See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/241685145

Review of EPC Projects Cost Estimation and Minimum Error Technique

Introduction

Article · January 2013

CITATION READS

1 11,074

1 author:

Saeed Rouhani
University of Tehran
74 PUBLICATIONS 1,075 CITATIONS

SEE PROFILE

All content following this page was uploaded by Saeed Rouhani on 15 May 2014.

The user has requested enhancement of the downloaded file.

 International Journal of
Science and Engineering Investigations
Review of EPC Projects Cost Estimation and Minimum
Error Technique Introduction
Kamran Jamali Firoozabadi1, Saeed Rouhani2, Nazanin Bagheri3
123
Industrial Engineering Department , Islamic Azad University of firouzkoh, Tehran , Iran
(1mitkamm@yahoo.com, 2rouhani.saeed@gmail.com, 3nazanin_bagheri@yahoo.com)
Abstract— One of the most influential factors of why a
project becomes successful is estimating the cost. Though
there are bidding and auctions held in my country, Iran, based
on the cost estimation, they end costing an arm and a leg
because of inappropriate cost estimations. To avoid cost
overrun, there should be a cost estimation approach to
minimize the risk, errors associated with techniques and to
provide the most appropriate usages to be applied in all EPC
projects. Here, the techniques measuring these standards and
goals are applied and studied based on the scope of PMBOK
and PRINCE 2 standards. Analogous, bottom-up and
parametric approaches were used to stochastic and
deterministic modeling. Stochastic models were simulated
with Monte Carlo process. The results helped to comparisons
of three approaches with utilizing the qualitative and
quantitative criteria and MCDM and finally best cost
estimation approach is selected.
Keywords- Project Management, Cost Estimation, MCDM,
Minimum Error Approach
I. INTRODUCTION
Cost overrun is a chronic problem across most projects. almost
one third of sampled projects experienced cost overruns of
150–200%, with an average overrun of 189% of the original
cost estimate[i] . This ratio has been 178% for corporations
and large companies, 182% for medium-sized companies and
214% for small companies and firms. Unfortunately, there is
no sufficient data or time and appropriate resources are not
available to estimate the costs. And more surprisingly, when
the cost estimation is correctly measured, the CEO or
managers may order the cost to be reduced without a decrease
in the work scope. And the consequences associated with cost
overrun are often paid by the final consumers.
The inappropriate computation of cost and time results in false
implementation of plans based on irrational estimations;
therefore, these plans are unrealistic and impractical. In
addition, such wrong estimations raise unrealistic expectations
vol. 2, issue 12, January 2013
ISSN: 2251-8843
for stakeholders, and when they are not met, these people
become dissatisfied.
II. REVIEW OF THE PREVIOUSLY CONDUCTED
RESEARCHES INTO THIS FIELDS
A. The applied approaches for practicing the cost estimation
Though there were some interests in measuring the cost prior
to the project before, our main focus is on today, especially
after 2000, when two researchers in England, Fitzgerald and
Akintoye, named the main reasons of being careless in cost
estimations[2].Wallace and Jogensen proposed an algorithm
for simulating the overall cost (2000)[3]. Wang proposed a
model to identify the rational sale of a project using the SIM-
UTILITY method for simulating cost estimation [4].M.
Elkjaer also proposed a stochastic budget simulation approach
[5]. In the U.S., Flybjerg and others understood the costs in
transportation project are estimated considerably low because
of political and psychological aspects (2002)[6]. Nassar and
others used Statistical process control (SPC) (2005)[7]. In
Australia, Pasco and Aibinu found the manufacturing is
mainly influenced by project size in Australia (2008)[8]. Later
in 2010, Hemanta Kumar Doloi devised a meaningful model
SSM to create a standard for comparing the realistic situations
(2010)[9].In 2011, Xiangbai Gu in 'Hierarchy probability cost
analysis model incorporate MAIMS principle for EPC'
introduced HICP model, regarding to MAIMS principle[10].
Meanwhile, since the deterministic value method does not
compute the risks in a project efficiently, analyzing the risks
has gained a remarkable amount of interest. In 2011, Jui-
Sheng Chou applied Monte Carlo simulation to estimate the
costs[11].
American National Standards Institute(ANSI), as an
international and remarkable standard, propose impeccable
hints to calculate the cost during running a project, which
deals with the project like a live entity.
B. Practice standard for project estimating
Principles underlying the life cycle stages of project
estimating include prepare to estimate, create estimates,
1
manage estimates and improving estimating process[12].
Baseline is an approved plan for a project, plus or minus
approved changes. It is compared to the actual performance to
determine if performance is within acceptable variance
thresholds[13].
The stages of project estimating can be seen in the context of
the project life cycle structure. Figure 1 shows that there is
some overlap of project estimating stages within the project
life cycle[14]. The process considered in this research
specified separately by blue color.
Starting Carryin Closing
Organizin
the g and g out the the
Project Preparing Work Project
Prepare to Estimate
Create Estimates
Manage Estimates
Improve Estimating Process
