Scotrail 334 becomes uncoupled near Uphall (26/11)

[Ben Glasgow]

This is what happened. After station duties were completed the driver was undertaking a "pan drill/battery reset" (basically a reboot/switch off and on again) as a result of spurious faults on the train, very common on 334's. This is why there were reports of a "power outage." For some reason (and which is the focus of the investigation) the units detached themselves during this process. As a result the front unit rebooted as expected so the driver continued unaware, leaving the rear unit detached and dead in the platform.

Out of curiosity, what sort of ratio do modern trains run via software versus direct user input via switches etc which bypass any sort of onboard software or TMS System? With how you’ve described it (also thanks as I’m genuinely fascinated how this happened) do you reckon that in doing a reboot of the system, it’s then cycled the coupling mechanisms?

Because the uncouple button was pressed

As others have mentioned, this doesn’t seem very likely but would suggest that it would have to be knowingly done by crew who’d have been able to energise the rear cab which again seems very bizarre and unlikely.

Yes, this is the big problem in this case. After 5 days of investigating, they have so far not been able to recreate the fault. The question I'd now be asking if I was still involved in engineering and not a train driver is "do we think it will or could happen again? and if we do, what will we do?". The company will have to either ground the entire 334 fleet until such time as they find the cause OR immediately ban 334 units from operating in multiple OR come up with a solution that completely isolates the couple/uncouple circuit, thus allowing them to keep multiple units already coupled in service. Its a tough decision to make, especially if grounding an entire fleet of 40 units has to be done. This could seriously harm the December timetable change.

I reckon they will have to change the procedure so that should a 3+3 consist show abnormalities requiring a reboot of the train, the driver will have to inspect the connection between the two trains. Is this reasonable to ask of a driver?

 

[ComUtoR]

As others have mentioned, this doesn’t seem very likely but would suggest that it would have to be knowingly done by crew who’d have been able to energise the rear cab which again seems very bizarre and unlikely.

What is more likely. A spurious fault, that cannot be replicated, which randomly created the perfect set of circumstances that everything failed in such a way that the front unit didn't register any faults and effectively a serious wrong side failure occurred. Or the couple button was accidentally pressed ? Sometimes the most reason explanation is the right one.

Now why do I, an experienced Driver, suspect the uncouple button ? Unit behavior is important here. The front unit proceeded as if it was completely normal and more importantly, whole. For that to have happened the unit must have uncoupled both electrically AND mechanically. Even if we accept the a fault could have electrically uncoupled the train and reconfigured the units as whole how do you account for the mechanical aspect ? When I split with a tightlock, you have to push on to help release the mechanical connection. With a Delner you can hit the button and then just pull away, although the unit needs to reconfigure on the TMS. That button press is part of the mechanical connection as it throws the drumswitch and uses air to release the coupler. I have, on more than one occasion, pressed the uncouple button by mistake. The slightest distraction and the repetitive nature of the task leads to silly errors like pressing the wrong button.

So we have a fault that both mechanically and electrically uncoupled a train in a remote cab and affected both units. That's not a scenario I see as plausible. There is little evidence presented as to any sequence of events and the actions of the Driver are not discussed and even that I find suspicious. What is also missing is any statement about this being investigated by the RAIB. As posted up thread an UNEXPECTED detachment must be reported. As this appears to be limited to an internal investigation I again suspect, that they know the cause already and that any actions carried out are already known. Not forgetting that the rear unit would have to have been recovered too. Did the Driver who recovered the unit put their key on and need to release the Drum switch as that should have been in the coupled position. Again, this information is not been released either. Any full and proper investigation will need both units interrogated and a full sequence of events, as well as what happened post incident, will be important.

I reckon they will have to change the procedure so that should a 3+3 consist show abnormalities requiring a reboot of the train, the driver will have to inspect the connection between the two trains. Is this reasonable to ask of a driver?

I do not know the procedure for 334s as I don't sign them but when I carry out a reboot or pan up/down I do it from the front cab. As the units are electrically coupled both pans drop then both pans go back up. I then need to check that both pans have gone up again. I used to have to carry out a physical check but on 700 you can now do this via the pan camera. Again, I'm not familiar with the TMS on a 334 but ours will show the formation so you can check this via the TMS. If this is indeed a serious fault with 334's they will need to pull the fleet. The problem with a walk back and physical check is that the units would not have moved and a visual check potentially would miss any issues; especially with Delners. After a reboot the Driver would of had to reset the cab. Again, the Drivers actions here are not known and for me, checking the TMS is an integral part or that procedure.

