Fit for Purpose. Do road safety barriers work?

< Back

Fit for Purpose. Do road safety barriers work?

Helen Strong, 2015


Dictionary definitions will tell you that “fit for purpose” means that the item you have been sold is good enough to do the job that it is designed to do. Alternatively that it is of a necessary standard for its intended use. Great — if you know the job, the standards and the design specifications!



“Why standards?” you may ask.  Standards provide a quality level reference and ensure systems are compatible (interoperability). In general one needs a benchmark for comparison of the outcomes or performance of an activity. They provide a common defined level of safety and reliability. We know what to expect. We can reduce risk, avoiding products that could be dangerous for our well-being.  On the business front standards save time as suppliers don’t need to calculate new levels for each job. According to the experts, standards encourage innovation and awareness of developments.


In the context of road safety, national roads agencies throughout the world have been working for decades to determine what the industry should expect when installing road safety barriers. Here the introduction of standards has the objective of providing design guidelines to help reduce road crashes and their consequent fatalities.  For road barrier systems standards are meant to specify minimum levels of containment and deflection; to guide the manner of installation, repairs and maintenance; and provide a point of reference when testing new products.  Finally the presence of standards allows clients to specify the performance level that they judge necessary when constructing or modifying our roads.


South African road barrier standards are guided by European (EN1317) and American (NCHRP) standards. The standards cover not only the longitudinal barrier systems, but include specifications for barrier terminals and transitions from one system to another. There are a couple of issues regarding recognition of standards. At one time SABS accreditation was bandied about as an acceptable standard. NO. The SABS mark is recognition of a quality of manufacturing, not a performance standard.  Secondly, whatever the type of barrier, it needs to be manufactured and mounted to be in accordance with prevailing SANS requirements. Thirdly, and crucially, in the road safety barrier world, to know that a system conforms and be sure of its performance, you need to make sure that the system has been installed correctly according to the standard against which it has been crash tested.


The next piece of the puzzle is that road safety barriers need to be “fit for purpose”.

Essentially this implies that the road designer knows what the purpose is!


Since the approach to building roads is scientific, engineers first need to have hard facts about road conditions.  To this end a road safety audit is almost mandatory.  How does the environment create difficult conditions for road users?  What mitigating actions need to be taken to increase the safety level for travellers?


How is “purpose” established in the context of keeping vehicles, pedestrians and road workers safe?   Engineers will tell you that there are at least four elements that need to be interrogated:


–       What types use the road

–       How many

–       How fast do they travel

–       What time of day


–       How many

–       Routes

–       Type (adults / children)

Pavement characteristics

–       Straight or winding  (sight distance and reaction time for curves)

–       Alignment (horizontal and vertical)

–       Pavement materials or surface

–       Number of lanes (e.g. single or dual carriage)

–       Lane width

–       Edge break (depth)

Hazards / Other

–       Intersections / interchanges

–       Utility structures / lighting masts

–       Drainage or water requirements

–       Signs

–       Bridge parapets

–       Trees

–       Weather conditions

–       Provision for emergency vehicles and stopping

–       Other

And when the road is being built, road construction managers will be concerned about the risks to their site. They will be conscious of the need to provide a safe working area, which will allow people and equipment to survive an impact to the barriers which are meant to be protecting them.


Each of these elements needs to be taken into account when defining the purpose of road safety barriers. What is “fit for purpose” under given environmental conditions? Men and women involved in the provision of SAFER ROADS are both professionally and legally responsible to do all that they can to reduce the possibility of road accidents. It all adds up to the serious need to abide by best practice in provision of road safety elements.


The first decision is whether a barrier is necessary at all. Is there a need to keep vehicles apart; to prevent them from leaving the road; or to stop them from damaging assets on the side of the road?  Yes? Then what type of barrier should be installed?


The most obvious factor to be catered for by safety barriers is the size and speed of vehicles which normally use the road.  Barriers need to contain the impact energy so as to prevent the vehicle from leaving the road or crossing over into the path of on-coming traffic. They need perform in such a way as to:

  • Minimise damage to occupants of the vehicle
  • Not cause the vehicle to roll, to
  • Avoid pushing the barriers into equipment or people working on the side of the road, and
  • Prevent the vehicle from vaulting into a hazard or unlucky pedestrian on the other side of the barrier.


Internationally standards have been developed by the entities responsible for national highways. South Africa has adopted the NCHRP framework within which barrier strength and containment is related to the expected impact. Given the traffic type and volume the most appropriate containment level can easily be chosen. But how can one know that the barrier will live up to its promises?  This is the tricky part.


