High-risk rural roads guide

Interim draft for use April 2011

NZ Transport Agency
Foreword - High-risk rural roads guide

Crashes on rural roads are a serious issue in New Zealand. The social and economic costs are high. The
Government’s road safety strategy Safer Journeys signals that more must be done to improve safety on our
high risk rural roads. Crashes on rural open roads (state highways and local roads with speed limits of
80km/h or more) accounted for 53 percent of all New Zealand fatal and serious road crashes for the five
year period to 2009.

Safer Journeys vision is "a safe road system increasingly free of death and serious injury". The strategy gives
us a road map for focusing our efforts where the greatest gains can be made. Roads and roadsides are an
area of high concern, and high risk rural roads are identified as requiring early action under the strategy.

Safer Journeys introduces the Safe System approach that represents a fundamental shift in the way we
think about, and act upon, road safety. Human beings make mistakes and although crashes are inevitable,
in a safe system they are less likely to result in death and serious injury.

This High-risk rural roads guide is a flagship Safer Journeys initiative. It is a practical guide to making our
roads safer, intended for use by all road controlling authorities that manage high risk rural roads. This is a
draft Guide and we are seeking feedback to finalise it.

The High-risk rural roads guide introduces a new way to identify high risk road sections and, using the
Safe System approach, provides best practice guidance on choosing effective countermeasures.

Our traditional approach to road safety has helped achieve our current levels of road safety. We now need
to add to this mix the Safe System approach, where road designers and users share responsibility for a
system to protect road users from death and serious injury.

If you are involved in managing a rural road network, I encourage you to consider how applying the High-
risk rural roads guide will change what you do, and to send us your feedback on the Guide using the forms
provided on our website at www.nzta.govt.nz/consultation by 30 May 2011.

The finalised Guide will be published in June 2011. While the High-risk rural roads guide is currently a
draft guide, approved organisations can begin using it now to inform the 2012-15 National Land Transport
Programme.

Geoff Dangerfield
Chief Executive
New Zealand Transport Agency
 High-risk rural roads guide

Contents
1 INTRODUCTION AND OBJECTIVES _____________________________________________________ 1
1.1 Purpose ____________________________________________________________________ 1
1.2 Scope ______________________________________________________________________ 1
1.3 Target audience ______________________________________________________________ 2
1.4 Definitions __________________________________________________________________ 3
1.5 Structure of the document _____________________________________________________ 4
2 STRATEGIC CONTEXT _______________________________________________________________ 5
2.1 Safer Journeys – Road Safety Strategy 2010 _______________________________________ 5
2.2 Safe System _________________________________________________________________ 5
2.2.1 Safe System principles __________________________________________________________ 5
2.2.2 Human tolerance to physical force _________________________________________________ 6
2.2.3 Safe System components ________________________________________________________ 7
2.3 Key Safer Journeys initiatives ___________________________________________________ 8
2.3.1 Introduction __________________________________________________________________ 8
2.3.2 Safe Road and Roadsides ________________________________________________________ 9
2.3.3 Safe Speeds ___________________________________________________________________ 9
2.4 Source material _____________________________________________________________ 10
3 CRASH PRIORITIES: STRATEGIC CONTEXT ______________________________________________ 11
3.1 Crash severity on New Zealand’s rural road networks ______________________________ 11
3.2 State highways and local roads network performance ______________________________ 12
3.3 Key crash types _____________________________________________________________ 12
3.3.1 Head-on crashes ______________________________________________________________ 14
3.3.2 Run-off road crashes ___________________________________________________________ 15
3.3.3 Intersections _________________________________________________________________ 16
3.3.4 Vulnerable road users __________________________________________________________ 17

4 IDENTIFYING HIGH-RISK RURAL ROADS _______________________________________________ 19

4.1 High-risk rural road definitions _________________________________________________ 19
4.2 Summary of process _________________________________________________________ 19
4.3 Step 1: Process to calculate risk metrics__________________________________________ 21
4.3.1 Risk metrics __________________________________________________________________ 21
4.3.2 Selection of risk metrics for individual RCAs _________________________________________ 22
4.3.3 Using CAS and RAMM __________________________________________________________ 23
4.3.4 Collective and personal risk metrics calculations _____________________________________ 23
4.3.5 Star rating and RPS risk metrics derivation __________________________________________ 24
4.4 Step 2: Process to determine a high-risk rural road using risk metrics __________________ 24
4.4.1 Crash density (collective risk) and crash rate (personal risk) comparison on all rural roads ____ 24
4.4.2 KiwiRAP collective and personal crash risk maps _____________________________________ 26
4.4.3 KiwiRAP RPSs and star ratings (state highways) ______________________________________ 30
4.5 Treatment of high-risk rural roads ______________________________________________ 33
4.5.1 Process _____________________________________________________________________ 33
4.5.2 Safety Improvement Strategy ____________________________________________________ 33
4.5.3 Treatment examples ___________________________________________________________ 35
4.6 Performance and safety improvement potential of high-risk rural roads _______________ 45
4.7 Addressing other network safety issues__________________________________________ 46
5 COUNTERMEASURES ______________________________________________________________ 50

high risk rural roads guide interim draft for use april 2011 final_.docx i
 High-risk rural roads guide

5.1 Section layout ______________________________________________________________ 50

5.1.1 Crash types __________________________________________________________________ 51
5.2 Treatment philosophy strategy ________________________________________________ 51
5.2.1 Safe System Transformation Treatments ___________________________________________ 55
5.2.2 Safer Corridor improvements ____________________________________________________ 67
5.2.3 Safety management treatments __________________________________________________ 89
5.2.4 Safety maintenance treatments _________________________________________________ 101
5.2.5 Crash clusters (blackspot) treatments ____________________________________________ 101
5.2.6 Treatments based on collective and personal risk ___________________________________ 101
5.2.7 Treatments based on key crash types (themes) _____________________________________ 102
5.3 Road safety action plans _____________________________________________________ 105
5.4 Summary of countermeasures ________________________________________________ 106
5.4.1 Brief description _____________________________________________________________ 106
5.5 Case studies _______________________________________________________________ 110
5.5.1 Centennial Highway __________________________________________________________ 110
5.5.2 Rangiriri ____________________________________________________________________ 111
5.6 Source Material ____________________________________________________________ 112
5.6.1 iRAP Road Safety Toolkit _______________________________________________________ 112
5.6.2 Austroads Road Safety Engineering Toolkit ________________________________________ 112
5.6.3 KiwiRAP Assessment Tool (KAT) _________________________________________________ 112
5.6.4 Rune Elvik – Handbook of Road Safety Measures ___________________________________ 113
5.6.5 PIARC Countermeasures _______________________________________________________ 113

6 PROGRAMME IMPLEMENTATION, MONITORING AND EVALUATION_______________________ 114

6.1 Introduction _______________________________________________________________ 114
6.2 Programme development ____________________________________________________ 115
6.2.1 Programme Prioritisation – Focus on Infrastructure Transformation Countermeasures ______ 117
6.2.2 Challenges to Implementation __________________________________________________ 117
6.2.3 Programme Implementation ___________________________________________________ 118
6.2.4 Focus on Incremental Improvements Across Network ________________________________ 118
6.2.5 Consistency _________________________________________________________________ 119
6.3 Monitoring and Evaluation ___________________________________________________ 119
6.3.1 Monitoring _________________________________________________________________ 119
6.3.2 Evaluation __________________________________________________________________ 119
6.3.3 Evaluation of Treatments ______________________________________________________ 120
6.3.4 Monitoring and Evaluation Performance Measures __________________________________ 124
6.3.5 Goals and Targets ____________________________________________________________ 127
6.3.6 Responsibilities for monitoring and evaluation _____________________________________ 128

7 OTHER INFORMATION SOURCES____________________________________________________ 129

APPENDIX A DISTRIBUTION OF HIGH-SEVERITY RURAL CRASHES ___________________________ 138
APPENDIX B RAMM SQL FOR CALCULATING PERSONAL AND COLLECTIVE RISK ________________ 140
APPENDIX C ALLOCATION OF ROAD NETWORKS TO CLIMATE ZONES________________________ 144
APPENDIX D PERSONAL REPORTED VS PREDICTIVE RISK CORRELATION CHARTS _______________ 145

high risk rural roads guide interim draft for use april 2011 final_.docx ii
 High-risk rural roads guide

Glossary of terms

3 Es engineering, education and enforcement

AADT annual average daily traffic
ATP markings audio-tactile profiled markings
Austroads National Association of Australian Road Authorities
CMS changeable message sign
EMP edge marker post
gTKP global transport knowledge partnership
high-severity crashes fatal and serious crashes
HRRRG High-risk rural roads guide
IL investigation levels for skid resistance
iRAP International Road Assessment Programme
KAT KiwiRAP assessment tool
KiwiRAP The New Zealand joint agency Road Assessment Programme
LoS level of service
MoT Ministry of Transport
NSC Network Safety Coordination
NZTA NZ Transport Agency
OECD Organisation for Economic Co-operation and Development
PIARC World Road Association
RCA road controlling authority
RISA Road safety infrastructure assessment
RoNS road of national significance
RPS road protection score
RRPM retro-reflective raised pavement marker
SAWS speed-activated warning signs
SCRIM Sideway-force Coefficient Routine Investigation Machine
SHGDM State highway geometric design manual
SWOV Dutch national road safety research institute
TERNZ Transport Engineering Research New Zealand
the Guide the HRRRG
TLA territorial local authority
VMS variable message sign
WRB wire rope barrier

high risk rural roads guide interim draft for use april 2011 final_.docx A
 High-risk rural roads guide

1 Introduction and objectives

1.1 Purpose

The High-risk rural roads guide (HRRRG or ‘the Guide‘) has been prepared by the NZ Transport Agency
(NZTA) to provide guidance on the government’s Safer Journeys 2020 Strategy initiative to focus efforts on
high-risk rural roads. The objective of the Guide is to provide practitioners and policy makers with best
practice guidance to identify, target and address key road safety issues on high-risk rural roads. The Guide
provides links to a number of road safety resources and guidance for planning, funding and evaluating
safety projects and programmes. Specifically, the Guide is intended to provide:

• details of a Safe System approach to Safe Roads and Roadsides and to a lesser degree Safe Speeds
in New Zealand

• identification of key crash issues on New Zealand rural roads

• tools to help identify and analyse high-risk rural roads

• a range of countermeasures for key crash types occurring in rural environments, to help develop
best-value remedial treatments

• guidance for developing, prioritising and funding road safety infrastructure and speed management
programmes

• references to further tools and resources to evaluate implemented countermeasures.

The Guide has also been developed to provide national consistency regarding the identification of high-risk
rural roads and the application of proven countermeasures.

The Guide provides a mechanism for road controlling authorities (RCAs) to manage the safety of their road
networks. Although there are many ways in which high-risk rural roads can be identified, regions will still
need to identify and prioritise their own issues regardless of whether they conform to those identified in
the Guide.

In addition to this Guide, the High-risk intersection guide (currently under development) could be
referenced for more detailed information and recommendations on both high-risk urban and high-risk rural
intersections.

1.2 Scope

The HRRRG refers and directly links to the Austroads Guides and to a number of appropriate policies,
standards and guidelines applicable to New Zealand practice.

The HRRRG will support and reference:

• the New Zealand Ministry of Transport’s (MoT) Safer Journeys 2020, New Zealand’s Road Safety
Strategy 2010–2020 (March 2010)

high risk rural roads guide interim draft for use april 2011 final_.docx 1
 High-risk rural roads guide

• the MoT’s cross-agency Safer Journeys Implementation Action Plan 2011/12 (released March 2011)

• the NZTA’s Road Safety Strategy Plan (December 2010)

• New Zealand legislation and, in particular, the Land Transport Act 1998 and rules made pursuant to
that act, including the Land Transport (Road User) Rule, the Land Transport Rule: Traffic Control
Devices, and the Land Transport Rule: Setting of Speed Limits

• general polices contained in Austroads Guides (Guides to Traffic Management, Road Design, Road
Safety) and other Austroads Technical Guides.

• New Zealand and, as appropriate, Australian standards codes of practice and guidelines

• published standards of various organisations and authorities.

The Guide provides rules, standards and guidance on measures to improve safety on high-risk rural roads.
However, practitioners must always apply sound judgement when identifying and installing any
countermeasures to ensure the best possible safety outcomes. Any departures from recommended
practice must be supported by documentation of the principles behind the departures.

1.3 Target audience

The principles presented in the Guide are relevant to both state highway and territorial local authority (TLA)
rural road networks. The HRRRG is intended to provide guidance to:

• RCAs

• state highway and local roads engineers

• planners

• funders

• policy makers.

It may also be useful to other industry practitioners, developers and private landowners when they would
like to identify road safety risks and develop appropriate risk-reducing measures.

high risk rural roads guide interim draft for use april 2011 final_.docx 2
 High-risk rural roads guide

1.4 Definitions

For the purposes of using the Guide:

• a rural road is a road with a speed limit of 80km/h1 or more. It can be a motorway, state highway,
expressway, local road or private road

• a high-risk rural road is classified as:

– a rural road where the fatal and serious crash rate (personal risk) or crash density (collective
risk) is high compared with other roads as defined in section 4

– a high or medium–high collective; or a high or medium high personal risk route (as defined by
KiwiRAP2)

– a rural road that has features that are likely to increase the potential for fatal or serious injury
crashes along a route as determined by the KiwiRAP star rating or road protection score (RPS), ie
1 or 2 star road or an RPS greater than 10 ( section 4)

– an equivalent process such as the Road Safety Infrastructure Assessment (RISA) where the risk
score is greater than 3.0 (section 4).

1
Note that, for the purposes of the Guide and any relevant crash analysis, the definition of a rural road is a road with a posted
speed limit of 80km/h or more. However, in some documents, such as the NZ Transport Agency’s Traffic control devices manual, this
has been defined as being 70km/h or more.[2]
2
KiwiRAP is New Zealand’s joint agency Road Assessment Programme.[3] The Ministry of Transport, NZTA, Police, ACC and AA
developed the programme to assess the risk of New Zealand roads and targeted it at decision makers and the wider public.
high risk rural roads guide interim draft for use april 2011 final_.docx 3
 High-risk rural roads guide

1.5 Structure of the document

The Guide is divided into six main sections, as described below:

high risk rural roads guide interim draft for use april 2011 final_.docx 4
 High-risk rural roads guide

2 Strategic context

2.1 Safer Journeys – Road Safety Strategy 2010

The New Zealand government released its Safer Journeys – Road Safety Strategy in March 2010. Safer
Journeys is a national strategy to guide improvements in road safety over the period 2010 to 2020. The
Strategy sets out a long-term vision for New Zealand of ‘A safe road system increasingly free of death and
serious injury‘. [1]

To support the vision, Safer Journeys introduces, for the first time in New Zealand, a Safe System approach
to road safety (section 2.2).

Safer Journeys also lists a number of key initiatives that have been identified as having the greatest impact
on road trauma. These initiatives will be implemented through a series of action plans relating to Safe
Roads and Roadsides, Safe Speeds, Safe Road Use and Safe Vehicles.

2.2 Safe System

2.2.1 Safe System principles

A Safe System approach to road safety represents a fundamental shift in the way New Zealanders think
about road safety. It works on the principle that it is not acceptable for a road user to be killed or seriously
injured if they are involved in a crash. The Safe System approach also acknowledges that road users are
fallible and will continue to make mistakes.

Scandinavian research [4] indicates that, even if all road users complied with all road rules, fatalities would
only fall by around 50% and serious crashes by 30%. Putting this in a New Zealand context, if everybody
obeyed all the road rules, there would still be around 200 road deaths each year (based on present
fatalities).

The traditional 3 Es approach to road safety – engineering, education and enforcement – has helped
achieve our current levels of road safety and these elements remain important. However, the 3 Es
approach tends to blame and try to correct the road user. Continuing with this approach will not achieve
the desired gains in road safety in New Zealand. A Safe System approach recognises the need for system
designers and road users to share responsibility, with the ultimate aim of protecting road users from death
and serious injury.

high risk rural roads guide interim draft for use april 2011 final_.docx 5
 High-risk rural roads guide

The Safe System approach is about acknowledging that:

1. Human beings make mistakes However, the current consequences of those mistakes and crashes should not
and crashes are inevitable be regarded as acceptable. A Safe System aims to reduce the likelihood of
crashes with a focus on removing the potential for death or serious injury.

2. The human body has a limited The human body has a limited tolerance to crash forces. A Safe System aims
ability to withstand crash forces to manage the magnitude of crash forces on the human body to remove the
potential for death or serious injury. [refer to figure 2-1]

3. System designers and road The aim of the system designer is to deliver a predictable (self-explaining)
users must all share road environment to the road user that is also forgiving of mistakes. The Safe
responsibility for managing crash System relies on the principle of shared responsibility between system
forces to a level that does not designers and road users. System designers include planners, engineers,
result in death or serious injury policy makers, educators, enforcement officers, vehicle importers, suppliers,
utility providers, insurers, etc.

4. It will take a whole-of-system Everyone plays a part in providing a safe transport system. Road designers
approach to implement the Safe will design safe roads and roadsides that will encourage safe behaviour and
System in New Zealand be forgiving of human error. Vehicle technology (safe vehicles) will vastly
improve communication with the road environment to ensure appropriate
speeds that respond to real–time conditions (safe speeds). Road users need
to understand and play their part in the system, including an acceptance of
the skills required to get a driver licence as well as maintaining their vehicles
to appropriate standards.

2.2.2 Human tolerance to physical force

The fundamental principle of a Safe System is the relationship between road users, vehicles, speeds and
road infrastructure, and how much force the human body can withstand when each of these four elements
interacts in the event of a crash. The OECD [5] states that ‘the human body’s tolerance to physical force is
at the centre of the Safe System approach’. In addition the Draft Australian Road Safety Strategy (2011–
2020) states that ‘the chances of surviving a crash decrease rapidly above certain impact speeds, depending
on the nature of the collision’ [119]. This is illustrated in figure 2-1, which presents the risk of a fatality
occurring as a result of five key crash types.

high risk rural roads guide interim draft for use april 2011 final_.docx 6
 High-risk rural roads guide

Figure 2-1: Survivable impact speeds for different scenarios) (Figure 7 [119])

Note: The range of impact speeds for each crash type is considered to be survivable in most cases. [119]

The OECD (2008) recognises that safe speeds are paramount in achieving a Safe System. However,
achieving operating speeds that are safe speeds (figure 2-1) on rural roads throughout New Zealand will in
some cases adversely affect transport efficiency. Other measures, such as providing median separation,
would be needed to reduce crash severity where safe speed thresholds cannot be appropriately provided.
The need to balance efficiency desires and safety often leads to a harm reduction philosophy as opposed to
more rigid harm minimisation philosophies.

2.2.3 Safe System components

Under a Safe System, designers create and operate a transport system where road users who are alert and
compliant are protected from death and serious
injury. The four key components of a Safe System are
illustrated in Figure 2-2 and include:

• Safe Roads and Roadsides that are

predictable and forgiving of mistakes – their
design should encourage appropriate road
user behaviour and speeds

• Safe Speeds that suit the function and level of

safety of the road – road users understand
and comply with speed limits and drive to the
conditions
Figure 2-2: The Safe System
• Safe Vehicles that help prevent crashes and
protect road users from crash forces that cause death and serious injury

high risk rural roads guide interim draft for use april 2011 final_.docx 7
 High-risk rural roads guide

• Safe Road Use that ensures road users are skilled, competent, alert and unimpaired, and people
comply with road rules, choose safer vehicles, take steps to improve safety and demand safety
improvements.

HOW IS THE SAFE SYSTEM APPROACH

DIFFERENT?
When the driver of a vehicle travelling on a
rural road in wet weather loses control on
a bend, a crash into a solid roadside object
such as a power pole is likely to result in
death.

UNDER A SAFE SYSTEM

The road user has a much lower risk of
death or suffering serious injury because:

Source: NZ Herald Thursday 21 December 2006

vehicles will increasingly have advanced safety features, including electronic stability control (ESC),
front and side airbags, and head restraints

road surfaces will be improved and roadside objects removed or barriers installed

speed is managed to safe levels through more appropriate speed limits, self-explaining roads that
encourage safe speeds and devices such as Intelligent Speed Assist

road users are alert and aware of the risks and drive to the conditions.

2.3 Key Safer Journeys initiatives

2.3.1 Introduction

The Safer Journeys Strategy contains a number of road safety initiatives; the HRRRG provides guidance on
how to implement a number of the key initiatives for Safe Roads and Roadsides and Safe Speeds (to a
degree).

Specifically, the HRRRG provides guidance on the following action under the Safe Roads and Roadsides plan
with a particular focus on high-risk rural roads:

‘Focus safety improvement programmes on high-risk rural roads and high risk urban intersections’

Guidance is also provided on a number of speed management initiatives presented under the Safe Speeds
plan, including:

‘Create more speed zones on high-risk rural roads to help make roads more self explaining, and to
establish criteria for what roads with different speed limits should look like’

Speed management is an important tool for improving the safety of high-risk rural roads and is therefore
included in the Guide, alongside typical infrastructure improvement measures.

high risk rural roads guide interim draft for use april 2011 final_.docx 8
 High-risk rural roads guide

2.3.2 Safe Road and Roadsides

‘Focus safety improvements on high-risk rural roads…’

The greatest safety gains on high-risk rural roads are expected to be achieved by focusing on reducing fatal
and serious outcomes on key crash types. This approach is also consistent with the Safer Journeys long-
term vision:

‘A safe road system increasingly free of death and serious injury”

For instance, around 8 out of 10 fatal and serious crashes on the nation’s state highways occur on rural
roads and, of those state highway crashes3, 85–90% of fatal and serious crashes are due to one of three
predominant crash types:

• run-off road (on a curve or straight)

• head-on

• at intersections.

Detailed information on these crash types, for both state highways and local roads, is included in section 3.

The NZTA and local government need to ensure that road safety efforts focus on these key areas to help
achieve Safe Roads and Roadsides.

2.3.3 Safe Speeds

Safe Speeds are closely linked to Safe Roads and Roadsides – especially for rural road and highway
networks. The Guide describes how Safe Speeds can be achieved to complement Safe Roads and Roadsides
in order to improve safety for all road users. Appropriate speed management related countermeasures are
proposed that relate to all aspects of the Safe System, ie. Safe Roads and Roadsides, Safe Road Use, Safe
Speeds and Safe Vehicles.

‘The likelihood of being involved in a serious casualty crash rises significantly with even minor
changes in travelling speed (figure 2.3). For example; Australian research has shown that the risk
of a serious casualty crash doubles with just a 5km/h speed increase on 60km/h urban roads or
with a 10km/h increase on rural highways.’ [118]

3
Note that a local road analysis will be undertaken.
high risk rural roads guide interim draft for use april 2011 final_.docx 9
 High-risk rural roads guide

Figure 2-3: Relationship between change of mean speed and crashes (Figure 6 [118])

The default speed limit on New Zealand open/rural roads is 100km/h and it is generally applied to all rural
roads with only limited exceptions at the present time. A more suitable speed limit for these roads might in
future be one that more closely matches the design speed and the present safety features. The NZTA
recognises that there is some merit in applying a safer operating speed limit or speed zone4 for roads on
which the standard rural speed limit is inappropriate. This is further described in section 5.

Harm minimisation and harm reduction speeds are used within Safe Systems to reduce high-severity
crashes. These are discussed later in section 5.2.1 (c).

The Safer Journeys Strategy refers to other speed management initiatives, such as speed cameras, adoption
of lower speed limits in urban areas, GPS-based speed advice systems and road improvements to address
speed-related crashes.

2.4 Source material

The Guide recognises the availability of several other high-risk rural roads guidance documents and web-
based tools to apply relevant countermeasures. These are described in sections 4 and 5.

4
A speed zone takes into account the alignment of a route or section of road and in particular the 85th percentile operating speed
of vehicles. This is in contrast to the historical (and still the current – 2010) method of setting speed limits, which is based primarily
on the amount of frontage development.
high risk rural roads guide interim draft for use april 2011 final_.docx 10
 High-risk rural roads guide

3 Crash priorities: strategic context

Prioritising safety improvement measures for high-risk rural roads requires a focus on reducing the number
of fatal and serious crashes; this involves specifically focusing on three key crash types as summarised in
figure 3-1. Details of crash severity, road type and key crash types within a New Zealand context are further
described in this section.

Figure 3-1: Strategic context of crash priorities

SAFER JOURNEYS 2020 ROAD SAFETY STRATEGY

‘Focus safety improvement programmes on high-risk rural roads and

high-risk intersections’

Priority crash types (targeting high-risk rural roads and high-risk intersections)

FATAL AND SERIOUS CRASHES

Head-on crashes Run-off road crashes Intersection crashes

3.1 Crash severity on New Zealand’s rural road networks

Rural (open road) crashes5 accounted for 25% of all reported motor vehicle crashes on New Zealand roads
over the five-year period 2005–2009. Approximately 30% of injury crashes on all rural roads are recorded as
fatal or serious, which would contribute to a large portion of the social cost of crashes.

Fatal and serious injury crashes impose significant costs in terms of grief and suffering, as well as economic
costs. It is for this reason that Safer Journeys and the Safe System approach that underpins Safer Journeys
focuses on these high-severity crash types.

However, these high-severity crashes are typically highly dispersed. Nationwide, somewhere in the order of
56% of fatal and serious crashes on local authority rural roads have occurred at locations where no other
injury crash has been recorded in the past five years. That said, the proportions vary from TLA to TLA as
does the actual number (see appendix A). There will, however, be an increasing move to corridor or route
treatments rather than focusing solely on crash clusters, thereby ensuring road users are provided with a
consistent level of safety, appropriate to the route on which they are travelling.

5
State highway and local roads combined – and with speed limits of 80km/h or more. Excludes motorways. Note that even though
excluded from the overall data analysis in section 3 (types of crashes occurring on motorways would skew results), motorways can
still be included when defining a high-risk rural road.
high risk rural roads guide interim draft for use april 2011 final_.docx 11
 High-risk rural roads guide

If significant gains in road safety are to be made, priority should be given to addressing these routes and
locations where crashes occur of high-severity and high social cost on rural roads. This is the intention of
the Guide.

3.2 State highways and local roads network performance

As illustrated in figure 3-2, state highways comprise only 12% of the nation’s road network. However, they
account for over half of all fatal and serious injury crashes. Figure 3-3 shows that the crash density, and
especially the density of fatalities (the numbers per unit road length), is proportionally higher on state
highways than local roads. The amount of travel on state highways represents approximately half of all
vehicle kilometres travelled. Therefore, in relation to travel activity, fatal and serious crash rates are higher
for state highways than for local roads.

