C0tc constant term of the tyre tread wear model (dm3).
Ctcte the wear coefficient of the tyre tread wear model, (dm3/J-m)
The default values for C0tc and Ctcte are given in Table E2.10.
Table E2.10 Default tyre consumption model values
Vehicle Vehicle type NR0 C0tc Ctcte VOL

Number
3
(dm )
1 Motorcycle 1.30 0.00639 0.00050 0.35
2 Small car 1.30 0.02616 0.00204 1.40
3 Medium car 1.30 0.02616 0.00204 1.40
4 Large car 1.30 0.02616 0.00204 1.40
5 Light delivery vehicle 1.30 0.02400 0.00187 1.60
6 Light goods vehicle 1.30 0.02400 0.00187 1.60
7 Four wheel drive 1.30 0.02400 0.00187 1.60
8 Light truck 1.30 0.02400 0.00187 1.60
9 Medium truck 1.30 0.02585 0.00201 6.00
10 Heavy truck 1.30 0.03529 0.00275 8.00
11 Articulated truck 1.30 0.03988 0.00311 8.00
12 Mini-bus 1.30 0.02400 0.00187 1.60
13 Light bus 1.30 0.02173 0.00169 1.60
14 Medium bus 1.30 0.02663 0.00207 6.00
15 Heavy bus 1.30 0.03088 0.00241 8.00
16 Coach 1.30 0.03088 0.00241 8.00
Source: Bennett (1996)
7.2.1 Tangential energy

The tangential energy of each tyre is calculated as a function of the forces acting upon the
tyre. This is executed separately for the uphill segment and the downhill segment, using the
general expression:
TE kp =
(CFT kp
2
+ LFTkp
2
) ...(7.7)
NFT
where:

Version 1.0
CFTkp the circumferential force acting on a tyre (N)

LFTkp the lateral force acting on a tyre (N)
NFT the normal force per tyre (N)
7.2.2 Forces acting on a tyre

There are three components of forces acting upon a tyre: circumferential, lateral and normal.
These are calculated for each of the idealised road segments (that is, uphill and downhill).
! Circumferential force
The circumferential force CFTkp is quantified as follows:
(1 + CTCONk * dFUEL kp ) (FA kp + FG kp + FR kp )

CFTkp = ...(7.8)
NUM_WHEELS
where:
CTCONk the incremental change of tyre consumption related to additional fuel,

dFUEL, (default = 0.1)
dFUELkp additional fuel consumption factor due to speed change cycle effects (see
Section 5.4)
FAkp aerodynamic resistance to motion (N) (see Section 5.2.1)
FGkp gradient resistance to motion (N) (see Section 5.2.1)
FRkp rolling resistance to motion (N) (see Section 5.2.1)

Since the acceleration effects will result in additional longitudinal wear on the tyres, the
circumferential component is increased based on the additional fuel factor (dFUEL).
! Lateral force
The lateral force LFTkp is quantified as follows:
FCVkp
LFTkp = ...(7.9)
NUM_WHEELS
where:
FCVkp curvature resistance to motion (N) (see section 5.2.1)

NUM_WHEELSk number of wheels per vehicle type k
! Normal force
The normal force per wheel NFT is given by the expression:

Version 1.0
WGT_OPER * g
NFT = ...(7.10)
NUM_WHEELS
where:
WGT_OPER vehicle operating weight (kg)

g acceleration due to gravity taken as equal to 9.81 m/s2
7.3 Tyre consumption per 1000 vehicle-km
7.3.1 Two-way traffic sections

The total tyre consumption expressed in terms of equivalent number of new tyres per 1000
vehicle-kilometres for each vehicle type k, for each traffic flow period p, is calculated from
the expression:
[0.5 * (EQNTkpu + EQNTkpd ) * NUM_WHEELS ]

TC kp = ...(7.11)
MODFAC kp
MODFAC kp = VEHFAC k * TYPEFAC * CONGFAC kp ...(7.12)
where:
TCkp number of tyres consumed per 1000 veh-km by vehicle type k during traffic
flow period p
MODFACkp tyre life modification factor for vehicle type k during traffic flow period p
VEHFACk vehicle type modification factor (see Table E2.11)
TYPEFAC tyre type modification factor (see Table E2.12)
CONGFACkp Congestion effects modification factor for vehicle type k during traffic flow
period p
The default values of these modification factors are supplied in Section 7.4.
The life of a single tyre of vehicle type k operating in traffic flow period p, TLIFE, (1000s
kilometres) is given as:
1000
TLIFE kp = ...(7.13)
0.5 * (EQNTkpu + EQNTkpd )
7.3.2 One-way traffic sections

For uphill segments, the total tyre consumption expressed in terms of equivalent number of
new tyres per 1000 vehicle-kilometres for each vehicle type k, for each traffic flow period p, is
calculated from the expression:

Version 1.0
[EQNTkpu * NUM_WHEELS ]
TC kp = ...(7.14)
MODFAC kp
For downhill sections, the total tyre consumption expressed in terms of equivalent number of
new tyres per 1000 vehicle-kilometres for each vehicle type k, for each traffic flow period p, is
[EQNTkpd * NUM_WHEELS ]
TC kp = ...(7.15)
MODFAC kp
Tyre life (TLIFEkp) is calculated using Equation 7.13 above with the denominator replaced
with EQNTkpu for the uphill segment and EQNTkpd for the downhill segment.
7.4 Modification to the tyre consumption model

Harrison and Aziz (1998) recommended the following adjustments to the tyre consumption
model:
! Road roughness effects
The values of adjusted average roughness (RImod) used in Equation 7.2 above, and of
VEHFAC used in Equation 7.12 above are given in Table E2.11.
Table E2.11 Roughness effects and vehicle type modification factor
Vehicle number Adjusted roughness VEHFAC
(RImod)
1, 2, 3, 4, 5, 6, 7, 8, 12, 13 = RIav 2.0
10, 11, 15, 16 = min[7, RIav] 1.0
9, 14 =7 1.0
Source: Harrison and Aziz (1998)
Note: A vehicle number is related to a vehicle type as given in Table E2.10
! Tyre type effects
The values of tyre type modification factor, TYREFAC, used in Equation 7.12 above is
given in Table E2.12.
Table E2.12 Tyre type modification factor (TYREFAC)
Tyre type Paved roads Unpaved roads
IRI ≤ 6 m/km IRI > 6 m/km

Bias 1.00 1.00 1.00
Radial 1.25 1.20 1.00
Source: Harrison and Aziz (1998)

Version 1.0
! Traffic congestion (or acceleration) effects
The values of congestion effects factor (CONGFACkp) used in Equation 7.2 above are
obtained as follows:
❏ for Qp < Qo
CONGFAC kp = 1 ...(7.16)
❏ for Qo ≤ Qp ≤ Qnom
CONGFAC kp = 1 − 
0.3 * (Qp - Qo) 
...(7.17)
 (Qnom - Qo) 
❏ for Qnom < Qp ≤ Qult
CONGFAC kp = 0.7 − 
0.2 * (Qp - Qnom) 
...(7.18)
 (Qult - Qnom) 
7.5 Annual average tyre consumption

The annual average number of tyres consumed per 1000 vehicle kilometres for each vehicle
type k is calculated from the formula:
n
∑ HRYR
p =1
p * HVp * TC kp
TC kav = n
...(7.19)
∑ HRYR
p =1
p * HVp
where:
TCkav annual average number of tyres consumed per 1000 veh-km by vehicle type k
TCkp number of tyres consumed per 1000 veh-km by vehicle type k during traffic
flow period p

Version 1.0
8 Vehicle utilisation and service life

Vehicle utilisation and service life are required for calculating parts consumption, capital
costs, and overhead costs.
8.1 Utilisation
Vehicle utilisation is expressed in terms of the annual kilometreage driven during the annual
working time. Working time as defined by Hine (1996) is the time spent undertaking the
essential tasks of making a complete round trip, in normal circumstances. This excludes time
spent idle, when the driver is eating, sleeping or otherwise resting, but includes time spent
driving, loading, unloading and refuelling. In certain circumstances it may be appropriate to
include the time that the vehicle must spend waiting to move forward in a queue.
8.1.1 Annual number of kilometres driven

Road investment may affect vehicle annual kilometreage through changes in journey times
and trip distances. The utilisation of commercial vehicles can be quite sensitive to these
changes. Although passenger car utilisation is not absolutely fixed in most situations the
assumption of constant utilisation is probably appropriate.
The baseline average annual utilisation (AKM0) is either entered as a constant value by the
user or calculated if the user specifies an age distribution along with the percentage of
vehicles at each age:
n
AKM0 = ∑ AKMVi * PCTVi 100
i =1
...(8.1)
where:
AKM0 baseline average number of kilometres driven per year , input by the user
(km/year)
AKMVi average number of kilometres driven per vehicle of age i per year, input by the
user (km/year)
PCTVi percentage of vehicles of age i in the fleet (for i = 1, 2, ..., n)
8.1.2 Annual number of working hours

The baseline average annual working time (HRWK0) is either input as a constant value by the
user or calculated if the user specifies an age distribution along with the percentage of
vehicles at each age:
n
HRWK0 = ∑ HRWKVi * PCTVi 100
i=1
...(8.2)
where:

Version 1.0
HRWK0 baseline average number of vehicle working hours per year, input by the user
(hrs/year)
HRWKVi number of working hours per vehicle of age i per year
PCTVi percentage of vehicles of age i in the fleet (for i = 1, 2, ..., n)
8.2 Service life

The following two methods of calculating vehicle service life are provided:
! Constant vehicle life method (see Section 8.2.1)
! Optimal vehicle life method (see Section 8.2.2)
The user chooses which of the two methods should be used for calculating vehicle parts
consumption and for modelling capital costs.
8.2.1 Constant vehicle life method

This method uses straight-line depreciation, in which the vehicle service life, LIFE, is
assumed to be constant irrespective of vehicle speed and equal to the user-specified value. For
details, refer to Watanatada et al. (1987a).
8.2.2 Optimal vehicle life method

This is the expected service life defined as the distance at which it becomes appropriate to
scrap the vehicle (see Bennett, 1996b). The optimal life of a vehicle under the condition of
different road roughness values is determined as follows:
LIFEKM0 * LIFEKMPCT
LIFEKM = ...(8.3)
100
where:
LIFEKM the predicted optimal lifetime vehicle utilisation in kilometres

LIFEKM0 baseline average vehicle service life in kilometres
LIFEKMPCT optimal lifetime kilometreage as a percentage of baseline vehicle service
life
The baseline average vehicle service life is calculated from the expression:
LIFEKM0 = AKM0 * LIFE0 ...(8.4)
where:
AKM0 baseline average number of kilometres driven per vehicle per year, input by
the user (km/year)
LIFE0 baseline average vehicle service life in years, input by the user

Version 1.0
The optimal life as a percentage of the user defined baseline vehicle service life is given by:
 100 
LIFEKMPCT =  
( )
...(8.5)
1 + exp a0 * RI adj a1 
where:
RIadj the adjusted road roughness (IRI m/km), see Section 9.2
a0, a1 regression coefficients. The default values (for all vehicle types) are as
follows:
a0 = -65.8553
a1 = -1.9194
The default values for vehicle utilisation parameters are shown in Table E2.13.
Table E2.13 Default vehicle utilisation model values
Vehicle Vehicle type AKM0 LIFE0 HRWK0

Number
(km/year) (years) (h/year)
1 Motorcycle 10000 10 400
2 Small car 23000 10 550
3 Medium car 23000 10 550
4 Large car 23000 10 550
5 Light delivery vehicle 30000 8 1300
6 Light goods vehicle 30000 8 1300
7 Four wheel drive 30000 8 1300
8 Light truck 30000 8 1300
9 Medium truck 40000 12 1200
10 Heavy truck 86000 14 2050
11 Articulated truck 86000 14 2050
12 Mini-bus 30000 8 750
13 Light bus 34000 8 850
14 Medium bus 70000 7 1750
15 Heavy bus 70000 12 1750
16 Coach 70000 12 1750

Version 1.0
9 Parts consumption

Spare parts costs constitute a significant component of vehicle operating costs. The
requirements for spare parts depend on vehicle age in kilometres and road surface roughness.
Speed-change cycles (or acceleration and deceleration) due to traffic congestion, road
alignment, the presence of NMT, roadside friction and driver behaviour also affect the rate of
wear and tear of vehicle and components. The parts consumption model therefore considers
the effect of vehicle age, roughness and speed-change cycles.
Parts consumption cost is expressed as a fraction of the replacement vehicle price. For each
vehicle type, parts consumption is predicted for the particular operating conditions of each
traffic flow period.
1 For each vehicle type k calculate:
(a) The adjusted road roughness to be used in the model (see Section 9.2)
(b) The age in terms of cumulative number of kilometres driven (see Section 9.3)
(a) The incremental change in parts consumption due to speed change cycles (see
Section 9.4)
(b) The parts consumption per 1000 vehicle-kilometres as a fraction of replacement
vehicle price
3 For each vehicle type k calculate the annual average parts consumption
(see Section 9.5)
The parts consumption for each vehicle type k, and for each traffic flow period p, is calculated
as follows:
PC kp = K0pc * [CKMKP * (a0 + a1 * RIadj ) + K1pc ] [1 + CPCONk * dFUEL kp ]
...(9.1)
where:
PCkp parts consumption per 1000 veh-km, expressed as a fraction of the average
new (or replacement) vehicle price, NVPk,
CKM average cumulative number of kilometres driven per vehicle type (km)
KP the age exponent in parts consumption model
RIadj the adjusted road roughness (IRI m/km)
CPCONk incremental change factor in parts consumption due to vehicle speed change
cycles effects (default value = 0.10)
dFUELkp additional fuel consumption factor due to vehicle speed-change cycles (that is,
accelerations and decelerations)

Version 1.0
a0 the constant term model parameter

a1 roughness dependent model parameter
K0pc parts consumption rotational calibration factor (default value = 1.0)
K1pc parts consumption translational calibration factor (default value = 0)
The default values for the model parameters are shown in Table E2.14, and are based on the
default CKM values shown.
Table E2.14 Proposed default parts consumption model values
Vehicle Vehicle type CKM KP a0 a1

Number
x10-6 x10-6
1 Motorcycle 50,000 0.308 9.23 6.20
2 Small car 115,000 0.308 36.94 6.20
3 Medium car 115,000 0.308 36.94 6.20
4 Large car 115,000 0.308 36.94 6.20
5 Light delivery vehicle 120,000 0.308 36.94 6.20
6 Light goods vehicle 120,000 0.308 36.94 6.20
7 Four wheel drive 120,000 0.371 7.29 2.96
8 Light truck 120,000 0.371 7.29 2.96
9 Medium truck 240,000 0.371 11.58 2.96
10 Heavy truck 602,000 0.371 11.58 2.96
11 Articulated truck 602,000 0.371 13.58 2.96
12 Mini-bus 120,000 0.308 36.76 6.20
13 Light bus 136,000 0.371 10.14 1.97
14 Medium bus 245,000 0.483 0.57 0.49
15 Heavy bus 420,000 0.483 0.65 0.46
16 Coach 420,000 0.483 0.64 0.46
9.2 Roughness effects

Roughness does not significantly influence the parts consumption below a certain level. The
parts consumption model needs to be adjusted to reflect this by limiting the roughness using
the following equation:
RIadj = [
MAX RIav , MIN(RI0, RIMIN + a2 * RIa3 ) ] ...(9.2)
where:

Version 1.0
RIav the average roughness of the road (IRI m/km)

RIMIN represents the minimum roughness to be used in the model . It has a default
value of 3.0
RI_SHAPE shape factor; it has a default of 0.25
Thus, the calculations are as follows:
RI0 = RIMIN + RI_SHAPE ...(9.3)
RI_SHAPE
a2 = RI0
...(9.4)
RI_SHAPE
RI0
RI0
a3 = ...(9.5)
RI_SHAPE
where:
a2 and a3 model shape parameters
9.3 Vehicle age effects

Vehicle age is given in terms of the cumulative number of kilometres driven CKM. This is
calculated using one of the following three methods:
1 For the constant vehicle life method
CKM k = 0.5 * AKM0 k * LIFE0 k ...(9.6)
where:
CKMk average cumulative number of kilometres driven per vehicle type k (km)
AKM0k baseline average number of kilometres driven per vehicle type k per year ,
input by the user (km/year)
LIFE0k baseline average service life of vehicle type k in years, input by the user
2 For the optimal vehicle life method
CKM k = 0.5 * LIFEKM k ...(9.7)
where:
LIFEKMk the predicted optimal vehicle service life in kilometres

Version 1.0
3 User specifies the age spectrum distribution and the percentage of vehicles
If the user specifies the age spectrum distribution along with the percentage of vehicles at
each age i, then the value of CKMk is calculated from the following expression:
n
CKMk =
∑ AKMV * VEHAGE
i=1
i
100
* PCTV i i
...(9.8)
where:
CKMk cumulative number of kilometres driven, for vehicle type k (km)

AKMVi average number of kilometres driven per vehicle of age i per year, input
by the user (km/year)
VEHAGEi vehicle age i in years (for i = 1, 2, …., n)
PCTVi percentage of vehicles of age i in the fleet
The CKMk value is used in Equation 9.1 above to calculate parts consumption PCkp per 1000
vehicle-km during each traffic flow period p. Note that the value of CKMk is the same for all
the traffic flow periods.
9.4 Acceleration effects

The effect of speed fluctuations on increasing parts consumption is modelled using the
additional fuel factor dFUEL, which is calculated as a function of acceleration noise and
vehicle speed (see Section 5.4). It is assumed that the fractional change in parts consumption
due to vehicle accelerations and decelerations (dPARTS) is related to the fractional change in
fuel consumption due to the same effect (dFUEL). Thus, the fractional change in parts
consumption is expressed as:
dPARTS kp = CPCON * dFUEL kp ...(9.9)
where:
dPARTSkp fractional change in parts consumption per 1000 veh-km (PC)

The expression given in Equation 9.9 above forms the last part of Equation 9.1 above. Using
the default value of CPCON = 0.1 implies that for every 10% increase in fuel consumption
there will be a 1% increase in the parts consumption.
9.5 Annual average parts consumption

The annual average parts consumption as a fraction of the new vehicle price per 1000 veh-km

Version 1.0
∑ HRYR
p =1
p * HVp * PC kp
PC kav = n
...(9.10)
∑ HRYR
p =1
p * HVp
where:
PCkav annual average parts consumption per 1000 veh-km, expressed as a fraction of
the average new vehicle price
HVp hourly traffic flow in period p expressed as a proportion of AADT
PCkp parts consumption per 1000 veh-km by vehicle type k during traffic flow
period p, expressed as a fraction of the average new vehicle price

Version 1.0
10 Labour hours
10.1 The model

Maintenance labour hours are predicted to determine the labour component of fitting spare
parts and repairing vehicles. Labour wage rates (input by the user) are applied to the predicted
number of labour hours to obtain maintenance labour costs. Maintenance labour hours are
calculated as a function of the parts consumption.
1 For each vehicle type k and for each traffic flow period p calculate maintenance
labour hours per 1000 vehicle-kilometres
2 For each vehicle type k calculate the annual average labour hours
The number of labour hours per 1000 vehicle-kilometres is calculated for each vehicle type k,
and for each traffic-flow period p as follows:
(
LH kp = K0lh * a0 * PC kp
a1
) + K1lh ...(10.1)
where:
LHkp number of labour hours per 1000 veh-km of vehicle type k in traffic flow
period p
PCkp parts consumption per 1000 veh-km of vehicle type k during traffic flow
period p, expressed as a fraction of average new vehicle price
a0 constant term of the maintenance labour model
a1 parts exponent of the maintenance labour model
K0lh rotational calibration factor (default value = 1.0)
K1lh translational calibration factor (default value = 0)
The default values for the labour hours model parameters are shown in Table E2.15, and are
based on the default CKM values given in Table E2.14.

Version 1.0
Table E2.15 Proposed default labour hours model parameter values
Vehicle Vehicle type a0 a1

Number
1 Motorcycle 77.14 0.547
2 Small car 77.14 0.547
3 Medium car 77.14 0.547
4 Large car 77.14 0.547
5 Light delivery vehicle 77.14 0.547
6 Light goods vehicle 77.14 0.547
7 Four wheel drive 77.14 0.547
8 Light truck 242.03 0.519
9 Medium truck 242.03 0.519
10 Heavy truck 301.46 0.519
11 Articulated truck 301.46 0.519
12 Mini-bus 77.14 0.547
13 Light bus 242.03 0.519
14 Medium bus 293.44 0.517
15 Heavy bus 293.44 0.517
16 Coach 293.44 0.517
10.2 Annual average labour hours

The annual average number of labour hours per 1000 veh-km are calculated as follows:
n
∑ HRYR
p =1
p * HVp * LH kp
LH kav = n
...(10.2)
∑ HRYR
p =1
p * HVp
where:
LHkav annual average number of labour hours per 1000 veh-km of vehicle type k
HRYRp number of hours in traffic flow period p
LHkp number of labour hours per 1000 veh-km for vehicle type k during traffic flow
period p

Version 1.0
11 Capital costs

Capital costs comprise of depreciation and interest costs. These constitute a significant
component of the total vehicle operating cost. Capital costs, and their allocation, are sensitive
to both the utilisation of a vehicle and its service life, which in turn depends on vehicle speed
and road condition. The modelling of capital costs is therefore performed for each particular
set of operating conditions pertaining to a traffic flow period.
For each analysis year, the capital cost per 1000 vehicle-kilometres for each vehicle type k,
and for each traffic flow period p, is calculated using the formula:
CAPCSTkp = DEPCSTkp + INTCSTkp ...(11.1)
where:
CAPCSTkp capital cost per 1000 veh-km incurred during traffic flow period p
DEPCSTkp depreciation cost per 1000 veh-km incurred during traffic flow period p
INTCSTkp annual interest cost per 1000 veh-km incurred during traffic flow period p
The depreciation and interest costs are computed separately as described below.
! Depreciation cost is given as:
DEPCSTkp = DEPkp * NVPLTk ...(11.2)
where:
DEPkp depreciation cost factor per 1000 veh-km for traffic flow period p
NVPLTk average new (or replacement) vehicle price less tyres
For each vehicle type k, the average new vehicle price less tyres (to avoid double counting) is
NVPLTk = NVPk - NUM_WHEELS k * NTPk ...(11.3)
where:
NVPk average new (or replacement) vehicle price

NUM_ number of wheels of vehicle type k
WHEELSk
NTPk average new tyre price

Version 1.0
! Interest cost is given as:
INTCSTkp = INTkp * NVPk ...(11.4)
where:
INTkp annual interest cost factor per 1000 veh-km for traffic flow period p
NVPk average new (or replacement) vehicle price less tyres
11.2 Depreciation
Vehicle depreciation arises mainly due to use, time/ageing and technical obsolescence. There
are two methods of calculating depreciation costs based on vehicle service life:
1 Constant life method (see Section 11.2.2)
2 Optimal life method (see Section 11.2.3)
Both methods calculate depreciation cost over the service life of a vehicle using a straight-line
method. The residual value at the end of the vehicle’s service life is deducted from the vehicle
price before calculating depreciation.
11.2.1 Residual vehicle value

The residual value at the end of a vehicle’s life is a function of road roughness, with user-
definable parameters. Vehicles operated on rougher roads will have a lower residual value
since they will have suffered more wear and tear. Residual vehicle value is calculated from the
RVPLTPCT = MAX[a2, a3 - MAX(0, (RIav - a4 ))] ...(11.5)
where:
RVPLTPCT residual vehicle price less tyres at the end of its service life (%)
RIav average road roughness (IRI m/km)
a2 minimum residual value of the vehicle (%) (default = 2)
a3 maximum residual value of the vehicle (%) (default value = 15)
a4 average roughness, IRI, below which the maximum value arises (default
value = 5)
11.2.2 Constant life depreciation method

The depreciation cost factor per 1000 vehicle-kilometres, DEPkp, for each traffic flow period
1000 * (1 - 0.01 * RVPLTPCTk )

DEPkp = ...(11.6)
SS kp * HRWK0 k * LIFE0 k

Version 1.0
where:
RVPLTPCTk residual vehicle price less tyres of vehicle type k (%)

SSkp vehicle operating speed (km/h) during traffic flow period p
HRWK0k baseline average number of working hours per year for vehicle type k, input
by the user (hrs/year)
LIFE0k baseline average service life of vehicle type k in years, input by the user
(years)
For passenger cars the depreciation cost factor per 1000 vehicle-kilometres, DEPkp, for each
traffic flow period is calculated as follows:
! If PP (that is, the percentage of vehicle use on private trips) is greater than 50:
1000 * (1 - 0.01 * RVPLTPCTk )

DEPkp = ...(11.7)
AKM0 k * LIFE0 k
! otherwise use Equation 11.6 above
11.2.3 Optimal life depreciation method

The depreciation cost factor per 1000 vehicle-kilometres for each traffic flow period is
1000 * (1 - 0.01 * RVPLTPCTk )

DEPkp = ...(11.8)
LIFEKM k
where:
RVPLTPCTk residual vehicle price less tyres of vehicle type k (%)

LIFEKMk predicted optimal service life of vehicle type k (km)
Note that the value of DEPk is the same for all the traffic flow periods.
11.3 Interest
Interest costs are the opportunity cost of vehicle ownership. These consist of the income that
would have been received had the capital invested in the vehicle been invested elsewhere.
The interest cost factor per 1000 vehicle-kilometres INTkp for each traffic flow period is
1000 * AINVk
INTkp = ...(11.9)
2 * SS kp * HRWK0 k * 100
where:

Version 1.0
AINVk annual interest charge on the purchase cost of the average new vehicle type k
(%)
11.4 Annual average capital cost

The annual average capital cost per 1000 vehicle-kilometres is calculated as follows:
n
∑ HRYR
p =1
p * HVp * CAPCSTkp
CAPCSTkav = n
...(11.10)
∑ HRYR
p =1
p * HVp
where:
CAPCSTkav average capital cost per 1000 veh-km for vehicle type k
HRYRp number of hours in traffic flow period p
HVp hourly traffic flow in period p expressed as a proportion of AADT
CAPCSTkp capital cost per 1000 veh-km for vehicle type k during flow period p

Version 1.0
12 Crew hours
12.1 The model

Crew cost is included as a vehicle operating cost rather than as a time cost. It is obtained from
the product of the number of crew hours and the crew wage rate. In HDM-4, the cost of crew
labour is considered to be a variable cost rather than a fixed cost. This means that the time the
crew spends on non-driving activities such as loading, unloading and layovers is not charged
against this cost category. Thus, the number of crew hours required per 1000 vehicle-
kilometres (or distance-dependent annual vehicle hours) for each vehicle type k, during each
traffic flow period p is calculated as a function of the vehicle operating speed, as follows:
1000 * (100 - PPk )

CHkp = ...(12.1)
100 * SS kp
where:
CHkp number of hours per crew member per 1000 veh-km for traffic flow period p
PPk percentage of vehicle use on private trips (%)
12.2 Annual average number of crew hours

The annual average number of hours for each crew member per 1000 vehicle-kilometres is
n
∑ HRYR
p =1
p * HVp * CHkp
CHkav = n
...(12.2)
∑ HRYR
p =1
p * HVp
where:
CHkav average number of hours per crew member per 1000 veh-km for vehicle type
k
CHkp number of hours per crew member per 1000 veh-km for vehicle type k

Version 1.0
13 Overhead costs
13.1 The model

This covers all other cost elements including administration, insurance, parking/garaging, and
any overheads associated with the crew (for example, training, uniform, etc.). Overhead costs
are calculated as a function of the annual vehicle utilisation and average operating speed.
For each analysis year, and for each vehicle type k, the overhead cost per 1000 vehicle-
kilometres incurred during traffic flow period p, is calculated using the formula:
1000 * OA k * (100 − PPk )

OC kp = ...(13.1)
100 * SS kp * HRWK0 k
where:
OCkp overhead cost per 1000 veh-km for vehicle type k incurred during traffic
flow period p
OAk overhead cost per year, for vehicle type k, input by the user
PPk percentage of vehicle use on private trips
13.2 Annual average overhead costs

The annual average overhead costs per 1000 vehicle-kilometres is calculated as follows:
n
∑ HRYR
p =1
p * HVp * OC kp
OC kav = n
...(13.2)
∑ HRYR
p =1
p * HVp
where:
OCkav annual average overhead costs per 1000 veh-km for vehicle type k
OCkp overhead costs per 1000 veh-km for vehicle type k during traffic flow period
p

Version 1.0
Module C: Travel Time
14 Passenger travel time

The number of passenger-hours is calculated as a function of the vehicle operating speed.
The number of passenger-hours spent in travelling for each vehicle type k, during each traffic
flow period p is calculated separately for travel during working hours and for travel during
non-working hours. This makes it possible to assess the delays associated with the particular
operating conditions of each traffic flow period.
14.1 Working passenger-hours

The number of passenger-hours per 1000 vehicle-kilometres spent travelling during working
time is given as:
1000 * PAX k * Wk
PWH kp = ...(14.1)
100 * SS kp
where:
PWHkp number of working passenger-hours per 1000 veh-km, for vehicle type k
PAXk number of passengers (non-crew occupants) in vehicle type k
Wk percentage of passengers on work-purpose journey (%)
14.2 Non-working passenger-hours

The number of passenger-hours per 1000 vehicle-kilometres spent in travelling during non-
working time is given by the following expression:
1000 * PAX k * (100 - W k )

PNH kp = ...(14.2)
100 * SS kp
where:
PNHkp number of non-working passenger-hours per 1000 veh-km for vehicle type k
PAXk number of passengers (non-crew occupants) in vehicle type k
Wk percentage of passengers on work-purpose journey (%)

Version 1.0
14.3 Annual average number of passenger-hours

The annual average number of working and non-working passenger-hours is calculated as
shown in Sections 14.3.1 and 14.3.2.
14.3.1 Working passenger-hours

The annual average number of working passenger-hours per 1000 vehicle-kilometres is given
by the expression:
n
∑ HRYR
p =1
p * HVp * PWH kp
PWH kav = n
...(14.3)
∑ HRYR
p =1
p * HVp
where:
PWHkav annual average number of working passenger-hours per 1000 veh-km, for
vehicle type k
PWHkp number of working passenger-hours per 1000 veh-km, for vehicle type k
14.3.2 Non-working passenger-hours

The annual average number of non-working passenger-hours per 1000 vehicle-kilometres is
given by the expression:
n
∑ HRYR
p =1
p * HVp * PNH kp
PNH kav = n
...(14.4)
∑ HRYR
p =1
p * HVp
where:
PNHkav annual average number of non-working passenger-hours per 1000 veh-km

PNHkp number of non-working passenger-hours per 1000 veh-km, for vehicle type k

Version 1.0
15 Cargo holding time

This refers to the number of vehicle-hours spent in transit, and it is calculated as a function of
the vehicle operating speed.
The number of cargo holding hours per 1000 vehicle-kilometres for each vehicle type k,
during each traffic flow period p, is calculated using the formula:
1000
CARGOH kp = ...(15.1)
SS kp
where:
CARGOHkp annual number of cargo holding hours per 1000 veh-km, for vehicle type k
The annual average number of cargo holding hours per 1000 vehicle-kilometres is calculated
as follows:
n
∑ HRYR
p =1
p * HVp * CARGOH kp
CARGOH kav = n
...(15.2)
∑ HRYR
p =1
p * HVp
where:
CARGOHkav annual average number of cargo holding hours per 1000 veh-km, for vehicle
type k
CARGOHkp number of cargo holding hours per 1000 veh-km, for vehicle type k during
traffic flow period p

Version 1.0
16 Road impassability costs

The cost due to impassability of a seriously damaged unsealed road is calculated as follows:
CPASS k = VC kav * (FPASS k - 1) ...(16.1)
  GH 
FPASS k = 1 + (FPLIM k - 1) * MAX 0, 1 −  ...(16.2)
  GHMIN 
where:
CPASSk cost due to impassability of vehicle type k, in currency per 1000 km

VCkav unit annual average vehicle operating cost (and travel time cost)
per 1000 veh-km of vehicle type k
FPLIMk user-specified vehicle-specific dimensionless maximum value of FPASS
(default = 1.0)
GH mean gravel thickness for the analysis year (mm)
GHMIN minimum gravel thickness (mm), determined as:
GHMIN = MIN (100.0, MAX (40.0, 2*D95))
where:
D95 = maximum particle size (mm)
Note that the physical explanation for increasing FPASS for gravel thickness less than the
minimum is that there is greater risk of weak spots and of increased vehicle costs, in this range
than for roads with adequate gravel cover thickness.
The factor FPLIM ranges in value, from:
1 for subgrade materials with soaked CBR greater than 10% (that is, fully
passable)
to
3 for heavy vehicles on soft soils
A default value of 1.0 is used which the user may override. By definition, CPASS is zero for
paved roads, and GH is zero for earth roads.

Version 1.0
Module D: Costing Vehicle Resources
17 Vehicle resource costs
17.1 Unit costs

In previous sections the analysis has dealt, wherever feasible, with physical quantities of
resources used, so that fundamental physical relations would not be obscured by price
variations. Once the physical quantities are determined, they are multiplied either by unit costs
or prices. The user provides the unit costs or prices.
Physical concepts prove difficult to define and quantify for maintenance parts and for
overhead and miscellaneous costs, and are not relevant for depreciation and interest, which are
financial in nature. For three of these elements, that is, all except overhead and miscellaneous
costs, it is convenient and valid to deal with the ratio of the element’s cost to the price of a
new vehicle, another cost factor to be supplied by the user. Overhead is treated as a lump sum
per year. Miscellaneous costs are also treated as a fraction of other costs.
Table E2.16 shows the units in which each element of resource consumption is measured and
the dimensions of the price, unit cost, or other factor by which each has to be multiplied to
obtain its value as a component of vehicle operating cost per 1000 vehicle-kilometres.
17.2 Vehicle-trip costs over the road section

For each section alternative j, for each vehicle type k, and for each year of the analysis period,
the costs per vehicle-trip over the road section are obtained from the following expression:
KCOST jk * L j
TRIPCOST jk = ...(17.1)
1000
where:
TRIPCOSTjk cost per vehicle-trip over the road section under investment alternative j
KCOSTjk cost per 1000 vehicle-kilometres of vehicle type k under section alternative j
Lj road section length under investment alternative j (km)
The annual average costs per vehicle trip are required as inputs for economic analysis and
comparisons of different investment options. They are used for calculating the net benefits of
each pair of section options to be compared (see Part G).

Version 1.0
Table E2.16 Predicted vehicle resource consumption
Resource Units of measurement Unit cost or other

component multiplying factor
Fuel Litres per 1000 vehicle-kilometres, FC Cost per litre
Oil Litres per 1000 vehicle-kilometres, OIL Cost per litre
Tyres Number of equivalent new tyres per 1000 vehicle- Cost per tyre
kilometres, TC
Parts Proportion of new vehicle cost per 1000 vehicle- Cost of new vehicle
kilometres, PC
Labour Labour-hours per 1000 vehicle-kilometres, LH Wage cost per hour
Depreciation Fraction of new vehicle cost less tyres per 1000 Cost of new vehicle minus cost of
vehicle-kilometres, DEP tyres
Interest Fraction of new vehicle cost per 1000 vehicle- Cost of new vehicle
kilometres, INT
Crew Hours per 1000 vehicle-kilometres, CH Crew wages per hour
Overhead Cost per 1000 vehicle-kilometres, OH No factor needed
Passenger working Working passenger-hours per 1000 vehicle- Value per hour of passenger
time kilometres, PWH working time
Passenger non- Non-working passenger-hours per 1000 vehicle- Value per hour of passenger non-
working time kilometres, PNH working time
Cargo holding time Vehicle-hours per 1000 vehicle-kilometres, Cargo holding cost per vehicle-hour
CARGOH
Unsealed road Fraction of above costs Sum of costs above

impassability
17.3 Annual cost streams

The annual road user costs for each vehicle type are obtained by multiplying the cost per
vehicle-trip by the annual traffic volume of the particular vehicle type. Summing these values
over all the vehicle types used in the analysis yields the total vehicle operating cost and travel
time cost. These annual total costs may also be computed by:
! Traffic flow periods
! Vehicle resource component
! Traffic categories
(that is, normal and diverted, and generated)
17.4 Annual vehicle-kilometres

For each section alternative j, the annual vehicle-kilometres (VKMjk) of vehicle type k is
obtained by multiplying the annual traffic volume of vehicle type k (Tjk) by the section length
(Lj).

Version 1.0
18 References
ISOHDM Publications, (1994 - 1996)
NDLI, (1995)
Modelling Road User Effects in HDM-4
Final Report to the Asian Development Bank (RETA: 5549)
N. D. Lea International Ltd., Vancouver, Canada
Bennett C.R., and Greenwood I.D., (1996)
Specifications for the HDM-4 Road User Effects Model. Third and fourth Draft
Bennett C.R., (1996b)
Modelling Capital Costs HDM-4
Report to the International Study of Highway Development and Management Tools
Bennett C.R., (1996c)
Modelling the effects of traffic congestion on fuel consumption in HDM-4: Overview of
methodology and specifications
Bennett C.R., (1998)
Structured Testing of the HDM-4 Road User Effects Model for Australia
Highway and Traffic Consultants Ltd., UK
Biggs D.C., (1988)
ARFCOM – Models for Estimating Light to Heavy Vehicle Fuel Consumption
Research Report ARR 152, Australian Road Research Board, Nunawading, Australia
du Plessis H.W., editor (1989)
An Investigation of vehicle Operating Cost Relationships for use in South Africa
NITRR Report DPVT-C96.1, CSIR, Pretoria
Harrison R., and Aziz S., (1998)
HDM-4 Tyre Relationships, Memorandum to ISOHDM Secretariat
Centre for Transportation Research, The University of Texas at Austin, USA
Hine J. L., (1996)
Briefing Paper on Vehicle Utilisation and Service Life for the December 1996 Workshop
on HDM-4, Unpublished Project Report, PR/ORC/590/96 R6472
For International Study of Highway Development and Management Tools
Overseas Unit, TRL, Crowthorne Berkshire, UK
World Bank Publications. Washington D.C.

Version 1.0
McLean J.R., (1991)

Adapting the HDM-III Vehicle Speed Prediction Models for Australian Rural Highways
Working Document TE 91/014, Australian Road Research Board, Nunawading,
Australia
Pienaar W.J., (1984)
Olieverbruik van Padvoertuie in Suid-Afrika
NITRR Report RT/5, CSIR, Pretoria
St. John A.D. and Kobett D.R., (1978)
Grade Effects on Traffic Flow Stability and Capacity
National Cooperative Highway Research program Report 185
Transportation Research Board, National Research Council, Washington D.C.
HDM-III Model.
Tsunokawa K, (1987b)
Yuli P., (1996)
Development of Speed and Fuel Consumption Models for Chinese Vehicles
PhD Thesis, School of Civil Engineering, University of Birmingham, UK

Version 1.0
E3 Non-Motorised Transport
1 Introduction
This chapter describes the implementation of Road User Effects (RUE) models for calculating
non-motorised vehicle speeds, operating costs and travel time (see Figure E3.1). It provides an
overview of the modelling concepts and logic, a description of the relationships, and default
parameter values for each of the RUE components and the HDM-4 representative Non-
Motorised Transport (NMT) types.
Road
RoadUser
UserEffects
Effects
Overview Vehicle
VehicleSpeeds
Non-Motorised Road
RoadSafety
Safety
Overview
and
andOperating
Operating Transport
Transport
Chapter
ChapterE1
E1 Chapter
ChapterE4
E4
Costs
Costs Chapter
ChapterE3
E3
Chapter
ChapterE2
E2

Version 1.0
PART E ROAD USER EFFECTS E3 NON-MOTORISED TRANSPORT
2 Modelling concepts and logic

Non-motorised modes of transport account for the vast majority of the movement of people
and goods in many developing countries. For this reason the inclusion of NMT in the
appraisal of transport projects and policies in countries with significant volumes of NMT is
essential. For example, the presence of NMT can influence the speed of motorised transport
(MT), thereby affecting the operating costs of motorised vehicles. In addition, policies such as
road improvements influence the costs and benefits to both motorised and non-motorised road
users.
2.1 NMT vehicle types

Non-motorised transport is considered as separate vehicles used for moving passengers and
goods in their own right, and not only as side-friction or nuisance to motorised vehicles.
Hence, NMT effects can be included within the analytical framework for road projects,
programmes and strategies. The NMT costs and benefits are calculated separately for different
types of NMT vehicles.
The NMT category includes the following representative classes (see Chapter E1):
! Pedestrian
! Bicycle
! Cycle rickshaw
! Animal cart
! Farm tractor (not included in this release)
Users can define their own set of NMT vehicles within each class by calibrating the default
NMT class. Table E3.1 gives the default values of the key NMT characteristics required for
the analyses.
The calculation of NMT speeds utilises the performance and physical size of these NMT
types. The operating costs are calculated separately according to the utilisation of the NMT
vehicle (for example, private or commercial use, passenger or freight use).

Version 1.0
Table E3.1 Default values of key NMT characteristics
Key parameters Units NMT type

Bicycle Rickshaw Animal cart Pedestrian
(Bullock)
Wheel type Pneumatic Pneumatic Wood
Number of wheels, 2 3 2
NUM_WHEELS
Wheel diameter, m 0.7 0.7 1.0

WHEEL_DIA
Operating weight, kg 100 300 1200 80

WGT_OPER
Payload, PAYLD kg 35 235 900 15
Average life, LIFE0 years 10 6 3
Baseline annual km 2500 7200 4000

number of kilometres
travelled, AKM0
Baseline annual hours 150 500 1300

number of working
hours, HRWK0
Number of passengers, 1 3 0 1
PAX
Source: Odoki and Kerali (1999)
2.2 Modelling issues

The types of road improvements that are directly relevant to NMT user costs and benefits
include the following:
! Addition of NMT lanes
! NMT designated shoulders
! Pavement widening
! Improvement of pavement surface condition
! Improvement of road geometry characteristics
These improvements would affect the performance characteristics of NMT; thereby affecting
NMT user costs and benefits in terms of the following:
! Travel speed and time
! Wear and tear of NMT vehicles and components
! Fares/user charges
! Degree of conflicts with MT traffic
! Accident rates
In modelling terms, the effects on NMT that need to be considered in the economic analysis of
road investments can be quantified separately as follows (PADECO, 1996):

Version 1.0
1 MT flow and speed
2 MT operating costs
3 NMT flow and speed
4 NMT operating costs
5 NMT energy consumption
6 MT and NMT safety-related costs
7 Road deterioration and maintenance
8 NMT travel demand
The first six effects (1 to 6) are modelled in HDM-4, and are discussed further in the
following sections. The NMT effects on road deterioration and maintenance, and NMT travel
demand are currently not considered.
2.3 Data requirements

The data required for modelling NMT can be grouped as follows:
! Physical characteristics of NMT
(for example, operating weight, type of wheels).
! NMT utilisation
(for example, average service life, annual number of kilometres travelled, annual number
of working hours).
! Unit costs
(for example, purchase cost, interest rate, hourly crew wages, passenger and freight value
of time, energy). It is recognised that energy costs may be difficult to obtain, therefore
the default values given in Table E3.2 may be used.
! Road characteristics
(for example, length, vertical alignment, surface condition).
! Model calibration parameters
! NMT traffic data
(in terms of AADT, composition, and growth).
Table E3.2 Default energy unit costs in US dollars per MJoule
NMT type
0.10 0.10 0.07 0.10
2.4 Computational logic

For each road investment option, and for each analysis year, the NMT models are applied in
the following sequence:
1 Calculate average daily flow for each NMT type
2 Calculate the operating speed for each NMT type

Version 1.0
3 Calculate time costs and operating costs
4 Perform economic analysis, and energy balance analysis

Version 1.0
3 Impact of NMT on motorised transport
3.1 Impact on MT speed

The impact of Non-Motorised Transport (NMT) on Motorised Transport (MT) speed is
modelled through the side-friction or speed reduction approach (Hoban, 1987). In this
approach, it is assumed that the reduction in the steady-state speed of MT vehicles is directly
proportional to the degree of conflicts between MT and NMT. This is user defined through the
MT speed reduction factor (XNMT) for each road section being analysed. The XNMT value
is used in the relationship for calculating the free speeds of MT vehicles (see Chapter E2
Section 3.2.5). The XNMT value for a road section remains constant over the analysis period
until a road improvement alters the characteristics or degree of conflicts between NMT and
MT. The benefits to motorised traffic of improving the flow conditions for NMT traffic can
therefore be quantified from the effects of changing the XNMT value before and after the road
improvement.
3.2 Impact on MT operating costs

The impact of NMT on MT operating costs is estimated through speed change cycles or
acceleration effects models. Under ideal conditions drivers would maintain a steady-state
speed without any decelerations and accelerations. However, this is not possible in reality
since drivers are forced to adjust their speeds in response to traffic congestion, road alignment,
pavement surface condition, the presence of NMT and other roadside activities (Greenwood
and Bennett, 1996; Bennett, 1996; NDLI, 1995).
The standard deviation of accelerations, referred to as the acceleration noise, gives an
indication of the severity of speed changes. Low values of acceleration noise indicate that
there are minor speed changes, large values indicate major speed changes. For each road
section, it is considered that the additional MT operating costs due to speed change cycles
arising from the presence of NMT is proportional to the magnitude of acceleration noise (see
Chapter E2 Section 4.6). An appropriate intervention (for example, road improvement) to
reduce the degree of conflicts between NMT and MT would lead to lower values of
acceleration noise, and hence a benefit to motorised traffic in terms of savings in vehicle
operating costs (VOC).

Version 1.0
4 NMT speeds
4.1 Factors influencing NMT speeds

There are several factors that influence NMT speeds (PADECO, 1997):
! MT traffic volume and speed
! NMT traffic volume
! Roadside activities
! Roadway grade
! Rolling resistance
! Road width (where NMT can travel safely) and/or number of lanes
! Method of separating NMT/MT traffic (for example, markings, physical separation)
! Roughness of road surface (particularly shoulder roughness)
! Inclement weather
However, to capture the effects of all these factors would necessitate the formulation of a
complex NMT speed model and calibration procedure. A simplified speed model has
therefore been adopted, based on the minimum limiting velocity approach used in HDM-4
(Watanatada et al., 1987b). The steady-state speed of NMT vehicles is regarded as the
minimum of potential speed constraints generated by the interaction of road severity factors
with relevant characteristics of the vehicle. The limiting factors include road roughness,
desired speed, and the road gradient. With the exception of NMT traffic volume, all of the
factors given above have been considered either explicitly or implicitly in the speed model or
in the calculation of forces opposing motion and their impact on energy use.
4.2 The speed model

The NMT speed over a road section is calculated separately for each direction of traffic flow,
the uphill segment and the downhill segment, and the results are then averaged for a round
trip.
The uphill and downhill speeds are calculated using Equations 4.1 below and 4.2 below
respectively:
VS ku = MAX[0.14, MIN(VDES ks , VROUGH k , VGRAD ku )] …(4.1)
VS kd = MAX[0.14, MIN(VDES ks , VROUGH k , VGRAD kd )] …(4.2)
The average speed for a round trip is calculated as the harmonic mean of the uphill and
downhill speeds from the expression:
7.2 * XMT
Sk = …(4.3)
 1   1 
 VS  + VS  
 ku   kd 
where:

Version 1.0
Sk the average speed for NMT type k (km/h)

VSku the predicted speed of NMT type k for the uphill segment (m/s)
VSkd the predicted speed of NMT type k for the downhill segment (m/s)
VDESks desired speed of NMT type k on a smooth, level road (s = paved or unpaved
surface) (m/s) (see Table E3.3)
VROUGHk the speed limited by roughness (m/s) (see Section 4.2)
VGRADk the speed limited by the roadway gradient (m/s), (the uphill and downhill
directions are denoted by the subscripts u and d respectively) (see Section 4.3)
XMT speed reduction factor due to motorised traffic and roadside activities,
allowable range:
MIN 0.4 to MAX 1.0 (default = 1.0)
For the analysis of one-way traffic, the average speed is calculated as follows:
S k = 3.6 * VS ku * XMT …(4.4)
S k = 3.6 * VS kd * XMT …(4.5)
The benefits to NMT users of improving traffic flow conditions, for example by separating
NMT and MT traffic flows can be assessed by changing the value of the parameter XMT in
Equations 4.3 above to 4.5 above.
4.3 VROUGH
The limiting speed due to road roughness and the associated ride severity on NMT is
estimated from:
VROUGH k = VDES ks + a_rgh * RIav ...(4.6)
where:

a_rgh roughness dependent model coefficient for NMT type k
Table E3.3 shows typical default parameter values for the NMT speed-roughness model. It is
expected that a proportion of NMT traffic will use part of the carriageway and the shoulders
(if provided). This will need to be specified, so that carriageway roughness is applied to the
former and the shoulder roughness applied to the latter. The NMT benefits arising from
improvement of road surface condition can be assessed in terms of increased operating speeds

Version 1.0
Table E3.3 Default values of NMT speed model parameters
Key parameters Units NMT type

(Bullock)
Desired speed on km/h 21.26 18.60 3.83 5.11
paved roads VDESp)
Desired speed on km/h 18.00 15.40 3.20 4.60

unpaved roads
(VDESu)
Roughness dependent -0.225 -0.197 -0.036 -0.048

speed model
coefficient (a_rgh)
Gradient dependent -49.00 -47.00 -6.00 -9.20

speed model
coefficient (a_grd)
Critical gradient -0.04 -0.04 -0.04 -0.04

(CRGR)
4.4 VGRAD
The effect of gradient on NMT speed is calculated separately for uphill and downhill travel.
For the uphill direction, the limiting speed due to gradient effects is given by:
VGRAD ku = VDES ks + a_grd * GR ...(4.7)
where:
GR average road gradient (as a fraction)

a_grd gradient dependent model coefficient for NMT type k
For the downhill direction, the limiting speed depends on the critical gradient (CRGR). It is
assumed that below the critical gradient, there is no effect of downgrade on NMT speed. The
downhill speed is therefore calculated as follows:
if: abs(GR ) > abs(CRGR )
VGRAD kd = VDES ks + a_grd * ABS(GR - CRGR k ) ...(4.8)
otherwise:
VGRAD kd = VDES ks ...(4.9)
where:
CRGRk the critical gradient for NMT type k (default = -0.04 for all NMT types, range:
-0.15 < CRGR < 0)

Version 1.0
The average road section gradient (GR) is estimated from the following expression
(Watanatada et al., 1987a):
RF
GR = ± ...(4.10)
1000
where:
GR For uphill speed: GR = + [RF/1000]

For downhill speed: GR = - [RF/1000]
RF road rise plus fall (m/km)
Table E3.3 includes typical default parameter values for the NMT speed-gradient model. The
benefits to NMT of improving the vertical alignment of roadway, measured in terms of
increased speeds, can be estimated using Equations 4.7 above and 4.8 above.
4.5 Resistance to motion

Speeds of motorised vehicles include a component derived from the main forces opposing
motion; that is, aerodynamic resistance, gradient resistance, and rolling resistance. These
forces could be used to derive a limiting speed (VDRIVE) based on the driving power of
NMT vehicles from mechanistic principles, and could then be used to substitute VGRAD in
Equations 4.1 above and 4.2 above. However, this formulation would require the estimation of
used driving power for each NMT type. Since this is difficult to determine, a simple form of
the NMT speed model is used incorporating gradient effects only.
The effect of rolling resistance on NMT speed can be significant on unsealed roads,
particularly on soft and sand surfaces. This together with gradient resistance has therefore
been incorporated in the NMT energy cost model described in Section 5.6. The benefits to
NMT in terms of savings in energy used would therefore result from the upgrading of
pavement types, and improving road geometry characteristics. The gradient and rolling
resistance forces are calculated as described below.
4.5.1 Gradient resistance

The gradient resistance is calculated separately for the uphill and downhill segments using the
FG = WGT_OPER * g * GR ...(4.11)
where:
FG gradient resistance (N)

WGT_ NMT operating weight (kg)
OPER
g acceleration due to gravity, taken as equal to 9.81 (m/s2)

Version 1.0
For the uphill travel direction, FGu is calculated using the positive value of GR, and for the
downhill travel direction, FGd is calculated using the negative value of GR. The subscripts u
and d denote uphill and downhill, respectively.
4.5.2 Rolling resistance

Several studies including those by Cenek (1994), Bester (1981), and CRRI (1985), has found
that a relationship exists between motorised vehicle speeds and the rolling resistance
characteristics of roads. This finding has been extended to apply to NMT. The rolling
resistance to NMT (excluding pedestrians) is calculated using the model formulation by Biggs
(1988), as follows:
[ (
FR = FCLIM * CR2 * b1 * NUM_WHEELS + CR1 * b2 * WGT_OPER + b3 * v 2 )]
...(4.12)
where:
FR rolling resistance to motion (N)

NUM_ number of wheels of NMT
WHEELS
v NMT speed (m/s), taken as VSku or VSkd
CR1 wheel dependent factor of rolling resistance
CR2 pavement dependent coefficient of rolling resistance
FCLIM climatic/inclement weather factor
b1, b2, b3 model parameters
The wheel factor CR1 depends on the wheel type as follows:

! wheel type is steel or wood
CR1 = 0.9
! wheel type is pneumatic
CR1 = 1.0
The pavement dependent coefficient of rolling resistance CR2 is given by:
CR2 = Kcr2 * [CR2_a0 + CR2_a1 * TD av + CR2_a2 * RIav ] ...(4.13)
where:

TDav average sand patch texture depth (mm)
Kcr2 rolling resistance factor
Table E3.4 gives the default values for the various rolling resistance model parameters. The
diameter and number of wheels influence the rolling resistance parameters as follows:

Version 1.0
b1 = = WD_a0 * WHEEL_DIA ...(4.14)
WD_a1
b2 = ...(4.15)
WHEEL_DIA
WD_a2 * NUM_WHEELS
b3 = ...(4.16)
WHEEL_DIA 2
where:
WHEEL_ wheel diameter (m)

DIA
WD_a0 model coefficients (defaults = 37, 0.064, and 0.012 respectively)
WD_a1
WD_a2
Table E3.4 Rolling resistance model parameters
Pavement WGT_OPER < = 2500 kg WGT_OPER > 2500 kg

Surface Surface CR2 CR2 CR2 Kcr2 CR2 CR2 CR2 Kcr2
class type _a0 _a1 _a2 _a0 _a1 _a2
Bituminous AM or ST 0.90 0.022 0.022 1 0.84 0.03 0.03 1
Concrete JP, JR or CR 0.90 0.022 0.022 1 0.64 0.03 0.03 1
Unsealed Gravel 1.00 0.00 0.075 1 1.00 0.00 0.075 1
Unsealed Earth 0.80 0.00 0.10 1 0.80 0.00 0.10 1
Unsealed Sand 7.50 0.00 0.00 1 7.50 0.00 0.00 1
Block CB, BR or SS 2.00 0.00 0.00 1 2.00 0.00 0.00 1
Source: NDLI (1995)
Notes: (see Part C, Chapter C1 for the definitions of surface types)

The following abbreviations were used in Table E3.4:
AM = Asphalt Mix
ST = Surface Treatment
JP = Jointed Plain
JR = Jointed Reinforced
CR = Continuously Reinforced
CB = Concrete Block
BR = Brick
SS = Set Stone

Version 1.0
The effect of inclement weather is incorporated in the rolling resistance model in a similar
manner to that for motorised vehicles. The rolling resistance factor depends upon the
percentage of time travelled on water covered roads (PCTDW), and if applicable, on snow
covered roads (PCTDS) as follows:
FCLIM = 1 + 0.002 * PCTDW + 0.003 * PCTDS ...(4.17)

Version 1.0
5 NMT time and operating costs

The total time and operating costs of each NMT type are calculated separately then added:
TOC k = TMC k + VOC k ...(5.1)
where:
TOCk total time and operating cost of NMT type k per veh-km
TMCk travel time cost of NMT type k (cost/km)
VOCk operating cost of NMT type k (excluding pedestrians) (cost/km)
5.1 Travel time cost

The cost of travel time is directly related to average speeds. The time cost comprises
passenger time value and cargo holding cost, and this is expressed as follows:
TMC k = PAXC k + CARGC k ...(5.2)
where:
PAXCk passenger time value for NMT type k per veh-km

CARGCk cargo holding cost for NMT type k (cost/km)
The value of passenger time is given by:
PAXVk
PAXC k = ...(5.3)
Sk
where:
PAXVk average hourly value of passenger time for NMT type k (cost/h), and is equal
to the number of passengers per vehicle (PAX) multiplied by the value of
passenger time (PTV)
Sk annual average speed of NMT type k (km/h)
The cost of cargo holding is given by:
CAGVk
CARGC k = ...(5.4)
Sk
where:
CAGVk average hourly value of cargo holding time for NMT type k (cost/h)

Version 1.0
5.2 Operating cost

The cost of operating each NMT type is obtained from the costs of capital depreciation, repair
and maintenance, crew (if any), energy and overheads:
VOC k = CAPC k + RMC k + CRWC k + ENC k + OVHD k ...(5.5)
where:
VOCk total operating cost of NMT type k per veh-km

CAPCk capital cost of NMT type k per km (excluding pedestrians) (cost/km)
RMCk repair and maintenance cost of NMT type k (excluding pedestrians) (cost/km)
CRWCk crew cost of NMT type k (excluding pedestrians) (cost/km)
ENCk energy cost of NMT type k (cost/km)
OVHDk overhead cost of NMT type k (excluding pedestrians) (cost/km)
5.3 Capital cost

The capital cost is derived from the purchase cost depreciated over the average service life of
each NMT vehicle, the frequency of utilisation, and the interest charge on the purchase cost.
Hence, the capital cost per km, CAPC, is given by:
CAPC k = DEPC k + INTC k ...(5.6)
where:
DEPCk depreciation cost of NMT type k per veh-km

INTCk interest cost of NMT type k (cost/km)
The depreciation cost per km, DEPC, is calculated using the expression:
PCHC k
DEPC k = ...(5.7)
LIFE0 k * AKM k
where:
PCHCk average purchase cost (price) of NMT type k

LIFE0k baseline average service life of NMT type k (years), input by the user
AKMk annual kilometres travelled by NMT type k (km/year)
Interest cost per km, INTC, is calculated from the following expression:
PCHC k * AINVk
INTC k = ...(5.8)
2 * 100 * AKM k

Version 1.0
where:
AINVk the annual interest charge on the purchase cost of NMT type k (%)
For each analysis year, the annual kilometres travelled, AKM, is calculated as follows:
AKM k = S k * HRWK0 k ...(5.9)
where:
Sk annual average speed of NMT type k (km/h)

HRWK0k baseline average number of NMT working hours per year, input by the user
5.4 Repair and maintenance cost

The repair and maintenance cost per km, RMC, includes costs of replacing tyres, braking
devices and other components, lubricating oil and the cost of maintenance labour. This can be
estimated using the model form in Equation 5.10 below comprising two components:
1 Road roughness (RI)
2 NMT vehicle age measured in terms of cumulative kilometres travelled (CKM).
RMC k = (a_rmc + b_rmc * RIav ) * CKMk * PCHCk * 10 -3 ...(5.10)
CKM k = 0.5 * AKM0 k * LIFE0 k ...(5.11)
where:
CKMk average cumulative kilometres travelled by NMT type k (km)

AKM0k baseline average annual kilometres travelled by NMT type k (km/year), input
by the user
a_rmc and model calibration coefficients
b_rmc
Table E3.5 gives the default values of the repair and maintenance cost model calibration
coefficients.

Version 1.0
Table E3.5 Default parameter values for NMT repair and maintenance cost
model
NMT type Model parameter

a_rmc b_rmc
-6
(x 10 ) (x 10-6)
Bicycles 1.600 0.267
Rickshaws 0.712 0.178
Bullock carts 2.780 0.617
Pedestrians
5.5 Crew cost

The crew cost per km, CRWC, of each NMT vehicle type is calculated from the average crew
wages per hour as follows:
CRWVk
CRWC k = ...(5.12)
Sk
where:
CRWCk average crew wages per hour for NMT type k (cost/hr)
Sk annual average speed of NMT type k (km/hr)
5.6 Energy cost

Methods for modelling energy used by motorised and non-motorised transport is described in
Kerali et al. (1997) and in Part F of this document. Therefore only a summary of the
quantitative methods used is given here:
Energy cost per km, ENC, is calculated as follows:
ENC k = ENUSD k * UCEN k ...(5.13)
where:
ENCk energy cost per km for NMT type k (cost/km)

ENUSDk average energy consumption of NMT type k (Joules/km)
UCENk unit cost of energy used by NMT type k (cost/Joule)
Energy consumption per NMT veh-km, ENUSD, is calculated for the uphill and downhill
directions of traffic flow using Equations 5.14 below and 5.15 below, respectively:
ENUSD ku = (FR ku + FG ku ) * DIST ...(5.14)

Version 1.0
ENUSD kd = (FR kd + FG kd ) * DIST ...(5.15)
The average energy consumption for a round trip is then calculated using Equation 5.16
below, which incorporates an energy factor to account for energy used to overcome other
forces opposing motion:
ENUSD k = 0.5 * Kef k * (ENUSD ku + ENUSD kd ) ...(5.16)
where:
ENUSDk the average energy consumption (Joules/km) for NMT type k

FRku uphill rolling resistance to NMT type k (N)
FGku uphill gradient resistance to NMT type k (N)
DIST unit distance travelled by NMT (m), taken as equal to 1000
FRkd downhill rolling resistance to NMT type k (N)
FGkd downhill gradient resistance to NMT type k (N)
Kefk energy efficiency factor for NMT type k, to account for energy used to
overcome other forces opposing motion (default = 1.1 for all NMT types)
Walking
Replogle (1992) estimates the energy used in walking as 1.8 kJ/km/kg. This is equivalent to
144 kJ/km for an 80kg person. Thus, the average energy consumption used by walking is
estimated from the following expression:
ENUSD k = 1.8 * WGT_OPER * 10 3 ...(5.17)
5.7 Overheads
This covers all other cost elements including; taxes, administration, insurance,
parking/garaging, and any overheads associated with the crew (for example, training,
uniforms, etc.). The overhead cost per NMT vehicle-km, OVHD, is calculated using the
expression:
OHC k
OVHD k = ...(5.18)
S k * HRWK0 k
where:
OHCk overhead cost per year for NMT type k

Version 1.0
6 Estimation of economic benefits

For a given road section, the annual economic benefits in terms of savings in NMT user costs
are calculated separately for normal traffic and diverted traffic, and for generated traffic. The
economic analysis method is described in Part G - Section 5.2.2.
The key input parameter for the economic analysis is the annual average NMT time and
operating cost per vehicle-trip over the road section. This is determined from the following
expression:
UTOC jsk = TOC jsk * L js ...(6.1)
where:
UTOCjsk annual average NMT total time and operating cost per vehicle-trip over the
road section s for vehicle type k under investment option j
TOCjsk total time and operating cost of NMT type k for section s under investment
option j (cost/km)
Ljs road section length under investment option j (km)
The types of outputs for Non-Motorised Transport are similar to that for Motorised Transport.
The outputs include the following:
! NMT speeds
! NMT traffic flows
! Time and operating cost
The amounts of energy consumption are considered as renewable energy in Energy Balance
analysis within Social and Environmental Effects module (see Part F).

Version 1.0
7 References
The New HDM-4 Model
Proceedings, Combined 18th ARRB
Transport Research Conference and Transit New Zealand Symposium, Christchurch,
New Zealand, Part 4 (Asset Management)
Bester C.J., (1981)
Fuel Consumption of Highway Traffic
Ph.D. Dissertation, University of Pretoria, South Africa
Biggs D.C., (1988)
ARFCOM – Models for Estimating Light to Heavy Vehicle Fuel Consumption
Research Report ARR 152
Australian Road Research Board, Nunawading, Australia
Cenek P.D., (1994)
Rolling Resistance Characteristics of New Zealand Roads
Transit New Zealand Research Report PR3-001
Wellington, New Zealand
CRRI, (1985)
Traffic Simulation Modelling Study: Part 1 – Development of Simulation Models
Central Road Research Institute, New Delhi, India
Greenwood I.D., and Bennett C.R., (1996)
The Effects of Traffic Congestion on Fuel Consumption
Road & Transport Research, Journal of Australian and New Zealand Research and
Practice, ARRB Transport Research, South Victoria, Australia, Vol. 5, No. 2, June 1996
Hoban C. J., (1987)
Evaluating Traffic Capacity and Improvements to Road Geometry
World Bank Technical Report, Paper Number 74
Kerali H.G.R., Odoki J.B., and Collings S., (1997)
Energy Balance Framework for Road Transport Analysis
January 1998, Paper No. 980819
Transportation Research Board, Washington D.C., USA
NDLI (1995)
Asian Development Bank Project RETA 5549
Odoki J.B., and Kerali H.G.R., (1999)
Modelling Non-motorised Transport in HDM-4 - TRB, Paper No. 991129
Transportation Research Board, 78th Annual Meeting, Washington D.C., USA

Version 1.0

Non-Motorised Transport (NMT) Modelling in HDM-4, Draft Final Report (second
Version)
Transport Division, World Bank, Washington D.C., USA
Recommended Short-Term Applications to Implement Non-motorised Transport (NMT)
Modelling in HDM-4
Transport Division, World Bank, Washington, USA
Replogle M., (1992)
Non-Motorised Vehicles in Asian Cities
World Bank technical Paper 162
Washington D.C., USA
The Highway Design and Maintenance Standards Model - Volume 1 - Description
Watanatada T., Dhareshwar A.M., and Rezende Lima P.R.S., (1987b)
Vehicle speeds and operating costs - HDM-III Series, 1987

Version 1.0
E4 Road Safety
1 Introduction
This chapter describes the specification for Road Safety analysis in HDM-4 (see Figure E4.1).
The HDM-4 system allows users to define a series of look-up tables for accident rates. These
are basically broad, macro descriptions of the expected accident rates which can be defined in
several different ways according to a particular set of road and traffic attributes (for example,
road type, traffic level and flow pattern, presence of Non-Motorised Transport (NMT), and
geometry class). This tabular approach to implementing road safety analysis in HDM-4 was
recommended (ISOHDM, 1995), following a detailed review of various road safety studies,
modelling and analysis methods.
Road
RoadUser
UserEffects
Effects
Overview Vehicle
VehicleSpeeds
Non-Motorised Road
RoadSafety
Safety
Overview
and
andOperating
Operating Transport
Transport
Chapter
ChapterE1
E1 Chapter
ChapterE4
E4
Costs
Costs Chapter
ChapterE3
E3
Chapter
ChapterE2
E2

Version 1.0
PART E ROAD USER EFFECTS E4 ROAD SAFETY
2 Modelling logic
2.1 Accident types

An accident is an event involving one or more road vehicles, which results in death, personal
injury and/or damage to property. In HDM-4, road safety effects are analysed according to the
following accident severity/types:
! Fatal
An accident is considered as fatal if death occurs within a fixed period (for example, 31
days) following the accident. The fixed period may vary from one country to another.
! Injury
An accident causing injury but not resulting in fatalities.

