Highways Agency

Chapter 4 - A Comparison With Some International Practice

4.1 The Group were asked to look at the relevant international experience and whether there were any features that would be worth considering for introduction into UK advice. TRL were asked to investigate the practice in a selection of countries, where transport infrastructure is similar to the UK and report on some of the main features in their standards. Historical information was also collected from library sources.

4.2 This chapter considers the international practice and looks at how it compares with the application of the standard on major roads in the UK. The data on barrier provision collected by TRL for the M3, M4, A34 and A339 has been used to estimate the level to which the standard is implemented on the road network. Additionally, data on the lengths of barriers at trunk road and motorway bridges over rail lines has been collected by the Highways Agency to aid the comparison. The Group had already considered the overall safety record on UK roads in relation to other countries (see paragraph 2.21).

International Standards

4.3 In approaching a range of countries TRL asked for information on the following:

1. Current Standards for safety barrier systems.
2. Documents that control or advise on where, if at all, particular safety systems should be installed.
3. Documents that relate to the specific provision of safety systems where roads pass over rail tracks.

The response was varied with returns being received from a dozen or so countries. No country provided details of how widely their standards are applied or the effectiveness of their measures at reducing the number of injuries arising from road accidents.

4.4 International experience shows that other countries consider a variety of factors (some of which are combined) in order to determine not only the length of safety barrier, but also the containment capability of the barrier and the overall layout of the barrier. These factors include traffic speed and density, the geometry of the site, the distance of the hazard from the main carriageway and the physical dimensions of the hazard. The UK standard also considers such factors and provides advice on the basis of the minimum acceptable design with general advice on where additional provision should be considered.

4.5 International experience has shown that there are a number of differences between the methodologies surrounding the installation of safety barriers in the UK and those used in other countries. However, it is not currently known whether the provision of safety barriers 'in the field' is the same as that specified in their standards. For example, standards will generally apply to new works and hence existing roads may not necessarily be retrospectively upgraded to the requirements of new standards. It is also unclear how effective these alternative approaches to safety barrier installation are at reducing injury rates. Accident figures from other countries, where sufficient detail was available, would give a guide to this, but time restraints have not allowed the Group to consider this.

4.6 TRL assessed the international information they received and produced comparisons against the areas connected with the provision of barriers that the Group were interested in, broadly being the length and strength of barriers, the choice of barrier type and risk assessment methods. A summary of the detailed comparisons is at Annex 6. It is important to note that not all countries provided the same type of information, however where information has been provided, it has been included, where relevant. The sections below therefore are not comprehensive.

Barrier Length

4.7 In the UK the minimum length of barrier to be installed in advance of any hazard is 30m. This stands towards the top of the range of countries that use a defined minimum length. Switzerland define a minimum length of 50m, while the shortest minimum length comes from California at 15m (see Annex 6 Table A6.1).

4.8 The road survey of sections of the M3, M4, A339 and A34 (totalling approximately 160 miles (257km) in length), has found that all of the 379 safety barrier lengths present on the nearside, and at hazards, were at least 40m. These lengths included twenty on the approach to bridges over a live railway line. Lengths of up to 2300m were found in-situ on both the M3 and M4; a full table of the lengths is in Annex 4 Section 1. This highlights the fact that within the UK there are notable differences between the minimum provision quoted in standards and the lengths that are actually installed 'in the field'.

4.9 The Group were interested in the length of safety barriers provided at rail bridges internationally. The UK specifies 30m as the minimum length of barrier to be installed at the approach to a bridge over railways. Other minimum lengths range from 16m in Austria to 98m in New York (see Table A6.2 in Annex 6). In Germany two parallel rows of safety barriers (of different lengths and different levels of containment) are recommended on bridge approaches.

4.10 Other countries specify a method of calculating the length of safety barriers, including those at bridge approaches, although a minimum may also be specified. The approach is varied but is generally based on the anticipated angle of approach of a vehicle. The application of such methods to UK major roads could produce bridge approach lengths in a range from about 105m (using Australia's method of calculation) to 140m (using Canada and New York's method of calculation).

4.11 The Group also considered an analysis of the lengths of safety barrier at the approaches to motorway and trunk road bridges over rail lines in England, provided by the Highways Agency. This showed that the majority of approaches to rail bridges have safety barriers installed in excess of 30m in length. The specified minimum length of 30m at bridge approaches is provided to protect the end of the bridge parapet. Additional lengths of barrier ought to be provided according to the hazard that the bridge passes over or to the roadside conditions in advance of the bridge approach, leading to total lengths well in excess of 30m. There are locations where the barrier length is shorter than the 30m standard due to local physical conditions (such as requirements for access) and at these locations a formal departure from the standard will have been approved (see Chapter 2 for an explanation of Departures from Standard). Table 4.1 below provides a breakdown of the lengths of safety barrier found on a large sample of sections of the motorway and trunk road network in advance of rail bridges (see Annex 4 Section 2).