Figure 3-2: All roads (rural and urban 2005–2009): Network length, usage and crashes
Figure 3-3: Number of deaths and serious injuries on rural state highways and local roads (2005–2009),
excluding motorways)

3.3 Key crash types

Many high-volume rural roads have known crash problems. All crashes are used when identifying high-risk
routes, but the guide focuses on run-off road, head-on and intersection crashes as they are the most
common crash types for fatal and serious crashes on rural roads.

high risk rural roads guide interim draft for use april 2011 final_.docx 12
 High-risk rural roads guide

Further analysis of fatal and serious crashes shows that 86% of crash types6 occurring on New Zealand rural
roads over the period 2005–2009 are one of the three main crash types: run-off roads – lost control on
curves 7 and lost control on straights (45%); head-on (25%); and at intersections (16%).

Analysis also shows that vulnerable road users account for 3% of fatal and serious crashes on rural roads,
with the remaining 11% comprising a variety of crash types.

‘Run-off road’ crashes are the highest in terms of both fatal and serious crashes and injuries (figure 3-4).
However, when comparing deaths across crash types, the number of deaths in head-on crashes is of similar
magnitude to those in ‘run-off road’ crashes. In addition, there is a significant increase in the injuries of
those involved in head-on crashes (1.7 times the amount of fatal and serious crashes), due mainly to more
than one vehicle being involved.

Figure 3-4: Key crash types – crashes and severity of crashes

When comparing crashes and casualties across the key crash types, you can see that, although run-off road
crashes are the predominant crash type, the significance lowers when the number of casualties is
calculated (table 3-1).

6
Key crash types are defined in the NZTA’s Economic evaluation manual.
7
For the purposes of this Guide, it has been assumed that fatal and serious injuries due to loss of control on curves equate to run-off
road on curves.
high risk rural roads guide interim draft for use april 2011 final_.docx 13
 High-risk rural roads guide

Table 3-1: Key crash type percentages

Key crash type % of crashes for key crash type % of casualties of key crash
on New Zealand rural roads types on New Zealand rural
roads
Run-off road 61 56
Head-on 25 31
At intersections 14 13

3.3.1 Head-on crashes

Head-on crashes represent the second highest fatal and serious crash type (approximately 25%) on all rural
roads. The main type of movement is loss of control on a curve, which represents 33% of all rural fatal and
serious head-on crashes. ‘Head-on – cutting corner’ and ‘head-on – lost control on straight’ are both 11% of
all head-on rural fatal and serious crashes, with the rest of the movement types making up 11%.

Overtaking
and lane Head-on
change

AB BA BB BC BD BE BF BO

Other

Figure 3-5: Main movement types for head-on crashes (2005–2009)

high risk rural roads guide interim draft for use april 2011 final_.docx 14
 High-risk rural roads guide

In order to rate the safety on rural state highways, the KiwiRAP8 road assessment programme (discussed
later) takes into account the risk of head-on crashes, based largely on the level of median protection
provided as well as other related factors such as traffic volume and roadway alignment and width. To
achieve a low risk rating (4 stars or greater), higher-volume roads will typically need physically divided
carriageways with a central median barrier, thus reducing the potential for head-on crashes. Varying types
of median-separation measures and their application to all rural roads are discussed further in section 5.

The NZTA intends to initially target the highest-risk rural roads (those that carry over 15,000 vehicles per
day), in particular the roads of national significance (RoNS). A road with 15,000 vehicles per day has roughly
five head-on crashes per 10km every five years. Some New Zealand roads carry 15,000–20,000 vehicles per
day but do not have median barriers. Installing median barriers9 on all high-risk high-volume rural roads is
estimated to save 8 to 10 lives per year and 102 to 119 injuries per year. [1]

3.3.2 Run-off road crashes

Loss of control (run-off road) crashes are the most common rural road crash type and account for 45% of
fatal and serious crashes on all rural roads for the period 2005–2009. These types of crashes occur on both
curves and straights. The main type of movement is loss of control turning right, which represents 40% of
all rural fatal and serious run-off road crashes. The next most common movement types for run-off road
crashes are lost control turning left (29%), lost control off roadway to the left (14%) and lost control off
roadway to the right (10%), with the remainder equalling 7% of different crash movement types.

8
KiwiRAP is New Zealand’s joint agency Road Assessment Programme.[3] The Ministry of Transport, NZTA, Police, ACC and AA
developed the programme to assess the risk of New Zealand roads and targeted it at decision makers and the wider public.
9
Other countries require median barriers on all high-speed routes that have over 10,000–15,000 vehicles per day. [1] Sweden is
targeting roads with greater than 4000 vehicles per day for installing median barriers.
high risk rural roads guide interim draft for use april 2011 final_.docx 15
 High-risk rural roads guide

Overtaking
Lost Control or off road (straights) Cornering
and lane
change
AD CA CB CC CO DA DB DC DO

Other Other

Figure 3-6: Main movement types for run-off road crashes (2005–2009)
KiwiRAP (discussed later) has specifically examined the level of road and shoulder width, clear zones and
barrier treatments on state highways, in order to rate the run-off road risks associated with those roads. In
order to achieve a low risk rating (4 stars or greater), sufficient clear zones need to be provided, along with
some form of barrier treatment, a good standard of alignment or an increase in shoulder width for recovery
to address both the number and severity of run-off road crashes.

3.3.3 Intersections

Intersection crashes represent the third key crash type. The main type of intersection movement on rural
roads is right turning out traffic colliding with straight through traffic; these represent 31% of all rural fatal
and serious intersection type crashes. Those vehicles turning right across approaching traffic make up 22%
and those travelling at right angles equals 17% of rural fatal and serious intersection crashes. These
proportions are higher than for all injury intersection crashes shown below as they typically involve higher
impact speeds. The remainder of intersection movement types equals 7%.

Figure 3-7: Main movement types for intersection type crashes (2005–2009)

high risk rural roads guide interim draft for use april 2011 final_.docx 16
 High-risk rural roads guide

Turning vs Crossing
Rear end Crossing (no turns) Merging
same (vehicle
direction turning)
FB GD HA HO JA KA KB KC

Other

Right turn Manoeuvring

against
LB MA MB MC MD MG MO

Other

In order to achieve KiwiRAP’s highest 5 star rating (ie. the lowest safety risk); a route must have grade-
separated intersections. Various forms of intersection countermeasures for high-risk rural routes are
discussed in section 5.

Comment: A High-risk intersection guide is currently being developed. Once this guide is completed,
further information can be referenced.

3.3.4 Vulnerable road users

(a) Pedestrians and cyclists

Due to the lower numbers of users present, vulnerable road user (pedestrians and cyclists) crashes are less
prevalent on rural or open road networks, than on urban roads (less than 2% of crashes in rural areas).
They are not one of the three main crash types. However, where pedestrians and cyclists are involved in
open road crashes, the outcomes are typically severe, due to the often high speeds of traffic and the
human body’s limited tolerance of crash forces at speeds above 40km/h.

The inclusion of all fatal and serious crash numbers when identifying high-risk rural roads is also likely to
identify road sections that have high numbers of pedestrian and cyclist crashes.

(b) Motorcyclists

Although motorcyclists travel at higher speeds than ‘typical’ vulnerable road users, they are still at high risk
of severity of injury. On some routes, this may be higher and any treatments proposed should recognise
their vulnerability.

The Safe System seeks to provide Safe Roads and Roadsides and Safe Speeds that in turn improve safety for
vulnerable road users on rural roads. Section 5 discusses the ways in which good design practice for
pedestrians; cyclists and motorcyclists can be appropriately adapted to rural road networks.

high risk rural roads guide interim draft for use april 2011 final_.docx 17
 High-risk rural roads guide

The inclusion of all fatal and serious crash numbers when identifying high-risk rural roads will identify road
sections that have a high over-representation of motorcycle crashes.

Comment: A ‘High-risk motorcycle guide’ is currently being developed. Once this guide is completed,
further information can be referenced.

high risk rural roads guide interim draft for use april 2011 final_.docx 18
 High-risk rural roads guide

4 Identifying high-risk rural roads

The safety performance of a road is a function of:

• the likelihood of each user travelling on the road being involved in a crash

• the traffic volume, ie. the number of vehicles using the route.

High-risk rural roads are essentially lengths of road with a higher than ‘average‘ crash risk, and by
implication are roads where targeted safety improvements are most likely to achieve reductions in trauma.
High-risk rural roads generally have the greatest safety improvement potential, which is why they
represent one of the highest Safe Roads and Roadsides priorities for investigation. However, there are also
likely to be benefits from improving roads with moderate risks, or riskier spot locations (such as crash
clusters/blackspots) on road lengths not formally classified as high risk by this Guide. Cost-effective
solutions may be available for such sites and it is not the intention of this Guide to stop blackspot studies
and treatments, but rather to focus most attention on high-risk routes.

This section of the Guide defines risk metrics and what constitutes a high-risk rural road and outlines how
the various risk metrics that make up the definition of a high-risk rural road can be derived. Guidance has
also been provided on how these metrics can be used to determine an appropriate treatment strategy,
together with some examples of the process.

4.1 High-risk rural road definitions

A high-risk rural road is defined as:

• a rural road where the fatal and serious crash rate (personal risk) or crash density (collective risk) is
classified as high compared with other roads (section 4.4.1 and figures 4-1 and 4-2); and/or

• a high or medium–high collective risk and/or high or medium–high personal risk (as defined by
KiwiRAP risk maps) (section 4.4.2); and/or

• a rural road that has features that are likely to increase the potential for fatal or serious injury
crashes along a route as determined by the KiwiRAP star rating or RPS, ie. 1 or 2 star road or an RPS
greater than 10 (section 4.4.3); and/or

• an equivalent process such as the Road Safety Infrastructure Assessment (RISA) where the risk
score is greater than 3.0 (section 4.4.4).

4.2 Summary of process

The process of identifying high-risk rural roads in your network should be completed every three years to
provide information in support of maintenance and renewal works included in the three-year National Land
Transport Programme (NLTP). Table 4-1 is a summary of the processes you can use to determine high-risk
rural roads. Detailed information on each of these is provided, with some description of treatment
examples in section 4.5.

high risk rural roads guide interim draft for use april 2011 final_.docx 19
 High-risk rural roads guide

Table 4-1: Summary of process to identify high-risk rural roads

Use Treatment Philosophy Strategy (figure 4-6) to determine appropriate scale of treatments for the route or site

Comment: Further analysis on how RISA correlates with other processes will be completed and added
to table 4-1.

high risk rural roads guide interim draft for use april 2011 final_.docx 20
 High-risk rural roads guide

4.3 Step 1: Process to calculate risk metrics

4.3.1 Risk metrics

Firstly, it is important to understand what a risk metric is. Generally, a measure is an operation for assigning
a number to something. A metric is our interpretation of the assigned number. For example, we can record
how tall someone is by obtaining a measure in metres from the soles of their feet to the top of their head.
We interpret the number obtained as the person’s height. Height (in this case) is a metric.

In the case of risk, risk measurement is the process by which we measure risk and a risk metric is the unit
corresponding to the resulting value.

Generally, risk can be determined in two ways:

• actual crash risk; or

• predicted crash risk.

Actual crash risks are based on the reported crashes in recent years and are, in theory, the primary
measure of performance. They are most reliable where crashes are frequent, which happens on the busier
roads where traffic volumes are also more reliable. However, on quieter roads, the crash density of fatal
and serious crashes becomes too low to provide a reliable picture.

KiwiRAP has already published some actual crash risk metrics. These are:

• collective risk maps (section 4.4.2)

• personal risk maps (section 4.4.2).

Metrics for predicting risk have also been developed for the local roads RISA programme (RISA scores –
section 4.4.4.) and as part of KiwiRAP and RISA, such as the:

• road protection score (RPS) (section 4.4.3)

• star rating (published maps – section 4.4.3)

Predictive risk scores are most useful for roads with fewer crashes and lower traffic volumes, where the
random nature of crashes can be misleading when dealing with small crash numbers. Because Safer
Journeys and the Safe Systems approach focus on the less common fatal and serious crashes, predictive risk
scores are important to ensure we don’t implement measures that simply chase random crashes around
the network.

Given the higher traffic volumes using the state highway network and the relatively high proportions of
high-severity crashes, the initial release of KiwiRAP has focused on rural state highways. However, all RCAs
are able to calculate collective and personal risk from their crash histories and traffic volumes.

high risk rural roads guide interim draft for use april 2011 final_.docx 21
 High-risk rural roads guide

The definitions for each of the risk metrics are as follows:

• Collective risk (also known as crash density) is a measure of the number of high-severity (fatal and
serious) crashes that have happened per kilometre of road per year. Additional information and
calculations are provided in section 4.3.4.

• Personal risk (or crash rate) is a measure in terms of the number of high-severity (fatal and serious)
crashes that have happened per 100 million vehicle kilometres of travel on the road. Additional
information and calculations are provided in section 4.3.4.

• Road protection score (100m) is a predictive measure of the personal safety of a road based on the
presence or absence of road infrastructure features that are associated with the major crash types
on the New Zealand rural road network, ie. head-on run-off road and intersection crashes (section
4.4.3).

• Star rating (5km) is a predictive measure of the personal safety of a longer length of road based on
the 100m RPSs and averaged over 5km lengths (section 4.4.3).

• By combining predictive scores for personal risk with traffic volumes, it is possible to create
predictive metrics that represent collective risk or crash density.

4.3.2 Selection of risk metrics for individual RCAs

To help RCAs identify high-risk rural roads, the following risk metrics can currently be used.
• For local roads:
– collective and personal risk (formulas provided in section 4.3.4)
– RISA scores (section 4.4.4)
Note: At this stage, no KiwiRAP predicted crash risk metrics are available for local roads.
• For state highways:
– collective and personal risk (formulas provided in section 4.3.4) or the KiwiRAP risk maps (section
4.4.2); and/or
– the RPSs and associated star ratings (section 4.4.3).
The followings sections (sections 4.3.3, 4.3.4 and 4.4) provide guidance to RCAs on how to use both
calculated crash risk parameters and KiwiRAP (risk maps, RPSs and star ratings) to identify high-risk rural
roads and from these metrics how to determine the appropriate treatment philosophy.

Analysis of high-risk rural roads should be completed in the year leading up to the three-year NLTP.

high risk rural roads guide interim draft for use april 2011 final_.docx 22
 High-risk rural roads guide

4.3.3 Using CAS and RAMM

(a) Crash Analysis System (CAS)

Using CAS, map the crashes on your identified section of road and then use the measurement tool to obtain
the length of the road section. This approach works well when looking at predefined links, but is
cumbersome when seeking to screen a network. RAMM data is better suited to network screening.

(b) Road Asset Maintenance Management database (RAMM)

RAMM is an alternative to CAS for calculating collective risk and is more appropriate for network-wide
screening. Using the traffic crash data held in the databases operated by RCAs, an annual report from CAS
can be created and exported. This is linked to the RAMM road_id. Appendix B describes how to perform
this calculation.

One of the key issues with this approach is the proliferation of relatively short links in many RAMM
databases. Once plotted in RAMM Map or another GIS system, it is generally necessary to ‘join‘ up
sequential sections of road to define a publicly known route.

4.3.4 Collective and personal risk metrics calculations

Having identified links of interest using CAS and/or RAMM, the personal and collective risk needs to be
calculated. The following sections further discuss the definition and the calculation of collective and
personal risk.

(a) Collective risk

Of the two crash risks, collective risk or crash density is the easiest to quantify, and is simply the number of
high-severity crashes divided by the length of road under consideration.

Collective risk = (Fatal crashes + serious crashes) / number of years of data

Length of road section

There are two ways of collecting data to calculate collective risk: CAS or the RAMM databases.

Having calculated the collective risk, plot this value (figure 4-1) against the length of road you are
investigating to determine whether you have a high-risk rural road.

(b) Personal risk

Personal crash risk (or crash rate) is in effect a measure of the likelihood of an individual road user being
involved in a crash as they travel the road in question.

Personal risk = Fatal crashes + serious crashes

(length of highway in km × number of years × 365 days × AADT) / 108

Calculating the personal crash risk is more complicated because you need to establish the volume of traffic
on the section under consideration (annual average daily traffic – AADT). Personal risk is typically calculated
high risk rural roads guide interim draft for use april 2011 final_.docx 23
 High-risk rural roads guide

for a five-year period, although in networks with lower traffic volumes and crash numbers a 10-year period
may be more appropriate, provided there have not been substantial changes during this period.

Plot the calculated personal risk value for the length of road you are investigating (figure 4-2) to determine
whether you have a high-risk rural road.

AADT data for the mid-year is preferred. If traffic data is not available for the mid-year, it is generally
possible to use an appropriate growth factor to adjust flows from other years. As with collective risk,
personal risk can be determined using RAMM, for example, provided the traffic volume data is reasonably
reliable and does not vary much along the route being assessed.

4.3.5 Star rating and RPS risk metrics derivation

The star rating uses the RPSs to determine a rating over a 5km length. These ratings have been published as
part of the KiwiRAP risk maps. To determine a rating for a particular length of road, reference can be made
to the maps of the KiwiRAP Analysis Tool (KAT). Further detailed information on these maps is provided in
section 4.4.3.

The RPS risk metrics can only be derived from KAT.

4.4 Step 2: Process to determine a high-risk rural road using risk metrics

4.4.1 Crash density (collective risk) and crash rate (personal risk) comparison on all rural roads

The collective and personal risk comparison (illustrated in figures 4-1 and 4-2) is based on an analysis of the
state highway rural network, as this network has relatively complete traffic volume coverage; however, the
figures can be used for all roads. Bands of collective and personal risk have been defined for each risk level.
Each band contains approximately 20% of the routes when ordered according to both Collective High-
severity Crash Risk (figure 4-1) and Personal High-severity Crash Risk (figure 4-2). Routes defined as high
risk are the highest or worst 20% of roads when ordered in terms of crash rate and crash density
respectively.

Having determined the collective and personal crash risks, as outlined in section 4.3.4, we need to seek to
identify what types of improvement strategy are likely to be worthwhile (section 5).

high risk rural roads guide interim draft for use april 2011 final_.docx 24
 High-risk rural roads guide

Figure 4-1: Collective high severity crash risk (crash density)

0.30

0.25 Constitutes a
High High-risk rural

0.20
High Severity Crashes
km per year

0.15

Medium - High
0.10

Medium
0.05

Medium - Low

0.00 Low
4000

8000

12000

16000

20000

24000

28000

32000
1

Length (m)

Figure 4-2: Personal high severity crash risk (crash rate)

20

18

16 Constitutes a
High
High-risk rural
14
High Severity Crashes
100 millon veh. Km.

12

10
Medium - High
8

4 Medium

2 Medium - Low

Low
0
12000

16000

20000

24000

28000

32000
4000

8000
1

Length

high risk rural roads guide interim draft for use april 2011 final_.docx 25
 High-risk rural roads guide

4.4.2 KiwiRAP collective and personal crash risk maps

As part of developing KiwiRAP, a rating system to describe personal and collective risk in terms of a five-
category ordinal scale was created using the following descriptors:

Table 4-2: KiwiRAP risk map ratings

Collective risk Personal risk

(average annual fatal and (average annual fatal and
Risk descriptions
serious injury crashes serious injury crashes per 100
per/km) million vehicle/km)

≤0.0390 <4
Low

Low–medium 0.04<+0.069 4≤4.9

Medium 0.07≥0.10 5≤6.9

Medium–high 0.11≤0.189 7≤8.9

High 0.19+ 9+

The five-category scale provides practitioners with an appreciation of where the collective and personal
risks on the road they are considering fit within the national picture. A five-category scale has also been
developed for each of the KiwiRAP star ratings and RPSs (section 4.4.3, table 4-3).While each uses a five-
category scale, they are not the same.

Note that the figures in table 4-2 provide general information on risk description for collective and personal
risk. More detailed information on where a particular length of road may fit into a risk category can be
completed using calculations in section 4.1.1 and figures 4-1 and 4-2.

The KiwiRAP crash risk maps (figures 4-4 and 4-5) present both the personal and collective risks for 172
state highway routes, based on historic crash performance. These maps
(www.kiwirap.org.nz/risk_maps.html generated in 2007 but released in January 2008) used fatal and
serious crash data for the period 2002 to 2006 inclusive, and it is expected that the maps will be updated in
2011. At that time, it is likely that the risk band thresholds (figure 4-3) will also be updated to reflect
improvements in safety performance.

The state highway network has been divided into five bands, each containing 20% of the routes (eg. 20% of
the routes are rated ‘high’, 20% ‘medium–high’). However, the segment length used in the KiwiRAP risk
maps average some 60km. This is because the maps were initially aimed at informing the general public
about travel risk and hence the links were selected primarily between major town centres or intersections
of state highways.

The way in which these state highway links have been selected has two effects:

1. Many of the high collective risk links are shorter higher-volume sections typically located in the
North Island.

high risk rural roads guide interim draft for use april 2011 final_.docx 26
 High-risk rural roads guide

2. The higher personal risk lengths tend to be longer lengths with lower traffic volumes and are
typically in the South Island.

The result of these biases is that only 22% (2372km) of the rural state highway network has been mapped
as ‘high’ or ‘medium–high’ in terms of collective risk, while 46% (4962km) has been mapped as ‘high’ or
‘medium–high’ in terms of personal risk. Within any particular link, there will be sections, sub-routes or
corridors that may have higher risk ratings than the link itself and these sub-sections, which will typically be
greater than 5km, may be high-risk rural roads (sections) in their own right. Similarly, there will be lengths
with lower risk ratings. That said, the KiwiRAP risk maps do provide a starting point for investigating high-
risk rural routes as shown in examples (a) and (b) of section 4.3.5.

high risk rural roads guide interim draft for use april 2011 final_.docx 27
 High-risk rural roads guide

Figure 4-3: KiwiRAP North Island risk maps (collective risk – left, personal risk – right)

high risk rural roads guide interim draft for use april 2011 final_.docx 28
 High-risk rural roads guide

Figure 4-4: KiwiRAP South Island risk maps (collective risk – left, personal risk – right)

high risk rural roads guide interim draft for use april 2011 final_.docx 29
 High-risk rural roads guide

4.4.3 KiwiRAP RPSs and star ratings (state highways)

The second KiwiRAP protocol – the KiwiRAP RPSs and associated star ratings – also provides a means of
identifying high-risk rural roads through predicted crash risk. The KiwiRAP RPS and associated star rating
measure the safety of a road based on the presence or absence of road infrastructure features that are
associated with the major crash types on the New Zealand rural road network: head-on, run-off road and
intersection crashes.

Risk Descriptions Star Rating (5km) Road Protection Score (100m)

5 <1.05
Low

Low-Medium 4 1.04-4.5

Medium 3 4.5-10

Medium-High 2 10-25

High 1 >25

Table 4-3: KiwiRAP Star Rating and Road Protection Score

The strong relationship between the RPS and crash rate (injury crashes per 100 million vehicle kilometres of
travel) means that KiwiRAP RPS or star ratings can be used as a surrogate but more proactive measure of
personal crash risk.

More importantly the KiwiRAP RPSs have been calculated for every 100m section of rural state highway.
These have then been averaged over 5km lengths to produce the star ratings (see figure 4-5).

A more detailed discussion of the KiwiRAP RPS process and the resulting star rating maps can be found at
www.kiwirap.org.nz/pdf/KIWIRAP%202010%20book%20low%20res.pdf.

As the KiwiRAP star rating provides, in effect, a measure of personal risk, these can be combined with crash
plots from CAS to provide an assessment of both personal and collective risk.

Due to the enormous volume of data underlying the KiwiRAP RPSs and star ratings, the NZTA has created
the KiwiRAP Analysis Tool (KAT), which allows practitioners to identify and investigate sections of the state
high network that meet particular criteria. For example, KAT would enable a user to find sections of
highway at least 5km in length, with an RPS of more than 10 (these are 2 star sections) and carrying more
than 5000 vehicles per day. Using the Road Safety Infrastructure Assessment (RISA)

Road Infrastructure Safety Assessment (RISA) has been developed by the NZTA to ensure that it has factual
information about the risks to road users as a result of the infrastructure on an RCA’s road network.

The system involves visual assessment of infrastructure features, which crash research has shown to result
in crashes, on randomly selected road sections within an RCA network.

RISA outputs have been reviewed against actual road crashes and shown to be a very good indicator of
crash risk potential.

RISA can generate both personal and collective risks for each road section assessed.
high risk rural roads guide interim draft for use april 2011 final_.docx 30
 High-risk rural roads guide

These results are then factored by vehicle kilometres of travel (VKT) on the section assessed and then
across the network, based on the sample VKT, to give a Network Risk score.

This assessment process can be adapted to focus on higher-volume roads for an RCA and to provide
personal risk scores for each. These scores are an indication of the ‘relative’ risks between road sections
assessed and are comparable between RCAs.

Comment: Note that there is likely to be additional RPSs developed for higher-volume local roads in the
future.

high risk rural roads guide interim draft for use april 2011 final_.docx 31
 High-risk rural roads guide

Figure 4-5: KiwiRAP Star ratings

high risk rural roads guide interim draft for use april 2011 final_.docx 32
 High-risk rural roads guide

4.5 Treatment of high-risk rural roads

This section provides guidance on how to use the above risk metrics to determine an appropriate
treatment strategy, together with some examples of the process.

4.5.1 Process

Using the process explained in section 4.2, you have now determined what your level of risk is, be it high
risk or low risk. Using your calculated risk levels for collective and personal risk, star rating, RPS, volume
data and/or using figures 4-1 and 4-2, you can use the ‘treatment philosophy strategy’ (figure 4-6) to
determine the appropriate treatment for your route or site. A detailed explanation is provided in section
4.5.2.

4.5.2 Safety Improvement Strategy

Figure 4.6 provides a schematic of the general treatment philosophy strategy that has been developed to
guide the selection and implementation of various improvement measures based on the various metrics
that define the risk of a particular route under consideration, ie. collective risk shown on the horizontal
axis, and personal risk shown on the vertical axis. Both can be based on crash histories (as shown at the left
and top) or predictive scores (as shown on the right or across the bottom). In addition, a correlation chart
(appendix D) has been provided to convert KiwiRAP personal predictive risk scores to crash rates if needed.

The upper right portion circle quadrant of figure 4-6 shows those routes with both high personal risk and
high collective risk. The high personal risk provides scope for potentially large reductions in personal risk.
When high personal risk occurs on higher-volume routes, the result is a high crash density. There is
considerable scope to reduce personal risk, and there are likely to be sufficient crash reduction benefits to
justify larger infrastructure improvements.

high risk rural roads guide interim draft for use april 2011 final_.docx 33
 High-risk rural roads guide

Figure 4-6: Treatment Philosophy Strategy

At the other extreme, the lower left, both the personal crash risk and the resulting crash density are low;
there is in effect no serious safety problem. That said there may still be scope for treating crash clusters or
shorter sections of the route for which crash data or other tools such as KiwiRAP predict higher levels of
risk.

The lower right quadrant comprises routes with relatively low personal risk but higher traffic volumes. In
these situations safety improvements are less likely to result in major reductions in personal risk but the
high volumes can generate significant benefits, although probably not sufficient to see a complete
transformation of the road environment. In these situations, incremental improvements such as hazard
management, side barriers, median treatments or other theme-based interventions along the route are
likely to be the most appropriate approach.

