AGRD06-20 Guide To Road Design Part 6 Roadside Design Safety and BarriersEd3.1

Guide to Road Design Part 6: Roadside Design,
Safety and Barriers
Sydney 2020
Guide to Road Design Part 6: Roadside Design, Safety and Barriers
Third edition prepared by: Professor Rod Troutbeck Publisher
Third edition project managers: Peter Ellis and David Bobbermen Austroads Ltd.
assisted by Richard Fanning Level 9, 287 Elizabeth Street
Sydney NSW 2000 Australia
Abstract
Phone: +61 2 8265 3300
The Guide to Road Design Part 6: Roadside Design, Safety and Barriers
provides guidance on roadside design and in particular guidance on austroads@austroads.com.au
evaluating the risk of a roadside and the selection and use of road safety www.austroads.com.au
barrier systems.
About Austroads
Roadsides have to accommodate many features that support the road and
the safe and efficient operation of traffic, and have to be designed with Austroads is the peak organisation of Australasian
regard to environmental requirements. Part 6 should therefore be read in road transport and traffic agencies.
conjunction with other Parts of the Guide to Road Design.
Austroads’ purpose is to support our member
Part 6 provides information to enable designers to understand the organisations to deliver an improved Australasian
principles that lead to the design of safer roads, identify hazards, road transport network. To succeed in this task, we
undertake a risk assessment process of roadside hazards, establish the undertake leading-edge road and transport research
need for treatment of hazards and determine the most appropriate which underpins our input to policy development and
treatment. Methods of evaluating the effectiveness of treatment options published guidance on the design, construction and
are summarised. A comprehensive design process, guidance and design management of the road network and its associated
considerations are provided for the selection of a suitable barrier and for infrastructure.
the lateral and longitudinal placement of barrier systems.
Austroads provides a collective approach that
delivers value for money, encourages shared
Keywords knowledge and drives consistency for road users.
Roadside design, designing for safety, hazard, hazard identification, Austroads is governed by a Board consisting of
hazard mitigation, risk assessment, treatment options, evaluation, road senior executive representatives from each of its
safety barrier systems, road safety barrier design process, deflection eleven member organisations:
width, medians, containment level, working width, run-out length, length of
need, points of redirection, terminal treatments, transitions, vulnerable • Transport for New South Wales
road users, steep downgrades, arrester beds, barriers at intersections.
• Department of Transport Victoria
• Queensland Department of Transport and Main
Edition 3.1 published November 2020
Roads
This edition corrects the risk score formula for rollover events in Tables • Main Roads Western Australia
B.11, B.12, B.14, B.15, B.16 and B.18. • Department of Planning, Transport and
Infrastructure South Australia
Edition 3.0 published August 2020
• Department of State Growth Tasmania
This third edition contains changes throughout Sections 1, 2 and 3. The
major change has been to the risk assessment process including the • Department of Infrastructure, Planning and
Network Risk considerations in Section 2. Other changes include: Logistics Northern Territory
• the use of graphical techniques to evaluate roadside risk, including the • Transport Canberra and City Services Directorate,
Network Roadside Risk Intervention threshold (NRRIT) Australian Capital Territory
• the use of jurisdictional policies and corridor safety visions to define • Commonwealth Department of Infrastructure,
appropriate roadside treatments Transport, Regional Development and
• the recognition that an agency may use and alternative network-wide Communications
risk assessment methodology to identify where to treat the risk • Australian Local Government Association
• Sections 4 to 6 have had minor changes and it is expected that
significant changes will be made to these sections in the next edition • New Zealand Transport Agency.
• a new section on Work Zone Safety Barrier Systems is included in
Section 7 but will be reviewed in the next edition
• new appendices on the risk assessment process have been added.
Edition 2.1 published August 2018 [format update only]
Edition 2.0 published October 2010
Edition1.0 published November 2009
ISBN 978-1-922382-44-3 © Austroads Ltd 2020 | This work is copyright. Apart

from any use as permitted under the Copyright Act
Austroads Project No. TP2056 Pages 235
1968, no part may be reproduced by any process
Austroads Publication No. AGRD06-20 without the prior written permission of Austroads.
This Guide is produced by Austroads as a general guide only. Austroads has taken care to ensure that this publication is correct at
the time of publication. Austroads does not make any representations or warrant that the Guide is free from error, is current, or,
where used, will ensure compliance with any legislative, regulatory or general law requirements. Austroads expressly disclaims all
and any guarantees, undertakings and warranties, expressed or implied, and is not liable, including for negligence, for any loss
(incidental or consequential), injury, damage or any other consequences arising directly or indirectly from the use of this Guide.
Where third party information is contained in this Guide, it is included with the consent of the third party and in good faith. It does
not necessarily reflect the considered views of Austroads Readers should rely on their own skill, care and judgement to apply the
information contained in this Guide and seek professional advice regarding their particular issues.
Acknowledgements
The author acknowledges the role and contribution of the Austroads Road Design Task Force in providing guidance and
information during the preparation of the third edition of this Part. The task force comprised the following members:
Mr Peter Ellis Transport for New South Wales
Mr Richard Fanning Department of Transport, Victoria
Mr Mike Whitehead/Mr Bernard Worthington Queensland Department of Transport and Main Roads
Mr Albert Wong Main Roads Western Australia
Mr Ben McHugh/Mr Ken Marshall Transport Canberra and City Services Directorate
Mr Edi Winkler Department of Planning, Transport and Infrastructure South Australia
Mr Sam Hatzivalsamis Department of Infrastructure, Planning and Logistics Northern Territory
Mr Tony Napoli/Mr Geoff Armstrong/Mr Michael Hogan Australian Local Government Association
Mr James Hughes NZ Transport Agency
Mr Andrew Baker Consult Australia
Mr Michael Tziotis Australian Road Research Board
David Bobbermen Austroads Safety Program Manager
It is also acknowledged that much of the information for Section 7 – Work Zone Safety Barrier Systems has been provided by Main
Roads Western Australia.
The Austroads Traffic Management Working Group, Austroads Safety Barrier Assessment Panel and the Roads and Roadsides
Theme Group are also acknowledged for their involvement in consultation activities.
Edition 2.0 prepared by Gary Veith and project managed by Graeme Nichols.
Contents
1. Introduction to Roadside Design .......................................................................................................... 1

1.1 Context Sensitive Designs ....................................................................................................................... 1
1.2 Purpose .................................................................................................................................................... 2
1.3 Reading this Part in the Context of Part 1 ................................................................................................ 3
1.3.1 Combining Design Parameters and Consistent Design .......................................................................... 4
1.3.2 Progressive Safety Updates to the Guide to Road Design...................................................................... 4
1.4 Scope of this Part ..................................................................................................................................... 5
1.4.1 AGRD Part 6A: Paths for Walking and Cycling ....................................................................................... 6
1.4.2 AGRD Part 6B: Roadside Environment................................................................................................... 6
1.5 Principles Considered in Roadside Design to Achieve the Safest System .............................................. 6
1.5.1 Safe System Principle ............................................................................................................................. 7
1.5.2 Road Network Efficiency Principle .......................................................................................................... 7
1.5.3 Community Wellbeing Principle .............................................................................................................. 8
1.5.4 Environmental Sustainability Principle .................................................................................................... 8
1.5.5 Utility Services Principle .......................................................................................................................... 8
1.5.6 Investment Benefit Principle.................................................................................................................... 8
1.5.7 Safety in Design Principle ....................................................................................................................... 8
1.6 Roadside Safety Design ........................................................................................................................... 8
1.6.1 General ................................................................................................................................................... 8
1.6.2 Road Environments that Promote Safer Travel....................................................................................... 9
1.6.3 Design for Risk Reduction....................................................................................................................... 9
1.6.4 Appropriate Barriers and Other Treatments .......................................................................................... 10
1.7 Terminology ............................................................................................................................................ 10
1.8 Overview of the Roadside Risk Assessment Process ........................................................................... 10
1.8.1 Network Risk Assessment .................................................................................................................... 10
1.8.2 Assessment of the Road Segment against National Practices, Jurisdictional Policies and the
Corridor Safety Vision ........................................................................................................................... 11
1.8.3 Road Program or Project Risk Assessment .......................................................................................... 11
1.9 Calculating a Risk Score ........................................................................................................................ 12
1.9.1 Overview of the Simplified Process for Calculating the Risk Score for a Roadside Cross-section........ 13
1.9.2 Description of Significant Hazards ........................................................................................................ 13
1.9.3 Description of Less Significant Hazards ................................................................................................ 14
1.9.4 Measuring the Lateral Distance to a Hazard ......................................................................................... 15
1.9.5 Operating and Design Speed ................................................................................................................ 16
1.9.6 Terrain Type .......................................................................................................................................... 16
1.9.7 Background and Isolated Hazards ........................................................................................................ 17
1.9.8 Risk Score Charts for Undivided Rural Roads ...................................................................................... 18
1.9.9 Grade and Curve Radii Factors for Rural Roads Risk Score Values .................................................... 22
1.9.10 Risk Score Charts for Divided Urban Roads ......................................................................................... 23
1.9.11 Risk Score Charts for Roadside Safety Barriers ................................................................................... 24
1.9.12 When should the process in Appendix B be used? ............................................................................... 26
1.9.13 Hazards for Motorcyclists and Other Vulnerable Road Users ............................................................... 26
1.9.14 Hazards for Heavy Vehicle Occupants.................................................................................................. 27
2. Network Risk Assessment .................................................................................................................. 28

2.1 General ................................................................................................................................................... 28
2.2 Corridor Safety Visions ........................................................................................................................... 28
2.2.1 Adverse Crash History .......................................................................................................................... 28
2.3 Treatment of Roads Based on Policies and Practices ........................................................................... 29
2.3.1 Treatment of Roads Based on National Practices................................................................................. 29
2.3.2 Treatment of Roads Based on Jurisdictional Policies ........................................................................... 29
2.4 The Network Roadside Risk Intervention Threshold (NRRIT) ............................................................... 30
2.4.1 Application to Greenfield and Brownfield Sites ..................................................................................... 30
2.4.2 Establishing the NRRIT ......................................................................................................................... 30
Austroads 2020 | page i

2.4.3 Setting a NRRIT Based on Two-lane Undivided Rural Roads .............................................................. 31

2.4.4 Setting a NRRIT Based on Urban Roads .............................................................................................. 31
2.5 Example of Setting a NRRIT .................................................................................................................. 32
3. Program and Project Risk Assessment ............................................................................................. 35

3.1 Overview of the Risk Evaluation Process .............................................................................................. 35
3.2 Concepts Used in Evaluating the Risk at Particular Sites ...................................................................... 35
3.3 Step 1: Assess Against National Practices, Jurisdictional Policies and Corridor Visions ......................37
3.4 Step 2: Compare the Risk Score with the NRRIT .................................................................................. 37
3.4.1 Adverse Crash History .......................................................................................................................... 38
3.4.2 Examples of the Use of the Procedure.................................................................................................. 38
3.5 Step 3: Identify, Evaluate and Rank Risk Mitigation Options ................................................................. 46
3.5.1 Identify the Options ............................................................................................................................... 46
3.5.2 Evaluate the Risk Associated with a Roadside Treatment Using a Qualitative Assessment................. 48
3.5.3 Evaluate the Risk Associated with a Roadside Treatment Using a Quantitative Assessment .............. 48
3.5.4 Rank Treatment Options ....................................................................................................................... 49
3.6 Step 4: Design the Recommended Roadside Treatments ..................................................................... 49
4. Treatment Options................................................................................................................................ 50
4.1 General ................................................................................................................................................... 50
4.2 Summary of Treatment Options ............................................................................................................. 50
4.3 Effectiveness of Treatment Options ....................................................................................................... 51
4.4 Types of Treatments............................................................................................................................... 52
4.4.1 Treatments for Trees............................................................................................................................. 52
4.4.2 Treatments for Steep Downgrades ....................................................................................................... 52
4.4.3 Treatments for Medians ........................................................................................................................ 52
4.4.4 Treatments for Drains ........................................................................................................................... 53
4.4.5 Treatments for Drainage Features ........................................................................................................ 54
4.4.6 Treatments for Rock Face Cuttings....................................................................................................... 56
4.4.7 Treatments for Minor Roadside Hazards .............................................................................................. 56
4.4.8 Treatments for Road Furniture .............................................................................................................. 56
4.4.9 Treatments for Poles ............................................................................................................................. 57
4.4.10 Treatments at Active Controlled Level Crossings ................................................................................. 60
4.4.11 Weather Warning Systems ................................................................................................................... 60
5. Road Safety Barriers ............................................................................................................................ 61

5.1 Introduction ............................................................................................................................................. 61
5.1.1 General ................................................................................................................................................. 61
5.1.2 General Requirements for Road Safety Barrier Systems ...................................................................... 61
5.1.3 Road Safety Barriers for Vulnerable Road Users .................................................................................. 62
5.2 Factors Considered in Barrier Selection................................................................................................. 65
5.2.1 Site Conditions ...................................................................................................................................... 66
5.3 Road Safety Barrier Design Process...................................................................................................... 69
5.3.1 Outline of Process ................................................................................................................................. 69
5.3.2 Collect Site Information (Step B1) ......................................................................................................... 72
5.3.3 Determine the Objectives of the Safety Barrier (Step B2) ..................................................................... 72
5.3.4 Determine the Lateral Position of the Barrier (Step B3) ........................................................................ 73
5.3.5 Offset to Traffic Lane............................................................................................................................. 75
5.3.6 Support Width ....................................................................................................................................... 81
5.3.7 Deflection Width .................................................................................................................................... 81
5.3.8 System Width ........................................................................................................................................ 81
5.3.9 Barrier Location in Medians .................................................................................................................. 81
5.3.10 Narrow Medians .................................................................................................................................... 83
5.3.11 Wider Medians ...................................................................................................................................... 85
5.3.12 Barrier-to-hazard Clearance (Step B4).................................................................................................. 85
5.3.13 Barrier Containment Level Required (Step B5) ..................................................................................... 85
5.3.14 Barrier Type (Step B6) .......................................................................................................................... 87
5.3.15 Dynamic Deflection (Step B7) ............................................................................................................... 96
5.3.16 Vehicle Roll Allowance and System Width (Step B8) ............................................................................ 96
Austroads 2020 | page ii

5.3.17 Working Width (Steps B9 and 10) ......................................................................................................... 99

5.3.18 Barrier Length of Need (Step B11) ...................................................................................................... 100
5.3.19 Minimum Length of Barrier (Step B12) ................................................................................................ 105
5.3.20 Sight Distance (Step B13) ................................................................................................................... 107
5.3.21 Terminal Treatments (Step B14) ......................................................................................................... 108
5.3.22 Transitions between Barriers (Step B15) ............................................................................................ 116
5.3.23 Confirm that the Barrier Meets the Objectives (Step B16) .................................................................. 117
5.4 General Access Through Road Safety Barriers ................................................................................... 117
5.5 Aesthetic Road Safety Barriers ............................................................................................................ 117
5.6 Other Road Safety Barrier Design Considerations .............................................................................. 118
5.6.1 Barriers at Intersections ...................................................................................................................... 118
5.6.2 Stepped Offset .................................................................................................................................... 120
5.6.3 Excessive Offset ................................................................................................................................. 121
5.6.4 Delineation .......................................................................................................................................... 121
5.6.5 System Height..................................................................................................................................... 121
6. Roadside Design for Steep Downgrades ......................................................................................... 122

6.1 Purpose and Need................................................................................................................................ 122
6.2 Containment Facilities .......................................................................................................................... 122
6.2.1 Gravity Safety Ramps ......................................................................................................................... 122
6.2.2 Arrester Beds ...................................................................................................................................... 122
6.2.3 Dragnets ............................................................................................................................................. 123
6.3 Warrants for Investigation .................................................................................................................... 123
6.4 Location and Spacing ........................................................................................................................... 123
6.5 Key Design Considerations .................................................................................................................. 124
6.6 Design Process .................................................................................................................................... 125
6.6.1 Outline of Process ............................................................................................................................... 125
6.6.2 Step S1 – Determine Vehicle Entry Speed ......................................................................................... 126
6.6.3 Step S2 – Evaluate Truck Stability on Approach ................................................................................. 126
6.6.4 Step S3 – Design Entry Alignment ...................................................................................................... 126
6.6.5 Step S4 – Determine Type of Facility .................................................................................................. 126
6.6.6 Step S5 – Determine Pavement Surface of Facility ............................................................................ 126
6.6.7 Step S6 – Design Facility Length ........................................................................................................ 127
6.6.8 Step S7 – Design the Facility .............................................................................................................. 128
6.6.9 Step S8 – Design End Treatment........................................................................................................ 130
6.6.10 Step S9 – Design Vehicle Recovery Facilities .................................................................................... 131
6.6.11 Step S10 – Design Delineation ........................................................................................................... 131
6.6.12 Step S11 – Design Truck Parking Areas ............................................................................................. 131
7. Work Zone Safety Barrier Systems .................................................................................................. 133
References .................................................................................................................................................... 134
Appendix A Terminology......................................................................................................................... 139

A.1 Vehicle Movement Terminology ........................................................................................................... 139
A.2 Road Safety Barrier Terminology ......................................................................................................... 139
Appendix B Detailed Risk Evaluation Procedure ................................................................................. 144

B.1 Exposure .............................................................................................................................................. 144
B.1.1 Determine the Future Traffic Flow ....................................................................................................... 145
B.2 Likelihood ............................................................................................................................................. 148
B.3 Likelihood of Colliding with a Point Hazard and a Length of Road Safety Barrier ...............................152
B.4 Severity ................................................................................................................................................. 153
B.4.1 Consistent Background Hazards ......................................................................................................... 154
B.4.2 Isolated Hazards ................................................................................................................................. 156
B.4.3 Roadside Barriers ............................................................................................................................... 157
B.4.4 Oncoming Vehicles ............................................................................................................................. 157
B.4.5 Trauma Indices for Roadside Features on Roads with Other Operating Speeds................................ 157
B.5 Evaluation of Risk Associated with Vehicle Rollovers.......................................................................... 158
Austroads 2020 | page iii

B.6 Risk Score ............................................................................................................................................ 158

B.6.1 Collective Risk .................................................................................................................................... 159
B.6.2 Individual Risk ..................................................................................................................................... 159
B.6.3 Risk of Short Lengths of Roadside ...................................................................................................... 159
B.7 Limitations of the Risk Assessment Process ....................................................................................... 160
B.8 Worksheets ........................................................................................................................................... 160
B.9 Example 1 – Culvert Headwall ............................................................................................................. 164
B.10 Example 2 – Centreline Treatments on a Rural Two-lane Road .........................................................172
B.11 Number of Fatal and Serious injuries ................................................................................................... 173
Appendix C Treatment of Roads Based on Jurisdictional Policies ....................................................174

C.1 Installation of Barriers on Particular Road Segments .......................................................................... 174
C.2 Installation of Barriers at Sites with High Consequence Infrastructure and Land ................................174
C.3 Installation of Barriers Between Adjacent Carriageways ..................................................................... 175
C.4 Installation of Barriers or Other Safety Measures on Other Defined Road Types and Locations........175
Appendix D Risk Score Charts ............................................................................................................... 176
Appendix E Cost of Impacts ................................................................................................................... 187

E.1 Determine Crash Costs ........................................................................................................................ 187
Appendix F Treatments for Brownfield Sites........................................................................................ 188

F.1 Treatments for Roads ........................................................................................................................... 188
F.1.1 Treatments for Pavement Edge Drop-off ............................................................................................ 188
F.1.2 Treatments for Opposing Vehicles ...................................................................................................... 188
F.2 Treatments for Bridges ......................................................................................................................... 189
F.2.1 General ............................................................................................................................................... 189
F.2.2 Treatments for Bridge Piers, Abutments, End Posts and Tunnel Portals ............................................ 189
F.3 Barrier Placement in Constrained Situations ....................................................................................... 189
F.3.1 Location on Embankments.................................................................................................................. 189
F.3.2 Location on Urban Footpath Corners .................................................................................................. 190
F.4 Wire Rope Barriers in Narrow Medians ................................................................................................ 191
Appendix G Angle of Departure Method ................................................................................................ 193
Appendix H Examples of Length of Need Calculations ....................................................................... 198

H.1 General ................................................................................................................................................. 198
H.2 Run-out Length Method ........................................................................................................................ 198
H.3 Angle of Departure Method .................................................................................................................. 199
H.4 Worked Examples to Determine Road Safety Barrier Length of Need ................................................199
H.4.1 Example 1 ........................................................................................................................................... 200
H.4.2 Example 2 ........................................................................................................................................... 203
Appendix I Types of Safety Barrier Terminals .................................................................................... 207

I.1 General ................................................................................................................................................. 207
I.1.1 Gating End Treatments ....................................................................................................................... 207
I.1.2 Non-gating End Treatments ................................................................................................................ 208
I.1.3 Trailing Terminal ................................................................................................................................. 211
Appendix J Transitions between Barrier Types ................................................................................... 212

J.1 General ................................................................................................................................................. 212
J.2 Design Criteria – Physically Connected Barriers ................................................................................. 212
J.3 Typical Interfaces between Barrier Types ............................................................................................ 213
J.3.1 General ............................................................................................................................................... 213
J.4 W-beam to Thrie-beam ........................................................................................................................ 213
J.5 W-beam to Concrete ............................................................................................................................ 213
J.6 Thrie-beam to Concrete ....................................................................................................................... 214
J.7 Wire Rope Safety Barrier to Semi-rigid Barrier .................................................................................... 214
J.8 Wire Rope Safety Barrier to Concrete Barrier ...................................................................................... 214
Austroads 2020 | page iv

Commentary 1 ............................................................................................................................................... 217

Commentary 2 ............................................................................................................................................... 224
Commentary 3 ............................................................................................................................................... 224
Commentary 4 ............................................................................................................................................... 225
Commentary 5 ............................................................................................................................................... 226
Commentary 6 ............................................................................................................................................... 227
Commentary 7 ............................................................................................................................................... 234
Commentary 8 ............................................................................................................................................... 234
Tables
Table 1.1: Road characteristics for the charts in Figure 1.6 ....................................................................... 19
Table 1.2: Rural road characteristics for the Risk Score charts in Figure 1.7 ............................................21
Table 1.3: Risk Score adjustment factors for curve radii on two-lane rural roads ......................................22
Table 1.4: Risk Score adjustment factors for gradient of two-lane rural roads ..........................................23
Table 1.5: Road characteristics for the charts in Figure 1.12 ..................................................................... 25
Table 4.1: Crash types and effectiveness of treatments ............................................................................ 51
Table 5.1: Selection criteria for roadside barriers ....................................................................................... 65
Table 5.2: Key considerations in barrier selection ...................................................................................... 68
Table 5.3: Minimum offsets from kerb to barrier face ................................................................................. 78
Table 5.4: Shy-line offset values ................................................................................................................ 79
Table 5.5: Flare rates.................................................................................................................................. 80
Table 5.6: Typical support width ................................................................................................................. 81
Table 5.7: Vehicle roll allowance ................................................................................................................ 99
Table 5.8: Run-out lengths for barrier design ........................................................................................... 101
Table 5.9: Selection factors for terminal treatments ................................................................................. 110
Table 6.1: Typical warrants for analysis for runaway vehicles ................................................................. 123
Table 6.2: Approximate distance from summit to safety ramp ................................................................. 124
Table 6.3: Maximum decrease in speed between successive geometric elements ................................124
Table 6.4: Rolling resistance .................................................................................................................... 127
Table 6.5: Design features of arrester beds ............................................................................................. 129
Table 6.6: Arrester bed material specification .......................................................................................... 130
Table A.1: Vehicle movement terminology ............................................................................................... 139
Table A.2: Road safety barrier terminology .............................................................................................. 139
Table B.1: Exposure adjustment factors for undivided roads ................................................................... 146
Table B.2: Exposure adjustment factors for divided and one-way roads .................................................147
Table B.3: Exposure adjustment factors for urban undivided roads with operating speeds less than
or equal to 90 km/h .................................................................................................................. 147
Table B.4: Trauma Indices for continuous features – operating speed = 110 km/h .................................154
Table B.5: Trauma Indices for embankment slopes on fills and in cuts – operating speed =
110 km/h .................................................................................................................................. 155
Table B.6: Trauma Indices for vertical drops and water depths – operating speed = 110 km/h ..............155
Table B.7: Trauma Indices for parallel V drains with a height less than 1.2 m – operating speed =
110 km/h .................................................................................................................................. 156
Table B.8: Trauma Indices for point hazards – operating speed = 110 km/h ...........................................156
Table B.9: Trauma Indices for slopes parallel to the road – operating speed = 100 km/h .......................157
Table B.10: Speed factors for the likelihood of a rollover ........................................................................... 158
Table B.11: Risk Score calculation worksheet for verges on divided and undivided roads .......................160
Table B.12: Risk Score calculation worksheet for medians on divided roads ............................................162
Table B.13: Risk Score calculation worksheet for oncoming vehicles on undivided rural roads ................164
Table B.14: Risk Score calculation for the background hazards being the trees at the toe of the
embankment ............................................................................................................................ 165
Austroads 2020 | page v

Table B.15: Risk Score calculation for the background hazard being the embankment ............................167
Table B.16: Risk Score calculation for the background hazards being the trees at the toe of the
embankment and shielded with a barrier ................................................................................ 168
Table B.17: Risk Score calculation for a 200 m length of road containing the culvert end wall .................169
Table B.18: Risk Score calculation for reduced speeds and a continuous barrier .....................................171
Table B.19: Risk Score for a median treatment of a two-lane two-way rural road .....................................172
Table D.1: Risk evaluated in the figures in this appendix ......................................................................... 176
Table D.2: Road characteristics for the charts in Figure D 1 .................................................................... 178
Table D.3: Rural road characteristics for the Risk Score charts in Figure D.2 .........................................179
Table D.4: Road characteristics for the charts in Figure D.5 and Figure D.6 ...........................................181
Table D.5: Road characteristics for the charts in Figure D.7 .................................................................... 182
Table D.6: Road characteristics for the charts in Figure D.8 and Figure D.9 ...........................................183
Table D.7: Road characteristics for the charts in Figure D.11 .................................................................. 185
Table D.8: Significant and less significant and minor hazards.................................................................. 185
Table D.9: Risk Score correction factors for curve radii on two-lane rural roads......................................186
Table D.10: Risk Score correction factors for the gradient of two-lane rural roads ....................................186
Table E.1: Trauma Indices and predicted crash outcomes ...................................................................... 187
Table F.1: Issues for wire rope barriers located centrally in medians ......................................................192
Table G.1: Angles of departure from the road ........................................................................................... 194
Table C1.1: Considerations in relation to designing roads for safety .........................................................218
Table C1.2: Key considerations for consistent design ................................................................................ 221
Table C1.3: Road design parameters for consideration in relation to heavy vehicles ................................222
Table C1.4: Other factors contributing to errant vehicles ............................................................................ 223
Table C6.1: An example of bumper trajectory data over embankments at 100 km/h .................................229
Table C6.2: An example of bumper trajectory data over specific kerbs .....................................................230
Table C6.3: Bumper trajectory data over cutting slopes ............................................................................. 232
Figures
Figure 1.1: Flow chart of the Guide to Road Design ..................................................................................... 3

Figure 1.2: Risk assessment flow diagram .................................................................................................. 12
Figure 1.3: Calculation of an effective lateral distance to a hazard ............................................................. 16
Figure 1.4: View of road showing trees at different offsets.......................................................................... 17
Figure 1.5: View of an urban road showing background and isolated hazards ...........................................18
Figure 1.6: Risk Score for undivided rural roads with significant background hazards at different
offsets and for different operating speeds and road types ........................................................19
Figure 1.7: Risk Score for undivided rural roads with significant isolated hazards at different
average spacings along a roadside and for different operating speeds and evaluated
over 1 km ................................................................................................................................... 20
Figure 1.8: Risk Score for isolated significant hazards on undivided rural roads without background
hazards ...................................................................................................................................... 21
Figure 1.9: Rollover Risk Score for undivided rural roads without background hazards and an
operating speed of 110 km/h ..................................................................................................... 22
Figure 1.10: Risk Score for divided urban roads with significant hazards at different offsets and for
different operating speeds (Scenario 1) .................................................................................... 23
Figure 1.11: Risk Score for divided urban roads with significant hazards at different average
spacings along a roadside and for different operating speeds (Scenario 2) .............................24
Figure 1.12: Risk Score for roadside barriers on undivided rural roads with different operating speeds
and terrain (refer to Table 1.5 for road details) .......................................................................... 25
Figure 1.13: Risk Score for roadside barriers on divided carriageway urban roads with different
operating speeds and barrier offsets (refer to Table 1.6 for road details) .................................25
Figure 1.14: Risk Score for roadside barriers on divided carriageway urban and rural roads with an
operating speed of 110 km/h ..................................................................................................... 26
Figure 2.1: Estimates of the NRRIT using lower section Figure 1.6 ............................................................ 32
Figure 2.2: Estimates of the NRRIT using Figure 1.10 ................................................................................ 33
Figure 2.3: Estimates of the NRRIT using Figure 1.11 ................................................................................ 33
Figure 3.1: Diagrammatical representation of a roadside ............................................................................ 36
Figure 3.2: Using the Risk Score values...................................................................................................... 36
Austroads 2020 | page vi

Figure 3.3: Using assumed characteristics to evaluate the Risk Score over a 1 km segment ....................37
Figure 3.4: Road view for example 1 ........................................................................................................... 39
Figure 3.5: Risk Score for example 1 .......................................................................................................... 39
Figure 3.6: Example 2 Highway in Western Australia ................................................................................. 40
Figure 3.7: Risk Score for example 2 using charts 11 and 12 in Figure D.3 ...............................................40
Figure 3.8: Rollover Risk Score for example 2 using chart 13 in Figure D.4...............................................41
Figure 3.9: Risk Score for example 2 using chart 5 ..................................................................................... 41
Figure 3.10: Example 3 – culvert headwall on outside of horizontal curve ...................................................42
Figure 3.11: Equivalent cross-section for example 3 ..................................................................................... 42
Figure 3.12: Unadjusted Risk Score for example 3 ....................................................................................... 43
Figure 3.13: Risk Score for culverts classified as a significant hazard ..........................................................44
Figure 3.14: View of a typical urban road ...................................................................................................... 45
Figure 3.15: Risk Score for example 4 using chart 18 ................................................................................... 45
Figure 3.16: Risk Score for example 3 using charts 17 and 18 ..................................................................... 46
Figure 3.17: Risk Scores for continuous barriers on undivided rural roads with different operating
speeds and in different terrain ................................................................................................... 47
Figure 4.1: Example of a driveable culvert end wall for a small pipe under a driveway ..............................54
Figure 4.2: Example of a traversable culvert end treatment (under construction) for a culvert under
the road...................................................................................................................................... 54
Figure 4.3: An example of a slip-base pole mechanism .............................................................................. 59
Figure 4.4: Examples of impact absorbing poles ......................................................................................... 60
Figure 5.1: Road safety barrier design process ........................................................................................... 71
Figure 5.2: Verge barrier location ................................................................................................................ 74
Figure 5.3: An illustration of bumper height trajectory characteristics over a fill embankment ...................75
Figure 5.4: Barrier offset at kerb .................................................................................................................. 78
Figure 5.5: Detail of flare rate ...................................................................................................................... 80
Figure 5.6: Examples of road safety barriers for use in narrow medians ....................................................83
Figure 5.7: An example of a barrier layout for shielding a rigid object in a median .....................................84
Figure 5.8: Encroachment angle .................................................................................................................. 88
Figure 5.9: Maximum angle of impact for cars on straights ......................................................................... 89
Figure 5.10: Angle correction factor ............................................................................................................... 89
Figure 5.11: Examples of flexible (i.e. wire rope) barrier ............................................................................... 90
Figure 5.12: Semi-rigid barrier ....................................................................................................................... 90
Figure 5.13: Rigid barrier ............................................................................................................................... 90
Figure 5.14: Examples of profiles of semi-rigid barriers ................................................................................ 91
Figure 5.15: Examples of profiles of rigid barriers ......................................................................................... 91
Figure 5.16: Preferred installation of rigid barrier on a superelevated roadway ............................................94
Figure 5.17: Rigid barrier in narrow median with independently graded carriageways .................................94
Figure 5.18: Two-stage shielding layout ........................................................................................................ 95
Figure 5.19: Vehicle roll allowance ................................................................................................................ 98
Figure 5.20: Length of need – single direction............................................................................................. 100
Figure 5.21: Length of need – two directions............................................................................................... 100
Figure 5.22: Run-out length method of determining length of need ............................................................ 103
Figure 5.23: Length of need on outside of curve using run-out length method ...........................................104
Figure 5.24: Length of need on inside of curve using run-out length method .............................................105
Figure 5.25: Road safety barrier lengths – anchored systems .................................................................... 106
Figure 5.26: Road safety barrier lengths – unanchored systems ................................................................ 107
Figure 5.27: Space required for crash cushions in gore areas .................................................................... 111
Figure 5.28: Run-out area for gating terminals ............................................................................................ 113
Figure 5.29: Gating and non-gating systems ............................................................................................... 114
Figure 5.30: Redirective and non-redirective systems ................................................................................ 115
Figure 5.31: An example of barrier access details ...................................................................................... 117
Figure 5.32: An example of a curved barrier at a major road intersection (radius 2.5 to 9.9 m) .................119
Figure 5.33: An example of a curved barrier at a major road intersection (radius ≥ 10 m) .........................120
Figure 6.1: Design process for steep downgrade ...................................................................................... 125
Figure 6.2: An example of an arrester bed ................................................................................................ 128
Figure 6.3: An example of an arrester bed layout ..................................................................................... 129
Figure A 1: Vehicle movement terminology ............................................................................................... 139
Austroads 2020 | page vii

Figure A.2: Road safety barrier terminology – curvature ........................................................................... 142

Figure A.3: Road safety barrier terminology – impact ................................................................................ 142
Figure A.4: Road safety barrier terminology – placement on two-lane/two-way road ...............................143
Figure A.5: Road safety barrier terminology – placement on multilane/one-way road ..............................143
Figure B.1: Base exposure (enc/km/year) for each side of an undivided road ..........................................145
Figure B.2: Base exposure (enc/km/year) for each side of each carriageway of a divided road...............145
Figure B.3: Gradient correction factor ........................................................................................................ 147
Figure B.4: Curve correction factor ............................................................................................................ 148
Figure B.5: Lateral distance measures ...................................................................................................... 148
Figure B.6: Likelihood of reaching a lateral distance for a road with an operating speed of 110 km/h
and with curves of different radii .............................................................................................. 149
Figure B.7: Likelihood of reaching a lateral distance for road with a 90 km/h operating speed and
with curves of different radii ..................................................................................................... 150
Figure B.8: Likelihood of reaching a lateral distance for road with a 70 km/h operating speed and
with curves of different radii ..................................................................................................... 150
Figure B.9: Likelihood of a vehicle crossing a lane line and colliding with an oncoming vehicle on an
undivided rural road with a speed limit of 100 km/h ................................................................ 151
Figure B.10: Lateral distance to a hazard .................................................................................................... 152
Figure B.11: Impact envelopes for a hazard 1 m by 1 m (not to scale) .......................................................153
Figure B.12: Example 1 – culvert headwall on outside of horizontal curve .................................................165
Figure D 1: Risk Score for undivided rural roads with significant background hazards .............................178
Figure D.2: Risk Score for undivided rural roads with significant isolated and background hazards ........179
Figure D.3: Risk Score for undivided rural roads with isolated and no background hazards ....................180
Figure D.4: Risk Score for vehicle rollover on roadsides on undivided rural roads ...................................180
Figure D.5: Risk Score for divided urban roads with significant background hazards ...............................181
Figure D.6: Risk Score for divided urban roads with significant isolated and background hazards ..........181
Figure D.7: Risk Score for undivided lower speed urban roads with background hazards .......................182
Figure D.8: Risk Score for roadside barriers on undivided rural roads (refer to Table D.6 for road
details) ..................................................................................................................................... 183
Figure D.9: Risk Score for roadside barriers on divided carriageway urban roads (refer to Table D.6
for road details) ........................................................................................................................ 183
Figure D.10: Risk Score for roadside barriers on divided roads with an operating speed of 110 km/h .......184
Figure D.11: Risk Score for culvert headwalls classified as a significant hazard, measured over a 100
m road section ......................................................................................................................... 184
Figure F.1: Wire rope barrier in a narrow median on a sharp bend ...........................................................188
Figure F.2: Barrier post locations on constrained sites .............................................................................. 190
Figure F 3: Road safety barriers on footway corners ................................................................................. 191
Figure G.1: Angle of departure method of determining length of need ......................................................193
Figure G.2: Angle of departure method on curves where leading angle meets the rear of hazard ...........196
Figure G.3: Angle of departure method where leading angle does not meet the rear of hazard ...............197
Figure H.1: Run-out length method of determining length of need ............................................................ 199
Figure H.2: Road cross-section .................................................................................................................. 200
Figure H.3: Culvert headwall (plan view).................................................................................................... 200
Figure H.4: Establishing the leading point of need ..................................................................................... 201
Figure H.5: Establishing trailing point of need ............................................................................................ 202
Figure H.6: Barrier layout ........................................................................................................................... 203
Figure H.7: Establishing leading point of need ........................................................................................... 204
Figure H.8: Establishing the trailing point of need ...................................................................................... 205
Figure I.1: An example of a non-proprietary bull-nose attenuator ............................................................ 210
Figure I.2: Trailing terminal ....................................................................................................................... 211
Figure J.1: An example of a transition between a wire rope barrier and W-beam barrier .........................215
Figure J.2: An example of a transition between a wire rope barrier and rigid barrier ...............................216
Figure C6.1: An example of bumper height trajectory characteristics over fill embankments .....................228
Figure C6.2: An example of bumper height trajectory characteristics over kerbs........................................230
Figure C6.3: An example of bumper height trajectory characteristics on cutting slopes .............................231
Figure C6.4: Summary of barrier locations – preferred and not recommended...........................................233
Figure C8.1: Types of vehicle escape ramps ............................................................................................... 235
Austroads 2020 | page viii

1. Introduction to Roadside Design
1.1 Context Sensitive Designs
The basis of road design in Australia and New Zealand is the incorporation of context-sensitive principles
into the design of road and roadside facilities (refer to the Guide to Road Design Part 1). The design process
should adopt a principles-based decision-making approach for responding to competing objectives. This
process provides a robust method for determining the suitability of designs that fall within and outside the
normal design domain (NDD) and the extended design domain (EDD).
Projects that consider Context Sensitive Design (CSD) are more able to meet transportation needs,
recognise the need for cost-effective and at times innovative solutions, while enhancing the benefits to the
community. This is achieved through preservation efforts, sensitivity to local values and recognising the
needs of all road users (including people, goods distributors and those who provide essential services).
Concepts such as Movement and Place, described in the Guide to Traffic Management Part 4: Network
Management Strategies (Austroads 2020a), can be used to assess transport objectives and priorities. Any
solution should be one that balances safety, mobility, reliability, community acceptance, stakeholder
acceptance, project and maintenance costs, and aesthetic, historical, environmental and other community
values within an emphasis on maximising safe mobility (Austroads 2018). The greatest challenges and
opportunities in designing a transportation solution that balances those values occur early in the design
process, during project planning and development of design alternatives (AASHTO 2018).
When developing designs, there are generally no off-the-shelf solutions that fully address all situations.
Every location is unique. Therefore, well-founded design criteria must be chosen for each location. The
context of the particular network link or project is of great significance in determining these parameters. This
context is formed from the amalgamation of many factors, such as:
• what is physically possible to construct at that location
• what is reasonable to expect
• what operational and safety performance can be achieved
• what costs are involved
• what social, community and environmental effects might result.
CSD offers flexibility for jurisdictions to decide what criteria are appropriate to a particular situation while
aligned to the vision for the network or link and considers infrastructure as more than just a transportation
facility, emphasising as well the importance that it be integrated with the local community. CSD takes an
interdisciplinary approach, promoting collaboration between project developers, designers, delivery teams
and the community. It asks questions about the need and purpose of the project and fosters appropriate
thinking and analysis during development of solutions. This is essential to achieve good design.
Criteria must be applied only after careful analysis of what must be designed and should be considered in
conjunction with engineering judgement and common sense based on knowledge and experience. CSD
promotes efficiency in design and construction by developing economic solutions that are appropriate for the
network link and its purpose. It also promotes safety for road users.
Those involved in making decisions during design should exercise critical engineering judgement to ensure
that the unique features of a location are considered. They must be able to produce strong, defensible
evidence in support of their judgements.
It is usually the responsibility of the person managing the project to ensure that the use of particular design
criteria is appropriate in consideration of jurisdictional practices and policy. It is usually the responsibility of
the designer to provide advice on the risks involved in adopting any particular criterion. Designers must fully
understand the effect on safety and operations of combining different design elements under different
circumstances. They must also inform relevant stakeholders of the potential implications of design decisions,
particularly if adopted criteria are less than optimal.
Austroads 2020 | page 1

Road design is often a compromise between what is considered an ideal solution and what is a reasonable
outcome with regard to safety, driver expectation, economic imperatives, environmental impacts and
community values.
1.2 Purpose
Austroads Guide to Road Design seeks to assimilate the contemporary road design practice of member
organisations (for an overview refer to the Guide to Road Design Part 1: Introduction to Road Design
(Austroads 2015a)). It provides guidance to designers to produce safe, economical and efficient road
designs.
There are three parts of the Guide to Road Design (AGRD) that collectively deal with the design of
roadsides:
• Part 6: Roadside Design, Safety and Barriers (AGRD Part 6) (this Part)
• Part 6A: Paths for Walking and Cycling (Austroads 2017a)
• Part 6B: Roadside Environment (Austroads 2015b).
The collective purpose of these documents is to:

• promote a uniform approach to roadside design and safety by road agencies throughout Australia and
New Zealand
• provide road design and road safety practitioners with
– an understanding of roadside safety issues including the assessment of risk
– guidance on the design of the roadside and infrastructure that must be accommodated within the
road reservation.
Figure 1.1 shows the broad context in which AGRD Part 6 (including AGRD Parts 6A and 6B) is applied. It
indicates that AGRD Part 6 is the sixth in a series of Guides that comprise the Austroads Guide to Road
Design and provide information on a range of disciplines including geometric design, intersections and
crossings, drainage and geotechnical design, all of which may influence the space available within a
roadside and the design of features and infrastructure within it.

Figure 1.1: Flow chart of the Guide to Road Design
While Figure 1.1 outlines the structure of the Guide to Road Design, designers should be aware that there
are other subject areas spanning the range of Austroads publications that may also be relevant to roadside
design and safety and can be accessed via <www.austroads.com.au>.
1.3 Reading this Part in the Context of Part 1
The content of Guide to Road Design Part 1 describes the broad basis and concepts of road design and
aspects of network management and planning, corridor and project design.
Road design input is necessary in all phases of the life cycle of infrastructure but predominantly:
• Network strategy and planning – consideration of the purpose and need of the links in the network with a
view to achieving the best network safety outcome through setting consistent standards and achieving a
sustainable and self-explaining road environment. This requires a strategic statement on the level of
investment to meet the future requirements for safety and mobility of a network or corridor. In the context
of roadside safety, this means setting a Network Roadside Risk Intervention Threshold (NRRIT) to
identify the risk of roadside crashes and the need for mitigating treatments. The NRRIT is explained in
detail in Section 2.

• Program investment and development – The scope of the program and component projects should align
with the vision in the network strategy and planning phase. The NRRIT, set as part of the network
strategy and planning phase, identifies the improvements required in the program(s).
• Project development and delivery – understanding the project objective (for which stakeholder
engagement is essential), the context of the site, the alternative solutions considered in meeting design
principles, the design risks and controls, the costs of options, and alignment with the link and network
vision. Project locations need to align with NRRIT set in the network strategy and planning phase.
Road design considerations need to be integrated into the management phases of network strategy and
planning, program investment and development, project development and delivery and subsequent
operations and maintenance activities. These processes cannot be effectively undertaken without this
integration. Possible treatments need to be considered as early as possible in the network strategy and
planning phase to maximise the outcomes.
It is essential that the design brief for a road project sets out not only the design standards for alignment and
cross-section, and other engineering requirements, but also the location and extent of roadside safety
facilities or infrastructure that must be accommodated. This information should result from a planning study
or liaison with other authorities and, in some instances, may take the form of a conceptual layout. With a
clear brief, the designer should be able to develop a roadside design that accommodates and coordinates all
the requirements of road users and other stakeholders or alternatively, identifies areas of conflict that require
further investigation and/or negotiation.
1.3.1 Combining Design Parameters and Consistent Design
A safer road does not necessarily require a wide pavement and an alignment designed to accommodate
high speeds but is one where on-road and roadside features clearly show drivers the path that a road takes
and helps them keep their vehicles in the running lane.
Roads must be located within the topography in a cost-effective way, and this may require that a design
speed and cross-section be adopted to suit the function of the road, traffic characteristics and topography.
The road design in these circumstances should enable the driver to travel safely at the intended design
speed on a consistent alignment. In summary, the following considerations are important:
• Combinations of design parameters – the adoption of lower-order values for a number of design
parameters in combination may create an unsafe design even though the individual design parameters
comply with guidelines.
• Consistent design environment – on-road and roadside features that clearly show drivers the path that a
road takes and helps them keep their vehicles in the running lane.
• Vehicle mix considerations – it is important to consider the impact and additional risk of a higher than
normal percentage of heavy vehicles, particularly where steep grades are involved.
• Other design elements and features – e.g. horizontal and vertical alignment, lane widths and drainage.
Further information is provided in Commentary 1.

[see Commentary 1]
1.3.2 Progressive Safety Updates to the Guide to Road Design
Consistent with the Australian National Road Safety Strategy 2011–2020 (Australian Transport
Council 2011) and New Zealand’s Safer Journeys: Road Safety Strategy 2010–2020 (NZ Ministry of
Transport 2010), this Guide is being progressively updated to facilitate the practical implementation of Safe
System thinking. While there is information in relation to the potential treatments for a Safe System (refer to
the Austroads Guide to Road Safety), the challenge is to provide the necessary warrants, criteria and
dimensional guidance to support most decision-making situations, scenarios and options for designers.
Future updates will progressively provide the information to make the Safe System improvements given the
project, program and network considerations.

There are numerous innovative treatments being considered to improve road safety with many from
overseas. For Australia and New Zealand, these treatments need to be assessed and translated to local
conditions to ensure an improved outcome for all road users. Assessment and translation require full and
robust engineering processes to assess risk, safety, operational and whole-of-life cost impacts of these
treatments. Where no supporting dimensional guidance and quantitative evidence is available, necessary
design exception analysis and reporting are required before implementation of any trials.
All safety improvements will fall within the overarching objective to deliver a context sensitive, network-wide
treatment to develop a consistent road so that drivers are not impacted by unexpected changes in road
standard or functionality. Also, network-wide consistent treatment should be planned and implemented over
a timeframe matched to the asset life to minimise rework and wastage of infrastructure treatments and
investment. An integrated road design approach is important in achieving Safe System transformation of the
road network across all areas of activity.
1.4 Scope of this Part
This Part describes the elements of roadside design and the many features and objects that may have to be
accommodated and coordinated in the space between the edge of the carriageway and the reservation
boundary, and within medians. The major focus of Part 6 is to provide guidelines for hazard identification and
risk mitigation processes and to give a clearly defined process for designing roads for safety.
Specifically, this Part provides:

• guidelines on the rationale of errant vehicle management
• guidelines for assessment and treatment of hazards on the roadside
• guidance on the selection and location of road safety barriers
• a road design process that implements errant vehicle management and risk management principles.
While some consideration is given to trucks, motorcyclists, cyclists and pedestrians with respect to road
safety barriers, it should be understood that the hazard mitigation processes discussed in this Part largely
relate to the occupants of cars.
This Part also covers the:

• assessment of the risk associated with vehicles running off the road and crashing into roadside hazards,
based on traffic conditions, the geometry of the road and roadside and the presence of hazards
• assessment of engineering risk mitigation treatments for roadside hazards, including:
– the provision of roadside barriers
– the provision of a hazard-free area alongside the road
– the use of audio-tactile line-marking and other treatments that guide and warn drivers
– minor design changes, such as flattened batters
• the type of road safety barriers to shield roadside hazards including the performance level, length and
clearances required
• design of other road safety related devices such as runaway vehicle ramps and heavy vehicle arrester
beds.
Features associated with the design of the road itself and which exist within the roadside are concerned with
cross-sections (e.g. verges, embankments) or drainage (e.g. open drains, inlets and outlets to transverse
culverts) and are covered in the Guide to Road Design Part 3: Geometric Design (Austroads 2016a) and the
Guide to Road Design Part 5B: Drainage: Open Channels, Culverts and Floodways (Austroads 2013b).
However, the design of elements associated with cross-sections and drainage may also have to consider an
extensive range of other requirements within the road reserve.

This Part does not cover events in which drivers deliberately drive off the road, such as an act of terrorism,
the management of crashes on the carriageway or at intersections (apart from cross-centreline head-on
crashes on two-lane rural roads).
1.4.1 AGRD Part 6A: Paths for Walking and Cycling
AGRD Part 6A (Austroads 2017a) covers the need for paths and the geometric design of pedestrian, shared
and bicycle paths both within roadsides and in reservations that are remote from roads but may intersect with
them.
1.4.2 AGRD Part 6B: Roadside Environment
AGRD Part 6B (Austroads 2015b) covers the roadside facilities and infrastructure that are not directly
associated with roadside safety, safety barriers, or pedestrian and cyclist paths. The purpose of the Part is to
provide the guidance to a road designer to integrate these facilities into a road design. It is expected that the
designer will have to involve experts in various fields to resolve requirements for:
• environmental aspects such as stormwater run-off, fauna management and noise control
• landscaping
• roadside amenity including visual amenity and rest facilities
• roadside infrastructure such as road furniture, lighting, emergency/help telephones, off-street parking and
utilities.
1.5 Principles Considered in Roadside Design to Achieve the Safest System
Roadside design includes the design of all features and infrastructure that need to be accommodated in the
area between the road reserve boundary and the nearest road shoulder (or kerb) and within medians.
A transparent, principles-based approach to decision-making should be adopted during design. This

approach encourages independent designs tailored to particular situations and in compliance with
engineering principles. The responsibilities and liability of designers, who deviate from the guidance provided
here, are discussed in other Parts of the Guide to Road Design.
As well as providing robust engineering justification for many design decisions, standards and guidelines of
jurisdictions, Austroads also provides discretionary principles to help designers deal with multi-objective
integrated planning. A comprehensive approach should be adopted when applying discretionary principles,
seeking technical advice from subject matter experts where appropriate. Safety performance must always be
considered, especially the Safe System Principle (see 1.5.1).
Each core principle applied to assist with decision making should have quantitative support where
appropriate. For example, the project benefits of a design element or design decision should be quantified,
as should the risks. Quantification provides a measure for later monitoring and evaluating the success of the
design.
Road design aims to provide the safest transport network for all users within available resources by:
• minimising the likelihood and severity of conflict between vehicles
• reducing the likelihood that vehicles collide with roadside objects
• reducing the forces on road users and occupants when their vehicles collide with roadside objects
• minimising the likelihood and severity of conflict between vulnerable users and other users of roads
• meeting drivers’ expectations (including reasonable driver capability) by using self-explaining road
designs
• providing consistency of character along the road corridor.

Designers must thoroughly investigate the types of crashes at those sites and their relationship to the design
of the site. Designers should also consider the substantive safety, being the measured or expected crash
frequency and severity, provided for similar sites. The substantive safety should be documented, along with
any existing substandard design elements and elements designed to minimum criteria.
1.5.1 Safe System Principle
The Safe System approach considers the human factors in the road system, recognising that humans make
mistakes and the human body can withstand only limited forces. A Safe System approach considers how
roads and roadsides can be made more forgiving of human error, looks at how vehicles can contribute to
saving lives and reducing harm, and ensures that travel speeds are appropriate for the roads and for all who
use them. Such an approach is not only about reducing the likelihood of crashes, but also about reducing
their severity when they do occur. In particular, it aims to reduce the potential for fatalities and serious
injuries.
The potential for death or serious injury is directly related to the likelihood of a crash and the likely severity of
forces occupants are subjected to during a crash. This potential for harm reduction is affected by the speed
environment, the expected and actual safety record, traffic composition, road facility type, geometry,
roadside characteristics and the likely success of various mitigation options. The risk evaluation process
outlined in this Part assists in undertaking assessments.
While it is unlikely that death and serious injury will be completely eliminated from the entire system in the
near future, substantial gains and even virtual elimination are possible in components of the system over
varying timeframes.
As more is understood about Safe System implementation, a hierarchy of treatments (refer to the Guide to
Road Safety) that achieve the best harm reduction outcomes will become more apparent and able to be
quantified. Practitioners need to be aware of these hierarchies and why treatments are aligned with Safe
System principles.
While this Part is focussed on certain aspects of the road infrastructure pillar, it is important to note that
infrastructure alone cannot be expected to achieve Safe System outcomes and contributions from the other
pillars will still be required. Speed management can achieve rapid and significant outcomes. Automated
technologies have massive potential to reduce serious injuries and fatalities, but their deployment is affected
by the time taken to turn over the fleet. Behavioural measures can also be effective, but it is difficult to
optimise outcomes.
From an infrastructure perspective, the Safe System approach involves:

• designing a road system through planning, programming, constructing and maintenance stages for roads,
vehicles and operating conditions, so that forces on the human body generated in crashes do not result in
fatality or serious injury
• improving roads and roadsides to reduce the potential for crashes and minimise harm (by, for example,
dividing the traffic, designing forgiving roadsides and providing clear driver guidance)
• setting speed limits that consider vulnerable road users and the variability of risks on different parts of the
road system.
1.5.2 Road Network Efficiency Principle
Consistent with the Movement and Place decision-making framework in Austroads (2020a), designs should
not adversely affect the efficient movement of vehicles on roads designated as significant traffic or freight
routes. Transport network decisions should be aligned with land-use decisions so that transport infrastructure
will meet the accessibility and operational needs of all transport modes in both the present and the future.
Consideration should be given to vulnerable road users and also those who may not directly use the road but
whose transportation needs may be affected by the proposed works.

