Probability of road trauma in NorthConnex highway 03 Sep 2020

The 9km long NorthConnex road tunnel in north west of Sydney is being promoted by the New South Wales Government and the toll road operator Transurban as a safe, reliable, free flowing motorway alternative to the currently congested Pennant Hills Road which is carrying up to 8,400 multi-combination vehicles per day. It is proposed that the tunnel will carry 5,000 heavy vehicles/day and a larger volume of passenger cars.

For safety in operation concerns, the total number of accidents in the tunnel and their severity will be determined by the traffic volume and the percentage of heavy vehicles. Remarkably, in the official environmental impact assessment, issued by the New South Wales Department of Planning, omits consideration of road safety⁽¹⁾. In fact, the Director of Planning did not include road safety as a required item in his instructions to the Department of Planning.

Investigation of road safety on toll roads operated by Transurban reveals a different view of performance to that published in the Transurban annual reports. These reports show a record of continual improvement in safety which is not confirmed by data of road accidents published by the New South Wales CRS, Centre for Road Safety, an agency of the State Department of Transport. The CRS measures trauma by the ratio of the total of killed and injured persons to the total number of crash events.

The increasing costs of trauma, now at an estimate of $30 billion/annum according to the ECON (Economic Connections) report of 2017, a figure that suggests that safety has not improved on Australian roads. The extent of road trauma in Australia can be judged by the detailed data on accidents in New South Wales (NSW) published since 1997 by the CRS.⁽²⁾ An example is the data representing six motorways in the Sydney regions, five of them that include long lengths of toll road underground twin tube alignment with two or three lanes in each direction. In the seven years from 2012 to 2018, data for the seven Sydney motorways M2, M4, M5, M7, and the Cross City Tunnel and underground Eastern Distributor highway, totalled 3,563 persons killed and injured (Fig 1).

CRS data on tunnel accidents in the Sydney region is generally not extensive enough to enable meaningful statistical results to be derived, as accidents in tunnels are not always recorded separately to accidents on surface highway carriageways. The exception is the underground Eastern Distributor highway, which has been used to demonstrate the difference in road trauma resulting from accidents on a surface road compared with those in a tunnel (Fig 2). Investigation has shown that congestion plays a key role in increasing the ratio particularly beyond the year 2015.

Fig 2. Trauma in the Eastern Distributor tunnel compared with that on the M7 open motorway
(Compiled from data published by the NSW Centre for Road Safety 2009-2018⁽²⁾)

It is generally accepted that when accidents occur in road tunnels they can have more severe and economic consequences than in open road sections.⁽⁵⁾ Certain factors such as the proximity of the tunnel walls, limited sight distance, differences in lighting at the entrances and exits, increase the probability of crashes in a tunnel. Rear end crashes are also found to be more likely in tunnels which can lead to fires which may release toxic fumes. The risk is increased by traffic conflicts induced by driver behaviour such as tailgating in the same direction and other manoeuvres likely to reduce time to collision (TTC). TTC is a key variable in the prediction of crash frequency.

The response to road trauma in Australia can be compared with a review of policies in other parts of the world.

In the European Union, the occurrence of catastrophic tunnel accidents that have occurred, involving serious injuries and deaths, resulted 15 years ago in the issue of a formal directive 2004/54/EC, requiring member states to harmonise their responses to this situation.⁽⁶⁾ The requirement of member states is to carry out a quantitative risk assessment for road tunnels as experience showed that safety in tunnels was inferior to that of associated motorways.

Evidence from Italy

Results of a four-year investigation into the safety performance of road tunnels in Italy were reported in 2012 at the 5th International Congress on the topic of Sustainability of Road Infrastructure. The authors obtained separate data for tunnels and associated motorway sections. They based their analysis on 195 tunnels with unidirectional traffic, with two or three lanes and tunnel lengths of between 500m and 3.25km. The database consisted of 762 severe crashes involving 775 injuries and 18 deaths.

