Road hazard management is an important component of any road design project. Concepts including forgiving roadside design and risk management are important to provide a safe environment for all road users. The intention of this guide is to address road safety elements that focus on "Keeping vehicles on the road" and "Dealing with errant vehicles" on the occasion that vehicles leave the carriageway. This document aims to address many of the issues related to road safety in road design, and direct practitioners to appropriate design standards and practices.
It is expected that most work done with reference to this document will involve maintenance and upgrading of existing roads. While this may limit the road designer's scope for implementing some of the recommendations listed, a comprehensive upgrade project should involve reference to the guides listed in this Document.
Road Hazard Management Guide,(PDF 1.44MB)
- Keeping Vehicles on the road
- Linemarking (centre line / edge line)
- Raised pavement markers
- Warning signs
- Weather warning systems
- Road Design Elements
- Lane widths
- Shoulder widths (sealed/unsealed)
- Horizontal curvature and localised curve widening
- Vertical alignment
- Road surface
- Sight distance
- Dealing with errant vehicles
- Clear zones
- Types of hazards and their treatments
- Rigid objects
- Open drains
- Bodies of water
- Types of hazards and their treatments
- Safety Barriers
- Decision to install a safety barrier
- Barrier types
- Layout of safety barriers
- Appropriate lengths
- End treatments
- Transitions between barrier types
- Works zones
This document was prepared by ARRB Transport Research for the Department of Infrastructure, Energy and Resources. Acknowledgement is also given to Milan Prodanovic from the Department of Infrastructure, Energy and Resources for his major contribution to this guide.
Keeping vehicles on the road
This section discusses on-road and roadside features aimed at showing drivers the path that a road takes and helping them keep their vehicles in the running lane.
Roads should be designed with the objective of making it as easy as possible for drivers to keep their vehicles on path. This would be straightforward if the landscape always suited the desired path of the road and there were no financial constraints, but this is rarely the case. In reality a road has to be accommodated within the topography in a cost effective way and this may lead to situations that require departures from the standard. In such cases it becomes necessary to provide additional features designed to help drivers follow the line of the road.
Australian Standard AS1742.2-1994, Manual of uniform traffic control devices, Part 2 - traffic
control devices for general use and its amendment, AS1742.2-1994/Amdt 1-1997 sets out requirements of all traffic control devices for use on roads other than freeways. It includes sections on signage for intersection and mid-block locations, pavement markings, and appendices on topics such as calculating advisory speeds for curves.
Centre lines should be marked to separate opposing directions of traffic flow on sealed pavements at least 5.5m wide. On pavements narrower than this the provision of centre lines is not usually provided other than where sight lines for overtaking are deficient. Centre lines may be of the following types:
- separation lines - continuous or broken
- barrier lines - either continuous double lines or single continuous with parallel broken line.
Barrier lines should not be used on pavements of insufficient width where it is not practicable for all vehicles to travel on their side of the line.
Edge lines are used to delineate the edge of the travelled lane. The aim of this delineation is both to discourage travel on the road shoulder and to assist drivers to track the edge of the road. Edge lines are particularly helpful on bends and at night. The minimum sealed pavement width on which edge lines should be installed in Tasmania is 6.4m, although narrower pavements may be provided with edge lines in special circumstances where the lines are considered essential. Edge lines must always be used in conjunction with a centre line (see Clause 4.3.5 of AS 1742.2-1994).
Audiotactile line marking
Edge lines can be provided with raised, transverse bars of thermoplastic material placed at short intervals (see Figure 1). This practice creates edge lines that provide an audible and tactile warning whenever a vehicle tyre runs over them. The intention of the treatment is to alert the driver to the fact that their vehicle is straying onto the road shoulder and thereby give them a chance to take appropriate action before the vehicle leaves the pavement. Audiotactile edge lines should be considered where there is a recorded history of fatigue related crashes, and may be considered on roads prone to fog.
Some international studies on the effectiveness of using such a treatment have shown run-off-road crash reductions of at least 20% (Ligon et al (1985) cited Dravitzki et al 1998). The effectiveness of the treatment in practice would vary according to individual site conditions.
While the noise generated by audiotactile line marking is usually easily audible in normal passenger cars, it is harder to hear in large vehicles such as four-wheel drives, and is often impossible to hear in trucks. The treatment therefore should not be relied upon to provide warning to heavy vehicle drivers.
Audiotactile edge lines are not mandatory but may be of benefit when it is necessary to warn drivers that they are straying off the carriageway. The absolute minimum seal width required for installation of audiotactile edge lines is 6.6m. This is because at least 0.1m of seal is required outside the edge line. Consult AS 1742.2-1994 , Section 4 for detailed requirements of various types of pavement markings.
It should be noted that while the noise made by audiotactile edge lines can be of help to drivers, it can also be an annoyance to nearby residents. If these devices are to be installed in areas where roads run close to houses, it may be helpful to inform residents of the proposed installation and its expected safety benefits. It is suggested that such edge lines are not used within 200m of residences.
Audiotactile edge lines on a freeway
Raised pavement markers can be used in conjunction with, or sometimes used instead of painted line markings. Section 4.6 of AS 1742.2-1994 discusses the use of non-retro-reflective and retro-reflective raised pavement markers. Non-retro-reflective pavement markers (NRPMs) may be used in conjunction with raised retro-reflective pavement markers (RRPMs) where it is intended that they simulate marked lines, for example, for lane lines on freeways. NRPMs may also be used at intersections to provide drivers guidance when negotiating the intersection.
RRPMs may be used to augment painted lines or instead of painted lines for the provision of lane lines, separation and barrier lines, edge lines and traffic islands and medians. RRPMs are not obscured at night under wet conditions as the retro-reflective panels sit above the surface and are more prominent than reflectorised painted markings (i.e.. paint incorporating glass beads for added reflective capability). In addition, they provide an audible and tactile signal when traversed by vehicle wheels.
It should be noted that while the noise made by RRPMs can be of help to drivers, it can also be an annoyance to nearby residents. If these devices are to be installed in areas where roads run close to houses, it may be helpful to inform residents of the proposed installation and its expected safety benefits.
RRPMs should be used to highlight centre line marking on all Tasmanian rural roads that have a sealed width of at least 6.4m. They should also be applied to the centre lines and outside the edge lines on all National Highway routes and on dual carriageways with no street lighting.
RRPMs are used in various colours as follows:
- White markers are used to augment lane lines, separation lines, markings at traffic islands and freeway ramp gore areas.
