West Coast Railway Company ECS catches fire near Salford Crescent

[Titan]

The upshot of it all is the fact that several contributory factors very unfortunately came together to cause the derailment.

I think it would be unwise to blame it on one single cause alone, eg, the track.

The problem is that all of the contributary effects outside of the track faults, were all things that the track should be capable of dealing with, and if the track was not faulty would not have been an issue at all. You can't say that just because an unusual set of circumstances came together, especially since the track should handle them anyway, that the track is in any way less to blame?

[TheSignalEngineer]

We have an extremely sharp curve going into Liskeard yard from the mainline. Its so sharp that any passenger train that traverses it must not be carrying passengers when it does, ie, railtours etc. Passengers have to disembark and walk across to the Looe platform and wait until the train is reversed into the platform before they rejoin it.

Up until recently, PCA tanks would use it which are right mares for derailing at the best of times!

The curve has zero cant, no check rail and no lubricator yet we have never had a derailment on it.

Yes, an occasional train under very controlled conditions. Deansgate has about 180 trains per day each way.

[The Stationmaster]

Read paragraph 73 of the report.

 

As I said, all the points were considered and the only cause was the track.

That's not exactly what para 73 says - nor (all too typical of RAIB Reports I'm afraid) does it mention any detail in relation to other matters, it simply dismisses them.  And surely from what you've already said the only cause was not 'the track' as you stated in an earlier post that the newly turned tyres were a factor and, to quote your words 'were more susceptible to derailment'.  There is therefore already a combination of circumstances plus the sheer fact that it happened.

 

I think any of us with any experience on investigating derailments and who cut our teeth on simple ones before we got to more complex ones know that almost inevitably it is a combination of factors which causes a particular vehicle to derail instead of others.

[newbryford]

The curve has zero cant, no check rail and no lubricator yet we have never had a derailment on it.

 

And probably a speed limit to match!

Cheers,

Mick

[The Stationmaster]

And probably a speed limit to match!

Cheers,

Mick

Nah, it's on the Western   (although in reality of course all sorts of vehicels and locos have been barred from using it for more years than anyone round here could remember)

[newbryford]

Nah, it's on the Western   (although in reality of course all sorts of vehicels and locos have been barred from using it for more years than anyone round here could remember)

 

Gauge widening on the curve to Brunel standards?

Cheers,

Mick

[Titan]

That's not exactly what para 73 says - nor (all too typical of RAIB Reports I'm afraid) does it mention any detail in relation to other matters, it simply dismisses them.  And surely from what you've already said the only cause was not 'the track' as you stated in an earlier post that the newly turned tyres were a factor and, to quote your words 'were more susceptible to derailment'.  There is therefore already a combination of circumstances plus the sheer fact that it happened.

 

I think any of us with any experience on investigating derailments and who cut our teeth on simple ones before we got to more complex ones know that almost inevitably it is a combination of factors which causes a particular vehicle to derail instead of others.

It seems were mixing up factors and causes, yes many factors, but only one cause - lack of check rail.

 

Its a bit like saying that if someone misjudged their braking and lightly bumped back of the car in front which was stationary at a red traffic light, causing a small crack in its number plate, that the stationary car was a cause since if it was not there they would have stopped before getting to the line and there would have been no accident.

 

Likewise the wheel profiling was a factor but as it was perfectly entitled to be as it was - like the car above was entitled to be where it was. it was not a cause. The only cause (as opposed to factors) was the lack of check rail.

 

Indeed the RAIB specifically differentiates them as such - the wording they use being the check rail as a 'causal factor' ie the one that caused the derailment to happen (paragraph 74)  Everything else is merely a 'factor' (including the lack of lubrication now that I re-read it!)  not what caused the derailment, but a factor required for the derailment to happen.

Edited May 11, 2014 by Titan

[The Stationmaster]

Well whenever I have investigated anything in the past or taken part in a joint inquiry factors were simply that - factors which led to whatever it was occurring and if one of them had not been present the incident or whatever would not have happened - simple (and as plain English) as that.  The expression 'causal factor' is in some respects a contrived nonsense when applied to a derailment if you think about it logically -  for the simple reason that any and every factor is a contributory part of the cause otherwise it would not be identified as a factor.

 

Incidentally I wonder what the braking rate of 47500 was compared with the vehicles formed between it and the train engine? (anyone who has ridden on a train with an unpowered  loco attached rear will know why I mention that ).