Personally I hope I'm wrong but in some ways a fault of this nature has a serious implications and I find that the more scary scenario to accept. Without an RAIB investigation I doubt the public will ever know what actually happened.

 

[Ben Glasgow]

What is more likely. A spurious fault, that cannot be replicated, which randomly created the perfect set of circumstances that everything failed in such a way that the front unit didn't register any faults and effectively a serious wrong side failure occurred. Or the couple button was accidentally pressed ? Sometimes the most reason explanation is the right one.

Now why do I, an experienced Driver, suspect the uncouple button ? Unit behavior is important here. The front unit proceeded as if it was completely normal and more importantly, whole. For that to have happened the unit must have uncoupled both electrically AND mechanically. Even if we accept the a fault could have electrically uncoupled the train and reconfigured the units as whole how do you account for the mechanical aspect ? When I split with a tightlock, you have to push on to help release the mechanical connection. With a Delner you can hit the button and then just pull away, although the unit needs to reconfigure on the TMS. That button press is part of the mechanical connection as it throws the drumswitch and uses air to release the coupler. I have, on more than one occasion, pressed the uncouple button by mistake. The slightest distraction and the repetitive nature of the task leads to silly errors like pressing the wrong button.

So we have a fault that both mechanically and electrically uncoupled a train in a remote cab and affected both units. That's not a scenario I see as plausible. There is little evidence presented as to any sequence of events and the actions of the Driver are not discussed and even that I find suspicious. What is also missing is any statement about this being investigated by the RAIB. As posted up thread an UNEXPECTED detachment must be reported. As this appears to be limited to an internal investigation I again suspect, that they know the cause already and that any actions carried out are already known. Not forgetting that the rear unit would have to have been recovered too. Did the Driver who recovered the unit put their key on and need to release the Drum switch as that should have been in the coupled position. Again, this information is not been released either. Any full and proper investigation will need both units interrogated and a full sequence of events, as well as what happened post incident, will be important.




I do not know the procedure for 334s as I don't sign them but when I carry out a reboot or pan up/down I do it from the front cab. As the units are electrically coupled both pans drop then both pans go back up. I then need to check that both pans have gone up again. I used to have to carry out a physical check but on 700 you can now do this via the pan camera. Again, I'm not familiar with the TMS on a 334 but ours will show the formation so you can check this via the TMS. If this is indeed a serious fault with 334's they will need to pull the fleet. The problem with a walk back and physical check is that the units would not have moved and a visual check potentially would miss any issues; especially with Delners. After a reboot the Driver would of had to reset the cab. Again, the Drivers actions here are not known and for me, checking the TMS is an integral part or that procedure.

Personally I hope I'm wrong but in some ways a fault of this nature has a serious implications and I find that the more scary scenario to accept. Without an RAIB investigation I doubt the public will ever know what actually happened.

Thanks for giving your input as a driver as I’m just speculating (again just genuinely interested which I hope doesn’t come across the wrong way).

Yeah it would really mean excrement hitting the fan if that route had its consists pulled if it was found to be major fault/design flaw rather than a singular fault isolated to one unit.

I’m curious what trains will be going on to the glasgow central Shotts line and if it’ll be the 334s.

 

[Southsider]

Thanks for giving your input as a driver as I’m just speculating (again just genuinely interested which I hope doesn’t come across the wrong way).

Yeah it would really mean excrement hitting the fan if that route had its consists pulled if it was found to be major fault/design flaw rather than a singular fault isolated to one unit.

I’m curious what trains will be going on to the glasgow central Shotts line and if it’ll be the 334s.

385s planned

 

[Tomnick]

What is more likely. A spurious fault, that cannot be replicated, which randomly created the perfect set of circumstances that everything failed in such a way that the front unit didn't register any faults and effectively a serious wrong side failure occurred. Or the couple button was accidentally pressed ? Sometimes the most reason explanation is the right one.