- Yellow markers can be used on the right-hand edge lines of one-way carriageways.
- Red markers are used where appropriate to augment left-hand edge lines of two-way and one-way carriageways.
- Blue markers are used to mark the location of fire hydrants on roads. In this case a single marker is placed near the road centre line opposite the position of the hydrant on that side of the road.
In areas above the snow line there is a risk that raised pavement markers could be damaged by snow ploughs. For this reason it is recommended that orange snow poles be used instead. Snow poles are designed to be high enough to protrude at least 300mm above expected snow drift levels and their orange colour aids visibility in snow.
Guideposts are used to show the edge of the road and enhance the delineation of the road's path for drivers. They should be installed at a uniform distance from the edge of the road and should be fitted with delineators (see Figure 2). On narrower or lower volume roads where there is insufficient road width to mark a centre line, guideposts may be the only form of delineation provided.
Requirements for the size, spacing and location of guide posts are detailed in Section 3.2.4 of AS1742.2-1994 document. On straight road sections, guide posts should be arranged in pairs at a spacing of 150m, although this spacing may be amended according to conditions outlined in the standard. The standard also specifies the spacing of guide posts on curves, crests, cuttings, bridges and culverts. Guideposts should be installed to be 1m high above ground level and have a white 100mm wide area for at least the upper 300mm of their height. This 100mm wide area should face oncoming traffic and incorporate a retro-reflective delineator.
Requirements for delineators, including details for mounting on guide posts, safety barrier and snow poles, are described in section 3.2.5 of AS1742.2-1994. A red delineator is used for the left side of the carriageway and white for the right side of the carriageway. Delineators should be of Class 1 material (AS/NZS1906.1-1993) or Class 1A (AS1906.2-1981/Amdt1-1988) where maximum delineation is required.
Guide post and retro-reflective delineators providing delineation of a curveWarning signs
While guide posts and line marking can be used to delineate the path of a road, some of the more unexpected aspects of the road's geometry will require additional signage to convey their severity and nature to drivers.
Curve warning signs, advisory speed signs and chevron alignment markers (CAMs) are all appropriate treatments for substandard curves (see Figure 3). Warning and advisory speed signs should be used in the first instance, with CAMs only installed at locations where other signage is deemed to give insufficient warning. Section 3.4 of AS1742.2-1994 sets out the requirements for various levels of sign provision.
Warning sign, advisory speed sign and chevron alignment markers (CAMs) at a sub standard curve
Figure 3: Warning sign, advisory speed sign and chevron alignment markers (CAMs) at a sub standard curve
On sections of road that have curved alignment, an accident history, and that may pass through an environmentally sensitive landscape, it may be desirable to provide an enhanced warning sign at both approaches to the road section (see Figure 4).
Red background used to emphasise potential hazard
A range of conditions related to weather can have an adverse affect on vehicles staying on the road through their impact on drivers or the road surface. Common conditions include:
- heavy rain
- ice and/or snow
- water on road
- strong winds.
Weather warning systems may be used on freeway and non-freeway arterial roads where the adverse effects of weather increase the risk of road crashes. Such a system may be as simple as permanent signs, but more complex systems are also possible. For example, an ice warning system can consist of warning lights and signs that are activated by inputs from temperature and humidity sensors. Similarly, a fog warning system could activate advance warning signs and lights in response to inputs from a visibility detection device.
A weather warning system will generally be applicable to a specific location where a particular weather condition, unique to that location, has been identified as a contributor to crash risk. If a system is proposed as a crash countermeasure for a particular site, it is recommended that several years' crash data first be examined to determine that the relevant weather conditions did influence crashes at the site.
Road design elements
In order to give motorists the best chance of keeping their vehicles on the road, it is necessary to provide a geometric design conducive to safe travel. The principal factor influencing a vehicle's ability to traverse and remain on a particular section of road is the speed of the vehicle. Accordingly, it is necessary to take into account the operating speed of a road section when setting such parameters as curve radii, lane widths, shoulder widths, seal types and drainage.
For full information on the topics in this section the practitioner is referred to the Austroads publications, Rural Road Design (2003b) and Urban Road Design (2002).
The width of a traffic lane influences the ease with which vehicles can operate in that lane. Higher traffic volumes and higher speeds demand wider lanes to allow more space between passing vehicles, and between vehicles and any roadside objects. Austroads specifies appropriate lane widths for urban and rural environments in Section 12 of Urban Road Design and Section 11 of Rural Road Design respectively.
However, the recommended lane widths for the various categories of Tasmanian roads differ from the Austroads requirements and are shown in Table 1. The information in the table has been taken from the draft Tasmanian State Road Hierarchy Target Standards document (October 1999).
Table 1: Recommended lane widths
|Traffic volume (AADT)||Recommended lane widths|
|5000 plus||3.5m||3.5m||3.25m||3.0m||As warranted|
Shoulder widths (sealed/unsealed)
Apart from its function as a lateral support for the road pavement, a road shoulder is provided as additional road width for a number of traffic-related reasons. The most important of its traffic functions in relation to road management is the provision of a degree of separation between moving traffic and roadside objects. This extra road space, while not intended for regular travel, allows drivers more room to bring their vehicles back under control after inadvertently leaving the traffic lane.
To reduce the incidence and severity of run-off-road crashes it may be desirable to ensure that shoulders are sealed. This will assist errant vehicles to recover should they leave the carriageway. A decision to seal shoulders will depend on the road category, traffic volume and the accident record of the section of road.
The actual width of shoulder sealing will depend on traffic speed, volume and composition, environmental conditions and the nature of the roadside area. Ideally the sealed shoulder width should be between 1.5 and 2.4m as this range was found to be the safest in a recent Austroads investigation carried out by ARRB TR. Section 11.5 of Austroads' Rural Road Design (2003b) provides further advice on sealed and unsealed shoulder widths. The information therein is applicable to both rural and urban locations, although shoulders are not usually required in urban settings except for drainage or the storage of broken down vehicles.
However, the draft Tasmanian State Road Hierarchy Target Standards document (October 1999) states specific sealed and unsealed shoulder widths applicable to the various categories of Tasmanian roads, as shown in Table 2.
|Traffic volume (AADT)||Recommended lane widths|
|0-1000||N/A||0.6m sealed||0.3m sealed||0.6m unsealed||As warranted|
|1000-2500||N/A||1.0m sealed||0.6m sealed||0.3m sealed||As warranted|
|2500-5000||2.0m sealed||1.5m sealed||1.0m sealed||0.6m sealed||As warranted|
|5000 plus||2.0m sealed||2.0m sealed||1.0m sealed||1.0m sealed||As warranted|
Section 11.5 of Rural Road Design provides further advice on sealed and unsealed shoulder widths. The information therein is applicable to both rural and urban locations, although shoulders are not usually required in urban settings except for drainage or the storage of broken down vehicles.