1.5.3 Community Wellbeing Principle
Transport infrastructure should support healthy liveable communities, provide high-quality roadside amenity
and accommodate active travel, such as walking and cycling. It is usually necessary to gain broad
stakeholder and community acceptance of road infrastructure proposals.
1.5.4 Environmental Sustainability Principle
The transport system should protect the natural environment, in particular the flora and fauna, by enhancing
the aesthetics of roadside amenities. This principle extends to respecting and preserving cultural heritage
values and assets, both Indigenous and non-Indigenous.
1.5.5 Utility Services Principle
Road reserves need to accommodate utility services without adversely affecting road safety. Works should
not make the roadside environment any less safe than it currently is (and provide appropriate mitigation if
necessary). This includes providing easy access and a safe workplace for utility maintenance.
1.5.6 Investment Benefit Principle
A proposed design element should be objectively reviewed to establish whether the associated capital
expenditure represents the best use of community resources. This evaluation should be based on
whole-of-life costs including the initial development costs, mitigation costs and maintenance costs.
Project planners must also ensure the sustainability of proposals so they align with long-term network plans
to provide a self-explaining and consistent level of treatment and to avoid redundant works in the future.
1.5.7 Safety in Design Principle
While the maintenance task should not dictate the road safety outcome, appropriate assessment of safety in
design is a mandatory workplace health & safety requirement for the construction, maintenance and
operation of the network asset. The safety of workers constructing and maintaining the treatment is integral
to design decisions.
1.6 Roadside Safety Design
1.6.1 General
In every situation, the road should be designed to minimise the likelihood that any aspect of the design will
contribute to vehicles leaving the road. In the event that a vehicle does leave the road the roadside should
be designed to eliminate or minimise the likelihood of a crash occurring and to minimise the severity of any
crash that does occur.
Road and roadside designers should:

• consider the safety of all road users and produce a road environment that promotes safer travel
• design for risk reduction
• choose appropriate barriers and other treatments for potential impacts at the site.
This Part applies to both greenfield and brownfield sites. While greenfield sites may offer designers greater
flexibility it is unlikely that designers will be able to avoid, remove, or relocate all potential roadside hazards.
The design may therefore require assessment, treatment and shielding of roadside hazards.

1.6.2 Road Environments that Promote Safer Travel
The initial approach to hazard management is to design roads to keep vehicles on the road and prevent
vehicles from becoming errant.
To assist drivers to keep their vehicles on the road, it is necessary to provide geometric designs that improve
a driver’s ability to anticipate events. Visual cues, appropriate signing and line-marking, sealed shoulders
and well-maintained roads minimise factors that cause vehicles to leave the road (Austroads 2008).
Practitioners should refer to the Guide to Road Safety Part 9: Roadside Hazard Management
(Austroads 2008) and Guide to Road Design Part 3 (Austroads 2016a) when designing a new road or
considering the adequacy of an existing road. All practical and economically feasible measures should be
taken to prevent vehicles from leaving the road in addition to providing a safer roadside. The likelihood that a
vehicle will leave either side of the road may depend on many factors relating to the driver and environment
as described in Commentary 2. It is not only the safety of the car occupants that should be considered, but
that of other road users such as pedestrians, cyclists and motorcyclists, as well as persons occupying
properties that may be impacted by traffic crashes.
[see Commentary 2]
1.6.3 Design for Risk Reduction
The systematic approach to risk reduction in design involves:

• reduction of the inherent risk
• prevention of an incident
• limiting potential for serious injuries and fatalities
• being aware of the risk of vehicle rollover.
Reduce inherent risk
The objective of design is to ensure that the level of roadside risk is as low as practical. While the risk
associated with hazards can be reduced through infrastructure treatments, safety barriers and other safety
devices are also a form of roadside hazard as they can result in significant injuries to the occupants, and can
be particularly severe for errant motorcyclists. Risk cannot be entirely eliminated and so there is always a
‘residual risk’.
Prevent an incident
Prevention of an incident is a key step in risk elimination. For example, preventing the loss of control by
matching horizontal curve radii to the operating speed reduces incident frequency.
Limit potential injuries
If a vehicle leaves the road, the incident should have limited consequences. A prerequisite to the risk
mitigation process is that no element of the road design is a contributing factor to run-off-road incidents. The
use of road safety barriers to reduce impact severity is an example of limiting potential injuries. Drivers on
two-way roads can leave the road on both sides, or may be involved in head-on crashes. Designers should
therefore ensure that roadside hazards on these roads are shielded from impacts from both directions and
the risk of head-on crashes assessed.

The risk of vehicle rollover
Even if rigid hazards (poles, trees, etc.) could be removed from a roadside area that is sufficiently wide to
ensure that vehicle occupants are protected from the likelihood of severe injury, there is still a risk of an
errant vehicle rolling over, particularly if the ground surface is rutted. This is discussed further in
Appendix B.5 in Appendix B.
1.6.4 Appropriate Barriers and Other Treatments
Road design aims to achieve the best practical safety outcomes, using practical and economic measures to
mitigate the severity and likelihood of hazardous incidents.
Most impacts with hazards, or a combination of hazards, can be treated in a number of ways. The choice of
treatment methods should aim to provide a solution to remove, or if this is not practicable, to reduce the
potential for impacts with hazards.
The choice of a barrier should balance the harm to vehicle occupants and the barrier’s ability to contain and
redirect vehicles.
Where there is the potential for a high head-on crash rate on two-lane two-way roads, a road agency may
consider the use of:
• wide centreline treatments
• central barriers
• audio-tactile line-marking (ATLM)
• reduced speed limits.
1.7 Terminology
The Austroads Glossary of Terms (Austroads 2015f) provides a comprehensive listing of terms that relate to
its Guides. However, it is important for practitioners to be familiar with the terms associated with roadside
safety and road safety barriers and for convenience these terms are defined in Appendix A.
1.8 Overview of the Roadside Risk Assessment Process
The process, used in this Part, to evaluate the risk to the occupants of vehicles that leave the road is
different to others used around the world although it is based on similar concepts. This Section provides an
overview of the process, which is described in later sections of this Part.
1.8.1 Network Risk Assessment
The network risk assessment is presented in two forms. The first is to use corridor safety visions, as part of
the Network Safety Plan, to identify typical roadside cross-sections, which are used to develop a common
target level or levels of risk across the network (refer to Sections 2.1 and 2.2). By establishing a common
target, efforts to improve roadside safety will address the road segments with poorer levels of safety. This
target level of risk is termed the Network Roadside Risk Intervention Threshold (NRRIT) and is set by a
jurisdiction for a network independently of project requirements and within the economic capability of the
jurisdiction (refer to Section 2.4).
The second form of network risk assessment is based on accepted national practices and jurisdictional
policies, which relate to the installation of safety barriers and other safety treatments on roads and roadsides
with prescribed geometries and road types. These policies and practices should be developed through a risk
assessment (refer to Section 2.3).

Methods to establish the NRRIT through a ‘Risk Score’ evaluation and potential jurisdictional policies are
described in Section 2.
1.8.2 Assessment of the Road Segment against National Practices, Jurisdictional Policies and the
Corridor Safety Vision
The primary method of mitigating risk due to vehicles leaving travel lanes is to install safety barriers. National
practices and a jurisdiction’s policies describe the locations where the decision to install barriers is
independent of the road and roadside cross-section details and over-ride a further assessment of site risk
refer to Section 2.3).
1.8.3 Road Program or Project Risk Assessment
A road segment in an investment program or project, inherits the NRRIT. It has already been set in the
network strategy and planning phase (see Section 1.2). The NRRIT would have been documented by a
jurisdiction before a roadside or median of the road segment or project is assessed.
The program or project manager should ensure that the NRRIT has been made available from the
jurisdiction’s network strategy and planning group.
The broad process is to evaluate the risk associated with a roadside and if the risk is greater than the
NRRIT, then the road segment or the site should be treated. The process is not concerned with how much
greater the risk is above the NRRIT, only that it is or not. The comparison of the risk for different treatment
options will need to be more detailed, although often the installation of a safety barrier is the only feasible
treatment.
A graphical method to calculate the Risk Score for a roadside cross-section is described in Section 1.9. This
method assists designers to make decisions more quickly and is based on the process described in
Appendix B.
Figure 1.2 illustrates a flow diagram of the processes for both the national risk assessment and the process
to evaluate risk of roadsides at road program and project segments.

Figure 1.2: Risk assessment flow diagram
* An alternative process can be used by a jurisdiction if a network risk assessment is used to determine the application of
treatments.
1.9 Calculating a Risk Score
The Risk Score is used to inform the NRRIT covered in Section 2 and the risk evaluation of roadsides at road
program and project segments is covered in Section 3.
The Risk Score is based on the geometry of the road and the cross-section of the roadside or median. The
procedure for calculating the Risk Score is aligned to the concepts in the Safe System Risk Assessment
Framework (Austroads 2016c). However, rather than using qualitative indices, quantitative risk values are
utilised in the Risk Score procedure.
For a particular cross-section, the procedure evaluates the collective risk of run-off-the-road crashes on the
roadside using an assessment of:
• Exposure: The frequency of vehicles leaving the traffic lane and encroaching onto the shoulder. The
exposure is dependent on the carriageway AADT, the number of lanes, the lane width, terrain type, grade
and curve radii.
• Likelihood: The proportion of drivers that leave the traffic lane and then collide with a roadside hazard.
The likelihood is dependent on the operating speed and the lateral distance to a hazard.
• Severity: The expected severity of a collision as measured by a Trauma Index (refer to Appendix B).
Different hazards have different Trauma Indices.

The Risk Score is the product of the exposure, likelihood and severity and is evaluated using:
• a detailed procedure outlined in Appendix B
• a simplified process outlined below. This process is also based on the procedure in Appendix B.
1.9.1 Overview of the Simplified Process for Calculating the Risk Score for a Roadside
Cross-section
A graphical process used in this simplified approach assumes a number of cross-section dimensions. If the
assumed dimensions are used in the process described in Appendix B, the same values are obtained. To
enable the Risk Score to be plotted against the carriageway AADT, the severity of the impact has been
defined by impacts with significant hazards. An allowance is later made for less significant hazards. The
assumptions in this simplified approach allow a Risk Score to be developed easily, but they do not cover all
possible roadsides.
The risk to errant vehicle occupants increases as the length of road under consideration increases.
Comparable Risk Score measures are obtained if a nominal one kilometre of road is used.
The Risk Score defined above represents the collective (or total) risk to the broader community from
run-off-the-road crashes. The risk to an individual road user is independent of the number of vehicles using
the road and is assessed with some elements of exposure and the product of likelihood and severity (refer to
Appendix B.6.2).
The following sub-sections describe:

• significant hazards
• less significant hazards
• measuring the lateral distance to a hazard
• operating and design speed
• terrain type definitions
• grade and curve radii factors
• background and isolated hazards
• Risk Score charts for undivided rural roads
• Risk Score charts for divided urban roads
• Risk Score charts for roadside barriers
• the use of the process in Appendix B.
• hazards for motorcyclists and other vulnerable road users
• hazards for heavy vehicle occupants.
1.9.2 Description of Significant Hazards
The risk evaluation process described in Appendix B, uses a ‘Trauma Index’ to define the likely outcome of
collisions with hazards. Collisions with hazards with a higher Trauma Index are more severe. Significant
hazards for the purpose of evaluating the NRRIT have a Trauma Index of approximately 6 and higher and
include:
• tree lined edges
• isolated trees
• utility poles (excluding slip base poles and energy absorbing poles)
• fixed base lighting columns

• traffic signal poles

• rocks protruding more than 300 mm above the ground surface
• ruts in the ground surface more than 300 mm deep
• rock cuttings
• 2:1 fill batters more than 5 m high
• 1.5:1 fill batters more than 2 m high
• vertical drops of more than 2 m
• water courses more than 2 m deep
• down slopes, parallel to the road, higher than 5 m and with a slope of 4:1 or steeper
• down slopes, parallel to the road, between than 2 m and 5 m high and with a slope of 2:1 or steeper
• up slopes, parallel to the road, 1.5:1 or steeper
• significant drainage structures, exposed culvert headwalls and wing walls, with vertical height drops of
more than 2 m or a watercourse more than 2 m deep.
Embankment slopes are specified as the ratio of the horizontal to the vertical dimension (that is H:V).
Other urban significant hazards include control boxes for traffic lights, brick pillars, large veranda posts and
so on.
For roads designed for vehicles travelling at 110 km/h, a tree lined edge, a utility pole and an isolated tree
have Trauma Indices of 6.0, 6.4 and 6.7, respectively. These hazards were all considered to have a similar
impact on crash severity outcome and have Trauma Indices of approximately 6 and were the basis of
selecting the Trauma Index of 6 as a measure of significant hazards. Roads are designed to suit the speed
of the vehicles using the road.
This list of significant hazards is not exhaustive and may be added to. The NRRIT is based on these
significant hazards to define the risk. It is most likely that a corridor, without safety barriers, will have
examples of these hazards.
1.9.3 Description of Less Significant Hazards
The evaluation of roadside risk is generally controlled by the presence of these significant hazards. However,
at some sites these hazards may not be present and other hazards are classified as 'less significant’ hazards
or ‘minor’ hazards. The hazards are classified based on their Trauma Indices listed in Appendix B. Again,
these lists are not exhaustive and other hazards may be included as knowledge is obtained.
Collisions with the following hazards are likely to be less serious. These hazards have a Trauma Index of
between 1.5 and 4. (Refer to Appendix B for more details about the Trauma Index and their values.):
• slip base or energy absorbing poles
• 2:1 fill batters between 2 m and 5 m high
• 1.5:1 fill batters less than 2 m high
• rocks protruding between 200 mm and 300 mm above the ground surface
• ruts in the ground surface between 200 and 300 mm deep
• watercourses less than 1 m deep and with a drop between 0.3 m and 2 m
• down slopes, parallel to the road with slopes of 3:1 and between 2 m and 5 m high
• down slopes, parallel to the road with slopes of 4:1 and between 2 m and 5 m high

• down slopes, parallel to the road with slopes of 6:1 or flatter and higher than 2 m
• up slopes, parallel to the road, with slopes steeper than 4:1 but flatter than 1.5:1.
Sections 1.9.8, 1.9.10 and 1.9.11 contain the Risk Score charts based on the impact severity having a
Trauma Index of 6. Conservatively, the Trauma Index of these less significant hazards can be considered to
be 3. If these less significant hazards are present rather than a significant hazard, then the graphed Risk
Score is reduced by half.
The following hazards are classed as being ‘minor hazards’. The severity of impacts with these hazards is
demonstrated by their Trauma Index being less than 1.5. Impacts with these minor hazards are more severe
than impacts with safety barriers:
• 3:1 fill batters between 2 m and 5 m high
• 2:1 fill batters less than 2 m high
• watercourses less than 1 m deep and with a drop of between 0.3 m and 2.0 m
• down slopes, parallel to the road, less than 2 m high
• up slopes, parallel to the road, with slopes of 4:1 or flatter.
Conservatively, the Trauma Index for these less significant hazards can be considered to be 1. If only these
minor hazards are present on a roadside then the graphed Risk Score in Sections 1.9.8, 1.9.10 and 1.9.11 is
divided by 6, as the charts are based on a hazard Trauma Index of 6.
1.9.4 Measuring the Lateral Distance to a Hazard
The effective lateral distance (or offset) from the road to a hazard is measured as shown Figure 1.3. If there
is an embankment slope 6:1 or flatter, then the width of the embankment is included in the measure as car
drivers are generally able to regain control and trucks can traverse these embankments. Drivers who
proceed down slopes that are steeper than 6:1 are more likely to collide with hazards on the embankment or
at the toe of the embankment.
If the embankment slope is steeper than 6:1, but not steeper than 4:1, the lateral distance is the distance to
the significant hazard excluding the width of the embankment.
Embankment slopes steeper than 4:1 are considered to be a hazard in their own right and the area beyond
the embankment is not included in the lateral distance measure. There is an increased likelihood of vehicles
rolling over on steeper embankments.
In most cases, additional hazards on the embankment and beyond further define the crash severity.
It is important that the designer look beyond the right of way boundary as hazards in abutting properties
might also constitute a significant hazard to motorists and to the broader community. There is no specific
guidance on how far beyond the right of way should be included in the area of interest as this can depend on
the type of risk being considered.

Figure 1.3: Calculation of an effective lateral distance to a hazard
1.9.5 Operating and Design Speed
Roads are designed to suit the speed of the vehicles using the road. The operating speed of roads is set at
approximately the 85th percentile speed, which is usually 10 km/h faster than the speed limit in rural areas
and equal to the speed limit on urban roads. A jurisdiction could adopt other relationships between operating
speed, design speed and the speed limit (refer to Austroads 2016a).
Roadside risk assessments centre on the management of energy. Accordingly, the term operating speed is
used in this Part to represent the 85th percentile speed and to act as an indicator for the kinetic energy of
errant vehicles. For example, an operating speed of 110 km/h generally relates to high speed roads,
highways and freeways with a speed limit of 100 km/h. Other operating speeds used in the procedure are 70
and 90 km/h in rural areas and 70, 80 and 90 km/h for urban roads.
1.9.6 Terrain Type
Different terrain types have not been defined by the procedure in Ray et al. (2012b), which is the basis of this
procedure, and is generally not defined in other road design guidelines. The definitions have been
subjective. Giummarra (2001) has defined different terrain types as follows:
Flat terrain – based on 0–10 five-meter contours per kilometre. Roads generally follow the
ground contours. Typical gradients being up to 1: 20 or 5% (3 degrees).
Rolling terrain – based on 11–25 five-meter ground contours per kilometre. Roads can have
substantial cuts and fills. Typical gradients being up to 1: 20 to 1:8 or 5–12.5% (3–7 degrees).
Mountainous terrain – based on more than 25 five-meter contours per kilometre. Roads in
rugged hilly or mountainous ground with substantial restrictions to both horizontal and vertical
alignments. Typical gradients being > 1:8 or > 12.5% (7 degrees).
These definitions could be used if required.

1.9.7 Background and Isolated Hazards
Many roadsides have hazards at various offsets from the road. Often these include trees near property
boundaries and additional trees closer to the road. The left side of the rural road illustrated in Figure 1.4,
shows trees in approximately two lines. The trees closer to the property boundary could be considered to be
the background hazards and the trees closer to the road considered to be isolated hazards. It is assumed
errant vehicles that do not collide with the isolated hazards, will collide with the background hazards.
Figure 1.4: View of road showing trees at different offsets
The analysis of risk should include an analysis of the background hazards and the additional risk of the
isolated hazards in front.
The background hazards are considered to be consistent and continuous. If the background hazards are
trees, then they should be less than 20 m apart to be classified as continuous. The background hazards
could be any other continuous hazard. If the isolated hazards are fairly close together and a relatively
consistent offset from the traffic lane, they should be considered to be background hazards and there would
be no additional isolated hazards.
Some roads in the interior of the country, may have no property boundary fences, no consistent background
hazards and few trees or other significant hazards near the road. Then the Risk Score for the background
hazards can be assumed to be related to the risk of vehicles rolling over.
The process is similar in urban areas. Figure 1.5 shows a fence on the property boundary and trees or other
hazards in front. The fence is the background hazard and the trees are isolated hazards.

Figure 1.5: View of an urban road showing background and isolated hazards
1.9.8 Risk Score Charts for Undivided Rural Roads
For rural roads, with significant hazards listed in Section 1.9.2, the risk is dependent on the traffic volume
and the lateral offset to these hazards from the edge of the lane. The offset of the hazards is measured in
accordance with Section 1.9.4.
The charts for the rural roads are copied in Appendix D and grouped together so that pages may be copied
and annotated to assist a program or project manager. The axes in the charts generally have the same
scales. In these figures, the chart number is shown in the top left corner.
The terrain influences the geometric design of the road. Accordingly, the mountainous and rolling terrain
types are also applicable to roads with more restricted sight distances, narrower pavements, tighter curves
and so on. Roadsides in rolling terrain have a slightly greater exposure than those for mountainous terrain,
all other dimensions being the same. Consequently, roads in mountainous or rolling terrain have similar Risk
Scores. In this simplified approach, either ‘rolling’ or ‘mountainous’ terrain will be used and can be applied to
both terrain types.
The hazards are considered to be ‘continuous’ if trees, poles and the like have an average spacing less than
20 m.
The charts in this Section are considered to be representative and conservative but do not cover all
possibilities.
Rural roads with background hazards and with no isolated hazards in front
Figure 1.6 illustrates the Risk Score for rural roads with an operating speed (and design speed) of 70 km/h
(in mountainous terrain), and 90 and 110 km/h in flat and rolling terrain. Table 1.1 lists the road
characteristics for each road type.
It is appreciated that Figure 1.6 does not cover all rural situations, but the charts are sufficient for most
situations. The process in Appendix B could be used for other situations.
The risk of the background hazards includes the risk of rollovers.

Figure 1.6: Risk Score for undivided rural roads with significant background hazards at different offsets and
for different operating speeds and road types
Table 1.1: Road characteristics for the charts in Figure 1.6
Chart
Chart description Terrain Operating speed Lane width
number
1 Rural mountainous 70 km/h Mountainous 70 km/h 3.0 m
2 Rural 90 km/h Rolling 90 km/h 3.3 m
4 Rural 90 km/h Flat 90 km/h 3.5 m
Note: There are no isolated hazards.

Rural roads with background hazards and with isolated hazards in front
When there are isolated trees or other significant hazards in front of a continuous line of significant
background hazards, specified in Section 1.9.7, the isolated hazards in front shield the background hazards
behind and so reduce the risk of vehicles colliding with them. The combined risk of background and isolated
hazards increases with an increasing number of isolated hazards. The Risk Score is given in Figure 1.7 for
rural roads with an operating speed of 70, 90 or 110 km/h and in different terrains. The Risk Score is
evaluated over 1 km. Table 1.2 lists the road characteristics for each road type represented by the charts in
Figure 1.7.
Figure 1.7: Risk Score for undivided rural roads with significant isolated hazards at different average spacings
along a roadside and for different operating speeds and evaluated over 1 km
Note: Refer to Table 1.2 for road and hazard characteristics and dimensions. The road segments are straight or winding
without sharp curves.

Table 1.2: Rural road characteristics for the Risk Score charts in Figure 1.7
Chart Operating Lane Background Isolated

Chart description Terrain
number speed width hazards at hazards at
6 Rural mountainous 70 km/h Mountainous 70 km/h 3.0 m 3m 1m
7 Rural 90 km/h Rolling 90 km/h 3.3 m 3m 1m
9 Rural 90 km/h Flat 90 km/h 3.5 m 4m 2m
These charts give the Risk Scores for both the background hazards and the isolated hazards in front, and
account for any shielding by the isolated hazards.
Rural roads without background hazards but isolated hazards
Figure 1.8 shows the risk of isolated significant hazards with an average spacing of 50 m and 200 m. The
road is undivided in flat terrain and has two 3.5 m lanes and there are no background hazards. The operating
speed for these roads is 110 km/h. The isolated hazards are at various distances from the edge of the traffic
lane.
If there are no background hazards, then there is still a risk of vehicles rolling over. Figure 1.9 illustrates this
risk. Note that the Risk Score axis has a difference scale. The charts in this figure particularly apply to roads,
in the western and central areas of Australia that have few significant hazards. The risk of a rollover (from
Figure 1.9) is added to the risk for isolated hazards (from Figure 1.8).
Figure 1.8: Risk Score for isolated significant hazards on undivided rural roads without background hazards
Note: Lanes are 3.5 m wide. The terrain is flat. The operating speed is 110 km/h.

Figure 1.9: Rollover Risk Score for undivided rural roads without background hazards and an operating speed
of 110 km/h
Note: Lanes are 3.5 m wide. The terrain is flat. The operating speed is 110 km/h. The Risk Score axis has a different
scale to other charts.
1.9.9 Grade and Curve Radii Factors for Rural Roads Risk Score Values
Figure 1.6, Figure 1.7 and Figure 1.8 refer to the straighter segments of road with upgrades or slight
downgrades. The procedure in Appendix B accounts for the gradient and the curve radii. The limit of curve
radii that do not increase the Risk Score depends on the operating speed. Risk Score adjustment factors for
roads with 300 m, 600 m or 900 m radius curves are given in Table 1.3. This table gives the Risk Score
adjustment factors for significant hazards being either 3 m or 6 m from the road. The factors in Table 1.3 are
used to increase the Risk Score from Figure 1.6, Figure 1.7 and Figure 1.8. These factors are applied to the
analysis of the risk for background hazards and the risk of background and isolated hazards.
Table 1.3: Risk Score adjustment factors for curve radii on two-lane rural roads
Operating speed
70 km/h 90 km/h 110 km/h

Curve
Location
radius Hazard location from the road •
3m 6m 3m 6m 3m 6m
300 m Inside curve 2.6 3.4 2.6 3.2 2.4 2.8

Outside curve 5.3 6.8 5.1 6.5 4.7 5.5
600 m Inside curve 1.2 1.5 1.1 1.3 1.2 1.3
Outside curve 1.2 1.5 1.1 1.3 1.2 1.3
900 m Inside curve 1.1 1.2 1.1 1.3 1.1 1.2
Outside curve 1.1 1.2 1.1 1.3 1.1 1.2
Note: This table can also be applied to higher speed urban divided roads. If applied to undivided urban roads, then the
designer should acknowledge that using this table is an extrapolation of the data.
The hazard location and the curve radii, listed in Table 1.3, are the values that were used to evaluate the
numbers in the body of the table. Users should use the hazard location and the curve radii that are closest to
the site characteristics and not interpolate the numbers. If more detail is required then the process in
Appendix B should be used.

Table 1.4 lists an appropriate adjustment factor for downgrades. The procedure in Appendix B may need to
be consulted if the gradients and curve radii are significantly different. Both a curve radii factor and a grade
factor should be applied in the one analysis.
Table 1.4: Risk Score adjustment factors for gradient of two-lane rural roads
Gradient Upgrade 2% downgrade 4% downgrade 6% downgrade

Factor 1.0 1.0 1.5 2.0
Note: This table can also be applied to higher speed urban divided roads. If applied to undivided urban roads, then the
designer should acknowledge that using this table is an extrapolation of the data.
1.9.10 Risk Score Charts for Divided Urban Roads
A Risk Score for divided urban roads is evaluated using two different scenarios. The first is when there are
closely spaced (less than 20 m apart) significant hazards along the road such as large trees, light poles,
utility poles, rock walls and so on. These create a continuous background hazard. In this case the risk is
dependent on the traffic volume and the lateral offset to these hazards. The arterial roads have an operating
speed of 70, 80 or 90 km/h (see Figure 1.10).
The second scenario is when there is a fence or a property wall 4 m from the edge of the road and isolated
significant hazards at various spacings that are 1 m from the edge of the traffic lane. Again, the arterial roads
have an operating speed of 70, 80 and 90 km/h. There is an increased risk when the average spacing
between significant hazards decreases. If these significant hazards are further from the edge of the traffic
lane then the risk is reduced. The results presented in Figure 1.11 are considered to be representative of
most situations and to be conservative.
For Figure 1.10 and Figure 1.11, it was assumed that there were two 3.5 m wide lanes in each direction and
the terrain is flat. For Figure 1.11, it is assumed that there is a generic wall 4 m from the edge of the road
(creating a background hazard line) and isolated trees 1 m from the edge of the road in front of the wall.
The adjustment factors in Table 1.3 and Table 1.4 can be applied to divided urban roads. The AADT is the
expected traffic volume for the period of the corridor safety vision.
Figure 1.10: Risk Score for divided urban roads with significant hazards at different offsets and for different
operating speeds (Scenario 1)
Note: There are no isolated hazards. The background hazards are at various distances from the edge of the traffic lane.
The terrain is flat.

Figure 1.11: Risk Score for divided urban roads with significant hazards at different average spacings along a
roadside and for different operating speeds (Scenario 2)
Note: There are isolated hazards 1 m from the edge of the traffic lane and a generic wall is the background hazard and is
4 m from the edge of the traffic lane. The terrain is flat.
Charts 20 to 22 in Figure D.7 in Appendix D show the Risk Score for undivided urban roads with operating
speeds of 50, 60 and 70 km/h. It must be noted that these charts relate risk to occupants in vehicles that run
off the road. This does not represent the complete risk to road users and the charts should be seen in this
context. Designers using the charts in Figure D.7 should be aware that they are an extrapolation of the
procedure. Although the adjustment factors in Table 1.3 and Table 1.4 have not been calibrated for these
undivided urban roads, they can be used cautiously. These charts are provided in Appendix D for
information.
These charts give the Risk Scores for both the background hazards and the isolated hazards in front, and
account for any shielding by the isolated hazards.
1.9.11 Risk Score Charts for Roadside Safety Barriers
The Risk Score for roadside barriers on undivided rural roads is given in Figure 1.12. The barrier offsets
used in these Risk Score charts are given in Table 1.5. Figure 1.13 gives the Risk Score for divided urban
roads with two barrier offsets. The 0.5 m offset corresponds to a barrier being placed at the kerb line and the
3.5 m offset relates to a parking area or cycle lane adjacent to the running lane. Figure 1.14 applies to both
urban and rural divided carriageway roads with 110 km/h vehicle operating speeds. The user should note
that some of these charts are drawn to different scales to other charts in this section.

Figure 1.12: Risk Score for roadside barriers on undivided rural roads with different operating speeds and
terrain (refer to Table 1.5 for road details)
Chart number Rural or urban Terrain Operating speed Lane width Offset of the barrier
23 Rural undivided Flat/rolling 70 km/h 3.0 m 1.0 m*
* The barrier offsets used produce conservative (higher) Risk Scores. If barriers were located further from the road, the
Risk Score would be smaller.
Figure 1.13: Risk Score for roadside barriers on divided carriageway urban roads with different operating
speeds and barrier offsets (refer to Table 1.6 for road details)

Chart number Rural or urban Terrain Operating speed Lane width Offset of the barrier*
26 Urban divided Flat 70 km/h 3.0 m 0.5 and 3.5 m
Note:* The barrier offsets used produce conservative (higher) Risk Scores. If barriers were located further from the road,
the Risk Score would be smaller.
Figure 1.14: Risk Score for roadside barriers on divided carriageway urban and rural roads with an operating
speed of 110 km/h
Note: The barrier offset of 2.0 m is used. This produces conservative (higher) Risk Scores. If barriers were located
further from the road the Risk Score would be smaller.
1.9.12 When should the process in Appendix B be used?
The process in Appendix B should be used if the designer is concerned that the road geometry and its
characteristics assumed for each chart do not reflect the study site. For instance, divided rural roads should
be analysed using the process in Appendix B if they are not addressed by national or jurisdictional policies
described in Section 2.3.
For cases where a design exception is required or when the design is outside the normal design
domain (NDD), the designer should use the method described in Appendix B to support the assessment of risk.
The process in this section does not cover the risk associated with oncoming vehicles on two-lane rural
roads and this should be analysed using the process in Appendix B.
1.9.13 Hazards for Motorcyclists and Other Vulnerable Road Users
In this Part, the following objects are not considered to be hazardous fixed objects for vehicle occupants:
• small-size steel and timber sign support posts that comply with AS 1742.2-2009
• slip-base poles and frangible posts
• objects located beyond the deflection area of a safety barrier
• trees less than 70–100 mm diameter (depending on the species) at their base
• tubular thin-walled traffic signal posts at urban intersections.

While these objects may not be hazardous to vehicle occupants, they are likely to be hazardous for errant
motorcyclists and so consideration should be given to impacts by all road users.
Roads that have a high number of motorcyclists will need to be evaluated using a different process to that
described in Section 2.5 and Section 3.3. At this time, the process for deciding where to install barriers and
motorcyclist protection devices has not been well documented and it is often managed through jurisdictional
policies and practices.
1.9.14 Hazards for Heavy Vehicle Occupants
Many objects that are hazardous to the occupants of a car are not as hazardous to heavy vehicles due to
their greater mass and stiffer construction. There is insufficient evidence to classify the outcomes of impacts
by heavy vehicles for different hazards.
Carrigan, Ray and Johnson (2014) reported that the exposure measure for heavy vehicles is approximately
30% of the value for all vehicles. The severity of impacts for the occupants of heavy vehicles are also likely
to be less than for occupants of passenger vehicles. Consequently, the risk to heavy vehicle occupants is
also lower. When heavy vehicles leave the road, the risk to others is often increased.

2. Network Risk Assessment
2.1 General
As discussed in Section 1.5, a primary objective of road design is to provide drivers with a consistent view of
the road and roadsides through self-explaining roads (refer to Guide to Road Design Part 1). Associated with
this aspect, the level of protection to errant drivers should also be consistent and at a level of risk
commensurate with the road’s function in the hierarchy, the community requirements and aspirations and the
potential funding for the road in the intermediate and longer terms.
A safety assessment of the network should be undertaken before evaluating the safety of individual road
segments. The risk of roadsides across a network is informed by corridor safety visions and documented
with the Network Roadside Risk Intervention Threshold (NRRIT) for run-off-the-road crashes.
These concepts and processes are new to this edition of Part 6 and are discussed below. This Section
describes the process to set the NRRIT from the corridor safety vision and its consistent application for the
assessment of road projects and sites discussed in Section 3. While this efficient process is recommended, a
jurisdiction is able to use an alternative procedure if a network risk assessment is used to determine the
application of treatments (also see the flow diagram in Figure 1.2.
2.2 Corridor Safety Visions
A corridor safety vision is part of a Network Safety Plan as required by Action A of the National Road Safety
Action Plan 2018-2020 (Commonwealth of Australia 2018). These corridor safety visions are a jurisdiction’s
policy for roadside treatments and are based on a targeted level of risk defined by the iRAP star ratings and
the expected number of fatal and serious injuries (FSIs) per kilometre per year. A corridor safety vision is a
broad clear statement of the vision of a network or a corridor to create the safest road network with the
available resources.
The corridor safety vision is a high-level document that should not be static. A road agency should monitor
and update it from time to time. The corridor safety vision allows for a consistent network level of safety
provided by a road and consequently a roadside. Adopting a more consistent level of safety will allow the
expenditure on roadside safety improvements to be more efficiently utilised.
The corridor safety vision enables a generalised targeted level of safety to be identified. Road stereotype
tables, Austroads Guide to Road Design Part 2 (Austroads 2019), and other comprehensive safety analysis
tools provide a means to identify a targeted level of safety and the broad characteristics of the road and
roadside geometry to achieve this target.
The decision process to define the targeted level of safety for a particular road is not described here and is a
function of the road network planning process. It is likely that the process will be different for different
jurisdictions.
Until a Network Safety Plan has been prepared, a jurisdiction must define an appropriate roadside
cross-section that would reflect the requirements of the user, the community and resources for particular
segments of road. This road segment would then be used to define a NRRIT using the procedures in
Section 2.4.
A corridor safety vision may also document safety treatments for a particular corridor. These treatments are
restricted to specific corridors and are not applied across a jurisdiction.
2.2.1 Adverse Crash History
An adverse crash history should be used to re-evaluate the corridor safety vision for the road or network.

Many factors may need to be examined as possible contributory causes, even when they are not the primary
cause of a crash, as they may indicate that treatments other than road safety barriers are appropriate at
sites. Detailed guidance on investigating crash locations, diagnosing crash problems and developing
solutions is contained in the Guide to Road Safety Part 8: Treatment of Crash Locations (Austroads 2015d).
2.3 Treatment of Roads Based on Policies and Practices
The primary method of mitigating risk due to vehicles leaving travel lanes is to install roadside barriers. This
section describes the locations where the decision to install barriers is independent of the detailed road and
roadside cross-section details and are based on over-riding national practices, jurisdictional policy or corridor
safety vision.
2.3.1 Treatment of Roads Based on National Practices
Barriers should be installed on the following greenfield road segments regardless of the roadside or median
dimensions:
• on verges and in medians of rural freeways with a speed limit of 100 km/h or higher and with access
control and divided carriageways
• on verges and in medians of urban freeways with a speed limit of 100 km/h or higher and with access
control and divided carriageways
• in medians on rural highways with divided carriageways with a speed limit of 100 km/h or higher.
The run-off-the-road risk associated with these road types would not require further evaluation.
An exception report should be required to justify if it is proposed that a barrier be not installed in the median
or verge of a greenfield rural highway or a greenfield freeway specified above.
2.3.2 Treatment of Roads Based on Jurisdictional Policies
A jurisdictional policy may include:

• the installation of barriers to shield and to protect high consequence infrastructure that cannot feasibly be
removed or relocated and if impacted by an errant vehicle has high community costs, for instance,
barriers to shield rail infrastructure
• the installation of barriers to shield and to protect high consequence land that is not a public utility and is
located outside of the road corridor, for instance, some school yards, child care centres and fuel storage
facilities
• the installation of continuous barriers
• the installation of wide centreline treatments with or without a barrier
• the installation of barriers or barrier types for defined road types and locations
• the installation of other safety treatments including audio-tactile line-marking on rural roads and the
sealing of shoulders
• the installation of barriers to shield sites that present a high real or perceived risk to motorists.
A corridor safety vision can also list requirements for installation of barriers and other safety devices. These
are considered along with the jurisdictional policies.
Further details of potential jurisdictional policies are provided in Appendix C.

2.4 The Network Roadside Risk Intervention Threshold (NRRIT)
A description of generalised typical target cross-sections is developed from the corridor safety visions
outlined in Section 2.2 for all crash types. These typical cross-sections can be on existing roads or can be
desired cross-sections. Cross-sections on existing roads will usually relate to straight road segments with
level or a slight uphill gradient. Their description should not include sharper curves, steep down grades,
isolated hazards in front of continuous hazards unless these define a typical cross-section for a road. These
typical cross-sections would be expected to apply over a number of road segments.
Typical cross-sections for the evaluation of the NRRIT will usually be from higher volume two-lane single
carriageway rural roads, and higher volume urban roads. Divided carriageway rural roads could be included;
however, these roads generally have better geometries and the level of safety may not be typical of the
greater network.
The risk of run-off-the-road crashes is evaluated by Risk Scores for the generalised cross-sections on a
number of different road segments and types and these Risk Scores then inform the NRRIT for these road
types. The procedure to calculate these Risk Scores is described in Section 2.5.
Different jurisdictions may have different corridor safety visions for similar roads. One jurisdiction’s corridor
safety vision may be to install continuous barriers 2.0 m from the traffic lane on two-lane single carriageway
rural roads and this cross-section will inform the NRRIT. Another jurisdiction may consider that, given their
level of resources and traffic volumes, a cross-section with trees no closer than 3 m from the road is its
corridor safety vision. (Different projected traffic volumes or different financial commitment to the two
corridors may have affected the different decisions).
The extent of the network reviewed to evaluate a NRRIT may include all corridors or specific targeted
corridors. While the NRRIT will be set for a remediation timeframe of probably 10 years, it is expected that
successive reviews of the NRRITs for a network should have lower Risk Score values as the network
improves. In cases where the NRRIT has not been set at a sustainable level, the NRRIT may be raised so as
to target the highest risk locations through project delivery.
The NRRIT can apply to the rural zones of a regional network to align with regional resource allocation. The
NRRIT should be applied over the region where resource and funding equitability is to be applied.
The NRRIT is standardised by evaluating the risk over a standard 1 km of roadside.
2.4.1 Application to Greenfield and Brownfield Sites
The NRRIT determines when designers need to intervene to mitigate the risk from a roadside in a greenfield
or brownfield site. For greenfield sites, the objective is to have the Risk Score below the intervention level
and as low as possible within economic constraints.
The notion of the NRRIT is to identify road segments where a roadside treatment is required. This concept
can also be applied to the situation where additional hazards, for instance additional trees, are to be installed
beside a road. In this case a jurisdiction could select another threshold level, so that the risk with the
installation of additional hazards does not extend above this lower threshold. In effect, Risk Scores can be
used to establish policies both for the mitigation of existing risk and for the introduction of additional risks. It
is a jurisdiction’s decision whether to apply the Risk Scores in this way.
2.4.2 Establishing the NRRIT
The NRRIT is informed by the Risk Score associated with typical cross-sections. From these typical
cross-sections, the collective risk of run-off-the-road crashes on the roadside are defined by a Risk Score
evaluated using:
• a simplified process described in Section 1.9
• a detailed procedure outlined in Appendix B.

There are a number of methods which can be used to help select these typical cross-sections and,
consequently, the NRRIT. These methods allow a jurisdiction to understand the upper and lower levels of the
NRRIT, and whether different NRRITs are appropriate for different environments (such as urban or rural).
The essential requirement is that typical cross-sections are identified so that Risk Scores can be evaluated
for the roadsides.
Potential methods to establish a NRRIT include:

• Network Aspirational Level Method
This method starts by defining the roadside with a low risk aspirational level to make significant road
safety gains and resolves the commitments and constraints needed to achieve this level.
• Jurisdictional Constrained Method
This method relies on a list of the lengths of road (in km) with different scenarios defined by traffic volume
ranges, hazard type and offset of significant hazards (defined in Section 1.9.2) and the road environment.
Each scenario is a broad description of a typical roadside and can be used to evaluate a Risk Score using
the process in Section 1.9 or the charts in Appendix D. An initial NRRIT is chosen and using unit rates for
treatments, the cost to reduce the risk level (defined by the Risk Score for each scenario) to be below the
initial NRRIT is evaluated. A number of iterations may be necessary to raise or lower the NRRIT to match
jurisdictional resources and constraints.
• Historic Treatment Level and Predominant Environments
This option identifies a risk level which has been attained using predominant historic treatments. For
example, a common treatment has been the use of a 9 m clear area alongside the road in flat terrain.
Again, this allows a typical cross-section to be identified. This level can then be checked for other
predominant road types to gauge the level of acceptance. While this option is not recommended it would
provide a short-term interim basis for consequential decision-making as part of projects and programs.
In any case, it is important to select the typical cross-sections that reflect the roadside safety described in the
corridor safety vision or jurisdictional policy statement.
2.4.3 Setting a NRRIT Based on Two-lane Undivided Rural Roads
The graphical Risk Score charts provided in Section 1.9.8 are applicable to rural undivided roads. As the
typical cross-sections on rural roads would generally not have isolated hazards, charts in Figure 1.6 are
more applicable. Using the design AADT, the operating speed-terrain combination and the offset of
significant hazards (listed in Section 1.9.2) from the road, the user can identify the Risk Score for a particular
cross-section. The Risk Score should be modified when there are frequent curves or grades extending over
the road segment. These factors are described in Section 1.9.9.
The NRRIT is a composition of the Risk Scores for a number of typical cross-sections.
2.4.4 Setting a NRRIT Based on Urban Roads
A Risk Score, to define the NRRIT for divided urban roads, is evaluated in the charts in Section 1.9.10.
A single NRRIT could be applied to urban roads with different functions. This threshold value is more usefully
evaluated using the characteristics of road segments with higher operating speeds or traffic volumes (AADT).
It is appreciated that figures in Section 1.9.10 do not cover all urban situations, but they are sufficient for
most. Generally, there is little additional value in trying to describe the cross-section more accurately, as the
results do not vary widely. The process in Appendix B could be used for other situations.
When setting the NRRIT, it is useful to consider its implications across urban and rural roads with different
cross-sectional characteristics. Appendix D provides the Risk Score charts in one location to assist.

2.5 Example of Setting a NRRIT
In this example, it is assumed that a local government jurisdiction has a network of two-lane single
carriageway rural and single and multiple carriageway urban roads. The NRRIT is the Risk Score typically
evaluated with using both rural and urban roads carrying higher traffic volumes.
A typical roadside on a two-lane two-way rural road corridor has a significant number of trees that are within
a lateral distance of 3 to 5 m from the road. The jurisdiction is seeking to improve the level of safety, but
within the expected funding levels. Using the jurisdictional constrained method, the NRRIT is the Risk Score
associated with a rural road, in flat terrain with an operating speed of 110 km/h, an AADT of 2000 veh/day
and with trees and other similar significant hazards that are 5 m laterally from the road. Using Figure 1.6 the
Risk Score for this road is 1.6. The jurisdiction may decide that the NRRIT is this Risk Score and that it can
be applied to this and other similar existing rural corridors. This is illustrated as point A in Figure 2.1, which is
a copy of Figure 1.6. The choice of the NRRIT value is a jurisdiction’s decision.
Other two-lane rural roads in the jurisdiction have an operating speed of 90 km/h, and with less than
500 veh/day and with trees approximately 1 m from the road. The Risk Score for these roads is less than 0.6
as illustrated by point B in Figure 2.1.
Figure 2.1: Estimates of the NRRIT using lower section Figure 1.6
The jurisdiction has developed a policy that significant road hazards should be more than 2 m from the
edge of a divided urban road based on the research by Dixon et al. (2008). The Risk Score of these urban
roadsides with high traffic volumes (greater than 12 000 veh/day) and an operating speed of 80 km/h is 1.5.
See point C in Figure 2.2.

Figure 2.2: Estimates of the NRRIT using Figure 1.10
For a more equitable use of safety funds, a single NRRIT should be used across both urban and rural roads.
Consequently, using these examples, the NRRIT of 1.5 should be selected for the network of urban and rural
roads, in this example.
Hazards should not be unnecessarily introduced into the roadside. However, there is always pressure to
plant additional trees in urban areas. The increase in risk should be as small as possible. This process can
also be used to define whether the level of risk after the introduction of additional hazards (like additional
trees) is tolerable. Using the same example, if hazards on the urban divided road are, on average, 50 m
apart and within 1.0 m from the edge of the lane, then the Risk Score would be 0.9. See point D in Figure
2.3. If this urban roadside was seen to be the maximum acceptable level of risk when introducing new
hazards, then this Risk Score would not be a NRRIT, but used to identify unacceptable risk for introduced
hazards. A policy statement could be written so that the Risk Score of a road is compared to the maximum
tolerable Risk Score for introduced hazards (0.9 in this example). If the roadside Risk Score is over 0.9 with
additional hazards, then they should not be allowed to be planted or installed.
Figure 2.3: Estimates of the NRRIT using Figure 1.11

As a guide, using a NRRIT of 2.0 gives roadside designs that are consistent with the current level of risk on
long lengths of network in reasonably flat terrain where barriers have not been installed. On some
mountainous roads, the currently acceptable roadsides would have higher Risk Scores and a jurisdiction
may need to adopt a higher NRRIT on mountainous sections in the first instance.
The use of a NRRIT between 1.5 and 2.0 is considered to be reasonable initially, until better information, is
obtained from the corridor safety vision and associated acceptable roadside cross-sections. A jurisdiction is
free to choose any NRRIT.
Similar charts can be obtained for narrower urban roads with lower operating speeds, but these would be
unlikely to define a NRRIT (for run-off-the-road crashes) that could be applied across a wide road network.
Appendix D provides these charts to allow Risk Score comparisons to be made.

3. Program and Project Risk Assessment
3.1 Overview of the Risk Evaluation Process
The hierarchy of control is to first remove a hazard, then treat a hazard that may result in a crash or
contribute to the severity of a crash and finally to install a safety barrier to shield a hazard. Safer roadsides
often involve the installation of a safety barrier. Regardless, there is a level of risk with any roadside. The
objective is to minimise the risk to below the Network Roadside Risk Intervention Threshold (NRRIT) at both
greenfield and brownfield sites. The NRRIT needs to be determined before assessing the risk at particular
road projects and other sites. Refer to Section 2 for the process to establish the NRRIT.
Risk Scores are used to evaluate the risk of roadsides and medians of individual road segments for
run-off-the-road crashes. The graphical process described in Section 1.9 is used to calculate Risk Scores
and is the same as the one used to inform the NRRIT described in Section 2.4. This process is applicable to
rural roads and highways and urban roads with higher flows. The process is applied to roadside design in
greenfield and brownfield sites.
This Section describes:

• the concepts used in evaluating the risk at particular sites (Section 3.2)
• the process to determine whether risk mitigating treatments are required at a particular site. This process
has the following four steps.
– Step 1: Assess the road against national and jurisdictional policies. A barrier or other treatment should
be installed on road segments described by a jurisdiction’s policies, which could also include safety
treatments specified in the corridor safety vision. These safety treatments are not evaluated further
with steps 2 and 3 (Section 3.3).
– Step 2: Compare the Risk Score with the NRRIT. The Risk Score (evaluated from charts), which
represents the risk to errant vehicles on the existing roadside, is compared with the NRRIT. If the Risk
Score exceeds the NRRIT, then a roadside treatment is required to mitigate the risk to be below the
NRRIT. Section 3.2 describes the concepts for this process. Section 3.4 describes the process of
establishing a Risk Score for a particular road segment and Section 3.4.2 gives examples.
– Step 3: Identify, evaluate and rank risk mitigation options. The options may include installing a safety
barrier, sealing shoulders and installing ATLMs, or providing a hazard-free area alongside the road
(Section 3.5).
– Step 4: Design the recommended roadside treatments (Section 3.6).
Also see Figure 1.2 for a flow diagram of the procedure.
Risk Scores, used in Steps 2 and 3, can be evaluated using the simplified approach presented in this
Section or the procedure outlined in Appendix B. The simplified approach, which is also based on the
procedure in Appendix B, does not cover all roadsides but it does address the majority of roadsides and
verges.
3.2 Concepts Used in Evaluating the Risk at Particular Sites
This Section provides an overview of the process to mitigate the risk at particular sites.
The background hazards are those that exist along the whole road segment and are therefore considered to
be continuous. The isolated hazards are those that generally exist in front of the background hazards. The
Risk Score is for a 1 km road segment and so the frequency of isolated hazards is based on this distance.
Figure 3.1 is a diagrammatical representation of a roadside and some essential dimensions. Further
discussion on the location of the background and for the isolated hazards is provided in Section 1.9.7.

Figure 3.1: Diagrammatical representation of a roadside
From these dimensions a Risk Score is obtained for both the background and the isolated and background
hazards. A plot of these Risk Scores against the NRRIT produces three possible outcomes:
• Treat the whole road length.
• Treat the isolated hazards.
• No treatment is required.
Figure 3.2 illustrates this concept. If there was a need to treat isolated hazards only, then all isolated hazards
should be treated.
Figure 3.2: Using the Risk Score values
The Risk Score is based on a standard or uniform road segment length. The process assumes 1 km road
segments and requires that uniform geometry exists over each segment. At some locations, a road segment
will have a part with an increased risk: for example, roads with tighter curves and steeper grades as in the
left hand illustration in Figure 3.3. In these instances, the attributes of the shorter part of road segment are
assumed to occur over the full 1 km length of road, as in the right hand illustration in Figure 3.3. This enables
a comparable Risk Score to be obtained. However, this Risk Score is only applied to the shorter part of the
road segment. (Remaining road lengths within the segment in the left illustration in Figure 3.3 would have a
lower Risk Score).

Figure 3.3: Using assumed characteristics to evaluate the Risk Score over a 1 km segment
In Figure 3.3, the average spacings of isolated hazards are the same in both cases.
If the action is to treat only the isolated hazards, then those that present a higher risk should be treated first.
These will be the isolated hazards that are closer to the road. The procedure presented in this section and
described in Appendix B, will assist if a more detailed analysis is required.
3.3 Step 1: Assess Against National Practices, Jurisdictional Policies and

Corridor Visions
If the road and road segment is covered by national practices, jurisdictional policies or a safety treatment
specified in the corridor safety vision then the risk needs mitigation without further risk assessment. Refer to
Section 1.8, Section 2.3 and Appendix C. The mitigation treatment options are described in Section 4 and
information regarding the design of roadside barriers is provided in Section 5.
3.4 Step 2: Compare the Risk Score with the NRRIT
The charts and process described in Section 1.9 are used to calculate a Risk Score for a program road
segment or a particular cross-section in a road project. The same Risk Score calculation process is used to
inform the NRRIT and to assess the risk at a particular cross-section. This allows for risk to be compared on
the same basis.
The designer should not interpolate between charts for different operating speeds, but rather use the next
higher operating speed chart.

These figures provide Risk Scores for the majority of roadsides. If the roadside and road dimensions are
significantly different from those used in the compilation of the charts, then the procedure in Appendix B
should be used.
If the existing roadside has a Risk Score that is greater than the NRRIT, then the risk needs to be mitigated
and the process advances to Step 3. The mitigation treatment options are described in Section 4 and
information regarding the design of roadside barriers is provided in Section 5.
It is not necessary to establish how much the Risk Score exceeds the NRRIT only that it does.
Consequently, if the Risk Score for significant background hazards exceeds the NRRIT then the analysis
does not need to continue any further as a risk mitigating treatment is required. Similarly, if there are
downgrades or sharper curves on the road then these may cause the Risk Score to be increased above the
threshold and only these areas may require treatment. Finally, other less significant and isolated hazards
may cause the Risk Score to exceed the NRRIT. The process is stopped once it can be shown that the
NRRIT is exceeded, and the road segment needs a safety barrier or other treatment to mitigate the risk to a
level below the NRRIT. Once this outcome is achieved, the process should proceed to Step 4.
The process to determine the Risk Scores is used to evaluate the risk associated with the background
hazards and the background and isolated hazards. The examples in the text follow this approach for both
rural and urban roads.
Figures in Appendix D list a complete set of Risk Score charts that are used in this step.
On roads where the design standard decreases, that is in a transition zone, the risk may be increased.
Designers should pay particular attention to interpret the Risk Scores for these sites.
3.4.1 Adverse Crash History
If a jurisdiction determines that a site has an adverse crash history, there may be a need to identify an
appropriate treatment in Step 3 and continue through to Step 4 to recommend an appropriate treatment
without considering the network risk.
3.4.2 Examples of the Use of the Procedure
In these examples, charts copied from those in Appendix D will be annotated and used in the analysis. The
chart number is shown in the top left corner of each chart.
Example 1 – Rural road in rolling terrain
A general view of the highway is shown in Figure 3.4. The road is through flat to gently rolling terrain with
significant side slopes and in some places the slopes are wooded. The lanes are 3.3 m wide and the
unsealed shoulder and verge width is 3 m wide. The operating speed at the site is 110 km/h, although it is
acknowledged that in some places along this road segment the operating speed is less than 110 km/h. The
traffic volume is 1400 veh/day.
The jurisdiction had previously decided that the NRRIT for this road was 2.0.

Figure 3.4: Road view for example 1
The background significant hazards are trees and steep embankments 3 m from the carriageway. Using
chart 5 in Figure 3.5 which is a copy of Figure D 1, this road segment has a Risk Score of approximately 1.8,
and accordingly does not require treatment. The terrain was described as rolling terrain. A generalised rural
chart is used to limit the number of charts in Appendix D.
Figure 3.5: Risk Score for example 1
Note: This figure is a copy of the lower part of Figure D 1 in Appendix D.
If the traffic volume was to increase, then the Risk Score could exceed the NRRIT and the site should be
treated. As an aside, the scene in Figure 3.4 is considered to be ‘flat’ and the results in Figure 3.5 apply. If
the process in Appendix B had been used then the Risk Score would be 1.9 for flat terrain and 3.3 m lanes.
Embankments steeper than 6:1 are not considered to be recoverable by both cars and trucks. Errant
vehicles traversing these slopes have a greater propensity to rollover and this site would warrant shielding
before other sites.
If a lower NRRIT of 1.5 was used, then this site should be treated.

Example 2 – Rural road in flat terrain with no background hazards and isolated trees in front
The verges on this road are relatively flat and wide and the road alignment is flat and straight (see
Figure 3.6). The lanes are 3.5 m wide and the sealed shoulders are 600 to 800 mm wide. The roadside has
shrubbery at approximately 7 m from the edge of the lanes. There is an occasional isolated tree that is
approximately 6 m from the edge of the lane; the assumed spacing is 200 m. The shrubbery has small trunk
diameters (less than 100 mm) and is not considered to be a hazard. Accordingly, there are no ‘background
hazards’. The operating speed is 110 km/h and the traffic volume is 1200 veh/day.
The NRRIT for this road is 1.5. There are no significant hazards forming a continuous background hazard.
Using the 110 km/h charts in Figure 3.7 (copied from charts 11 and 12 in Figure D.3 in Appendix D), with the
larger trees being 6 m from the road, the roadside Risk Score is 0.1 and clearly less than the NRRIT value of
1.5. The roadside does not require treatment. If the average spacing was reduced to 50 m, then the Risk
Score would be 0.3 and still less than the NRRIT.
Figure 3.6: Example 2 Highway in Western Australia
Source: Main Roads Western Australia.
Figure 3.7: Risk Score for example 2 using charts 11 and 12 in Figure D.3
Note: This figure is a copy of Figure D.3 in Appendix D. Chart numbers are shown in the top left of each chart.

