Collision Plymouth 03/04/16

[83G/84D]

Deep inhale...




Totally agree. This is certainly something I can't quite grasp. This incident should not have happened. There was a series of unfortunate events that led to the collision but the Driver should have stopped short.



Not sure I agree with this part. I see this as more a series of assumptions that were incorrect. Route knowledge appears lacking as he had to count the platforms rather than knowing he was routes into 6 and the HST.

Training I'm not so sure on. I know the requirements that are needed (at least my TOC) and that there are many opportunities that require permissive working. He assumed the location of where he was expected to stop and drove accordingly. He aimed for a place that he assumed he was going to stop at rather than going at caution.

The only defence of the Driver I can think of is that it takes a few seconds to realise what is about to happen. When this happened to me at Victoria I came round the corner into the platform and the platform was occupied with a unit I wasn't expecting to be there (much longer than advised)

My mindset was :

'Are you forking kidding me..'
'seriously signaller wtf..'
'ahh crap this is gonna end badly..'
'better make an announcement'

Only because I was at caution did I manage to stop. Its very easy to doubt when your new and you are reliant on training and 'normal' workings. Its also very easy to say out of experience and I have no doubt that this Driver will change his approach to permissive working.

The illustration on page 12 shows what P15 signal was showing, white dots and a number 6. It is perfectly clear from that, what platform the driver is taking the train into.

Paragraph 33 says what is expected of the driver.

 

[SpacePhoenix]

Would the 150s involved had either ASDO or SDO installed at the time? When the doors are released does it open up all the doors automatically on the released side or is it down to passengers to press the local open door button for a released door?

 

[BestWestern]

Would the 150s involved had either ASDO or SDO installed at the time? When the doors are released does it open up all the doors automatically on the released side or is it down to passengers to press the local open door button for a released door?

There is no SDO. It is worth pointing out that, whilst the train being overlength would of course been an issue for the Driver, the matter of door release would not have been any responsibility of his at all; the Guard takes care of that. Having landed in too short a platform, the Guard could have used a 'local door', locked out the overhanging doors prior to a full release, or made appropriate announcements about the train being replatformed and kept them locked.

 

[Crossover]

I don't see how it is? You're right that he should have driven such that he could stop short of an obstruction. But equally his train should never have been signalled into platform which didn't have sufficient space for his train to even fit fully into it! Both elements are surely concerning? 1) That the driver's knowledge of the relevant part of the rule book wasn't good enough and 2) That the signaller signalled a move which couldn't be safely completed in the space available.

Surely we shouldn't just be relying on the driver's knowledge to keep the train safe in this situation? The signaller has a role to play to surely? Therefore both parts are very relevant?

As I understand it, permissive working means that when "called on" one doesn't presume anything and must be prepared to stop short of any obstruction which I guess could reasonably be a completely full platform, effectively driving on line of sight

I have seen similar happen at Huddersfield. In theory, there was plenty space for a train to come on to platform 4 from Deighton on top of another - until the timetable change a year or so back, in this case it was always a 158 onto a 158 for the particular service involved. However, the first driver hadn't taken his unit far enough down the platform (I had thought that as I watched the second one approach, but presumed the involved would know what they were doing, as we often told here) so the second one came on top and stopped short, leaving half the train hanging off the platform until the whole lot was coupled and shunted about.

 

[ComUtoR]

The illustration on page 12 shows what P15 signal was showing, white dots and a number 6. It is perfectly clear from that, what platform the driver is taking the train into.

Paragraphs 67-69 states the Drivers actions when entering the station. He route knowledge was lacking and he did not believe he was going on top of the HST. Considering the unit was on P6, he was indicated for P6 and he was heading towards P6 the Driver still discounted that. The Driver had doubts, didn't believe he was wrong, didn't understand, etc. etc.

There is also a moment when something is going wrong that you doubt what you saw. Maybe he thought he misread the signal.

It's very easy to see in hindsight that the Driver had the indication and that it may be obvious to us but you still doubt it when it happens. He checked the point work and still discounted it. I think we can all see the issue with the Driver checking the point work.

The Driver lost focus for a split second and made an error. It happens and the Driver will learn from the mistake. Part of the reasons why the RAIB exist and investigate is so that we learn from it.