Figure 1.Relationship of the Estimating Life Cycle[15]
While the estimating life cycle stages are the same for all
types of projects, there are some differences in estimation
metrics and models that are employed across industries. These
variations are primarily related to the nature of the project's
product[16].
C. PRINCE2 INTRODUCTION
PRINCE2 focuses on the products to be produced by a project,
not the processes to produce them[17].
PRINCE2 gives a pragmatic tool through which one can put
the flesh onto the introduction of PRINCE2. The perspectives
PRINCE2 deal with are the PRINCE2 process model, the
PRINCE2 components, the PRINCE2 principles and the
PRINCE2 documents[18].PRINCE2 offers only a few
techniques. You may already gave a technique that is
satisfactorily covering that need. The exception is the product-
based planning technique[19]. Because of this attitude of
PRINCE2, the techniques introduced in it are not considered
in this paper.
D. Getting to know Pertmaster[20]
Pertmaster allows definition of both a riskless program and
a program with different aspects of risks. Moreover, it can be
applied different tools, like Monte Carlo simulation. Using the
project risk analysis allows the possibility of accomplishing a
project in a certain time and with a certain budget to be
measured. A program only shows one plausible way for
accomplishing the project without risk analysis; however,
using the risk analysis allows us to evaluate different outputs
and ways of accomplishing a project. It should also be noted
the most practical distribution for simulating time and cost of
an activity is the triangle distribution.
International Journal of Science and Engineering Investigations, Volume 2, Issue 12, January 2013
www.IJSEI.com ISSN: 2251-8843
E. Mean Squared Error, the least flawed approach[21]
MSE is a risk function and measures the mean square of
errors. The mean squared error (MSE) of an estimator is one
of many ways to quantify the difference between values
implied by an estimator and the true values of the quantity
being estimated. MSE is the second moment of error and both
estimated variance and diagonal variance surround it. The
lowest square root or root-mean-square deviation is the
typically used amount for measuring the defined deviations
above.
MSE and RMSE for the estimator which estimated the
parameter are defined as follow:
= E [( − ) ]
= = E [( − ) ]
III. EXPLAINING THE STUDIED PROJECT AND SIMULATION
ANALYSIS
EPC project of an electrical substation in different voltages
of 400KV, 360KV, 230KV and 132KV and extension of
66KV is created. Electrical substations are created to
transform the high voltage flowing through the lines into low
voltage used in the industries. The finally transformed voltage
cannot be used for domestic consumers and another
transformation is required to make it useable for them. These
substations are divided into outdoors and indoors.
Here in this paper, 20 electrical substations were studied
under the same circumstances. By 'under the same
circumstances', we mean the designs, equipment supplies,
buildings and installing the equipment were the same with the
same time intervals, resulting in planning and doing the
projects with the same costs. In the implementation phase, the
actual costs of doing a project differs because of the different
reasons and situations which are not under control of the
organization. Here, we applied the mean distribution of the
actual costs in measuring the MSE.
A. The primary data of the project for preparing the
estimation
At first, the primary data of the project should be obtained
to use the aforementioned techniques. The entries for
practicing the estimations win this project were gathered from
the experts and project managers. It should firmly be noted
that all the following data, assumptions and stated points were
entered into the software the same, because the primary data
should be the same for all points to compare the data (because
of having more homogeneous and valid outcomes of
comparing different techniques). Moreover, the WBS level
was chosen to highlight the limits of estimation, regarding to
Xiangbai Gu research in 2011.
1. The scheduling of activities were considered as same
as the similarly conducted project.
2. the links between activities considering predecessors
and successors are demonstrated via FS (Finish to
Start) ،SS(Start to Start ) ،FF (Finish to Finish).Then,
press short F 9 key to create the primary schedule
structure.
3. The required sources of the projects in the
engineering sector (E) include the expert work force
and the designing work force were predicted similar
2
Paper ID: 1213-06
 to the previously conducted projects. In the
procurement sector (P), expert human resource and
the budget resource were predicted. In the
construction sector (C), since this sector is done
outsource, the mateDollars and machines were
considered as limitless.
4. The risks in this project was considered as
follow:
• the time risk for all activities was considered.
• the resource risk for the activities in the procurement
and construction sectors was considered, because
there is need to use a resource except the expert
human resource. Therefore, this risk is not considered
for the engineering sector, as there are sufficient
numbers of human resource.NB. In this project, a
resource as the cost was assigned to the activities and
the cost risk was specified to this resource; meaning,
the cost risk was fully considered.
• The contingencies relevant to the physical situation