Note that there is a risk of vehicles rolling over, on wider flatter verges. This risk has been included in the risk
diagrams for the background hazards. As there are no background hazards in this example, the risk
associated with rollover needs to be calculated separately. Using Figure 3.8, which is a copy of Figure 1.9
and Figure D.4, illustrates that the risk of rollover is relatively low at around 0.012. This is added to the other
Risk Score.
Figure 3.8: Rollover Risk Score for example 2 using chart 13 in Figure D.4
If on the other hand, trees were 3 m from the road and were much closer together, so as to be considered a
‘continuous hazard’, then they should be shielded. Figure 3.9 is a copy of Charts 4 and 5 in Figure D 1, and
illustrates that the Risk Score would be 1.6.
Figure 3.9: Risk Score for example 2 using chart 5
Note: This figure is a copy of the lower part of Figure D 1 in Appendix D.
Example 3 – Culvert headwall
A two-lane rural road in rolling terrain has two 3.3 m lanes, an operating speed of 90 km/h and a design
AADT of 2000 veh/day. The cross-section of the roadside is shown in Figure 3.10. There is a culvert
headwall 3 m long (in the direction of traffic) 6.4 m from the traffic lanes. There is a 4:1 embankment that is
9 m wide. In the vicinity of the culvert there is a 700 m radius horizontal curve to the right and a downgrade
of 6%. This puts the culvert on the outside of the curve. Note the headway of the culvert protrudes above the
embankment slope and it is considered to be a significant hazard.

Figure 3.10: Example 3 – culvert headwall on outside of horizontal curve
There is a line of trees at the toe of the embankment and these can be considered to be a background
continuous hazard with the culvert headwall as an isolated hazard. Many drivers would not recover on this
embankment because the 4:1 embankment is reasonably steep. Consequently, it is assumed that drivers
who travel over the embankment will collide with the trees beyond. The Risk Score will then be largely
independent of the presence of the culvert headwall and is calculated as if the trees were at the
embankment hinge point, 2.4 m from the traffic lane. This cross-section, in Figure 3.11, is then equivalent to
one shown in Figure 3.10.
Figure 3.11: Equivalent cross-section for example 3
If the trees were not at the toe of the embankment, then there is a chance of drivers being able to regain
control in the area beyond the embankment. In these circumstances the trees and the embankment are
treated as separate hazards (see Appendix B.9 Example 1).
Using Figure 3.12, the Risk Score for a significant hazard 2.4 m from the road is 3.1. This Risk Score is then
modified using Table 1.3 and Table 1.4. These tables do not cover all possibilities, and in this case the curve
adjustment factor is based on an operating speed of 90 km/h, the hazards being 3 m from the road, a curve
radius of 600 m and with the hazard on the inside of the curve. The curve adjustment factor from Table 1.3 is
1.1. Table 1.4 gives the adjustment factor for 6% downgrades as 2.0. Combining these values, the Risk
Score for this roadside, with the hazards 2.4 m from the road, is 3.1*1.1*2.0 which is 6.8.

Figure 3.12: Unadjusted Risk Score for example 3
If there were no trees at the toe of the embankment or on the embankment, then the embankment can be
considered to be the background hazard. The Risk Score of the embankment would be evaluated using
Figure 1.6 and the adjustments in Table 1.3 and Table 1.4. As the offset to the embankment is the same as
in the example above, Figure 3.11 and the curve and grade adjustment factors would provide a preliminary
Risk Score, that is 6.8 if the embankment was a ‘significant hazard’ as listed in Section 1.9.2. If the
embankment was a less significant hazard and had characteristics listed in Section 1.9.3 the Risk Score is
halved, if the embankment is listed as a minor hazard, the preliminary Risk Score would be divided by 6.
In this case the embankment slope is 4:1 and the embankment height is a little less than 2 m. It would be
considered to be a minor hazard and the Risk Score for the embankment would be (3.1)/6 or 0.5.
The risk of the culvert also needs to be determined. This could be done using Figure 1.8. The Risk Score is
evaluated for a road length of 1 km. If there is one culvert in a 1 km length of road, then the risk of the culvert
would be swamped by the other 950 m length of roadside. If the measurement window is reduced the
effective risk of the road length increases. As a comparison between a hazard and an installation of a barrier,
a window of 100 m is recommended. This is done by assuming culverts are spaced at 100 m intervals over a
1 km length of road.
Figure 3.13 shows the Risk Scores for culverts that are classified significant hazards having:
• protruding end walls
• vertical drops of more than 2 m or
• watercourses more than 2 m deep.
Culverts that are less significant hazards have watercourses less than 1 m deep with a drop between 0.3 and
2 m. The Risk Score for these culverts is half that shown in Figure 3.13.
Culverts with watercourses less than 1 m deep and with a drop less than 0.3 m, are minor hazards and have
a Risk Score that is 1/6th of the value shown in Figure 3.13.
In this case the culvert is assumed to be a significant hazard and its Risk Score is 0.24. The total Risk Score
is 0.74 being the Risk Score for the embankment (0.5) plus the Risk Score for the culvert (0.24). Note that in
this case, the isolated hazard, the culvert, does not partially shield the background hazard, and that the
background hazard is in front of the isolated hazard.

Figure 3.13: Risk Score for culverts classified as a significant hazard
Note: 100 m measurement distance used.
Chart
number
30 Rural mountainous 70 km/h Mountainous 70 km/h 3.0 m
Example 4 – Urban road
The guidelines for urban roads can only be indicative as urban roads often have street furniture that are
hazards which cannot be conveniently shielded without causing significant disruption to pedestrians and
other road users.

Figure 3.14 shows typical views with trees and other significant hazards within a metre of the lane edge and
with an average spacing of 50 m. The traffic volume is 10 000 veh/day. The lanes are 3.5 m wide and the
operating speed is 80 km/h. The property boundary is 3 m from the road. For this example, the NRRIT for
this road had been evaluated to be 1.5.
Figure 3.14: View of a typical urban road
Note: This figure is illustrative for the example described above.
Figure 3.15 which is a copy of chart 18 of Figure D.6 in Appendix D, shows that the Risk Score for this road
is 0.9 for a background hazard of a generic wall (or fence in this example) 4 m from the roadside and isolated
hazards 1 m from the roadway.
Figure 3.15: Risk Score for example 4 using chart 18
Note: This figure is a copy of Figure D.6 in Appendix D.
As a check, using this offset the Risk Score, calculated using the process in Appendix B, is 0.85. The
property boundary is 3 m from the edge of the lane, rather than the 4 m used to develop chart 18 in
Figure D.6. If a 3 m wall offset was used in the process in Appendix B then the Risk Score would be higher at
0.87. Given either Risk Score, this roadside does not require treatment at this time.

If more trees were planted, then the Risk Score could be increased to be over 1.5, and they would need to
be shielded or the operating speed reduced. This evaluation is shown by point A in Figure 3.16 when the
average spacing is reduced to 20 m. If the operating speed is reduced to 70 km/h, then the Risk Score is 1.2
(see point B in Figure 3.16). However, this discussion depends on the NRRIT used. Planting trees,
introduces additional risk and a jurisdiction could establish not only a NRRIT, but also a maximum risk level
for the introduction of additional hazards. This additional risk level should be well below the NRRIT (see
Section 2.5).
Figure 3.16: Risk Score for example 3 using charts 17 and 18
Note: This figure is a copy of Figure D.6 in Appendix D.
3.5 Step 3: Identify, Evaluate and Rank Risk Mitigation Options
3.5.1 Identify the Options
A prerequisite to assessing both new and existing roadsides is to ensure that the road is designed and
maintained in a way that should enable drivers to keep their vehicles on the road when travelling at an
appropriate speed (refer to Section 1.6.3). Previously, the clear zone concept was used to define the area
beside the road to be evaluated for roadside safety. It was considered that hazards outside the clear zone
were acceptable. This is no longer appropriate. The designer must consider all hazards in the road
reservation and that a clear area be now considered a mitigating treatment option (see Commentary 3).
Austroads (2018) stated ‘no matter how wide a clear zone or central median, the risk of incursion cannot be
eliminated’. This is evident from Figure B.7 in Appendix B. Even if the hazards were further than 20 to 30 m
off the road, there will still be a possibility of vehicles colliding with them. In addition, the risk of vehicles
rolling over increases as the width of verge increases. All roadsides have a non-zero risk, no matter how flat
or smooth or wide. Accordingly, the concept that the clear area alongside the road will be sufficient to provide
a responsible roadside is now generally not accepted for roads with heavy traffic flows.
[see Commentary 3]
Whatever option is chosen, there will be a risk of injury if vehicles leave the road. The process is to identify
this risk and to choose the option that minimises the risk and reduces the Risk Score to be below the NRRIT.
The basic options include:

• installing barriers
• installing audio-tactile line markings (ATLM)
• changing the cross-section (including the use of wide medians and median barriers)
• removing the hazard

• relocating the hazard to a position where it is less likely to be struck (ideally as far from the road as
possible)
• reducing the impact severity posed by the hazard (e.g. redesign so that a hazardous feature can be
safely traversed, use of frangible poles)
• improving the delineation of the road
• accepting the risk of the untreated hazard where the frequency of hitting the hazard and severity are both
low. In this case, the risk should be monitored.
A combination of options listed above may provide the best solution.
The Risk Scores for continuous safety barriers on rural undivided roads are given in Figure 3.17. The Risk
Scores recorded are based on a barrier being 1.0 m from the traffic lane for roads with a 70 km/h operating
speed and 2.0 m from the road for the 90 and 110 km/h operating speed roads. The Risk Scores decrease if
the barrier is located further from the road; these values are considered to be conservative. Figure D.8 in
Appendix D contains the same figures.
As an example, if a barrier is installed on a rural undivided rural road on rolling terrain with an operating
speed of 90 km/h, and an AADT of 2000 veh/day then the Barrier Risk Score is 0.48.
Figure 3.17: Risk Scores for continuous barriers on undivided rural roads with different operating speeds and
in different terrain
These figures can be used to establish the change in risk if a barrier is installed.
Austroads (2014b) lists the reduction in crashes from the installation of ATLMs. They recorded a crash
modification factor of 0.8. The Risk Score when ATLMs are used should be multiplied by a factor of 0.8 as
the ATLMs reduce the exposure. While each treatment acts to reduce crashes, the impact of each
successive treatment on reducing the numbers of crashes is diminished. A number of techniques exist to
calculate the cumulative benefit of more than one treatment (refer to Austroads 2015d).
Treatment options are discussed in Section 4 and barrier selection is described in Section 5.

3.5.2 Evaluate the Risk Associated with a Roadside Treatment Using a Qualitative Assessment
An evaluation of treatment options is undertaken using a quantitative assessment and a qualitative

assessment. Before a treatment option is selected for prioritisation and implementation, its suitability in terms
of environmental and engineering factors must be considered.
Environmental considerations
The environmental issues to be considered include:

• environmental impact
• recognition of unique vegetation (e.g. environmentally sensitive areas or national parks). If the clearing of
trees is unacceptable on environmental grounds, alternative treatment options will have to be considered.
• the retention of watercourses in their natural state adjacent to the road
• reduction of clearing
• visual pollution.
Engineering and other considerations
The factors to be considered include:

• traffic growth
• pedestrian and cyclist traffic (particularly children)
• vehicle mix including motorcyclists
• crash history
• other geometric influences
• social justice/equity
• road is used as a school bus route
• access requirements
• road is used as a freight route
• aesthetics.
3.5.3 Evaluate the Risk Associated with a Roadside Treatment Using a Quantitative Assessment
A quantitative evaluation includes an assessment of the risks associated with hazards and computation of
Risk Score as described in Section 1.9. Some treatments will reduce the exposure, such as the use of
ATLMs and signage. Some treatments may reduce severity by changing the hazard characteristics. The
same method can be used to analyse the risk and determine the change in the Risk Score associated with
different treatment options.
Other factors that should be considered in a quantitative assessment are:

• design and construction costs
• maintenance costs
• whole-of-life costs
• time to implement
• constructability.

The quantitative method described in Section 1.9, is to be used to evaluate the treatment options by
evaluating the change in risk (measured as the product of exposure, likelihood and the Trauma Index). This
approach is not based on traditional benefit-cost analysis, which is now not a recommended practice.
Appendix E provides a procedure to predict the injury costs associated with different Trauma Indices. This
will enable a benefit-cost analysis of roadside hazards and treatments to be undertaken within the Australian
Transport Assessment and Planning Guidelines (ATAP 2018) if it is considered appropriate.
3.5.4 Rank Treatment Options
Often, the only reasonable option is to install a safety barrier. Options should be ranked when selecting the
preferred one amongst a number of possible options and when a number of risk reduction options have been
identified and funds are limited. It is suggested that the options should be ranked in risk reduction order first
and then the qualitative factors.
3.6 Step 4: Design the Recommended Roadside Treatments
The intent of this Step is to determine the dimensional limits and geometric requirements for treatments
identified for location(s) identified in Steps 1 to 3. This step uses the content in Sections 4 and 5.
This step may involve the design of a treatment for an isolated site or the preparation of a road design plan
that shows a number of treatments along a road segment that are designed to address different types of
hazard. In the latter case a detailed design may be prepared for each treatment.
In some cases, jurisdictional standard drawings will provide the necessary detail (e.g. culvert end treatments
and road safety barriers) whereas the road design layout will show the location of the treatment (e.g. lateral
and longitudinal location (i.e. chainage) of a barrier as well as information not covered by the standard
drawings.
Draft road designs of options should be prepared to assist in estimating costs and to facilitate approvals,
together with documentation on:
• all the hazards that were considered for treatment
• the type and location of all treatment options considered
• the possibility of grouping hazards for treatment.
The final design and documentation should consolidate the design as a whole, considering:
• all hazards for which a treatment is warranted
• the treatment options chosen for the hazards
• the priority of the treatment options.
The final design should be consistent with the cross-sections described in the corridor safety vision.

4. Treatment Options
NOTE: This section is largely unchanged in this edition and will be updated in the next edition of
this Part.
4.1 General
This section expands on the summary in Section 3.5.4. It describes treatment options that can be used to
effectively reduce risk on an existing road or in the design of a new road. As discussed in Section 1.6.3 the
systematic approach to risk reduction in design involves:
• reduction of the inherent severity of a hazard
• prevention of an incident
• limiting damage.
It is important to understand that there may also be risks associated with treatment options and that the
comparative risk of the treatment should be assessed in relation to risks associated with an untreated
hazard.
4.2 Summary of Treatment Options
Research has shown that flexible barriers can provide a better safety outcome compared to other treatments
(Austroads 2014a, 2018). However, flexible barriers may not be suitable in every situation e.g. the currently
available systems cannot contain all heavy vehicles. A site-specific assessment should be made considering
the operating speeds, roadside environment, road users (e.g. motorcyclists and pedestrians), and the
benefits and costs of the installation of any safety barrier.
The first choice should be to install a barrier to shield hazards in the verge or median. A safer roadside may
involve measures such as:
• installation of a barrier
• removal of hazards
• provision of sealed shoulders
• gentle slopes with firm, even surfaces and rounded batter hinge points
• traversable open drains
• extension of culverts; however, care must be taken not to cause excessive warping of the embankment
slope that may affect the stability of an errant vehicle
• traversable culvert ends
• frangible supports for road furniture and road lighting
• adequate clearances to structures
• provision of underground utility services.
All occupants in vehicles that leave the road are at risk. The risk of any roadside is not zero and the risk
should be understood if it is to be managed. The risk is different for different road users and minimising the
risk to one user group might be at the expense of another group. It is not possible at this stage to have a
safety barrier that provides optimum performance for motorcyclists, car drivers and truck drivers.

4.3 Effectiveness of Treatment Options
The Guide to Road Safety Part 8: Treatment of Crash Locations (Austroads 2015d) provides information on
the effectiveness of crash treatments and lists crash modification factors (CMFs) that have been used in
Table 4.1 to provide a broad rating of the effectiveness of various treatments in reducing the risks associated
with specific types of crashes. The table is intended to provide designers with some general guidance on the
types of treatments that are likely to be most effective as a countermeasure for the crashes shown in the
table. The table also indicates the way in which the hazard reduction is achieved (e.g. shielding or
prevention). Some of these treatments can be used in combination, which may provide a greater safety
outcome, e.g. sealing of shoulders and installing a flexible safety barrier system.
Table 4.1: Crash types and effectiveness of treatments
Effectiveness of treatment by crash type
Out of
Treatment Type of hazard reduction Off Off Cross- Impact
control
path, on path, on median with
on
straight curve head-on hazard
curve
Install flexible barriers on Very Very Very Very Very

Provide shielding
median high high high high high
Install flexible barriers along Very Very Very Very Very

Provide shielding
the verge high high high high high
Install rigid/semi-rigid
Provide shielding Medium Medium Medium Medium Medium
barriers on median
Install rigid/semi-rigid
Provide shielding Medium Medium Medium Medium Medium
barriers along verge
Very
Remove roadside hazards Prevent an incident
high
Very
Duplicate carriageway Reduce likelihood
high
Widen median Reduce likelihood High
Improve alignment Reduce likelihood Medium Medium Medium Medium Medium
Seal shoulder Reduce likelihood Medium Medium Medium Medium Medium
Widen shoulder Reduce likelihood Medium Medium Medium Medium Medium
Resurface road-high skid

Reduce likelihood Low Low Low Low Low
resistance
Provide overtaking lane Reduce likelihood Medium
Widen or replace bridge or

Reduce inherent hazard Medium Medium Medium Medium
culvert
Provide line-marking and

Reduce likelihood Low Low Low Low Low
guideposts
Advisory speed sign Reduce likelihood Medium Medium Medium
Source: Based on Roads and Traffic Authority (2008) and Austroads (2014a).

4.4 Types of Treatments
This section discusses treatments that may be applied to elements of road and roadside design at greenfield
sites to address road safety issues that may emerge throughout the design process. These treatments may
also be applied at brownfield sites. Treatments that generally relate to brownfield sites are described in
Appendix F.
4.4.1 Treatments for Trees
Trees feature prominently as impacted hazards in run-off-road crashes and account for a large proportion of
fatalities. Trees are a particular hazard when located within and close to curves. Trees create a hazard when
they drop branches that can end up on the shoulder or road.
Trees greater than 70 to 100 mm diameter (depending on the species) that are located within the area of
interest pose a particular hazard to motorists. Tree stumps more than 100 mm above the ground are also a
hazard.
There are two possible treatments for dealing with trees:

• Tree removal. This is not always an option because of environmental considerations relating to the
intrinsic values of many trees and the habitat they provide. The removal of individual trees should be
considered when they are in particularly hazardous locations, and maintenance patrols should ensure
that naturally seeding saplings that are in hazardous locations are not allowed to mature.
• Installation of road safety barriers. Provision of barriers will depend on a number of factors relating to site
conditions, crash history, economics and the environment. However, such action should only be taken
where it is determined that collision with the barrier is less severe than collision with the existing hazard
(i.e. trees).
Assessment of significant trees
Significant trees should be assessed in accordance with jurisdiction guidelines before removal is proposed.
New trees
The hazard potential of newly planted trees should consider them as fully grown, and they should only be
planted in areas most unlikely to be hit by errant vehicles as demonstrated by the risk assessment using the
Risk Score.
4.4.2 Treatments for Steep Downgrades
Where a warrant has been established for treatment of a steep downgrade (refer to Section 6.3) the
treatment options may include:
• a gravity safety ramp
• an arrester bed
• a dragnet
• a combination of systems.
Design of treatments needs to follow the process shown in Section 6.6.
4.4.3 Treatments for Medians
Medians provide an opportunity to separate opposing traffic flows and the installation of a safety barrier
system can prevent errant vehicles crossing the median into the opposing traffic flow (refer to Sections 5.3.9,
5.3.10 and 5.3.11).

4.4.4 Treatments for Drains
Deep unshielded drains should not be installed at the base of cut batters. Effective redirection of vehicles
requires a flat, even surface approaching the batter.
Open drains are present on the majority of rural roadsides and may exist on urban freeways. Open drains
constructed close to the road may be the most efficient way of removing water but, unless they are of a
suitable shape, they are a hazard for vehicles that leave the road.
Typical drains can be classified by whether they are designed with abrupt or gradual slope changes. Abrupt
slope change designs include vee drains, drains with a rounded bottom and a width less than 2.4 m, and
trapezoidal drains with bottom widths less than 1.2 m.
Vehicles leaving the roadway and encroaching into a drain are faced with three hazard areas:
• Drain front slope – if the front slope is 4:1 (14°) or steeper, the majority of vehicles entering the ditch will
be unable to stop and can be expected to reach the bottom.
• Drain bottom – abrupt slope changes at the bottom of the drain can cause errant vehicles to roll or stop
abruptly and increase the severity of the impact.
• Drain back slope – vehicles travelling through the ditch bottom or becoming airborne from the front slope
can collide with the back slope.
AGRD Part 3 (Austroads 2016a, Section 4.6) provides some preferred design cross-sections for channels
with abrupt and gradual slope changes with preferred foreslopes and backslopes for basic ditch
configurations indicated.
Some crash testing of drain shapes indicates that, depending on the angle, shapes outside of those
described in AGRD Part 3 are traversable (Thomson & Valtonen 2002); refer also to Commentary 4 for more
information.
[see Commentary 4]
As a general guide, longitudinal drains and kerbs are hazardous where:

• the foreslope and the backslope form a V-shape, as it is possible to crash into the back slope (It is the
change in embankment slope as well as the slopes themselves that affect safety.)
• they are located at the foot of a fill slope
• there is an object with a Trauma Index located in the drain
• the kerb shape and height are not appropriate for the speed environment and may affect the performance
of the roadside infrastructure installed behind it.
Transverse open drains are usually provided outside of the road formation to carry water into culverts and
unless designed correctly with a suitable cross-section, they may also present a hazard to vehicle occupants.
In addition, a depth of water as little as 0.3 m in a traversable drain shape may make the drain
non-traversable, particularly for passenger cars.
Drain sections that are assessed as hazardous should be:

• reshaped
• converted to a closed system (culvert or pipe)
• shielded with a road safety barrier where appropriate.
Drains can funnel a vehicle along the drain bottom, and this increases the probability of impact with any
object present on the bottom or side slopes of the drain. For this reason, such objects should not be located
within drains; their presence also has the potential for water flow to be adversely affected.

4.4.5 Treatments for Drainage Features
The ends of culverts that cross under the road or are located parallel to the road constitute hazards. Road
design should aim to eliminate all non-essential drainage features. Where drainage features are
unavoidable, they should be designed as follows:
• Drains parallel to the road (e.g. under a driveway or side road) – traversable culvert end treatments
should be installed wherever a culvert exists parallel to the road (Figure 4.1).
• Perpendicular to the road (headwall treatment) – culverts that run perpendicular to the road (i.e. run under
the road) should be designed to be traversable (Figure 4.2) or present a minimal obstruction to an errant
vehicle. The slope of the fill batter should be drivable or be shielded with an appropriate road safety
barrier. Alternatively, the culvert can be extended to a location further from the travelled way where the
end is less likely to be impacted by errant vehicles.
Figure 4.1: Example of a driveable culvert end wall for a small pipe under a driveway
Source: Department of Planning Transport and Infrastructure (2017).
Figure 4.2: Example of a traversable culvert end treatment (under construction) for a culvert under the road
Source: Provided by VicRoads (n.d.).

Cross-drainage of road reserves is achieved by the provision of culverts that may vary in size from a single
375 mm pipe to large multiple pipes or box culverts. The preferred open drain cross-sections described in
AGRD Part 3 (Austroads 2016a) apply to longitudinal open drains that may convey water to transverse
culverts. Transverse open drains are usually provided outside of the road formation to carry water into
culverts and, unless designed correctly (i.e. with flat foreslopes and backslopes), they may also present a
hazard to vehicle occupants.
Traditionally culverts have been designed with concrete headwalls and wing walls that either have resulted in
a potential roadside hazard or required shielding with road safety barrier. In such cases, the options to
remove or reduce the hazard caused by these obstacles are (AASHTO 2011a):
• eliminate the structure
• design the culvert end to be traversable
• shield the culvert with a road safety barrier
• delineate the culvert if the previous options are not cost-effective or practicable.
If a foreslope (embankment or drain) is traversable the preferred option is always to extend (or shorten) the
culvert to intercept the roadway embankment and to match the inlet or outlet slope to the embankment slope.
For small culverts, no other treatment is required. A small culvert may be defined as a single pipe that has a
diameter of 900 mm or less, or multiple pipes each having a diameter of 750 mm or less.
Matching culvert ends to embankment slopes is also desirable because it:

• results in a smaller obstacle for an errant vehicle
• reduces erosion problems
• simplifies mowing operations.
If a front slope is not traversable it may not be appropriate to provide a traversable end treatment, and an
evaluation of alternative treatments must be undertaken (e.g. flatten the embankment to make it traversable,
shield the embankment with a road safety barrier).
As an errant vehicle may travel a substantial distance from the road, an obstacle at a large offset may still be
a hazard. Extending culverts without providing a traversable end is therefore not preferred, particularly on
high-speed roads, as this option may create discontinuities in an otherwise traversable slope. A risk
assessment, described in Section 3, should be used to evaluate options.
Single culverts and end treatments wider than 1 m can be made traversable for passenger-size vehicles by
using bar grates. Full-scale crash tests have shown (AASHTO 2011a) that cars can cross grated culvert end
treatments on slopes as steep as 3:1, at speeds as low as 30 km/h or as high as 100 km/h, when steel pipes
spaced at 750 mm centres are used across the opening. Although this treatment does not significantly
change the hydraulic performance of the culvert, during the design process due consideration should be
given to the likely accumulation of debris and the level of maintenance required.
Culverts equal to or less than 900 mm in diameter or box culverts with a span equal to or less than 900 mm
can be made safer by the use of a traversable headwall. Based on Trauma Indices, culverts perpendicular or
skewed to the road alignment greater than 900 mm in diameter or box culverts with a span greater than
900 mm pose a greater hazard to the road user than a safety barrier.
Where culverts are placed parallel to the road they can be made safe for road users by installing gratings
over the culverts. For this treatment to be effective it must be used on slopes of 4:1 or flatter with a pipe
diameter or box width less than or equal to 900 mm.
In some instances, it may be appropriate not to treat the end of a culvert at all, and to simply provide
adequate delineation. Provision of barriers on low-volume roads should not result in a higher risk to road
traffic than not providing a barrier, given that all other things are equal.

Designers should refer to AS 5100.1-2017 for guidance regarding road safety barriers on bridges and should
note that culverts are also referred to in that standard.
4.4.6 Treatments for Rock Face Cuttings
Cuttings and rock faces are generally expensive to construct. Economic and environmental constraints often
result in cuttings being as narrow as possible and prevent the provision of a cutting wide enough to allow a
clear, flat verge beside the road. Therefore, cuttings and rock faces should be designed to provide a smooth
face that will act as a rigid barrier, allowing errant vehicles to slide along and stop gradually. Uneven batter
surfaces may present a hazard to vehicles that happen to run off the road (e.g. snagging and rolling). If a
smooth face and approach surface cannot be provided, it may be appropriate to install a barrier to prevent
vehicles colliding with an uneven rock surface.
There are no guidelines available for the acceptable roughness of rock faces. However, the degree of
roughness that can be tolerated is minimal as indicated by guidelines accepted in a US Federal Highway
Administration (FHWA) approval letter regarding the provision of vertical relief on the face of rigid barriers
(FHWA 2002). The guideline (refer to Commentary 5) specifies a maximum height of 64 mm for the
irregularities in the surface of the barrier above the height where wheel contact would occur.
[see Commentary 5]
The risk of snagging may be reduced by making the cutting smooth for at least the first 1 m of the cutting
height (based on the centre of gravity of a 2000 kg vehicle). This could be achieved in the case of a rock
cutting by applying a shotcrete treatment to the cut batters.
The requirement for treatment should be determined on a case-by-case basis depending on the speed
environment, traffic volume, offset to the cutting, surface condition and the length of the cutting.
4.4.7 Treatments for Minor Roadside Hazards
Minor roadside obstacles such as fences, fire hydrants, mailboxes and other hazards can pose a serious risk
to an errant vehicle that may strike the object. Objects containing horizontal rails capable of spearing
vehicles (such as post-and-rail fences) can be particularly hazardous. Such objects should be located where
an impact with the object should not result in a serious crash. It is essential that objects located close to the
road are designed to minimise risk to road users and this will often require them to be frangible.
4.4.8 Treatments for Road Furniture
As for other roadside furniture, traffic signal poles can pose a hazard for any errant vehicles. They are often
necessarily located close to the travelled path at intersections, which could lead to a higher risk of impacts,
although some measures can be taken to minimise this risk. Such measures include not locating a traffic
signal pole on the outside of a curve, setting poles as far back from the travelled path edge as practicable,
minimising the number of poles and installing joint-use poles wherever practicable. Provision of high skid
resistance at intersections can also reduce the risk of a vehicle losing control at an intersection and skidding
into traffic signal poles or other roadside hazards.
Small road signs are usually supported by small diameter and thin-walled metal conduits that are frangible
under vehicle impact. However, larger signs require substantial supports and should either be provided with
frangible mechanisms at the base of the supports (e.g. weakened timber, slip-bases with hinge points just
below the sign) or be shielded by a road safety barrier or crash attenuator. Frangible bases are often not
suitable in urban areas in which case the support should be located as far as possible from the travelled way
or shielded. AS 1742.2-2009 discusses aspects of longitudinal and lateral placement, mounting height and
orientation for signs, post type and selection.

4.4.9 Treatments for Poles
General
The hazard presented by a roadside pole is related to both its location and type of construction; these factors
contribute to the crash hazard the pole may pose and the consequences of an errant vehicle hitting the pole.
Poles that are present in road reservations to reticulate electricity are problematic in that they are generally
very expensive to remove and replace with an underground supply. However, this option should be
considered in appropriate situations.
Poles are a common road furniture item used to support signs (regulatory, warning, guidance, informative),
road lighting and various devices. In line with the preferred treatment for roadside hazards (i.e. removal), the
practitioner’s aim should be to minimise the number of poles in the area of interest.
Avoid placing poles close to the roadway
Any roadway improvement that involves reconstruction of utility services should take the opportunity to avoid
placement of poles close to the roadway. This proactive approach will avoid problems rather than having to
rectify them in future.
Minimum lateral setback distances for signs and for road lighting poles are specified in AS 1742.2-2009 and
AS/NZS 1158.1.2-2010, respectively. Where possible, poles should be located so the risk to the occupants
of an errant vehicle is low.
Undergrounding cables
Relocation of utility services to underground ducts and removal of the poles is the most effective option for
the treatment of hazardous poles.
Rationalisation of pole functions
It may be possible to rationalise the number of poles along a road corridor by combining separate functions
and services onto common poles. For example, traffic signals, road lighting and large signing may be
supported by the same poles. Power cables, telecommunication services and streetlights can share common
poles.
It may be possible to place all poles on the side of the road that has the better safety performance or least
risk. This may involve changing the poles from side to side as the crash risk changes along a curved route.
Relocation
Pole relocation needs to target areas where the run-off-road crashes are likely, for example on the approach
to curves, the outside of curves, near lane merges, lane terminations, adjacent to exits from roundabouts and
intersections. Research (Zegeer & Cynecki 1984; Zegeer & Parker 1984) has confirmed the belief that the
number of crashes decreases as poles are moved further from the roadway.

Frangible poles
The use of frangible poles may be effective in reducing the severity of pole-related crashes, if pole removal
or relocation is not feasible. These types of poles are designed to collapse or breakaway on impact and
thereby reduce the severity of injuries to the occupants of an impacting vehicle, compared to those that could
occur if the pole was rigid. Rigid poles do not deform to a great extent but are designed so that they remain
upright after an impact. Alternatively, frangible poles are designed to deform upon vehicle impact and are
usually used for road lighting as the lighting needs to be close to the road. Types of frangible poles include:
• slip-base poles that breakaway at the base upon impact, allowing the vehicle to pass beneath the pole in
order to minimise or avoid injury to vehicle occupants
• impact-absorbing poles that collapse over the colliding vehicle and are designed to bring the vehicle to a
controlled stop at the base of the pole. These deformable poles are designed to remain in the ground
after being hit.
The following issues need to be considered when specifying frangible poles to reduce impact severity:
• Removing or relocating the pole should be considered before specifying frangible poles.
• The area behind the pole should be free of other hazards and in the case of breakaway poles a run-out
area may be required.
• There should be limited pedestrian activity in the vicinity of the pole.
• The damaged pole and any elements that detach under impact should not pose a risk to other road users.
Impact-absorbing poles should be favoured over slip-base poles where there is closely abutting
development, pedestrian and parking activity and a low traffic speed environment.
Signposts should be designed to be frangible in the event of impact (i.e. posts that are designed to fracture,
breakaway, give way or bend), so that the damage to a colliding vehicle and risk of injury to vehicle
occupants upon impact is minimised. Small signs are usually supported by posts that deform in a way that
causes minimum damage to cars, whereas larger posts and supports (for larger signs) may be provided with
mechanisms that are designed to yield in a controlled manner upon impact.
Aspects to be considered in the selection of pole type and setback from the roadway include:
• surrounding land use
• pedestrian activity
• speed limit
• whether the road has kerbs or not
• location (mid-block or at an intersection)
• whether the pole is to be located behind a road safety barrier
• maintenance.
This may involve locating them at the property line (urban and rural) or in an easement (rural).
Frangible poles – slip-base
Slip-base poles consist of a standard pole stem, mounted on two base plates that are clamped together with
bolts that release on impact thus allowing the pole stem to break away from its foundation. An example is
shown in Figure 4.3. A disadvantage with slip-base poles is that the dislodged pole may create a secondary
incident by falling on bystanders or adjacent vehicles.

The decision to use slip-base poles will depend on the space available and the resultant likelihood that a
falling pole would cause injury to other users of the road or roadside area. For example, a slip-base pole will
usually be inappropriate where pedestrian or cyclist traffic is common because a falling pole may pose an
unacceptable risk to those road users.
Slip-base poles are designed to operate when impacted at certain speeds. If the poles are installed within
the deflection zone of road safety barriers, then the slip-base may not operate as intended and the
redirection of the vehicle by the barrier may be affected.
Lack of maintenance is a significant problem with slip-base poles. They should be checked regularly to
ensure they are free to slide and the bolt tension is correct. Wind vibration can cause poles to move the
assembly and jam the bolts.
Figure 4.3: An example of a slip-base pole mechanism
Source: Based on Roads and Traffic Authority (2008).
Frangible poles – impact absorbing
Impact-absorbing poles remain attached to the base structure and absorb impact energy by progressively
deforming and entrapping the impacting vehicle. The deformation of the pole is controlled by a designed
weakening of the pole stem. As with slip-base poles, an impact-absorbing pole can fall onto the impacting
vehicle causing significant roof deformation. Figure 4.4 illustrates the concept of an impact-absorbing pole.

Figure 4.4: Examples of impact absorbing poles
4.4.10 Treatments at Active Controlled Level Crossings
Longitudinal barriers generally are required to have a minimum length of 20–30 m (between terminals) to be
effective, so their use to shield level crossing equipment is limited. Short lengths may present a greater
hazard to road users than the equipment behind the barrier. Alternatives are to use crash cushions, to install
level crossing equipment as far from the edge of the travelled way as possible and to define the approaches
with enhanced delineation.
4.4.11 Weather Warning Systems
A range of conditions related to weather can have an adverse effect on vehicles staying on the road through their
impact on drivers (e.g. reduced visibility and strong crosswind) or the road surface. Common conditions include:
• heavy rain
• ice and/or snow
• fog
• water on road
• strong winds.
Weather warning systems may be used in locations where the adverse effects of weather are known to
increase the risk of road crashes. This may be as simple as permanent signs, but more complex systems are
also possible. For example, an ice warning system can consist of warning lights and signs that are activated
by inputs from temperature and humidity sensors. Similarly, a fog warning system could activate advance
warning signs and lights in response to inputs from a visibility detection device.
If a system is proposed, several years of crash data should be examined to determine that the weather
conditions did influence crashes at the site.

5. Road Safety Barriers
NOTE: This section is largely unchanged in this edition and will be updated in the next edition of
this Part.
5.1 Introduction
5.1.1 General
A possible outcome of a roadside safety assessment (using the hazard mitigation process described in
Section 3) during the design of a new road or upgrade of an existing road, is that roadside hazards will need
to be shielded by a road safety barrier system.
This section provides a general description of the types of road safety barriers, crash cushions and other
devices available and discusses factors that should be considered in selecting an appropriate type of barrier
and designing suitable layouts for barriers. It also describes the process and considerations involved in
designing barriers. However, designers should read this Part in conjunction with AS/NZS 3845.1:2015 which
covers standards relating to the use of barriers in Australia and New Zealand.
A barrier should only be installed when the barrier reduces the risk associated with the roadside. However, in
many, if not most locations, the installation of a roadside barrier is likely to be the safest design option
(Austroads 2018). Generally, the likelihood of striking a barrier is greater than striking the hazard (e.g. a tree
some distance further from the road). However, the severity of an impact with the barrier is usually much less
than that associated with striking the hazard.
Only road safety barriers and end treatments accepted by the relevant national, state, territory or local road
agency should be used. The barriers and end treatments covered in this Part generally comply with
AS/NZS 3845.1:2015. However, other devices may be used with the approval of the relevant agency. When
using proprietary products, it is important that reference is made to the manufacturer’s manuals and
specifications.
Note that AS/NZS 3845.1:2015 has adopted the MASH (AASHTO 2016) test standards for crash testing of
barriers and end treatments. In line with the USA, the Austroads Safety Barrier Assessment Panel is moving
towards MASH accepted products. The accepted safety barrier product list for each jurisdiction should be
consulted to determine the installation and replacement status of a product.
New barrier systems will have to be tested under these guidelines before their use is accepted. However,
some jurisdictions have continued to use barriers that have not been crash tested but have a satisfactory
field performance.
Road safety barriers are more hazardous when they are improperly maintained, too short to adequately
shield a hazard, incorrectly installed, installed close to kerbs and drains, do not use effective terminals and
anchorages and when they restrict sight lines for drivers.
5.1.2 General Requirements for Road Safety Barrier Systems
Barrier systems should be accepted or approved for use by the national, state or territory road agency. It is
generally required that the use of barrier systems should be:
• supported by technical literature and assembly instructions that clearly demonstrate the essential mode of
operation and prominently show the test level achieved in crash testing carried out in accordance with
AS/NZS 3845.1:2015

• selected and located in accordance with a recognised design procedure that is professionally applied; the
procedure is to take account of risk management techniques that address the community of road users
and neighbours that may be affected by the installation; the hazard mitigation and design procedures in
this Part meet these requirements
• erected in accordance with manufacturers’ instructions
• maintained in a manner that reflects specified requirements
• returned into service only after professional evaluation and execution of repairs
• fitted with end treatments and interface devices that are appropriate.
5.1.3 Road Safety Barriers for Vulnerable Road Users
Barriers and some of their features can be very hazardous to motorcyclists and cyclists who crash into
barriers, whether on the vehicle or not. The provision of inappropriately located and designed barriers in
close proximity to footpaths and bicycle facilities can also cause injury to these road users.
Motorcyclists
Safety barriers are effective in containing and redirecting design vehicles, but their effects on the safety of
other road users, especially motorcyclists, remain unclear (Duncan et al. 2000).
In motorcycle crashes, injuries result from the rider and/or pillion passenger contacting the barrier system.
While the objective of safety barriers in motorcycle crashes is to prevent the rider (or pillion passenger) from
coming into contact with the roadside hazard, the support structure (posts) may potentially cause serious
injury. This is different from passenger vehicle crashes where a well-designed barrier system contacts and
engages the errant vehicle and redirects it but does not transfer excessive forces to the occupants.
A hazardous aspect of barriers with respect to motorcyclists is exposed barrier posts, as their edges
concentrate the impact forces resulting in more severe injuries (Australian Transport Safety Bureau 2000).
Other barrier features that may be hazardous to motorcyclists include:
• upper and lower W-beam edges
• block or block-out edges, impact extruder heads
• the upper portion of the steel posts protruding above the top edge of the horizontal section of semi-rigid
barriers
• protruding delineators utilising metal componentry
• barrier systems that are too low as motorcyclists can be catapulted over barriers of insufficient height
• semi-rigid and flexible barriers that are too high to prevent motorcyclists from sliding beneath the
horizontal rail
• discontinuous or jagged barrier surfaces, such as concrete barriers with decorative designs, which
present edges to concentrate the forces of impact
• rigid barriers (likely to be involved in front-on collisions), which require an impacting rider to absorb
virtually all of the kinetic energy at impact.
Note that delineator brackets on a barrier should be frangible not sharp and should not be located in the
W shape of steel beams.
These issues should be considered when designing roadsides and barriers on roads that carry significant
numbers of motorcyclists (e.g. popular motorcycling routes). Some of these features can similarly be
hazardous (e.g. sharp edges) to cyclists and pedestrians who travel in close proximity to the barriers.

There has been no comprehensive crash-testing program undertaken that has compared the safety
performance of a number of different barrier types in controlled conditions with respect to motorcyclists. It is
therefore difficult to make comparisons between barrier types regarding their relative safety.
Roads should be designed so that the risk of motorcyclists running off the road is minimised. Measures that
should be considered include:
• adequate and consistent skid resistance and elimination of loose gravel on road surfaces
• avoiding variations in superelevation through curves
• provision of a clear and smooth roadside to assist errant riders to recover or stop without serious injury
• ensuring that utility service covers are constructed and maintained so that they are flush with the road
surface
• minimising the number of rigid road furniture supports adjacent to the road.
Road safety barrier systems designed to minimise the severity of motorcycle crashes
A number of methods designed to improve existing road safety barriers to better protect motorcyclists have
been developed (Koch & Schueler 1988, Sala & Astori 1998). The methods generally involve use of a
proprietary product that may provide:
• additional rails or attenuation cushions on the lower section or other components of the barrier system so
that riders do not impact hazardous features including the posts
• a specifically designed covering of energy absorbing material for existing posts
• devices to shield sharp edges (e.g. post caps).
The use of enhancements to barriers is a matter for each jurisdiction and may be conditional on crash testing
and the proposed devices not creating other problems (e.g. related to debris or drainage). The use of
barriers and devices to improve motorcyclist safety may be considered by jurisdictions particularly on popular
motorcycling routes and areas considered to be high-risk (e.g. on the outside of curves).
Pedestrians and cyclists
Need for a safety barrier
Pedestrians or cyclists may require shielding by a road safety barrier in situations where they are considered
to be exposed to a higher than normal risk of being struck by an errant vehicle. Where a pedestrian/cyclist
facility either exists or is proposed for an existing site that has a run-off-road crash history, an assessment of
pedestrian, cyclist and bystander exposure should be undertaken so that crash reductions for alternative
treatments can be considered.
For new works, the protection of pedestrians and cyclists from passing traffic may also be considered and
should be investigated by undertaking a risk assessment to assess the likelihood of the encroachment of
errant vehicles into proposed pedestrian/cyclist facilities. Consideration also needs to be given to providing
protection for the pedestrians and cyclists from the barrier as they travel along the path.
When considering the need to protect pedestrians and cyclists the designer should consider the combination
of factors including the:
• number and type of path users (e.g. whether large numbers of people congregate in or pass through the
area, the presence of primary school children)
• factors that make the site more hazardous than other sites along the road (e.g. road geometry and
characteristics that would increase the risk of run-off-road events)
• type of traffic that may cause a run-off-road event to be particularly severe (e.g. high numbers of heavily
laden freight vehicles).

Situations where a road safety barrier may be appropriate are:

• intermediate and high-speed roads (i.e. posted speed limit of > 60 km/h) where a pedestrian or bicycle
path is close to the road
• shared-use paths separated by less than 4 m from an adjacent heavily trafficked lane, especially if the
geometry is substandard. However, designers should refer to Section 5.3.4 for discussion on lateral
location issues.
• sites where there is expected to be large numbers of bystanders congregated adjacent to the road
(e.g. schoolyard, sporting facilities) and the consequences of a crash are expected to be high.
Treatment options
Where practicable the preferred options for treatment are:

• design and management of the road to minimise the likelihood of encroachment into the roadside by
motor vehicles
• location (or relocation for existing facilities) of the pedestrian/bicycle facility away from the road where it
has a low probability of encroachment by errant vehicles
• provision of a road safety barrier. The deflection of the barrier should be taken into account, and the
barrier should be placed as close as practical to the vehicle travelled way.
Cyclists and pedestrians may require a barrier to prevent them inadvertently running onto a traffic lane from
an adjacent shared path (e.g. footpath on a bridge with high numbers of young pedestrians/cyclists). In
cases where there is no need to protect path users from errant vehicles, or errant vehicles from roadside
hazards, a pedestrian fence of a suitable height for cyclists should be adequate.
Design considerations
Where there is a need to provide a road safety barrier between a path and road traffic it is important that the
rear of the road safety barrier is not a hazard for pedestrians and cyclists. Designers should ensure that:
• adequate clearance is provided between the rear of the road safety barrier and the path, refer to AGRD
Part 6A (Austroads 2017a)
• no sharp edges, burrs or other potential hazards (e.g. protruding bolts) exist
• where sufficient clearance cannot be provided, suitably designed rub rails should be used to shield
cyclists and pedestrians from ‘snagging’ on posts
• where sufficient clearance cannot be achieved, consideration is given to the need to increase the height
of the barrier either to prevent errant cyclists from falling over the barrier and into a traffic lane or to
discourage pedestrians from jumping over the barrier to cross the road at an unsafe location.
Where sufficient space is available a frangible pedestrian fence may be erected behind the road safety
barrier at a distance that would accommodate the likely deflection of the barrier under impact by an errant
vehicle. Adequate clearance is also required between pedestrian fences and bicycle paths and shared paths.
In situations where space is restricted, it may be necessary to consider provision of a higher rigid barrier.
Designers should ensure that any modification or attachments to a barrier would not be detrimental to its
performance under vehicle impact or result in components being hazardous to motorists or path users in the
event of a crash with the barrier (e.g. horizontal rails spearing vehicles). AS/NZS 3845.1:2015 requires that if
an attachment is connected to a barrier, the barrier with the attachment should be crash tested.
Where pedestrian facilities are incorporated behind a road safety barrier system, the desirable minimum
height is to be 1200 mm above the surface of the pedestrian path. Where provision for cyclists is required,
the desirable minimum height above the surface of the path should be 1400 mm. Refer also to
AS 5100.1-2017: Bridge Design: Scope and General Principles and AGRD Part 6A (Austroads 2017a).

Separate rails may be provided to meet these requirements provided they do not have the potential to spear
through an impacting vehicle, create debris that poses a serious hazard, or change the characteristics of the
system to the extent that crash outcomes are significantly altered.
Pedestrian and cyclist access through barriers
Preferred practice is to avoid providing breaks in a road safety barrier. However, it may be necessary to
consider breaks at locations where pedestrians cross the road and where breaks are necessary, barriers
may be overlapped and/or safe end treatments must be provided.
Bridges and overpasses
AS 5100.1-2017 and the Bridge Manual (NZ Transport Agency 2013) provide information on barriers for
bicycle and pedestrian bridges and for some design elements for bicycle/pedestrian paths as they relate to
bridges.
5.2 Factors Considered in Barrier Selection
This section provides general guidance for the initial selection of longitudinal barrier systems, remembering
that the best solution is one that provides the required degree of shielding at the lowest whole-of-life cost for
the specific application (AASHTO 2011a). It should also be noted that end treatments are covered in
Section 5.3.21 and Appendix I.
The various factors that should be considered in the selection of the type of barrier to be adopted are
summarised in Table 5.1.
Table 5.1: Selection criteria for roadside barriers
Criteria Comments
1. Performance Barrier must possess sufficient structural integrity to contain and redirect the design vehicle.
capability
2. Deflection Expected maximum deflection of a barrier should not exceed available room to deflect.
3. Site conditions Slope approaching the barrier and distance from the carriageway may preclude use of
some barrier types.
4. Compatibility Barrier must be compatible with the planned end anchor and be capable of having suitable
transition segments (i.e. of adequate stiffness) installed to join to other barrier systems
(such as bridge railings).
5. Cost The cost of barrier systems varies but high-performance barriers designed to contain and
redirect heavy vehicles generally cost significantly more.
6. Maintenance The cost of repair has to be assessed independently.
a. Routine Few systems require a significant amount of routine maintenance.
b. Collision Generally, flexible systems require significant repair after a collision, semi-rigid systems
have fewer repair requirements and rigid systems or high-performance railings require an
even smaller amount of repair.
c.Materials storage The fewer different systems used, the fewer inventory items and the less storage space
required.
d. Simplicity Simpler designs, besides costing less, are more likely to be constructed and repaired
properly by field personnel.
7. Aesthetics Occasionally, barrier aesthetics is an important consideration in its selection.
8. Field experience The performance and maintenance requirements of existing systems should be monitored
to identify problems, especially those which could be lessened or eliminated by using a
different barrier type.
Source: Adapted from AASHTO (2011a).

The appearance of a road safety barrier system may also be an important issue at some sites where
compatibility with other architectural or geological features is essential.
The first two factors from Table 5.1 are discussed in more detail in Section 5.3 as part of the design process.
However, the following information is provided for the other factors that may need to be considered.
5.2.1 Site Conditions
The key site factors that need to be assessed include road geometry, offset distances and cross-slopes.
Road geometry
• Some systems, such as wire rope barriers, have restrictions in regard to their use where the horizontal
and vertical alignment standards are less than that specified by the manufacturer.
Offset
• To prevent an errant vehicle from travelling into the roadside area and minimise the number of incidental
impacts on the safety barrier, it is desirable that the face of a barrier be located at the back of the
shoulder (to a maximum of 4 m from the edge of the traffic lane). This will maximise the chance of the
driver being able to regain control of the vehicle and also minimise the length of barrier required and the
hazard it presents. It should be understood that a greater offset from the edge of the lane could result in
larger impact angles, higher impact severity and a higher probability of the barrier being penetrated.
• It is essential that the most appropriate barrier be selected to suit the particular site. Rigid barriers should
generally be located between 1 m and 3 m (and no more than 4 m) from the edge of the through lane as
the angle of impact for errant vehicles may increase with the offset. At increasing impact angles, the rigid
barrier profile becomes ineffective and injury severity increases. Also refer to Section 5.3.14.
• When located on the inside of horizontal curves, barriers may need to be offset further from the edge of
the traffic lane so that they do not impede horizontal sight distance, refer to AGRD Part 3
(Austroads 2016a).
• Sufficient width should be provided between the barrier and the traffic lane to enable stationary vehicles
to park clear of through traffic in an emergency stopping situation. Consideration should be given to
sealing the shoulder.
• Sufficient width should be provided between the rear of the barrier and a hazard so that the barrier can
deflect without the barrier or the impacting vehicle coming into contact with the hazard or infrastructure.
• When barriers are used to shield embankments, consideration needs to be given to the provision of
adequate ground support as over time softening of the verge may occur. For example, where the
restraining mechanism is supported on posts, a clearance of not less than 500 to 600 mm from the rear of
the post to the top hinge point of a fill embankment should be provided, although this may vary due to soil
conditions, batter slope, post depth, and other factors. In situations where post restraint is of concern,
deeper post embedment, closer post spacing or the use of soil plates may be considered. A soil plate is
attached to the bottom end of the post to increase the area of post available to resist moment forces
arising from vehicle impact. Reference should also be made to the manufacturer’s specifications,
although, road agency requirements supersede them.
As a starting point, the offset to the face of the barrier should be the same as the shoulder width.
Cross-slopes
Irrespective of the type of barrier being used it is preferable that the approach slope be a projection of the
crossfall because barriers perform best when they are impacted by vehicles with their centre of gravity at or
near the normal position.
In general, semi-rigid and flexible barriers should not be used where the slope in front of the barrier is steeper
than 10:1. An exception is where it is necessary to provide a drain in a median and also to locate a barrier
within the median in which case a 6:1 slope may be acceptable, but only if it is not practicable to achieve 10:1.

Compatibility
As a general practice, road agencies use a limited number of different, proven barrier systems on new
construction and reconstruction. This practice has advantages in that maintenance personnel need to be
familiar with only a few systems and stocks of replacement parts are more easily managed. Non-standard or
special barrier designs need only be considered when site characteristics or performance requirements
cannot be met with standard systems.
Cost
The selection of a barrier should consider the life-cycle cost of the systems and their safety performance,
including injury, property damage and maintenance costs. The initial capital cost of the barrier is only one
component of economic evaluation; however, this is not to say that the initial cost of the system is not an
important budgetary and project management consideration.
The choice of a terminal treatment may also be a significant factor with respect to the cost of the system.
Maintenance
A barrier can perform as intended only if it is properly installed and maintained. Maintenance factors that
need to be considered are:
• routine maintenance of the barrier
• impact repair
• effect of the barrier on adjacent road and roadside maintenance (e.g. pavement overlays)
• material and component storage requirements.
Aesthetics
Aesthetics are not normally an over-riding factor in the choice of barrier. However, greater importance is now
being placed on the aesthetics of barriers, especially in recreational and tourist areas. Section 5.5 provides
some information on the aesthetics of barriers. It may also be preferred for aesthetic reasons that a particular
type of barrier is used consistently along a road or road segment.
Field experience
There is no substitute for documented evidence of a barrier’s performance in-service on the road. This
information provides feedback to designers and construction personnel on various types of barrier in typical
situations. It is particularly important that road agencies learn from both observing the results of impacts with
barriers and examining crash reports.
Environmental impact
Apart from the aesthetic appeal of the barrier, other environmental factors that may require consideration are:
• Barriers that have a larger frontage area may contribute to a build-up of drifting snow or sand, thereby
affecting operation of the road and potentially the effectiveness of the barrier.
• The use of certain preservatives in some wooden barriers or barriers that have wooden components may
be an issue.
• Some types of steel railing may deteriorate rapidly in highly corrosive environments.
• Barriers may block tourists’ views of scenic panoramas, or a driver’s sight distance.
• Fauna migration patterns may be affected.