Table 4.1 Lengths of Safety Barriers on approaches to rail bridges on English Motorway and Trunk Roads (see Annex 4 Section 2 Table A4.2.1).

4.12 In the USA regulations indicate that the length of safety fence may be shortened by the introduction of a long flare, angled away from the main carriageway. The regulations state that "this enables a hazard to be protected, without the costly application of long sections of safety fence. This method is particularly effective at reducing the probability of a vehicle entering a hazardous area after leaving the carriageway at a narrow angle. It may have adverse consequences for the vehicle occupant(s)".

Barrier Height

4.13 In the UK most types of metal safety barrier are required to be 0.61m high, measured from the ground to the centre of the barrier beams. Many other countries specify barrier heights measured from the ground to the top of the barrier beams. These range from 0.61m (USA lower limit) to 1.0m (Belgium), although the distance between the top and centre of a barrier beam will depend on the design of the barrier system involved. A comparison between different countries is in Annex 6 Table A6.3.

Selection of Barrier Type

4.14 In the UK there are several approved barrier types available for use, which have different performance characteristics, whilst still meeting the required criteria. For example, rigid concrete barriers, flexible wire rope and semi-rigid steel safety fences perform in different ways during impact and may have different overall levels of containment, in addition to different installation and maintenance costs. The information collected during testing, especially that relating to the barrier's performance during impact, can then be used to determine which type of safety barrier, and in what configuration, is most suitable for a particular application.

4.15 UK standards define the containment levels required in particular locations, but do not specify particular safety barrier products. As a result, they identify some locations where a greater level of protection for the road user may be required. Other countries select the type of barrier to be used according to an assessment of various criteria, such as traffic conditions, distance of the hazard from the carriageway, field experience and so on. Australia, New Zealand, Canada and USA, for example, have an established set of criteria that cover barrier performance, barrier deflection, site conditions, cost and so on, in taking decisions about what type of barrier to use.

Barrier Strength

4.16 A comparison between the barrier containment (strength) used in different countries is in Annex 6, Table A6.4. The comparison shows that most countries use barriers with the same strength as the UK for the containment of cars ('normal' containment). For the containment of heavier vehicles ('high' containment), other countries use either the same or a lower barrier strength than the UK.

Risk Assessment

4.17 In the UK there are specific conditions under which safety barriers are installed on the nearside of major roads. Several other countries use different methods ranging from the individual assessment of each case to an assessment of the hazard using standard data.

4.18 The degree of risk assessment varies between countries (see Table A6.6 of Annex 6). France, for example, uses severity indices and looks at the probability of leaving the carriageway, the consequences for the vehicle occupants and the consequences to third parties. Australia has a methodology that uses a flowchart which defines the problem (in terms of the hazard, and an associated severity index, traffic flows and speeds and roadside geometry), determines the risk (using probabilities) and then develops options. Canada and the USA produce lists of roadside hazards that normally require protection. They then consider whether the consequences of a vehicle striking the hazard or running off the road are more serious than the consequences of a vehicle hitting a barrier, and then apply cost benefit analysis.

Road Geometry

4.19 TRL's comparisons showed that most countries have criteria for determining the provision of barriers on embankments using embankment height, the gradient of the embankment slope, or a combination of the two (see Table A6.7 in Annex 6). In the UK, it is only the first of these criterion that is used. A number of countries, including the UK, also apply criteria for providing barriers when there is a curve in the road of less than a certain radius.

Use of Clear Zones

4.20 A clear zone is a zone from the edge of the carriageway, which is traversable and obstacle free. The aim of a clear zone is to allow errant vehicles to be brought under control and appropriate action taken. This zone varies widely between countries, due in part to the available space and the approach taken to the geometric design of the road.

4.21 TRL's comparisons (see Annex 6 Table A6.8) determined that there was a range of clear zones internationally, from 2.5m in Hungary to 13m in the Netherlands, with some countries defining greater clear zones for motorways than for other high speed roads. The UK defines a clear zone of 4.5m for motorways only.

4.22 Australia relies on a graphical method for determining a clear zone. There, the width of the clear zone is dependent upon the gradient of the embankment at the side of the road, and the average daily flow of traffic. Canada also uses an approach based on the gradient of the embankment.