[TheSignalEngineer]

 

Incidentally I wonder what the braking rate of 47500 was compared with the vehicles formed between it and the train engine? (anyone who has ridden on a train with an unpowered  loco attached rear will know why I mention that ).

Given the momentum of over 100 tons of loco the back of the train would be catching up with the front. In that case the dead loco will only go in two directions, upwards or sideways, usually both if it's on a curve.

[Titan]

Well seeing as the train was coasting through the curve when the derailment occured,(18mph entering the curve, falling to 14mph at point of derailment)traction/braking does not really come in to it. The driver said the train inexplicably came to a halt just before he was about to apply some power to get through the junction.

[Gary H]

The problem is that all of the contributary effects outside of the track faults, were all things that the track should be capable of dealing with, and if the track was not faulty would not have been an issue at all. You can't say that just because an unusual set of circumstances came together, especially since the track should handle them anyway, that the track is in any way less to blame?

I do respect the points you make but you could also turn the above on its head, thus-

If the loco was in 100% condition, (ie no frame or bogie twist pre incident of which we will never really know for sure)  it to should have been able to deal with a curve with no rail / flange lubricator and check rail, especially at a mere 15 MPH or so just like the lead loco did and the dozens of trains before it.

I think we'll have to agree to disagree on this one!

[Titan]

Interesting you should bring that up, I do find it rather ironic that a 100% loco should have a greater tendancy to derail than one that has seen some service. On the one hand a modern loco with a different profile might not have had a problem, on the other its the same profile that 47's have always been turned to without them falling off the tracks because of it.

[Gary H]

Interesting you should bring that up, I do find it rather ironic that a 100% loco should have a greater tendancy to derail than one that has seen some service. On the one hand a modern loco with a different profile might not have had a problem, on the other its the same profile that 47's have always been turned to without them falling off the tracks because of it.

No no no, I said IF the loco was in 100% condition, it should have been able to deal with a curve not fitted with a lubricator or check rail.

Please try and keep up!

As I said, lets just just agree to disagree eh!

[TheSignalEngineer]

For years 47s ran round there with Class 1s and freights without incident. There are about 15 loco hauled trains and 170 MUs per day over that track and it's the only one to have fallen off as far as I am aware.

Edited May 12, 2014 by TheSignalEngineer

[Catkins]

Well seeing as the train was coasting through the curve when the derailment occured,(18mph entering the curve, falling to 14mph at point of derailment)traction/braking does not really come in to it. The driver said the train inexplicably came to a halt just before he was about to apply some power to get through the junction.

Any one who has learnt to ride a motorcycle, or taken advanced/race driving tuition will know that you need to gradually apply power during when you are going round a curve purely to maintain your speed.

If you allow the speed to fall as you go round the curve, the centripedal forces will want the vehicle to go in a straight line, and -

 

Given the momentum of over 100 tons of loco the back of the train would be catching up with the front. In that case the dead loco will only go in two directions, upwards or sideways, usually both if it's on a curve.

Which obviously happened here, the fire is a separate event, and as such really should be considered separately.

[TheSignalEngineer]

Wasn't 47500 involved in an incident at Pershore about 30 years ago? The train partly derailed due to a broken fishplate but I can't remember if the loco came off or just had a bumpy ride.

[Titan]

Any one who has learnt to ride a motorcycle, or taken advanced/race driving tuition will know that you need to gradually apply power during when you are going round a curve purely to maintain your speed.

If you allow the speed to fall as you go round the curve, the centripedal forces will want the vehicle to go in a straight line, and -

 

 

Which obviously happened here, the fire is a separate event, and as such really should be considered separately.

 

Yes, its so obvious that a train behaves like a motorbike its a wonder that RAIB missed it!!!

 

Its like playing alternative theory bingo, I wonder which one will be next? Eyes down...

Edited May 12, 2014 by Titan

[The Stationmaster]

Wasn't 47500 involved in an incident at Pershore about 30 years ago? The train partly derailed due to a broken fishplate but I can't remember if the loco came off or just had a bumpy ride.

Spot on SE - the loco was not derailed in the Pershore incident but part of its train was

 

http://www.railwaysarchive.co.uk/documents/MoT_Pitlochry1983.pdf

 

 

Yes, its so obvious that a train behaves like a motorbike its a wonder that RAIB missed it!!!

 

Its like playing alternative theory bingo, I wonder which one will be next? Eyes down...

