Incident at Talerddig, Wales - 21/10/2024

[12LDA28C]

Would you want the sander to actually operate as part of the test - wouldn’t that be an issue in areas where this might happen with sand going places you don’t want it?

The sanders would usually be tested on Depot (or remote stabling location) as part of a train prep prior to entering service. The whole point of testing the sanders is to see some sand applied to the railhead.

 

[edwin_m]

Would you want the sander to actually operate as part of the test - wouldn’t that be an issue in areas where this might happen with sand going places you don’t want it?

The sanders would usually be tested on Depot (or remote stabling location) as part of a train prep prior to entering service. The whole point of testing the sanders is to see some sand applied to the railhead.

Yes, seeing the full operation from start to end is the only fully comprehensive test of functionality. Conceivably a low voltage could be applied to the circuit so current detection would demonstrate the integrity of the circuit itself including the solenoid valve which might be susceptible to burning out. However, that wouldn't prove the mechanical integrity of the sanding system, such as whether the pipes are blocked as was initially reported. I wonder if that is still thought to be the case for this incident, and the reported electrical fault is separate.

 

[fishwomp]

New NIR (4134) out from TfW today, in summary the sander on the leading vehicle of the sliding train has been found with two separate electrical faults, ...

I'd thought that those kinds of reports would have been issued by RAIB as they are investigating?. They have so far only done their initial report which spoke of a blocked sander pipe., no update on the website.. their reports are usually very good and thorough.

 

[millemille]

I'd thought that those kinds of reports would have been issued by RAIB as they are investigating?. They have so far only done their initial report which spoke of a blocked sander pipe., no update on the website.. their reports are usually very good and thorough.

No, NIR's are completely separate to RAIB. NIR's are almost issued on a daily basis, well on average at least (you normally see a flurry of them on a Friday afternoon as fleet engineers realise they've still got them to do for the week's list of things that went wrong so bang them out, shut down their computer and go home and the rest of the industry gets a deluge of them and are left clueless as to what they should do...).

NIR's are an immediate warning mechanism, or at least they're supposed to be. Sometimes they get a "bit" delayed (like the NIR issued in April for a sander system defect that caused a station over run 7 months earlier...).

 

[D365]

It depends on how much of the system it tests. Generally, the further up-stream you go, the more complex it gets. You could, for example, connect the test button in parallel with the WSP output that controls the sanders. But that wouldn't prove that the WSP output was working. Proving that might need you to simulate all the WSP inputs.

I could envisage a situation where the WSP appears to be working correctly, because it correctly controls the traction motors, and the sanders appear to be working correctly when tested manually, but for some reason the WSP fails to operate the sanders.

That is why electronic outputs and the electrical ’coupling’ interfaces (i.e. relay and switch) need to tested in separate steps! In a similar vein to what @edwin_m has written above.

 

[RichardAsh1981]

The WSP is a single, quite compact, box. If it is modern and electronic, it can be required to have self-monitoring / self test features built in. Like most digital systems, it will work fine until it doesn't, unless there is a design/type fault - which is not what testing the sanders on a unit is there to find. At this point it either works as intended, or can send a big flag to tell the driver that it is not going to work at all (which would include input sensing errors). There would be no case for simulating inputs to the WSP for in-service testing (development and type testing is different - see https://railtechnicalstrategy.co.uk...ction-evaluation-rig-brake-system-technology/).

What would be sensible would be a test switch input to the WSP, to provide a way to instruct the WSP device to try and drop sand. Fit that control down near the sand pipes, and how have the best of both worlds - you know the sand has actually come out(!) and that the control from the WSP to the sander is working because the test signal has gone all the way up to the WSP and back. The down side is twice as much cabling to route through the vehicle, and needing the input on the WSP - fine if you put it in at the specification stage, a problem to add later. It all comes down to the risk of failure vs. the cost, as always.

Taking this further, electrical power systems protection usually inverts the signals to make them fail-safe - current flows all the time in the "do nothing" state, and is broken to operate the circuit breaker, raise the alarm etc. This means any fault would immediately become obvious! How you would avoid emptying the sand box every time you switch the master switch off would take some thinking about - and is almost certainly overkill for this case.