Limitations on the use of barrier types
There are some key factors (Table 5.2) that must be considered in the selection of a barrier type to ensure
that it is suitable for the particular circumstances. However, it is important that designers also consult the
relevant national or state/territory guidelines in choosing a barrier. It should be noted that some products may
also be assessed and considered to be suitable to use on Australian and New Zealand roads by the
Austroads Safety Barrier Assessment Panel and some road agencies may also have a product acceptance
process for barriers proposed for use on their roads.
Table 5.2: Key considerations in barrier selection
Type of
Consideration
barrier
All When selecting the type of safety barrier system to be installed, designers should consider the types of
road users that are expected to travel along the road, e.g. heavy vehicles and vulnerable road users such
as motorcyclists and take them into account in assessing the most appropriate system.
Flexible Ground slope:
• The maximum lateral slope on which wire rope road safety barriers (WRSB) should be installed is
typically 10:1.
• If it is proposed to install WRSB on steeper slopes, all relevant factors must be considered including
confirmation from the distributor/manufacturer that the proposal is acceptable.
It should be noted, however, that flexible barriers have the advantage of smaller exit angles compared to
rigid and semi-rigid barriers. Recent developments have seen relatively larger post spacing proposed for
flexible barriers (i.e. > 4 m), which may be advantageous in some situations and may result in a significant
increase in deflection. However, this is only one of many factors to be considered.
Horizontal curves:
• Careful consideration is required where the horizontal radius is less than 200 m because the required
rope tension and height may not be maintained during or after an impact.
• Designers should consult with WRSB manufacturers where it is proposed to install it on curves less
than 600 m radius.
Vertical curves:
• Road designers should be aware that there might be limitations regarding the use of a flexible barrier
on vertical crest and sag curves. For example, on sharp sag curves the tension in the ropes may cause
the posts at the bottom of the dip to lift out of their sockets, especially in cold weather. As specific
requirements may apply to particular products, designers should refer to product information and
jurisdictional guidelines.
• A sag curve, combined with the possibility of the suspension of an errant vehicle being compressed at
the bottom of a vertical sag curve, may lead to an occurrence where the vehicle body passes under the
ropes, instead of being caught on them. The ropes may then encroach into the turret of the vehicle
causing injury to the occupants.
Transitions:
• WRSB systems should not be installed so that they connect directly to any other barriers or bridge
parapets. The deflection inherent in the design cannot ensure that vehicles colliding in the transition
area between the rope barrier system and another system will be redirected safely.
• WRSB may be installed in close proximity to rigid or semi-rigid barriers provided that there is sufficient
distance between the barriers to accommodate the dynamic deflection.
Barrier length:
• The minimum length of WRSB at full height should comply with the road agency’s acceptance
conditions. This length does not include the transition from full height to the end anchors.
• In assessing the maximum length of WRSB between end anchorages and the spacing of intermediate
anchorages, the designer should consider the effect of barrier length on the maximum deflections and
the risk of long lengths of barrier being made ineffective due to an impact at the barrier terminal. The
manufacturer and road agency should be consulted when determining anchorage spacing.

Type of
Consideration
barrier
Semi- Horizontal curves:
rigid • W-beam and thrie-beam barriers perform well on the outside of curves, even those of relatively small radius,
as the concave shape (in plan view) supports the development of tension in the rail.
• The convex (plan view) when used on the inside of small radius curves can mitigate against the
development of tension in the rail. However, this is usually only a problem for very small radii such as those
on the corners of intersections (refer to Section 5.6.1 for which appropriate designs have been developed).
Barrier length:
• As a general guide, 30 m can be taken as the minimum length of semi-rigid barrier that should be installed.
Kerbs:
• Where a kerb exists at the edge of the road, semi-rigid barrier must either be placed within 200 mm of the
face of the kerb or a distance behind it to ensure that impacting vehicles do not vault over the barrier (see
Commentary 6).
[see Commentary 6]
Rigid Horizontal curves:
• Rigid barrier should generally not be used in situations where it is likely to result in impacts occurring at
high angles as this could subject vehicle occupants to high-severity crashes.
• Where practicable, the use of rigid barriers on the outside of small radius horizontal curves should be
avoided for similar reasons. However, it is acknowledged that this is not possible in all situations,
particularly adjacent to ‘loop' ramps at urban freeway interchanges (although impact speed should be
relatively low in this situation).
Length:
• The minimum length requirement for a rigid barrier in order to provide adequate contact length is 20–30 m.
Drainage:
• Outlets should be provided to discharge storm water from the road pavement.
5.3 Road Safety Barrier Design Process
5.3.1 Outline of Process
Once a road safety barrier has been clearly established as the preferred treatment using the hazard mitigation
process in Section 3.5, a design can be undertaken in accordance with Figure 5.1. In designing a system, it is
important that all system components meet the required test level over the entire length of the barrier.
Figure 5.1 shows that the design process includes:

1. collection of information about the site such as geometry, speed zoning, AADT
2. a clear understanding of the objectives of the proposed barrier
3. choice of a trial lateral location for the barrier
4. confirmation that the clearance between the barrier and the hazard can accommodate the required
working width for the chosen barrier type
5. the choice of a type of barrier for further consideration
6. where necessary, choice of an alternative barrier type and a repeat of points 3 and 4
7. development of detailed aspects of the barrier design such as the:
a. transverse location of the barrier and any site modifications necessary to ensure that impact height
criteria are met
b. points of redirection and length of need
c. treatment of leading and trailing terminals
d. details of interfaces where different types of barrier systems meet (such as a road safety barrier
meeting a bridge barrier).

At all sites where hazards have been identified, it is important to realise that drivers expect a consistency of
treatment (in accordance with current design practice), along segments of road that have similar features.
Steps B1 to B3 in Figure 5.1 are discussed in Sections 5.3.2 to 5.3.4, respectively and Steps B4 to B16 are
discussed in Sections 5.3.12 to 5.3.23, respectively.

Figure 5.1: Road safety barrier design process
Source: Adapted from Roads and Maritime Services (2016).

5.3.2 Collect Site Information (Step B1)
The information that may be required to commence the process includes:

• general site details
• the size and position of identified hazards requiring protection or shielding (length, width and offset from
the carriageway) and their risk levels
• detailed topographic information such as embankment details, lateral widths, crossfalls, topography at the
leading end of the barrier system and any restrictions on the use of gating terminals
• features of the site which could pose difficulties such as public utilities, access to property, drainage
installations, site geology, maintenance access requirements to road furniture and the like
• locations where restrictions to sight distance could be critical, such as curves and intersections
• the location, type and condition of any existing barriers
• traffic volume and mix, including pedestrians, bicycles, motorcycles, heavy vehicles
• design vehicle (e.g. mass and principal dimensions)
• intended design speed for the road
• impact speed and impact angle
• the nature of the ground in front of the barrier system, sufficient to allow the likely approach elevation of
an errant vehicle to be established
• provisions for access (e.g. emergency vehicles, pedestrians, fauna crossings)
• period of time the barrier system will be required to operate
• existing delineation
• operational temperature range
• flooding.
5.3.3 Determine the Objectives of the Safety Barrier (Step B2)
From knowledge of the site, the designer should determine the objectives of the proposed barrier, by
considering who (persons-at-risk) or what vulnerable site is to be protected or shielded.
In the majority of cases, the primary persons-at-risk are the occupants of errant vehicles. The vast majority of
barrier installations are provided for the purpose of redirecting errant vehicles away from hazards thereby
reducing the risk of severe consequences.
In some cases, however, the primary persons-at-risk might be the occupants of a property (e.g. children in a
child care centre) in a vulnerable location such as at a curve or at the bottom of an embankment. In this case
the occupants of the errant vehicle are the secondary persons-at-risk.
In setting the objectives of a proposed barrier system, designers should consider the possible crash events
at a site and identify the range of possible outcomes for impacts by all types of road users. Traditionally,
most barrier systems have been tested with typical passenger cars striking the device at relatively small
angles of impact. However, systems may have reduced performance if impacted by:
• cars at larger angles of impact and at speeds higher than barrier test speeds
• heavier vehicles
• bicycles or motorcycles.
Similarly, on low-volume and low-speed roads, a system that is less expensive than one designed to cope
with the higher loadings may be appropriate.

5.3.4 Determine the Lateral Position of the Barrier (Step B3)
Factors considered in lateral location
Barriers are a roadside object that can be hazardous to errant vehicles and the lateral location of the barrier
can affect the safety outcome. Designers should understand that a barrier located:
• relatively close to the traffic lane has an advantage in that vehicles will impact it at a low angle and hence
the impact will be less severe
• a greater distance from the traffic lane is likely to result in less impacts but those that do occur will be at a
higher angle and hence result in a more severe impact
• too close to the traffic lane is likely to be impacted more frequently, cause more damage to vehicles and
the barrier and hence result in higher property damage costs for road users and higher maintenance
costs for road agencies.
In addition to operational considerations, a barrier and its footings should not:

• interfere with any utilities, drainage conduits or structures
• impair access of personnel or machinery to any utilities, drainage conduits or installation, or structures.
Barriers that pass the test requirements in AS/NZS 3845.1:2015 and are assessed by the Austroads Safety
Barrier Assessment Panel (ASBAP) may be approved for use by road agencies. The tests involve standard
test vehicles impacting a barrier with a generally flat surface in front of the barrier and behind it. While it is
preferable that this condition is replicated, in practice this is not always possible (e.g. where it is necessary to
provide kerbs).
The lateral position of a barrier is influenced by the:

• road cross-section (e.g. need for shoulder and/or kerb)
• barrier-to-hazard clearance
• trajectory of vehicles when crossing kerbs and slopes
• desire to avoid nuisance damage
• maintenance access.
Figure 5.2 shows a typical lateral location of a barrier in the verge of a road. It can be seen that the position
of the barrier may be dependent on the:
• offset to traffic lane
• system width
• support width.
In some cases, the distance between the edge of the traffic lane and the hazard may be limited in which
case the designer will have to consider how the available space will be best utilised and what type of barrier
is most suitable for the particular situation.
Sections 5.3.5 to 5.3.12 discuss the key factors to be considered in determining the appropriate lateral
location for a barrier situated in road verges and medians. Figure 5.2 illustrates some of the cross-section
elements relating to verges.
The area between the traffic lane and the front face of a non-rigid barrier should be a trafficable surface with
a crossfall ≤ 10:1. The area between the face of the barrier, to the full extent of the working width, should be
in accordance with the manufacturer’s specifications.

Figure 5.2: Verge barrier location
1 This slope is usually determined by the crossfall of the shoulder or lane. Details of crossfall and batter rounding are
provided in Section 4 of AGRD Part 3 (Austroads 2016a).
2 The offset to the hinge point may be reduced where there is no other option.
When a vehicle passes over a slope or kerb it may follow a trajectory that might influence the height required
for the safety barrier, should it be necessary to place a barrier on the embankment slope or some distance
behind the kerb. The lateral position of the barrier may therefore be restricted when it is located:
• on embankment slopes
• adjacent to kerbs
• on cutting slopes
• within medians.
Figure 5.3 illustrates an example of a vehicle trajectory over an embankment slope where a barrier placed
within the distance ‘L’ may need to be higher in order to contain a car. Commentary 6 provides further
information on the subject and examples of trajectories and heights of the trajectories where cars traverse
embankments, kerbs and cutting slopes. It also illustrates suggested locations where a barrier should
preferably not be erected.

Figure 5.3: An illustration of bumper height trajectory characteristics over a fill embankment
S1 H1
HM
HM
H1
S1
Source: Based on Roads and Maritime Services (2016).
5.3.5 Offset to Traffic Lane
Road user requirements
The offset from the edge of the traffic lane to the face of the barrier will depend on the type of road and its
use. Rural roads generally require shoulders that have several purposes while urban roads may require the
provision of parking lanes, bus lanes or bicycle lanes. Guidance on cross-section requirements is provided in
AGRD Part 3 (Austroads 2016a).
When roadside features such as bridge railings, parapets, retaining walls, fences or roadside road safety
barriers are located too close to traffic, drivers in the adjacent traffic lane tend to reduce speed, drive
off-centre in the lane, or move into another lane. The distance from the edge of the traffic lane beyond which
a roadside object will not be perceived as an obstacle and result in motorists changing their behaviour is
called the shy-line. Where possible, barriers should be located outside of the shy-line, particularly where
relatively short lengths are used.
Depending on circumstances it may be preferable to provide the same shoulder width (if applicable) adjacent
to barriers as is provided elsewhere along a road. However, jurisdictional practice may require consideration
to be given to the provision of a wider shoulder (e.g. 3 to 4 m from the edge of the adjacent traffic lane to the
barrier) in order to provide space for nearside vehicle doors to be opened or to provide space for
maintenance vehicles to stand clear of the traffic lane. This width enables the doors of vehicles to be opened
clear of traffic lanes in the case of non-discretionary stops.
Where space is limited, and discretionary parking or emergency stopping is not essential, it may be
preferable to provide a reduced shoulder width in front of the barrier, provided that the shy-line principle is
given adequate consideration. Designers should comply with any requirements that jurisdictions may have
regarding the minimum offset to barriers. Consideration should be given to sealing the shoulder for its full
width where the barrier is installed at the edge of the shoulder.

The offset from the edge of the traffic lane to the face of the barrier should be such that the available
stopping sight distance and intersection sight distance are not impeded. It is not possible for drivers to see
through roadside barriers.
Designers should be aware of the following risks if barriers are placed close to the traffic lane:
• There is an increased risk of sideswipe crashes if narrow lane widths are used next to barriers. This
occurs because drivers tend to move away from the barrier and may encroach into adjoining traffic lanes.
• Drivers will travel at moderate speed close to long lengths of barrier; however, this is generally only
successful in a high-stress driving environment (e.g. tunnels and bridges) where drivers are attentive and
ready to react quickly to risks.
• Driving close to barriers increases the stress of the driving task and cannot be sustained for long periods.
• Barriers close to the traffic lane have a high probability of nuisance impacts that will increase the need for
maintenance and repair. Barriers that remain operational after an impact should be used where there is a
high risk of repeat impacts before repairs can be made.
• Short lengths of barrier close to the traffic lane have a high risk of being impacted because of driver
fatigue or inattentiveness.
In summary, the underlying principle when designing the layout of a roadside barrier is to provide a distance
between the barrier and the running lane that:
• allows the driver to avoid colliding with the barrier in minor encroachments
• provides some space to reduce speed before impact in major encroachments
• ensures better sight distance at intersections, accesses and around horizontal curves
• allows vehicles to stand clear of the adjacent traffic lane after impact
• provides an opportunity for disabled vehicles to stop clear of the running lanes.
Maintenance access
In some situations (e.g. high-volume, high-speed roads) consideration may be given to offsetting the safety
barrier a sufficient distance from the road (e.g. 4 m) to enable a road maintenance truck to stop clear of
general traffic while inspecting or undertaking routine maintenance of the barrier or other roadside features.
Barrier setback from kerb
As a general principle, it is preferable that surface conditions in front of and beneath barriers should be
similar to the conditions under which barriers are tested (i.e. relatively flat). While this is usually possible on
roads that have a rural cross-section, there are instances where kerbs are required, and urban roads almost
always have kerbs.
In rural situations drainage should be designed so that it is not necessary to place a kerb under or in front of
a barrier.
Kerbs should not be placed in proximity to barrier systems, particularly in environments with speed limits
above 70 km/h. Kerbs in front of barrier systems will have a significant effect on the performance of the
vehicle as it approaches the barrier with potentially hazardous consequences. It is essential that the barrier
system should be located and designed to ensure that the errant vehicle does not:
• vault over the barrier
• go under the barrier causing snagging of the barrier supports and other problems.

The decision about location is made more difficult by the wide range of factors influencing the behaviour and
trajectory of errant vehicles (e.g. suspension stiffness, vehicle weight, speed of impact, angle of impact).
Locating kerbs in front of barriers makes it more difficult to ensure that the two conditions noted above are
achieved.
Further, on high-speed facilities, placing kerbs close to the traffic lane introduces an additional hazard with
little benefit for the traffic stream as they:
• do not influence driver behaviour prior to the deviation of the vehicle
• do not redirect errant vehicles after impact with them
• may cause a driver to lose control after impact
• may cause the vehicle to leave the ground after impact thereby changing the trajectory of the exit path.
Where it is necessary to place a steel beam barrier close to a kerb it is preferable that the face of the barrier
is aligned vertically with the face of the kerb. Where this is not practicable because of design constraints, or
there is a high probability of minor nuisance vehicle impacts (e.g. where the kerbside lane is narrow) the
barrier may be offset a small distance, with the preferred option being 200 mm, behind the kerb as shown in
Figure 5.4. The 200 mm offset minimises nuisance impacts while still reducing the possibility of a vehicle
snagging under a barrier element. This offset may be determined by footing requirements or the need to
avoid nuisance impacts as detailed in Table 5.3.
The offset for wire rope safety barrier (WRSB) depends on the kerb type adopted and the size of the footing
required. The preferred kerbing type adjacent to flexible and semi-rigid barriers is Type A 100 mm mountable
kerbing. A semi-mountable kerb may be suitable in some situations (typically speeds < 70 km/h).
For new roadworks where road safety barrier is used in close proximity to a kerb it is desirable that a
semi-mountable kerb profile is used.
Offsets greater than those shown in Table 5.3 may have an advantage in that they can:
• reduce nuisance impacts on the barrier
• avoid the situation where vehicles that have impacted the barrier are disabled in a high-speed traffic lane,
which increases the risk of a secondary crash.
However, kerbs located in the vicinity of a barrier can affect the vehicle trajectory and hence the
effectiveness of a barrier. For this reason, it is suggested that, as a general guide, a barrier (other than a
rigid barrier) should not be located between 0.2 and 3 m (upper limit depends on speed) behind the kerb.
Designers should refer also to Commentary 6 where the behaviour of vehicles crossing kerbs is discussed
and further guidance is provided on the location of barriers in relation to kerbs.
Concrete barriers should not be located behind kerbs.
If there is no alternative to placing the kerb in front of a barrier, further guidance regarding the position of
kerbing is provided in Commentary 6.

Figure 5.4: Barrier offset at kerb
(a) Steel beam barrier (b) Wire rope barrier
Table 5.3: Minimum offsets from kerb to barrier face
Situation Offset from kerb face to barrier (m)

Wire rope barrier 0.3 m(1) generally over length of barrier
Larger offset may be required at terminal to accommodate foundation(1) (e.g. say
0.7 m – based on half the anchor width + width of kerb)
Steel rail barrier (e.g. W-beam, 0.0 m (increased nuisance hits may occur)
thrie-beam), constrained situation
Steel rail barrier (e.g. W-beam, 0.2 m
thrie-beam barrier), normal situation
Concrete barrier 0.0 m (barrier is the kerb)
Barrier on traffic island 0.2 m + bus overhang when a wheel is at the kerb
1 Varies depending on product and foundation required – consult manufacturer’s drawings.
Source: Roads and Traffic Authority (2008).
Shy-line offset
Generally, there should be uniform clearance between traffic and roadside features, particularly in urban
areas where there are many roadside features. Uniform alignment enhances road safety by providing the
driver with a certain level of expectation, thus reducing driver concern for and reaction to the roadside
features. The distance from the edge of the traffic lane beyond which a roadside feature will not be perceived
as an obstacle and result in motorists reducing speed or changing lanes on the road is called the shy-line
offset (adapted from AASHTO 2011a).
Where long continuous lengths of barrier are used, this shy-line effect is not so critical, especially if the
commencement of the barrier can be gradually transitioned from beyond the shy-line.
AASHTO (2011a) provides the offset values shown in Table 5.4 and suggests barriers with a lateral offset
less than the shy-line offset should have a flare on the approach as shown in Figure 5.5.
There is also some evidence that the presence of a safety barrier may influence the operating speed of
traffic. Research (Tay & Churchill 2007) has shown that the mean traffic speed on roads with median barriers
is higher than similar roads without median barriers. Experience in Sweden (Bergh & Carlsson 1999) is that
traffic speeds increased on a narrow road after construction of a median wire rope road safety barrier. This
research has called into question previous assumptions that drivers will slow down near barriers.

Table 5.4: Shy-line offset values
Design speed (km/h) Shy-line offset (m)

50 1.1
60 1.4
70 1.7
80 2.0
90 2.2
100 2.4
110 2.8
120 3.2
130 3.7
Source: AASHTO (2011a).
Flaring
Motorists are less likely to perceive roadside barriers to be a hazard if the barrier is introduced gradually to
the roadside environment through the use of a ‘flare’. Consequently, some end treatments for semi-rigid
barrier (i.e. W-beam) are designed to be flared away from the approaching traffic as shown in Figure 5.5.
The flare rate is the ratio of the length of the flared part of the barrier (measured parallel to the road) to the
barrier offset.
Designers should provide a flare where barriers are within the shy-line offset, provided that other design
requirements are not compromised (e.g. some barrier terminals need to be parallel to the roadway).
Flaring should be used to:

• locate a barrier terminal further from the travelled path
• minimise shy-line effects where a hazard is close to the travelled path
• provide a gradual transition to a major hazard close to the roadway (such as a bridge parapet or railing)
• provide a transition from the centre median placement to the dual side median placement.
By adopting a flare, the length of barrier required can be significantly reduced. In some situations, placing the
barrier on a flare may require widening of the road embankment. The additional cost of widening of the
embankment should be considered against the extra cost of the barrier and the ongoing maintenance costs.
The maximum flare rates that should be used on an approach to a barrier are shown in Table 5.5. Caution
should be used in applying flaring as barriers are designed to work best with a glancing impact. Flaring may
lead to vehicles impacting a barrier at a high angle that will lead to hazardous consequences.
It should be noted that if a flare is used the terminal specified must suit the flare. Practitioners should refer to
the manufacturers’ published information for proprietary systems and the road agency information for public
domain devices.

Figure 5.5: Detail of flare rate
Notes:
 For angle ‘a’ refer to Table G.1.
 Flare rates expressed as d:1 and shown in Table 5.5.
Source: Adapted from Roads and Maritime Services (2016).
The flare rates applied should not enable impacts with the barrier to occur at an angle greater than 25°.
Barrier systems are not tested at impact angles greater than 25°, as the barrier may rupture or cause
excessive vehicle damage and occupant injury.
Table 5.5: Flare rates
Flare rate for rigid Flare rate for non-rigid

Flare rate for barrier
Design speed barrier outside the barrier outside the
within the shy-line offset
(km/h) shy-line offset shy-line offset
(d:1)
(d:1) (d:1)
50 13:1 8:1 7:1
60 16:1 10:1 8:1
70 18:1 12:1 10:1
80 21:1 14:1 11:1
90 24:1 16:1 12:1
100 26:1 18:1 14:1
110 30:1 20:1 15:1
Notes:
 Non-rigid is equivalent to semi-rigid or flexible.
 Flare rate applies to approach flare only.
Source: Based on AASHTO (2011a).
Table 5.5 shows that the flare rate adopted depends on whether the barrier is located within or outside the
shy-line, and on the type of barrier. The flare rate values indicate a smaller flare angle for both types of
barrier when located inside the shy-line. Smaller flare angles should be used where extensive grading would
be required to ensure a low-angle approach to the barrier from the carriageway (AASHTO 2011a).
Flaring of barriers can have the following disadvantages:

• The greater the flare angle the higher the impact angle and the subsequent severity of crashes into that
part of rigid and semi-rigid barriers.
• The likelihood of a vehicle being redirected back onto the roadway following an impact with the flared
section is increased.
• Higher flare angles may also increase the need for additional earthworks and slope flattening in the area
between the roadway and the barrier.

5.3.6 Support Width
The minimum support width required between the rear of the barrier and the hinge point of the batter is
shown in Table 5.6. However, on batters steeper than 3:1 it may be necessary to provide structural support
to the barrier foundations.
Situations sometimes arise where a safety barrier must be provided to shield a foreslope at a site where
there are severe lateral constraints. In such cases the designer may consider the options for brownfield sites
described in Appendix F.
Table 5.6: Typical support width
System type Support width

Wire rope barrier 600 mm
W-beam barrier and thrie-beam 600 mm
Permanent concrete barrier 0 mm where barrier is part of a structure (e.g. retaining wall)
Where a concrete barrier is not part of a structure such as a noise wall it must have
adequate lateral support either from the pavement and/or its surfacing or the
surrounding ground. Depending on circumstances options may include:
• setting the base of the barrier into the road pavement
• setting the barrier into the asphalt road surfacing
• attaching the barrier to an existing road surface through a system of dowels
• provision of a foundation to support the barrier profile.
5.3.7 Deflection Width
The deflection width varies depending on the type of system and the particular product. The significance of
deflection width is discussed in Section 5.3.15.
5.3.8 System Width
The width of barrier systems varies depending on the type of system and in some cases the strength
required. Details of systems and products approved by the road agency, including the system width, may be
obtained from manufacturers’ specifications or information published by the road agency.
5.3.9 Barrier Location in Medians
The application of road safety barriers to medians depends on the median width and the cross-section.
Section 4.7 of AGRD Part 3 (Austroads 2016a) provides guidance on medians and median width.
Median barrier selection
Median barriers need to take into account the:

• impact severity at high speed which is a measure of the possible damage to vehicles and injury to occupants
• crash costs per accident including human and incident costs based on information from the road agency
• width required for system hardware which has an influence on the minimum median width required
• sight distance requirements and aesthetics, i.e. the effect that the barrier will have on driver sight distance
and the visual impact of the barrier
• drainage including the effects on surface water run-off
• requirements for barrier terminal treatments.
The safety benefits of safety barriers in narrow medians are described in Guidance on Median and
Centreline Treatments to Reduce Head-on Casualties (Austroads 2016b).

Minimum median width
The minimum median width required to accommodate a safety barrier depends on the overall width of the
barrier and the clearance required between the barrier and the edge of the traffic lane. The clearance will
depend on the dynamic deflection expected under impact by the design vehicle or a nominal minimum
clearance necessary for drivers to feel comfortable travelling adjacent to the barrier. Safety barriers in narrow
medians are illustrated in Figure 5.6. In general, it can be expected that concrete barriers will experience
virtually no deflection and that steel and wire rope barriers will deflect to varying degrees depending on the
system used (e.g. post spacing, stiffness). Deflection is discussed in Section 5.3.15.
Additional width will be required where a median barrier is located within a curve and the barrier will impede
sight distance to objects on the road or the brake lights of vehicles preparing to stop in the median lane
(e.g. due to an incident such as congestion or a crash). Some agencies may consider additional clearance is
desirable because of the likelihood of vehicles shying away from a barrier.
An important consideration with respect to the clearance to median barriers, particularly the more rigid
systems, is accessibility for maintenance crews undertaking repairs and the occupational health and safety
issues surrounding such activities.
Encroachment onto opposing carriageway
A vehicle hitting a WRSB may encroach beyond the line of the barrier, or it may cause barrier wires or posts
to encroach beyond this line. On narrow medians, this may allow the vehicle or part of the barrier to
encroach into the opposing carriageway and it is known that such incidents have occurred in Sweden. While
experience shows that the WRSB reduces the consequences of a head-on collision by reducing the speed of
the crash-causing vehicle, it is desirable to provide sufficiently wide medians to limit encroachment into an
opposing carriageway.
The probability of a collision due to encroachment after impact with a WRSB in a narrow median is related to
the probability of a vehicle being adjacent to the impact site during this short period of time and the design
deflection being exceeded.
Issues
Issues that need to be resolved in considering median treatments include:

• incident clearance
• width of median
• provision of additional lanes in future
• sight distance
• median break influence on median width (i.e. design requirements)
• road safety barrier options
• environmental impacts
• construction and maintenance
• delineation
• road user issues (e.g. pedestrians crossing roads)
• cost.
From a barrier treatment point of view, medians are classified in two categories: narrow medians in the range
1 to 4.5 m, and wider medians in the range 4.5 to 10 m.

Narrow medians
Narrow medians in urban areas are usually raised kerb medians and medians painted on the pavement. The
4.5 m limit is loosely based on the practice in urban areas to provide a minimum 1.0 m residual median and a
3.5 m right-turn bay within the median area.
Wider medians
Numerous studies (e.g. Austroads 2016b & 2017c) have shown that wider medians improve safety and that
90% of errant vehicles deviate less than 15 m from the edge of the carriageway. However, the marginal
effect of increased width drops off rapidly (80% of errant vehicles deviate less than 10 m) and, where land is
expensive, it is hard to justify widths greater than the minimum.
In most rural areas the cost of the wide median is small, and wide medians (≥ 15 m) are more usual as they
give more scope to the road designer in the treatment of the median and in the location of a barrier. Barriers
should always be considered on high-speed roads.
Wider medians are classified in three groups:

• depressed medians, or medians with a ditch section
• stepped cross-section medians where the separate carriageways are individually graded
• raised medians.
The use of barriers in narrow and wider medians is discussed below.
5.3.10 Narrow Medians
General
At many locations, the appropriate treatment is to provide a centrally located barrier, immediately behind the
shoulder, capable of being impacted from either side (e.g. urban roads where space is limited). Rigid barriers
are often chosen in these situations but back-to-back barriers or flexible barriers may be used where the
dynamic deflection can be contained within the median width so that an unacceptable risk does not
eventuate for opposing traffic. Typical profiles of median barriers for use in narrow medians are shown in
Figure 5.6. Appropriate end treatments must be used to suit each type of barrier and situation (i.e. width
available and behaviour of end treatment on vehicle impact).
Figure 5.6: Examples of road safety barriers for use in narrow medians
F type Single slope Back-to-back Wire rope

concrete concrete steel beam

Figure 5.7 details how a rigid object such as a bridge pier can be shielded in a narrow median. The treatment
will differ depending on whether the barrier only shields an isolated object or whether it is to be incorporated
into a longer barrier system.
Figure 5.7: An example of a barrier layout for shielding a rigid object in a median
Notes:
 Designers should also investigate the use of a crash attenuator to shield the hazard.
 The barrier tapers should be aligned to be consistent with the flare rates, as suggested in Table 5.5.
Source: Queensland Department of Transport and Main Roads (2013).
Minimum width of median
The factors to consider in the determination of minimum widths for medians with barriers are:
• the width of the system or width of the terminal, whichever is greater
• an allowance for permanent deformation of the system on straights or curves
• whether or not the damaged barrier components and debris will be contained within the median area
following impact
• the shoulder width required for delineation and accommodation of drainage grates
• an allowance for sight distance on curves
• the deflection width of the barrier
• stormwater capacity of the median drain.
It is preferable that deflection is contained within the median. However, in constrained situations where this is
not possible the deflection may be allowed to encroach into the opposing traffic lanes but only if a risk
calculation demonstrates that the level of risk is acceptable to vehicles travelling in the opposing direction.
Appendix F provides some information on the use of wire rope safety barriers in narrow medians.
When specifying barriers for narrow medians it is essential that appropriate approach delineation and signing
is included. For guidance on minimum median width designers should refer to AGRD Part 3
(Austroads 2016a).

5.3.11 Wider Medians
The most desirable median is one that is relatively flat (slopes of 10:1 or less), free of hazards and wide
enough to enable virtually all errant vehicles to come safely to rest without encroaching into the opposing
carriageway or having to be contained by a barrier. To fulfil this objective, a median would have to be very
wide, with Austroads (2014a) reporting that some vehicles were found to travel in excess of 50 m across
roadside areas. A method of preventing this cross-over movement is with the installation of a safety barrier
system on both sides of the median.
The location of the barrier should be 4 to 5 m from the traffic lane as this offset provides the greater safety
benefits compared with other offsets (Austroads 2014a).
A wire rope barrier located on both sides of the median has the advantage that it maximises the opportunity
to contain deflections within the median. However, a central location has advantages in that:
• Debris from damaged barriers is less likely to encroach into the carriageway (however may be more likely
to encroach onto the opposing-flow carriageway).
• Sight distance past the barrier on curves is maximised.
• The barrier sustains less nuisance impacts than a barrier on the side of the median.
• The cost is less than a barrier on both sides of a median.
5.3.12 Barrier-to-hazard Clearance (Step B4)
The barrier-to-hazard distance is the minimum clearance available between the proposed face of the barrier
and the face of the hazard. It is established either from the design plans or site measurements. The
barrier-to-hazard clearance defines the space available for designing a barrier installation.
5.3.13 Barrier Containment Level Required (Step B5)
General
The containment level relates to the objectives for providing the barrier (Section 5.3.3), the class of vehicle
that must be contained and corresponding barrier test levels (AS/NZS 3845.1:2015). For example, the
containment level may be chosen to protect:
• errant vehicle occupants
• a vital piece of infrastructure such as an electrical sub-station or a bridge pier from impact by a heavy vehicle
• an establishment such as a child-minding centre that is in a vulnerable situation with respect to errant vehicles.
Designers should refer to AS/NZS 3845.1:2015, policy information from jurisdictions and to barrier product
information available from suppliers to ascertain the required containment level and the type of barrier
needed.
The test level criteria for barriers specified in AS/NZS 3845.1:2015 is based on the criteria used in
AASHTO (2016) Manual for Assessing Safety Hardware (MASH).
In the absence of jurisdictional policy on the containment level required for particular situations, the level may
be determined by taking into account the:
• design vehicle to be retained by the road safety barrier
• speed environment of the barrier site
• consequences of penetration e.g. whether the consequences of a barrier penetration are limited to a
single vehicle
• adjacent land use e.g. whether there are generators of pedestrian or cyclist traffic such as schools,
playgrounds, parks, shopping centres that might increase the probability of an errant vehicle hitting a person.

Road databases or traffic measurements can be used to determine the volume of heavy vehicles that use
the road and whether or not heavy vehicles are to be used as the design vehicle. This decision will also be
based on the consequences of a truck or bus impact.
Design for heavy vehicles
When considering the adoption of a heavy vehicle as the design vehicle, it should be noted that:
• Road safety barriers suitable for all trucks are rarely used and are expensive.
• A heavy vehicle will not be contained by a Test Level 3 (TL3) barrier.
• A car may be extensively damaged by impact with a barrier designed for trucks.
Except for barriers associated with bridges (refer to AS 5100.1-2017), and other situations where the
consequences of vehicles leaving the road are extreme, barriers are not normally designed to contain van or
tanker-type semi-trailers (TL5 and TL6). This design limitation has been practised primarily because of the
relatively low volumes of these vehicles on many roads and the high cost of providing barriers to contain
them. The increased severity of passenger car crashes into high-containment barriers is also an important
consideration. Designers should therefore consult any jurisdictional policy that may be available with respect
to the use of high-containment barriers.
Where a risk assessment indicates that the run-off-road risk associated with heavy vehicles is particularly
high, a barrier meeting TL5 or TL6 may be considered. If available, local information on truck encroachment
frequency should be considered. However, a decision to design for heavy vehicles should be largely based
on consideration of the existence of particular hazards where the consequences of a heavy vehicle running
off the road would be catastrophic, for example:
• a heavy vehicle falling from a bridge or embankment onto a passenger railway line or onto a major road,
causing multiple fatalities in many vehicles plus significant societal disruption
• a heavy vehicle impacting a water or steam pipeline, electrical transmission tower or the like where major
disruption to supply, or other consequential damage, such as fire or contamination could occur
• a heavy vehicle impacting the supports of a structure, such as a bridge or pedestrian overpass, causing
the structure to collapse onto the road. This would be catastrophic if there were numbers of people on the
structure or if the collapse impacted vehicles or pedestrians on the road and/or caused long-term
transport disruption.
• a high-occupancy vehicle, such as a bus, falling into deep water from a bridge or embankment or over a
drop of sufficient height, endangering or killing many of the passengers in the vehicle
• a heavy vehicle leaving the road at a curve and impacting a community, commercial or residential
building (e.g. a school or playground located beside the road). This would be catastrophic if there were
significant numbers of people in or around the building.
• a heavy vehicle unable to stop at the base of a long downgrade colliding with other vehicles or buildings
with significant numbers of people in or around the building.
Designers should consider the likelihood that a heavier barrier, although not required initially, may be
required at particular sites during the life of a project. It may be the case that the initial stage of a project
requires barriers that will contain a passenger car, but future widening of the carriageway will require a
heavier barrier to shield the bridge piers from heavy vehicles. This may also apply where it is known that land
use will change adjacent to the road in future and the level of risk may then be such that a high-containment
barrier will be required. In such cases designers should ensure that the ultimate barrier requirements can be
accommodated.
In some situations, there is a need for two levels of protection, the first to protect the errant vehicle
occupants, the second to protect a vital piece of infrastructure from impact by a heavy vehicle or to address
the risk of high-severity crashes involving buses. Two levels of protection can be achieved by placing a
flexible barrier in front of a rigid barrier as discussed in Section 5.3.14.

5.3.14 Barrier Type (Step B6)
Key barrier requirements
Barriers used in Australia and New Zealand should comply with the guidelines or product acceptance criteria
described in AS/NZS 3845.1:2015 and administered by the Austroads Safety Barrier Assessment Panel
(ASBAP) or state/territory road agencies. In addition, only barrier products accepted by the road agency
should be used.
The main parameters to consider when selecting a barrier are:

• the speed environment of the barrier site
• the containment level required
• that it has a dynamic deflection that will fit within the working width
• the terminals are suitable for the site.
There are other factors that may be considered in selecting an appropriate type of barrier, either from a
general asset management or environmental perspective, or in relation to a particular site (refer to
Section 5.2).
Barrier design criteria
Barrier design criteria must be obtained from information published or provided by the road agency. Design
values that need to be matched against the site constraints include:
• design speed
• tested containment
• maximum deflection
• system width
• point-of-redirection
• minimum length between terminals
• allowable terminals
• anchor requirements
• allowable use on medians
• minimum median width
• minimum offset to traffic lane
• embankment slope limit
• ability to contain multiple impacts
• foundation conditions
• vulnerable road user limitations
• emergency vehicle access
• allowable use in pedestrian areas
• allowable use in gore areas.

Designers should be aware that the deflection values published for different types of barriers may be the
values of test deflections recorded under controlled conditions. The vehicle mass, speed and angle of
impact, along with the barrier length and condition are factors that are controlled as part of the testing.
Designers should be aware that the deflection figure published as a test result may not be the deflection
value achieved in the field for all impacts by errant vehicles.
Types of barrier
Barrier systems can generally be divided into three broad types comprising flexible, semi-rigid and rigid
barriers. As a general principle, if it is practicable to meet the requirements of the guidelines described
below, the more flexible barrier should always be used as this minimises the severity of any vehicle impacts
with the barrier. However, flexible barriers have relatively large deflections which may render them unsuitable
in some situations. In some cases, a barrier that will contain large heavy vehicles may be required and this
will be reflected in the containment level specified.
In determining the likely impact angles and the suitability of the barrier types, the designer should consider
the angle of impact as determined in Figure 5.8. In assessing the angle of impact, the designer should only
consider the lane closest to the barrier. Figure 5.9 shows angles for a range of speeds and offsets. Where
impact angles (including correction for horizontal curvature, shown in Figure 5.10, exceed the angles used in
crash testing, the designer should review the barrier type to confirm that it is still the best alternative.
Figure 5.8: Encroachment angle
Source: AASHTO (1996).

Figure 5.9: Maximum angle of impact for cars on straights
Source: Main Roads Western Australia (2017).
Figure 5.10: Angle correction factor
In the cases of flexible or semi-rigid systems, where the angle of impact is greater than in the crash tests, it
may be necessary to increase the allowance made for deflection. Placing of a barrier at the shy-line on a
tangential section of road results in impact angles less than or equal to the crash-tested angles.
The use of concrete barrier is accepted, provided that, from the lane closest to the barrier, the angle of
impact is no greater than 20°.
Special barrier designs and modifications to existing designs have been developed for use where:
• there is a need to cater for vulnerable road users (i.e. motorcyclists and cyclists)
• the aesthetic appearance of the roadside is important.

In general, flexible barriers comprise tensioned wire ropes, semi-rigid barriers have horizontal steel beams
and rigid barriers are constructed with concrete. Examples are shown in Figure 5.11, Figure 5.12 and
Figure 5.13, respectively while the profiles of semi-rigid and rigid barriers are illustrated in Figure 5.14 and
Figure 5.15. For details of barrier profiles and materials (e.g. concrete strength and composition) designers
should refer to jurisdictional or manufacturers’ standard drawings and specifications.
The type of rigid barrier profile used varies between jurisdictions. The single slope and vertical wall barriers
have an advantage in that the road can be resurfaced without affecting the profile or requiring expensive
resetting of the base level of the barrier. This a key reason for their use in some jurisdictions. There are two
types of profile that have the same height but different slopes on the wall.
To function correctly, rigid barriers require some lateral restraint which may require a foundation (refer to
Table 5.6 regarding support width). Where drainage is required at the barrier location (e.g. in a narrow
median) it is important that the drainage design is coordinated with the barrier design (e.g. location and
depth of pipes, pits).
Figure 5.11: Examples of flexible (i.e. wire rope) barrier
Figure 5.12: Semi-rigid barrier Figure 5.13: Rigid barrier

Figure 5.14: Examples of profiles of semi-rigid barriers
Steel post
W-beam Thrie-beam Modified Thrie-beam
Figure 5.15: Examples of profiles of rigid barriers
Wire rope safety barriers (WRSB)
WRSBs are the preferred barrier type because they have the lowest severity associated with an impact. The
critical consideration in the use of such systems is adequate room to allow for deflection of the barrier and a
suitable traversable area (slope of 10:1 (horizontal:vertical) or flatter) over this width of deflection.
For single-lane carriageways (i.e. one lane in each direction) where the full verge width cannot be provided
due to project constraints, a 1 m verge is acceptable provided that:
• The design vehicle is a light vehicle (refer to Section 5.3.13).
• The slope of this verge is 10:1 or flatter; this width should enable at least two wheels to remain on the
verge for the design light vehicle.
• The area behind the barrier to the full width of the dynamic deflection is clear of any fixed hazards and
slip-base roadside furniture (as these may not work due to the barrier restraining the vehicle) and
desirably free of frangible objects.
• Beyond the 1 m acceptable verge, embankment slopes are up to 4:1 desirable slope and 2.5:1 as the
maximum slope.
At anchor locations additional width may be required so as to provide adequate resistance for the anchor
block to work during impact. The width requirements for the anchor block need to be in accordance with the
manufacturer's specification.

It should be noted that WRSBs have some limitations with regard to horizontal and vertical curves and the
systems should be designed in accordance with the manufacturer's specification.
The amount of deflection expected should be based on the relevant test-level condition (TL1–TL4). The
deflections published may be based on a single test result achieved under controlled conditions. Deflections
of WRSBs in real-life situations may differ greatly, depending on factors such as barrier length and post
spacing.
Only tensioned wire rope systems that have been approved by the local jurisdiction should be used.
The order of magnitude of deflection of these systems is 2 m under TL3 conditions but is dependent on the
actual system used and the post spacing adopted. Expected deflections will depend on factors such as
vehicle size, speed, angle of impact and length of barrier. Designers should seek advice from WRSB
manufacturers regarding deflection.
All WRSB systems should be installed with concrete footings/sockets for post installations rather than driven
posts as the footing reduces the deflection of the system and also reduces whole-of-life costing (as during
maintenance activities any damaged post can be simply removed from the concrete footing and replaced).
For different soil types, the size of footing should be confirmed by the WRSB manufacturer. If the post is
driven, special plant is required to be mobilised to drive a new post and the time taken to drive the
replacement post is longer than simply removing the damaged post from the concrete footing and inserting a
new post.
Water-filled barriers
Water-filled barriers are only to be used as temporary systems on worksites.
Temporary steel barriers
Temporary steel barrier systems are only to be used as work zone barriers on worksites.
Semi-rigid barrier systems W-beam (public domain)
For limitations on its use refer to the requirements of local jurisdictions.
Thrie-beam (public domain)
Thrie-beam barriers are acceptable for use in transitions between semi-rigid systems and rigid barriers or
longitudinal barriers. Refer to each jurisdiction’s accepted barrier list.
Rigid barrier systems
Where no deflection can be permitted, or where a higher level of shielding is required, a rigid concrete barrier
should be used. The preferred concrete barrier shapes are the constant slope and the AASHTO F shape.
Crash testing indicates that both concrete barrier types give a comparable level of performance.
Concrete barriers can be provided to either TL4 or TL5 conditions by changing the height of the barrier,
footing and reinforcement details.
There is also a vertical wall barrier, which is crashworthy but does result in additional vehicle damage. It is
preferable for vertical walls to have one of the concrete shapes on the lower part of the wall to reduce vehicle
damage.

When using concrete barriers careful attention needs to be paid to drainage. The use of drainage openings
in concrete barriers needs to be carefully considered by the designer to ensure that the barrier performance
is not compromised. The use of any openings in the barrier and any maintenance considerations must be
taken into account.
Barriers on structures
Careful attention needs to be given to the extent of barrier required on the approach and departure of the
bridge and to the roadside barrier system to ensure that the vertical drop associated with the bridge and
associated wingwalls as well as any roadside hazards (e.g. fill embankments or objects) are adequately
shielded.
The same level of shielding provided on the bridge should be provided on the approaches for a length
necessary to adequately shield the hazard where the consequences of the barrier being penetrated are
similar to that of the bridge barrier being penetrated. Only when the hazard is adequately shielded should the
level of the barrier be reduced.
An acceptable methodology to determine the length of barrier required is that presented in Appendix B of
AS 5100.1-2017.
At increments away from the bridge the designer should undertake the process outlined in Appendix B of
AS 5100.1-2017 to determine where or if the level of the barrier can be reduced.
Barrier height
Barrier height is critical in that a rail installed too low is likely to cause errant vehicles to pass over the top of
it (i.e. vault), whereas a rail that is too high is likely to cause errant vehicles to snag on posts or even pass
under the rail. The heights of the various barrier systems have been established through crash testing using
appropriate test vehicles and therefore all barriers should be installed at a height that complies with the
requirements of the road agency.
The height of a barrier is measured as follows:

• Steel rail barrier height is measured from the pavement level or ground surface to the top of the rail.
• Wire rope barrier height is measured from the surface immediately below the wire ropes. When a wire
rope barrier is erected above, or next to a pavement without kerbs, the height is measured from the
pavement. When a wire rope barrier is erected on a median or verge, behind a kerb, the distance behind
the kerb will determine whether the height of the barrier is measured from the pavement or from the
finished surface below the barrier (see Commentary C6.3 Vehicle Trajectory Over Kerbs)
• Concrete barriers should be mounted on the pavement and the height is measured from the pavement
level at the bottom of the barrier to the top of the barrier to ensure the angles on the face are at the
correct height in relation to the pavement. If retrofitting a concrete barrier on an existing kerb line the kerb
should be removed to ensure the heights of the angles in the barrier face are at the correct height above
the pavement.

Where a rigid barrier is erected on a superelevated roadway the options illustrated in Figure 5.16 should be
considered. Practice may vary between jurisdictions, but it is suggested that a rigid barrier should be
installed vertically where the superelevation (or crossfall) is 5% or less. Where the superelevation is steeper
than 5%, and the barrier is to be on the high side of the pavement, it should be installed perpendicular to the
pavement, unless otherwise specified by system design or the jurisdiction.
Figure 5.16: Preferred installation of rigid barrier on a superelevated roadway
Axis at right angles

Vertical axis to pavement
Pavement
superelevation
(a) Installed vertically (≤ 5%) (b) Installed at right angles (> 5%)
Where a concrete barrier is erected in a narrow median on independently graded duplicated carriageways
the arrangement shown in Figure 5.17 may be considered. This treatment requires a foundation designed to
provide for the difference in level.
Figure 5.17: Rigid barrier in narrow median with independently graded carriageways
Barriers on structures
Barriers on structures in Australia should be designed in accordance with AS 5100.1-2017. In New Zealand,
designers should refer to the Bridge Manual (NZ Transport Agency 2013). Refer to Step B15 for guidance on
the transition from a road barrier to the bridge structure barrier.

Double barrier system
Factors that should be considered in the determination of the required containment level for barriers are
discussed in Section 5.3.13, including the option of a two-stage barrier system. In most cases a single barrier
to contain the design vehicle will be chosen. However, where a two-stage barrier system is appropriate a
layout similar to that illustrated in Figure 5.18 may be used.
Figure 5.18: Two-stage shielding layout
Direction of travel (1)
Flexible safety barrier
Observe maximum flare rates Rigid safety barrier

Median
Hazard
Rigid safety barrier Observe maximum flare rates
Flexible safety barrier
Direction of travel (2)
The following issues must be considered when designing a two-stage shielding:

• Design requirements for working width, length of need and minimum length of barrier must be met. The
traffic face of the rigid barrier must be offset from the face of the bridge pier to accommodate working
width.
• The structural design of the bridge pier may require a gap between the rigid barrier and the bridge pier, or
the rigid barrier may be an integral part of the bridge pier, if the pier has been designed for high-mass,
high-speed impact loadings.
• The rigid barrier must be designed to ensure it does not overturn when impacted by errant large mass
vehicles.
• It is preferable that the rigid barrier be aligned parallel to the carriageway, rather than flared. This will
facilitate extension of the rigid barrier if additional lanes are added in the future.
• In some situations, the end of the rigid barrier may require a crash attenuator (refer to road agency
requirements).
• Wherever possible, the offset between the flexible barrier and the rigid barrier should be more than the
working width of the flexible barrier.
• A cross-slope of 10% or flatter should be provided between the flexible and the rigid barrier.
Barriers across culverts
Options that may be considered for placing barriers across culverts where posts cannot be used are:
• thrie-beam with a span up to 4 m with a 3.5 mm thick rail or nested rail
• bridge-style barrier
• strengthening the rail and omission of posts.
Where there is less than 1 m cover to culverts and/or underpasses then normal post and rail systems cannot
be used.

Options available to the designer are:

• nested W-beam layout, for examples refer to Main Roads Western Australia (2017)
• direct connection of the barrier system to the culvert (applicable only for box culverts).
In setting the test level of the barrier required, the designer needs to consider the test level of adjacent
barriers.
Posts may be omitted and a stronger rail used if the integrity of the barrier system can be verified. For
example, FHWA acceptance letter HMHS-B58 agrees with the use of a W-beam nested rail over a span of
7.62 m and designers may refer to the letter for the details of the treatment, particularly post spacing either
side of the span and the length of nesting (FHWA 1999).
It should be noted that the culvert behind the barrier should extend beyond the barrier to accommodate the
dynamic deflection of the options described.
5.3.15 Dynamic Deflection (Step B7)
When a vehicle strikes a barrier, the dynamic deflection varies according to the characteristics of the
impacting vehicle, including vehicle mass, impact speed, angle of impact and the characteristics of the
barrier system. Sufficient dynamic clearance should be provided between the face of a barrier and a hazard
to accommodate the appropriate dynamic deflection. For design purposes, dynamic deflections should be
determined from information available from jurisdictional publications and other advice (e.g. product-specific
specifications that may be based on tests and manufacturers’ advice).
Designers may refer to dynamic deflection information resulting from tests on a particular barrier. However,
these tests are often conducted on a relatively short length of barrier and the effect of longer lengths and
other factors on deflection should be taken into account wherever possible. Designers may therefore refer to
research on particular types of barriers. For example, in the case of wire rope barriers research is available
that provides correction factors to take account of aspects such as barrier length, horizontal curvature and
temperature (Alberson et al. 2003).
It should be noted that some semi-rigid systems can be strengthened locally by adding additional posts or by
reinforcing the rail element (i.e. using a double beam or nested rails) to shield individual fixed hazards that
are within the deflection distance for a single-beam barrier. In addition, the deflection of wire rope road safety
barrier can also be reduced by adopting closer post spacing. However, it should be noted that the practice of
using closer post spacing to reduce deflections is based on a limited number of tests with light European test
vehicles. It is not clear how far in advance of the hazard that the reduced post spacing is required for the
smaller deflection to be achieved and if the same magnitude of reductions would be achieved with heavier
test vehicles.
5.3.16 Vehicle Roll Allowance and System Width (Step B8)
All barrier systems deflect under impact by vehicles except for rigid concrete barriers. The behaviour of
barriers under impact varies depending on the type of system. The posts in semi-rigid systems act to reduce
deflection but may induce more vehicle roll whereas the weaker posts in flexible systems do not restrain the
cables under impact and vehicle roll is likely to be less.
Vehicle roll allowance
The vehicle roll allowance concept is an indicative line of the roll experienced by a vehicle impacting a
barrier.
The clearance required between the face of a barrier and a hazard includes the dynamic deflection of the
barrier system and a vehicle roll allowance and is described with the term ‘working width’ as shown in
Figure 5.19.

The vehicle roll allowance values shown in Table 5.7 are for a height of 4.3 m above the pavement. These
values are based on the vehicle dynamics of a 4.3 m high van-type rigid vehicle. The vehicle roll allowance
values may be interpolated where the hazard is less than 4.3 m high but caution should be applied in their
use. The capacity of the barrier to accommodate the roll allowance should also be checked.
Zone of intrusion
The Roadside Design Guide (AASHTO 2011a) also gives some indication of how much a heavy vehicle
could roll over a barrier on impact and defines that region above and behind the face of the barrier as the
zone of intrusion. There is an indication that, for TL4 barriers and impacts, which would involve an 8 t single
unit truck impacting at 80 km/h and angle of impact of 15°, the truck cabin could roll laterally about 860 mm
at a height of 2.4 m, and the cargo box could roll 2 m laterally at a height of 3 m.
If the structure of the bridge is such that it cannot accommodate impact from a heavy vehicle, consideration
should be given to installing a taller barrier, such as a 1070 mm or 1370 mm barrier. It is suggested that this
tall wall extend 3 m in advance of the pier.

Figure 5.19: Vehicle roll allowance
Object
requiring
protection
from impact
e line
ro ll s lo p
V e hic le 4.3m
Initial position of barrier
r
sition of ba rrie
Limit of dynamic
deflection
Safety
De fle cted po
barrier
System width Allowance for

system width
Dynamic deflection and/or vehicle roll
Dynamic deflection
(may be 0 mwith
be 0m withrigid barriers
rigid )
barrier)
Working width

Table 5.7: Vehicle roll allowance
Vehicle roll allowance at 4.3 m height above pavement (m)
Crossfall
Design
(%)
speed
(km/h)
–7 –6 –5 –4 –3 –2 –1 Flat +1 +2 +3 +4 +5 +6 +7
40 0.70 0.65 0.60 0.55 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50
50 0.75 0.70 0.65 0.60 0.55 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50
60 0.80 0.75 0.70 0.65 0.60 0.55 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50
70 0.85 0.80 0.75 0.70 0.65 0.60 0.55 0.55 0.55 0.50 0.50 0.50 0.50 0.50 0.50
80 0.90 0.85 0.80 0.75 0.70 0.65 0.60 0.60 0.60 0.55 0.50 0.50 0.50 0.50 0.50
90 1.00 0.95 0.90 0.85 0.80 0.75 0.70 0.70 0.70 0.65 0.60 0.55 0.50 0.50 0.50
100 1.10 1.05 1.00 0.95 0.90 0.85 0.80 0.80 0.80 0.75 0.70 0.65 0.60 0.55 0.50
110 1.20 1.15 1.10 1.05 1.00 0.95 0.90 0.90 0.90 0.85 0.80 0.75 0.70 0.65 0.60
Notes:
 Interpolate vehicle roll slope line for objects that are less than 4.3 m high.
 The vehicle roll allowance will depend on the height and type of barrier. The values in this table apply to a 710 mm
high (approximately) W-beam barrier and rigid barriers.
System width
The system width is the overall width of the barrier at the top as shown in Figure 5.2. If the system width is
greater than the vehicle roll allowance, the system width will be used in the calculation of working width
(Section 5.3.17).
5.3.17 Working Width (Steps B9 and 10)
Working width is defined in Appendix A.2 and illustrated in Figure 5.19. The working width is simply the sum
of the dynamic deflection and vehicle roll allowance (or system width if it is larger than the vehicle roll
allowance).
The working width is checked (Step B10) against the barrier-to-hazard clearance. If the working width is
greater than the barrier-to-hazard clearance available, the barrier may not prevent an impact with the hazard.
In this case designers should consider:
• changing the barrier to a more rigid type
• changing the lateral position of the barrier
• using a double barrier as referred to in Section 5.3.14 under Double barrier system.

