Safety at level crossings - Transport Committee Contents

Annex: Level crossing and road risk compared

Note by Professor Andrew Evans, Specialist Advisor

1. In order to place the risk at level crossings in perspective, it is desirable to find a means of comparing LC risks with road risk.

2. One way of doing this would be first to estimate the road safety fatality risk of a typical car trip or walking trip without a level crossing, to form a baseline. Then it could be supposed that the journey required the traverse of one level crossing, and the additional risk imposed by the level crossing could be compared with the baseline risk.

3. This can be done for both typical car journeys and typical walk journeys, though various simplifications have to be made. Table 1 gives the data and calculations. The column headed "source" gives either the external source of the data or a calculation from other figures in the table. The alphanumeric sources such as NTS0409 are references to DfT statistical tables (NTS = National Travel Survey).

4. The bottom line of the table gives the additional risk from the presence of one level crossing on an average car or walk journey as a percent of the baseline road risk. For car journeys, the additional risk from the level crossing is estimated at 7.6% of the baseline road risk, which is modest. For walk journeys, the additional risk from the presence of a level crossing is much larger. It is estimated to be 114% of the baseline road risk, implying that the LC roughly doubles the risk of the journey.

5. This in turn is because the risk per level crossing traverse is much higher for pedestrians than for vehicle occupants. Based on LC fatalities over the decade to March 2013, row (k) of Table 1 estimates that there were 1.35 fatalities per billion LC traverses for vehicle occupants and 30.1 fatalities per billion LC traverses for pedestrians, which was about 22 times greater. Pedestrians make about one eighth of the number of LC traverses as vehicle occupants, but they have just under three times the number of fatalities. However, it is notable that the road fatality risk per kilometre of travel is also about 20 times greater for pedestrians than vehicle occupants - see row (d) of Table 1.

6. In conclusion, if a typical car journey includes a level crossing, the crossing imposes an additional fatality risk estimated at about 7.6% of the baseline road risk. If a typical walk journey includes a level crossing, the crossing imposes an additional fatality risk estimated at about 114% of the baseline road risk. However, because a typical walk journey is much shorter than a typical car journey, it is less likely to include a level crossing. The fatality risk per LC traverse is estimated to be 22 times greater for pedestrians than for vehicle occupants. This ratio is roughly in line with the road fatality risk per kilometre for pedestrians relative to vehicle occupants.

Table 1: Journeys and fatalities with car and walk as main mode (2012)

 Source Car Walk Road journeys (a) Trips/person/year with given main mode NTS0409 614 212 (b) Km/person/year with given main mode NTS0410 8359 243 (c) Average length of trip with given main mode (km) (b)/(a) 13.61 1.146 Road fatalities (d) Road fatalities per billion person-km RAS53001 1.3 23 (e) Road fatalities per billion trips (c)*(d) 17.70 26.36 (f) Road fatalities per single trip (e)*1E-09 1.77E-08 2.64E-08 Level crossing traverses (g) LC vehicle or pedestrian traverses per year NR LC database 12.38E+08 2.43E+08 (h) Average car/van occupancy NTS0905 1.56 (i) LC person-traverses per year. (g)*(h) 19.31E+08 2.43E+08 Level crossing fatalities (j) LC Fatalities per year 2003/04-2012/13j ASPR Chart 187 2.6 7.3 (k) LC fatalities per person-traverse (j)/(i) 1.35E-09 3.01E-08 LC /road fatalities comparison