and local conditions of the project are influential;
thus, a coefficient in the costs as contingency reserve
was considered.
5. In this project, it was assumed that the resource
calendar was definite and limitless, and all the
resources were available due to the working
calendar.
6. It was assumed that there are all the quality
documents, like ISO1000, ISO9001and HSE, and
organization properties, such as processes,
instructions and procedures which have to be done so
that the organization process becomes assuring,
regarding to this organization work.
7. The correlation between the activities (risks) can be
an amount from -100% to +100%. When the
prolonging of one activity affects the other one and
makes the activity longer, the correlation between
them is considered as positive. By adding the
correlation to the project avoids planning in an
unrealistic condition significantly. The correlation
can be computed by estimation. At any situation, this
amount has been measured between correlated
activities in a team consisting of professional experts.
When the correlation coefficient is identified in the
software, it calculates the correlation amount relying
on Pearson and Spearman correlation.
8. The assumptions in this project are as followed:
• The activities are definite and probabilistic or risky
activities are not considered.
• The links (the predecessors and successors) are not
probabilistic, but definite.
• The calendars are definite and there is no probable
calendar.
To create the model of the estimation, it is decided to
make the introduced work breakdown structure
thresholds, in which the estimation is calculated via its
thresholds transparently. Considering all the
aforementioned principles and assumptions, the models
were created in Pertmaster software. After receiving the
entry data for practicing the cost estimation, it is
necessary to clarify the estimation approach relying on the
available information. Since it is possible that different
International Journal of Science and Engineering Investigations, Volume 2, Issue 12, January 2013
www.IJSEI.com ISSN: 2251-8843
data is required to use different approaches, which may
not be available in this phase, clarifying the estimation
approach limits the approaches which can be chosen.
Here, we tried to use all the mentioned approaches in
PRINCE2 standard and Practice standard for project
estimating in PMI, elaborated sufficiently in the theatrical
principles. And among them, the one which shows the most
appropriate and accurate results for estimation was chosen and
we hope it will be applied in the future estimations.
B. Practicing estimation approach
In this process, the following steps should be done relying on
accessing to the previously identified entries
• Define the estimation threshold for the project.
• Consider the points which should be updated during
accomplishing the project.
• Plan the contingency reserves.
Based on the following proposed theoretical principles,
different models for practicing the estimation in different
projects were defined. Some models are only effective in one
project, as any project cannot be modeled via them. Some
models are not general enough to cover all the considered
points in the cost estimation. Some models are merely used for
the definite estimation practice, while others are not accurate
enough to be applied in indefinite estimation practice. In this
paper, we tried that the chosen model to be a general approach
as much as possible; meaning, it can be modeled in all
projects. Moreover, the chosen approach was tried to be
accurate and in the estimation management process can be
controlled after running the project. Hence, the three common
approaches for estimating including all standards were used,
after we had studied and evaluated different models. They are
Analogous, bottom-up and parametric cost estimating
approaches. To practice the estimation via applying each
approach, the practice confine is different.
C. Contingency Reserve analysis
Contingency reserves are estimated costs to be used at the
discretion of the project manager to deal with anticipated, but
not certain, events. These events are "known unknowns" and
are part of the project scope and cost baselines. This type of
contingency reserve is defined based on the chosen approach22
In the analogous and bottom-up approaches, all the known
unknowns are defined in the Pertmaster as risks. But in the
parametric approach, the contingency reserve is considered as
3% of the total estimated cost of the project due to its definite
estimation. This number was defined in PMO.
There is another type of contingency reserve, called as
'management contingency reserves'. They are budgeted
reserved for unplanned, but potentially required, changes to
project scope and cost. These are "unknown unknowns" and
the project manager must obtain approval before obligating or
spending this reserve. Management contingency reserves are
not part of the project cost baseline, but included in the budget
for the project[23]. Here, the management contingency reserve
for all techniques was defined.
3
Paper ID: 1213-06
D. Analogous approach probability of 50%? 45,448,725,024 Dollars
What is the total cost of finishing a project at the
In this method, a combination of three points estimation and probability of 80%? 46,723,737,699 Dollars
coefficient estimation was used. In this technique, for all Tarbiat 132-20KV GIS Substation
Entire Plan : Cost
activities of third level of WBS, the costs, taken from the 100% 50,510,868,824

