The Precursor Indicator Model measures the underlying risk from train accidents by tracking changes in accident precursors, and is calibrated against the
Safety Risk Model.
Train accidents are rare. The rate of train accidents that resulted in passenger or workforce fatalities has been less than one per year over the last twenty years. Because they are so rare, we can understand the risk properly only by looking at the underlying things that might cause them. These precursors can indicate the risk of accidents happening, even though they do not often result in an actual accident. RSSB has developed the Precursor Indicator Model (PIM) to quantify changes in that underlying risk over recent years.
To compare the harm arising from various incidents, we weight the outcome according to how severe they are. For example, we count one fatality as having the same statistical weight as ten major injuries, which means that we can make a meaningful comparison between different outcomes. Within the overall harm that rail passengers have suffered (expressed as these Fatalities and Weighted Injuries – FWI), the amount from train accidents is very small.
The PIM provides a day by day review of the main elements of train accident risk, which includes the risk of collisions, derailments and fires. It is scaled to take account of the increasing exposure due to increasing rail traffic.
We measure the frequencies of operational occurrences which have the potential to cause a train accident. These occurrences are called precursors in the model.
Each precursor has its frequency measured and is then weighted to place its relative contribution correctly within the total train accident risk. This involves referring to the Safety Risk Model (SRM), which is updated every few years in a major reassessment of the industry's risks but, in the interim, the PIM provides a measure of how large an effect each precursor is having on train accident risk.
Much of the train accident risk is to members of the public – that is, people who are neither passengers nor employees. This is particularly so in relation to accidents at level crossings. Some collisions between trains and road vehicles derail the train and can lead to many passengers being harmed, but almost all will lead to the road vehicle’s occupants suffering a very severe outcome.
To illustrate clearly the part of the risk that is borne by passengers, the PIM has been divided by person type. The "Passenger PIM" is presented as a subsection of the overall PIM, showing the precursor groups as they relate just to passengers. The overall shape of the chart is visible, showing the overall risk trend, but only the risk to passengers is broken down into the main precursor groups.
The chart above (click to enlarge) shows the model’s risk estimate, daily since April 2010, with coloured areas for the risk to passengers from each of the main precursor groups. The wider the coloured strip, the higher the risk estimate at any point. The pale and dark sections stacked on top then add estimates of the risk to the public and to the workforce, so that the total may easily be seen.
Results from the PIM can be found in publications such as the
2015/16 Annual Safety Performance Report.
For more information on the outputs of the PIM, please contact the