Historic and significant accidents Recent reports, news and views on rail accidents and safety Free email account Free eCards Newsletter Bulletin Board E-mail Danger Ahead! Search Danger Ahead!; RailServe and the Web Help
The premier source on the web for information about rail safety and accidents

Historic Accidents
Free Newsletter
Conference Diary
Book Shop

User name:

Get your FREE email account

Hidden Dangers
Hidden Dangers
by Stanley Hall

this book
*  * * 

Selby Disaster 28 February 2001
Push - Pull: the Hidden Dangers

R H State : Wednesday February 28, 2001
The tragedy of the collision which has occurred on the East Coast Mainline today will be investigated and the prime causes determined.However, the contribution to the severity of the accident by the presence of the locomotive at the rear of the GNER train instead of at the front where most locomotives are located may not be one of the most immediate factors which comes to light.

The operation of locomotives at the rear of passengers trains is not new. In steam days the coupling of a locomotive to an "Auto trailer" so that it could be driven from a driving compartment in the coach itself whilst only the fireman remained on the steam locomotive footplate was used throughout the rail network. There were operational advantages for short journeys which reduced the need for the locomotive to run round its train. Such trains were used on branch lines where between one and three "Auto trailer cars" would be employed. Essentially these were slow speed and the brakes used were automatic vacuum. This concept was know as "Push-Pull".

When steam was replaced by diesel and electric locomotives technology did not exist to permit the safe operation of locomotives remotely from remote passenger cabs. It was not until BR faced increasing pressure from road competition that "push-pull" operation was re-introduced fro the Edinburgh to Glasgow services in the 1970's. At first Type 2 locomotives were used one at each end and later a single class 47/7 attached to one end usually that facing Edinburgh. For this service a number of coaches were modified to have a driving compartment at one end and it was from this end that the driver drove the train in one direction (usually towards Glasgow) with no-one in the cab of the locomotive at the rear.

Pushing a rake of coaches is little different from pulling them. The speed and control over the train is the same and they are quite safe PROVIDING the train stays on the track. As an example of this, take a toy wooden train with no track on a table top. Hold the locomotive and pull the train along. The locomotive will go where it is pulled and the coaches will duly follow. Now put the locomotive at the rear and push the train with the locomotive. Without the guiding influence of the track the coaches go anywhere almost immediately.

The inherent weakness in the "push-pull" arrangement is therefore ready for exploitation should the front of the train be lifted from the track.

On the 10th July 1984 such an accident happened. The 17:30 Edinburgh to Glasgow express rounded a bend between Polmont and Falkirk to find a cow standing on the track. The train comprised of 8 coaches pushed by class 47/7 47707 with a driving trailer at the front. The driver immediately shut off power and braked but the 34ton driving cab hit the cow so hard that it was lifted off the track. With the heavy 125ton locomotive pushing in the rear the leading coaches spun round throwing the passengers violently from side to side and even through the windows. In the carnage 13 died and 17 were seriously injured.

The attitude of the train contrasts markedly with the derailment at Pershore on the 30th November 1984 just 4 months after Polmont. Here at train travelling at or about the same speed as that at Polmont was derailed by a broken fishplate. Again the locomotive was a class 47 this time class 47/4 47500. However, the locomotive was at the front and the train remained upright and in line and no deaths or serious injuries resulted.

The problems with the locomotive at the rear for the Polmont accident can be described by consideration of the dynamics of the train at the point of application of the brakes from the front. At the moment of throwing the controller to close and the brakes to emergency there is a delay in the transmission of the drop in brake pressure to the rear. In addition some locomotives including class 47's had "proportionality", that is there was designed small delay in application of the locomotive brake which was introduced to allow the train brakes to be fully applied first. This was to prevent shocks to passengers due to surging which would take place if the locomotive brakes were applied first. This was fine when the locomotive was at the front but in the rear increased the time for the full retardation of the locomotive brakes to come into effect.

The combined effect of these small delays caused the Polmont train to be violently pushed from the rear just as the front wheels were lifted off the track by the impact with the cow.

For later designs of Driving Trailers the weight was increased to 45tons. When the Mark 4 coaches and class 91 locomotives were built push-pull was again specified. These are the vehicles which were involved in the Sandy accident and in both Hatfield and Great Heck.

At Sandy on the 17th June 1998 a wheel broke on the bogie of the last passenger vehicle on a northbound GNER train. Behind this vehicle was the driving trailer and the train was being PULLED by a class 91. Although derailed the train stayed upright and in line and only 9 passengers complained of minor injuries.

It is very clear that had the wheel broken on the southbound leg then a completely different story would be told as the derailed car would have again been pushed from the rear and the whole train would have gone into catastrophic derailment mode. In this case fortune favoured the railway.

At Hatfield on the 17th October 2000 the complete failure of the rail derailed the complete train. This again was a mark 4 set hauled from the front by a class 91 locomotive. Despite being completely liberated from the track the front half of the train stayed upright and in line and if it had not been that the couplings parted between the front portion and the restaurant car it is likely that no deaths would have resulted.

So we come to Great Heck. Yet again we have the class 91 locomotive at the rear when the front of the train is liberated from the track by a collision with an automobile. Again the front of the train is lifted off the track. The subsequent path of the train has yet to be determined but at some point it moved sufficiently to the right to foul the down line on which a Freightliner hauled by nearly new class 66 locomotive 66521. In the ensuing collision the latter ended up on its left side.

The position of the coaches of the GNER train show obvious signs of being "squeezed" between the force of the 126ton class 91 in the rear and the impact of the 130ton class 66 in the front. Only the enquiry will determine the true facts but I for one will be amazed if "push-pull" does not become a major issue.

Join the discussion

Eriks Rail News
ERN Headlines
*  * * 
Related Items
Sandy: British high-speed train in 100mph derailment
17 June 1998

Polmont, Scotland 1986
[an error occurred while processing this directive]
Seven days . . .
. . . web focus on rail safety and accidents in the last week

Search currents

powered by FreeFind

Click Here!

This file last updated: Wednesday, 28-Feb-2001 21:35:34 EST
Copyright © David Fry 1999, 2000