Paragraph 33 says what is expected of the driver.

I am fully aware of the Drivers duties and responsibilities.

 

[MichaelAMW]

From the opening paragraphs of the RAIB report into the Plymouth collision:

The purpose of a Rail Accident Investigation Branch (RAIB) investigation is to improve railway safety by preventing future railway accidents or by mitigating their consequences. It is not the purpose of such an investigation to establish blame or liability. Accordingly, it is inappropriate that RAIB reports should be used to assign fault or blame, or determine liability, since neither the investigation nor the reporting process has been undertaken for that purpose.


From page 18 of that report:

Identification of the immediate cause

34 The driver did not apply the brakes early enough to stop train 2E68 before it collided with the stationary train in the platform.


However...


From page 4 of the March 2017 ASLEF Journal, a copy of which I found lying on the table of a Newark - Matlock train today:

Signaller to blame for Plymouth crash

Investigators from the Rail Accident Investigation Branch have blamed signalling errors for a train crash at Plymouth station in April last year when a service from Penzance to Exeter collided with an empty train already waiting at platform 6.


(It would be wrong not to add that a causal factor in the report was the signaller's incorrect estimation of the space available in the platform for the second train.)

 

[Chris M]

Just because it is inappropriate to use RAIB reports to assign blame does not stop people doing just that.
Indeed the wording in the reports was strengthened a couple of years back, presumably to make it this clearer. For example from the 2010 Romford report:

1 The sole purpose of a Rail Accident Investigation Branch (RAIB) investigation is to prevent future accidents and incidents and improve railway safety.
2 The RAIB does not establish blame, liability or carry out prosecutions.

 

[Parallel]

It's been nearly a year and half of 150219 is still out of service too. Does anyone know when it will return?

 

[158747]

What a load of nonsense from ASLEF.
The signaller's actions were one link in the chain of events leading to the collision, but he was not to blame for the collision.

 

[bramling]

What a load of nonsense from ASLEF.
The signaller's actions were one link in the chain of events leading to the collision, but he was not to blame for the collision.

Agreed.

The signaller may have, albeit with the best of intentions, laid a trap. But the belt & braces nature of railway safety should have taken that in its stride, namely the driver should have been able to stop short of any obstruction. Whilst some lessons to learn all round, the signaller 's mistake was a performance and delay one, the driver's was the safety mistake which resulted in the collision. Evidently his training was deficient as he should not have been in that position.

 

[ComUtoR]

The signaller's actions were one link in the chain of events leading to the collision,

Was he just a link in the chain or the first domino that pushed the rest over ? If he hadn't cleared the road would the incident still have happened.

There was absolutely a chain of events to led to the incident. It doesn't absolve the Signaller or the Driver of their actions.

 

[Taunton]

I remember, in the old mechanical signalling days, that "calling on" signals were regularly used at Taunton, to bring a branch train from Minehead etc into the relief line platforms behind an express stopped there.

But it was done more sensibly. There were three up home signals in the half mile before the station, and the train would be progressively slowed by the signalman, on alert, pulling them off at the right moment. The last was immediately before the platform, from where the train ahead was visible, it being a straight run in. And there was never a collision, despite the limited outlook from a steam loco, vacuum brakes, or whatever. The signalbox directly overlooked the manoeuvre.

In contrast, at Plymouth this was all controlled by just one signal, well before the station throat. The train ahead was not visible from it. Furthermore the curved approaches would not give a good view of whether the HST was on the same track or the immediately adjacent one, one track to the left, especially as that was where it was timetabled to be, it having been changed on the day. I was disappointed that the report did not comment on the similarity of forward view of a train ahead standing on either track.

This whole approach of a being called on round sharp connections to a platform which you would never expect to already be fully occupied such that your own train will never fit is just a parallel to the recent Norwich accident.

So, the driver approaches, having had all the memos about ensuring punctuality in arrivals, knowing that the HST is timetabled on the adjacent track, is called on to a platform where his train will not fit, round a sharp connection with limited visibility, and when surprised by this is told "it's all your fault, mate". Such an approach to the operation would normally fail any meaningful risk assessment outright.