5.3.18 Barrier Length of Need (Step B11)
General
The length of need is the length of barrier required to redirect an errant vehicle and shield the hazard (i.e. the
distance between the leading and trailing points of need). Points B and D in Figure 5.20 and Figure 5.21
represent the leading and trailing points of need respectively, while points A and C represent the extremities
of the hazard relating to points B and D. Figure 5.20 illustrates the concept when a barrier that could be
impacted from a single direction of travel and Figure 5.21 when a barrier that could be impacted from two
directions of travel. The method to establish these points is further explained in the text relating to the run-out
length method of establishing the length of need. There needs to be an effective barrier in the length of need
(between points B and D).
The appropriate length of a barrier system is related to its points of redirection documented in Section 5.3.19.
The points of need for a hazard should be within the barrier system’s points of redirection. The minimum
length of a barrier is when the points of need for a hazard coincide with the points of redirection of a hazard.
Figure 5.20: Length of need – single direction
Source: Roads and Maritime Services (2016).
Figure 5.21: Length of need – two directions
Where the test level of a barrier is determined by the level of performance required for a specific hazard, it is
implicit that the same level of performance should be provided over the whole length of need associated with
that hazard. For example, where a hazard requires a TL5 barrier, a barrier successfully tested to TL5 must
be provided over the full length of need associated with that hazard.

There are two geometric methods used to determine the likely trajectory of a vehicle that leaves the road in
the vicinity of a roadside hazard and the length of need required, a method based on run-out length and a
method based on angle of departure. The methods may result in different lengths. Designers should consult
the road agency to establish local practice.
The run-out length method is described below and the angle of departure method is described in
Appendix G.
For a non-gating terminal point of redirection, refer to manufacturers’ requirements.
See Appendix H for worked examples of the run-out length method.
Run-out length method
This is the method favoured by most Australian and New Zealand road agencies and AASHTO (2011a). The
run-out length (LR) is shown in Table 5.8 and is the length of clear run-out area that should be made
available as a passageway for deceleration between the start of the barrier and a hazard. It is the theoretical
distance needed for a vehicle that has left the roadway to come to a stop and is therefore dependent on
vehicle speed. It is measured from the upstream extent of the obstruction along the roadway to the point at
which a vehicle is assumed to leave the roadway, although the actual distance travelled is along the vehicle
departure path. Examples of calculations are provided in Appendix H.
Table 5.8: Run-out lengths for barrier design
Run-out length LR (m) for AADT range

Design speed
(km/h)
> 10 000 5 000–10 000 1 000–5 000 < 1 000
110 110 101 88 76
100 91 76 64 61
90 80 67 56 54
80 70 58 49 46
70 60 49 42 38
60 49 40 34 30
50 34 27 24 21
Note: The values in Table 5.8 are lower, based on research and observations, than in previous editions of this Part.
Straight roads
The application of the run-out length method to establish barrier ‘length of need’ for both traffic approaching
in the left lane, and for opposing traffic, is illustrated in Figure 5.22. On a two-lane two-way road, and for
medians, these requirements are combined to develop a design layout that shields hazards from both
directions. The layout of barriers on straight or nearly straight roads is established by applying the following
formulae (Equations 1, 2 and 3):
For installations where the barrier is flared (refer to Section 5.3.5):

b
�LA + � � (L1 ) -L2 � 1
a
X=
b L
�� + � A ��
a LR

For parallel installations that have no flare:
[LA - L2 ] 2
X=
L
� A�
LR
The lateral offset, Y, from the edge of the running lane to the beginning of the length of need may be
calculated from:
LA 3
Y = LA - (X)
LR
where
X = the required length of need in advance of the area of concern (hazard)
LR = run-out length (Table 5.8)
b/a = flare rate (Table 5.5)
LA = lateral extent of the area of concern
Y = lateral distance from the edge of the traffic lane to point of need
These parameters are illustrated in Figure 5.22. The barrier length is a function of the distance that it is
located from the edge of the driving lane and can most readily be obtained geometrically by drawing the
length of need chord from the edge of the running lane at distance LR from the hazard to the rearmost point
of the hazard. The barrier should cross this chord as shown in Figure 5.22(a) and (b).

Figure 5.22: Run-out length method of determining length of need
Notes:
 LR is the run-out length for the barrier.
 LA is the lateral extent of the hazard (edge of traffic lane to rear of hazard).
 L1 is the tangent length of the barrier upstream from the area of concern.
 L2 is the barrier's lateral distance from the edge of the traffic lane.
 L3 denotes the distance from the edge of the traffic lane to the nearest point on the hazard.
 LS is the shy-line offset.
It should be noted that:

• The distance between the edge of the traffic lane and barrier affects the length of need; placing a flexible
or semi-rigid barrier further from the road can result in a shorter barrier and lower installation and
maintenance costs associated with shielding hazards. However, designers should refer to Section 5.3.4
for discussion on lateral location issues.
• The influence of roadside batter slopes on the design may be considered by completing the layout
procedure on a scale plan, highlighting the hazard and showing the contour lines.
• LA is the distance from the edge of the running lane to the far side of the fixed object, to the road-property
boundary line, or to a point beyond the road-property boundary line to shield a hazardous fixed object.
Depending on site characteristics the designer may choose to shield only that portion of a hazard that lies
within the road reserve only.

• L1 is chosen by the designer. For the situation where a semi-rigid railing is connected to a rigid barrier, it
is suggested (AASHTO 2011a) that the tangent length should be at least as long as the transition section.
This measure reduces the possibility of pocketing at the transition and increases the likelihood of smooth
redirection if the barrier is struck immediately adjacent to the rigid barrier.
The result of these calculations is the required length of need of an approach barrier for traffic in the lane
immediately next to the barrier. For opposing traffic, an approach longitudinal barrier length of need is
calculated in the same manner. In this case, all lateral dimensions are measured from the edge of the
opposing traffic lane that is nearest to the hazard (Figure 5.22(b)).
Curved roads
The length of need formula is applicable only to straight roads. For barrier designs on the outside of
horizontal curves, it is assumed that a vehicle’s exit path from the road will follow a tangential run-out path if
the area outside the roadway is flat and traversable. Therefore, rather than using the theoretical LR distance
to determine the barrier length of need, a line from the outside edge of the hazard (or the road-property
boundary for a continuous non-traversable feature) to a tangent point on the curve should be used to
determine the appropriate length of need for the barrier (Figure 5.23).
The barrier length then becomes a function of the distance it is located from the edge of the driving lane and
can most readily be obtained graphically by scaling (AASHTO 2011a). Depending on the radius of the curve,
a flare may not be required on the barrier but a properly designed and installed, crashworthy end treatment
will be required.
Figure 5.23: Length of need on outside of curve using run-out length method
LA
L2
Note: In the case depicted, both the culvert headwall and creek are hazards.
Source: Adapted from Queensland Department of Transport and Main Roads (2013).
For barrier designs on the inside of curves the length of need is based on the length of run-out (LR) projected
from the edge of the traffic lane to the rear of the hazard (Figure 5.24). This is based on the premise that a
vehicle leaving the road in advance of or at the departure point will be able to stop before reaching the
hazard or pass to the rear of it. The various possible vehicle trajectories beyond this departure point will be
shielded from the hazard.

Figure 5.24: Length of need on inside of curve using run-out length method
Note: In the case depicted, both the culvert headwall and creek are hazards.
Source: Adapted from Queensland Department of Transport and Main Roads (2013).
5.3.19 Minimum Length of Barrier (Step B12)
In order to perform satisfactorily, barrier systems must have sufficient length to enable the strength to be
developed through the system and into the posts as impact occurs.
MASH requires the minimum tested barrier length to be at least three times the length in which deformation
is predicted, but not less than 30 m for steel beam systems and 180 m for WRSBs.
While shorter lengths than the tested length are possible, the designer must consider how this will affect
other performance values (e.g. deflection).
The lengths to be considered in the design of barriers are the:

• terminal length
• transition length
• minimum length
• development length.
These lengths are shown conceptually for anchored and unanchored barrier systems in Figure 5.25 and
Figure 5.26, respectively. The development length applies to unanchored barriers and is the length in
advance of the point of redirection that is necessary to provide sufficient mass for the barrier within the length
of need to perform in accordance with its design parameters.

Figure 5.25: Road safety barrier lengths – anchored systems
Notes:
 Point of redirection is defined in Table A.2 of Appendix A.
 Point of redirection varies for different proprietary products, refer to manufacturers’ specifications or design details.

Figure 5.26: Road safety barrier lengths – unanchored systems
Notes:
 Point of redirection (PoR) is defined in Table A.2 of Appendix A.
 For some barriers, designers may consider whether the development length should contribute to the minimum length
of the barrier.
5.3.20 Sight Distance (Step B13)
Barriers on horizontal curves can impede stopping sight distance. Barriers located close to intersections can
also impede the safe intersection sight distance and minimum gap sight distance available to drivers
attempting to select a safe gap in traffic on the major road. This issue applies to barriers located on the verge
and barriers located in medians.
For sight distance requirements in mid-block situations including horizontal curves, designers are referred to
Section 5 of AGRD Part 3 (Austroads 2016a).
For sight distance requirements associated with intersections and interchanges designers are referred to the:
• Guide to Road Design Part 4A: Unsignalised and Signalised Intersections (Austroads 2017b)
• Guide to Road Design Part 4B: Roundabouts (Austroads 2015e)
• Guide to Road Design Part 4C: Interchanges (Austroads 2015c).

5.3.21 Terminal Treatments (Step B14)
General
Once the barrier has been located longitudinally (points of redirection) and laterally to accommodate dynamic
deflection and sight distance, suitable leading and trailing terminal treatments must be selected for use.
Crashworthy terminals are used on safety barriers located in areas where they may be hit by an errant
vehicle.
Terminal treatments and crash cushions or impact attenuators are used to terminate a barrier. These devices
are designed to ensure that the ends of barriers provide safe conditions for occupants of vehicles that may
impact this area of a barrier. They must be used on all rigid and semi-rigid barrier systems. Flexible barriers
have an end anchorage and when impacted head-on the posts and cables collapse as the vehicle
decelerates.
Key performance aspects (AS/NZS 3845.1:2015) are that barrier terminals should:
• where necessary, incorporate an anchor to the barrier system to enable the full tensile strength of the
system to be developed during impacts with the barrier at locations away from the terminal
• not cause an impacting vehicle to roll, vault or yaw in an inappropriate manner (applies to leading and
trailing terminals)
• not spear the impacting vehicle or cause undue problems with debris
• perform acceptably when impacted from either direction, except when erected on a single direction
carriageway where there is a low probability of a vehicle impacting the terminal from the reverse direction
• not distinguish between a temporary or permanent situation. The only exception may be an allowance for
the reduction of impact forces where the site is effectively managed (i.e. if the maximum traffic speed is
controlled, then a lower performance end terminal treatment may be used than would otherwise be
required).
• be tested in accordance with AS/NZS 3845.1:2015 and as the tests are undertaken in a controlled
environment, actual site conditions need to be considered when selecting an end terminal treatment.
Where a barrier is located some distance from the edge of the road it may be possible to flare the barrier and
terminate further from the road. In such cases, because a significant percentage of errant vehicles may
reach the terminal, it is preferable that a crashworthy end treatment is provided. A non-crashworthy end
treatment should only be considered where a detailed assessment concludes that the likelihood of an end-on
impact with the barrier is very low (i.e. negligible).
Where kerbing is present at the site and the kerbing type is to be modified for the barrier then the
modification needs to be extended upstream of the terminal to a point determined where an errant vehicle
would cross the kerb line at 1 in 20 measured from the start of the terminal. A crash involving a vehicle
impacting an untreated or inappropriately treated end of a barrier can have serious consequences for the
occupants because the:
• vehicle is stopped abruptly
• barrier may penetrate into the occupant space of the vehicle
• vehicle may be launched and roll over.
It is therefore imperative that terminal treatments are appropriate for the type of barrier and installed in
accordance with the manufacturer’s specifications and road agency guidelines. The type of terminal
treatment used depends on the type of barrier and its performance level. Some treatments function only to
provide a safe terminal for the barrier, while others also function as an anchor for the system.

A barrier terminal treatment may fulfil its function by:

• permitting controlled penetration by the vehicle into an area behind the device
• decelerating a vehicle to a safe stop within a relatively short distance
• containing and redirecting the vehicle
• a combination of the above.
Road safety barrier terminals are generally classified as either a gating/non-gating terminal or as a crash
cushion/impact attenuator (refer to paragraph Gating and non-gating terminals).
Selection factors for terminal treatments
The selection of the most appropriate crashworthy terminal treatment for a barrier should take into account
the:
• need for gating or non-gating characteristics
• need for redirective or non-redirective characteristics
• speed environment
• space available for installation and deformation of the terminal
• need for a run-out area behind the barrier
• width required for accommodation and deformation of the terminal
• capacity to absorb nuisance crashes
• compatibility with barrier type
• cost and maintenance factors.
The gating and non-gating characteristics are discussed in Gating and non-gating terminals and the
redirective and directive characteristics of barriers are discussed in Crash cushions and impact attenuators
(the sections follow Figure 5.28) whereas the other selection factors are summarised in Table 5.9.
For crash cushions or impact attenuators the following aspects should also be considered:
• Type – redirective or non-redirective.
• Classification – if a redirective cushion is required then it needs to be specified if it is to be gating or
non-gating.
• Performance level – some cushions have achieved multiple test levels and for some products the system
owner can provide configurations for different design speeds rather than various test levels.
• Configuration – crash cushions may be available in different configurations including width, anchoring in
terms of rigid backstops, different colours of nose cones and only certain configurations may be
acceptable to the road agency. Side panel/rail laps may vary depending on whether the adjacent passing
traffic is one-way or two-way. For some systems, the side panels are aligned to accommodate the
direction of travel.
• Transitions – there may be a number of options available depending on the direction of impact
(unidirectional or bidirectional) and the hazard or barrier system to which the cushion is to be connected.
• Foundation options – some cushions have a range of foundation options and only certain options may be
acceptable to the road agency or applicable to the site in which the cushion is to be installed.
• Site conditions – there are design limitations (e.g. maximum crossfall) which may limit the use of certain
devices.

A number of terminal treatments that have been used in the past are no longer suitable because they enable
the barrier to penetrate (or spear) the cabin space of light vehicles and/or cause vehicles to vault or roll.
These terminal treatments include:
• splayed ends (fishtail ends) on W-beam barrier
• sloped (turned down into the ground) ends on semi-rigid or rigid barrier (although a sloped concrete
terminal treatment may be suitable where speeds are low (e.g. 60 km/h or less) and space is limited by
right-of-way constraints or the presence of other features preclude the use of one of the tested terminal
treatments)
• a narrow double bull-nose terminal treatment on back-to-back W-beam
• breakaway cable terminals (slotted and non-slotted).
Table 5.9: Selection factors for terminal treatments
Factor Considerations
Gating and non- Refer to discussion under Gating and non-gating terminals.
gating characteristics
Redirective and Refer to discussion under Crash cushions and impact attenuators.
non-redirective
characteristics
Speed environment The terminal should be suitable for the speed environment at the location; terminals may
have been tested for different speeds.
Particular terminals and different configurations of the same crash attenuator will be
suitable for particular speed environments.
The operating speed should usually be taken to represent the speed environment. Where
the operating speed is not known the speed limit +10 km/h may be used provided that the
speed zone exists some distance upstream of the installation.
The length of some crash attenuators can be varied depending on the speed environment
and likely maximum impact speed. Manufacturers’ advice should be sought.
Space availability The installation should comply with all of the manufacturers’ recommendations with respect
to space.
The space available for the attenuator or end treatment will also influence the type of
terminal to be installed, for example:
• Some crash attenuators are more suited for use in narrow medians while others are
suitable to shield wider hazards.
• Some crash attenuators and end treatments may require a large run-out area free of
hazards for gating of the end, while others may require space to accommodate
displacement of the attenuator clear of traffic.
Consideration may be given to selecting a physically smaller system on the basis that a
smaller size will reduce the number of crashes, especially nuisance crashes, thereby
reducing the maintenance that must be undertaken following an incident.
Space requirements of terminals should be allowed for during all stages of road design and
construction (e.g. preliminary design for new works or rehabilitation of existing roads) to
ensure compatibility between the final design and the terminal that is to be installed.
Figure 5.27 is a guide, for planning purposes only, to the area that should be made
available for crash attenuator installation.
Susceptibility to Terminals are susceptible to nuisance crashes.
nuisance crashes The system should be capable of performing satisfactorily following a number of minor
crashes without requiring repair. It may be the case that a non-gating system would perform
better than a gating system in this respect.
Compatibility with The terminal should be suitable for use with the proposed barrier type. Manufacturers’
road safety barrier specifications should therefore be consulted.
type In some instances, a transition section will be required to ensure adequate stiffness is
provided at the connection between the terminal and the barrier. This is required to
minimise vehicle snagging and pocketing of the barrier, and to limit the change in deflection
occurring between the barrier and the end treatment.

Factor Considerations
Cost and The whole-of-life cost should be taken into account when selecting a terminal including:
maintenance • capital costs
• risks associated with maintenance repair times.
Site preparation costs to accommodate some systems can also be significant.
Crash attenuators are relatively costly to install and to repair after impact, so they are
generally used only where it is likely that errant vehicles will hit a hazard with severe
consequences, and either:
• It would be very difficult or costly to remove or relocate the hazard, make it frangible, or
realign the traffic path away from the hazard.
• There is insufficient room for a normal barrier and its terminals, or normal barrier ends
would form unacceptable hazards (e.g. in some narrow medians).
At locations where frequent hits are expected, life-cycle costs for repairing or replacing an
attenuator system may be a significant factor in the selection process. The repair and
replacement time for an attenuator system following an impact is also an important
consideration as this can cause significant losses to road users through delays. The direct
costs associated with worker safety and traffic management also need to be considered.
Figure 5.27: Space required for crash cushions in gore areas
For flare rates

see Note 1
F
N
F
L
Design speed Dimensions for crash cushions, reserve area (metres)

on main Minimum Desirable
carriageway
(km/h) Restricted conditions Unrestricted conditions
N L F N L F N L F
50 2 2.5 0.5 2.5 3.5 1 3.5 5 1.5
80 2 5 0.5 2.5 7.5 1 3.5 10 1.5
110 2 8.5 0.5 2.5 13.5 1 3.5 17 1.5
130 2 11 0.5 2.5 17 1 3.5 21 1.5
Notes:
1 For flare rates refer to Table 5.5.
2 The information provided in this table is generic and should therefore be used only for planning purposes. Detailed
product information should be used for design purposes.
3 Although the figure depicts a gore location, the same recommendations will generally apply to other types of fixed
objects that require shielding.
4 The unrestricted conditions represent the minimum dimensions for all locations except for those sites where it can be
demonstrated that the increased costs for obtaining these dimensions (as opposed to those for restricted conditions)
will be unreasonable. The desirable dimensions should be considered optimum.

Gating and non-gating terminals
Gating terminals and non-gating terminals can be either a public domain product or a proprietary product
(e.g. extruding head terminal).
Gating terminal systems are designed to allow a vehicle impacting the nose, or the side of the terminal at an
angle near the nose, to pass through the terminal and behind the barrier. They may breakaway, hinge or
pivot when impacted. Gating terminals are therefore not suitable for use where there is a high potential that
an errant vehicle may travel through the treatment and into a hazard or into opposing traffic lanes (e.g. in
narrow medians). A gating terminal is considered to have functioned properly if the vehicle remains stable
during and after impact and is kept away from the hard point of the road safety barrier system.
Gating treatments for semi-rigid barriers cause the beam to deform or absorb the kinetic energy of the
vehicle. Crash cushions and impact attenuators (see under the following section Crash cushions and impact
attenuators) that are designed to be used as a terminal for concrete barriers or to shield other fixed objects,
may also be designed to ‘gate’. For gating end treatments, the length of need usually starts about one panel
of rail from the impact head of the unit, but this can vary depending on the terminal used.
Vehicles that pass through a gating treatment are directed into the area behind the end treatment (i.e. on the
side of the barrier opposite the travelled lane). Figure 5.28 illustrates both a flared gating terminal and a
parallel gating terminal and the required area behind the barrier. It is necessary to ensure that this run-out
area should:
• contain no fixed hazards (e.g. poles and trees)
• be traversable, with a lateral slope of 6:1 or flatter
• extend parallel to the barrier/terminal at least for a distance of 18 m beyond the point of redirection for the
barrier/terminal.
• be at least 6 m wide.
AS/NZS 3845.1:2015 refers to a requirement for an 18 m long run out area (hazard-free area) measured
from the point of redirection of any gating treatment) which is based on crash tests on older public domain
terminals.
As end treatments are designed and tested on flat and level terrain with a vehicle impacting at normal height,
it is imperative that these conditions be replicated in practice. Failure to do so may result in the device failing
to perform as intended. Terminals must therefore be placed on a relatively flat surface (10:1 maximum slope)
and the path between the road and the attenuator must be clear of any irregularities or obstructions, such as
excessive slopes or kerbs. These features can cause a vehicle to become airborne and ride over the barrier
or roll over on impact. Maximum crossfalls are recommended for various types of proprietary terminal
systems.
The 18 m x 6 m dimensions are minimum figures and may not be sufficient for all collisions. If a run-out area
cannot be provided or would be smaller than these dimensions, a non-gating terminal should be used.
For flexible and semi-rigid barrier types, terminal treatments must be properly anchored so that the design
operational requirements are achieved in practice. Any redirective capability required by the design will only
be achieved by the end treatment developing the same full tensile strength as the barrier upon impact.
Non-gating terminals do not allow vehicles to pass through the leading section of the terminal as they are
designed to contain an impacting vehicle and redirect it along the length of the terminal towards the barrier.
Most non-gating terminal treatments are crash attenuators that do not allow a colliding vehicle to pass behind
the terminal. On colliding with the end of the terminal, the vehicle will be redirected away from the barrier or
be arrested by the barrier. The point of redirection for a non-gating system is at the nose. A barrier with a
non-gating terminal does not require a run-out area.

Figure 5.28: Run-out area for gating terminals
Notes:
 Designers should note that the dimensions shown are a minimum requirement and there are benefits in providing a
longer and wider run-out area.
 In constrained circumstances, it may not be possible to provide the run-out area. In these situations, designers should
assess the risk involved with the use of a gating end treatment and no suitable run-out area versus other options such
as the use of a non-gating end treatment.
 This figure indicates grading requirements for a parabolic flared end treatment and for a parallel end treatment.
Wire rope safety barrier (WRSB) terminal treatments should be provided in accordance with the
manufacturer's specification. The end anchors are frangible and designed to ensure that when impacted the
wire ropes are restrained and not a hazard to adjacent traffic. An errant vehicle running into the end of a
WRSB straddles the cables and may be arrested by them as the vehicle progressively flattens posts and
comes to rest.
Crash cushions and impact attenuators
In some situations, a crash cushion will be the most appropriate device. Crash cushions and impact
attenuators are protective devices that prevent errant vehicles from impacting fixed hazards. This is achieved
by absorbing energy at a controlled rate to decelerate a vehicle in a short distance to a safe stop before
impact with the hazard. Some crash cushions redirect the vehicle away from the hazard when impacted at an
angle.
Crash cushions are suited to protect larger hazards which cannot be removed, relocated or shielded by a
conventional safety barrier. They are proprietary products and therefore the dimensions and installation
requirements should be sourced from information published by the manufacturer and the road agency should
be consulted regarding the use of the device.

The principles on which crash cushions and impact attenuators operate are described in Commentary 7.
[see Commentary 7]
Crash cushions and impact attenuators can be classified as:

• gating or non-gating, depending on their behaviour when impacted on the side near the leading end of the
barrier
• redirective or non-redirective depending on their ability to redirect impacting traffic away from the hazard.
Figure 5.29 illustrates the behaviour of gating and non-gating systems while Figure 5.30 illustrates the
behaviour of redirective and non-redirective systems.
Figure 5.29: Gating and non-gating systems

Figure 5.30: Redirective and non-redirective systems
A gating/redirective crash attenuator can function well in both head-on and side angle impacts. The kinetic
energy of the vehicle is absorbed by crushable or plastically deformable materials, or by the use of hydraulic
energy absorbers placed in front of an obstacle (AASHTO 2011a). Most impact attenuators are based on this
concept and need a rigid support to resist the vehicle impact force as the energy-absorbing material is
deformed.
When impacted head-on, these impact attenuators have the energy-absorbing ability to slowly bring the
vehicle to a safe stop. Angle impacts in the leading section cause the device to ‘gate’ and, when subjected to
glancing or side angle impacts beyond the point of redirection, they redirect vehicles back into their originally
intended direction of travel.
A non-redirective crash attenuator performs most effectively when hit head-on. These attenuators comprise
barrels or containers, typically filled with variable masses of sand, and their performance is based on the
transfer of momentum of a moving vehicle to an expendable mass of material located in the vehicle’s path.
This is the only type of crash attenuator for which the design can be analytically determined. They are
designed in accordance with the principle of conservation of momentum whereby the kinetic energy of the
impacting vehicle is transferred to the mass of sand. These devices require no rigid back-up or support.
In a crash, a non-redirective crash attenuator stops a vehicle in head-on impacts, and in side angle impacts
is unable to redirect the vehicle back into its intended direction. This limitation results in continued forward
motion at a high speed, with the consequence that the vehicle penetrates the attenuator. The impacting
vehicle continues in the same direction until it either is arrested by the device or impacts an object (e.g. near
the rear end of the device, refer to Figure 5.30(b)).
Crash attenuators should be orientated so that they face the most likely direction of impact which is
particularly important where the approach is on a tight curve (a situation that may be encountered in
construction zones).

Types of terminals
Gating and non-gating terminals may be public domain or proprietary products. The accepted product list
from the jurisdiction should be consulted before use. The design of proprietary terminals may change over
time due to product development and therefore designers should remain familiar with the products that are
available (e.g. through relevant websites) and the specifications relating to the products and their use.
Appendix I also provides some generic information on proprietary products.
5.3.22 Transitions between Barriers (Step B15)
General
Wherever it is necessary to change from one type of barrier to another, or to physically join them together
(e.g. a bridge barrier to a road barrier), the interface must be designed to ensure that the overall system will
perform safely when impacted by a design vehicle.
Interfaces are designed to provide a smooth, snag-free transition between different types of barriers where
they meet, such as at bridge parapets. Inappropriate, incorrectly installed or missing interfaces present a
hazard to vehicles impacting the barriers at or near the interface point (i.e. the end of the concrete or steel
bridge barrier).
Different profiles of semi-rigid steel barrier and different profiles of rigid barrier can all be interfaced with a
properly designed continuous transition, whereas interfaces between flexible barriers and more rigid systems
can only be effected by overlapping the different systems. The overlap should be designed on the basis that
the terminating system will overlap in front of a system that is beginning, irrespective of the system type. The
barriers should be separated by a clearance at least equivalent to the dynamic deflection of the terminating
system.
For interface details designers should refer to manufacturers’ specifications, standard drawings and
guidelines on appropriate use of interfaces that are available from the road agency. Transitions are
described in detail in Appendix J.
Bridges
Bridge barrier systems, including the transitions between bridge barriers and road barriers, should be
designed in accordance with AS 5100.1-2017.
The design procedure in the standard is focussed on bridge railings. However, the procedure could also be
applied to sites on the approaches to bridges and at other (non-bridge) locations on roads where similar
conditions exist. The procedure comprises:
• a selection method that leads to a recommendation for a low (TL2), regular (TL4) or medium (TL5)
performance-level barrier
• descriptive advice on the assessment of individual medium-risk to high-risk sites, using risk assessment
and benefit-cost analysis to determine whether a higher performance-level barrier should be provided.
The selection method takes into account factors for road type, downgrade, curvature, deck height and
under-structure conditions, commercial vehicle percentage, and speed environment.
AS 5100.1-2017 also provides information on barriers for bicycle and pedestrian bridges and for some
design elements for bicycle/pedestrian paths as they relate to bridges.

Bridge approaches
Transitions from approach barriers to bridge barriers should conform to the following requirements
(AS 5100.1-2017):
• A transition barrier should be provided on the approach to all bridge safety barriers.
• The strength and stiffness of the approach barrier should vary to provide a transition from the flexible
barrier to the rigid or semi-rigid bridge barrier.
• A smooth face and tensile continuity should be maintained throughout, for example, exposed rail ends,
kerbs, posts and sharp changes in the geometry of the barrier components and kerbs, should be avoided
or transitioned out with a taper that reflects the change to flare rates in Table 5.5.
Additional information on transitions between bridge and road barriers is provided in Section 16.1 of the
Guide to Bridge Technology Part 3: Typical Superstructures, Substructures and Components
(Austroads 2018b).
5.3.23 Confirm that the Barrier Meets the Objectives (Step B16)
When the design has been completed it is prudent to revisit the original objectives of providing the barriers
(Section 5.3.3) and check that the design will perform the functions that led to the barrier being proposed. If
there is doubt that the design is the best outcome to achieve the objectives the designer should consider
whether an alternative type of barrier should be considered and return to Step B11 (Section 5.3.18) to
continue the design process.
5.4 General Access Through Road Safety Barriers
Preferred practice is to avoid providing breaks in a road safety barrier. However, it may be necessary to
consider breaks at locations where pedestrians cross the road, intersections, points of access to property
and access points in medians. Where breaks are necessary, barriers may be overlapped and where
necessary, safe end treatments must be provided. Authorised emergency or temporary access through
concrete barriers can be provided by the installation of a suitably crash tested system (e.g. fabricated steel
gate). An example of an overlapping-type access through a barrier is shown in Figure 5.31.
Figure 5.31: An example of barrier access details
5.5 Aesthetic Road Safety Barriers
In areas such as parks, historical communities and scenic areas, roads must not only provide safe and
efficient access but also preserve the environmental and aesthetic qualities of the area. To achieve this
balance some barriers are able to have textured finishes applied. Textures that do not result in excessive
vehicle damage may be considered acceptable for concrete vertical wall barriers or constant slope barriers.
Alternative textures have been tested in the USA and found to be acceptable. Some guidelines for
acceptable textures have been developed by the Federal Highway Administration (FHWA 2002).

5.6 Other Road Safety Barrier Design Considerations
5.6.1 Barriers at Intersections
Intersections present special problems for barrier design because the corner radius is relatively tight and
rather than impacting at acute angles typical of barriers adjacent to highway alignments, the impacts may be
at any angle, including a right angle. Rigid and standard semi-rigid barriers will result in a high-severity crash
whereas a flexible barrier is unsuitable for such small radii. However, some designs have been developed to
reduce the severity of such treatments.
The number of suitable choices becomes limited when continuing a barrier around a corner such as at the
intersection of an overpass bridge and a freeway ramp. Special proprietary treatments may be suitable and
should be considered. Wire rope safety barriers cannot be used on tight radius situations less than 200 m.
Intersection corners often accommodate road furniture such as signs, utility and signal poles and traffic
control boxes, and any fixed hazards should be moved as far away from the traffic lane as practicable. The
barrier systems that could be installed to shield these fixed objects may represent as much or even more of a
hazard than the shielded objects themselves.
Where the intersection is adjacent to an overpass, consideration should be given to the protection of the
traffic on the road or rail below the overpass. If the volume of traffic on the lower road is so great that an
errant vehicle would be likely to be involved in a secondary crash, then it may be appropriate to provide a
strong barrier on the corner to minimise this risk. A concrete barrier may be preferred in this situation.
Prior to adopting such a treatment, alternative options should be considered such as closure or relocation of
the intersecting road. Sight distances to and from side roads must not be impeded by barriers.
Figure 5.32 and Figure 5.33 show examples of designs for curved barriers at major road intersections. In
order to be effective, the treatments must be installed in accordance with the notes in the figures. The
principle of the designs is that the barrier forming the corner radius is designed so that a design car
impacting at a high angle is contained and decelerates at an acceptable rate. This is achieved:
• using breakaway posts at 2.0 m spacing
• by omitting blockouts
• by not providing washers on the mushroom-headed (coach) bolts connecting the rail to the blockouts.
An additional measure in the case of radii < 10 m is to omit the bolts that attach the rail to the post at the
centre of the curve. This creates a curved rail that has been shown to contain vehicles that impact at high
angles.

Figure 5.32: An example of a curved barrier at a major road intersection (radius 2.5 to 9.9 m)

Figure 5.33: An example of a curved barrier at a major road intersection (radius ≥ 10 m)
The requirements described and noted in Figure 5.32 and Figure 5.33 are essential for safe operation of
these curved barrier sections. A designated run-out area behind the barrier should be kept free of hazardous
objects.
In these situations, it is essential that the barrier is located in a position where it does not impede sight
distance. Sight distance requirements at intersections are covered in Section 3 of the Guide to Road Design
Part 4A: Unsignalised and Signalised Intersections (Austroads 2017b).
These treatments should be provided with a flat area graded at 10:1 or less and free of fixed hazards as
shown in the figures. If these criteria cannot be met, then a non-weakened barrier is to be installed.
The use of the weakened W-beam is more critical at higher-speed locations (i.e. > 80 km/h).
5.6.2 Stepped Offset
Varying offsets to barriers along a length of road may cause guidance problems at night. Varying offsets of
delineators on barriers may be confusing in the dark because steps in the barrier offset create a broken line
of reflectors that is not consistent with the lane marking. In such circumstances designers should consider
omitting the delineators on barriers and use alternative delineation (e.g. raised reflective pavement markers
and guide posts).

5.6.3 Excessive Offset
At night, a large offset between a barrier and the edge of the travelled way can give the impression that an
extra lane is available between the edge line and the offset barrier. This may lead to crashes where drivers
have moved onto the shoulder and verge in the mistaken belief that an extra lane is available. Alternative
delineation should be provided closer to the road. Refer to AS 1742.2-2009 for details.
5.6.4 Delineation
The terminal ends of barriers should be clearly delineated to avoid impact when drivers pull off the road,
particularly at night. Delineation of the terminal ends of wire rope barriers is especially important because
they are difficult to see both night and day.
Delineation on barriers can conflict with the guidance provided by guide posts and raised pavement markers.
As noted above this may be a problem where a barrier is offset at about a lane width from the traffic lanes.
If a barrier is located beyond the edge of a shoulder, then consideration should be given to the need for
guide posts to define the area where it is safe to pull off the road.
5.6.5 System Height
An issue associated with barriers is the reduction in their height above pavement level when overlays or
resurfacing are implemented. Reduction in the height of a barrier can adversely affect its operation, by
increasing the risk of a vehicle vaulting the system.
A further consideration that needs to be assessed is the crossfall slope in front of the barrier, which should
not be steeper than 10:1. Refer also to Figure 5.2 and Section 5.3.4 – Determine the Lateral Position of the
Barrier (Step B3).
Concrete barriers must not be placed on top of any kerb profile, as this will raise the height of the system,
generating the potential to roll small to medium-sized vehicles.

6. Roadside Design for Steep Downgrades
NOTE: This section is unchanged and may be updated in the next edition of Part 6.
6.1 Purpose and Need
Long, steep downgrades can result in the drivers of heavy vehicles losing control and it may therefore be
desirable to take measures to prevent the occurrence and limit the consequences of runaway heavy
vehicles. Out-of-control vehicles result from drivers losing control because of the loss of brakes through
overheating or mechanical failure or because the driver failed to change down gears at the appropriate time.
When considering the provision of runaway vehicle facilities, it is suggested that road agencies liaise with
stakeholders with respect to the location, spacing and design of the facilities. Measures aimed at managing
errant vehicles on steep descents include:
• alerting drivers of a steep descent on the approach to the downgrade
• regulating the use of a low enough gear to control the descent speed of heavy vehicles
• providing containment facilities for runaway vehicles.
Standard traffic signs exist to warn drivers of steep descents and to instruct drivers to use a low gear (refer to
AS 1742.2-2009 or NZ Transport Agency (2010a)).
6.2 Containment Facilities
Runaway vehicle containment facilities include the:

• gravity safety ramp
• arrester bed
• dragnet.
A combination of these facilities may be needed to suit a particular site. In addition, in some cases it may be
desirable to place an energy absorbing barrier at the end of a safety ramp or arrester bed to cover an event
where a vehicle has not totally decelerated within the ramp or bed (e.g. natural compaction of bed material
reduces its effectiveness).
6.2.1 Gravity Safety Ramps
Gravity safety ramps use an ascending grade to reduce the speed of a runaway vehicle. Ramps are normally
hard surfaced and take advantage of naturally occurring grades on a mountain range.
6.2.2 Arrester Beds
Arrester beds are long trenches filled with small, round gravel particles that are designed to stop runaway
trucks. The truck is stopped by drag and friction as the vehicle sinks into the gravel in the bed.
Arrester beds are classified as a:

• direct entry arrester bed
• side entry arrester bed – full width
• side entry arrester bed – half width.

6.2.3 Dragnets
A dragnet vehicle-arresting barrier consists of a chain link net that is attached to energy absorbing poles.
Several nets in series are needed to capture heavy vehicles. Design of dragnet systems needs to be in
accordance with the manufacturers’ design parameters.
6.3 Warrants for Investigation
Downgrades have the potential to cause brake fade in heavy vehicles and can be considered for treatment
to reduce the risk of runaway vehicles. Grade and length combinations that warrant investigation are shown
in Table 6.1.
Table 6.1: Typical warrants for analysis for runaway vehicles
Minimum continuous length

Grade
(km)
–3% 8.0
–5% 3.1
–7% 1.9
–9% 1.4
–12% 1.0
Where a warrant has been established for investigation for treatment of a steep downgrade the design of
treatments should follow the process in Section 6.6.
6.4 Location and Spacing
Runaway vehicle facilities should not be constructed where an out-of-control vehicle would need to cross
oncoming traffic. On undivided roads, safety ramps should ideally be located at the start of a right-hand
curve as the runaway vehicle can readily negotiate a tangential path into the ramp. On divided roadways
where adequate space is available in the median, safety ramps can be located on either side of the
carriageway provided that adequate advance warning signs are erected prior to the safety ramp exit.
For safety ramps to be effective their location is critical. They should be located prior to or at the start of the
smaller radius curves along the alignment. For example, an escape ramp after the tightest curve will be of
little benefit if trucks are unable to negotiate the curves leading up to it. Vehicle brake temperature is a
function of the length of the grade, therefore escape ramps are generally located within the bottom half of the
steeper section of the alignment.
Lack of suitable sites for the installation of ascending-type ramps may necessitate the installation of
horizontal or descending arrester beds. Suitable sites for horizontal or descending arrester beds can also be
limited, particularly if the downward direction is on the outside or fill side of the roadway formation.
For new projects Table 6.2 may be used as a guide when considering the need for escape exits on grades
greater than 6% and with numbers of commercial vehicles exceeding 150 per day.
The distances in the table are not absolute and greater distances could be acceptable, as site location is
dependent on other factors. The need for a facility will be increased if the number of commercial vehicles is
more than 250 per day and the maximum decrease in operating speed between successive geometric
elements is approaching the limits set in Table 6.3.

Table 6.2: Approximate distance from summit to safety ramp
Approximate distance from summit to the

Grade (%)
ramp (km)
6–10 3.0
10–12 2.5
12–15 2.0
15–17 1.5
17 1.0
Note: Actual distances will depend on site topography, horizontal curvature and costs.
Table 6.3: Maximum decrease in speed between successive geometric elements
Maximum decrease in speed between

Grade (%)
successive geometric elements (km/h)
<6 10
6–10 8
> 10 6
6.5 Key Design Considerations
The design, construction and maintenance of runaway vehicle facilities should ensure that the:
• length of the escape ramp is sufficient to dissipate the kinetic energy of the vehicle
• alignment of the ramp is straight or of very gentle curvature to relieve the driver of undue vehicle control
problems
• width is wide enough to accommodate two vehicles if it is considered likely that a second vehicle will need
to use the ramp soon after the first one
• adequate work space is available for heavy vehicle removal (e.g. lifting cranes)
• arrester bed material is clean, not easily compacted or consolidated and has a high coefficient of rolling
resistance
• full depth of the arrester bed is achieved in the first 50 m of the entry to the bed using a tapering depth
from 50 mm at the start to the full depth at 50 m
• bed is properly drained
• entrance to the ramp is designed so that a vehicle travelling at high speed can enter it safely. A 5° angle
of departure or less is required, and as much sight distance as possible should be provided. The leading
edge of the arrester bed must be normal to the direction of entry to ensure that the two front wheels of the
vehicle enter the bed simultaneously
• signing is in accordance with the appropriate standard to alert the driver to the presence of the escape
ramp. The location of signs, street lighting poles and overhead power lines should not obstruct the
operation of the arrester bed or retrieval operations. Routine maintenance of any light poles should not
impose any entry restriction to the arrester bed at any time.
• facility, where necessary, has an emergency roadside phone with connection to an operations centre or
emergency service placed in a visible and easily accessible location.

In addition, the following operational factors should be considered:

• The alignment of all curves preceding the ramp should be checked to ensure that a runaway vehicle can
safely negotiate them at the speeds estimated to be likely.
• Vehicles that enter the ramp will have to be retrieved, as it is unlikely that they will be able to be driven
from the arrester bed. An appropriate service road adjacent to the ramp is required to enable retrieval. An
alternative and/or enhancement to the service road is the provision of anchorage points/blocks for
winching vehicles out.
• When the location of the ramp is such that the length is inadequate to fully stop an out-of-control vehicle,
a positive attenuation (or ‘last chance’) device may be required. Care is required to ensure that the device
does not cause more problems than it solves – sudden stopping of the truck can cause the load to shift
with potentially harmful consequences to the driver and the vehicle. Judgement will be required on
whether the consequences of failing to stop are worse than these effects. Crash cushions or piles of sand
or gravel have been used as last chance devices.
6.6 Design Process
6.6.1 Outline of Process
The steps in the design process for treatment of steep downgrades are shown in Figure 6.1. Steps S1 to S11
in the process are discussed in Sections 6.6.2 to 6.6.12 respectively.
Figure 6.1: Design process for steep downgrade
Step
S1 Determine the vehicle entry speed
S2 Evaluate truck stability on approach
S3 Design entry alignment
S4 Determine the types of facility
S5 Determine the pavement surface of the facility
S6 Design facility length
S7 Design the facility
S8 Design end treatment
S9 Design vehicle recovery facilities
S10 Design delineation
S11 Design truck parking areas

6.6.2 Step S1 – Determine Vehicle Entry Speed
The recommended design vehicle should be identified as part of the design process and should be used in
determining the vehicle entry speed to the facility. Heavy runaway vehicles attain high speeds but speeds in
excess of 130 to 140 km/h will rarely, if ever, be attained. An escape ramp should therefore be designed for
a minimum entering speed of 130 km/h, a 140 km/h design speed being preferred. Several formulae and
software programs have been developed to determine the runaway speed at any point on the grade. These
methods can be used to establish a design speed for specific grades and horizontal alignments
(AASHTO 2011b).
6.6.3 Step S2 – Evaluate Truck Stability on Approach
The designer should check that the truck can reach the facility at the calculated speed and will not roll over
on curves uphill of the runaway containment facility. The maximum cornering speed is given by Equation 4
and can be expressed as:
v= �(agR) 4
where
v = speed of vehicle (m/s)
a = maximum lateral acceleration 0.3 g
g = gravitational constant 9.81 m/sec2
R = radius of curvature
If trucks are likely to roll over before reaching the containment facility then relocation of the facility should be
considered, if the terrain allows.
6.6.4 Step S3 – Design Entry Alignment
The entry speed of a runaway vehicle is used for designing the approach and entry to safety ramps and
arrester beds. The alignment of the escape ramp should be at a tangent or very flat curvature to reduce the
likelihood that the driver will experience vehicle control problems. Designers should refer to AGRD Part 3
(Austroads 2016a) in designing the vertical and horizontal alignment of the treatment.
6.6.5 Step S4 – Determine Type of Facility
The constraints imposed by the terrain will largely determine the type of facility to be implemented. Several
iterations of design may be necessary if a combination of facility types proves to be necessary. Changes to
the type of facility and pavement type may be necessary to determine the best fit to the site constraints.
6.6.6 Step S5 – Determine Pavement Surface of Facility
The rolling resistance of the facility pavement will have a significant influence on the length required for the
containment facility. The values shown in Table 6.4 are used for length calculations.
From field tests and other research studies, rounded particles such as uncrushed river gravel with uniform
gradation produce higher deceleration than the more angular crushed aggregate. This is because the
vehicles sink deeper into the river gravel, transferring more energy to the stones over a shorter length. The
use of a material with low shear strength is desirable in order to permit tyre penetration.
Crushed stone has been used but is not considered effective as it will require longer beds and will need
regular ‘fluffing’ or de-compaction.

Sand also has problems of drainage, compaction and contamination – it consolidates with time and moisture
ingress – and should not be used unless alternative materials are unavailable. Beds using sand will require a
strict maintenance regime to ensure their continued effectiveness. However, all arrester beds and bedding
materials require regular maintenance.
Nominal 10 mm river gravel has been used satisfactorily in testing. The gravel should be predominantly
rounded, of uniform gradation, free from fine fractions and with a mean particle size ranging between
12 and 20 mm. In general, gravels with a smaller internal friction angle will perform better than those with
larger angles.
An appropriate crush test such as the Los Angeles abrasion test (or equivalent) should be used to evaluate
the durability of the stone. Stones with a high crush test will not deteriorate and will therefore not produce
fines.
Table 6.4: Rolling resistance
Rolling resistance
Equivalent grade
Material (kg/1000 kg gross vehicle mass)
(%)
(R)
Cement concrete pavement 10 1.0
Asphaltic concrete pavement 12 1.2
Compacted gravel 15 1.5
Earth, sandy, loose 37 3.7
Crushed aggregate, loose 50 5.0
Gravel loose 100 10.0
Sand 150 15.0
Pea gravel 250 25.0
Note: Pea gravel is rounded gravel having a uniform particle size of about 10 mm.
6.6.7 Step S6 – Design Facility Length
The length of a containment facility will vary depending on entry speed, grade, pavement surface and the
type of facility.
The vehicle entry speed described in Section 6.6.2 is used as the initial velocity for determining the length of
an arrester bed. The length of an arrester bed (Roads and Traffic Authority 2000) is given by Equation 5 and
is expressed as:
V2i - V2f
L= 5
2.54(R+G)
where
L = length travelled (m)
Vi = initial velocity (km/h)
Vf = final velocity (km/h)
R = grade in per cent
G = rolling resistance expressed as a grade in percent from Table 6.4
Where there is a grade change in the arrester bed, Equation 5 can be used to calculate the length required
on each grade. The final velocity of a section becomes the initial velocity of the next section.
The length of a half-width arrester bed is double the length of a full-width arrester bed.

6.6.8 Step S7 – Design the Facility
General
This step requires the preparation of the layout and design of the facility. Several iterations with different
combinations of facility types may be necessary.
Safety ramp design features
The grade of the safety ramp will be largely determined by the terrain. Safety ramps need steep-sided cut
batters on both sides. When a runaway vehicle stops in a ramp it will begin to roll back because the brakes
are not functional. In this situation drivers must jack-knife the vehicle against the sides of the ramp to prevent
it rolling down the ramp.
Arrester bed design features
Arrester beds aim to provide deceleration similar to an emergency braking situation to avoid the risk of the
truck cabin being crushed by a shifting load. Arrester beds can be constructed on up, level or downgrades
depending on the topography at the site (refer to Commentary 8). Arrester beds on downgrades require
additional length to bring out-of-control vehicles to rest. An example of an arrester bed is shown in Figure 6.2
while Figure 6.3 shows an example of a layout.
[see Commentary 8]
Figure 6.2 shows that a working area for a retrieval vehicle or crane is provided only on one side of the bed
whereas the example layout in Figure 6.3 shows a service road on both sides. An access area on both sides
of the arrester bed will not be necessary in many cases but may be required where heavy vehicles on a route
carry very heavy or difficult loads that require retrieval vehicles or cranes to work from both sides.
Figure 6.2: An example of an arrester bed

Figure 6.3: An example of an arrester bed layout
Source: Austroads (2003).
Table 6.5 provides a summary of the features associated with arrester beds and key considerations required
in design.
Table 6.5: Design features of arrester beds
Feature Consideration
Horizontal Some steering is possible in a gravel arrester bed.
alignment Where a curve is necessary the radius should be generous, well in excess of standard travel speed
to radius ratio.
Lateral The round gravel sprayed or dislodged from the arrester bed may be a hazard to passing vehicles
location as it is likely to cause crashes due to loss of steering and traction on adjoining traffic lanes.
An arrester bed should be located more than 4 m from traffic lanes. This offset provides an access
area for recovery vehicles and provides a space for containment of sprayed gravel. An alternative is
to provide a kerb at an offset to the traffic lane to contain gravel that would be swept onto traffic
lanes.
Kerbs should not prevent a grader with blade extension being used to recover and grade gravel after
a vehicle has been removed from the arrester bed.
Width A width of 5 m gives some room for steerage yet should control a heavy vehicle if it starts to get out
of control within the bed.

Feature Consideration
Depth A gradual or staged increase in the depth of the bed should be provided on the entry ramp. There
should be a gradual increase in aggregate depth in the first 30 m although the initial depth of the
aggregate need not start at zero. This gradual increase also assists in vehicle extraction.
A maximum bed depth of 350 mm provides adequate deceleration without causing damage to the
vehicle.
Higher deceleration rates can be achieved by increasing the bed depth up to 450 mm; however,
driver safety may be jeopardised and damage may be caused to the vehicle.
An increase in depth to 450 mm at the end of the bed will provide for higher-speed vehicles to be
arrested at the point where vehicle speed has been reduced by the treatment.
Base The base of an arrester bed should be concrete with a crossfall of 2% falling towards the concrete
barrier and graded to a drainage system.
Steeper crossfalls should not be used as they cause trucks to veer off-line as they pass through the
arrester bed.
Barrier A vertical concrete barrier should be placed on the edge of the bed furthest from the traffic lane to
assist in keeping vehicles travelling along the bed.
Drainage Stormwater should be directed away from the bed. The base of the bed should be designed to
accommodate drainage to help protect the bed from freezing and avoid contamination of the arrester
bed material by accumulation of fines that would compact the bed material.
Installation of perforated drains in the base of the bed and lining the bed base and sides with asphalt
or cemented material are required.
Fuel spill Truck fuel lines may be ruptured when impacting the gravel in an arrester bed. The drainage system
containment of the arrester bed should be fitted with a fuel spill containment facility.
Arrester bed Rounded pea gravel in loose condition is essential to make an arrester bed effective. The aggregate
material should be predominantly single sized and uniform. It should be clean, free of fines and have smooth
rounded surfaces. The recommended specification is shown in Table 6.6. Deceleration
characteristics of the bedding material may be affected by wet weather.
Table 6.6: Arrester bed material specification
Criteria Percent passing AS sieve (%)

Per cent passing 19 mm AS sieve 100%
Per cent passing 9.5 mm AS sieve 0–5%
Per cent passing 0.075 mm AS sieve Maximum 2%
Fractured faces Maximum 10%
2/1 Misshapen particles Maximum 10%
Crushing Not greater than 5%
Cracking Not greater than 5%
Slump angle Not greater than 30°
Bulk density Not greater than 3.4 tonnes/cu m
Note: Washed and screened uncrushed river gravel could meet this specification.
6.6.9 Step S8 – Design End Treatment
The consequences of a vehicle passing through and out of the ramp or arrester bed should be considered.
Crash cushions are designed for cars and have limited effectiveness for trucks. They should only be
considered where they would act as a cushioning device before a rigid object such as a rock face.
A dragnet system may be needed if a ‘fail-safe’ end treatment is not available.

6.6.10 Step S9 – Design Vehicle Recovery Facilities
Access and anchors for cranes and/or tow trucks should be designed and provided to facilitate removal of
the disabled vehicle from the containment facility. The design of removal facilities should ensure the
occupational health and safety of removal workers.
Safety ramp recovery facilities
If separate access is available to the top of the ramp a tow truck can be used for vehicle recovery. A large
anchor block should be buried below the surface at the top of the ramp as recovery will be assisted if the tow
truck can be chained to an anchor while winching the runaway vehicle up the steep slope.
If access is not available to the top of the ramp it will be necessary to use a bulldozer to engage the rear of
the runaway vehicle and lower it backwards down the slope and hence the treatment should be designed for
this loading.
Arrester bed recovery facilities
To facilitate recovery of vehicles from an arrester bed a service road adjacent to the bed must be provided
with a minimum width of 3.5 m. Access to the service road should be available for either two heavy-duty tow
trucks or two 50 tonne capacity cranes and therefore the pavement of the service road should be capable of
supporting these cranes. The service road should also be designed so a grader with blade extension can
grade the gravel after a vehicle has been removed.
Anchor blocks are required to secure tow trucks while winching vehicles out of the arrester bed. Anchor
blocks are to be located at 35 m intervals along the service road and 10 m from the entry and end of the
arrester bed. Anchors should be designed to a 35 tonne winching force through an attachment shackle rated
to withstand the design load. Attachment shackles should be recessed flush with pavement levels.
It is preferable that recovery is made from the exit end of the arrester bed, as articulated vehicles will
jack-knife if dragged backwards through the bed.
To enable drivers to get assistance, CB radio frequency or telephone numbers for emergency service may
be advised on signposting adjacent to the bed or service road.
6.6.11 Step S10 – Design Delineation
The existence and location of a containment facility must be made obvious by signing to give the operator of
an out-of-control vehicle time to react and decide to enter the facility. Standard signs should be provided and
located in accordance with AS 1742.2-2009 or NZ Transport Agency (2010a). Signs that are likely to be
required are those that:
• warn or advise of a steep descent
• provide advance notification of the facility
• indicate direction at the facility entrance
• advise truck and bus drivers to use a low gear.
Adequate delineation should also be provided so that the entrance to a containment facility is not mistaken
for the through carriageway and the entry path to the facility is clear by day and night.
6.6.12 Step S11 – Design Truck Parking Areas
Truck parking areas before steep grades provide an area for truck drivers to stop and check the brakes of
the vehicle. This area is also called a brake check area. A brake rest area, however, is an area set aside
part-way down or at the bottom of the descent.

These areas enable truck drivers to stop and allow brakes to cool before the descent is negotiated. Brake
check areas would naturally be on the top of a hill and should be easily accessible to heavy vehicles. Good
sight distance should be available at both entry and exit.
The area available will vary with the number of trucks using the route. Long stays in these areas should be
discouraged. Other truck rest stops should be available within a reasonable distance of the brake check
parking area.
These facilities should be provided on routes that have long, steep downgrades and commercial vehicle
numbers that exceed about 100 per day, especially on National Highways and principal traffic routes. These
areas:
• ensure that drivers begin the descent at zero velocity and in a low gear that may make the difference
between controlled and out-of-control operation on the downgrade
• provide an opportunity to display information about the grade ahead, escape ramp locations and
maximum safe descent speeds
• may need to be large enough to store several prime mover and semi-trailer combinations, the numbers
depending on volume and predicted arrival rate
• should desirably have a sealed surface, or at least a well-compacted gravel surface.
Good visibility to the areas and adequate acceleration and deceleration tapers should be provided.
Adequate signing should be provided to advise drivers in advance of the facilities. Special signs, specific to
the site, may need to be designed.

