David Shirres reports
Two types of railway safety
At Quintishill on 22nd May 1915, signalman George Meakin forgot the location of a local train and accepted a troop train that ploughed into the back of the first train. A third train then ran into the wreckage which caught fire killing 227 people. It was Britain’s worst rail disaster. Since then the introduction of track circuits and signal interlocking prevent such accidents.
In 2011, track circuits are rarely used to reduce the risk of human error associated with work on track. Although the history of rail operational safety is one of the progressive introduction of engineered safeguards, few are used to protect work on track which is almost entirely dependent on track workers following rules and not making mistakes.
How much of an issue
Since 1990, 52 track workers have been killed due to a failure of the Safe System of Work.
Figure 1 shows the awful number of fatalities in the early nineties, with a significant reduction since the mid nineties.
This may be due to improved discipline, including a strict Drugs and Alcohol Policy together with hard hitting safety campaigns. Since 1994 there have been 25 fatalities. Figure 2 is an analysis by Safe System of Work which shows that:
- 32% (8 of the 25) fatalities were those providing the SSoW
- When used, lookouts account for more than half the fatalities. i.e. 5 fatalities out of 9 occasions (6 static red zone plus 3 of the mobile zones requiring lookouts)
- 28% of the fatalities (7 out of the 25) were in mobile red zones, of which 16% (4 of the 25) were working alone in a mobile site.
Data in the RSSB risk model shows that the probability of an individual track worker being killed by a train is 1 in 15850 compared with the HSE’s unacceptable level of 1 in 1,000. This RSSB figure is an average as illustrated by mobile red zones accounting for 28% of fatalities for what is a tiny amount of work. If this proportion is less than 1.75%, mobile red zones present an unacceptable risk and should be banned. It would seem this high level of risk is not recognised.
As setting up protection takes a fair proportion of available engineering access time, significant inefficiencies occur if risk controls are not optimised for work practicalities. The best use of possessions often involves many activities within a worksite. Although this can be safely managed with suitable controls, the Rule Book does not adequately address this. Perhaps this is one of the reasons that the ORR consider Network Rail’s infrastructure work is 25 to 50% less efficient than in Europe and that in part this is due to the way possessions are undertaken (1). Engineered safeguards for work on the open line would also give cost savings as well as risk reduction.
To err is only human
Network Rail, the employer of all who work on track, has a legal duty to determine risk controls by risk assessment. Yet track work rules for this are produced by a 19-member RSSB committee primarily concerned with train movements that has only 3 Network Rail members. With this lack of a cohesive approach, some tasks have inadequate risk controls, safety arrangements do not always take account of the practicalities of the work and the principles of risk assessment are not followed. For example, Rule Book Handbook 8 specifies that protection to block a line is required “if the work will affect the safety of the line” whereas the legal requirement is to protect the line if it is reasonably practicable to do so.
Any risk assessment must consider human error. For this reason, in 1995 an HSE publication (2) recommended the use of on-site backup protection for line blockages. Except for signal disconnection, the TCOD is the only such protection. However, its use is limited due to type approval issues, lack of clarity of where it can be used and signallers not being involved. Thus risk controls require an effective implementation strategy. Other engineered safeguards not implemented are Automatic Track Warning Systems and LOWS including a low cost, easy to use Train Alert Handswitch favourably reviewed by a 2006 RSSB report (3).
Risk assessment of the impact on maintenance by project designers (i.e. anyone whose decision changes a project, including project sponsors) is a requirement of the CDM Regulations. Thus rail infrastructure projects should consider trackside facilities such as access points, safe cess and track warning systems which can often be provided at a minimum cost if specified at outline design. Constructing new assets without access points or creation of bi-directional lines without track worker protection systems would seem to breach the CDM Regulations.
What about culture?
Few would disagree that safety culture is the most important factor in accident prevention and about the need for safety culture initiatives. It is right that Network Rail considers safety to be everyone’s responsibility and the line management is directly accountable for it. What is not clear is who is accountable for safety systems. However good the culture, human beings will make mistakes. Effective action to reduce injuries and fatalities requires an understanding of the reasons for human failure. Figure 3 shows this is often not due to poor safety culture.
The way forward
Although many competent and committed individuals are involved, there has been little change in systems of work. This is certainly no reflection on those concerned, but is due to cultural and organisational factors that need to be addressed if there are to be significant improvements.
Network Rail’s culture made national news with RSSB’s report on accident under-reporting identifying a “culture of fear”. This is also a culture that discouraged identification of problems. Although Network Rail’s new Chief Executive is committed to changing this culture, it may take time. The industry’s strong culture of rules compliance is a great strength but can lead to a blame culture with individuals held solely responsible for rule breaking. This does not allow for genuine mistakes or poor working environments and so masks the need to improve risk controls.
Classifying track safety rules as Group Standards gives a disjointed approach with track safety risk controls split between Network Rail and RSSB. This also requires extensive consultation for rules changes which can take up to 2 years. Network Rail should take control of the risks for which it is legally responsible to provide a cohesive approach including a strategy for track safety assets and systems. To make this happen, Director level commitment is required. With many issues facing the industry, the risks and inefficiencies associated with poor track safety risk controls needs to be recognised as a priority. This requires suitable performance indicators.
(1) ORR – International cost efficiency benchmarking of Network Rail – September 2010
(2) Railway Safety: the prevention of risk to workers on the track” published by HSE in 1995
(3) Technology to support practical decision-making and competence assessment (T323) – Final Report Volume 3 – Addendum published by RSSB October 2006
Join David and other great Safety Speakers at the Rail Safety Summit on May 18th 2011.