The upper left quadrant is characterised by high levels of personal risk but lower traffic volumes result in
low crash density. On these roads, the potential crash reduction benefits will be limited, and strategies
focused around ensuring the highest levels of signage, delineation and road surface maintenance and
management will be most common. Specific attention should be paid to speed management recognising
that appropriate speeds will reduce both the likelihood and severity of crash outcomes.

When considering the crash pattern on routes, there is almost a continuum of safety performance that
needs to be investigated.

high risk rural roads guide interim draft for use april 2011 final_.docx 34
 High-risk rural roads guide

What type of safety problem do you have?

Possible Safety
Yes Maintenance (figure 4-6)

Safety maintenance
(section 5.2.4)
Yes
Safety Management and
site specific treatments
(section 5.2.3)

Yes Safer Corridors (section

5.2.2)

Safe System
Yes Transformation Works
(section 5.2.1)

Table 4-4: Safety Problem and possible treatment

Figure 4-6 provides guidance on the overall form of the corridor improvement strategy but not the specific
measures that may be most appropriate. The first step in such an investigation is to determine what type of
safety problem you have (table 4-4), whether the current crash patterns have either geographical or
thematic commonality, whether they are clustered (black or grey spots) or whether there is a common
theme, eg. lost control on curve in dark. Although there may not be specific black or grey spots, there may
be subsections of the route where there appears to more crashes than on other sections.

Further analysis and treatments of crash clusters (or blackspots) can also be found in the New Zealand
guide to the treatment of crash locations.

4.5.3 Treatment examples

In this section we provide three examples: the first two involve state highways and using various
components of KiwiRAP to identify and investigate high-risk corridors; the third example is based on a local
authority network.

(a) Using KiwiRAP Crash Risk Maps – SH2 for Katikati to Tauranga

The 2008 KiwiRAP crash risk mapping (based on 2002 to 2006 crash data) identified 32 Black Routes,
sections of state highway with the highest collective crash risk. One of these was a 27km section of State
Highway 2 from Tauranga to Katikati, which ranked 26th worst in New Zealand.

high risk rural roads guide interim draft for use april 2011 final_.docx 35
 High-risk rural roads guide

Figure 4-7: KiwiRAP Crash Risk Maps (2002 to 2006 data)

We further clarified and updated the crash data (using CAS) and determined that, over the period 2005 to
2009 inclusive, there have been 30 high-severity crashes over the 27km route which carried around 11,500
vehicles per day in 2007. Using the calculations in sections 4.3.4, we obtained the following results for risk:

Personal Risk = 30 high-severity crashes / ((27km x 11,500 AADT x 5 years x 365 days) / 108)
= 30 / 5.66
= 5.29 high-severity crashes per 100 million vehicle kilometres of travel

This route is a medium personal risk (using figure 4-2) when using only this factor
and the definitions described in section 4.1. This route does not constitute a high-
risk rural road.

Collective Risk = 30 high-severity crashes / 27km / 5 years

= 0.22 high-severity crashes per km per year

This is a high collective risk (using figure 4-1), and therefore the route is a high-risk
rural road as described in section 4.1 (even though personal risk does not constitute
a high risk rural road).

Using the treatment philosophy strategy (figure 4-6) shows that this section of highway lies on the
boundary between Safe Systems Transformation Works and Safer Corridors treatments (see figure 4-8 red
star). However, using the KiwiRAP star rating of 2.8 (obtained from the KAT tool – figure 4-9) would suggest
a greater focus on the road infrastructure improvements (see figure 4-8 green star).

high risk rural roads guide interim draft for use april 2011 final_.docx 36
 High-risk rural roads guide

Figure 4-8: Treatment Philosophy Figure 4-9: KAT Search Map

Plotting the high-severity crashes in CAS (Figure 4-10: CAS-reported high-severity crashes (2005 to 2009
inclusive), we find the crashes are distributed along the route, as opposed to being clustered, suggesting a
corridor approach may be warranted. There is however a higher density of crashes over the 9km section
beginning around RS 116/10. While treating the whole corridor may be worth considering, given there is a
further high-density section to the south, focusing attention on the section around RS 116/10 may be
worthwhile bearing in mind the discussion regarding shorter sections within KiwiRAP risk mapping
discussed in section 4.4.3.

Figure 4-10: CAS-reported high-severity crashes (2005 to 2009 inclusive)

Investigating the entire route further in the KiwiRAP analysis tool (KAT) we can explore in more detail the
engineering attributes that contribute to the overall star rating of 2.8. The unfactored run-off road RPS is
high risk rural roads guide interim draft for use april 2011 final_.docx 37
 High-risk rural roads guide

7.6, while the unfactored head-on RPS is 12.5. These are then factored by a ratio of 65% to 35% for run-off
road to head-on, and the intersection score is added to give an overall RPS score. The ratio is supported by
the fact that 9 of the 30 high-severity crashes (30%) involved crossing the centreline. The high head-on RPS
indicates that a treatment that provides a central median barrier may be worth considering.

A further investigation in KAT ‘what-if’ analysis suggests that dividing this carriageway with a median
barrier would see the star rating increase from 2.8 to 3.3 which equates to a 30% reduction in the crash
rate.

(b) Using Collective and Personal Risk – SH27 Tatuanui to Matamata

Using a combination of CAS data, high-severity crashes and RAMM, a safety engineer sets out to find 20km
lengths of network on which there have been recorded 12 or more high-severity crashes (this is the
minimum number of crashes that determines whether a route is a high collective risk). It has been
recognised that a number of sections identified on SH27 are also programmed to be upgraded to a dual
carriageway; however, a section of SH27 (RS 46/5010 to RS 67/4554) has had 19 high-severity crashes in
the past five years and should be investigated further. The highway carries an average of 7760 vehicles per
day. Using the calculations in sections 4.3.4, we obtained the following results for risk:

Personal Risk = 19 high-severity crashes / (20km x 7760 AADT x 5 years x 365 days) / 108
=
19 / 2.83
= 6.8 high-severity crashes per 100 million vehicle kilometres of travel

This route is a medium personal risk (using figure 4-2). When using only this factor
and the definitions described in section 4.1, this route does not constitute a high-
risk rural road.

Collective Risk = 19 high-severity crashes / 20km / 5 years

= 0.19 high-severity crashes per km per year

This is a high collective risk (using figure 4-1) and is therefore defined as a high-risk
rural road as described in section 4.1 (even though personal risk does not constitute
a high risk rural road).

This corridor is deemed a 3 star route when using KAT (figure 4-11) which confirms that the personal risk is
moderate (figure 4-12). However, the route is also a high collective risk and therefore there should be some
potential Safe System Transformation Works that could improve safety. However, given that this is only a
medium personal risk; works should focus more on Safer Corridors treatments as shown in figure 4-12.

In addition, further analysis shows that the number of lost control crashes, combined with KiwiRAP
indicating that a large proportion of the route is bounded by severe roadside hazards within 4m of the
carriageway, suggests there is scope to treat this route by undertaking roadside improvements.

high risk rural roads guide interim draft for use april 2011 final_.docx 38
 High-risk rural roads guide

Figure 4-11: KAT map of route with star ratings displayed

Figure 4-12: Safety strategy for SH27 Tatuanui to Matamata

(c) Using collective and personal risk – local authority roads

(I) PORIRUA CITY

Porirua City has undertaken a crash study investigating fatal and serious injury crashes on their rural roads.
The study has identified two potential high-risk rural roads: Grays Road and Paekakariki Hill Road.

high risk rural roads guide interim draft for use april 2011 final_.docx 39
 High-risk rural roads guide

Paekakariki
Hill Road

Grays Road

Grays Road is approximately 5.5km long with a small section of 50km/h at the start. On the remaining
4.8km of 80km/h rural road eight high-severity crashes have occurred over the five-year period 2004 to
2009. This rural section is relatively flat, but follows a winding alignment and carries approximately 5800
vehicles per day.

Personal Risk = 8 high-severity crashes / (4.8km x 5800 AADT x 5 years x 365 days / 108)

= 8 / 0.51

= 15.7 high-severity crashes per 100 million vehicle kilometres of travel

This route is a high personal risk (using figure 4-2), and therefore defined as a high-risk
rural road.

Collective Risk = 8 high-severity crashes / 4.8km / 5 years

= 0.33 high-severity crashes / km / year

This route is a high collective risk (using figure 4-1), and therefore defined as a high-risk
rural road.

high risk rural roads guide interim draft for use april 2011 final_.docx 40
 High-risk rural roads guide

In addition, the crash reduction study identified Grays Road as a route within the district that has an
overrepresentation of crashes, particularly lost control and head-on type crashes, crashes in the wet and a
number of hit objects crashes.

In some cases, higher cost infrastructure works (such as shoulder widening and side median barriers)
cannot be justified and therefore other treatments are considered. The recommendations provided from
the safety study cover some of Safety Management treatments, including improved skid resistance,
delineation (signs and edge marker posts – EMPs) and some site-specific engineering measures. These
measures will help provide good road safety gains for the route.

Further consideration could be given to determining the gap between the current operating speeds and the
harm minimisation speed (figure 2-1) to determine whether any additional measures could be completed
to reduce overall operating speeds along the route.

Figure 4-13: Grays Road Treatment Strategy

Paekakariki Hill Road is 13km long, has had 11 high-severity crashes in the five years from 2004 to 2009 and
carries 2500 vehicles per day.

Personal Risk = 10 high-severity crashes / (13km x 2500 AADT x 5 years x 365 days / 108)

= 10 / 0.59

= 16.9 high-severity crashes per 100 million vehicle kilometres of travel

This is a high personal risk (using figure 4-2) and therefore defined as a high-risk rural road.

Collective Risk = 10 high-severity crashes / 13km / 5 years

= 0.15 high-severity crashes per km per year

This is a high collective risk route (using figure 4-1) and is defined as a high-risk rural road.

high risk rural roads guide interim draft for use april 2011 final_.docx 41
 High-risk rural roads guide

Figure 4-14: Paekakariki Treatment Strategy

(II) HASTINGS DISTRICT COUNCIL

In this example the Hastings District Council is interested in identifying potential high-risk rural routes on
their network. The process begins with plotting the high-severity crashes on rural roads in CAS, as shown in
figure 4-15. A review of the data suggests a number of possible corridors that may be worthy of further
investigation, as illustrated in figure 4-16. Although the local knowledge is needed to support the selection
of possible corridors, we have selected two for this example.

Figure 4-15: High-severity crashes on a network Figure 4-16: Possible route of interest

high risk rural roads guide interim draft for use april 2011 final_.docx 42
 High-risk rural roads guide

Route A

Route A is a section of Middle Road 24km long and carrying around 500 vehicles per day. There have been
eight high-severity crashes in five years. Using the calculations in sections 4.3.4, we obtained the following
results for risk:

Personal Risk = 8 high-severity crashes / ((25km x 500 AADT x 5 years x 365 days) / 108)
= 8 / 0.22
= 36 high-severity crashes per 100 million vehicle kilometres of travel

This is a high personal risk as per figure 4-2 and is deemed to be a high-risk rural
road using the definitions described in section 4.1.

Collective Risk = 8 high-severity crashes / 25km / 5 years

= 0.06 high-severity crashes per km per year

This is a medium collective risk as per figure 4-1, but as it has a high personal risk
this is still deemed to be a high-risk rural road.

When mapping high personal and medium collective risk onto the treatment philosophy strategy chart
(figure 4-17) for Route A, the ‘red star’ sits in the middle of Safety Management and Safe System
Transformation Works; however, the most effective treatment for this route is expected to revolve around
Safety Management as it has a low number of crashes and low traffic volume. For example, for a site where
there has been an identified issue with loss of control on bend type crashes, consideration could be given
to providing higher levels of delineation (a safer corridors treatment), including a consistent application of
curve signage along this route.

Figure 4-17: Safety Strategy for Middle Road

high risk rural roads guide interim draft for use april 2011 final_.docx 43
 High-risk rural roads guide

Route B

Route B is a 5km section of Lawn Road. Although the five high-severity crashes (2005–2009) are more
concentrated on a 3km central section, it is worth considering the route as a whole. The following analysis
relates to the 5km section, which carries an average of 3500 vehicles per day. Using the calculations in
section 4.3.4, we obtained the following results for risk:

Personal Risk = 5 high-severity crashes / ((5km x 3500 AADT x 5 years x 365) / 108)
= 5 / 0.31
= 16.12 high-severity crashes per 100 million vehicle kilometres of travel

This is a high personal risk as per figure 4-2 and is deemed to be a high-risk rural
road using the definitions described in section 4.1.

Collective Risk = 5 high-severity crashes/5km/ 5 years

= 0.2 high-severity crashes/km/year

This is a high collective risk (using figure 4-1) and is therefore defined as a high-risk
rural road as described in section 4.1.

When mapping high personal and high collective risk onto the treatment philosophy strategy chart (figure
4-18) for Route A, the ‘red star’ sits in upper quadrant in Safe System Transformation Works.

Figure 4-18: Safety Strategy for Lawn Road

Without investigating this route in detail, it is difficult to say just what major safety transformation works
may be most appropriate. However, as the carriageway is narrow (in the order of 6.6m), widening the
carriageway may well be an option. This is in keeping with the treatments that would address the majority
of the high-severity crashes and it is worth noting that, excluding the intersection crashes (which may
require other treatment options), this route would still score medium–high on both personal and collective
crash risk.

high risk rural roads guide interim draft for use april 2011 final_.docx 44
 High-risk rural roads guide

4.6 Performance and safety improvement potential of high-risk rural roads

The safety improvement potential can arise through one of two/three mechanisms:

1. The route has a high or medium–high personal crash risk, ie. the likelihood of an individual user
being involved in a crash or crash rate (crashes / 100 million vehicle kilometres of travel) is high and
improvements to the road have the potential to significantly improve the personal crash risk.

2. Large numbers of vehicles travel the route, and as a consequence the number of crashes per
kilometre is high; collective crash risk or crash density (crashes per kilometre of road) is high
despite the personal crash risk not being extreme. Improvements that generate a marginal
improvement in collective risk can still significantly reduce crash numbers.

These mechanisms are best explained with an example comparing results between the previously discussed
Route A and Route B in the above section.

• Route A carries 1000 vehicles per day over 10km and has had two high-severity crashes in the past
five years equating to 11 high-severity crashes per 100 million vehicle kilometres of travel. This
compares to a medium risk crash rate of typically 6–7 high-severity crashes per 100 million vehicle
kilometres of travel. As a result there is considerable potential for highway improvements to
reduce the personal risk. This could be in the order of a 36% reduction in personal crash risk.

Note: Extreme caution would need to be exercised in calculating the crash rate based on such a low
number of reported severe/high-severity crashes, and a 10-year period should probably be
considered.

• Route B is also 10km and carries 9000 vehicles per day and has had 11 high-severity crashes,
equating to a high-severity crash rate of 6.7 fatal and serious crashes per 100 million vehicle
kilometres of travel. Undertaking works that reduce the crash rate to a slightly lower rate of 6 high-
severity crashes per 100 million vehicle kilometres of travel would result in a 10% reduction in
personal crash risk.

Table 4-5 shows the expected benefits of improvements to these routes.

So while treatments to Route A may significantly reduce personal risk, a more moderate reduction in the
crash rate on the higher-volume Route B could give a better safety outcome, ie. a greater reduction in the
number of people killed or seriously injured. However, this will not always be the case and the purpose of
this example is to demonstrate the need to assess both collective risk and personal risk.

Combining both metrics guides the analyst toward the types of treatment philosophy strategy (figure 4-7)
that may best suit the length of highway under consideration; there is effectively a matrix of solutions.
Further detailed information on route treatments is discussed in section 4.5

high risk rural roads guide interim draft for use april 2011 final_.docx 45
 High-risk rural roads guide

1 2
Route Length AADT vkt / 5 yrs Situation Personal Risk Collective Risk High-severity
High-severity High-severity crashes / 5
(100 crashes/100 crashes/km/yr
2 years
million) million vkt

Before 11 0.040 2.0

A 10km 1000 0.1825 After 6 0.026 1.1

Reduction 36% 0.014 0.9

Before 6.7 0.220 11

B 10km 9000 1.6425 After 6 0.198 9.9

Reduction 10% 0.022 1.1

1
Annual average daily traffic
2
Vehicle kilometres of travel
Table 4-5: Example of crash reductions

4.7 Addressing other network safety issues

The identification of high-risk rural routes has, to this point, focused on overall crash risk (both personal
and collective). However, a route which has an abnormally high instance of a particular crash type, eg. loss
of control in the wet, may also have a high safety improvement potential and should be considered for
further investigation. Addressing an abnormally high instance of a particular crash type will usually provide
better than average benefits.

Two key crash types are worthy of additional consideration: wet weather crashes and those occurring in
the dark conditions. This information is shown in tables 4-6 and 4-7 respectively. These tables group both
the local road and state highway networks according to the nine climatic zones shown in figure 4-19. A list
of the allocations is contained in appendix C.

While the relative proportions of crashes occurring in different conditions will vary according to travel
volumes and patterns and operating speeds, comparing the relative proportions can provide valuable
insight into potential problems and issues.

Two examples of sites that have an abnormally high instance of either wet or dark crashes are described:

• Example (a): A section of SH29 has 16 bend lost control/head-on high-severity crashes, of which 10
(63%) occurred in the wet. SH29 is in the northern New Zealand climate zone where we would expect
about 37% of bend lost control/head-on crashes to have occurred in the wet (table 4-6).

Further investigation into the crash data on SH29 shows that 6 of the 7 bend lost control/head-on
crashes in 2008 and 2009 occurred in the wet. A review of the SCRIM10 values for this highway (in the
past two years) indicates that they have been at or below Investigatory Level.

• Example (b): A section of SH3 south of Whanganui shows 30 high-severity crashes over a 10-year
period (2000–2009), of which 14 (47%) occurred in dark conditions. Compare this to the percentage of

10
Sideway-force Coefficient Routine Investigation Machine.
high risk rural roads guide interim draft for use april 2011 final_.docx 46
 High-risk rural roads guide

high-severity crashes on open state highways in dark conditions for all crash types for this south-west
North Island area – 36% (table 4-7). This route has 30% more dark crashes than would be expected.

Further analysis could be completed on the types of crashes and then what associated treatments
could help reduce those crashes, ie. if there is a high percentage of loss of control crashes on bends in
night-time conditions, delineation treatments (signs, makings, etc) along the route should be checked.

Figure 4-19: New Zealand Climatic Zones

Climate Network All Severe Bend – Lost Straight – Lost Crossing All other
Zone Crashes Control/Head- Control/ Head- /Turning Crashes
on on

Northern local roads 24% 26% 23% 19% 17%

New Zealand
state highways 27% 37% 19% 15% 18%

Central local roads 25% 27% 28% 20% 12%

North Island
state highways 31% 41% 17% 26% 25%

Eastern local roads 16% 19% 4% 13% 16%

North Island
state highways 24% 28% 17% 24% 21%

South-west local roads 23% 24% 32% 16% 5%

North Island
state highways 28% 36% 23% 18% 24%

Northern local roads 17% 18% 8% 20% 17%

South Island
state highways 20% 25% 17% 4% 17%

high risk rural roads guide interim draft for use april 2011 final_.docx 47
 High-risk rural roads guide

Climate Network All Severe Bend – Lost Straight – Lost Crossing All other
Zone Crashes Control/Head- Control/ Head- /Turning Crashes
on on

Western local roads 18% 22% 0% 0% 29%

South Island
state highways 40% 42% 43% 33% 28%

Eastern local roads 16% 18% 21% 14% 9%

South Island
state highways 22% 25% 21% 24% 18%

Inland South local roads 21% 26% 10% 33% 13%

Island
state highways 18% 27% 14% 8% 6%

Southern local roads 32% 35% 36% 18% 22%

New Zealand
state highways 36% 43% 35% 15% 29%

All New local roads 22% 25% 23% 17% 14%

Zealand
state highways 28% 36% 22% 18% 21%

Table 4-6: Proportion of rural road fatal and serious injury crashes occurring in the wet

Climate Network Total % of high- Bend Straight Crossing All other

Region severity /Turning Crashes
Lost Control Lost Control
crashes on
Head-on Head-on
open road

Northern local roads 41% 44% 49% 20% 34%

New Zealand
state highways 37% 42% 37% 20% 35%

Central local roads 29% 33% 38% 8% 23%

North Island
state highways 32% 35% 33% 20% 29%

Eastern local roads 37% 40% 47% 17% 27%

North Island
state highways 34% 36% 35% 21% 38%

South-west local roads 36% 41% 38% 6% 32%

North Island
state highways 36% 40% 38% 26% 33%

Northern local roads 38% 43% 42% 10% 33%

South Island
state highways 28% 27% 40% 17% 25%

Western local roads 26% 35% 13% 0% 14%

South Island
state highways 32% 33% 38% 0% 22%

Eastern local roads 37% 48% 49% 17% 25%

South Island
state highways 34% 37% 41% 28% 23%

high risk rural roads guide interim draft for use april 2011 final_.docx 48
 High-risk rural roads guide

Climate Network Total % of high- Bend Straight Crossing All other

Region severity /Turning Crashes
Lost Control Lost Control
crashes on
Head-on Head-on
open road

Inland South local roads 33% 37% 34% 0% 13%

Island
state highways 33% 33% 45% 25% 24%

Southern local roads 36% 38% 34% 18% 41%

New Zealand
state highways 34% 35% 43% 12% 31%

All New local roads 37% 41% 43% 155 30%

Zealand
state highways 34% 37% 38% 215 31%

Table 4-7: Proportion of Open Road Fatal and Serious Injury crashes occurring in the dark

high risk rural roads guide interim draft for use april 2011 final_.docx 49
 High-risk rural roads guide

5 Countermeasures

Under a Safe System, system designers create and operate a transport system where people are protected
from death and serious injury. The countermeasures for high-risk rural roads relate specifically to the Safe
Roads and Roadsides and Safe Speeds components of the Safe System concepts (section 2).

This section describes a number of Safe System treatments and incremental countermeasures that have
been proven to reduce both the number and severity of crashes. A Safe System approach focuses on
providing higher-level infrastructure measures or applying Safe Speed thresholds to achieve the Safer
Journeys Strategy’s vision of ‘A safe road system increasingly free of death and serious injury’. However,
bigger-scale infrastructure projects may not be achievable in the short term, nor practicable/feasible for
low-volume high-risk rural roads. This is where lower cost and best value for money incremental
treatments may be more suitable.

5.1 Section layout

This section is divided into specific countermeasures. Each countermeasure includes the following:

Description Describes the countermeasures

Application How the countermeasures can be applied

Issues What issues are associated with using the

countermeasures

Crash reduction The effectiveness of the countermeasure.

Crash reduction percentages are sourced
from a variety of references and therefore
there are a range of values

Other benefits What other benefits do you get by using the

countermeasures

Cost Cost can be site specific (eg. grade-

separated interchange) or cost per kilometre
(such as length of median barrier). Costs are
sourced from a variety of references and
therefore there are a range of values

Treatment life Describes range of years as deterioration

can be site specific. Treatment life is sourced
from a variety of references and therefore
there are a range of values

References and References any sources of information used

guidelines/guidance to describe or evaluate the countermeasure
documents

high risk rural roads guide interim draft for use april 2011 final_.docx 50
 High-risk rural roads guide

5.1.1 Crash types

The countermeasures have been selected as they are specific to reducing the three key crash types: head-
on, run-off road and intersection crashes. There will be other crashes addressed as a result of these
countermeasures as well and other site-specific, crash-specific or safety deficiency associated treatments
can be sourced from various locations (section 5.5). In addition, motorcycle crashes are not specifically
addressed within this document; however in some cases where relevant, consideration has been given to
how specific treatments may impact on these types of road users. In addition, detailed information will be
provided within a separate high risk motorcycle guide that is currently being developed.

The safe system infrastructure countermeasures are referenced first and show the largest reduction in the
three key crash types compared to other Safer Corridors and Safety Management countermeasures.

5.2 Treatment philosophy strategy

There are five key treatment philosophies when considering countermeasures for high-risk rural roads. The
treatments fall into five main categories:

Treatment
philosophy Description
strategy

Safety Maintaining rural roads to an appropriate standard in accordance with

Maintenance specified standard criteria. Example measures include maintaining skid
resistance levels to current specifications.

Safety Measures aimed at optimising safety levels through maintenance of the

Management existing road network such as skid resistance. Generally, high personal risk
roads with low traffic volumes will not warrant significant infrastructure
investment. It will therefore be important to consider supplementing safety
management on these routes with additional speed management (speed
activated warning signs, etc), education and enforcement measures.

Safer Corridors Infrastructure and speed management measures that improve safety, though
to a lesser extent and generally at a lower cost compared to Safe System
Transformation Works. Example measures include delineation, curve warning
signs, seal widening and audio tactile profiled (ATP) markings.

Safe System Measures that eliminate or significantly reduce the potential for fatal and
Transformation serious injuries. These include infrastructure measures that physically separate
Works road users and/or speed management measures that reduce impact speeds to
survivable human tolerance limits. Example infrastructure measures include
median barriers, roadside barriers, clear zones and roundabouts.

Site-specific Although not included in the treatment philosophy strategy (figure 5-1), these
treatments measures are used where you have crash clusters (blackspots) along a route,
or just one site. Depending on where the crash cluster is located and to be
consistent with other measures along the route, the types of treatments can
be from a range of measures covering Safe System Transformation Works,
Safer Corridors, Safety Management and Safety Maintenance.

Table 5-1: Types of treatments

high risk rural roads guide interim draft for use april 2011 final_.docx 51
 High-risk rural roads guide

Table 5-1 gives a general description of the treatment philosophy, with detailed information provided in
sections 5.2.1 to 5.2.4. How to identify whether a high-risk rural road fits into a treatment category, which
is largely based on the relationship between Personal Risk, Collective Risk11, the KiwiRAP Road Protection
Score (RPS) and traffic volume as illustrated in figure 5-1. Although the RPS is currently only calculated for
State highways, other Road Controlling Authorities can calculate collective and personal risk, or define high
potential risk as outlined in section 4.

In addition, a separate addendum to this guide will be developed on an annual basis that lists sections of
high-risk rural roads in New Zealand

High-risk rural Road

High-risk rural

High-risk rural
Road

Road

Figure 5-1: Selection of treatments (treatment philosophy strategy)

The treatment philosophy strategy chart (figure 5-1) shows that where collective and personal risk is high,
Safe System Transformation Works are recommended. In this case, a treatment like a review of the speed
limit may not necessarily be the most effective strategy as it does not achieve the efficiency, function and
user expectations of the road. However, regardless of the function of the road, consideration must be given
to applying interim treatments (such as harm reduction speeds) where there is risk. Where one of the risk
metrics is lower (collective, personal or RPS), then other approaches such as Safer Corridors and Safety
Management treatments (such as a centreline treatments and roadside improvements) may be more
appropriate.

