Working to a safer railway
Accident investigation is about prevention, not blame
During 2015, some 296 accident notifications were advised to the Rail Accident Investigation Branch (RAIB), an organisation whose primary task is to investigate the causes of accidents and incidents where the findings will bring safety learning to the industry, in order to prevent a similar accident happening again or removing conditions where one might occur.
The RAIB was established in 2005, following criticism of technical deficiencies in the previous arrangements for accident investigation by the Health and Safety Executive (HSE). These deficiencies could be caused by staff refusing to co-operate with a prosecuting authority, for fear of incriminating themselves.
The RAIB shares expertise with other modes - a Board of Transport Accident Investigators has been established that consists of the three Chief Inspectors of accident investigation (Rail, Marine and Air), and which identifies common strategic issues to improve outcomes.
The focus is to conduct impartial investigations that are well communicated within the industry and which result in improvement action, covering UK main line railways, light rail operations and heritage railways.
An issue has emerged (and here RAIB acknowledges the need for improvement) with the length of time taken to publish findings - in one case last year it took 19 months, and the average for the 20 formal investigations undertaken was 11 months. To counter this, a procedure to distribute Urgent Safety Advice has been established, whereby preliminary findings that reveal shortcomings are briefed out to the industry.
Crucially, it is not the purpose to find evidence to support a prosecution, or to apportion blame or liability. In a previous era this was the role of the Railway Inspectorate, but the flawed decision to incorporate this activity within the HSE changed the focus from finding out what had gone wrong to searching for someone to blame.
The pattern of accidents that has emerged in the recent past is dominated by events at stations and level crossings, although the number of freight train incidents and track worker fatalities or near misses remains significant.
With experience now held of the responses to incidents where improved working practices are identified, there is evidence that changes are not made with urgency. This suggests that a more robust system for updating the Rule Book, making staff aware of changes, is required.
One example is incidents involving empty shipping containers falling from wagons. On March 1 2008 at Cheddington, two containers were blown off a train operating from Isle of Grain to Doncaster while it was travelling at 75mph. The detached containers blocked the running lines and caused damage to the track and to overhead line equipment (OLE).
On the same day five empty containers were blown off a Tilbury to Coatbridge service at Hardendale ( between Tebay and Penrith). The train was again running at 75mph, and the detached containers blocked running lines and caused infrastructure damage.
A common feature was the use of FEA container flats, where the spigots that keep the containers in place are not compliant with UIC standards. This was established as the cause when the incidents were investigated.
The subsequent report recommended that freight operating companies should review the action they take when conveying lightweight or empty containers in windy conditions, and that where wagons are fitted with non-compliant UIC spigots measures should be developed and implemented to retain the containers.
No solution was implemented, however, and the same thing happened again twice last year. In the early hours of March 7 an empty 30ft container became detached at Scout Green on the West Coast Main Line. It passed over the adjacent track and came to rest down an embankment, about four minutes after a Sleeper train had passed the site in the opposite direction.
In a separate incident on March 31, an empty 40ft container was displaced near Deeping St Nicholas (Lincolnshire). The container was dragged a short distance by the train, causing extensive damage, before coming to rest on the adjacent track where a freight train had passed five minutes earlier.
Again, the detachments were caused by strong winds combined with the speed of the trains, and inadequate container retention. It was revealed that work to implement the 2008 accident recommendations had not yet been completed. Implementation would have prevented the accidents, both of which had the potential for disastrous consequences.
FEA wagons with spigots that do not fully comply with standards on container retention are finally being modified, with the aim of preventing containers being blown off in windy conditions.
I’m devoting space to the issues that continue to occur with the platform train interface, as this is relevant to a current industrial dispute at Southern about the removal of guards.
Regular incidents have occurred where passengers and their clothing become trapped in closing doors - as the train departs, this results in the passenger being dragged along the platform, with the risk of falling under the train.
As a result of work undertaken by the RAIB there is greater understanding about the design of power doors where there is a lack of sensitivity to closure (where items have been trapped, but drivers are still able to depart). It is usefully pointed out that people used to using lifts would expect the doors to re-open if full closure is prevented by an obstruction.
Many London-area train services now operate as 12- car formations, and given door design, leaving it to the driver to ensure it is safe to depart is unrealistic - particularly where platforms are curved or crowded with people.
It does not have to be a guard to perform this function, and there is no reason why the driver cannot operate the doors, but at many locations there is a need for train despatch staff to provide a positive indication that it is safe to leave the platform.
The RAIB is right to call for a systematic assessment of the dispatch risk at each platform, and for improvement to the design of train door obstruction detection.
“It is not the purpose to find evidence to support a prosecution, or to apportion blame or liability.”