Horizontal curvature and localised curve widening
The careful design of horizontal curves is one of the primary considerations in designing to minimise the danger of roadside hazards. In order for a vehicle to travel around a bend at a certain speed, the horizontal friction between the vehicle and the road pavement must be sufficient to counteract the inertial force that would maintain the vehicle's initial direction. Constructing a bend with as large a radius as the landscape allows is therefore the first step in providing a driveable path. However, it is desirable to have a consistent alignment standard over a section and well designed transitions from generous to tighter alignments.
For the construction of a new road or realignment of an existing road, Chapter 9 of Rural Road Design provides guidance on the entire process of designing a road's horizontal alignment.
Widening of the road pavement may be required at curves in the road, dependent on curve radius, lane width and vehicle sizes. One reason for this localised widening is that a vehicle (particularly a truck or bus) travelling around a curve will occupy more of the lane width than the same vehicle travelling on a straight. This increased width occupied by vehicles also reduces the clearance between vehicles travelling in opposing directions. Extra lane width at curves maintains an acceptable clearance.
The second reason for localised widening on curves is that vehicles typically do not maintain the same lateral position in a curve that they would on a straight. This is due to the requirement that a driver steer through the curve. Some deviation from a perfect path must be expected.
Vertical alignment is an important consideration in road design. Flat or almost flat grades should generally be provided where possible. Steep grades become prohibitive or even non-negotiable for heavy vehicles. Flat grades allow all vehicles sharing a road to travel at the same speed. Steep grades, on the other hand, cause different vehicles to travel at different speeds, which creates a higher risk of rear end crashes. Differences in vehicle speeds also contribute to queuing, which may be frustrating to drivers within the queue. Where vertical curves occur in conjunction with horizontal curves extra care in design needs to be taken.
The definition of a 'flat' grade varies according to a road's operating speed. For a speed environment of 100 km/h, a slope of up to five per cent can be considered flat. At lower speeds, steeper grades can be considered flat. The grades listed in Table 3 are taken from Austroads' Rural Road Design (2003b) and are suggested maximum grades for various terrain and speed environments. The practitioner must always beware of vertical grades or changes in grade that may impede sight distance. Refer to Section 2.2.6 for further information on sight distance.
Table 3: Maximum grades (%) by speed and terrain
Vertical grades of zero may be provided, although consideration needs to be given to type of drainage provided. Adequate drainage must be provided to prevent water pooling on the road surface during normal levels of rain. Longitudinal drains need to have adequate fall, generally accepted as 0.5 per cent grade.
Chapter 10 of Austroads' Rural Road Design (2003b) provides guidance on the design of aspects of vertical alignment issues including grades and
It is recommended that where it is necessary to have a length of steep grade, the length of the section be minimised. Vertical grades of zero to three per cent are considered to have little effect on the operation of all vehicles, while grades in excess of six per cent can have a significant effect on heavy vehicles for travel both uphill and downhill. Safety considerations need also to be addressed in relation to long downhill grades with regard to the risk of a crash due to break failure. Where it is necessary to have long steep grades, consideration for the provision of safety features such as passing bays and descending lanes to allow light vehicles to safely overtake slower moving heavy vehicles, or the installation of safety ramps and arrester beds to safely bring a runaway vehicle to rest should be considered. Provision of such features may be particularly relevant to roads with a reasonably high proportion of heavy vehicles. Chapter 13 of Austroads' Rural Road Design (2003b) provides guidance on auxiliary lanes for safety and capacity reasons. Section 13.7 provides guidance on the provision of runaway vehicle facilities.
A road surface needs to be constructed and maintained to a sufficient standard to ensure adequate skid resistance. The skid resistance of a particular surface is a result of the interaction of surface texture and the presence of moisture. For example, a situation where a pavement can hold water instead of draining properly can contribute to vehicles aquaplaning.
To be sure of the condition of an existing pavement it is necessary to conduct skid resistance measurements as well as an assessment of the level of rutting and the occurrence of potholes. Measurement of skid resistance and rutting can be undertaken using a number of methods, some of which are highly automated and efficient. The decision to act on the results of such measurements is left to the experienced practitioner, however a guide to the use of skid resistance values can be found in the Austroads publication Guide to the selection of road surfacings (Austroads 2003a).
Roads with a comparatively high volume of heavy vehicle traffic (usually major link routes) may require a higher standard of construction and maintenance than roads that predominantly carry light vehicles such as cars and vans.
It is important that adequate sight distance is provided whenever possible to allow drivers and other road users to safely negotiate the road. Sight distance can be affected by the road geometry (horizontal and vertical alignment), terrain (particularly on the inside of horizontal curves) and roadside objects (such as trees and signs). Section 8 of Rural Road Design discusses sight distance in general, while section 9.11 looks more specifically at sight distance on horizontal curves. Section 3.4 of AS1742.2-1994 discusses signage of substandard curves.
A number of sight distance types can be calculated, dependant on the driving environment. These include:
- Stopping sight distance - the minimum sight distance that should be available for a driver. As defined by Austroads' Rural Road Design (2003b), this is the distance that it takes for a "normally alert driver, travelling at the design speed on a wet pavement, to perceive, react and brake to a stop before reaching a hazard on the road ahead". Section 8.3 of Austroads' Rural Road Design (2003b) looks at stopping sight distance, outlines the method for calculation and provides tables for the minimum car and truck stopping sight distances required for various operating speeds and grades.
- Overtaking sight distance - the distance required by a driver "to safely overtake a slower moving vehicle without interfering with the speed of an oncoming vehicle". This is only considered for two-lane two-way roads, where an overtaking manoeuvre requires a driver to travel on the wrong side of the carriageway. Section 8.4 of Austroads' Rural Road Design (2003b) focuses on overtaking sight distance.
- Manoeuvre sight distance - is the distance required for a driver of a vehicle to react and manoeuvre around an obstacle. Manoeuvre sight distance is less than stopping sight distance and is the absolute minimum sight distance that may be provided. Section 8.5 of Austroads' Rural Road Design (2003b) discusses the circumstances in which the use of manoeuvre sight distance instead of stopping sight distance may be acceptable.