7. Work Zone Safety Barrier Systems
NOTE: This section has been adopted from MRWA Guide to the Design of Workzone Barriers
(2018) and may be updated in the next edition of Part 6.
Occupational health and safety is an important aspect in the management of road systems and it is
necessary to protect the workplace, and the road users travelling through the work site or adjacent to it, as
far as is practicable.
Contributing to hazards at work sites are increasing traffic volumes, larger vehicles and in some cases higher
speeds through and adjacent to work sites. Inappropriate speed limits or operating speeds can also increase
the risk for road users, both during and after work-site operating hours. Where long-term construction sites
are created, enforcement of speed can be a problem even with active police participation. While publicity
campaigns and enforcement (often utilising radar devices and speed cameras) may assist, generally these
measures are outside the control of the personnel on a construction site.
Separating the traffic from the workplace is one way of reducing the hazards. This separation can be
achieved by the consistent and appropriate usage of safety barriers on work sites. Appropriate standards,
and education and training of the people that are responsible for the erection and maintenance of the
barriers are necessary to ensure their proper usage.
The design requirements for temporary roads and appropriate geometric design to control speeds through
work sites must be implemented to ensure that the temporary barriers used will be appropriate. Barrier
installations must meet the requirements for permanent installations where high speeds are allowed through
the work site (≥ 80 km/h).
Construction and maintenance managers, engineers, supervisors and workers must take an active role in
ensuring that work sites are safe for workers and road users. It is particularly important that where safety
barriers are necessary they are used in a consistent and appropriate way. This can only be achieved if
personnel are educated and trained and have knowledge of guides and standards, and experience in their
application. It is equally important that designers focus on guiding traffic safely through work sites by:
• the development of effective temporary traffic management plans including standards for sidetracks and
deviations for traffic within sites
• provision of effective and well-maintained signs and markings (including regulatory, warning and guide
signs and delineation) that meet the requirements of AS 1742.3-2009.
It is very important that work zone barriers are located so that they do not place road users at risk by
restricting sight distance for traffic entering, crossing or moving through the work site. Care must also be
taken to ensure that drainage is adequate so that stormwater does not form ponds adjacent to the barrier or
flow across the road at a depth that could cause vehicles to aquaplane.
Work-site barriers are therefore used to contain and redirect errant vehicles to prevent them from leaving the
roadway and/or entering the work site. Like permanent barriers, they should only be used if they reduce the
severity and adverse consequences of potential crashes, as they are a hazard in themselves. As well as
enhancing site safety, they may improve job productivity and reduce road user delays.
Guidance on the use of barriers at work zones is provided in AS 1742.3-2009.

References
AASHTO 1989, Roadside design guide, 1st edn, American Association of State Highway and Transportation
Officials, Washington, DC, USA.
AASHTO 1996, Roadside design guide, 2nd edn, American Association of State Highway and Transportation
AASHTO 2011a, Roadside design guide, 4th edn, American Association of State Highway and Transportation
AASHTO 2011b, A policy on geometric design of highways and streets, 6th edn, American Association of
State Highway and Transportation Officials, Washington, DC, USA.
AASHTO 2014, Highway safety manual first edition with 2014 supplement, American Association of State
Highway and Transportation Officials, Washington, DC, USA.
AASHTO 2016, Manual for assessing safety hardware, 2nd edn, American Association of State Highway and
Transportation Officials, Washington, DC, USA.
AASHTO 2018, A policy on geometric design of highways and streets, 7th edn, American Association of
State Highway and Transportation Officials, Washington, DC, USA.
Alberson, DC, Bligh, RP, Buth, CE & Bullard, DL 2003, ‘Cable and wire rope barrier design considerations
review’, Transportation Research Record, no. 1851, pp. 95-104.
Albuquerque, FDB, Sicking, DL & Stolle, CS 2010, ‘Roadway departure impact conditions’, Transportation
Research Record, no. 2195, pp. 106-14.
Australian Transport Assessment and Planning Guidelines (ATAP) 2018, T2 cost benefit analysis,
Commonwealth Department of Infrastructure and Regional Development, Canberra, ACT.
Australian Transport Council 2011, National road safety strategy 2011 – 2020, ATC, Canberra, ACT.
Australian Transport Safety Bureau 2000, Review of wire rope safety barriers, ATSB, Canberra, ACT.
Austroads 2003, Rural road design: a guide to the geometric design of rural roads, AP-G1-03, Austroads,
Sydney, NSW (superseded).
Austroads 2004, Guide to traffic engineering practice part 12: roadway lighting, AP-G11.12/04, Austroads,
Austroads 2008, Guide to road safety part 9: roadside hazard management, AGRS09-08, Austroads,
Sydney, NSW.
Austroads 2010, Guide to road design part 6: roadside design, safety and barriers, AGRD06-10, Austroads,
Austroads 2011, Improving roadside safety: stage 2: interim report, AP-R387-11, Austroads, Sydney, NSW.
Austroads 2013a, Guide to road design part 5: drainage, general and hydrology considerations, edn 3.0,
AGRD05-13, Austroads, Sydney, NSW.
Austroads 2013b, Guide to road design part 5B: drainage: open channels, culverts and floodways,
AGRD05B-13, Austroads, Sydney, NSW.
Austroads 2014a, Improving roadside safety: summary report, AP-R437-14, Austroads, Sydney, NSW.
Austroads 2014b, Australian National Risk Assessment Model, AP-R451-14, Austroads, Sydney, NSW.

Austroads 2015a, Guide to road design part 1: introduction to road design, edn 4.0, AGRD01-15, Austroads,
Sydney, NSW.
Austroads 2015b, Guide to road design part 6B: roadside environment, edn 2.0, AGRD06B-15, Austroads,
Sydney, NSW.
Austroads 2015c, Guide to road design part 4C: interchanges, edn 2.0, AGRD04C-15, Austroads, Sydney,
NSW.
Austroads 2015d, Guide to road safety part 8: treatment of crash locations, edn 2.0, AGRS08-15, Austroads,
Sydney, NSW.
Austroads 2015e, Guide to road design part 4B: roundabouts, edn 3.0, AGRD04B-15, Austroads, Sydney,
NSW.
Austroads 2015f, Austroads glossary of terms, edn 6.0, AP-C87-15, Austroads, Sydney, NSW.
Austroads 2016a, Guide to road design part 3: geometric design, edn 3.3, AGRD03-16, Austroads, Sydney,
NSW.
Austroads 2016b, Guidance on median and centreline treatments to reduce head-on casualties, AP-R519-
16, Austroads, Sydney, NSW.
Austroads 2016c, Safe System assessment framework, AP-R509-16, Austroads, Sydney, NSW.
Austroads 2017a, Guide to road design part 6A: paths for walking and cycling, edn 2.0, AGRD06A-17,
Austroads, Sydney, NSW.
Austroads 2017b, Guide to road design part 4A: unsignalised and signalised intersections, edn 3.0,
AGRD04A-17, Austroads, Sydney, NSW.
Austroads 2017c, Verification of Austroads road design criteria based on objective safety evidence, AP-
T320-17, Austroads, Sydney, NSW.
Austroads 2018a, Towards safe system infrastructure: a compendium of current knowledge, AP-R560-18,
Austroads, Sydney, NSW
Austroads 2018b, Guide to bridge technology part 3: typical superstructures, substructures and components,
AGBT03-18, Austroads, Sydney, NSW.
Austroads 2018c, Guide to asset management technical information part 15: technical supplements,
AGAM15-18, Austroads, Sydney, NSW.
Austroads 2019, Guide to road design part 2: design considerations, edn 2.1, AGRD02-19, Austroads,
Sydney, NSW.
Austroads2020a, Guide to traffic management part 4: network management strategies, AGTM04-20,

Austroads, Sydney, NSW.
Austroads2020b, Background to the development of the 2020 edition of Guide to Road Design Part 6, AP-
R628-20, Austroads, Sydney, NSW.
Bergh, T & Carlsson, A 1999, ‘2+1-Roads with and without cable barriers: speed performance’,
Transportation Research Board annual meeting, 78th, Washington, DC, TRB, Washington, DC, USA.
Burbridge, AZ 2019, ‘The road safety barrier dilemma revisited: balancing capacity against flexibility’, IPWEA
SWQ branch conference, 2019, Gatton, Qld, Institute of Public Works Engineering Australasia, Albion,
Qld.

Carrigan, CE & Ray, MH 2017, ‘A new approach to run-off road crash prediction’, Transportation Research
Board annual meeting, 96th, Washington, DC, TRB, Washington, DC, USA.
Carrigan, CE, Ray, MH & Johnson, TO 2014, ‘Understanding of heavy vehicle encroachment frequency’,
Transportation Research Record, no. 2437, pp. 20-6.
Commonwealth of Australia 2018, National Road Safety Action Plan 2018-2020, INFRA3507, Transport and
Infrastructure Council, Canberra, ACT.
Department of Planning, Transport and Infrastructure 2017, Technical standards and guidelines, DPTI,
Adelaide, SA.
Dixon, KK, Liebler, M, Zhu, H, Hinbter, MP & Mattox, B 2008, Safe and aesthetic design of urban treatments,
report 612, National Cooperative Highway Research Program, Transportation Research Board,
Washington, DC, USA.
Doecke, S & Woolley, JE 2010, ‘Effective use of clear zones and barriers in a safe system's context’,
Australasian road safety research, policing and education conference, 2010, Canberra, ACT,
Conference Logistics, Kingston, ACT, 10 pp.
Duncan, C, Corben, B, Truedsson, N & Tingvall, C 2000, Motorcycle and safety barrier crash testing:
feasibility study, report CR 201, Australian Transport Safety Bureau, Canberra, ACT.
FHWA 1998, ‘Approval letter CC-53’, July 1998, Federal Highway Administration, Washington, DC, USA.
FHWA 1999, ‘Approval letter HMHS-B58’, 8 October 1999, Federal Highway Administration, Washington,
DC, USA.
FHWA 2001, ‘Approval letter CC-53A’, April 2001, Federal Highway Administration, Washington, DC, USA.
FHWA 2002, ‘Approval letter HSA-10/B110’, 20 December 2002, Federal Highway Administration,
Washington, DC, USA.
Giummarra, G 2001, Road classifications, geometric designs and maintenance standards for low volume
roads, ARR 354, ARRB Transport Research Ltd, Vermont South, Vic.
Hutchinson, JW & Kennedy, TW 1966, Medians of divided highways: frequency and nature of vehicle
encroachments, bulletin 487, University of Illinois Engineering Experiment Station, IL, USA.
Jurewicz, C & Pyta, V 2011, Improving roadside safety: stage 2: interim report, AP-R387-11, Austroads,
Sydney, NSW.
Kloeden, C & McLean, J 1999, ‘Roadside hazard involvement in severe and fatal car crashes’, Road safety
research, policing, education conference, 2nd, 1999, Canberra, ACT, Australian Road Safety Bureau,
Canberra, ACT, pp. 957-64.
Knuiman, MW, Council, FM & Reinfurt, DW 1993, The association of median width and highway accident
rate, HSIS summary report, FHWA-RD-93-046, Federal Highway Administration, McLean, VA, USA.
Koch, H & Schueler, F 1988, ‘Reduction of injury severity involving guardrails by the use of additional W-
Beams, impact attenuators and ‘sigma-posts’ as a contribution to the passive safety of motorcyclists’,
International technical conference on experimental safety vehicles, 11th, Washington, DC, National
Highway Traffic Safety Administration, Washington, DC, USA, pp. 878-83.
Mak, K, Sicking, DL & Coon, BA 2010, Identification of vehicular impact conditions associated with serious
ran-off-road crashes, report 665, National Cooperative Highway Research Program, Transportation
Research Board, Washington, DC, USA.

Main Roads Western Australia 2017, Supplement to Austroads Guide to Road Design: Part 6, MRWA, Perth,
WA.
Main Roads Western Australia 2018, Guide to the design of workzone barriers, D11#38509, MRWA, Perth,
WA.
NZ Ministry of Transport 2010, Safer journeys: New Zealand’s road safety strategy 2010-2020, Ministry of
Transport, Wellington, NZ.
NZ Transport Agency 2010a, Manual of traffic signs and markings (MOTSAM): part 1: traffic signs, NZTA,
Wellington, NZ.
NZ Transport Agency 2010b, Manual of traffic signs and markings (MOTSAM): part 2: markings, NZTA,
Wellington, NZ.
NZ Transport Agency 2011 High risk rural roads guide, NZTA, Wellington, NZ.
NZ Transport Agency 2013, Bridge manual, 3rd edn, SP/M/022, NZTA, Wellington, NZ.
Queensland Department of Transport and Main Roads 2013, Road planning and design manual, 2nd edn,
TMR, Brisbane, Qld.
Queensland Department of Transport and Main Roads 2016, RISC version 6 user manual, TMR, Brisbane
Qld.
Queensland Department of Transport and Main Roads 2017, Roadworks drainage culverts and geotechnical,
TMR, Brisbane, Qld, viewed 12 June 2020, <https://www.tmr.qld.gov.au/business-industry/Technical-
standards-publications/Standard-drawings-roads/Roadworks-drainage-culverts-and-
geotechnical#ConcreteBarriersAndGuardrails>.
Ray, MH, Carrigan, CE, Plaxico, CA, Miaou, SP & Johnson, TO 2012a, Roadside Safety Analysis Program
(RSAP) update: version 3, produced for the National Cooperative Highway Research Program Project
22-27, Transportation Research Board, Washington, DC, USA.
Ray, MH, Carrigan, CE, Plaxico, CA, Miaou, SP & Johnson, TO 2012b, Roadside Safety Analysis Program
(RSAP) update: version 3: appendix B: engineer’s manual, produced for the National Cooperative
Highway Research Program Project 22-27, Transportation Research Board, Washington, DC, USA.
Roads and Maritime Services 2016, Roads and Maritime supplement to Austroads Guide to Road Design
Part 6: Roadside Design, Safety and Barriers (2009), RMS, Sydney, NSW.
Roads and Traffic Authority 1996, Road design guide: section 6: safety barriers for roads and bridges, RTA,
Sydney, NSW.
Roads and Traffic Authority 2000, ‘Current practice for design and maintenance of truck arrestor beds’,
report no. 002209/1, Pavement Projects Branch, RTA, Sydney, NSW.
Roads and Traffic Authority 2003, ‘Wire rope road safety barrier installed on a 200 m radius’, preliminary
barrier report BR2003/067, Crashlab, RTA, Sydney, NSW.
Roads and Traffic Authority 2008, ‘Road design guide: section 6: design for errant vehicles: hazard mitigation
and safety barrier design’, unpublished draft, RTA, Sydney, NSW.
Ross, HE, Sicking, DL, Zimmer, RA & Michie, JD 1993, Recommended procedures for the safety
performance evaluation of highway features, report 350, National Cooperative Highway Research
Program, Transportation Research Board, Washington, DC, USA.

Sala, G & Astori, P 1998, ‘New concepts and materials for passive safety of motorcyclists’, International
IRCOBI conference on the biomechanics of impacts, Göteborg, Sweden, International Research
Council on the Biomechanics of Impact, Bron, France, pp. 425-36.
Stonex, KA 1960, ‘Road design for safety’, Proceedings of the Highway Research Board, vol. 39, pp. 120-57.
Tay, R & Churchill, A 2007, ‘Effect of different median barriers on traffic speed’, Canadian Journal of
Transportation, vol. 1, no. 1, pp. 56-66.
Thomson, R & Valtonen, J 2002, ‘Vehicle impacts on v-shaped ditches’, Transportation Research Record,
no. 1797, pp. 82-8.
VicRoads 2005, Road Design Guidelines – Safety Barriers, VicRoads, Melbourne, Victoria
Washington State Department of Transportation 2014, Design manual, M 22-01.11, Washington State
Department of Transportation, Olympia, WA, USA.
Zegeer, CV & Cynecki, MJ 1984 Selection of cost-effective countermeasures for utility pole accidents: user
manual, FHWA-IP-84-13, Federal Highway Administration, Washington, DC, USA.
Zegeer, CV & Parker, MR 1984, ‘Effect of traffic and roadway features on utility pole accidents’,
Transportation Research Record, no. 970, pp. 65-76.
Australian and New Zealand Standards
AS 1742.2-2009, Manual of uniform traffic control devices: traffic control devices for general use.
AS 1742.3-2009, Manual of uniform traffic control devices: traffic control for works on roads.
AS 5100.1-2017, Bridge design: scope and general principles.
AS/NZS 1158.1.2-2010, Lighting for roads and public spaces: vehicular traffic (category V) lighting: guide to
design, installation, operation and maintenance.
AS/NZS 3845:1999, Road safety barrier systems (superseded).
AS/NZS 3845.1:2015, Road safety barrier systems and devices: part 1: road safety barrier systems.

Appendix A Terminology
A.1 Vehicle Movement Terminology
The vehicle movement terminology is described in Table A.1 and illustrated in Figure A 1.
Table A.1: Vehicle movement terminology
Term Meaning
Pitch The rotation of a vehicle around its transverse axis.
Roll The rotation of a vehicle around its longitudinal axis.
Yaw The rotation of a vehicle around its vertical axis.
Spin Uncontrolled yaw.
Lurch The acceleration of a vehicle along its transverse axis.
Surge The acceleration of a vehicle along its longitudinal axis.
Bounce The acceleration of a vehicle along its vertical axis.
Figure A 1: Vehicle movement terminology
Yaw
Roll
Offside
Pitch
F
Lurch
Surge
on of Nearside
Directi l
Trave Bounce
A.2 Road Safety Barrier Terminology
Table A.2 summarises road safety barrier terminology.
Table A.2: Road safety barrier terminology
Term Explanation
Concave Barrier curvature away from the adjacent traffic lane, i.e. inside the curve (Figure A.2).
Containment The maximum tested vehicle mass used in a set of standard crash tests.
Convex Barrier curvature towards the adjacent traffic lane, i.e. outside the curve.Figure A.2

Term Explanation
Crash Devices that prevent an errant vehicle from impacting hazardous objects by gradually decelerating
attenuator the vehicle to a safe stop or by directing the vehicle away from the hazard. They are often used as
the end treatment on the leading end of a barrier system.
Crash cushion An energy absorption device installed in front of a rigid object to reduce the severity of impact.
Departure angle The angle at which the vehicle leaves the barrier after initial impact (Figure A.3).
Development A length of unanchored barrier, in advance of the point of redirection, that is necessary to provide
length sufficient mass for the barrier within the length of need to perform in accordance with its design
parameters.
Double-sided A barrier designed for impact on both sides.
barrier
Dynamic The largest transverse deflection of a barrier system during a crash or during a full-scale impact
deflection test (i.e. the amount the barrier deflects from its initial position during impact (Figure A.3).
Exit trajectory The path followed by the vehicle after last impact with the barrier (Figure A.3).
Flare The change in the offset of a barrier to move it further from the travelled way or closer to the
travelled way.
Flare rate The curvature applied near the end of a barrier installation (Figure 5.5). Expressed as the ratio of
the longitudinal distance to the transverse offset, by which a barrier flares away from, or towards,
the edge of the travelled way.
Flexible barrier A barrier system where the barrier elements under an impact respond and absorb kinetic energy,
by substantial movement, deformation and deflection.
Gating terminal A barrier terminal designed to allow an impacting vehicle to pass through the device. Terminals
that are designed to break away, pivot or hinge, and allow a vehicle to pass through when
impacted at an angle to the end, or at a point upstream of the beginning of the length of the
associated barrier system.
Impact angle The minimum angle at which a vehicle at speed leaves the road (Figure A.3).
Initial lateral The lateral position of the vehicle prior to initial change of direction (Figure A.3).
position
Interface The length of barrier system used to connect systems with different operating characteristics,
commonly used to connect a non-rigid barrier system to a rigid barrier system, such as a bridge
safety barrier.
Lateral The lateral position of the vehicle after impact (Figure A.3).
redirection
Leading point of In relation to a roadside hazard, the first point at which the barrier is needed to prevent an errant
need vehicle from striking the hazard (Figure A.4 and Figure A.5).
Leading The terminal treatment at the end of a barrier that faces vehicles approaching in the adjacent
terminal traffic lane (Figure A.4 and Figure A.5).
Length of need The length of a barrier system, needed to prevent errant vehicles from impacting a hazard,
representing the length over which a barrier will redirect an impacting vehicle. It is the distance
between the leading and trailing points of need (Figure A.4 and Figure A.5).
Median A barrier system when it is installed in a median location. Can be impacted from both sides.
application
Nearside The side of a vehicle closest to the kerb on the left-hand side of the road when the vehicle is
travelling in the normal direction of travel.
Non-gating A barrier terminal that is designed to redirect or contain an impacting vehicle and absorb part of
terminal the energy of the impacting vehicle at any point along the terminal without allowing it to pass
through the device.
Non-redirective A crash cushion designed to contain and capture an impacting vehicle.
crash cushion
Non-rigid barrier A barrier system where elements are designed to move substantially in a crash, and where energy
system is absorbed by movement of the barrier system and deformation of the vehicle.
Offside The side of a vehicle furthest away from the kerb on the left side of the road when the vehicle is
travelling in the normal direction of travel (i.e. it corresponds to the driver side of the vehicle).

Term Explanation
Permanent The permanent deformation of the barrier that remains after impact (Figure A.3).
deformation
Permanent A barrier that is installed permanently at the roadside.
barrier
Point of impact The point where the vehicle first impacts a barrier (Figure A.3).
Point of need The start or end of the length of need, defining the length over which an errant vehicle needs to be
redirected by the barrier and would otherwise strike the hazard if a barrier was not provided.
Post-impact The speed of the vehicle following impact (Figure A.3).
speed
Pre-impact The speed of the vehicle before a change of direction (Figure A.3).
speed
Proprietary A barrier system that is the subject of patent or other intellectual property rights within Australia
system and New Zealand.
Public domain A barrier system that is not the subject of patent or other intellectual property rights within Australia
system and New Zealand.
Redirective A crash cushion designed to contain and redirect an impacting vehicle.
crash cushion
Rigid barrier A barrier where there is no observable dynamic deflection during a vehicle impact. The
system deformation is contained within the impacting vehicle.
Risk Score The product of exposure, likelihood and crash severity (defined by the Trauma Index).
Road safety A roadside device that provides a physical restriction to penetration of a vehicle in a way that
barrier system reduces the risk to vehicle occupants and other traffic. Its purpose is to redirect or contain an
errant vehicle. It is used to shield roadside obstacles or non-traversable terrain features.
Occasionally, it may be used to protect people from vehicular traffic.
Secondary The angle at which the vehicle impacts the barrier for the second time (Figure A.3).
impact angle
Semi rigid A barrier where the barrier elements under an impact, manage and absorb kinetic energy by
barrier limited movement, yielding deformation and deflection.
Single-sided A barrier designed for impact on one side only.
barrier
System width The front-to-back dimension of the barrier including its supporting posts, etc. (Figure A.3). This
dimension should be less than the working width so that the system will not impact the hazard.
Temporary A barrier that is readily removable and used at roadworks, emergencies or similar situations.
barrier
Terminal A device designed to treat the end of a barrier. The terminal may function by decelerating a vehicle
to a safe stop within a relatively short distance, or permit controlled penetration of the vehicle
behind the device, or contain and redirect the vehicle, or a combination of these performance
characteristics.
Test level (TL) A set of conditions, defined in terms of vehicular type and mass, vehicular impact speed and
vehicular impact angle that quantifies the impact severity of a matrix of tests.
Thrie-beam A triple-corrugated steel-rail barrier supported on steel posts.
Trailing point of In relation to a roadside hazard, the last point at which the barrier is needed to prevent an errant
need vehicle from striking the hazard (Figure A.4 and Figure A.5).
Trailing terminal The terminal treatment at the departure end of a barrier in the direction of travel in the adjacent
traffic lane (Figure A.4 and Figure A.5).
Transition The connection of two barriers of different designs and/or performances.
Vehicle roll The lateral distance between the deflected face of a barrier and the maximum extent of vehicle
allowance body roll during impact.
W-beam A double-corrugated steel rail barrier supported on steel posts.
Wire rope A barrier system consisting of wire rope cables under high tension that are supported on posts and
barrier anchored at the ends.

Term Explanation
Working width The minimum width that is required to prevent an impacting design vehicle from colliding with an
object behind a barrier system. This includes both the dynamic deflection of the barrier (if any) and
the extra width to allow for the roll (vertical rotation) of an impacting vehicle. This ensures that the
system width can be accommodated between the deformed barrier and the hazard during impact
(Figure A.3) and that the top of a high heavy vehicle will not impact a high hazard during impact.
The terminology associated with barrier curvature is shown in Figure A.2.
Figure A.2: Road safety barrier terminology – curvature
The terminology associated with barrier impact is shown in Figure A.3.
Figure A.3: Road safety barrier terminology – impact
For detail see inset

Point of Hazard Safey barrier
impact Possible secondary
impact angle
Lateral re -direction
Initial lateral Departure angle
Shoulder Impact angle
position
Edge line
Errant
Exit
vehicle Lane Errant trajectory
vehicle
Centre line
Pre impact Post impact

speed (S1) Lane speed (S2)
Edge line
Inset: Barrier dimension details L Working width

Dynamic
Hazard W deflection
System
width
Initial position
Working
of barrier
Hazard
width
Dynamic
Permanent
deflection
deformation
Safety barrier
The terminology associated with barrier placement is shown in Figure A.4 and Figure A.5.

Figure A.4: Road safety barrier terminology – placement on two-lane/two-way road
Figure A.5: Road safety barrier terminology – placement on multilane/one-way road

Appendix B Detailed Risk Evaluation Procedure
This Appendix describes a process to analyse the level of risk associated with roadside hazards along a road
segment with a defined road and roadside cross-section and traffic volume. The process in Section 1.9 is
based on the procedure described in this Appendix.
The design procedure uses the principles developed in Austroads (2016c) to evaluate a roadside based on
the geometry of the road and the roadside, the traffic characteristics and the presence of hazards. The
process involves an evaluation of:
• Exposure: The frequency of vehicles leaving the traffic lanes.
• Likelihood: The proportion of drivers that have left the traffic lanes and then collide with a roadside
hazard.
• Severity: The expected severity of a collision, and reported as the function of fatal and serious injury
crashes (FSI crashes). The severity is defined by Trauma Indices.
The process is used to evaluate the risk of roadsides and medians of rural roads and highways with divided
carriageways, the left-hand roadsides on single carriageway rural highways, and the potential for a barrier in
the centre of a two-lane road to reduce head-on crashes. The process is used both to quantify the Network
Roadside Risk Intervention Threshold (NRRIT) (see Section 2) and to evaluate the risk of short lengths of
roadside (see Section 3).
The process is based on the US procedure Roadside Safety Analysis Program version 3 (RSAPv3) and
published in Ray et al. (2012a). RSAPv3 is a computer routine and the background to the algorithms is
documented in an engineering manual (Ray et al. 2012b). More information about the background to the
process presented in this Appendix is given in Austroads (2020b).
The process can also be applied to urban roads. However, the ability to treat hazards on urban roads is often
restricted by the width of the road reserve.
This Appendix describes the principles used to evaluate the risk of a roadside.
B.1 Exposure
Exposure is an indication of the number of vehicles that would be expected to leave the traffic lanes and
move onto the shoulder. In past procedures, the exposure has been called the encroachment frequency. The
units used are encroachments/km/year. Exposure uses attributes and dimensions that are common to longer
lengths of a road link.
Exposure is a function of geometry of the road, the terrain, the AADT and the operating speed of the road as
defined by the design speed. The evaluation of the exposure is based on the concepts in RSAPv3 and
documented in Ray et al. (2012b). RSAPv3, as did previous models of the encroachment rate, predicted a
reduction in the rate for increasing traffic volumes in the mid-range of AADTs. This aspect has been
eliminated from the process presented here and constant exposure is predicted for mid-range AADTs.
Figure B.1 and Figure B.2 are plots of the base exposure rate, measured in encroachments per kilometre per
year against AADT. For higher AADTs there is no data to predict the exposure and Ray et al. (2012b)
recommend a linear exposure against AADT as illustrated by the dotted line in Figure B.1 and Figure B.2.
The data for encroachment rates were recorded in the USA and Canada, but re-analysed by Ray et
al. (2012b).
Note that the base exposure rate for divided roads gives the number of encroachments per km per year on
the verge and the median for each carriageway based on the carriageway AADT, and not the total AADT.
Both verge and median exposures are the same and are presented in Figure B.2.

Figure B.1: Base exposure (enc/km/year) for each side of an undivided road
1.5
1.2
Exposure (enc/km/year)
0.9
0.6
0.3
0
0 4,000 8,000 12,000 16,000 20,000
AADT (veh/day)
Source: Adapted by Austroads (2020b) from processes in Ray et al. (2012b).
Figure B.2: Base exposure (enc/km/year) for each side of each carriageway of a divided road
2.1
1.8
Exposure (enc/km/yr)
1.5
1.2
0.9
0.6
0.3
0
0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000
Carriageway AADT (veh/day)
Source: Adapted by Austroads (2020b) from processes in Ray et al. (2012b).
B.1.1 Determine the Future Traffic Flow
The current traffic volume of the road can be determined from traffic survey counts. The traffic volume used
in the analysis is the carriageway volume. For example, for a two-lane two-way road, the traffic volume on
the road is the carriageway volume. For a divided facility, the volume is proportioned by the directional split;
usually a 50/50 split is used.

To allow for growth in traffic flow, the future traffic flow is calculated using Equation A1.
g n
Qn =AADT �1+ � A1
100
where
Qn = traffic flow on roadway in ‘n’ number of years
AADT = annual average daily traffic on the road (both directions) in current year
g = annual percent growth in traffic
n = number of years to project into the future.
The base exposure rate is adjusted using factors from RSAPv3 for the number of lanes, the average lane
width, the terrain, the grade and curve radius and whether the road is divided, undivided or in urban areas.
These factors are listed in Table B.1, Table B.2, and Table B.3 and in Figure B.3 and Figure B.4.
The factors for the average lane width are from the US Highway Safety Manual (AASHTO 2014). The user
should use the value for the width that is slightly lower than the field measurement if necessary. For instance,
if the average lane width is 3.4 m use the value for 3.3 m.
The procedure predicts slightly lower exposure rates for roads in mountainous than in rolling areas. This is
expected to be a result of greater driver awareness of the risk. The grade and the curve correction factors
are the same as those used in the previous version of Part 6 (Austroads 2010) and are also included in the
RSAPv3 procedure.
Table B.3 refers to undivided urban roads with an operating speed of 90 km/h. For urban roads with a higher
operating speed or with divided carriageways, Table B.2 should be used.
Table B.1: Exposure adjustment factors for undivided roads
Attribute Condition Factor
Number of lanes in the direction of travel 1 lane 1.00

More than 1 lane 0.76
Lane width 2.7 m and narrower 1.50
3.0 m 1.30
3.3 m 1.05
3.5 m and wider 1.00
Terrain Flat  1.00
Rolling 2.58
Mountainous 2.27
Source: Adapted by Austroads (2020b) from data in Ray et al. (2012b).

Table B.2: Exposure adjustment factors for divided and one-way roads
Number of lanes in the direction of travel 2 lanes 1.00

More than 2 lanes 0.91
3.0 m 1.15
3.3 m 1.03
Rolling 1.66
Mountainous 1.51
Table B.3: Exposure adjustment factors for urban undivided roads with operating speeds less than or equal
to 90 km/h

3.0 m 1.85
3.3 m 1.49
Rolling 2.58
Figure B.3: Gradient correction factor
Source: Ray et al. (2012b) and TMR (2016).

Figure B.4: Curve correction factor
Source: Ray et al. (2012b) and TMR (2016)
The overall exposure is then the base encroachment rate (from Figure B.1 for example), multiplied by factors
for the number of lanes, the lane width and the terrain, road gradient and curve radii.
B.2 Likelihood
The likelihood of colliding with a hazard in the verge after the vehicle has left the road depends on the
roadside cross-section and the speed of vehicles on the road. The likelihood of a collision with a hazard is
the probability of a vehicle travelling beyond the lateral distance from the road to the hazard. RSAPv3 does
not explicitly use a function to describe the likelihood of travelling beyond a lateral distance, but rather uses
890 trajectories established in the National Cooperative Highway Research Program (NCHRP) project 17-22
Crash Reconstruction Database (Albuquerque, Sicking & Stolle 2010) to predict impacts. The procedure,
presented here, uses a function that describes a likelihood of travel beyond a lateral distance from the edge
of the travel lanes (Figure B.5). This lateral distance to a hazard includes the road shoulder, whether paved
or not.
Figure B.5: Lateral distance measures
The likelihood model is based on the lateral extent model documented in the AASHTO Roadside Design
Guide Edition 4 (AASHTO 2011a), which includes a relationship for multi-lane divided roads and for two-lane
undivided roads. A re-evaluation of data from Jurewicz and Pyta (2011) together with encroachment data
from RSAPv3 led to the conclusion that the multi-lane curve can be used for all road types. (This curve is
also more conservative for two-lane roads in that it predicts that drivers travel further from the road.)

The likelihood of a vehicle colliding with a hazard uses the model documented in the AASHTO Roadside
Design Guide Edition 4 for roads with an operating speed of 110 km/h. For roads with a lower operating
speed (and typically a lower design speed) the expected distance travelled by drivers is related to the square
of the operating speed.
On shorter radii corners, the lateral distance travelled by a vehicle increases. AASHTO (2011a) allowed for
this effect by increasing the clear zone requirement by a factor based on the operating speed and the radius
of the curve. This effect has been included in the likelihood function shown in Figure B.6, Figure B.7 and
Figure B.8 for roads with different operating speeds.
Likelihood curves are given for only three operating speeds because the level of science and data supporting
these relationships is limited. As more comprehensive research data becomes available, there will be more
confidence in extending the data. The designer would use the more conservative graph. For instance, if the
operating speed is 100 km/h, then use the 110 km/h operating speed instead of interpolating between
graphs. On the other hand, if the radii are between the values in a single figure, then the user should
interpolate. The process is based on the best available information and a high level of accuracy is not
implied.
Figure B.6: Likelihood of reaching a lateral distance for a road with an operating speed of 110 km/h and with
curves of different radii
100
Likelihood of reaching lateral distance
90
80 Curve radius
70 Straight
60 1200 m
700 m
(%)
50
≤450 m
40
30
20
10
0
0 5 10 15 20 25 30
Lateral distance (m)
Source: Adapted from Austroads (2020b).

Figure B.7: Likelihood of reaching a lateral distance for road with a 90 km/h operating speed and with curves
of different radii
100
90
80
70
Curve radius
60
Straight
(%)
50 1200 m
40 450 m
30 ≤300 m
20
10
0
0 4 8 12 16 20
Figure B.8: Likelihood of reaching a lateral distance for road with a 70 km/h operating speed and with curves
of different radii
100
90
80
70 Curve radius
60 Straight
(%)
50 900 m
40 300 m
30 ≤150 m
20
10
0
0 4 8 12 16
For rural roads, the likelihood of colliding with an oncoming vehicle once the errant vehicle has crossed the
lane line is a function of the opposing traffic flow which has been assumed to be half the total traffic flow
(AADT). Refer to Figure B.9. At this stage a similar graph for urban roads is not available.

Figure B.9: Likelihood of a vehicle crossing a lane line and colliding with an oncoming vehicle on an undivided
rural road with a speed limit of 100 km/h
10
9
8
7
Likelihood (%)
6
5
4
3
2
1
0
0 2000 4000 6000 8000 10000 12000
Bi-directional AADT
As described in Section 1.9.4, the assessment process involves measuring the effective lateral distance from
the road to a hazard. If there is an embankment slope 6:1 or flatter, then the width of the embankment is
included in the measure. If the embankment slope is steeper than 6:1, but not steeper than 4:1, the lateral
distance is the distance to the hazard excluding the width of the embankment. Embankment slopes of 3:1
and steeper are considered to be a hazard in their own right and the area beyond the embankment is not
included in the lateral distance measure. See Figure B.10 which is the same as Figure 1.3. Drivers of both
cars and trucks are able to regain control on embankments that are 6:1 or flatter.
In practice, a number of hazards are often in close proximity to each other and the designer should assess
whether they can be considered collectively as one large (or longer) hazard, given that they have similar
Trauma Indices. Otherwise treat the hazards as being independent with each having a different Trauma
Index.
The ground surface to the hazard should be traversable by vehicles. Ground surfaces in the verge with
ditches, rock spalls, obstacles or steep slopes should be considered a hazard in themselves. They may
cause a vehicle’s undercarriage to become snagged with the result that drivers lose control. The propensity
for vehicles to roll over increases with the roughness of the ground and with the slope of the embankment.
Refer to Appendix B.4.1 for additional guidance.
Where there is an auxiliary lane adjacent to the through lane it is appropriate to consider the auxiliary lane
width as part of the lateral distance to the hazard for vehicles in the through lane. However, the available
lateral distance for drivers using the auxiliary lane should also be considered. The lateral distance to a
hazard for turning lanes should be determined by allowing for appropriate deceleration and a reduced speed,
or in the case of an acceleration lane by determining the likely speed adjacent to the hazard being
considered. A separate analysis should be done for through lanes and auxiliary lanes.

Figure B.10: Lateral distance to a hazard
Vehicles are considered errant when they leave the travelled path onto either the median or the verge. For
this reason, the lateral distance to hazards apply to both sides of the travelled path and each direction must
be independently derived according to the road conditions. The implications for multilane divided roads and
two-lane, two-way roads are illustrated in Figure B.5.
The product of the exposure and the likelihood of a vehicle travelling beyond the lateral distance to a hazard
gives the expected frequency of impacts per year per kilometre.
B.3 Likelihood of Colliding with a Point Hazard and a Length of Road Safety
Barrier
Mak et al. (2010) reported on the impact conditions associated with serious run-off-road crashes. They
reported the mean departure angles of between 17° and 18° for speeds between 70 and 115 km/h.
Burbridge (2019) has reviewed the characteristics of vehicle departure from the roadway. He found that
approximately one third of the crashes observed by Doecke and Woolley (2010) were drift-off-the-road
crashes. The mean departure angles for these vehicles was 7°. In the remaining two-thirds of the crashes,
vehicles driven off the road following a loss of control had a mean departure angle of 17°.
AASHTO (1989) describes a process for estimating collision frequency. This was incorporated into the 2010
edition of Part 6 of the Austroads Guide to Road Design (Austroads 2010) 1. As the vehicle is not tracking,
the swath width of the vehicle is assumed to be greater than the width of the vehicle and AASHTO (1989)
assumed a swath width of 3.6 m. Using a departure angle of 17° and a swath width of 3.6 m for the
drive-off-the-road crashes, the distance from A to D is 17 m. See Figure B.11. Vehicles driven off the road
departing the road between A and D are assumed to collide with the hazard.
1
This process is also incorporated in the Queensland Department of Transport and Main Roads Roadside Impact Severity Calculator
(RISC) program.

For vehicles that drift off the road, the swath width is assumed to be the width of a vehicle or 1.9 m. the
length between A and D is 25 m. Using the approximate ratio between crashes involving vehicles driving off
the road and yawing or off-tracking and vehicles drifting off the road, the average distance over which
vehicles leave the road and collide with a hazard is 19 m.
Consequently, the likelihood of an impact with a single tree is 0.019 times the likelihood for a continuous
hazard measured over 1 km.
If the hazard was larger with a length of L (measured parallel to the road) and with the same width then the
length of road over which an errant vehicle is assumed to collide with the hazard is 18 + L. These lengths
should be used in the analysis of the risk due to a point hazard. The likelihood of an impact with a longer
hazard is (18 + L)/1000 times the likelihood for a continuous hazard measured over 1 km.
The same process is used to establish the number of drivers who are expected to impact a barrier where X is
the length of the barrier installation, including terminals. The likelihood of an impact with a safety barrier is
(18 + X)/1000 times the likelihood for a continuous safety barrier measured over 1 km. Note that a barrier
extends beyond the hazard.
Figure B.11: Impact envelopes for a hazard 1 m by 1 m (not to scale)
The ’drive-off-the-road’ condition is on the left and the ‘drift off’ condition is on the right.
B.4 Severity
The likely severity of impacts with a hazard uses similar concepts to other approaches. In the previous
version of AGRD Part 6, the impact severity was related to the average crash costs for impacts with a hazard
type. Unfortunately, the basis for these severity indices has not been well documented and would appear to
have developed over time by ‘experience and judgement’. Ray et al. (2012b) re-evaluated the severity
indices based on reported analyses in the literature and used the Equivalent Fatal Crash Cost
Ratio (EFCCR) to define the severity of impacts with a hazard. The EFCCR is the average crash cost
associated with the impacts divided by the average cost of a fatal crash 2.
In the procedure described here, the EFCCR values used in RSAPv3 and their associated severity indices
were used to construct a Trauma Index that is linearly related to the expected number of fatal and serious
injures (FSIs) (Austroads 2020b).
2
Note this is based on the number of fatal crashes and not the number of fatalities.

The Trauma Index for a hazard provides a useful estimate of the likely number of FSI crashes. If one hazard
has twice the Trauma Index of another hazard, it is expected to have twice the number of FSI crashes from
impacts on average. The following tables and figures list the Trauma Indices for an operating speed of
110 km/h and are listed for isolated and continuous hazards along the length of the road. The Trauma
Indices for a lower operating speed are a function of these values.
Trauma Indices (and severity indices used in the past) are difficult to measure, as the data need to include all
crashes with an object and not just those that had a fatal or serious injury outcome. The Trauma Index is not
to be confused with the Severity Index in older publications or the expected FSI from the stereotype tables
(Austroads 2020b).
B.4.1 Consistent Background Hazards
The Trauma Indices for continuous hazards are listed in Table B.4 to Table B.7. These generally apply to the
background hazards (refer to Section 1.9.2).
Frequently there will be trees on embankments or at the bottom of the embankments. The trees will be then
the defining hazard. Kerbs with a vertical face over 100 mm high on roads with operating speeds of 80 km/h
or greater and a minor hazard should be given a Trauma Index of the ‘Generic rigid wall’ (0.77).
Other hazards, which do not have a Trauma Index, will need to be assigned one based on judgement. A rock
cutting can be used for significant continuous hazards in most situations.
Table B.5 and Table B.7 refer to relatively smooth batter slopes. Rocks and erosion gullies are hazards in
their own right and can be treated as a generic fixed object or a rock cutting. These tables account for
rollover on embankments. Refer to Appendix B.5 for a discussion of rollovers on flat terrain or on slopes
flatter than 6:1.
Some steep embankment slopes may not constitute a direct hazard but prevent errant vehicles from
recovering when they leave the road and run onto the embankment. A 6:1 embankment will allow a majority
of both car and truck drivers to regain control. The condition of the ground surface may increase the
hazardousness of an embankment. If the embankment surface has ruts deeper than 0.2 m or has large
stones protruding more than 0.2 m above the surface, the Trauma Index for these slopes should be doubled.
The propensity for vehicles to rollover is increased on steeper embankments.
Table B.4: Trauma Indices for continuous features – operating speed = 110 km/h
Hazard Trauma Index – 110 km/h

Generic rigid wall 0.77
Rock cutting 6.0
Tree lined edge 6.0
Rocks protruding between 200 mm and 300 mm above the ground surface 6.0
Ruts in the ground surface more than 300 mm deep 6.0*
Rocks protruding between 100 mm and 150 mm above the ground surface 3.0*
Ruts in the ground surface between 200 and 300 mm deep 3.0*
* Assumed values
Source: Austroads (2020b).
Fast-moving bodies of water are considered to be more hazardous than still water. In general, designers
should carefully consider the risk associated with bodies of water over 0.6 m deep, or watercourses with a
normal base flow depth greater than 0.6 m, as these could cause a stunned, trapped, or injured occupant to
drown. Refer to the Guide to Road Design Part 5B: Drainage: Open Channels, Culverts and Floodways
(Austroads 2013b) for more details.

The entries in Table B.5 to Table B.7 should not be interpolated but rather the larger Trauma Index should
be used. For instance, if the water depth is 3 m and the height is 1 m, use the values for a 4 m water depth
and a height of 2 m, giving a Trauma Index of 30.
Table B.5: Trauma Indices for embankment slopes on fills and in cuts – operating speed = 110 km/h
Cuts Fills
All Slope Height

heights
0 to 2 m 2 to 5 m 5 to 10 m 10 to 20 m More than 20 m
0.35 6:1 * 0.63 0.63 0.63 0.63 0.63
0.35 4:1 0.69 0.89 0.89 0.89 0.89
0.35 3:1 0.83 1.4 1.5 1.5 2.1
0.61 2:1 1.3 3.5 4.5 6.7 9.9
0.74 1.5:1 1.7 7.0 14 27 39
0.90 1:1
0.80 Vertical
Note: * For embankments flatter than 6:1, the rollover risk should be evaluated using Appendix B.5.
Table B.6: Trauma Indices for vertical drops and water depths – operating speed = 110 km/h
Water depth
Height
1 m and lower 2.0 m 4.0 m 6 m and higher
0.0 m 1.2 7.2 17 35
0.3 m 2.0 9.9 22 38
2.0 m 5.2 20 30 40
4.0 m 12 28 35 43
6.0 m 19 33 39 45
8.0 m 28 36 42 46
10 m and higher 45 48 51 52

Table B.7: Trauma Indices for parallel V drains with a height less than 1.2 m – operating speed = 110 km/h
Backslope
Foreslope
6:1 4:1 3:1 2:1
6:1 0.55 0.59 0.61 0.68
4:1 0.68 0.69 0.80
3:1 0.73 0.81 0.90
2:1 0.90 1.0
Notes:
 It is recognised that V drains are not a preferred cross-section. This table is included because these drains exist on
the network.
 Some values in this table are lower than similar values in Table B.5.
B.4.2 Isolated Hazards
The Trauma Indices for isolated (point) hazards are contained in Table B.8 and Table B.9.
Culverts can be considered to be a point hazard and the Trauma Index depends on their design. If the
headwall is proud of the embankment, then the culvert end treatment should be treated as a generic fixed
object unless there is a grate over the entry that allows vehicles to drive across it. Otherwise, values for
embankment slopes, from Table B.5, can be used.
Generic fixed object includes large planter boxes, protruding footings, non-traversable driveway headwalls
and non-traversable culvert headwalls.
It should be noted that while trees less than 75 mm in diameter are not considered to be hazards, they can
grow to become hazards in the future. Designers should consider the final trunk diameter.
Table B.8: Trauma Indices for point hazards – operating speed = 110 km/h
Hazard Trauma Index – 110 km/h

Tree with a trunk of 75 to 100 mm in diameter 0.53
Signs breakaway 0.75
Small wooden sign 0.75
Slip base or energy absorbing poles 1.9
Tree with a trunk of 150 to 200 mm in diameter 3.1
Generic fixed object 6.0
Utility pole 6.4
Tree with a trunk 250 mm and larger in diameter 6.7
Traffic signal pole 8.6
Bridge pier column 44

Other hazards, which do not have a Trauma Index, will need to be assigned a Trauma Index based on
judgement. As a guide, mailboxes, signs on inappropriate slip-bases, hydrant bases more than 100 mm high,
could be conservatively assumed to have a Trauma Index equivalent to a tree with a trunk diameter of 150 to
250 mm.
Table B.9: Trauma Indices for slopes parallel to the road – operating speed = 100 km/h
Upslope Downslope height

All heights Slope
0 to 2 m 2 to 5 m 5 to 10 m More than 10 m
0.73 6:1 and flatter 1.0 1.9 2.7
1.09 4:1 1.2 3.7 11 14
1.6 3:1 1.2 4.4 17 25
3.3 2:1 1.2 4.5 21 38
5.0 1.5:1 1.2 4.5 21 42
6.5 1:1
B.4.3 Roadside Barriers
The Trauma Index for safety barriers is 0.43 to 0.84 for operating speeds of 110 km/h. A range is given here
as the guidance for the final selection of a barrier is in Section 5. For the risk evaluation, a Trauma Index of
0.84 should be used.
B.4.4 Oncoming Vehicles
The Trauma Index for oncoming vehicles is 87 for rural roads with an operating speed of 110 km/h. This
ensures that the number of fatal and serious injuries from head-on crashes under a generalised
cross-section and the number of fatal and serious injuries from vehicles running off the road are
approximately the same at an AADT of 5700 veh/day as demonstrated in NZ Transport Agency (2011).
(Refer to Austroads 2020b).
B.4.5 Trauma Indices for Roadside Features on Roads with Other Operating Speeds
Trauma Indices for roadside features on roads with different operating speeds are evaluated by multiplying
the Trauma Index for roads with an operating speed of 110 km/h by a factor. The factor for roads with an
operating speed of 90 km/h is 0.548, for roads with an operating speed of 80 km/h it is 0.385 for roads and
with an operating speed of 70 km/h or less it is 0.258. These factors are based on the third power law.
For example, if the Trauma Index is 5.0 for an operating speed of 110 km/h, then the Trauma Index would be
5*0.548 or 2.7 on roadsides were the operating speed is 90 km/h.

B.5 Evaluation of Risk Associated with Vehicle Rollovers
Rollovers on relatively flat terrain had previously been considered to be not hazardous. Jurewicz and
Ptya (2011) have identified that a considerable number of serious and fatal crashes involved the vehicle
rolling over. Austroads (2020b) developed a method for predicting the likelihood and severity of rollover
events. This later work considered the Trauma Index of a rollover to be equal to the Trauma Index of
vehicles traversing a 6:1 embankment, which is 0.63. The severity of a rollover is not changed with operating
speed.
The likelihood is proportional to the width of the area alongside the road that has a slope flatter than 6:1. The
average probability of a rollover was reported in Ray et al. (2012b) and was applied at the average distance
a vehicle would travel from the traffic lane in flat terrain. This is explained by the curves in Figure B.6 to
Figure B.8. The likelihood is also affected by the square of the operating speed of the rural road. The
equation for likelihood is given by Equation A2.
Likelihood = 0.592 fS X
A2
where
Likelihood = Likelihood of a rollover on relatively flat terrain (per cent)
The lateral width of the flat area beside the traffic lane. This width includes the
X =
shoulder and an embankment or area flatter than 6:1 (m)
Operating speed factor listed in
fS =
Table B.10
Table B.10: Speed factors for the likelihood of a rollover
Operating speed Factor
70 km/h 0.40
90 km/h 0.67
110 km/h 1.00
While the ground condition has been included as a factor in Table B.4, as it is a safety issue if a vehicle
impacts large rocks, a rutted ground also poses additional risk of rolling over. This has not been accounted
for in this evaluation process as there is no data available to quantify it.
The likelihood function is only applied to rural roads and highways.
B.6 Risk Score
Risk is a function of the frequency of an event occurring and the expected outcomes. In this analysis, the
Risk Score is the product of the exposure, the likelihood and the severity (described by the Trauma Index).
Exposure, likelihood and severity are evaluated using the processes described in Appendix B.1, B.2 and B.3.
The exposure is the number of encroachments per kilometre per year; the likelihood is a probability of a
collision with a hazard given an encroachment. Exposure times likelihood gives an indication of the expected
number of collisions with a hazard per year. The Trauma Index is linearly related to the number of fatal and
serious injury crashes resulting from impacts with a hazard. The Risk Score is then linearly related to the
number of FSI crashes per kilometre per year. This is consistent with other methods used to review the
safety of roadsides.

If a constant FSI crash rate per million vehicle kilometres of travel (FSI/MVKT) is required, the Risk Score
should be modified by the AADT.
B.6.1 Collective Risk
The Risk Score is the collective risk and represents the total risk of run-off-the-road crashes to the broader
community. It represents the community risk for the roadside configuration over one kilometre. The
procedure to estimate the risk for an isolated tree or short length of road safety barrier is described in
Appendix B.6.3.
The Risk Score for rollover events uses the exposure (from Appendix B.1) and the severity and exposure
from Appendix B.5. The Risk Score for rollovers is added to the Risk Score for background hazards, isolated
hazards and rollovers to give a combined Risk Score.
The risk evaluation process has been based on the US procedure ‘Roadside Safety Analysis Program’
Version 3 (RSAPv3) (Ray et al. 2012a, 2012b). This process is based on the best information available, yet
there are many aspects that are not well defined. Accordingly, the reader should always treat the Risk
Scores as useful estimates and not differentiate between roadsides with similar scores.
The stereotype tables in Austroads (2020b) also provide an estimate of the expected number of fatal and
serious injuries based on the Australian National Risk Assessment Model (ANRAM) model
(Austroads 2014b). ANRAM provides a broad evaluation of the level of safety for a network within a similar
framework and algorithms to those used in the US Highway Safety Manual (HSM) (AASHTO 2014). The
stereotype tables provide a means to establish broad design parameters for budgeting and planning. These
are appropriate tools to establish a corridor safety vision as part of a Network Safety Plan but are insufficient
to provide a detailed roadside design.
B.6.2 Individual Risk
The individual risk is the relative risk to a road user who runs off the road. It is independent of the traffic on
the road and can be evaluated assuming that the base exposure rate is 1.00 and then applying the relevant
factors from Table B.1 to Table B.3 and Figure B.3 and Figure B.4.
B.6.3 Risk of Short Lengths of Roadside
At times the risk of a short length of roadside will need to be estimated. Examples are:
• a short length of safety barrier
• an isolated hazard when the background hazards alone have a Risk Score lower than the NRRIT, but the
isolated hazards warrant shielding.
The risk is the collective risk calculated using the procedure outlined above multiplied by the length of
roadside edge which errant vehicles cross. For single longer isolated hazards or short lengths of barrier that
extend along the roadside for a distance L, this distance is L + 18 where L is in metres. This concept is
illustrated in Figure B.11 and Appendix B.3. The Risk Score for this short length is then the Risk Score over a
kilometre multiplied by the factor (L + 18)/1000.
The appropriate length of barrier required to shield a hazard is given in Section 5.3.18.

