A MAJOR report published by Britain's Rail Safety & Standards Board on July 31 draws together the findings of several research programmes extending over five years into the case for using seat belts on trains, and the apparent conflict that exists between the risk of passengers being ejected through windows and the desire to use them as an escape route after an accident.

The most important conclusion is that the three issues of seat belts, the strength of window glazing, and escape or access for rescue must be looked at together if the optimum policy is to be adopted. The conclusions that have emerged are clear-cut, and RSSB feels that seat belts on trains would actually increase the risk of injury in a collision. In addition, windows should not be identified or used as escape routes and laminated window glass would save lives but not impede rescue.

The report distils the results of three research projects initiated by RSSB, with much of the work carried out by DeltaRail and its predecessor, AEA Technology Rail. The first project was a response to a derailment at Potters Bar in 2002 where, not for the first time, there were strident calls for seat belts to prevent passengers from being hurled around inside carriages during an accident and ejected through windows.

This project, completed in 2005, looked at lap belts as used in aircraft which are anchored at two points. The conclusion was that they actually increase the risk of injury, and are incompatible with crashworthy seats now being installed.

The findings of the two other projects were published by RSSB on July 31 together with the overall report. One examined the case for three-point seat belts as used in cars, while the other considered the use of laminated as opposed to toughened glass in carriage windows.

Much of the data was obtained from detailed study of injuries sustained during seven UK accidents over the last decade. At Watford Junction (1996), Southall (1997), Ladbroke Grove (1999) and Great Heck (2001) the injuries resulted from collisions between trains. At Hatfield (2000), Potters Bar (2002) and Ufton (2004), passengers were injured as carriages derailed at speed.

Significantly, Ladbroke Grove was the only accident where fire was involved, creating a genuine need to escape quickly. This is very much the exception, unlike aviation safety measures where the aim is usually to evacuate passengers from aircraft as quickly as possible.

The researchers found, however, that images of vehicles damaged at Ladbroke Grove and fears of a train fire appear to have prompted passengers at Great Heck, Potters Bar and Ufton to climb out through windows when there was no need to do so. In total darkness, 27% of the passengers involved in the Ufton derailment climbed out through the windows 'although in every case there were alternative and safer means of exit', according to RSSB.

Why seat belts are unsafe

Both of the research projects into seat belts anchored at two or three points used standard crash test techniques at the Transport Research Laboratory’s test facility. Trials with dummies took place with face-to-back ‘crashworthy’ seating currently used on trains, which is designed to deform when struck by the knees or head of the passenger seated behind it.

In simple terms, the observation in the case of two-point lap belts was that the rear passenger’s body rotated around the hips so that the head struck the back of the seat in front in a way that could break the neck.

When three-point belts were tried out, it was realised that the crashworthy seats were not strong enough to provide anchorage for the diagonal belt. The result was that a choice had to be made: belts or crashworthy seats?

In identical longitudinal collisions, the injury levels of unrestrained passengers in crashworthy seats were well within acceptable limits, but if the seat backs were rigid the situation for unrestrained passengers was much worse. And it would be quite impractical to enforce the wearing of seat belts in trains where standing is routine.

The other issue is whether seat belts would have prevented passengers being ejected through windows. Six recent accidents were analysed and the conclusion was that although 11 lives of ejected passengers might have been saved by belts, they would have caused 88 lives to be lost by keeping some passengers in ‘lost survival space’. It seems that many passengers who are thrown around in a carriage thereby escape being crushed.

Laminated glass saves lives

If the objective is to stop passengers being thrown out, and also to stop debris from an accident coming in, there is no question that laminated glass in which a sheet of polyvinyl butyl is sandwiched between two panes of glass is the optimum choice for carriage windows. Like a metal sheet, it deforms under impact but does not shatter.

The move to use laminated glass on UK rolling stock was originally triggered by the accident at Polmont in 1984 where a train of Mk II coaches being pushed by a locomotive derailed after hitting a cow on the line. The leading coach rotated violently and was thrown up the side of a cutting killing 13 passengers, some ejected through the toughened glass windows favoured at that time.

The eventual result was Railway Group Standard GM/TT0122, which from 1993 required all windows in new carriages that were not designated for escape to be laminated. However, all vehicles involved in the seven fatal accidents between 1996 and 2004 which formed the basis of the RSSB studies pre-dated this standard, and 12 of the 20 passengers ejected from the train in these accidents suffered fatal injuries.

A review of major accidents since 1973, undertaken specially for this latest study, confirmed that fatal injuries sustained as a result of being ejected through a train window are a comparatively modern phenomenon. Prior to Polmont, most fatalities inside trains were the result of crushing as the bodies of Mk I coaches were penetrated, often by the underframes of adjacent vehicles.

The first major accident involving carriages built to the new standard with mostly laminated windows occurred at Grayrigg last February. Despite the fact that several cars left the track at speed, hitting obstructions such as catenary masts, there was remarkably only one fatality and not a single passenger was ejected through a train window.

The recommendations now are that 'all future new vehicles should have laminated glass (or equivalent) only', and that 'windows on existing vehicles should be considered for progressive replacement with laminated glass, particularly at refurbishment'.

The advice to train operators is that hammers to break windows in an emergency should be removed, and signs amended to direct passengers to external doors or gangways for evacuation. However, RSSB also feels that the principle to be established is that in the majority of accidents the safest place for passengers to be is inside the carriages until the emergency services arrive.

The report also lays to rest the idea that rescuers will not be able to get in through windows. It says 'tests by rescue services demonstrate that any extra time taken to gain access through laminated windows is insignificant when compared with the time to arrive on site'.