- Headlight sight distance - the distance for which a small unilluminated object is visible in a vehicle's headlights. It is generally limited to 120 to 150m, which corresponds to a safe stopping sight distance for 80 to 90 km/h. Section 8.6 of Austroads' Rural Road Design (2003b) looks at headlight sight distance.
- Horizontal curve perception distance - where a horizontal curve exists, drivers need to be aware of the curvature of the road ahead, react and slow down (if necessary) to safely navigate the curve. It needs to be ensured that a driver can see a sufficient length of curve in order to judge its curvature and safely navigate the curve. It is recommended that a curve should not commence just over the crest of a hill. However, where this situation is unavoidable, it needs to be ensured that drivers are aware of the curved road alignment ahead.
- Roadside features (such as embankments and vegetation) that limit sight distance should be removed or modified to ensure sufficient stopping sight distance on curves. If this is not practical, speeds should be reduced through such sections to compensate (for example, with warning signs). It is important that roadsides are maintained to ensure that sight distance requirements are sustained, for example by regularly pruning trees and cutting grass.
On substandard curves it may be appropriate to cut benches in high batters (see Figure 5) in order to improve sight distance. Sections 11.7 and 9.11.1 of Austroads' Rural Road Design (2003b) discusses design and use of batters and benching.
Benching on inside of horizontal curve
Drainage of the road surface and surrounding areas is an important consideration for road design. A number of different aspects need be considered with regard to drainage. These include:
- drainage of the road pavement - by providing adequate grade and crossfall so that the pavement is able to drain and pooling of water is avoided, which allows maintenance of skid resistance
- appropriate infrastructure to collect and transfer the water from the pavement, which may include kerb and channel or table drains
- a road reservation that can accommodate water run-off from adjacent land uses.
If constructed along a flow path, a road may need to be designed to be able to handle the run-off from adjacent land for a flood event.
Drainage design at the road design stage requires consideration of flood estimation. Chapter 16 of the Austroads' Rural Road Design (2003b) looks into drainage and lists a number of rural flood estimation methods for gauged and ungauged catchments. Similar information is available for urban areas in Chapter 15 of Austroads' Urban Road Design (2002).
The ideal roadside environment would be completely free of any obstructions to the safe passage of errant vehicles. Such a roadside would prevent injuries in run-off-road crashes by providing drivers with enough space to regain control of their vehicles and stop safely without colliding with any objects or the vehicle rolling over. However, it is usually not possible to construct a road environment completely free of hazards. There is usually a requirement for signage, utility poles and other roadside furniture, and often the topography of the landscape necessitates the provision of cut or fill embankments.
A clear zone is an area adjacent to the traffic lane that should be kept free from features that would be potentially hazardous to errant vehicles. The clear zone is a compromise between the recovery area for every errant vehicle, the cost of providing that area and the probability of an errant vehicle encountering a hazard. Where economically viable, the clear zone should be kept free of non-frangible hazards and all features that would 'snag' a vehicle or cause it to roll. Alternatively, hazards within the clear zone should be treated to make them safe or be shielded by a safety barrier. Clear zones are measured from the edge of the traffic lane.
Clear zones are intended as a guide to by which practitioners can assess sites, not a prescriptive value. Practitioners may provide a greater or lesser width depending on the risk factors applying to a particular site.
The Austroads method of calculating clear zone widths takes into account traffic volume, 85th percentile speed, curve radius and roadside slope. The use of this method is presented in Section 17.3 of Austroads' Rural Road Design (2003b) and the charts that make up the method are reproduced here as Figures 6, 7 and 8. Note that Figure 7 provides a multiplying factor for the output of Figure 8.
Clear zone width chart (Austroads 2003b)
Figure 7: Curve adjustment factors for clear zones (Austroads 2003b)
The following section identifies hazards that may be found in the roadside environment and possible treatments to reduce the risk of these hazards to drivers. For the purpose of hazard identification, the types of hazard that may be encountered in roadsides can be divided into five broad categories:
- rigid objects - trees, utility poles, culvert end-walls etc
- medians (cross median crashes)
- open drains
- bodies of water.
Notwithstanding that there are physical, environmental and economic constraints, the preferred treatments (in order of preference) of roadside hazards are:
- relocation to reduce the chance of them being hit
- redesign so that they can be safely traversed
- redesign to be frangible or break away, or to otherwise reduce severity
- shielding with a traffic barrier or impact attenuator
- delineation of the hazard.
Each option for hazard reduction is to be ranked according to benefit cost analysis techniques and engineering judgement.
Because of the number of variables and complexity of the analysis, computer software such as the following may be used to perform the quantitative analysis associated with the evaluation process:
- Roadside Incident Severity Calculator (RISC) developed by Main Roads Queensland.
- Road Safety Risk Manager (RSRM), a prioritisation program developed by ARRB Transport Research in association with Austroads.
- USA Roadside Safety Analysis Program (RSAP), refer TRB web site (http://www.trb.org/trbnet/projectdisplay.asp?projectid=678)
Section 11.7 of Austroads' Rural Road Design (2003b) provides guidance on use and design of batters. It is important that batters are constructed to allow errant vehicles to negotiate the slope safely in the event of a run off road crash. If there is a risk that a batter slope is severe enough to cause an errant vehicle to overturn during a crash, installation of a safety barrier should be considered. Design slopes for both cut and fill batters are listed for arterial and local roads in Table 11.7 of Rural Road Design.
One aspect of batter design not covered by the Austroads document is the concept of recoverable, non-recoverable and critical fill batter slopes. These terms refer to the likelihood of a vehicle overturning on various slopes. Practitioners should be aware of these definitions.
After running off the road onto a recoverable batter slope, a driver will usually be able to regain control of the car and return to the road or stop safely. On a non-recoverable slope, the driver is unlikely to be able to return to the road but will be able to stop safely at the bottom of the slope. A critical slope will probably cause the vehicle to overturn.
The AASHTO Roadside Design Guide (AASHTO 2002) defines recoverable slopes as flatter than 1 on 4 (ie. a fall of one metre for every four metres of width), non-recoverable as between 1 on 4 and 1 on 3, and critical as steeper than 1 on 3.
Cuttings and rock faces
Cuttings and rock faces are generally expensive to construct. Economic and environmental constraints often result in cuttings being as narrow as possible and prevent the provision of a cutting wide enough to allow for clear zones. Therefore, cuttings and rock faces should be cut to provide a smooth face to act as a rigid barrier, allowing errant vehicles to slide along and stop gradually. Uneven surfaces, may present a hazard to vehicles that happen to run off the road. If a smooth face and approach surface cannot be provided, it may be appropriate to install a barrier to prevent vehicles colliding with an uneven rock surface.