The results of this research allowed the development of a crash prediction model for tunnels and associated social cost rates. The model was able to show that severe accident rates in tunnels are likely to be up to two times the rates for the associated motorways and that this result applied to 136 of the 195 tunnels studied. The costs of severe accidents were shown to be more adverse in 164 of the 195 tunnels. A general conclusion from this investigation is that total accidents and accident severity in tunnels is affected by traffic volume and a high percentage of trucks, a finding relevant to the operation of NorthConnex in Sydney.

Evidence from the Netherlands

The Netherlands has 14 motorway tunnels with a total tube length of 10km. In 2012, the Netherlands introduced a regulation that prescribes a standard set of safety measures for every tunnel type. Experience of road crashes in the Netherlands also suggests a higher rate of accidents in tunnels than for open road sections.

Fig 3. Compiled from BITRE data comparing deaths in collisions between light and heavy vehicles⁽⁸⁾

Evidence from Switzerland

Experience in Switzerland reveals that although the risk of accidents is lower in longer tunnels, the risk increases in proportion to AADT, average annual daily traffic, volumes and the percentage of heavy vehicles. The finding agrees with Italian experience.

Evidence from Singapore

Singapore has two major highway tunnels, the 12km long Kallang/Paya Lebar Expressway with a 9km underground section and the 17km Central Expressway of which 2.4km is underground. It is reported that approximately 70% of crashes in these tunnels have been rear end crashes which are a major cause of fires and the release of toxic fumes.⁽⁷⁾

Singapore has developed a methodology that can estimate the frequency of rear end crashes by measurement of time to collision (TTC) assisted by real time video data collected at 200m intervals along the tunnels. TTC is defined as the time that remains until a collision would have occurred if the collision course and speed difference between the two vehicles is maintained. TTC is correlated significantly with traffic volume. Observations have shown that to avoid rear end collisions, TTC should be in the range 2 to 4 seconds. The crash frequency is then estimated by applying statistical methodology to the TTC data which involves deriving its probability distribution. If TTC is less than the lower value, the traffic would be defined as exposed to traffic conflicts.

Trauma resulting from collision between light and heavy vehicles

Another major concern for accidents in tunnels is the probability of road trauma resulting from crashes between heavy vehicles and passenger cars. The NorthConnex tunnel will carry a mix of vehicles ranging in size from passenger cars to 16 axle road train trucks. The total number of accidents and their severity will be determined by the traffic volume and the percentage of heavy vehicles.

Portal of the NorthConnex underground highway

Data compiled by Federal Bureau of Infrastructure Transport and Regional Economics (2018) through a nine-year period has been used to estimate the probability of deaths in road trauma crashes.⁽⁸⁾ It was found that a passenger car driver is between 15 and 25 times more likely to be killed than a heavy vehicle driver in a one to one collision (Fig 3). The result emphasises the danger to passenger vehicles in a mix of heavy vehicles. The proposal to encourage or compel 5,000 heavy vehicles per day to use the NorthConnex toll tunnel, as well as passenger cars, is likely to exacerbate the incidence of road trauma. Assessment of road trauma incidence and costs required a cost benefit analysis, but this analysis was not undertaken as part of the of the NorthConnex project assessment.

Travel time claims

Claims of a 15-minute transit time for vehicles through the 9km NorthConnex tunnel route are also subject to scrutiny. The claim does not take into account the delays that occur to multi-combination vehicles at signaled intersections at the entrance and exit of a tunnel that can cause queues backing up into the tunnel and delays passing into the tunnel. A report by Ramsay and Brunker (2003) at the Queensland University of Technology, shows that MCVs have poorer acceleration capabilities which will lead to a lower acceleration for all vehicles in the traffic queue.⁽⁹⁾

These on road safety findings raise a number of serious implications for the operation of the NorthConnex tunnel:

• The failure to carry out an assessment of road safety means that no guidance is available to tunnel users as to how a tunnel accident can be avoided;
• The claimed travel time savings will not be achieved, which means that the trucking industry will be paying tolls for no gain while submitting to the risk of an accident;
• The risk of accidents to toll paying passenger car drivers from collisions with heavy vehicles is shown to be much greater than that for truck drivers.