 

[Clip]

In contrast, at Plymouth this was all controlled by just one signal, well before the station throat. The train ahead was not visible from it. Furthermore the curved approaches would not give a good view of whether the HST was on the same track or the immediately adjacent one, one track to the left, especially as that was where it was timetabled to be, it having been changed on the day. I was disappointed that the report did not comment on the similarity of forward view of a train ahead standing on either track.

This whole approach of a being called on round sharp connections to a platform which you would never expect to already be fully occupied such that your own train will never fit is just a parallel to the recent Norwich accident.

So, the driver approaches, having had all the memos about ensuring punctuality in arrivals, knowing that the HST is timetabled on the adjacent track, is called on to a platform where his train will not fit, round a sharp connection with limited visibility, and when surprised by this is told "it's all your fault, mate". Such an approach to the operation would normally fail any meaningful risk assessment outright.

I was interested to read this part of your post and had a look on good old google maps
to understand better with what you were saying, its been many a decade since I was at Plymouth you see.

Now I dont know which one is platform 6 but from the map view their looks to be a HST at one of the platforms.

Also I was interested in how you claim there to be sharp curves approaching the station as when I zoomed out they didnt look that sharp to me and at one end the approach is nearly staright, but Ill bow to your better knowledge as a driver but Im still pretty sure they couldve seen a very long train thaat was in the platform they were going into? Of course ill bow to yoru better knowledge on the sighting from the cab here - im just making an observation and asking questions and by no means am i apportioning blame to anyone - just asking.

 

[edwin_m]

Was he just a link in the chain or the first domino that pushed the rest over ? If he hadn't cleared the road would the incident still have happened.

There was absolutely a chain of events to led to the incident. It doesn't absolve the Signaller or the Driver of their actions.

There's nearly always a chain of events, and if any of them had been different the accident wouldn't have happened. Do you blame the people who decided to refurbish the lift, which was the reason for the re-platforming so without it the accident would not have happened?

However according to the report the driver had a fundamental misunderstanding of the nature of permissive platform working - due to poor training rather than any fault of the driver himself. I can't help thinking that if this accident hadn't happened something similar would have happened before too long, either to this driver or to any others who came out of their training with similarly faulty understanding. An accident waiting to happen in fact.

If they want to have a pop at someone, ASLEF ought to take a close look at GWR's training programme.

 

[ComUtoR]

There's nearly always a chain of events, and if any of them had been different the accident wouldn't have happened. Do you blame the people who decided to refurbish the lift, which was the reason for the re-platforming so without it the accident would not have happened?.

Root cause is always looked for and lessons learned. That is the point of investigation. The refurbishment of the life is a contributory factor just as much as everything else. Maybe in the future it would be better timed or arrangements put in place before last minute decisions are made. There is still a lesson to be learned.

Hopefully people will put more robust procedures in place all round.

 

[MarkyT]

So, the driver approaches, having had all the memos about ensuring punctuality in arrivals, knowing that the HST is timetabled on the adjacent track, is called on to a platform where his train will not fit, round a sharp connection with limited visibility, and when surprised by this is told "it's all your fault, mate". Such an approach to the operation would normally fail any meaningful risk assessment outright.

I appreciate your point, and there are many contributory factors identified in the report, but the driver was also relatively inexperienced, and it is claimed not FULLY trained and fluent in the meaning and practice of permissive movements. It's very easy to fall into the trap of believing that the PL aspect means you've clearance to a preconceived position when it very specifically does not. I hasten to add that the lack of experience, training etc in the particular scenario was not the driver's fault either. A more experienced driver or a signaller who had spoken to the driver before the movement would probably have avoided the incident altogether or at least reduced its severity.

Note the signalling layout at Plymouth dates from the cost cutting 1970s, but is probably not that much different to what might be provided there if resignalled today. In response to changes in standards there was a flurry of schemes in the early noughties when concerns about permissive working added additional signals very close to platforms to reduce the length of permissive movements, but the additional main signals were very expensive and often had to be positioned right in the middle of throat junctions (think Ashford International, Portsmouth Harbour). Those layouts are a nightmare to operate, especially if a train stops in the middle of the throat waiting for the PL then can't get going again. The current standard is more pragmatic I believe, not requiring the whole move to be length limited, but requiring any potential obstruction to be visible on approach with sufficient warning to be able to stop. The problem I have is there is no upper speed limit on a permissive movement defined in the rule book so how can the 'sufficient warning' be determined. I contend that PL moves should be limited to 15MPH max. That would encourage designers to place home signals closer in to limit the length and time taken for the permissive movements, but not require the counterproductive expensive additional signals within the throat junctions.