! Damage only
An accident in which no personal injuries occur is considered as a damage (to property)

only accident.
2.2 Accident rates

The term accident rate is defined as the average number of reported accidents per year,
measured over a period of time (for example, 5 calendar years), divided by the exposure. This
is expressed as follows:
ACCYR
ACCRATE = ...(2.1)
EXPOSURE
where:
ACCRATE accident rate

ACCYR number of accidents per year
EXPOSURE annual exposure to accidents
The annual exposure to accidents is calculated as follows:

! Road sections
The annual exposure is expressed in terms of one hundred million vehicle-kilometres as:
365 * AADT * L
EXPOSSEC = ...(2.2)
10 8
! Road intersections
The annual exposure is expressed in terms of one hundred million vehicles as:
365 * AADTin
EXPOSINT = ...(2.3)
10 8

Version 1.0
where:
EXPOSSEC annual accident exposure on a road section (100 million veh-km)

AADT annual average daily traffic on the section (veh/day)
EXPOSINT annual accident exposure at an intersection (100 million veh-km)
AADTin annual average daily traffic entering the intersection (veh/day)
For each road type or intersection type, the user is required to enter the rate for each accident
severity (that is, fatal, injury or damage only), in terms of the number of accidents per 100
million vehicle-kilometres or per 100 million vehicles. For analyses at the aggregate data
level, the user has the option of entering a single rate for All accident types. This value is
equal to the sum of the values of the rates for the three accident types. When a road is
improved (that is, providing separate NMT lanes, widening of road shoulders) a change to the
road type or intersection type occurs, and different accident rates need to be defined for the
new road type. Thus, it is possible to analyse the change in accident rates and costs resulting
from improving a road section or an intersection node.
Note: Modelling of intersection nodes is not included in this release.
2.3 Primary data

The following primary data are required for modelling accident effects:
! Traffic volumes
The annual average daily traffic (AADT) on a road section, in vehicles per day, (or the
total AADT entering an intersection) and the growth rates.
! Length of road section
! Accident rates
Different accident rates may apply in different analysis years depending upon the
changes in road type (or intersection type) resulting from the road works carried out over
the analysis period.
! Unit costs of accidents
The unit costs can be defined either by accident types, or a weighted mean value may be
specified for All accident types.

Road safety analysis is included within the Road User Effects (RUE) module. For each
section (or investment) option, the computational procedure for each analysis year can be
summarised as follows:
1 Initialise input data
2 Calculate the annual exposure to accidents
3 Calculate the annual number of accidents (by accident type)
4 Calculate the annual accident costs (if applicable)

Version 1.0
5 Store results for economic analysis, comparisons and the reporting phase
2.5 Number of accidents

The annual number of accidents for each investment option is given by:
ACCYR i = EXPOSSEC j * ACCRATEi ...(2.4)
where:
ACCYRi the annual number of accidents of type i

EXPOSUREj the annual exposure to accidents under investment option j
ACCRATEi the accident rate of accident type i
2.6 Accident costs

The annual accident costs for each investment option are calculated from the following
expression:
ACCOSTji = ACCYR ji * UNITCOSTi ...(2.5)
where:
ACCOSTji the annual cost of accident type i under investment option j

UNITCOSTi the unit cost of accident type i (currency)
The total annual accident cost is given as:
ACj = ∑ ACCOST
i
i ...(2.6)
where:
ACj the annual accident cost under investment option j

Version 1.0
3 Economic analysis and comparisons
3.1 Economic analysis

The user can specify whether or not to include accident costs together with vehicle operating
costs, travel time costs, road agency costs, and exogenous benefits and costs in economic
analysis. For each pair of investment options to be compared (that is, option m and the base
case option n), the annual difference in accident costs is calculated as follows:
∆ACC (m -n) = AC n - AC m ...(3.1)
where:
∆ACC(m -n) the annual difference in accident costs
For the details of economic analysis and comparisons, see Part G.
3.2 Net change in the number of accidents

For each pair of investment options, the user may choose to compare only the number of
predicted accidents for one investment option against that predicted for the base case option.
The annual net number of accidents (by accident type) is given by:
∆ACCYR (m -n)i = ACCYR ni - ACCYR mi ...(3.2)
where:
∆ACCYR(m -n)i the annual difference in the number of accidents (for accident type i)
3.3 Outputs
The standard reports for safety analysis are supplied in Applications Guide. These include:
1 Annual number of accidents, by accident type, for each investment option
2 Annual accident costs, by accident type, for each investment option
3 Net annual streams of accident numbers for each pair of investment options

Version 1.0
4 References
ISOHDM, (1995)
Predicting changes in accident rates in developing countries following modifications in

road design
International Study of Highway Development and Management
Unpublished Report for the ISOHDM
School of Civil Engineering, The University of Birmingham, UK

Version 1.0
Contents

Descriptions: Part F
F1 Overview
1 Introduction F1-1
2 Energy balance F1-2
3 Vehicle emissions F1-4
4 Applications scope F1-5
5 References F1-6
F2 Energy Balance Analysis

1 Introduction F2-1
1.1 Background to Energy Balance Analysis F2-1
1.2 Objectives of the energy balance analysis F2-2
2 Analytical requirements F2-3
2.1 Energy used by motorised vehicles F2-3
2.2 Energy used by non-motorised vehicles F2-9
2.3 Energy used during road construction and maintenance F2-11
3 Energy balance framework F2-12
3.1 Outline of the energy balance framework F2-12
3.2 Energy use factors for a motorised vehicle F2-12
3.3 Energy use factors for non-motorised transport F2-17
3.4 Total energy use F2-17
3.5 Comparison of investment options F2-20
4 References F2-22
F3 Vehicle Emissions
1 Introduction F3-1
2 Emissions model F3-2
2.1 Types of pollutants F3-2
2.2 The relationships F3-2
3 Modelling logic F3-8

Version 1.0
PART F SOCIAL AND ENVIRONMENTAL EFFECTS CONTENTS
3.1 Primary data F3-8

3.2 Emissions quantities F3-8
4 Comparisons of investment options F3-11
5 References F3-12

Version 1.0
Part F Road Map
Analytical
AnalyticalFramework
Framework
and
andModel
Model
Descriptions
Descriptions
Introduction
Introduction
Part
PartAA
Traffic
Traffic
Part
PartBB
RD
RDModel WE
WEModel RUE SEE
SEEModel
Model Model RUEModel
Model Model
Part
PartCC Part
PartDD Part Part
PartFF
PartEE
Economic
EconomicAnalysis
Analysis
PartGG
Nomenclature
Nomenclature
PartHH

Glossary
Part
PartI I
Figure F Analytical Framework and Descriptions Road Map

Version 1.0
PART F SOCIAL AND ENVIRONMENTAL EFFECTS
F1 Overview
1 Introduction
The Social and Environmental Effects (SEE) models (see Figure F1.1) are for the analysis of:
! Energy balance (see Section 2)
! Vehicle emissions (see Section 3)
It is widely recognised that energy and environmental effects need to be considered in the
assessment of alternative investment policies and projects. By adopting projects and policies
that minimise total life cycle energy use and vehicle exhaust emissions, related benefits such
as reduced vehicle operating costs, reduced pollution, reduced dependency on imports of
energy and reductions in the balance of payments deficits can be maximised. Planners and
decision makers need to be able to understand the energy implications and environmental
impacts of alternative road transport projects and policies.
Social
Socialand
and
Environmental
Effects
Overview Energy
EnergyBalance Vehicle
Overview Balance VehicleEmissions
Emissions
Analysis
Analysis
Chapter
ChapterF1
F1 Chapter
ChapterF3
F3
Chapter
ChapterF2
F2
Figure F1.1 Social and Environmental Effects module
The assessment of alternative investment policies and projects requires that the various impact
measures be translated or reduced into common policy-sensitive units, which can be
considered under a framework of multi-criteria analysis. It is expected that the evaluation of
other aspects of environmental impacts that arise from road transport (for example, noise
pollution, damage to crops and buildings, etc.) will be included in a later release.
Analytical Framework and Modl Descriptions F1-1

Version 1.0
PART F SOCIAL AND ENVIRONMENTAL EFFECTS F1 OVERVIEW
2 Energy balance
The energy used in the road transport sector forms a significant share of the total energy
consumption in most countries. The impact of road investment strategies and projects, and of
road transport policies on energy use has become an important aspect of the appraisal process.
The appraisal of road transport projects is based primarily on the assessment of economic
benefits which are estimated by comparing the total discounted costs calculated for a base
case against that for the project case. This is essentially an economic balance analysis. A
similar analytical framework is implemented for comparing the total energy used by different
modes of road transport for the base case and project case. The proposed Energy Balance
framework is used to calculate:
! Life-cycle energy consumption
Total life-cycle energy consumption at both project and network level analyses of road
investment policies.
! Differences in consumption
Differences in the consumption of renewable and non-renewable fuels by non-motorised

(NMT) and motorised transport (MT) modes.
! Total national and global energy use
This permits comparisons to be made between different modes of road transport and thereby
influence policy on long term investment in the road sector (Kerali et al., 1998).
The analytical framework, for assessing the energy implications of road investment projects
and strategies, considers a number of parameters. These have been classified into three energy
use categories:
! Energy used by motorised vehicles
The energy consumption associated with the use of motor vehicles can be split into two
broad categories:
❏ energy used to produce and deliver both the fuel and vehicle
❏ energy used to power and operate the vehicle
Together, the energy used in each of these categories constitutes the full life-cycle energy
consumption. The energy consumed by motorised vehicles depends on a wide range of
factors, related to:
❏ vehicle size
❏ weight
❏ design
❏ age
❏ road characteristics and condition
❏ traffic characteristics
The energy used by motorised vehicles is generally from non-renewable sources. The
following resource components, used in vehicle operation are included in the energy
balance analysis:

Version 1.0
❏ fuel consumption
❏ lubricating oil consumption
❏ tyre wear
❏ vehicle parts consumption
! Energy used by non-motorised vehicles
Non-motorised modes of transport (NMT) account for the vast majority of the movement
of people and goods in many countries. For this reason the inclusion of NMT in the
appraisal of transport projects and policies in developing countries is essential. For
example the presence of NMT can influence the speed of motorised transport, thereby
affecting the operating costs of motorised vehicles. In addition, policies such as road
improvements influence the costs and benefits to both motorised and non-motorised road
users. The energy used by NMT is calculated separately for animal-drawn carts, cycle-
rickshaws, bicycles and pedestrians.
! Energy used during the construction and maintenance of road networks
This is a significant aspect of the complete energy balance picture for road transport
investments. Thus, when comparing the energy implications of alternative policy or
project alternatives, it is important that this type of energy use is considered.
The energy balance framework implemented provides an efficient and neutral mechanism for
assessing the benefits of road transport investments. This avoids the possible distortion in
economic balance methods that may be influenced by fuel prices. Whilst some of the
parameters have had to be estimated from a desk study, the concept of energy balance
framework can be applied to road investment appraisal using a similar method to that used in
economic analyses.

Version 1.0
3 Vehicle emissions
The objective of modelling vehicle emissions is to assess the effects (Hammerstrom, 1995), in
terms of pollutant quantities and changes in the following:
! Road characteristics
! Traffic congestion
! Vehicle technology
The model predicts the different components of vehicle exhaust emissions as a function of
fuel consumption. Fuel consumption is a function of vehicle speed, which in turn depends on
road characteristics and the characteristics of the vehicle itself. Thus, it is possible to analyse
the change in emissions effects as a result of implementing different road maintenance and
improvement options, or the implications of major changes to the vehicle fleet (for example,
due to improved vehicle technology). The different components of emissions modelled are:
! Hydrocarbon
! Carbon monoxide
! Nitrous oxide
! Sulphur dioxide
! Carbon dioxide
! Particulates
! Lead
In this release, the effects of exhaust emissions are not costed for inclusion in economic
analysis of road investments, only the net differences in the quantities of pollutants are
assessed for each pair of investment options. It is intended that the scope of vehicle emissions
modelling will be extended (Collings and Watkiss, 1998) to include impacts on:
! Air quality
! Health
! Damage costs
! Global warming

Version 1.0
4 Applications scope
The analysis of energy balance and environmental impacts has two levels of application:
1 Project level analysis
Allows users to compare the life-cycle energy and emissions implications of a range of
project alternatives with the base line, do-minimum, alternative. The results of these
comparisons will assist in deciding which road investment alternatives should be
implemented.
2 Network level analysis
Enables decision makers to understand the energy and emissions implications of broader
transport policy objectives which have an impact on specific road networks (for example,
urban roads) as well as at the national level (for example, rural roads).

Version 1.0
5 References
Collings S.A., and Watkiss P.R., (1998)
Development of Environmental Impacts and Energy Balance Models for HDM-4
Final project report for the Department for International Development, UK
Hammerstrom U., (1995)
Proposal for a Vehicle Exhaust Model in HDM-4
ISOHDM Supplementary Technical Relationships Study Draft Report
Swedish National Road Administration, Road and Traffic Management Division,
Borlange, Sweden
Kerali H.R., Odoki J.B., and Collings S.A., (1998)
Energy Balance Framework for Road Transport Analysis
Transportation Research Board, Paper no. 980819, Washington D.C., USA

Version 1.0
F2 Energy Balance Analysis
1 Introduction
1.1 Background to Energy Balance Analysis

The energy used in the road transport sector forms a significant share of the total energy
consumption in most countries. The impact of road investment strategies and projects, and of
road transport policies on energy use has become an important aspect of the appraisal process
(see Figure F2.1). By adopting projects and policies that minimise total life-cycle energy use,
related benefits such as reduced operating costs, reduced pollution, reduced dependency on
imports of energy and reductions in balance of payments deficits can be maximised. For this
reason it is important for planners and decision-makers to be able to understand the energy
implications of alternative transport projects and policies.
Social
Socialand
and
Environmental
Effects
Overview Energy
Emissions
Analysis
Analysis
Chapter
ChapterF1
F1 Chapter
ChapterF3
F3
Chapter
ChapterF2
F2
The HDM-4 system can be used for assessing technical, economic, social and environmental
impacts of road investment. The assessment of alternative investment policies and projects
requires that the various impact measures be translated or reduced into common policy-
sensitive units. Examples of policy-sensitive units include monetary value, amount of time,
safety level, quantity of pollutant, and energy used.
Economic analyses use monetary values of resources in order to derive economic indicators
that serve in decision making on different alternatives for road investments. However, the
results of economic analyses depend to some extent on the relative monetary values placed on
the different components of vehicle resources. For example, a project whose justification is
primarily dependent on savings in petroleum consumption shows relatively less economic
benefits in an oil rich country than it would in a country which imports its fuel at a high cost.
Viewing the benefits of such projects in terms of their efficiency in energy use would place
them on an equal footing, and this yields more useful information to planners and decision-
makers.
This chapter describes a methodology for assessing the energy implications of different road
transport projects and policies within the context of the HDM-4 model. The assessment
technique is referred to here as Energy Balance analysis.
Analytical Framework and Model Descriptions F2-1

Version 1.0
PART F SOCIAL AND ENVIRONMENTAL EFFECTS F2 ENERGY BALANCE ANALYSIS
1.2 Objectives of the energy balance analysis

The principal objective of the energy balance analysis is to be able to compare total life-cycle
energy consumption as a result of different transport policies. A related objective is to
determine the relative efficiency of the different modes of transport, both motorised and non-
motorised. This efficiency should be measured with respect to the productivity of each mode
of transport, that is, the energy used to move a certain number of people or tonnage of goods.
In addition to an overall total energy balance calculation, it is also important to distinguish
between different sources of energy. Firstly, energy from renewable and non-renewable
sources should be distinguished since the environmental and socio-economic implications of
these are different. Secondly, energy expended outside of the country being studied (energy
use in the manufacture of imported vehicles, for example) is unlikely to be of concern to the
national policy maker, but is significant in terms of global sustainability.
The HDM-4 system offers tools in separate modules for project analysis, programme analysis
and network strategy analysis. The energy balance analysis has two levels of application:
1 Project appraisal level
2 Network strategic analysis level
The project analysis allows users to compare the life-cycle energy implications of a range of
project alternatives with the base line, do-nothing alternative. The results of these
comparisons assist in the decision of which investment project alternatives to implement. For
example; the decision over which maintenance techniques to use for a particular stretch of
road is influenced by the trade off between:
! Energy benefits
The energy benefits gained from lower vehicle fuel consumption, oil use and tyre wear,
due to the improved flow characteristics of the road, and
! Energy requirements
The actual energy requirements of carrying out the maintenance.

The strategic level analysis should enable decision-makers to understand the energy
implications of broader transport policy objectives, which impact both specific road networks
(for example, a particular urban area), as well as at a national level. For example, at a network
level, the energy impacts of various regional public transport investment strategies may be
investigated, or similarly, at a national level, the energy implications of fiscal policies to
promote alternative fuel types may be a concern.

Version 1.0
2 Analytical requirements
Vehicles have been categorised as Motorised Transport (MT) and Non-Motorised Transport
(NMT). The development of a model framework for assessing the energy implications of road
investment projects and strategies requires the consideration of a number of parameters.
These have been broadly classified into three energy use categories:
1 Energy used by motorised vehicles (see Section 2.1)
2 Energy used by non-motorised vehicles (see Section 2.2)
3 Energy used during road construction and maintenance (see Section 2.3)
Sections 2.1, 2.2 and 2.3 describe the parameters that need to be considered for each category
of energy use.
2.1 Energy used by motorised vehicles

The energy consumption associated with the use of motor vehicles can be separated into two
broad categories; that used to:
1 Power and operate the vehicle
2 Produce and deliver the fuel and vehicle
Together, the energy used in each of these categories encompasses the full life-cycle energy
consumption. Table F2.1 shows the various life-cycle stages that could be considered in
determining the full life-cycle energy use.
Table F2.1 Energy use stages associated with vehicle production and use
Major sections Sub-sections

Fuel production ! Raw material extraction
! Feed stock transportation
! Processing
! Fuel distribution
Vehicle manufacture ! Raw material extraction
! Processing
! Component manufacture
! Component transportation
! Assembly
! Vehicle distribution
Vehicle use ! Fuel consumption
! Oil consumption
! Tyre wearing
Vehicle maintenance and support ! Component manufacture
! Distribution
Source: ETSU, RYCA 18825001 (1997)

Version 1.0
2.1.1 Energy use due to vehicle operation

The energy consumed by motorised vehicles is dependent upon a wide range of factors
relating to the:
! Vehicle size, weight, design and age
! Road characteristics and condition
! Traffic characteristics
The energy used by motorised vehicles is from non-renewable sources.

The following resource components used in vehicle operation are considered in energy
balance analysis:
! Fuel consumption
! Lubricating oil consumption
! Tyre wear
! Vehicle parts consumption
The prediction of a MT vehicle’s energy consumption under specific driving conditions can
be considered in two ways; either by using:
1 A mechanistic approach based on physical and mechanical first principles, or
2 Actual measured data to give energy use factors for a range of vehicles operating
under different conditions
The former mechanistic approach provides a greater degree of accuracy but has far greater
data requirements, whereas the energy use factor approach can be applied relatively easily.
Fuel consumption
The fuel consumption model follows a purely mechanistic approach that allows:
! The flexibility to model individual vehicle and road characteristics
! The ability to alter the model as new technologies are introduced
The mechanistic approach has been well researched, providing a theoretical and experimental
justification for all of the numerical assumptions that are made for each of the parameters. The
standard default vehicles are based on those using petrol and diesel fuel.
Suitable models for other vehicle technologies using different fuel types (for example,
electricity, LPG, CNG, etc.) are still being developed. For analyses involving comparison with
these vehicles, the energy consumption of alternative fuelled vehicles can be estimated by
applying scale factors which relate the calculated energy consumption of petrol or diesel
vehicles to the energy consumption of various alternative fuelled vehicles.
The fuel consumption model does not include the treatment of cold starts on fuel
consumption. In order to provide the needed mixture strength for satisfactory combustion at
the start of a journey, a large excess of fuel must be delivered to compensate for the
condensing losses due to the cold engine. This is more of a problem for petrol vehicles than
for diesel.
For the majority of road project appraisals the cold start issue is not considered to be
important, with the assumption being made that all journeys are run under hot engine
conditions. However, the influence of cold starting on fuel consumption may be of particular

Version 1.0
significance in relation to some transport policies, particularly in urban areas where there are a
great number of shorter trips.
To overcome the problem of cold starts, scale factors relating the calculated fuel consumption
under normal hot conditions to fuel consumption under cold conditions can be used. The
proportion of the journey, which is run under cold conditions, can be estimated using the
CORINAIR methodology (Eggleston et al., 1993) which takes into account the average trip
length of the vehicle journeys and the average ambient temperature.
The average fuel use per vehicle kilometre for a full trip is calculated from the sum of the cold
start fuel use and hot conditions fuel use, weighted by the proportion of journey run cold and
hot, respectively.
Different transport fuels have different calorific values. Thus in order to compare like with
like the average fuel use per vehicle kilometre should be converted to energy factors per
vehicle kilometre using the energy content values given in Table F2.2.
Table F2.2 Energy content of transport fuels
Fuel Energy content

(MJ/litre)
Petrol 34.7
Diesel 38.7
LPG1 25.5
2
CNG 40
Ethanol 23.9
Methanol 18.1
Biodiesel 32.8
Source: ETSU (1996)
Notes:
1 Assumes 90% propane, 10% butane
2 Units are MJ/m3
Lubricating oil consumption

Oil consumption is calculated in terms of litres per 1000 vehicle kilometres as a function of
fuel consumption. During the procedure for energy balance analysis, the amount of
lubricating oil consumed is converted to an energy value using a conversion factor of 47.7 MJ
per litre.
Tyre wear
The tyre consumption of a vehicle is proportional to the energy requirements (see Watanatada
et al., 1987). It is calculated using a model based on slip energy theory.
Tyre consumption, in terms of number of tyres per 1000 vehicle kilometres, is calculated for
light and heavy vehicles. These figures can be converted to an energy value using a
conversion factor of 32 GJ/tonne of tyres (Department of Trade and Industry, 1996). The
weight of tyres can be estimated using the factors given in Table F2.3.

Version 1.0
Table F2.3 Tyre weights by vehicle type
Vehicle Vehicle type Tyre weights

number (kg)
1 Motorcycle 2.0
2 Small car 3.0
3 Medium car 3.5
4 Large car 4.0
5 Light delivery vehicle 4.0
6 Light goods vehicle 4.0
7 Four wheel drive 5.0
8 Light truck 7.0
9 Medium truck 12.4
10 Heavy truck 12.4
11 Articulated truck 13.7
12 Mini-bus 4.0
13 Light bus 7.0
14 Medium bus 9.8
15 Heavy bus 11.2
16 Coach 11.2
Source: ETSU (1996)
Vehicle repair and parts consumption

Maintenance and support services are very difficult to consider given the fragmented nature of
the vehicle repair and spares businesses. Their contribution to overall energy use is very small
in comparison with the other life-cycle stages. For example, one study, based on transport in
Switzerland, estimated maintenance and support services energy use to be approximately 4%
of total life-cycle energy use (Maibach et al., 1995). However, in developing countries,
where vehicle maintenance assumes a lower priority than it would in Switzerland, this
proportion is likely to be much lower. For this reason the energy use associated with
maintenance and support services is not considered to be a significant aspect of the overall
energy balance.
Nevertheless, the RUE (Road User Effects) model does calculate vehicle parts consumption.
This is measured in terms of the fraction of the new vehicle price per 1000 km. The energy
use associated with this vehicle parts consumption can be estimated by multiplying the energy
used to produce the vehicle (see Table F2.4) by the fraction of the parts price to the new
vehicle price. To calculate vehicle parts energy consumption per vehicle kilometre, the
energy used to produce the vehicle parts should be divided by the vehicle cumulative
kilometreage, at the time the parts are replaced. The typical energy used during a year can be
divided by annual average vehicle kilometres, to give vehicle maintenance energy use per
vehicle kilometre.

Version 1.0
Table F2.4 Energy use for vehicle production
Vehicle Vehicle type Vehicle mass Vehicle

number production
energy
(kg) (GJ)
1 Motorcycle 200 20
2 Small car 800 80
3 Medium car 1000 100
4 Large car 1200 120
5 Light delivery vehicle 1400 140
6 Light goods vehicle 1600 160
7 Four wheel drive 1800 180
8 Light truck 4000 400
9 Medium truck 6000 600
10 Heavy truck 10000 1000
11 Articulated truck 15000 1500
12 Mini-bus 3000 300
13 Light bus 5000 500
14 Medium bus 7000 700
15 Heavy bus 10000 1000
16 Coach 7000 700
Source: ETSU, RYCA/18825001 (1997)
2.1.2 Fuel production and vehicle manufacture energy use

An accounting framework is needed to estimate the cradle to grave energy use for the various
life-cycle stages. Under this category, the energy balance analysis framework considers
energy use associated with fuel production and vehicle manufacture as described below:
Fuel production
Figure F2.2 illustrates the stages that must be considered for petrol and diesel fuel production.
The energy use values from each stage of the fuel cycle are calculated on an-energy delivered
basis, and then aggregated to give the total energy use per unit of energy delivered for the full
fuel-cycle. In this way energy use for fuel production can be calculated per vehicle kilometre.

Version 1.0
Feed
Feedstock
stock Processing
Processing Distribution
Distribution End-use
End-use
Crude oil and

Distribution Distribution
natural gas Depot storage
Pipeline terminal
liquids
Petrol and
Pipeline Oil refinery diesel road Diesel vehicle
tankers
Gas
Crude oil Service
separation Petrol vehicle
tanker station
plant
Source: ETSU, RYCA/18825001 (1997)
Key: - - - - - means optional (storage)
Figure F2.2 Petrol and diesel fuel production cycle
The energy use associated with each of the fuel production stages (shown in Figure F2.2 for
petrol and diesel) varies considerably from country to country. However, if no local data
exists and neither do the resources to carry out a fuel-cycle analysis, then default data such as
that shown in Table F2.5 could be used to estimate the energy use associated with fuel
production. Fuel production energy use on a vehicle kilometre basis is derived from the
energy consumption during vehicle use.
Table F2.5 Fuel-cycle energy use factors
Fuel Energy use

(MJ/MJ of fuel delivered)
Petrol 0.169
Diesel 0.122
1
LPG 0.122
CNG 0.061
2
Electric 2.857
Biomethanol 0.514
Bioethanol 0.510
Biodiesel 0.655
Source: ETSU (1996)
Notes:
1 Assumes 40% of LPG comes from refineries and 60% is directly extracted
2 Assumes a higher calorific value (HCV) average generating efficiency of 35%

Version 1.0
Vehicle manufacture
Data concerning the energy used to manufacture and deliver a motor vehicle is very difficult
to obtain due to the commercial sensitivities of the motor manufacturing industry. A recent
survey of what data is available revealed a typical value of 100 GJ per medium-sized car with
a weight of 1 tonne (ETSU, 1995). This represents approximately 15% of total life-cycle
energy use and is therefore a significant part of the overall energy balance.
To estimate the energy used to manufacture other vehicle types, a first order approximation
can be obtained by scaling 100 GJ by the ratio between the weight of a medium sized car and
the weight of the other vehicle. Based on this approach Table F2.4 shows some estimates of
energy use for vehicle production for each of the 16 standard vehicle types. In addition, it
shows estimates of total life-time vehicle kilometres which can be used to calculate default
values for vehicle production energy use on a per vehicle kilometre basis. The user can input
their estimates of vehicle service life in the Vehicle Fleet folder.
The energy used in the production process of each vehicle can simply be divided by the
average total lifetime vehicle kilometres to give vehicle production energy use per vehicle
kilometre.
2.2 Energy used by non-motorised vehicles

Non-motorised modes of transport (NMT) account for the vast majority of the movement of
people and goods in developing countries. For this reason the inclusion of NMT in the
appraisal of developing country transport projects and policies is essential. For example, the
presence of NMT can influence the speed of motorised transport, thereby affecting the
operating costs of motorised vehicles. In addition, policies such as road improvement
influence the costs and benefits of not only motorised road users, but also non-motorised road
users.
With animal-drawn transport, the energy used can be justifiably included on the grounds that
possibly the only reason that the animals are kept, and fed, is so that they can provide
transport services to their owners. Thus the feed used can be looked upon as an energy loss
equal to the opportunity cost of forfeiting its energy input for another purpose. However, with
human powered transport such as cycling and walking, it is very difficult to say that the
energy consumed would not have been consumed if the trip had not taken place. It is unlikely
that people consciously fill-up by consuming say 10% more calories when they know they are
going to be making a trip.
Nevertheless, walking and cycling expend energy, and it could be argued where food is a
particularly scarce resource that this energy should be included as a cost to society. In
addition, where individuals are using their energy for powering their business (for example,
cycle-rickshaws) and are therefore very active for most of the day, they probably need to
consume several times the calories of a sedentary person.
Sections 2.2.1, 2.2.2 and 2.2.3 describe the principles behind calculating energy use for
various types of NMT.
2.2.1 Animal-drawn vehicles

The energy required for animal-drawn carts can be split into two parts:
1 Energy required for the animal to walk the required distance
2 Energy required to move the load (if a load is being pulled)
Data on energy use by animals is sparse. As a broad assumption, a relationship derived for
human walking activity may be used:

Version 1.0
1.8 kJ/km/kg (Replogle, 1992)

The two main loads that have to be overcome in pulling a laden cart are rolling resistance and
hill climbing. Thus the energy required to pull a laden cart can be estimated from the
following equation:
Power = C R Mgν + Mgνsinθ …(2.1)
where:
CR coefficient of rolling resistance

M cart mass including the load
g acceleration due to gravity
ν Speed
θ slope gradient
Table F2.6 shows a range of energy use factors derived using this formula. The animal and
load weights have been chosen to reflect those typically seen in certain parts of the world. For
example the weight of a laden cart pulled by two oxen (common in more humid zones of West
Africa (Starkey, 1993) is estimated to be 1000kg (Dennis, 1995). An ox could be assumed to
weigh 400kg. For comparison a donkey pulling a 100 kg laden cart, common in the drier
zones of West Africa, weighs around 150 kg (Starkey, 1993). A pack animal, commonly used
in hilly areas, would tend to carry loads in the region of 30-70 kg.
The values in Table F2.6 have assumed a rolling resistance for carts of 0.04 (Dennis, 1995),
an average speed of 6.4 km/h and a zero gradient.
Table F2.6 Typical energy use factors for animal transport (kJ/km)
Animal weight Cart weight, including load

(kg) (kg)
(including any load) No cart 100 500 1000
200 360 399 556 752
400 720 759 916 1110
600 1080 1119 1276 1470
800 1440 1479 1636 1830
2.2.2 Cycling
There is over 800 million bicycles worldwide (United Nations, 1993). Of these, approximately
400 million are in Asia, with 300 million in China (Replogle, 1992).
Published figures regarding the energy used in cycling vary widely. Naturally, the specific
energy use depends upon the weight of the rider and bicycle, the friction due to the cycle and
the speed of motion. Table F2.7 shows a range of cycling energy use factors quoted by Lewis
(1995).

Version 1.0
Table F2.7 Energy use factors for cycling (kJ/km)
Weight Speed
(kg) (km/h)
19.2 24 27.2 30.4
50 64.12 70.73 78.63 91.50
59 73.97 82.11 91.45 106.84
68 83.82 93.49 104.28 122.05
77 93.45 104.87 117.09 137.25
86 103.30 116.25 129.92 152.60
91 108.33 122.03 136.41 160.20
2.2.3 Walking
Replogle (1992) estimates the energy used in walking as:
1.8 kJ/km/kg
This is equivalent to 144 kJ/km for an 80-kg person (see Chapter E3).
2.3 Energy used during road construction and maintenance

Energy use during the construction and maintenance of road networks is a significant aspect
of the complete energy balance picture for road transport investments. Thus when comparing
the energy implications of alternative policy or project options it is important that this type of
energy use is considered.
Data on energy use during road construction and maintenance is sparse and, as yet, none has
been obtained. The energy used in performing the various road works activities can be broadly
considered under the following:
! Production of materials - (for example, bitumen, cement, lime, stone-aggregates, etc.)
! Delivery of materials to work sites

! Operation of equipment
However, for the purpose of energy balance analysis, such a detailed treatment is not
appropriate. Instead, a framework is implemented using aggregate level data for the energy
used in performing each of the different types of road works modelled. For example, an
average representative value of energy used per cubic metre of overlay can be specified. This
value is then multiplied by the total quantity of overlay performed on the road section to give
the total energy used. It is also necessary to distinguish between labour intensive works and
mainly mechanised works.

Version 1.0
3 Energy balance framework

Based on the approach described in Section 2, the following sections specify the calculations
required to compare the energy consumption implications of alternative transport policy
options. The outputs that are required from an energy balance analysis are:
! Total energy consumption
! Total consumption of renewable and non-renewable energy
(This is essentially a distinction between Non-Motorised Transport (NMT) energy use

and all other energy use, except if biofuels are used).
! Total consumption of energy used nationally and energy used globally
! Specific energy consumption
These can be reported by vehicle types or aggregated by vehicle class.

The following indicators usually measure the specific energy consumption:
! Average energy use, per kilometre, by mode
! Energy use, per passenger kilometre, for passenger transport modes
! Energy use, per tonne kilometre, for freight transport modes
3.1 Outline of the energy balance framework

The proposed methodology for assessing the energy implications of transport policies splits
into four main elements:
1 Generation of the energy use characteristics - for each vehicle type (both motorised
and non-motorised)
2 Incorporation of life-cycle effects
3 Calculation of total energy use
4 Generation of indicator results
The average energy use factors are combined with the estimated total annual vehicle-
kilometres operated by each mode to give total annual energy use for the policy or measure
being considered. These totals are then used to give a range of indicator results for
comparative analysis.
3.2 Energy use factors for a motorised vehicle
3.2.1 Fuel
Vehicle fuel use factors are split into factors for hot and cold operation. The fuel use per
vehicle kilometre under hot conditions comes as output from the RUE (Road User Effects)
model.
The cold start fuel use is related to the basic hot fuel use, the ambient temperature and the
average vehicle trip length. The level of cold start fuel use is related to the hot fuel use by
factor CRAT, which is a function of the ambient temperature. This is expressed as follows:
FCOLD k = CRATk * FHOTk …(3.1)

Version 1.0
where:
FCOLDk cold start fuel use for vehicle type k (litres per km)
CRATk cold start ratio at a given ambient temperature
FHOTk hot fuel use for vehicle type k (litres per km)
The hot fuel use is obtained from the following equation:
FHOTk = FC kav * 10 -3 …(3.2)
where:
FCkav annual average fuel consumption for vehicle type k (litres per 1000 vehicle-
km)
The cold start ratios (CRAT) are for passenger cars of petrol and diesel technologies, but are
also applicable to two-wheelers and vans. These ratios can be estimated from the following
CORINAIR relationships (Eggleston, 1993):
! Petrol engine cars
CRAT = 1.47 - 0.009 * TEMP …(3.3)
! Diesel engine cars
CRAT = 1.34 - 0.008 * TEMP …(3.4)
where:
TEMP the average day temperature (degrees Celsius)
Buses and trucks are usually considered to operate permanently under hot conditions, to a
good approximation, since their average trip lengths are very high.
This fuel use under cold conditions only occurs in the initial stages of a journey. The
proportion of any journey run cold is calculated as follows:
CRUNk = MAX {0, [0.698 - 0.051 * L k - (0.01051 - 0.000770 * L k ) * TEMP] }

…(3.5)
where:
CRUNk the proportion of the journey run under cold conditions

Lk average trip length for vehicle type k (km) (default=15)
Taking into account the proportion of the journey run cold, the average vehicle fuel use per
kilometre for the full trip can be calculated as follows:

Version 1.0
FAVE k = CRUNk * FCOLD k + (1 - CRUNk ) * FHOTk …(3.6)
where:
FAVEk average fuel use for vehicle type k (litres per kilometre)
The average vehicle fuel use, FAVE, is converted to energy use factor per kilometre by
applying the energy content of fuel given in Table F2.2. Thus,
ENFUEL k = FAVE k * FEC fk …(3.7)
where:
ENFUELk annual average fuel energy consumption of vehicle type k (MJ/km)

FECfk energy content of fuel type f used in vehicle type k (MJ/litre). Defaults are
given in Table F2.2
3.2.2 Lubricating oil

The annual average oil consumption, in litres per 1000 vehicle-kilometres, output from the
RUE (Road User Effects) model is converted to energy use factor using the energy conversion
factor given in Section 2.1.1 as follows:
ENOIL k = OIL kav * OEC * 10 -3 …(3.8)
where:
ENOILk annual average oil energy consumption of vehicle type k (MJ/km)

OILkav annual average oil consumption of vehicle type k (litres per 1000 vehicle-km)
OEC energy content of lubricating oil (MJ/litre), default = 47.7
3.2.3 Tyre
The annual average number of equivalent new tyres consumed per 1000 vehicle-kilometres
output from the RUE (Road User Effects) model is converted to energy use factor using the
energy conversion factor given in Section 2.1.1 as follows:
ENTYRE k = TC kav * TEC * TWGTk * 10 -3 …(3.9)
where:
ENTYREk annual average tyre energy consumption of vehicle type k (MJ/km)

TCkav annual average number of equivalent new tyres consumed per 1000 veh-km
TEC energy content of tyre (MJ), default = 32 MJ/Kg
TWGTk tyre weight of vehicle type k (kg per tyre) (see Table F2.3)

Version 1.0
3.2.4 Vehicle repair and parts

The annual average parts consumption per 1000 vehicle kilometres expressed as a proportion
of the new vehicle price output from the RUE (Road User Effects) model is converted to an
energy use factor using the energy conversion factor given in Section 2.1.1 as follows:
ENPARTk = PC kav * ENVPk * 10 -3 …(3.10)
where:
ENPARTk annual average parts energy consumption of vehicle type k (MJ/km)

PCkav annual average parts consumption per 1000 veh-km as a proportion of the new
vehicle price
ENVPk vehicle production energy use (MJ per km)
and:
ENVPROD k
ENVPk = …(3.11)
LIFEKM k
where:
ENVPRODk total energy used in the production of vehicle type k (MJ) (see Table F2.4)
LIFEKMk the predicted vehicle service life (km)
3.2.5 Global life-cycle energy use factors

To incorporate certain life-cycle aspects into the energy use factors, as discussed in Sections
3.1 and 3.3, the following relationships should be used:
EGLICYk = ENFUEL k + ENOIL k + ENTYRE k + ENPARTk + ENVPk + (ENFUEL k * FPf )

…(3.12)
where:
EGLICYk annual average life cycle energy use factor for vehicle type k (MJ/km)
FPf fuel production factor of fuel type f (MJ per MJ of fuel used), defaults are
shown in Table F2.5
The annual average global energy use per passenger-kilometre is calculated as:
EGLICYk
EGPAXKM k = …(3.13)
PAX k
where:

Version 1.0
EGPAXKMk annual average global energy use per passenger-kilometre for vehicle type k
(MJ/passenger-km)
PAXk average number of passengers per vehicle type k
The annual average global energy per tonne-km is calculated as:
EGLICYk
EGGDSKM k = …(3.14)
PAYLD k
where:
EGGDSKMk annual average global energy use per tonne-km for vehicle type k (MJ/tonne-
km)
PAYLDk average payload per vehicle type k (tonnes)
The average payload for each vehicle type k is calculated from the difference between the
average operating weight and the tare weight as follows:
PAYLD = MAX [0, (WGT_OPER - WGT_TARE )] …(3.15)
3.2.6 National life-cycle energy use factors

To incorporate certain life-cycle aspects into the energy use factors, as discussed in Sections
2.1 and 2.3, the following relationships should be used:
ENFUEL k + ENOIL k + ENTYRE k + (PNPk * ENPARTk )

EGLICYk = 
( ) (
+ PNVk * ENVPk + PNFf * ENFUEL k * FPf ) 

…(3.16)
where:
ENLICYk annual average national life cycle energy use factor for vehicle type k (MJ/km)
PNPk proportion of parts for vehicle type k produced within the country (as a
fraction)
PNVk proportion of vehicle type k produced within the country (as a fraction)
PNFf proportion of fuel type f produced within the country (as a fraction)
The annual average national energy use per passenger-kilometre is calculated as:
ENLICYk
ENPAXKM k = …(3.17)
PAX k
where:
ENPAXKMk annual average national energy use per passenger-km for vehicle k
(MJ/passenger-km)
PAXk average number of passengers per vehicle type k

Version 1.0
The annual average national energy per tonne-km is calculated as:
ENLICYk
ENGDSKM k = …(3.18)
PAYLD k
where:
ENGDSKMk annual average national energy use per tonne-km for vehicle type k
(MJ/tonne-km)
PAYLDk average payload per vehicle type k (tonnes)
3.3 Energy use factors for non-motorised transport

Energy use factors EGLICYk, ENLICYk for each Non-Motorised Transport (NMT) mode k
can be obtained as follows:
The calculation of energy used by NMT is shown in Chapter E3.
EGLICYk = ENUSDk * 10 6 …(3.19)
ENLICYk = EGLICYk …(3.20)
where:
EGLICYk annual average global life cycle energy use factor for NMT type k (MJ/km)
ENUSDk average energy consumption for NMT type k (Joules/km)
ENLICYk annual average national life cycle energy use factor for NMT type k (MJ/km)
3.4 Total energy use
3.4.1 Total global energy use

The annual global energy use for each vehicle type k is calculated by multiplying the average
energy use factor EGLICYk by the total kilometres travelled by the vehicle. Thus:
EGLOB k = EGLICYk * VKM k …(3.21)
where:
EGLOBk annual global energy use for vehicle type k (MJ)

VKMk annual vehicle kilometres operated by vehicle type k (km)
The annual total global energy use is then the sum of the energy use for each vehicle type k
(of MT and NMT, for k = 1, 2, …, K) plus the energy used for road works performed on the
particular section(s) in that analysis year:

Version 1.0
K
EGTOT = ∑ EGLOB
k =1
k + ENROAD …(3.22)
where:
EGTOT annual total global energy use (MJ)

ENROAD energy used for road works performed in the given analysis year (MJ)
The energy used for road construction and maintenance in the given analysis year is calculated
from the following equation:
W
ENROAD = ∑ QTY
w =1
w * WEFw …(3.23)
where:
QTYw amount of works type w

WEFw energy used for a unit amount of works type w (MJ/unit). (Default data is not
yet available)
The total global energy used over the analysis period for each investment option is given by
the expression:
Y
GLOENGY = ∑ EGTOT
y =1
y …(3.24)
where:
GLOENGY total global energy use over the analysis period (MJ)
EGTOTy total global energy use in analysis years y (y= 1, 2, . . Y) (MJ)
3.4.2 Total national energy consumption

Energy use within a country is associated with vehicle use (both MT and NMT) together with
the energy use associated with any fuel, oil, vehicle and parts production that occurs within
the country.
The annual national energy use for each vehicle type k is calculated by multiplying the
average energy use factor ENLICYk by the total kilometres travelled by the vehicle. Thus:
ENATk = EGLICYk * VKM k …(3.25)
where:
ENATk annual national energy use for vehicle type k (MJ)

VKMk annual vehicle kilometres operated by vehicle type k (km)

Version 1.0
The annual total national energy use is then the sum of the energy use for each vehicle type k
(of both MT and NMT, for k = 1, 2, …, K) plus the energy used for road works on the
section(s) involved in that analysis year:
K
ENTOT = ∑ ENAT
k =1
k + ENROAD …(3.26)
where:
ENTOT annual total national energy use (MJ)

ENROAD energy used for road works performed in the given analysis year (MJ)
The total national energy used over the analysis period for each investment option is given by
the expression:
Y
NATENGY = ∑ ENTOT
y =1
y …(3.27)
where:
NATENGY total national energy use over the analysis period (MJ)
ENTOTy total national energy use in analysis years y (y= 1, 2, . . Y) (MJ)
3.4.3 Total renewable and non-renewable energy consumption

Assuming that renewable and non-renewable energy use can be split between energy use by
Non-Motorised Transport and Motorised Transport respectively, then total renewable and
non-renewable energy consumption is simply the total energy use for NMT modes and MT
vehicles, respectively.
Thus, annual renewable energy is calculated as:
ERNWi = ∑ EYi
k∈NMT
k …(3.28)
where:
EYik annual energy use for NMT vehicle type k (MJ) (that is, EGLOBk or ENATk)
i global (g) or national (n)
The total renewable energy (RNWTEi) used over the analysis period for each investment
option is obtained by summing ERNWi over the years.
The annual non-renewable energy is calculated as:
ENONRWi = ∑ EYi
k∈MT
k …(3.29)

Version 1.0
where:
EYik annual energy use for MT vehicle type k (MJ)

i global (g) or national (n)
The total non-renewable energy (NORNTEi) used over the analysis period for each
investment option is obtained by summing ENONRWi over the years.
3.5 Comparison of investment options

The true benefit of the assessment outlined above is in comparing results from before a
transport measure or policy has been implemented (base case option n) with results after
implementation (option m), so that the impact of the policy or measure can be seen.
The basic indicator of the performance of a measure is simply the difference between the base
case results and the alternative case results:
∆ENERGYM (m-n) = ENERGYM n - ENERGYM m …(3.30)
Indicators are computed and implemented in the reports for each comparison of investment
options m and n. The indicators are changes in the:
! Annual average global and national life-cycle energy use factors
∆EGLICYk(m-n) and ∆ENLICYk(m-n), respectively, for vehicle type k (MJ/km).

! Annual global and national energy use
∆EGLOBk(m-n) and ∆ENATk(m-n), respectively, for vehicle type k (MJ).

! Annual average global and national energy use per passenger-km
∆EGPAXKMk(m-n) and ∆ENPAXKMk(m-n), respectively, for vehicle type k (MJ/passenger-

km).
! Annual average global and national energy use per tonne-km
∆EGGDSKMk(m-n) and ∆ENGDSKMk(m-n), respectively, for vehicle type k (MJ/tonne-km).

! Annual total global and national energy use
∆EGTOT(m-n) and ∆ENTOT(m-n), respectively (MJ).

! Total global and national energy use over the analysis period
∆GLOENGY(m-n) and ∆NATENGY(m-n), respectively (MJ).

! Annual total global and national renewable energy use
∆ERNWg(m-n) and ∆ERNWn(m-n), respectively (MJ).

! Annual total global and national non-renewable energy use
∆ENONRNWg(m-n) and ∆ENONRNWn(m-n), respectively (MJ).

! Total global and national renewable energy use over the analysis period
∆RNWTEg(m-n) and ∆RNWTEn(m-n), respectively (MJ).

Version 1.0
! Total global and national non-renewable energy use over the analysis period
∆NORNTEg(m-n) and ∆NORNTEn(m-n), respectively (MJ).

Version 1.0
4 References
Dennis R., (1995)
IT Transport, Personal Communication
Department of Trade and Industry, (1996)
Digest of United Kingdom Energy Statistics, 1996
HMSO, London, UK
Eggleston H.S., Gaudioso D., Gorissen N., Jourmard R., Rijkeboer R.C., Samaras Z., and
Zierock K-H., (1993)
CORINAIR working group on emissions factors for calculating 1990 emissions from
road traffic - Volume 1: Methodology and emissions factors, B4-3045 (91) 10PH
Commission of the European Communities, Brussels, Belgium
ETSU, (1995)
Life-Cycle Analysis of Motor Fuel Emissions - Final report to COST 319 Sub-group
A4.C. ETSU Ref: RYCA/18691001/Issue 1
ETSU, (1996)
Alternative Road Transport Fuels - A Preliminary Life-Cycle Study for the UK
HMSO, London, March 1996
Hughes P.S., (1992)
A Strategy for Reducing Emissions of Greenhouse Gases from Personal Travel in Britain
PhD Thesis Open University, 1992
Lewis C.A., (1995)
Energy Use in Bicycle and Animal-Drawn Transport in Developing Countries
ETSU Working Paper 18400304, 1995
Maibach M., Seiler B., and Seiler P.D., (1995)
Okoinventar Transporte
Verlag Infras, Zurich
Replogle M., (1992)
Non-Motorised Vehicles in Asian Cities
World Bank Technical Paper 162
Washington D.C., 1992
Starkey P., (1993)
Animal-Powered Transport in Africa
Appropriate Technology, Vol.20 No.1, June 1993 pp 9-10
United Nations, (1993)
Energy Efficiency in Transportation
Alternatives for the Future, 1993

Version 1.0

Version 1.0
F3 Vehicle Emissions
1 Introduction
This chapter describes the implementation of Vehicle Emissions analysis in HDM-4 (see
Figure F3.1). The objective of modelling vehicle emissions is to assess the effects, in terms of
pollutant quantities, of changes in road characteristics, traffic congestion, and vehicle
technology.
Social
Socialand
and
Environmental
Effects
Overview Energy
Emissions
Analysis
Analysis
Chapter
ChapterF1
F1 Chapter
ChapterF3
F3
Chapter
ChapterF2
F2
The emissions model is based on the model proposed by Hammerstrom (1995). Originally a
variety of units were used. The relationships and model coefficients have been adjusted so that
all emissions predictions are in terms of grammes per vehicle-kilometre, Bennett (1996).
The model predicts the different components of vehicle exhaust emissions as a function of
fuel consumption and speed. Fuel consumption is a function of vehicle speed, which in turn
depends on road characteristics and the characteristics of the vehicle itself. Thus, it is possible
to analyse the change in emissions effects as a result of implementing different road
maintenance and improvement options, or when there are major changes to the vehicle fleet
using the road network (for example, due to improved vehicle technology).
In this release, the effects of exhaust emissions are not costed for inclusion in economic
analysis of road investments, only the net differences in the quantities of pollutants are
assessed for each pair of investment options.

Version 1.0
PART F SOCIAL AND ENVIRONMENTAL EFFECTS F3 VEHICLE EMISSIONS
2 Emissions model
2.1 Types of pollutants

The following seven different components of exhaust emissions are considered:
1 Hydrocarbon (HC)
2 Carbon monoxide (CO)
3 Nitrous oxide (NOX)
4 Sulphur dioxide (SO2)
5 Carbon dioxide (CO2)
6 Particulates (Par)
7 Lead (Pb)
2.2 The relationships

The quantities of the different emission components are predicted using the relationships,
together with default parameter values (see Table F3.1, Table F3.2 and Table F3.3) for the 16
standard motorised vehicle types, given below:
1 Hydrocarbon
3.6 * Kehc0 * (a0 + a1 * Kehc1 * IFC ) * (1 + 0.5 * a2 * LIFE ) * 10 3

E_HC = ...(2.1)
SPEED
where:
E_HC hydrocarbon emissions (g/veh-km)

IFC instantaneous fuel consumption (ml/s)
LIFE vehicle service life (years)
SPEED vehicle speed (km/h)
a0 to a2 model parameters
Kehc0 calibration factor (default = 1.0)
Kehc1 calibration factor (default = 1.0)
The default values of the model parameters are given in Table F3.1.

Version 1.0
Table F3.1 Model parameters for HC and CO emissions
Veh Vehicle type HC CO

No
a0 a1 a2 a0 a1 a2
-2 -2 -2 -2 -2 -2
(x 10 ) (x 10 ) (x 10 ) (x 10 ) (x 10 ) (x 10 )
1 Motorcycle 0.000 21.0000 7 0.000 66.500 6
2 Small car -0.866 2.4400 7 -3.020 18.500 6
3 Medium car -0.866 2.4400 7 -3.020 18.500 6
4 Large car -0.866 2.4400 7 -3.020 18.500 6
5 Light delivery vehicle -3.510 5.6000 7 -21.400 36.400 6
6 Light goods vehicle -3.510 5.6000 7 -21.400 36.400 6
7 Four wheel drive -3.510 5.6000 0 -21.400 36.400 0
8 Light truck -3.510 5.6000 0 -21.400 36.400 0
9 Medium truck 2.400 -0.0439 0 2.020 0.949 0
10 Heavy truck 2.400 -0.0439 0 2.020 0.949 0
11 Articulated truck 2.810 0.0446 0 -0.715 0.107 0
12 Mini-bus -0.866 2.4400 7 -3.020 18.500 6
13 Light bus -3.510 5.6000 0 -21.400 36.400 0
14 Medium bus 2.400 -0.0439 0 2.020 0.949 0
15 Heavy bus 2.400 -0.0439 0 2.020 0.949 0
16 Coach 2.400 -0.0439 0 2.020 0.949 0
Source: Hammerstrom (1995)
2 Carbon monoxide
3.6 * Kec0 * (a0 + a1 * Kec1 * IFC ) * (1 + 0.5 * a2 * LIFE ) * 10 3

E_CO = ...(2.2)
SPEED
where:
E_CO carbon monoxide emissions (g/veh-km)

Kec0 calibration factor (default = 1.0)
Kec1 calibration factor (default = 1.0)
All other variables are as defined previously.


Version 1.0
3 Nitrous oxide
3.6 * Kenox0 * (a0 + a1 * Kenox1 * IFC ) * (1 + 0.5 * a2 * LIFE ) * 10 3

E_NOX =
SPEED
...(2.3)
where:
E_NOX nitrous oxide emissions (g/veh-km)

Kenox0 calibration factor (default = 1.0)
Kenox1 calibration factor (default = 1.0)

The original model had the emissions increasing with increasing vehicle age. For HDM-
4, it is assumed that the emissions at 0.5*LIFE (that is, halfway through the vehicle’s
life) are suitable for representing the emissions (Bennett, 1996). The model parameter a2
(given for HC, CO and NOX) represents the degradation of a vehicle equipped with a
catalytic converter over time.

Version 1.0
Table F3.2 Model parameters for NOX and SO2 emissions
Veh Vehicle type NOX SO2

No
a0 a1 a2 a0 a1
-2 -2 -2 -3
(x 10 ) (x 10 ) (x 10 ) (x 10 )
1 Motorcycle 0.00 0.25 2 0.012 15.0
2 Small car -3.92 4.92 2 0.012 15.0
3 Medium car -3.92 4.92 2 0.012 15.0
4 Large car -3.92 4.92 2 0.012 15.0
5 Light delivery vehicle -2.93 6.01 2 0.012 15.0
6 Light goods vehicle -2.93 6.01 2 0.012 15.0
7 Four wheel drive -2.93 6.01 0 0.031 16.6
8 Light truck -2.93 6.01 0 0.031 16.6
9 Medium truck 1.39 2.90 0 0.031 16.6
10 Heavy truck 1.39 2.90 0 0.031 16.6
11 Articulated truck 13.70 2.94 0 0.031 16.6
12 Mini-bus -3.92 4.92 2 0.012 15.0
13 Light bus -2.93 6.01 0 0.031 16.6
14 Medium bus 1.39 2.90 0 0.031 16.6
15 Heavy bus 1.39 2.90 0 0.031 16.6
16 Coach 1.39 2.90 0 0.031 16.6
4 Sulphur dioxide
3.6 * Keso0 * a0 * a1 * IFC * 10 3

E_SO2 = ...(2.4)
SPEED
where:
E_SO2 sulphur dioxide emissions (g/veh-km)

a0, a1 model parameters
Keso0 calibration factor (default = 1.0)

The default values of the model parameters are given in Table F3.2. The model parameter
a0 in the E_SO2 equation is the percentage content, by weight, of sulphur in fuels. For
petrol the default is 0.012 %.

Version 1.0
5 Carbon dioxide
3.6 * Keco0 * a0 * IFC * 10 3

E_CO2 = ...(2.5)
SPEED
where:
E_CO2 carbon dioxide emissions (g/veh-km)

a0 model parameter
Keco0 calibration factor (default = 1.0)

The default value of the model parameter is given in Table F3.3.
Table F3.3 Model parameters for carbon dioxide, particulates and lead
emissions
Veh Vehicle type CO2 Particulates Pb

No
a0 a0 a1 a0 a1
-4 -4
(x 10 ) (x 10 )
1 Motorcycle 2.36 0.000 0.000 0.15 7.5
2 Small car 2.36 0.455 0.436 0.15 7.5
3 Medium car 2.36 0.455 0.436 0.15 7.5
4 Large car 2.36 0.455 0.436 0.15 7.5
5 Light delivery vehicle 2.36 -1.440 2.550 0.15 7.5
6 Light goods vehicle 2.36 -1.440 2.550 0.15 7.5
7 Four wheel drive 2.61 -1.440 2.550 0.00 0.0
8 Light truck 2.61 -1.440 2.550 0.00 0.0
9 Medium truck 2.61 -11.900 18.400 0.00 0.0
10 Heavy truck 2.61 -11.900 18.400 0.00 0.0
11 Articulated truck 2.61 60.300 5.330 0.00 0.0
12 Mini-bus 2.36 0.455 0.436 0.15 7.5
13 Light bus 2.61 -1.440 2.550 0.00 0.0
14 Medium bus 2.61 -11.900 18.400 0.00 0.0
15 Heavy bus 2.61 -11.900 18.400 0.00 0.0
16 Coach 2.61 -11.900 18.400 0.00 0.0
6 Particulates
3.6 * Kepar0 * (a0 + a1 * Kepar1 * IFC ) * 10 3

E_PAR = ...(2.6)
SPEED

Version 1.0
where:
E_PAR particulates emissions (g/veh-km)

Kepar0 calibration factor (default = 1.0)
Kepar1 calibration factor (default = 1.0)

7 Lead emission
3.6 * Kepb0 * a0 * a1 * IFC * 10 3

E_PB = ...(2.7)
SPEED
where:
E_PB lead emissions (g/veh-km)

Kepb0 calibration factor (default = 1.0)

The default values of the model parameters are given in Table F3.3. The parameter a0 in
the E_PB equation is the percentage content, by weight, of lead in fuels and has the
default value of 0.15 % for petrol.

Version 1.0
3 Modelling logic
For each section (investment) option and for each analysis year, the quantities of each
component of exhaust emissions are computed separately for each vehicle type k and for each
traffic flow period p. The annual total quantities of emissions (by component) are obtained by
summing over all the vehicle types.
3.1 Primary data

The following primary data is required for modelling vehicle emissions:
! Traffic volume on the road section
The annual traffic volume during each flow period (vehicles per year).
! Length of road section
! Vehicle speeds
Calculated within the RUE module.

! Fuel consumption
The instantaneous fuel consumption, for each vehicle type, in each traffic flow period is
calculated within the RUE module.
! Vehicle service life and model parameters
Defined with other Vehicle Fleet data.
3.2 Emissions quantities

For each section option, the quantities of emissions for each vehicle type k and for each traffic
flow period p are calculated using the following expression:
EYRi kp = Tpk * L j * EMi kp * 10 -9 ...(3.1)
where:
EYRikp the annual quantity of emission component i from vehicle type k in traffic flow
period p (tonnes)
Tpk the annual traffic volume of vehicle type k in traffic flow period p (vehicles per
year)
Lj length of road section under investment option j (km)
EMikp the average quantity of emission component i (g/1000 veh-km), from vehicle
type k during traffic flow period p
The value of the average quantity of emission (EMikp) is obtained from the following
expression:
EMi kp = 500 (Ei kpu + Ei kpd ) ...(3.2)
where:

Version 1.0
Eikpu the quantity of emission component i (g/veh-km) for the uphill trip on the
section. It is calculated using the respective equations for each emission
component given in Section 2.2 (see Equations 2.1 above to 2.7 above), using
the following parameters (also see Part E):
IFC taken as IFCkpu - the instantaneous fuel consumption (ml/s) for the
uphill road segment
SPEED taken as SUkp (km/h)
Eikpd the quantity of emission component i (g/veh-km) for the downhill trip on the
section. It is calculated using the respective equations for each emission
component given in Section 2.2 (see Equations 2.1 above to 2.7 above), using
the following parameters:
IFC taken as IFCkpd - the instantaneous fuel consumption (ml/s) for the
downhill road segment
SPEED taken as SDkp (km/h)
For the analysis of one-way, the values of EMikp are obtained as follows:
! One-way uphill
EMi kp = Ei kpu * 10 3 ...(3.3)
! One-way downhill
EMi kp = Ei kpd * 10 3 ...(3.4)
The annual average quantities of vehicle emissions (by component i) per 1000 vehicle-
kilometres is given by the following expression:
n
∑ HRYR
p =1
p * HVp * EMi kp
EAVi k = n
...(3.5)
∑ HRYR
p =1
p * HVp
where:
EAVik annual average quantity of emission component i by vehicle type k (g/1000

veh-km)
HRYRp the number of hours in traffic flow period p (p = 1, . . ., n)
The annual quantities of emissions (by component i) for each vehicle type k using the road
section under investment option j is calculated from the following expression:
n
EYRi jk = ∑ EYRi
p =1
kp ...(3.6)

Version 1.0
where:
EYRijk the annual quantity of emissions of component i by vehicle type k for section
option j (tonnes)
The total annual quantities of emissions (by component i) for all vehicles using the road
section are calculated from the following expression:
n
EYRi j = ∑∑ EYRi
k p =1
kp ...(3.7)
where:
EYRij the annual quantity of emissions of component i (tonnes), for section option j

Version 1.0
4 Comparisons of investment options

In this release, the predicted quantities of exhaust emissions are not costed for inclusion in
economic analysis. The comparison of each pair of investment options is based on the changes
in the annual net difference in the predicted quantities of emissions (by component). Thus, for
each pair of investment options m and the base case n the annual net difference in the
predicted quantities of emissions of component i is calculated as follows:
∆EYRi (m-n ) = EYRi n - EYRi m ...(4.1)
where:
∆EYRi(m -n) the annual net difference in the quantity of emissions component i
The standard reports for emissions analysis include:

! Quantities of vehicle emissions (g/1000 veh-km)
By component i and by vehicle type k for each traffic flow period p (EMikp). These
quantities will be reported for each section option j.
! Annual average quantities of vehicle emissions (g/1000 veh-km)
By component i and by vehicle type k (EAVijk). These quantities will be reported for
each section option j.
! Annual quantities of vehicle emissions (tonnes)
By component i and by vehicle type k (EYRijk). These quantities will be reported for each
section option j.
! Annual total quantities of vehicle emissions (tonnes)
By component i (EYRij) for each section option j.

! Annual net quantities of vehicle emissions
By component i (∆EYRi(m-n)) for each pair of investment options m and n being

compared.