Note that, within the treatment philosophy strategy (figure 5-1), some measures will cross boundaries. Also
note that this is a guide to the types of treatments that are the most appropriate for the level or risk. It
does not mean you should discount all options and treatments when determining the best measures for

11
Definitions to calculate these risks are included in section 4.
high risk rural roads guide interim draft for use april 2011 final_.docx 52
 High-risk rural roads guide

your site or route. Cost–benefit analysis needs to be undertaken and the most cost-effective treatments
considered.

Also note that, even though the focus of this Guide is the treatment of high-risk rural roads with Safe
System infrastructure measures (mostly engineering type works), consideration needs to be given to ensure
that other Safe System initiatives are considered, in particular speed management and safer road use
projects. The following types of measures are discussed in more detail in sections 5.2.1, 5.2.2 and 5.2.3.
Note that this list is not exhaustive and some measures can fall within the safe system transformation
works and safer corridors treatments or safer corridors treatments and safety management categories, and
measures appropriate for only crash cluster (blackspot) sites may fit into all of these categories. This Guide
offers a range of countermeasures for various issues and good judgement should be applied. In summary:

Treatment Types of measures

philosophy

Safety Maintenance The legal or minimum required standard and in accordance with current specifications of:

skid resistance intervention levels

signs and markings

prioritisation for treatment of safety deficiencies and treatment of deficiencies
using conventional ‘good’ maintenance practice
Safety Management Skid resistance enhancements

increased intervention levels

high friction surfacing
Intersections

auxiliary turn lanes

sight distance

priority control
Variable signs and information

active signs

variable message signs (VMS)
Hazard maintenance

vegetation maintenance and planting policies
Safer Corridors 2+1 treatments12
Delineation12

line marking

edge marker posts

curve warning

RRPMs

ATP markings
Median treatments

flush medians

other median and centreline treatments

ATP markings – centrelines

12
Delineation will be relevant to all of the treatment philosophies.
high risk rural roads guide interim draft for use april 2011 final_.docx 53
 High-risk rural roads guide

Treatment Types of measures

philosophy

Seal widening

lane widening

shoulder widening
Passing Lanes
Geometry

Consistent Super-elevation

Curve Radius and Alignment Consistency
Speed Management Treatments

Speed Activated Warning Signs (SAWs)

Speed Thresholds

Lower the posted and operating speed
Hazard Treatment (Clear zones or Side Barriers)

Poles/trees

Open Drains and Steep Slopes
Safe System Expressways (4-laning and 2+1 treatments)13
Transformation Median Barriers
Works
Wide Medians
Roadside Barriers
Clear Zones
Intersections:

Grade Separation

Roundabouts

Speed Management
Crash Clusters All of the above treatments can be considered
(Blackspots)

13
Note that, although these are treatment strategies, they are not defined further within the countermeasures section as they are
an overall concept rather than a specific countermeasure.
high risk rural roads guide interim draft for use april 2011 final_.docx 54
 High-risk rural roads guide

5.2.1 Safe System Transformation Treatments

(a) Introduction

This section focuses on Safe System transformation treatments; these are likely to address 80–90% of the
fatal and serious crashes of the three key crash types: run-off road, head-on and intersections (section 3).
Safe System treatments are generally the higher cost infrastructure countermeasures and are developed
and implemented over a longer term; however, they can also include speed reduction measures.

The Safe System countermeasures identified in the Guide are linked to the key crash types. These are
generally presented in the order of the main crash types to be addressed, ie. median barriers address head-
on type crashes, roadside barriers and clear zones address run-off road crashes, and grade separation and
roundabouts address intersection type crashes. However, these countermeasures are not limited to
reducing only the key crash types. Where text and crash reduction figures are used from other source
material, this is referenced.

high risk rural roads guide interim draft for use april 2011 final_.docx 55
 High-risk rural roads guide

(b) Safe System Infrastructure Measures

Median barriers
Description Median barriers are generally of three types

Flexible barriers
(wire rope)

[Source: Google Maps 2010 Pro Licence]

Semi-rigid barriers
(typically steel
beam)

[Source: Google Maps 2010 Pro Licence]

Rigid barriers
(concrete)

[Source: Google Maps 2010 Pro Licence]

Application For the type of median treatments, consideration needs to be given to the traffic volumes. Where
volumes are large (ie. greater than 12,000–15,000 vpd) and head-on risk is high, then a wire rope
barrier or solid median should be used depending on the site. Where they are between 8000 vpd
and 15,000 consider wider central medians with treatments. Where they are lower than 8000 vpd
then ATP markings could be used. Where they are lower than 5000 vpd, then road marking (flush
medians) could be used. Each RCA could develop their own level of treatments.
Issues • Can restrict entry to and exit from accesses and can sometimes compromise the safety of
the object
• Restricts location of turnaround points for enforcement purposes
• Adequate end treatments and good delineation are crucial to ensure the barrier ends do
not become significant hazards
• Barriers can have significant maintenance costs that need to be compared with expected

high risk rural roads guide interim draft for use april 2011 final_.docx 56
 High-risk rural roads guide

benefits
• Often requires carriageway widening and thus ancillary effects
• Consider combining with ATP markings to reduce impacts
• While all barrier types are successful with respect to head-on fatalities, rigid barriers
seem to be less successful with serious injuries and minor injuries
Crash reduction • 30% reduction in injury crashes with the installation of a median barrier on a multi-lane
divided highway [16]
• 40% reduction of injury crashes if installing a guardrail median barrier [16]
• 30% reduction in injury crashes if installing a wire rope barrier [16]
• 14–27% reduction in total crashes [11]
• 51% decrease in mid-block injury crashes and 63% decrease in fatal and serious injury
crashes as a result of installation of a 2+1 wire rope median barrier [12]
• 100% reduction in fatal and serious crashes following installation of a 1+1 wire rope
median barrier [13]
• 40–60% reduction in head-on and run-off road crashes [3]
Other benefits Deterrent to pedestrians crossing. This can be a positive effect in situations where the location is
unsafe to cross or a negative effect in locations where pedestrians desire to cross and it would be
safe to cross without the barrier

Cost Moderate to high depending on barrier type and whether carriageway widening is required
Treatment life 10+ years
References and
guidelines
[3],[11],[12],[13],[14],[15],[16]

high risk rural roads guide interim draft for use april 2011 final_.docx 57
 High-risk rural roads guide

Wide medians

Description A wide (>9m) grassed traversable median in the centre of the road for errant vehicles to
recover.

SH1N – Tamahere, New Zealand (source: Google Earth Pro licence 2010)

Application Used on rural high traffic volume roads with more than 2 lanes in each direction
A depressed median configuration should be traversable. Median side slopes:
• should preferably be ≤1:20
• should not exceed 1:10, particularly where a median barrier is installed
• must not exceed 1:6
Issues Ongoing mowing and associated traffic management costs
Does not totally eradicate high-speed vehicle conflicts as some vehicles still traverse the
whole distance, so a barrier is still desirable. If a barrier is installed the wide space is no
longer beneficial and land cost can be saved

Crash reduction A percentage reduction in crashes by increasing the clear zone width is provided in the figure
below. [99]

Other benefits Visually more pleasing than a sealed surface with median barrier.
Assist with stormwater drainage

Cost Medium (depends on length)

Treatment life 10+ years

References and [99],[100]

guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 58
 High-risk rural roads guide

Roadside barriers
Description Roadside safety barriers include:
• flexible barriers (wire rope)
• semi-rigid barriers (typically steel beam)
• rigid barriers (concrete).
Well-designed roadside barriers reduce the severity of crashes involving errant vehicles leaving
the road and colliding with more severe roadside hazards.

Side and central wire rope barrier, SH1 Rangiriri

Application Traditionally, safety barriers have been developed for speed environments in excess of 70km/h,
where the crash severity without a barrier outweighs the severity associated with colliding with
the barrier [16]

Issues • Safety barriers are roadside hazards. Therefore, all other options for hazard reduction
should be examined before choosing to install a barrier. Barriers are designed to reduce
the severity of a collision but may also increase the collision frequency because they
are closer to the roadside than the hazard being protected and often extend over a
longer length than the hazard being protected
• Length of need must be adequately calculated and designed for
• Adequate end treatments are crucial to ensure the barrier ends do not become
significant hazards
• Barriers can have significant maintenance costs that need to be compared with
expected benefits
Crash reduction • Side barrier = 45% reduction in run-off road injury crashes [16]
• 40% reduction in total crashes [18]
Other benefits Protection of valuable or dangerous assets on roadside
Adds to the delineation of road environment, particularly on curves

Cost Moderate

Treatment life 10+ years

References and
guidelines [15],[17],[18],[19]

high risk rural roads guide interim draft for use april 2011 final_.docx 59
 High-risk rural roads guide

Clear zone
Description The clear zone is the space outside the road carriageway available for an errant vehicle to
recover or come to a rest. Where clear zones cannot be provided, roadside safety barriers
may be considered to reduce crash severity, along with measures that reduce the risk
likelihood of a vehicle running off the road.

Application • Provision of clear zones is particularly important near intersections or bends, where
the complexity of the driving task and interaction with other vehicles add to the
likelihood of run-off-road crashes. [16]
• Side slopes must not be steeper than 1:4 on embankments and 1:3 in cuttings [100]
• While full clear zone widths require in excess of 9m, the provision of 4–5m still
provides significant benefits in most locations, as shown in the figure below.
Figure 5-2: Relationship between the distance of the edge of lane and proportion of
vehicles within that distance (Source: Austroads Part 6: Roadside Design Safety and
Barriers)

Issues • Difficult to provide in many situations as full-width clear zones require space
outside most road reservations. Some situations can be high cost
• Widening the look of the road environment can create increases in operating
speeds
• Comparative costs and benefits of roadside barriers should be considered where
full clear zone width cannot be achieved
• Creating shallow drainage ditches can sometimes create land or subsurface
drainage issues

high risk rural roads guide interim draft for use april 2011 final_.docx 60
 High-risk rural roads guide

Crash reduction • Clear zones reduce the likelihood of errant vehicles striking roadside hazards by
providing clear areas for vehicles to recover
• Studies have indicated that, on high speed roads, a clear traversable width about
9m from the edge of the traffic lane allows about 80% of vehicles that run-off the
road to regain control [100]. The relationship between the distance from the edge
of the lane and proportion of drivers/vehicles that regain control is shown in the
figure above.
A crash reduction for increasing the clear zone width by a certain amount is shown in the
figure below.

Source: [19], [109] (figures relate to all casualty crashes)

25–40% reduction in run-off road crashes [3]
Other benefits Reduction in maintenance costs as roadside furniture is not hit by errant vehicles

Cost High (over $100,000) to achieve full clear zone width [distance]

Treatment life 5–20 years

References and
guidelines [3],[17],[15],[18],[19],[100]

high risk rural roads guide interim draft for use april 2011 final_.docx 61
 High-risk rural roads guide

Grade separation
Description Grade separation can be in the form of an overpass or an interchange.

Application Used where there is a high exposure to risk of death and serious injury through potential conflict
between large volumes of through traffic and large volumes of crossing/turning traffic.
Issues • Structures and ramps can be hazards if not correctly designed with adequate clearance,
adequate merge areas, forward visibility to structures and safety devices such as guard rails.
• May not be visually appealing; aesthetic design needs to be considered.
• Can create community severance.
• High cost
• Facilities should be considered for pedestrians and cyclists.
Crash • 50% reduction of injury crashes by changing an at-grade crossroads intersection to a grade
reduction separated intersection [18]
• 40–60% of intersection crashes [3]
• 100% of intersection approaches and opposing vehicles turning type crashes [101]
Other • Improved traffic flow
benefits • Reduced cost of maintaining and operating at-grade traffic control hardware
Cost High
Treatment 25+ years
life

References
[3],[18],[22],[101]
and
guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 62
 High-risk rural roads guide

Roundabouts
Description Rural roundabouts are typically high-speed roundabouts.

Oropi Road, Western Bay of Plenty (source: Google map Pro Licence)
Application Roundabouts generally provide a safer alternative to signalised and other unsignalised intersections.
Crash reductions at roundabouts are primarily attributed to two factors; reduced traffic speeds and
elimination of high-energy conflicts that typically occur at other types of at-grade intersections.

Issues • Approach volumes and movements should be reasonably balanced to ensure all
approaches function efficiently and safely
• Can be difficult for heavy commercial vehicles if not appropriately designed
• Not appropriate where there are high levels of pedestrians and cyclists; however, this is not
usually an issue in a rural location
• May require substantial land acquisition when compared with other intersection forms as a
result of having to provide appropriate alignments that manage speeds.
• They need to be carefully engineered with regards to high approach speeds.
Crash • Up to 70% reduction of all injury crashes in rural areas [18]
reduction • 60% reduction in intersection crashes [3]
• Upgrading an intersection from a rural single-lane stop sign (T-junction) to single-lane rural
roundabout produces 58% reduction in total crashes and 82% reduction in injury crashes
[25]
• 50–70% reduction in intersections, head-on, opposing vehicles and U-turn type crashes in
high-speed areas [101]
Other • Can improve traffic flow
benefits • Low maintenance requirements
• Can act as threshold to complement other speed management measures
Cost High (over $100,000)

Treatment
10+ years
life

References
and [3],[18],[23],[24],[25],[26],[115]
guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 63
 High-risk rural roads guide

(c) Safe System Speed Management

Under a Safe System, designers create and operate a transport system where road users who are
alert and compliant are protected from death and serious injury. Safe Speeds are a component of the
Safe System and should suit the function and level of safety of the road – road users understand and
comply with speed limits and drive to the conditions.

We need to consider several types of speed:

1. Speed Limits (determined by Land Transport Setting of Speeds limits rule in New Zealand)

2. Speed Zones

3. Harm Minimisation Speeds

4. Harm Reduction Speeds.

(I) SPEED LIMIT SETTING IN NEW ZEALAND

The current way to set speed limits in New Zealand is based primarily on the level of roadside
development. The higher the level of roadside development, the lower the speed limit. Some
recognition is given to road geometry and facilities, but this is secondary to the development factor.
This is the philosophy in Speed Limits New Zealand, which is part of the Land Transport Rule: Setting
of Speed Limits 2003.

Speed limits in rural areas, where there is little or no development, are 100km/h. In these areas road
geometry, terrain or other operating conditions that require a driver to slow down may not be
adequately explained to a driver by simply lowering the speed limit. It has been argued that correctly
using warning signs, delineating or changing the road environment to meet traffic demands are
better ways of managing these situations. This philosophy assumes that drivers who have sufficient
information about the road geometry, terrain and other matters will make correct decisions about
the safe and appropriate speed for any section of road they are driving along. [118]

(II) SPEED ZONES

In 2004, the NZTA developed a draft speed zoning procedure that takes into account the alignment
of the route and determines a speed limit based on the driver’s 85th percentile operating speed. This
is in contrast to the historical and still current way to set speed limits, which is based primarily on the
amount of frontage development. Speed zone procedures were developed due to some RCAs
wanting lower speed limits in situations where the safe operating speed for a road is lower than the
standard rural speed limit. There is also evidence from overseas that speed limits that match the
characteristics of the road contribute to a safer road environment. [118]

Several trials are being undertaken around the country and are to be evaluated by the NZTA for
future use.

high risk rural roads guide interim draft for use april 2011 final_.docx 64
 High-risk rural roads guide

(III) HARM MINIMISATION SPEEDS

As stated in Austroads [97]:

‘The harm minimisation speed limit targets [table 5-1] are based on the impact speeds at
which chances of a fatal outcome for those involved in the different crash types increase
rapidly. Translation of these safe speeds into a speed limit setting model for different road
categories requires an understanding of the effect of changed speed limits on crash
outcomes. There has been a substantial body of research published over the years relating
change in speed limits to change in travel speeds to changes in crashes outcomes.
Understanding these relationships will assist in analysing the role of road infrastructure in
changing speeds, and thus, the crash outcomes. Elvik et al. (2004) examined the magnitude
of a change in mean speed associated with different speed limit changes in the sub-set of
studies that evaluated such initiatives. Generally they found that the mean speed change in
km/h was about one-quarter of the speed limit change (also in km/h) as shown by the slope
of the line in best fit in Figure [5-3].’

Figure 5-3: Relationship between change in speed limit and change in mean speed
(source: Austroads AP-T140/10 – referenced from Elvik et al, 2004)

As described in a paper by C Jurewicz [27], ‘the Safe System approach seeks to regulate driver’s
speeds so that drivers respond to the level of protection offered by the road infrastructure. Under
Safe System, speed limits should be set to maximise mobility consistent with safe travel – that is, to
achieve safe mobility’. Jurewicz goes on to explain that there are four principles of speed limit setting
within a Safe System of which the ‘prime objective is harm minimisation while maintaining mobility
appropriate to road class and function’. The four principles and their application are described in
table 5-1.

high risk rural roads guide interim draft for use april 2011 final_.docx 65
 High-risk rural roads guide

1. Mobility – What There are a wide range of road classifications. It is important to select the
speed limit does the mobility-based speed limit for a section of road that matches expectations
community expect for a already held by the community. These include:
given road class and Road class and function Typical speed limit
function Rural undivided roads of low design standard and urban freeways 80km/h
Rural arterial and sub-arterial roads 100km/h
Rural freeways (m/ways) and arterials of high design standard 110km/h

2. Harm minimisation – Involves determining the maximum speed that vehicles could travel on any road
What are the safe section under consideration without the occupants or other road users risking
speeds for a road given death or serious injury.
the existing conditions Crash Type Max. Impact Harm Applicability
speed minimisati
tolerance on speed
limits
Car/motorcycle or 20–30km/h 30km/h Where vulnerable road users are
vulnerable road present in high numbers
user
Car/tree or pole 30–40km/h 40km/h Where unprotected road hazards
exist within defined clear zone
Car/car (side 50km/h 50km/h Where car/car side impact is
impact) possible >50km/h
Car/car (head-on) 70km/h 70km/h Where there is no separation
between opposing traffic streams

3. Gap analysis – Safe Gap analysis concerns the difference between the road class and function typical
System analysis speed limit, and the harm minimisation speed limit.
evaluation of the
A selected harm minimisation speed limit may no longer be applicable if the
existing level of
effect of providing road safety features is expected to raise safety to the level
protection offered by
where the revised harm minimisation speed limit matches the mobility speed
the road to identify
limit.
speed limit and
infrastructure The RCA needs to weigh up the capital investments for improved road features
improvement options against the loss of mobility due to a lower speed limit.

4. Driver perception – If the new speed limit is more than 10km/h lower than the existing mean speed,
Management of the road it is likely to require additional measures, such a road narrowing, streetscaping
environment and traffic or planting, education, publicity and enforcement
speeds if necessary

Table 5-1: Speed Limit Setting Principles

(IV) HARM REDUCTION SPEEDS

’In the short to medium term, the recommended road infrastructure features are not likely to be
provided immediately on all roads in the system to achieve harm minimisation. Because in the
interim there is gap between the driver perceptions of safe speeds and the harm minimisation safe
speeds, a compromise is necessary called harm reduction. A harm reduction approach may be
applied while road authorities move towards the Safe System, Safer Corridor Improvements.’ [27]

To determine the harm reduction speed, a ‘gap analysis’ (table 5-1) between the posted speed limit and the
harm minimisation speed can be undertaken.

high risk rural roads guide interim draft for use april 2011 final_.docx 66
 High-risk rural roads guide

5.2.2 Safer Corridor improvements

Introduction

Safer corridor improvements are those that are medium to low cost and can be implemented in a relatively
short time frame.

Consistency

It is important that safety improvements implemented on New Zealand roads are consistent along the
corridors as much as possible, and consistent with the Safe Roads and Roadsides Infrastructure objectives.

Driver awareness measures/self-explaining roads

Driver awareness measures for self-explaining roads provide clear direction and unambiguous information
to all road users which drivers can use to make decisions and modify their behaviour depending on the
design and function of a road and the associated risks. These measures are more likely on routes where
there are higher levels of personal risk but low to medium levels of collective risk.

As with the Safe System treatments, this section on Safer Corridor measures follows those treatments that
address head-on (eg. line marking), run-off road (eg. line marking, edge marker posts and other delineation
treatments) and intersection (eg. speed-activated warning signs) type crashes. However, these
countermeasures are not limited to reducing only the key crash types.

high risk rural roads guide interim draft for use april 2011 final_.docx 67
 High-risk rural roads guide

(a) Delineation

Line marking
Description Line marking is generally limited to painted midblock line marking (ie. paint) on roads. ATP
markings can be added to line marking. ATP is discussed further in this section. Other centre
treatments are described later in this section.

Marking and re-marking painted line markings can have numerous benefits:

• Centrelines can discourage overtaking and drifting from the lane and reduce head-on
type crashes by shifting lane position
• Edgelines can reduce run-off road crashes and sealed shoulder damage
Application Centrelines
• Should be used where a road is greater than 5m wide and minimum AADT of 250 vpd
[2] [89]
• May be marked on a road that is wider than 5.1m with a centreline [2a]
Edgelines
• May be marked if it is desirable [2a]
• Shall be used where the seal width is greater than 7.4m or the seal width is greater
than 6.6m and the AADT is greater than 750vpd. [2]
• Should be marked where seal width is greater than 6m and AADT is greater than
250vpd [89]
Issues Wide lines
• Marking centrelines on narrow roads can increase travel speeds and decrease the
level of safety. Marking edgelines only may be more beneficial on narrow roads
• May present a hazard to cyclists and motorcyclists depending on the type, thickness,
skid resistance, etc
Crash reduction Centreline
• 30% reduction in all crashes [29]
• 25–40% reduction in casualty crashes [16]
Edgeline
• 30% reduction in crashes on curves and straights [102]
• 25% reduction in loss of control crashes [29]

high risk rural roads guide interim draft for use april 2011 final_.docx 68
 High-risk rural roads guide

Line marking
• 8–35% reduction of total accidents [11]
• Widened edgelines (200mm) in high-risk locations (such as on curves) have been
shown to reduce crash rates.
Other benefits Edgelines can reduce shoulder damage, reducing maintenance costs

Cost Low

Treatment life 1–5 years

References and
[2],[2a],[11],[16],[29],[30],[31][89],[102]
guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 69
 High-risk rural roads guide

Edge marker posts

Description Retro-reflective edge marker posts (EMPs) give guidance to road users of the alignment of the
road ahead, especially at horizontal and vertical curves.

Application EMPs are to be Installed on the side of the road in the shoulder or attached to a guardrail
They should be used where other sources of delineation (such as line marking) are not
sufficient and cannot be correctly placed
Roads with greater than 500 vpd; however, where there are unfavourable conditions they can
be applied on any road [89]
EMPs shall be installed on all rural state highways [2]

Issues Maintenance costs can increase due to need for frequent cleaning, weed spraying and repair /
replacement of breakages
Any gaps in the sequence of EMPs reduces the overall effectiveness of the delineation
Speeds may increase at night

Crash reduction • 32–67% reduction in loss of control crashes at night [89]

• 15–18% reduction in total crashes at night [89]
• 30% crash effectiveness [16]
Other benefits Nil

Cost Low

Treatment life 1–5 years

References and
[2],[16],[32],[33],[34],[89]
guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 70
 High-risk rural roads guide

Curve warning
Description This section only relates to standard curve warning signs (including chevron signs), not speed-
activated curve warning signs. For information regarding speed-activated warning signs, see
section 5.2.2(g).

‘Warning signs inform drivers of the nature of a hazard they are approaching. Advisory signs,
including advisory speed signs, tell drivers how to navigate the hazard safely, eg. hazardous
bend signs placed on the approach to the bend can inform the driver of how the road
alignment changes. Hazardous bend signs can be mounted above an advisory speed limit sign
which states a safe speed for the bend.’ [18]

Application Used on horizontal curves. Specific installation requirements can be found in the NZTA’s Traffic
control devices manual. Note that these are known as Chevron Alignment Markers (CAM) in
Australia and other countries.
They should be consistent along a route, ie. it is undesirable to have a mixture of colours and
styles along a route.

Issues Vandalism, maintenance (dust on sign etc), correct placement

Visibility of the chevron signs in both directions needs to be considered and a sign for one
direction should not be visible to traffic travelling in the opposite direction

Crash reduction • 25–40% reduction in run-off road, head-on and intersection type crashes [3]
• 30% reduction in crashes [45]
• 40.8% reduction in crashes with the use of both curve warning and chevron signs
[103]
• 20–57% reduction in total crashes [11]
• 25% reduction in rural night-time crashes [104]
Other benefits Potential maintenance benefit as there would be reduced collisions with roadside furniture on
the curve due to drivers being better able to read the curve.

Cost Low

Treatment life 5–10 years

References and [3],[11],[18],[30],[35],[36],[45], [104]

guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 71
 High-risk rural roads guide

RRPMs
Description Retro-reflective raised pavement markers (RRPMs) or ‘cats eyes’ use retro-reflection to
improve night-time delineation along a road and are usually used in conjunction with painted
road markings to advise drivers of the alignment of the road ahead. RRPMs are most helpful
when line marking becomes difficult to see (eg during wet night-time conditions).

Application RRPMs can be used for roads where there are

• frequent horizontal and/or vertical or substandard curves
• frequent fogs or high rainfall
• hazards like narrow bridges
• high numbers of wet or night crashes
• edge marker posts that cannot be consistently located.
Issues • RRPMs have a large initial loss in retro-reflectivity due to factors such as abrasion and
build-up of road film. This improves when wet. [105]
• RRPMs can be noisy if close to residential areas – this may be a concern
• Maintenance costs increase due to need for maintenance and replacement
Crash reduction • 15–20% reduction in lost control and head-on crashes at night and during wet road
conditions [89]
• 6–18% reduction in total crashes [11]
• 5% reduction in crashes [16]
• 5.7% reduction in total crashes, and a 6.2 % reduction in daytime crashes. [106]
Other benefits Can provide audible and tactile signal when traversed by vehicle wheels

Cost Low ($1000–$5000 per km (source: TERNZ) based on 10m spacing

Treatment life 4 years (source: TERNZ)

References and
[11], [16],[30],[37],[38],[105],[106]
guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 72
 High-risk rural roads guide

Audio tactile profiled (ATP) markings (edgelines)

Description ATP markings can be provided along the edgeline and/or centreline of a roadway and provide
audio and tactile feedback to road users. Centreline ATP are discussed further in this section
under ‘median treatments’. This section focuses on edgeline ATP (also known as rumble strips,
profiled edgelines or audio tactile profiled edgelines).

Application ATP edgeline marking may replace or supplement standard edgeline markings on sections of
road where:
• traffic volumes are high
• there is a significant number of run-off road crashes in which fatigue or driver inattention
is identified
• there are specific site problems such as poor visibility, frequent or heavy rain, or night-
time crash history
As run-off road crashes resulting from fatigue or other factors can occur anywhere along a
route, ATP edgelines should be installed as a corridor treatment rather than be site specific.