Deep, unprotected drains should not be provided at the base of the cut batter. Effective redirection of vehicles requires a flat even surface approaching the embankment.
Poles are a common road furniture item used to support signs (regulatory, warning, guidance, informative), road lighting and various devices. In line with the preferred treatment for roadside hazards (removal), the practitioner's aim should be to minimise the number of poles in the clear
The hazard presented by a roadside pole is related to both its location and type of construction. Both of these contribute to the hazard the pole may pose and the consequences of an errant vehicle hitting the pole.
Where possible, poles should be located such that an errant vehicle is unlikely to hit them. Minimum lateral set back distances for signs and for road lighting poles are specified in Australian Standards AS1742.2 and AS/NZS1158.1.3-1997 respectively.
Sign posts should be designed to be frangible in the event of collision with an errant vehicle, i.e. posts that are designed to fracture, break away, give way or bend such that the damage to a colliding vehicle and risk of injury to vehicle occupants upon impact is minimised. Small signs are usually supported by posts that deform in a way that causes minimum damage to cars, whereas larger posts and supports (for larger signs) may be provided with mechanisms that are designed to yield in a controlled manner upon impact.
Appendix C of AS1742.2-1994 discusses aspects of longitudinal and lateral placement and mounting height for signs, orientation, post type and selection. Signs should be erected such that sight distance is not compromised. Longitudinally, signs should be located to provide enough warning for a driver to be able to make a decision and respond as necessary. It is also important that signs are spaced far enough apart longitudinally that drivers are able to process the information before encountering another sign.
In a rural setting (unkerbed roads), signs should generally be placed between 2m and 5m from the edge of the outside lane, and at least 600 mm from the road shoulder. Signs should be mounted at a height at least 1.5m above the carriageway level.
In an urban setting (kerbed roads), signs should be located at least 300 mm behind the face of barrier kerbing and 500 mm from mountable or semi-mountable kerbing. Signs should generally be mounted at least 2m above the top of kerb or 2.5m above a footway, to allow for pedestrians and parked vehicles. In some circumstances, signs may be mounted lower than 2m.
Overhead mounted signs are often employed for multi-lane carriageways and freeways. They should generally be mounted a minimum of 5.3m above a carriageway.
Appendix B of AS/NZS1158.1.3-1997 discusses the use and placement of rigid and frangible road lighting poles.
Rigid poles do not deform to a great extent, but are designed such that they remain upright after an impact. Alternatively, frangible poles are designed to deform upon vehicle impact. Types of frangible poles include:
- slip base poles - break away at the base upon impact, allowing the vehicle to pass beneath the pole and causing aiming to minimise or avoid injury to vehicle occupants (see Figure 9)
- deformable poles - collapse over the colliding vehicle and are designed to bring the vehicle to a controlled stop at the base of the pole (see Figure 10). Deformable poles are designed to remain in the ground after being hit.
The decision to use slip base poles will depend on the space available and the resultant likelihood that a falling pole would cause injury to other users of the road or roadside area. For example, a slip base pole will usually be inappropriate where pedestrian or cyclist traffic is common because a falling pole would pose an unacceptable risk to those road users.
Slip-base pole (Austroads 1988)
Figure 9: Slip-base pole (Austroads 1988)
Aspects involved in the selection of pole type and set back from carriageway include:
- surrounding land use and pedestrian activity
- speed limit
- whether road is kerbed or unkerbed
- location - mid-block or intersection
- road alignment
- whether pole is to be located behind safety fence or on front of/ behind an obstruction.
Refer to section B6 of Appendix B of AS/NZS1158.1.3-1997 for detailed information.
Deformable pole (Austroads 1988)
Figure 10: Deformable pole (Austroads 1988)
Other - service poles
Service poles, such as electricity poles, are generally rigid as the consequences of one being toppled over can be very serious. The ideal treatment of service poles is to remove them and relocate services underground. Where this is not possible, service poles should be located where it is unlikely an errant vehicle may hit them. This may involve locating them at the property line (urban and rural) or in an easement (rural).
Trees greater than 100 mm in diameter located within the clear zone pose a potential hazard to motorists.
New trees should be located outside of the clear zone so that they do not pose a serious roadside hazard risk. Where existing trees are within a clear zone, and are deemed to pose a risk, the first option is to remove the tree. Where this is not feasible it may be appropriate to install safety barrier. Provision of a safety barrier will depend on a number of factors relating to site conditions, accident history, economics and the environment. However, such action should only be taken where it is determined that collision with the barrier is less severe than collision with the existing hazard (i.e. trees).
It is also important that trees are pruned regularly enough to ensure that any growth does not restrict sight distance.
Minor roadside obstacles (fire hydrants, mail boxes and other roadside hazards)
Minor roadside obstacles should not pose a serious risk to an errant vehicle that may strike the object. Objects containing horizontal rails capable of spearing vehicles (such as post-and-rail fences) can be particularly hazardous. Such objects should be located outside the clear zone or such that impact with the object should not result in a serious crash.
As for other roadside furniture, traffic lights can pose a hazard for any errant vehicles. They are often necessarily located close to the carriageway of intersections, which could lead to a higher risk of impact, although some measures can be taken to minimise this risk. Such measures include not locating a light pedestal on the outside of a curve, setting poles as far back from the carriageway edge as practicable, minimising the number of poles and joint use of poles. Provision of high skid resistance at intersections can also reduce the risk of a vehicle losing control at an intersection and skidding into traffic pedestals or other roadside hazards.
The ends of culverts that cross under the road or are located parallel to the road constitute hazards for motorists. They should be relocated, treated or shielded if within the clear zone.
Parallel to road (driveable treatment)
Driveable treatments need to be installed wherever a culvert exists parallel to the road and within the clear zone (see Figure 11).
Driveable culvert situated parallel to road
Figure 11: Driveable culvert situated parallel to road
Perpendicular to road (headwall treatment)
Culverts that run perpendicular to the road (i.e. run under the road) need to be:
- driveable if the fill batter is of a low enough slope
- protected with an appropriate barrier if the slope is not driveable.
Bridge end posts
Bridge ends need to be designed to prevent vehicles from running into end support posts, being speared by any horizontal bridge members or simply crashing through any approach barrier and being exposed to a hazard (eg roll-over, water course).