Version 1.0
5 References
ISOHDM Publications, (1994 - 1996)
Hammerstrom U., (1995)
Proposal for a Vehicle Exhaust Model in HDM-4
ISOHDM Supplementary Technical Relationships Study Draft Report
Swedish National Road Administration
Road and Traffic Management Division
Borlange, Sweden
HDM-4 Emissions Model - Draft Specifications
NDLI, (1995)
Asian Development Bank RETA 5549
N. D. Lea International, Vancouver

Version 1.0
Contents

Descriptions: Part G
G1 Economic Analysis
1 Introduction G1-1
2 Background G1-2
2.1 Economic analysis G1-2
2.2 Optimisation G1-2
2.3 Classification of benefits and costs G1-3
3 Benefits and costs considered in HDM-4 G1-4
3.1 Summary of benefits and costs G1-4
3.2 Costs incurred by the road administration G1-4
3.3 Road user costs G1-4
3.4 Environmental effects G1-5
3.5 Other benefits and costs G1-5
3.6 Unit costs G1-5
4 Outline methodology G1-6
4.1 Basic unit of analysis G1-6
4.2 Life cycle analysis G1-6
4.3 Models G1-8
4.4 Analysis sequence G1-8
5 Economic analysis G1-14
5.1 Comparison of investment options G1-14
5.2 Determination of costs and benefits G1-14
5.3 Economic decision criteria G1-19
5.4 Comparison of environmental effects G1-21
5.5 Diverted traffic G1-21
6 Optimisation G1-22
6.1 Total enumeration G1-22
6.2 Incremental benefit/cost ranking G1-24
7 References G1-27

Version 1.0
Part G Road Map
Analytical
AnalyticalFramework
Framework
and
andModel
Model
Descriptions
Descriptions
Introduction
Introduction
Part
PartAA
Traffic
Traffic
Part
PartBB
RD
RDModel WE
WEModel RUE
Model SEE
SEEModel
Model
Part
PartCC Part
PartDD Part
PartEE Part
PartFF
Economic
EconomicAnalysis
Analysis
PartGG
Nomenclature
Nomenclature
PartHH

Glossary
Part
PartI I
Figure G Analytical Framework and Model Descriptions Road Map

Version 1.0
PART G ECONOMIC ANALYSIS
G1 Economic Analysis
1 Introduction
HDM-4 caters for three applications levels commonly used in decision making within the road
sub-sector. The different applications, which are described in more detail in the Applications
Guide, are:
1 Strategic planning
For estimating medium and long-term budget requirements for the development and
preservation of a road network under various budgetary and economic scenarios.
For preparing single or multi-year work programmes under budget constraints, in which
those sections of the network likely to require maintenance, improvement, or new
construction, are identified in a tactical planning exercise.
3 Project analysis
For estimating the economic or engineering viability of different road investment projects
and associated environmental effects. Typical projects include the maintenance and
rehabilitation of existing roads, widening or geometric improvement schemes, pavement
upgrading and new road construction.
For all the three applications, the underlying operation of HDM-4 is based on the concept of
life cycle analysis under a user-specified scenario of circumstances. This involves the analysis
of pavement performance, road works effects and costs, together with estimates of road user
costs and environmental effects, and economic comparisons of different project alternatives.
This chapter describes how HDM-4 is used to determine the benefits and costs associated with
a road investment, and how these are applied in economic analysis and optimisation
procedures to find the best use of available resources.
Analytical Framework and Model Descriptions G1-1

Version 1.0
PART G ECONOMIC ANALYSIS G1 ECONOMIC ANALYSIS
2 Background
2.1 Economic analysis

Economic analysis of the time stream of costs and benefits is used to compare the economic
viability of different alternatives, and to provide the criteria needed for economic decision
making. Decisions can be made about which option to implement, and when is the most
favourable time for implementation. Economic analysis can also be used to investigate the
technical standards and strategies to be followed by a particular investment decision.
Economic analyses involves the following tasks:
1 Identification of the problem to be solved and the formulation of alternatives.
2 Identification and quantification of the life-cycle costs to be incurred and benefits to be
realised.
3 Modelling future impacts of the proposed alternatives on road performance and traffic
flow.
4 Economic comparison of the different alternatives, involving:
(a) discounting the annual costs and benefits streams to a chosen base year
(b) Comparing the time stream of costs for each pair of alternatives
(c) Calculating the economic indicators such as the net present value, internal rate of
return, benefit-cost ratio, and first year benefits
A project analysis usually involves a small number of road links or sections and the results of
economic analysis would provide adequate information for decision making, since a budget
would normally already have been approved for these activities.
2.2 Optimisation
The purpose of the Strategy and Programme applications is to calculate the economic benefits
derived from maintenance or improvement options, and to select the set of investments to be
made on a number of road sections within a network which will optimise an objective
function.
Programme analysis is concerned with short to medium term planning and preparation where
budget levels are known with reasonable certainty and the objective is to select a set of road
sections and road works within the budget constraint.
Strategy analysis involves the analysis of an entire road network (or sub-network). The
objective is either to determine which types of road works should be applied in order to
maximise economic benefits, or it may also be applied to determine the budget required for a
given long term target road network condition. Thus, the problem can be posed as one of
searching for the combination of investment alternatives that optimises the objective function
under a budget constraint or a road network condition constraint. Note that the set of
investment options to be optimised is user-defined and is not the set of all possible options for
the particular network; hence the problem is not true optimisation since all possible solutions
are not normally considered. Note also that the investment options on any one road section are
mutually exclusive.
The three alternative objective functions provided for the Strategy and Programme
applications are:

Version 1.0
1 Maximisation of economic benefits (that is, NPV)
This option is used when the problem can be defined as the selection of a combination of
investment options applied on several road sections which maximises the NPV (net
present value) for the whole network subject to the sum of the financial costs being less
2 Maximisation of the improvement in network condition
The roughness reduction on each road section multiplied by the section length
(∆IRI*Length) is used instead of NPV. Consequently, the arithmetic procedure is similar
to that used for maximisation of economic benefits.
3 Minimise costs of road works to achieve a given target road network condition
This option is used mainly in the Strategy analysis application. The target road condition
defined in terms of the long-term average roughness (IRI) over the whole analysis period
must be specified for each road section. The optimisation procedure is then reduced to a
simple selection of the road work options for which the average IRI (over the analysis
period) is equal to or just below the target IRI and has the lowest total financial cost.
2.3 Classification of benefits and costs

Costs and benefits due to road investments may be classified into the following three broad
categories:
1 Benefits and costs expressed in monetary terms
For example, vehicle operating costs, savings in travel time, accident costs.
2 Quantified benefits and costs not expressed in monetary terms
For example, road safety, pollution from vehicle emissions and traffic noise.
3 Non-quantified benefits and costs
For example, better social welfare, ecological impacts.

An economic analysis considers directly only benefits and costs expressed in monetary terms.
Other costs and benefits may also need to be considered, and this is sometimes done within
the framework of a multi-criteria analysis.

Version 1.0
3 Benefits and costs considered in HDM-4
3.1 Summary of benefits and costs

HDM-4 considers quantified benefits and costs that can be expressed in monetary terms, and
has some scope for considering those that cannot be expressed in this way. The benefits and
costs considered are:
! Costs incurred by the road administration (see Section 3.2)
! Road user costs (see Section 3.3)
! Environmental effects (see Section 3.4)
! Other benefits and costs (see Section 3.5)
3.2 Costs incurred by the road administration

These costs are also referred to as road agency costs and include the following:
! Road development
! Pavement maintenance
! Road-side or off-carriageway activities
The cost of works is derived from the product of the physical quantities involved in the
activity and the unit cost. These are determined for each road section and investment option,
and for each year of the analysis period. The resulting costs are assigned to budget categories
that are user-definable. The following default categories are used in HDM-4:
! Capital (or periodic)
! Recurrent (or routine)
! Special
Budget constraints can be applied separately to each category when required by the economic
analysis and optimisation.
3.3 Road user costs

The following components of Road User Costs are modelled:
! Motorised vehicle operating costs
These costs include:

❏ Fuel and lubricant consumption
❏ Tyre and parts consumption
❏ Labour
❏ Capital
❏ Crew

Version 1.0
❏ Overheads
! Travel time costs
These costs include passenger travel time costs, and cargo holding time costs.
! Non-motorised transport (NMT)
These costs include time and operating costs.

! Accident costs
These costs are evaluated both in monetary and non-monetary terms, and separated into
several different types (for example, fatal, injury, and damage only, and all accidents). Note
users are allowed the flexibility to include or exclude accident costs from an economic
analysis.
3.4 Environmental effects

The following environmental impacts are determined:
! Vehicle emissions
! Energy use
! Traffic noise (not included in this release)
3.5 Other benefits and costs

The user can specify those benefits and costs that are not modelled for each year of the
analysis period. These benefits and costs are discounted and added to those that are calculated
internally. These other benefits and costs are sometimes termed exogenous.
3.6 Unit costs

Unit costs are applied to the calculated physical and operational quantities to produce the cost
estimates used in investment decisions and budget preparation. Unit costs should be expressed
in economic terms when economic analysis is being undertaken, and in financial terms for
financial analysis. Financial unit costs are the market prices of resources. Economic unit costs
are the real value or opportunity costs of resources, and they are found by removing
distortions such as taxes, subsidies and other miscellaneous costs from the market prices.
Unit costs are required for the following:
! Road development, maintenance and road-side activities
These unit costs are specified by the user (see Part D).
! Road user costs
These unit costs include vehicle resources, travel-time values, and road accident
resources (see Part E).
In most cases, unit costs are specified in units-per-quantity. However, some costs are specified
as a proportion of other costs, or as a lump sum.
In addition to calculating economic costs, financial costs are also computed if the user gives
appropriate inputs (for example, unit costs in financial terms).

Version 1.0
4 Outline methodology
4.1 Basic unit of analysis

The basic unit of analysis in HDM-4 is the homogeneous road section. Several investment
options can be assigned to a road section for analysis. One or more vehicle types that use the
road must also be defined together with the traffic volume specified in terms of the annual
average daily traffic (AADT).
4.2 Life cycle analysis

The underlying operation of HDM-4 is common for the project, programme or strategy
applications. In each case, HDM-4 predicts the life cycle pavement performance and the
resulting user costs under specified maintenance and/or road improvement scenarios. The
broad concept of the life cycle analysis is illustrated in Figure G1.1. The agency and user costs
are determined by first predicting physical quantities of resource consumption and then
multiplying these by the corresponding unit costs.
Two or more options comprising different road maintenance and/or improvement works
should be specified for each candidate road section with one option designated as the do
minimum or base case (usually representing minimal routine maintenance). The benefits
derived from implementation of other options are calculated over a specified analysis period
by comparing the predicted economic cost streams in each year against that for the respective
year of the base case option. The discounted total economic cost difference is defined as the
net present value (NPV). The average life cycle riding quality measured in terms of the
international roughness index (IRI) is also calculated for each option.

Version 1.0

geometry, pavement
characteristics, unit costs



Road geometry and surface Social and Levels of emissions and energy
texture, vehicle characteristics Environmental Effects used, and number of accidents


by section
Figure G1.1 Life cycle analysis procedure in HDM-4

Version 1.0
4.3 Models
Total life-cycle conditions and costs of sections or road networks can be simulated over a
user-defined period into the future. The inter-dependence between the costs incurred by the
road administration and the road user is recognised, and models are used to predict cost
streams under the various headings.
The models incorporated in HDM-4 contain technical relationships for the following
purposes:
1 Calculation of traffic volumes and flows, and vehicle loading over the road section.
2 Prediction of road deterioration, and works effects and costs, that are incurred in response
to traffic flows, time and the surrounding environment.
3 Prediction of the costs of road use incurred as road condition and traffic flow change over
time.
4 Prediction of accident rates as a function of the road and traffic characteristics, and the
evaluation of accident costs.
5 Evaluation of vehicle emissions and energy use due to different road investment projects.
6 Economic analysis by comparison of the impacts or effects of different road investment
project alternatives.
4.4 Analysis sequence

The overall logic sequence for economic analysis and optimisation is illustrated in Figure
G1.2a and represented below by pseudo codes. This shows the following:
1 The outer analysis loop
Enabling economic comparisons to be made for each pair of investment options, using
the effects and costs calculated over the analysis period for each option, and indicates
that generated and diverted traffic levels may vary depending on the investment option
considered.
2 Costs
How annual costs to the road administration and to the road users are calculated for
individual road section options.
3 Optimisation procedures
These are performed after economic benefits of all the section options have been
determined.
The pseudo code that represents the outer analysis loop is given below:

Version 1.0
START
Define input data
Loop for each option
Loop for each section
Loop for each analysis year
Calculate traffic over the road section
Model annual effects and costs (see Figure G1.2b)
Store results for evaluation and reporting phase

End loop
End loop
End loop
Loop for each pair of options to be compared
Loop for each analysis year
Loop for each section
Calculate non-discounted net benefits
Calculate discounted net benefits
Calculate net environmental effects and energy used
End loop
Calculate total non-discounted net benefits over all the sections (see Figure
G1.2c)
Calculate total discounted net benefits over all the sections (see Figure G1.2c)
Calculate total environmental effects and energy used over all the sections (see
Figure G1.2c)
End loop
Calculate economic indicators (NPV, IRR, BCR, and FYB see Section 5.3.1 and
Figure G1.2c)
End Loop
Perform budget optimisation (for strategy and programme analysis)
Output results
END

Version 1.0
Define Input Data
LOOP
For each option
LOOP
For each section
LOOP
For each analysis year
CALCULATE
This year’s traffic
over the road section
See Figure G1.2b

MODEL
Annual effects and costs
STORE
Results for evaluation
and reporting phase
Yes
More
Years?
No
Yes
More
Sections?
No
Yes
More
Options?
No
See Figure G1.2c
Economic Analysis and
Comparisons
Optimisation
procedures
Output Results
Figure G1.2a Overall analysis sequence logic – Part A

Version 1.0
MODEL
Annual effects and costs
CALCULATE
Road Deterioration
[RD Module]
CALCULATE
Road User Effects
(Speeds, VOC, time costs,
NMT, accident costs)
[RUE Module]
CALCULATE
Road Works Effects
[WE Module]
CALCULATE
Social and Environmental Effects
Social
(emissions
and Environmental
and energy balance)
Effects
(emissions andModule]
[SEE energy balance)
ADD
This year’s exogenous
benefits and costs
Return
Figure G1.2b Overall Analysis sequence logic - Part B

Version 1.0
Economic Analysis and

Comparisons
LOOP
For each comparison
LOOP
For each analysis year
SET NPVsr = 0
Total net effects TNEsn (1, . . .,n) = 0
LOOP
For each section
CALCULATE
non-discounted Net Benefits
CALCULATE
Discounted Net Benefits at different
discount rates (r)
CALCULATE
Net effects (n)
Yes
More
Sections?
No
CALCULATE
Total net benefits and effects
Yes
More
Years?
No
CALCULATE Economic indicators
CALCULATE
Total net effects
Yes
More
comparisons?
No
Return
Figure G1.2c Overall Analysis sequence logic - Part C

Version 1.0
The procedure for calculating annual road agency costs and road user effects for individual
section options is illustrated in Figure G1.2b and summarised by the following steps:
1 Calculate road deterioration - in the RD module (see Part C)
2 Calculate road user costs
VOC, travel time costs, NMT time and operating costs, and accident costs - in the RUE
module (see Part E).
3 Calculate quantities and costs for road works - in the WE module (see Part D)
4 Calculate environmental effects
For example, emissions and energy use - in the SEE module (see Part F).
5 Add exogenous benefits and costs
Figure G1.2c illustrates the inner analysis loops for economic analysis and comparison of each
pair of road alternatives.

Version 1.0
5 Economic analysis
5.1 Comparison of investment options

Economic indicators are computed at different user-specified discount rates using the time
streams of benefits or costs resulting from the various comparative pairs of investment
options. The term investment options has been used in this document to refer to both project
options (or alternatives) and section options (or alternatives).
For each pair of investment options to be compared, the net benefits (or costs) of
implementing one option relative to the other is calculated year by year. The various methods
of comparison are described in sub-sections within Sections 5.2 and 5.3. In all cases,
investment option m is compared against option n (that is, option n is the base case).
5.2 Determination of costs and benefits
5.2.1 Costs to the road administration

The cost differences between a pair of investment options, m and n, in a given year, are
 
∆C (m−n)i = 

∑C
s
mis - ∑C
s
nis 

...(5.1)
where:
∆C(m-n)i the difference in road administration cost of investment option m relative to

base option n for budget category i
Cjis the total costs to road administration incurred by investment option j (where j
= n or m) for budget category i, for road section s (see Part D)
The difference in annual costs to road administration is given by the expression:
∆RAC (m −n ) = ∑ ∆C
i
(m −n)i ...(5.2)
where:
∆RAC(m-n) the difference in annual costs to road administration of investment option m

relative to base option n. (The summation is over all the budget categories)
These cost differences provide a relative measure of the increase in costs to the road
administration, of implementing investment option m over base option n.
The difference in the salvage values of works performed under investment options m and n is
a component of the net economic benefits to be included in the last year of the analysis period
(see Section 5.2.4), and is given as:
∆SALVA (m-n) = [SALVA m - SALVA n ] ...(5.3)

Version 1.0
where:
∆SALVA(m-n) the difference in salvage value of implementing investment option m

relative to base option n
SALVAj salvage value of the works performed under investment option j (where
j = n or m) (see Part D)
5.2.2 Savings in road user costs

The annual economic benefits in terms of savings in road user costs are calculated separately
by components and traffic categories as follows:
! Savings in motorised vehicle operating costs
Vehicle operating benefits due to normal and diverted traffic is calculated as follows:
 
∆VCN (m−n) = 

∑ VCN - ∑ VCN
s
ns
s
ms 

...(5.4)
VCNns = ∑ TN
k
nsk * UCnsk ...(5.5)
VCN ms = ∑ TN
k
msk * UC msk ...(5.6)
Vehicle operating benefits due to generated traffic is calculated as follows:
 
∆VCG (m−n) = 

∑∑ {0.5 * [TG
s k
msk + TG nsk ] * [UC nsk − UC msk ]}

...(5.7)
The summations are over all the motorised vehicle types (k = 1, 2, ..., K) specified by the
user, and all road sections (s = 1, 2, ... ., S) being analysed.
The annual saving in vehicle operating costs is given by the expression:
∆VOC (m −n ) = [ ∆VCN (m −n ) + ∆VCG (m −n ) ] ...(5.8)
where:
∆VCN(m-n) vehicle operating benefits due to normal and diverted traffic of investment
option m relative to base option n
VCNjs annual vehicle operating cost due to normal and diverted traffic over the
road section s with investment option j
TNjsk normal and diverted traffic, in number of vehicles per year in both
directions on road s, investment option j, for vehicle type k
UCjsk annual average operating cost per vehicle-trip over road section s, for
vehicle type k under investment option j (where j = n or m)
VCGjs annual vehicle operating cost due to generated traffic over road section s
with investment option j

Version 1.0
∆VCG(m-n) vehicle operating benefits due to generated traffic of investment option m

TGjsk generated traffic, in number of vehicles per year in both directions on road
s, for vehicle type k, due to investment option j
∆VOC(m-n) savings in vehicle operating costs due to the total traffic of investment
! Savings in travel time costs – motorised vehicles
Vehicle travel time benefits due to normal and diverted traffic are calculated as follows:
 
∆TCN (m−n) = 

∑ TCN - ∑ TCN
s
ns
s
ms 

...(5.9)
TCN ns = ∑ TN
k
nsk * UTnsk ...(5.10)
TCN ms = ∑ TN
k
msk * UTmsk ...(5.11)
Vehicle travel time benefits due to generated traffic are calculated as follows:
 
∆TCG (m−n) = 

∑∑ {0.5 * [TG
s k
msk + TG nsk ] * [UTnsk − UTmsk ]}

...(5.12)
The annual savings in travel time costs are given by the expression:
∆TTC (m −n ) = [∆ TCN (m −n ) + ∆TCG (m −n ) ] ...(5.13)
where:
∆TCN(m-n) travel time benefits due to normal and diverted traffic of investment
TCNjs annual vehicle travel time cost due to normal and diverted traffic over
UTjsk annual average travel time cost per vehicle-trip over the road section s, for
vehicle type k, under investment option j (where j = n or m)
TCGjs annual vehicle travel time cost due to generated traffic over road section s
with investment option j
∆TCG(m-n) travel time benefits due to generated traffic of investment option m
relative to base option n on the given road section in the given year
∆TTC(m-n) savings in travel time costs due to total traffic of investment option m
! Savings in NMT time and operating costs
Non-Motorised Transport (NMT) time and operating benefits due to normal and diverted
traffic are calculated as follows:

Version 1.0
 
∆TOCN(m−n) = 

∑ TOCN - ∑ TOCN
s
ns
s
ms 

...(5.14)
TOCNns = ∑ TN
k
nsk * UTOC nsk ...(5.15)
TOCNms = ∑ TN
k
msk * UTOC msk ...(5.16)
NMT time and operating benefits due to generated traffic are calculated as follows:
 
∆TOCG (m−n) = 

∑∑ [0.5 * (TG
s k
msk + TG nsk ) * (UTOC nsk − UTOC msk )]

...(5.17)
The summations are over all the NMT types (k = 1, 2, ..., K) specified by the user, and all
road sections (s = 1, 2, ... ., S) being analysed.
The annual savings in NMT time and operating costs are given by the expression:
∆NMTOC (m −n ) = [∆ TOCN (m −n ) + ∆TOCG (m −n ) ] ...(5.18)
where:
∆TOCN(m-n) NMT time and operating benefits due to normal and diverted traffic of
investment option m relative to base option n
TOCNjs annual NMT time and operating costs due to normal and diverted
traffic over the road section s with investment option j
TNjsk NMT normal and diverted traffic, in number of vehicles per year in
both directions on road s investment option j, for vehicle type k
UTOCjsk annual average NMT time and operating cost per vehicle-trip over road
section s, for vehicle type k, under investment option j (where j = n or
m)
TOCGjs annual NMT time and operating costs due to generated traffic over
TGjsk NMT generated traffic, in number of vehicles per year in both
directions on road s, for vehicle type k, due to investment option j
∆TOCG(m-n) NMT time and operating benefits due to generated traffic of investment
∆NMTOC(m-n) annual savings in NMT time and operating costs due to total traffic of
investment option m relative to base option n
! Reduction in accident costs
The benefits from reduction in total accident costs are given by the expression:
∆ACC (m-n) = [AC n - AC m ] ...(5.19)

Version 1.0
where:
∆ACC(m-n) the accident reduction benefits due to implementing investment option

m relative to base option n
ACj the total accident costs under investment option j (where j = n or m)
! Road user benefits
The annual savings in road user costs are given by the expression:
∆RUC (m −n ) = [∆VOC (m-n) + ∆TTC (m-n) + ∆NMTOC (m-n) + ∆ACC (m-n) ] ...(5.20)
where:
∆RUC(m-n) the total road user benefits of investment option m

5.2.3 Other benefits and costs

The difference in other (exogenous) benefits and costs, for each pair of investment options m
and n in a given year, are calculated as follows:
∆NEXB = EXB - EXC - EXB + EXC  ...(5.21)

y(m - n)  ym ym yn yn 
where:
∆NEXBy(m-n) the annual net exogenous benefits of investment option m relative to base
option n, in year y
EXBjy exogenous benefits for investment option j, in year y, (where j = n or m)
EXCjy exogenous costs for investment option j, in year y
5.2.4 Annual net economic benefits

For each pair of investment options, the annual net economic benefits of implementing option
m relative to option n is obtained by combining the differences in costs to the road
administration, road user costs, and other benefits and costs, as follows:
NB y (m-n ) = [∆RUC y(m-n) + ∆NEXB y(m-n) - ∆RAC y(m-n) ] ...(5.22)
where:
NBy(m-n) net economic benefit of investment option m relative to base option n in

year y, and the parameters on the right-hand side are as defined earlier, but
with subscript y added to indicate year

Version 1.0
In the last year of the analysis period, the net economic benefits of implementing option m
relative to option n is calculated as:
NB Y (m-n) = [∆RUC Y(m-n) + ∆NEXB Y(m-n) - ∆RAC Y(m-n) + ∆SALVA (m-n) ]
...(5.23)
where:
NBY(m-n) net economic benefit of investment option m relative to base option n in the
last year of the analysis period Y, and the parameters on the right-hand side
are as defined earlier, but with subscript Y added to indicate the last year of
the analysis period
5.2.5 New road sections (or links)

For the analysis of a new road section, the following variables used in the equations given in
Sections 5.2.1 to 5.2.3 are set to zero: Cnis, SALVAn, UCnsk, TNnsk, TGnsk, UTnsk, UTOCnsk,
ACn, EXByn, and EXCyn.
5.3 Economic decision criteria
5.3.1 Indicators determined

The following economic indicators are computed from the time streams of benefits or costs at
the user-specified discount rate:
! Net Present Value - NPV (see Section 5.3.2)
! Internal Rate of Return - IRR (see Section 5.3.3)
! Benefit/Cost Ratio - BCR (see Section 5.3.4)
! First Year Benefits - FYB (see Section 5.3.5)
The determination of these indicators is described in the sections referenced.
5.3.2 Net present value

The Net Present Value (NPV) of investment option m relative to base option n is the sum of
the discounted annual net benefits or costs, calculated from the relationship:
Y
∑ [1+ 0.01 * r]
NB y(m-n)
NPV(m −n) = (y -1)
...(5.24)
y =1
where:
NBy(m-n) net economic benefit of investment option m relative to base option n in

year y
r discount rate (%)
y analysis year (y = 1, 2, ... ., Y)

Version 1.0
The higher the NPV, the greater the benefits from investment option m relative to base option
n. If there are no budget constraints, then the choice between the two alternative investments
should be based on NPV. Larger investments will tend to have larger NPVs.
5.3.3 Internal rate of return

The Internal Rate of Return (IRR) is the discount rate at which NPV is zero. It is calculated by
solving the implicit relationship for r°:
Y
NB y(m−n)
∑ [1 + 0.01 * r°]
y =1
(y -1)
= 0 ...(5.25)
This equation is solved for r° by evaluating the NPV at 5 percent intervals of discount rates
between -95 and +900 percent, and determining the zero(es) of the equation by linear
interpolation of adjacent discount rates with NPV of opposite signs. Depending on the nature
of the net benefit stream, NBy(m-n), it is possible to find one solution, multiple solutions, or
none at all.
The IRR gives no indication of the size of the costs or benefits of an investment; it acts as a
guide to the profitability of the investment - the higher the better. If the computed IRR is
larger than the planning discount rate, then the investment is economically justified.
5.3.4 Benefit cost ratio

The Benefit Cost Ratio (BCR) of investment option m, relative to base option n, is the ratio is
NPV(m-n)
BCR (m-n) = +1
Cm
...(5.26)
where:
BCR(m-n) benefit cost ratio of investment option m relative to base option n
NPV(m-n) discounted total net benefit of investment option m relative to base option n.
This is the Net Present Value at discount rate r
Cm discounted total agency costs of implementing investment option m
If the NPV(m-n) is zero, then (NPV/C)(m-n) is zero. These ratios give an indication of the
profitability of investment option m relative to base option n at a given discount rate. These
measures eliminate the bias of NPV towards larger project options but, like the IRR, they give
no indication of the size of the costs or benefits involved.
5.3.5 First-year benefits

The First-Year Benefits (FYB) is defined as the ratio, in percent, of the net benefit realised in
the first year after construction (or improvement) completion to the increase in total capital cost:
100 * NB y °(m-n)
FYB (m-n) =
∆TCC (m−n)
...(5.27)

Version 1.0
where:
FYB(m-n) first-year benefits of investment option m relative to base option n (%)
NBy°(m-n) net economic benefit of investment option m relative to base option n in

year y°, where:
y° is the year immediately after the last year in which the capital cost
for improvement or construction is incurred in option m
∆TCC(m-n) the difference in total capital cost (non-discounted) of investment option m
FYB gives a rough guide to project timing: if it is greater than the discount rate, then the
project should go ahead; otherwise it should be delayed until it satisfies the criterion.
5.4 Comparison of environmental effects

Where it is not possible to model costs directly, the effects of alternative investments can be
evaluated. This information could be used as a decision tool for screening projects. For
example; investment alternatives can be selected that are more effective in reducing the
number of fatal accidents, or which are more effective in reducing the number of persons
disturbed by a high level of traffic noise. Effects may also be a useful input for multi-criteria
analysis.
The approach to carrying out a comparative study of environmental effects, for a pair of
investment options, is similar to that used for economic analysis (see Section 5.2). The annual
net quantities of vehicle emissions, number of accidents, and levels of traffic noise that are
determined are compared with the benefit of implementing one investment option relative to
the other.
5.5 Diverted traffic

Traffic diversion reduces or increases traffic on the roads that are affected. Therefore, in a
situation where a road works causes traffic to divert significantly to a new or improved road
section, a direct economic comparison of section options is not valid since the normal traffic
flows on the road with and without the works are not identical.
Economic comparisons of investment options involving diverted traffic can only be performed
meaningfully at the project analysis level, if the following conditions are met:
! All the road sections from and to which traffic diverts must be analysed together with the
as a result of carrying out the road works.
! In any given analysis year, the total traffic volume entering the study area equals the total
The analysis of a new road section (or link) in an entirely new location always involves
diverted traffic. The normal traffic in the first year of road opening is diverted traffic from
nearby routes (and from other transport modes, which may complicate matters further). The
economic analysis and comparison involving a new road section, therefore, should always
comply with the conditions described above.

Version 1.0
6 Optimisation
The two methods for budget optimisation provided for road works programming and network
strategic analysis are:
1 Total enumeration
2 Incremental benefit/cost ranking
[Further methods may be added in later versions of the HDM-4 software.]

If the number of roads to be analysed is less than 100, and there are no more than five budget
periods and 16 alternatives per road, total enumeration (see Section 6.1) can be used. This will
be externally done in the EBM-HS model of HDM-III. If the above mentioned constraints are
exceeded, incremental benefit/cost ranking (see Section 6.2) will be used.
6.1 Total enumeration

This is the method used by the EBM-HS model of HDM-III. It requires the user to specify the
following parameters:
! Name of data set
For example, CAPROG97.

! Length of analysis period
For example, 20 years.

! Budget periods
For example, 1, 2, 3, 4-20 years.

! Objective function
Either: maximise NPV or maximise the improvement in roughness.

! Constraints on resources for each budget period
For example, 10, 10, 10, 200.

The analysis period is given in terms of the number of years over which the overall analysis
should be performed, together with the initial calendar year. Budget periods are shorter time
periods for which the budget constraints are given. The objective function defines which
parameter is to be optimised. The default is the maximisation of NPV over the analysis period,
but the user can also choose the maximisation of the improvement in roughness.
The program is run for all the road sections defined with positive economic return, and for all
budget and investment options. The budget requirements from any committed projects will be
deducted from the available budget and the balance is used for optimisation.
The optimisation problem is then defined as an integer programming problem of maximising
the total objective function (TOBJ) for the network (extracts from EBM documentation):
S Ms
Maximise TOBJ[ Xsm] = ∑∑
s=1 m=1
OBJsm Xsm ...(6.1)
where:

Version 1.0
s a road section (s = 1, 2, ... , S)

Ms the number of alternatives for road section s
m an investment alternative on a road section
OBJsm the objective function to be maximised which may be the discounted net
present value of economic benefits, or the average reduction in roughness due
to the investment alternative
sm subscript denoting alternative m for road section s
Xsm the zero-one decision variable:
Xsm 1 if alternative m of investment unit s is chosen
Xsm 0 otherwise
m 1,...,Ms
The above is subject to the following resource constraints:

S Ms
∑∑ R
s =1 m =1
smqt X sm ≤ TR qt , q = 1,...., Q; t = 1,...., T ...(6.2)
where:
Rsmqt Non-discounted amount of resource of type q incurred by the sectoral agency

within a budget period t
TRqt maximum amount of resource type q available for budget period t
Q the total number of resource types
T the total number of budget periods (the duration of t may be one or more years
and need not be equal for different budget periods)
The above is subject to the constraint of mutual exclusivity:
Ms
∑m =1
X sm ≤ 1, s = 1,...., S ...(6.3)
that is, for each road section s, no more than one alternative can be implemented.
If M is the average number of alternatives for the roads, the problem then has SM (= S x M)
zero-one variables, QT (= Q x T) resource constraints and S interdependency constraints. The
parameters that define the problem size are S, M and QT. Depending on the solution method
used, different problem-size parameters determine whether the method is suitable for the
problem in terms of the computational effort needed.
The total enumeration method provides the user with an unconditionally optimal solution. It
computes the total net present values of all feasible programme selections, and chooses the
one with the highest value. The computational effort required for this may be considerable, so
the method is only feasible when the number of alternatives per investment unit is relatively
small.

Version 1.0
Total enumeration is performed externally in the EBM-HS software of HDM-III. The

procedure is as follows:
1 Create an input file for the EBM-HS in the programme analysis. The format of this file is
defined in the HDM-III EBM documentation.
2 The user starts the EBM-HS, imports the file to the EBM-HS, runs it, and exports results
to an output file.
3 The output file is imported to the programme analysis for reporting.
6.2 Incremental benefit/cost ranking

With many applications of HDM-4, a large number of road sections will need to be
prioritised. In these cases, the incremental benefit/cost method is the most appropriate. This
involves searching through investment options on the basis of the incremental NPV/cost ratio
of one alternative compared against another. The incremental NPV/cost is defined as:
  NPV - NPV  
 j i 
E =  ...(6.4)
  cost - cost  
ji
 j i
where:
Eji the incremental NPV/cost ratio

NPVj the net present value of the more expensive alternative j
NPVi the net present value of the cheaper alternative i
costj the economic cost of the more expensive alternative j
costi the economic cost of the cheaper alternative i
In Equation 6.4 above, the incremental NPV/cost can be replaced by the incremental
∆IRI*Length/cost where ∆IRI*Length is the weighted average change in roughness obtained
by comparing the project alternatives using IRI instead of NPV.
The objective of the incremental method is to select road sections successively starting with
the largest NPV/cost ratio (Eji), since this maximises the NPV (net present value) for any
given budget constraint. Where there is more than one investment option on any individual
road section, that with the lowest discounted investment costs is designated the base case
alternative. This method considers all possible options, and compares these incrementally
starting against the base case, by using the incremental algorithm to select the combination
that maximises the selected objective function.
An incremental search technique is used to select the options with successively lower
incremental NPV/cost ratios, ensuring that at any time there is no more than one option per
road section. The process continues until the budget is exhausted for each budget period. The
method is often referred to as the efficiency frontier, which is a line that joins investments
with the highest NPV along the cost axis in a plot of NPV against investment cost (Harral and
Faiz, 1979). In essence, the method seeks out those options that are close to the boundary of
the efficiency frontier. The algorithm is illustrated in Figure G1.3, and is defined in the
following steps:

Version 1.0
1. Determine the pre-defined investment options for pre-selected sections and deduct the
financial costs of these options from the available budget in corresponding years. Exclude
these sections from any further optimisation.
2. Determine possible investment options for the remaining sections. If the life cycle
analysis option is being used, set the user-defined base alternatives as the do minimum
for each road section. For the multi-year forward programme, the do minimum option is
that with the delayed capital works.
3. If the total financial cost of the do minimum investment alternatives on each section is
greater than the available budget for any period, then the investment options or budget
constraints must be redefined.
4. Deduct the financial cost of the do minimum investments from the available budget to
determine the remaining budget for each period. Set the do minimum as the first Base
option for each section.
5. Calculate the incremental NPV/Cost ratio for all remaining section-options compared
against the Base option, and all other option pairs with higher economic cost. For
example, consider the following investment options for a particular section arranged in the
ascending order of discounted total economic costs:
options: A, B, C, D, E
The incremental NPV/Cost ratios for these are given by:
Eba Eca Eda Eea ; Ecb Edb Eeb ; Edc Eec ; Eed
6 Delete incremental NPV/cost ratios that are less than the user specified minimum
incremental value (MIV).
7 List the remaining incremental NPV/cost ratios in decreasing order (with the associated
section-option pair codes) and, within each incremental NPV/cost, in the order of
decreasing economic cost. For example, if Eeb = Edb then Eeb is ranked higher.
8 Select the next incremental NPV/cost ratio from the top of the list. If the lower cost
section-option is not the current Base Option for that section, continue searching until one
is found.
9 If the remaining budget is insufficient in any of the periods for the financial costs of
works required for the section-option selected in Step 8 above, then the selected option
should be rejected, and continue searching by repeating Step 8.
10 If the section-option can fit within the remaining budgets for all periods, deduct the net
increase in financial cost of capital works from all corresponding budget periods. Set the
Base option for this section to be that corresponding to the lower cost option for the
incremental NPV/Cost ratio chosen in Step 8. Providing that the remaining list is not
empty, return to Step 8.
The process described above continues until the budget is exhausted or there are no more
section-options remaining in the list. The resulting list of selected section-alternatives
constitutes the optimal work programme.

Version 1.0
NPV
E
D
B
C A, B, C, D, E : Section-Alternatives
Priority for funds:

1. Eba
2. Edb
3. Eed
A
ECONOMIC COST
(relative to Base Option)
Figure G1.3 Efficiency frontier concept

Version 1.0
7 References
Harral, C.G. and Faiz, A. (1979)
Warwick, 13-16 July 1979. PTRC Education and Research Services, London, UK
PIARC, (1991)
Methods for Selecting Road Investment, Economic and Finance Committee of PIARC,
Paris, France
TRRL Overseas Unit, (1988)
A Guide to Road Project Appraisal. Transport and Road Research Laboratory Overseas
Road Note 5, Crowthorne UK

Version 1.0
Part H Nomenclature
Contents

Descriptions: Part H
H1 Nomenclature

Version 1.0
Part H Road Map
Analytical
AnalyticalFramework
Framework
and
andModel
Model
Descriptions
Descriptions
Introduction
Introduction
Part
PartAA
Traffic
Traffic
Part
PartBB
RD
RDModel WE
WEModel RUE
Model SEE
SEEModel
Model
Part
PartCC Part
PartDD Part
PartEE Part
PartFF
Economic
EconomicAnalysis
Analysis
PartGG
Nomenclature
Nomenclature
PartHH

Glossary
Part
PartI I
Figure H Analytical Framework and Model Descriptions Road Map

Version 1.0
PART H NOMENCLATURE
Part H Nomenclature
H1 Nomenclature
To be included in a subsequent edition of this document.
Analytical Framework and Model Descriptions H1-1

Version 1.0
Part I Glossary
Contents

Descriptions: Part I
I1 Glossary

Version 1.0
Part I Road Map
Analytical
AnalyticalFramework
Framework
and
andModel
Model
Descriptions
Descriptions
Introduction
Introduction
Part
PartAA
Traffic
Traffic
Part
PartBB
RD
RDModel WE
WEModel RUE
Model SEE
SEEModel
Model
Part
PartCC Part
PartDD Part
PartEE Part
PartFF
Economic
EconomicAnalysis
Analysis
PartGG
Nomenclature
Nomenclature
PartHH

Glossary
Part
PartI I
Figure I Analytical Framework and Model Descriptions Road Map

Version 1.0
PART I GLOSSARY
Part I Glossary
I1 Glossary
Analytical Framework and Model Descriptions I1-1

Version 1.0
volume five
A Guide to Calibration
and Adaptation
Christopher R. Bennett
William D. O. Paterson Association
mondiale
de la Route
Wo r l d R o a d
Association
5
Part A
About This Manual

This Version1.0 edition of A Guide to Calibration and Adaptation provides details on
calibrating the HDM model. It is one of five manuals comprising the suite of HDM-4
documentation (see Figure 1). It is intended to be used by specialists interested in technical
issues or responsible for setting up the HDM model.
Overview of HDM-4
Volume 1
Applications
Applications Guide
Guide Software
Software User
User Guide
Guide
Volume
Volume 22 Volume
Volume 33
Analytical
Framework andand Model
Model
Descriptions
Descriptions
Volume
Volume 44
A
A Guide
Guide to
to Calibration
Calibration and
and Adaptation
Adaptation
Volume
Volume 55

! Overview (Volume 1)

Describes the HDM-4 software. It is a general-purpose document that provides an

A Guide to Calibration and Adaptation i

Version 1.0
ABOUT THIS MANUAL
Notes:
Structure of ‘A Guide to Calibration and Adaptation’

The information in this document is structured as follows:
! Chapter 1 - Introduction
Provides a general description of HDM-4 and its scope.

! Chapter 2 – Calibration issues
Describes the need for model calibration, different levels of calibration based on
available time and resources.
! Chapter 3 - Reliability concepts
Addresses bias and precision of model predictions, input data accuracy, and ways of
assessing the reliability of HDM predictions.
! Chapter 4 – Sensitivity of HDM
Presents the results of analyses with HDM to identify the sensitivity of HDM’s output to
changing the input data or model parameters.
! Chapter 5 – Adapting data to the model
A discussion of how one adapts local data to the HDM model. Appendices A to H
address the specifics of how one collects data for use in HDM.
! Chapter 6 – RUE model calibration
Presents details on the calibration of the HDM road user effects model.
! Chapter 7 – RDWE calibration
Presents details on the calibration of the HDM road deterioration and maintenance effects
model.
A Guide to Calibration and Adaptation ii

Version 1.0
ABOUT THIS MANUAL
! Chapter 8 – Summary and conclusions
Summarises the main conclusion of the report.

! Chapter 9 – References
Lists the reports referenced in the document.

! Appendix A - HDM data items and calibration factors
Lists HDM data, model parameters and calibration levels.

! Appendix B - Parameter values used in HDM studies
Lists parameter values used in different HDM-III studies.

! Appendix C - Road user data
References road user data

! Appendix D - Road and pavement data
Provides descriptions and discussions of general road data, pavement characteristics,

pavement condition, and the maintenance strategies.
! Appendix E - Traffic data
Describes traffic volumes, growth rates and other associated issues.

! Appendix F - Unit cost data
Defines unit cost data for RUE and works effects (WE).
! Appendix G - Economic data

! Appendix H - Determining sample sizes
Describes the two approaches (the t Distribution and the normal distribution) that can be
used. The approach is dependent upon the number of samples.
! Appendix I - Survival curve analysis
Describes the calculation of the survival curve.

! Appendix J - Orthogonal regression
Describes orthogonal regression data.

! Appendix K - HDM Tools user guide
Describes HDM Tools, a set of software applications, which is designed to assist in the
calibration of HDM-4.
Figure 2 shows the process readers should follow in using this manual.
A Guide to Calibration and Adaptation iii

Version 1.0
ABOUT THIS MANUAL
Calibrate Selected
Establish Calibration
Setup RUE Model
Level
HDM Model Parameters
(Chapter 2)
(Chapter 6)
Calibrate Selected
Review Reliability
RDWE Model
Concepts
Parameters
(Chapter 3)
(Chapter 7)
Based on Time,
Resources, Type of Establish Input Data
Analysis Select Most for HDM Model
Critical Parameters (Appendix C to G)
(Chapter 4)
Establish Appropriate
Level of Data Detail
(Chapter 5)
Figure 2 HDM calibration process
ISOHDM products
Customer contact
HDM-4. Comments on A Guide to Calibration and Adaptation should be sent to the
following address:
Telephone: +44 - 121 - 414 6717 (5049)
Fax: +44 - 121 - 414 3675 (5060)
A Guide to Calibration and Adaptation iv

Version 1.0
ABOUT THIS MANUAL

Edgbaston
Birmingham B15 2TT
United Kingdom
The authors can be contacted at:
Dr. Christopher R. Bennett chris@htc.co.nz
Dr. William D.O. Paterson wpaterson@worldbank.org
Change details
HDM-4 documents
comprises:
A Guide to Calibration and Adaptation v

Version 1.0
ABOUT THIS MANUAL

Edgbaston
Birmingham
B15 2TT
England
Tel: +44 - 121 - 414 6717 (or 5049)
Fax: +44 - 121 - 414 3675 (or 5060)
La Grande Arche
France
Tel: +33 1 41 02 05 84
Fax: +33 1 49 00 02 02
Many of the tools, background reports and additional material for calibration is available from
the HDM links available at the following web site:
! HTC Infrastructure Management Ltd.
Web: http://www.htc.co.nz
Acknowledgements
The work on this Guide to Calibration and Adaptation was initially sponsored by N.D. Lea
International Ltd. of Canada. Additional work was sponsored by the Asian Development
Bank, HTC Infrastructure Management Ltd., and finally the ISOHDM Study through PIARC.
The authors would like to express their appreciation to all those who assisted in developing
this manual:
! Ian Greenwood of Opus International Consultants Ltd. (N.Z.)
Wrote the HDM Tools calibration applications.
! Rodrigo Archondo-Callao of the World Bank
Assisted in generating elasticities for the HDM-III model.
! Technical comments were received from a number of reviewers including, but not
limited to:
❏ Rodrigo Archondo-Callao
A Guide to Calibration and Adaptation vi

Version 1.0
ABOUT THIS MANUAL
❏ Clive Daniels
❏ Chris Hoban
❏ Michael J. Riley
❏ Robert Thurlow
! The World Bank

including:
! Finnra

! FICEM

concrete roads.
Birmingham
! Road Research Institute (IKRAM) in Malaysia supported by N.D. Lea International
(NDLI)
costs.

! SNRA
A Guide to Calibration and Adaptation vii

Version 1.0
ABOUT THIS MANUAL
Copyright statement
A Guide to Calibration and Adaptation viii

Version 1.0
Part A A Guide to Calibration and Adaptation
Contents
Volume 5 - A Guide to Calibration and
Adaptation
1 Introduction 1
2 Calibration issues 4
2.1 The need for calibration 4
2.1.1 Model development considerations 5
2.1.2 Case examples 5
2.2 Input data 7
2.3 Calibration 7
2.3.1 Introduction 7
2.3.2 Level 1 - Application 9
2.3.3 Level 2 - Calibration 11
2.3.4 Level 3 - Adaptation 11
2.4 Report scope 11
3 Reliability concepts 13
3.1 Introduction 13
3.2 Bias and precision 13
3.3 Correction factors 15
3.4 Input data accuracy 15
3.5 Limits on bias 17
3.6 Assessing the reliability of HDM predictions 18
3.6.1 Pavement performance: simulation of the past 18
3.6.2 Pavement performance: controlled studies 19
3.6.3 Road User Effects: controlled survey 20
3.6.4 Road User Effects: tariff survey 20
3.6.5 Road User Effects: fleet survey 22
4 Sensitivity of HDM 24
4.1 Introduction 24
4.2 Road User Effects 25
4.3 Sensitivity classes 30
4.3.1 High impacts, Class S-I (> 0.5) 30
4.3.2 Moderate impacts, Class S-II (0.2 - 0.5) 30
A Guide to Calibration and Adaptation i

Version 1.0
CONTENTS
4.3.3 Low impacts, Class S-III (0.05-0.2) 31

4.3.4 Negligible impacts, Class S-IV (< 0.05) 31
4.4 Road deterioration and works effects 31
4.4.1 High impacts, Class S-I (> 0.5) 32
4.4.2 Moderate impacts, Class S-II (0.2 - 0.5) 32
4.4.3 Low impacts, Class S-III (0.05-0.2) 33
4.4.4 Negligible impacts, Class S-IV (< 0.05) 33
4.5 Economic models 35
5 Adapting data to the model 36

5.1 Introduction 36
5.2 Concept of Information Quality Level (IQL) 37
5.3 Relating the local IQL to the HDM model 39
5.4 Adaptation of local condition data units 41
5.5 Transforming road input data 43
5.5.1 Road geometry 44
5.5.2 Pavement condition 44
5.5.3 Pavement structure 44
5.5.4 Environment 48
5.6 Transforming traffic input data 48
5.6.1 Traffic volume 48
5.6.2 Traffic flow 48
5.6.3 Traffic safety 50
5.6.4 Vehicle emissions 50
5.7 Transforming output data and producing performance Indicators 50
5.8 Data accuracy 50
5.9 Checklist for data adaptation 50
6 RUE model calibration 52

6.1 Introduction 52
6.2 Selecting representative vehicles 53
6.3 Level 1 - Basic application 53
6.3.1 Mass and vehicle damage factor: S-II 53
6.3.2 Capacity and speed flow data 56
6.3.3 Average service life: S-II 59
6.3.4 Vehicle utilisation: S-II/S-III 59
6.3.5 Speed prediction model parameters: S-II/S-III 61
6.3.6 Vehicle driving power: S-III 63
6.3.7 Tyre type, wheel diameter and number of wheels: S-IV 64
A Guide to Calibration and Adaptation ii

Version 1.0
CONTENTS
6.3.8 Volume of wearable rubber: S-III 65

6.3.9 Depreciation parameters 68
6.3.10 Aerodynamic drag coefficient and projected frontal area: S-III 68
6.3.11 Braking power: S-IV 68
6.3.12 Engine speed 69
6.4 Level 2 - Calibration of primary relationships 69
6.4.1 Speed prediction 69
6.4.2 Capacity and speed flow 72
6.4.3 Passenger car space equivalencies: S-II/S-III 77
6.4.4 Fuel consumption 78
6.4.5 Acceleration noise 82
6.4.6 Parts and labour costs 85
6.4.7 Used power 86
6.4.8 Braking power 88
6.4.9 Service life 88
6.4.10 Utilisation 88
6.4.11 Oil consumption 89
6.5 Level 3 - Adaptation of model 89
6.5.1 Acceleration noise 89
6.5.2 Speed β 90
6.5.3 Rolling resistance 90
6.5.4 Idle fuel consumption 90
6.5.5 Fuel efficiency 90
6.5.6 Critical gradient length 91
6.5.7 Tyre stiffness 91
6.5.8 Engine accessory power 91
6.5.9 Engine drag power 91
6.5.10 Drivetrain efficiency 91
6.5.11 Tyre wear parameters 91
6.5.12 Parts consumption 91
6.5.13 Labour hours 91
7 RDWE calibration 92
7.1 Bituminous paved road deterioration 92
7.2 Level 1 - Basic application 93
7.2.1 Roughness-age-environment adjustment factor: S-I 93
7.2.2 Cracking initiation adjustment factor: S-I 96
7.2.3 Cracking progression adjustment factor: S-I 98
7.2.4 Rut depth progression adjustment factor: S-I 98
7.2.5 General roughness progression factor: S-III 100
A Guide to Calibration and Adaptation iii

Version 1.0
CONTENTS
7.2.6 Ravelling initiation factor: S-III 100

7.2.7 Potholing progression adjustment factor: S-IV 100
7.3 Level 2 - Calibration of primary relationships 101
7.3.1 Roughness-age-environment deterioration factor: S-I 101
7.3.2 Crack initiation adjustment factor: (S-II or S-I) 102
7.3.3 Cracking progression factor: S-III 104
7.3.4 Rut depth progression factor: S-II/S-III 105
7.3.5 General roughness progression factor: S-III 106
7.3.6 Ravelling initiation factor: S-IV 107
7.3.7 Potholing adjustment factor: S-III 108
7.4 Level 3 - Adaptation of model 109
7.4.1 General design and methods of experimental study 110
7.4.2 Distress initiation models 111
7.4.3 Distress progression models 112
7.4.4 Rut depth progression models 112
7.4.5 Roughness progression models 112
7.5 Works effects on bituminous paved roads: S-II 112
7.5.1 Level 1 113
7.5.2 Level 2 113
7.6 Concrete pavement deterioration 113
7.7 Concrete pavement works effects 113
7.8 Unpaved roads deterioration and works effects 113
8 Summary and conclusions 114
9 References 115
Appendix A HDM data items and calibration factors 122
Appendix B Parameter values used in HDM studies 132
Appendix C Road user data 156
Appendix D Road and pavement data 157

D.1 Introduction 157
D.2 Pavement classifications 157
D.3 General road data 160
D.4 Pavement characteristics 161
D.5 Pavement condition 162
D.6 Maintenance alternatives 163
D.7 Homogeneous sections 164
A Guide to Calibration and Adaptation iv

Version 1.0
CONTENTS
Appendix E Traffic data 167

E.1 Introduction 167
E.2 Traffic volume and composition 167
E.3 Hourly distribution of traffic volume 169
E.4 Traffic growth rates 170
E.5 Traffic volume and classification surveys 173
E.6 Vehicle loading 177
Appendix F Unit cost data 180

F.1 Introduction 180
F.2 Economic and financial costs 180
F.3 Road user effect unit costs 182
F.4 Works effects unit costs 189
Appendix G Economic data 190

G.1 Introduction 190
G.2 Discount rate 190
G.3 Analysis period 190
Appendix H Determining sample sizes 191
Appendix I Survival curve analysis 193
Appendix J Orthogonal regression 195
Appendix K HDM Tools user guide 198

K.1 Introduction 198
K.2 Road user effects 198
K.3 HDM Tools Launchpad 199
K.4 Pavement deterioration and maintenance effects 214
K.5 Analysis 214
A Guide to Calibration and Adaptation v

Version 1.0
A Guide to Calibration and Adaptation
1 Introduction
The Highway Development and Management Model (HDM), originally developed by the
World Bank, has become widely used as a planning and programming tool for highway
expenditures and maintenance standards. HDM is a computer model that simulates physical
and economic conditions over the period of analysis, usually a life cycle, for a series of
alternative alternatives and scenarios specified by the user.
HDM is designed to make comparative cost estimates and economic evaluations of different
construction and maintenance options, including different time-staging alternatives, either for
a given road project on a specific alignment or for groups of links on an entire network. It
estimates the total costs for a large number of alternative project designs and maintenance
alternatives year by year, discounting the future costs if desired at different postulated
discount rates so that the user can search for the alternative with the lowest discounted total
cost.
Three interacting sets of costs (related to construction, maintenance and road use) are added
together over time in discounted present values, where the costs are determined by first
predicting physical quantities of resource consumption and then multiplying these by unit
costs or prices.
The broad concept of HDM is illustrated in Figure 1.1. The user defines a series of
alternatives that describe different investment and preservation options for the road. The
investments influence the condition of pavement over time and road maintenance costs. The
pavement and traffic conditions have an influence on Road User Effects (RUE). The model
predicts traffic speeds and the consumption of the RUE components, such as fuel, tyres etc.
Multiplying these by the unit costs of the individual components gives the RUE over time.
Comparing the cost outputs from various investment alternatives allows assessment of the
relative merits, cost savings and benefits of the different alternatives using economic
principles.
Aternative n
Aternative 3
Process
Process
Aternative 2
Process
Aternative 1 Process Economic
Economic
Comparison
Comparison
Process
Process Total Costs
Pavement condition Road maintenance Total Costs
over time costs over time Net Present Value (NPV)
Net Present Value (NPV)
Internal Rate of Return (IRR)
Internal Rate of Return (IRR)
Vehicle operating
cost component
Vehicle operating
consumption as a
costs over time
function of road
condition
Figure 1.1 Overview of the HDM approach
A Guide to Calibration and Adaptation 1

Version 1.0
Introduction
As illustrated in Figure 1.2, HDM consists of a series of sub-models that address different
aspects of the analysis. Each of these sub-models requires certain input data and each
produces its own output. In order to apply the model correctly, one needs to ensure that HDM
is given the appropriate input data and has been suitably calibrated.
This report presents guidelines for the quality assurance of HDM applications through control
of data quality and calibration of the HDM model. It describes the range of options for data
collection and input to HDM, as well as accuracy considerations. The sensitivity of HDM to
the input data is used to establish those data items that are most critical in the analysis.
Similarly, the report discusses how one calibrates HDM through its various calibration factors
and the sensitivity of the model to these factors. It presents three levels of effort to achieve
calibration based on the available time and resources.
The focus of the report is on HDM-4, although many of the broad principles are applicable to
the earlier releases HDM-III and HDM-95.
Box 1.1 Some definitions
• Vehicle Operating Costs (VOC) are the costs of vehicle operation. They are
comprised of fuel, tyres, maintenance, labour, depreciation, interest, lubricants,
crew, cargo and travel time.
• Road User Effects (RUE) are the VOC along with safety, vehicle emissions,
noise and energy balance.
• Road Deterioration and Works Effects (RDWE) encompass pavement
deterioration and the effects of works improvements. The latter include
maintenance as well as improvements such as widening.

Version 1.0
Introduction

geometry, pavement characteristics,
unit costs
Pavement type, strength, age, Cracking, ravelling, pot-holes,

condition, and ESAL Road Deterioration rut depth, faulting (paved);
gravel thickness (unpaved);
roughness
Road geometry and roughness; Fuel, lubricant, tyres,

vehicle speed, type; congestion Road User Effects maintenance, fixed costs, speed,
parameters; unit costs travel time, road user costs
Road works standards and strategies Reset cracking, ravelling, pot-

holes, rut depth (paved); gravel
works quantities and agency
costs
Road geometry and surface texture, Social and Levels of emissions and energy
vehicle characteristics Environmental Effects used, and number of accidents

environmental, and other exogenous Economic Analysis exogenous benefits
costs and benefits

by section
Figure 1.2 Structure of the HDM-4 model

Version 1.0
Calibration issues
2 Calibration issues
2.1 The need for calibration

As part of the International Study of Highway Development and Management Tools
(ISOHDM), a compendium was compiled of the countries where HDM had been applied.
HDM or its relationships has been applied in over 100 developed and developing countries
having markedly different technological, climatic and economic environments.
Since the model simulates future changes to the road system from current conditions, the
reliability of the results is dependent upon two primary considerations:
! How well the data provided to the model represent the reality of current conditions and
influencing factors, in the terms understood by the model; and,
! How well the predictions of the model fit the real behaviour and the interactions between
various factors for the variety of conditions to which it is applied.
Application of the model thus involves two important steps:
! Data input - A correct interpretation of the data input requirements, and achieving a
quality of input data that is appropriate to the desired reliability of the results.
! Calibration of outputs - Adjusting the model parameters to enhance how well the
forecast and outputs represent the changes and influences over time and under various
interventions.
The steps in modelling, from data to predictions to the model results are illustrated in Figure
2.1. In the context of this report, the input data and the calibration requirements of the
predictive models are dealt with.
Predictive
Input Data Analysis Results
Models
This Report
Figure 2.1 HDM modelling process
Calibration of the HDM model focuses on the two primary components that determine the
physical quantities, costs and benefits predicted for the analysis, namely:
! Road User Effects (RUE) - comprised of vehicle operating costs (VOC), travel time,
safety and emissions, and
! Road deterioration and works effects (RDWE) - comprised of the deterioration of the
pavement and the impact of maintenance activities on pavement condition and the future
rate of pavement deterioration.

Version 1.0
Calibration issues
2.1.1 Model development considerations

Early versions of HDM (HCM and HDM-II) relied on simple empirical regression models
that were built on field data collected at specific study sites. However, these simple models
lacked transferability because they could not show how the results would change when there
was a change in the assumed conditions. In the development of HDM-III and HDM-4, a high
degree of transferability across different technological and climatic conditions was built into
the model. This was achieved through the use of a structured mechanistic-empirical approach
in deriving the underlying predictive relationships. This powerful approach combined various
insights provided by the theories of motion and vehicle technology, and of pavement material
and structural behaviour under traffic loading, with a rigorous and advanced statistical
analysis of real data gathered from a wide range of vehicle types and road conditions. The
validity of the vehicle speed and operating costs models was built up from the four major
field studies conducted in Kenya, the Caribbean, Brazil and India (Chesher and Harrison,
1987; and Watanatada et al., 1987a). The validity of the various road deterioration models
was demonstrated on independent field data from up to eight different countries and states in
climates ranging from hot arid to cold wet (Paterson, 1987).
This approach to model building has two principal implications for the user of HDM:
! Size
The number and variety of data items required to feed structured models are potentially
large and complex. The HDM modelling chose a middle path, avoiding parameters that
require research-level tests and, for road user effects, setting some parameters to have
default values that the user could change when analytical resources permitted.
! Non-modelled effects
Even though the field experiments covered wide ranges of conditions, and even though
state-of-the-art theory was applied in the model’s development, there remain some
factors that could not be introduced, because:
❏ they were not measured, or
❏ would have made the model’s input too complex, or
❏ their effects could not be determined within the ranges observed
Also, there have been advances in vehicle technology in the period since the studies were
done which have lowered consumption and operating costs. For these reasons, some
calibration of the HDM model to local conditions is both sensible and desirable.
2.1.2 Case examples

The needs for careful application and some local calibration are discussed in Curtayne et al.
(1987). As an example of the needs, Figure 2.2 illustrates the differences arising between
costs observed for trucks in Canada and those predicted by the HDM-III model with
unadjusted truck default values (Lea, 1988). Not only were the total VOC predicted by HDM-
III significantly different from those observed, but the relative contributions of the various
components to the total operating costs were also very different; which would also arise with
HDM-4 were it not properly calibrated. Thus, inadequate local application or calibration can
distort both the costs and their allocation that ultimately could distort the indicated allocation
of highway expenditures.

Version 1.0
Calibration issues
180
2-Axle Truck 5-Axle Truck 7-Axle Truck
160
140
120
Crew
Cost in cents/km
100 Interest
Depreciation
80 Maintenance
Tires
60 Fuel & Oil
40
20
0
Observed HDM Observed HDM Observed HDM
Figure 2.2 Comparison between observed and non-calibrated HDM-III

predictions
Chesher and Harrison (1987) also point to the need for local calibration through their
comparisons of the predictions of the models developed in the various road users cost studies.
The relationships not only show significant variations in the magnitudes of the various costs,
but also in their sensitivity to pavement and operating conditions. Finlayson and du Plessis
(1991), in a comprehensive report containing a number of different papers on VOC, illustrate
the importance of calibrating VOC equations to local conditions and discuss how it was done
for Southern Africa.
For road deterioration, environmental conditions have a strong influence, affecting
deterioration rates by factors of up to 2 or 3 between the extremes of hot arid and cold wet
climates. Local construction materials, practices and quality also affect deterioration rates and
the effectiveness of maintenance. Both influences can be controlled through calibration
factors in the inputs to HDM, in addition to the main traffic and pavement input parameters.
This report presents the ways in which the level of confidence of the HDM predictions can be
raised through differing levels of effort that can be matched to the level of application. The
advantages of a universal model such as HDM are that it can be used quickly with relatively
little or no investment in extensive empirical and statistical research, and that it represents a
comprehensive techno-economic framework. Enhancing the reliability thus involves first
ensuring the validity of input data and primary parameter values, and second, refining but not
reconstructing the predictive relationships to conform to local data. A third stage may be
undertaken which sees relationships replaced by new ones developed from local research.

Version 1.0
Calibration issues
Box 2.1 Data versus calibration
An analogy of the sea is useful to illustrate the roles of data and calibration. Data
determine the order and magnitude of costs and effects, so these must be of the
same order as baseline information, much like the depth of the sea and other
attributes like density. Calibration ensures that the height of the waves will be
correct under the influence of winds, currents and depths of water.
Calibration
'The Depth of the Sea and

Data
the Height of the Waves'
This report deals with both issues. In the context of HDM-4 data are those items
which, when you install the model, require input data. This includes unit costs, road
attributes, maintenance alternatives, etc. Where default values are supplied, these
are model parameters or coefficients that can be calibrated.
The main text of the report deals with calibration of model parameters and
coefficients; Appendix C to Appendix G deals with how one establishes the input
data for HDM-4.
2.2 Input data

Input data are the basic data items required to run HDM. These consist of parameters that
describe the physical characteristics of the pavements and the network, road user data, traffic
data, unit costs and economic data. Appendix A summarises the input data items that are
required for HDM applications. Appendix B gives some examples of parameter values
adopted in different HDM studies. Appendix C to Appendix G describes how the data are
quantified.
In establishing input data, the accuracy required is dictated by the objectives of the analysis.
If one is doing a very approximate analysis there is no need to quantify the input data to a
very high degree of accuracy. Conversely, if one is doing a detailed analysis it is important to
quantify the data as accurately as is practical given the available resources. This is discussed
in Chapter 5.
2.3 Calibration
2.3.1 Introduction
Calibration differs from input data since calibration is aimed at adjusting the model
predictions. As shown in Appendix A, the HDM RUE and RDWE sub-models contain a large
number of parameters that can be adjusted. It will be noted from Appendix A that HDM-4

Version 1.0
Calibration issues
has many more calibration factors than HDM-III. This is because many values that were hard
coded in the HDM-III source code can now be altered in HDM-4.
The degree of local calibration appropriate for HDM is a choice that depends very much on
the type of application and on the resources available to the user. For example, in planning
applications the absolute magnitude of the RUE and road construction costs need to match
local costs closely because alternative capital projects with different traffic capacities or route
lengths are evaluated on the comparison of the total road transport costs. In road maintenance
programming, on the other hand, the sensitivity of RUE to road conditions, particularly
roughness, and all the road deterioration and maintenance predictions are the most important
aspects.
There are three levels of calibration for HDM, which involve low, moderate and major
levels of effort and resources, as follows:
! Level 1 - Basic Application (see Section 2.3.2)
Determines the values of required basic input parameters, adopts many default values,
and calibrates the most sensitive parameters with best estimates, desk studies or minimal
field surveys.
! Level 2 - Calibration (see Section 2.3.3)
Requires measurement of additional input parameters and moderate field surveys to

calibrate key predictive relationships to local conditions. This level may entail slight
modification of the model source code.
! Level 3 - Adaptation (see Section 2.3.4)
Undertakes major field surveys and controlled experiments to enhance the existing
predictive relationships or to develop new and locally specific relationships for
substitution in the source code of the model.
In terms of effort, these three levels can be viewed as weeks, months and years. An analyst
should be able to undertake a Level 1 calibration in about one-week. For a Level 2 calibration
there is an increase in the amount of effort required so it will take at least a month. Level 3
calibrations require a long-term commitment to basic data collection so their extent spans for
a year or more.
Every HDM analysis requires at least a Level 1 calibration.
Figure 2.3 illustrates this concept of increasing effort and increasing resources. It must be
appreciated that there is a direct relationship between the time and effort expended in setting
up HDM and the reliability and accuracy of its output.

Version 1.0
Calibration issues
Time Required
Experimental
Years
Surveys and
Research
Months
Field Surveys
Weeks
Desk Studies
Resources
Limited Moderate Significant Required
General Planning Project Appraisal Research and

Quick Prioritisation Detailed Feasibility Development
Preliminary Screening
Coarse Estimates Reliable Estimates
Figure 2.3 Resources and time required for calibrations
As shown in Table 2.1, a related concept of a hierarchy of calibration with increasing levels
of activity was suggested by Curtayne et al. (1987) in regard to VOC. They noted that
calibration activities could be based on both primary and secondary data sources. Primary
sources are based on direct comparisons between costs and highway characteristics whereas
secondary data provide information about the economic or operating conditions of a region.
The calibration of VOC results varies from the selection of available relationships on the
basis of few data to the estimation of local relationships using especially collected data.
The following sections outline the components of Level 1 to Level 3 calibrations.
Appendix A gives the data items and model parameters that should be investigated in
calibration. It should be noted that not all the items need to be considered, Chapter 4 presents
the results of a sensitivity analysis with HDM which shows which are the most important data
and parameters in terms of influencing the model predictions. These are where the emphasis
should be focused with less sensitive items receiving attention only if time and resources
allowing.
2.3.2 Level 1 - Application

In order to run HDM it is always necessary to undertake at least a Level 1 calibration; this can
be viewed as a set-up investment for the model. Once this has been done, it generally does not
need to be repeated for most of the input data files during future applications in the same
country since many data items and most model parameters are relatively stable over time.
A Level 1 calibration is largely based on secondary sources; that is, it is a desk study. For
example, the RUE parameters can be estimated using data from sources such as government
and industry publications, operator organisations or various RUE reports from previous
studies. For road deterioration, the sources would include climate statistics, road traffic and
condition statistics, geometric standards, maintenance programmes and budgets.
It can be assumed that the bulk of the default HDM model parameters are appropriate for
local conditions so only the most critical ones need to be addressed.