 

[MadMac]

We had a case on Los Angeles Metro where the sanders made a train disappear…..

First empty train of the day into the terminus at Norwalk was automatically routed into one platform. It was quickly followed by a second empty train, which should have been routed into the other platform, but was routed into the same platform as the first one, which had disappeared off the track circuits. It was 3 AM on a damp morning. What we found out and were subsequently able to replicate was that the vehicle sanders “self-tested” every time the vehicles were booted up - this feature was supposed to have been disabled some years previously. There was a nice pile of damp sand at every wheel from three trains booting up in exactly the same place every hour all day.

 

[edwin_m]

Taking this further, electrical power systems protection usually inverts the signals to make them fail-safe - current flows all the time in the "do nothing" state, and is broken to operate the circuit breaker, raise the alarm etc. This means any fault would immediately become obvious! How you would avoid emptying the sand box every time you switch the master switch off would take some thinking about - and is almost certainly overkill for this case.

If this was done, the sander might go off unsolicited when some fault appeared. This could itself be hazardous, for example if a maintenance worker was near the sand pipes.

We had a case on Los Angeles Metro where the sanders made a train disappear…..

First empty train of the day into the terminus at Norwalk was automatically routed into one platform. It was quickly followed by a second empty train, which should have been routed into the other platform, but was routed into the same platform as the first one, which had disappeared off the track circuits. It was 3 AM on a damp morning. What we found out and were subsequently able to replicate was that the vehicle sanders “self-tested” every time the vehicles were booted up - this feature was supposed to have been disabled some years previously. There was a nice pile of damp sand at every wheel from three trains booting up in exactly the same place every hour all day.

Another reason not to have the sanders operate when failed. A manual test in the depot avoids this problem as it won't be track circuited, and also allows any fault to be found where there's some scope to do something about it.

 

[rower40]

Another reason not to have the sanders operate when failed. A manual test in the depot avoids this problem as it won't be track circuited, and also allows any fault to be found where there's some scope to do something about it.

And you can then send the trainee/student out with a shovel and a bucket to retrieve the pile of sand.

 

[merry]

If this was done, the sander might go off unsolicited when some fault appeared. This could itself be hazardous, for example if a maintenance worker was near the sand pipes.

Of course, that is a risk if the circuits energise separately...but a bit of design thought (ie. control and status interlocking) could reduce the risk.

e.g. if the sander actuation line is actually 'sander inhibit' when energised, then a control logic can be derived which only activates sanding when both the sander inhibit is not energised, and the sander & WSP are enabled (on).

If the train is not 'prepared' (ready for service, cab key in, any other necessary conditions satisfied), then the sander and WSP would be disabled, and the sander could not be activated.

Only when the sander & WSP supply are on, and the WSP has ceased to inhibit sanding, would the sander activate.

A 'sanding active' indicator would of course also be sensible, showing the driver when the sander valve/solenoid is open (operated from a sensing contact on the sander, ideally), as well as an indication of when the WSP is active. And you'd still need a 'manual sand' button for those cases it is required & to test, and the ability to isolate auto sanding if it has a wiring fault that causes continual activation.

Like one poster said, this requires quite a bit more wiring to achieve. Is it justified? Depends on how often it's an issue. We are here talking of older vehicles, and it's questionable how much modification would be justifiable given the service life remaining.

In more modern digitally managed trains, the interaction of systems like this would likely be managed & systems monitored by the TCMS, with the driver informed of any problem state if need be.

 

[158841]

Do we have any (if) possible return dates for these units as yet?
841 is a stern favourite of mine!

 

[yorkie]

According to an informed source, the units were still stored with their damaged ends inside a shed at Machynlleth depot as of last week, and there has been no known road movements of any units out of the depot, so presumably they are still there.

 

[newtownmgr]

Suspect they won’t return to service with the 197’s due onto the Cambrian soon.

 

[Towers]

Suspect they won’t return to service with the 197’s due onto the Cambrian soon.