 

[Taunton]

We have expensively installed cab radio across the system.

Then someone decided that it is not to be used when under way.

If you look at aviation, pilots, even those flying aircraft single pilot, with a notably more complex set of conditions to handle, are routinely instructed, whether in the air or on the ground, about all events going on around them by the controller. Just think of how aviation would have handled this event. They would certainly have been told "you will be called on into platform x, when you get the signal there is a HST in there, pull up just behind it, you will just about fit. Caution, reverse curves approaching that train. Attention guard, pay attention in case you need to lock out any doors".

But the railway doesn't do this.

 

[MarkyT]

We have expensively installed cab radio across the system.

Then someone decided that it is not to be used when under way.

If you look at aviation, pilots, even those flying aircraft single pilot, with a notably more complex set of conditions to handle, are routinely instructed, whether in the air or on the ground, about all events going on around them by the controller. Just think of how aviation would have handled this event. They would certainly have been told "you will be called on into platform x, when you get the signal there is a HST in there, pull up just behind it, you will just about fit. Caution, reverse curves approaching that train. Attention guard, pay attention in case you need to lock out any doors".

But the railway doesn't do this.

Excellent points. I don't understand why 'one way' instructions like this can't be passed from a signaller to driver with a simple requirement to acknowledge understood in response. Pemissive moves are always approach released in the signallng almost to a stand at the home before the PL clears, so there's a perfect opportunity for this kind of exchange to take place.

 

[Juniper Driver]

Agreed.

The signaller may have, albeit with the best of intentions, laid a trap. But the belt & braces nature of railway safety should have taken that in its stride, namely the driver should have been able to stop short of any obstruction. Whilst some lessons to learn all round, the signaller 's mistake was a performance and delay one, the driver's was the safety mistake which resulted in the collision. Evidently his training was deficient as he should not have been in that position.

Don't see why the training is blamed all the time by the "experts" on here...An incident like this can be caused by a wide variety of factors...The best and most experienced of drivers can make the most basic of mistakes...

If you want to see mistakes I advise you watch Air Crash Investigation which has the same sort of catalogue of errors leading to an unfortunate incident that no one wants.

I don't quite understand the signallers "mistake" but it's anything that I "assume" it to be it may be similar to early one morning when I was signalled onto an occupied platform at Guildford (plat 5 or 4...can't remember it was a long time ago) and not being able to fit in due to there already being an 8 car (or possibly 12) in there.I couldn't even see the platform from the previous signal,either..Of course I'm just assuming what happened here,so sorry.

I suppose my incident that morning could have been looked at as a "mistake" but at the time I dealt with it the best I could.Everything went well and we all lived happily ever after.

 

[contrad!ction]

Excellent points. I don't understand why 'one way' instructions like this can't be passed from a signaller to driver with a simple requirement to acknowledge understood in response. Pemissive moves are always approach released in the signallng almost to a stand at the home before the PL clears, so there's a perfect opportunity for this kind of exchange to take place.

From a signaller's perspective you can do this, albeit not for this situation, using the berth triggered messages on the latest GSM-R update. I don't like using it though so usually I'll just speak to the driver directly (call me old fashioned...)

When doing unplanned permissive moves, if I'm not 100% sure how long a train is I'll give the driver a bell and double check and let them know they'll be coming in on top of another unit and that I've confirmed there's enough room for them, either by track circuit occupation or liaison with platform staff. It's rather embarrassing for all involved to have to do unnecessary wrong direction movements etc...

From reading the report, it sounds like a series of events that individually shouldn't have led to an incident but on this occasion did. It's impossible to attribute (sole) blame to either the driver or the signaller and illustrates perfectly how a momentary lapse of concentration/judgement from either side can lead to potentially very serious accidents, even at relatively low speeds.

 

[Taunton]

potentially very serious accidents, even at relatively low speeds.