Version 1.0
Calibration issues
Table 2.1
RUE calibration data sources and hierarchy of resources
Type of data Sources or Increasing resources required

needs
Secondary Government Choose primary Choose data sets Examine intercept
available publications study results which and vehicle types and slope values
correspond to local for vehicle types
Industry technical conditions Calibrate with available cost
reports mechanistic models and road condition
for technology data
RUE research changes
reports
Manufacturer's
literature
Consultant's reports
Operator
organisations
Road condition
inventories
Survey calibration Small-scale Confirm intercept Estimation of tyre Estimate new local
research collecting values and assess and depreciation relations, especially
cost data from slope magnitudes costs for maintenance
companies costs
operating over a Determine vehicle Confirm slope
good range of utilisation by age values for total Compare predicted
highway and road condition costs RUE with rates
characteristics
Personnel with
industry knowledge
Rates survey
Route classification
Experimental Small-scale Speed calibration Fuel consumption Estimate new

calibrations research of vehicle calibration relations, for
performance and example for rolling
roadway resistance and road
characteristics roughness
Trained personnel
Analytical
capabilities
Source: Curtayne et al. (1987)
As shown in Appendix A, the following input data should be obtained for a Level 1
calibration:

Version 1.0
Calibration issues
! Unit costs (RUE and RDWE)
! Certain characteristics of representative vehicles
! Economic analysis data (discount rates and analysis period)
! Pavement characteristics (RDWE studies)
! Traffic composition and growth rates

! Regional climatic type
While HDM often calls for a wide range of input data and calibration parameters, but only the
most important need to be established for use with a Level 1 calibration, so the HDM default
values should be used almost exclusively.
2.3.3 Level 2 - Calibration

A Level 2 calibration uses direct measurements of local conditions to verify and adjust the
predictive capability of the model. It requires a higher degree of data collection and precision
than in a Level 1 calibration, and extends the scope. For RUE, it concentrates on speed, fuel
consumption, tyre consumption, parts consumption and the fixed costs relating to utilisation
and vehicle life. For RDWE, it concentrates on the initiations of surface distress modes,
rutting progression, and maintenance effects, and enhances the estimate of environmental
impacts. For the economic analysis, it ties cost data more closely to observed cost and price
levels through data collection surveys.
With Level 2 calibrations, more detailed input data are also collected than with Level 1.
2.3.4 Level 3 - Adaptation

A Level 3 calibration is generally comprised of two components:
! Improved data collection
! Fundamental research
Some data items can be estimated with reasonable accuracy using short-term counts, for
example the hourly distribution of traffic volume, but the reliability is greatly enhanced by
collecting data over more sites over a longer period.
Fundamental research considers the relationships used in HDM. This consists of structured
field surveys and experimental studies conducted under local conditions which lead to
alternative relationships. For example, alternative functions may be developed for predicting
fuel consumption or new pavement deterioration and maintenance effects functions for
different pavement types. Such work requires a major commitment to good quality, well-
structured field research and statistical analysis over a period of several years. Pavement
deterioration research is a particularly long-term endeavour, typically requiring a minimum of
5 years.
2.4 Report scope

This report has been prepared as a guide to those responsible for preparing analyses with the
HDM model. It mainly deals with the first two levels, that is Application and Calibration
hereinafter referred to as Level 1 and Level 2 calibrations respectively which constitute the
great majority of all uses.
Users interested in a Level 3 calibration are referred to the HDM-III background references
(Watanatada et al., (1987a); (1987b); (1987c); Paterson, 1987; and Chesher and Harrison,

Version 1.0
Calibration issues
1987) detailing the original research. These contain material on experimental design, theory,
model forms, analytical methods and empirical limitations.

Version 1.0
Reliability concepts
3 Reliability concepts
3.1 Introduction
The objective of an HDM analysis is to model roads. This entails predicting the deterioration
of the pavement under time and traffic, the road user effects, and the effects of maintenance
on the pavement condition and rate of deterioration. As with any model, HDM is a
representation of reality. How well the model predictions reflect reality is dependent upon a
combination of the:
! Validity of the underlying HDM relationships
! Accuracy and adequacy of the input data
! Calibration factors used in the analysis
Since the underlying HDM relationships have proven to be robust and applicable in a number
of countries, the reliability of most HDM analyses depends on the input data and the
calibration factors.
This chapter addresses reliability issues in the context of HDM. It gives an overview of the
commonly encountered problems and describes statistical methods to improve reliability.
3.2 Bias and precision

The only way of assessing the adequacy of HDM’s predictions is by comparing the model
predictions to known data. For example, one may have data on the current roughness of a
number of pavements of known ages. By using HDM to predict the condition of pavements of
the same age with the same attributes from when they were new, one could assess whether
HDM was giving appropriate predictions.
There are two considerations when comparing predicted and observed data:
! Bias
A systematic difference that arises between the observed and predicted values. For
example, if the predictions are always 10 per cent lower than the observed data. The
formal definition of bias is the difference between the mean predicted and mean
observed values.
! Precision
A measure of how closely the observed and predicted data are to each other1. It is
represented by the reciprocal of the variances (σobs2/σpre2); that is, it is reflected by the
scatter when plotting the observed versus predicted data. Precision is influenced the
inherent stochastic variations of most natural processes, measurement and observational
errors, and unexplained factors omitted from relationships in the model.
Figure 3.1 illustrates both of these concepts for four scenarios:
1
Precision here should not be confused with precision of a measurement which refers to the closeness of
repeated measurements to each other, for example, the weight is 10 kg + 0.1 kg.

Version 1.0
! Low Bias - High Precision (see Figure 3.1A)
! Low Bias - Low Precision (see Figure 3.1B)
! High Bias - High Precision (see Figure 3.1C)
! High Bias - Low Precision (see Figure 3.1D)
In Figure 3.1 the shaded ellipse represents observed data which has been plotted against
predicted data. The solid line at 45° is the line of equality, where the observed and predicted
are equal.
Low Bias Low Bias

A High Precision
B Low Precision
Data
Predicted
Predicted
Data
Observed Observed
High Bias High Bias

C High Precision
D Low Precision

Predicted
Predicted
Data
Data
Observed Observed
Figure 3.1 Example of bias and precision
When there is low bias, the data will fall around, or close to, the line of equality. As
illustrated in the Figure 3.1-A (low bias and high precision), when there is high precision
there is little scatter in the data. However, a reduction in precision sees an increase in the
scatter and, thus, the standard deviation (Figure 3.1-B (low bias and low precision)).
When there is high precision, the change from low to high bias sees the slope of the observed
data systematically different to the line of equality. This is illustrated in Figure 3.1-C (high
bias and high precision). This figure also shows what happens when there additionally is a
systematic difference, the slope of the observed versus predicted line does not pass through
the origin.

Version 1.0
The most difficult situation is illustrated in Figure 3.1-D (high bias and low precision). In this
instance it is often difficult to verify the reasonableness of the model since the differences
could be equally due to the poor precision as to the bias.
Box 3.1 Important considerations
• Calibration needs to extend over the full range of values likely to be

encountered
• The sample size must be sufficient to detect the nature of the bias and the level
of precision
• A high r2 (correlation) does not always mean high accuracy, there can still be
significant bias
3.3 Correction factors

Bias arises because of systematic differences in the observed versus predicted values.
Correction factors are used to correct for the bias. As shown in Figure 3.2, there are two types
of calibration factors: rotation and translation. Either or both of these may be present.
In the simplest case, the bias is established as the ratio of the mean observed to mean
predicted to the mean observed. Thus, the correction factor is:
Mean Observed
CFrot = …(3.1)
Mean Predicted
This is referred to as the rotation correction factor because, as illustrated in Figure 3.2-A, the
predictions are rotated down to where they correspond to the observed data.
The translation factor is used when there is a constant difference between the observed and
predicted values across all conditions (see Figure 3.2-B). An example of this is where the
vehicle operating costs are overestimated due do overheads being improperly included. In this
instance the correction factor is:
CFtrans = Mean Observed - Mean Predicted …(3.2)
Figure 3.2-C also shows the third, and common, scenario where there is a combination of
rotation and translation.
3.4 Input data accuracy

Bias and precision also come into play in establishing the input data for HDM. This is
because when quantifying input data, one obtains a sample of measurements whose properties
are assumed to be the same as that of the underlying population. The bias is often expressed
as:
Mean Sampled
Bias = …(3.3)
Mean Population
This is analogous to the rotation correction factor presented in Section 3.3.

Version 1.0
A Rotation B Translation
Translation
Rotation
Predicted
Predicted
Data Data
Observed Observed
Translation
Rotation and
C Translation
Rotation
Observed = Predicted
Predicted
Data
Observed
Figure 3.2 Correction factors
As illustrated in Figure 3.3, when the mean of the sample is different to the mean of the
population, the sample is biased. If there is the same mean but different variances, there is a
problem with the precision (see Figure 3.4).
In quantifying input data the objective is to ensure that the distribution of the sampled data is
similar as that for the population. This is achieved by using standard statistical sampling
techniques that ensure that there is a sufficient sample size to limit random sampling errors.
Appendix H describes the method for establishing the required sample sizes for different
levels of confidence.

Version 1.0
frequency
Bias
Population Sampled
value
Figure 3.3 Bias in sampled data
frequency Precision
Sampled
Population
value
Figure 3.4 Precision in sampled data
Box 3.2 Bias and precision

The principal objective is to minimise the bias.
There is always a cost/resource trade-off when it comes to precision.
High precision is costly and resource intensive whereas lower cost
techniques usually result in lower precision.
If the bias is low, one can usually accept lower precision data since the
mean observed and predicted will be similar (see Figure 3.1-B). If the
bias is high, it is important to have high precision data in order to
ascertain the nature of the bias (see Figure 3.1-C). High bias and low
precision data (see Figure 3.1-D) makes it impossible to accurately apply
the model.
3.5 Limits on bias

It is important that the bias in the model predictions be limited so as to ensure that the results
are accurate. Since each of the many data items and model parameters may have their own
biases, and the processes being modelled by HDM are complex, it is extremely difficult to
determine the net bias effects of multiple parameters.
The goal of calibration is defined here with respect to the impact of an individual parameter,
but very approximate in respect of the impact of all parameters. The goals of calibration,
excluding the fundamental lack of fit of the model, are to reduce any bias of the predictions

Version 1.0
by the model to acceptable levels. Level 2, being of a higher standard, is intended to produce
less residual bias in the model’s predictions than a Level 1 calibration.
A Level 3 calibration aims to improve both precision and bias by reconfiguring one or more
relationships within the model. It thus requires a comprehensive scope and database in order
to adequately detect and determine the many factors, effects and interactions.
3.6 Assessing the reliability of HDM predictions

There are a number of techniques that can be used to assess the reliability of HDM
predictions. The most appropriate one depends upon the available data and what the objective
is. Specific guidance is given in later chapters.
3.6.1 Pavement performance: simulation of the past

One of the easiest methods for assessing the overall reliability of RDWE predictions is to
simulate the past condition of the road. This can be done in a Level 1 or Level 2 calibration
and is always a good check on the model.
In Bangladesh the evolution of the network over the previous 15 years was simulated
(Transroute, 1992). It was assumed that the initial condition of the network was equivalent to
the current condition and that the traffic growth rates over the period were equivalent to the
forecast traffic in the next 10 years. The simulation was tested using two premises, the:
! Predicted network roughness distribution was similar to the current roughness
distribution.
! Average annual expenditure was close to what was observed.
Using the above criteria and appropriate maintenance standards it was found that “the
simulation [gave] a fairly good picture”.
In Nepal, NDLI (1993) analysed a specific section of road that was opened in 1970. Estimates
were made of the opening year traffic and HDM was run with a range of assumptions about
pavement strength and initial roughness. The predicted conditions were then compared to the
measured roughness and surface distress surveys assuming minimum maintenance.
The predicted roughness was dependent upon the assumed initial roughness and on the
pavement strength. Discussions with local engineers indicated that the road had a smooth
surface and using 3.5 IRI m/km in the modelling resulted in a predicted mean roughness only
slightly higher than the measured roughness. HDM under-predicted patching but this was
ascribed to the fact that HDM only patched potholes whereas local practices included badly
cracked areas and local depressions. Overall, the model was considered to give reasonable
predictions.
In terms of roughness progression, one method for testing the model predictions has been
through the use of a "slice-in-time" analysis. This technique uses network roughness data to
investigate the roughness-time and/or roughness-pavement strength relationship. Given the
variations present in network roughness data it is unlikely that this method will yield useful
results. NDLI (1992) applied this technique without success to Thailand. Several factors
contributed to the failure of the analysis:

Version 1.0
! The ages of pavements were difficult to accurately estimate which meant that the traffic
loadings were inaccurate.
! It was difficult to estimate the post-construction roughness.
! The pavement strength was estimated from construction records. However, the variations
in roughness along a section of road indicated that there were local variations in strength
which were not reflected in the construction records.
! Variations in traffic growth over the life of the road were not accurately captured.
3.6.2 Pavement performance: controlled studies

The only way of completely calibrating the HDM-III pavement deterioration model is by
conducting a study into the rate of pavement deterioration. Those considering such an
undertaking should consult the material in Hodges et al. (1975), GEIPOT (1982) and
Paterson (1987).
It is important that the experiment be designed to gather data on the following items:
! The effect of traffic on pavement deterioration
! The effect of the environment on pavement deterioration
! Deterioration rates by pavement type
! Pavement strength effects
! Surface distresses
The sites selected for the study should cover the full range of pavement types and strengths
within the country. For each pavement type, pavements covering the complete range of
strengths should be included. If there are major differences in the climate over the country,
the experiment should be designed to account for this.
The data to be collected will depend upon the objectives of the study. For an adequate
calibration the following items need to be collected as a minimum:
! Roughness
! FWD/Benkelman beam deflection
! Cracking
! Rut depth
In selecting the sections it should be appreciated that if the pavements have been properly
designed for the traffic level it should be difficult to observe traffic loading effects as these
have been catered for in the design. One should therefore try and use pavements under-
designed for their traffic levels.
It is important that the test sections be continually monitored and that all the data items are
collected at the same time. In a number of studies this was not done and it created problems
with the subsequent analysis.
It is recommended that a vehicle mounted roughness meter (RTRRMS) not be used in
measuring roughness. This is because the measurement errors with such instruments are of
the same or greater magnitude to the incremental changes in roughness over time. These
meters are also prone to calibration problems that may lead to additional errors. The
roughness should thus be measured using one of the direct profiling techniques used for
roughness meter calibration (for example, Dipstick, Walking Profilometer). This will ensure
that the data are of the greatest possible accuracy.

Version 1.0
It is important to monitor crack initiation and progression. Because of this, the test sections
should not have any maintenance done to them over the course of the study. This is often
difficult to arrange but is very important, particularly insofar as monitoring crack progression
is concerned.
3.6.3 Road User Effects: controlled survey

Controlled surveys for RUE often concentrate on fuel consumption, as this is the easiest
component to measure. Tyre wear can be investigated but it is often difficult to execute such
a study effectively1.
du Plessis and Meadows (1990) tried to investigate maintenance costs through a controlled
survey. Three rental cars were driven over two pre-selected routes, one paved and one gravel,
for a period of about three months during which they covered 40,000 km. Thorough
maintenance checks were effected every two weeks and fuel consumption and tyre wear were
monitored closely. The results showed that road roughness markedly affected both
maintenance and tyre wear, but not fuel consumption. The latter was ascribed to the
unavailability of speed and acceleration data. The maintenance and tyre results were used to
check the validity of the HDM predictions.
3.6.4 Road User Effects: tariff survey

One of the easiest ways of evaluating the accuracy of HDM RUE predictions is by comparing
the predictions to market tariffs. These can be done in Level 1 or Level 2 calibrations.
The underlying hypothesis is that tariffs reflect the total VOC of a vehicle and will therefore
give a reasonable estimate not only of the magnitude of VOC but also of the marginal costs
under different operating conditions. Generally, freight operations are more independent than
passenger operations since the latter often have price regulations/subsidies.
In collecting data on tariffs for such purposes, the following should be considered:
! The use of freight tariff data for model calibration is most appropriate when there is a
competitive market, that is, where there is a minimum of government control on
operation. When regulation or monopolies affect the supply of transport services, the
data must be carefully scrutinised.
! It is better to gather data from a single company operating over a range of operating
conditions than from many companies. As described in Chesher and Harrison (1987),
the variation in costs between companies is often greater than the variations due to
operating conditions.
! The data should be standardised in such a fashion that effects such as load levels are
accounted for.
! The HDM predictions should be made using financial instead of economic costs.
! The predicted costs should be lower than the tariff due to the need for the owner to make
a profit on the trip.
Hine and Chilver (1991) give a good example and discussion on using freight tariffs for VOC
model calibration. Using comprehensive data collected in Pakistan, a comparison was made
1
One major problem lies in being able to accurately establish the amount of tyre wear. Since this is a small
amount weighing the tyre before and after a set of measurements often does this. However as described in
Transit (1997), even when using the most controlled conditions significant errors may arise.

Version 1.0
between observed costs and the VOC predictions from the RTIM2, HDM-III Brazil and
HDM-III India VOC models with observed costs.
There are two components which are reflected in the tariff, the:
! Non-productive cost (for example, loading/unloading, repairs, looking for work)
! Per-km distance cost
Thus, the cost can be expressed as:
COST = a0 DIST + a1 …(3.4)
where:
COST is the total cost of the trip

DIST is the trip distance
a0 is the marginal cost due distance travelled
a1 is the fixed cost
The tariff rate can then be expressed as:
a1
RATE = a0 + …(3.5)
DIST
where:
RATE is the cost per km
NDLI (1997) used medium truck tariffs to check VOC predictions in India. Data were
collected from nine operators in the form of the tariffs charged for a standard load to travel
various distances. The data were quite variable, reflecting the different charges by operators
for trips of the same length. A regression analysis of the data gave the following equations:
COST = 6.62 DIST + 562 …(3.6)
562
RATE = 6.62 + …(3.7)
DIST
The analysis indicated that the marginal financial cost for truck travel was 6.6 Rupees/km.
The model was run using the financial unit costs instead of the economic costs. The predicted
total cost was 6.2 Rupees/km, a difference of 6 per cent. To evaluate the predicted costs
against the raw tariff data the total trip cost was calculated using the same constant cost of
562 Rupees/trip as from the tariff data. The results are plotted in Figure 3.5 against the
original tariff data (NDLI, 1997). It was concluded that the model was giving a reasonable
representation of the observed costs.
For passenger cars, taxi rates are often a ready source of data for evaluating the model
predictions. The rates need to be adjusted for the value of driver time.

Version 1.0
3.6.5 Road User Effects: fleet survey

Fleet surveys are often a good source of data, particularly when there are operators available
with good records. For example, Hine and Chilver (1991) used driver logbooks to investigate
truck costs in Pakistan while du Plessis et al. (1991) used bus company records in South
Africa.
Transit (1995) gives a good discussion of potential sources of data for a study in New
Zealand. It was found that there were a number of public and private organisations with data
covering a range of vehicles and operating conditions. There were two commercial companies
managing and recording vehicle operating cost data for their clients, and one of these was
used to investigate maintenance costs. It was also found that many private operators had good
databases with the costs recorded to a very disaggregate level.
7000
Regression to
6000 tariff data
5000
Tariff in Rs/trip
4000
Model prediction for:

Roughness: 6.75 m/km
3000
VCR: 0.21
Width: 7.0 m
2000
1000
0
0 100 200 300 400 500 600 700 800
Trip Length in km
Figure 3.5 Comparison of predicted and tariff survey trip costs in Gujarat
du Plessis and Schutte (1991) contain a number of papers describing several studies
conducted by the CSIR into South Africa that involved fleet surveys. These provide good
guidance on how to approach a fleet study and analyse the data.
du Plessis et al. (1991) studied a single bus company since this eliminated any inter-company
effects, although it was noted that the seven local depots had some latitude with their
maintenance practices to reflect local conditions. The company operated on a range of terrain
and road types. The data were grouped into depots in flat, rolling and mountainous terrain and
average attributes were established for each depot (for example, roughness, and operation on
different road types).
It was noted that “any direct attempt to relate costs to roughness may yield misleading results
because of other factors such as bus age and maintenance policy that may play a role”. The
study was therefore initiated at each depot where the actual policies were investigated. This
also allowed for discussions with the local maintenance managers that provided valuable
insights into local conditions that were influencing the costs. For example, it was found that
the tyre consumption was influenced by the angularity of gravel through casing penetrations.
The data were used to develop relationships that were compared to HDM. The results broadly
confirmed the validity of the HDM predictions.

Version 1.0
Findlayson and du Plessis (1991) conducted a similar study into trucks that developed
maintenance, tyre and depreciation relationships. These also gave similar trends to HDM.

Version 1.0
Sensitivity of HDM
4 Sensitivity of HDM
4.1 Introduction
It is important for users to be aware of the general level of sensitivity of the model to each
parameter1 so that appropriate emphasis can be given to important parameters and less
emphasis to second or third order effects. The influences of individual parameters differ
according to the particular parameter, the particular result being considered, and the values
assigned to other parameters in the particular analysis. The sensitivity of results to variations
in a parameter therefore varies somewhat under different circumstances.
Sensitivity analyses were conducted with the HDM RUE2 and RDWE sub-models so as to
determine the levels of sensitivity and to rank them3. Sensitivity was quantified by the impact
elasticity, which is simply the ratio of the percentage change in a specific result to the
percentage change of the input parameter, holding all other parameters constant at a mean
value.
For example, if a 10 per cent increase in traffic loading causes a 2.9 per cent increase in
roughness developed after 15 years, the impact elasticity term of traffic loading for that
roughness result is 0.29. If there were a 2.9 per cent decrease, the value would be -0.29.
As described by Mrawira et al. (1998), there are different approaches which can be used for
undertaking sensitivity analyses. The approach used here is the traditional ceteris paribus
method: changing a single factor while holding all others constant. The alternative approach,
using factoral experiments which combine all the levels of one factor with all levels of all
other factors, were not used due to the large number of combinations to consider. Thus, the
analysis here does not consider factor interactions. Mrawira et al. (1998) describes the results
of using a factor approach for an HDM sensitivity analysis.
On the basis of the analyses, four classes of model sensitivity have been established as a
function of the impact elasticity. The higher the elasticity, the more sensitive the model
predictions. These classes are listed in Table 4.1. Throughout the remainder of this report the
terms S-I to S-IV will be used to refer to the various sensitivity classes.
1
In order to run HDM it is necessary to supply the basic input data and values for the model coefficients; for
simplicity both of these are termed parameters.
2
As described in Bennett (1999), the analyses were done using a stand-alone version of the HDM-4 RUE
sub-model developed for this purpose.
3
The authors gratefully acknowledge the assistance of Rodrigo Archondo-Callao of the World Bank who
performed the RUE sensitivity analysis for HDM-III.

Version 1.0
Sensitivity of HDM
Table 4.1
HDM sensitivity classes
Impact Sensitivity class Impact elasticity

High S-I > 0.50
Moderate S-II 0.20 - 0.50
Low S-III 0.05 - 0.20
Negligible S-IV < 0.05
Practitioners, as a guide to where their efforts should be directed, should use the results of
these sensitivity analyses. Those data items or model coefficients with moderate to high
impacts (S-I and S-II) should receive the most attention. The low to negligible impact (S-III
and S-IV) items should receive attention only if time or resources permit. One usually
assumes the default HDM values for S-III and S-IV items since these will generally give
adequate results.
4.2 Road User Effects

RUE are comprised of the VOC, travel time, vehicle emissions (noxious gases and noise),
safety, energy use along with developmental effects. This release of the report only considers
VOC.
As described in Chapter 2, the HDM RUE sub-model predicts the amount of resources
consumed; for example the amount of fuel and tyres. These are multiplied by the unit costs of
the resources to obtain the total cost. There are a number of different components modelled
and, as shown in Figure 4.1 and Figure 4.2, they are influenced by different factors. These
figures also highlight the increased sophistication of RUE modelling in HDM-4 over HDM-
III.
RUE are mainly influenced by vehicle speeds and roughness. The influence of these factors
varies depending upon the RUE component. For example, fuel consumption is very sensitive
to speed but relatively insensitive to roughness. By comparison, parts consumption is
insensitive to speed and heavily influenced by roughness.
Analysts need to be aware of these differences since they can have a marked bearing on the
data collected in a study. If the study is principally dealing with capacity improvements that
will influence speeds, then factors such as roughness will be of lesser importance. However,
road maintenance studies that will lead to major changes in roughness will need to have good
data on roughness.
When considering RUE there are two situations that may be of interest, namely the:
! Magnitude of the total RUE
! Effect of operating conditions on RUE
The former is important in situations such as when there are different route lengths. The
effects of operating conditions on RUE are important when you are comparing changes to
road condition.
As described in Bennett (1999), sensitivity analyses were conducted which considered both of
these issues. The first set of analyses considered the sensitivity of the total VOC on a per-km
to changes in the input data. The second considered the effects of roughness changes. For the

Version 1.0
Sensitivity of HDM
latter, the sensitivity of the difference in the total VOC between 3 and 5 IRI m/km to changes
in the input data was tested.
In order to cover the range of conditions encountered in different studies, the analyses were
done with unit cost data adapted from actual HDM studies from Australia, India, and
Thailand. These three data sets represent countries with markedly different levels of
economic development. Using these different studies meant that the relative contribution of
each VOC component was different thereby influencing the sensitivities. This ensured that
the results are broadly representative of the full range of conditions where the model will be
applied.
The individual data items were varied under a range of operating conditions and their impact
elasticities (see Table 4.1) were determined. On the basis of the results of these analyses, the
variables were assigned to the different sensitivity classes shown in Table 4.2 for HDM-4.
For reference purposes, the results from a similar analysis with HDM-III are given in Table
4.3.

Version 1.0
Sensitivity of HDM
Traffic influenced speeds

Traffic Influenced are in HDM-95. HDM-III
Speed only used free speeds.
(km/h) This figure shows the
HDM-95 interactions.
Free Speed Travel Time

Road Geometry
(km/h) (h/1000 km)
Passenger Time
Roughness
(h/1000 km)
Crew Time
(h/1000 km)
Cargo TIme
(h/1000 km)
Power
Fuel Consumption
Requirements
(l/1000 km)
(kW)
Tyre Consumption
(Tyres/1000 km)
Oil Consumption
(l/1000 km)
Parts Consumption
(%NVP/1000 km)
Labour Hours
(h/1000 km)
Service Life Depreciation

(km/veh) (%NVP/1000 km)
Annual Utilisation Interest

(km/yr) (%NVP/1000 km)
Figure 4.1 HDM-III/95 Road User Effects components and their interactions

Version 1.0
Sensitivity of HDM
Driver and Road

Alignment
Traffic Influenced
Traffic Volume
Speed
and Capacity
(km/h)
Non-Motorised
Transport
Free Speed Travel Time

Roadside Friction
(km/h) (h/1000 km)
Road Width,
Geometry, Passenger Time
Pavement Texture, (h/1000 km)
Speed Limit
Crew Time
Roughness
(h/1000 km)
Cargo Time
Acceleration Noise
(h/1000 km)
Overhead Cost Emissions

(cost/1000 km) (kg/1000 km)
Power
Fuel Consumption Oil Consumption
Requirements
(L/1000 km) (l/1000 km)
(kW)
Tyre Consumption
(Tyres/1000 km)
Parts Consumption Labour Hours

(%NVP/1000 km) (h/1000 km)
Service Life Depreciation

(km/veh) (%NVP/1000 km)
Residual Value
Annual Utilisation Interest Depends

(km/yr) (%NVP/1000 km) on Option
Selected
Figure 4.2 HDM-4 Road User Effects components and their interactions

Version 1.0
Sensitivity of HDM
Table 4.2
Sensitivity classes for HDM-4 RUE model1

2 3 4
class elasticity for total VOC for VOC savings
Replacement vehicle price Replacement vehicle price
S-I > 0.50 Parts model exponent kp Parts model exponent kp
Roughness in IRI Roughness in IRI

S-II 0.20 - 0.50 Parts model a1 Parts model a1
Aerodynamic drag coefficient Aerodynamic drag coefficient

Aerodynamic drag multiplier Cdmult Aerodynamic drag multiplier Cdmult
Annual utilisation Annual utilisation
Base fuel-power factor ξb Base fuel-power factor ξb
Cost of fuel Cost of fuel
Cost of maintenance labour Cost of maintenance labour
Cost of passenger work time Cost of passenger work time
Desired speed Cost of tyre
Driveline efficiency Driveline efficiency
Driving power Driving power
Engine accessories power factor Engine accessories power factor
Fuel ξ Engine idle speed
Interest rate Fuel ξ
Labour hours calibration factor Interest rate
Mass density of air Labour hours calibration factor
Maximum average rectified velocity Mass density of air
S-III 0.05 - 0.20 NMT friction factor XNMT Maximum average rectified velocity
Number of passengers Number of passengers
Number of wheels Number of wheels
Parts model a0 Parts model a0
Posted speed limit Posted speed limit
Projected frontal area Projected frontal area
Rolling resistance CR1 Rolling resistance CR1
Rolling resistance CR2 a0 Rolling resistance CR2 a0
Rolling resistance factor CR2 Rolling resistance factor CR2
Roughness-speed a0 Roughness-speed a0
Side friction factor XFRI Speed β
Speed bias correction factor Speed limit enforcement factor
Speed limit enforcement factor Vehicle mass
Vehicle mass Vehicle service life
Vehicle service life Volume of wearable rubber
Wheel diameter Wheel diameter
Notes:
1 The variables listed here are defined in Bennett and Greenwood (1999).
2 This is the highest class found over all common applications.
3 These are the variables important in determining the total RUE.
4 These are the variables important in determining the effect of roughness on RUE.

Version 1.0
Sensitivity of HDM
4.3 Sensitivity classes
4.3.1 High impacts, Class S-I (> 0.5)

For both sets of analyses the results indicate that the only high impact parameters are the
replacement vehicle price and the parts model exponent kp. The former is used to calculate
parts, depreciation and interest costs. The latter governs the magnitude of the parts
consumption and the impact of vehicle age on parts consumption.
Table 4.3
Sensitivity classes for HDM-III RUE model1

2 3 4
class elasticity for total VOC for VOC savings
kp - parts model exponent kp - parts model exponent
New vehicle price New vehicle price
S-I > 0.50 CSPQI - parts model roughness term
C0SP - parts model constant term
Roughness E0 - speed bias correction

E0 - speed bias correction ARVMAX - max. rectified velocity
S-II 0.20 - 0.50 Average service live CLPC - labour model exponent
Average annual utilisation
Vehicle weight
Aerodynamic drag coefficient Beta - speed exponent

Beta - speed exponent Vehicle age in km
BW - speed width effect C0LH - labour model constant term
Calibrated Engine Speed Labour cost
CLPC – labour model exponent Hourly utilisation ratio
C0SP - parts model constant term BW - speed width effects
CSPQI - parts model roughness term Number of tyres per vehicle
S-III 0.05 - 0.20 Crew/Cargo/Passenger cost New tyre cost
Desired speed Lubricants cost
Driving power Crew/Cargo/Passenger cost
Energy efficiency factors Vehicle weight
Fuel cost Number of passengers
Hourly utilisation ratio
Interest rate
Projected frontal area
Notes:
1 The variables listed here are defined in Watanatada et al. (1987a).
2 This is the highest class found over all common applications.
3 These are the variables important in determining the total VOC.
4 These are the variables important in determining the effect of roughness on VOC.
4.3.2 Moderate impacts, Class S-II (0.2 - 0.5)

The moderate impact parameters are the roughness and the parts consumption model
parameter, a1, that governs the roughness effects.

Version 1.0
Sensitivity of HDM
4.3.3 Low impacts, Class S-III (0.05-0.2)

A large number of parameters fall into the low impact category. These cover a range of
attributes and include most of the unit costs. The only major differences between the two
analyses, total costs versus roughness effects, is in the importance of the side friction factor
(XFRI) and non-motorised factor (XNMT) in the total cost analysis. This is because of the
importance of the speeds in the total costs whereas the roughness-speed effects were
important with the roughness costs.
4.3.4 Negligible impacts, Class S-IV (< 0.05)

By far the majority of the HDM-4 data items have a negligible impact on the results. Under
some circumstances this changes; for example, gradient was S-III when the grades were
significant, but for most HDM-4 analyses the default values for these should be adopted.
4.4 Road deterioration and works effects

The RDWE sub-model predicts the deterioration of the pavement over time and under traffic.
When maintenance is applied, the pavement condition is improved and the volumes of
material applied are multiplied by their unit costs to establish the cost of treatment.
In common with the RUE sub-model, there are a series of interdependencies in the modelling
of RDWE. These are illustrated in Figure 4.3. It will be noted that surface distresses have
separate initiation and progression models. This is illustrated in Figure 4.4.
Structural
Crack Initiation Crack Progression Rutting
Deformation
Potholing Roughness
Ravelling Environmental
Ravelling Initiation
Progression Effects
Figure 4.3 RDWE distress interactions

Version 1.0
Sensitivity of HDM
100 Initiation Progression
80
Area Cracked in per cent
60
40
20
0
0 5 10 15 20 25
Age in years
Figure 4.4 Surface distress initiation and progression
The sensitivity of road deterioration and maintenance predictions to variations in individual

input parameters is summarised in the extreme right-hand column (see Table 4.4). The
generalised ranking is made in the same four ranges of input elasticity as are used for RUE
effects. For the RDWE, however, they are based on the impacts on physical conditions as well
as on costs and economic returns because they influence the performance of the pavement, the
timing of maintenance, and the economic returns and likely priority of the various treatments.
4.4.1 High impacts, Class S-I (> 0.5)

The most sensitive inputs include pavement structural variables, traffic and roughness, which
are all measured data items. Variations in the structural variables, Modified Structural
Number or deflection, and annual traffic loading (in million ESA), affect most major results
including periodic maintenance and rehabilitation alternatives and the economic returns. They
are sensitive to this high degree only when the pavement structural adequacy (PSA – see
Section 4.2) is low or moderate compared with the loading (under other conditions when the
PSA is high to moderate the impacts are moderate, that is, Class S-II). Variation of the traffic
volume, which determines the number of users deriving benefits, has strong influence on the
economic returns but low influences (S-II or S-III) on physical impacts. Variation of the
pavement roughness, which affects the unit savings gained by each user, has high impacts on
all economic results and on rehabilitation needs, but little on surface distress (S-III).
4.4.2 Moderate impacts, Class S-II (0.2 - 0.5)

Variations in pavement and surfacing ages have moderate impacts on the needs and timing of
periodic maintenance, and low impact on economic returns. Variation in the amounts of all
and wide cracking have moderate impacts on the economic returns for maintenance, strong
impacts on maintenance needs, but low impacts on economic returns for rehabilitation. The
calibration factors for adjusting predictions of environment-roughness effect, roughness
progression, cracking initiation and cracking progression have mainly moderate impacts
(some high and some low) on maintenance and rehabilitation needs and economic returns.
Variations in the overlay thickness and the unit costs of all treatment types have moderate
impacts on economic returns.

Version 1.0
Sensitivity of HDM
4.4.3 Low impacts, Class S-III (0.05-0.2)

Variations in the amount of potholing and the volume of heavy vehicles (as distinct from the
loading) have high direct impacts on the amount of potholing and the economics of patching,
but generally low impacts on maintenance and rehabilitation needs. Note, however, that the
presence or not of potholing has moderate impacts in the absence of a patching alternative.
Variations in the mean rut depth and standard deviation, and the rut depth progression
adjustment factor, have either low or negligible impacts on cracking maintenance
intervention, roughness intervention, and economic returns. Reseal thickness has low impacts
on economic returns of maintenance.
4.4.4 Negligible impacts, Class S-IV (< 0.05)

Variations in the subgrade compaction and rainfall parameters have low impacts on rut depth
progression and negligible impacts overall on interventions and economic returns. Other
rainfall effects are implicit in the pavement strength (Structural Number or deflection) and are
not linked to the rainfall parameter. Variations in the area of ravelling and the ravelling
progression factor affect potholing but overall have negligible impacts on cracking or
roughness intervention and economic returns. Variations in pavement deflection have
negligible impacts when the Modified Structural Number is also provided.

Version 1.0
Sensitivity of HDM
Table 4.4
Sensitivity classes for RDWE variables
Sensitivity Impact Parameter Outcomes most impacted

class elasticity
Pavement Resurfacing Economic

performance and surface return on
distress maintenance
Structural • • •
number 1/
Modified structural • • •
number1/
S-I > 0.50 Traffic volume •
3/
Deflection • • •
Roughness • •
Annual loading • • •
Age • •
All cracking area • •
Wide cracking area • •

S-II 0.20-0.50 Roughness-environment • •
factor
Cracking initiation factor • • •
Cracking progression •
factor
Subgrade CBR (with SN) •
Surface thickness (with • •

SN)
Heavy axles volume • •
Potholing area • •
Rut depth mean •

S-III 0.05-0.20
Rut depth standard •
deviation
Rut depth progression •

factor
Roughness general factor • •
S-IV < 0.05 Deflection (with SNC) •
Subgrade compaction • •
Rainfall (with Kge) •
Ravelling area •
Ravelling factor •

Version 1.0
Sensitivity of HDM
Notes:
1 Only one of the structural parameters is required, the other two are optional. When the
structural capacity is high relative to the traffic loading these parameters are in class S-II.
4.5 Economic models


Version 1.0
Sensitivity of HDM
5 Adapting data to the model
5.1 Introduction
Data represent the particular information that describes the case being analysed and are the
specific inputs to the model that are required of the user. Examples include traffic volume and
composition, road geometry, pavement type and condition, unit costs, etc. These data items
are thus fundamentally different from the internal parameters and coefficients of the
underlying simulation model which need calibration; an error in a data item means that the
model is analysing a different case from the one intended, much like entering a wrong
destination in a flight instruction.
For example, the accuracy of input data can have substantial impact on the timing of future
interventions, sometimes more important than the deterioration rate. This is because HDM
uses incremental models and the existing condition is the start point for the modelling. This is
illustrated in Figure 5.1 which shows for the same intervention criteria a difference in the
initial condition has markedly different times for future interventions.
Condition Confidence
Interval
Existing
Condition Criteria
for
1 Intervention
Condition Condition
2 1 2
Trigger Trigger
Interval Interval
Time
Figure 5.1 Effect of existing condition on triggering maintenance
Just as information can be either imprecise or very precise, so the HDM model can be
operated with either very simple information or much more detailed information. However,
by the nature of simulation models and software, the input parameters of the model are fixed
and explicit. So this chapter describes how the user adapts various levels of locally available
data (whether simple or complex) to the required input parameters, and later reconverts the
results into local formats, where desired.
Finally, we look at how the user handles the related issues of accuracy and approximation.
These often have significant cost implications for the operation of a road management system
but also have an impact on the reliability of results and decisions.

Version 1.0
Sensitivity of HDM
5.2 Concept of Information Quality Level (IQL)

An item of information can be presented in either simple or detailed terms. Viewed through a
lens, the image of an object from a distance or great height will be seen as an outline and in
general features. Close-up or at low heights, the amount of detail seen increases and other
features or attributes of the object can be identified. The object, or information, is the same
but the quality of information has been enhanced. In some instances the general outline or
overall situation is the quality of information which is required; that is, the high-level or
macro-level information, whereas in other instances the greater detail (micro-level) is what is
required.
The concept of Information Quality Levels (IQL), defined by Paterson and Scullion (1990),
allows us to structure road management information in ways that suit the needs of different
levels of decision making and the variety of effort and sophistication of methods for
collecting and processing data. In the IQL concept, very detailed information at a low level
(low-level data) can be condensed or aggregated into progressively fewer items at
successively higher levels of IQL (high-level data) as shown in Figure 5.2.
HIGH LEVEL DATA
System Performance
IQL-5 Performance
Monitoring
Planning and
IQL-4 Structure Condition
Performance Evaluation
Programme Analysis or
IQL-3 Ride Distress Friction
Detailed Planning
Project Level or
IQL-2 Detailed Programme
Project Detail or
IQL-1
Research
LOW LEVEL DATA
Figure 5.2 Information Quality Level concept
In road management, five levels have been identified for general use, as defined in Table 5.1:
! IQL-1
Represents the following fundamental type data:

❏ Research
❏ Laboratory
❏ Theoretical
❏ Electronic
where many attributes may be measured or identified.

Version 1.0
Sensitivity of HDM
! IQL-2
Represents a level of detail typical of many engineering analyses for a project-level

decision.
! IQL-3
A simple level of detail (simpler than IQL 1 or IQL 2), typically two or three attributes,
which might be used for large production uses like network-level survey or where
simpler data collection methods are appropriate.
! IQL-4
A summary or key attribute which has use in planning, senior management reports, or
alternatively in low effort data collection.
! IQL-5
Represents a top level such as key performance indicators, which typically might
combine key attributes from several pieces of information. Still higher levels can be
defined when necessary.
Table 5.1
Classification of Information Quality Level and detail
Level Amount of detail

1 Most comprehensive level of detail, such as would be used as a reference benchmark for
other measurement methods and in fundamental research. Would also be used in detailed
field investigations for an in-depth diagnosis of problems, and for high-class project
design. Normally used at project-level in special cases, and unlikely to be used for network
monitoring. Requires high level of staff skills and institutional resources to support and
utilise collection methods.
2 A level of detail sufficient for comprehensive programming models and for standard design
methods. For planning, would be used only on sample coverage. Sufficient to distinguish
the performance and economic returns of different technical options with practical
differences in dimensions or materials. Standard acquisition methods for project-level data
collection. Would usually require automated acquisition methods for network surveys and
use for network-level programming. Requires reliable institutional support and resources.
3 Sufficient detail for planning models and standard programming models for full network
coverage. For project design, would suit elementary methods such as catalogue-type with
meagre data needs, and low-volume road/bridge design methods. Able to be collected in
network surveys by semi-automated methods or combined automated and manual methods.
4 The basic summary statistics of inventory, performance and utilisation, of interest to

providers and users. Suitable for the simplest planning and programming models, but for
projects is suitable only for standardised designs of very low-volume roads. The simplest,
most basic collection methods, either entirely manual or entirely semi-automated, provide
direct but approximate measures, and suit small or resource-poor agencies. Alternatively,
the statistics may be computed from more detailed data.
Table 5.3 serves to introduce the concept. At IQL-1, pavement condition is described by
twenty or more attributes. At IQL-2, these would be reduced to 6-10 attributes, one or two
for each mode of distress. At IQL-3, this reduces to 2-3, namely roughness, surface distress,
and texture or skid resistance. At IQL-4, this reduces to one attribute, Pavement condition (or
state or quality) which may have been measured by class values (good, fair, poor) or by an
index (for example, 0-10). An IQL-5 indicator would combine pavement quality with other
measures such as structural adequacy, safety aspects, and traffic congestion; that is
representing a higher order information such as road condition.

Version 1.0
Sensitivity of HDM
From these definitions three observations arise:

! It can be observed that as the decision-level rises, so the IQL that is appropriate also
rises. Information at IQL-4 or IQL-5 is appropriate for performance indicators and road
statistics that are of interest to senior management and the public because they tend to
be, or should be, easily understood without much technical background. At project-
level, however, the appropriate IQL depends much more on the standard of the project
and the resources of the agency:
❏ For a rural road or a small local agency, IQL-3 is usually sufficient, being simple
but effective for the purpose
❏ For most agencies and main roads IQL-2 is typical, but
❏ For expressways or a high-powered well-funded agency IQL-1 may be used in some
instances.
The criterion to use in selecting the appropriate IQL is to ask ‘is the decision likely to be
altered by having more detailed information?’
! The second observation is that primary data collection at a low-level or detailed IQL
typically costs more and involves more complex or sophisticated equipment than
collection of higher IQL data. Thus, the IQL for primary data collection which is
appropriate to a given agency and situation depends on:
❏ Financial and physical resources
❏ Skills
❏ Cost
❏ Speed or productivity
❏ Degree of automation
❏ Complexity
all summed up in the need for the method to be sustainable for the intended purpose,
such as the regular operation of a road management system.
! A third observation is that a higher level IQL represents an aggregation or transformation
of the lower level IQL. When there is a specific rule or formula for conversion, from say
IQL-2 into IQL-3, then the information is reproducible and reliable. Thus, when the
appropriate IQL is chosen, the data can be re-used through transformation to the higher
IQLs as the decision-making moves up the project cycle, this avoids the need for
repeating surveys and saves cost.
5.3 Relating the local IQL to the HDM model

The HDM model, in both versions HDM-III and HDM-4, operates internally at a level that is
primarily IQL-2. This fairly detailed level of modelling was necessitated by the demands to
make the model as universally applicable as possible and that could only be achieved by
adopting fundamental, mechanistic and structured empirical formulations that would operate
as close to first principles as would be practical. The downside of this fairly detailed level is
that many users of HDM-III over the past decade felt the need to collect all the input data at
that same IQL-2 level, or rejected the model as too complex, when this was not in fact
necessary. In HDM-4, some of the more common data input simplifications have been
provided in the form of buttons for built-in conversions, which provide for example an IQL-3
approximation to the IQL-2 internal inputs, so it can be used at either level.

Version 1.0
Sensitivity of HDM
Local data can be adapted for use with HDM by determining a conversion for transforming
the data into the HDM parameters, and vice-versa. The basic approach for adapting local data
to HDM is first to transform the local data (from whichever IQL they may have) into IQL2
input parameters, and later to transform outputs from IQL-2 to the user’s desired output form,
typically IQL-3 or IQL-4. Additional data collection would only be required when essential
information items were missing. The approach is summarised as follows, and illustrated in
Figure 5.3.
1 The local data items are sorted into groups that relate to the HDM input parameter
groups (these groups are addressed in the following four sections).
2 Identify and record the IQL of each local data group (for information purposes), note that
where the IQL of the data appears to be mixed it would be worthwhile to consider a
separate review to normalise them around the most appropriate IQL.
3 If the local data is IQL-1, then the items will need to be combined or aggregated into the
HDM inputs, usually this is a process of selection rather than conversion since the HDM
inputs are directly measurable parameters such as area. These transformations are
external to the model.
4 For local data of IQL-2, there may be a need for conversion of measurement units to the
international units used in HDM, this is external to the model.
5 For local data of IQL-3, the information is less detailed than the HDM inputs; that is, one
attribute needs to be subdivided into two or more attributes. Thus, it is only possible to
estimate what the other attribute values might be and the answer is not unique. The
approximations will use an average or mid-point value, assuming that values above and
below will compensate each other and the final result will be reasonably accurate. The
conversion can be made outside HDM. However, for those items that have an internal
transformation built into HDM-4, the conversion will need to be checked against real
local data to ensure that it is reasonable and to adjust the conversion as needed.
6 For local data of IQL-4, the information is even less detailed than HDM and the
approximation is somewhat greater, but the procedures for adaptation are similar to those
for IQL-3. This is unimportant if the data item is in Sensitivity classes S-III or S-IV, but
deserves review if the data item is S-I or S-II, in which case consideration should be
given to changing the data collection to a more detailed IQL.
7 For adaptation of the output data to local items and terminology, the reverse process may
need to be applied. The same conversion formula should be used (in reverse) to ensure
that the model’s results will be reflected correctly in the local terminology. Adaptation
is generally less of a problem for output data since there are fewer variations in the IQL-
3 and IQL-4 definitions.

Version 1.0
Sensitivity of HDM
Selection of Local External Selection of Output

HDM Model
Data IQL Transformations Results IQL
IQL - 1
A IQL - 2
HDM
HDM Standard
IQL - 2 Detailed
Analysis Reports
Data Entry
IQL - 3
B
IQL - 3 HDM Class
Data Internal Transformations
Options
IQL - 4
IQL - 4
Option A Transform or input local data directly to HDM inputs
Transform or input local data to HDM internal transforms

Option B (Class Data Options)
Figure 5.3 Data conversion process
5.4 Adaptation of local condition data units

When condition data have been collected using different units to those in HDM it is necessary
to develop a transfer function to convert these data. This function can either be a
mathematical relationship or a table.
The function is developed by conducting parallel surveys with both the local and HDM
measures on a common sample of road sections. A total of 15-20 sections for paved roads and
10-15 sections for unpaved roads constitute an adequate sample size for most conversions.
The total length of sections should be at least 20 km for each pavement type (that is,
bituminous paved, rigid paved, unpaved).
An example of the conversion of units from Niger is shown in tabular form in (Paterson,
1986) and in graphical form in Figure 5.4. The data were based on a study of many sections
with a total length of 120 km of paved roads. On each section the values of the appropriate
HDM parameter and the local condition score were recorded. The rating for surface distress
(Enduit) included ravelling, potholing and bleeding collectively. This was matched with
representative areas of ravelling and potholing as shown in Table 5.2. The rating for cracking
(Fissuration) was done from a moving vehicle so it was only related to wide cracks since
narrow cracks cannot be observed in this manner. The local Deformation score included
depressions, average rut depth and the subjective rating of roughness so these were broken
into the individual components.

Version 1.0
Sensitivity of HDM
Ravelling to "Enduit" Score
120
100
80
Area of Ravelling (%)
E-0
E-1
E-2
60 E-3
E-4
E-5
Means
40
20
0
0 1 2 3 4 5 6
"Enduit" (E) score
Figure 5.4 Graphical form of conversion units - Niger investigation

Version 1.0
Sensitivity of HDM
Table 5.2
Conversions between condition units and HDM-III input parameters - example
from Niger office of road management
Distress mode Symbol Approximate conversion between units

Surface distress (Enduit) E 0 1 2 3 4 5
Area of ravelling (%) ARAV 0 0 5 30 70 50
Area of potholing (%)1 APOT 0 0 0 0 0.08 1.0
Cracking (Fissuration) F 0 1 2 3
Area wide cracking (%)2 ACRW 0 5 25 50
Deformation D 0 1 2 3
(deformation)
Area of depressions (%) ADEP 0 5 10 30
Roughness (m/km) IRI 2 3 4 5
Rut depth3 RDM - -1 3 5
Source: Paterson (1986)
Notes:
1 The incidence of potholes is recorded only as 4 or 5.
2 Predominantly only wide cracks are visible from a moving vehicles. If the cracks are
widely spaced, these values should be reduced applying a factor of 0.4.
3 The rut depths were generally indiscernible (< 5 mm), even for high classes of
deformation.
5.5 Transforming road input data

In Table 5.3 the four groups of input data for the road infrastructure:
! Road geometry
! Pavement structure
! Environment
are presented at four levels of IQL:

Version 1.0
Sensitivity of HDM
! IQL-2 - the actual level of direct inputs into HDM-III and HDM-4
! IQL-2B - a special level of simplified inputs for HDM-4
! IQL-3 - class-type information for the key attributes within each data group
! IQL-4 - group-level information
The more detailed level, IQL-1, for each parameter is not shown for reasons of clarity but
comprise the fundamental parameters that describe the detailed attributes of an IQL-II data
parameter.
Table 5.3 shows the recommended grouping and aggregation of data items. Details of
definitions, classification ranges and conversion relationships are given in Appendix F .
5.5.1 Road geometry

HDM-4 requires four primary parameters defining rise and fall and horizontal curvature, plus
five that define the speed regime on the road. Continuous devices would measure the IQL-2
but simpler methods of estimation that are suited to visual means are available to estimate
those for IQL-2B. A more approximate method in IQL-3 is to classify the vertical alignment
into say 4 classes (for example, flat, rolling, moderate, steep) and the horizontal alignment
into say 4 classes (for example, straight, fairly straight, curvy, and winding). The most
aggregate at IQL-4 combines these into say 6-8 combinations of vertical and horizontal
alignment.
5.5.2 Pavement condition

There are 12 data inputs for pavement condition at the full IQL-2, which are typically
measured. Those can be simplified and approximated reasonably well by six indices, one for
each key mode of distress (roughness, cracking, deformation, disintegration, texture and
friction), which are typically estimated by a score or class. A higher level method at IQL-3
simplifies this to three (riding quality class, surface distress index and friction class) which
can all be estimated by a trained observer (or alternatively, may be generated from the more
detailed measures). At IQL-4, for performance indicators, these are combined into one
pavement condition rating, which can be by class values (for example, good, fair, and poor)
or an index (for example, pavement quality index).
5.5.3 Pavement structure

There are numerous parameters to be specified for HDM-4 at IQL-2, for example, 15 for
bituminous pavements, 3 for concrete pavements, and 14 for unpaved roads. These can be
simplified to 8, 2 and 9 respectively for a simpler, class-type method of estimating the inputs
at IQL-2B. At IQL-3, these can be reduced to 3 (structural adequacy, construction quality
and previous intervention) for bituminous pavements, 1 (structural adequacy) for concrete
pavements, and 3 (gravel standard, earth passability and load rating) for unpaved roads. As
performance indicators at IQL-4, these can be summarised to one for bituminous and concrete
pavements, and one for unpaved roads.

Version 1.0
Sensitivity of HDM
Table 5.3
IQL examples for road data
Information Quality Level

IQL-2 IQL-2B IQL-3 IQL-4
Rise and fall Average absolute
(m/km) gradient (%) Gradient class
Number of rises and No. of gradient (4 classes) Geometry class
falls (no./km) changes (no./km) (6-8 classes)
Average horizontal No. of curves by
curvature (deg/km) speed class (class- Curvature class
Super-elevation freq/km) (4 classes)
Road
geometry Speed limit (km/h)
Speed limit Desired speed
enforcement factor (km/h)
Roadside friction Speed environment n/a
(factor) (6 classes)
Non-motorised Speed reduction
transport speed factors
reduction (factor)
Motorised transport
speed reduction
(factor)
Lane roughness Roughness Ride quality (class)

(m/km IRI) (6 ranges)
All cracks area (%

area) Cracking (score, or
Wide cracks area Universal Cracking
(% area) Index, UCI)
Transverse thermal
Pavement cracks (no./km) Pavement condition
condition Ravelled area (% (class)
area)
Surface Distress Index
Potholes number Disintegration (SDI)
(units/lane-km) (score)
Edge-break area
(m2/km)
Patched area (%
area)
Rut depth mean
(mm) Deformation (score)
Rut depth standard
deviation (mm)
Macro-texture depth Surface texture
(mm) (class) Surface friction
Skid resistance Friction (class) (class)
(SF50)
Pavement Pavement type (class) Pavement type (class) Pavement class (class)
structure
… Continued

Version 1.0
Sensitivity of HDM

Pavement Structural
Number (adjusted
SNP) Deflection (mm BB)
Deflection (mm BB)
Thickness of Surface thickness Pavement structural
surfacing (mm) (class) adequacy (index)
Thickness of base Pavement depth
(mm) (class)
Shoulder effect Shoulder type
(factor)
Construction
defects surfacing
Bituminous (index) Remaining service life
Construction quality
Construction (yrs)
(index) Construction quality
defects base (index) (class)
Relative compaction
(%)
Drainage (index)
Pavement Pavement
environment (index) environment (index)
Pavement age (yr)
Surfacing age (yr) Surfacing age
(range)
Previous all Previous intervention
cracking area (%) (class)
Previous wide Previous condition
cracking area (%) (class)
Previous thermal
cracking (number)
Thickness of slab Slab thickness
(mm) (class) Pavement Structural
Concrete Modulus of rupture Adequacy (index)
(MPa) Material strength
Reinforcing steel (class)
(%)
… Continued

Version 1.0
Sensitivity of HDM

Surfacing Material:
Material Type Material Type
(class)
Max. Particle Size Material Size (class)
(mm) Gravel standard
Material passing 2.0 (class)
mm (%)
Material passing Material Gradation
0.425 mm (%) (class)
Material passing
0.075 mm (%)
Unpaved Durability standard
Plasticity Index Plasticity (class)
(class)
Subgrade Material:
Material Type Material Type
(class)
Max. Particle Size Material Size (class)
(mm) Earth passability
Material passing 2.0 (class)
mm (%)
Material passing Gradation (class)
0.425 mm (%)
Material passing
0.075 mm (%)
Plasticity Index Plasticity (class)
Structure:
Gravel depth (mm) Load rating (class)
Cross-section Surface depth
(class) (class)
Rainfall monthly
mean (mm) Rainfall class
Dry season duration
(fraction)
Moisture
classification (class)
Temperature Climate (class)
classification (class)
Temperature mean
annual (deg C)
Environment Climate classification Climate classification
Temperature range
(class) (class)
(deg C)
Time above freezing
(days)
Cold climate
Freezing Index classification (class)
(degC-days)
De-icing salt use
(class)
Studded tire use (%)
Snow-driving time
(%)
Wet-road driving
time (%)
Air density (kg/m2)

Version 1.0
Sensitivity of HDM
5.5.4 Environment
There are 13 parameters defining aspects of environment for the various pavement types at
IQL-2. These can be approximated and estimated by three class parameters at IQL-2B:
! Rainfall class
! Climate class
! Cold climate classification
At IQL-3 and IQL-4, these can all be summarised by one climate classification with several
classes.
5.6 Transforming traffic input data
5.6.1 Traffic volume

Full inputs at IQL-2 involve an AADT-adjusted daily volume for each vehicle class adopted
in the fleet classification (see Table 5.4). These could be approximated at IQL-2B by
applying appropriate, perhaps estimated, percentages for each class to the AADT. An IQL-3
method would provide only two parameters; for example, the AADT volume and the
percentage heavy vehicles. At IQL-4, the volume would be grouped by class for example,
preferably ranging by factors of three (for example, 30, 100, 300, 1000, 3000, 10,000, etc.)
because these also approximate to key decision thresholds for capacity.
5.6.2 Traffic flow

Full inputs at IQL-2 range from 14 to 20 measured parameters, depending on the fineness of
the flow-bands chosen. These could be simplified to seven class parameters that could be
estimated at IQL-2B. At IQL-3, these would be summarised by two parameters (volume-
capacity ratio and flow-type class). An IQL-4 measure could be a congestion classification,
or a performance indicator such as veh-hours delay per day.

Version 1.0
Sensitivity of HDM
Table 5.4
IQL examples for traffic data

Volume (AADT Volume (AADT
veh/day) veh/day)
Volume – veh. Percent veh. class 1

class 1 (veh/day)
Volume – veh. Percent veh. class n

class n (veh/day)
Traffic Volume – veh. Percent veh. class 5 Traffic class (class)

volume class 5 (veh/day)
Heavy traffic (%
Volume – veh. Percent veh. class n AADT)
class n (veh/day)
Volume growth
(%/yr) – class 1
Volume growth Volume growth (%
Volume growth (%/yr.) (all classes) AADT/yr)
(%/yr) – class n
Ultimate capacity Ultimate capacity

(pcse/la/h) (class)
Free-flow capacity Free-flow capacity

(%) (class)
Nominal capacity Nominal capacity

(fraction) (class)
Volume-capacity ratio
Jam speed at Jam speed at capacity
capacity (km/h) (class)
Acceleration noise Acceleration noise

Traffic flow max. (m/s2) max. (class) Congestion class
(class)
Flow-frequency
periods (no.)
Duration of F-F
period 1 (h)
Duration of F-F Flow type (3 classes)

period n (h)
Flow type (3 classes)
Percent AADT in
F-F period 1
Percent AADT in
F-F period n
Intersection type Intersection type

(class) (class)
Safety To be completed
Emissions To be completed

Version 1.0
Sensitivity of HDM
5.6.3 Traffic safety

5.6.4 Vehicle emissions

5.7 Transforming output data and producing performance

Indicators
For data groups that are also used as output data, the parameters defined for IQL-4 are
generally suitable for use as performance indicators, such as the physical condition and
functional condition of the assets; for example, pavement condition, traffic congestion, etc.
Other indicators that are important are those relating to the effectiveness of the work
alternatives being evaluated, for example, NPV per unit cost, etc.
5.8 Data accuracy

The accuracy of measurement of any of the parameters mentioned in this chapter would be a
function of the data collection method and the quality control applied. In general, for the
multiple analyses that are typically done with HDM for road management, the control of bias
(that is, the reproducibility of the measurement by different instruments or at different times)
is more important than the precision (the repeatability of the measurement between
successive runs). Thus, calibration of the equipment being used and independent verification
of non-equipment-based methods is vital for controlling the potential for bias in the input
data.
Using the sensitivity function in HDM can test the impact of a bias error in the input data. An
approximate estimate can be gained from the sensitivity classification given earlier in this
book.
When an IQL-3 or IQL-4 method is used to estimate the HDM input parameters, average
values need to be estimated for the various parameters that have been combined into the
simplified ones that are collected. If all the values range widely it is unlikely that there will
be much bias. However, it is useful to conduct a verification exercise that compares a
substantial sample of the simple measures with the more detailed, formal parameters
measured on the same road section or in same traffic streams.
5.9 Checklist for data adaptation

The steps for choosing an appropriate IQL for local data and establishing a means for
adapting those data to the HDM model inputs include the following:
! Choose or verify that the local IQL is appropriate to the needed decision level and
available collection resources
! Sort the local data into a format suitable for transformation to the HDM inputs (IQL-2)
! Determine a suitable transformation between the local data and HDM-input data, using:
❏ desk-top estimates, or
❏ field conversion sites, or
❏ determining conversion relationships

Version 1.0
Sensitivity of HDM
! Apply transformation using conversion relationships

Version 1.0
RUE model calibration
6 RUE model calibration
6.1 Introduction
The RUE model calibration focuses on ensuring that the key RUE model parameters and
calibration factors are appropriate for the conditions under which the model is to be applied.
As described in Chapter 2, there are three levels of calibration that entail different levels of
resources and time. These are:
! Level 1 - Basic application (see Section 6.3)
Determines the values of required basic input parameters, adopts many default values,
and calibrates the most sensitive parameters with best estimates, desk studies or minimal
field surveys.
! Level 2 - Calibration (see Section 6.4)
Requires measurement of additional input parameters and moderate field surveys to

calibrate key predictive relationships to local conditions. This level may entail slight
modification of the model source code.
! Level 3 - Adaptation (see Section 6.5)
Undertakes major field surveys and controlled experiments to enhance the existing
predictive relationships or to develop new and locally specific relationships for
substitution in the source code of the model.
Figure 6.1 shows the recommended priorities for the RUE calibration. This shows the data
which are required, of first and second priority, and which should have the defaults assumed
for.
Assume
Required Priority 1 Priority 2
Defaults
Mass
Power
Capacity
Tyre Parameters
Level 1 Unit Costs Speed
Depreciation
All Others
Service Life
Parameters
Utilisation
Priority 1 Priority 2 Priority 3
Fuel
Speed
Power
Level 2 Capacity
Service Life
All Others
PCSE
Utlilisation
Priorities
Depend on
Level 3 Resources and
Objectives
Figure 6.1 Calibration priorities

Version 1.0
This chapter commences with a discussion of selecting representative vehicles. This is then
followed by procedures for calibrating the model parameters to each of the three levels.
6.2 Selecting representative vehicles

As it is not possible to model the operating costs of individual vehicles, the analyst must
resort to the use of representative vehicles. These are vehicles whose characteristics can be
considered to be representative of all vehicles within a certain class. Depending upon the
nature of the study, any number of representative vehicles may be adopted. The number will
be influenced by factors such as the composition of traffic, functional differences between
different types of vehicles, the objectives of the study, and the availability and quality of data.
Table 6.1 lists vehicles adopted in HDM-III studies from a number of different countries
Bennett and Greenwood (1999).
HDM-4 allows an unlimited number of representative vehicles, and using a large number of
representative vehicles to describe the vehicle fleet may seem to be advantageous. However,
the intrinsic difficulties associated with accurately describing the characteristics of the
representative vehicles and the continually changing composition of the fleet means that there
will always be errors, irrespective of the number of vehicles selected. It is therefore
recommended that for most studies a minimum of seven vehicles be used:
! Motorcycle
! Small passenger car
! Utility/light commercial vehicle
! Light truck
! Medium/heavy truck
! Minibus
! Heavy bus
If adequate data are available, the medium and heavy trucks should be further disaggregated
since these vehicles tend to have the widest variations in the RUE and the greatest impact on
pavement performance.
6.3 Level 1 - Basic application
6.3.1 Mass and vehicle damage factor: S-II
Mass
The vehicle mass influences the vehicle speeds, fuel and tyre consumption and, through the
associated heavy vehicle damage factor, has a major impact on the rate of pavement
deterioration.
The influence of mass on RUE is not a major issue for pavements in flat terrain. However, the
presence of gradients will result in a major increase in fuel consumption and this increase is
proportional to the vehicle mass.