similar project, were considered as most likely costs of the 95% 47,947,012,925
90% 47,409,768,334
project with consideration of 30% inflation ratio. Then, 10000
85% 47,041,028,492

professional experts were asked to estimate the lower and 80% 46,723,737,699

75% 46,458,534,264
higher costs than the most likely ones as 'optimistic costs' and 8000 70% 46,211,278,271

'pessimistic costs'. Moreover, the most likely costs were 65% 45,990,106,066

Cumulative Frequency
60% 45,784,819,338
balanced by these experts. 55% 45,591,212,261
6000

Hits
In this project, some activities did not have the previous 50% 45,393,973,134
45% 45,202,550,916
documentations; therefore, coefficient was applied to estimate 40% 45,012,131,635

them. In the coefficient estimation different coefficients, like 4000 35% 44,816,515,566
30% 44,616,316,574
local coefficients, and other coefficients in the price list were 25% 44,401,634,242

used[24]. When analogous approach was applied, because of 2000

20% 44,168,436,503

15% 43,898,628,421
the estimation practice of likelihood, the contingency reserve 10% 43,576,943,460

for all of them was considered. 5% 43,132,411,410

0 0% 40,975,600,326
To practice the estimation using the analogous technique, the 45,000,000,000 50,000,000,000

third level of WBS in the Pertmaster was defined. This is Distribution (start of interval)