Issues • May present a hazard to cyclists and motorcyclists

• Should be implemented over a continuous length rather than isolated sites
• Drainage may be a problem in high rainfall areas if associated with a raised long life line
• The auditory effect is less noticeable for larger vehicles, especially trucks
• May cause noise disturbance for adjoining land users
• Adequate shoulder width is required for cyclists outside of the ATP
• Maintenance costs
Crash reduction • Average 27% reduction in crashes, 32% reduction in run-off road crashes and 42% in fatal
run-off road crashes [41]
• 10% reduction of single vehicle run-off road crashes and 17% reduction of single vehicle
run-off road fatal crashes on rural freeways [108]
• 16% reduction of single vehicle run-off road crashes and 36% reduction of single vehicle
run-off road fatal injury crashes on rural two lane roads [108]
Other benefits Reduced shoulder maintenance (but additional cost of rumble strip maintenance)

Cost Low–moderate. The typical cost for a lane length of 10 kilometres is under $20,000. [107]

Treatment life 3–8 years

References and
[2b],[39],[40], [41],[42],[29],[43],[19],[104], [107], [108]
guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 73
 High-risk rural roads guide

(I) MEDIAN TREATMENTS

Median treatments (as opposed to median barriers) are the lower-cost countermeasures for the centre of
the road. These treatments include marking and planted/grassed median countermeasures. For other
median barrier treatments such as wire rope barriers and solid medians, see section 5.2.1.

Flush medians
Description Flush medians are white diagonal lines that are painted down the centre of the road and
mark an area that is generally as wide as one car.
Flush medians can also be narrow (ie they only provide separation between opposing traffic).
Some examples of this are also included in the centreline treatment countermeasure.

Rural flush median – SH27 Raungaiti

Application In rural areas a flush median provides separation between opposing traffic streams and a
refuge for vehicles turning into and out of side roads or driveways.
Useful where a head-on crash risk is evident or predicted

Issues Should be limited to 80km/h areas only

Potential for use of the flush median as a passing lane which may lead to rear end collisions
or lane change collisions where the flush median is also used as a turning lane. Where the
flush median is used as an area from which to turn, sight distance needs to be considered

Crash reduction • 30% reduction in injury crashes for a narrow flush median [110]
• 44% reduction in all crashes for less than 5000 vehicles per lane per day [110]
• 52% reduction in all crashes for greater than 5000 vehicles per lane per day [110]
• 90% reduction in fatal crashes [110]
• A 47% reduction in all head-on type crashes [19], [109]
• Install flush median = 25% reduction of total crashes [104]
Other benefits Improved flow – reduced delays if flush median is used as turning lane
Provision of painted medians may result in narrowing of wide lanes, encouraging slower
speeds [111]

Cost Moderate ($5,000, – $20,000)

Treatment life 1–5 years

References and
[19],,[67],[68],[69],[70],[71],[72],[73],[19], [104][109],[110],[111]
guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 74
 High-risk rural roads guide

(b) Other median and centreline treatments

Rural centreline treatments

Description Rural wide centreline (trial) markings are two centrelines (yellow or white) placed
approximately 1m apart which are used to further separate opposing flows of traffic

Application Wide centreline trial markings

(Source: NZTA Memo: 13 October 2010; centreline marking trial)

(Source: example of centreline treatment ‘Centreline Treatment countermeasures

to address cross over crashes; Levett S.P et al 2009]

Issues The carriageway width has to be increased to accommodate a central treatment. Depending
on the site, this may reduce the shoulder width to less than the ideal.

Crash reduction Separation will reduce head-on crashes.

20% reduction in casualty crashes [112]

Other benefits Provision of painted medians may result in narrowing of wide lanes, encouraging slower
speeds [111]

Cost Low [112]

Treatment life 5–10 years [112]

References and
[111],[112]
guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 75
 High-risk rural roads guide

Audio tactile profiled (ATP) markings: centrelines

Description This section discusses the impact of no-passing lines and centreline rumble strips. Edgeline
rumble strips are discussed above.
No-passing lines are solid yellow and white (dashed centreline trial) lines that indicate to a driver
that passing (or overtaking) is not allowed along the road section. No-passing lines are
increasingly being supplemented with centreline rumble strips (ATP markings) to assist where
there are number of head-on crashes.

Figure 5-4: No-passing centreline with rumble strips

Application ATP centreline marking may replace or supplement standard centreline markings on sections of
road where:
• traffic volumes are not high enough for median barrier treatments
• a significant number of road crashes are attributed to fatigue or driver inattention
• there are specific site problems such as poor visibility, frequent or heavy rain, or night-time
crash history.
ATP no-passing lines should be installed as a corridor treatment rather than be site specific and
should be used in conjunction with profiled edgelines.
Issues • May present a hazard to cyclists and motorcyclists if centreline is crossed
• Should be implemented over a continuous length
• The auditory effect is less noticeable for larger vehicles, especially trucks
• May cause noise disturbance for adjoining land users
• Insufficient passing opportunities can increase travel times and frustrate drivers. Consider
implementing passing lanes or sign posting upstream passing lanes
• Different types have different effectiveness or wear off more quickly.
Crash • 21–37% reductions in head-on and sideswipe crashes ranging from 21% to 37% of reported
reduction crashes. [41]
• On two-lane rural roads:
- 12% reduction in fatal and injury crashes [108]
- 44% reduction in fatal and injury head-on and sideswipe (opposite direction) crashes
[108]
- 25% reduction in head-on injury crashes [113]
Other benefits Nil
Cost Moderate ($5,000 – $20,000)
Treatment life 1–10 years
References and
[39],[74],[40],[75],[41],[43],[19],[26],[108]
guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 76
 High-risk rural roads guide

(c) Seal widening

Lane widening
Description In New Zealand, the typical rural road lane varies in width from 2.5m to 3.5m. There are
many instances where a narrow lane width leads to an increase in head-on and run-off road
type crashes.

Application Only a minor increase in width such as 0.5–1.0m is usually needed.

Note that any increase to above 3.6m has little benefit in reducing crash frequency [59]

Issues • Research indicates that the safety benefits are from the overall carriageway width
increase irrespective of whether it is in the land or the shoulder. Designers need to
consider (if retrofitting to an existing road) whether the benefits of an increase in
lane width outweighs the dis-benefits of a reduction in shoulder width.
• Can be costly
• Increasing lane width (with the exception of widening on curves) can increase
vehicle speeds and therefore should only be used if there is an existing crash record
related to narrow lane widths.
Crash reduction • Increase from 2.7–3.0m (13%) [58]
• Increase from 3.0–3.3m (19%) [58]
• Increase from 3.3–3.6m (5%) [58]
Other benefits Improved traffic flow

Cost Medium

Treatment life 10+

References and
[57],[58],
guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 77
 High-risk rural roads guide

Shoulder widening
Description A sealed or unsealed shoulder provides drivers with an appropriate surface on which to
regain control of an errant vehicle.

Application • Historically have aimed for consistent corridor shoulder widths, ie link width.
• Greatest benefits may come from widening on curves. Particularly on the outside of
curves.
Issues Shoulders should not be too wide (greater than about 2m) or drivers may use them as an
additional lane and benefits can reduce.

Crash reduction • 25% casualty reduction for widening shoulder to less than 1.2m [104]
• 35% reduction of casualty crashes for widening sealed shoulder to greater than
1.2m [104]
• 30% reduction of casualty crashes for shoulder sealing [104]
• Seal 1m shoulder (rural) [104]

22% reduction of property damage crashes, 19% reduction in total crashes and 14%
reduction of fatal, serious and minor injury accidents

Other benefits Allows drivers to pull off road in emergency situations or for emergency vehicle access
Sealed shoulder can be used by cyclists and pedestrians. Reduces edge break and water
ingress – hence lengthens pavement life.

Cost High

Treatment life 10–20 years

References and
[59],[60],[104]
guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 78
 High-risk rural roads guide

(d) Passing lanes

Passing lanes
Description This section describes passing lanes (also known as overtaking lanes) and slow vehicle bays

Application

(Source: NZTA Passing and Overtaking Policy)

Issues Advance signage advising motorists that a passing lane is ahead will reduce the likelihood of
drivers making passing manoeuvres in less safe areas.
There are areas where passing lanes should not be installed including sites which include
significant intersections and access ways and sites within poor geometry immediately
downstream of the passing lane.
Sight distance considerations and the length of tapers need to be considered in relation to
the operating speeds. [93]

Crash reduction • 10–25% reduction in head-on and run-off road crashes [3]
• 25% reduction in casualty crashes [92]
• 30% reduction in head-on and 10% increase in lane change crashes with passing/
overtaking lanes [93]
• 33 % reduction in fatal and injury crashes (US. Dept of Transport Federal Highway
Administration)
Other benefits Reduced driver frustration and stress

Cost High (over $100,000)

Treatment life 10+ years

References and [3], [91],[92],[25],[93],[98]

guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 79
 High-risk rural roads guide

(e) Geometry

Consistent super-elevation
Description Super-elevation is applied to a curve to improve the centripetal force keeping a vehicle on
the road surface. Camber or crossfall refers to the normal side slope of the surface used for
drainage. The following section discusses the effect super-elevation improvements can have
on road safety.

Application Super-elevation shall be applied in accordance with Austroads guidelines.

Super-elevation on curves is required where the longitudinal gradient is steeper than 8% to
achieve the design speed for the road.

Issues Super-elevation on bridges can increase construction cost of the bridge

Super-elevation around curves may increase the length of drainage paths and therefore alter
drainage requirements

Crash reduction • 50% reduction in head-on, run-off road and loss of control crashes with
reconstruction of super-elevation on curve [95]
• 40% for all crashes, 50% for run-off road crashes [110]

Note: AMF=accident modification factor. Crash reduction factor = 1 – AMF ([113]

Other benefits Improved drainage

Cost High

Treatment life 10+ years

References and [94],[98],[110]

guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 80
 High-risk rural roads guide

Curve radius and alignment consistency

Description To be completed

Application

Issues

Crash reduction

Other benefits

Cost

Treatment life

References and
guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 81
 High-risk rural roads guide

(f) Speed management treatments

Speed-activated warning signs (SAWS)

Description Speed-activated warning signs (SAWS) are electronic signs that display a message when approached by
a driver exceeding a speed threshold. They are typically used to warn the motorist of an upcoming
hazard, eg a bend, crossroad, school or worksite.

Figure 5-5: Curve advisory SAWS

(source: NZTA, 2010, p32 Specification for electronic warning signs on state highway)

Application Use to highlight and draw drivers’ attention to a particular type of hazard at a site where standard
reflectorised warning signs have been tried and have been found not to be sufficiently effective in
warning drivers to reduce their speeds and modify their behaviour so they safely negotiate the
hazardous site [114]

Issues • Vandalism
• Power supply in rural areas (solar-powered devices are available).
Crash • 35% reduction in all crashes [44]
reduction
• Up to an 11km (7mph) reduction in speeds on approach to a curve [47]

Other Speed reduction without enforcement.

benefits
SAWS can collect speed data for monitoring, although only on sign approach, not at the hazard

Cost Moderate ($20,000 – $50,000)

Treatment
5–10 years
life

References
[44],[45],[46],[47],[114],[116]
and
guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 82
 High-risk rural roads guide

Speed thresholds
Description Threshold treatments (gateways) are used to alert road users of a change in speed limit or road
environment. They are often used where a speed limit sign alone is not effective in ensuring driver
compliance with the speed limit on the approach to a town.

Figure 5-6: Threshold treatment example (source: LTSA, RTS 15, 2002) and SH3 Hamilton

Application According to the guidelines for urban-rural speed thresholds [50], thresholds are a potential traffic
management technique when one or more of the following conditions are present:
• vehicle speeds on the town outskirts or through the urban areas are inappropriately high
• all reported injury crash rates are higher than average or need to be reduced
• when vulnerable road users such as pedestrians and cyclists feature in the crash analysis.
They should only be installed on roads that have a difference in the warranted speed limits of 20km/h
or more at the rural-urban interface.
Issues The speed reduction produced by a threshold may dissipate within 250m if there are no downstream
changes in road conditions, such as decreases in road width or an increase in urban density. [48]
A threshold needs to be clearly visible with adequate sight distance to be effective.
Crash • 15–27% reduction in crashes with high visibility and physical features [48]
reduction • 11% reduction in crashes with the use of dynamic or active signs [48]
• 11–20% reduction in crashes with visual narrowing treatments [48]
Other Visually appealing entrances/gateways into smaller rural towns
benefits
Cost Low – moderate (<$50,000)
Treatment 5–10 years
life
References [48],0[49],[50],
and
guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 83
 High-risk rural roads guide

Lower the posted and operating speed

Description The default posted speed limit on New Zealand open/rural roads is 100km/h and is generally
applied to all rural roads with only limited exceptions at the present time (2010). A more suitable
speed limit for these roads might be one that more closely matches the design speed and the
present safety features, ie a speed that reflects Safe System harm minimisation speeds (section
2.3.2 and 5.2 (d)).

Application To lower the posted speed limit, surveys must be undertaken to first determine the current
operating speed. This will provide the platform from which to make a decision. If there is already an
operating speed limit that is lower than the posted speed limit, then consideration could be given
to implementing a speed limit that closely reflects the 85th percentile speed of road users, ie a
speed zone.
Safe threshold/harm minimisation speeds are discussed in section 2.3.2 and 5.2(d); this type of
speed limit should be carefully considered and consulted on prior to implementation as a typical
rural road in New Zealand. At-grade intersections and a head-on crash risk would require a 50km/h
speed limit to be introduced to eliminate most deaths and serious injuries.

Issues Where speed limits are introduced on routes where the operating speeds are higher than the limit,
then additional measures should be considered to achieve compliance. In most cases a posted
lower speed limit where one is not warranted or, where it is not supplemented with engineering
measures and enforcement, is unlikely to be complied with.

Crash • For every 10km/h reduction in operating speed, a 15–40% reduction in head-on, run-off
reduction road and intersection crashes [3]
• Change in posted and operating speed limit
• All reductions in speed limit – 15% reduction in crashes [54]
• Change in operating speed
• % reduction in crashes = 1 – ( speed before/speed after) 2 [54]

high risk rural roads guide interim draft for use april 2011 final_.docx 84
 High-risk rural roads guide

Lower the posted and operating speed

Relationship between change of mean speed and crashes (Figure 6 [118])

Other
benefits

Cost Low–medium (depends on the associated treatments needed)

Treatment
5–10 years
life

References
and [51],[52],[53],[54]
guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 85
 High-risk rural roads guide

(g) Hazard removal

Roadside hazards – poles/trees

Description Utility poles and trees are commonly located within the clear zone on New Zealand roads.
These create severe spot hazards to errant vehicles. See also section 5.2.1 on clear zones.

Application For power poles, consideration needs to be given to providing frangible poles or relocating,
undergrounding, moving outside of the clear zone or providing barriers to protect road users
from colliding with the pole.

Issues • After roadside hazards are removed, the roadside should be left in a safe condition.
Large stumps and deep holes are hazards that may remain after removal of a tree.
[66]
• Replacement of removed trees with more appropriate plants should be considered,
otherwise re-growth or soil erosion may affect the site. It is not always possible to
remove, replace or put barriers around roadside hazards. Reducing vehicle speeds is
an alternative solution. [66]
• In some Australian jurisdictions, the utility company cannot reinstate poles that
have been hit (ie need for undergrounding)
Crash reduction ‘
General removal of roadside hazards create a reduction in run-off road injury crashes of
25–40% [3]
A percentage reduction in crashes by increasing the clear zone width is provided in the figure
below. [66]

high risk rural roads guide interim draft for use april 2011 final_.docx 86
 High-risk rural roads guide

Roadside hazards – poles/trees

Providing a safety barrier to protect the hazard results in an 80% reduction in run-off road
injury crashes [98]

Other benefits Reduced maintenance costs if not being hit

Cost Low to medium (depends on whether removing, relocating, undergrounding or protecting)

Treatment life 10+ years

References and
[3],[66],[98]
guidelines

Roadside hazards – open drains and steep slopes

Description Steep slopes (eg 1 in 4) and open drains which are commonly located within the clear zone
on New Zealand roads are hazards to errant vehicles, particularly if located within the clear
zone.

Application With regards to open drains, consideration should be given to piping or providing a side
barrier to protect errant vehicles from the hazard. Steep slopes should be flattened to a
traversable gradient or a side barrier installed to protect errant vehicles from the hazard

Issues • It is not always possible to pipe drains, flatten slopes or put barriers around
roadside hazards. Reducing vehicle speeds is an alternative solution.
• High cost to reduce the hazard where drains are quite large
Crash reduction General removal of roadside hazards creates a 25–40% reduction in run-off road injury
crashes [3]
A percentage reduction in crashes by increasing the clear zone width is provided in the figure
below. [66]

high risk rural roads guide interim draft for use april 2011 final_.docx 87
 High-risk rural roads guide

Providing a safety barrier to protect the hazard results in an 80% reduction in run-off road
injury crashes [ 98]

Other benefits Nil

Cost Low to medium (depends on whether piping drains, providing flatter slopes or installing side
barriers)

Treatment life 5–10 years

References and [3],[65],[98],[110]

guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 88
 High-risk rural roads guide

5.2.3 Safety management treatments

Safety Management treatments are lower-cost measures (such as making sure what is on the road is
adequate for the environment and risk) and the most appropriate on lower-volume roads where higher
cost infrastructure measures such as solid median barriers and grade-separated intersections are not
feasible. Although not considered purely Safe System treatments, other lower-cost measures can create
some safety benefits. These low-cost measures (are not likely to address 80–90% of rural fatal and serious
crashes but may reduce the overall level of severity.

Maintain the existing asset

Figure 5-8 illustrates the difference and procedures for curves on state highways. Part of this figure shows
general maintenance and reactive and proactive improvements in relation to treating, identifying and
programming works for treatment.

high risk rural roads guide interim draft for use april 2011 final_.docx 89
 High-risk rural roads guide

Figure 5-8: Prioritise curves according to maintenance, proactive and reactive improvements (MWH)

high risk rural roads guide interim draft for use april 2011 final_.docx 90
 High-risk rural roads guide

(a) Skid resistance

Routine resurfacing of rural roads generally reduces wet-weather crashes by 15% (initially) and increases
dry weather crashes by 10% (initially) probably because of the increased speeds. The net effect is small (less
than 5% initially) and diminishes over time.

Skid resistance and intervention levels

Description Skid resistance is a very complex issue that includes factors such as speed, water and/or
detritus, micro texture, stone shape, etc, to name just a few.
A wealth of research demonstrates the strong relationship between skid resistance levels
and crash risk. These relationships support skid resistance policies such as NZTA T10/2010.
Refer to the Davies, Cenek, Henderson (2005) graph shown below.

all crashes selected crashes wet road crashes selected wet road crashes
40
Crash rate (per 10 8 vehicle-km)

35

30

25

20

15

10

0
0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70
Scrim coefficient

Research undertaken in a number of countries consistently indicates that a

disproportionately high number of crashes occur on road surfaces that have a low level of
skid resistance (particularly below 0.4–0.5) and/or low surface texture (below 1mm in Sand
Patch Texture Depth), particularly in higher-speed locations.
The strongest skid resistance/crash relationships are typically found on two-lane undivided
roads and at high demand areas such as curves and intersection approaches, which is why
higher levels of skid resistance are recommended at these locations in the NZTA T10/2010
specification. However, these areas are also subjected to the highest levels of stress and
consequently are often the hardest for which to maintain good skid resistance surfaces.
Due to the large body of evidence supporting the effectiveness of measures to improve skid
resistance and their net economic benefits, there can be high confidence in improving skid
resistance through a variety of methods.

The measurement of skid resistance can be undertaken via a variety of methods. Refer
Austroads Guide to Asset Management Part 5F: Skid Resistance (2009). The NZTA undertakes
annual surveys of the entire state highway network using the SCRIM machine. Some other
Application New Zealand RCAs also undertake periodic SCRIM surveys.
As a minimum the levels of skid resistance on the state highway network should be in
accordance with the NZTA T10/2010 requirements. Particular attention should be given to
the high-demand, high-risk areas and intersection approaches. The KiwiRAP analysis tool can

high risk rural roads guide interim draft for use april 2011 final_.docx 91
 High-risk rural roads guide

also be used to identify the higher-risk areas, evaluating the run-off road and head-on RPSs,
in conjunction with the Curve Risk Rating levels developed by the T10 procedure and held
within the RAMM database.
Methods of improving skid resistance include:
• resurfacing, particularly with a stone capable of providing a high level of skid
resistance
• slag surfacing
• high Polished StoneValue surfaces treatments, eg epoxy-based products such as
SafeGRIP
• grooving, scabbling, waterblasting, although some of these provide short-term
temporary relief only.
Consideration needs to be given to what the treatment life will be and what crash migration
might occur when high skid resistance treatments are used at some sites but not adjacent to
Issues similar situations. The desire to equalise skid resistance provision versus demands across the
network needs to be considered etc. Skid resistance will again deteriorate over time,
especially in high demand, high volume sites

Crash reductions will vary depending on the base state, level of improvement etc. Typically a
20% reduction in all crashes can be achieved by improving skid resistance levels by 0.1.
Crash reduction Higher savings of about 35% can be expected in wet road crashes.
The effect of skid resistance on crash risk, Davies, Cenek, Henderson (2005)

Other benefits xxx

Cost: varies from as low as $5/m2 for high PSV Open Graded Asphalt mixes to >$70/m2 for
Cost
high PSV surfacing.

Treatment life 3–10 years

References and
[117]
guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 92
 High-risk rural roads guide

(b) Intersections

Auxiliary turn lanes

Description Auxiliary turn lanes include right-turn lanes and left-turn lanes.
Application Auxiliary turn lanes on curves should be carefully designed with visibility to the back of the
queue in mind. It should also be clear to drivers which lane is the major through lane and
which is the auxiliary turn lane. Often it is beneficial to extend the auxiliary lane through the
curve so that it begins on the preceding straight; this improves visibility of the auxiliary lane
and reduces driver confusion.
Intersections on curves are not ideal because visibility is often limited. However, where an
intersection is required on a curve it may be beneficial to include an auxiliary turn lane so
that queuing traffic can queue away from through traffic. This reduces the risk of rear-end
collisions as a result of poor visibility.
Right turn bay: SH33 Rotorua District
(source : Google Pro licence 2010)

Issues Turn lanes should be designed long enough to accommodate vehicle deceleration clear of
through traffic, thus reducing the potential for rear-end crashes.
If a turning lane is excessively long, through drivers may enter the lane by mistake without
realising it is a turning lane. Effective signing and marking at the upstream end of the turning
lane may remedy such problems.
Auxiliary lanes are sometimes illegally used for overtaking manoeuvres, particularly where
they are located prior to a marked passing lane.
At crossroads, right turn lanes widen conflict area and lead to more crossing crashes. Where
there is enough side road traffic, roundabouts are a better option.
Typical left turn auxiliary lanes increase right turn out crashes as the left turner hides
following traffic. This is fixed by a left turn splitter island.
So we need to be very careful we are not reducing low-severity crashes, only to replace them
with high-severity types.
Right turn bays at T-junctions don’t suffer from any of these problems.
Crash reduction • 25–40% reduction in intersection crashes [3]
• 30% reduction of casualty crashes with construction of right turn (rural) and/or left
turn auxiliary lane [79]
• 33% reduction in overall injury crashes [109]
Other benefits • Improved traffic flow
• Increased intersection capacity
Cost High (over $100,000)
Treatment life 10+ years

high risk rural roads guide interim draft for use april 2011 final_.docx 93
 High-risk rural roads guide

Auxiliary turn lanes

References and
[3], [78],[79],[19],[109]
guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 94
 High-risk rural roads guide

Auxiliary turn lanes

Sight distance
Description Sight distance at an intersection is needed to allow intersecting traffic to identify gaps in the
through traffic stream and to allow through traffic to anticipate and accommodate traffic
turning in or out of an intersection. Adequate site distance is a key part of the safety
performance of an intersection. (Further analysis on safety performance will be completed as
part of the High-risk intersection guide that is currently under development.)

Sight distance restricted by vegetation looking from side road onto SH27

Application The following low-cost solutions may be implemented to restore/improve the sight distance
at intersections: [80], [83]
• Remove/cut back the vegetation.
• Relocate structures that impede sight distance (signs, safety barriers).
• Flatten embankment or batter.
• Bring forward the limit line, if this can be done safely.
Issues Can be difficult to achieve in rural areas as a low-cost measure due to nature of the road
If too much sight distance is achieved this can sometimes create ‘rolling start’ type
movements where drivers become complacent and make an early decision to pull out of an
intersection before correctly assessing the distance of approaching traffic.

Crash reduction • 30% reduction in casualty crashes [80]

• 28% reduction in total crashes [11]
Other benefits Improved lighting

Cost Low (<$5000) for minor works (such as vegetation maintenance). Likely to be undertaken by
a maintenance crew.

Treatment life 5–10 years

References and [11],[80],[81],[82]

guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 95
 High-risk rural roads guide

Priority control
Description Priority control is either a stop or give way control at an intersection. Traffic signals, although
a type of control, are not commonly used on rural roads and therefore are not discussed in
the Guide.

Application As stated in the Land Transport Rule: Traffic Control Devices, an intersection that has four or
more approaching roadways must be controlled by:

stop or give way signs; or

roundabout (section 5.2.1); or

traffic signal.
Where there are three approaching roadways, then discretion on their use with regards to
the function and traffic volumes of the road is considered by the RCA.

Issues The use of a stop sign where not warranted (ie where there is sufficient sight distance). A
stop sign should not be used to reinforce a road hierarchy or as a routine response to an
actual or perceived problem as this can decrease the effectiveness of the control type.
Further analysis on safety performance will be completed as part of the High-risk intersection
guide (currently under development).

Crash reduction

15% reduction in crashes for give way signs [85]

35% reduction in crashes for two-way stop signs at a four-legged cross intersection [85]

20% reduction in crashes for a one-way stop sign at a T-intersection [85]

Other benefits Improved traffic flow (may also be a cost due to increased delays to major road flow)

Cost Low (<$10,000)

Treatment life 5–10 years

References and [83],[84],[85],[86]

guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 96
 High-risk rural roads guide

(c) Signs and information

Active signs (vehicle activated and variable speed)

Description An active sign is a warning sign that has an electronic display component which becomes
active when the activity or hazard described by the sign (eg children on the road, out of
context curves, slow down, queues ahead) is likely to be occurring on or close to the road.
They can also include:

vehicle-activated signs

speed-activated warning signs (SAWS) see section 5.2.2(f)

variable speed signs.

Application Should be restricted to sites where the RCA considers that none of the standard warning signs
will provide adequate warning to approaching drivers.

Issues • Ownership and responsibility – eg is a ‘cattle ahead’ electronic warning sign or

flashing light the farmer’s responsibility to operate and maintain or the RCA’s
responsibility?
• Legal liability in event of power or equipment failure
• Vandalism, especially in rural areas
• Power source (solar-powered signs are available)
• Daylight saving tune adjustment
• Enforcement
Crash reduction • 35% reduction in all crashes [44]
• 30–35% reduction in crashes at rural curves and intersections [91]
Other benefits • Reduced traffic speed with speed activated and dynamic speed signs
Cost Moderate ($20,000 – $50,000)

Treatment life 5–10 years

References and [44],[87],[88],[89],[90]

guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 97
 High-risk rural roads guide

Variable message signs

Description A variable message sign (VMS) is an electronic sign in which the message can be changed in
content, form, shape, layout and/or colour. Such signs may be illuminated or otherwise. They
can be permanently located or portable.