B.7 Limitations of the Risk Assessment Process
The evaluation of a roadside for a number of years has revolved around a risk assessment approach
essentially trying to predict the number of collisions and the likely outcomes. Previous editions of this Part
have used these techniques, which have been based on the various editions of the US Roadside Safety
Analysis Program. These models have all considered three aspects of a collision; namely vehicle leaves the
roadway, impacts a hazard, with various outcomes. The data on which these models are based, are limited.
However, the models do provide estimates of the consequences of engineering design changes on the likely
crash rates and the models have been subjected to limited validation (Ray et al. 2012b).
With the development of the Highway Safety Manual (HSM) (AASHTO 2014) there is now considerable
research effort to develop greater alignment between the risk assessment models and the HSM crash
modification factors and functions. This research will improve the risk prediction models. Refer to Carrigan
and Ray (2017).
Although this process is based principally on US research, the process does not consider that providing a
clear zone is first choice treatment; it is one of many treatments that may be considered.
B.8 Worksheets
Worksheets are provided in Table B.11, Table B.12 and Table B.13.
These worksheets are provided on separate pages so that they may be copied and used in the field or
drawing office.
Table B.11: Risk Score calculation worksheet for verges on divided and undivided roads
Group Measure Item Reference Result

General road Terrain (flat, rolling, mountainous) A Plans
characteristics
Environment (urban, rural) B Plans
Operating speed (70 km/h, 80 km/h, 90 km/h, C Plans

110 km/h)
Divided road? (yes/no) D Plans
Number of lanes in the direction of travel E Plans
Lane width (m) F Plans
Design AADT (veh/day) G Plans
Length of road segment (m) H Plans
Particular road Chainage where roadside is assessed I Plans

characteristics
Side of travelled way (verge/left or median/right) J Plans Verge
Grade (%) (upgrade positive) K Plans
Curve radius (m) (Use 2000 m if straight) L Plans
Curve direction (left/right in the direction of M Plans

travel)
Background Description N Plans
hazard
characteristics
Lateral distance to the background hazards from O Plans

the edge of lane (m) (Refer to Figure B.10)


Isolated hazard Description P Plans
characteristics
Length of the hazard (m) Q Plans
Average distance between isolated hazards (m) R Plans
Lateral distance to the isolated hazards from the S Plans

edge of lane (m) (Refer to Figure B.10)
Exposure Base exposure (encr/km/yr) T Figure B.1 or
Figure B.2
Number of lanes factor U Table B.1 or
Table B.2
Lane width factor V Table B.1, Table B.2
or Table B.3
Terrain factor W Table B.1, Table B.2
or Table B.3
Grade factor X Figure B.3
Curve correction factor Y Figure B.4
Exposure (encr/km/yr) Z Product T to Y
Likelihood Likelihood for background hazards (considering AA Figure B.6,

curve radius) Figure B.7 or
Figure B.8
Modification factor for likelihood if the isolated AB 1-(Q+18)/MIN(R,H)
hazards shield the background hazards (Use
1.00 if shielding does not exist.)
Modified likelihood for background hazards AC AA*AB
Likelihood of impacts with isolated (point) AD Figure B.6,

hazards (based on being continuous) Figure B.7 or
Figure B.8
Likelihood of impacts with isolated hazards AE AD*(Q+18)/MIN(R,H)
Likelihood of a rollover AF Equation A2
Severity Trauma Index for background hazards AH Table B.4 to

(Trauma Index) (operating speed of 110 km/h) Table B.7
Trauma Index for background hazards at the AJ Appendix B.4.5
appropriate operating speed
Trauma Index for isolated hazards (operating AK Table B.8 and
speed of 110 km/h) Table B.9
Trauma Index for isolated hazards at the AL Appendix B.4.5
Trauma Index for a rollover AM 0.63
Risk Score Calculate Risk Score (product of the exposure, AP Z*AC*AJ

likelihood and Trauma Index) for background
hazards
Calculate Risk Score (product of the exposure, AQ Z*AE*AL
likelihood and Trauma Index) for isolated
hazards
Calculate Risk Score (product of the exposure, AR Z*AF*AM
likelihood and Trauma Index) for rollover events
Combined Risk Score AS AP+AQ+AR

Table B.12: Risk Score calculation worksheet for medians on divided roads

characteristics
Environment (urban, rural) B Plans

110 km/h)
Divided road? (yes/no) D Plans Yes
Length of road segment (m) H Plans

characteristics
Side of travelled way (verge/left or median/right) J Plans Median

travel)
Background Description N Plans
hazard
characteristics
Width of median (m) (Refer to Figure B.10) O Plans
Isolated hazard Description P Plans

characteristics
Length of the hazard (m) Q Plans
Average distance between isolated hazards (m) R Plans
Lateral distance to the isolated hazards from the S Plans

Exposure Base exposure (encr/km/yr) T Figure B.1 or
Figure B.2
Number of lanes factor U Table B.1 or
Table B.2
Lane width factor V Table B.1, Table B.2
or Table B.3
Terrain factor W Table B.1, Table B.2
or Table B.3


Likelihood Likelihood for background hazards (considering AA Figure B.6,
curve radius) Figure B.7 or
Figure B.8
Modification factor for likelihood if the isolated AB 1-(Q+18)/MIN(R,H)
Modified likelihood for background hazards AC AA*AB
Likelihood of impacts with isolated (point) AD Figure B.6,

hazards (based on being continuous) Figure B.7 or
Figure B.8
Likelihood of impacts with isolated hazards AE AD*(Q+18)/MIN(R,H)
Likelihood of a rollover AF Equation A2
Trauma Index Trauma Index for background hazards AH Table B.4 to

(operating speed of 110 km/h) Table B.7
Trauma Index for background hazards at the AJ Appendix B.4.5
Trauma Index for isolated hazards (operating AK Table B.8 and
Trauma Index for isolated hazards at the AL Appendix B.4.5
Trauma Index for a rollover AM 0.63
Risk Score Calculate Risk Score (product of the exposure, AP Z*AC*AJ

hazards
Calculate Risk Score (product of the exposure, AQ Z*AE*AL
hazards
Calculate Risk Score (product of the exposure, AR Z*AF*AM
Combined Risk Score AS AP+AQ+AR

Table B.13: Risk Score calculation worksheet for oncoming vehicles on undivided rural roads

characteristics
Environment (urban, rural) B Plans Rural

110 km/h)
Divided road? (yes/no) D Plans No
Length of road segment (m) H Plans 1000

characteristics
Side of travelled way (verge/left or median/right) J Plans Median

travel)
Background Description: Oncoming vehicles N Plans
hazard
characteristics Distance between oncoming vehicle lanes (m) O Plans
Exposure Base exposure (encr/km/yr) T Figure B.1
Number of lanes factor U Table B.1
Lane width factor V Table B.1

Terrain factor W Table B.1
Likelihood Likelihood of impacts with oncoming vehicles AG Figure B.9
Trauma Index Trauma Index for an operating speed of AN Appendix B.4.4 87

110 km/h
Trauma Index at the appropriate operating AO Appendix B.4.5
speed
Risk Score Calculate Risk Score (product of the exposure, AT Z*AG*AO
likelihood and Trauma Index)
B.9 Example 1 – Culvert Headwall
A two-lane rural road in rolling terrain has two 3.5 m lanes, an operating speed of 110 km/h and a design
AADT of 4000 veh/day. The cross-section of the roadside is shown in Figure B.12. There is a culvert
headwall 3 m long (in the direction of traffic) 6.4 m from the traffic lanes. There is a 4:1 embankment that is
9 m wide. In the vicinity of the culvert there is a 700 m radius horizontal curve to the right and a downgrade
of 6%. This puts the culvert on the outside of the curve.

There is a line of trees at the toe of the embankment and these can be considered to be a background
continuous hazard with the culvert headwall as an isolated hazard. The embankment could be considered to
be a hazard and the additional Risk Score for the culvert head wall added to the Risk Score for the
embankment. In addition, because the 4:1 embankment is reasonably steep and it is considered that many
drivers would not ‘recover’ on this embankment. Consequently, it is assumed that drivers who travel over the
embankment will collide with the trees beyond.
This is a similar example to the one discussed in Section 3.4.2. Here the terrain is rolling and the operating
speed is 110 km/h.
Figure B.12: Example 1 – culvert headwall on outside of horizontal curve
The calculations are presented in Table B.14 for the case where the trees are the background hazard and
the culvert is the isolated hazard. In this case the isolated hazard shields the background hazard.
Table B.14: Risk Score calculation for the background hazards being the trees at the toe of the embankment

General road Terrain (flat, rolling, mountainous) A Plans Rolling
characteristics
Operating speed (70 km/h, 80 km/h, 90 km/h, C Plans 110 km/h

110 km/h)
Number of lanes in the direction of travel E Plans 1
Lane width (m) F Plans 3.5
Design AADT (veh/day) G Plans 4000
Length of road segment (m) H Plans 1000
Particular road Chainage where roadside is assessed I Plans –

characteristics
Side of travelled way (verge/left or median/right) J Plans Verge
Grade (%) (upgrade positive) K Plans –6%
Curve radius (m) (Use 2000 m if straight) L Plans 700
Curve direction (left/right in the direction of M Plans Right

travel)
Background Description: Trees at the toe of the embankment N Plans
hazard
characteristics Lateral distance to the background hazards from O Plans 2.4


Isolated hazard Description: Drainage culvert headwall P Plans
characteristics
Length of the hazard (m) Q Plans 3
Average distance between isolated hazards (m) R Plans 1000
Lateral distance to the isolated hazards from the S Plans 2.4

Exposure Base exposure (encr/km/yr) T Figure B.1 0.82
Number of lanes factor U Table B.1 1.0
Lane width factor V Table B.1 1.0

Terrain factor W Table B.1 2.58
Grade factor X Figure B.3 2.0
Curve correction factor Y Figure B.4 1.0
Exposure (encr/km/yr) Z Product T to Y 4.22
Likelihood Likelihood for background hazards (considering AA Figure B.6 0.74

curve radius)
Modification factor for likelihood if the isolated AB 1-(Q+18)/MIN(R,H) 0.98
Modified likelihood for background hazards AC AA*AB 0.72
Likelihood of impacts with isolated (point) AD Figure B.6 0.74

hazards (based on being continuous)
Likelihood of impacts with isolated hazards AE AD*(Q+18)/MIN(R,H) 0.02
Likelihood of a rollover AF Equation A2 0.014
Severity Trauma Index for background hazards AH Table B.4 6.0

(Trauma Index) (operating speed of 110 km/h)
Trauma Index for background hazards at the AJ Appendix B.4.5 6.0
Trauma Index for isolated hazards (operating AK Table B.8 and 6.0
Trauma Index for isolated hazards at the AL Appendix B.4.5 6.0
Trauma Index for a rollover AM 0.63 0.63
Risk Score Calculate Risk Score (product of the exposure, AP Z*AC*AJ 18.3
hazards
Calculate Risk Score (product of the exposure, AQ Z*AE*AL 0.4
hazards
Calculate Risk Score (product of the exposure, AR Z*AF*AM 0.04
Combined Risk Score AS AP+AQ+AR 18.7
An alternative way to look at this issue is to consider the embankment as a background hazard and the
culvert as an isolated hazard. In this case the culvert does not shield the background (or continuous) hazard.
The calculations for this case are presented in Table B.15. Essentially this assumes that there are no trees at
the toes of the embankment,

Table B.15: Risk Score calculation for the background hazard being the embankment

General road See Table B.14
characteristics
Particular road See Table B.14
characteristics
Background Description: 4:1 embankment N Plans
hazard
Isolated hazard Description: culvert wall P Plans
characteristics

Exposure Exposure (encr/km/yr) (From Table B.14) Z Product T to Y 4.22

curve radius)
Likelihood of impacts with isolated (point) AD Figure B.6 0.74


Severity Trauma Index for isolated hazards (operating AK Table B.8 6.0
(Trauma Index) speed of 110 km/h)
hazards
hazards
The governing Risk Score is from Table B.14 as this produces the higher combined Risk Score. In both
cases the risk of the background hazard dominates and it may be sufficient to demonstrate that the risk is
above the NRRIT.

Suppose a barrier is used to shield the culvert and the trees at the toe of the embankment (Figure B.12). The
barrier is assumed to be 130 m long including terminals for this example. The details of the barrier would
need to be developed through the discussion in Section 5.3.18. Table B.16 describes the Risk Score
evaluation. Here the barrier is the isolated hazard, with a Trauma index of 0.84 (see Appendix B.4.3).
Table B.16: Risk Score calculation for the background hazards being the trees at the toe of the embankment
and shielded with a barrier

General road See Table B.14
characteristics
Particular road See Table B.14
characteristics
Background Description: Trees at the toe of the embankment N Plans
hazard
Isolated hazard Description: safety barrier P Plans
characteristics

Exposure Exposure (encr/km/yr) (From Table B.14) Z Product T to Y 4.22

curve radius)
Likelihood Likelihood of impacts with isolated (point) AD Figure B.6 0.83


Trauma Index for isolated hazards (operating AK Appendix B.4.3 0.84
speed of 110 km/h)
hazards
hazards

The risk has been reduced from 18.7 to 16.4 (for a 1 km of road). The road still has a high risk as the
majority of the road has not been treated. The barrier is only 130 m long. There may be other areas that can
be treated with short lengths of barrier, but it is unlikely that they will be effective in reducing the Risk Score
to a low level.
The analysis in Table B.14 and Table B.16 was for a 1 km segment of road. If an analysis was to be
undertaken on a 200 m length of road, then the additional risk of the culvert end wall is more significant.
Table B.17 lists the results for an unshielded culvert and a shielded culvert based on a 200 m length of road.
A 200 m length was chosen as it is longer than the length of the barrier.
Table B.17: Risk Score calculation for a 200 m length of road containing the culvert end wall
Unshielded Shielded
Group Measure Item
roadside roadside
General road Terrain (flat, rolling, mountainous) A Rolling
characteristics
Environment (urban, rural) B Rural
Operating speed (70 km/h, 80 km/h, 90 km/h, C 110 km/h

110 km/h)
Divided road? (yes/no) D No
Number of lanes in the direction of travel E 1
Lane width (m) F 3.5
Design AADT (veh/day) G 4000
Length of road segment (m) H 200
Particular road Chainage where roadside is assessed I –

characteristics
Side of travelled way (verge/left or median/right) J Verge
Grade (%) (upgrade positive) K –6%
Curve radius (m) (Use 2000 m if straight) L 700
Curve direction (left/right in the direction of M Right

travel)
Background Description: N Trees at the toe of the embankment
hazard
characteristics Lateral distance to the background hazards from O 2.4
Isolated hazard Description P Drainage culvert Safety barrier
characteristics headwall
Length of the hazard (m) Q 3 130
Average distance between isolated hazards (m) R 200 200
Lateral distance to the isolated hazards from the S 2.4 1.5

Exposure Base exposure (encr/km/yr) T 0.82
Number of lanes factor U 1.0
Lane width factor V 1.0

Terrain factor W 2.58
Grade factor X 2.0
Curve correction factor Y 1.0
Exposure (encr/km/yr) Z 4.22

Unshielded Shielded
Group Measure Item
roadside roadside
Likelihood Likelihood for background hazards (considering AA 0.74 0.74
curve radius)
Modification factor for likelihood if the isolated AB 0.90 0.26
Modified likelihood for background hazards AC 0.66 0.19
Likelihood of impacts with isolated (point) AD 0.74 0.83

Likelihood of impacts with isolated hazards AE 0.08 0.61
Likelihood of a rollover AF 0.01 0.01
Severity Trauma Index for background hazards AH 6.0 6.0

Trauma Index for background hazards at the AJ 6.0 6.0
Trauma Index for isolated hazards (operating AK 6.0 0.84
speed of 110 km/h)
Trauma Index for isolated hazards at the AL 6.0 0.84
Risk Score Calculate Risk Score (product of the exposure, AP 16.7 4.9
hazards
Calculate Risk Score (product of the exposure, AQ 2.0 2.2
hazards
Calculate Risk Score (product of the exposure, AR 0.04 0.04
Combined Risk Score AS 18.7 7.1
The results in Table B.17 show a much greater improvement, simply because of the shorter road segment.
Note that the results have been normalised as if to apply over 1 km length of road. This can be visualised as
5 identical 200 m road segments used to produce a 1 km segment. The risk when the culvert is shielded is
still high as there is still 70 m of unshielded trees. Table B.17 does indicate the value in shielding the end
wall, but it has not reduced the risk to be below the NRRIT.
If a continuous barrier is installed 1.5 m from the traffic lane and the operating speed is reduced to 90 km/h
then the risk is as calculated in Table B.18. Note that in this case there are no isolated hazards and the
background hazard is the safety barrier. This has greatly reduced the risk to errant vehicles at this roadside.

Table B.18: Risk Score calculation for reduced speeds and a continuous barrier

General road Terrain (flat, rolling, mountainous) A Plans Rolling
characteristics
Operating speed (70 km/h, 80 km/h, 90 km/h, C Plans 90 km/h

110 km/h)
Number of lanes in the direction of travel E Plans 1
Lane width (m) F Plans 3.5 m
Design AADT (veh/day) G Plans 4000
Particular road As in Table B.14

characteristics
Background Description: safety barrier N Plans
hazard
characteristics Lateral distance to the background hazards from O Plans 1.5 m
Isolated hazard No isolated hazards
characteristics
Exposure Exposure (encr/km/yr) (see Table B.14) Z 4.22

curve radius)
Modification factor for likelihood if the isolated AB 1-(O+18)/P 1.00
Modified likelihood for background hazards AC Y*Z 0.74
Trauma Index Trauma Index for background hazards AH Table B.4 0.84
(operating speed of 110 km/h)
hazards
Combined Risk Score AS AP+ AR 1.44
Note: Rows have been omitted when they were not relevant or were previously calculated.
Note that the analysis does not consider only the isolated hazards, but a combination of background and
isolated hazards. Also note that the risk relates to a longer road segment not just to a short length around a
hazard. This enables the risk to be standardised (over 1 km) and for a more global view of risk to be
obtained.
This example assumed rolling terrain. Table B.1 indicated that in rolling terrain, there are 2.58 more
encroachments over the number on level terrain. For most well-designed roads this would seem to be an
excessive factor and, in most cases, rolling and mountainous should be used for poorer design standard
roads in hilly and mountainous regions. It is important to note that roads in rolling terrain have more
encroachments than roads in mountainous terrain, other things being equal.

B.10 Example 2 – Centreline Treatments on a Rural Two-lane Road
In this example, a two-lane rural road has 3.5 m lanes, 3 m shoulders and a line of trees 5 m from the road.
The current AADT is 3000 vehicles per day. The road is flat and the operating speed is 110 km/h. In
10 years, development will see the traffic increase. It is questioned whether the road should have a barrier in
the centre of the road or not, both now and when the AADT is 5000 veh/day.
The process, outlined above, gives a Risk Score of 2.0 for an AADT of 3000 veh/day. The risk associated
with oncoming vehicles is evaluated using Table B.13. For this example, the result, shown in Table B.19, is a
Risk Score of 1.6.
When the Traffic has increased to 5000 veh/day, then the Risk Score for the roadside increases to 2.20 and
the cross-median Risk Score has increased to 2.78. This is also shown in Table B.19.
If the NRRIT is 2.0, then the verges should be treated when the AADT is 3000 veh/day. The median needs to
be treated when the traffic is 5000 veh/day and perhaps before. To establish when the median needs to be
treated requires the process, described here, to be used iteratively. This has not been done for this example.
Table B.19: Risk Score for a median treatment of a two-lane two-way rural road
Item Result for AADT = Result for AADT =

Group Measure
3000 veh/day 5000 veh/day
General road Terrain (flat, rolling, mountainous) A Flat
characteristics
Environment (urban, rural) B Rural
Operating speed (70 km/h, 80 km/h, C 110 km/h

90 km/h, 110 km/h)
Divided road? (yes/no) D No
Number of lanes in the direction of travel E 1
Lane width (m) F 3.5
Design AADT (veh/day) G 3000 5000
Length of road segment (m) H 1000
Particular road Chainage where roadside is assessed I –

characteristics
Side of travelled way (verge/left or J Median
median/right)
Grade (%) (upgrade positive) K 0
Curve radius (m) (Use 2000 m if straight) L Straight
Curve direction (left/right in the direction M –

of travel)
Background Description: Oncoming vehicles N
hazard
characteristics Distance between oncoming vehicle O 0
lanes (m)
Exposure Base exposure (encr/km/yr) T 0.76 0.83
Number of lanes factor U 1.0
Lane width factor V 1.0

Terrain factor W 1.0
Grade factor X 1.0
Curve correction factor Y 1.0
Exposure (encr/km/yr) Z 0.74 0.83

Item Result for AADT = Result for AADT =

Group Measure
3000 veh/day 5000 veh/day
Likelihood Likelihood of impacts with oncoming AG 0.025 0.039
vehicles
Trauma Index Trauma Index for an operating speed of AN 87 87
110 km/h
Trauma Index at the appropriate AO 87 87
operating speed
Risk Score Calculate Risk Score (product of the AT 1.6 2.8
exposure, likelihood and Trauma Index)
B.11 Number of Fatal and Serious injuries
The procedure here is based on the expected notional number of FSI crashes. The Risk Score has been
developed to be proportional to the number of FSI crashes. Double the Risk Score and the number of FSI
crashes are expected to double.
The NZ Transport Agency (2011) indicated that there were 1.6 times as many fatal and serious injuries as
there were FSI crashes for head-on crashes. For run-off-the-road crashes there were 1.2 times as many fatal
and serious injuries as there were FSI crashes. These ratios were used in developing the likelihood function
for the risk of collisions with oncoming vehicles on two-lane rural roads.

Appendix C Treatment of Roads Based on Jurisdictional

Policies
A road agency may establish policies for the installation of safety treatments on roads in a uniform fashion. A
jurisdiction might have policies that included:
• the installation of barriers on selected road segments regardless of roadside and median dimensions
(Appendix C.1)
• the installation of barriers to shield and to protect critical infrastructure or land (Appendix C.2)
• the installation of barriers between adjacent carriageways (Appendix C.3)
• the installation of barriers or barrier types or other treatments for other defined road types and locations.
This includes a Safe System policy of installing continuous barriers along defined rural roads
(Appendix C.4).
While the locations described in these dot points are considered appropriate for all jurisdictions, it is up to
each jurisdiction to establish their own policies. The following sections discuss these locations in more detail.
The requirements for bridge parapets and hence associated bridge approaches are described in the Bridge
Design Code (AS 5100.1-2017).
Roadsides that are subject to a particular jurisdictional policy are not evaluated by the risk assessment
method described in Section 5.
C.1 Installation of Barriers on Particular Road Segments
The requirements for installing barriers on freeways and high-speed divided carriageway roads in greenfield
areas are documented in Section 2.3. The NDD requirements for barriers on freeways and rural highways in
brownfield sites should follow the requirements for greenfield sites.
The AASHTO Roadside Design Guide (AASHTO 2011a) indicates that a barrier is optional in medians 15 m
and wider or with an AADT less than 20 000 veh/day. Consideration could be given to establishing a
jurisdictional EDD policy to allow for barriers to be not installed in a median of a rural highway, with an
operating speed of 110 km/h, if the median is wider than 15 m or the AADT is less 20 000 veh/day. (A barrier
is required if the median is narrower than 15 m and the AADT is more than 20 000 veh/day.)
Roadside safety for many roads, with divided carriageways and operating speeds less than 100 km/h, would
benefit from the installation of barriers. This can be addressed through a jurisdictional policy or a corridor
safety vision.
C.2 Installation of Barriers at Sites with High Consequence Infrastructure and

Land
High consequence infrastructure is an artificial object for a public utility that cannot feasibly be removed or
relocated and if impacted by an errant vehicle the consequences could include:
• significant personal injury (including fatalities) to the occupants of the errant vehicle and other road users
• significant personal injury (including fatal injury) to others in or around the incident site
• events that indirectly cause a significant disruption to a sector of the community.
Examples include bridges, bridge piers, gantries, rail infrastructure, mass transport system components,
airport infrastructure and electricity transmission towers.

High consequence land has a primary use that is not for a public utility and is located outside of the road
corridor. If encroached into by an errant vehicle, the consequences could include:
• significant personal injury (including fatalities) to the occupants of the errant vehicle and other road users
• significant personal injury (including fatalities) to a predictably high concentration of people in or around
the incident site.
Examples may include some schools, child care centres and fuel storage facilities, but generally do not
include residential or commercial land uses.
High consequence infrastructure and land may warrant protection regardless of the number of road users.
This is often termed third party protection. These sites may be shielded without further evaluation if the need
for protection can be justified on the criticality of the infrastructure or land to the community.
C.3 Installation of Barriers Between Adjacent Carriageways
At times, adjacent carriageways present a hazard to occupants of errant vehicles on one carriageway and
road users on the other carriageway. Section 2.3.1 covers the medians on roads with divided carriageways.
There are some additional road configurations where barriers between traffic streams should be considered.
These include:
• between a high-occupancy vehicle lane and normal traffic lanes
• on freeway/motorway carriageways that require separation for same direction traffic to manage access
to/from multiple destinations
• on carriageways where there is traffic adjacent to the through traffic where there is a speed differential
equal to or greater than 20 km/h.
C.4 Installation of Barriers or Other Safety Measures on Other Defined Road

Types and Locations
A jurisdictional policy may be developed to include:

• the installation of continuous barriers
• the installation of wide centreline treatments with or without a barrier
• the installation of barriers or barrier types for other defined road types and locations for instance the
treatment on high fill embankments or culverts where the risk to an individual is high and the jurisdiction is
responding to a perceived risk
• the installation of audio-tactile line-marking on rural roads
• the sealing of shoulders
• a mass action plan to remove a particular hazard. For instance, the plan might be to replace old style
breakaway cable terminals (BCTs) with compliant terminals.
Policies may be written as a generic jurisdiction-wide policy or they may be written as a corridor specific
policy and included in the corridor safety vision. For instance, Safe System solutions (Austroads 2018) could
be applied to a corridor in this manner. Refer to Section 2.3.
If such a policy exists then the barrier can be installed without further analysis, although the policy needs to
be justified through the application of the guidelines in this Part.

Appendix D Risk Score Charts
These Risk Score charts are provided on consecutive pages so that a user is able to quickly gain an
assessment of the risk of run-off-the-road crashes on the roadside. These charts were developed using the
procedure in Appendix B. Assumed values for the roadside geometry and characteristics are used for each
chart and these are listed in tables below each chart. Table D.1 is a summary of the charts presented in this
Appendix.
Table D.1: Risk evaluated in the figures in this appendix
Reference road
Figure Risk evaluated Charts
characteristics
D.1 Undivided rural roads with background hazards 1 to 5 Table D.2
D.2 Undivided rural roads with isolated and background hazards 6 to 10 Table D.3
D.3 Undivided rural roads with isolated and no background hazards 11 and 12 Text
D.4 Vehicle rollover on roadsides on undivided rural roads 13 Text
D.5 Divided urban roads with significant background hazards 14 to 16 Table D.4
D.6 Divided urban roads with isolated and background hazards 17 to 19 Table D.4
D.7 Undivided lower speed urban roads with background hazards 20 to 22 Table D.5
D.8 Roadside barriers on undivided rural roads 23 to 25 Table D.6
D.9 Roadside barriers on divided carriageway urban roads 26 to 28 Table D.6
D.10 Roadside barriers on divided roads with an operating speed of 110 km/h 29 Text
D.11 Culvert headwalls 30 to 34 Table D.7

Figure D 1 establishes the Risk Score for undivided rural roads with significant background hazards at
different offsets and for different operating speeds and road types. Table D.2 lists the road characteristics
assumed in each chart. In rural areas, the operating speed is taken as the design speed, while in urban
areas the operating speed is taken as the speed limit. (These are copied from Figure 1.6 and Table 1.1).
Figure D.2 establishes the Risk Score for undivided rural roads with significant isolated hazards at different
average spacings along a roadside and for different operating speeds and is evaluated over 1 km. Table D.3
lists the road characteristics assumed in each chart. (These are copied from Figure 1.7 and Table 1.2).
Figure D.3 establishes the Risk Score for isolated significant hazards on undivided rural roads without
background hazards and is a copy of Figure 1.8.
Figure D.4 establishes the Rollover Risk Score for undivided rural roads without background hazards and an
operating speed of 110 km/h, and is a copy of Figure 1.9.
Figure D.5 illustrates the Risk Score for divided urban roads with significant hazards at different offsets and
for different operating speeds and Figure D.6 illustrates the Risk Score for divided urban roads with
significant hazards at different average spacings along a roadside and for different operating speeds. (These
figures are copies of Figure 1.10 and Figure 1.11). Table D.4 lists the road characteristics for Figure D.5 and
Figure D.6.
Figure D.7 illustrates the Risk Score for single carriageway urban roads with significant continuous hazards
at different offsets from the road and for different operating speeds (refer to Table D.5 for road details). This
figure is provided to illustrate the expected risk although designers should use this figure in the knowledge
that this is an extrapolation of the data and the Ray et al. (2012b) procedure.
Figure D.8 illustrates the Risk Score for roadside barriers on undivided rural roads with different operating
speeds and terrain and Figure D.9 illustrates the Risk Score for roadside barriers on divided carriageway
urban roads with different operating speeds and barrier offsets. Table D.6 lists the road details for these
charts. (These are copied from Figure 1.12 and Figure 1.13 and Table 1.5 and Table 1.6).
Figure D.10 illustrates the Risk Score for roadside barriers on divided carriageway urban and rural roads with
an operating speed of 110 km/h and is a copy of Figure 1.14.
Figure D.11 illustrates the Risk Score for culvert headwalls which are classified as being significant hazards.
The road section length is 100 m, rather than the 1 km length used in other charts in this Appendix.
Table D.7 lists the road details for these charts.
Table D.8 lists the significant hazards (from Section 1.9.2) and less significant hazards from Section 1.9.3. If
the road has background hazards in the less significant list then the Risk Score should be reduced by using
the factors shown in Table D.8. (Also see Section 1.9.3).
Table D.9 lists the Risk Score adjustment factors for curve radii on two-lane rural roads. This is a copy of
Table 1.3. Table D.10 lists the Risk Score adjustment factors for the gradient of two-lane rural roads. This is
a copy of Table 1.4.

Figure D 1: Risk Score for undivided rural roads with significant background hazards
Table D.2: Road characteristics for the charts in Figure D 1
Chart
number
1 Rural mountainous Mountainous 70 km/h 3.0 m
70 km/h
Note: There are no isolated hazards.

Figure D.2: Risk Score for undivided rural roads with significant isolated and background hazards
Note: Refer to Table D.3 for road and hazard characteristics and dimensions. The road segments are straight or winding
without sharp curves and with a nominal length of 1 km.
Table D.3: Rural road characteristics for the Risk Score charts in Figure D.2
Chart Operating Background Isolated

Chart description Terrain Lane width
number speed hazards at hazards at
Rural mountainous
6 Mountainous 70 km/h 3.0 m 3m 1m
70 km/h

Figure D.3: Risk Score for undivided rural roads with isolated and no background hazards
Note: Lanes are 3.5 m wide. The terrain is flat. The operating speed is 110 km/h.
Figure D.4: Risk Score for vehicle rollover on roadsides on undivided rural roads
Note: Lanes are 3.5 m wide. The terrain is flat. The operating speed is 110 km/h. The Risk Score axis has a different
scale to other charts.

Figure D.5: Risk Score for divided urban roads with significant background hazards
Note: Refer to Table D.4 for road characteristics.
Figure D.6: Risk Score for divided urban roads with significant isolated and background hazards
Table D.4: Road characteristics for the charts in Figure D.5 and Figure D.6
Offset of
Chart Chart Operating Lane Isolated
Terrain background
number description speed width hazards
hazards
14 Urban 70 km/h Flat 70 km/h 3.5 m
15 Urban 80 km/h Flat 80 km/h 3.5 m Various nil
Various average
18 Urban 80 km/h Flat 80 km/h 3.5 m 4.0 m* spacings at 1 m
offsets
* The background hazards is a generic wall 4.0 m from the edge of the traffic lane.

Figure D.7: Risk Score for undivided lower speed urban roads with background hazards
Table D.5: Road characteristics for the charts in Figure D.7
Offset of Offset in
Operating Lane
Chart number Rural or urban Terrain background isolated
speed width
hazards hazards
20 Urban undivided Flat 50 km/h 3.5 m Various Nil
Notes:
 These charts are provided for comparison purposes and are an extrapolation of the method.
 The Risk Score between 5 000 and 10 000 AADTs is constant.

Figure D.8: Risk Score for roadside barriers on undivided rural roads (refer to Table D.6 for road details)
Figure D.9: Risk Score for roadside barriers on divided carriageway urban roads (refer to Table D.6 for road
details)
Table D.6: Road characteristics for the charts in Figure D.8 and Figure D.9
Chart number Rural or urban Terrain Operating speed Lane width Offset of the barrier
* The barrier offsets used produce conservative (higher) Risk Scores. If the barrier was located further from the road the
Risk Score would be smaller.

Figure D.10: Risk Score for roadside barriers on divided roads with an operating speed of 110 km/h
Note: The barrier offset of 2.0 m is used. This produces conservative (higher) Risk Scores. If the barrier was located
further from the road the Risk Score would be smaller.
Figure D.11: Risk Score for culvert headwalls classified as a significant hazard, measured over a 100 m road
section
Note: 100 m measurement distance used. Refer to Table D.7 for road characteristics.

Table D.7: Road characteristics for the charts in Figure D.11
Chart
number
30 Rural mountainous Mountainous 70 km/h 3.0 m
70 km/h
Table D.8: Significant and less significant and minor hazards
Significant hazards Less Significant hazards Minor hazards
Factor =1 Factor =1/2 Factor =1/6

Generic fixed object
Rock cuttings
Rocks protruding more than 300 mm Rocks protruding between 200 mm
above the ground surface and 300 mm above the ground surface
Ruts in the ground surface more than Ruts in the ground surface between
300 mm deep 200 and 300 mm deep
Tree lined edge
Water
Utility pole Slip base or energy absorbing poles
Tree
Traffic signal pole
Bridge pier ,
2:1 fill batters more than 5 m high 3:1 fill batters more than 2 m high 4:1 fill batters more than 2 m high
2:1 fill batters between 2 and 5 m high 3:1 fill batters between 2 m and
1.5:1 fill batters more than 2 m high
5 m high
Vertical drops more than 2 m 1.5:1 fill batters less than 2 m high 2:1 fill batters less than 2 m high
Watercourses less than 1 m deep and Watercourses less than 1 m deep
Watercourses more than 2 m deep.
with a drop of between 0.3 and 2.0 m and with a drop less than 0.3 m
Down slopes, parallel to the road, Down slopes, parallel to the road, Down slopes, parallel to the road,
higher than 5 m and with a slope of 4:1 between 2 and 5 m with a slope flatter less than 2 m high
or steeper than 2:1
Down slopes, parallel to the road, Down slopes, parallel to the road,
between 2 and 5 m high and with a higher than 2 m and with a slope of 6:1
slope of 2:1 or steeper or flatter
Up slopes, parallel to the road, 1.5:1 or Up slopes, parallel to the road, steeper Up slopes, parallel to the road,
steeper than 4:1 but flatter than 1.5:1 with slopes of 4:1 or flatter
Significant drainage structures,
exposed culvert headwalls and wing
walls, with vertical height drops of
more than 2 m or watercourse more
than 2 m deep

Table D.9: Risk Score correction factors for curve radii on two-lane rural roads
Curve radius Location 70 km/h 90 km/h 110 km/h

300 m Inside corner 2.6 to 3.5 2.6 to 3.3 2.3 to 2.8
Outside corner 5.3 to 7.0 5.1 to 6.5 4.7 to 5.6
Table D.10: Risk Score correction factors for the gradient of two-lane rural roads
Gradient Upgrade 2% downgrade 4% downgrade 6% downgrade

Factor 1.0 1.0 1.5 2.0

Appendix E Cost of Impacts
E.1 Determine Crash Costs
The risk analysis procedure described in Section 1.9 does not include costs as it is considered that road
safety should concentrate on trauma and not costs. However, some jurisdictions may prefer a monetary
approach in a benefit-cost analysis and this Appendix will assist them.
The risk analysis procedure presented in this Part is based on the research by Ray et al. (2012b). Their
procedure tabulated a severity for impacts into different hazards. The severity index is a function of the
Equivalent Fatal Cash Cost Ratio (EFCCR), which is the average cost of a crash with a particular hazard
divided by the average cost of a fatality. For instance, the costings used in the Queensland Department of
Transport and Main Roads program RISC are $9 775 for property damage only, $40 930 for crashes
involving possible minor injuries, $125 395 for crashes in which occupants require medical treatment,
$633 244 for crashes in which occupants are hospitalised and $9 077 270 for fatal crashes. These costs are
in 2016 dollars. Changing the crash costs for the different injury classifications will affect the outcome.
Table E.1 lists the Trauma Index for the different severity indices in the RSAPv3 procedure. For RSAPv3
severity indices of 1 and 2, Ray et al. (2012b) listed no fatalities and serious injuries. It was considered in
developing this process that there is always a small probability of a serious injury and so Trauma Index
values of 0.1 and 0.5 have been included in Table E.1. The proportion of different injuries for a particular
hazard impact severity (and Trauma Index) were established in AASHTO (1989) and the background to
these ratios is unknown.
Table E.1: Trauma Indices and predicted crash outcomes
Property
Moderate Serious
Trauma Index Severity damage only Minor injury Fatal crashes
injury injury
(%) index (PDO) crashes (%) (%)
crashes (%) crashes (%)
crashes (%)
0 0 100
0.28 1 90.4 7.3 2.2 0.1
0.55 2 71 22 6.5 0.4 0.1
2 3 43 34 21 1 1
8 4 30 30 32 5 3
18 5 15 22 45 10 8
38 6 7 16. 39 20. 18
60 7 2. 10 28 30 30
77 8 4. 19 27 50
93 9 7 18. 75
100 10 100
Source: Ray et al. (2012b) and Austroads (2020b).

Appendix F Treatments for Brownfield Sites
F.1 Treatments for Roads
F.1.1 Treatments for Pavement Edge Drop-off
Pavement edge drop-offs can lead to loss of control under certain circumstances, where inattentive or
inexperienced drivers return to the traffic lane by oversteering to overcome the resistance from a continuous
pavement edge and tyre-scrubbing condition. This can result in drivers crossing the opposing traffic stream
sometimes with disastrous consequences.
Pavement edge drop-offs can occur during highway repair or resurfacing work. When not properly
addressed, drop-offs can lead to loss of vehicle control with a high potential for a serious crash.
Pavement edge drop-offs can be rectified by edge patching or sealing of shoulders.
F.1.2 Treatments for Opposing Vehicles
Where the opposing lane or lanes are within the area of interest, there is a danger of collision with other
vehicles, especially oncoming vehicles. In this case all available treatment options should be considered
including the following options:
• Provide sufficient median width to place the opposing carriageway outside the area of interest.
• Provide a road safety barrier in the median.
• Where traffic flows warrant only a two-lane, two-way road it may not be possible to introduce a central
median. In this situation ensure the roadside is free of hazards so vehicles can take evasive action if an
opposing vehicle crosses the centre of the road.
• However, in some instances it may be possible to provide a narrow, painted median to increase the
separation or a barrier in a narrow median (Figure F.1).
Figure F.1: Wire rope barrier in a narrow median on a sharp bend

F.2 Treatments for Bridges
F.2.1 General
The structural limitations of old bridges often prevent upgrade of their barriers to current standards.
Engineering expertise and judgement must be used to design the best possible upgrade if a risk assessment
shows that the existing barrier has an unacceptable risk. The reasons for not meeting current standards
need to be well documented and justified.
If the bridge and its barriers present a very high risk, and upgrade is not structurally possible, it may be
necessary to program the replacement of the bridge.
F.2.2 Treatments for Bridge Piers, Abutments, End Posts and Tunnel Portals
Bridge ends should be designed to prevent vehicles from running into end support posts, being speared by
any horizontal bridge members or simply crashing through any approach barrier and being exposed to a
hazard (e.g. rollover, railway track, watercourse).
Stiffening needs to be provided on the transition from the semi-rigid approach barrier to the rigid bridge
structure, otherwise the excessive local deformation will cause errant vehicles to snag on the end of the
bridge barrier.
The piers of bridges over roads (at overpasses) should be shielded by a crash cushion or barrier. It may be
necessary to provide a barrier that can shield piers from heavy vehicle crashes, which may involve a
two-stage shielding system (Figure 5.18).
F.3 Barrier Placement in Constrained Situations
F.3.1 Location on Embankments
Barriers perform best on slopes flatter than 10:1 (10% maximum crossfall). Barriers should not be placed on
embankments steeper than 6:1. On steep embankments, the vehicle trajectory may impact the barrier too
high and cause the vehicle to launch. Vehicle trajectory on embankments and cut batters is discussed and
illustrated in Commentary 6.
The hinge point of batters steeper than 4:1 should be located outside the deflection width of the barrier. In
low-speed areas with constrained formation widths (e.g. existing mountain passes) the post locations shown
in Figure F.2 may be necessary but should only be adopted where there are no other options.

Figure F.2: Barrier post locations on constrained sites
Non-standard
Made-to-order post Non-standard
Made-to-order post Non-standard
Made-to-order post
Source: Based on Washington State Department of Transportation (2014).
F.3.2 Location on Urban Footpath Corners
It is generally preferable that barriers are not provided on urban footway corners to protect pedestrians
because:
• The encroachment of errant vehicles at these locations is a relatively rare event.
• It is necessary to provide openings in the barrier for pedestrian movement and difficult to accommodate
safe end treatments (which are very expensive).
• The short lengths of barrier are too short to develop the required strength and operate as intended when
impacted.

Where it is necessary to install a barrier to shield a vulnerable site (e.g. childcare centre) it is preferable to
locate the barrier on the fence line rather than over the kerb. This will avoid the problem of shorter than
minimum barrier lengths, lack of terminals, locating a terminal on the kerb line, conflict of posts with drainage
conduits and avoid problems with pedestrian access across the barrier, as shown in Figure F 3.
Figure F 3: Road safety barriers on footway corners
F.4 Wire Rope Barriers in Narrow Medians
The use of wire rope barriers in the centre of narrow medians is a relatively new practice. As mentioned
previously it is preferable to contain the dynamic deflection within the median. However, it may be
appropriate to allow partial encroachment of the deflected barrier into the opposing traffic lane. Table F.1 lists
the effects that a centrally located wire rope barrier will have after impact for various median widths and
which should be considered if the barrier is proposed for a narrow median.
Containment of deflections within the median will influence the width of the median to be provided. Roads
and Maritime Services considers that 1.6 m is the minimum median width for installation of wire rope barriers
as the half-median width of 0.8 m is generally sufficient to contain dislodged cables and bent posts from
damaged installations.

The cables of a severely impacted wire rope barrier system form a chord across the inside of a curve
because the cables are released from the supporting posts. Tests at 80 km/h and 25° impact angle on a
200 m radius curve show that the cables will lie about 0.8 m off the line of the barrier after impact and may
become a hazard for oncoming traffic. It is likely that this offset would be larger at higher impact speeds due
to dislodgement of longer lengths of wire rope.
Table F.1: Issues for wire rope barriers located centrally in medians
Consequences of 1.7 m deflection at

Median width Debris after impact on 200 m curve
100 km/h
2.8 m Deflection will encroach 0.3 m into Bent posts and cables lie within median.
2.0 m Deflection will encroach 0.7 m into Bent posts and cables lie within median.
1.6 m Deflection will encroach up to 0.9 m into Bent posts lie within median. Cables may lie on
opposing carriageway. edge of carriageway.
The deflection of a wire rope barrier may be greater than the width of the median, which means that the
impacting vehicle may encroach onto the opposing carriageway. A crash test on a 200 m radius curve at
80 km/h impact speed and 25° angle of impact showed that the 1500 kg vehicle was past the original line of
the barrier for about 1.5 seconds, after which the vehicle would be back in the traffic lane (Roads and Traffic
Authority 2003). The probability of a collision due to running off the road after impact is related to the
probability of a vehicle being adjacent to the impact site during this short period.
Refer also to Austroads (2016a & b).

Appendix G Angle of Departure Method
Angle of departure method
The angle of departure method is preferred by some road agencies as it determines the length of barrier
required from angles at which vehicles are assumed to leave the road.
Straight roads
The angle of departure of vehicles leaving the road varies over a range of values. In this method vehicle
trajectories are plotted based on angles at which most vehicles are likely to depart from the traffic lane, in
order to establish the barrier points of need and the length of barrier required. This method is illustrated in
Figure G.1. The angle of departure is related to the posted speed limit and values are shown in Table G.1.
Figure G.1: Angle of departure method of determining length of need
Source: Based on Roads and Maritime Services (2016).

Table G.1: Angles of departure from the road
Sign-posted speed limit 15th percentile angle (1:X) 85th percentile angle (1:X)
(km/h) use as leading angle (‘a’) use as trailing angle (‘b’)
60/70 5.7° (1:10) 22° (1:2.5)
80/90 3.8° (1:15) 22° (1:2.5)
100/110 2.9° (1:20) 22° (1:2.5)
Note: For operating speeds less than 60 km/h use the values for 60/70 km/h.
The method to establish points of need for a two-direction carriageway (Figure G.1(a)) is:
1. Identify the first possible point of contact with the hazard for direction 1 (the lane adjacent to the barrier)
(Point A).
2. Using the impact angles from Table G.1, project a line at the leading impact angle until it intersects the
offset line of the barrier (Point B).
3. Record this as the leading point of need.
4. Identify the first possible point of contact with the hazard for direction 2 (the opposite lane) (Point C).
offset line of the barrier (Point D). Record this as the trailing point of need.
From these points establish the longitudinal position and length of the barrier installation. The points of
redirection for the barrier system should be outside the length of need.
The method to establish points of need for a single-direction carriageway (Figure G.1(b)) is:
1. Identify the first possible point of contact with the hazard (Point A).
offset line of the barrier (Point B).
3. Record this as the leading point of need.
4. Identify the last possible point of contact with the hazard (Point C).
5. Using the impact angles from Table G.1, project a line at the trailing impact angle until it intersects the
offset line of the barrier (Point D). Record this as the trailing point of need.
If the hazard is located on a median and is in the area of interest (for the opposing direction of traffic, repeat
the process for the opposing direction of traffic).
Curved roads
When determining the leading point of need for a road safety barrier, the angle of departure of an errant
vehicle should be taken from the outer edge of the traffic lane in all cases. Working back from the obstacle
will give the same result if the lane/road alignment is straight, but when the alignment is curved, the leading
and trailing angles of departure should be determined from a tangent on the outside of the edge of the traffic
lane.
For a curve, the leading angle of departure from Table G.1 (2.9° for speeds equal to or greater than
100 km/h) is taken off a tangent to determine where the initial point of need lies when this angle meets with
the back of a hazard. The trailing angle of departure at 22° is then taken from a tangent in front of the hazard
to determine the final point of need for a one-way road. Figure G.2(a) to (d) illustrates the situations for
hazards on the outside and inside of a curve, and for two-way and one-way carriageways.

In determining the length of need for a barrier, there is a range of angles of departure that are considered
between the leading angle of 2.9° (at 100 km/h) and the trailing angle of 22° (for all speeds). These are
general limits and when applied in cases where the leading angle from Table G.1 does not meet with the
hazard, a departure angle that is somewhere between the leading and trailing limits must be considered.
On the inside of a horizontal curve, a slightly different procedure is required if the leading angle of departure
does not meet with the back of the hazard (i.e. the line passes through or in front of the hazard), and as a
consequence the initial point of need for the barrier does not relate to the rear of the hazard. However, the
leading and trailing angles cover a range and an angle within these limits can be used as a leading angle for
establishing the initial point of need. Therefore, in these situations a chord to the curve should be drawn
across the back of the hazard, square to the centre of the curve. This process is illustrated in Figure G.3(a)
and (b) for two-way and one-way carriageways. The chord should be extended to intersect with the edge of
traffic lanes at points A and B. Point A is where the leading angle of departure begins for traffic in the lane
adjacent to the hazard, and B is the corresponding point for the opposing traffic. The leading angle of
departure is the angle between the chord and the tangent to the curve at A. It can be calculated and will be
somewhere in the range of 2.9° to 22° for a speed limit of 100 km/h or greater.

Figure G.2: Angle of departure method on curves where leading angle meets the rear of hazard
Note: A–B is the length of need.

Figure G.3: Angle of departure method where leading angle does not meet the rear of hazard
Notes:
 A – B is the length of need.
 C – D is the chord across the rear of the hazard.

Appendix H Examples of Length of Need Calculations
H.1 General
The two methods currently used in Australia and New Zealand for determining the length of road safety
barrier required to shield a hazard are presented in Section 5.3.18. AASHTO (2011a) suggests the use of the
run-out length method for determining the length of need of a barrier and acknowledges that some
jurisdictions choose to use an alternative method based on the vehicle angle of departure from the road.
In Australia and New Zealand jurisdictions may use either the run-out length method or the angle of
departure method. Each example in this appendix is calculated using both methods.
H.2 Run-out Length Method
This method is presented in Section 5.3.18 and figures, tables and formulae are reproduced below for
convenience with respect to the examples provided. Application of the run-out length method to establish
barrier length of need for both traffic in the adjacent lane, and for opposing traffic, is illustrated in Figure H.1.
On a two-lane two-way road, and for medians, these requirements are combined to develop a design layout
that shields hazards from both directions.
The layout of barriers on straight or nearly straight roads is established by applying the following formulae.
Dimension X is the required length of need in advance of the hazard and can be calculated from the
following equations.
For installations where the barrier is flared (refer to Section 5.3.18) Equation 1 (repeated here) is used.
𝑏𝑏 A3
�𝐿𝐿𝐴𝐴 + � � (𝐿𝐿1 ) − 𝐿𝐿2 �
𝑋𝑋 = 𝑎𝑎
𝑏𝑏 𝐿𝐿
�� + � 𝐴𝐴 ��
𝑎𝑎 𝐿𝐿𝑅𝑅
For parallel installations that have no flare (Equation 2 repeated here).
[𝐿𝐿𝐴𝐴 − 𝐿𝐿2 ] A4
𝑋𝑋 =
𝐿𝐿
� 𝐴𝐴 �
𝐿𝐿𝑅𝑅
The lateral offset, Y, from the edge of the running lane to the beginning of the length of need may be
calculated (from Equation 3 repeated here).
𝐿𝐿𝐴𝐴 A5
𝑌𝑌 = 𝐿𝐿𝐴𝐴 − (𝑋𝑋)
𝐿𝐿𝑅𝑅
where
X = the required length of need in advance of the hazard
LR = run-out length (refer to Table 5.8)
b/a = flare rate (refer to Table 5.5)
LA = lateral extent of the area of concern
L1 = tangent length of the barrier upstream from the area of concern
L2 = barrier's lateral distance from the edge of the running lane
Y = lateral distance from edge of traffic lane to point of need.

Figure H.1: Run-out length method of determining length of need
H.3 Angle of Departure Method
The angle of departure method is described in Section 5.3.18 and the relevant figure and table are
reproduced in the Appendices for convenience. The method is shown in Figure G.1 and appropriate angles
of departure for use in calculations are shown in Table G.1.
H.4 Worked Examples to Determine Road Safety Barrier Length of Need
The length of barrier required in the following examples is computed using both the run-out length method
and the angle of departure method. The examples assume a straight road alignment. While it is possible to
perform similar computations for a curved alignment, the graphical method described in Section 5.3.18 is
normally used.

H.4.1 Example 1
A two-lane, two-way road is being designed. It will have a design speed of 100 km/h. The predicted two-way
traffic volume (i.e. AADT) in 20 years’ time is 6500 vehicles per day. The predicted traffic volume includes
only 3% heavy vehicles.
A straight segment of the road crosses a significant waterway, which will require a culvert consisting of
4 m x 1.2 m diameter reinforced concrete pipes.
The typical road cross-section is shown in Figure H.2.
Figure H.2: Road cross-section

Sealed shoulder
2.0 m wide Unsealed shoulder
Unsealed shoulder Sealed shoulder
1.0 m wide
1.0 m wide 2.0 m wide
Lane 3.5 m Lane 3.5 m
wide wide
6:1 Slope 6:1 Slope

Natural surface
A detail of the plan view of the culvert headwall is shown in Figure H.3:
Figure H.3: Culvert headwall (plan view)
6.0 m
A B
5.5 m
4.0 m Edge of sealed shoulder
Edge of lane
Adjacent traffic
Centreline
Opposing traffic
Edge of lane
Edge of sealed shoulder
The culvert headwall is a significant roadside hazard and is to be shielded by a barrier. The barrier is to be
unflared and located 2 m offset from the edge of the lane, at the edge of the sealed shoulder.
Determine:
1. the length of need in the leading direction required to shield the culvert headwall only
2. the length of need in the trailing direction required to shield the culvert headwall only
3. the distance between the leading and trailing points of need.

Solution:
1. determine length of need in the leading direction.
Refer to Figure H.4 (part of Figure 5.22 reproduced here).
Figure H.4: Establishing the leading point of need
Notes:
 LS is shy-line offset.
From Table 5.8:
LR = 76 m (design speed = 100 km/h, traffic volume [AADT] = 6500 vehicles per day) A6
5.5 m (measured from the edge of the lane for adjacent traffic to the corner of the
LA =
culvert apron marked (A) on the plan view)
L1 = not applicable, as barrier is not flared
L2 = 2.0 m (it is noted that this offset will place the barrier within the shy-line)
L3 = 4.0 m
The length of need in the leading direction, X
(LA - L2 )
X =
(LA /LR )
X = (5.5 – 2.0)/(5.5/76)
X = 48.4 m.
Determine length of need in trailing direction.

Figure H.5: Establishing trailing point of need
Notes:
From Table 5.8:
LR = 76 m, design speed = 100 km/h, traffic volume [AADT] = 6500 vehicles per day) A7
5.5 m + 3.5 m = 9.0 m (measured from the edge of the lane for opposing traffic
LA =
(i.e. centreline) to the corner of the culvert apron marked (B) on the plan view)
L1 = not applicable, as barrier is not flared
L2 = 2.0 m + 3.5 m = 5.5 m
L3 = 4.0 m + 3.5 m = 7.5 m
The length of need in the trailing direction, X
LA - L2
X =
LA /LR
X = (9.0 – 5.5) / (9.0 / 76)
X = 29.6 m.
Determine distance between the leading and trailing points of need.
The distance between the leading and trailing points of need is:
leading length of need (48.4 m) +
trailing length of need (29.6 m) +
culvert apron (A–B) (6.0 m)

= 84.0 m.
This represents the length of redirective barrier that is required to shield the culvert headwall hazard.
It should be noted that a barrier system with footings that did not clash with the reinforced concrete pipe
culvert would be required.

H.4.2 Example 2
For the situation described in Example 1 (H.4.1), if the barrier was installed on a flare rate of 30:1, with the
tangent extending 2 m past the points of concern, what would be the effect? The barrier layout is shown in
Figure H.6, in red.
Figure H.6: Barrier layout
6.0 m
A B
30:1 flare (typ)

5.5 m
4.0 m Edge of sealed shoulder
Edge of lane
Adjacent traffic
Centreline
Opposing traffic
Edge of lane
Edge of sealed shoulder
Determine:
1. the length of need in the leading direction required to shield the culvert headwall only
2. the length of need in the trailing direction required to shield the culvert headwall only
3. the distance between the leading and trailing points of need.
Solution:
1. Determine length of need in the leading direction.

Figure H.7: Establishing leading point of need
Notes:
 LS is shy-line offset.
From Table 5.8:
LR = 76 m, (design speed = 100 km/h, traffic volume [AADT] = 6500 vehicles per day) A8
5.5 m (measured from the edge of lane for adjacent traffic to the corner of the
LA =
culvert apron marked (A) on the plan view)
L1 = 2.0 m
L2 = 2.0 m (it is noted that this offset will place the barrier within the shy-line)
L3 = 4.0 m
b/a = 1/30.
The length of need in the leading direction (Equation 1 reproduced here):
b A9
�LA + � � (L1 )- L2 �
a
XX =
b L
�� + � A ��
a LR
where
X = [5.5 + (1/30)(2.0) – 2.0] / [(1/30) + (5.5/76)]
X = [5.5 + 0.0667 – 2.0] / [0.0333 + 0.0724]
X = [3.5667] / {0.01057]
X = 3.8 m.
Determine length of need in trailing direction.
Refer to part Figure H.5 (reproduced here):

Figure H.8: Establishing the trailing point of need
Notes:
From Table 5.8:
LR = 76 m (design speed = 100 km/h, traffic volume [AADT] = 6500 vehicles per day) A10
5.5 m + 3.5 m = 9.0 m (measured from the edge of the lane for opposing traffic
LA = [i.e. centreline] to the corner of the culvert apron marked (B) on the plan view,
refer to Figure H.6)
L1 = 2.0 m
L2 = 2.0 m + 3.5 m = 5.5 m
L3 = 4.0 m + 3.5 m = 7.5 m
b/a = 1/30
X = The length of need in the trailing direction
b
� LA + � � (L1 )- L2 �
a
X =
b LA
�� + ��
a LR
X = [9.0 + (1/30)(2.0) – 5.5] / [(1/30) + (9.0/76)]
X = [9.0 + 0.0667 -5.5] / [0.0333 + 0.1184]
X = [3.5667] / [0.1517]
X = 23.5 m.
Determine distance between the leading and trailing points of need (measured along the barrier alignment).
The distance between the leading and trailing points of need is:
leading length of need (33.8 m) +
trailing length of need (23.5 m) +
culvert apron (A–B) (6.0 m)
= 63.3 m.