Version 1.0
Table 6.1
Representative vehicles adopted in different HDM-III studies
Country MC PC LDV LGV LT MT HT AT LB MB HB Comments

S M L
Barbados • • • • • • •
Botswana • • • • • • • •
Canada • • • • • • • • •
Ethiopia • • • • • • • • LGV = 4WD
India • • • • • • • •
India • • • • •
India • • • • • • • •
India • • • • • •
Indonesia • • • • • • •
Indonesia • • • • • • • •
Lesotho • • • • • • •
Malaysia • • • • • •
Myanmar • • • • •
Myanmar • • • • • • • Private/Gov’t HT
Nepal • • • •
Nepal • • • • Empty/Full MT
New Zealand • • • •
New Zealand • • • • • • • • •
Romania • • • • • •
Tanzania • • • • •
Thailand • • • • • •
Trinidad • • • • • • •
Uganda • • • • • • • LDV = 4WD
Uganda • • • • •
South Africa • • • • •

Version 1.0
It is impossible to obtain a reliable estimate of vehicle mass without conducting a field

survey, which is a Level 2 calibration activity. This is particularly so in countries with poor
axle load enforcement practices: vehicles will often be overloaded well beyond the
manufacturer’s rated GVW.
If it is impossible to measure weights, a Level 1 calibration of the average mass should be
estimated from the manufacturer's tare and rated gross vehicle weights. The analyst must
estimate the percentage of vehicles that are travelling empty, half-full, full and overloaded.
The average mass is then calculated as:
[Pe TARE + Ph (0.5 TARE + 0.5 GVW ) + Pf (GVW ) + Po (zo GVW )]

M= …(6.1)
100
where:
M is the average vehicle mass (km)

TARE is the vehicle tare (empty) mass (kg)
GVW is the manufacturer's gross vehicle mass (kg)
Pi is the percentage of vehicles empty, half-full, full and overloaded (%)
zo is the overloaded weight relative to GVW (as a decimal)
With this approach, for container trucks standard container weights of 2 t and 4 t can be
assumed.
Vehicle Damage Factor

The vehicle damage factor (VDF) is a measure of the damage caused to the pavement by a
heavy vehicle. It is a function of the axle configuration and its mass. The VDF is calculated
using the equation (Watanatada et al., 1987a):
n
∑
4
 AX i 
VDFVEH k =   …(6.2)
 SX i 
i=1
∑ VDFVEH
k =1
k
VDF = …(6.3)
z
where:
VDFVEHk is the vehicle damage factor for vehicle k (ESA/vehicle)

VDF is the vehicle damage factor for a stream of vehicles (ESA/vehicle)
AXi is the load on axle i (tonnes)
SXi is the standard axle load for the axle group j (tonnes)
n is the number of axles on the vehicle

Version 1.0
z is the number of vehicles in the stream
It is common to divide the stream into similar vehicle class, for example, medium, heavy and
articulated trucks, and calculate a VDF for each class.
The standard axle loads (SXi for different configurations are (Watanatada et al., 1987c):
6.60 tonne single wheel, single axle

8.16 tonne dual wheel, single axle
9.00 tonne dual wheel, per tandem axle
15.1 tonne dual wheel, per tandem axle group
10.0 tonne dual wheel, per triple axle
22.9 tonne dual wheel, per triple axle group
There are several points to recognise with regard to the VDF:

! Because of the 4th power for the exponent in the VDFVEH equation (see Equation 6.2
above), the VDF must be calculated as the average VDF per vehicle as opposed to the
VDF for a vehicle with an average mass. The average VDF is always higher than the
VDF of the average load.
! The VDF is the sum of all axle load factors for all axle groups of a vehicle, not the
damage factor of the average axle load.
! The VDF must represent the average VDF of all vehicles in the class in the traffic,
inclusive of empty, partially-laden and fully or over-laden vehicles.
Since the calculations use the 4th power rule, some vehicles with high axle loads may have
very high values for their VDFVEH. Care should be taken about including these when doing
surveys, particularly with small samples, since they may distort the results. High VDFs may
also be rendered invalid if there are changes in policy; for example, axle loads enforcement.
For a Level 1 calibration the VDF is estimated using the assumed percentages of vehicles
empty, half-full, full and overloaded. A sensitivity analysis should be conducted with HDM to
test the sensitivity of the results to the estimated values.
It must be emphasised that the Level 1 approach of estimating the vehicle mass and VDF will
yield only the coarsest values. It is only by conducting a field survey that an accurate estimate
of these parameters can be obtained. These surveys are discussed in Appendix E .
6.3.2 Capacity and speed flow data
Introduction
HDM-4 and HDM-95 uses the speed-flow model proposed by Hoban (1987) and shown in
Figure 6.2. This requires the user to provide five key parameters:

Version 1.0
! Qult - the ultimate capacity of the road
! Qnom - the nominal capacity of where all vehicles are travelling at the same speed
! Qo - the flow where interactions commence
! Snom - the speed at nominal capacity
! Sult - the speed at ultimate capacity
Table 6.2 gives the default HDM parameter values. In the discussion that follows the term
capacity is applied to the ultimate capacity, unless otherwise noted.
S1
S2
Speed in km/h
S3
Snom
Sult
Qo Qnom Qult
Flow in pcse/h
Figure 6.2 HDM speed-flow model
Table 6.2
Default HDM speed-flow model parameters
Road type Width Qo/Qult Qnom/ Qult Sult

Qult
(m) (PCSE/h) (km/h)
Single lane road <4 0.0 0.70 600 10
Intermediate road 4 to 5.5 0.0 0.70 1800 20
Two lane road 5.5 to 9 0.1 0.90 2800 25
Wide two lane road 9 to 12 0.2 0.90 3200 30
Four lane road >12 0.4 0.95 8000 40
Estimating capacity
The capacity of a facility is defined as the maximum hourly flow rate at which vehicles can
reasonably be expected to traverse a point or uniform section of road under prevailing road,
traffic and control conditions. While often considered to be a single fixed value, there are

Version 1.0
always inherent variations due to factors such as driver behaviour, vehicle performance and
prevailing conditions. One should therefore calculate an average capacity that will
encompass these variations.
It is beyond the scope of this report to address capacity issues in detail and readers should
refer to reports such as the Highway Capacity Manual (HCM) for a fuller discussion (TRB,
1997). McLean (1989) gives a good discussion of capacity issues and provides insight into
the background behind the two-lane highway values adopted in the HCM.
For a Level 1 calibration the values in Table 6.2 should be adopted as defaults. However, it
must be appreciated that these are ideal capacities and they must therefore be adjusted to
reflect typical operating conditions. For example, the HCM reduces the ideal capacity to
account for different directional splits (for example, 60/40 instead of the ideal 50/50) and the
presence of no-passing zones. These reductions can be quite significant, reducing the ideal
capacity by 20 per cent or more.
Table 6.3 to Table 6.5 give capacity reduction factors from the Highway Capacity Manual
(TRB, 1997) for two-lane highways. These factors should be applied in a Level 1 calibration,
albeit with caution as they may not be entirely appropriate for conditions outside of those
from which they were developed. The ideal capacity is multiplied by these factors to reduce
it to a value that reflects actual operating conditions. Multi-lane factors are given in TRB
(1997).
Table 6.3
Two-lane highway capacity reduction factors - no passing zones
Terrain Reduction Factor by Percentage No Passing Zones

0 20 40 60 80 100
Flat 1.00 1.00 1.00 1.00 1.00 1.00
Rolling 0.97 0.94 0.92 0.91 0.90 0.90
Mountainous 0.91 0.87 0.84 0.82 0.80 0.78
Table 6.4
Two-lane highway capacity reduction factors - directional split
Reduction factor by directional split

100/0 80/10 80/20 70/30 60/40 50/50
0.71 0.75 0.83 0.89 0.94 1.00

Version 1.0
Table 6.5
Two-lane highway capacity reduction factors - lane and shoulder width
Usable shoulder Capacity reduction factor by lane width (m)

width 3.6 3.3 3.0 2.7
(m)
> 1.8 1.00 0.94 0.87 0.76
> 1.2 0.97 0.92 0.85 0.74
> 0.6 0.93 0.88 0.81 0.70
0 0.88 0.82 0.75 0.66
Note: Where shoulder width is different on each side use the average width.
6.3.3 Average service life: S-II

A vehicle, or any physical property, has three measures of its life, namely the:
! Service life - is the period over which the vehicle is operated
! Physical life - is the period which the vehicle exists (even if it is not being used)
! Economic life - is the period which the vehicle is economically profitable to operate
With HDM, the service life is of interest. It is used by HDM in calculating the depreciation
costs of vehicles that can have a significant impact on the RUE.
In HDM-4 the user needs to define the expected service life in kilometres for a vehicle
operating on a smooth pavement. This value is then used to determine the effect of roughness
on service life when using the Optimal Life technique. This expected service life is the
distance at which it becomes appropriate to scrap the vehicle.
There are a number of different techniques available for calculating the service life and an
overview of these may be found in Winfrey (1969). For a Level 1 calibration the ages of a
sample of vehicles should be obtained, either from a small survey or by sampling
advertisements of vehicles for sale. Daniels (1974) indicates that the service life will be
double the mean age. This was also found to be the case in New Zealand (Bennett, 1985)
where several different techniques were tested for estimating the service life.
6.3.4 Vehicle utilisation: S-II/S-III
Annual utilisation - number of kilometres driven

Annual kilometreage; that is, the number of kilometres driven per year - data are used in
calculating the parts consumption and the interest costs.
In order to determine the annual kilometreage it is necessary to have information detailing the
ages of vehicles and the distances that they have travelled. The utilisation of a vehicle
generally varies with age. In several studies older vehicles have been found to have lower
utilisation than newer ones (Daniels, 1974; Bennett, 1985). It is therefore important that any
data collected not be biased in favour of vehicles of a given age.
For a Level 1 calibration a suitable data source for utilisation is advertisements of used
vehicles for sale. Newspapers and other similar sources provide data on the year and total

Version 1.0
kilometreage of a sample of used vehicles. Dividing the kilometreage by the age gives the
average kilometreage of the vehicle over its life. Problems may be encountered with older
vehicles wherein the odometer may have cycled beyond 100,000 km (or miles) or the ages
may be in error. However, if a sufficiently large sample of data is obtained these problems
will be minimised.
Annual hourly utilisation - the number hours per year

There are three definitions for the hourly utilisation:
! HAV - the number of hours the vehicle is available per year
This is the number of hours per year (8760), less the time allowed for crew rest, time lost
loading, unloading, refuelling, finding cargo, repairs, etc.
! HRD - the numbers of hours driven
This is the hours that the vehicle is operated. It can be calculated from the annual
kilometreage divided by the average annual speed.
! HWK - the number of hours worked
This is similar to the hours driven (HRD), except it includes the time spent loading,
unloading and refuelling.
The HDM-III adjusted utilisation model was based on the hours drove. HDM-4 is based on
the hours-worked approach.
Using a standard working week, a vehicle is typically available for approximately 1800 hours
per year. However, since there are substantial periods of time when the vehicle is not in use,
for example due to loading/unloading, the driving time would often be less than 50 per cent of
this value. In the Brazil study, for example, the vehicles were available 839 - 2414 h, but only
driven 652 - 1863 h (Watanatada et al., 1987c). Trucks and buses had the highest utilisations;
utilities the lowest.
As discussed in Bennett (1995), it is important that the value adopted for hours driven be
consistent with the annual utilisation and the average speed. If not, the predicted costs could
be distorted. To this end it is recommended that in the absence of more detailed data, the
hours driven be calculated using the following equation:
AKM
HRD = …(6.4)
S0
where:
AKM is the average annual utilisation in km

S0 is the average operating speed in km/h
To calculate hours worked it is necessary to have sufficient data to identify the times spent
“undertaking the essential tasks of making a complete round trip, in normal circumstances.
Time spent idle, where the crew is eating, sleeping or otherwise resting should not be
included. Time in repair should also, in general, be excluded because it is not part of a regular
trip. Driving, loading and unloading should be included together with refuelling” (Hine,
1996).

Version 1.0
One may also include the administration time that a driver had to spend finding loads or the
time that must be spent waiting with the vehicle to move it up the queue, if the driver can’t do
anything else. It should be appreciated that in some circumstances not all loaded time will be
working time, for example if the driver stops the vehicle to sleep while it is loaded.
Working time is established by conducting small surveys covering vehicle activities over
several days. The work of Hine (1996) in Pakistan gives valuable insight into issues arising
from these surveys and special considerations that must be made in approaching the issue.
Percentage private use

The percentage private use can only be established through a small survey of users. This will
differentiate between those on work trips and private trips. The resulting values are used to
calculate the value for travel time.
6.3.5 Speed prediction model parameters: S-II/S-III

The HDM speed prediction model is mechanistic, being based on physical and kinematic
principles, as well as behavioural constraints. Consequently, the basic physical model is
highly transferable and the focus of Level 1 calibration should be on the behavioural
constraints, defined by:
! VDESIR - the desired speed of travel
This can be expected to differ considerably between countries, and even regions within
the same country. VDESIR represents the maximum speed of travel adopted by the
driver of a vehicle when no other physical constraints, such as gradient, curvature,
roughness or congestion, govern the travel speed. The value of VDESIR is influenced by
factors such as speed limits and enforcement, road safety, cultural and behavioural
attitudes.
! β - the ‘draw down’
Indicates how far from the constraining speeds the predicted speed will be. β is the
Weibull Shape Parameter. As described in Watanatada et al. (1987c), it is functionally
related to the dispersion of the underlying distribution of the constraining speeds.
This assumes that the physical performance of the vehicle has been properly calibrated by
identifying valid representative values of other important vehicle characteristics, namely:
! Vehicle mass
! Used driving power
! Braking power
For calibration purposes it is important to understand the role of β. The HDM speed model
predicts that speeds are the probabilistic minimum of five constraining speeds based on:
! Power
! Braking
! Curvature
! Roughness
! Desired speed
As shown in Figure 6.3 (Bennett and Greenwood, 1999), when β approaches zero, the mean
speed for a given road section would be equal to the minimum of the five constraining speeds.

Version 1.0
The greater the value for β, the further away the predicted mean speed will be from the
constraining speeds. Thus, if the desired speed is 100 km/h, with a value of β = 0, the
predicted speed would be 100 km/h. However, if a non-zero value for β is used, the predicted
speed would be less than 100 km/h.
In calibrating the model, Watanatada et al. (1987c) found value of β=0.24-0.31 for Brazil. A
similar calibration from India found β=0.59-0.68. Watanatada et al. (1987c) considered that
the higher values for β reflected the more congested traffic conditions in India. It was
recommended that “For environments, such as India, which have congested rural roads and
low level of traffic discipline, somewhat higher values of the β parameter (than Brazil) may
be used”. They note that the quantification of β would require a major field study, that is, a
Level 2 calibration, so it is discussed later in Section 6.4.1.
30
β = 0.01
25
β = 0.2
β = 0.4
20
Speed in m/s
15 β = 0.6
10
0
-10 -8 -6 -4 -2 0 2 4 6 8 10
Gradient in per cent
Figure 6.3 Effect of β on predicted speed
In undertaking a calibration of VDESIR, it is therefore suggested that this be considered in

tandem with the value of β. If the default values of β result in values for VDESIR which are
markedly higher than the predicted speed, unless there is heterogeneous traffic such as in
India, the value for β should be increased from the default.
For a Level 1 calibration when the value of travel timesaving is not required, subjective
estimates of the average free speeds of each vehicle class are adequate. Ideally, these would
be for several road sections with different conditions, but one will suffice.
The HDM model should be run on a road section with average characteristics reflecting the
conditions for the average speed estimates. The predicted free speed should be compared to
the average speed estimate and the value for VDESIR used in the modelling adjusted by the
ratio of the predicted speed to the estimated speed. After several runs of the model the
predicted and estimated speeds should be the same. If a number of road sections are to be
used in the analysis VDESIR should be iteratively established using this method for each

Version 1.0
section and then these values averaged1. As described above, it may also be necessary to
adjust the value for β to ensure a reasonable value for VDESIR.
6.3.6 Vehicle driving power: S-III

The driving power only has a significant effect on speeds when the gradient is positive and
higher than about 4 per cent for light vehicles and 2-3 per cent for heavy vehicles. For a Level
1 calibration the used power can be estimated from the vehicle attributes using the following
equations:
Diesel vehicles
HPDRIVE = 0.70 HPRATED …(6.5)
where:
HPDRIVE is the used driving power (kW2)

HPRATED is the SAE maximum rated engine power (kW)
This equation is from Watanatada et al. (1987c) and so applies to older technology vehicles
(that is, pre-1985). Bennett (1994) indicated that for modern vehicles the factor should be
0.75. His data showed that larger vehicles tended to use more power and the following
alternative equation was developed relating power to mass:
HPDRIVE = (8.5 x 10 -6 M + 0.53 ) HPRATED …(6.6)
Petrol vehicles
HPDRIVE = 1.8 HPRATED -0.3 Pre-1985
HPDRIVE = 2.0 HPRATED -0.3 Post-1985
The modern technology petrol equation is an update of the old technology equation from
Watanatada et al. (1987c) using data from Bennett (1994). For use in HDM-III the values will
need to be divided by the factor of 0.736 to convert to MPH.
The SAE rated engine power is generally available from motor publications and
manufacturers. A weighted-average should be calculated based on the frequency of different
vehicle types, and thus engine powers, in the vehicle fleet.
In applying these equations it should be noted that there is evidence that there is a
relationship between the gradient and power usage, with vehicles having higher power usage
on higher grades (Bennett, 1994). It may therefore be necessary to increase the driving power
from those predicted using the above equations for vehicles operating in hilly conditions.
1
If there are marked differences in the values of VDESIR it may be appropriate to eliminate the greatest
outliers - as well as reviewing the estimated speeds.
2
Engine power conversions: 1 BHP = 0.746 kW; 1 MPH = 0.736 kW; 1 BHP = 0.987 MPH

Version 1.0
6.3.7 Tyre type, wheel diameter and number of wheels: S-IV

The tyre type and the number of wheels are used in HDM-4 for establishing the rolling
resistance. Bias ply tyres have greater rolling resistance than radial tyres, and the resistance
increases with an increasing wheel diameter and number of wheels.
Tyre sizes have a standard typology. The two most common types are shown in Figure 6.4
along with a description of what each term means. The top typology is common with truck
tyres and is based on the nominal section width being expressed in inches. The second is used
with for all vehicles and has the nominal section width in mm along with the aspect ratio1.
For clarity, the section width is separated from the aspect ratio by a slash (/)2. The discussion
that follows is based on the second definition, that is, a metric nominal section width. The
imperial section width can be converted from inches to mm using the factor 1” = 25.4 mm.
10.00 R 20 14 (G)
Load Range
Ply Rating
Rim Diameter in Inches
Radial Construction
Nominal Section Width in Inches
285/75 R 24.5 14 (G)

Load Range
Ply Rating
Rim Diameter in Inches
Radial Construction
Aspect Ratio
Nominal Section Width in Millimetres
Figure 6.4 Standard tyre typology
The ply rating and load range usually only apply to heavy vehicle tyres so the standard
typology reduces to:
xxx/yyRzz …(6.7)
where:
xxx is the nominal width of the tyre (mm)

yy is the aspect ratio
zz is the rim size (inches)
For light vehicles the aspect ratio of 82 is often omitted so they are specified, for example, as
175/R13 instead of 175/82R13. The value of 82 can be assumed for heavy vehicles when
omitted.
1
The aspect ratio is the tyre’s section height, which is the distance from the bead to the centre of tread, to the
section width. An aspect ratio of 65 means that the tyre’s section height is 65% of the tyre’s section width.
2
A less common typology uses the nominal width in inches and the aspect ratio separated by the slash, for
example, 14/80R20. Because of the much lower magnitude of the nominal width, it is readily apparent
when this case arises that the data should be converted to mm.

Version 1.0
The tyre type can be established from manufacturer’s specifications since motorists tend to
purchase the same replacement tyres as new vehicles. Discussing with tyre retailers can check
this information. Alternatively, a small study can be conducted which records the tyres on a
sample of vehicles, for example in a parking lot or at a truck stop.
For Level 1 calibration the wheel diameter can be estimated from the tyre typology using the
following equation (Greenwood, 1997):
yy
DIAM = 25.4 zz + 2 xxx …(6.8)
100
The number of wheels can be estimated from the manufacturer’s specifications or from a
small study.
6.3.8 Volume of wearable rubber: S-III

The volume of wearable rubber is used in the tyre consumption modelling and dictates the
tyre life. The costs are directly proportional to the volume of wearable rubber. For a Level 1
calibration it is estimated from the tyre size using the following procedure (Bennett, 1998).
With reference to Figure 6.5, the tread cross-section is treated as a trapezoid with the upper
and lower bases representing the upper and lower parts of the tread (the lower part being in
contact with the road).
Tyre Casing
Length of Upper Trapezoid

Base (LU)
Grooves
Minimum Legal
Tread Depth
Maximum Depth
of Grooves
Length of Lower Trapezoid
Base (LL)
Figure 6.5 Factors influencing volume of wearable rubber calculation
The area or wearable rubber is calculated as:
LL + LU
AREA = (DE − DEmin ) ARUB …(6.9)
2
where:
AREA is the area of the tread cross-section (mm2)

LL is the length of the lower trapezoid base; that is, at road contact (mm)
LU is the length of the upper trapezoid base (mm)
DE is the depth of the grooves (mm)
DEmin is the minimum depth of grooves (mm)
ARUB is the area of rubber versus grooves (decimal)

Version 1.0
The volume of wearable rubber is calculated from the tread area as:
PI DIAM AREA
VOL = …(6.10)
1,000,000
where:
VOL is the volume of wearable rubber (dm3)

DIAM is the tyre diameter (mm)
To calculate the tyre volume it is therefore necessary to predict the tyre diameter, the tread
depth, the tread width and the area of rubber versus grooves.
The tread width as a function of the nominal section width is predicted as (Bennett, 1998):
LL = a0 xxx - a1 …(6.11)
where:
LL Tread width corresponding to the lower length of the trapezoid base

a0, a1 Equation coefficients (see Table 6.6)
It was found that the coefficients for Equation 6.11 above varied by vehicle class so separate
equations were developed for light and heavy vehicles. The coefficients are given in Table
6.6.
The tread width LL corresponds to the lower length of the trapezoid base in Figure 6.5. For
the upper width (UL) Bennett (1998) recommended an increase of 10 mm for heavy vehicles;
6 mm for light vehicles and 4 mm for motorcycles. Table 6.6 gives these recommended values
by vehicle class.

Version 1.0
Table 6.6
Wearable rubber model parameters
Vehicle class Tread width Tread Width of Area of

model depth UL over rubber
parameters (DE) LL (ARUB)
a0 a1 (mm) (mm) (decimal)
Motorcycles 1.05 -52.5 5 4 0.90
Passenger cars 1.05 -52.5 8 6 0.85
Utilities 1.05 -52.5 8 6 0.85
4WD 1.05 -66.7 9 6 0.85
Light trucks 1.05 -66.7 11 10 0.80
Medium trucks 1.05 -66.7 15 10 0.70
Heavy trucks 1.05 -66.7 17 10 0.70
Articulated trucks 1.05 -66.7 17 10 0.70
Light buses 1.05 -66.7 11 10 0.80
Medium buses 1.05 -66.7 15 10 0.70
Heavy buses 1.05 -66.7 15 10 0.70
Heavy truck driven axles 1.05 -66.7 26 10 0.70
Heavy truck non-driven axles 1.05 -66.7 15 10 0.70
Heavy truck trailers 1.05 -66.7 12 10 0.70
Super single tyres 0.76 -3.8 16 10 0.70
Tread depths are available from manufacturer’s specifications, or can be easily measured
from a sample of tyres, or the values given in Table 6.6 can be used. This table also contains
the values for ARUB: the tread area as a decimal (Bennett, 1998). The latter can be easily
measured by recording the groove widths and frequency of grooves on a tyre.
The volume of rubber is calculated as follows:
1 Establish the tyre size for the representative vehicle (for example, 175SR13).
2 Calculate the tyre diameter from the tyre typology using Equation 6.8 above.
3 Calculate the tread width LL from the tyre typology using Equation 6.11 above.
4 Establish the value for UL relative to LL from Table 6.6 (for example, for passenger cars
UL = LL + 6).
5 Establish the value for ARUB from Table 6.6 (for example, for passenger cars ARUB =
0.85).
6 Establish the tread depth DE from a survey or from Table 6.6.
7 Establish the legal minimum tread depth DEmin or, if there is no enforcement, assume a
value of 0 mm.

Version 1.0
8 Substitute the values for ARUB, DE, DEmin, LL and LU into Equation 6.9 above to
establish the tyre area.
9 Substitute the tyre area and diameter into Equation 6.10 above to establish the volume of
wearable rubber.
6.3.9 Depreciation parameters

HDM-4 contains two methods for depreciation predictions:
! Constant life
! Optimal life
With the constant life method, the depreciation is calculated as the replacement value less
residual value1 divided by the vehicle life. When less than 50 per cent of the trips are for
private use, this life is adjusted by the number of hours worked. The calibration of this
method is therefore achieved using the calibrated values for annual utilisation, service life
and hours worked.
For the optimal life calibration, the user needs to establish the average service life of the
vehicle at an average roughness. Using the HDM Tools software (See Appendix K ), the
service life as a function of roughness is predicted. A regression equation is then fitted to
these data and used in HDM-4. Details of how this is done are given in Appendix K .
6.3.10 Aerodynamic drag coefficient and projected frontal area: S-III

The aerodynamic drag coefficient and projected frontal area should be considered in tandem
since their product is used to calculate the aerodynamic resistance. Aerodynamic drag
coefficients are difficult to obtain since the values reported in the literature are highly
dependent upon the testing conditions. Aerodynamic drag coefficients are also influenced by
the operating conditions; the value changes as a function of the wind angle (Biggs, 1987). The
default values given in HDM-4 should thus be used unless more appropriate values are
readily available.
The projected frontal area is obtained by subtracting the area under the vehicle from the
product of the maximum height with the maximum width. For light vehicles manufacturer's
specifications may be used to determine the appropriate measurements. For trucks and buses
it is recommended that the data be gathered in a small field survey due to the wide variety of
body configurations and loading practices.
In HDM-4 there is an additional value termed the CD Multiplier (CDmult). This is used to
reflect the effect of wind on the aerodynamic drag coefficient (CD) with Bennett and
Greenwood (1999) discussing its quantification. Its calculation is done using the HDM Tools
software that is described in Appendix K . The CD Multiplier does not need to be
requantified unless there is a significant change in the value for CD.
6.3.11 Braking power: S-IV

In HDM-III the braking power was used to establish the speed on downgrades. HDM-4
differentiates between speeds on short and long grades (NDLI, 1995a). On short grades, the
gradient has no effect on speed. When the gradient exceeds a critical length the speed is
governed by the braking power so the vehicle slows down to maintain control.
1
The residual value is assumed to be 1% of the replacement vehicle price.

Version 1.0
For a Level 1 calibration it can be assumed that the default HDM values are adequate,
although if the analyses include steep gradients it may be desirable to calculate new values.
This can be done using the following equations:
HPBRAKE = 10.3 GVW HDM-III …(6.12)
HPBRAKE = 9.3 GVW + 13 HDM-4 …(6.13)
where:
HPBRAKE is the braking power in Kw
The HDM-III equation is from Watanatada et al. (1987c); the HDM-4 is from an analysis of
the default parameter values from NDLI (1995a). For use in HDM-III the values will need to
be divided by the factor of 0.736 to convert from kW to MPH.
6.3.12 Engine speed

Engine speed is used in HDM-4 to predict the fuel required to maintain engine operation. The
engine speed depends on the road speed, the gear selected and the differential ratio.
To calibrate the engine speed a simulation model has been developed. Bennett and
Greenwood (1999) give a description of its principles and operation. Appendix K describes
the model’s operation. This simulation model should be used for a Level 1 calibration of the
engine speed model.
Using manufacturer’s specifications, a set of gear and differential ratios for a selection of
representative vehicles should be established. These are used as input to the simulation model
that applies a set of rules for driver behaviour to estimate the engine speed at a range of road
speeds. The model output is a set of engine versus road speeds for each set of vehicle
characteristics supplied. These data should then be analysed using regression techniques to
develop the necessary set of model parameters for HDM-4. Appendix K illustrates the use of
this model.
6.4 Level 2 - Calibration of primary relationships
6.4.1 Speed prediction
Measurement of desired speed (VDESIR)

A Level 2 calibration calls for VDESIR to be measured. This should ideally be done on a
number of sections of roads that are straight, level or with a minor downgrade, and with a low
roughness. The free speed should be measured using either manual or automated methods.
It is important that a sufficient number of vehicles be observed to minimise statistical error.
For the same level of accuracy, the sample size is proportional to the standard deviation of
speeds. Thus, developed countries with homogeneous traffic will require smaller sample sizes
than underdeveloped countries with heterogeneous traffic. Appendix I presents a technique
for estimating sample sizes. With indirect measurements, the sample size should be sufficient
to ensure that there is 90 per cent confidence that the estimate is + 2.5 km/h. With speeds,
one can assume the coefficient of variation is approximately 0.12 for estimating the sample
size.

Version 1.0
Having established the mean operating speed, one then uses this in conjunction with HDM-4
to determine the appropriate value for VDESIR. HDM-4 is run iteratively using the same
road section parameters as prevailed in the speed survey varying VDESIR until the predicted
and observed speeds agree.
It should be recognised that the desired speed is a function of bendiness. This effect was
observed in several studies when vehicles travelling on roads with severe alignments had
lower desired speeds (McLean, 1991). NDLI (1995a) proposed a negative exponential model
should be used to reduce the desired speed as a function of bendiness within HDM-4.
However, on testing it was found that this model gave unreasonable predictions at extreme
levels of bendiness so it was not included in HDM-4.
If HDM will be applied at different levels of bendiness, it would be prudent to establish
separate values of VDESIR to apply at these different levels.
Width
The HDM-4 desired speed-width model is illustrated in Figure 6.6. It is based on the work of
Hoban et al. (1994) and Yuli (1996). The underlying assumption is that there is critical width
below which speeds will be unaffected by width (CW1). Between this minimum speed
(VDESMIN) and the desired speed on two-lane highways, there is a linear increase in the
speed. On roads wider than two-lanes there is a continued increase in speed, but at a much
lower rate.
The minimum desired speed, in HDM-4, is assumed to be 75 per cent that of the two-lane
highway desired speed. Conducting free speed studies on narrow and two-lane roads can
check this value. It is important that the roads studied have similar levels of roughness and, if
possible, be on flat, tangent sections. By holding all other factors constant the differences in
speeds will be due to width. A similar approach is used when roads are wider than two lanes
to obtain the speed increase slope.
VDESIR' in m/s
a3
1m
a2
1m
a1
a0
Hoban, et al. (1984) HDM-Q Pan (1995) PhD Thesis
CW1 CW2 Road Width in m
Figure 6.6 HDM-4 desired speed - width model
Roughness
The speed-roughness model is calibrated through the maximum rectified velocity.
Data for calibrating the maximum rectified velocity are best obtained from a before and after
study on a pavement selected for maintenance which will reduce the roughness. If there are
no changes to the width or alignment, any increases in speed after the maintenance is
attributable to the roughness reduction. Section 7.5 describes special considerations in

Version 1.0
undertaking such a study, particularly the need for control sections to check the roughness
meter calibration.
Table 6.7 shows the change in speeds before and after overlays from such a study in India
(NDLI, 1997). In all cases there was an increase in speed accompanying the decrease in
roughness.
Calibration of the model is done by running HDM-4 at each of the roughnesses in the before
and after study and recording the predicted speeds. The other speed model parameters should
reflect the conditions of the speed survey sites.
The speed-roughness slope (km/h/IRI) should be calculated and compared to that from the
studies. The ratio of the means is used to modify the maximum average rectified velocity used
as input to HDM-4. Since the speed-roughness model is non-linear, the process should be
repeated at least two times to ensure that consistent results are obtained.
Table 6.7
Data on speed-roughness study
Vehicle Site Dir. Speed in km/h Roughness in IRI m/km Speed/

Roughness
Slope
(km/h/IRI)
Before After Change Before After Change By site Mean
1 1 61.1 65.4 4.3 5.7 4.0 1.7 2.5
1 2 61.0 66.5 5.5 7.9 4.3 3.6 1.5

PC
2 1 66.0 72.5 6.5 7.4 3.7 3.7 1.8
2 2 59.7 70.0 10.3 6.7 3.7 3.0 3.4 2.3
1 1 48.3 52.3 4.0 5.7 4.0 1.7 2.4
1 2 51.4 54.8 3.4 7.9 4.3 3.6 0.9

MCV
2 1 50.6 55.8 5.2 7.4 3.7 3.7 1.4
2 2 53.1 54.8 1.7 6.7 3.7 3.0 0.6 1.3
1 1 57.0 57.3 0.3 5.7 4.0 1.7 0.2
1 2 55.0 56.7 1.7 7.9 4.3 3.6 0.5

BUS
2 1 57.2 60.6 3.4 7.4 3.7 3.7 0.9
2 2 56.7 57.1 0.4 6.7 3.7 3.0 0.1 0.4
1 1 44.3 49.0 4.7 5.7 4.0 1.7 2.8
1 2 44.1 49.0 4.9 7.9 4.3 3.6 1.4

MC
2 1 44.0 50.1 6.1 7.4 3.7 3.7 1.6
2 2 43.1 48.9 5.8 6.7 3.7 3.0 1.9 1.9
Source: NDLI (1997)
Curve speed
In order to calibrate the curve speed model it is necessary to undertake a series of
measurements on roads with different radius of curvature. HTC (1999) and Bennett (1994)

Version 1.0
describe in detail how to undertake such studies. Data are collected on speeds at the
midpoints of curves over a range of curve radii and analysed to develop a non-linear model
relating curve speed to the radius of curvature. Figure 6.7 is an example of the data for
passenger cars from Thailand (HTC, 1999).
90
80
70
60
Curve Speed in km/h
50
40
30
20
10
0
0 100 200 300 400 500 600
Curve Radius in m
Figure 6.7 Effect of curve radius on speed from Thailand
The data are fitted to the following model:
VCURVE = a0 × R a1 …(6.14)
where:
VCURVE is the limiting speed due to curvature (m/s)

R is the radius of curvature (m)
a0, a1 are model parameters
HTC (1999) discusses in detail the analytical requirements for developing such a model.
Speed limit enforcement factor

The speed limit enforcement factor is the speed by which traffic travels above the posted
speed limit under ideal conditions. It can be estimated by conducting a speed survey on a
tangent road and comparing the mean speeds to the mean posted speeds. The default value is
1.1 that indicates that traffic will travel up to 10 per cent above the posted speed limit.
6.4.2 Capacity and speed flow
Capacity
As described earlier with Figure 6.2, HDM requires the nominal and ultimate capacities, and
a Level 2 calibration undertakes field trials to establish these capacities.

Version 1.0
As shown in Figure 6.8, there are several different methods available for estimating capacity
(Minderhoud et al., 1997). These authors make an important observation: “Attempts to
determine the capacity of a road by existing methods will generally result in a capacity value
estimate, but the validity of this value is hard to investigate because of the lack of a reference
capacity value, which is supposed to be absolutely valid. A clear, reliable method does not
appear to be available at this time.”
Capacity
Estimation
Direct Empirical Indirect Empirical
Observed
Observed
Observed Observed Volumes, Guidelines
Volumes and Simulation Models
Headways Volumes Densities and (eg HCM)
Speeds
Speeds
Figure 6.8 Capacity estimation methods
Irrespective of what method is used, it is necessary to sample over sufficient time intervals to
obtain a reliable estimate. One normally measures the flow in 15-minute periods and
converts this to an hourly flow. In most situations capacity is a difficult condition to reliably
encounter so one must often extrapolate data collected at lower flows to estimate the capacity.
The capacity estimates must consider the range of factors influencing capacity. In Section
6.3.2 a series of default capacity reduction factors were given. The capacity analyses should
consider these to establish local values.
Width is a particularly important factor. The HDM approach results in a step-function of
capacity versus width. However, as shown in Figure 6.9, other studies have treated capacity
as a continuous function of width. This has the advantage of avoiding discontinuities that can
arise at the boundaries of step functions. Figure 6.9 also shows the different capacities found
in different countries, and highlights the importance of local calibration.

Version 1.0
3500
Indonesia (Sweroad, 1997)
3000
HDM-95
HDM-95 Wide Two-Lane
Two-Lane
2500
Australia
(McLean, 1989)
Capacity in pcse/h
2000
India (Kadiyali, 1991)
1500
HDM-95
Narrow Two-Lane
1000
500
HDM-95
Single Lane
0
0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0
Road WIdth in m
Source: Sweroad (1997) and Kadiyali (1991)
Figure 6.9 Effect of width on capacity
From a practical perspective, it is important to focus on the pavement widths that are most
relevant. Many applications consider incremental widening in the range of 6-8 m and in this
band minor width changes can have a marked impact on capacity. If pavements are less than
6 m wide, they often deal with a major widening of several metres that will have a substantial
increase in capacity, analogous to the step functions in HDM. Conversely, small changes in
width below 6 m have little impact on capacity. One should therefore establish realistic
capacity estimates that will reflect the operational impacts expected.
In the context of HDM speed modelling, both the nominal and ultimate capacities are of
interest. As described in HTC (1999), the ultimate capacity prevails over short time periods so
is best represented by the maximum flow observed on a road. The nominal capacity is the
maximum flow rate that can reasonably be expected to be maintained so it should be based on
a longer period. In Thailand HTC (1999) used 5 minute intervals for the ultimate capacity and
10-15 minute intervals for the nominal capacity.
Given the limited data available for most Level 2 calibrations, there are two techniques that
should be considered for estimating the capacity:
! Headways
! Observed flows
Headways are the time difference, in seconds, between successive vehicles (usually measured
from rear bumper to rear bumper although if using axle detectors from front axle to front
axle). The capacity of the road is defined as:
3600
Qult = …(6.15)
hc
where:

Version 1.0
hc is the mean headway of constrained vehicles (s)
The method is based on the theory that at high flows there are still two populations of
vehicles:
! Constrained (followers)
! Unconstrained (leaders)
The distribution of following headways is expected to be the same as for constrained drivers
in any stationary traffic stream.
Although relatively easy to collect, it is considered that this method tends to overestimate
capacity due to the assumption that the distribution of constrained drivers can be compared at
capacity with that below capacity.
With observed flows, data are collected on roads that will reach capacity at some point during
the study period. The road capacity is taken to be the maximum flow, or the mean of several
very high flows, observed during the analysis period. This is illustrated in Figure 6.10.
Flow Rate
Maximumum Flow
Averaging Interval
Cycle [hour]
(15 min)
Figure 6.10 Estimating capacity from observed flows
Figure 6.11 is an example of speed-flow study results from Thailand (HTC, 1999). This figure
shows the capacity of 2121 PCSE/h. This was based on 5-minute peak flows measured using
data loggers.
The disadvantage of this approach is that the roads may not always reach capacity. However,
if the sites are carefully selected this situation can be minimised. One way of ensuring high
flows is to plan the data collection around special events, such as holidays when traffic flows
can be expected to be high.
Speed at nominal capacity

HTC (1999) estimated the speed at nominal capacity by fitting a regression function to speed-
flow data for flows beyond the breakpoint flow (Qo). The nominal capacities were then
substituted into these equations to establish the corresponding speeds. The results were fairly
consistent between different vehicle classes and suggested that the HDM approach of
assuming 85 per cent of the free speed was too high and it should be 77-82 per cent.

Version 1.0
120
Site 19A - 6MW

100
Speed in km/h 80
60
40
20
0
0 500 1000 1500 2000 2500 3000
Flow in pcse/h
Figure 6.11 Thailand speed-flow results for 6-lane motorway
Ultimate capacity
As noted by HTC (1999): “The speed at ultimate capacity is notoriously difficult to estimate.
This is because the flows become unstable in that region when the demand exceeds capacity
there is a breakdown in the flow that leads to markedly lower speeds.” Measurements of
speeds at flows such as shown in Figure 6.11 should be used to determine an appropriate
speed.
Low flow period

There are two techniques that can be used to investigate the low flow period (HTC, 1999):
! Regression approach
Linear regressions can be performed for a number of flow ranges:
<100; <200, <300, <400, <500, etc.
If there are no traffic interactions one would not expect to find that the resulting equation
was statistically valid1. The upper limit of the last range before a statistically valid
equation was taken to be the breakpoint where the low-flow period ends.
! Maximum speed approach
For each site, average speeds at different flows are established. The flow where the
maximum speed was observed was taken to be the breakpoint. To account for variability
in the data it is common to average the flows in bands of, say, 100 PCSE/h.
HTC (1999) found that the regression method did not give valid results at all sites whereas the
maximum speed approach did. However, the regression method was considered to give more
reliable results.
1
There are a variety of methods for assessing the validity of a regression equation. The easiest is to ensure
that the 't' statistics of the regression coefficients are significant at 95% confidence. HTC (1999) found that
this yielded the best results.

Version 1.0
6.4.3 Passenger car space equivalencies: S-II/S-III

Traffic streams are comprised of a range of vehicles, from passenger cars to heavy trucks. For
the purposes of capacity it is necessary to convert these into a homogeneous traffic stream.
This is done in HDM-4 through the use of Passenger Car Space Equivalents (PCSE). The
PCSE differs from the tradition passenger car unit/equivalency (PCU/PCE) in that it is based
on the area occupied by the vehicle. The PCU is based on the area occupied as well as the
vehicle’s performance. Since HDM explicitly models performance, the use of PCE/PCE
would lead to double counting. For this reason, PCSE values are always lower than PCU
values.
Hoban et al. (1994) describes the recommended PCSE values for HDM. This work builds
upon the earlier work of Hoban (1987). The PCSE were established based on the assumption
that each vehicle has a typical length as well as typical leading and following headways1.
Using an assumed speed of 72 km/h, Hoban et al. (1994) calculated the basic PCSE values
shown in Table 6.8. These basic values only accounted for the longitudinal space occupied by
vehicles. Additionally, larger vehicles tend to impact adjacent lanes, with this “adjacent lane”
effect being greater for larger vehicles and for narrower roads (Hoban et al., 1994). This led
to the PCSE values varying by width as shown in Table 6.8.
Table 6.8
PCSE values by vehicle class
Recommended values
(includes “basic” plus

adjacent lane effects)
Vehicle Avg. Time Space Total Basic 2-lane Narrow 1-lane
class length headway headway space PCSE 4-lane 2-lane
(m) (s) (m (m)
Car 4.0 1.6 32 36.0 1.0 1.0 1.0 1.0
Pickup 4.5 1.8 36 40.5 1.1 1.0 1.0 1.0
Heavy Bus 14.0 2.2 44 58.0 1.6 1.8 2.0 2.2
Light Truck 5.0 2.0 40 45.0 1.3 1.3 1.4 1.5
Medium Truck 7.0 2.2 44 51.0 1.4 1.5 1.6 1.8
Heavy Truck 9.0 2.4 48 57.0 1.6 1.8 2.0 2.4
Truck and 15.0 2.5 50 65.0 1.8 2.2 2.6 3.0

Trailer
Notes:
1 Basic data from Hoban et al. (1994)
2 Time headway calculated from space headway using 72 km/h speed
3 Truck and Trailer average length increased to 15 from 11 given in Hoban et al. (1994)
based on difference between total space and space headway
1
The PCSE is analogous to truck equivalencies calculated using the headway method. This defines the
equivalency of a truck as, the ratio of the average headway for trucks in the stream (in s), to the average
headway for cars in the stream (in s). See McLean (1989) for a detailed discussion of equivalency factors
and their determination.

Version 1.0
The PCSE values in Table 6.8 were calculated by firstly establishing the total space occupied
by passenger cars (default = 36.0 m). The Basic PCSE values for the other vehicle classes
were then determined from the ratio of their total space to the passenger car space. This basic
value was then subjectively adjusted for width effects to obtain the recommended values in
the final three columns.
Local calibration of the PCSE values can be done by adjusting the basic assumptions in Table
6.8 to reflect local vehicles. The easiest data to obtain is that on vehicle lengths. Field studies
could focus on establishing the average headways for vehicles thereby giving the total space.
Axle detectors are quite useful in this type of work since the time between the first axles of
successive vehicles divided by the velocity (in m/s) gives the total space occupied.
6.4.4 Fuel consumption
Overview
The HDM-III and HDM-4 fuel consumption models are based on mechanistic principles.
HDM-III was developed using a regression-based approach while HDM-4 is more of a pure
mechanistic formulation that more effectively reflects the different components affecting fuel
consumption. To calibrate the fuel consumption model parameters it is either necessary to
have a series of raw fuel consumption measurements or, alternatively, existing fuel
consumption models, such as those developed using multiple linear regression techniques.
Each of these will be discussed individually. The discussion will focus on the HDM-4 fuel
model, with some reference to HDM-III.
HDM-4 fuel consumption model

The HDM-4 fuel consumption model predicts the fuel consumption as a function of the
power as follows:
IFC = MAX (α, ξ Ptot) …(6.16)
which can be expressed as:
IFC = MAX (α, ξ Ptr + ξPengaccs) …(6.17)
where:
IFC is the instantaneous fuel consumption (ml/s)

α is the idle fuel consumption (ml/s)
ξ is the fuel-to-power efficiency factor (ml/kW.s)
Ptot is the total power requirements (kW)
Ptr is the total tractive power requirements (kW)
Pengaccs is the total engine and accessories power (kW)
The fuel model calibration therefore focuses on the parameters α and ξ. It is assumed that the
models used to predict the total power have been correctly calibrated elsewhere. The
equations used to predict the total power are as follows:
Ftr = Fa + Fg + Fr …(6.18)

Version 1.0
Fa = R1 v 2 …(6.19)
Fg = R2 GR …(6.20)
Fr = R3 + R4 v 2 …(6.21)
The variables R1 to R4 are simplifications of the variables in the actual HDM-4 equations:
R1= 0.5 RHO CD CDMULT AF …(6.22)
R2 = M g …(6.23)
R3 = FCLIM CR2 (b11 Nw + CR1 b12 M) …(6.24)
R4 = FCLIM CR2 CR1 b13 …(6.25)
The total forces are:
Ftr = R3 + (R1 + R4 ) v 2 + R2 GR …(6.26)
v
Ptr = Ftr …(6.27)
1000
Ptr =
[(R3 + R2 GR ) v + (R1 + R4 ) v 3 ] …(6.28)
1000
where:
v is the vehicle speed (m/s)

RHO is the mass density of air (km/m3)
CD is the aerodynamic drag coefficient
CDMULT is the aerodynamic drag coefficient multiplier
AF is the projected frontal area
CR1 and are rolling resistance model parameters
CR2
b11 to b13 are rolling resistance model parameters
GR is the gradient (%)
FLCIM is a climatic adjustment factor
Calibration using raw fuel consumption data

This is the best way of calibrating the model, although the data are seldom readily available.
There have been numerous studies conducted into fuel consumption and these are readily
found in the literature. Yuli (1996) gives a good description of calibrating the HDM-III fuel
model, and the conversion of the results to the HDM-4 fuel model. There are two parameters
to be calibrated:

Version 1.0
! ξb)
Base engine efficiency (ξ
! Change in engine efficiency at higher power
The calibration procedure is as follows:

1 Undertake observations with the engine idling. These give the idle fuel rate.
2 Undertake a series of on-road measurements travelling at different speeds. This will
result in a set of data that are pairs of speed and fuel. Ensure that the fuel is expressed in
terms of ml/s. If, for example, it was measured in terms of ml/km it is converted to ml/s
by multiplying by the factor v/1000.
3 It is necessary to calculate the power required by the vehicle at each speed. Using the
HDM-4 mechanistic model formulation calculate the total tractive power requirements
(Ptr) for the speeds where fuel measurements are available. This is done using the
equation given above for Ptr.
4 Using the default HDM-4 parameters, calculate the engine and accessories power
Pengaccs.
5 The tractive power is corrected for driveline efficiency using the default HDM-4
parameter to give the following equation for total power:
Ptr
Ptot = + Pengaccs …(6.29)
edt
6 Divide the fuel by the total power (Ptot). This gives the power-to-fuel conversion factor
ξ for that power.
7 Undertake a regression analysis and fit the model parameters ξb and ehp in the model:
ξ = ξ b1 + ehp
Ptot 
 …(6.30)
 Pmax 
where:
Pmax is the maximum rated power for the vehicle.
8 If the analysis does not allow for the simultaneous estimation of ξb and ehp, assume the
HDM-4 default relationship for ehp.
Calibration from existing relationships

There two ways by which existing relationships can be used to develop the calibrated
parameters for α and ξ:
! Generating data
! Via the model coefficients
Generating data
1 Use the existing models to generate a series of fuel consumption predictions covering the
full range of operating conditions likely to be encountered, for example: speeds,
roughness, gradients.

Version 1.0
2 Make an assessment of the range of conditions typically encountered and weight the data
set accordingly. For example:
70% operate on flat terrain between 50 and 80 km/h;
20% on rolling terrain between 40 and 70 km/h;
10% in mountainous terrain between 20 and 50 km/h.
3 Create a data set that contains data reflecting the assessments from (2). This will ensure
that the results have the appropriate weightings.
4 From this data set, follow the instructions given above for calibration with raw fuel data.
Model coefficients
As described in NDLI (1997), a commonly used model form that has been fitted to fuel
consumption data is:
+ (A2 S 2 ) + (A3 IRI) + (A4 RS ) + (A5 FL )

A1
SFC = A0 + …(6.31)
S
where:
SFC is the specific fuel consumption (l/1000 km)

A0 to A5 are regression coefficients
S is the speed (km/h)
RS is the rise (m/km)
FL is the fall (m/km)
It can be shown that this model is compatible with the HDM-4 model. Ignoring the fall term,
the following are the terms that correspond between the two models:
Regression HDM-4
A0 + A3 IRI (ξ/edt) R3
A1 3600 ξ Pengaccs
A2 (ξ/edt) (R1 + R4)/(3.62)
A4 (ξ/edt) R2
The terms A0 and A3 are thus mainly related to rolling resistance; A1 to the fuel required to
operate the engine; A2 to aerodynamic resistance, and A4 to the gradient resistance.
Calibration is done as follows:
1 Calculate the SFC for a range of conditions similar to those described above under
Generating Data.
2 The idle fuel consumption α is given by A1/3600, or ξ Pengaccs.

Version 1.0
3 Subtract A1/S from the total SFC to obtain the fuel consumption due to forces opposing
motion. Convert this to the IFC in ml/s by dividing SFC by the speed.
4 Calculate the total tractive power using the HDM equations with the vehicle
characteristics and divide by the driveline efficiency.
5 Calculate ξ by dividing IFC by the tractive power, corrected for driveline efficiency.
In applying this method in India, NDLI (1997) found some inconsistencies between the results
from Indian equations and the default values in HDM-4. For some vehicles the values were
outside of the range of values considered viable, while for others they agreed well. This
method should therefore be applied prudently.
6.4.5 Acceleration noise
Overview
Acceleration noise is used in HDM-4 to predict the effects of speed variations due to
congestion or road conditions on vehicle operating costs. It is a relatively simple parameter to
measure and can be done through a small controlled experiment. Once the appropriate
equipment has been obtained, the work should only take a few days.
Figure 6.12 is an example of the acceleration profile of a vehicle on a motorway (NDLI,
1995a). It can be observed that the accelerations before entering the motorway were much
greater than once on the motorway. If a vehicle has the same start and end speeds over a
section, the mean acceleration will be zero. HDM-4 therefore uses the standard deviation of
acceleration, called the acceleration noise to represent the magnitude of accelerations; the
higher the acceleration noise, the greater the magnitude of the accelerations.
3
Entering Motorway Once on Motorway
1
Acceleration in m/s/s
-1
-2
-3
0 50 100 150 200 250 300 350 400 450 500
Time in s
Figure 6.12 Example of acceleration data on motorway
In HDM-4 the acceleration noise is a function of:

Version 1.0
! Traffic interactions (that is, the volume-to-capacity ratio)
! Driver (σadr)
! Roughness (σairi)
! Road alignment (σaal)
! Non-motorised traffic - NMT (σanmt)
! Side friction (σasf)
The total acceleration noise is calculated as:
σa = σat 2 +σan 2 …(6.32)
where:
σa is the acceleration noise (m/s2)
σat is the traffic noise (m/s2)
σan is the natural noise (m/s2)
The natural noise is calculated as:
σan = MAX((σadr 2 + σaal 2 ), σasf 2 , σanmt 2 , σairi 2 ) …(6.33)
HDM-4 uses a sigmoidal function for the traffic interactions and a linear function for the
others. The acceleration noises are combined to obtain the total acceleration noise for a given
road section.
A Level 2 calibration of any of the model consists of making measurements at two extremes
and assuming that the existing model applies. Level 3 calibration sees the model formulations
requantified.
It should be noted that it is not necessary to calibrate all components of the model. If one is
looking only at the effects of capacity improvements, the traffic interaction model should be
calibrated and possibly the NMT/side-friction model. If roughness improvements are of
interest then only the roughness model needs to be calibrated.
Measurement of acceleration noise

To measure acceleration noise it is necessary to have a vehicle equipped with some form of a
data logger. This data logger must measure speed or acceleration at 1 s intervals. Direct speed
measurement is preferable to an accelerometer as with accelerometers it is necessary to
identify flat sections of road to eliminate gradient effects.
There are a variety of devices available, ranging from low to high cost. A suitable device
usually incorporates a dedicated data logger which will monitor the speed or acceleration on
an instantaneous basis. Speed based systems are easier to use than accelerometers since the
data can usually be directly analysed to establish the acceleration noise.
One such device with purpose-built features, ROMDAS, measures the speed of the vehicle
and contains data processing routines designed for calibrating the HDM-4 acceleration noise

Version 1.0
model. Further information is available through the HTC Infrastructure Management web site
given in the About This Manual section of this document.
Irrespective of which component(s) of the acceleration noise model are being investigated,
the basic approach is the same:
! Outfit a vehicle with the measurement equipment. The vehicle should be as
representative of the vehicle class as possible. Ideally, the measurements should be made
with several vehicles and several drivers, but NDLI (1995a) found that the results were
broadly similar so this is not critical.
! A test route should be established which would allow for the two critical points to be
monitored:
❏ low impact
❏ high impact
For example, if roughness is of interest the measurements should be made on a low and
high roughness road, for traffic impacts on non-congested and highly congested roads. It
is vital that all the other road attributes (for example, width, alignment, traffic
composition) be similar. That way, any differences in acceleration noise can be
attributed to the parameter of interest.
! It is important that the route be long enough for the process to stabilise. Accelerations
are assumed to be random and if too short a route is used the process will not stabilise.
Figure 6.13 shows the effect of sampling interval on data based on an analysis of the
NDLI (1995a) measurements (Greenwood, 1998). On the basis of this it is recommended
that a minimum sample of 5 minutes be collected in a single run for each attribute of
interest.
1.0
0.9
0.8
A cceleration Noise Results (m/s²)
0.7
0.6
0.5
Max
0.4
Min
Mean
0.3
0.2
0.1
0.0
0 30 60 90 120 150 180 21 0 240 270 300 330 360 390 420 450 480 510 540
Le ngth o f Interval A cceleratio n Noise C ollected Over (seco nds)
Figure 6.13 Variation of acceleration noise with sample length
Driver noise calibration

This must always be calibrated. The driver noise is the noise that arises when a vehicle is
operated under normal operating conditions without traffic, roughness, alignment, NMT or
side friction effects.

Version 1.0
The measurements should be made on as good a standard road as is available for the type of
road of interest. HDM allows data to be input for different road classes so measurements can
be made, for example, on two-lane and multi-lane highways.
If no other measurements are made (for example, alignment) it is assumed that the natural
noise is entirely due to the driver: σan = σadr.
Traffic noise calibration

A Level 2 calibration of the traffic noise model entails taking measurements under severely
congested conditions. This gives the maximum noise (σamax). Since the measurements also
include the natural noise, the maximum traffic noise (σatmax) is calculated as:
σatmax = σamax 2 − σan 2 …(6.34)
The value for σatmax is used in conjunction with a sigmoidal relationship to predict the
traffic noise at intermediate volume-to-capacity ratios.
Roughness, road alignment, NMT and side friction

All of these assume a linear relationship between severity and acceleration noise.
Measurements are taken at the extreme conditions, for example a roughness of 20 IRI m/km,
and the noise is linearly interpolated from a value of 0 to this extreme condition. The
measurements are done as follows. The example applies to roughness, but can be done the
same way for the other factors.
1 Identify road sections with low and high roughnesses which are otherwise identical (that
is, alignment, traffic, roadside friction etc.).
2 Measure the acceleration noise on the low and high roughness sections in the absence of
any traffic interactions.
3 Establish the difference between these values. This represents the maximum change due
to roughness.
6.4.6 Parts and labour costs

A Level 2 calibration can focus on the magnitude of costs as well as the effects of roughness.
The magnitude of the costs is done by surveying a sample of users and establishing an
average cost and then adjusting the HDM parts and labour model predictions so that the
predicted cost is the same as the surveyed cost at the average conditions. Kampsax (1986)
use this approach in Pakistan for HDM-III. A National Trucking Study had established a total
maintenance and repair cost of 0.35 Rs/km for medium trucks. Using an average age of
300,000 km and an estimated average roughness of 6 IRI m/km, the parts cost was calculated
using the default HDM-III parameters and the rotation factor was reduced to 32.5 per cent of
the default value to obtain the same costs.
Opus (1999) presents the results of a study using accelerometers to investigate the effects of
roughness on parts consumption. This work offers great scope for a relatively rapid and low-
cost method for verifying the shape of the roughness model (that is, linear) and the
roughnesses below which parts consumption will not be influenced.
Accelerometers (10 g strain gauge with 0-120 Hz frequency response and low signal to noise
ratio) were fitted to vehicles. For cars, an accelerometer was fixed to the rear axle, as close to
the wheel assembly as possible. Another was firmly fixed to the boot floor, as close as
possible to being directly over the axle accelerometer. For trucks, an accelerometer was fitted

Version 1.0
to the rear axle and another immediately above on the cargo tray. By mounting the axle and
body accelerometers in close proximity to each other, the degree to which roughness induced
axle vibrations were transmitted to the vehicle body could be determined.
The vehicles were driven at different speeds over sites with different roughnesses and the
accelerations were recorded. Figure 6.14 is an example of the body accelerations versus
roughness for medium trucks1. It can be seen that the body accelerations increase with
increasing roughness and this can be used to calibrate the roughness effects in the HDM parts
consumption model. It will also be noted that at lower roughnesses the accelerations do not
have much of an impact. This can be used to confirm the roughness level below which there
is no impact on parts consumption.
3.0
mcbdv.grf
MCV - Body Acceleration
2.8 Legend Title
50km/h
2.6 70km/h
Fit: 50km/h
2.4
RMS Acceleration (m/s )
Fit: 70km/h
2
2.2
2.0
1.8
1.6
1.4
1.2
1.0
20 40 60 80 100 120 140 160

NAASRA (100m averages)
Figure 6.14 Effect of roughness on body acceleration (Opus (1999))
One finding of this research which is not considered in HDM is the speed-roughness
dependence. The results clearly show that the higher the speed, the greater the impact of
roughness on body accelerations.
6.4.7 Used power

For a Level 2 calibration the driving power is established by measuring speeds on gradients.
As described in Bennett (1994), there are two approaches that can be used:
! Crawl speeds
! Speed differences (or the spatial method)
1
The roughness units in this figure are NAASRA counts/km. A conversion to IRI is 1 IRI = 26.4 NAASRA
counts/km.

Version 1.0
Crawl speeds
On a long, steep upgrade vehicles will travel at a constant crawl speed where all the forces
are in balance. Since there is no acceleration, the observed speed can be substituted into the
following equation to calculate the used power:
Pu 0.5 RHO CD CDmult AF + R3 3 R4 M g GR

=av+ v + v+ v …(6.35)
M' M' M' 100 M'
where:
Pu is used power (W)

v is the vehicle speed (m/s)
M’ is the effective mass (km)
RHO is the mass density of air (km/m3)
CD is the aerodynamic drag coefficient
CDmult is the aerodynamic drag coefficient multiplier
AF is the projected frontal area
R3 and R4 are rolling resistance components (Equations 6.24 above and 6.25 above)
GR is the gradient (%)
g is the acceleration due to gravity (m/s2)
Speed differences
When the gradients are insufficient for vehicles to reach their crawl speeds the speed
reduction between two points on the grade can be used to establish the driving power. Called
the speed difference, or spatial method, this uses the following equation to establish the used
power:
__
Pu v 1 − v o 0.5 RHO CD CDmult AF + R3  SL  3 R4 SL
2 2
M g HC
= +   + + …(6.36)
M' 2T M'  T  M' T M' T
where:
__ is the mean power used over the section

Pu
v1 is the speed at the exit to a section (m/s)

vo is the speed at the entry to a section (m/s)
SL is the section length (m)
T is the time taken to travel the section length (s)
HC is the height change over the section (m)

Version 1.0
It should be noted that this method gives the average power used over the section, as opposed
to the crawl speed method which gives the maximum used power.
Bennett (1994) compared these two methods and concluded that both will give similar results,
with the spatial method giving the best results when measurements are made as far along the
grade as possible. When using the spatial method there is almost a linear relationship between
the predicted power used and the distance measured. Thus, any errors in measured distances
will lead to a corresponding over or under estimation of the used power.
Used power will likely be proportional to the gradient, with vehicles using more power on
steeper gradients (Bennett, 1994). However, HDM only allows for a single value to be input
so either the average or a value based on the most typical gradient should be adopted.
6.4.8 Braking power

For a Level 2 calibration the used power can be calculated from speed in the same manner as
the driving power. Either the crawl speed or speed difference approach can be used
(Bennett, 1994). This entails measuring the speed of the vehicle and then calculating the used
braking power (which is expressed as -Pu).
6.4.9 Service life

The Level 2 method of estimating the service life is through the use of a survival curve. This
is a curve showing the proportion of the original number of vehicles remaining in service of a
given age. The curves for each original year of registration are unique although a reasonable
representation may be obtained by evaluating the age structure of the vehicles in use at a
particular date. By comparing this age structure with the number of units originally put into
service each year, a quasi-survival curve may be prepared from which the average service life
may be calculated. Appendix I illustrates how the survival curve is derived and how the
average service life is calculated from the curve.
In order to employ a survival curve it is necessary to have an age distribution of the existing
vehicle fleet and records of the number of vehicles originally registered. In certain
circumstances the registration data may not be available or non-applicable. A survival curve
can still be constructed by obtaining two age distributions one year apart. This will show the
number of vehicles retired for each age group from which a survival curve can be calculated.
These techniques will give the average service life for the fleet as a whole, or a sub-section of
the fleet operating in a certain area. It is necessary to estimate the traffic-weighted roughness
for this area so as to adjust the value during the Optimal Life calculations to the roughness on
a smooth pavement. This is done by through calculating the product of the roughness on
major links and the traffic volume.
6.4.10 Utilisation
For a Level 2 calibration a field survey is required. This entails stopping a sample of vehicles
and recording their odometer readings or, for commercial vehicles, recording data from the
driver’s logbook . In some countries the data are available from fleet management companies.
The data should be grouped by age with the average annual kilometreage of all vehicles of a
given age being calculated. If a survival curve is available, the average lifetime kilometreage
can be calculated by multiplying the average annual kilometreage for each year by the
survival ratio. The total of these products represents the average lifetime kilometreage.
Appendix I illustrates this technique using data from Bennett (1985). The average annual
kilometreage is then calculated as the average lifetime kilometreage divided by the average
service life.