Unless the driver went to, and opened up, a middle cab as part of the 'turn it off and on again', I don't understand how the uncouple button could've accidentally been pressed in a cab where it'd have any effect?

 

[Sirius]

I’m curious what trains will be going on to the glasgow central Shotts line and if it’ll be the 334s.

Almost every single 334 is required to run the Helensburgh/Edinburgh and Milngavie/Edinburgh service at as close to 6 carriages all day, with no other unit cleared for the Airdrie to Uphall stretch in service.

You see the odd one running Airdrie/Balloch and Dumbarton/Cumbernauld, but to move the 334s to a new route would require a lot of planning.

 

[sqrtc]

Almost every single 334 is required to run the Helensburgh/Edinburgh and Milngavie/Edinburgh service at as close to 6 carriages all day, with no other unit cleared for the Airdrie to Uphall stretch in service.

Why is that? Seems like an odd bottleneck in fleet flexibility.

 

[Sirius]

Why is that? Seems like an odd bottleneck in fleet flexibility.

I think at the time the restriction was in cab CCTV not available on the other 20m units (320, 318) to allow DOO at the new stations. I could be wrong though.

 

[380101]

What is more likely. A spurious fault, that cannot be replicated, which randomly created the perfect set of circumstances that everything failed in such a way that the front unit didn't register any faults and effectively a serious wrong side failure occurred. Or the couple button was accidentally pressed ? Sometimes the most reason explanation is the right one.

Now why do I, an experienced Driver, suspect the uncouple button ? Unit behavior is important here. The front unit proceeded as if it was completely normal and more importantly, whole. For that to have happened the unit must have uncoupled both electrically AND mechanically. Even if we accept the a fault could have electrically uncoupled the train and reconfigured the units as whole how do you account for the mechanical aspect ? When I split with a tightlock, you have to push on to help release the mechanical connection. With a Delner you can hit the button and then just pull away, although the unit needs to reconfigure on the TMS. That button press is part of the mechanical connection as it throws the drumswitch and uses air to release the coupler. I have, on more than one occasion, pressed the uncouple button by mistake. The slightest distraction and the repetitive nature of the task leads to silly errors like pressing the wrong button.

So we have a fault that both mechanically and electrically uncoupled a train in a remote cab and affected both units. That's not a scenario I see as plausible. There is little evidence presented as to any sequence of events and the actions of the Driver are not discussed and even that I find suspicious. What is also missing is any statement about this being investigated by the RAIB. As posted up thread an UNEXPECTED detachment must be reported. As this appears to be limited to an internal investigation I again suspect, that they know the cause already and that any actions carried out are already known. Not forgetting that the rear unit would have to have been recovered too. Did the Driver who recovered the unit put their key on and need to release the Drum switch as that should have been in the coupled position. Again, this information is not been released either. Any full and proper investigation will need both units interrogated and a full sequence of events, as well as what happened post incident, will be important.




I do not know the procedure for 334s as I don't sign them but when I carry out a reboot or pan up/down I do it from the front cab. As the units are electrically coupled both pans drop then both pans go back up. I then need to check that both pans have gone up again. I used to have to carry out a physical check but on 700 you can now do this via the pan camera. Again, I'm not familiar with the TMS on a 334 but ours will show the formation so you can check this via the TMS. If this is indeed a serious fault with 334's they will need to pull the fleet. The problem with a walk back and physical check is that the units would not have moved and a visual check potentially would miss any issues; especially with Delners. After a reboot the Driver would of had to reset the cab. Again, the Drivers actions here are not known and for me, checking the TMS is an integral part or that procedure.

Personally I hope I'm wrong but in some ways a fault of this nature has a serious implications and I find that the more scary scenario to accept. Without an RAIB investigation I doubt the public will ever know what actually happened.

The couple button was NOT pressed. This is mentioned up thread by @scotraildriver and his comments are pretty much what has happened. The forward facing cctv from both intermediate cabs along with internal saloon cctv show that there was no persons in either cab at the time of the incident. The 334s are fitted with Delner couplers now.

A full electrical and mechanical uncouple took place in this incident and as far as I know, fleet are still trying to replicate the fault. I suspect that the pan drill carried out by the driver has caused an electrical fault of some sort.