This represents the length of redirective barrier that is required to shield the culvert headwall hazard.
It should be noted that a barrier system with footings that did not clash with the reinforced concrete pipe
culvert would be required.
It should also be noted that while flaring of barriers may reduce the length of barrier required to shield a
hazard, it may also require widening of the embankment to provide sufficient support for the flared barrier
system. In some situations, the additional earthworks’ cost associated with widening the embankment may
exceed the cost saving due to the reduction in length of barrier created by flaring.

Appendix I Types of Safety Barrier Terminals
I.1 General
Road safety barrier terminals may be public domain or proprietary products. At sites that are unsuitable for a
public domain terminal treatment it will often be necessary to provide a proprietary terminal treatment.
However, in suitable situations it may be appropriate to terminate a barrier in a cutting face or a back slope.
Such public domain, buried terminal treatments can be effective and may be used provided that they are
designed and crash tested (including the anchor) to meet the requirements of the appropriate test level. This
type of treatment may be appropriate where:
• A road passes through a series of cut-to-fill lines, the cuttings are steep (e.g. say 0.5:1 or steeper),
smooth and able to redirect vehicles, and barrier is required between the cuttings. This may require the
use of a suitably designed end treatment/transition to anchor the barrier (e.g. perhaps in the shape of a
concrete barrier) that does not require significant disturbance of the cutting face.
• However, in deciding to adopt this technique designers should be confident that the batter approaching
the barrier system will redirect an errant vehicle and not result in the vehicle travelling up the batter and
behind the barrier.
• A suitably designed flat-bottomed drain or V drain exists at a site and it is desirable that the barrier
passes through the drain and is buried in a 4:1 back slope (refer to AGRD Part 3 (Austroads 2016a) for
suitable drain profiles). These buried terminals have been successfully tested at test level TL3; refer to
the USA Federal Highway Administration website, approval letter CC-53 (FHWA 1998) and CC-53A
(FHWA 2001). Key design considerations include:
– the height of the W-beam remaining constant relative to the roadway grade until the barrier crosses
the flow line of the drain
– a flare rate, appropriate until the barrier reaches the flow line
– adding a rubbing rail
– using an appropriate anchor (concrete block or steel post) that is capable of developing the full
tensile strength of the W-beam rail.
I.1.1 Gating End Treatments
Only end treatments (terminals) that have been accepted by the jurisdiction can be used.
Proprietary products
Most crash attenuators including those that gate (Figure 5.30) are generally patented proprietary products
and the manufacturer’s specifications and representative should be consulted to establish the availability of
new and improved products that have passed the required testing procedures. In addition, the designer
should consult with the road agency to determine acceptance criteria relating to new and improved products.
Up-to-date information, including in-service reports about crash attenuator features should also be used for
the selection/design procedure, it being recognised that specific and highly controlled crash tests are not
always adequate indicators of how crash attenuators will perform in different situations.
The number and complexity of factors that enter the selection process for crash attenuators preclude the
development of a simple selection procedure. Each operational system has its own unique physical and
functional characteristics. In some cases, one crash attenuator will stand out as the most appropriate, but in
most cases two or more types will provide satisfactory protection to an errant motorist, and the designer must
choose between them (AASHTO 2011a). The designer must therefore refer to the manufacturer’s
specifications and literature to develop a good understanding of the installation requirements of each device
and its behaviour under impact, so that the most appropriate product can be selected for any given situation.
The FHWA web site and approval letters should also be consulted.

Many attenuators provide a gating function and use various principles and mechanisms through which a safe
end treatment is achieved. In all cases site preparation and the installation of the device must meet the
manufacturer’s specifications. In particular, the point of redirection may vary between products. Products that
provide a gating terminal may use the methods of operation described below.
Displacement of sand
The containers are held together (e.g. by cables or other devices) and upon impact the deceleration of the
vehicle is controlled through displacement of the sand contained in deformable containers. The energy of the
vehicle impact is transmitted to the weights of sand in the barrels, thus dissipating the collision energy based
on the principle of conservation of momentum. It is essential that the sand used meets material grading
requirements.
The force of impact is not transmitted through the barrels so backup structures or walls are not required for
these systems. The systems can be used as either a crash attenuator placed directly in front of the hazard,
or as a road safety barrier end treatment. However, they will not redirect some side crashes, particularly
those occurring toward the rear of the installation. Damaged modules must be replaced after each impact.
These systems can be used to protect hazards of any width and are particularly suited to gore areas. They
can be used on the left side of the road or in medians. The site must be well compacted and be able to
accommodate a concrete or asphalt foundation pad and the transverse slope should not exceed 20:1.
Designers should note that the water content (typically 3%) in the sand might freeze if systems are installed
in mountainous regions and cold weather continues for several days. In this situation, the attenuators will not
work as designed. Mixing rock salt (5 to 25% by volume) with the sand will help reduce the possibility of
errant vehicles hitting barrels of frozen sand.
Collapsible steel beams and posts
These systems generally involve a structure of W-beams, posts and cable anchorages. Collision energy is
dissipated by the breaking away of the posts and shearing as W-beams telescope into each other.
If the first post is damaged in any way, a system may lose its redirective characteristics. Some systems may
also include containers filled with sand, liquid or other crushable material that contribute to the attenuation
qualities of the end treatment.
An entire system or some of its components may not be able to be salvaged and used again after a major
crash, and nuisance crashes may result in a system not operating as it should.
Deformation of a steel beam
These end treatments employ a steel impact-head mounted at the leading end of the system. On impact, the
head is pushed along the W-beam, causing the rail to deform, curl around or shred, thus dissipating the
collision energy.
These systems require sufficient width in the verge to accommodate the discarded rail sections and it is
important to establish whether the rail is extruded onto the traffic side of the system or to the back of the
system. An obstacle-free area for a distance of 18 m beyond the point of redirection of the barrier system
(parallel to the rails) and at least 6 m behind the rails is generally necessary. Systems may be designed to
suit straight or flared barrier alignments.
I.1.2 Non-gating End Treatments
Most non-gating end treatments are crash attenuators that do not allow a colliding vehicle to pass behind the
terminal (Figure 5.30). On colliding with the end of the terminal, the vehicle will be redirected away from the
barrier or be arrested by it.

Because non-gating end treatments do not require a clear, level area behind the barrier, their application is
suited to:
• median barrier ends where it is important to prevent colliding vehicles from encroaching onto the opposite
carriageway
• situations where a run-out area is not available, thus precluding the use of a gating terminal.
Non-gating end terminals are appropriate for shielding:

• barrier ends, including those in medians
• exit ramp gore areas
• fixed objects
• bridge rail ends
• bridge piers.
Non-gating terminals employ similar principles to gating terminals whereby crushable containers or cylinders,
collapsible structures and other mechanical devices (e.g. guide cables) may be employed. Some systems
may dissipate the energy of the impact through a braking mechanism and the nesting of barrier rails. Others
may employ rubber components or crushable materials that are capable of being reused after impact.
All impact attenuation systems available at present are patented products and must be installed in
accordance with the manufacturer's specifications.
Public domain end treatments
Where there is a need to install parallel semi-rigid road safety barriers, for example to shield a bridge pier in
a median, the public domain ‘bull-nose’ treatment shown in Figure I.1 may be suitable. The bull-nose is
constructed of a circular section of slotted thrie-beam supported by breakaway posts. The end treatment is
suitable for use with W-beam or thrie-beam barriers through the use of appropriate transition sections.
When a vehicle crashes into this bull-nose the posts in the nose breakaway and the rail deforms inward,
arresting the vehicle in the process. For the bull-nose to deform as intended under impact, the rail in the
nose section should not be bolted to the posts, and the bolt heads in the first section of rail at the sides
should not be provided with washers. A similar arrangement with splayed sides can be used to shield objects
in gore areas of off-ramps. Where necessary, a sign support may be installed behind the bull-nose provided
that it has a breakaway support.
The area within the bull-nose barrier system for a distance of 19.0 m beyond the nose must be free of
hazards (refer to Diagram A in Figure I.1). This requirement is based on a 100 km/h test. It should also be
noted that the original system was crash tested with timber breakaway posts.
Proprietary products
Non-gating crash attenuators are also available for use in situations where it is not desirable to allow vehicles
to pass through the nose section of the attenuator. They are generally patented, and the manufacturer’s
specifications and representative should be consulted to establish the availability of new and improved
products that have passed the required testing procedures. In addition, the designer should consult with the
road agency to determine acceptance criteria relating to new and improved products.
As is the case for gating proprietary products, designers must refer to the manufacturer’s specifications and
literature for non-gating products to develop a good understanding of the installation requirements of each
device and its behaviour under impact. This should ensure that the most appropriate product and design are
used for any given situation.
The point of redirection for non-gating attenuators is at the nose.

Figure I.1: An example of a non-proprietary bull-nose attenuator
Note:
Original system was crash tested with timber breakaway posts.
Source: Based on Queensland Department of Transport and Main Roads (2017) Standard Drawing No. 1488.

I.1.3 Trailing Terminal
This terminal is shown in Figure I.2 and incorporates a cable anchorage. Its main function is to provide a
cable anchorage at the trailing end of the system. While this terminal will gate, it is not a crashworthy
terminal and should only be used on installations where there is no possibility of an errant vehicle from the
opposing traffic flow striking the trailing end of the barrier end-on (e.g. on a wide duplicated road). Where the
trailing end of an installation is located so that it can be impacted by opposing traffic, an appropriate
crashworthy terminal should be used.
Figure I.2: Trailing terminal

Appendix J Transitions between Barrier Types
J.1 General
Transitions are used to provide a safe interface whenever it is necessary to change from one type of barrier
to another. A satisfactory interface may be achieved by:
• providing a structurally designed and tested physical connection between the systems; the connections
are facilitated through transition sections of barrier that are designed to provide gradually increasing
lateral stiffness and hence continuity of protection for vehicles that impact the barrier in the vicinity of the
interface; transitions can be used only between semi-rigid systems (i.e. steel to steel) or between
semi-rigid and rigid systems (i.e. steel to concrete)
• overlapping the barriers by commencing the more rigid system behind the less rigid system.
The purpose of a transition section is to produce a gradual increase in stiffening between the barrier systems
so that vehicular pocketing, snagging or penetrations are prevented at any position along the transition. The
overlapping of the barriers achieves a similar outcome by providing adequate lateral separation between
them.
In practice, transitions are achieved by:

• increasing the stiffness of a barrier system
• decreasing the post spacing
• nesting one rail behind another
• using another steel section behind the W-beam
• using a heavier rail for the transition (e.g. thrie-beam).
Overlapping different types of barrier is only possible where adequate space is available to accommodate
deflections. This may be used for any system but is the only way of achieving a transition from wire rope
barrier to a more rigid barrier.
Specially designed barrier sections or connections are used for situations where W-beam is to be connected
to thrie-beam, or where either of these semi-rigid barriers is to be connected to a rigid barrier (refer to
AS/NZS 3845:1999). The latter situation typically arises on the approach to bridges that have rigid barriers
but may also occur at other locations.
J.2 Design Criteria – Physically Connected Barriers
Several criteria are important when designing a transition section or connection (AASHTO 2011a). Although
AASHTO provides this guidance in relation to bridge approaches, the following principles apply where any
semi-rigid barrier system is connected to a rigid barrier:
• The connection point of the two systems must be as strong as the approach barrier to ensure that the
connection will not fail on impact by pulling out. The use of a cast-in-place anchor or through-bolt
connection is recommended.
• The transition must be designed to minimise the likelihood of snagging an errant vehicle, including one
from the opposing lane on a two-way facility.
• When providing a transition section to a bridge railing end it is highly desirable to taper the bridge railing
end behind the approach transition to prevent pocketing on vehicle impact.
• The transition should be long enough to ensure that changes in deflection do not occur over a short
distance. The change in stiffness from the less rigid barrier to the more rigid barrier, over the transition
length, should increase with a high degree of continuity. This may be achieved by reducing the post
spacing, strengthening the rail element or a combination of these techniques.

• As with longitudinal barriers, kerb and slope features must be addressed. The slope between the edge of
the road and the barrier should not be steeper than 10:1.
• Drainage features such as kerbs, kerb inlets, raised inlets or open drains should not be constructed in
front of barriers or a transition area, as they may initiate vehicle instability and adversely affect the
crashworthiness of the barrier or transition.
J.3 Typical Interfaces between Barrier Types
J.3.1 General
AS/NZS 3845:1999 provides detailed illustrations of transitions between semi-rigid and rigid barriers. These
transitions have been tested or are deemed to be acceptable with respect to NCHRP 350 (Ross et al. 1993).
The transitions are achieved through stiffening of the steel safety barrier by the use of special sections and
connectors, and reduced post spacing and nesting of the beams (i.e. two sections of rail, one inside the
other).
Wire rope safety barriers (WRSB) are not designed to be connected to semi-rigid or rigid safety barriers or
bridge ends. However, WRSB may be transitioned to more rigid barriers provided that the WRSB overlaps
the more rigid barrier by an adequate longitudinal distance and the lateral separation is sufficient to
accommodate the maximum likely deflection of the WRSB. Such arrangements should enable the two
systems to work independently while providing continuous shielding of hazards.
The WRSB manufacturer should be consulted with regard to any proposed design transitions between
WRSB and semi-rigid or rigid barriers to seek assurance that they have either been tested or have been
otherwise demonstrated to be acceptable.
Guidance on various transitions is provided in the following sections.
J.4 W-beam to Thrie-beam
The transition is achieved through the use of a product that bolts to the W-beam at one end and to the
thrie-beam at the other end. This transition is 2 m between post centres and is illustrated in Figure F5 and
Figure F15 of Appendix F of AS/NZS 3845:1999.
J.5 W-beam to Concrete
W-beams are connected to a concrete barrier either through the use of a thrie-beam transition (Figure F5 in
AS/NZS 3845:1999) or by connecting the W-beam directly to the concrete using an acceptable direct
transition (Figure F9 in AS/NZS 3845:1999). Both treatments provide a structurally sound connection and a
smooth and stiffened transition to prevent snagging and pocketing of impacting vehicles.
The thrie-beam transition involves:

• the use of a prefabricated product to connect the W-beam to the thrie-beam (Figure F15 of
AS/NZS 3845:1999)
• the post spacing being reduced from the standard spacing of 2 m to 1 m for five spaces and then to
500 mm for the two spaces prior to the concrete barrier
• nesting of the thrie-beam over the last 4 m prior to the concrete barrier
• the use of a structure connector to bolt the thrie-beam into a recess in the concrete barrier (Figure F27 of
AS/NZS 3845:1999).

The W-beam transition (directly to concrete barrier) involves the:

• W-beam being recessed into the concrete barrier to provide a flush barrier face at the connection
• transition being strengthened by the post spacing being reduced progressively from the standard spacing
of 2 m to 1 m and then 500 mm over the last 10 m of the beam
• transition being further strengthened by nesting of the W-beam over the last 5 m
• concrete barrier being flared away from the W-beam, the latter being stiffened by circular hollow sections
bearing on the face of the concrete at the rear of the beam.
J.6 Thrie-beam to Concrete
The transition between the thrie-beam barrier and concrete barrier is achieved through the use of a structure
connector, as shown in Figure F27 of AS/NZS 3845:1999, which enables the thrie-beam to be bolted into a
recess in the concrete barrier. Details of the transition are shown in Figure F6 of AS/NZS 3845:1999. The
thrie-beam is stiffened in the manner described in Appendix J.5 above.
J.7 Wire Rope Safety Barrier to Semi-rigid Barrier
These transitions involve the WRSB overlapping the W-beam or thrie-beam barrier by a nominal longitudinal
distance based on site conditions. Where space is available the barriers can be separated laterally so that
they operate independently. An alternative acceptable arrangement (refer to Figure J.1) involves a design
that ensures that each barrier does not adversely affect the performance of the other.
J.8 Wire Rope Safety Barrier to Concrete Barrier
The transition between WRSB and the concrete barrier also requires a longitudinal overlap and lateral
separation adequate to accommodate deflections under impact. This transition has not been tested.
However, the principle of having each barrier separated by a distance that should enable them to operate
independently under impact is considered to be a sound and safe practice. An example of such a transition
is shown in Figure J.2.

Figure J.1: An example of a transition between a wire rope barrier and W-beam barrier

Figure J.2: An example of a transition between a wire rope barrier and rigid barrier
Notes:
 X is the dynamic deflection of the WRSB related to vehicles travelling along the road adjacent to the WRSB should
they impact it.
 A is the dynamic deflection for the WRSB for vehicle impacts from the opposite direction. The dynamic deflection X
varies depending on the WRSB used and the post spacing. Refer to Figure 5.7 for an approximate guide to deflections
of WRSB for concept design and planning purposes only.
Source: Based on VicRoads (2005).

Commentary 1
C1.1 Combinations of Design Parameters
The adoption of lower-order values for a number of design parameters in combination may create an unsafe
design even though the individual design parameters are in compliance with guidelines. For example,
combining a minimum radius horizontal curve with a minimum length vertical curve and narrow lanes may
produce a design combination that has a low factor of safety, even though the individual elements comply
with guidelines.
Designers should ensure that the combination of design elements makes it easy for drivers to keep their
vehicles on the road, especially at night and in inclement weather. Particular attention needs to be paid to
combinations of:
• vertical alignment
• horizontal curvature
• lane widths
• shoulder width
• sight distance
• medians
• road surface
• road surface drainage
• delineation
• verges.
Considerations on these design elements are provided in Table C1.1.

Table C1.1: Considerations in relation to designing roads for safety
Element Considerations
Vertical alignment Flat grades:
• Allow all vehicles sharing a road to travel at the same speed.
• Vertical grades of zero to 3% have little effect on the operation of all vehicles.
Uphill grades:
• Steep grades become prohibitive for heavy vehicles.
• Where steep grades are required, the design should minimise their length.
• Grades in excess of 6% have a significant effect on heavy vehicles for travel uphill.
• Vehicle speed differences also contribute to queuing, which is frustrating to drivers in the
queue.
• Where vertical curves occur in conjunction with horizontal curves extra care in design needs
to be taken. Vertical curves or changes in grade may impede sight distance.
Downhill grades:
• Grades in excess of 6% have a significant effect on heavy vehicles for travel downhill.
• Long downhill grades increase the risk of a crash due to brake failure.
• Very long downgrades, particularly those containing horizontal curves and reverse curves,
cause vehicles (especially heavy vehicles) to suffer brake fade, and reach critical roadside
features at such high speeds and impact angles that roadside protection is very difficult or
impossible, even with the best road safety barriers.
General:
Steep grades cause different vehicles to travel at different speeds, introducing a higher risk of
rear-end crashes.
Where it is necessary to provide long, steep grades, consider providing:
• passing bays and descending lanes to allow light vehicles to overtake slower-moving vehicles
safely
• safety ramps and arrester beds to bring a runaway vehicle to rest.
These facilities are important where there is a high proportion of heavy vehicles.
Adequate drainage must be provided to prevent water pooling on the road surface during
normal levels of rain. Longitudinal drains need to have adequate fall, generally not less than
0.5%.
Horizontal Design for a radius as large as the landscape allows is a first step in providing a driveable path.
alignment Provide a consistent alignment standard over the relevant section of road with well-designed
transitions where reductions from generous to tighter alignments are necessary.
For a vehicle to travel around a bend at a certain speed, the horizontal friction between the
vehicle and the road pavement must be sufficient to counteract the inertial force that tends to
maintain the vehicle’s initial direction.
Provide localised curve widening where required (refer to AGRD Part 3 (Austroads 2016a)).
Extra lane width at curves maintains an acceptable clearance between vehicles. Road
pavement widening may be required at curves depending on curve radius, lane width and
vehicle sizes for the following reasons:
• A vehicle (particularly a heavy vehicle) traversing a curve occupies more lane width than
when travelling straight.
• When a driver steers through a curve, the vehicle does not maintain the same lateral road
position that it did on the straight. Some deviation from the ideal path must be expected.
Both of these factors reduce the clearance between opposing vehicles and increase the risk of
collision. Recommended curve widening for various categories of roads is given in AGRD Part 3
(Austroads 2016a).

Lane width Traffic lane width influences the ease with which vehicles can operate in that lane.
Higher traffic volumes and higher speeds require wider lanes to allow more space between
passing vehicles, and between vehicles and any roadside objects.
If lane width is insufficient, vehicles may be forced into the roadside because of:
• the blast of air pressure created by large vehicles passing in the opposing traffic lane
• vehicles (particularly articulated vehicles) in the adjoining or opposing traffic lane swaying into
the incorrect traffic lane.
In some instances, lightweight vehicles have been sucked into the slipstream of passing
vehicles, and the evasive action by drivers of the lightweight vehicles has caused the lightweight
vehicles to encroach into the roadside. Recommended lane widths for various categories of
roads are given in AGRD Part 3 (Austroads 2016a).
Shoulder width Shoulders not only provide a lateral support for the road pavement but also provide additional
separation between traffic and roadside objects.
Shoulders are not intended for regular travel but allow drivers more room to bring their vehicles
back under control after inadvertently leaving the traffic lane.
Sealed shoulders are desirable to assist errant vehicles to recover should they leave the
travelled path and also reduce the incidence and severity of run-off-road crashes.
The width of shoulder sealing depends on traffic speed, volume and composition, environmental
conditions and the nature of the roadside area. Sealing part of the shoulder to reduce pavement
drop-off may reduce errant vehicle incidents. Recommended shoulder widths for various
categories of roads are given in AGRD Part 3 (Austroads 2016a).
Sight distance Adequate sight distance should be provided to allow road users to safely negotiate the road.
Sight distance can be affected by road geometry (horizontal and vertical alignment), terrain
(particularly inside of horizontal curves) and roadside objects (such as trees and signs).
At a horizontal curve, drivers need to be aware of the road curvature ahead and be able to react
and slow down (if necessary) to safely navigate the curve.
Drivers should be able to see a sufficient length of horizontal curve in order to judge its
curvature and safely navigate the curve.
A curve should not commence just over the crest of a hill. However, where this situation is
unavoidable, measures should be taken to ensure that drivers are made aware of the alignment
ahead (e.g. sight distance, warning signs, delineation).
Roadside features such as cutting slopes and vegetation may limit sight distance and should be
modified or removed to ensure sufficient stopping sight distance on curves. If this is not
practical, the speed limit may have to be reduced to compensate.
It is important that roadsides be maintained to ensure that sight distance requirements are
sustained, for example by regularly pruning trees and cutting grass.
Where substandard curves are unavoidable, consider cutting benches in high batters to improve
sight distance. Recommended sight distance requirements are given in AGRD Part 3
(Austroads 2016a).
Medians Median width influences the crossover crash rate on medians without road safety barriers.
Cross-median crashes are often high severity, head-on crashes (Knuiman, Council &
Reinfurt 1993). Jurisdictional policy should be consulted regarding the current approach for
median design and protection by barriers.

Road surface A road surface should be constructed and maintained to a sufficient standard that ensures
adequate skid resistance. The skid resistance of a particular surface results from the surface
texture and the presence of moisture (e.g. a pavement surface that holds water instead of
draining properly can contribute to vehicles aquaplaning).
The condition of an existing pavement can be determined by conducting skid resistance
measurements as well as assessing the level of rutting and occurrence of potholes.
Measurement of skid resistance and rutting can be undertaken using a number of methods,
some of which are highly automated and efficient. The decision to act on the results of such
measurements is left to the experienced practitioner; however, a guide to the use of skid
resistance values can be found in the Guide to Asset Management Technical Information Part
15: Technical Supplements (Austroads 2018c).
Roads with a comparatively high volume of heavy vehicle traffic (usually major link routes) may
require a higher standard of construction and maintenance than roads that predominantly carry
light vehicles such as cars and vans.
Unevenness and rutting of road surfaces can cause motorcyclists to abruptly change from their
intended cornering line which may result in crashes.
Road surface A number of different aspects need to be considered with regard to drainage. These include:
drainage • drainage of the road pavement by providing adequate grade and crossfall so that the
pavement is able to drain and pooling of water is avoided, which allows maintenance of skid
resistance
• appropriate infrastructure to collect and transfer the water from the pavement, which may
include kerb and channel or table drains
• a road reservation that can accommodate water run-off from adjacent land uses.
Drainage design at the road design stage requires consideration of flood estimation. If
constructed along a flow path, a road may need to be designed to accommodate the run-off
from adjacent land for a flood event (refer to the Guide to Road Design Part 5: Drainage:
General and Hydrology Considerations (Austroads 2013a)).
Where concrete barriers are installed it is essential that adequate provision is made for water to
drain from the road (i.e. under or through the barrier) so that ponding of water does not occur
adjacent to the barrier.
Delineation and The more unexpected aspects of a road’s geometry will require additional signing and
signposting delineation to convey information to drivers in accordance with AS 1742.2-2009 or NZ Transport
Agency (2010b) and jurisdictional guides.
Guide posts show the edge of the road and enhance delineation of the path to be travelled by
drivers. They should be installed at a uniform distance from the edge of the road and should be
fitted with delineators. On narrower or lower-volume roads where there is insufficient road width
to mark a centreline, guide posts may be the only delineation provided.
In areas above the snow line, there is a risk that raised pavement markers could be damaged by
snowploughs or obscured by snow. For this reason, it is recommended that orange snow poles
are used for delineation. Snow poles are designed to protrude above snow drifts and their
orange colour aids visibility in snow.
Verges It is preferable that verges and roadside areas be free of hazards and are traversable. Where
this is not possible a risk assessment should be undertaken to determine the most appropriate
treatment (e.g. removal, modification or shielding of the hazard).
C1.2 Consistent Design Environment
Safety on roads is closely related to the driver’s ability to anticipate events and react to them. Where drivers
are uncertain of what lies ahead their perception and reaction times will be longer than in situations where a
consistent design environment gives them confidence in what to expect.
A road has features, both on-road and on roadsides that clearly show drivers the path that a road takes and
helps them keep their vehicles in the running lane. Road design should therefore be based on appropriate
driver characteristics with the objective of making it as easy as possible for the driver to keep the vehicle on
path. This may be straightforward if the landscape is always suited to the desired path of the road and there
is no space or financial constraints, but this is rarely the case.

Roads must be contained within the topography in a cost-effective way and this may lead to situations that
require departures from the standard. In such cases it becomes necessary to provide additional features
designed to help drivers follow the line of the road (e.g. vegetation, signs and delineation).
In order to give motorists the best chance of keeping their vehicles on the road, it is necessary to provide a
geometric design conducive to safe travel. The principal factor influencing a vehicle’s ability to traverse and
remain on a road is the speed of the vehicle. Accordingly, it is necessary to take into account the operating
speed of a road when setting such parameters as curve radii, lane widths, shoulder widths, seal types,
drainage and vertical alignment. Designers should refer to AGRD Part 3 (Austroads 2016a) for guidance on
geometric design elements.
Design consistency needs to be considered in relation to all relevant design elements and considerations.
Key examples are summarised in Table C1.2.
Table C1.2: Key considerations for consistent design
Consideration Comment
Cross-section Cross-section dimensions should be compatible with horizontal and vertical alignment.
consistency (e.g. improving cross-section dimensions while retaining poor alignment can create the
hazardous illusion that the road can be driven at a higher speed than is safely possible).
Where cross-sections change (e.g. where a divided road link joins an undivided link) generous
tapers and advance signing should be provided to make the change obvious.
Operating Differing speeds in the traffic stream can be caused by:
speed • an unclear road hierarchy
consistency
• drivers being unsure of what lies ahead
• drivers having differing levels of confidence in negotiating road geometry with low design
values.
Road networks that do not provide a hierarchy of road functions can cause speed differentials
because local, short-trip traffic is mixed with high-speed through traffic.
The greater and more frequent the speed differentials between vehicles, the greater is the
chance of crashes.
Driver workload Abrupt changes in driver workload may influence crashes because driver response to
consistency situations may be slow or inappropriate. If driver workload is:
• too low then drivers may become inattentive
• too high then drivers begin to shed information (look but not see). Some of the shed
information may be critical (e.g. other vehicles entering the travel path).
Increases in driver workload may be caused by:
• limited sight distance
• inconsistent design, causing surprise (e.g. a sharp curve at the end of a long straight)
• the driver being unfamiliar with the road (e.g. on infrequently travelled highways).
Design aspects that affect driver performance to be considered include:
• Avoid low arousal straight alignments. A curve with a very large radius (i.e. almost straight)
will be monotonous for drivers leading to a lack of concentration on the steering task).
• Avoid the concentration of decisions into a short time frame as this will create information
overload (e.g. excessively complex intersection layouts).
• Stage speed changes (e.g. change the design speed of geometry in steps rather than
implement abrupt changes from high speed to low speed).
• Provide rest areas to cater for fatigued drivers.
C1.3 Vehicle Mix Considerations
Total travel by trucks is growing at a rate twice that for cars and recent traffic growth estimates indicate that
this growth is likely to continue for at least another 15 years. A significant proportion of trucks are articulated
heavy vehicles. It is important to consider the impact and additional risk of a higher than normal percentage
of heavy vehicles.

The consequences of crashes involving heavy vehicles are much greater than for cars. A run-off-road crash
involving a truck may kill or injure the truck driver and passengers, but in a catastrophic crash involving a
truck and cars the consequences to the car occupants and other people outside the truck (e.g. pedestrians)
are the major effects.
In the case of a bus, a crash is more likely to injure persons outside the bus. However, a catastrophic crash
involving a bus may kill or injure many of its passengers.
When designing for heavy vehicles, designers should examine the road design parameters shown in
Table C1.3.
Table C1.3: Road design parameters for consideration in relation to heavy vehicles
Parameter Consideration
Grades Weight and low power-to-weight ratio cause heavy vehicles to slow below the speed limit on
grades and cause problems with faster traffic.
Acceleration lanes Heavy vehicles accelerate slower than cars and need longer to reach a target speed.
Forcing trucks to merge with main traffic lanes too early can adversely affect faster traffic.
Lanes that merge with high-speed roads should be long enough to allow for heavy vehicle
acceleration. A downgrade on these lanes will help.
Curve radii Longer vehicles may encroach into the adjacent lane on corners that have small radii or
narrow lanes.
The roll stability of heavy vehicles is less than cars, because of their weight and higher centre
of gravity. A heavy vehicle is more likely to roll than to skid in a tight corner.
Designers should provide the largest curve radii consistent with the environment.
Stopping and sight Although the eye height of a heavy vehicle driver is higher than that of a car driver, truck
distances stopping distances are considerably longer because of the relatively inferior braking of heavy
vehicles thus requiring longer sight distances.
C1.4 Other Factors Contributing to Errant Vehicles
Other road design issues that may contribute to errant vehicles are outlined in Table C1.4.

Table C1.4: Other factors contributing to errant vehicles
Factor Contribution to
Lack of overtaking lanes Drivers overtaking ill-advisedly and causing:

• vehicles to swerve into the roadside to avoid head-on crashes
with oncoming vehicles
• head-on crashes
• vehicles hitting objects in the median.
Unsealed shoulders Causing a vehicle to lose control and travel further into the roadside
making it more likely to impact a hazard or overturn.
Rounded pebbles encroaching on sealed or
unsealed shoulders
Clear roadside areas with surface that is either

rutted or covered with rounded pebbles
Pavement edge drop-off Vehicles becoming errant or over-correcting to regain the pavement
and swerving across the road into oncoming traffic.
Crack filling with slippery, hard, raised lines Loss of friction and loss of control.
Motorcycles being thrown off line.
Forming slippery, hard, abrupt humps causing motorcycles to bump.
Settlement of the roadway behind bridge Deep depressions in the roadway and sharp transverse edges at
abutments abutments, causing a change in the line of travel.
Road patching with uneven, raised or Motorcycles having to change their line of travel.
depressed edges and surfaces
Mounding and/or cracking of road surface due

to tree roots
Raised service covers (e.g. water, sewerage,

communications)
Non-standard raised pavement markers
Gravel on the road Loss of friction causing vehicle slide.
Sun glare and dust Blinding of drivers causing swerving, or ill-advised overtaking and:
• causing vehicles to swerve into the roadside to avoid head-on
crashes with oncoming vehicles
• causing vehicles to hit objects in the median.
Cross-winds Throwing vehicles seriously off line, causing a crash.
Overhanging vegetation May:

• obstruct sight distance so the driver has no warning of the road
alignment ahead or of slower-moving vehicles in the lane ahead
• be struck by a motorcycle rider’s head.
This in turn can cause:
• vehicles to brake excessively and veer into the roadside
• rear-end crashes.
[Back to body text]

Commentary 2
Key factors that may contribute to run-off-road crashes are:

• Road geometry, including sight distance. Vehicles are more likely to leave the road at curves that have
small radii or inadequate pavement crossfall, particularly at curves with radii inconsistently smaller than
those of preceding curves or at curves with restricted approach sight distance.
• Traffic volume and speed. Drivers are more likely to leave the road when performing avoidance
manoeuvres on high-speed, high-volume roads, especially two-lane rural roads that have limited
overtaking opportunities.
• Driver inattentiveness, fatigue and lack of awareness of the road environment. Drivers who are tired,
inattentive or unfamiliar with the road are more likely to leave the road than alert drivers. Thus, long
distance routes in monotonous terrain or roads that are inconsistent with the terrain require special
consideration.
• Adequacy of visual cues of road alignment, including delineation. Lack of adequate edge delineation or
misleading cues because of gaps in vegetation or lines of service poles may increase the risk of drivers
leaving the road.
• Number and frequency of decisions required of the driver. Drivers are more likely to make mistakes and
leave the road in complex situations requiring many decisions in rapid succession, especially if visual
cues are insufficient or misleading.
• Road surface condition. Drivers are more likely to leave the road if a large part of their attention is
devoted to negotiating a poor road surface, or if they suddenly encounter loose or slippery surfaces.
• Weather. Rain, frost, snow, fog, wind gusts and sun glare reduce the effective control drivers can exert on
the paths of their vehicles and thus increase the risk of encroachments.
• Mechanical failure.
However, the likelihood of a vehicle leaving the road may be minimised by implementing appropriate
measures.
[Back to body text]
Commentary 3
C3.1 Background to Clear Zone Width
The clear zone concept was first introduced in the USA in the 1970s. The concept was if a clear area was
provided alongside the road then most drivers would be able to regain control and come to a safe stop. While
this is true in most cases there are a significant number of cases where vehicles have travelled beyond this
clear area.
Reports on studies into roadside run-off-road crashes (e.g. Hutchinson & Kennedy 1966) stated that an
unencumbered 9 m wide corridor beside the travelled lane permits about 80% of the out-of-control vehicles
leaving a high-speed roadway to recover. It is understood that the original dimensions were obtained from
incidents at the US General Motors proving ground (Stonex 1960). This width did not take into account
roadside geometrical factors that may determine the extent of vehicle run-off-road incidents, such as slopes,
bends or individual road characteristics.

Kloeden and McLean (1999) conducted a study of roadside hazard involvement in fatal and serious injury
crashes in South Australia. Analysis of fatal crash records for the 12-year period from 1985 to 1996 revealed
that 95% of fatal crashes involving a collision with a roadside object occurred between 0 and 10 m adjacent
to the road.
Austroads (2018) have noted that research by Jurewicz and Pyta (Austroads 2011) indicated that objects are
continuing to be struck at very large clear zone widths and a high threshold for the road departure to the left
casualty crashes is observed. For rollover casualty crashes, the proportion appears to increase with very
wide clear zones.
Austroads (2011) concluded that:

Clear zones should now be considered in the following light:
• Clear zones cannot deliver Safe System outcomes in isolation and should be
regarded as a supporting treatment.
• Some clear zone is better than none at all when continuous lengths of barrier
cannot be installed.
• Clear zones should be regarded as having the potential to be a hazard in their

own right in the same way that barriers are afforded this attention.
It is now considered that the clear area alongside the road should be treated as a mitigating treatment and
that designers do not consider only those hazards within this area, as this does not adequately evaluate the
risk to motorists.
[Back to body text]
Commentary 4
C4.1 Embankment Slopes
Some steep embankment slopes may not constitute a direct hazard but prevent errant vehicles from
recovering when they leave the road and run onto the embankment. The condition of the ground surface
may increase the hazardousness of an embankment. The propensity for a vehicle to rollover is increased on
steeper embankments.
Thomson and Valtonen (2002) describe the results of some crash testing of drain shapes which indicates
that, depending on the angle of departure of the vehicle from the road, shapes outside of those shaded in
Figure 4.18 and Figure 4.19 in AGRD Part 3 (Austroads 2016a) are traversable.
Thomson and Valtonen (2002) reported that the collision severity of vehicles travelling in a V-shaped ditch
was not appreciably worse than the loading measured in standardised testing of load restraint systems, as
long as a rollover did not occur. The rollovers observed tended to be quite violent even for the lowest speed
tests (80 km/h).
The impact severity of a V-shaped ditch is dependent on the change of slope at the V. Vehicles will plough
into the back slope if this change in slope is too great.

A significant risk not measured in these (or similar) tests was the consequence of a vehicle travelling over
the backslope and continuing into the roadside terrain. The backslope used in these tests was 1 m higher
than the road, which was not sufficient to contain vehicles to the ditch. The speed was observed to be not
significantly reduced as the vehicle exited the ditch. Often the vehicle was airborne as the backslope acted
as a ramp. Subsequent impact with a pole, tree, or rock located beyond the ditch could have severe
consequences for the vehicle trajectories observed in the tests.
Transverse embankments with relatively flat side slopes may cause vehicular vaulting with the vehicle and
becoming airborne. For steeper slopes, the vehicle bumper may ‘catch’ in the slope resulting in an abrupt
stop and high occupant decelerations. For these reasons, transverse slopes should be as flat as practical.
A preferred option may be to flatten the slope to 6:1 to make it ‘recoverable’. An economic evaluation of
flattening the embankment, compared to installing a barrier, may inform the designer. This evaluation should
include the costs associated with crashes, maintenance and installation or construction for each option.
[Back to body text]
Commentary 5
While it is preferable to have a smooth, flat face on crash barriers there is evidence to suggest that some
vertical relief to enhance the appearance of barriers is acceptable (FHWA 2002). Based on a review of
submitted information and test results, the FHWA accepted that the guidelines for concrete barrier texturing
described below are acceptable and would not adversely affect the NCHRP Report 350 (Ross et al. 1993)
test level of the barrier to which a texture or pattern is applied.
On the basis of the FHWA acceptance, road agencies may consider the use of textured surfaces on concrete
barriers provided that they comply with the following requirements:
• Sandblast textures with a maximum relief of 9.5 mm.
• Images or geometric patterns inset into the face of the barrier 25 mm or less and having 45 degree or
flatter chamfered or bevelled edges to minimise vehicular sheet metal or wheel snagging.
• Textures or patterns of any shape and length inset into the face of the barrier up to 13 mm deep and
25 mm in width.
• Any pattern or texture with gradual undulations that have a maximum relief of 20 mm over a distance of
300 mm.
• Gaps, slots, grooves or joints of any depth with a maximum width of 20 mm and a maximum surface
differential across these features of 5 mm or less.
• Any pattern or texture with a maximum relief of 64 mm, if such pattern begins 610 mm or higher above
the base of the barrier and all leading edges are rounded or sloped to minimise any vehicle snagging
potential. No part of this pattern or texture should protrude above the plane of the lower, untextured
portion of the barrier.
The FHWA approval also concluded that:

• Texture or pattern meeting the guidelines can be applied to all crashworthy single slope or vertical wall
designs.
• It was clear from the crash test results that textured barriers can result in more vehicular body damage in
a crash due to increased friction even if their crash performance remains within acceptable limits.
• Although the barriers tested were 1220 mm and 1422 mm high, review of the crash and post-crash
vehicle trajectories indicated that these guidelines may also be applied to vertical walls as low as 685 mm
and to any single-sloped barrier at the standard 813 mm height or higher.

• These treatments may prove acceptable on New Jersey and F-shape concrete barriers if the treatment is
applied only to the upper sloped face of the barriers, but some crash testing would be advisable to verify
good performance with these shapes.
[Back to body text]
Commentary 6
C6.1 General
Section 5.3.4 discusses the lateral placement of road safety barriers in relation to the road. It also introduces
the issues surrounding the placement of barriers on embankments, behind kerbs and on cutting slopes.
When cars pass over embankment slopes, kerbs or cutting slopes their front bumper height may follow a
trajectory that is lower or higher than the normal static bumper height.
The trajectory is important should it be necessary to locate a barrier within particular lateral limits with
respect to the back of a shoulder or kerb of a road because the barrier may have to be set at a height
(i.e. higher or lower than normal height if installed on a relatively flat surface) that will contain an errant car. If
set too high the car may become snagged or pass under the barrier and if set too low it may vault over the
barrier.
The Guide to Road Design Part 2: Design Considerations (Austroads 2019) discusses the concept of normal
design domain (NDD) and extended design domain (EDD). It is emphasised that the placement of a barrier
on embankment slopes, behind kerbs and on cutting slopes within the lateral distances described as not
recommended in Figure C6.4 falls within the realm of EDD.
In constrained locations at greenfield sites (and particularly at brownfield sites), it may not always be
practical or possible to achieve all the relevant NDD values (and practice). In these constrained locations,
road agencies may consider the use of values outside of the NDD.
In applying this Part, NDD values (and practice) given in the body of this Part should be used wherever
practical.
Design values (and practice) outside of the NDD are only to be used if approved in writing by the delegated
representative from the road agency. The road agency may be a state/territory road agency, municipal
council or private road owner.
If using EDD values, the reduction in the standard associated with their use should be appropriate for the
prevailing local conditions. Generally, EDD should be used for only one parameter in any application and not
be used in combination with any other minimum or EDD value for any related or associated parameters.
The following sections provide examples of vehicle trajectories over embankments, kerbs and cutting slopes.
The examples are obtained from Roads and Maritime Services (2016). A summary of limitations on barrier
location in these situations is provided in Commentary C6.5.

C6.2 Vehicle Trajectory Over Embankments
When vehicles pass over embankments, even at moderate speed, the bumper trajectory rises above normal
bumper height, as illustrated in Figure C6.1, and this can cause a vehicle to vault over an incorrectly placed
road safety barrier. The rise in bumper level is not significant for embankment slopes of 10:1 but becomes
significant at steeper slopes. Barriers should therefore be located between the traffic lane and the
embankment hinge point. If this is not possible, the barrier may be placed up to 0.5 m beyond the hinge
point.
If there is no alternative than to place a barrier on an embankment, it must be located beyond distance L in
Figure C6.1, the point at which the bumper returns to its static height. This distance varies with design speed
and batter slope and an example of distances relating to embankment slopes and encroachment angles are
shown in .
Table C6.1. It is also desirable that the batter be rounded at the hinge point to reduce the effect of the
change in slope on vehicle dynamics.
Figure C6.1: An example of bumper height trajectory characteristics over fill embankments
Note: See Table C6.1 for measurements.

Table C6.1: An example of bumper trajectory data over embankments at 100 km/h
Encroachment Embankment
L ΔHO ΔHM LM d
angle slope S2
(m) (mm) (mm) (m) (m)
(degrees) (S:1)
25 10:1 5.0 30 30 4.0 2.0
25 6:1 6.8 60 130 4.9 3.8
25 4:1 9.2 100 340 6.1 6.2
25 3:1 11.5 140 650 7.3 8.5
25 2:1 16.2 230 1550 9.6 13.2
15 10:1 3.7 10 10 3.4 0.7
15 6:1 4.4 40 50 3.7 1.4
15 4:1 5.3 90 130 4.2 2.4
15 3:1 6.2 130 240 4.6 3.2
15 2:1 8.0 210 580 5.5 5.0
Note: See Figure C6.1 for description of the measurements.
C6.3 Vehicle Trajectory Over Kerbs
When vehicles pass over kerbs at speed they are subjected to an upward force such that pitch and roll will
be developed. The combination of these effects will cause the vehicle bumper to follow a trajectory that will
lead it to being higher or lower than its normal position relative to the wheels and the bumper trajectory may
rise above normal bumper height as illustrated in Figure C6.2, and this can cause a vehicle to vault over an
incorrectly placed road safety barrier. Barriers should therefore be located close to the back of the kerb or at
a sufficient distance further behind the kerb where the bumper height has returned to normal level.
The trajectory of the bumper depends on the:

• size and suspension characteristics of the vehicle
• vehicle impact speed and angle
• height and shape of the kerb.
Table C6.2 shows an example of data relating to Figure C6.2. However, it should be noted that the trajectory
profile shown in the figure and the data in the table are not based on current vehicle fleet characteristics
which may behave differently in traversing kerbs.
An understanding of the vehicle behaviour (i.e. bumper trajectory) is important in locating the barrier because
the:
• Lowering of a bumper may cause a vehicle to snag on the underside of the barrier rail within the distance
L1 in Figure C6.2.
• Rise of the bumper may cause it to ramp and vault over the rail. This effect (i.e. rise) is greater for vertical
faced kerbs than for semi-mountable kerbs and can be in excess of 200 mm depending of the type of
kerb, and the speed and impact angle of the vehicle.

Figure C6.2: An example of bumper height trajectory characteristics over kerbs
Table C6.2: An example of bumper trajectory data over specific kerbs
Roads and
Traffic Speed Angle L1 ΔH1 (max) L2 ΔH2 (max)
Authority (km/h) (degrees) (mm) (mm) (mm) (mm)
kerb type
80 10 575 150 1550 140
80 25 1275 150 4500 180

SA/SM/SL(1)
100 10 750 150 1750 190
100 25 1625 150 5400 190
90 12.5 1525 140 2900 50
90 20 1950 165 4200 85

SE/SF(2)
100 12.5 1775 165 3750 85
100 20 2075 160 4950 85
1 These are forms of vertical faced kerb.

2 These are forms of semi-mountable kerb.
The following guidance should be considered in the use of barriers in conjunction with kerbs:
• The kerb should not be located in front of or under semi-rigid or flexible barriers on high-speed roads; a
drain located behind the barrier, a shallow gutter immediately in front of the barrier, or a subsurface
grated drainage system are the preferred drainage solutions. Crash tests have shown that the use of any
barrier/kerb combination where high-speed, high-angle impacts are likely should be discouraged. Where
there are no feasible alternatives, AASHTO (2011a) suggests that designers should consider using a kerb
no higher than 100 mm and consider stiffening the barrier to reduce potential deflection.
• Rather than locating a kerb close to the face of a rigid barrier, drainage should be facilitated by the face of
the barrier.

• Where a kerb must be used in conjunction with semi-rigid or flexible barrier, as is often the case in urban
situations, it is desirable that it is placed either within the distance L1 of the kerb or beyond distance L2
shown in Figure C6.2; however, the latter location may be impractical in urban situations.
• To ensure satisfactory barrier performance, it is preferred that the barrier is set back no more than the
distances shown in Table 5.3. An offset of this magnitude should also minimise nuisance damage to
barriers in low-speed urban situations. A semi-mountable kerb is preferable in these situations, and a
vertical-faced kerb should preferably only be used in speed zones ≤ 60 km/h.
• In spite of the above guidance, it is sometimes necessary in urban areas where the speed zone is
≤ 80 km/h to place a barrier behind a footpath and this results in the barrier being located a relatively
large distance (and perhaps within distance L2) behind the kerb. Furthermore, in placing barriers on these
urban roads, consideration should also be given to the possible adverse effect on traffic flow of a long
barrier being placed immediately behind the kerb.
C6.4 Vehicle Trajectory Over Cutting Slopes
When a vehicle runs up a cut batter, the momentum of the body on the front suspension causes the bumper
height to be significantly lower than the normal bumper height, as shown in Figure C6.3. The reductions in
bumper height can be significant enough (e.g. 200 to 300 mm depending on the vehicle type, speed and
batter slope) to cause a vehicle to run under a semi-rigid or flexible barrier. A barrier should therefore not be
located in the area defined by L in Figure C6.3 and an example of the distances for specific kerb types is
shown in Table C6.3.
Figure C6.3: An example of bumper height trajectory characteristics on cutting slopes

Table C6.3: Bumper trajectory data over cutting slopes
Speed Angle Batter slope of L ΔH (max)

(km/h) (degrees) cutting (mm) (mm)
80 10 2.6:1 1375 285
80 15 2.6:1 1675 320
90 7 8:1 1775 115
90 15 8:1 2450 170
100 7 8:1 1925 120
100 15 8:1 2600 185
C6.5 Summary of Limitations on Barrier Location
Figure C6.4 shows an example of the preferred locations of a barrier and locations where barriers are not
recommended with respect to the behaviour of vehicles passing over embankments, kerbs and cutting
slopes.

Figure C6.4: Summary of barrier locations – preferred and not recommended
[Back to body text]

Commentary 7
Commonly used crash cushions and impact attenuators use one of two principles to absorb the energy of
impacting vehicles at a controlled rate:
• The kinetic energy principle whereby the kinetic energy of an impacting vehicle is absorbed by crushable
or plastically deformable materials or by other energy absorbers. Some of the energy is also dissipated by
the crushing of the front end of the colliding vehicle. This type of system requires a rigid back-up or
support to resist the collision force of the vehicle, usually in the form of a ground anchor or other linkage
back-up (such as part of the barrier), or both. This type of system is generally referred to as a
compression system (AASHTO 2011a).
• The conservation of momentum principle where the end treatment design involves the transfer of the
momentum of an impacting vehicle to an expendable mass (usually sand) located in the vehicle’s path.
This type of system is generally referred to as an ‘inertial road safety barrier’ (and is a gating device). No
rigid back-up is required for this type of system since the energy of the vehicle is not absorbed but
transferred to other masses such as sand (AASHTO 2011a).
[Back to body text]
Commentary 8
An escape ramp may be provided on a descending, horizontal or ascending grade as illustrated in

Figure C8.1. It requires the use of single sized or uniform graded aggregate to prevent compaction in an
arrester bed to increase rolling resistance and, therefore, slow the vehicle. The descending-grade ramps can
be rather long because the gravitational effect is not acting to help reduce the speed of the vehicle.
For the horizontal-grade ramp, the effect of the force of gravity is zero and the increase in rolling resistance
has to be supplied by an arrester bed composed of single sized or uniform graded aggregate to prevent
compaction. This type of ramp will be longer than those using gravitational force acting to stop the vehicle.
The ascending-grade ramp uses both the arresting bed and the effect of gravity, in general reducing the
length of ramp necessary to stop the vehicle. The loose material in the arresting bed increases the rolling
resistance, as in the other types of ramps, while the force of gravity acts downgrade, opposite to the vehicle
movement. The loose bedding material also serves to hold the vehicle in place on the ramp grade after it has
come to a safe stop. Ascending grade ramps without an arresting bed are not encouraged in areas of
moderate to high commercial vehicle usage as heavy vehicles may roll back and jack-knife upon coming to
rest.

Figure C8.1: Types of vehicle escape ramps
Each one of the ramp types is applicable to a particular situation where an emergency escape ramp is
desirable and must be compatible with the location and topography. The most effective escape ramp is an
ascending ramp with an arrester bed. On low-volume roads of less than approximately 1000 vehicles per
day, clear run-off areas without arrester beds are acceptable.
[Back to body text]

Guide to Asset Management Technical Information Part 15: Technical Supplements (Austroads
2018b)

AGRD06-20 Guide To Road Design Part 6 Roadside Design Safety and BarriersEd3.1

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

AGRD06-20 Guide To Road Design Part 6 Roadside Design Safety and BarriersEd3.1

Uploaded by

Copyright:

Available Formats

Guide to Road Design Part 6: Roadside Design,

Safety and Barriers

Third edition prepared by: Professor Rod Troutbeck Publisher

ISBN 978-1-922382-44-3 © Austroads Ltd 2020 | This work is copyright. Apart

1. Introduction to Roadside Design .......................................................................................................... 1

2. Network Risk Assessment .................................................................................................................. 28

Austroads 2020 | page i

2.4.3 Setting a NRRIT Based on Two-lane Undivided Rural Roads .............................................................. 31

3. Program and Project Risk Assessment ............................................................................................. 35

5. Road Safety Barriers ............................................................................................................................ 61

Austroads 2020 | page ii

5.3.17 Working Width (Steps B9 and 10) ......................................................................................................... 99

6. Roadside Design for Steep Downgrades ......................................................................................... 122

7. Work Zone Safety Barrier Systems .................................................................................................. 133

References .................................................................................................................................................... 134

Appendix A Terminology......................................................................................................................... 139

Appendix B Detailed Risk Evaluation Procedure ................................................................................. 144

Austroads 2020 | page iii

B.6 Risk Score ............................................................................................................................................ 158

Appendix C Treatment of Roads Based on Jurisdictional Policies ....................................................174

Appendix D Risk Score Charts ............................................................................................................... 176

Appendix E Cost of Impacts ................................................................................................................... 187

Appendix F Treatments for Brownfield Sites........................................................................................ 188

Appendix G Angle of Departure Method ................................................................................................ 193

Appendix H Examples of Length of Need Calculations ....................................................................... 198

Appendix I Types of Safety Barrier Terminals .................................................................................... 207

Appendix J Transitions between Barrier Types ................................................................................... 212

Austroads 2020 | page iv

Commentary 1 ............................................................................................................................................... 217

Austroads 2020 | page v

Figure 1.1: Flow chart of the Guide to Road Design ..................................................................................... 3

Austroads 2020 | page vi

Austroads 2020 | page vii

Figure A.2: Road safety barrier terminology – curvature ........................................................................... 142

Austroads 2020 | page viii

1. Introduction to Roadside Design

1.1 Context Sensitive Designs

Austroads 2020 | page 1

The collective purpose of these documents is to:

Austroads 2020 | page 2

Figure 1.1: Flow chart of the Guide to Road Design

1.3 Reading this Part in the Context of Part 1

Austroads 2020 | page 3

1.3.1 Combining Design Parameters and Consistent Design

Further information is provided in Commentary 1.

1.3.2 Progressive Safety Updates to the Guide to Road Design

Austroads 2020 | page 4

1.4 Scope of this Part

Specifically, this Part provides:

This Part also covers the:

Austroads 2020 | page 5

1.4.1 AGRD Part 6A: Paths for Walking and Cycling

1.4.2 AGRD Part 6B: Roadside Environment

1.5 Principles Considered in Roadside Design to Achieve the Safest System

A transparent, principles-based approach to decision-making should be adopted during design. This

Austroads 2020 | page 6

1.5.1 Safe System Principle

From an infrastructure perspective, the Safe System approach involves:

1.5.2 Road Network Efficiency Principle

Austroads 2020 | page 7

1.5.3 Community Wellbeing Principle

1.5.4 Environmental Sustainability Principle

1.5.5 Utility Services Principle