August 2006

The effect of geometric road design

standards on road safety
This is the fourth in a series of newsletters describing the results
of the Road Safety Engineering Risk Assessment project which
Study objectives and
is being funded by Austroads to assess risk involving the road, approach
traffic and roadside environment.
The objectives of the research project were to:
Introduction • identify mid-block road design elements that affect
Geometric road design standards have a direct bearing on road safety
the level of risk for drivers and other road users. The effective • identify how a variation in standards for design
management of road safety risk involving different road design
elements (e.g. speed, sight distance, cross section)
elements requires an understanding of:
affects the safety of roads in different environments
• the factors that contribute to risk and how to reduce them
(e.g. rural and urban)
• the extent and distribution of these factors in the
• identify key gaps in knowledge requiring
road network.
further research
This knowledge allows comparison of the effects of higher
design standards and other measures on safety. It also provides a • develop models for practitioners to determine the
framework for road designers to measure the road safety impact appropriate balance between road design standards,
of applying minimum standards where funds are limited. road safety benefits and costs.

www.arrb.com.au
The study was based on a review of Australian and
international research with an emphasis on geometric design
elements that were expected to provide cost-effective safety
benefits. Information was also sought from Australian road
authorities to identify and include the results of any research
undertaken by them.

Results
Key road design elements
The geometric road design elements that are critical to road
safety are well established in research literature and in road
design manuals. They comprise:
• horizontal alignment
• vertical alignment
• sight distance
• cross section (pavement width, shoulder width and type,
lane width) • No clear relationship has been shown between crash
frequency and sight distance restrictions on crests over
• roadside design (width, slopes and condition).
the whole range of sight distance values; there is some
Horizontal realignment is particularly important in the context agreement that low sight distance crests (e.g. <140 m)
of road safety treatments, major changes to existing roads have a substantially higher crash rate than comparable
locations with greater sight distance (e.g. > 215 m) above
and the design of new roads. The combination of cross
which the effect remains constant.
section elements is also very important with respect to the
overall sealed width, the optimum combination of lane • Crash rates have been found to decrease with increases
in total seal width up to widths of 10 to 11 m. The effect
width and sealed/unsealed shoulder width, and the roadside
of adding width to a sealed shoulder or to a lane is
environment. Improvements to cross section and the roadside
similar, although, crash risk is likely to increase on wide
are usually very cost-effective treatments. shoulders (> 2.5 m).
• On rural two-lane highways the crash risk reduces
The effects of road design elements substantially with increasing lane width up to 3.6 m
The review of research led to the following main conclusions but is likely to increase on wider lanes.
on the link between road design elements and safety: • Lane width has only a small effect on crash rates for
• On rural highways the crash risk associated with urban arterial roads within the range of normal lane
horizontal curvature increases substantially at radii widths (2.75 m to 3.75 m).
below about 400 m. • On urban arterial roads the provision for right turning
• No information is available on risk associated with vehicles to stand clear of through traffic substantially
horizontal curves on urban roads. reduces crash risk; the density of access and type of
abutting development can substantially affect crash risk.
• Crash rates increase only slightly with increases in
gradient below about 6%, but increase rapidly • Guidance typically identifies the benefits from the
above 6%. provision of roadside clear zones with a width of up to
9 m. However, additional benefits can be gained with
• There is a need to investigate the combined effects of
clear zones of up to 13 m and beyond, especially in high
horizontal and vertical alignment on crash rates and risk.
speed environments.

2
Adequacy of current information
There is agreement in the research literature on the Future developments
relationships between various geometric design standards Full details of this research will be provided in a
and crash risk. However, the values assigned to road safety
report due later this year. Further research will
effects or crash risks vary markedly between studies. Many
continue at ARRB on the effect of geometric
studies aggregate data or present results in a way that
limits their usefulness with respect to specific situations
road design standards on road safety.
or treatment options. As a result, there is insufficient Given the need for improved information,
consistent information about the level of risk associated it is recommended that data on geometric
with specific road design elements in Australia and New alignment and cross-section that is currently
Zealand. The results nevertheless provide a basis for some collected using instrumented vehicles be used
minor improvements in the values used in current models. to develop comprehensive risk assessment
models that relate to Australian and New
No research information was found relating crash risk to
Zealand conditions. The data should cover
standards of alignment for urban roads. Findings relating
all jurisdictions and be analysed in relation to
to width standards and cross-section treatments for urban
roads were also more variable than for rural arterial roads.
crash data.

The key areas of geometric road design that require

improved information involve determining the effect of: For further information
• various standards of horizontal curvature for rural on this research, contact
roads including variables such as length of curve and Gary Veith
tangent conditions (e.g. alignment or approach speed) +61 3 9881 1526
• combinations of horizontal and vertical alignment email [email protected].

• different combinations of lane width and shoulder For further information

width for rural roads on the overall research
• shoulder type at various traffic levels program, contact
• downhill gradient particularly for grades above 6% Blair Turner
+61 3 9881 1661
• narrow lane widths and horizontal alignment for
email [email protected]
typical situations in urban and outer urban areas.

To receive further editions

If you are interested in receiving future editions of this newsletter, contact us with your email address at
[email protected] Newsletters will also be placed on the ARRB website.

Published by ARRB Group ARRB would like to acknowledge the continued support and input of Graphic Design
500 Burwood Hwy Austroads members in the development and delivery of this project. Vicki Jaeger
Vermont South VIC 3133
3