Stiffening needs to be provided on the transition from the semi-rigid approach barrier to the rigid bridge structure.
The piers of bridges over roads (at overpasses) should desirably be protected by a crash cushion or safety barrier.
A drainage channel is defined as an open drain usually parallel to the highway and within the limits of the highway right of way (AASHTO, 2002). Open drains are present on the majority of rural roadsides and may also exist on urban freeways. Their primary function is to collect and carry the surface water away from the roadway. Open drains are designed to accommodate run-off from heavy rain storms with minimal highway flooding or damage. Deep drains constructed close to the road may be the most efficient way of removing water but, unless they are of a suitable shape, they are a hazard for errant vehicles that leave the road.
Typical drains can be classified by whether they are designed with abrupt or gradual slope changes. Abrupt slope change designs include vee drains, rounded drains with bottom widths less than 2.4m, and trapezoidal drains with bottom widths less than 1.2 m.
Vehicles leaving the roadway and encroaching into a drain face three hazard areas:
- Drain front slope. If the front slope is 1:4 or steeper, the majority of vehicles entering the ditch will be unable to stop and can be expected to reach the bottom.
- Drain bottom. Abrupt slope changes can result in errant vehicles colliding with the bottom of the ditch.
- Drain back slope. Vehicles travelling through the ditch bottom or becoming airborne from the front slope can collide with the back slope.
The AASHTO Roadside Design Guide (AASHTO 2002) contains figures describing the preferred design for abrupt and gradual change slopes. These figures are presented in this Document at Figure 12 and Figure 13 respectively. Drain cross sections that fall within the shaded region of each of the figures are considered as traversable. These preferable drain designs are not considered hazardous and need not be constructed at or beyond the clear zone distance for a specific roadway.
Drain sections that fall outside the shaded area of the figures are considered non-traversable. As a general rule, these drains should either be:
- converted to a closed system (culvert or pipe)
- located beyond the clear zone
- where appropriate, shielded with a traffic barrier.
Preferred cross sections for channels with abrupt slope changes (from AASHTO (2002))
Figure 12: Preferred cross sections for channels with abrupt slope changes (from AASHTO (2002))
Preferred cross sections for channels with gradual slope changes (from AASHTO (2002))
Figure 13: Preferred cross sections for channels with gradual slope changes (from AASHTO (2002))
If the drain bottom and slopes are free of any fixed objects, then non-preferred drain sections may be acceptable for roads or projects where a better treatment is impracticable or uneconomical because of factors such as:
- restrictive right-of-way
- rugged terrain
- resurfacing, restoration or rehabilitation projects where these works result in an unavoidable change to the shape of a drain and it is not feasible to provide a compliant shape
- low volume, low speed roadways.
Drains of both the abrupt and gradual slope designs can funnel a vehicle along the drain bottom. This increases the probability of impact with any fixed objects present on the bottom or side slopes of the drain. Breakaway hardware may not operate correctly if the vehicle is airborne or sliding sideways when contact is made. For these reasons, non-yielding fixed objects should not be located on the side slopes or bottom of drains.
Back slopes typically occur when roadways are constructed by cutting the existing terrain away to develop the roadbed. If the slope between the roadway and the base of the back slope is 1: 3 or flatter, and the back slope is obstacle free, then the back slope may not be a significant hazard regardless of its distance from the roadway. Back slopes that will not provide a relatively smooth redirection or that can cause vehicle snagging should begin outside the clear zone or be shielded. This usually includes rough sided rock cuts when the rough face can cause excessive vehicle snagging.
Bodies of water
Bodies of water should be evaluated with respect to the degree of potential hazard they pose. This will be a combination of the amount of water and its accessibility. The depth may be ranked according to whether
- a vehicle can completely submerge, resulting in the drowning of uninjured non-swimmers, disabled or elderly persons, or infants
- water could fill an upright car to a point where an unconscious or injured driver or passenger would drown (typically / assumed to be a depth of 0.6 m)
- an upside down car would be in water deep enough that an unconscious person would drown (a depth of 0.3 m).
Fast moving bodies of water should are considered to be more hazardous than those that are still. In general, practitioners should carefully consider the risk associated with bodies of water over 0.6 m deep, or water courses with a normal base flow depth greater than 0.6 m, as these could cause a stunned, trapped, or injured occupant to drown.
Other factors to consider include the:
- slope of the vehicle path to the water
- total distance available in which to stop
- persistent or intermittent presence (flooding potential) of the water hazard
- presence of intervening obstructions that would reduce the likelihood of an errant vehicle reaching the water.
The practitioner should visualise the paths that errant vehicles are likely to take in reaching the water. If the water hazard substantial and likelihood of errant vehicles reaching that water is high enough, the practitioner should consider providing shielding to prevent access to that course.
In instances where a roadside hazard cannot be made safe, removed or relocated, it may be necessary to provide physical protection from the hazard. Safety barriers are available for a variety of applications and this section provides advice on selecting, installing and maintaining safety barriers.
The Australian Standard document "AS/NZS 3845:1999 Road Safety Barrier Systems" discusses various methods of roadside hazard protection and provides direction on the correct use of the different systems. The Standard has been the basis for a number of guidelines written by individual road authorities for use within their jurisdiction.
It should also be noted that the Austroads publication Safety Barriers (1987) is currently being reviewed and updated.
Decision to install a safety barrier
Safety barriers are a form of roadside hazard. When considering whether to install a safety barrier, it is important to remember that the barrier will present some danger to the occupants of errant vehicles, and especially to unprotected road users such as motorcyclists. A barrier should only be installed if collision with it will present less of an injury risk to vehicle users and occupants than would result from collision with the roadside hazard that is to be shielded by the barrier.
It is important to consider specifically the danger posed to motorcyclists by both the hazard and the intended safety barrier. As essentially unprotected road users, motorcyclists are particularly vulnerable to unforgiving roadside environments; any obstacle in the path of an errant motorcyclist has the potential to cause severe injury. If it is decided that a safety barrier is necessary at a site, attention should be paid to the design of the barrier to ensure that it poses as little risk as possible to colliding motorcyclists.
The following sections describe a number of roadside safety barriers and end-treatments. This list does not contain all available types of barrier, and the practitioner should be aware that manufacturers continually develop new or improved barrier designs. Accordingly, the information presented here refers to testing procedures, the results of which can be used to determine the suitability of proposed barriers. No barrier should be installed unless it has been shown to meet the applicable standards and can therefore be expected to perform satisfactorily.