Version 1.0
If a survival curve is not available, a reasonable estimate of the average annual kilometreage
will be obtained by calculating the average kilometreage of all the survey data. If the sample
is unbiased, this should yield a similar value to that from the survival curve technique.
Bennett (1985) used this approach and it yielded an average of 10,800 km/year that compared
with 9,850 km/year using the survival curve approach with the same data. The differences are
probably largely due to the biasing of the Bennett (1985) data towards newer vehicles that had
higher average utilisations.
6.4.11 Oil consumption

Oil consumption is broken into two components:
! Oil loss due to contamination, and
! Oil loss due to operation.
Oil loss due to contamination is a function of the distance between oil changes. Oil loss due
to operation is a function of the fuel consumption.
The contamination losses can be calibrated by establishing the typical distances between oil
changes. These may be based on the manufacturer’s specifications but most likely are
different. The losses are calculated as:
OILCAP
OILCONT = …(6.37)
DISTCHG
where:
OILCONT is the contamination loss (l/1000 km)

OILCAP is the engine oil capacity (l)
DISTCHG is the distance between oil changes (km)
Oil loss due to operation is more difficult to measure and, as it likely has limited impacts on
the total costs, should not usually be calibrated. If calibration is desired it should be done as
follows for a sample of vehicles:
1 Change the oil and replace with a known quantity of oil.
2 Operate the vehicles for at least 5-10,000 km, keeping exact records of the amount of
fuel used and, if necessary, additional oil added.
3 At the end of the period drain the oil completely and measure the amount remaining. The
difference between this volume and the initial volume is the oil lost due to operation.
Dividing this by the total fuel consumption gives the operation loss rate.
6.5 Level 3 - Adaptation of model
6.5.1 Acceleration noise
Overview
Compared with a Level 2 calibration, a Level 3 calibration of acceleration noise can address:

Version 1.0
! Quantifying parameters for additional vehicle classes or road types; or,

! Developing alternative relationships
Expanding the data to additional vehicle classes would see experiments done in a similar
manner to Level 2, but with a larger matrix of vehicles, drivers, and road conditions. To
develop alternative relationships would address the key assumptions in the traffic noise and
roughness/road alignment/NMT/side friction relationships.
Traffic noise calibration

The traffic noise model is a sigmoidal function that gives different predictions depending
upon the volume-to-capacity ratio (VCR) and the volume where speeds begin to be affected
by traffic interactions (Qo). The latter integrates the traffic noise model with the HDM-4
speed flow model so that when VCR < Qo there is only natural noise; above that the
combination of natural and traffic noise. The model form in HDM-4 is:
1.04
σat = σatmax (a0 + a1 VCR)
…(6.38)
1+ e
The effort should focus on calibrating the parameters a0 and a1 and then developing a new
model form around the calibrated parameters. This will entail taking measurements of σat at
different levels of VCR and then fitting a curve to the data. Because of the difficulties in
estimating VCR, it is recommended that surveys be conducted at VCR levels of 1, 0.5 and
0.3. This will provide an indication of the trend in the data. Additional measurements can be
then be made to supplement the data as necessary.
Roughness, road alignment, NMT and side friction

The assumption is that there is a linear relationship between these factors and acceleration
noise. Collecting data over a range of road conditions would test the validity of this
assumption. It is important that all the other factors be held constant so that the only factor
influencing the results is the factor of interest.
6.5.2 Speed β
The quantification of the speed β requires a major field study. A series of speed data are
required covering the full range of conditions under which the constraining speed models will
be applied.
This data set should then be analysed using the same statistical methods as those described in
Watanatada et al. (1987c). The analysis should consider the same issues, such as logarithmic
transformations, and requantify the model from first principles.
6.5.3 Rolling resistance

6.5.4 Idle fuel consumption

6.5.5 Fuel efficiency


Version 1.0
6.5.6 Critical gradient length

6.5.7 Tyre stiffness

6.5.8 Engine accessory power

6.5.9 Engine drag power

6.5.10 Drivetrain efficiency

6.5.11 Tyre wear parameters

6.5.12 Parts consumption

6.5.13 Labour hours


Version 1.0
RDWE calibration
7 RDWE calibration
7.1 Bituminous paved road deterioration

The HDM flexible pavement deterioration and works effects (RDWE) model (as
implemented in HDM-III) has six deterioration adjustment factors. Table 7.1 shows the
impact elasticity class for various applications combined with the typical range of values of
the factors to give a potential net impact. Six criteria representing different applications were
used in the analysis and these are listed in the footnote to Table 7.1.
Table 7.1
Ranking of impacts of road deterioration factors
Deterioration factor Impact class for Impact Typical Net Sensi-

1
given criteria Elast- values Impact tivity
icity of class
Factor (%)
1 2 3 4 5 6
Roughness-age- D D B C B B 0.20 0.2 - 5.0 10
environment
Cracking initiation A C B B C B 0.25 0.5 - 2.0 6 High
Cracking progression A C C C C C 0.22 0.5 - 2.0 6
Rut depth progression D A B D C C 0.10 0.5 - 2.0 3
Roughness progression D D B D C B 0.09 0.8 - 1.2 1 Low

general
Potholing progression D D C B C D 0.03 0.3 - 3.0 2
Ravelling initiation D D D C D D 0.01 0.2 - 3.0 1
Note:
1 Criteria 1 = Amount of cracking
2 = Rut depth
3 = Roughness
4 = EIRR for patching
5 = EIRR for reseal
6 = EIRR for overlay
Impact sensitivity A = S-I
B = S-II
C = S-III
D = S-IV

Version 1.0
RDWE calibration
The roughness-environment factor is clearly the most important, due to the wider range of its
values, followed by the cracking initiation and progression factors. The general roughness
progression factor has low priority, despite its moderate sensitivity, because its range is small
based on many inter-country validation studies. As shown in the table, for convenience, these
adjustments can be grouped into just two classes:
! High impact
! Low impact
The following sections describe calibrating these factors for the three levels of calibration.
In HDM-4, the flexible pavement deterioration and works effects models have been
generalised to allow more specific adaptations to local materials, and adding a few improved
relationships. The changes include an increase in the number of calibration factors available
for adjusting the bituminous RDWE model from 6 to 20.
Consequently, while this chapter discusses mainly the determination of calibration
adjustments for the HDM-III RDWE model, there are only relatively brief and incomplete
references to specific HDM-4 calibration factors. Cross-references are given to HDM-4
model descriptions published in Volume 4 of this document series, Analytical Framework and
Model Descriptions.
In most cases, the guidelines provided may be directly transposed to apply to the HDM-4
model, although such approach requires a good knowledge of the specific relationships. The
HDM-4 aspects of these guidelines will be published in a future update to this Guide.
7.2 Level 1 - Basic application

A Level 1 calibration thus adjusts the top three deterioration factors from Table 7.1 and
retains the default values of unity for the remaining factors in most instances.
7.2.1 Roughness-age-environment adjustment factor: S-I

This factor, which determines the amount of roughness progression occurring annually on a
non-structural time-dependent basis, is related to the pavement environment and is effectively
an input data parameter rather than a calibration adjustment. The factor adjusts the
environment coefficient, m, which has a base value of 0.023 in the model, representing 2.3
per cent annual change independent of traffic, that is:
∆R te = K ge 0.023 R t …(7.1)
where:
Kge is the roughness age-environment calibration factor (HDM-III)

∆Rte is the change in the roughness component due to environment in the 1-year analysis
time increment
Rt is the roughness at the beginning of the year t
In HDM-4, the calibration factor Kgm is equivalent to the HDM-III factor Kge. The following
discussion uses Kge.

Version 1.0
RDWE calibration
Method 1
For a Level 1 Calibration, the values are established based on the general environmental
conditions and the road construction, drainage standard. This is done as follows:
Step 1 Identify the environment applicable to the immediate vicinity of the road
project in terms of the classifications provided in Table 7.2.
Step 2 Select the appropriate value of m from Table 7.3 according to the
environmental classification.
Step 3 Determine the effective m-value, meff, by multiplying m by a factor km
according to the standard of road construction and drainage Table 7.4, as
follows:
m eff = m k m …(7.2)
Step 4 Calculate Kge from meff as follows:

m eff
K ge = …(7.3)
0.023

Version 1.0
RDWE calibration
Table 7.2
Classification of road environment
Temperature Description Typical temperature range (°°C)

classification
Tropical Warm temperatures in 20 to 35
small range
Subtropical - hot High day cool night -5 to 45

temperatures, hot-cold
seasons
Subtropical - cool Moderate day -10 to 30

temperatures, cool winters
Temperate - cool Warm summer, shallow -20 to 25

winter freeze
Temperate - freeze Cool summer, deep winter -40 to 20

freeze
Moisture Description Typical moisture Typical annual

classification index precipitation (mm)
Arid Very low rainfall, high -100 to -61 < 300
evaporation
Subhumid Moderate rainfall, or -20 to 19 800 to 1600

strongly seasonal rainfall
Humid Moderate warm season 20 to 100 1500 to 3000

rainfall
Perhumid High rainfall, or very > 100 > 2400

many wet-surface days
Table 7.3
Recommended values of environmental coefficient, m
Moisture Temperature classification

classification Tropical Subtropical Temperate - Temperate -
non-freezing shallow freeze extended
freeze
Arid 0.005 0.010 0.025 0.040
Semi-arid 0.010 0.016 0.035 0.060
Subhumid 0.020 0.025 0.060 0.100
Humid 0.025 0.030 0.100 0.200
Perhumid 0.030 0.040

Version 1.0
RDWE calibration
Table 7.4
Modifying factor of environmental coefficient for road construction and
drainage effects
Construction and drainage Non-freezing Freezing

environments environments
High standard materials and drainage; for example, motorways, 0.6 0.5
raised formation, free-draining or non-frost-susceptible materials,
special drainage facilities.
Material quality to normal engineering standards; drainage and 1.0 1.0

formation adequate for local moisture conditions, and moderately
maintained.
Variable material quality in pavement, including moisture or 1.3 1.5

frost-susceptible materials; drainage inadequate or poorly
maintained, or formation height near water table.
Swelling soil subgrade without remedial treatment 1.3 - 2.0 1.2 - 1.6
Method 2
The values for Kge estimated using this approach could be enhanced with limited field data, as
follows:
A small number (two to five) of strong old pavements should be selected in one environment
for which accurate roughness, traffic loading and age data are already available, and for
which:
! Pavement structural adequacy: PSA > 1; and,
! Pavement surfacing age > 10 years.
The observed estimates of m are then calculated for the selected pavements (following the
method in Section 7.3.1), averaged and compared with the values selected from the tables. If
the values are similar, no changes should be made to the originally chosen values. If the
differences are consistently different, such as mostly higher or lower, a partial correction may
be made (say about one half of the difference in the means). But care should be taken not to
attach too much significance to these few results until a more thorough application of a Level
2 calibration can be made.
7.2.2 Cracking initiation adjustment factor: S-I

Cracking initiation is predicted in terms of the time to the first visible crack, and when the
surfacing age first exceeds this time, cracking is deemed to begin. The adjustment factor is a
simple multiplier of the time to first crack, that is:
TYcra = K ci TY (Predictive relationship for relevant surfacing type) …(7.4)
where:
TYcra is the time to first visible crack in years
The predictive relationships for each surfacing type take account of the interactive fatigue
effects of pavement strength and traffic loading, and the durability effects of ageing. What the

Version 1.0
RDWE calibration
relationships could not do without requiring significantly more complex data entry was to
define how satisfactory were the material design, manufacture and construction quality, or
define the oxidising power of the environment, except in average terms. The calibration
adjustment therefore compensates for these in specific situations. Cracking behaviour of
bituminous materials tends to be variable, even under fairly identical situations, so it can be
dangerous to make general adjustments based on only one or two cases.
The Level 1 calibration attempts to estimate the durability properties by considering the
actual or approximate behaviour of lightly-trafficked or strong pavements. It assumes that the
predicted structural effects are correct. Thus, it is necessary to estimate the quality of each
bituminous surfacing type and likely adjustment required.
Step 1 Evaluate the quality of the available refined bitumens:

! High quality (HB)
Refined by international oil companies for specific uses in the highway

industry, from selected crude sources with low wax content.
! Low quality (LB)
Produced by local refineries or from high-wax crude sources, with poor

oxidation resistance.
Step 2 Evaluate the likely oxidation by the atmosphere on exposed road surfaces given
the local climate:
! Highly oxidising (HO)
Low cloud cover, high incidence of sunshine, high altitudes, depleted

ozone area.
! Moderately oxidising (MO)
Mixed conditions of sunshine hours and cloud cover, low to medium

altitudes.
! Low oxidising (LO)
Frequent cloud cover, low altitudes, cool subhumid or high rainfall

climate.
Step 3 Evaluate the construction quality:
! High (HC)
Careful binder temperature control, adequate binder content, low air voids
in asphalt mixtures (or good compaction), high standard of mixing plant
and compaction equipment, use of medium or soft bitumens (for example,
80/100 penetration or higher).
! Fair (FC)
Moderate or variable adherence to the qualities above.

! Low (LC)
Frequent over-heating of binder, low binder content, high air voids in

asphalt mixtures (or poor compaction), extensive use of hard bitumens
(40/50 or 60/70 penetration).

Version 1.0
RDWE calibration
Step 4 Select adjustment factor based on binder quality, oxidising climate and
construction quality from Table 7.5.
Table 7.5
Level 1 adjustment factor for cracking initiation
Construction Bitumen quality Oxidising climate

quality High Medium Low
High 1.0 1.2 1.5
High
Low 0.8 1.0 1.1
High 0.8 1.0 1.1

Medium
Low 0.6 0.8 0.9
High 0.6 0.8 0.9

Low
Low 0.4 0.6 0.7
7.2.3 Cracking progression adjustment factor: S-I

The rate of cracking progression in the analysis year is a function of the area of cracking,
surfacing type and other factors. The adjustment factor multiplies the amount of increase in
area of cracking, so factor values greater than 1 accelerate the progression of cracking.
For a Level 1 calibration, it is recommended that the factor be taken as the inverse of the
cracking initiation factor, that is:
1
K cp = …(7.5)
K ci
7.2.4 Rut depth progression adjustment factor: S-I

In HDM-III, the predicted increase in rut depth in an analysis year is a strong function of the
existing:
! Rut depth (-)
! Traffic loading (+)
! Pavement strength (-)
and slightly sensitive to the amounts of:

! Cracking and rainfall (+)
and whether the surfacing is an:

! Overlay (-)
Progression at the early stages is a strong positive function of relative compaction, which is
the average ratio of layer density to reference compaction standards. After the first 1-3 years,
the rate of progression slows down and only accelerates in the presence of cracking and rain.
The adjustment factor is a direct multiplier of the predicted increase so a higher factor
accelerates rut depth progression.

Version 1.0
RDWE calibration
If the predicted rut depth is low when the pavement is young (< 5 years) but the observed rut
depth is much higher, then the adjustments should first be made through decreasing the value
for the relative compaction.
The HDM-III prediction model is a good representation of the effects of structural
deformation and densification through the pavement and subgrade. For badly cracked
pavements in wet climates, the model may underestimate the rate of progression. The model
does not represent the:
! Plastic deformation - which can occur in thick bituminous surfacings under high
temperatures and heavy loading, nor
! Abrasion - that occurs under studded tyres used in some freezing climates
In HDM-4, the rut depth progression model has been decomposed to separately represent the
four components of rutting:
! Initial densification
! Structural deformation
! Plastic deformation
! Surface wear due to studded tyres
Each component may be independently calibrated using separate K adjustment factors (see
Chapter C2, Section 10 of Analytical Framework and Model Descriptions).
The procedures given in this edition of this Guide relate primarily to calibrating the
prediction of rut depths in HDM-III. Further guidance on calibrating the HDM-4 rut depth
component models will be given in a future edition of this Guide. In the meantime, the
procedures given in this Guide for calibrating rut depth progression will assist in planning an
HDM-4 rut depth model calibration study.
In most cases for a Level 1 calibration of HDM-III or HDM-4, adjustment is not considered
necessary. Economic results in the model are barely sensitive to rut depth and the small
benefits derive from reduction of the rut depth variation that is associated with rut depth.
However, if one of the exceptional situations mentioned above is significant and dominant,
the user should consider some adjustment, as follows:
1 High temperature - plastic flow deformation in asphalt
Effects are visible as wavey surface on upgrades or at intersections or as ridge alongside

rut, formed by lateral displacement.
❏ Used pavements:
Krp = 1.5 – 4; or,
❏ New pavements:
Krp = 1.2 – 3.
2 Abrasive wear - from studded tyre use
Effects are visible as pock-marked surface wear in a rut without side ridges.
❏ All pavements:
Krp = 1.2 – 2
3 Cracked pavement - clayey or silty base materials, and high to moderate rainfall:

Version 1.0
RDWE calibration
❏ All pavements:
Krp = 1.5(1.1 – 2)
As the available adjustment is approximate for traffic volume-related distress (cases (1) and
(2)), a wide degree of latitude is provided and iterative adjustments may be needed until the
results meet a test of reasonableness (if the user has some data for verification). If calibration
of this mode of distress is important to the user, a Level 2 or 3 calibration is advisable.
7.2.5 General roughness progression factor: S-III

The structure and coefficient values of the roughness progression prediction have proved to
be very reliable throughout many countries and climates, so no adjustment is considered
necessary in this factor in a Level 1 calibration.
7.2.6 Ravelling initiation factor: S-III

As the adjustment factor for ravelling initiation has low impact on most applications, it is
usually reasonable to retain the default value of 1 for it. However, if ravelling and potholing
are extensive in practice, the user may wish to make adjustment.
The ravelling initiation factor in HDM applies only to surface treatments. Its value is likely to
be similar to the cracking initiation factor when the ravelling is caused by oxidation and
hardening of the bituminous binder. A high incidence of hydrophilic mineral aggregates,
siliceous minerals such as quartzite or chert, may cause a high incidence of stripping and thus
early ravelling. Poor quality construction practices, such as contaminated stone, poor
compaction or wet weather, have a strong influence but are allowed for separately in the input
variable Construction Quality (HDM-III) or Construction Defects Indicator for Surfacing
(CDS) in HDM-4. The HDM-III factor Krv is equivalent to the HDM-4 factor Kvi for ravelling
initiation.
Typical values may vary widely (but have low importance in the model), for example:
1 Oxidation-sensitive binder or climate
Krv = Kvi = Kci

2 Siliceous aggregates and moderate-high rainfall
Krv = Kvi = 0.5 (0.1 – 0.8)

3 Binders modified with adhesive agent or anti-oxidant
Krv = Kvi = 1.3 (1.0 – 1.6)
7.2.7 Potholing progression adjustment factor: S-IV

As the adjustment factor for potholing progression has low impacts on most maintenance
alternatives except patching and extremely low maintenance, it is reasonable in most cases to
adopt the default value of 1. However, if surface disintegration is dominant in practice, if
patching alternatives are expected to dominate, and if there is evidence from preliminary runs
of the model that the predicted area of potholing is very different from observed data, the user
may wish to make adjustment.
The progression of potholing is highly variable and unpredictable. It depends not only on
factors such as the amount of potholing and cracking which are in the model, but also on
factors that are difficult to quantify, such as how sensitive the base material is to
disintegration and to moisture, which are not in the model. Potholing area is typically small
(from <0.01 to 1.0 %) (1 % is about 50-100 potholes per 100 m) and has a fairly strong

Version 1.0
RDWE calibration
influence on roughness inside the model, so the user is advised to be careful not to exaggerate
the effects.
The adjustment factor directly multiplies the incremental area of potholing of the analysis
year. Adjustments may vary widely, for example, Kph = 0.3 - 3.0.
7.3 Level 2 - Calibration of primary relationships
7.3.1 Roughness-age-environment deterioration factor: S-I
Pavement environment coefficient (m)

The pavement environment coefficient, m, can be estimated directly from samples of
pavements selected in several different environmental zones. The calibration segments should
be selected by the following criteria:
Field sampling
About five pavement segments are selected in each of 2-4 climatic zones as follows:
1 Select 2-4 climatic zones representative of the study area in which each climate zone is
typical of one (or, at most, two adjacent zones) of the moisture-temperature zones
defined in Table 7.2, for example, semi-arid/subtropical, or humid-perhumid/tropical.
2 Select about 5 (between 3 and 10) pavement segments in each zone as follows:
(a) Segment
Select segments about 5 km long (minimum 3 km, maximum 10 km), which may
comprise non-contiguous sections of a minimum 1 km length on the same road and
reasonably homogeneous in the basic data (type 1) in Field measurements below.
(b) Pavement
Select pavement (rehabilitation) age preferably > 10 year (minimum 6 year.)
Note: AGE3 is the age since the last overlay, reconstruction or construction
whichever is the most recent, but excluding surface treatments.
Field measurements
For HDM-4 roughness model relationships, refer to Chapter C2, Section 11 of Analytical
Collect the following data:
1 For all sections
Pavement age (AGE3), cumulative traffic loading (NE ESA), roughness (m/km IRI),
pavement structural number (SNP), pavement type, and drainage environment type.
2 If pavement surface distress exceeds average PSD 2.0 - then also collect:
RDS, ACRX, APAT, APOT.

3 Determine average values of each parameter - for each calibration segment.

Version 1.0
RDWE calibration
Evaluation
The parameter m is estimated directly from the summary model for roughness progression
(Paterson and Attoh-Okine, 1992), assuming the rest of the summary predictive model is
correct, as follows:
1 Estimate original roughness, RI0 (m/km IRI) for each pavement type within either the
climate zone or the study zone, whichever is considered most practicable. This is best
estimated from pavements of the given type less than 3 year old with mean PSD < 1.0,
but also may be made from other estimates (this estimate does not need to be restricted to
the same pavements as selected for the m-calibration).
2 Estimate m from the summary model rewritten as follows depending on which data are
available, and using a spreadsheet or the worksheet provided in HDM Tools:
m=
{ ln[RI ] − ln[RI
t 0 + 263 NE(1 + SNP )
-5
]} …(7.6)
AGE3
m=
{ln[1.02 RI - (0.143 RDS + 0.0068 ACRX + 0.056 APAT )] − ln[RI
t 0 + 263 NE(1 + SNP )
-5
]}
AGE3
…(7.7)
Adjustment factor Kge (HDM-III) or Kgm (HDM-4)

Once the observed values of m have been determined for the representative climate zones, the
factors Kge or Kgm will normally be set to 1.0 and adjusted only for the Road Construction
and Drainage quality factor, km, according to Table 7.4. In order to calibrate those km factor
values as well as m, the calibration study sampling should include construction and drainage
type (in accordance with Table 7.4) as a selection parameter in addition to climate zone in
Field Sampling (1) above.
If the observed values of m differ from the standard values (Table 7.3) by more than half of
the difference between m-values for adjacent environmental zones, the assumptions and work
of the calibration study should be carefully reviewed to identify possible errors before
adopting the observed values.
7.3.2 Crack initiation adjustment factor: (S-II or S-I)

Cracking initiation is predicted as the surfacing age when fatigue cracking becomes visible,
with a minimum area of 0.5 per cent of the carriageway area (or about 1.8 m2 of 100 m lane
length). The Level 2 calibration takes account of both the implicit and explicit predictive
parameters through direct field observations. Implicit parameters include the material design
and climate-related ageing which both affect fatigue behaviour - cracking predictions are
therefore calibrated specific to the surfacing type and the climate.
Major surfacing types are represented by separate equations, but in some instances the user
may also wish to define subtypes where the behaviour is expected to be significantly
different. For example, to distinguish asphalt concrete and gap-graded or hot-rolled asphalt
within Asphalt Mixture; or double chip seal, cape seal, and penetration macadam within
Surface Treatment. Explicit parameters include pavement structural parameters (such as
structural number, deflection, surfacing thickness, etc.) and traffic loading parameters (for
example, annual loading (YE4), etc.) - these are measured on the sample pavements.

Version 1.0
RDWE calibration
Field sampling
First, stratify the sample by:
1 Surfacing type and subtype (material type); and
2 Climate - based on oxidising potential (see Section 7.2.2).
This stratification needs to be made realistically into the minimum number of surfacing-
climate groups that is considered necessary and practical to be applied (typically 1-4). Note
that original surfacings need to be classed as separate groups from second and subsequent
surfacings (for example, overlays and reseals).
Second, in each surfacing-climate group, identify a minimum of 15 pavement sections (of
300 m lane length) with Low surface distress (less than 5 per cent area of cracking), selected
from a wide range of annual traffic loading (ESA/lane/year) representing the range found on
the network. The requirement for low surface distress is ideal (since it relates to the initiation
state). It should not include sections with nil cracking unless:
The surfacing age is Medium (6-15 yrs) or Old (> 15 yrs) and they comprise less than 20 per
cent of the sample; and sections of Medium distress (5-30 per cent cracking) may be included
where necessary, but with adjustment to the observed surfacing age (for example, reduced by
one year for each 10 per cent of cracked area).
For a high precision calibration (S-I), to be applied when cracking is considered the primary
intervention parameter or when higher precision is generally preferred, the size of the sample
should be increased to a minimum of 30 pavement sections in each surfacing-climate group.
Field measurements
On each identified cracking calibration pavement section, the following data should be
determined and recorded:
1 Surfacing age (years)
2 Percentage area of all cracking (more than 1 mm width)
3 Percentage area of wide cracking (more than 3 mm width or spalled narrow cracks)
4 The explicit independent parameters relevant to each surface type relationship
Evaluation
For cracking model relationships, refer to Chapter C2, Section 5.1 of Analytical Framework
and Model Descriptions.
For each pavement section data record, the predicted all cracking initiation age should be
calculated by the model relationship using a spreadsheet or the HDM Tools software (see
Appendix K). The calibration adjustment factors and prediction errors should be determined
separately for each surfacing-climate group from the mean predicted (mean PTCI) and
observed initiation (mean OTCI) ages as follows:
mean OTCI
K ci = …(7.8)
mean PTCI
{ [
RMSE = SQRT mean (OTCI j - PTCI j )
2
]
j=1,n
} …(7.9)

Version 1.0
RDWE calibration
7.3.3 Cracking progression factor: S-III

Cracking progression is predicted as a function of the area cracked and surfacing type in a
sigmoidal function. The variance of progression data is typically very high so the required
sample size is similar to cracking initiation though the sensitivity is lower. In general, a
calibration derived from the initiation adjustment factor (similar to Level 1) is considered
adequate without further field data collection, that is:
1
K cp = …(7.10)
K ci
Alternatively, the following explicit field sampling based approach may be adopted:
Field sampling
First, stratify the sample by surfacing type. In most cases, further stratification by surfacing
material subgroup and climate type (as for cracking initiation) is unlikely to add reliable
differences, but may be done optionally.
Second, in each surfacing group, identify a minimum of 15 pavement sections (of 300 m lane
length) with either medium surface distress (5-30 per cent area of cracking), or high surface
distress (>30 per cent area of cracking).
Field measurements
On each identified cracking calibration pavement section, the following data should be
determined and recorded:
1 Surfacing age (years)
2 Percentage area of all cracking (more than 1 mm width)
3 Percentage area of wide cracking (more than 3 mm width or spalled narrow cracks)
4 Values of the explicit independent parameters required for predicting cracking

initiation and progression
Evaluation
For cracking model relationships, refer to Chapter C2, Section 5.1 of Analytical Framework
For each pavement section data record, determine:
1 The predicted cracking initiation age, calculated by the model relationship using a
spreadsheet or the HDM Tools software and adjusted by the calibration adjustment
factors determined under Section 7.3.2
2 The estimated age since cracking initiation, calculated by subtracting the predicted
cracking initiation age from the observed surfacing age
3 Fit a sigmoidal curve to the observed cracking area versus the estimated age since
initiation data, and determine the estimated age at 30 per cent cracking area (ET30) by
interpolation or extrapolation
4 Calculate the predicted age at 30 per cent cracking area (PT30) using the HDM-4
equation with coefficients appropriate for the pavement and surface type

Version 1.0
RDWE calibration
5 Calculate the adjustment factor from the mean values of ET30 and PT30 across all
calibration sections separately for each surface type.
mean PT30
K cp = …(7.11)
mean ET30
7.3.4 Rut depth progression factor: S-II/S-III

The progression of mean rut depth is predicted by a single relationship for all flexible
pavement types. For HDM-III, calibration is achieved just by comparing the observed and
predicted mean rut depths for a range of pavements. For HDM-4, the prediction is determined
from four components that can be calibrated separately. A Level 2 calibration focuses on the
plastic deformation component and can be applied for either version of HDM as follows:
Field sampling
Identify a minimum of:
! 20 pavement sections - (200 m lane-length) if the flexible pavements are
predominantly of the thin surfacing type (less than 50 mm bituminous material), or
! 30 sections - (about half thin and half thick) if thick asphalt pavements are common
(more than 50 mm thickness of asphaltic layers).
At least 50 per cent of the sample should have medium to high mean rut depth (greater than 6
mm) and the range should include the highest level of rutting prevailing on the road network.
In high standard road networks this sampling requirement may be difficult to meet, in which
case the calibration of rut depth prediction is unlikely to be important.
Field measurements
On each identified section, the following data should be measured or determined:
1 Mean and standard deviation of rut depth values - measured in each wheelpath at 10
m intervals under a 2 m straight-edge (or computed from automated non-contact survey
data for a 2 m straight-edge simulation); and
2 The IQL-2 explanatory parameters - required by the HDM-4 predictive relationships,
for example, SNP, YE4, COMP, ACX, HS etc.
Evaluation
For rut depth model relationships, refer to Chapter C2, Section 10 of Analytical Framework
For each calibration section j, compute the predicted mean rut depth (PRDMj) and the
predicted standard deviation of rut depth. Calculate the adjustment factor for mean rut depth
progression, by geometric means or from log values (LORDMj and LPRDMj) as follows:
K rp = Geometric Mean [ORDMj] / Geometric Mean [PRDMj] or

…(7.12)
K rp = [Sum (log ORDMj)] / [Sum(log PRDMj)]
Alternatively, determine Krp and the prediction error by linear regression of LORDMj v.
LPRDMj.

Version 1.0
RDWE calibration
7.3.5 General roughness progression factor: S-III

Since the roughness progression relationship predicts the change in roughness under the
influence of several factors including distress, a Level 2 calibration of Kgp requires a time-
series of at least four years of reliable roughness data on a wide range of pavement segments.
If such historical data are already available the following guidance can be used to determine
what adjustment should be made to Kgp if any. Once again, the user is advised that it is
unusual for this general factor to need adjustment and it is more likely that one of the internal
factors need correction; for example, the environmental coefficient (m or kge), structural
parameters (SNP or YE4), or the patching and potholing coefficients. If adjustment is found
to be necessary, then plans should be made to conduct the more rigorous analysis described
for a Level 3 adaptation (see Section 7.4)
Step 1 Select 20 or more pavement segments (minimum 10 per pavement type), 1 km

length, uniformly distributed in a matrix of age group (young, medium, old),
and annual traffic loading (light, medium, heavy), for each pavement type
(AMGB, STGB, etc. as appropriate)
Step 2 Using IQL-2 data from at least three consecutive applicable surveys spanning 4-
5 years, process the data to determine:
The mean and incremental values of each condition parameter over the duration
of the observed period (ACRX, ARAV, APOT, APAT, RDM, RDS, RI) and the
mean values of the pavement and traffic parameters.
The incremental values should be determined preferably by linear regression
between the first and last applicable survey (or alternatively simplified by the
difference between the averages of the first and last pairs of values, adjusted to
the full applicable period by extrapolation), and the mean values by arithmetic
averaging, for example:
∆ORI t = (EORI tn - EORI t1 ) …(7.13)
where:
EORItn is the regression estimate of observed roughness RI at time n, and tn is
the last applicable time and t1 is the first applicable survey time; or
 [AVG(ORI3, ORI4 ) - AVG(ORI1, ORI2) ] (OT4 - OT1) 
∆ORI t =   …(7.14)
 [AVG(OT3, OT4 ) - AVG(OT1, OT2 )] 
where:
ORI4 is the observed roughness RI at the t=4th (or last) observation, and
OT4 is the corresponding time in years of the t=4 observation.
SUM(ORI1 : ORIn )
MORI = …(7.15)
n
where:
the notation is similar to the above for RI.
The incremental and mean values of the other parameters would be determined
similarly.
Step 3 The unadjusted predicted value of incremental roughness (∆PRIt ) should be

calculated for each calibration segment using the primary prediction

Version 1.0
RDWE calibration
relationship (refer to Chapter C2, Section 11 of Analytical Framework and

Model Descriptions) using a spreadsheet or the HDM Tools software, for
example:
∆PRI t = f {MORI, SNP, YE4, ∆OCX t , ∆ORDS t ∆OPOT, m, MOT, etc.} …(7.16)
Step 4 Calculate the residual errors by differencing the observed and predicted values
of incremental roughness for each calibration section j, for example:
RESRIj = ∆(PRIj t - ∆ORIj t , ) …(7.17)
Determine correlation and slope (b) without intercept between RESRI and
MORI. If the correlation and the determination of ‘b’ are significant, then
determine the adjustment factor, Kgp, as follows:
K gp = 1 + b …(7.18)
Check for other influences by examining the bivariate correlations of RESRIj

with other key explanatory variables from the equation in Step 3; for example,
MOCX, SNP, MORD, etc.
7.3.6 Ravelling initiation factor: S-IV

Ravelling initiation is predicted as the surfacing age when erosion of surface material
becomes evident on at least 0.5 per cent of the carriageway area. Since this factor has
relatively low impact on results except when disintegration and potholing are prevalent, the
relationship would only be calibrated at Level 2:
1 As part of a comprehensive calibration study, or
2 If surface treatment is a prevalent pavement surface type; or
3 If ravelling is the prevalent distress for a surfacing material not specifically represented
by the surface treatment relationship.
Since ravelling is evidence of a shortcoming in material design, specifications or
construction, and is a function of binder oxidation, ravelling distress may be concentrated in
certain districts, climatic zones or in projects of a certain period. If this tendency is evident,
the field sampling should group similar problem areas together so those effects can be
quantified, and keep other problem-free areas in a separate group.
In the HDM-III ravelling initiation model, this disparity of performance is captured by the
Construction Quality parameter, CQ, which is set to CQ=0 for normal performance and CQ=1
for premature failures. In the HDM-4 ravelling initiation model, construction defects in the
surfacing are captured by the parameter CDS (refer to Chapter C2, Section 4 of Analytical
Framework and Model Descriptions).
Calibration may be performed on the combined groups if the CQ or CDS values have been
assigned, or may be performed on the separate groups if specific, fractional values of CQ or
CDS are to be determined by the calibration.
Field sampling
The sampling should include about 15 pavements in each stratum as follows:
! Samples should be stratified by:
❏ surface type and material type; and

Version 1.0
RDWE calibration
❏ problem and non-problem areas, defined as described above, where problem areas
have been assigned a non-zero CQ value, or where separate adjustment factors are
to be determined for each area.
! In each group, select a minimum of 15 pavement calibration sections, of 300 m lane
length, either with:
❏ low positive incidence of ravelling (0 < area < 10 %), or
❏ nil incidence of ravelling if the surfacing age is six years or older (provided that this
group does not exceed 20 per cent of the total sample).
Field measurements
On each calibration section, the following data should be measured or recorded, in
accordance with the HDM-4 relationships:
! Surface type
! Material type
! Surfacing age
! Annual vehicle axle passes (YAX)
! Assessed construction defects indicator for bituminous surfacings (CDS)
! Assessed ravelling retardation factor (RRF)
Evaluation
For ravelling initiation model relationships, refer to Chapter C2, Section 6.1 of Analytical
For each calibration section, the observed time to ravelling initiation (OTRVj) for the groups
should be set as shown below:
! Low area of ravelling
Should be set equal to 10 per cent less than the surfacing age, that is, 0.9 AGES.
! Nil ravelling (OTRV)
Should be set to 20 per cent more than the surfacing age, that is, 1.20 AGESj.
The predicted time to ravelling (PTRVj) should be determined for each section by applying
the appropriate HDM-4 relationship and coefficients, using a spreadsheet or the worksheet in
HDM Tools. The calibration factor for ravelling initiation is determined for each strata group
(g) as follows:
mean[OTRVj(g)]
K vi (g) = …(7.19)
mean[PTRVPj(g)]
HDM-4 also provides for a separate calibration factor for ravelling progression Kvp. It may be
derived in a similar way to the initiation factor, or else set to equal the inverse of the initiation
factor.
7.3.7 Potholing adjustment factor: S-III

In HDM-III, the rate of potholing progression is adjusted by the factor Kph, but potholing
initiation and the contribution of potholing to roughness are fixed. As mentioned in Section

Version 1.0
RDWE calibration
7.2.7, potholing progression is highly variable and generally difficult to predict. A local
calibration is particularly valuable when potholing is highly prevalent and severe in
substantial portions of the road network, because potholing progression is dependent on the
sensitivity of the base material to disintegration and moisture and is a function of construction
and maintenance practices. Otherwise, if potholing is rare or quickly repaired by routine
maintenance, a detailed calibration of potholing progression is not worthwhile and may be
omitted.
In HDM-4, the conceptual approach to modelling of initiation and progression of potholes has
been expanded from that of HDM-III, and allows for the separate internal adjustment of
contributions to potholing from structural cracking, ravelling and enlargement of potholes
(refer to Chapter C2, Section 7 of Analytical Framework and Model Descriptions).
The following procedure relates only to calibrating the prediction of potholes in HDM-III.
Adjustment of any of the HDM-4 coefficients should not be attempted in a Level 2
calibration.
Field sampling
Select about 10 pavement sections of 100 lane-m area with incidence of potholing ranging
from 10 to 500 pothole units (10 litre volume, about 0.1 m2 area), that is, from 1 m2 or 0.3 per
cent, up to about 50 m2 or 15 per cent of carriageway area.
Field measurements
On each calibration section, measure or estimate the IQL-2 parameters used in the HDM-III
predictive relationship, that is, pavement surface thickness, annual average heavy axles,
surfacing age (AGES), area of open potholes and estimation of number of pothole units.
Evaluation
For each calibration section, estimate the time for initiation of cracking (PTCI), the time for
initiation of potholing (PTPI), and the time for progression of potholing to X units (PTPX) up
to 500 potholing units. Compute the observed and predicted potholing times as follows:
OTPXj = AGES - PTPIj
PTPXj = PTPXj - PTPIj
Determine the potholing adjustment factor, either by linear regression of OTPXj against
PTPXj, or as follows:
mean(OTPXj)
K ph =
mean(PTPXj)
7.4 Level 3 - Adaptation of model

A Level 3 study adopts a fundamental approach to the development of predictive models, for
the purposes of:

Version 1.0
RDWE calibration
1 Improving the accuracy of specific predictions;

2 Determining coefficients or new relationships for new distress types, specific materials
or pavements not directly covered by the model; or
3 Enhancing the capability of the model to assess the influence of additional factors in
existing relationships.
The Level 3 approach uses the framework of the model forms in the HDM-4 RDWE as the
starting point for identifying either a new set of coefficients or also some modifications to the
set of explanatory parameters in individual predictive relationships. In this way, the user is
guided to make maximum utilisation of best practice and existing knowledge. However, users
are not constrained by the existing model forms and may evaluate other potential model forms
as they wish.
7.4.1 General design and methods of experimental study

The keys to successful development of predictive models that are robust and reliable lie in the
adoption of formal experimental design principles and advanced statistical techniques. The
strength of the HDM road deterioration and maintenance effects models and their ability to be
transferable to many different countries and conditions is due to several fundamental
characteristics as follows:
! Factorial design
The main factorial matrix should comprise the primary explanatory variables and cover a
range of values as close as possible to the range over which they will be used. The
HDM-III and HDM-4 relationships, and the sensitivity classes assigned in Chapter 4, are
the starting point to identify which parameters are primary and which secondary. When
important interactions are to be determined, especially between parameters that may
exhibit collinearity, there must be sufficient range for the cross-effects to be estimated.
But to estimate the effect statistically it is necessary to have data with high, normal and
low traffic loading on a given SNP, and strong, normal and weak pavements under a
given loading YE4. It is not necessary to fill all cells of a factorial for all primary
parameters; a partial factorial can be used to reduce the number of sections to be found.
A summary of factorial designs for road deterioration studies is given in Paterson (1987)
and GEIPOT (1982), and design principles can be found in relevant textbooks.
! Cross-sectional and longitudinal analyses
A cross-sectional analysis, or slice-in-time across pavements at differing stages of the

life-cycle and differing factor values will only be valid when all relevant factors are
combined in the correct model form. The experience of the HDMS study shows that
complex interactions and even basic pavement performance models can only be achieved
by a combination of longitudinal analysis, or time-series, of data from pavements
monitored continuously over a period of four or more years, and cross-sectional analysis.
The impact is illustrated in Figure 7.1, and pertinent examples are found in the
discussion of roughness progression and rut depth progression models in Paterson
(1987) - (Chapters 7 and 8).
! Discrete events and analysis of censored data
For discrete events such as the initiation of cracking, ravelling or potholing, studies
which are limited to only pavements showing the initial stages of distress can risk being
biased if the set excludes pavements that out-perform the norm or that have premature
failures. The application of maximum likelihood statistical estimation (MLE) enabled

Version 1.0
RDWE calibration
the HDM-III analyses to determine the concurrent effects of ageing and fatigue, and the
MLE techniques have become readily accessible now in currently available advanced
statistical software.
! Analysis of residuals
Analysis of the residual errors of a statistical model should always be used to determine
whether they correlate with any of the possible explanatory variables and to reveal any
bias trends that may remain in the model. This can lead to the testing of further model
formulations and improving their explanatory power.
! Model form
The formulation of predictive models was derived in the first instance from:
❏ The received knowledge of mechanistic and empirical research,
❏ Selecting and clustering explanatory parameters on the basis of a mechanistic model
where this was available, and
❏ Adding other parameters in the forms in which they were expected to interact, for
example, in multiplicative, additive, or power functions.
Advanced statistical methods were used to test the validity of the relationships, derive
the observed coefficients and determine the real significance of individual parameters or
parameter clusters. Thus they represent structured empirical models, combining
mechanistic and logical theory with empirical estimation. The resulting model forms are
therefore sound starting points for further research. It is strongly advised that Level 3
adaptation always begin with a statistical re-estimation of the coefficients of the HDM
relationship directly from the local data.
7.4.2 Distress initiation models

The field sampling procedure outlined in the Level 2 section (sections 7.3.2, 7.3.6 and 7.3.7)
should be applied with two modifications, that is:
1 The sample sizes should be doubled or tripled to a minimum of 30 per surfacing type-
material type combination; and
2 The sections selected for modelling of distress initiation should be monitored over a 4-6
year period to derive distress progression data on the same sections, so that the direct
interactions between initiation and progression behaviour can be identified and
characterised.
Two sets of models should be evaluated:
1 The structured model defined for HDM (that is, the HDM-4 road deterioration model for
bituminous pavements described in Chapter C2 of Analytical Framework and Model
Descriptions), specifying the same model parameters and estimating the coefficients
from the local data; and
2 Variations on the structured model, including for example the mechanistic (strain-based)
relationships tested in (Paterson, 1987 - chapters 5 and 6) and others.
Some tips from previous research:
1 Structural and traffic loading variables must always be clustered interactively and not
additively.

Version 1.0
RDWE calibration
2 In relationships which use a log transform to represent an interaction by an additive

function, care must be taken that the rate of trafficking (for example, ESA per year)
preserves dimensional consistency in both the dependent and independent parts of the
relationship
7.4.3 Distress progression models

7.4.4 Rut depth progression models

7.4.5 Roughness progression models

7.5 Works effects on bituminous paved roads: S-II

In modelling the life cycle pavement deterioration it is vital that the effect of works on
pavement condition be adequately modelled. These effects may vary between pavement types
and construction practices in different countries. This is illustrated in Figure 7.1, which
compares the effects of overlays on roughness compares models from HDM-III (regular
paver), Thailand, Indonesia, and Sweden for an overlay thickness of 50 mm (NDLI, 1995b).
5
IRI After Overlay (m/km)
4 HDM-III
Thailand
Indonesia
Sweden
3
0
0 2 4 6 8 10 12
IRI Before Overlay (m/km)
Figure 7.1 Effect of overlays on roughness (NDLI (1995b))
The effect of overlays on roughness is a critical parameter as it dictates the future

deterioration rate and thus, maintenance activities. It is therefore an S-I parameter.
In HDM-III the overlay-roughness effects were either predicted using an internal relationship
or the roughness after maintenance was specified on the C-Series data cards. Unfortunately,
the relationship was hard coded in the model and could not be adapted without modifying
the source code. In HDM-4 the coefficients of the relationship can be specified.

Version 1.0
RDWE calibration
7.5.1 Level 1
A Level 1 calibration applies a constant value for the roughness after an overlay that is
applied to all pavements receiving a particular treatment. The default values in HDM-4 may
be used. However, since this is an S-II or S-I sensitivity, it is valuable to use some simple
field measurement. A number of pavements which have recently been overlaid and for which
the current roughness is available should be identified. On the basis of their roughness levels
an average roughness after overlay is established. If different overlay thicknesses are used the
sample can be stratified by overlay thickness and different values determined as a function of
thickness.
7.5.2 Level 2
A Level 2 calibration conducts a small study into the effects of overlay on roughness and
develops a local relationship for the effects of overlay on roughness. A series of sections
scheduled for overlays should be selected. There should be at least five sections, with a
minimum length of 500 m each. The roughness should be measured on each section. If a
manual method is used (for example, Dipstick, Walking Profilometer) only a single run is
required in each wheelpath; if using a vehicle a minimum of five runs is recommended.
In addition to measuring the roughnesses of the sections selected for maintenance,
measurements should also be made on untreated sections on the same road near the
maintenance sections. These provide control sections: since no maintenance is scheduled for
them, and the time between surveys is short, it can be expected that there will be little, if any
change in the roughness. They can therefore be used to correct the data for any roughness
measurement instrument drift over time.
The before and after roughness data should be analysed to establish a general relationship
between these roughnesses. If different overlay thicknesses are available the relationship
should take this into account.
If possible, speed measurements should be made at the same time, as this will allow for the
effects of roughness on speed to be established as well.
7.6 Concrete pavement deterioration

7.7 Concrete pavement works effects

7.8 Unpaved roads deterioration and works effects


Version 1.0
Summary and conclusions
8 Summary and conclusions


Version 1.0
References
9 References
Arup (1992). Maintenance Strategy for Songea-Makambako Road Tanzania 1993-1997.
Report to ODA, Arup Aikins, London.
Bennett, C.R. (1985). A Highway Economic Evaluation Model for New Zealand. School of
Engineering Report 368, Department of Civil Engineering, University of Auckland,
Auckland, New Zealand.
Bennett, C.R. (1989). The Depreciation of Motor Vehicles in New Zealand. Road Research
Unit Occasional Paper, Transit New Zealand, Wellington, New Zealand.
Bennett, C.R. (1994). A Speed Prediction Model for New Zealand. School of Engineering
Report 368, Department of Civil Engineering, University of Auckland.
Bennett, C.R. (1995). Modelling Depreciation, Interest, Utilisation, Crew and Overhead
Costs in HDM-4. Report to the International Study of Highway Development and
Management Tools, University of Birmingham.
Bennett, C.R. (1998). Modelling Road User Effects in HDM-4. International Study of
Highway Development and Management Tools, University of Birmingham.
Bennett (1999). Generating RUE Elasticities for the HDM-4 Calibration Reference Manual.
Internal Report, HTC Infrastructure Management Ltd., Auckland.
Bennett, C.R. and Dunn, R.C.M. (1992). Error Considerations in Speed Surveys. Proc. ARRB
Conference.
Bennett, C.R. and Greenwood, I.D. (1999). Modelling Road User and Environmental Effects
in HDM-4. Final Report Asian Development Bank Technical Assistance Project
5819-REG. To be published as Volume 7, the Highway Development and
Management Series, the World Road Association (PIARC) and the World Bank,
Paris.
Biggs, D. C. (1987). Estimating the Fuel Consumption of Light to Heavy Vehicles. Australian
Road Research Board Internal Report AIR 454-1, Nunawading, Australia.
Button, K. and Ngoe, N. (1991). Vehicle Ownership and Use Forecasting in Low Income
Countries. TRL Contractors Report 278. Transport Research Laboratory,
Crowthorne.
CES (1989). Vehicle Fleet Modernisation and Road Use Charges Study. Report to
MOST.Consulting Engineering Services (India) Private Limited, New Delhi, India.
CES (1991). Development of Long Term Plan for Expressways in India. Report to the
Ministry of Surface Transport, New Delhi.
Cenek, P. D. and Patrick, J. E. (1991). Prediction of Road Roughness Prediction. Central
Laboratories Report 91-29301. Mworks Consultancy Services, New Zealand.
CESTRIN (1994). Vehicle Operating Costs: Derivation of Vehicle Fleet Characteristics and
Unit Costs. Report to Centre for Road Engineering Studies, Bucharest.
Chamala, R. (1993). Economic Maintenance Intervention Standards. Report to the
Pavements and Asset Strategy Branch, Queensland Transport, Brisbane.
Chesher, A., and R. Harrison (1987). Vehicle Operating Costs: Evidence from Developing
Countries. World Bank Publications, Washington, D.C.

Version 1.0
References
CID. (1993) Projet Routier Complement du Programme oes Routes A Elargir, Rapport de
Presentation. Direction des Routes et de la Circulation Routiere. Ministere des
Travaux Publics de la Formation Profressionnelle et de la Formation de Cadres.
Royaume du Maroc.
Cox, B.E. (1983). The Value of Travel Time. RRU Occasional Paper. Transit New Zealand,
Wellington.
Curtayne, P.C., Visser, A.T., du Plessis, H.W. and Harrison, R. (1987). Calibrating The
Relationships Between Operating Costs of Buses and Road Roughness on Low-
Volume Roads. Transportation Research Record 1106, Transportation Research
Board, Washington, D.C.
Daniels, C. (1974). Vehicle Operating Costs in Transport Studies. EIU Technical Series No.
1, The Economist Intelligence Unit, London.
du Plessis, H.W. and Meadows, J.F. (1990). A Pilot Study to Determine the Operating Costs
of Passenger Cars as Affected by Road Roughness. In Report PR 88/070/3, Roads
and Transport Technology, CSIR, Pretoria, South Africa.
du Plessis, H.W., Visser, A.T. and Harrison, R. (1991). The Effects of Road Condition on the
Operating Costs of Buses. In Report PR 88/070/3, Roads and Transport Technology,
CSIR, Pretoria, South Africa.
du Plessis, H.W. and Schutte, I.G. (1991). Road Roughness Effects on Vehicle Operating
Costs: Southern Africa Relations for use in Economic Analyses and in Road
Management Systems. Report PR 88/070/3, Roads and Transport Technology, CSIR,
Pretoria, South Africa.
Estudio (1993). Estudio de Implantacion de los Modelos HDM-III y EBM para la
Preparacion de Programas de Mantenimiento y Mejora de la red de Carreteras de la
Generalitat de Catalunya. Report to Department de Politica Territorial I Obres
Publiques Direccio General de Carreteres, Generalitat de Catalunya. Estudio HC,
SA., Barcelona.
Findlayson, A.M. and du Plessis, H.W. (1991). Operating Costs of Medium to Heavy Trucks
as Affected by Road Roughness. In Report PR 88/070/3, Roads and Transport
Technology, CSIR, Pretoria, South Africa.
Fonseca, V. M. R., Queroz, C. and dos Santos A. B. (1991). Economic Analysis of Guinea-
Bissau Roads Program Using HDM-III and EBM Models. 5th International
Conference on Low Volume Roads, North Carolina State University, Raleigh, North
Carolina, USA.
Gaete, R., Echeverria, G., Videla, C. and Gutierrez, P. (1991). Sistema de Autocalibracion de
los Modelos de Deterioro del HDM-III. Proc. Sexto Congreso Ibero - Latino-
Americano del Asfalto, Santiago, Chile.
GEIPOT (1982). Research on the Interrelationships Between Costs of Highway Construction,
Maintenance and Utilization. Final Report on the Brazil-UNDP Highway Research
Project. (12 Volumes). Brazilia, Brazil.
GITEC (1992). Estudio de Factibilidad Para El Programa De Mantenimiento Vial. Report
to the Ministry of Communications, Republic of Guatemala.
Greenwood, I.D. (1997). Calculating Tyre Diameter. Memo to the ISOHDM Study,
University of Birmingham.

Version 1.0
References
Greenwood, I.D. (1998). Correspondence with the International Study of Highway

Development and Management Tools. University of Birmingham.
Hamilton, L.C. (1990). Modern Data Analysis - A First Course in Applied Statistics.
Brooks/Cole Publishing Company, Pacific Grove, California.
Hine, J.L. (1991). Pakistan Road Freight Industry: The Productivity and Time Use of
Commercial Vehicles. TRL Report 333, Transport Research Laboratory, Crowthorne.
Hine, J.L. (1996). Brtiefing Paper on Vehicle Utilisation and Service Life for the December
1996 Workshop on HDM-4. Report to the International Study of Highway
Development and Management Tools, University of Birmingham.
Hoban, C.J. (1987). Evaluating Traffic Capacity and Improvements to Road Geometry. World
Bank Technical Paper 74, The World Bank, Washington, D.C.
Hoban, C.J., Reilly, W. and Archondo-Callao, R. (1994). Economic Analysis of Road Projects
with Congested Traffic. World Bank Technical Paper, Transport Division, The World
Bank, Washington, D.C.
Hodges, J.W., Rolt, J. and Jones, T.E. (1975). The Kenya Road Transport Cost Study:
Research on Road Deterioration. Department of the Environment, Transport and
Road Research Laboratory Report LR673, Crowthorne, UK
Hoff & Overgaard (1992). Road User Cost Model. Report to the Directorate General of
Highways, Jakarta.
Hoff & Overgaard (1994). Vehicle Operating Costs. Report to the Directorate General of
Highways, Jakarta.
Houcine, L. (1982). Experience Tunisienne du Modele H.D.M. Report to Direction Generale
des Ponts et Chaussees, Ministere de L’Equipement et de L’Habitat.
HTC (1999). Calibration of HDM Speed Prediction Model to Thailand. Report to Dessau-
Sophrin. HTC Infrastructure Management Ltd., Auckland.
IBRD (1990). Application du Model HDM a la RN8. Report to the World Bank,
Washington, D.C.
INDEC & Assoc. and Kampsax International (1988). Five Year Development Plan for the
National and Provincial Road Network. Report to Republic of Indonesia Ministry of
Public Works Directorate General of Highways, Jakarta.
ISOHDM (1997). Specifications for the HDM-4 Road Deterioration Model for Bituminous
Pavements – Fourth Draft. Report to the International Study of Highway
Development and Management Tools, University of Birmingham.
JBP (1990). Pavement Management Project ADB/TA 1058 - IND : Report on Economic
Studies. Report to ADB. John Burrows & Partners, Surrey.
JKR (1991). Axle Load Study Phase 2 (Peninsular Malaysia) - Final Report. Public Works
Department, Kuala Lumpur.
Kadiyali (1991). Updating of Road User Cost Study. Report to the Ministry of Surface
Transport, L.R. Kadiyali and Associates Ltd., Delhi.
Kampsax International (1986). National Highway Maintenance Study - Pakistan: Final
Report. Report to Pakistan National Highways Board, Islamabad.
Kampsax International & Associates (1992). Pakistan Farm to Market Roads Study. Asian
Development Bank Report, Karachi.

Version 1.0
References
Kampsax - Beca (1990) Provincial Roads and Bridges Investment Programme, Phase II,
Extract Ch 4: Vehicle Operating costs. Final Report. The Independent State of
Papua New Guinea, Department of Transport, Asian Development Bank, New
Guinea.
Kannemeyer, L. and Visser, A.T. (1994). Calibration of HDM-III Performance Models for
Use in the Pavement of South African National Roads. Paper offered for Presentation
at the January 1995 Annual Meeting of the Transportation Research Board,
Washington, D.C.
Kinhill (1998). Tamil Nadu Project Co-ordinating Consultancy Project. Correspondence with
authors. Madras.
Lea Associates (1988). Saskatchewan Highways and Transportation Pavement Management
Information System: PMIS Technical Manual Appendix I - Vehicle Operating Costs.
Lea Associates Ltd., Vancouver, B.C., Canada.
Liautaud, G. and Archondo-Callao, R. (1994). Brazil Federal Highway Network- A Study to
Define Appropriate Policies, Standards, Budget Allocations and Rehabilitation
Programs and Priorities for the Next Five Years 1995-2000.
Louis Berger International (1990). The Federal Trunk Roads Study - Final Report. Report to
the Federal Ministry of Works and Housing, Nigeria.
McLean, J.R. (1989). Two-Lane Highway Traffic Operations: Theory and Practice. Gordon
and Breach Science Publishers, Melbourne.
McLean, J.R. (1991). Adapting the HDM-III Vehicle Speed Prediction Models for Australian
Rural Highways. Working Document WD TE 91/014, Australian Road Research
Board, Nunawading.
Minderhoud, M.M., Botma, H., and Bovy, P.H.L. (1997). Assessment of Roadway Capacity
Estimation Methods. Transportation Research Record 1572, pp. 59-67.
Transportation Research Board, Washington, D.C.
Mrawira, D., Haas, R. and Paterson, W.D.O. (1998). Streamlining the World Bank’s HDM-III
Model for Network Level Application. Proc. 4th International Conference on
Managing Pavements, Durban.
MRCU (1992). Vehicle Operating Costs in Nepal - Working Paper. Maintenance and
Rehabilitation Co-ordination Unit, Ministry of Works and Transport Department of
Roads, Katmandu.
MVA Consultancy, Institute of Transport Studies (Leeds), Transport Studies Unit (Oxford)
(1987). The Value of Travel Time Savings. Policy Journals, Newbury, Berkshire, UK
NDLI (1991). Calibration of Vehicle Operating Costs to Thailand. Report to the Department
of Highways, N.D. Lea International Ltd., Vancouver, B.C., Canada.
NDLI (1992). Thailand Road Maintenance Project: A.D.B. T.A. 1106-THA Final Report.
N.D. Lea International Ltd., Vancouver, B.C., Canada.
NDLI (1993). Nepal Road Maintenance Programme: Final Feasibility Report. N.D. Lea
International Ltd., Vancouver, B.C. Canada.
NDLI (1994a). Road Network Rationalisation Study - Draft Final Report. Report to Public
Works Division, Barbados, N.D. Lea International, Vancouver.

Version 1.0
References
NDLI (1994b). Comprehensive Highway Maintenance Study - Phase I Final Report. Report
to Ministry of Works and Transport, Trinidad and Tobago, N.D. Lea International,
Vancouver.
NDLI (1994c). Four States Pavement Management Project - Draft Interim Report. Report to
Ministry of Surface Transport (MOST), New Delhi, India. N.D. Lea International,
Vancouver, Canada.
NDLI (1995a). Modelling Road User Effects in HDM-4. Report to the Asian Development
Bank, N.D. Lea International Ltd., Vancouver, B.C., Canada.
NDLI (1995b). Modelling Road Deterioration and Maintenance Effects in HDM-4. Report to
the Asian Development Bank, N.D. Lea International Ltd., Vancouver, B.C., Canada.
NDLI (1997). Gujarat Project Co-ordinating Consultancy: Calibration of Road User Effects.
N.D. Lea International, Vancouver, B.C., Canada.
Opus (1999). Review of VOC Pavement Roughness Relationships Contained in Transfund’s
Project Evaluation Manual. Central Laboratories Report 529277.00. Opus
International Consultants Ltd., Wellington.
Paterson, W.D.O. (1986). Data Collection for HDM-III. Internal Report. Transportation
Department, The World Bank.
Paterson, W.D.O. (1987). Road Deterioration and Maintenance Effects: Models for Planning
and Management. World Bank Publications, Washington, D.C.
Paterson, W.D.O. and Scullion, T. (1990). Information Systems for Road Management: Draft
Guidelines on System Design and Data Issues. World Bank Technical Paper INU 77,
Infrastructure and Urban Development Department, The World Bank, Washington,
D.C.
Riley, M.J. (1998). Correspondence with authors. Riley Partnership, Launceston, UK
Riley, M.J., Petersen, H.H. and Miller, L.H. (1987). Optimizing Pavement Rehabilitation and
Maintenance Policies for Rural Roads. Asia-Pacific Conference on Roads, Highways
and Bridges, Jakarta, Indonesia.
RITES (1992). Feasibility Study for Rehabilitation and/or Upgrading of Quelimane-
Namacurra Road. Report to DNEP, Maputo, Mozambique. RITES, New Delhi, India.
RITES (1994). Gujarat Pavement Management System. New Delhi.
RPT (1990). Transport Policy and Planning Project UGA/90/018 Vehicle Operating Costs
Report. Report to Ministry of Works, Transport & Communications, Republic of
Uganda, Rendel Palmer & Tritton, London.
RPT (1993). Internal Report National Highway PMS Project. Rendel Palmer & Tritton,
London.
Sammour, A. (1992). Feasibility Study Report: Swaileh-Queen Alia International Airport
Road Using Highway Design and Maintenance Standards Model, HDM-III Model.
Report to Ministry of Planning, The Hashemite Kingdom of Jordan.
Sayers, M.W., Gillespie, T.D., and Paterson. W.D.O. (1986). Guidelines for Conducting and
Calibrating Road Roughness Measurements. World Bank Technical Paper Number
46, World Bank Publications, Washington, D.C.

Version 1.0
References
Slavik, M.M. (1998). Weighing of Trucks in Motion: Calibration of Equipment and

Correction of Measurements. Proc. 4th International Conference on Managing
Pavements. Vol (1) pp. 68-82. Transportation Research Board, Washington, D.C.
Solminihac, H. E., Echeverria and Salgado, J. (1989). Pavement Maintenance Management
Research: The Chilean Experience on Asphalt Pavements. Proc. 11th IRF World
Meeting, Seoul, Korea.
Sweroad (1997). Indonesian Highway Capacity Manual (IHCM) February 1997. Report to
the Directorate General Bina Marga, Sweroad, Bandung.
SWK (1993a). Road Maintenance Study. Report to Roads Department, Ministry of Works,
Transport and Communications, Republic of Botswana. Scott Wilson Kirkpatrick,
U.K.
SWK (1993b). Rehabilitation of the Modjo-Awassa Road Project Assessment Study. Report
to Ethiopian Roads Authority of the Transitional Government of Ethiopia, Scott
Wilson Kirkpatrick. Surrey.
SWK (1993c). Feasibility Study of Mubende to Fort Portal Road. Report to the Republic of
Uganda Ministry of Finance & Economic Planning. Scott Wilson Kirkpatrick, Surrey.
Symonds (1997). Valuation of Travel Time Savings. Transfund New Zealand Research Report
93. Symonds Travers Morgan (NZ) Ltd., Wellington.
Transit (1995). Vehicle Repair and Maintenance Costs—Literature Survey and Report on
Operator Costs Survey. Transit New Zealand, Wellington.
Transit (1997). Friction & Tyre Abrasion Characteristics of New Zealand Road Surfaces.
Research Report 94. Transit New Zealand, Wellington.
Transroute (1992). Road Master Plan Project – Final Report, Appendix Volume II. Report to
the Ministry of Communications Roads and Highways Department by
Transroute/Stup/BCL Consortium, Dhaka, Bangladesh.
TRB (1997). Highway Capacity Manual. Transportation Research Board, Washington, D.C.
TRDF (1982). Vehicle Operating Costs, Fuel Consumption and Pavement Type Condition
Factors. Report to the FHWA, Texas Research and Develoment Fund.
TRL (1978). A Guide to the Measurement of Axle Loads in Developing Countries Using a
Portable Weighbridge. Road Note 40. Overseas Centre, Transport Research
Laboratory, Crowthorne.
TRL (1988). A Guide to Road Project Appraisal. Overseas Road Note 5. Overseas Centre,
Transport Research Laboratory, Crowthorne.
TSA (1995). Presentation of Input Data for HDM-III Run in 1994. Report to the Ministry of
Transport, Communication and Water Management Road Department. Transport
Science Association. Budapest.
TSPC (1992). Sri Lanka Road User Charges Study. Transport Studies and Planning Centre,
Colombo.
Watanatada, T., Harral, C.G., Paterson, W.D.O., Dhareshwar, A.M., Bhandari, A., and K.
Tsunokawa (1987a). The Highway Design and Maintenance Standards Model:
Volume 1 - Description of the HDM-III Model. World Bank Publications,
Washington, D.C.

Version 1.0
References
Watanatada, T., Harral, C.G., Paterson, W.D.O., Dhareshwar, A.M., Bhandari, A., and K.
Tsunokawa (1987b). The Highway Design and Maintenance Standards Model:
Volume 2 - User's Manual for the HDM-III Model. World Bank Publications,
Washington, D.C.
Watanatada, T., Dhareshwar, A.M., and P.R.S. Rezende-Lima (1987c). Vehicle Speeds and
Operating Costs: Models for Road Planning and Management. World Bank
Publications, Washington, D.C.
Winfrey, R. (1969). Economic Analysis for Highways. International Textbook Co., Scranton,
Pennsylvania.
World Bank (1995). Hashemite Kingdom of Jordan Third Transport Project Staff Appraisal
Report. The World Bank, Washington, D.C.
Wyley, W. J., Petts, R. C. and Brooks, J. (1986). Evaluation of the World Bank’s Highway
Design and Maintenance Standards Model on United Kingdom Highways. Contractor
Report 27, Transport and Road Research Laboratory, Crowthorne, Berkshire, U.K.
Yuli, P. (1996). Development of Speed and Fuel Consumption Models for Chinese Vehicles.
Ph.D. Dissertation, Department of Civil Engineering, University of Birmingham.
Zaniewski, J.P., Butler, B.C., Cunningham, G., Elkins, G.E., Paggi, M.S., and Machemal, R.
(1982). Vehicle Operating Costs, Fuel Consumption, and Pavement Type and
Conditions Factors. FHWA Report FHWA-PL-82-001. Federal Highways
Administration, Washington, D.C.
Zukang, Y., Zhongren, W. and Xinghua, L. (1992). Calibration of Vehicle Operating Cost
Model for Pavement Management System. Selected Papers of Tongji University,
1992. Tongji University Press, China.

Version 1.0
HDM data items and calibration factors
Appendix A HDM data items and calibration

factors
This appendix presents data items, model parameters and their calibration levels.

