Collision between two trains at Plymouth Station - 3 April 2016 - merged topic

[whole diary]

[ChrisB]

................worth bearing in mind the next time the standard GWR^▸ (Great Western Railway) line is trotted out about overcrowded trains with multiple standees not presenting a risk to health and safety...........

Not really....it's the crash that is always the risk to H&S^▸ (Health and Safety). Otherwise there is no risk

[SandTEngineer]

I'm free to attribute blame to many, one, none. Individuals or entities. Based on the evidence presented thus far. That attribution of blame may differ from that of the BTP^▸ (British Transport Police), ORR^» (Office of Rail and Road formerly Office of Rail Regulation - about) and the employers of the staff involved. I suspect it won't in this case though, unless the sequence of events is vastly different to that laid out in the RAIB^▸ (Rail Accident Investigation Branch) Bulletin.

I wasn't remotely suggesting that you are not entitled to your opinion (we all are).  I just wanted to understand how you felt you had reached a conclusion before all of the facts come to light.  No offence intended.

[Chris from Nailsea]

May I thank you all, gentlemen, for the polite manner of your discussion here. 

[TonyK]

The line I often trot out is that there is seldom a single cause to a transport accident, usually a combination of at least three factors. In this case the RAIB^▸ (Rail Accident Investigation Branch) news release identifies as facts the unusual working of the platform, the misjudgment of the signalman in estimating the room available for the train to stop behind the HST^▸ (High Speed Train), the apparent confusion of the driver when faced with an unfamiliar situation, and standing pax at the time of the collision. The RAIB's job is to pick a path through all of that and find what should not be done again.

[chrisr_75]

Can also be rather effectively explained with the Swiss cheese model:

https://en.m.wikipedia.org/wiki/Swiss_cheese_model

[Chris from Nailsea]

Hence, "Emmental, my dear Watson!" said Sherlock Holmes. 

[ChrisB]

The RAIB^▸ (Rail Accident Investigation Branch) has released updated summary of this accident - seems signaller & driver of local train at fault

At 15:34 hrs on Sunday 3 April 2016, a train entering platform 6 at Plymouth station, collided with a stationary train. Around 42 people, including the drivers of both trains, were injured; one passenger seriously.

The stationary train, 1A91, had arrived in platform 6 around 34 minutes before the collision. This train would normally have used platform 7. On this occasion it could not because the platform lifts were closed for planned maintenance and the train required its catering supplies to be restocked before its onward journey to London Paddington at 15:41 hrs. At the time of the collision, train 1A91 had two members of staff on board.

Train 2E68 was the 13:39 hrs service from Penzance to Exeter. It was formed of two class 150 diesel multiple units coupled together; a total of four vehicles. There were approximately 76 passengers and 4 members of staff on the train as it approached Plymouth. This train was booked into platform 8 at Plymouth but the signaller decided to signal the train into platform 6 behind train 1A91 so that passengers intending to catch train 1A91 could do so easily, as the lifts were not working on both platforms 7 and 8. Allowing trains to share a platform is known as permissive working and was allowed for passenger trains using platform 6. The signaller could see the rear of train 1A91 from the signal box window and estimated that there was enough room for train 2E68 to fit on the platform behind train 1A91, however, there was insufficient room.

Travelling towards Plymouth station from the west, the data recorder fitted to train 2E68 showed the train slowed to about 8 mph (14 km/h) as it approached a red (stop) aspect displayed by signal P15 which is located around 330 metres from where the rear of train 1A91 stood. Two white lights and a numeral were then displayed adjacent to this red aspect (which remained lit) as an indication that the train was permitted to proceed to platform 6 with the driver prepared to stop short of any train, vehicle or obstruction. There is no evidence of any signalling malfunction.

The train then passed signal P15 and accelerated to a speed of around 21 mph (34 km/h). As it approached Plymouth station through a relatively tight, left-hand curve, the driver saw train 1A91 in platform 6 ahead. The driver was surprised to see the rear of train 1A91 so close to the west end of the platform. He looked at the tracks to confirm which route the train was taking and, realising a collision was imminent, applied the emergency brake. This was around three seconds before the collision, which occurred at about 15 mph (24 km/h). Many passengers were standing in preparation to leave the train and were thrown into the train’s fixtures and onto the floor.

Our investigation will include examination of:
•the actions of staff involved
•the signalling and platform working arrangements at Plymouth station
•the performance of the train during the collision
•any underlying management factors

[Chris from Nailsea]

Most of the wording is unchanged from the version issued by the RAIB^▸ (Rail Accident Investigation Branch) in April:

18 July 2016 9:59am
Text updated around passenger/injury numbers and to clarify the meaning around the signal aspect.

[Chris from Nailsea]

RAIB^▸ (Rail Accident Investigation Branch) has today released its report into a collision at Plymouth station, 3 April 2016.

Image showing trains after the accident: 2E68 on the left (courtesy BTP^▸ (British Transport Police))

Summary

At 15:34 hrs on Sunday 3 April 2016, the 13:39 hrs passenger train service from Penzance to Exeter collided with an empty train which was already waiting in platform 6 at Plymouth station. The collision occurred at a speed of about 15 mph (24 km/h) and resulted in injuries to 48 people and damage to both trains.

The signaller intended that both trains should share the platform because the empty train was to form a service to London and some passengers from the Penzance service were expected to join it. Lift refurbishment work meant that without platform sharing, passengers would have needed to use the stairs and a subway when changing trains. Permissive signalling arrangements were in place at Plymouth to permit two trains to share the same platform.