because the ability of Pertmaster to analyze the time and cost

risk simultaneously. Using the project risk analysis, the Figure 2: Cost distribution using analogous estimation
possibility of finishing project is defined at the determined
time and cost. To apply analogous, at first, we entered the Now, by looking at the graph, we can identify the probability
most likely cost into the software and applied all the available in which the project with the primary estimated cost was
risks, based on what defined in 3-1-4.The triangle distribution finished without considering the cost risk. This cost in the
was used for entering the time and cost risk in the Pertmaster, second figure is 44, 700, 000, 000 Dollars. It can be implied
which is recommended as the most applicable distribution for the project will finish with the probability of 32%, a low
all stochastic simulations. probability, and the primary definite estimation, without
The analysis iterations must be at least 76,833, regarding to an considering the contingencies, if we look at the yellow line.
article Jui-Sheng Chou by (2011). Hence, the iterations in this The another conclusion reached by the carried out analysis is
project were set at 80,000 and then the simulation was run. the project with the probability of 0% is the project cannot be
The data provided by the software in terms of cost project ended at any cost lower than 40,975,600,326 Dollars.
simulation based on the analogous technique are shown in
table 1 E. Bottom-up approach
It allows to estimate the consumption amount of each source
Cost of Entire Plan in the project details, and then these estimations are added
together to measure the middle levels and all levels of the
Analysis project, which produces a clear and structure-based estimation.
Iterations: 80000 The requirements of this approach are list of the works in
which activities are defined and detailed specifications of the
Statistics project.
Minimum: 40,975,600,326 In the bottom-up approach, the three points estimation or
Maximum: 50,510,868,824 PERT can be used instead of single point estimation at both
Mean: 45,448,725,024 levels of activity and work package. Consequently, after
adding the estimations at project level, a more accurate and
Bar Width: 500,000,000 clear picture of all the required works and costs for the project
Highlighters is computed, helping to have a more accurate control[25].
Deterministic 32% To apply this method, at first, WBS of the project was entered
(44,700,000,000) in the Pertmaster to the level 4. The times and links based on
the project management team's view and experience are
50% 45,393,973,134
defined and entered in the software. Then, the identified
80% 46,723,737,699 sources based on 3-1-3 are coordinated with the activities and
the intended risks at different sectors are entered. Finally, the
Table 1: Total cost using analogous estimation correlations among different risk in the software are defined as
follow:
The average costs required for finishing the project, or the
• The time risk of engineering activities is correlated
cost which all activities with all of their probabilities are done,
is 45,448,725,024 Dollars. The followings are two possibilities with their cost risk.
which are the most important ones in statistics analysis, • The time risk of provision activities is correlated with
regarding to the software suggestion: their cost risk.
• The time risk of manufacturing activities of the
What is the total cost of finishing a project at the