Version 1.0
Parameter values used in HDM studies
Table A.1
HDM data and model parameters and calibration levels
1
Sub-model Description Units Discussed Calibration level where local values
4
in section or should be established in place of defaults
appendix Level 1 Level 2 Level 3
Aerodynamic drag coefficient 6.3.10 •
Aerodynamic drag coefficient multiplier 6.3.10 •
Average annual utilisation km/year 6.3.4 •
Average service life years 6.3.3 •
Base number of retreads – NR0 •
Desired speed m/s 6.3.5 •
Engine speed – a0 6.3.12 •
RUE Engine speed – a2 6.3.12 •
Engine speed - Idle 6.3.12 •
Equivalent standard axles ESA/veh 6.3.1 •
Hours driven h/yr •
Number of axles 6.3.7 •
Number of wheels 6.3.7 •
Operating weight t 6.3.1 •
…Continued

Version 1.0
1
4
Optimal life depreciation parameters 6.3.9 •
Percentage of private use % •
Power - braking kW 6.3.11 •
Power - driving kW 6.3.6 •
Power - rated kW 6.3.6 •
Projected frontal area m2 6.3.10 •
Travel on wet roads % •
Travel on snow covered roads % •
RUE cont. Tyre type 6.3.7 •
Utilisation method 6.3.4 •

3
Volume of wearable rubber dm 6.3.8 •
Wheel diameter m 6.3.7 •
Acceleration noise - max. alignment noise 6.4.5 •
Acceleration noise - max. driver noise 6.4.5 •
Acceleration noise - max. NMT noise 6.4.5 •
Acceleration noise - max. noise m/s2 6.4.5 •
Acceleration noise - max. roughness noise 6.4.5 •
Acceleration noise - traffic noise – a1 6.4.5 •
…Continued

Version 1.0
1
4
Acceleration noise -traffic noise – a0 6.4.5 •
Fuel - fuel to power efficiency factor ml/kW/s 6.4.4 •
Fuel - ldle fuel rate ml/s 6.4.4 •
Labour model rotation – K0lh 6.4.6 •
Labour model translation – K1lh 6.4.6 •
Oil loss due to contamination l/1000 km 6.4.11 •
Oil loss due to operation 6.4.11 •
Parts model rotation - K0pc 6.4.6 •
RUE cont. Parts model translation - K1pc 6.4.6 •
Speed flow - low flow interaction level - pcse/h 6.4.2 •

Qo
Speed flow - nominal capacity - Qnom pcse/h 6.4.2 •
Speed flow - speed at nominal Capacity km/h 6.4.2 •
Speed flow -speed at ultimate Capacity km/h 6.4.2 •
Speed flow - ultimate capacity - Qult pcse/h 6.4.2 •
Speed flow - vehicle equivalency factors pcse/veh 6.4.3 •
Speed limit enforcement factor 6.4.1 •
Speed roughness - a0 6.4.1 •
…Continued

Version 1.0
1
4
Speed roughness - maximum rectified mm/s 6.4.1 •
velocity
Speed width critical width factor - CW1 6.4.1 •
Speed width critical width factor - CW2 6.4.1 •
Speed width desired speed-width factor - a1 6.4.1 •
Speed width desired speed-width factor - a2 6.4.1 •
Drivetrain efficiency •
Labour model constant •
Labour model parts coefficient •
RUE cont. Parts model - congestion factor •
Parts model - constant •
Parts model - minimum roughness •
Parts model - roughness •
Parts model age term - kp •
Parts model - roughness exponent •
Parts - roughness shape •
Power - engine accessory power - Kpea •
Power - engine power - a0 •
Power - engine power - a1 •
…Continued

Version 1.0
1
4
Power due to engine drag % •
Rolling resistance - a0 •
Rolling resistance factor - Kcr2 •
Speed curve - a0 •
Speed curve - a1 •
Speed gradient - critical gradient length - a0 •
RUE cont. Speed gradient - critical gradient length - a1 •
Speed gradient - critical gradient length - a2 •
Speed parameter - Beta •
Tyre model constant - C0TC dm3 •
Tyre stiffness - a0 •
Tyre stiffness factor - Kcs •
Tyre wear coefficient - CTCTE dm3/J-M •
…Continued

Version 1.0
1
4
Altitude m •
Area potholed % Appendix D •
Area with all cracking % Appendix D •
Area with wide cracking % Appendix D •
Average rainfall m/month Appendix D •
Base type Appendix D •
Benkelman beam deflection mm Appendix D •
Carriageway width m Appendix D •
RDWE Construction age yr Appendix D •
Effective number of lanes Appendix D •
Horizontal curvature deg/km Appendix D •
Mean rut depth mm Appendix D •
Number of surface layers Appendix D •
Posted speed limit Appendix D •
Preventative treatment age yr Appendix D •
Rise plus fall m/m Appendix D •
Roughness IRI Appendix D •
Roughness age term 7.2.1 •
…Continued

Version 1.0
1
4
Roughness progression 7.2.5 •
Sand patch texture depth mm Appendix D •
Shoulder width m Appendix D •
Structural number Appendix D •
Subgrade CBR % Appendix D •
Superelevation % Appendix D •
Surface type Appendix D •
RDWE cont. Surfacing age yr Appendix D •
Unit costs for construction and maintenance Appendix H •
Area of previous all cracks % Appendix D •
Area of Previous wide cracks % Appendix D •
Construction fault Appendix D •
Crack initiation factor 7.2.2 • •
Crack progression factor 7.2.3 • •
Cracking retardation time yr Appendix D •
Maximum acceleration noise m/s2 6.4.2 •
Number of base layers Appendix D •
Pothole progression 7.2.7 • •
…Continued

Version 1.0
1
4
Ravelling initiation 7.2.6 • •
Ravelling progression 7.2.6 • •
Ravelling retardation factor % Appendix D •
RDWE cont. Rut depth progression 7.2.4 • •
Soil cement resilient modulus GPa Appendix D •
Standard deviation of rut depth mm Appendix D •
Thickness of base layers mm Appendix D •
Thickness of surface layers mm Appendix D •
Average annual daily traffic veh/day Appendix E •
TRAFFIC Traffic growth rate %/year Appendix E •
Hourly distribution of traffic Appendix E •
Cost of cargo cost/h Appendix F •
Cost of crew cost/h Appendix F •
Cost of fuel cost/l Appendix F •
Cost of maintenance labour cost/h Appendix F •
Cost of oil cost/l Appendix F •
UNIT COSTS Cost of overhead cost/yr Appendix F •
Cost of retreaded tyre % Appendix F •
…Continued

Version 1.0
1
4
Cost of travel time cost/h Appendix F •
Cost of tyre cost/tyre Appendix F •
UNIT COSTS Cost of vehicle cost/vehicle Appendix F •
Interest rate % Appendix F •
Discount rate % Appendix H •

ECONOMIC
Analysis period yr Appendix H •

Version 1.0
Appendix B Parameter values used in HDM

studies
This appendix presents parameter values used in different HDM-III studies. It is from Bennett
(1995).

Version 1.0
Table B.1
Vehicle tare weights
Reference Country PC LDV LGV LT MT HT AT LB MB HB Comments

S M L
NDLI (1994a) Barbados 0.60 0.70 1.10 0.70 1.40 3.50 9.20 2.70 4.80 7.27
SWK (1993b) Ethiopia 0.96 1.32 1.32 4.00 7.80 14.70 2.40 8.10
GITEC (1992) Guatemala 1.00 1.50 5.40 12.70 3.10 8.10
RPT (1993) India 0.60 1.20 2.20 8.10 10.00 4.00
NDLI (1991) Myanmar 1.00 1.60 1.60 2.00 4.20 3.00
NDLI (1993) Nepal 1.00 1.30 5.70 3.89
Bennett (1989) New Zealand 1.00 1.30 1.35 2.00 2.10 12.00 14.50 14.50 10.30 10.30
Zukang et al. (1992) P.R. Chna 1.00 1.80 1.60 4.10 8.40 7.20 1.40 7.20
Kampsax - Beca (1990) Papua New Guinea 0.90 1.40 2.60 5.50 10.00 15.30
NDLI (1991) Thailand 1.00 1.30 3.40 7.60 1.40 3.40 12.00
NDLI (1994b) Trinidad 1.00 1.50 1.50 3.00 6.00 10.00
TRDF (1982) USA 1.00 1.20 1.20 1.80 1.50 4.90 7.20 11.20
Note: Tare weight in tonnes.

Version 1.0
Table B.2 Vehicle operating weights

S M L
Chamala (1993) Australia 0.95 1.15 1.35 2.50 4.50 13.00 21.00 41.00 18.00
Transroute (1992) Bangladesh 1.10 2.10 10.40 9.20
NDLI (1994a) Barbados 0.75 1.00 1.30 1.00 2.05 7.00 16.60 4.70 6.80 9.67
IBRD (1990) Burundi 1.30 3.50 2.20 8.00 15.50 29.40 2.60 LGV is Jeep
GITEC (1992) Guatemala 1.00 2.20 12.70 37.70 6.10 12.10
SWK (1993b) Ethiopia 1.40 2.30 2.30 8.40 14.80 29.30 4.40 16.50
TSA (1995) Hungary 1.16 1.16 7.00 7.00 7.00 39.70 13.40
JBP (1990) India 0.90 1.30 1.00 2.00 12.20 16.30 30.00 10.00
RPT (1993) India 0.90 1.50 6.00 16.30 26.00 10.00
INDEC (1988) Indonesia 1.60 2.20 5.50 9.50 14.50 5.50 10.80
Sammour (1992) Jordan 1.00 35.00 11.00
World Bank (1995) Jordan 1.00 35.00 45.00 50.00 11.00
NDLI (1991) Myanmar 1.20 2.10 2.90 5.00 8.20 7.90
NDLI (1993) Nepal 1.20 1.60 15.70 10.00
Louis Berger (1990) Nigeria 1.10 1.70 9.30 33.00 11.50
Bennett (1989) NZ 1.80 2.50 2.55 3.50 5.60 33.20 44.00
CESTRIN (1994) Romania 1.36 6.27 10.40 18.60 39.70 12.10
World Bank (1995) Russia 0.85 2.00 5.00 15.00 20.0 35.00 11.00
Estudio (1993) Spain 1.10 1.90 12.70 29.70 11.10
TSPC (1992) Sri Lanka 1.10 2.20 4.80 8.30 14.00 15.00 4.80 8.20
NDLI (1991) Thailand 1.20 1.60 5.70 14.40 2.00 4.50 15.00
NDLI (1994b) Trinidad 1.20 2.50 2.30 5.00 12.50 25.00
RPT (1990) Uganda 1.45 2.00 2.45 10.60 26.00 39.60 2.55 12.50
TRDF (1982) USA 1.10 1.40 1.40 2.00 1.70 7.80 16.30 25.50
Note: Operating weight in tonnes. It is defined as the sum of the tare weight plus load.

Version 1.0
Table B.3
Projected frontal area

S M L
NDLI (1994a) Barbados 1.80 2.20 2.72 3.25 5.20 5.20 3.25 3.25 6.30
IBRD (1990) Burundi 1.80 2.72 1.80 5.20 5.75 5.75 2.72
GITEC (1992) Guatemala 2.08 2.72 5.20 5.75 3.25 6.30
TSA (1995) Hungary 1.80 2.72 3.25 5.20 5.20 5.75 6.30
JBP (1990) India 1.90 2.00 2.72 4.42 5.20 5.20 7.22 6.30
INDEC (1988) Indonesia 1.80 2.72 3.25 5.20 5.20 3.25 6.30
NDLI (1991) Myanmar 2.05 2.75 2.75 4.70 6.60 5.20
NDLI (1993) Nepal 2.05 2.75 8.00 5.37
Bennett (1989) NZ 2.72 2.72 2.72 2.72 3.25 5.75 5.75 2.08 6.30 6.30
CESTRIN (1994) Romania 1.80 3.25 5.20 5.20 5.75 6.30
TSPC (1992) Sri Lanka 1.80 2.70 4.00 5.20 5.20 5.75 4.50 6.00
NDLI (1991) Thailand 2.05 2.75 4.70 6.60 3.25 7.25 8.50
NDLI (1994b) Trinidad 1.80 2.72 2.72 3.25 5.20 5.20
TRDF (1982) USA 1.92 2.05 2.05 2.62 2.87 3.60 5.32 8.90
Note: Projected frontal area in m2.

Version 1.0
Table B.4
Aerodynamic drag coefficient

S M L
NDLI (1994a) Barbados 0.45 0.46 0.45 0.46 0.70 0.85 0.85 0.70
0.65
IBRD (1990) Burundi 0.45 0.46 0.45 0.85 0.63 0.63 0.46
GITEC (1992) Guatemala 0.50 0.46 0.85 0.63 0.70 0.65
TSA (1995) Hungary 0.35 0.46 0.70 0.85 0.85 0.63 0.65
JBP (1990) India 0.45 0.45 0.46 0.52 0.58 0.85 0.85 0.58 0.65
INDEC (1988) Indonesia 0.40 0.46 0.70 0.85 0.85 0.70 0.80
NDLI (1991) Myanmar 0.35 0.40 0.40 0.70 0.85 0.70
NDLI (1993) Nepal 0.35 0.40 0.85 0.65
Bennett (1989) NZ
CESTRIN (1994) Romania 0.45 0.70 0.85 0.85 0.63 0.65
TSPC (1992) Sri Lanka 0.45 0.55 0.65 0.85 0.85 0.85 0.60 0.70
NDLI (1991) Thailand 0.35 0.40 0.70 0.70 0.70 0.70 0.70
NDLI (1994b) Trinidad 0.45 0.46 0.46 0.70 0.85

Version 1.0
Table B.5
Elasticity of vehicle utilisation

S M L
NDLI (1994c) Barbados 0.60 0.80 0.85 0.80 0.85 0.85 0.85 0.85 0.75
SWK (1993a) Botswana 0.10 0.15 0.20 0.35 0.40 0.50 0.35 0.40
IBRD (1990) Burundi 0.00 0.80 0.00 0.85 0.85 0.85 0.80
GITEC (1992) Guatemala 0.60 0.80 0.85 0.85 0.85 0.75
SWK (1993b) Ethiopia 0.10 0.20 0.20 0.20 0.30 0.40 0.20 0.30
JBP (1990) India 0.60 0.60 0.90 0.80 0.90 0.90 0.90 0.90
INDEC (1988) Indonesia 0.50 0.75 0.45 0.45 0.45 0.65 0.65 Sumatra
MRCU (1992) Nepal 0.00 1.00 1.00 1.00
CESTRIN (1994) Romania 0.00 0.31 0.42 0.45 0.50 0.80
TSPC (1992) Sri Lanka 0.50 0.50 0.50 0.50 0.50 0.50 0.60 0.60
Arup (1992) Tanzania
NDLI (1991) Thailand 0.10 0.17 0.21 0.36 0.15 0.17 0.23
NDLI (1994b) Trinidad 0.45 0.80 0.85 0.85 0.85

Version 1.0
Table B.6
Number of hours driven per year

S M L
Chamala (1993) Australia 400 500 500 600 600 700 800 1500 700
Transroute (1992) Bangladesh 500 1100 1250 1750
NDLI (1994a) Barbados 500 3000 1400 1600 1200 1200 1200 3500 3400
SWK (1993a) Botswana 188 279 390 695 833 1040 732 805
IBRD (1990) Burundi 0 900 0 800 800 1800 1800
GITEC (1992) Guatemala 730 1095 1460 1825 1825 1825
SWK (1993b) Ethiopia 250 450 450 450 650 850 450 650
TSA (1995) Hungary 285 285 875 875 875 1500 1750
JBP (1990) India 417 417 500 1250 1429 1429 1700 1900 2286
INDEC (1988) Indonesia 450 1555 1820 1820 1830 3150 3150 Sumatra
Sammour (1992) Jordan 4000 4000 4000
World Bank (1995) Jordan 2000 3000 3000 3000 2000
NDLI (1991) Myanmar 800 800 1440 1680 2400
Louis Berger (1990) Nigeria 500 1500 800 1500 2033
Kampsax - Beca (1990) Papua New Guinea 500 1200 2200 2200 2400 2400
CESTRIN (1994) Romania 400 740 1030 1140 1550 1580
World Bank (1995) Russia 230 825 825 825 1300 2400 2400
TSPC (1992) Sri Lanka 270 500 950 1100 800 700 950 1250
NDLI (1991) Thailand 354 462 615 1308 523 615 1077
NDLI (1994b) Trinidad 380 875 1250 875 875 980
RPT (1990) Uganda 579 824 1000 1200 1200 1450 1333 1765
SWK (1993c) Uganda 2000 2000 2000 2000 2000

Version 1.0
Table B.7
Vehicle service life

S M L
NDLI (1994a) Barbados 6 6 5 6 4 10 12 7 12 12
SWK (1993a) Botswana 8 8 8 7.5 7.5 6 4 8
IBRD (1990) Burundi 8 5 10 10 10 10 3
GITEC (1992) Guatemala 10 10 10 10 10 10
TSA (1995) Hungary 13 13 10 10 10 13 10
SWK (1993b) Ethiopia 12 8 8 15 12 10 12 14
JBP (1990) India 15 15 12.5 8 15 8 8 10
RPT (1993) India 12 12 12 12 12 12
INDEC (1988) Indonesia 8 4 10 13 13 6 6 Sumatra
Sammour (1992) Jordan 12 12 10
NDLI (1993) Nepal 15 10 12 12
CESTRIN (1994) Romania 15 15 15 15 15 15
Estudio (1993) Spain 10 12 12 12 10
TSPC (1992) Sri Lanka 16 16 14 20 20 22 14 20
NDLI (1991) Thailand 10 8 12 12 8 12
NDLI (1994b) Trinidad 10 10 10 10 10 10
RPT (1990) Uganda 6 5 6 5 7 7 7 6
SWK (1993c) Uganda 6 6 5 7 6
Notes: Service life in years.

Version 1.0
Table B.8
Annual utilisation

S M L
NDLI (1994a) Barbados 15 60 15 40 30 30 40 100 85
SWK (1993a) Botswana 18 24 30 50 60 75 60 66
IBRD (1990) Burundi 15 40 20 30 30 45 100
NDLI (1994a) Ethiopia 20 35 35 30 40 50 30 50
GITEC (1992) Guatemala 20 25 60 80 120 120
TSA (1995) Hungary 11.4 11.4 35 35 35 50 70
RPT (1993) India 20 16 40.2 50 80 80
INDEC (1988) Indonesia 23.7 70 80 80 75 125 150 Sumatra
NDLI (1991) Myanmar 20 20 20 36 42 77
NDLI (1993) Nepal 20 37.5 60 43.5
Louis Berger (1990) Nigeria 50 54 50 77.6 122
CESTRIN (1994) Romania 15 42 55 58 66 70
Estudio (1993) Spain 15 30 50 80 60
TSPC (1992) Sri Lanka 15 26 40 45 35 30 45 65
NDLI (1991) Thailand 23 30 40 85 34 70
NDLI (1994b) Trinidad 17.3 35 35 35 35 35
SWK (1993c) Uganda 14 20 31 44 37.5
Notes: Utilisation in ‘000 km/year.

Version 1.0
Table B.9
Value for fuel consumption model adjustment factor - α1

S M L
NDL I (1994) Barbados 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
SWK (1993a) Botswana 0.80 0.80 0.80 1.00 1.00 0.80 0.80 1.00
IBRD (1990) Burundi 1.00 1.00 1.00 1.00 1.00 1.00 1.00
SWK (1993b) Ethiopia 0.80 0.60 0.60 0.80 1.00 0.80 1.00 1.00
GITEC (1992) Guatemala 0.75 0.97 1.00 1.00 1.00 1.00
TSA (1995) Hungary 0.88 1.00 1.00 1.15 1.00 0.60 1.00
JBP (1990) India 0.70 1.00 1.00 0.60 0.70 0.76 0.60 0.60 0.87
INDEC (1988) Indonesia 0.90 0.90 0.90 0.90 0.90 0.90 0.90
CESTRIN (1994) Romania 1.00 1.00 1.00 1.00 1.00 1.00
TSPC (1992) Sri Lanka 0.85 0.85 0.90 0.90 0.90 0.90 0.90 0.90
Arup (1992) Tanzania 0.85 0.95 0.95 0.80 0.95
NDLI (1994b) Trinidad 0.90 1.00 1.00 1.00 1.00 1.00

Version 1.0
Table B.10
HDM-III fuel consumption model adjustment factor - α2

S M L
NDLI (1994a) Barbados 1.16 1.16 1.16 1.15 1.15 1.15 1.15 1.15 1.15
IBRD (1990) Burundi 1.16 1.16 1.16 1.15 1.15 1.15 1.16
GITEC (1992) Guatemala 1.16 1.16 1.15 1.21 1.15 1.15
JBP (1990) India 1.15 1.20 1.20 1.15 1.15 1.15 1.15 1.15 1.15
MRCU (1992) Nepal 1.50 1.50 1.5 1.5 1.5
CESTRIN (1994) Romania 1.16 1.15 1.15 1.15 1.15 1.15
TSPC (1992) Sri Lanka 1.16 1.16 1.15 1.15 1.15 1.15 1.15 1.15
Arup (1992) Tanzania 1.26 1.00 1.00 1.00
NDLI (1991) Thailand 1.15 1.15 1.15 1.15 1.15 1.15
NDLI (1994b) Trinidad 1.16 1.16 1.16 1.15 1.15 1.15

Version 1.0
Table B.11
Used driving power

S M L
NDLI (1994a) Barbados 60 40 85 40 72 160 350 90 120 180
SWK (1993a) Botswana 51 47 51 90 130 220 51 90
IBRD (1990) Burundi 30 40 30 100 210 210 40
SWK (1993b) Ethiopia 30 60 50 80 80 210 80 150
GITEC (1992) Guatemala 70 85 130 230 100 170
TSA (1995) Hungary 60 100 210 100
JBP (1990) India 26 28.5 30 62 121 102 75
RPT (1993) India 43 39 92.5 69 135 65
INDEC (1988) Indonesia 50 50 75 115 120 50 75
NDLI (1991) Myanmar 49 45 45 57 184 199
NDLI (1993) Nepal 48 63 78 78
Yuli (1996) P.R. China 58 67.5 98.8 135.3 86.5
CESTRIN (1994) Romania 33 62 95 151 179 134
Estudio (1993) Spain 43 56 120 210 170
TSPC (1992) Sri Lanka 30 40 54 78 140 160 66 88
Arup (1992) Tanzania 30 60 120 250 180
NDLI (1991) Thailand 48 63 88 133 63 225
NDLI (1994b) Trinidad 54.6 64.2 48.3 77.4 90.0 205
Notes: Used driving power in MPH.

Version 1.0
Table B.12
Values for braking power

S M L
NDLI (1994a) Barbados 17 30 27 30 100 250 250 100 100 160
IBRD (1990) Burundi 17 30 17 250 500 500 40
GITEC (1992) Guatemala 21 30 250 500 100 160
TSA (1995) Hungary 17 250 500 160
JBP (1990) India 13 19 15 84 177 236 336 145
INDEC (1988) Indonesia 25 35 110 250 270 110 180
NDLI (1991) Myanmar 35 42 42 70 195 111
NDLI (1993) Nepal 36 43 171 129
Yuli (1996) P.R. China 32.5 96.6 237.3 355 175.3
CESTRIN (1994) Romania 20 90 150 260 260 170
TSPC (1992) Sri Lanka 22 39 62 180 350 500 82 165
Arup (1992) Tanzania 50 100 160 310 220
NDLI (1991) Thailand 36 43 112 294 36 175
NDLI (1994b) Trinidad 17 30 30 100 250 250
Notes: Braking power in MPH.

Version 1.0
Table B.13
Parts consumption model parameter - C0SP

S M L
GITEC (1992) Guatemala 32.49 32.49 1.49 13.94 1.49 1.77
TSA (1995) Hungary 42.50 0.60 1.50 0.60
JBP (1990) India 15.20 2.20 2.20 0.70 0.70 0.70 5.90 8.50 1.00
INDEC (1988) Indonesia 25.04 25.04 1.08 1.08 4.71 1.08 1.34
Kampsax - Beca (1990) Papua New Guinea 40.61 40.61 1.86 1.86 10.76
CESTRIN (1994) Romania 32.49 1.49 1.49 8.61 13.94 1.77
TSPC (1992) Sri Lanka 25.00 6.30 5.50 4.50 6.00 6.50 1.50 0.90
Arup (1992) Tanzania 17.60 15.00 0.50 2.40 1.20

Version 1.0
Table B.14
HDM-III parts consumption model parameter - CSPQI

S M L
GITEC (1992) Guatemala 13.70 13.70 251.8 15.65 251.8 1.77
TSA (1995) Hungary 13.70 13.70 251.8 251.8 251.8 15.65 3.55
JBP (1990) India 13.70 30.00 30.00 251.8 35.31 15.65 3.56
INDEC (1988) Indonesia 13.70 13.70 251.8 251.8 35.31 251.8 3.56
Zukang et al. (1992) P.R. China 17.80 327.3 327.3 327.3 45.9 20.35 17.81 4.63
CESTRIN (1994) Romania 13.70 251.8 251.8 35.31 15.65 3.56
TSPC (1992) Sri Lanka 10.50 30.00 30.00 30.00 30.00 30.00 20.00 15.00

Version 1.0
Table B.15
Labour hours model parameter - C0LH

S M L
Default 77.14 77.14 242.0 242.0 301.0 242.0 293.0
GITEC (1992) Guatemala 77.14 77.14 242.0 652.5 242.0 293.4
TSA (1995) Hungary 77.14 77.17 242.0 242.0 242.0 652.5 193.4
JBP (1990) India 898.0 746.5 746.5 950.0 258.4 1000 1000 252.2
INDEC (1988) Indonesia 77.14 77.14 242.0 242.0 301.0 242.0 293.0
Kampsax - Beca (1990) Papua New Guinea 68.65 68.65 215.4 215.4 301 242.0
CESTRIN (1994) Romania 77.14 242.0 242.0 301.5 652.5 293.4
NDLI (1991) Thailand 274.0 166.0 276.0 440.0 151.0 786.0
TSPC (1992) Sri Lanka 100.0 150.0 150.0 300.0 350.0 400.0 120.0 200.0

Version 1.0
Table B.16
Labour hours model parameter - CLHPC

S M L
GITEC (1992) Guatemala 0.550 0.550 0.520 0.520 0.520 0.520
TSA (1995) Hungary 0.550 0.550 0.519 0.519 0.519 0.519 0.517
JBP (1990) India 0.547 0.700 0.700 0.520 0.520 0.439 0.397 0.820
CESTRIN (1994) Romania 0.547 0.519 0.519 0.519 0.519 0.517
TSPC (1992) Sri Lanka 0.500 0.519 0.519 0.519 0.519 0.519 0.470 0.470

Version 1.0
Table B.17
Tyre consumption model parameter - C0TC

S M L
Transroute (1992) Bangladesh 0.090 0.090
NDLI (1994a) Barbados 0.160 0.160 0.160 0.160 0.160 0.160
TSA (1995) Hungary 0.164 0.050 0.164 0.164 0.050
GITEC (1992) Guatemala 0.160 0.160 0.160 0.160
JBP (1990) India 0.160 0.160 0.190 0.080 0.160
INDEC (1988) Indonesia 0.164 0.164 0.164 0.164 0.164
World Bank (1995) Jordan 0.160 0.160 0.160 0.160
Zukang et al. (1992) P.R. Chna 0.068 0.068 0.134 0.134 0.068
CESTRIN (1994) Romania 0.164 0.164 0.164 0.164 0.164
TSPC (1992) Sri Lanka 0.164 0.164 0.164 0.164 0.164 0.164 0.164
NDLI (1991) Thailand .1008 .0849 0.072
NDLI (1994b) Trinidad 0.160 0.160 0.160

Version 1.0
Table B.18
Tyre consumption model parameter - CTCTE (x 10-3)

S M L
NDLI (1994a) Barbados 12.78 12.78 12.78 12.78 12.78
GITEC (1992) Guatemala 12.78 12.78 12.78 12.78
TSA (1995) Hungary 12.78 7.00 12.78 12.78 7.00
JBP (1990) India 12.78 12.78 9.00 9.00 12.78
INDEC (1988) Indonesia 12.78 12.78 12.78 12.78 12.78
CESTRIN (1994) Romania 12.78 12.78 12.78 12.78 12.78
TSPC (1992) Sri Lanka 12.78 12.78 12.78 12.78 12.78 12.78 12.78
NDLI (1991) Thailand 7.85 6.61 5.61
NDLI (1994b) Trinidad 12.78 12.78 12.78

Version 1.0
Table B.19
Volume of wearable rubber

S M L
NDLI (1994a) Barbados 4.30 7.60 7.30 4.30 6.85
GITEC (1992) Guatemala 7.60 8.39 4.30 6.85
TSA (1995) Hungary 4.30 15.00 7.60 15.0 15.0
JBP (1990) India 7.60 7.60 8.99 8.99 6.85
INDEC (1988) Indonesia 4.30 7.60 7.30 4.30 6.85
CESTRIN (1994) Romania 4.30 7.60 7.30 8.39 6.85
TSPC (1992) Sri Lanka 3.00 4.00 5.34 7.30 7.30 4.00 6.34
NDLI (1991) Thailand 12.16 14.30 13.80
NDLI (1994b) Trinidad 4.30 7.60 7.30
Notes: Volume of wearable rubber in dm3/tyre.

Version 1.0
Table B.20 HDM-III RDWE deterioration factors used in different countries
Country Pavement Pavement Y/N Deterioration factor SN CBR Traffic Rainfall

Rough- Mainten- Analysis Reference
type composition Kci Kcp Kvi Kpp Krp Kge Kgp ness ance period
AADT/ m/month policy
(thickness in Million m/km IRI Year
mm) eqv std.
axle
Bangladesh Existing Paved Road with N 0.50 2.00 1.00 10.00 1.00 1.30 1.10 NA NA NA NA NA NA NA Transroute (1992)
Pavement Herring Bone Brick
Subbase
Botswana Existing ST + Base, Crushed N 0.67 1.00 0.67 1.00 0.82-1.72 0.51 1.00 1.3-1.6 10-80 117-2438 NA 1786-3266 NA NA SWK (1993a)
Pavement Rock/ Stab. Grav/ mm/km BI
Natural Gravel
Brazil Existing Road Network Data Y 1.00 1.00 1.00 1.00 1.00 1.00 1.00 NA NA NA NA NA NA NA Liautaud and
Pavement Archondo-Callao
(1994)
Chile New Pavement ST30 + Gran Base240 N 0.50 NA 1.00 1.95 0.70 NA NA 13 Low Med NA NA NA NA Gaete et al. (1991)
New Pavement ST20 + Gran N NA NA NA NA 1.16 1.74 NA NA 25 Low Heavy NA NA NA
Base 290
Overlay On Dense Surf 35-45 + N 0.67 0.67 NA NA 0.80-1.06 0.43 3.00 NA 3-10 Medium Dry NA NA NA
Asphalt Open Asph Base40-
Pavement 60+ Old Surf 60+
Gran155-170
Overlay On Dense Asph Surf 55- N 0.58- NA NA NA 0.74-0.99 0.70 2.20 NA 4-13 Medium Med NA NA NA
Asphalt 70+ Binder 70-100 + 0.67
Pavement Dense Asph
Base120-130 + Old
Surf130 + Gran Base
130
Overlay On Dense Asph Surf60- N 0.33- 0.76 NA NA 0.51-1.09 NA NA NA 8 Heavy Med NA NA NA

Asphalt 100 + Dense/ Open + 0.75
Pavement AsphBase115+
Conc185+Gran
Base100-200
…Continued

Version 1.0

mm) eqv std.
axle
Overlay On Exist Open Graded Cold N 1.20 NA 1.20 NA 0.81 0.70 NA NA 6 Medium Med NA NA NA Gaete et al. (1991);
Pavement Mix Surf55 + Old Solminihac et al.
Surf90+ Gran Base (1989).
145
New Pavement Dense Asphalt Surf40 N 0.83 NA NA 0.51 NA 1.74 NA NA NA Medium Heavy NA NA NA
+ Dense Asph Base95
+ Gran. Base 320
Guinea Bissau Existing Road Network Data Y 1.00 1.00 1.00 1.00 1.00 1.00 1.00 All All NA NA NA NA NA Fonseca et al.
Pavement (1991)
India Reseal & Asph. Surfacing N 1.00 1.25 0.70 1.25 1.25 1.00 1.00 NA NA NA Low, Medium, NA Patch + Routine NA CES (1989)
Overlaid +Granular Base High
1.00 1.50 0.60 1.50 1.50 1.50 1.15
1.00 2.00 0.50 2.00 1.75 2.00 1.30
Reseal & Asph. Surfacing N 1.00 1.50 0.60 1.50 1.50 1.15 1.15 NA NA NA NA NA NA NA JBP (1990)
Overlaid +Granular Base
Reseal & Asph. Surfacing N 1.50 1.50 1.00 1.50 1.50 1.00 1.50 2 5 1-8 msa/year Medium 2500 mm/km Patch + Routine 15 RITES (1994)
Overlaid +Granular Base BI
Reseal & Asph. Surfacing Y 1.00 1.00 1.00 1.00 1.00 0.70 1.00 2.0 , NA 2 msa/year Medium 2500 mm/km Patch + Routine 20 NDLI (1994a)
Overlaid +Granular Base 3.0 BI
Jordan Existing AC Over Granular N 2.50 1.00 1.50 3.00 1.00 1.00 0.50 NA 13 NA 0.025 m/month NA NA 12 Sammour (1992)
Base
New AC Over Granular N 0.50 1.00 2.00 1.00 1.00 2.00 1.00 NA 15 2.5msa/year NA NA NA
Base
Malaysia Reseal & Existing & New N 0.70- 1.20-1.40 1.00 1.00 1.00 1.00-1.30 1.00 All All NA NA NA Routine+ NA JKR (1991)
Overlaid Pavements 0.80
Patch+ Reseal &
Overlay
Morocco Existing Road Network Data Y 1.00 1.00 1.00 1.00 1.00 1.00 1.00 NA NA NA NA NA NA NA CID (1993)
Pavement
…Continued

Version 1.0

mm) eqv std.
axle
Mozambique Existing ST20 + Cement N 1.00 1.50 1.00 1.50 1.50 1.50 1.50 2.0-3.0 5 Low Medium 4000 mm/km Routine/NIL 20 RITES (1992)
Pavement BI
Stab. Base 150 +
Select Soil 150
Nepal Existing ST and AC over Y 1.00 1.00 1.00 1.00 1.00 1.00 1.00 SNC= NA Low & Medium/ Heavy 10m/km Routine + Patch NA NDLI (1993)
Pavement Granular Base 3 Medium
IRI
New Zealand Existing Overlays & Reseals NA NA NA NA NA 0.76 1.00 2-4 NA 500-6700, NA NA NA 7 years Cenek and Patrick
Pavement (1991)
0.01-0.09
msa/lane/ year
Nigeria Existing Road Network Data Y 1.00 1.00 1.00 1.00 1.00 1.00 1.00 NA NA NA NA NA NA NA Louis Berger
Pavement (1990)
Pakistan Existing Road Network Data Y 1.00 1.00 1.00 1.00 1.00 1.00 1.00 NA NA NA NA NA NA NA Riley et al. (1987)
Pavement
Peru Existing Road Network Data Y 1.00 1.00 1.00 1.00 1.00 1.00 1.00 NA NA NA NA NA NA NA Liautaud and
Pavement Archondo-Callao
(1994)
South Africa New Pavement Surface Treatment N 1.00- 0.10-0.30 NA NA 1.50-1.75 0.392(1) 0.8-1.2(4) NA NA NA NA NA Reseal/ Overlay 20 Kannemeyer and
1.50 (avg. @8y Visser (1994)
021) (avg. 1.57) 0.607(2) 0.6-1.4(5) (1975-1995)
(avg.
1.21) 0.886(3) 0.8-1.2(6)
Existing Overlays & Reseals N 0.40- 0.30- NA NA 1.00 0.392(1) 0.8-1.2(4) NA NA NA NA NA Reseal/ Overlay 20
Pavement 0.80 @8y
0.70 0.607(2) 0.6-1.4(5) (1975-1995)
(avg.
0.63) (avg. 0.886(3) 0.8-1.2(6)
0.21)
Spain Existing AC and ST over Y 1.00 1.00 1.00 1.00 1.00 1.00 1.00 3.1-4.7 NA NA 0.045 m/ 2253-3352 NA 15 years Estudio (1993)
Pavement Granular Base mm/km BI
month
…Continued

Version 1.0

mm) eqv std.
axle
Tanzania Existing SSD/DSD + Lime / N 0.50- 0.50-1.65 1.00 1.00 1.50-2.10 0.20-0.30 0.60-1.30 SNC= NA ADT (1991)= 0.09 m/ 2253-3352 Reseals as per Year Arup (1993)
Pavement Cement Stab. Base 1.30 mm/km BI Actually Applied
2.3-3.0 85-266 month off Constr.-
1992
Tunisia Existing Road Network Data Y 1.00 1.00 1.00 1.00 1.00 1.00 1.00 NA NA NA NA NA NA NA Houcine (1982)
Pavement
UK Existing AC60+Old Surf 50+ 1.00 1.00 NA NA NA NA NA 80 5535-7057 0.03 m/month 1500 mm/km Patch + Routine 10 Wyley et al. (1986)
Pavement Cement Stab. Base BI
120 (1974-1984)
Reseal on AC SD10+Old Surf 90 + NA NA NA NA 3.00 NA NA 2.3 80 5535-7057 0.03 m/month 1500 mm/km Cold Mix 20 @ 10
Gran Base150 BI 5Y + Patch +
Routine (1974-1984)
Surface ST30 +Old Surf. NA NA NA NA 2.70 NA NA 1.5 8 2892-3659 0.03 m/month 1500 mm/km AC30@ 15/ 7Y 10
Treatment 20+Granular Base 205 BI +Patch +Routine
(1974-1984)

Version 1.0
Road user data
Appendix C Road user data

The road user data were discussed in Chapter 6.

Version 1.0
Road and pavement data
Appendix D Road and pavement data
D.1 Introduction
The road and pavement data required for HDM consists of general road data, pavement
characteristics, pavement condition and the maintenance alternatives. Each of these data are
discussed in the following sections after a description of how one classifies pavements.
D.2 Pavement classifications

The HDM pavement deterioration model contains different pavement deterioration models
for different types of pavements. It is therefore necessary to establish a pavement
classification system at the onset of the HDM-III analysis.
HDM-III considered the following pavement types:
! Surface treatment
! Asphalt concrete
! Slurry seal on surface treatment
! Reseal on surface treatment
! Reseal on asphalt concrete
! Open graded cold mix surfacing
! Asphalt overlay or slurry seal on asphalt concrete
For HDM-4 the pavement classification system is much more comprehensive to reflect the
wider range of pavement types modelled.
Table D.1 lists the HDM-4 bituminous pavements classification system and Table D.2 defines the
pavement codes (ISOHDM, 1997). In addition to these there are concrete, block and unsealed
pavements in HDM-4.

Version 1.0
Table D.1
HDM-4 bituminous pavement classification system
Surface type Surface Base type Base material Pavement type

material
AC CRS
GB AMGB
HRA GM
PMA AB AB AMAB
AM RAC CS
SB AMSB
CM LS
PA TNA
SMA AP FDA AMAP
xx
CAPE CS
GB STGB
DBSD GM
SBSD AB AB STAB
ST SL CS
SB STSB
PM LS
TNA
AP STAP
FDA
Source: ISOHDM (1997)
Note:
1 AM and ST surfacings on concrete pavements, that is, AMCP & STCP, are modelled in
HDM-4 as concrete pavement types in the rigid pavements sub-model.

Version 1.0
Table D.2
Definitions of HDM-4 pavement codes
AM Asphalt Mix GB Granular Base
AC Asphaltic Concrete AB Asphalt Base
HRA Hot Rolled Asphalt AP Asphalt Pavement
PMA Polymer Modified Asphalt SB Stabilised Base
RAC Rubberised Asphalt Concrete CRS Crushed Stone
CM Soft Bitumen Mix (Cold Mix) GM Natural Gravel
PA Porous Asphalt CS Cement Stabilisation
SMA Stone Mastic LS Lime Stabilisation
ST Surface Treatment TNA Thin Asphalt Surfacing
CAPE Cape Seal FDA Full Depth Asphalt
DBSD Double Bituminous Surface Dressing
SBSD Single Bituminous Surface Dressing
SL Slurry Seal
PM Penetration Macadam
xx User Defined
The available HDM-4 models are based on different factors. Many are based on surface and
base type, while some are based on surface material. Accordingly, the modelling is initially
done in terms of surface material and base type, even though base materials can be specified.
HDM-4 has default coefficients for the bituminous pavement types given in Table D.3.
Table D.3
Pavement Surface type Base type Description

type
AMGB AM GB Asphalt Mix on Granular Base
AMAB AM AB Asphalt Mix on Asphalt (Dense Bitumen Macadam) Base
AMSB AM SB Asphalt Mix on Stabilised Base
AMAP AM AP Asphalt Mix on Asphalt Pavement
STGB ST GB Surface Treatment on Granular Base
STAB ST AB Surface Treatment on Asphalt (Dense Bitumen Macadam)

Base
STSB ST SB Surface Treatment on Stabilised Base
STAP ST AP Surface Treatment on Asphalt Pavement
While it is straightforward to classify new pavements as surface treatment or asphalt

concrete, once the pavement has received a maintenance treatment, the classification becomes
very complicated.

Version 1.0
Without adequate maintenance records the only way of classifying a pavement into one of the
post-maintenance categories is by conducting field surveys and gathering data on a sample of
pavements.
D.3 General road data

HDM-III requires the user to specify the following pavement geometric characteristics:
! Rise and fall (m/km)
! Horizontal curvature (degrees/km)
! Carriageway width (m)
! Shoulder width (m)
! Superelevation (m/m)
These data are usually straightforward to obtain. The first two data are available from
topographic maps, highway information sheets, or route surveys. For the carriageway and
shoulder widths a sample of data should be used to estimate average or median values. These
are often a function of traffic volume so it may be useful to stratify the data as a function of
volume. The superelevation can either be supplied, or it can be calculated in HDM-III as a
function of curvature using the following equation developed from the Brazil data
(Watanatada et al., 1987a):
e = a0 C …(D.1)
where:
e is the superelevation (m/m)

C is the curvature (degrees/km)
a0 is a constant (0.012 for paved roads; 0.017 for unpaved)
Another value termed the effective number of lanes is used in HDM-III. This is used to
reflect the practice of traffic on narrow roads to travel towards the middle of the pavement,
thereby increasing pavement deterioration. Watanatada et al. (1987a) give the following
recommended values:
Width (m) Number of effective lanes

< 4.5 1.0
4.5 – 6.0 1.5
6.0 – 8.0 2.0
8.0 – 11.0 3.0
> 11.0 4.0
It should be noted that in places where there is poor lane discipline where vehicles do not
tend to drive in set wheelpaths, such as India, it may be appropriate to increase the number of

Version 1.0
effective lanes above these recommended values due to the reduced levels of effective traffic
loading.
D.4 Pavement characteristics
D.4.1 Strength: modified structural number and deflection

HDM-III requires the pavement strength to be calculated in terms of the modified structural
number (SNC) and Benkelman beam deflection. In HDM-4 the deflection from a falling
weight deflectometer (FWD) can be supplied.
Ideally, data on both these characteristics should be supplied to HDM, however the SNC is
often difficult to accurately quantify.
The SNC is the structural number of the pavement increased to reflect the contribution of the
subgrade. When the pavement layer history is available, it is possible to calculate the SNC
from the thicknesses of the individual layers using appropriate strength coefficients from
Watanatada et al. (1987a). It is also necessary to supply an estimate of the in situ subgrade
CBR for calculating the subgrade contribution. However, it should be noted that these
coefficients are based on laboratory tests of materials whereas the SNC should reflect their
actual performance in the field.
In instances where the pavement layer history is unavailable, the SNC can be calculated from
the Benkelman beam deflection using the following equations (Watanatada et al., 1987a):
SNC = 3.2 DEF -0.63 If base not cemented …(D.2)
SNC = 2.2 DEF -0.63 If base cemented …(D.3)
Watanatada et al. (1987a) provide similar equations for calculating the deflection from the
SNC. Paterson (1987) who used an orthogonal regression to ensure that the equations give
consistent predictions when going to/from either attribute developed these equations.
Appendix J describes orthogonal regression.
NDLI (1995b) describe several techniques for estimating SNC from FWD deflection and
these have been adopted for HDM-4.
If the existing pavements are strong, SNC of 5.0 or greater, there is not a great need to
accurately specify this parameter. This is because there are limited differences in the
predictions of the HDM-III pavement deterioration model with strong pavements. Thus, it is
only when there are weak pavements that the SNC is a critical parameter.
D.4.2 Pavement history: age of surfacing and layers

For each pavement type it is necessary to supply the pavement history in terms of the age of
the last surfacing, the number of layers and their thicknesses.
Many organisations have maintenance histories so these data can often be used to provide the
data. Where these are unavailable, it is often necessary to undertake field surveys.
Pavement age can be estimated from discussions with maintenance staff. Another good
source is local residents adjacent to the roads, particularly those in the hospitality industry.
They often recall when a road was improved due to the increased level of business associated
with the works.

Version 1.0
The pavement layer information is more difficult. When there are set standards it is often
possible to assume a standard surface thickness based on an estimated average age and the
number of years between typical maintenance treatments.
D.4.3 Drainage and environmental factors

The HDM-III pavement deterioration model is predicated on the assumption that there is good
drainage of the pavements. Since in practice this is not always the case, it is necessary to
adjust the model to reflect poor drainage. Similarly, the model must be adjusted to reflect
different environmental conditions. The techniques for doing this are discussed in Section
4.1.6 of Watanatada et al. (1987a).
D.5 Pavement condition
D.5.1 Roughness
The pavement roughness is a very important parameter. It is not only a measure of the
condition of the pavement but it also has a major impact on the RUE.
When measuring sections of road, roughness is usually measured using a response-type road
roughness measuring system (RTRRMS) or a non-contact profilometer (laser/accelerometer).
For short sections or studies manual methods such as a rod-and-level survey, dipstick or
walking profilometer may be used.
RTRRMS consist of an instrument mounted on the floor of a vehicle that is connected to the
rear axle. The instrument records the displacement of the vehicle body relative to the axle and
usually expresses it in terms of a vertical displacement per unit distance travelled, such as
mm/km. The TRRL Bump Integrator is probably the most widely used RTRRMS.
Profilometers usually employ sensors to measure the elevation of the vehicle to the pavement.
Accelerometers are double integrated to get the movement of the vehicle through space. The
difference between these values gives the road profile elevation. These data are processed to
obtain the roughness.
Since the response of a vehicle to road roughness will depend upon the characteristics of the
suspension system, tyres and other factors, it is necessary to calibrate the vehicle against a
standard reference roughness. The International Roughness Index (IRI) was developed by the
World Bank (Sayers et al., 1986) and is based on a complex simulation of a vehicle
suspension. HDM-III used QI but since then IRI has become the international standard and all
roughnesses should be expressed in terms of IRI m/km. The conversion between QI and IRI is
1 QI = 13 IRI m/km.
In order to calculate the IRI it is necessary to obtain an accurate profile of the pavement
surface on a number of test sections. From this profile, one calculates the IRI. The vehicle is
then operated over these sections and a regression equation is developed between the
roughness meter readings and the reference roughness. This constitutes roughness meter
calibration.
Once the vehicle has been calibrated, it is necessary to measure the roughness of the
pavements. There are several points that should be considered in undertaking these
measurements:
! Sampling interval
It is best to record the roughness at the shortest practical interval - if possible, every 100
m. When longer lengths such as 1 km are used, short sections of poor pavements will be
disguised by other sections in better condition.

Version 1.0
! Measurement speed
The measurements should be made at a single standard speed. Where data loggers are
available and roughnesses are low, this can be upwards of 80 km/hr. If different speeds
are to be used, it is necessary to have calibration equations for each speed.
! Vehicle load
The vehicle should be calibrated and operated at standard loads. In addition, the tyre
pressures should also be standard.
D.5.2 Surface distress

HDM calls for the pavement condition to be specified in terms of the areas of all cracks, wide
(> 3 mm) cracks, ravelled, and potholed. It also requires the mean and standard deviation of
rut depth. It is important that the values input for these characteristics are based at least
somewhat on field surveys.
As with roughness, it is recommended that these parameters are measured at short intervals of
200 m. If longer intervals such as 1 km are used, it will often be found that there is at least
some cracking within the interval, even if it confined to a short section. In the HDM-III
analysis the entire section will be treated as cracked which will lead to markedly different
results than if the analysis were conducted, for example, with 80 per cent not cracked and
only 20 per cent cracked.
D.5.3 Rutting
D.5.4 Potholes
D.6 Maintenance alternatives

Table D.4 illustrates the range of maintenance alternatives that were adopted in a study in
India (NDLI, 1997). There are several points to note with this table.
! A range of treatments from overlays to reconstruction were adopted. This allowed for the
analysis to establish the mix of treatments given the available budget.
! In keeping with the pavement design standards, the treatments are a function of
increasing traffic loading.
! For new construction, there is a constant SNC for a given level of traffic. This reflects
pavement design standards wherein there is a given strength required for each level of
traffic loading. If the strengths were not constant, there would be different performances
between the pavements being tested which could distort the results.

Version 1.0
Table D.4
Example of maintenance treatments
Cumul Pavement Existing pavement New construction

ative structure Recon- Rehab- Medium Thin CBR CBR CBR CBR CBR
loading struction ilitation overlay overlay < 3 3-6 7-11 12-19 > 20
(ESA x
6
10 )
Asphalt (BC) 70 70 40 45 30 30 30 30 30
Asphalt (DBM) 0 0 0 0 40 40 40 40 40
Asphalt Levelling 0 0 30 20 0 0 0 0 0
Course
Base 200 200 0 0 200 200 200 200 200
<25 Sub-Base 200 0 0 0 250 300 240 150 100
Granular Levelling 75 50 0 0 0 0 0 0 0
Course
Select Subgrade 0 0 0 0 375 100 0 0 0
Total 545 320 70 65 895 670 510 420 370
SN (Existing) or 3.0 1.9 0.9 0.4 4.5 4.5 4.5 4.5 4.5
SNC (New)
Asphalt (BC) 30 30 40 40 30 30 30 30 30
Asphalt (DBM) 40 40 70 20 70 70 70 70 70
Course
Base 200 200 0 0 225 225 225 225 225
25 to 80 Sub-Base 200 0 0 0 300 300 300 350 275
Course
Select Subgrade 0 0 0 0 550 350 175 0 0
Total 545 320 140 80 1175 975 800 675 600
SN (Existing) or 3.2 2.1 1.1 0.6 6.0 6.0 6.0 6.1 5.9
SNC (New)
Asphalt (BC) 40 40 40 40 30 30 30 30 30
Asphalt (DBM) 60 60 100 50 70 70 70 70 70
Course
Base 200 200 0 0 250 250 250 250 250
> 80 Sub-Base 200 0 0 0 300 300 300 300 325
Course
Select Subgrade 0 0 0 0 625 400 225 100 0
Total 575 350 170 110 1275 1050 875 750 675
SN (Existing) or 3.5 2.4 1.4 0.9 6.9 6.4 6.4 6.4 6.4
SNC (New)
Source: NDLI (1997)
D.7 Homogeneous sections

All economic analyses use homogeneous sections. These are sections of the network where
the road characteristics are consistent. There are two approaches to defining sections:
! Fixed
! Dynamic

Version 1.0
Fixed sections are defined at a regular interval, for example between km stones. Dynamic
sections are defined on the basis of the data attributes.
When characterising the condition of a road, each measurement has its own appropriate
sampling interval:
! Roughness - may be measured at intervals of, say, 100 m,
! Deflection - at intervals of 50 m,
! Condition to a longer section - such as 1000 m.
When using fixed links it is necessary to aggregate the data over the arbitrary link length,
even if the interval is not the best for the data items being measured.
Dynamic links avoid this by creating links based on the condition (or any other parameter
such as traffic) of the pavement. This leads to a much more realistic representation of the
network since the treatments one applies will be based on the pavement condition. The
process by which these links are created is termed automatic or dynamic sectioning.
Dynamic sectioning is achieved by defining the allowable variation in an attribute, for
example roughness + 1.0 IRI m/km; width + 0.5 m. The condition data are then analysed and
when the condition deviates from these variations a new section is created. This is illustrated
in Figure D.1 for a hypothetical road section.
Roughness Measurements
Roughness Sections
Rut Measurements
Rut Sections
Analysis Sections
If Fixed Analysis Sections
Figure D.1 Fixed versus dynamic sections
It is not unusual for fixed section systems to lead to inappropriate treatments due to the
averaging of characteristics over the section disguising short, poor sections. This is illustrated
in Figure D.1 where the fixed sections would result in fundamentally different treatments
being applied to the dynamic analysis sections.
One can establish homogeneous sections based on a range of attributes. Attributes commonly
used to establish homogeneous sections include:

Version 1.0
! Roughness
! Strength
! Width
! Traffic volume
! Environment
The sectioning process can be visualised as driving along a road. Whenever an attribute
changed by a certain amount a new section would be created. For example, a:
! Roughness - change of more than x IRI m/km from the mean roughness of the section to
that point
! Strength - change of more than y from the mean strength of the section to that point
! Width - change of more than z m from the mean width of the section to that point
! Change in the AADT on the section

! Change from urban to rural, or vice versa
! Change from normal to black cotton soil, or vice versa
Having established the homogeneous sections it is necessary to prepare a single data file for
use in the HDM modelling. This is done using dynamic segmentation and is illustrated in
Figure D.2. Depending upon the data attribute, one will take the average, weighted average,
maximum, minimum, or dominant feature.
500 m 300 m
Section 1 Section 2
AADT 500 1000

Skid Number 5 7
Deflection (mm) 0.02 0.04
Pavement Type Flexible Rigid
Maintenance Cost $750 $1,000
Dynamic Segmentation
800 m
Analysis Section
AADT 688 Weighted Average

Skid Number 5 Minimum
Deflection (mm) 0.04 Maximum
Pavement Type Flexible Dominant
Maintenance Cost $1,750 Sum
Figure D.2 Example of dynamic segmentation

Version 1.0
Traffic data
Appendix E Traffic data
E.1 Introduction
Traffic volume is a critical input data item for all HDM analyses. Since the total transport
costs are dominated by road user costs which are proportional to the traffic volume. In
addition, the commercial vehicle volumes influence the rate of pavement deterioration. This
appendix describes traffic volumes, growth rates and other associated issues.
E.2 Traffic volume and composition

The traffic flow is the number of vehicles passing a point in time:
n
q= …(E.1)
t
where:
q is the traffic flow in vehicles during time t

n is the number of vehicles
t is the time interval t
Annual traffic figures are expressed in veh/day. There are two important measures, depending
on how the data are sampled:
! Average Annual Daily Traffic (AADT)
! Passenger Car Units (PCU)
Average Annual Daily Traffic (AADT) is the average number of vehicles on a road over the
year:
Annual Traffic
AADT = …(E.2)
365
This can only be calculated at locations where there is continuous traffic counting.
Since it is more common to conduct short-term traffic counts, most data are termed Average
Daily Traffic (ADT). This is dependent upon the length of the traffic count and it is important
to specify what it refers to; for example:
! 5 - day ADT - is used for urban streets
! 7 - day ADT - is used for rural highways
The base traffic data supplied to HDM must be an AADT. ADT are converted to AADT
using correction factors. These are often based on the time of year when the traffic is counted.
If, for example, the count is made in winter or during the monsoon season the factor usually
increases the ADT to convert it to an AADT. Conversely, summer or dry season counts are
often reduced to make them an AADT. Table E.1 shows seasonal correction factors from

Version 1.0
Traffic data
India (NDLI, 1997). The data in this table indicate, for example, that an ADT from a count
made at Location 1 in August would be multiplied by 1.13 to convert it to an AADT.
Hourly traffic flows are expressed in veh/h. This is the most common measure to use in traffic
evaluations. HDM uses hourly flows in its congestion analyses.
The traffic stream is comprised of a number of different vehicle types. It is therefore common
to convert these to more homogeneous measures.
Table E.1
Example of seasonal correction factors from India
Month Seasonal correction factor by month

Location 1 Location 2 Location 3
January 0.94 0.95 0.91
February 1.00 0.97 0.98
March 0.97 0.90 0.87
April 0.99 1.00 0.97
May 0.96 0.97 0.94
June 1.02 1.17 1.08
July 1.11 1.14 1.14
August 1.13 1.17 1.09
September 1.09 1.14 1.19
October 0.99 1.01 0.99
November 0.98 1.03 1.06
December 0.86 0.84 0.90
Source: NDLI (1997)
Passenger Car Units (PCU) is a measure that converts all vehicles to equivalent passenger
cars. These are most commonly applied on rural road appraisals where the larger vehicles
have a significant negative impact on traffic flow. The values for PCU conversions vary
between countries and depend upon factors such as the vehicle size and power to weight ratio.
Recently, the measure Bicycle Space Equivalent (BSE) has been proposed as a measure for
use in countries such as China and India where there are high levels of non-motorised traffic
(Yuli, 1996).
In HDM, traffic is expressed in terms of Passenger Car Space Equivalents (PCSE). The PCSE
differs from the PCU in that it is based on the area occupied by the vehicle1. A discussion of
PCSE and how they are calculated was given in Section 6.4.3.
It is important to consider the traffic composition as well as the total volume since different
vehicle classes have different operating costs and impacts on pavements. The traffic stream is
modelled in HDM using representative vehicles and so the default HDM representative
vehicles should be used as a guide when considering the traffic survey classes. Issues such as
1
PCSE are primarily based on length. However, as described in Section 6.4.3, Hoban et al. (1994)
recommend increased values of PCSE for narrower roads to reflect the greater effects of the vehicles on the
total road space under these conditions.

Version 1.0
Traffic data
the likely operating cost differences between vehicle classes, occupancies, loading levels, etc.
all should be assessed. As a minimum, the following are the recommended vehicle classes for
HDM analyses:
! Non-motorised traffic
! Motorcycles
! Passenger cars
! Light trucks
! Medium/heavy trucks
! Mini-buses
! Heavy buses
E.3 Hourly distribution of traffic volume

Over the space of a year most roads undergo different hourly flows. Some roads that are
congested during peak hours have greatly reduced flows at night-time. Other roads, such as
interurban routes in India, experience major flows at night when trucks prefer to travel.
These variations are considered in HDM through the provision of an hourly distribution of
traffic volume. This consists of the number of hours per year that the flow is at different
levels. As shown in Figure E.1, the distributions are markedly different between countries and
so should be established based on local data. They will also vary between road classes.
20
18
16 USA - Seasonal
14
Hourly Flow as % AADT
12
10
USA - Commuter
6 India
USA - Intercity
4
New Zealand
2
0
0 1000 2000 3000 4000 5000 6000 7000 8000 9000
Cumulative Number of Hours per Year
Figure E.1 Comparison of traffic flow distributions
For HDM it is necessary to establish the number of hours per year the traffic is at different
flow levels. It is common to adopt at least five levels. These should consist of a low and high
flow level, as well as an appropriate number of intermediate levels. When there are non-
motorised traffic (NMT), it is necessary to account for these in the flow bands. Analysing the
NMT traffic associated with the same flow bands as the motorised traffic does this.
Table E.2 is an example of how the hourly distribution is established (NDLI, 1997). Short-
term traffic counts were expanded to cover a full year (8760 h/year) and broken down into

Version 1.0
Traffic data
five flow bands. After reviewing the distribution five flow-bands were selected and the
number of hours in each band was established along with the mean flow.
Table E.2
Example of hourly distributions from India
Motorised traffic Non-

motorised
traffic
Flow band Hourly flow as Length of year at that Mean
percentage of AADT hourly flow hourly flow
as
Bin Mean Hours Days Percentage
range of NMT
AADT
1 0 - 2.5 1.59 2575 107.3 0.79
2 2.5 - 6.5 4.53 4805 200.2 4.20
3 6.5 - 8.5 7.14 1114 46.4 8.88
4 8.5 - 10.5 9.41 216 9.0 12.17
5 > 10.5 12.92 50 2.1 35.04
Source: NDLI (1997)
When establishing these distributions a check should be made that:

f
365 AADT =
∑ HRYR HRVOL
i =1
i I …(E.3)
where:
HRYRi is the number of hours per year for flow band I

HRVOLi is the hourly volume in veh/h
f is the number of flow bands
The values for HRVOLi should be adjusted to ensure this check is made.
E.4 Traffic growth rates

Establishing the traffic growth rate for an HDM analysis is a very important task. This is
because it will not only influence the total RUE, but also will influence capacity
improvements and other intervention alternatives.
Traffic grows over time, although the growth can be quite sporadic. Figure E.2 shows the
results from traffic counts taken on a road in India over a 10-year period. The data are
presented from April and October, which is before and after the monsoon period. It will be
noted that the post-monsoon flows are always lower than the pre-monsoon flows. This is due
to the major reduction in agricultural activities after the monsoon. While there is a continual
growth, the October data from 1992 and 1994 is out of context with the rest of the historical
data. This could be due to some unusual event at the site, for example a temporary road

Version 1.0
Traffic data
closure or a local catastrophe. If such drops are not recorded elsewhere, they should probably
be ignored.
30000
T raffic Count Data from Surat-Dhulia Road, Gujarat, India
25000
20000
7 Day ADT in PCU .
15000
10000
5000
0
April '88
April '89
April '90
April '91
April '92
April '93
April '94
April '95
April '96
May '87
Oct '87
Oct '88
Oct '89
Oct '90
Oct '91
Oct '92
Oct '93
Oct '94
Oct '95
Figure E.2 Example of variation in traffic over time
The two usual ways of expressing growth are as a geometric or as an arithmetic growth rate.
These are calculated as:
! Geometric
(YEAR −1)
AADTyear = AADTbase 1+
GROWTH 
 …(E.4)
 100 
! Arithmetic
 GROWTH
AADTyear = AADTbase 1+ (YEAR − 1) …(E.5)
 100 
where:
AADTyear is the traffic volume in the analysis year

AADTbase is the traffic volume in year 1
YEAR is the year of analysis
GROWTH is the traffic growth rate (%)
HDM uses a geometric growth rate.

The traffic growth rate is usually calculated using historical traffic trends, such as the data
illustrated in Figure E.2, economic trends, such as growth in GDP, vehicle ownership
trends, or a combination.

Version 1.0
Traffic data
Using historical traffic trends is the least accurate method for predicting growth rates. This is
because many developing countries are experiencing rapid motorisation and economic growth
which will alter the historical trends.
Calculating growth from economic trends is often a more sound method than historical data
since there are strong correlations between traffic volume and economic activity, something
that a straight historical model will not capture. The underlying approach is to collect data on
traffic growth rates and then to do a statistical analysis that fits a model to the data. The
variables often used are income, population, fuel use, industrial production, agricultural
production, or any other measure of the economy. These lead to elasticities which express the
change in traffic as a function of the economic measures. Typical equations would be:
GROWTH = dGNP EGNP + dPOP EPOP …(E.6)
GROWTH = dGNP EFUEL …(E.7)
where:
dGNP is the forecast change in the GNP (%)

dPOP is the forecast change in population (%)
EGNP is the GNP elasticity
EPOP is the population elasticity
EFUEL is the fuel use elasticity
The use of fuel is preferred by some since it gives a good overall view of the historical
growth in traffic. By comparing the changes in fuel use with changes in GDP one calculates
the effect of GDP changes on fuel use. Coupling this with the forecast GDP gives the total
forecast traffic growth. This method has advantages in that it implicitly considers the
population growth.
As an example of the elasticities, in India CES (1991) gave values for the GNP elasticity of:
! 1.75 and 1.0 for freight vehicles
! 1.0 and 0.5 for passenger vehicles
in the periods 1994-2005 and after 2005 respectively. Values of 1.43 and 2.71 were used for
the population elasticity for freight and passengers respectively. In Nepal (NDLI, 1993)
adopted a value of 1.5 for the elasticity of fuel. The use of declining elasticities with time is
common as this reduces the impacts of future uncertainties on the predictions.
One important feature of traffic growth is that it is different depending upon the level of a
country’s development. Developed countries, with high levels of vehicle densities have
markedly different car ownership growth rates than developing countries that are building up
their densities. The model that best reflects this is a sigmoidal (‘S’ shaped) model. This is
discussed in Button and Ngoe (1991) who present a generalised model with data covering a
number of developing countries to use with the model.

Version 1.0
Traffic data
WARNING
When forecasting traffic, always consider whether or not the predictions are
reasonable. It is easy to adopt what seems to be a relatively low growth rate,
such as 5 per cent, but this may mean that after a period of time your facility is
at capacity. For this reason it is often prudent to adopt several different growth
rates based on short, medium, and long-term considerations, with the rates
declining in future years.
There are four principal categories of future traffic which are forecast:
! Normal traffic
The future traffic that can be expected assuming the current trends (for example,
historical patterns) remains steady.
! Diverted traffic
The traffic that can be expected to divert to the road because of the improvement.
! Generated traffic
Traffic that would not have existed but is expected because of reduced travel times or
diversion from other modes.
! Induced traffic
Traffic expected because of the new development created by better access (for example,
building a road into a new area will open it up for development).
E.5 Traffic volume and classification surveys

The purpose of these surveys is to collect data on the number and types of vehicles passing a
point on a link (link counts) or making specified movements at a junction (turning counts).
The occupancy of the vehicles may also be recorded to provide data on the volumes of people
using the road space. Volume and classification surveys are carried out either by manual or
automated traffic counts.
E.5.1 Manual counting

Manual surveys are conducted with a surveyor standing by the road, counting and classifying
the vehicles as they pass, dividing the survey into fixed time periods (usually 15 minutes). It
is normal for the surveyor to record only one direction of flow.
For relatively low flows, count marks on a suitable form are adequate, but re-settable hand
counters mounted on boards are required for heavier flows. Another method is to use a hand
held data logger or personal computer for this purpose.
E.5.2 Moving vehicle surveys

As described in TRL (1988), a general indication of the traffic flows can be obtained through
a moving vehicle survey. As a vehicle drives along a section of road the number of vehicles
met and the number of vehicles overtaking the survey vehicle, and overtaken by the survey
vehicle, are recorded. The flow is calculated as:
x+y
q= …(E.8)
t
where:

Version 1.0
Traffic data
q is the total flow in both directions in time t

x is the number of vehicles travelling in the opposite direction
y is the number of vehicles that overtake the survey vehicle less the number
overtaken by the survey vehicle
This expression assumes that the flows in each direction are balanced. To apply this equation
one first converts the flows to veh/h for by-hour-of-day. Using data on the variation of flow
by-hour-of-day, the one-hour flow is converted to an ADT. This is then factored up to an
AADT using the procedure described earlier.
In the Tamil Nadu Project Co-ordinating Consultancy Study, Riley (1998) adjusted the hourly
flow from the moving survey using factors based on 7-day counts that were done using
manual methods. Figure E.3 shows the excellent agreement between 7-day count data
factored from the moving survey and those using data from the actual 7-day count stations.
8,000
7,000 Line of equality

y = 1.02x
AADT Derived from Moving Observer Count
R2 = 0.97
6,000
5,000
4,000
3,000
2,000
1,000
0
0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000
AADT: 7 - day Count
Figure E.3 Comparison of moving car and 7-day survey counts from India
E.5.3 Automatic counting

Automatic traffic counter equipment consists of detector(s) to detect the vehicle, and a
counter to record the information. There are a number of different types of detectors used:
! Pneumatic tubes
When a wheel crosses the rubber tube and air impulse is sent along the tube which
activates an air switch. They are very common and relatively reliable over several days
and simple to operate. Their disadvantages are that the tube in time wears and breaks and
they are unsuitable where there is heavy breaking or turning. They are mainly confined
to short-term count stations.
! Inductive loops
A wire loop carrying a pre-determined frequency signal is embedded in the pavement. A

vehicle passes over the loop and changes the inductance and thus the frequency, thereby

Version 1.0
Traffic data
detecting the vehicle. It is almost universally used for permanent count stations as it is
very reliable, relatively inexpensive and largely unaffected by the weather. It is used at
almost all traffic signals.
! Positive contact
The detector is activated by the weight of the vehicle causing two thin metal strips to
touch and so to complete an electrical circuit. Provides an excellent signal but fairly
short life so limited to special counting and research.
! Video
Analyses a video recording to estimate counts and speeds.