(Source: the NZTA’s Traffic control devices manual – General requirements for traffic signs.

Application A VMS can be used to:

• actively manage traffic flows
• complement changeable message signs (CMS) to enhance travel information
measures
• warn road users of unusual conditions that may affect traffic operations on the
roading network
• provide real-time travel information to road users
• complement the fixed warning signs for temporary traffic control
Issues Vandalism
Power supply can be an issue in rural areas

Crash reduction The use of VMS can be related to a range of issues along a route such as traffic control, traffic
diversion and hazards. Therefore it is difficult to provide a crash reduction figure.

Other benefits Can provide real-time information and therefore reduce travel times and driver frustration

Cost Moderate (<$50,000)

Treatment life 10 years

References and [2],[46]

guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 98
 High-risk rural roads guide

Vegetation maintenance and planting policies

Description Vegetation maintenance can include the trimming or removing vegetation as required,
maintaining sight distances and removing hazards from clear zones. Clear guidance should be
provided in planting policies to ensure that hazards are not created.

Application RCAs should develop planning policies to maintain a clear zone. Reference can also be made
to using shrubs and plants (those that are frangible, ie with a trunk that is less than 100mm
wide) to create visual vertical narrowing effects to reduce operating speeds where it would
not compromise safety and sight distances.

Issues • Effects on the environment and community values need to be considered before
removing vegetation

• The roadside needs to be left in a safe condition following vegetation maintenance

(eg no tree stumps should be left that may be hazards)

• Regrowth of vegetation and soil erosion from removal of vegetation need to be

considered
• Overgrown vegetation can obscure signs and markings and create maintenance
issues
• Plants may interfere with sight distances at intersections and on road curves
• Overhanging tree branches can interfere with truck and bus traffic and may cause
these vehicles to swerve into adjacent lanes to avoid damage to the vehicle or load

Crash reduction ‘Generally, the removal of a fixed object like a tree results in a 50% reduction in fatality
crashes and a 25–35% reduction in non-fatality crashes at that location. However, these
values depend on the distance of the object from the traffic; the further away the fixed
object is located, the less likely is a crash which will result in an errant vehicle hitting
the object, hence there is a lesser benefit from removing it.’ [62]
General reduction in crashes are shown in the chart below [66]

high risk rural roads guide interim draft for use april 2011 final_.docx 99
 High-risk rural roads guide

Vegetation maintenance and planting policies

Other benefits
Visually appealing landscape values

Manage stormwater runoff

Provide traffic calming

Promote biodiversity

Improve air quality
Cost Low–moderate (under $20,000)

Treatment life 1–10 years

References and
[61],[62],[63],[64],[65],[66]
guidelines

high risk rural roads guide interim draft for use april 2011 final_.docx 100
 High-risk rural roads guide

5.2.4 Safety maintenance treatments

The safety maintenance treatments are generally those standards and guidelines that set the legal or
minimum required standard in accordance with current specifications. This Guide does not go into specific
information on what is considered good maintenance practice. Local policy and guidance documents should
be referenced. In summary, safety maintenance treatments can relate to:

• skid resistance intervention levels

• signs and markings

• prioritisation for treatment of safety deficiencies and treatment of deficiencies using conventional
‘good’ maintenance practice.

5.2.5 Crash clusters (blackspot) treatments

The most important aspect of developing solutions is to link the specific countermeasures to the specific
problems identified.

Typically, a crash reduction study has focused on low to medium cost engineering solutions such as signs
and markings and minor intersection improvements and these have proven to be very effective with
excellent economic returns. However, in some cases a significant crash reduction may only be achieved
through larger-scale, more substantial improvements.

When developing solutions for crash clusters, all the recommended treatments in section 5 can be
considered.

Further information on identifying and treating crash clusters can also be referenced from the New Zealand
guide to the treatment of crash locations.

5.2.6 Treatments based on collective and personal risk

(a) Collective risk approach

Collective risk is a measure of the total number of fatal and serious injury crashes per kilometre (or crash
density). These high crash density routes can help determine where the greatest safety gains (best target
for $ per kilometre) can be made from investment in infrastructure. While large infrastructure
improvements do not fit into the low-cost safety management measures for collective routes, other
treatments such as median treatments that separate oncoming traffic and improved delineation that create
more self-explaining roads would help reduce the number and severity of crashes along these routes.

(b) Personal risk approach

Personal risk is a measure of risk to each individual using the road being assessed and is based on fatal and
serious injury crashes, and traffic volumes. Sites where there are high personal risks are likely to be in
environments with more difficult terrain and where traffic volumes and road standards are often lower.

It is recognised that the nature and function of this road does not justify that high-cost infrastructure
measures be implemented to reduce fatal and serious crashes. We need to recognise that risk cannot be

high risk rural roads guide interim draft for use april 2011 final_.docx 101
 High-risk rural roads guide

eliminated through infrastructure improvements alone. We are not going to engineer our way out and
other lower cost incremental measures are more feasible where larger cost projects are not. Drivers must
share responsibility for a Safe System and take care to minimise their risk until road improvements are
made. A risk-aware driver will be better informed and more able to identify situations where they need to
modify their behaviour to respond to the conditions.

5.2.7 Treatments based on key crash types (themes)

As discussed in section 3 of this Guide, three key crash types contribute to the majority of rural road
crashes and result in fatal or serious crashes: head-on, run-off road and Intersection type crashes. We have
included crashes involving vulnerable road users. Where pedestrians and cyclists are involved in open road
crashes, the outcomes are typically severe. This is due to the often high speeds of traffic and the human
body’s limited tolerance to crash forces at speeds above 40km/h (figure 2-1).

A description of how to address key crash types with safety management and safer corridor treatments are
provided in this section, a summary of which is shown in table 5-2.

Key crash type Recommended safe Recommended safer corridor Recommended safety
system treatments treatments management treatments

Head-on Median barriers (solid/ Marked median treatments Increased intervention levels
semi-rigid and flexible)
ATP markings Improved Skid resistance
Safe System Speeds delineation (signs and
Hazard removal
markings)
Active signs
Harm reduction speeds

Run-off road Roadside barriers Wider shoulders Increased intervention levels

Clear zones ATP markings Skid resistance
Safe System Speeds Improved Delineation Planting policies
Harm reduction speeds Hazard removal

Intersections Grade-separated Wider shoulders and Intervention levels

interchanges or overpasses separated turning facilities
Skid resistance
Roundabouts Improved delineation
Improved sight visibility
Safe System Speeds Active signs through various treatment
Harm reduction speeds

Vulnerable road Separated off-road facilities Wider shoulders Improved sight visibility
users through various treatment
Safe System Speeds Improved delineation
Active signs
Harm reduction speeds

Table 5-2: Summary of key crash types’ best-value treatments

high risk rural roads guide interim draft for use april 2011 final_.docx 102
 High-risk rural roads guide

(a) Head-on

Reducing head-on type crashes will predominantly occur by providing a form of median separation through
paint markings, solid islands or other median treatments such as wire rope barriers. Each of these
treatments will provide varying levels of crash reduction. An evaluation of flexible posts as a median
treatment compared the differences in levels of safety for certain types of median treatments14. Each of
these median treatments is discussed in more detail throughout section 5.

Median treatment Treatment Indicative AADT*

philosophy*
More safe Median Barrier Safe System >12–15,000 vpd
Transformation
Works
Flush Median with ATP & Flexible Posts Safer Corridors 12,000–15,000 vpd

Wide centreline treatment with ATP 8,000 –15,000 vpd

ATP on (or next to) centrelines 5,000–8,000 vpd

Less safe Painted yellow no-overtaking lines <5,000 vpd

* not part of the TERNZ report. Added to the table

Other low-cost measures may provide some, but not the best, safety benefits to reduce these types of
crashes such as widening shoulders, ATP road marking (edge and centreline), improved signs and markings
to help prevent drivers losing control in the first instance and to provide room for errant vehicles to
recover.

The graph below shows the way the head on severe injury density varies with traffic volume. At above
about 6,000 vehicles per day there are typically more people killed or seriously injured in head on crashes
than in run off the road crashes.

14
TERNZ ‘Peka Peka passing lane safe hit post median evaluation’, August 2010.
high risk rural roads guide interim draft for use april 2011 final_.docx 103
 High-risk rural roads guide

Run-off road and Head-on Injury Densities

(Open Road State Highways excluding Motorways, 2000-2009, excluding AADTs>20,000)
1.4
High Severity Injury Density (injuries per km)

1.2

1.0 y = 9E-05x + 0.0067

R² = 0.9991

0.8

0.6
y = 0.1716ln(x) - 0.904
R² = 0.995
0.4

0.2

0.0
0 2000 4000 6000 8000 10000 12000 14000 16000
Annual Average Daily Traffic (vehicles per day)
Run Off Road F+S Injury Density Head On F+S Injury Denisty
Log. (Run Off Road F+S Injury Density) Linear (Head On F+S Injury Denisty)

Note: this chart has been referenced from Thorsten Berg. Further information will be provided.

(b) Run-off Road

Run-off road crashes include loss of control on bends. Reducing both the number and severity of these
crashes can be reduced by providing, in the first instance, treatments that reduce the likelihood that
vehicles lose control and, if they leave the road, providing roadsides that are clear of hazards.

(c) Intersections

It is somewhat more difficult to achieve good reductions in rural intersection crashes without major
infrastructure type treatments such as grade separation and roundabouts. Applying a Safe Speed threshold
is an option under the Safe System but it must be recognised that posted speed limits should consider the
function and level of safety of the road and where road users understand and comply with speed limits and
drive to the conditions.

Other safety management measures that may help reduce speeds and crashes at intersections include
providing dedicated turning lanes, improving sight visibility and delineation, restricting movements and
therefore conflict points, and providing protection or removing hazards around the intersection to reduce
the severity of crashes.

(d) Vulnerable road users

The number of crashes involving vulnerable road users (pedestrians and cyclists) on rural roads is low.
Applying a Safe Speed is an option under the Safe System but it must be recognised that posted speed

high risk rural roads guide interim draft for use april 2011 final_.docx 104
 High-risk rural roads guide

limits should consider the function and level of safety of the road and where road users understand and
comply with speed limits and drive to the conditions.

For high-risk rural sites or routes where a number of these types of crashes occur, consideration could be
given to providing the following treatments:

• separated off-road facilities

• wider shoulders

• improved delineation

• active signs.

5.3 Road safety action plans

Road safety action plans provide a sense of urgency, focus and commitment to mitigate road safety risks.
The plans record agreed processes for local road safety risks, objectives and targets, actions, and
monitoring and reviewing. Each plan is the result of collaboration by key road safety partners (eg the NZTA,
local and regional authorities, NZ Police, ACC).

It is recognised that this guide is focused on engineering treatments; however, the practitioner needs to
consider a range of countermeasures to address the safety issues and concerns of key stakeholders.

The road safety action plans are the primary mechanism for coordination of education, engineering and
enforcement approaches to road safety problems at sub-regional levels. These plans can be referenced for
any additional information on agreed measures at sites or routes of interest or updated as a result of Safe
System investigations.

high risk rural roads guide interim draft for use april 2011 final_.docx 105
 High-risk rural roads guide

5.4 Summary of countermeasures

5.4.1 Brief description

Tables 5-3, 5-4 and 5-5 briefly describe each of the countermeasures, and the range of crash reduction
percentages based on severity and crash types, cost and treatment life. Full descriptions of each
countermeasures, including a range of crash reduction percentages based on severity and crash types, are
included throughout section 5.

There are limitations to quick implementation of larger infrastructure works. Consideration should be given
to providing lower cost measures in the interim, although they should still focus on providing a staged
approach to the original recommended infrastructure treatment.

(a) Summary of Safe System Transformation Works

Treatments Summary of description and key facts

Description %* $** L***

Safe System Infrastructure Works

Median Barriers Can include flexible, semi rigid and rigid 30–100 $$ 10+

Wide Medians Central grassed median 25–40 $$–$$$ 5–20

Roadside Barriers Can include flexible, semi rigid and rigid 40–45 $$ 10+

Clear Zone Clear recovery zones outside carriageway for

25–40 $$–$$$ 5–20
errant vehicles to recover

Grade Separation Can be in the form of an overpass or interchange 40–60 $$$ 10+

Roundabouts Typically high speed rural roundabouts 50–70 $$$ 10+

Speed Management Applying Harm minimisation speeds (with

xx $$ 5-20
associated enforcement)

* Potential Crash reduction; ** Potential Cost; *** Treatment Life (years)

2
$ ≤ $50,000, $$ = $50,000 to $500,000, $$$ = $500,000+ (similar to KiwiRAP)

Table 5-3: Summary of Safe System Transformation Works

high risk rural roads guide interim draft for use april 2011 final_.docx 106
 High-risk rural roads guide

(b) Summary of Safer Corridor treatments

Treatments Summary of description and key facts

Description %* $** L***

Delineation (midblock)
Line Marking Edgeline or centreline painted markings 25–40 $–$$ 1–5

Edge Marker Posts (EMPs) EMPs give guidance to the driver indicating 15–67 $–$$ 1–5
the alignment of the road ahead, especially at
horizontal and vertical curves.

Curve Warning Standard curve warning signs (including 20–57 $ 5–10

chevron signs)

RRPMs Retro-reflective road pavement markers 5–20 $ 1–5

(RRPMs) or road studs (‘cats eyes’) use retro-
reflection to improve night-time visibility

ATP Rumble strips can be provided along the 10–42 $–$$ 5–10
edgeline and centreline of a roadway which
provide an audible notice when traversed

Median Treatments

Flush Median White diagonal lines painted in the centre of 30–52 $ 1–5
the road, normally about one car width

Other Median and A central marked area, normally narrower 20 $ 5–10

Centreline Treatments than typical flush median. Does not
necessarily contain diagonal lines

ATP – Centrelines Audio tactile no-passing lines 12–44 $ 1–10

Seal Widening

Lane Widening The typical rural road lane varies in width 5–19 $$ 10+
from 2.8m to 3.5m in New Zealand. There are
many instances where a narrow lane
increases head-on and run-off road type
crashes

Shoulder Widening A sealed shoulder provides drivers with a 14–35 $$ 10–20

dependable surface to regain control of an
errant vehicle.

Passing Lanes

Passing lanes Also known as overtaking lanes. Includes slow 10–33 $$ 10

vehicle bays

Geometry

Consistent Super-elevation Super-elevation (or crossfall/camber) is 40–50 $$ 10+

applied to a road for drainage and to improve

high risk rural roads guide interim draft for use april 2011 final_.docx 107
 High-risk rural roads guide

Treatments Summary of description and key facts

Description %* $** L***

centripetal force

Speed Management treatments

Speed Activated Warning Digital signs that display a message when 35 $ 5–10
Sign (SAW) approached by a driver exceeding a speed
threshold.

Speed Thresholds Gateways treatments that are used to 11–27 $ 5–10

indicate to a driver a change of speed
environment.

Lower the posted Speed Applying Harm reduction speeds (with >30% $ 5-20
Limit associated enforcement) (figure
2-3

Hazard Removal

Roadside Hazards – Power poles and trees located close to edge of 10–40 $–$$ 5–10
Poles/Trees road create hazards for errant vehicles

Roadside Hazards – Open drains and steep slopes located close to 10–40 $–$$ 5–10
Open Drains/Gorges edge of road create hazards for errant vehicles

* Potential crash reduction; ** Potential Cost; *** Treatment Life (years)

$ ≤ $50,000, SS = $50,000–$500,000, $$$ = $500,000+ (similar to KiwiRAP)

Table 5-4: Summary of Safer Corridors Treatments

high risk rural roads guide interim draft for use april 2011 final_.docx 108
 High-risk rural roads guide

(c) Summary of Safety Management treatments

Treatments Summary of description and key facts

Description %* $** L***

Skid Resistance
Enhancements
Increased Intervention Minimum levels of skid resistance for roads 30–50 $–$$ 3–10
Levels

Intersections

Auxiliary Turn Lanes Auxiliary turn lanes include right and left turn 25–40 $$ 10+
lanes

Sight Distance Allows drivers sufficient time with which to adapt 28–30 $ 5–10
to other road users turning in/out of intersections

Controls Either stop or give way control at intersections 15–35 $ 5–10

Hazard Removal

Vegetation Includes trimming vegetation and providing 10–50 $ 1–10

Maintenance and planting policies to prevent hazard being created
Planting Policies

Variable Signs and

Information
Active Signs (vehicle Warning signs that have electronic display 30–35 $ 5–10
activated/variable components which become active with hazardous
speed) activity on the road

VMS Warning signs that have electronic display N/A15 $$ 10

components and the message can be changed

* Potential Crash reduction; ** Potential Cost; *** Treatment Life (years)

$ ≤ $50,000, SS = $50,000–$500,000, $$$ = $500,000+ ( similar to KiwiRAP)

Table 5-5: Summary of Safety Management Treatments

15
Difficult to determine crash reduction percentage as there are many different forms and messages contained within a VMS.
high risk rural roads guide interim draft for use april 2011 final_.docx 109
 High-risk rural roads guide

5.5 Case studies

Sections 5.4.1, 5.4.2 and 5.4.3 summarise historical high-risk rural routes, treatments and evaluation.

5.5.1 Centennial Highway16

This paper reports on the performance of a wire

rope barrier (WRB) on a narrow median
installation on Centennial Highway, New Zealand.
Since the speed limit was reduced from 100km/h
to 80km/h and the median WRB was installed, no
fatalities or serious injuries have occurred. The
median barrier was introduced in two stages
(front section 700m) and has been very successful
in reducing road trauma.

Prior to stage one the average annual social cost of crashes on Centennial Highway was $5,796,889. This
has since reduced to an average social cost of $65,400 per year. Surveillance of the Centennial Highway
median WRB showed that vehicles generally sustained relatively little damage when they struck the
barrier and were often observed to drive away after the impact. Drivers also tended to travel more
centrally within their lane with the barrier in place. While the narrow median on Centennial Highway has
resulted in an increase in maintenance costs due to impacts on the WRB, this cost has been significantly
offset by reductions in trauma costs.

The use of a narrow median WRB has proven to significantly reduce crash severity and is considered an
appropriate solution when retrofitting existing roads, particularly in constrained environments. However,
it is recommended that wider medians be adopted wherever possible to minimise the associated
maintenance costs. Ideally, the median width should provide at least sufficient space to fully
accommodate the design deflection of the selected barrier system.

16
‘Evaluation of Narrow Median Wire Rope Barrier Installation on Centennial Highway, New Zealand’ By F. Marsh and M. Pilgrim
{2010}

high risk rural roads guide interim draft for use april 2011 final_.docx 110
 High-risk rural roads guide

5.5.2 Rangiriri17

In 2004/05 a central median wire rope

barrier (WRB), based on the Swedish 2+1
system, was installed on a 9km section
of State Highway 1 between Auckland
and Hamilton.

While the installation required

compromises to be made, and there has
been a significant increase in the
number of crashes post-installation,
there has been a correspondingly
significant reduction in crash severity.

A number of issues are associated with the performance of the installed WRB, from which solutions have
been identified, that can be applied to existing and future WRB installations.

The frequency of barrier strike appears to be related to the relatively narrow median on which the WRB
was installed. A wider median could have reduced the potential for barrier strikes, accommodated
deflection of the barrier during a crash, improved visibility for drivers and better facilitated maintenance
operations. However, there were potentially significant construction cost implications associated with
this.

The key conclusion of the paper is that, even in cases where some design compromises are required, an
appropriately installed and maintained WRB system can significantly reduce the severity of crashes along
a section of highway -

Crash Severity Numbers and Social Costs per Year

WRB Installed 2004 - 2005
40 14

35 12

30
10

25
Social Cost ($M)
No. of Crashes

8
20
6
15

4
10

5 2

0 0
1999 2000 2001 2002 2003 2006 2007 2008 2009

Crash Year

Fatal Serious Minor Non-injury Cost

17
Case study referenced from paper ‘Longswamp to Rangiriri Wire Rope Barrier increased crash numbers but improved road safety.’
Crowther S, Swears, ARRB 2010 Conference

high risk rural roads guide interim draft for use april 2011 final_.docx 111
 High-risk rural roads guide

5.6 Source Material

A number of documents and guidelines are referenced in this Guide to provide more detailed information.
In addition to those documents, the following web-based tools and manuals are considered good sources
of road safety information for Safe System designers.

5.6.1 iRAP Road Safety Toolkit18

The international Road Assessment Programme (iRAP) Road Safety Toolkit is a web-based tool that allows
users to identify treatments, road users, crash types and management policies.

The Road Safety Toolkit provides free information on the causes and prevention of serious road crashes.
Building on decades of road safety research, the Toolkit helps engineers, planners and policy makers
develop safety plans for car occupants, motorcyclists, pedestrians, cyclists, heavy vehicle occupants and
public transport users.

The Road Safety Toolkit is the result of collaboration between iRAP, the Global Transport Knowledge
Partnership (gTKP), the World Bank Global Road Safety Facility and ARRB Group.

Further information can be found at www.toolkit.irap.org/

5.6.2 Austroads Road Safety Engineering Toolkit 19

The Road Safety Engineering Toolkit is a reference tool for road engineering practitioners. It outlines best-
practice, low-cost, high-return road environment measures to achieve a reduction in road trauma.

The Toolkit seeks to reduce the severity and frequency of crashes involving road environment factors. It
draws together existing road safety engineering knowledge as far as possible into one Toolkit for easy
access by practitioners. The presented knowledge has been updated with recent experience from local and
state government agencies, and with the results of comprehensive road safety research reviews.

The Toolkit is a ‘living’ document including updates and revisions, so that more recent safety ‘wins’ are
captured and disseminated.

The information included in the Toolkit is based on extensive research into the effectiveness of crash
countermeasures. Nonetheless, the Toolkit is not a replacement for sound engineering judgement or good
design. In-depth investigation is required at locations which have a crash history or high crash risk to
identify causes or potential causes of crashes. If necessary, seek professional advice from practitioners
specialising in road safety engineering.

Further information can be found at www.engtoolkit.com.au/

5.6.3 KiwiRAP Assessment Tool (KAT)

The KiwiRAP star rating process captured and evaluated a range of safety engineering features on over
10,000km of rural state highways and assigned relative levels of risk based on their presence, absence or
quality. This generates an RPS which in turn produces the star rating. The KiwiRAP analysis tool (KAT) is an

18
Information for the general description referenced from www.toolkit.irap.org/
19
Information for the general description referenced from www.engtoolkit.com.au/
high risk rural roads guide interim draft for use april 2011 final_.docx 112
 High-risk rural roads guide

interrogatable database that stores the base rated data and allows search queries to be undertaken on
regions, networks and highways, or allows road sections to be identified by a range of features or feature
conditions for review or comparison. KAT also allows analyses to be undertaken on the safety risk effects of
altering one or more of the features. The software tool also allows the base rated data to be updated as
changes occur, allowing performance monitoring of the network over time.

5.6.4 Rune Elvik – Handbook of Road Safety Measures

The handbook of road safety measures contains summaries of the effects of 128 road
safety measures.

This book covers various areas of road safety including: crash reduction; the results of
more than 1700 road safety evaluation studies, traffic control; vehicle inspection; driver
training; publicity campaigns; police enforcement; and, general policy instruments. It also
covers topics such as post-crash care and speed cameras.

5.6.5 PIARC Countermeasures

This catalogue presents a set of common design errors and suggests a range of measures
to overcome them; it also gives an indication of the comparative countermeasure costs to
facilitate prioritisation of the work. The catalogue can be used both as a proactive safety
tool to ensure the design faults do not arise in the first place, or as a reactive safety tool to
help design cost-effective countermeasures where problems already exist on the road
network.

high risk rural roads guide interim draft for use april 2011 final_.docx 113
 High-risk rural roads guide

6 Programme implementation, monitoring and evaluation

6.1 Introduction

The focus of this Guide is to identify high-risk rural roads and develop countermeasures that reduce fatal and
serious crashes along a route or at a site. Once these routes and measures have been identified a suitable
programme of implementation is important, along with a system to monitor the effectiveness of these
countermeasures. ‘The effectiveness of treatments guides investment in road safety programs and reliable and
accurate information will be necessary to determine the effectiveness of treatments’.20

In this section we look at issues associated with developing programmes for treating high-risk rural routes, and
then monitoring the effectiveness of those programmes to:

1. identify the benefits or rather the effectiveness of the various treatments

2. identify the most effective packages of treatments
3. assess the levels of funding that may be required to achieve various levels of crash reduction
4. ‘prove’ that funding has been spent wisely.

Figure 6.1 is a modified version of the safety management triangle. Working from the base up, the foundation
of this triangle is the identification and analysis of crash issues, which would include the means of identifying
high-risk rural roads, corridors or sites (section 4).

Figure 6-1: Road safety management triangle

20
National Risk Assessment Model, Program Development and Trials: Interim Report 2009/10; ARRB

high risk rural roads guide interim draft for use april 2011 final_.docx 114
 High-risk rural roads guide

Having identified our sites/routes and clarified our safety concerns, this Guide discusses some possible
treatments or strategies that could be used to improve the safety of our high-risk rural routes, and reduce the
risk of death or serious injury, the primary outcome.

In an ideal world, the analysis of the effectiveness of each treatment or programme item would be assessed by
applying only one specific treatment to a range of sites and monitoring the performance of the treatment over
time, before moving on to apply the next treatment. However in New Zealand the number of people killed or
seriously injured in any one location is too small and the risk of doing nothing is too severe. Combined with the
delays associated with the post-implementation data collection and the immorality of ‘playing with people’s
lives’, this precludes such a purest approach. So in order to facilitate the necessary analysis, the road safety
management triangle introduces the concept of intermediate and secondary outcomes.

In this section we begin by looking at the development of a programme of treatments, and how to establish the
appropriate intermediate measures. We then look at the monitoring site-specific secondary and primary
measures.

6.2 Programme development

While the focus of the Guide is on high-risk rural routes – those typically located in the upper and right side
parts of Figure 6-2 – it is important to remember low-cost safety management treatments still apply to the
bottom left quadrant.

Figure 6-2: Safety improvement strategies

high risk rural roads guide interim draft for use april 2011 final_.docx 115
 High-risk rural roads guide

The assessment of rural road risks in section 4 identifies the longer-term plan for a particular highway. In some
regions there will be no rural road sections in the upper and right side portions of figure 6-1, but that does not
mean a programme of ongoing safety improvement should be abandoned; it just needs to be tailored to fit the
appropriate end game.