The signaller misjudged the amount of space available behind the London train and wrongly believed there was room for the Penzance train. He was aware that the platform sharing arrangement required an unusual form of permissive working, but did not communicate this to the Penzance train driver, and the rules did not require him to do so.

The Penzance train driver incorrectly believed he would not be sharing a platform with the London train. There was insufficient distance to stop his train by the time he realised his mistake and had applied the emergency brake.

Great Western Railway, the operator of both trains, and Network Rail the owner of the infrastructure, had not identified the risk of a collision due to the combination of an unusual form of permissive working, the track alignment on the approach to Plymouth station, and an inexperienced driver.

Recommendations

The RAIB has made three recommendations. The first, addressed to Great Western Railway and possibly also relevant to other train operators, seeks improvements to the training and assessment of new drivers. The second, also addressed to Great Western Railway and possibly relevant to other train operators, arises from difficulties encountered during passenger evacuation and seeks improvements to emergency door release controls. The third recommendation, addressed to Network Rail and to be undertaken with the assistance of appropriate train operating companies, seeks a review of permissive working arrangements at stations.

Two learning points stress the care needed by drivers when undertaking permissive moves, and the value of preventing passengers boarding or alighting from trains when permissive movements are taking place in the same platform.

Simon French, Chief Inspector of Rail Accidents said:
This collision caused great distress to the large numbers of passengers and staff involved, particularly those who suffered injuries. It occurred when a train was routed into a platform that was already occupied by a stationary high speed train. This form of train working, known as permissive platform working is not unusual on the UK^▸ (United Kingdom)’s busy network and is usually performed safely. Platform sharing allows trains to be joined together or, as was intended in this case, can facilitate the easy movement of passengers between connecting train services.

This accident reinforces the need for drivers to take great care when signalled into an occupied platform - assumptions should never be made about the length of platform that is unoccupied. This learning applies to all drivers but is particularly applicable to those who are inexperienced or new to a route. For this reason we have today issued a recommendation to Great Western Railway concerning the training and assessment of new drivers to better prepare them for permissive platform working. I am also urging other train operators to think about how well they prepare their drivers for similar circumstances.

Although the RAIB recognises the need for permissive working in station platforms, we have recommended that Network Rail, in conjunction with train operators, carries out a review of the way it is implemented at all stations where permissive platform working is currently authorised. This should include an assessment of a range of risk factors, including the information provided to the signallers when deciding whether or not to route a train into an occupied platform.

[stuving]

29  Many people who were standing up waiting to get off train 2E68 were thrown into
each other, into the train’s fixtures, and onto the floor. The collision resulted in
injuries to 42 passengers and 4 members of staff on train 2E68, including one
passenger who was seriously injured...

This was a collision at about 15 mi/hr, and I imagine that most of those injuries would have been prevented, or at least much less severe, if people were warned and could brace themselves. But I don't think I have ever seen it suggested that drivers make the equivalent of the "brace brace" warning announcement on aircraft. Why not? I have a very hazy recollection that it may have happened, sometimes.

Obviously there would be some conditions - such as the driver having use of the PA^▸ (Public Address). But in such low-speed crashes there should always be a few seconds after the driver knows it's inevitable. All you'd need is the right form of words.

Assuming that word "collision" is best avoided, perhaps "brace yourselves for a hard stop" or "... for a bump" would do? It would not happen very often, so public familiarity with a standard wording would be pretty low. It wouldn't be possible to rely on it being a trigger with a well-know response.

[ChrisB]

Publicity needed re remaining seated until the train has come to a stop/stand, maybe?

[eightf48544]

I am surprised that no comment was made about the position of the signal 330 mts from the platform and the platform out of site after passing it.

Presumably the driver naturally assumed that there was enough space on the platform and drove accordingly so when the HST^▸ (High Speed Train) came in view it was too late to stop.


Perhaps the signal should have nearer in the platform so the driver of the approaching train could see what was occupying the platform.

[JayMac]

Wherever the signal is placed nothing changes the fact that a driver, when given a permissive 'calling on' signal, is expected to drive such that he can stop short of any obstruction.

[Chris from Nailsea]

That RAIB^▸ (Rail Accident Investigation Branch) Report says this:

15 Train movements approaching Plymouth station from the west (the route followed by train 2E68) are controlled by signal P15. This signal is located 270 metres from the western end of platform 6. It moved to its current position in the late 1980s and the signal’s head (displaying the coloured lights) was replaced in 2007 with a light emitting diode type signal head (known as LED). Signal P15 is capable of signalling trains into any of the platforms at Plymouth. When displaying a proceed indication, the appropriate platform number is displayed adjacent to the signal head (figure 5).

16 If a platform is already occupied, another train can be signalled into the same platform under permissive working arrangements (paragraph 32). In these circumstances signal P15 will display two white lights (known as position lights), at an angle of 45° adjacent to the red aspect.

17 There is no evidence that the functioning of signal P15 was a factor in the accident. Although the signal was re-positioned before the current Railway Group standard was introduced, the distance of 270 metres between the signal and the platform is compatible with Railway Group standard GK/RT0044 ‘Controls for signalling a train onto an occupied line’. The December 2014 version of this standard required the distance to be ‘minimised, as far as reasonably practicable’, and the previous version (February 2000) required that the spacing ‘shall not be greater than 400 metres’.