! Other types:
❏ Photoelectric
Uses infrared to detect the interruption of light. Not reliable.

❏ Pressure sensitive
A pressure sensitive switch is embedded in the roadway surface.

❏ Magnetic
Senses the change in the earth’s magnetic field.

❏ Radar
Uses radar signal.

❏ Ultrasonic
Transmits an ultrasonic signal.

❏ Tribo-electric cable
Consists of a two-core cable that creates a charge when it comes into contact with a
wheel.
❏ Piezo-electric film
Creates a voltage when comes into contact with a wheel.

Counters work on similar principles. The detector transmits a signal to the counter. Most
counters record the time when the axle (or vehicle) is detected. The data are used to establish
the vehicle class and, optionally, to calculate the speed. With axle detectors, two detectors are
required to establish these accurately.
As an example of how this is done, consider a two-axle vehicle which is detected by two
detectors at a distance ‘D’ metres apart. The following is how the speeds and axle lengths are
calculated:
D
v1 = …(E.9)
t 21 − t 11
D
v2 = …(E.10)
t 22 − t 12
SPACING 1 = v 1 (t 12 - t 11 ) …(E.11)

Version 1.0
Traffic data
SPACING 2 = v 2 (t 22 - t 21 ) …(E.12)
where:
D is the distance between the detectors (m)

tij is the time of detection, in s, at detector i and axle j
vi is the velocity of axle I
SPACINGi is the spacing, in m, of axle I
The values for v1 and v2 represent the velocity of axle 1 and the velocity of axle 2 (in m/s).
The spacings are the distances between axle 1 and axle 2 (in m), based on these velocities.
These values are usually very similar, with the differences due to timing errors in the
detectors. It is common to average the values or else to adopt only one.
It must be appreciated that there are errors associated with the measurement of all automated
traffic counters and even when installed perfectly there will still be errors due to the rounding
off of the timing data. For example, if the recorder is accurate to a millisecond, the time will
fall within the range of 0.001 + 0.0005 secs. While this may seem to be quite small, if the
detectors are placed close together (D in the above equations), it can result in a sizeable error
when measuring speeds. As described in Bennett and Dunn (1992), on the basis of this it is
recommended that detectors be spaced at 5 m intervals.
The axle spacing data allows a vehicle to be classified into a number of different classes. It
needs to be appreciated that all automatic classifiers follow certain logic in assigning the
vehicles to specific classes. These are usually based on the common vehicle types in the
country of development so may not be appropriate to every country. Unfortunately, suppliers
of traffic counters are often reticent to modify their systems for local conditions so the output
often needs to be scrutinised carefully.
There are a number of caveats to the use of automated counters, the:
! Count sites should not be located in places where they are likely to be interfered with
(for example, near schools).
! Road at the site should be straight and level.
! Detector must be at right angles to the traffic flow; and if two detectors are used they
must be parallel.
! Detector must be firmly fixed to the road surface.
! Detector or connections must not cross the footpath.
! Counter must be securely locked to a permanent object such as a telephone pole.
If only a single detector is used the counter will only record the total number of axles and the
counter assumes that two axles equals one vehicle. This overestimates the total volume and a
correction factor must be applied, calculated from manual classified counts as follows:
Number of Vehicles
Correction Factor = …(E.13)
0.5 Number of Axles on Vehicles
A correction factor is calculated for a specific site, but various sites can be averaged to
produce factors for an area, or different classifications of roads.

Version 1.0
Traffic data
E.6 Vehicle loading

Vehicle loading, or axle load, surveys are conducted using either static/low speed equipment
or high-speed weigh-in-motion (WIM) equipment. TRL (1978) gives a good discussion of
how to conduct static/low speed surveys.
The success of the survey, and the ease that it can be carried out, depends largely on the
choice of site. The site must be selected so that the traffic can be sampled easily and safely. It
should ideally be on a clear stretch of road with good visibility as it is important to give
traffic ample times to slow down and stop. It is often useful to survey at the crest of a hill
where vehicles will be travelling slowly due to the gradient. If both directions are being
sampled, it is not necessary that the measurements be made exactly opposite to each other.
The site should allow for vehicles to be safely weighed off the carriageway, and queue if
necessary. An ideal site is one where there are slip roads, although it is more common to use
the shoulders or a level area adjacent to the road.
The general area must be level and firm, with no high spots or risk of subsidence during the
survey. Depending upon the length of the survey and the type of equipment, a concrete pit
may be used for holding the scales. Alternatively, ramps may be used to ensure a smooth
transition of the vehicle onto the scales. Care should also be taken to ensure that vehicles
inadvertently driving over them couldn’t damage the cables. It is useful to have the approach
to the scales clearly defined using flags or rocks to guide the vehicles to the scales. If only
measuring on a single side of the vehicle it is best to position the scale on the driver’s side as
this facilitates positioning the vehicle.
Before the survey commences the scales should be calibrated using the manufacturer’s
recommended procedure. It is important to ensure that there are spare batteries, cables, etc.
available so that the survey will not be interrupted.
Experience has shown that it is best to monitor both sides of a road, as there are usually
differences in the axle loading. The sampling of vehicles should be made on a systematic
basis as opposed to random basis. Thus, every nth vehicle is weighed as opposed to selecting
vehicles when the scales are free (that is, a random approach). Since the heaviest vehicles
tend to travel the slowest, they have a higher likelihood of being sampled with a random
approach, thereby overestimating the axle loads. The sampling rate should be based upon the
traffic volume, assuming a measurement rate of 60 to 90 veh/h. The achievable measurement
rate is dependent upon both the equipment and the experience of the crew.
For each vehicle both the axle configuration and the axle loads should be recorded, as these
are required for calculating the vehicle damage factor (see Section 6.3.1). It is also useful at
the same time to collect additional data from the driver, such as the commodity carried, as
these data can be used in determining the road user effects.
With the advent of relatively inexpensive WIM technology, this is becoming more widely
used for collecting vehicle weight data. With most WIM technology, particularly portable
equipment, there is a trade-off between accuracy and sampling; one samples the entire traffic
stream but with less accuracy than one gets by weighing individual vehicles from the traffic
stream using static techniques. This reduction in accuracy arises because of the effects of
vehicle dynamics.
As with all measurements, WIM equipment is subject to random and systematic errors.
Slavik (1998) notes that calibration will limit the systematic errors, however, most procedures
will not always eliminate the random error. On-site calibration, which involves stopping and
statically weighing trucks and comparing their static loads to the dynamic loads, will consider
both systematic and random errors. But there has often been uncertainty as to how many

Version 1.0
Traffic data
vehicles one needs to weigh to achieve suitable confidence in the data from the WIM scale.
Slavik (1998) addresses this issue and shows that the sample size is much smaller for
estimating the average load as opposed to the number of equivalent standard axles. Figure E.4
gives the 90% confidence intervals for both of these based on a sample of 218 axles (Slavik,
1998).
The Slavik (1998) approach takes into consideration the properties of the WIM equipment
used, the condition of the road surface, the composition and loading of truck traffic which
makes the procedure site specific. The raw axle loads are multiplied by a correction factor to
convert them to adjusted axle loads:
a i = k ri …(E.14)
where:
ai is the adjusted axle load I

ri is the raw axle load I
k is the correction factor
80
70
ESA
60
90% confidence inteval, % of mean
50
40
30
20
Load
10
0
50 100 150 200
Number of Axles Used in Calibration
Figure E.4 90% confidence intervals for WIM equipment
This suppresses the systematic error but affects the distribution of axle loads. The
distribution is therefore corrected by converting the adjusted axle loads into corrected axle
loads using the equation:
(
c i = a i + f ai − a i ) …(E.15)
where:
ai is the mean adjusted axle load
f is a correction factor

Version 1.0
Traffic data
The values for k and f are established from the on-site calibration data using the following
equations. Their derivation is described in Slavik (1998).
 si − σ si si + σ si 
 n n 
k = 0.5  +  …(E.16)
 ai ai 
 
f = 1 − Ve i / Va i …(E.17)
where:
si is the mean static axle load
σsi is the sample standard deviation of the static axle load

n is the sample size
Vai is the variance of the adjusted axle loads
Vei is the variance of the difference between the adjusted and static axle loads

Version 1.0
Unit cost data
Appendix F Unit cost data
F.1 Introduction
The HDM road model operates by predicting the amount of resources consumed and
multiplying them by the unit costs. It is therefore necessary to supply unit cost data for RUE
and works effects (WE). These costs can be supplied as financial and/or economic costs,
which are defined below.
The RUE unit costs can be established from various sources that are listed in Table F.1. In
establishing these unit costs, they should reflect the cost over the life of the project, normally
10 - 20 years. For most items the current prices can be used as the basis since inflation can be
expected to influence the various components similarly so that they maintain the same cost in
relation to one another. An exception to this is often fuel and lubricants that are heavily
influenced by world supply and demand. As described later, it is therefore better to estimate
these based on the long-term average future oil prices1.
Table F.1
Sources for unit cost data
Data source Data item

Vehicle dealers and distributors Replacement vehicle price, finance charges
Motor vehicle magazines/trade publications Replacement vehicle price
Oil companies Fuel prices, discounts, regional variations
Customs and Excise Department Fuel prices, tax structures
Tyre companies Tyre prices, discounts, regional variations
Insurance companies Insurance and accident costs
Trade unions Wages and earnings
Licensing authorities Annual license and registration fees
For works effects, unit costs are best established from recent tender prices or contracts.
F.2 Economic and financial costs

HDM allows for two types of costs; economic and financial, although the economic costs are
those generally used in analyses2. The financial costs are the market costs. The economic
costs are the market costs net of (excluding) taxes and subsidies (for example, diesel is
subsidised in many countries). In establishing the economic costs shadow pricing is often
used. This sees the price of the commodity adjusted to reflect the real scarcity value of the
resources in the commodity. Daniels (1974) discusses the use of shadow pricing in a number
of different transport studies in developing countries and highlights some of the difficulties
1
For example, Hoff & Overgaard (1994) indicate that in early 1994 the prevailing price for a barrel of oil
(158.6 l) was $USD 15 whereas the long-term average price was expected to be $USD 20.
2
HDM-III also included foreign exchange costs.

Version 1.0
Unit cost data
encountered in using shadow pricing. While shadow pricing is desirable, for most studies it is
sufficient to quantify the economic costs, as the market costs net of taxes.
To illustrate how this is done, consider Figure F.1. This shows the economic and financial
cost components of petrol and a medium truck from Thailand. Petrol had a series of taxes and
levies added to the cost of production and supply. The taxes with the truck had two
components, the:
1 Import duty on the CKD chassis
2 Sales tax on the completed unit
Since these were levied at different stages of production one cannot just deduct them directly
from the financial cost.
Petrol Truck
6.2% Dealer
6.3% Sales Tax
8.9% Body
Tax
Tax
Tax 56.2%
17.9% Assembly
Financial Cost
8.9% Import Duty
Retailing 6.7%
Transport 5.6%
Economic Cost
Refining 4.5%
51.8%
CKD Chassis
Shipping 4.5%
Oil 22.5%
Figure F.1 Breakdown of 1992 Thailand petrol and medium truck prices
When deducing the taxes from the market cost, it is important to consider not only the direct
but also the indirect taxes. To illustrate this, consider the cost of petrol from Thailand shown
in Figure F.1.
The indirect tax component of this price amounted to 56.2 per cent. The cost of production
and supply 43.8 per cent, however, it is necessary to deduct the direct tax component from
this remaining cost. This would entail assessing profit margins and the direct taxes on these
profits. The foreign exchange component would consist of the refinery cost less production
costs and margins along with a component of distribution costs. For simplicity, most analyses
would ignore the direct taxes and just use the total indirect taxes.
Hoff & Overgaard (1992) as well as Hoff & Overgaard (1994), Kampsax (1992), NDLI
(1997), Kinhill (1998) used an alternative approach to that outlined above. Instead of starting
with the financial cost of fuel and deducting the taxes, this method begins with the cost of a

Version 1.0
Unit cost data
barrel of oil and increases the costs to reflect the refining, transport and marketing costs. It
has the advantage that it will identify any hidden taxes that may be missed using the above
approach. These hidden taxes may arise in situations such as Nepal where the Indian
government taxes fuel exports to Nepal and these taxes are later refunded as a credit to the
Nepal government. Since these taxes are external to Nepal several projects missed removing
them when calculating the economic cost of fuel. An example of this approach from Hoff &
Overgaard (1994) is given in Table F.2.
Table F.2
Example of calculating economic cost of fuel
Component Petrol ($) Diesel ($)

Crude oil price 20.00 20.00
Refining margin 6.20 4.80
Transport and distribution 3.00 3.00
Cost per barrel 29.00 27.80
Cost per litre 0.184 0.175
Source: Hoff & Overgaard (1994)
Another argument in favour of using the price of a barrel of oil for calculating fuel prices is
that oil exporting countries often price fuel below the prevailing world price for oil. As noted
by Hoff & Overgaard (1994), changes in fuel consumption as a result of road projects affect
the quantity of fuel available for export and the potential export as opposed to the market
price is therefore the appropriate value of fuel to use.
F.3 Road user effect unit costs
F.3.1 New vehicle price: S-I

Because of high sensitivity, the new vehicle price should always be quantified as accurately
as practical. The new vehicle price is used in calculating the parts, depreciation and interest
costs. It is the singularly most important unit cost, and probably data item, in any HDM
analysis.
The values should be those for the vehicle less tyres, since tyres have their costs calculated
separately. If the tyre cost was used there will be double counting in the analysis1.
Within a representative vehicle class there will be a range of different vehicle types, each
with different vehicle prices. The best method for calculating the representative vehicle price
is by using recent sales data to determine the number of different models of vehicles sold.
Weighting these frequencies by the costs of the individual models will give the weighted-
average cost. Since most commercial vehicles are sold in a cab and chassis configuration it is
necessary to increase this cost to reflect the additional costs of bodywork. The same often
applies to heavy buses as well.
As discussed earlier, it is necessary to convert the vehicle prices from financial costs to
economic costs. If there are different import duties and taxes on individual vehicles, for
example due to their engine capacity or country of origin, it is necessary to convert the prices
1
It should be noted that in HDM-III this would lead to an underestimation of the interest costs.

Version 1.0
Unit cost data
from financial to economic before calculating the weighted-average. If there are uniform
taxes for all vehicles within a class, the weighted-average financial cost can be directly
converted to the weighted-average economic cost with a single factor.
In many countries vehicles are completely rebuilt at a certain kilometreage. This sees major
components replaced or refurbished. If the cost of rebuilding is included in the new vehicle
price the depreciation costs will be higher over the life of the vehicle. It is better to have two
different vehicles, one for vehicles that have not been rebuilt and a second for those that have
been rebuilt. This will reflect the different parts costs while not biasing the depreciation costs.
In HDM-III there was no mechanism to include a residual value for the vehicle in the
depreciation cost. These can be quite high, particularly in countries where old vehicles are
assigned to different activities, such as port work. To circumvent this problem, Transroute
(1992) assumed that vehicles had a 30 per cent residual value. The residual value was
discounted at the assumed accounting rate of interest and deducted from the replacement
vehicle price. This was done to reflect the “cascading down of both types of vehicles in their
later years to short distance, low annual mileage, low maintenance, long life work”. Since this
procedure served to reduce the maintenance and repair costs, the latter were adjusted through
the calibration parameters. HDM-4 provides for a residual value in the analysis.
F.3.2 Fuel and lubricant costs: S-III/S-IV

In HDM-4 the fuel type is defined for each representative vehicle. If vehicles within the
representative vehicle class have petrol or diesel, it is recommended that these be modelled as
two different vehicles given the difference fuel consumption rates of these engine types.
When there are different types of the same fuel available, for example regular, premium and
unleaded petrol, the fuel cost for each of these types should be weighted by the proportion of
each fuel type sold, or an estimate of these proportions, to obtain a weighted-average cost.
In some countries there are major fluctuations in fuel prices by region due to transport and
distribution costs. Another problem arises when there are distortions created by artificial
exchange rates. This can see some vehicles, usually those owned by the government,
obtaining fuel at a much lower cost than privately owned vehicles. Under these
circumstances, for financial analyses it is necessary to define separate vehicle classes by tax
regime or ownership, and to use different financial fuel costs for the vehicle classes.
HDM requires the cost per litre of engine oil. This should be established in the same manner
as petrol. It should be noted that this is a minor cost component so does not warrant a large
effort.
F.3.3 Tyre cost: S-III

The average tyre cost is the cost of the tyre - not the entire wheel. Tyres can have a wide
range of costs and qualities. In some countries more expensive tyres are preferred because
there is a net advantage due to their longer lives. Some users prefer radial to bias ply tyres;
some tyres have inner tubes while many are tubeless. All of these need to be taken into
account when establishing the average tyre cost.
The average tyre cost is multiplied by the rate of tyre consumption (in tyres/1000 km) to
establish the cost of tyre wear. The cost should therefore exclude inner tube costs and be
restricted to the tyre carcass. If inner tubes are a major cost item associated with road use, for
example due to frequent damage and replacement, the tyre cost should be increased to reflect
this. This is done as follows:

Version 1.0
Unit cost data
1 Establish the cost of inner tubes per 1000 km (ITC) for an average road.
2 Run HDM using average conditions and establish the tyre cost for each representative
vehicle class in cost/1000 km (TYRE).
3 Multiply the average tyre cost by 100 ITC/TYRE to obtain the adjusted tyre cost. This
adjusted tyre cost should be used in all further modelling.
The issue of radial versus bias tyres; high quality, long-life versus lower-quality, shorter life
tyres is accounted for by calculating a weighted-average cost. The number of different tyre
sizes sold for each vehicle within a representative vehicle class should be weighted by the
average cost of the tyres to obtain a weighted-average cost. If there are significant differences
in prices between countries of origin, the averaging should also reflect this. Contacting tyre
retailers and asking them for an assessment of the percentages of sales of different tyre types
can readily collect these data. For example, in Nepal NDLI (1993) recorded the tyre types on
a sample of vehicles stopping at a customs checkpoint.
The HDM tyre model also requires the cost of retreads as a percentage of the new tyre price.
This is used in the calculations to establish an average tyre cost for vehicles that use a mix of
retreads and new tyres. There are two considerations here, the:
1 Average value
2 Probability of a retreaded tyre being used
The average value is calculated in the same manner as for a regular tyre. This value then
needs to be reduced based on the probability of a retreaded tyre being purchased since HDM
assumes an equal likelihood of a retread or a new tyre being on a vehicle. This is done as
follows:
1 Interview tyre retailers, other industry sources, truck drivers, etc. and establish the
percentage of tyres in use, which are retreads (PCTRET).
2 Express the average retread cost as a percentage of the new tyre cost (RETPCT).
3 Calculate the retread cost as:
RREC = PCTRET RETPCT
This is the value used as input to HDM.
F.3.4 Passenger time, crew and cargo costs: S-II/S-III

The value of passenger time is probably the singularly most difficult cost to quantify. There
have been a number of reports dedicated to this issue. Symonds (1997) gave a review of the
more recent work with Cox (1983) giving a good historical view. Symonds (1997) also gave a
good discussion of the issue of non-linearity of travel time savings1.
1
This relates to the question of to the appropriateness of treating a large number of small time savings as
equal to a single large time saving of the same magnitude. There are several arguments in favour of this, but
the most common is that since small time savings cannot be perceived, they should not be valued lower
than large time savings - hence the non-linearity of the value of time savings. Symonds (1997) reviews the
issue and concludes “… there are no compelling theoretical or empirical reasons to adopt non-linear
assumptions on values of time savings, and there are strong practical reasons in favour of a constant
[value]”. They note that the exception to this is leisure time where there are different levels of disutility but
that for practical reasons linear values should be used.

Version 1.0
Unit cost data
While some studies have simply ignored passenger time, this is not wise as it can bias the
results in favour of projects with high transport costs. Conversely, an unrealistically high
value for passenger time can bias projects in favour of those with large speed increases.
In quantifying the value it should be recognised that there are two sectors of the economy: the
formal and the informal sectors. The formal sector is comprised of wage earners. The
informal sector is those without salaries - such as many of the rural population in developing
countries.
For simplicity, travel is usually separated into work and leisure or non-work travel. Work
travel is comprised of those on business activities. Leisure, comprised of other activities (for
example, social, visiting family, personal business, going to school, etc.) is often a significant
component of travel in developing countries.
Savings in time when journeys are related to work clearly have a value; if less time is spent
travelling more time in the working day can be used for economically productive purposes.
Another way of looking at this is the employer pays the employee an hour’s wages for no
return. The employer would be willing to pay equal to an hour’s wages to reduce travel time
by one hour. It can be argued that due to overheads and social charges the employer would
be prepared to pay even more, but the common practice in developing countries is to equate
the value of work time to the earnings rate of the traveller. In developed countries, where
there are often large social costs, this gross wage is increased by the employer’s on-costs.
The use of wage rates is complicated by the fact that official statistics on wages will probably
underestimate the earnings of travellers. Wage statistics do not usually cover the earnings of
the highest paid workers, and the wages of those travelling during working time may be
higher than the average. Often, there are also regional variations in wages that make it
impractical to adopt a national average. For example, in Thailand the 1995 average wage was
Baht 9900/month in Bangkok versus 4500 in the north-eastern region of the country.
Those in the informal sector or travelling in leisure time are not considered to be productive
in the same way as those travelling in work time. Ultimately, the value of non-working time
should reflect Government policy. If the policy is to maximise GDP, ignoring leisure time
preferences and increasing the welfare of passengers, then a zero value should be placed on
non-work time. It must be recognised that assigning a zero value to the time for those in the
informal sector will serve to bias the results in favour of those who contribute to the modern,
cash economy.
There is evidence that the leisure time savings are valued, particularly since these travellers
still prefer their trips to be faster than slower and are often willing to pay more for this to
happen. How much a person is prepared to pay for a quicker trip is based upon their income
and wealth. It is therefore common practice to assume a value of personal time related to the
individual’s income. Various percentages have been assumed in different studies, usually in
the range 20-50 per cent, but 20-25 per cent seems to be the most common.
Many who travel in personal time do not earn any income and so using this approach would
have no value for their time. In affluent societies this would not be true, but it is argued by
some that in some countries a zero value of time is appropriate.
In these instances the mean income is used to calculate the value of time. The alternative
approach is to calculate the value of time based only on those working and then to apply the
value to all travellers. This will yield a higher value of time than using the mean income of
travellers.
To summarise, there are three sets of passenger time values to be considered:
1 Employed, travelling in work time,

Version 1.0
Unit cost data
2 Employed, not travelling in work time; and,
3 Unemployed or in non-paid activities.
There is evidence that travel time values are higher for traffic travelling under congested as
opposed to free-flow conditions. MVA et al. (1987) suggested a factor of 1.4 for the situations
they studied, and an even higher factor under more congested conditions.
When establishing the value of time, particularly for truck and bus operators, it is important
to include extra income that may be obtained above the base salary, for example:
! Daily allowances - to cover food and rest,
! Carrying passengers (trucks) or extra, non-reported passengers (buses); and,
! Backhaul of goods by trucks - where the operator instead of the owner keeps the
income.
These can be significant, for example in India NDLI (1997) estimated that the total salaries
for the truck driver and helper were Rs 4500 but that there was a further monthly income of
Rs 3000 from collecting passengers and Rs 3000 from daily allowances. About 20-25 per cent
of drivers managed to backhaul goods, earning a further Rs 7000 per month.
It is also common to differentiate between modes of travel. This, for example, sees different
passenger time costs for passengers in private transport and those in buses or other public
transport.
Table F.3 gives an example of calculating passenger time and crew costs, using basic data
from India NDLI (1997) along with some assumed additional values. The notes at the bottom
of the table detail how the individual values were calculated.

Version 1.0
Unit cost data
Table F.3 Example of calculating passenger time and crew costs
Two- Car Medium Heavy Heavy

wheeler truck truck bus
(1) Income - Work Trips (Rs/h) 15 32 - - 43
(2) Income - Non-Work Trips (Rs/h) 16 47 - - 31
(3) Time Value - Non-Work Trips (Rs/h) 3.2 9.4 - - 6.2
(4) Time Value - Not Employed (Rs/h) 0.6 1.8 - - 1.2
(5) Persons per vehicle 1.5 3.5 2.0 2.0 32.0
(6) Passengers per vehicle 1.5 3.3 - - 30.0
(7) Percentage in employment 75 75 - - 50
(8) Number employed 1.1 2.5 - - 15.0
(9) Number not employed 0.4 0.8 - - 15.0
(10) Percentage travelling in work time 36 19 - - 34
(11) Number travelling in work time 0.5 0.6 - - 10.2
(12) Number employed but not on work trip 0.6 1.9 - - 4.8
(13) Passenger Time cost Rs/h 9.7 38.5 0.0 0.0 486.4
(14) Crew per vehicle - 0.2 2.0 2.0 2.0
(15) Cost of Driver (Rs/month) - 600 3,000 3,000 3,000
(16) Cost of Helper(Rs/month) - 0 1,500 1,500 1,500
(17) Additional Income—Allowances - 0 3,000 3,000 3,000
(Rs/month)
(18) Additional Income—Phantom Passengers - 0 3,000 3,000 3,000
(Rs/month)
(19) Additional Income—Phantom Backhauls - 0 1,500 1,500 0
(Rs/month)
(20) Monthly Crew Time (Rs/month) - 600 13,000 13,000 10,500
(21) Crew Time Cost Rs/h (200 h/month) 0.0 3.0 65.0 65.0 52.5
(22) Cargo Cost (Rs/h) 0.0 0.0 2.5 5.0 0.0
Total Cost (Rs/h) 9.7 41.5 67.5 70.0 538.9
Notes:
(3) = (2) x 0.20. Assumed 20% for value of work time.
(4) = (3) x 0.20. Assumed 20% of value of non-work time.
(8) = (6) x (7)
(9) = (6) - (8)
(11) = (8) x (10)/100
(12) = (8) - (11)
(13) = (1) x (11) + (3) x (12) + (4) x (9)
For cars which had less than one crew, (15) = (14) x Rs 3,000
(19) is calculated from the extra revenue (here Rs 7,000/month) times the percentage of backhauls
(here, approximately 20 per cent).
(20) = (15) + (16) + (17) + (18) + (19)
(21) = (20)/200. Assumed 200 working h/month
(23) = (13) + (21) + (22)

Version 1.0
Unit cost data
If goods spend less time in transit the amount of goods held in inventory may be able to be
reduced. Some goods will not benefit from the time savings, for example, if they arrive before
the business opens and thus cannot be unloaded. Thus, only a portion of the goods in transit
should be included in the calculations. The value will depend upon the nature of operations in
the country, or even in the area of the country, but values of 50-75 per cent are common.
The value of cargo time is calculated using the opportunity cost and the following equation:
PCTCGT OPC VALCAR

CARGO = …(F.1)
365 × 24
where:
CARGO is the cargo cost in cost/h

PCTCGT is the fraction of vehicles whose cargo will benefit from time savings
OPC is the opportunity cost of the cargo as a decimal
VALCAR is the value of the cargo
F.3.5 Maintenance labour: S-III

The maintenance labour costs should reflect the costs of labour, tools and workshop
overheads. It is therefore not appropriate to base the costs only on the prevailing wage rates.
In some countries the wages will contain a number of transfer payments such as taxes, social
security, etc. These need to be taken into consideration when calculating the average cost. For
example, in Transroute (1992) assumed a 100 per cent overhead on the monthly wage and
250 h/month working time.
One of the simplest methods of estimating the average hourly labour costs is by conducting a
small survey. A number of typical maintenance activities should be defined such as replacing
the clutch, engine tuning, etc. Workshops should be surveyed to obtain their estimate of the
time it would take to perform the repairs and the total costs for the repairs. This will give an
hourly cost that includes overheads as well as labour and considers unproductive time. In
developing countries where a mechanic also has semi-skilled and unskilled assistants this
approach will implicitly include their costs as well1. The resulting value should be adjusted
for taxes etc. to convert it to an economic cost.
F.3.6 Interest rate: S-III

HDM requires an annual interest rate for calculating the opportunity cost of vehicle
ownership - also called the interest costs. For economic analyses, the interest rate should be
the same as the discount rate. For financial analyses, the value selected should be the real
long-term interest rate (that is, the market-borrowing rate less the underlying inflation rate).
The value selected must be based on long-term considerations since interest costs can
constitute a major component of the total vehicle operating costs.
1
Some analysts adopt the approach of calculating an average hourly rate based on a percentage of
skilled/semi-skilled/unskilled time spent working on a vehicle. This method is generally not appropriate
since in many countries the practice is to have more than one person working on the vehicle at a time. Thus,
the costs should be accumulated rather than weighted.

Version 1.0
Unit cost data
F.3.7 Overhead/standing costs

The overhead/standing costs are those costs associated with vehicle ownership. They consist
of costs due to licensing, insurance, and garaging. Overhead costs are often ignored in
analyses due to their having a minor impact on the total costs as influenced by road
conditions. Another view is that they should be entirely excluded since in an economic
evaluation the marginal overhead is zero.
If overhead costs are to be included in the analysis, it is best to use the overhead costs to
estimate an hourly value of vehicle time by dividing the total annual overhead costs by the
number of hours per year the vehicle is used. This hourly cost can then be treated in an
equivalent manner to travel time/crew costs.
F.4 Works effects unit costs


Version 1.0
Economic data
Appendix G Economic data

G.1 Introduction
G.2 Discount rate
G.3 Analysis period

Version 1.0
Determining sample sizes
Appendix H Determining sample sizes

The approach to use to determine the appropriate sample size depends upon the number of
samples. With few data items one deals with small samples (n < 30), so the t Distribution
should be used. With larger samples (n > 30), the normal distribution should be used. The
basic approach adopted is the same with each distribution.
The standard error of the mean for a sample of data is given by:
s
E= …(H.1)
n
where:
s is the standard deviation of the sample

n is the number of observations
The true mean of the population is aid to lie between x + E. Thus, the larger the sample size
the smaller the value of E and the closer the sample mean will be to the population, thereby
limiting the bias.
The necessary sample size for a given level of accuracy and confidence is given by:
s
e=K …(H.2)
n
where:
e is the error limit

K is the t statistic (n < 30) or the number of standard deviations about the mean of
the normal distribution (n > 30)
Rewriting the above equation gives the following equation for determining the required
sample sizes:
K 2 s2
n= …(H.3)
e2
Hamilton (1990) gives the values in Table H.1 for 90 and 95 per cent confidence intervals
using the t Distribution (n < 30). Table H.2 gives the values of K for the Normal Distribution
(n > 30).
In order to use the above equations for determining the required sample size, it is necessary to
know the standard deviation of the sample. Since this is not available until after the study has
been completed, it appears impossible to estimate the required sample sizes before the study.
However, the distributions of many data items are relatively stable between sites and between
countries and have a consistent value for the coefficient of variation. This is defined as:

Version 1.0
Determining sample sizes
σ
COV = …(H.4)
µ
where:
COV is the coefficient of variation

σ is the standard deviation
µ is the mean
Table H.1
Critical ‘t’ values (n < 30)
Number of samples Critical t value by confidence interval (K)

90 95
3 2.920 4.303
4 2.353 3.182
5 2.132 2.776
6 2.015 2.571
7 1.943 2.447
8 1.895 2.365
9 1.860 2.306
10 1.833 2.262
Table H.2
Critical K values (n > 30)
Confidence level K
68.3 1.00
90.0 1.65
95.0 1.96
95.5 2.00
99.0 2.58
99.7 3.00

Version 1.0
Survival curve analysis
Appendix I Survival curve analysis

This example of calculating the survival curve is based on the approach of Zaniewski et al.
(1982) and uses data from Bennett (1985).
Table I.1 presents the results of a survey of vehicle ages. A sample consisting of the number
of vehicles at different ages was obtained and expanded to represent the entire population.
This expanded sample is presented in column (2) of Table I.1 and column (3) contains the
number of the vehicles of each age, which were originally registered.
Dividing the expanded sample in column (2) by the number of original registrations in
column (3) gives a survival curve in column (4). Multiplying this column by 100 per cent
gives the percentage of each age surviving.
Zaniewski et al. (1982) recommends modifying this survival curve to improve the predicted
service life. In theory, the highest value in column (4) represents a year when all of the
registered vehicles are still in the population. As such, it is the sampling factor for the survey.
Dividing the survival curve in column (4) by this sampling factor gives the survival ratios in
column (5). These ratios are used in place of the values in column (4) for the survival curve1.
The sum of column (5) is the area under the survival curve and this corresponds to the
average service life. The data in Table I.1 indicate that the average service life is 19 years.
In column (6) of Table I.1 the average annual kilometreage for vehicles in each age range is
given. To determine the average lifetime kilometreage the product of the survival ratio
(column (5)) and the annual utilisation is calculated. These products are given in column (7).
The sum of the values in column (7) represents the average lifetime kilometreage - in this
example 187,202 km. The average annual kilometreage is given by the average lifetime
kilometreage divided by the average service life. Using the earlier derived service life of 19
years results in an average annual kilometreage of 9,850 km/yr.
1
For those years preceding the sampling factor the survival ratio is set to 1.0.

Version 1.0
Survival curve analysis
Table I.1
Example of establishing survival curve and lifetime utilisation
Vehicle Number in Original Sample/original Survival Average Survival

age expanded registrations [(3)/(2)] ratio annual Ratio x Av.
sample utilisation annual util
(km/yr) [(5)*(6)]
(1) (2) (3) (4) (5) (6) (7)
0- 1 65131 67149 0.97 1.00 11500 11500
1- 2 59261 61824 0.96 1.00 11500 11500
2- 3 71789 73527 0.98 1.00 11500 11500
3- 4 81653 83612 0.98 1.00 11500 11500
4- 5 91854 99213 0.93 1.00 11250 11250
5- 6 99569 102626 0.97 1.00 11000 11000
6- 7 89524 90302 0.99 1.00 10750 10750
7- 8 65413 74626 0.88 0.88 10500 9284
8- 9 68118 70426 0.97 0.98 10250 10000
9-10 53328 56438 0.94 0.95 10000 9531
10-11 46308 48808 0.95 0.96 9750 9331
11-12 50636 56484 0.90 0.90 9500 8590
12-13 53721 63768 0.84 0.85 9250 7860
13-14 54516 67189 0.81 0.82 9000 7366
14-15 52632 65742 0.80 0.81 8750 7066
15-16 44482 58347 0.76 0.77 8500 6536
16-17 27301 42680 0.64 0.65 8250 5323
17-18 24714 37663 0.66 0.66 8000 5295
18-19 18188 33859 0.54 0.54 7750 4199
19-20 13388 27214 0.49 0.50 7500 3722
20-21 13190 30985 0.43 0.43 7250 3113
21-22 15208 41323 0.37 0.37 7000 2599
22-23 11130 38980 0.29 0.29 6750 1944
23-24 12516 46237 0.27 0.27 6700 1829
24-25 8812 36610 0.24 0.24 6500 1578
25-26 2958 25724 0.12 0.12 6500 754
26-27 5645 37095 0.15 0.15 6500 998
27-28 2884 26595 0.11 0.11 6500 711
28-29 1378 20575 0.07 0.07 6500 439
> 29 8087 393744 0.02 0.02 6500 135
Notes:
Sampling factor:

Version 1.0
Orthogonal regression
Appendix J Orthogonal regression

Linear regression analysis fits the following model to data:
y=ax +b+ε …(J.1)
where:
y is the dependent variable

x is the independent variable
ε is the error term
a and b are regression coefficients
The analysis is done in such a way that the error term is minimised.
There are instances where the dependent and independent variables are interchangeable, for
example as in the HDM equations to predict modified structural number from the Benkelman
Beam deflection and vice versa. This gives rise to two sets of equations:
y=ax +b+ε …(J.2)
x=c y +d+ε …(J.3)
The issue here is one of orthogonal regression where instead of minimising the sum of
squares in the x or y plane the minimisation needs to be done in the orthogonal plane. To
illustrate the importance of this, consider Table J.1. Columns 1 and 2 contain sample data.
These data are plotted in Figure J.1 along with the linear regression equations.

Version 1.0
Table J.1
Orthogonal regression illustrative data
Data Standard regression Orthogonal regression

X Y Y X Residual Y Y
Predicted Predicted X Predicted Predicted
(1) (2) (3) (4) (5) (6) (7)
1.00 9.50 9.44 1.05 0.05 9.46 1.00
1.14 9.40 9.37 1.18 0.04 9.38 1.14
1.29 9.20 9.28 1.32 0.03 9.29 1.29
1.43 9.20 9.20 1.44 0.01 9.21 1.43
1.57 9.10 9.13 1.57 0.00 9.13 1.57
1.71 9.00 9.05 1.70 -0.01 9.05 1.71
1.86 9.05 8.97 1.84 -0.02 8.96 1.86
2.00 8.87 8.89 1.97 -0.03 8.88 2.00
2.14 8.70 8.81 2.10 -0.04 8.80 2.14
2.29 8.83 8.73 2.24 -0.05 8.71 2.29
9.6 2.5
9.5
9.4 2.0
Y = -0.5561X + 9.9987
2 X = -1.6568Y + 16.695
9.3
R = 0.92
R2 = 0.92
9.2 1.5
9.1
Y
9.0 1.0
8.9
8.8 0.5
8.7
8.6 0.0
0.0 0.5 1.0 1.5 2.0 2.5 8.6 8.7 8.8 8.9 9.0 9.1 9.2 9.3 9.4 9.5 9.6
X Y
Figure J.1 Fitted regression lines
The two regression equations developed were:
y = - 0.56 x + 10.00 …(J.4)
x = - 1.66 y + 16.70 …(J.5)
Column 3 in Table J.1 shows the value of Y predicted using Equation J.2 above along with
the original X data from Column 1. Using these values as input data for Equation J.3 above
gives the value of X. As shown in Column 5, there is a difference between these predicted
values of X and the original X values. This difference would not exist had an orthogonal
regression been done.
Orthogonal regressions require sophisticated statistical techniques. However, when the
residuals are similar in each direction the following simple solution may suffice.
y=ax +b …(J.6)
x=cy +d …(J.7)

Version 1.0
so:
(x − d)
y=
c
 a+ 1
c
( ) (
b−d
c
)
y= x+ …(J.8)
 2   2 
   
and:
1  b−d ( )
x= y− c
( ) ( )
…(J.9)
 a+  
 a+ c
1 1
 c 
 2 
 
Using this approach with Equations J.4 and J.5 above results in the following equations:
y = - 0.58 x + 10.04 …(J.10)
x = - 1.72 y + 17.31 …(J.11)
Column 6 in Table J.1 shows the predicted values of y using Equation J.4 above in
conjunction with the data from Column 1. Substituting these values into Equation J.5 above
gives the predicted values of x (Column 7). These values are the same as the original data in
Column 1.

Version 1.0
HDM Tools user guide
Appendix K HDM Tools user guide
K.1 Introduction
HDM Tools is a set of software applications designed to assist in the calibration of HDM-4.
The software is available from the Internet at:
! http://www.opus.co.nz
! http://www.htc.co.nz
The HDM Tools applications have been designed to run completely independently of the
HDM-4 software, although the data can be used to update the HDM-4 data files.
The applications fall into three groups:
! RUE
! RDWE
! Analysis Tools
Each are discussed separately below.
K.2 Road user effects

The current HDM Tools RUE applications are:
! ACCFUEL - Effect of acceleration on fuel and tyre consumption simulation;
! CDMult - Wind adjusted aerodynamic drag coefficient;
! GEARSIM - Engine speed simulation;
! OPTIMAL - Optimal life calibration; and,
! WORKZONE - Work zone effects.
These applications are run from a single launch button and output data that can be imported
to the HDM Vehicle Fleet.
If the default HDM-4 vehicle parameters are used it is not necessary to run the HDM Tools
applications. However, if the default values are changed then the guidelines in Table K.1
should be used to establish if the HDM Tools applications need to be re-run.

Version 1.0
Table K.1
Criteria for re-running RUE HDM tools
Program When should be run Calibration level where

should be run
Level 1 Level 2 Level 3
ACCFUEL If there is a change of more than 5 per cent to the • • •
vehicle mass; 10 per cent to the frontal area and/or the
aerodynamic drag coefficient
CDMult If there is a change of more than 10 per cent to the • • •

frontal area and/or aerodynamic drag coefficient
GEARSIM For Level 2 or 3 calibrations of engine speed • •
OPTIMAL For all HDM applications using the optimal life for • • •
capital costs
WORKZONE If work zone effects are of interest
K.3 HDM Tools Launchpad

The HDM Tools RUE component was developed by Ian Greenwood1 of Opus International
consultants Ltd. Running under Windows 95, the software is installed as per any standard
Windows application.
The applications are run from the HDM Tools Launch pad. Figure K.1 shows this Launch
pad and the individual applications. These may be changed in the future as additional
modules are added to the HDM Tools software. Selecting any button the Launch pad will
start the individual application. The lauchpad remains open in the background unless it is
exited.
1
Comments and suggestions on the software can be sent to: ian.greenwood@opus.co.nz.

Version 1.0
Figure K.1 HDM Tools Launch Pad
The use of these individual applications is discussed in the following sections.
K.3.1 Vehicle data

Since the HDM Tools software is run independently of HDM-4, HDM Tools comes with its
own vehicle database. This is the same database used with the HDM-4 standalone RUE
model.
The data can be exported from HDM-4 using the Vehicle Fleet (VF) export function and,
when copied to the HDM Tools Launch Pad directory, these data will be used as input to
HDM Tools. The user can alter these data, and Figure K.2 shows the data entry screen
presented when the VEH_DATA button is selected from the Launch Pad.

Version 1.0
Figure K.2 Vehicle database entry screen
K.3.2 Fuel-acceleration simulation

The Fuel-Acceleration Simulation (ACCFUEL) Tool is used to generate data on the effects of
accelerations on fuel and tyre consumption. As discussed in Section 6.4.5, these accelerations
may arise due to traffic interactions (that is, congestion), the road alignment, roughness, and
non-motorised traffic or side friction.
The acceleration is specified in HDM-4 as the acceleration noise. This is the standard
deviation of acceleration. This parameter has been adopted since the mean acceleration over a
homogeneous section will always be 0. However, the additional fuel and tyre consumption
are proportional to the magnitude of the accelerations and the higher the acceleration noise,
the higher the magnitude of the accelerations and thus, the higher the fuel and tyre
consumption.
There are two modes of operation:
1 Simulation of a vehicle under a range of speed and acceleration noise levels; or,
2 Simulation of a vehicle under a user defined speed cycle.
The first option generates a large matrix of results that enable the HDM-4 software to
increase the calculated steady speed fuel consumption and tyre wear values to account for
acceleration effects. Simulating a vehicle travelling along a section of road with different
initial speeds experiencing varying levels of acceleration does this. The additional fuel for
each of these speeds and levels of acceleration noise is calculated and stored as a matrix. This
matrix is read by HDM-4 and the additional fuel due to accelerations is applied in the
analyses.

Version 1.0
The second option enables for comparison of simulated versus observed fuel consumption
data such that the input vehicle parameters may be adjusted to yield the best match. Here, the
user defines a speed cycle and the software predicts the fuel consumption associated with
travelling that cycle. Comparing this to known fuel consumption enables key parameters to be
calibrated.
To run ACCFUEL the user supplies the number of vehicles to simulate and the minimum
length of road that each vehicle should travel over. It is recommended that a minimum of 250
vehicles travelling 5-10 km be simulated. For each mean initial speed, vehicles are simulated
travelling with acceleration noises from 0.05 to 1.00 in 0.05 m/s2 increments. This results in a
matrix as illustrated in Figure K.3 and Figure K.4 for fuel and tyres respectively. The fuel
data are also presented as Table K.2.
1.2
1.0
0.8
dFUEL 0.6
0.4
0.90
0.2
0.75
0.60
Acceleration Noise
0.45 in m/s/s
0.0
0.30
10
20
30
0.15
40
50
60
70
0.00
80
90
Speed in km/h
100
Figure K.3 Additional fuel due to accelerations - medium truck
While the values show consistent trends, the fuel results show the inherent variations that
arise with Monte Carlo simulations. The tyre results do not have these irregularities due to the
way in which the tyre consumption is calculated (Bennett and Greenwood, 1999). The
following points will be noted in the figures:

Version 1.0
! The highest additional fuel consumption arises in the area of the minimum fuel
consumption. This is most noticeable with heavier vehicles.
! There is a significant increase in congestion effects with heavier vehicles over light
vehicles for the same speed and acceleration noise. This is because of their higher mass
and, thus, inertial resistance.
! There is only a very limited effect of congestion on motorcycles - speed has a much more
significant impact. This is because motorcycles have such a low mass that the inertial
effects on fuel are minimal.
9.0
8.0
7.0
6.0
5.0
dTYRE
4.0
3.0
2.0
0.90
0.75
1.0
0.60
0.45 Acceleration Noise
0.0 in m/s/s
0.30
10
20
30
0.15
40
50
60
70
0.00
80
90
Speed in km/h
100
Figure K.4 Additional tyre consumption due to accelerations – medium truck
When generating a matrix there are two output files per vehicle, one for fuel consumption and
one for tyre wear. Each file has the format of vehicle speed down the rows and acceleration
noise across the columns. The name of the files is dfuel.xx and dtyre.xx, where xx is the
vehicle number within the vehicle database.
If you run the Combine files option from the File menu, then the program will combine all
the various files into two large files ready for inclusion into HDM-4. These files are called
dfuel.out and dtyre.out.
To simulate a known speed profile, the user must supply a text file with the vehicle number to
simulate followed by rows containing the time (seconds), velocity (m/s), speed (km/h),
acceleration (m/s/s), distance (m). An example of this is:

Version 1.0
1 0.0 0.0 0.0 0.0
2 0.0 0.0 0.0 0.0
3 0.0 0.0 0.0 0.0
4 0.0 0.1 0.0 0.0
5 0.4 1.6 0.4 0.5
6 1.1 4.0 0.7 1.6
7 1.7 6.0 0.5 3.3
8 1.9 7.0 0.3 5.2
9 2.0 7.1 0.0 7.2
10 2.2 8.0 0.2 9.4
This table is then used in the analysis and then a single output file is created with the same
name as the input file but the extension .OUT. This contains all the input data plus two extra
columns, the first being the instantaneous fuel consumption (ml/s) and the second the
cumulative fuel consumption (ml).

Version 1.0
Table K.2
Additional fuel due to accelerations – medium truck
Speed Additional fuel due to acceleration (dFUEL) by total acceleration noise

(km/h) 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00
10 0.007 0.009 0.020 0.042 0.065 0.098 0.134 0.171 0.212 0.252 0.289 0.328 0.373 0.413 0.452 0.493 0.535 0.582 0.620 0.662 0.689
15 0.011 0.012 0.020 0.030 0.046 0.069 0.102 0.138 0.183 0.220 0.266 0.303 0.339 0.388 0.437 0.480 0.510 0.562 0.602 0.640 0.684
20 0.012 0.013 0.019 0.028 0.045 0.069 0.097 0.122 0.156 0.195 0.245 0.281 0.339 0.371 0.422 0.464 0.517 0.538 0.603 0.624 0.682
25 0.016 0.017 0.025 0.037 0.055 0.082 0.115 0.144 0.182 0.228 0.262 0.315 0.364 0.401 0.455 0.486 0.532 0.593 0.625 0.665 0.726
30 0.023 0.022 0.025 0.041 0.063 0.098 0.127 0.179 0.205 0.263 0.307 0.359 0.410 0.453 0.520 0.559 0.600 0.651 0.717 0.752 0.809
35 0.026 0.029 0.029 0.052 0.074 0.113 0.147 0.191 0.229 0.274 0.342 0.415 0.454 0.532 0.562 0.621 0.678 0.722 0.790 0.842 0.892
40 0.032 0.035 0.040 0.045 0.084 0.118 0.154 0.227 0.263 0.315 0.378 0.435 0.484 0.556 0.614 0.687 0.740 0.794 0.846 0.896 0.978
45 0.039 0.041 0.041 0.063 0.083 0.109 0.170 0.220 0.272 0.332 0.403 0.465 0.529 0.578 0.651 0.704 0.777 0.832 0.897 0.969 1.017
50 0.045 0.048 0.054 0.065 0.086 0.127 0.172 0.232 0.273 0.349 0.383 0.468 0.535 0.605 0.659 0.717 0.809 0.851 0.909 0.972 1.035
55 0.052 0.058 0.062 0.076 0.100 0.128 0.162 0.228 0.293 0.338 0.410 0.463 0.535 0.599 0.660 0.720 0.799 0.840 0.929 0.960 1.031
60 0.061 0.067 0.085 0.074 0.094 0.132 0.179 0.242 0.280 0.344 0.411 0.459 0.527 0.583 0.671 0.694 0.761 0.832 0.872 0.974 1.008
65 0.069 0.066 0.067 0.082 0.103 0.145 0.175 0.227 0.270 0.324 0.393 0.456 0.521 0.574 0.618 0.707 0.769 0.808 0.871 0.913 0.979
70 0.093 0.084 0.082 0.089 0.112 0.130 0.164 0.211 0.261 0.320 0.380 0.421 0.459 0.548 0.608 0.663 0.722 0.754 0.833 0.877 0.923
75 0.090 0.096 0.099 0.101 0.100 0.137 0.161 0.206 0.237 0.311 0.352 0.398 0.458 0.520 0.587 0.614 0.673 0.728 0.781 0.837 0.881
80 0.095 0.086 0.090 0.102 0.121 0.121 0.158 0.185 0.236 0.268 0.327 0.373 0.433 0.464 0.532 0.565 0.624 0.677 0.722 0.778 0.822
85 0.102 0.103 0.100 0.101 0.132 0.126 0.145 0.183 0.216 0.259 0.306 0.344 0.399 0.453 0.501 0.511 0.614 0.629 0.676 0.749 0.765
90 0.107 0.120 0.124 0.111 0.132 0.123 0.153 0.178 0.210 0.254 0.298 0.335 0.365 0.408 0.445 0.500 0.537 0.600 0.629 0.669 0.710
95 0.124 0.108 0.131 0.125 0.118 0.144 0.159 0.166 0.199 0.250 0.273 0.312 0.338 0.382 0.404 0.472 0.504 0.547 0.596 0.625 0.682
100 0.138 0.137 0.120 0.121 0.129 0.131 0.159 0.163 0.198 0.217 0.256 0.289 0.326 0.341 0.391 0.430 0.497 0.508 0.539 0.549 0.617

Version 1.0
Specifications for the HDM-4 road deterioration model for bituminous pavements
K.3.3 Optimal life calibration

A full description of the Optimal Life (OL) calibration is found in Bennett and Greenwood
(1999). The program is used to develop a table of the effects of roughness on the OL for each
representative vehicle. In HDM-4 the standard equation for applying the OL method is:
 100 
LIFEKMPCT = MIN100, a1 
…(K.1)
 1 + EXP(a0 RI )
where:
LIFEKMPCT is the lifetime kilometreage as a percentage

a0 and a1 are regression coefficients
Figure K.5 shows the input screen for the program. The program uses the following data from
the VEHICLES data file:
! Replacement value of the vehicle
! Cost of maintenance labour, and
! Lifetime utilisation (km)
Figure K.5 Optimal life simulation
To run the program it is necessary to supply a base roughness level. This is the roughness at
which the lifetime utilisation applies. For example, if the average roughness on the network is
6 IRI m/km and the average lifetime utilisation is 100,000 km you would supply the value of
6. However, if you estimated that on smooth roads the lifetime utilisation would be 200,000
km and this value is entered in the data file, you would enter 3 IRI m/km for the roughness.
The user selects the vehicles to analyse and the optimal life at different roughnesses is
generated. Varying the parts consumption coefficient kp based on the user-supplied values
and this roughness level does this. Once it has the value of kp that matches the optimal life

Version 1.0
information, it then calculates the life at a roughness of the 3 IRI m/km (which if the base
roughness is 3 equals the lifetime utilisation). It then goes on to calculate the percentage of
life at various roughness levels relative to the life at a roughness of 3 IRI m/km (that is, not
the supplied lifetime utilisation).
Table K.3 is an example of the program output.
Table K.3
Optimal life analysis output
Representative vehicle Roughness (IRI m/km) Optimal life as a percentage

of baseline life
1 3.0 97.1
1 3.5 89.6
1 4.0 83.3
1 4.5 77.9
1 5.0 73.3
1 5.5 69.6
1 6.0 65.8
1 6.5 62.9
1 7.0 60.0
1 7.5 57.5
1 8.0 55.4
1 8.5 53.3
1 9.0 51.3
1 9.5 49.6
1 10.0 47.9
The data from OPTIMAL need to be analysed to quantify the coefficients a0 and a1. This is
done with the non-linear regression program supplied with HDM Tools through the script file
illustrated in Figure K.6.

Version 1.0
Title "Optimal Life Analysis";

Variable RI; // Roughness in IRI m/km
Variable Lifekm; // Optimal Life in km
Parameter Lifekm_a0; // Coefficient a0
Parameter Lifekm_a1; // Coefficient a1
Function Lifekm = min(100,100/(1 + EXP(lifekm_a0 * RI ^ lifekm_a1)));
Plot xlabel="Roughness in IRI m/km",ylabel="Optimal Life as Percentage";
Data;
3.0 97.1
3.5 89.6
4.0 83.3
4.5 77.9
5.0 73.3
5.5 69.6
6.0 65.8
6.5 62.9
7.0 60.0
7.5 57.5
8.0 55.4
8.5 53.3
9.0 51.3
9.5 49.6
10.0 47.9
Figure K.6 Script file for fitting HDM-4 OL equation to raw data
K.3.4 Engine speed simulation

The engine speed simulation program is used to generate a matrix of engine speed versus road
speed. These data are then analysed to establish the following polynomial equation. The
coefficients are supplied to HDM-41.
RPM = a0 + a1 SP + a2 SP 2 + a3 SP 3 …(K.2)
SP = MAX (20, SPEED ) …(K.3)
where:
RPM is the engine speed (revolutions per minute)

SPEED is the road speed (km/h)
Figure K.7 shows the input screen for the program. The user defines the gear ratio data,
which are readily available from manufacturer’s specifications, along with certain other key
attributes. Bennett and Greenwood (1999) discuss these values in detail. Selecting calibrate
the program then applies the simulation logic described in Bennett and Greenwood (1999) to
generate a table of the mean engine speed and mean effective mass ratio as a function of road
speed.
1
NDLI (1995a) proposed a different formulation based on three zones but as shown in Bennett and
Greenwood (1999), the above polynomial formulation gives more than adequate results.

Version 1.0
Table K.4 is an example of this output.
Figure K.7 Engine speed simulation
K.3.5 Wind adjusted aerodynamic drag coefficient

The basis for the wind averaged drag coefficient (CDmult) is described in Bennett and
Greenwood (1999).
CDmult is calculated for angles of χ between 0 and 359 degrees, in intervals of 1 degree.
From the 360 values of yaw angle, the value for CDmult and relative velocity are calculated
for each angle. The average1 of these values was then divided by the square of vehicle speed
to obtain the value for CDmult. The HDM Tools database is then updated with this value,
which in turn can be imported to HDM-4 to update its vehicles database. Figure K.8 shows
the input data screen for this application.
1
The weight of the product at a wind angle of 0 and 180 degrees was halved to account for the full 360
degrees of possible wind angles.

Version 1.0
Table K.4
Engine speed simulation results
Road speed Engine speed Effective Road speed Engine speed Effective mass
(km/h) (RPM) mass ratio (km/h) (RPM) ratio
20 1693 1.719 75 2347 1.114
25 1677 1.433 80 2462 1.111
30 1739 1.331 85 2592 1.109
35 1783 1.271 90 2722 1.108
40 1809 1.219 95 2850 1.108
45 1843 1.183 100 3029 1.108
50 1913 1.160 105 3171 1.108
55 1954 1.143 110 3324 1.107
60 2017 1.130 115 3466 1.107
65 2134 1.124 120 3618 1.107
70 2230 1.117
Figure K.8 HDM Tools CDmult calibration program screen
There are four data items required:

Version 1.0
! Average vehicle speed;
! Average wind speed;
! Critical yaw angle (ψc); and,
! Proportional increase in CD at the critical yaw angle (h).
It is recommended that ψc be assumed to be a constant value of 30o for all vehicle classes.
The vehicle and wind speeds will vary depending on applications. As shown in Table K.5,
the value for h varies between vehicle classes.
Table K.5
Aerodynamic resistance default parameters
1
Vehicle number Type h CDmult CD AF
2
(m )
1 Motorcycle 0.4 1.12 0.70 0.8
2 Small Car 0.4 1.12 0.40 1.8
3 Medium Car 0.4 1.12 0.42 1.9
4 Large Car 0.4 1.12 0.45 2.0
5 Light Delivery Vehicle 0.5 1.16 0.50 2.9
6 Light Goods Vehicle 0.5 1.16 0.50 2.8
7 Four Wheel Drive 0.5 1.16 0.50 2.8
8 Light Truck 0.6 1.19 0.55 4.0
9 Medium Truck 0.6 1.19 0.60 5.0
10 Heavy Truck 0.7 1.22 0.70 8.5
11 Articulated Truck 1.2 1.38 0.80 9.0
12 Mini-bus 0.5 1.16 0.50 2.9
13 Light Bus 0.6 1.19 0.50 4.0
14 Medium Bus 0.7 1.22 0.55 5.0
15 Heavy Bus 0.7 1.22 0.65 6.5
16 Coach 0.7 1.22 0.65 6.5
Notes:
1 Calculated assuming wind speed of 14.4 km/h and vehicle speed of 75 km/h
K.3.6 Work zone simulation model

Work zones are modelled using the program ROADWORK (Bennett and Greenwood, 1999).
This program calculates the additional delay and fuel due to traffic being interrupted for a
work zone.
The program works as follows. The user supplies:

Version 1.0
! A range of values for the work zone capacity, work zone length, work zone speed and
AADT.
! An hourly flow distribution over the 24-hour period.
For each AADT level and the hourly flow distribution, the program generates the arrivals at
the work zone. These arrivals will apply irrespective of the work zone capacity or length.
For each combination of work zone capacity, work zone length and work zone speed the
program will calculate the average delay and queue size, in two directions.
The WORKZONE simulation model is capable of analysing the impact of road closures on
road users. The user inputs the configuration of the lane closure from any of the following:
! Lane closure - typically motorway situations
! Directional closure - two lane road situation
! Complete closure - complete blockage of road
The start-up screen for the software is illustrated in Figure K.9. The first tab is used to
execute the analysis while the other tabs are used for data.
Figure K.9 Screenshot of Start-up screen to the Work Zone simulation model
As described in Bennett and Greenwood (1999), WORKZONE uses simulation to generate

arrival times of vehicles based on a headway distribution model, and then using the available
capacity of the work zone, determines when the vehicle will be able to depart. Based on this
information, and the configuration of the work zone, the model then produces estimates of
travel times and queue lengths.
To use the program, information needs to be supplied on the following items:
! Daily traffic flow, directional split, 24-hour profile

Version 1.0
! Normal lane configuration

! Speed-flow model data
! Type and duration of closure
Figure K.10 shows the traffic flow data. Instead of defining the number of individual vehicles
of each class to simulate, the program works in terms of an aggregate number of vehicles.
Accordingly, it is necessary to provide the average PCSE for each vehicle. This should be
based on the PCSE values for individual vehicles, weighted based on their frequency in the
traffic stream.
Figure K.10 Screenshot of Traffic Flow Information
The user must define speed-flow data, which is based on the HDM speed-flow model
presented in Chapter 6. The hourly flow profiles give the variation in flow over the course of
the day. Both of these data are used to create a unique traffic ID that consists of a volume,
speed-flow profile, hourly flow profile, and a capacity for the site.
As shown in Figure K.11, the user must define the work zone closure type, the length of the
work zone, the maximum speed through the work zone, the work zone capacity, and the
number of lanes in each direction. If the capacity will vary over a day (for example if the road
will be closed for periods of time) this can also be specified.

Version 1.0
Figure K.11 Screenshot of Closure Configuration Screen
Once all the input data have been defined the program is used to calculate the delays. The
output from the program is in a format suitable for performing a regression analysis to
determine various model coefficients for input to the HDM-4 road user effects model or to
apply as an exogenous cost.
K.4 Pavement deterioration and maintenance effects

K.5 Analysis
The non-linear regression program is NLREG. This is a shareware program and you are
expected to register it with the author should you decide to use the program. Registration
details are provided with the software.

Version 1.0
volume five
ISBN: 2-84060-059-5

ISBN: 2-84060-060-9

user interface.
ISBN: 2-84060-061-7

ISBN: 2-84060-062-5

ISBN-2-84060-102-8

ISBN-2-84060-103-6



Capitulo 4

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Capitulo 4

Uploaded by

Copyright:

Available Formats

A Training and Dissemination Strategy for HDM-4

Unpublished Project Report

PROJECT REPORT PR/INT/653/99

A TRAINING AND DISSEMINATION STRATEGY FOR HDM-4

by C C Parkman, J Lowe, A B Sterling, J Rolt and H R Kerali

Prepared for: Asian Development Bank

Copyright Transport Research Laboratory December 1999. All rights reserved.

Transport Research Foundation Group of Companies

Project Manager: Mr C C Parkman

Quality Audit Review Officer: Dr J Rolt

The circulation list for this report is as follows:

1-6. C Melhuish ADB

Draft Report March 1999 TRL 11.3.99

2. OVERVIEW AND BACKGROUND...................................................................... 1

3. OVERALL GOAL AND PURPOSE....................................................................... 3

6. THIS PROJECT ..................................................................................................... 19

• accumulated knowledge and experience developed during the lifetime of

HDM-4 is utilised in the “Planning”, “Programming” and “Preparation” processes

• Senior management and policy level groups (Planning)

2. OVERVIEW AND BACKGROUND

i) identification of target audiences and characterisation of their needs;

1 HDM-4 Training and Dissemination Strategy

2.2 Project workshop in Kuala Lumpur

‘To generate and sustain increasing application of HDM in road management

iii) Outputs were enlarged to

(a) ensure that dissemination to executives and policy makers was

2 HDM-4 Training and Dissemination Strategy

3. OVERALL GOAL AND PURPOSE

3.1 The strategy

• develop a sustainable dissemination and communication system, using

3 HDM-4 Training and Dissemination Strategy

4 HDM-4 Training and Dissemination Strategy

• The motivation of groups and individuals on the ground to adopt HDM-4

• Individuals and groups perceiving HDM-4 as a vehicle for technical and

4.1 Justification for dissemination

Hence we believe a suitable dissemination strategy comes before training, in so far as

5 HDM-4 Training and Dissemination Strategy

4.2 Existing modes of dissemination

6 HDM-4 Training and Dissemination Strategy

4.2.3 Conference papers and presentations and published articles

• Managing Pavements, Durban, May 1998

4.2.4 Website at University of Birmingham, PIARC and roadsource.com

7 HDM-4 Training and Dissemination Strategy

4.2.5 Training courses

4.2.6 Support from University of Birmingham and other institutions

A formalised system will need to be set up once the software is released, as it is

4.3 Options for dissemination

8 HDM-4 Training and Dissemination Strategy

4.3.2 Regional road management working groups

• selection of the most suitable ‘institution(s)’ to organise in-country training;

9 HDM-4 Training and Dissemination Strategy

4.3.3 Workshop sessions

4.3.4 Targeted conference sessions

10 HDM-4 Training and Dissemination Strategy

4.3.5 Published papers and articles

It might be considered worthwhile monitoring the quantity (quality is assumed to be

11 HDM-4 Training and Dissemination Strategy

Potential impact Medium.

4.3.8 Update information with new releases of program

Potential impact High.

4.3.9 Newsletter and brochure

12 HDM-4 Training and Dissemination Strategy

Potential impact High