A supply of spares will be pretty useful, I would think. That may help to ease a decision not to undertake costly repairs.

 

[newtownmgr]

A supply of spares will be pretty useful, I would think. That may help to ease a decision not to undertake costly repairs.

Yes, which I suspect is why they are still there, unless of course they are still waiting on insurance to be sorted.

 

[Wolfie]

A supply of spares will be pretty useful, I would think. That may help to ease a decision not to undertake costly repairs.

The ROSCO will ultimately decide if they are worth more to it as parts or as repaired trains. The shortage of DMUs could suggest the latter.

 

[Lurcheroo]

The units are still at Mach and will not be stripped of parts whilst they are there. My understanding is neither will see service again and are due to moved at some point.

 

[Wolfie]

The units are still at Mach and will not be stripped of parts whilst they are there. My understanding is neither will see service again and are due to moved at some point.

Service with TfW or fullstop?

 

[Lurcheroo]

Service with TfW or fullstop?

full stop.

 

[Wolfie]

full stop.

TY

 

[Class 170101]

The units are still at Mach and will not be stripped of parts whilst they are there. My understanding is neither will see service again and are due to moved at some point.

full stop.

Surprised they aren't being stripped for spares if they are not to see further service. Why would this be?

 

[millemille]

Surprised they aren't being stripped for spares if they are not to see further service. Why would this be?

Because ownership of the vehicles, in terms of Angel Trains and their insurers and potentially TfW and their insurers, is probably not yet finalized and because there are still ongoing investigations and potential prosecutions impending.

 

[Russel]

Has it already been said why the 2 units can't form 1 good unit from the undamaged coaches, or were all 4 too damaged?

 

[pokemonsuper9]

Has it already been said why the 2 units can't form 1 good unit from the undamaged coaches, or were all 4 too damaged?

While I don't know which carriages are damaged, if both the carriages with disabled toilets are damaged, forming 1 unit would mean no disabled toilet, and having no disabled toilet means you can't have normal toilets either, so no toilets for hours of journey.

 

[Krokodil]

While I don't know which carriages are damaged, if both the carriages with disabled toilets are damaged, forming 1 unit would mean no disabled toilet, and having no disabled toilet means you can't have normal toilets either, so no toilets for hours of journey.

More to the point, the ETCS equipment is divided between the two coaches. You must have a proper set, you can't mix and match or reform them at will like other operators have done. The damaged ends were both the same type of vehicle.

 

[61653 HTAFC]

While I don't know which carriages are damaged, if both the carriages with disabled toilets are damaged, forming 1 unit would mean no disabled toilet, and having no disabled toilet means you can't have normal toilets either, so no toilets for hours of journey.

I think (but could be misremembering) that the worst-damaged vehicles were the 52 car on one set and the 57 car on the other... So forming one good set from the less-damaged cars could be possible- though that does depend on just how damaged each of those vehicles are.
If (a) it's worthwhile repairing those vehicles; and (b) those vehicles are both the ones without the universal access toilet- then the only options for reuse would be to form a couple of 3-car sets. Not sure how well two oddball sets could be integrated into TfWs methods of working, such a move could cause more issues that it solves.

Note: ignore the struck-through text, thanks to the additional information in #445.

 

[pokemonsuper9]

then the only options for reuse would be to form a couple of 3-car sets

I think it was discussed at one point that Machynlleth depot won't work for 3 cars.

 

[61653 HTAFC]

I think it was discussed at one point that Machynlleth depot won't work for 3 cars.

Which I addressed in the very next sentence!

 

[Krokodil]

I suppose that it's possible that the undamaged vehicles will find a use as centre cars once TfW gets rid and the ETCS equipment is redundant. Depends upon the new operator really.

 

[HamworthyGoods]

Has it already been said why the 2 units can't form 1 good unit from the undamaged coaches, or were all 4 too damaged?

There weren’t any undamaged coaches in the collision, forces would have gone through both coaches in the set.

An example of this being the Clapham crash, the 12VEP that was on the Basingstoke train had a car written off in the middle of the train.