Peer review: Andrew Evans
Imperial College London
Improved railway safety is unquestionably a good thing. Therefore, however highly-placed Great Britain may be in the European league tables, where there is scope for further safety improvements, safety regulators have a duty to call for those improvements to be made.
Or do they? What concerns me about Paul’s article, and about the ORR’s Health and Safety Report 2013/14 on which it draws, is the absence of discussion about the scale of safety benefits to be expected from the various safety measures mentioned, or of their costs. It is therefore not possible to form a view about what safety measures give good value for money, and what do not. Instead, the argument is that if the railways are less than excellent in some respects, they should do something about it.
A good example is the number of signals passed at danger (SPADs). RSSB data show that there were 293 SPADs in 2013/14, an increase of 17% on the 250 recorded in the previous year. That sounds like a far from excellent performance, which moreover is getting worse.
However, the latest version of the RSSB’s Safety Risk Model estimates the current fatality risk from train collisions following SPADs as 0.55 fatalities per year. Given that the average fatal train accident causes about four fatalities, that implies a fatal SPAD accident about once every seven years.
In fact, we have done rather better than that. There has been no fatal SPAD since the Train Protection and Warning System (TPWS) provided a step-change in train protection in 2003. That implies that the gains from reducing SPADs are real, but limited.
If SPADs could be reduced by better driver management and training at no additional cost, then it would be a ‘no-brainer’, and the ORR would be right to press for it. However, if additional resources were required, these would need to be compared with the safety improvement.
Paul reports that the regulator is pressing for upgrades to TPWS equipment pending the rollout of ERTMS, which will not be widespread for many years. This appears to be a classic case where the costs need to be compared with the benefits - and given the data, it would be straightforward to carry out.
As for ERTMS itself, RSSB says its expenditure is “far in excess of the financial value of any safety benefit, and therefore it is accepted that the safety benefit should not drive the decision-making process”. The decision was driven by the non-safety business case.
To finish with a different point, the scope of rail safety reporting has been gradually increasing. A notable recent addition is the inclusion of road accidents and casualties to staff in the course of work. This is welcome, and reflects the acceptance by the industry of the need to manage road use.
However, it means that trends in casualties need careful interpretation. A series may rise not because of any real change, but simply because its scope has been widened.