International Journal of Science and Engineering Investigations, Volume 2, Issue 12, January 2013 4
www.IJSEI.com ISSN: 2251-8843 Paper ID: 1213-06
 products made from subproducts are correlated with
the time risk of producing them
• The time risk of construction engineering is
correlated with its performance time risk.
• The time risk of fundamental executive constructing
activities is correlated with the time risk of installing
equipment.
When the bottom-up approach was applied, due to the
practicing statistical estimation, the contingency reserve
for all contingencies was considered and there is no need
to recalculate it. The results of simulation are shown in
the table 2. In terms of risk analysis, the cost distribution
was shown as follow:
Cost of Entire Plan
Analysis
Iterations: 80000
Statistics
Minimum: 44,018,311,081
Maximum: 47,868,584,497
Mean: 45,910,291,597
Bar Width: 250,000,000
Highlighters
Deterministic 49%
(45,889,126,218)
50% 45,909,345,686
80% 46,378,649,779
The average costs required for finishing the project, or the cost
which all activities with all of their probabilities are done, is
45,910,291,597 Dollars. The followings are two possibilities
which are the most important ones in statistics analysis,
regarding to the software suggestion:
What is the total cost of finishing a project at the
probability of 50%? 45,909,345,686 Dollars
What is the total cost of finishing a project at the
probability of 80%? 46,378,649,779 Dollars
Now, by looking at the graph, we can identify the probability
in which the project with the primary estimated cost was
finished without considering the cost risk (figure 3) . This cost
in this graph is 45,889,126,218 Dollars. In this graph, the
probability is precisely shown 48% in the left side, which
implies the project will finish with the probability of 48% and
the primary definite estimation, without considering the
contingencies, if we look at the yellow line.
The another conclusion reached by the carried out analysis is
the project with the probability of 0% is the project cannot be
ended at any cost lower than 44,018,311,081 Dollars.
F. Parametric approach
Using this approach allows to estimate the details of fourth
and fifth level of WBS.
The accuracy of this approach is highly valid and reliable,
when the following criteria are met:
International Journal of Science and Engineering Investigations, Volume 2, Issue 12, January 2013
www.IJSEI.com ISSN: 2251-8843
The previous data used in the prepared model are
accurate.
The applied parameters in the model have the ability
of being quantitative.
The prepared model is capable of being measured in
order to be applied for project with lower scale, as it
can be applied for projects with larger scale.
In parametric approach, a definite model is devised by using
the links between previous data and other variables for
estimating the cost of an activity. The cost of activities can be
computed either by multiplying the required work amount by
the time length or by its cost in the previous projects. The
required work amount in this project is defined as follow:
• The required work amount in the engineering sector
equals the hours a person is designed. For each
designing sector, a designing expert and drawer is
needed.
• The required work amount in the provision sector
equals the amount of goods which should be bought
along with man-hour needed for it. The required
work amount in the construction sector was
forwarded by the contractor.
• After practicing all estimations relying on the work,
the overall estimated cost for the project is
45,753,976,647 Dollars.
G. Identifying the best approach for estimating the project
cost (EPC)
Three processes, create schedule model, prepare to estimate
and create estimates were applied so far. In three approaches,
analogous, bottom-up and parametric, were applied for
practicing the EPC of electrical substations. To find out all
results of each approach, we should know which one had the
lowest error. As mentioned previously, mean squared error
(MSE) is the best approach to do so; therefore, by applying
this in the research, the following RMSD for all estimators
are shown in the table 3.
Table 2: mean squared error
RMSD(Analogous) 942,681,976$
RMSD(Parametric) 637,430,353$
RMSD(bottom-up) 481,115,40$
According to the comparison between the estimated costs
amounts via different approaches and the actual cost of the 20
previous projects, it can be concluded that which approach
was the most accurate one. Moreover, other factors for an
appropriate estimator can be defined except the deviation,
which are as follow:
1. the Root-Mean-Square Deviation
2. Being user friendly
2.1 the approach should not be sophisticated.
2.2 the size of entries should be appropriate or not
too much.
5
Paper ID: 1213-06
 3. the approach should be controllable; meaning relying IV. CONCLUSION
on the approach, a baseline for identifying the budget and
the earned value management can be drawn A. results
4. cover the majority of known risks. In this article, the possible approaches were simulated using
5. does not depend on the experts in each field. In other the bottom-up and analogous approaches. They were
words, it should be practical without an expert in that simulated for 80,000 times in the Monte Carlo simulator
field, which decreases the costs significantly. inside the Pertmaster software, and their probable results were
6. does not depend on the previous data. This is because achieved. In addition, the definite estimation using parametric
the approach can be applied in the project which have not approach was carried out and the definite estimation of that
been done yet. was computed. In another step, the deviation of all estimators
Since the estimator team was consisted of professional from the realistic project information using the root-mean-
experts in different fields, AHP group decision making square deviation was computed. The results show the bottom-
technique was used for identifying the most tailored up cost estimation approach had lowest deviation comparing
approach for the estimation in the similar project in the to the actual data. Moreover, by considering the other factors
future. besides RMSD in the AHP group decision making approach,
To apply the AHP technique, modeling, data collection, the most appropriate approach for estimating the costs in EPC
processing and measuring the inconsistency rate should project was bottom-up cost estimation approach. This
be carried out. After drawing the decision making tree approach in the weighted deviation was also the best. It should
and collecting data, the coefficients and priorities of each also be taken into account that parametric approach in
attribute and alternative are computed. Then, the comparison with others was the best approach for managing
alternative earning the highest weight is considered as the the earned value in terms of controllability. However, it was
best. The alternatives considered are: weaker in terms of risk coverage standard and MSE than the
1. Analogous approach bottom-up.
2. Parametric approach
B. Suggestions
3. Bottom-up approach
First, it is worthy to find out the most appropriate approach
4. Table 3: Prioritized weighted attributes for managing estimation, because being unsure means a limit
for the baseline of project budget. Hence, estimation
Priority Weight Attributes management is obscure in terms of the probable estimations.
Next, as one of the major problems of project-oriented
1 0.285 .Root-Mean-Square Deviation
organizations is measuring the being thrifty in spending, the
2 0.270 Cover More risks actual cost of each sector and the baseline can be computed by
3 0.234 Being controllable estimating the costs in each sector of the project by the expert
4 0.097 Not depend on Experts team of the same sector. Finally, measuring the deviation by
5 0.067 Being user friendly the expert in each sector allows the calculation of the
5.1 0.750 Not be complicated expenditure and the cost conservation in each sector.
5.2 0.250 Not need much data entry
6 0.047 Not depend on previous data ACKNOWLEDGMENT
As shown in the table 4, RMSD earned the highest weight. The authors wish to thank an anonymous referee for
This standard presents the accuracy of the estimator, or in comments and suggestions that greatly improved the
other words, the estimation approach. In addition, the presentation of the paper.
consistency rate is 0.05, showing that the decision making is
rational and realistic.
Since the consistency rate of views of each DM was lower REFERENCES
1
than 0.1, it is clear the decision makers did not have a notable [ ]Hemanta Kumar Doloi,2011," Understanding stakeholders'
disagreement. In the next phase, each alternative with the perspective of cost estimation in project
above standards was examined, and finally, the weight that management",International Journal of Project Management 29
each alternative gained was shown in the table 5. (2011) 622–636
٢
[ ] Akintoye,A.,Fitzgerald,E.,2000,Asurvey of current cost
Table 4: Prioritized weighted alternatives estimating practices in the UK.,Construction Management
and Economics18,161-172.
٣
[ ] Trond Jùrgensen *, Stein W. Wallace,2000,"Improving
Priority Weight Alternative project cost estimation by taking into account managerial
1 0.411 Analogous approach flexibility ", Department of Industrial Economics and
2 0.392 Parametric approach Technology Management, Norwegian University of Science
3 0.197 Bottom-up approach and Technology, 7491 Trondheim, Norway, European Journal
of Operational Research 127,239-251
As shown in table 5, the approach earned the highest weight [4 ] Wang, W.-C., 2002. SIM-UTILITY: model for project
was the bottom-up approach. ceiling price determination.Journal of Construction
Engineering Management 128, 76–84.