The only way that an engineer can have confidence in barrier performance is to review the crash tests which have been conducted by reputable testing facilities. It could be argued that computer simulations are sufficient – in fact they can be very useful to allow product development and experimentation with different conditions. However, in the final analysis computer simulations need to be validated. How? Against the performance of the same barriers in an actual crash test.   One also has to evaluate the test conditions.  How long was the test barrier? Was it artificially “stronger” due to the proximity of anchors?  What was the type and speed of the vehicle? Angle of impact? Armed with this knowledge the engineer can then accurately match the performance level required to the type of barriers that should be used.

Road characteristics such as intersections have been shown to be associated with crashes. (“36% of vehicles were turning or crossing intersections”, according to NHTSA, 2008) This article will not address the need for signage and alerting drivers to upcoming “accident hot-spots”. It does however look at prevention of vehicles from leaving the road into hazardous conditions. The same report in critical pre-crash conditions identifies that

  • 22% of crash vehicles ran off the edge of the road;
  • Multi-lane driving is hazardous (51.9%); and that
  • 61% of the crashes occurred on roads that were “not physically divided”.
  • Curves are associated with 21% of crashes
  • Of events associated with road-way conditions, 46% were due to “slick roads” (ice or debris).

Remembering that the engineer has to meet the road conditions, the question becomes: What types of barriers are available? What features are necessary components of the barrier system?



Ideal for direction and delineation of traffic. However, not suitable in high speed areas – there they will not fully protect pedestrians from the moving vehicles. In his comparison of road safety barriers Grzebieta (2015) refers to them as temporary barriers. He points out that when first used in Australia they were required to be marked “NOT TO BE USED AS A SAFETY BARRIER”. Even with the addition of water ballast the officials there saw fit to create an additional “Level 0” to their test matrix to have some form of performance standard for comparison.  In spite of the criticism, there is an Australian supplier that claims that their plastic barriers have satisfied the evaluation criteria of NCHRP 350 Test Level 2 (TL-2). And a British company indicates that their water-filled system meets the entry level crash test safety standard: BS EN1317 – N1 (80Kmph)

Purpose conclusion: fit for delineation, not for large vehicles at speed.


Wire Rope

These so called flexible systems have been shown to be successful in reducing fatalities associated with crashes. Energy from impacts is dissipated along the length of the system, and redirection of vehicles accomplished in such a way as to reduce the impact on vehicle occupants.  However, according to Grzebieta (2015) they are not motor-cycle friendly and he says that cars have a tendency to under ride the ropes (due to maintenance issues in keeping the system “as installed”).  The third issue mentioned is that wire rope systems cannot redirect rigid and articulated trucks. He points out, however, that this is a common feature with W-beam guardrail and medium height concrete barriers.


Key findings in the Texas evaluation of cable barriers (Scott et al, 2009) included

  • Capital cost effectiveness compared with other systems.
  • Maintenance is an issue from the point of view of speed of repair, availability of trained technicians and cost of repairs. (Definitely applies in South Africa.)
  • Installation was affected by soil conditions, requiring special preparation of the soil when installing wire rope systems. (Same issues in South Africa.)


Purpose conclusion: Works well, but if the likelihood of impacts is high in an area, then other systems could be preferable to contain repair costs. In the South African context special attention needs to be given to the capacity for correct installation and repair of non-functional systems after impact.



The strength of this type of barrier depends on the shape of the railing; its height; and the type of posts (and the interval between them) that are used for mounting the steel.  Grzebieta (2015) provides information about USA crash tests on this type of barrier. “The typical post length is 1830 mm and the post spacing is 1905 mm. Strong-post W-beam barriers using wood or steel posts and wood blocks … have passed NCHRP Report 350 Test Level 3 crash tests, whereas strong-post W-beam barriers using steel posts and steel blocks … have only passed NCHRP Report 350 Test Level 2 crash tests”.


In spite of its ubiquitous presence, Lawrence Bank of the University of Wisconsin criticises guardrail for lacking innovation since its introduction in the 1960’s and for its reputation in not protecting trucks and today’s modern vehicles with their high centre of gravity. Smaller cars too have been shown to either vault over the barrier, or in other cases suffer undue damage due to their wheels sliding in below the railing.  Clearly the height of the railing and the type of posts are important considerations when installing such a barrier. Recently (August 2015) there has also been a controversy regarding the end barriers, where the State of Texas is being sued for allowing a new W-beam design without crash testing.