Having identified a road section, the crashes on that section must be investigated to identify the crash and risk
issues that must be addressed. Risk issues are road safety deficiencies issues which are not supported by a
crash history – in essence a predicted crash risk rather than actual crash risk. In these investigations the road
safety practitioner should look to understand:

• crash patterns for both:

– high-severity crashes, ie those resulting in death or serious injury

– all injury crashes (the inclusion of minor injury crashes will better highlight spatial, temporal
and crash movement or factor patterns)

• the spatial location of crashes, clustered or distributed

• key risk factors such as lengths of hazardous roadsides

• consistency of expectation and provision of road and roadside infrastructure.

As with other crash reduction studies, the treatments to address each actual crash (such as loss of control off
road on bend) or risk issue (such as KiwiRAP’s run-off road RPS) should be identified. One means of doing this is
to construct a table as shown in table 6-1.

Issue Treatment Strategy Application Effectiveness Cost

Loss of control Speed Safety management Route

off road on management/zone
bend ( actual
crashes) Delineation and Safety management Specific location &
signage Route

Surfacing Safety management Specific location &

OR Route

Sealed shoulder/ Safer corridors Specific location &

lane width Route
Low KiwiRAP
run-off road
Roadside hazard Safer corridors Specific location &
RPS score
management Route
(predicted
crash risk)
Realignment Transformation works Route

Table 6-1: Treatment table

high risk rural roads guide interim draft for use april 2011 final_.docx 116
 High-risk rural roads guide

With any treatments consideration needs to be given to the benefits of one against another to determine cost
effectiveness. Countermeasures can be applied to corridors and areas (ie Mass Action) as well as point specific
hazards. Certain countermeasures may be more effective in certain areas and more effective as network
treatments rather than smaller site-specific areas. For example, a speed zone is more likely to be applied to a
route and area rather than a shorter length or road.

If point hazard treatments are preferable it is desirable to programme the upgrade of those point hazards
along routes and within areas to maintain a consistent environment for road users.

6.2.1 Programme Prioritisation – Focus on Infrastructure Transformation Countermeasures

The main focus for a Safe System approach is to address fatal and serious crashes and the three key crash
types. A report by Monash University21 discusses infrastructure versus fundamental improvements, specifically
with regard to the pattern of key crash types on rural roads. The most strategically important transformational
countermeasures found within the literature review were:

Crashworthy Barrier system – when used over extended lengths of high-speed rural road, barrier systems
have the potential to reduce fatal and serious injuries to the occupants of errant vehicles by around 90%,
with conservatively estimated benefit–cost ratios of around eight. Flexible barrier systems can address
two major rural crash categories, namely single vehicle and head-on crashes, on straight or curved road
sections, without the need for costly road duplication and/or geometric improvements to rural
infrastructure.

Grade-separated Interchange – can virtually eliminate intersection crashes (potentially 100% effective)
but the high cost of grade-separation makes them less attractive than some other alternatives.

Roundabouts – can reduce casualty risk at intersections by between 70% and 80%, and crash costs by
around 90%. In addition, they have been found to result in benefit–cost ratios of around 19 when
constructed at rural intersections with a high crash record.

Although these measures cost more, they have significant benefits. It is desirable to plan for their
implementation in the long term.

6.2.2 Challenges to Implementation

A Safe System report22 identified some of the challenges to implementing a Safe System. These are:

Cost, particularly in relation to the main infrastructure type countermeasures (ie roundabout, grade
separation, median barriers).

Construction timeframes.

Community support: it would be particularly important to gain acceptance from the community with the
introduction of lower speed limits; otherwise compliance would be minimal.

Inter-agency planning: there is a strong need for organisations to work together to deliver a Safe System.

21
Cost effectiveness infrastructure measures on rural roads; Oxley, J; Corben B; et al, Monash University, April 2004.
22
Safe System Infrastructure – National Roundtable Report; Turner, B; Tziotis, M; Cairney, P; Jurewicz, C. ARRB Research Report AA 370

high risk rural roads guide interim draft for use april 2011 final_.docx 117
 High-risk rural roads guide

Incompatibility of travel modes: specific countermeasures may produce incompatibility between different
road users.

6.2.3 Programme Implementation

Consideration of the types of countermeasures and planning is important in providing the best possible
outcome in terms of reducing the number of fatal and serious crashes along a route, site or area.

Turner, Tziotis, Cairney and Jurewicz state:

‘the timeframe for implementation of a Safe System infrastructure is an important consideration. A step
process will most likely be required, and over a long term period (eg 20 years). Some initiatives can be
implemented immediately, but others require longer. With a focus on longer term objectives, total cost
can be divided over a larger number of years. The total costs per year may not be substantially more
than amounts currently spent on safety (including through maintenance and major projects budget)
although likely costs still need to be determined.’23

6.2.4 Focus on Incremental Improvements Across Network

The focus for a programme of works should concentrate on incremental improvements across networks to help
achieve larger benefit–cost ratios. So what are incremental improvements?

Having identified that a route requires larger infrastructure/capital projects to produce a Safe System
transformation, the end result has been confirmed. However, given the limited funding and associated
priorities, together with the lead time associated with getting major infrastructure projects to construction,
doing nothing until that project eventuates continues to place drivers at an increased risk of death or serious
injury.

As responsible road safety practitioners and network managers, we need to consider this risk. Incremental
improvements are viable if they:
• contribute to a reduction in the cost of the final project, ie providing incremental benefit and costs; or
• return an economic road safety benefit over the intervening period, ie between now and the realistic date
for delivery of the major project.

For example, the Safe System transformation of a high-volume section of highway may be the construction of a
physically separated dual carriageway in 10 years. If the long-term solution will be developed on essentially the
same alignment, an incremental approach could be adopted. Assuming the safety-related maintenance is being
undertaken, an incremental solution could involve, in the first instance, widening sections of the carriageway,
then installing a median barrier (1+1), improving roadside hazard management, installing passing lanes and
progressively moving to a 2+1 lane arrangement, before moving to a 2+2 lane arrangement. The works
undertaken at each step contribute, at least in part, to the overall Safe System Transformation Works, reducing
the costs associated with the final project.

23
Safe System Infrastructure National Roundtable Report; ARRB Research report 370; May 2009; Turner,B; Tziotis,M; Cairney, P; and
Jurewicz, C.

high risk rural roads guide interim draft for use april 2011 final_.docx 118
 High-risk rural roads guide

If, however, the final solution involves a completely new alignment, any proposed works will have a reduced
economic life and should be analysed over the pre-implementation period.
6.2.5 Consistency

Consistency of the road environment is important. Road users are generally unaware of driving over district
and regional boundaries and therefore it is essential that road users receive a consistent set of messages about
the road environment. This consistency is assisted by the application of legislation, policies and guidelines that
dictate the safe and efficient use of traffic signs, markings and other traffic control devices.

6.3 Monitoring and Evaluation

Monitoring and evaluation of Safe System treatments is important in gauging the effectiveness of different
treatments. This is also important when developing types of countermeasures for specific issues and
implementation procedures for future programmes. Specifically:

Monitoring involves an assessment of progress and collecting information through the course of a project,
can be before, during and after to gather results for which to do an evaluation (section 6.3.1).

An evaluation analyses the results of monitoring and determines the results and effectiveness of the types
of treatments used (section 6.3.2).

6.3.1 Monitoring

Monitoring and collection of data for evaluation will help to identify if road safety has been improved
‘Systematic recording of data and analysis of trends from which goals and targets (section 6.3.3) can be
calculated allows the most recent values of measures and their trends to be compared with target levels.’24

6.3.2 Evaluation

As stated in Austroads Report ST 1571 25 the role of evaluation is to:

ensure that recently-delivered programmes are effective and enable remedial action if they are not.

build up a reliable knowledge base about the effectiveness of different interventions, which will allow
more effective programmes to be developed in the future.

There are effectively two levels of monitoring and evaluation:

• Strategic monitoring and then evaluating the effectiveness of the overall programme or strategy, which is
made up of various projects or initiatives
• Individual monitoring and evaluating of specific projects or initiatives that make up the overall
programme or strategy.

While good monitoring and evaluation will support future road safety improvement programmes, the
monitoring and evaluation effort should not consume excessive amounts of staff time or other resources that

24
Guide to Road Safety part 2: Road Safety Strategy and Evaluation
25
National Risk Assessment Model, Program Development and Trials; Interim Report 2009/10; Turner, B; Jurewicz,C; et al

high risk rural roads guide interim draft for use april 2011 final_.docx 119
 High-risk rural roads guide

could be used to undertake more road safety initiatives. As a general observation, many people and
organisations undertake little or no monitoring, while others seek to monitor an extraordinary number of
items, arguing that the various measures do not take account of every minute impact.

In the following sections we look at the monitoring and evaluation of individual initiatives or projects and then
the monitoring of the overall strategy.

6.3.3 Evaluation of Treatments

A report by Austroads 26 summarises the role of evaluation, which is to:

• ensure that recently-delivered programmes are effective and to enable remedial action if they are not;

• build up a reliable knowledge base about the effectiveness of different interventions, which will allow
more effective programmes to be developed in the future.

In addition, ‘Only by monitoring the effects of real treatments which have been applied in real traffic situations
can a reliable picture of a countermeasure’s effects be obtained. Theoretical analyses form first principles and
simulation provides valuable insights. The quality of an Evaluation is measured by its “Validity”’ (Elvik et al,
2009). The Austroads document contains detailed information on validity and general trends, changes to traffic
flows, regression to the mean, crash risk mitigation and adjustment periods, which is located in section 6 of
that report. Sections 6.3.4–6.3.6 summarise section 6 of the Austroads document.

(a) Evaluation Methods

Evaluation is essential to determine the successfulness of individual types of countermeasures used. Any
evaluation completed will help develop future implementation programmes for a Safe System. The Austroads
document details information on three basic categories of evaluation of traffic studies:

• Observational Cross section studies (OCS)

• Observational before and after studies (OBAS)

• Experimental before and after studies (EBAS).

Details regarding the various analysis statistics are not covered here but can be found in Austroads Guide to
Road Safety ‘Part 8 Treatment of Crash Locations’ and include:

• chi-squared test of crash frequencies

• comparisons of crash rates using the paired t-test

• comparisons of proportions using z-test.

26
National Risk Assessment Model, Program Development and Trials: Interim Report 2009/2010

high risk rural roads guide interim draft for use april 2011 final_.docx 120
 High-risk rural roads guide

A summary of those studies and evaluation methods is outlined in table 6-2:

Description Study Type
OCS OBAS EBAS
Description Compare sites with and without Compare sites before treatment Same as OBAS, but the treatment
the treatment, usually over the installed with the same sites selection is determined by the
same period. after treatment has been priorities and operational
installed; treatment selection procedures of the RCA, ie crash
logic is as observed in the history; safety improvements.
studies.

Where used Often used where no suitable Most commonly used for road Should be applied in addition to
before or after data is available safety evaluations. observational studies. Designed to
for the purpose of performing control confounding factors across
an evaluation. treatment and control sites.

Other Three types: Naive CS, Three common types: Naive Also known as Randomised
information Regression CS & Matched CS. (simple), Before and after Controlled Trials (RTC). Most
Use control sites for comparison studies with control sites and effective evaluation method.
with treatment sites. Empirical Bayes Method.

Risks Avoid this method as it is Fundamental requirement is Key method in laboratory testing,
difficult to eliminate the that the treatment introduced but rarely used in road safety
influencing factors which can must retain much of the studies because:
lead to un-interpretable or original attributes of the study

treatment programmes are
seriously misleading results site.
subject to budget constraints,
(Elvik, 2008). Potential Issues
and therefore only sites having
include:
the highest expected BB ratios

bias would receive treatment

differences in traffic priority
volume, traffic composition
ethical issues arise, as it does
and annual driving distance not treat all of the high crash

differences in other locations
relevant risk such as roads
it typically results in lower
through mountains and crash reduction estimates
self-selection bias.
decision makers do not
understand the benefits of the
RTC method.
Validity Therefore validity depends on Valid only if we can be Eliminate all biases arising from
selecting the control sites that reasonably confident that no treating sites with the worst crash
would have same safety other factors that may impact history; regression to mean effects
performance as the treatment on safety apply in a biased
sites without the treatments manner. Need to account for
being present. regression to mean effects

Table 6-2: Summary of Austroads Evaluation Methods

high risk rural roads guide interim draft for use april 2011 final_.docx 121
 High-risk rural roads guide

(b) Data requirements

The key to effective evaluation of specific works is to ensure the data required for evaluation of individual
projects, treatments or initiatives is collected over the course of the programme and staff are not faced with
the arduous task of trawling back through project files to identify when and which works have been completed.

The best way of addressing this issue is to ensure the project monitoring is step up at the start of a project and,
as discussed above, the entering of monitoring data forms part of the contract, in-house service agreement or
task plan for the works. This is best done using the Crash Analysis System (CAS).

CAS has the ability to record to types of site:

– Sites of interest (figure 6-3) – these are simply locations that users can identify spatially and for
which crash data can be recalled. Once recalled the user can then analyse the effects of a
programme of works. Recording works as sites of interest relies on recording key data about the
works undertaken elsewhere, so sites of interest may be useful when monitoring areas to determine
ongoing trends, whether these are related to improvement programmes or not.

– Safety Improvement Projects or Crash Reduction Monitoring Sites (Figures 6-4 and 6-5) – these
two types of site are essentially the same in terms of the inputs required. The first data entry screen
(figure 6-3) allows the user to input site description data (the sites are spatially defined later in the
process).

high risk rural roads guide interim draft for use april 2011 final_.docx 122
 High-risk rural roads guide

Figure 6-3: CAS sites of interest

Figure 6-4: Monitoring site data entry screen 1

high risk rural roads guide interim draft for use april 2011 final_.docx 123
 High-risk rural roads guide

Figure 6-5: Monitoring site data entry screen 2

The second screen is used to identify the crash issues at the site and explicitly links the proposed solutions to
the problems and the expected crash savings. While entering projects as safety improvement projects or
monitoring sites involves a larger amount of more detailed data, monitoring site performance data
automatically adjusts for potential regression to the mean impacts.

It is, however, important to recognise that, under the Safe System approach, we are looking toward more
proactive treatment, rather than waiting for crash histories to develop, and implementing synergetic corridor
treatments to increase consistency. It is therefore quite likely that in some situations works will be undertaken
with a view to decreasing risks rather than to treat a documented crash history.

In such situations treatment performance monitoring may well be invalid because of a lack of a ‘before’ crash
risk. In these situations we need to monitor and evaluate our programme as a whole.

6.3.4 Monitoring and Evaluation Performance Measures

Referring back to Figure 6.1 three types of road safety measures are available for monitoring and evaluation:

Primary outcomes – the reduction in the number of people killed or seriously injured as a result of
road trauma.

high risk rural roads guide interim draft for use april 2011 final_.docx 124
 High-risk rural roads guide

Secondary performance measures, such as reductions in the collective and personal risk for all injury
crashes. They can be measured in terms of reported crash numbers and patterns of crash types and
factors. For Safer Roads and Roadside issues, reductions in predictive collective and personal risk
scores are most useful. The measures can also be expressed in terms of the amount of traffic exposed
to specified high-risk situations.

Lead performance indicators or intermediate measures describing the improvements to the road, road
environment, speed or other features that have a known impact on road safety, eg the increasing the
percentage of central median barriers on busier roads to reduce head-on type crashes. These output
measures are known to directly impact safety outcomes.

The latter are particularly important as stated in the OECD report27:

‘for a Safe System approach there is a need to switch from injury based data (final outcomes, such as
traditional performance measures) to performance data (intermediate outcomes, such as lead
performance indicators). Intermediate outcomes are on the basis that 100%

achievement of safety performance is required in various sub targets.’

(a) Primary Outcomes

The reduction in deaths and serious injuries across the network is the primary outcome target. Directly related
is the reduction in fatal and serious crashes over the highest-risk routes and intersections that contribute most
to the total. Where an RCA has a number of high and medium–high collective and personal risk routes, then
the target could be to reduce the risk on each of these routes over a period of time.

(b) Secondary Performance Measures

This performance measure relates to reducing the crash risks on the network and on each high-risk rural route
or intersection. Indicators could be reductions in all recorded crash types or particular subgroups:

• Key secondary performance measures based on crash data could include:

– Reduction in number and proportion of crashes on wet roads.

– Reduction in number and proportion of crashes in darkness.

– Reduction in number and severity of run-off road crashes.

– Reduction in number and severity of head-on crashes.

– Reduction in number and severity of intersection crashes.

– Reduction in injuries to road user groups such as cyclists and pedestrians. .

27
The Safe System Approach – Towards Zero Ambitious Road Safety Targets and the Safe System; OECD International Transport Forum –
Transport Research Centre, 2008.

high risk rural roads guide interim draft for use april 2011 final_.docx 125
 High-risk rural roads guide

• Key secondary performance measures based on predictive risk analysis may include:

– Reduction in overall personal and collective predictive risk scores.

– Reduction in predictive personal and collective risk scores for each main crash type.

– Reduction in traffic (VKT) exposed to risk scores above a threshold.

– Reduction in the length of route (through realignment)

(c) Lead Performance Indicators

The smartest and most relevant lead performance indicators will relate most directly to the change in collective
crash risk that is associated with improvements in the feature being assessed.

Key lead performance indicators may include:

• Proportion of highway (or travel on highways) on roads over 12,000 vpd with median barriers (this
could include just the increase from a nominal base).

• Proportion of highway (or travel on highways) with roadside barriers or hazard reduction (this could
include just the increase from a nominal base).

• Proportion of highway (or travel on highways) with lane widths of at least 3.5m. (This could include
just the increase from a nominal base. It could also include a measure of the width deficiency for
each length. ).

• Proportion of highway (or travel on highways) with sealed shoulder widths of at least 1m (this could
include just the increase from a nominal base).

• The number or percentage increase in roundabouts.

• The length of routes subject to speed zoning below the default limit or under active speed
management.

• The change in network mean and/or 85th percentile speed (measured regionally by the Ministry of
Transport).

For those networks for which KiwiRAP star ratings have been produced, this system provides a wealth of lead
performance indicators including:

• Length and travel weighted network average RPS scores.

• Reduction in length of, and travel on, 2 star roads.

high risk rural roads guide interim draft for use april 2011 final_.docx 126
 High-risk rural roads guide

6.3.5 Goals and Targets

(a) Goals – Primary outcomes

To achieve the vision of ‘A safe road system increasingly free of death and serious injury’, the key road safety
issues have been addressed with the following goals. The goals in this document are consistent with the
purpose and vision of providing a Safe System. The goals will contribute to achieving district, regional and
national targets as outlined in the Safer Journeys Strategy to 2020.

Goals are based largely on the primary outcomes, ie to reduce the:

• number of fatal and serious crashes

• collective and personal risk levels

• head-on, run-off road and intersection type crashes.

(b) Targets – Secondary outcome and lead performance indicators

Targets are those quantitative and measurable indicators that determine the effectiveness of the
countermeasures. In addition to meeting the goals of primary outcomes, the following targets are based largely
on the definitions of a high-risk rural road that have been determined through RPS and star rating road factors
Each RCA should develop a list of targets on key crash types and determine the appropriate reduction in those
types of crashes over a certain period of time. To achieve a Safe System, the focus should be on the following:

• Increase the proportion of highway (or travel on highways) with:

– median barriers
this could include just the
– roadside barriers or hazard reduction increase from a nominal
base)
– lane widths of 3.5m

– sealed shoulder widths of at least 1m

– more roundabouts (number or percentage increase).

• The change in network mean and/or 85th percentile speed (measured regionally by the Ministry of
Transport).

For those networks for which KiwiRAP star ratings have been produced, this system provides a wealth of lead
performance indicators including:

high risk rural roads guide interim draft for use april 2011 final_.docx 127
 High-risk rural roads guide

• Length and travel weighted network average RPS score.

• Reduction in length of, and travel on, 2 star roads.

Depending on which lead performance indicators are being used to monitor the effectiveness of the ongoing
programme of safety improvements, goals can be set for one or more lead indicators. However, in all cases the
goals should be:

– relevant

– measureable

– most importantly, achievable.

6.3.6 Responsibilities for monitoring and evaluation

The responsibility for monitoring and evaluation at the highest level lies with the Ministry of Transport, which
monitors the national trends in the numbers killed or seriously injured – the primary outcomes. However, the
various RCAs should also be monitoring these primary outcome for their respective networks. Where large
networks, eg the state highway network or Auckland City, have been divided into sub-networks, the roading
manager should also monitor the primary outcomes.

The various RCAs should also be monitoring the secondary outcomes, related to collective and personal risk,
patterns of crash types and factors and changes in the risk profile of the routes and intersections being
targeted.

RCAs will also focus on lead performance indicators as the measure of the work they are performing towards
Safe System goals.

high risk rural roads guide interim draft for use april 2011 final_.docx 128
 High-risk rural roads guide

7 Other Information Sources

[1] Safer Journeys 2020 Strategy, Ministry of Transport, March 2010

[2] Traffic Control Devices Manual, NZ Transport Agency,

[2a] Traffic Control Devices Rule and Traffic Note: NZ Transport Agency

[3] Kiwi Road Assessment Programme(KiwiRAP); New Zealand Joint Agency

www.kiwirap.co.nz/

[4] The handbook of road safety measures, Elvik, 2004

[5] Towards Zero: Ambitious Targets and Safe System Approach, OECD, 2008

http://www.internationaltransportforum.org/jtrc/safety/targets/08TargetsSummary.pdf

[6] OECD Report reference to P. Wramborg, 2005

[7] Safe Speed thresholds for different road types, SWOV, 2008

[8] Economic Evaluation Manual, NZ Transport Agency, January 2010

[9] Future Safe System application – Higher Speeds Roads (>80km/h) – a potential approach (Source
OECD),

[10] ‘Peka Peka passing lane safe hit post median evaluation; Transport Engineering Research NZ
(TERNZ) – August 2010].

[11]Catalogue of design safety problems and potential countermeasures, PIARC, 2009

http://publications.piarc.org/en/search/detail.htm?publication=6047

[12] Case study referenced from paper ‘Longswamp to Rangiriri Wire Rope Barrier increased crash
numbers but improved road safety. Crowther S, Swears, Opus International Consultants Ltd;
ARRB 2010 Conference Centennial highway paper, NZ Transport Agency,

[13] ‘Treatments > Median Barrier’ iRAP : International Road Assessment Programme (iRAP), the
Global Transport Knowledge Partnership (gTKP) and the World Bank Global Road Safety Facility

http://toolkit.irap.org/default.asp?page=treatment&id=13

[14]‘Case Studies > The Coast Road Median Barrier’ iRAP; International Road Assessment Programme
(iRAP), the Global Transport Knowledge Partnership (gTKP) and the World Bank Global Road
Safety Facility

http://toolkit.irap.org/default.asp?page=casestudy&id=1

[15] Treatment type: Safety barriers, Austroads Road Safety Engineering Toolkit,
www.engtoolkit.com.au/default.asp?p=treatment&i=30
high risk rural roads guide interim draft for use april 2011 final_.docx 129
 High-risk rural roads guide

[16]Treatment type: Median retrofit, Austroads Road Safety Engineering Toolkit

www.engtoolkit.com.au/default.asp?p=treatment&i=45

[17]‘Treatments > Roadside Safety – Barriers ‘, iRAP; International Road Assessment Programme
(iRAP), the Global Transport Knowledge Partnership (gTKP) and the World Bank Global Road
Safety Facility

http://toolkit.irap.org/default.asp?page=treatment&id=28

[18]Treatment type: Clear zone widening, Austroads Road Safety Engineering Toolkit,
www.engtoolkit.com.au/default.asp?p=treatment&i=38

[19]‘Toolbox of Countermeasures and Their Potential Effectiveness for Roadway Departure Crashes’,
US Department of Transportation Federal Highway Administration, 2007,

www.ite.org/safety/issuebriefs/Roadway%20Departure%20Issue%20Brief.pdf

[20] Toolbox of Countermeasures and Their Potential Effectiveness for Roadway Departure Crashes

www.ite.org/safety/issuebriefs/Roadway Departure Issue Brief.pdf

[21]‘Treatments > Intersection – Grade Separation’, iRAP,

http://toolkit.irap.org/default.asp?page=treatment&id=7

[22]‘Special Report 214: Designing Safer Roads – Practices for Resurfacing, Restoration &
Rehabilitation. Transportation Research Board – National Research Council, 1987,
www.scribd.com/doc/27302256/ispecial-Report-214

[23]‘Treatment type: Roundabouts’, Austroads Road Safety Engineering Toolkit,

www.engtoolkit.com.au/default.asp?p=treatment&i=29

[24]‘Treatments Intersection – Roundabout’; iRAP,

toolkit.irap.org/default.asp?page=treatment&id=10

[25]‘Research Results Digest 299’,National Cooperative Highway Research Program,2005,

http://onlinepubs.trb.org/onlinepubs/nchrp/nchrp_rrd_299.pdf

[26]Update and Enhancement of ODOT’s Crash Reduction Factors, Portland State University &
Oregon State University,2006,
www.oregon.gov/ODOT/TD/TP_RES/docs/Reports/Crash_Reduction_Factors.pdf?ga=t

[27]‘Speed Limits in the Safe System Concept’ (source Journal of the Australasian College of Road
Safety – May 2010)

[28] Using Police Enforcement to prevent road crashes: The Randomised Scheduled Management
System; Legget; LMW.