International Journal of Science and Engineering Investigations, Volume 2, Issue 12, January 2013 6
www.IJSEI.com ISSN: 2251-8843 Paper ID: 1213-06
5 12
[ ]Elkjaer, M., 2000. Stochastic budget simulation. [ ]Project management institute, 2011,practice standard for
International Journal of Project Management 18, 139–147. project estimating,2
6 13
[ ] Felyvbjerg, B., Holm., M.S., Buhl,S.,2002, Undrestanding [ ]Project management institute, 2011,practice standard for
costs in public works projects , Journal of the American project estimating,4
14
Planning Associa6on 68 (3) Summer. [ ] Project management institute, 2011,practice standard for
7
[ ] Nassar, K.M.,Nassar, W.M., 2005,Evalua6ng cost overruns project estimating,11
15
of asphalt paving project using statistical process control [ ]Project management institute, 2011,practice standard for
methods, Journal of Construction Engineering and project estimating,12
16
Management 131,1173-1178 NOVEMBER [ ] Project management institute, 2011,practice standard for
8 project estimating,15
[ ] Aibinu,A.A.,Pasco, T., 2008,The accuracy of pre-tender 17
building cost estimates in Australia .Construction [ ]Colin Bentley,PRINCE2TM revealed,2006, Elsevier
management and Economics 26, 1257-1269. Butterworth-Heinemann,1
9
[ ] Hemanta Kumar Doloi,2011, Understanding stakeholders' [18]Henny Portman. Prince2 in practice, 2009, Van Haren
perspective of cost estimation in project management, Publishing, Zaltbomel,2
19
International journal of project management 29, 622-638 . [ ]Colin Bentley,PRINCE2TM revealed,2006, Elsevier
[10 ] Xiangbai Gu,2011, Hierarchy probability cost analysis Butterworth-Heinemann,2
20
model incorporate MAIMS principle for EPC [ ]Radmehr Hossein, Project Management Risk (Pertmaster),
project cost estimation, Expert Systems with Applications, Hami Publica6on,54-56,97,222
21
8087–8098 [ ]http://en.wikipedia.org/wiki/Mean_squared_error
11 22
[ ] Jui-Sheng Chou,2011," Cost simulation in an item-based [ ]Project management institute ,Pmbok,166
23
project involving construction engineering and management", [ ]Project management institute , Pmbok,169
International Journal of Project Management 29 - 706–717. [24 ]Milanizade Alireza, Metering Engineering, Milan Afzar
publication, 140-214

International Journal of Science and Engineering Investigations, Volume 2, Issue 12, January 2013 7
www.IJSEI.com ISSN: 2251-8843 Paper ID: 1213-06

View publication stats