Barrier types include rigid barriers, semi-rigid barriers and flexible barriers. Semi-rigid and flexible barriers are preferred as they generally cause less damage to vehicles during a crash, while a rigid barrier is suitable where space is limited and it is placed relatively close to the traffic lane (eg. narrow median).
Where a barrier is essential, the practitioner should bear in mind that barrier posts are the main cause of injury to motorcyclists. Other barrier attributes that are considered to be dangerous to motorcyclists (ATSB 2000) include upper and lower edges (particularly if jagged edges exist), protruding reflectors, low barrier mounting height (as motorcyclists can be thrown over the barrier) and rigid barriers.
Guards have been designed to reduce the severity of motorcycle collisions with barriers. These are available in a range of designs for various types of barrier. They should be considered for use at sites where motorcyclists are subject to high risk of collision with barriers.
Wire rope safety fence
Flexible barriers cause the least damage to vehicles, and pose the smallest risk of injury to vehicle occupants, of all barrier types. Flexible barriers usually consist of cables, held in tension, suspended from closely-spaced posts. The cables may be arranged in a variety of ways, depending on the manufacturer (see cross-section diagrams in Figure 14 and Figure 15).
Wire rope safety fence, twisted array
Figure 14: Wire rope safety fence, twisted array
Wire rope safety fence, vertical array
Figure 15: Wire rope safety fence, vertical array
Such a configuration is commonly known as a wire safety rope barrier. The posts simply support the cables and provide little resistance to a colliding vehicle. When a vehicle strikes a wire rope barrier, the barrier catches the vehicle and brings it to a halt. During a collision a wire rope barrier may deflect by more than a metre, meaning that a wire rope installation requires more clear space behind it than is required by a semi-rigid or rigid barrier.
One of the most common types of roadside safety barrier on Australian roads is W-beam (Type G4) guard fence. It is a semi-rigid barrier constructed of a steel rail mounted on steel C-channel posts. The profile of the rail resembles the shape of a 'W' turned on its end (see Figures 14, 15 and 16)
W-beam safety barrier
Figure 16: W-beam safety barrier
Thrie-beam safety barrier
Figure 17: Thrie-beam safety barrier
This type of barrier is designed to deform when struck but to also retain its tensile strength and keep the vehicle from passing through. The deformation of the barrier gradually dissipates the energy of the vehicle impact and aids in redirecting and stopping the vehicle.
W-beam used as a median barrier
Figure 18: W-beam used as a median barrier
Square hollow section
Another type of semi-rigid barrier is the square hollow section barrier. They comprise continuous square hollow sections and weak support posts designed to break away at point of impact to allow the square hollow section to work in tension and allow the errant vehicle to come to rest gradually (see Figure 19). Section 22.214.171.124 of the AASHTO Roadside Design Guide (2002) discusses the box beam barrier.
This type of barrier is no longer recommended by Austroads for use in Australia and should be replaced with wire rope or W-beam safety barrier at the end of its life.
Square hollow section used as a median barrier
Figure 19: Square hollow section used as a median barrier
Concrete barriers belong to the group known as rigid barriers. Rigid barriers are designed to retain their shape and position when struck by a vehicle, thereby requiring no deflection space. Often a rigid barrier is the only appropriate choice when space is limited, such as the median treatment shown in Figure 21. However, less aggressive barriers should be used where possible.
Various types of rigid concrete barrier are shown in Figure 20 below.
New Jersey Barrier F Type Barrier
Vertical Face Barrier Constant Slope Barrier
Figure 20: Concrete barrier profiles
The mode of operation of sloping concrete barriers such as New Jersey, F-shape and constant slope barriers is to redirect an errant vehicle and slow it down by forcing it to ride up the side of the barrier.
Concrete barriers and, to a lesser extent, steel beam barriers, can reduce sight distance around bends if not carefully sited. Practitioners should take this into account when specifying barriers on bends and, if possible, take other measures to improve sight distance or warn drivers to take more care of approaching hazards.
Rigid median barrier
Figure 21: Rigid median barrier
An appropriate type of vegetation can slow vehicles down without injuring occupants. It may also be successfully used in conjunction with a manufactured barrier to improve aesthetics. However, vegetation can not be relied upon to act as a reliable continuous barrier, and hence provide protection for motorists. Vegetation can be successful as a headlight screen, particularly on dual carriageways.
Post-and-cable fencing is in place along many road sections in Tasmania.
The fencing consists of heavy stranded cable slung between steel I-section posts. Two cables are usually employed, with one cable close to the top of the posts and another cable approximately 250mm below the top cable. The cables are not tensioned.
This type of fencing has not been tested against the criteria in AS3845:1999 and should not be used in any new installations. Any existing installations that require renewal or repair should be removed and replaced with an approved type of safety barrier. W-beam is the suggested replacement for post-and-cable fencing. It is able to be curved to a minimum radius of 3m, making it suitable for the small-radius curves for which post-and-cable fencing has been used in the past.
Layout of safety barriers
An installation of longitudinal safety barrier will usually require a leading terminal, an intermediate section and a trailing terminal. The exact layout of these components will depend on the individual site and barrier type, but the general concept is illustrated in Figure 22.
Typical safety barrier layout - plan view (AS3845:1999)
The non-rigid barrier types rely on a degree of tensile strength to enable them to restrain vehicles. They must be of a certain minimum length in order to develop sufficient tension. A barrier that is too short will not be able to deform around a colliding vehicle without breaking off its posts. A recommended minimum length for any W-beam barrier is 20m plus appropriate end terminals. Practitioners will need to seek advice on minimum lengths directly from barrier manufacturers for specific installations as specifications may vary depending on design.
Road safety barrier terminals are covered in section B2.3.11 of AS/NZS 3845:1999. Terminals may be either gating or non gating. Gating terminals are designed to allow a vehicle to pass through and come to rest in the clear zone area behind the terminal. Installation of a gating terminal should only be used if a driveable clear zone exists behind the barrier terminal. In other words there should be no hazards behind the gating treatment. If the intended position for a gating terminal is such that a colliding vehicle would pass through the terminal and strike a hazard, the barrier needs to be longer so that the terminal is in front of a more forgiving roadside.
Non-gating terminals are designed to redirect the errant vehicle without allowing it to pass behind the safety barrier. This type of terminal is appropriate if a hazard exists behind the safety barrier.