Given the low speeds, I do feel that traditional buffers would have handled it. Goodness, loose shunting in the old days used to make contact at greater speeds than these with no problem.

Of course, buffers were another traditional aspect that someone decided to do away with ...

 

[BestWestern]

[Deleted]

 

[Right Away]

GWR now have an instruction in place stating that any train entering a platform on the authority of a position light/subsidiary signal must be doing a maximum of 10 mph at the start of the platform ramp.

Following the Newton Abbot incident in 1994, where a class 158 unit collided with an HST in the station, a couple of Western Region locations (Newport and Westbury) had their permissive working arrangements amended, requiring the signaller to advise the driver of the permissive move before clearing the signal. At Westbury, the instruction was withdrawn approximately 5 years ago owing to the delays that this caused. The previous arrangement also increased the risk of a SPAD as the signals approaching the station would have to be maintained at danger until the driver had been spoken to. I believe that the instruction still applies at Newport (it was the case last year during the Cardiff blockade).

In contrast, at Salisbury, when approaching from the Wilton direction, when the route is set the subsidiary signal clears before the signal is even in view of the driver, several hundred metres ahead.

'Closing up' signals are installed at stations such as Portsmouth Harbour, Reading and some of the platforms at Cardiff Central. These are a signal immediately before the start of the platform and serve to reduce the distance of a permissive move as well as reducing platform occupation headways. However, these can be a potential trap for an unwary driver. For example, at Portsmouth Harbour you receive your platform number at the previous signal. The driver has to remember that they need a double yellow to see them in to the platform, with a single yellow only taking them up to the closing up signal before the platform, even though you have already been given your platform number. (I believe there was a SPAD at Reading last year with similar circumstances).

 

[455driver]

In contrast, at Plymouth this was all controlled by just one signal, well before the station throat. The train ahead was not visible from it. Furthermore the curved approaches would not give a good view of whether the HST was on the same track or the immediately adjacent one, one track to the left, especially as that was where it was timetabled to be, it having been changed on the day. I was disappointed that the report did not comment on the similarity of forward view of a train ahead standing on either track.

This whole approach of a being called on round sharp connections to a platform which you would never expect to already be fully occupied such that your own train will never fit is just a parallel to the recent Norwich accident.

So, the driver approaches, having had all the memos about ensuring punctuality in arrivals, knowing that the HST is timetabled on the adjacent track, is called on to a platform where his train will not fit, round a sharp connection with limited visibility, and when surprised by this is told "it's all your fault, mate". Such an approach to the operation would normally fail any meaningful risk assessment outright.

What memos?
I work for GWR and have never had a memo about being on time!

He would not know where the HST would be parked because that isnt on our schedule cards, he might assume he knows but we all know what can happen.

Having driven from P15 into Plymouth on many occasions I can categorically state that platform visibility is fine if the train is controlled in an appropriate manner, if you cant see then go slow enough to be able to stop short, just because he has a platform number doesnt mean he is clear into the platform!

The train would have been almost at a stand (or indeed stopped) before P15 would clear and the driver would have first seen a big 6 illuminate followed by the 2 white lights (you do know what 2 white lights mean dont you?) and it is the drivers responsibility to stop in the distance they can see to be clear,

While the signaller did make a slight error (overlength train) there was no need for the collision to take place if the driver was doing his job properly.

I think the report is unnecessarily hard on the signaller.

 

[455driver]

We have expensively installed cab radio across the system.

Then someone decided that it is not to be used when under way.

If you look at aviation, pilots, even those flying aircraft single pilot, with a notably more complex set of conditions to handle, are routinely instructed, whether in the air or on the ground, about all events going on around them by the controller. Just think of how aviation would have handled this event. They would certainly have been told "you will be called on into platform x, when you get the signal there is a HST in there, pull up just behind it, you will just about fit. Caution, reverse curves approaching that train. Attention guard, pay attention in case you need to lock out any doors".

But the railway doesn't do this.

You cant just lock out doors like that!

 

[455driver]

Excellent points. I don't understand why 'one way' instructions like this can't be passed from a signaller to driver with a simple requirement to acknowledge understood in response. Pemissive moves are always approach released in the signallng almost to a stand at the home before the PL clears, so there's a perfect opportunity for this kind of exchange to take place.