[29]‘The Effectiveness of Delineation Treatments’, Charlton/Baas/Towler, 2004

www.nzrf.co.nz/techdocs/conferencepapers2005/papers/The%20effectiveness%20of%20delineat
ion%20treatments.pdf
high risk rural roads guide interim draft for use april 2011 final_.docx 130
 High-risk rural roads guide

[30]‘Treatments – Delineation’, iRAP, International Road Assessment Programme (iRAP), the Global
Transport Knowledge Partnership (gTKP) and the World Bank Global Road Safety Facility

http://toolkit.irap.org/default.asp?page=treatment&id=5

[31]‘Treatment type: Line marking improvements’, Austroads Road Safety Engineering Toolkit,
www.engtoolkit.com.au/default.asp?p=treatment&i=73

[32]‘Guidelines for rural road marking and delineation RTS 5’, Transit New Zealand, 1992,
www.nzta.govt.nz/resources/road-traffic-standards/docs/rts-05.pdf

[33]‘Delineation and Hazard Markers Section 5’,NZ Transport Agency,2010,

www.nzta.govt.nz/resources/motsam/part-2/docs/motsam-2-section-5.pdf

[34]‘Treatment type: Guide posts’, Austroads Road Safety Engineering Toolkit,

www.engtoolkit.com.au/default.asp?p=treatment&i=17

[35]‘Treatment type: Curve warning signs’, Austroads Road Safety Engineering Toolkit,
www.engtoolkit.com.au/default.asp?p=treatment&i=42

[36]‘Safety Evaluation of Curve Delineation Improvements Empirical Bayes Observational Before-and-

After Study’, Alfonso Montella,2___,
http://trb.metapress.com/content/n2274g3349t04025/fulltext.pdf

[37]‘Treatment type: Raised reflective pavement markers (RRPMs)’, Austroads Road Safety
Engineering Toolkit,www.engtoolkit.com.au/default.asp?p=treatment&i=28

[38]‘Delineation Improvement Cost Management Tool’, TERNZ,20__,

www.ternz.co.nz/DelineationCosts.xls

[39]‘Treatments > Rumble Strips’, iRAP; International Road Assessment Programme (iRAP), the
Global Transport Knowledge Partnership (gTKP) and the World Bank Global Road Safety Facility

http://toolkit.irap.org/default.asp?page=treatment&id=30

[40]‘Treatment type: Profile edge lines’, Austroads Road Safety Engineering Toolkit,
www.engtoolkit.com.au/default.asp?p=treatment&i=24

[41]‘The Usability and Safety of Audio Tactile Profiled Road Markings February 2009 Research Report
365’, John Edgar Consulting & TERNZ, 2009,
www.nzta.govt.nz/resources/research/reports/365/docs/365.pdf

[42]The cost effectiveness of delineation improvements for safety’ Research Report 322, Land
Transport New Zealand, 2007, www.nzta.govt.nz/resources/research/reports/322/docs/322.pdf

[43]‘Guidance on Design and Application of Rumble Strips’, DJ Torbic, JM Hutton, CD Bokenkroger,km

Bauer, ET Donnell, C Lyon, & B Persaud,2___,
http://trb.metapress.com/content/3vj887015hg44674/fulltext.pdf

[44]‘Treatment type: Vehicle activated signs’, Austroads Road Safety Engineering Toolkit,
www.engtoolkit.com.au/default.asp?p=treatment&i=34

high risk rural roads guide interim draft for use april 2011 final_.docx 131
 High-risk rural roads guide

[45]‘Assessment of Hazard Warning Signs used on New Zealand Roads’ Research Report 288, SG
Charlton, TERNZ Ltd & University of Waikato, 2006,
www.nzta.govt.nz/resources/research/reports/288/docs/288.pdf

[46]‘P32 Specification for Electronic Warning Signs on State Highways’, NZ Transport Agency, 2010,
www.nzta.govt.nz/resources/electronic-warning-signs/docs/p32-spec.pdf

[47]‘Vehicle-Activated Signs – A Large Scale Evaluation’, Road Safety Division, Department for
Transport, MA Winnett and AH Wheeler,2002, www.luno.se/swedish/products/VAS-
Evaluation.pdf

[48]‘Speed change management for New Zealand roads’ Research Report 300,SG Charlton, University
of Waikato & TERNZ Ltd, 2006, www.nzta.govt.nz/resources/research/reports/300/docs/300.pdf

‘Treatments > Speed Reducing Treatments’, iRAP, International Road Assessment Programme
(iRAP), the Global Transport Knowledge Partnership (gTKP) and the World Bank Global Road
Safety Facility, http://toolkit.irap.org/default.asp?page=treatment&id=33

[49]‘Guidelines for urban-rural speed thresholds RTS 15’Land Transport Safety Authority,2002,
www.nzta.govt.nz/resources/road-traffic-standards/docs/rts-15.pdf

[50]‘Traffic Advisory Leaflet 13/93 December 1993’,Department for Transport(UK),1993,

www.dft.gov.uk/adobepdf/165240/244921/244924/TAL_13-93

[51]‘Land Transport Rule: Setting of Speed Limits 2003’, Land Transport Safety Authority,2004,
www.nzta.govt.nz/resources/rules/setting-speed-limits-2003-index.html

[52]‘How Safe Are Our Roads? Rating New Zealand’s State Highways For Risk’, KiwiRAP, 2008,
www.kiwirap.org.nz/pdf/KIWIRAP%20Magazine%20final%20new.pdf

[53]‘Road Safety Risk Reporter’ No. 6, ARRB Group, 2006,

http://arrbcomau.ozstaging.com/admin/file/content13/c6/RiskReporter6.pdf

[54]‘Speed Management Summary Document’, OECD, 2006,

www.internationaltransportforum.org/jtrc/safety/SpeedSummary.pdf

[55]‘NZTA T10/2010’, Davis, Cenek, Henderson, 2005

[56]‘Austroads Guide to Asset Management Part 5F: Skid Resistance’, Austroads, 2009

[57]‘Treatments > Lane Widening , IRP; International Road Assessment Programme (iRAP), the Global
Transport Knowledge Partnership (gTKP) and the World Bank Global Road Safety Facility

www.toolkit.irap.org/default.asp?page=treatment&id=12

[58]‘Treatment type: Traffic lane widening’, Austroads Road Safety Engineering Toolkit,
www.engtoolkit.com.au/default.asp?p=treatment&i=71

[59]‘Treatments > Shoulder Sealing’, iRAP International Road Assessment Programme (iRAP), the
Global Transport Knowledge Partnership (gTKP) and the World Bank Global Road Safety Facility

high risk rural roads guide interim draft for use april 2011 final_.docx 132
 High-risk rural roads guide

http://toolkit.irap.org/default.asp?page=treatment&id=32

[60]‘Treatment type: Shoulder widening and/or sealing’, Austroads Road Safety Engineering Toolkit,
www.engtoolkit.com.au/default.asp?p=treatment&i=70

[61]‘Treatment type: Remove vegetation’, Austroads Road Safety Engineering Toolkit,

www.engtoolkit.com.au/default.asp?p=treatment&i=54

[62]‘Treatment type: Sight distance improvements – road sections’, Austroads Road Safety
Engineering Toolkit, www.engtoolkit.com.au/default.asp?p=treatment&i=58

[63]‘Safety deficiency: Vegetation – interference with driving task’, Austroads Road Safety
Engineering Toolkit, www.engtoolkit.com.au/default.asp?p=issue&i=31

[64]‘Specification for vegetation control’, Transit New Zealand, 1997,

http://www.nzta.govt.nz/resources/vegetation-control/docs/vegetation-control.pdf

[65]‘Treatments Roadside Safety – Hazard Removal’, iRAP; International Road Assessment

Programme (iRAP), the Global Transport Knowledge Partnership (gTKP) and the World Bank
Global Road Safety Facility

http://www.toolkit.irap.org/default.asp?page=treatment&id=29

[66]‘Guidelines for Highway Landscaping – Section 2: Transit’s Expectations for Highway

Landscaping’, Transit, 2006, http://www.nzta.govt.nz/resources/guidelines-highway-
landscaping/docs/highway-landscaping-section-2.pdf

[67]‘Treatments > Central Hatching’, iRAP: International Road Assessment Programme (iRAP), the
Global Transport Knowledge Partnership (gTKP) and the World Bank Global Road Safety Facility

http://toolkit.irap.org/default.asp?page=treatment&id=2

[68]‘Treatments Central Turning Lane Full Length’, iRAP; International Road Assessment Programme
(iRAP), the Global Transport Knowledge Partnership (gTKP) and the World Bank Global Road
Safety Facility

http://toolkit.irap.org/default.asp?page=treatment&id=4

[69]‘Treatment type: Painted/flush median’, Austroads Road Safety Engineering Toolkit,

http://www.engtoolkit.com.au/default.asp?p=treatment&i=46

[70]‘Treatment type: Median retrofit’, Austroads Road Safety Engineering Toolkit,

http://www.engtoolkit.com.au/default.asp?p=treatment&i=45

[71]‘Install Flush Median’, Land Transport Safety Authority, 1995,

http://www.nzta.govt.nz/resources/crash-reduction/docs/flush-median.pdf

[72]‘Guidelines for flush Medians RTS 4’, Land Transport Authority, 1991,
http://www.nzta.govt.nz/resources/road-traffic-standards/docs/rts-04.pdf

high risk rural roads guide interim draft for use april 2011 final_.docx 133
 High-risk rural roads guide

[73]‘Flush medians (Factsheet 52)’ Land Transport Authority, Land Transport Safety Authority, 2006,
http://www.nzta.govt.nz/resources/factsheets/52/flush-medians.html

[74]‘Case Studies Improved Line Marking and Rumble Strips’, iRAP; International Road Assessment
Programme (iRAP), the Global Transport Knowledge Partnership (gTKP) and the World Bank
Global Road Safety Facility

http://toolkit.irap.org/default.asp?page=casestudy&id=3

[75]‘Treatment type: Barrier lines’, Austroads Road Safety Engineering Toolkit,

http://www.engtoolkit.com.au/default.asp?p=treatment&i=13

[76]‘SH1 Peka Peka passing lane will improve safety’, NZ Transport Agency, 2010,
http://www.nzta.govt.nz/about/media/releases/575/news.html

[77]‘Development of Crash Reduction Factors’, Kentucky Transportation Centre, 1996, http://www.e-

archives.ky.gov/Pubs/transportation/tc_rpt/ktc_96_13.pdf

[78]‘Treatments Intersection – Turn Lanes (Unsignalised)’, iRAP; International Road Assessment

Programme (iRAP), the Global Transport Knowledge Partnership (gTKP) and the World Bank
Global Road Safety Facility

http://toolkit.irap.org/default.asp?page=treatment&id=9

[79]‘Treatment type: Turn lanes’, Austroads Road Safety Engineering Toolkit,

http://www.engtoolkit.com.au/default.asp?p=treatment&i=69

[80]‘Treatment type: Sight distance improvements – intersections’, Austroads Road Safety

Engineering Toolkit, http://www.engtoolkit.com.au/default.asp?p=treatment&i=57

[81]‘Right turn treatment’, NZ Transport Agency, 1994, http://www.nzta.govt.nz/resources/right-

turn-treatment/index.html

[82]‘Guide to Road Design Part 4A: Unsignalised and Signalised Intersections’, Austroads, 2009

[83]‘Guidelines for the implementation of traffic control at crossroads RTS 1’, Land Transport Safety
Authority, 1990, http://www.nzta.govt.nz/resources/road-traffic-standards/docs/rts-01.pdf

[84]‘Treatment type: Give Way/Stop signs’, Austroads Road Safety Engineering Toolkit,
http://www.engtoolkit.com.au/default.asp?p=treatment&i=63

[85]‘Low-cost engineering measures for casualty reduction. Application on the national road
network’, SV Gomes, 2___, http://www.ectri.org/YRS05/Papiers/Session-2bis/vieira.pdf

[86]‘Land Transport Rule – Traffic Control Devices 2004 Rule 54002’, Ministry of Transport, 2004,
http://www.nzta.govt.nz/resources/rules/traffic-control-devices-2004.html#10

[87]‘P32 Notes for Electronic Warning Signs on State Highways’, NZ Transport Agency, 2010,
http://www.nzta.govt.nz/resources/electronic-warning-signs/docs/p32-notes.pdf

high risk rural roads guide interim draft for use april 2011 final_.docx 134
 High-risk rural roads guide

[88]‘Traffic Note 56, Active school warning signs – Guidelines’, NZ Transport Agency, 2008,
http://www.nzta.govt.nz/resources/traffic-notes/docs/traffic-note-
56.pdfhttp://www.nzta.govt.nz/resources/traffic-notes/docs/traffic-note-56.pdf

[89]‘Traffic Note 57, Active warning signs (not at schools) – Guidelines’, NZ Transport Agency, 2008,
http://www.nzta.govt.nz/resources/traffic-notes/docs/traffic-note-57.pdf

[90]‘Engineering based approaches to reducing rural speed’, B Turner, ARRB Group Ltd, 2009,
http://www.rsconference.com/pdf/RS090005.PDF?check=1

[91]‘Treatments > Additional Lane’, iRAP: International Road Assessment Programme (iRAP), the
Global Transport Knowledge Partnership (gTKP) and the World Bank Global Road Safety Facility

http://toolkit.irap.org/default.asp?page=treatment&id=3

[92]‘Treatment type: Additional lanes for overtaking’, Austroads Road Safety Engineering Toolkit,
http://www.engtoolkit.com.au/default.asp?p=treatment&i=35

[93]‘Transit Planning Policy Manual, Appendix 3E- Passing and overtaking’, Transit NZ, 2007,
http://www.nzta.govt.nz/resources/planning-policy-manual/docs/planning-policy-manual-
appendix-3E.pdf

[94]‘Treatment type: Superelevation improvement’, Austroads Road Safety Engineering Toolkit,

http://www.engtoolkit.com.au/default.asp?p=treatment&i=19

[95]‘Safety deficiency: Curve – inadequate delineation’, Austroads Road Safety Engineering Toolkit,
http://www.engtoolkit.com.au/default.asp?p=issue&i=14

[96] Austroads : AP-T141/10: Infrastructure/Speed Limit Relationship in Relation to Road Safety

Outcomes

https://www.onlinepublications.austroads.com.au/items/AP-T141-10

[97] Austroads: Part 8; Guide to Road Safety

http://www.austroads.com.au/interest_safety.html

[98]IRAP website: International Road Assessment Programme (iRAP), the Global Transport
Knowledge Partnership (gTKP) and the World Bank Global Road Safety Facility

http://www.toolkit.irap.org;

[99]State Highway Geometric Design Manual, NZ Transport Agency

http://www.nzta.govt.nz/resources/state-highway-geometric-design-manual/index.html

[100] Road Transport Association (RTA)

http://www.nzrta.co.nz/

[101] Andressaon D;

high risk rural roads guide interim draft for use april 2011 final_.docx 135
 High-risk rural roads guide

[102] Safety Data, Analysis, and Evaluation 2009, Volume research report 2103 Transport
Research Board (TRB)

http://www.trb.org/Main/Blurbs/162222.aspx

[103] Crash Modifications Clearing House

http://www.cmfclearinghouse.org/

[104] RTA (2010) Delineation Section 15,

http://www.rta.nsw.gov.au/doingbusinesswithus/downloads/technicalmanuals/delineation
/delineationappendixaa_i.pdf

[105] NCHRP Report 500: Volume 20: A Guide for Reducing Head-on crashes on freeways

[106] The cost effectiveness of delineation improvements for safety; NZTA Research Report 322

http://www.nzta.govt.nz/resources/research/reports/322/

[107] Traffic Control Devices, Visibility, and Highway-Rail Grade Crossings 2010; Transport
Research Board (TBR) research report 2149

http://www.trb.org/Main/Blurbs/Data_Systems_and_Travel_Survey_Methods_163904.aspx

[108] Toolbox of countermeasures

Source:http://www.ite.org/safety/issuebriefs/Roadway%20Departure%20Issue%20Brief.pdf
[109] http://www.ite.org/safety/issuebriefs/Roadway%20Departure%20Issue%20Brief.pdf
[110] Austroads – General

[111] http://www.toolkit.irap.org/default.asp?page=treatment&id=2

[112] Update and Enhancement of Odot’s Crash Reduction Factor, Final Report 612, 2006

www.oregon.gov/ODOT/TD/TP_RES/docs/Reports/Crash_Reduction_Factors.pdf?ga=t
[113] Traffic Note 57: Active Warning Signs (not at schools) NZ Transport Agency

www.nzta.govt.nz/resources/traffic-notes/docs/traffic-note-57.pdf

[114] Accident Reduction Guide, RTA Technical Direction for Road Safety Practitioners
TD2004/RS01; March 2004

[115] Austroads Technical Report “Intelligent Transport Systems and Variable Message Signs for
Road Safety Applications: Current Status and Future Prospects”, AP-T133/09

[116] T10: 2010 Specification for State Highway Skid Resistance Management

http//nzta.govt.nz/resources/skid-resistance-investigation-treatment-selection/docs/skid-
resistance-investigation-treatment-selection-oct10.pdf

high risk rural roads guide interim draft for use april 2011 final_.docx 136
 High-risk rural roads guide

[117] The New Zealand Speed Zoning Policy Version 0.4.0 February 2005; NZ Transport Agency

[118] Australian National Road Safety Strategy (2011-2020) – Draft for consultation
www.infrastructuregov.au/roads/safety/national_road_safety_strategy/files/Draft_Nation
al_Road_Safety_Strategy_ext.pdf

high risk rural roads guide interim draft for use april 2011 final_.docx 137
 High-risk rural roads guide

Appendix A Distribution of High-severity Rural Crashes

high risk rural roads guide interim draft for use april 2011 final_.docx 138
 High-risk rural roads guide

The number of high-severity crashes (2005 to 2009 inclusive) by TLA and the proportion occurring in isolation, ie no other reported injury crash with 250m radius

Doc Ref HHRG Parts 1 2, 3, 4 and 5 Version 0.8

Date December 2010 139

 High-risk rural roads guide

Appendix B RAMM SQL for calculating Personal and Collective Risk

A user-defined table needs to be created in RAMM Manager.

To do this, open RAMM Manager. Go to Projects > User Defined Tables.

Click on Add to add a new table.
On step 1 we do not want to load any settings so click on Next.
Step 2 involves naming our table.
Type (in lower case) hrrr in the Name field.
In the description field, type in High-risk rural roads.
Set the permissions to Drainage.
It should resemble the picture below.

Click Next to move to Step 3.

In Step 3, select Length and uncheck all other boxes.
Click Next to move to Step 4.

Doc Ref HHRG Parts 1 2, 3, 4 and 5 Version 0.8

Date December 2010 140

 High-risk rural roads guide

In Step 4, select Do not include Offset then select Do not include a side column.
Click Next to move to Step 5.
Click Next on Steps 5 and 6 (do not select anything).
Step 7 is the hardest step.
Add fields to the table by clicking on the New button (blank white page with a yellow + symbol) under the
Custom Columns title.
The fields you need to add are described below:
Fatal Count
Name: fatal_count
Field Label: Fatal Count
Hint: Number of Fatal Crashes
Type: Integer
Default: 0
Minimum: 0
Serious Count
Name: serious_count
Field Label: Serious Count
Hint: Number of Serious Crashes
Type: Integer
Default: 0
Minimum: 0
VKT
Name: vkt
Field Label: VKT
Hint: Vehiclekm travelled
Type: Integer
Default: 0
Minimum: 0
Crash Rate
Name: crash_rate
Field Label: Crash Rate
Hint: Calculated Crash Rate (crashes per VKT millions) per year
Type: Decimal
Size: Large (16 digits, 4 decimal places)
Default: 0

Click Next to move to Step 8

Select Asset ID as the Description Column

Click Next to move to Step 9

Click on Finish (there’s no need to Save Settings).

The table is now set up ready for data.

Doc Ref HHRG Parts 1 2, 3, 4 and 5 Version 0.8

Date December 2010 141

 High-risk rural roads guide

Before we can populate the table, we need to make sure the latest crash data has been loaded into RAMM.
This step can be performed by following the instructions in RAMM Manager under Projects -> Crash Data -
> Import
The next step is to populate the HRRR table.
RAMM SQL needs to be opened and a new query created.
The following query needs to be type in to the query screen.
delete from ud_hrrr;
insert into ud_hrrr (road_id, start_m, end_m, carrway_start_m)
select road_id, min(carrway_start_m), max(carrway_end_m), min(carrway_start_m)
from carr_way
where urban_rural = “R”
and owner_type = “L”
group by 1;
update ud_hrrr
set vkt = (select sum((traffic_adt_est) * ((carrway_end_m-carrway_start_m)/1000) * 365 * 5)
from carr_way
where carr_way.road_id = ud_hrrr.road_id
and carr_way.urban_rural = “R”)
;

select max(crash_date) max_date

from cas_crash
into temp tbl_max_date;
select *
from cas_crash c , tbl_max_date m
where crash_date > (max_date – (365.25 * 5))
into temp crash;
update ud_hrrr
set fatal_count = (select count(crash_id)
from crash c
where crash_fat_cnt > 0
and ud_hrrr.road_id = c.road_id);
update ud_hrrr
set serious_count = (select count(crash_id)
from crash c
where crash_fat_cnt = 0
and crash_sev_cnt > 0
and ud_hrrr.road_id = c.road_id);
update ud_hrrr
set crash_rate = (fatal_count + serious_count)/(vkt/1000000);

Once you have type this in, go to the Transaction menu and select Begin Transaction
Next – click on the Run button (green triangle) below the SQL title
The table is now populated

Log into RAMM for Windows.

Doc Ref HHRG Parts 1 2, 3, 4 and 5 Version 0.8

Date December 2010 142

 High-risk rural roads guide

In the table list there will be a new table called High-risk rural Roads
Open this table by selecting it. On this table’s window, go to Options and select All Roads
When prompted, select All roads in the Entire Network
You can now view / filter / sort / export all the results of the query if you want to.
To map the results, firstly open up RAMM Map (the icon looks like a pink/purple globe with white lines on
it).
Next, switch back to the High-risk rural Roads window.
Go to Actions, then select Add to Map -> Add to Map Now
When the dialogue box pops up, click on the button to the left of the Settings box.
The Map Settings will appear.
Under the Appearance tab, make sure only the Line Style box is checked
Move to the Advanced tab.
We can add themes to this table by adding in conditions and filters on this screen.
Click the Add button to add conditions as required.
A sample is included in the screenshot below:

Note: You need to filter on the Crash Rate – the higher the crash rate, the more prominent the line should
be. The sample above is showing High Crash Rate (red solid line) for all roads with more than 8 crashes per
1M VKT. You may need to change this to your local traffic environment.
Click on OK to save the settings. When prompted for save name, call it HRRR
You will be returned to the Add to Map dialogue box.
Make sure Add all items is selected and click OK
RAMM will now colour all the lines (as you described) according to their crash rate.
There are a number of tools to help you navigate around RAMM Map shown above the map (zoom in/out,
pan, measure).
You can also alter layer settings (turn labels on/off etc) from the Layers Panel which can be accessed by

clicking the button.

Doc Ref HHRG Parts 1 2, 3, 4 and 5 Version 0.8

Date December 2010 143

 High-risk rural roads guide

You can also add crashes to this map by opening the Crash table, selecting All Roads, and adding all the
items to the map. Note, you may need to filter for crashes only in the last 5 years as that is what the HRRR
table is using (5 years back from the latest crash date available).

Appendix C Allocation of road networks to climate zones

CLIMATE REGION TLA NETWORKS NZTA NETWORKS

Northern New Zealand Far North District Thames-Coromandel Northland
Whangarei District District PSMC005
Kaipara District Hauraki District Auckland-NMMC
Rodney District* Waikato District East Waikato
North Shore City* Matamata-Piako Bay Roads
Waitakere City* District Tauranga City
Auckland City* Western Bay of Plenty BOP
Manukau City* Tauranga City
Papakura District* Whakatane District
Franklin District* Kawerau District
Opotiki District

Central North Island Hamilton City Rotorua District West Waikato

Waipa District Waitomo District Central Waikato
Otorohanga District Taupo District Rotorua Dist
South Waikato District Ruapehu District PSMC006

Eastern North Island Gisborne District Tararua District Gisborne

Wairoa District Masterton District Napier
Napier City Carterton District
Hastings District South Wairarapa
Central Hawkes Bay District
District

South-West North New Plymouth District Horowhenua District West Whanganui

Island Stratford District Kapiti Coast District East Whanganui
South Taranaki District Porirua City Wellington
Whanganui District Upper Hutt City
Rangitikei District Hutt City
Manawatu District Wellington City
Palmerston North City

Northern South Island Tasman District Marlborough District Nelson

Nelson City Kaikoura District Marlborough

Western South Island Buller District Westland District West Coast

Grey District

Eastern South Island Hurunui District Selwyn District North Canterbury

Waimakariri District Ashburton District South Canterbury
Christchurch City Timaru District

Inland South Island Mackenzie District Queenstown-Lakes Otago Central

Waimate District District
Waitaki District Central Otago District

Doc Ref HHRG Parts 1 2, 3, 4 and 5 Version 0.8

Date December 2010 144

 High-risk rural roads guide

Southern New Zealand Dunedin City Southland District Coastal Otago

Clutha District Invercargill City Southland
Gore District

* = Since 1 November 2010, part of the Auckland Supercity

Appendix D Personal Reported vs Predictive Risk Correlation Charts

80
4.6

70
y = 67.549652-7.891884*x2+1.084689*x3 (R2=0.8709)
4 and y set to constant for x=5+ and x<1
33.2
Reported Injury Crash Rate (per 100M VKT)

60

4
ChartLabels show:
18.8
# of Crashes per
50 18.8 Year in Each Star
56 Rating Category
6.6 52 44.8
20.6
29.2
4.4 39.4 45
40
72.6
113.8

30
139.4
151.2

20 115 94.8
103.6
113.6162.8143.6
105.8
84.4 103.654.243.6
26
35.4
10 34.8 43 67.448.627.8 7
30

0
1 1.5 2 2.5 3 3.5 4 4.5 5 5.5
1/10th Star Rating

Figure D-1: Reported injury crash rates associated with each 1/10th star rating category, based on 100m star rating
data for rated rural state highways (data for star rating categories with <2 reported injury crashes per year
associated with them have been removed)

Figure D-1 shows the relationship between the KiwiRAP star rating expressed to one decimal place. This is
based on the data reported for each 100m section. Figure D-2 provides the same relationship, but based on
the 5km KiwiRAP lengths. The relationships are subtly different, principally because each 5km length will
include 100m sections of highway that are better or worse than the other sections which make up the 5km
length.

The RPSs from which these star ratings are derived are essentially personal risk or crash rate scores based
on all reported injury crashes. However, this Guide is principally focused on high-severity crashes – those
resulting in death or serious injury – which typically make up approximately 30% of the reported injury
crashes.

When seeking to use these relationships, two approaches are possible:

Doc Ref HHRG Parts 1 2, 3, 4 and 5 Version 0.8

Date December 2010 145

 High-risk rural roads guide

1 Using the KiwiRAP Analysis Tool (KAT) the analyst can look at the impact that each improvement
has on the 100m section treated, quantify these and sum these over the total treated length. Using
the relationship from figure D-1 the analyst can determine the change in expected injury crashes by
applying the rate (before and after) from figure D-1 and adjusting this according to the length of
highway treated and the volume of traffic using this length of highway. The analyst can then take
30% of the result to represent the change in high-severity crashes.

2 The analyst can look at the 5km star rating reported to one decimal place in the NZTA spatial
viewer tool. The new 5km star rating, following improvements, can then be calculated using KAT
and the change in expected injury crashes can be calculated using figure D.2 and multiplying the
new difference in injury crash rate by the length of the section and the volume of traffic using the
section of highway. Again 30% of this change can be taken to represent the change in high-severity
crashes.

y=60.000082-8.445633*x2+1.281599*x3 (R2=0.8265)
Reported Injury Crash Rate (per 100M VKT)

60 and y set to = constant for x>4.5 and x<2

NB: the trendline relationship is based on curve fit to 1/10th (1dp) 5km star
50 rating data, which is only available for star ratings 2.4* to 4.7* as there is
insufficient data available outside of this range

14
Chart Labels show:
40
# of Crashes per
Year in Each Star
Rating Category
30
99 180
214 325343 287
29 234
20
120 97 47 37
9
30
11 134
10 5 3 61
6 5 18
6

0
1 1.5 2 2.5 3 3.5 4 4.5 5 5.5
5km Star Rating (rounded to 1dp)

Figure D-2: Reported injury crash rates associated with each 1/10th star rating category, based on 5km star rating
data for rated rural state highways

Doc Ref HHRG Parts 1 2, 3, 4 and 5 Version 0.8

Date December 2010 146