Terminals for rigid barriers
Unless the approach end of a rigid barrier can be buried in a cut embankment, it will usually be necessary to construct a semi-rigid terminal for the barrier. The junction between the semi-rigid and rigid barriers forms a gradually stiffening surface to guide vehicles onto the rigid barrier after they first make contact with the semi-rigid terminal. Semi-rigid terminals are discussed below.
Where it is necessary to protect road users in head-on impacts with rigid barriers, a crash cushion is the recommended treatment. A crash cushion consists of an arrangement of materials designed to reduce injury in head-on impact with the end of a barrier.
A large range of proprietary crash cushion products are available to suit various site conditions. Many such devices are designed to absorb the impact of a colliding vehicle and then return almost to their full pre-collision position. This design feature enables the unit to function as a crash cushion after a collision. While this feature should not be used as a reason to reduce maintenance, it does provide a degree of road user protection before repair is undertaken.
Note that it is not acceptable to use a sloped end as a rigid barrier terminal. Such designs have been found to increase the risk of colliding vehicles becoming airborne on impact. Where a ramp end exists it should be either removed and replaced with a more appropriate terminal treatment, or shielded by another terminal.
Terminals for semi-rigid barriers
Semi-rigid barriers such as W-beam can be terminated with a number of different terminal designs. The type used in a particular case will depend on the characteristics of the installation.
The breakaway cable terminal (BCTA) is common around Tasmania, although it has been superseded by the slotted breakaway cable terminal (SBCT), which is now the recommended terminal. The SBCT uses weakened timber posts and a slotted W-beam to cause the terminal to break and swing back behind the barrier when struck by a vehicle. The SBCT shown in Figure 23 has a clear, driveable batter behind it.
Slotted breakaway cable terminal (SBCT)
Figure 23: Slotted breakaway cable terminal (SBCT)
Various other types of terminal are available from a number of commercial manufacturers. The practitioner should consult barrier manufacturers for information regarding appropriate applications. Any terminal used must have met AASHTO's crash test requirements.
Terminals for flexible barriers
Flexible barriers do not require any special terminal treatments. The end of each length of barrier terminates at ground level and does not pose any more injury risk than the rest of the barrier.
Transitions between barrier types
Where two different types of barrier meet, it is necessary for a transition treatment to be constructed at the junction of the two barriers. For example, to connect a W-beam (semi-rigid) barrier to the concrete (rigid) barriers on a bridge requires a transition element stiff enough to ensure that a vehicle sliding along the deforming semi-rigid barrier does not suddenly become snagged on the unforgiving rigid barrier. The transition piece in this case is formed by a progressive stiffening of the W-beam for a short distance leading up to the rigid barrier. The additional stiffness is generated by closer spacing of support posts and nesting of two layers of W-beam. These features are shown in Figure 24.
Semi-rigid (W-beam) to rigid barrier (bridge barrier) transition
Figure 24: Semi-rigid (W-beam) to rigid barrier (bridge barrier) transition
A transition between flexible and semi-rigid barriers can be constructed by overlapping the flexible barrier in front of the semi-rigid barrier. A vehicle sliding along the flexible barrier will be travelling in a reasonably straight line and, at the end of the flexible barrier, will continue to slide along the semi-rigid barrier.
A transition from flexible to rigid can be achieved in a similar manner, by overlapping the departure end of the flexible barrier with the start of the rigid barrier. The posts of the flexible barrier will need to be positioned closer together on the approach to the rigid barrier to reduce deflection and thereby prevent a vehicle colliding with the end of the rigid barrier.
Wire rope safety barrier used as a median barrier - in transition with square hollow barrier
Figure 25: Wire rope safety barrier used as a median barrier - in transition with square hollow barrier
During works on or near a roadway, protection of workers needs to be ensured, as well as maintaining a safe environment for pedestrians and motorists. The purpose of any temporary safety barrieras at a works zone is therefore to redirect errant vehicles, preventing them from entering the works area, and minimise injury to the vehicle's occupants.
Safety barriers for works zones are designed to be portable, providing for quick installation and removal or relocation.
Works zone barriers may be necessary for a number of reasons:
- preventing vehicles from driving into works areas
- separating opposing flows of traffic on a temporarily constricted carriageway
- protecting incomplete structures from vehicle impact
- reducing or eliminating the need for temporary reductions in speed limit.
The decision to install temporary safety barriers at works zones must be made as part of a full assessment of the traffic management needs of an individual works zone. Such an assessment must be made in accordance with the Tasmanian Code of Practice for Traffic Control at Works Sites (DIER 2004). Practitioners should refer to that document for guidance in this area.
AASHTO (2002) Roadside Design Guide American Association of State Highway and Transportation Officials, Washington, D.C., United States of America.
AS/NZS 1158.1.3:1997 Road lighting - Vehicular traffic (Category V) lighting - Guide to design, installation, operation and maintenance, Standards Australia, Homebush, Australia.
AS 1742.2:1994 Manual of uniform traffic control devices, Part 2: Traffic control devices for general use, Standards Australia, Homebush, Australia.
AS/NZS 1906.1:1993 Retroreflective materials and devices for road traffic control purposes - Retroreflective materials, Standards Australia, Homebush, Australia.
AS/NZS 3845:1999 Road safety barrier systems, Standards Australia, Homebush, Australia.
Austroads (1988) Guide to Traffic Engineering Practice, Part 12 - Roadway Lighting, Austroads, Sydney, Australia.
Austroads (2002) Urban Road Design, Austroads, Sydney, Australia.
Austroads (2003a) Guide to the selection of road surfacings, Austroads, Sydney, Australia.
Austroads (2003b) Rural Road Design, Austroads, Sydney, Australia.
ATSB (2000) Review of Wire Rope Safety Barriers, ATSB Working Party Report, June 2000.
Department of Infrastructure, Environment and Resources (DIER) (2004) Tasmanian Code of Practice for Traffic Control at Works Sites, DIER, Hobart, Australia.
Dravitzki, V.K., Munster, D.E.C. and Wong-Toi, D. (1998): Safety Benefits of Using Rumble Strips in Road Safety Research, Policing and Education Conference Proceedings, Land Transport Safety Authority, Wellington, NZ.
Ligon, C.M., Carter, E.C., Joost, D.B. and Wolman, W.F. (1985): Effects of Shoulder Textured Treatments on Safety, Report No. FHWA RD-85/027, Federal Highway Administration, Washington DC, USA..
Transport Research Board (1998) USA Roadside Safety Analysis Program (RSAP), refer TRB web site (http://www.trb.org/trbnet/projectdisplay.asp?projectid=678)