 

[robbeech]

The bit that still concerns me is how on some stock the tail lights are reverted to the state they were in before coupling if a sucesful uncoupling is executed. Given this for all intents and purposes was a successful uncoupling (no brake application, no warnings etc) then had the train been a model where the tail lights do this (didn’t see anyone confirm this either way) and had this been an AB area (appreciate it’s not but the general point stands) then there would be huge potential for a collision.
I mentioned it before and the subject did seem to change quite quickly.

 

[Esker-pades]

The bit that still concerns me is how on some stock the tail lights are reverted to the state they were in before coupling if a sucesful uncoupling is executed. Given this for all intents and purposes was a successful uncoupling (no brake application, no warnings etc) then had the train been a model where the tail lights do this (didn’t see anyone confirm this either way) and had this been an AB area (appreciate it’s not but the general point stands) then there would be huge potential for a collision.
I mentioned it before and the subject did seem to change quite quickly.

The potential for collision bit has been answered fully. The signalling system would still have detected something in the block, meaning any protecting signals would be set to red.

 

[robbeech]

The potential for collision bit has been answered fully. The signalling system would still have detected something in the block, meaning any protecting signals would be set to red.

Is this always the way in absolute block signalling. Are there still track circuits in all of them?
I appreciate this area isn’t like this.

 

[Esker-pades]

Is this always the way in absolute block signalling. Are there still track circuits in all of them?
I appreciate this area isn’t like this.

I'm by no means an expert, but people with knowledge on this thread have stated that signals would be at red meaning a collision was never a risk.

 

[robbeech]

I'm by no means an expert, but people with knowledge on this thread have stated that signals would be at red meaning a collision was never a risk.

Yes. With track circuits there’s no real way a collision could occur as the track would still show a train on the line. But with absolute block signalling this isn’t necessarily the case as it is the signaller in the box that is accepting a train into their section when one has left. If one leaves with a tail light they will assume it is complete.
I’m even further from an expert in the railway and the workings of it but I would consider myself an expert in system control and safety in my field of work which whilst VERY different follows the same failsafe concepts.

 

[Tomnick]

I'm by no means an expert, but people with knowledge on this thread have stated that signals would be at red meaning a collision was never a risk.

Not necessarily in an AB section, no. Most aren’t track circuited throughout. The principle there is generally that the section is ‘proved’ clear by the signalman observing the tail lamp. It’s a Bad Thing if they can turn themselves on in the event of an unintended train division, in my view.

 

[robbeech]

Not necessarily in an AB section, no. Most aren’t track circuited throughout. The principle there is generally that the section is ‘proved’ clear by the signalman observing the tail lamp. It’s a Bad Thing if they can turn themselves on in the event of an unintended train division, in my view.

This confirms my thoughts. It’s incredibly unlikely of course.
My thoughts initially was that I cannot understand in any way how it would be deemed acceptable for tail lights to bebswitched on automatically. Surely the fail safe aspect of this would be a manual illumination of them. If the member of staff forgets to switch them on then this is fail safe as the signaller will assume the train is incomplete and not accept another train and contact the driver. Lights that have come on automatically, even if it is by operator error pressing an uncouple button is in my opinion (and hopefully others) not a failsafe operation. It has potential for a very dangerous situation.

 

[alangla]

The station & units in question have very comprehensive CCTV. If the driver had been in the middle cabs then it would have been captured, so it’s safe to exclude that possibility. My (uneducated) money is on some bizarre coding error in the TMS software, or something during the reset of the unit that somehow energised the uncouple mechanism. Bear in mind that these units were built around 2000 and have acquired WiFi, Dellner couplers and air conditioning in the last couple of years. We don’t know how or if any of that interfaces with the TMS.
On the reason for no other stock east of Airdrie, basically the stations east of Bathgate were sized for 6x23m before electrification but most of the Glasgow area ones are 6x20 with little prospect of extension, plus there’s no platform mirrors or monitors beyond Airdrie, so the 334 is the only thing suitable. 4x380 is theoretically possible, I think, but no traction knowledge & insufficient capacity for the peak traffic. 318/320 run ECS from Airdrie to Bathgate LMD every day, but can’t run in service obviously.