Mind you in view of their past fascination with umbrellas nothing would surprise me 

[TheSignalEngineer]

Although 47500 didn't derail at Pershore I would have thought It took a pretty substantial hit going over 5.5 metres of rail with the keys displaced and probalby hitting a rail end at 75mph.

[Michael Delamar]

Ive got some shots of the area in the 1960s somwehre but Im struggling to find them at the moment, I wanted to see if it did have a checkrail in the past. Does anyone have any historic shots of the area to see if it did have?

Edited May 12, 2014 by Michael Delamar

[ozzyo]

After 30 years do you think that it would still have the same bogies. Even if it did the bogies and the loco would have been through the B.R. works one or two times before being sold on to West coast.

 

I'm sorry to say that if you don't agree with what the RAIB have come up with, (and I do know that you have a lot of experience in working on the railway) what do we have then?

 

In one of the posts it mentioned about having a dead loco at the back of the train and it's weight. It could also be said about having a loco under power with no direct control from the front loco could be as bad, or worse (steam loco front Diesel at the back).

 

The amount of trains over the curve is a bit of a red herring, in that most (not all) of the trains are of 4 wheeled bogies so a shorter wheel base, 

 

Also IIRC the curve was laid using concrete sleepers and flat bottom rail. I've not seen that many curves that have this combination with check rails on them.

 

In my local area we had a curve that always had a check rail on it in the up direction and the down line did not. Was it because of the curve? But now both lines have flat bottom rail and no check rail, AFAIK the curve has not changed nor has the minimum radius for a check rail to be fitted.

 

OzzyO. 

[Dagworth]

Also IIRC the curve was laid using concrete sleepers and flat bottom rail. I've not seen that many curves that have this combination with check rails on them.

 

In my local area we had a curve that always had a check rail on it in the up direction and the down line did not. Was it because of the curve? But now both lines have flat bottom rail and no check rail, AFAIK the curve has not changed nor has the minimum radius for a check rail to be fitted.

 

OzzyO.

Concrete sleepers are not subject to gauge widening over time as wooden ones are. The check rail prevents the outer wheel pushing the rails apart as well as preventing wheels from climbing the rail.

 

Andi

[ozzyo]

Concrete sleepers are not subject to gauge widening over time as wooden ones are. The check rail prevents the outer wheel pushing the rails apart as well as preventing wheels from climbing the rail.

 

Andi

I've just had another look at the RAIB report.

Looking at the photos the sleepers were wooden ones, no check rail was in position, and the track was wide to gauge.

 

So some times a bit of gauge widening on a curve may not help!

 

IIRC a check rail should be in place on any passenger / main line curve under 10 chains radius.? The curve IIRC at Salford was approx. 8.9 chains (so that's 72'  under the recommend radius limit), so the use of concrete sleepers would have helped in this case as they would not permit any gauge widening with use (the only gauge widening you would get is the ware of the outer rail) .

 

The facts are that five flange greasers were removed and replaced by one (that was not working), the loco had just had it's wheels turned to the correct profile so the wheels were (to all intent and purpose new and clean), the bogie was seen fit for use (forget the indecent 30 years ago). The train was running under the speed limit for the curve.

 

It all boils down to Net Work Rail cutting costs. Wooden sleepers with no check rail, lack of flange greasers, the lack of checking for any gauge widening on the curve to take in to account for the lack of the check rail.     

 

OzzyO.

Edited May 13, 2014 by ozzyo

[LMS2968]

A friend of mine was a driver for Virgin at Longsight, but had been a driver on BR since the early 1970s and worked over the MSJ&A regularly; he was a bit surprised at the idea that a check rail was needed as he had no recollection of there ever being one. I've seen a photo by Jim Carter (copyrighted) of a Class 40 at the same spot, but it does not show one.

[Titan]

Ive got some shots of the area in the 1960s somwehre but Im struggling to find them at the moment, I wanted to see if it did have a checkrail in the past. Does anyone have any historic shots of the area to see if it did have?

 

The only thing I can find is that the RAIB states that not only the current standards, but the previous equivalents (track design handbook?) going back for at least 100 years specified that it should have one, and they presented no evidence that it ever had!

 

Gauge widening is really only an advantage on very very tight curves, on anything else they can allow a greater 'skew' of the bogie or short wheel base wagon, which can increase the angle of attack of the flange which in turn increases the risk of flange climb. To be fair, the rails on this curve had been replaced fairly recently (at which point the lack of check rail ought to have been corerected, but its seems it was just replaced like for like) and the amount of wear was well within acceptable limits.

Edited May 13, 2014 by Titan