Thats what the two white lights mean, there is no need for verbal communication to state the same thing.

 

[PHILIPE]

GWR now have an instruction in place stating that any train entering a platform on the authority of a position light/subsidiary signal must be doing a maximum of 10 mph at the start of the platform ramp.

Following the Newton Abbot incident in 1994, where a class 158 unit collided with an HST in the station, a couple of Western Region locations (Newport and Westbury) had their permissive working arrangements amended, requiring the signaller to advise the driver of the permissive move before clearing the signal. At Westbury, the instruction was withdrawn approximately 5 years ago owing to the delays that this caused. The previous arrangement also increased the risk of a SPAD as the signals approaching the station would have to be maintained at danger until the driver had been spoken to. I believe that the instruction still applies at Newport (it was the case last year during the Cardiff blockade).

In contrast, at Salisbury, when approaching from the Wilton direction, when the route is set the subsidiary signal clears before the signal is even in view of the driver, several hundred metres ahead.

'Closing up' signals are installed at stations such as Portsmouth Harbour, Reading and some of the platforms at Cardiff Central. These are a signal immediately before the start of the platform and serve to reduce the distance of a permissive move as well as reducing platform occupation headways. However, these can be a potential trap for an unwary driver. For example, at Portsmouth Harbour you receive your platform number at the previous signal. The driver has to remember that they need at least a double yellow to see them in to the platform, with a single yellow only taking them up to the closing up signal before the platform, even though you have already been given your platform number. (I believe there was a SPAD at Reading last year with similar circumstances).

Wasn't permissive working withdrawn at Newton Abbot itself, also. Another station was Cardiff Queen St following a Railtrack Risk Assessment (Westbury and Newport). Cardiff Queen St had recently been remodelled prior to which Permissive Working was not allowed. However, between the two events privatisation had come in so nobody knew from whom they should be able to claim back the costs involved.

 

[MarkyT]

GWR now have an instruction in place stating that any train entering a platform on the authority of a position light/subsidiary signal must be doing a maximum of 10 mph at the start of the platform ramp.

Following the Newton Abbot incident in 1994, where a class 158 unit collided with an HST in the station, a couple of Western Region locations (Newport and Westbury) had their permissive working arrangements amended, requiring the signaller to advise the driver of the permissive move before clearing the signal. At Westbury, the instruction was withdrawn approximately 5 years ago owing to the delays that this caused. The previous arrangement also increased the risk of a SPAD as the signals approaching the station would have to be maintained at danger until the driver had been spoken to. I believe that the instruction still applies at Newport (it was the case last year during the Cardiff blockade).

In contrast, at Salisbury, when approaching from the Wilton direction, when the route is set the subsidiary signal clears before the signal is even in view of the driver, several hundred metres ahead.

'Closing up' signals are installed at stations such as Portsmouth Harbour, Reading and some of the platforms at Cardiff Central. These are a signal immediately before the start of the platform and serve to reduce the distance of a permissive move as well as reducing platform occupation headways. However, these can be a potential trap for an unwary driver. For example, at Portsmouth Harbour you receive your platform number at the previous signal. The driver has to remember that they need at least a double yellow to see them in to the platform, with a single yellow only taking them up to the closing up signal before the platform, even though you have already been given your platform number. (I believe there was a SPAD at Reading last year with similar circumstances).

The Portsmouth ones are possibly the most futile as they can't have any other functionality other than reducing the length of the call-on (assuming the train already in the terminal platform doesn't clear the platform by taking a watery journey the wrong way through the bufferstops and into the harbour!)

I'm sure I recall another permissive incident at Newton Abbot some years earlier after introduction of the colour lights. A class 50 performing a run-round manoeuvre ran into a pacer, or skipper as they were known locally at the time. I can't find any accident report so I can only assume it was considered minor with little damage and no injuries. The skipper may not have been in passenger service at the time.

 

[PHILIPE]

The carriage from 150219 arrived back at SPM last night from Brodies, so 150938 looks to be back to 150238 and 150219 reformed

 

[455driver]

The carriage from 150219 arrived back at SPM last night from Brodies, so 150938 looks to be back to 150238 and 150219 reformed

Dont hold your breath on that one!