 

[Ben Glasgow]

Almost every single 334 is required to run the Helensburgh/Edinburgh and Milngavie/Edinburgh service at as close to 6 carriages all day, with no other unit cleared for the Airdrie to Uphall stretch in service.

You see the odd one running Airdrie/Balloch and Dumbarton/Cumbernauld, but to move the 334s to a new route would require a lot of planning.

Yeah it turns out we’re getting the 385s that also run on the Edinburgh to Glasgow queen street line! Exciting!

 

[GW43125]

Not necessarily in an AB section, no. Most aren’t track circuited throughout. The principle there is generally that the section is ‘proved’ clear by the signalman observing the tail lamp. It’s a Bad Thing if they can turn themselves on in the event of an unintended train division, in my view.

Any train in an absolute block section will be carrying a guard, who will be able to access the GSM-R to report such a division.

 

[gsnedders]

Any train in an absolute block section will be carrying a guard, who will be able to access the GSM-R to report such a division.

Will they, though? Are they required to be in the rear portion at any time in an AB section? What if just the rearmost coach split and the guard was in another coach? Or a three-unit train, would that have three guards?

 

[Tomnick]

Any train in an absolute block section will be carrying a guard, who will be able to access the GSM-R to report such a division.

Not necessarily. As per the previous reply, the guard won't necessarily be in the rear portion. There might not even be a guard at all on an ECS working - DOO(NP). Northern's current proposals almost certainly involve DOO(P) over AB lines without continuous train detection too, even if it's not currently allowed.

 

[Mathew S]

Not necessarily. As per the previous reply, the guard won't necessarily be in the rear portion. There might not even be a guard at all on an ECS working - DOO(NP). Northern's current proposals almost certainly involve DOO(P) over AB lines without continuous train detection too, even if it's not currently allowed.

That's quite the assertion. Any evidence for that?

 

[Tomnick]

That's quite the assertion. Any evidence for that?

Which bit? Assuming you mean the last bit, simply that the union claim that Northern state that they consider all bar two (?) of their routes suitable for DOO, and that most of their ‘Connect’ routes involve AB without continuous train detection at some point.

 

[DelW]

I know we're into hypothetical cases here, but are there any Absolute Block areas on electrified lines? I would have expected that AB would only be left in relatively quiet areas, and that electrification would (always?) have been accompanied by modernisation of the signalling to include track circuits or axle counters.

If that's correct, the possible danger would only occur if a diesel set could exhibit the same behaviour, and since it seems to have been triggered by a pantograph reset, that's even more improbable.

 

[Tomnick]

I know we're into hypothetical cases here, but are there any Absolute Block areas on electrified lines? I would have expected that AB would only be left in relatively quiet areas, and that electrification would (always?) have been accompanied by modernisation of the signalling to include track circuits or axle counters.

If that's correct, the possible danger would only occur if a diesel set could exhibit the same behaviour, and since it seems to have been triggered by a pantograph reset, that's even more improbable.

There are indeed. Stockport is the obvious one, but that is track circuited throughout so the risk is lower. Guide Bridge (-ish, actually Newton/Godley) to Dinting is still AB (without continuous train detection) though I believe. There's some Absolute Block on the independent lines around Crewe too, but that's maybe a little tenuous!

 

[matchmaker]

I know we're into hypothetical cases here, but are there any Absolute Block areas on electrified lines? I would have expected that AB would only be left in relatively quiet areas, and that electrification would (always?) have been accompanied by modernisation of the signalling to include track circuits or axle counters.

If that's correct, the possible danger would only occur if a diesel set could exhibit the same behaviour, and since it seems to have been triggered by a pantograph reset, that's even more improbable.

Before Yoker IECC opened in 1991, large parts of the Glasgow North lines were AB.

 

[DelW]

Thanks for the answers above, the Dinting line hadn't occurred to me but of course that was quite an early electrification. Fortunately it still looks pretty unlikely that any 334s will be working in AB areas for now.

 

[Edinburgh2000]

ScotRail has told The Scotsman its own investigation has found the cause to be “an electrical spike at the uncoupling point”. Does that make any sense?

ASLEF is quoted as saying "following the investigation by ScotRail’s fleet engineers and a full explanation of what has happened, we are content that these units are safe and it’s highly unlikely this incident will ever be repeated.”

The RAIB is still apparently "reviewing" the incident.