Purpose conclusion:  experience has shown that the barrier is functional, but where heavy vehicles are a factor on the road, traditional guardrail systems should probably be redesigned (and crash tested) or replaced with systems that can handle the impact.



This type of barrier has the advantage of being suitable for known containment levels, and when the working width is constrained, or available deflection is limited. (Grzebieta (2015) gives the examples of bridge barriers, where the hazardous conditions are close to the edge of the road and the median is narrow.)


Concrete safety barrier systems are gaining in popularity. They can be used as temporary or permanent safety barriers. With a known level of dynamic deflection they provide a greater level of protection for road construction workers. They are more likely to contain collision vehicles. Importantly where maintenance budgets are limited, they are almost maintenance free AND do not pose a risk to motorists because they remain functional even after severe collision.  An added benefit is the reduction of disruption to traffic from repair crews who are called out to repair barriers.


Concrete barriers come in different shapes. In 1955 the New Jersey barrier emerged in the United States. When it became evident that of small vehicle roll-overs were associated with this shape of barrier (EUPAVE, 2012), Britain introduced the single slope barrier and Europe developed the step-barrier shape.  In South Africa the F shape is preferred. As far as motor cyclists are concerned concrete safety barriers, since they do not have posts, pose less of an injury risk than the other systems.


Concrete barriers are sometimes criticised for inducing a higher level of injury to vehicle occupants. EUPAVE (2012) report that a study by Ove Arup & Partners found that no direct correlation of the higher ASI values with severity of injuries.  They also point out that “…concrete barriers also help to eliminate injury and deaths associated with cross-over accidents, barrier intrusions and deflections, and loss of vehicular control on soft verges, all of which are typical of steel barrier systems.”


Purpose conclusion: Concrete road safety barriers are “fit for purpose” in most areas where vehicle restraint is required, but especially when

  • Heavy vehicles are present
  • One wants to prevent vehicles from leaving the road and crashing into hazards or falling into ditches, embankments, culverts, ravines etc.
  • The possibility of cross-over accidents is high
  • Workers need protection when constructing or repairing roads.



Best practice in installing road safety barriers takes account of the road environment, evaluates the risk levels to all road users, and then installs systems (according to the suppliers) specifications which will minimise injury, maximise system availability and reduce the financial and social costs that are associated with the installation. Then the system is “fit for purpose”.




Australian Construction Products, date unknown, ArmorZone TL2 Plastic Barrier, New South Wales, ACP  Available at:  Accessed: 11 August 2015


Bank, L.C., 1999, New design could improve highway guardrail safety [Online], Madison, University of Wisconsin, Available at:  Accessed on 11th Aug 2015


Cooner S A, Yatin K. Rathod, Dean C. Alberson, Roger P. Bligh, Stephen E.

Ranft and Dazhi Sun, 2009, PERFORMANCE EVALUATION OF CABLE MEDIAN BARRIER SYSTEMS IN TEXAS, [Online], Austin, Texas Department of Transportation

Available at:

Accessed: 12 August 2015


EUPAVE, 2012, CONCRETE SAFETY BARRIERS: A SAFE AND SUSTAINABLE CHOICE, [Online] Brussels, European Concrete Paving Association. Available at: Accessed: 12 August 2015


Grzebieta, R.H., Zou R., Jiang, T. and Carey, A.,2015, Roadside Hazard And Barrier Crashworthiness Issues Confronting Vehicle And Barrier Manufactures And Government Regulators. Paper Number 05-0149, [Online], Monash University, Dept. of Civil Engineering

Available at:

Accessed 11 August 2015


NHTSA,  2008, National Motor Vehicle Crash Causation Survey: Report to Congress, [Online], Available at:  Accessed 11 August 2015


MacMillan Dictionary, date unknown, Fit for purpose, [Online] Available at: Accessed 10 August 2015


Oaklands Plastics, date unknown, Restraint / Crash Barrier – 50mph, [Online], Birmingham, Oaklands Plastics Ltd.

Available at:  Accessed 11 Aug 2015


Ochsner, N., 2015, Update: Expert Says Guardrail Test Failed , [Online] Norfolk, U.S.A., 13 News Now     Available from:  Accessed 12 August 2015


Sicking, D.L. and Mak, K.K., 1999, Improving Roadside Safety by Computer Simulation, [Online] Available from:  Accessed 11 Aug. 2015


Wiktionary, date unknown, Fit for purpose, [Online] Available at:, Accessed 10 Aug. 2015