Was the Cowden crash due to cost cutting?

[2HAP]

Cowden - multiple factors which due to the limitations of the time the role each may have played was never quite fully resolved

Cowden was down to BR cost-cutting. If the line hadn't been singled the accident could never have happened.

 

[D6130]

Cowden was down to BR cost-cutting. If the line hadn't been singled the accident could never have happened.

....or if it had to be singled, the provision of trap points at the end of the double line sections would have prevented the accident.

 

[21C101]

Cowden was down to BR cost-cutting. If the line hadn't been singled the accident could never have happened.

That is silly. Single Lines have as good a safety record as double track. You might as well blame the LBSCR because if they had not built the railway through Cowden, the accident would certainly not have happened.

 

[MarkyT]

That is silly. Single Lines have as good a safety record as double track. You might as well blame the LBSCR because if they had not built the railway through Cowden, the accident would certainly not have happened.

I blame Stephenson and his ilk. If they hadn't developed mining tramways into fully-fledged modern general-purpose steam railways in the early 19th century, non of this could have happened.

 

[tpfx89]

Singling created the opportunity for conflicting movements, clearly not responsible for the actual accident taking place. To add my own semi serious suggestiin, might as well blame the foundryman casting the rail itself. Single track railways are a safe concept when managed safely.

 

[21C101]

Singling created the opportunity for conflicting movements, clearly not responsible for the actual accident taking place. To add my own semi serious suggestiin, might as well blame the foundryman casting the rail itself. Single track railways are a safe concept when managed safely.

A better debate might be was it safe to abolish the use of physical tokens or staffs having to be carried by drivers on single line routes without a secure radio system being in place.

 

[Recessio]

I thought a contributing factor was also someone else (a guard, I think I recall?) was in the cab and possibly driving when they weren't meant to be? Could have distracted the driver/been a mistake by an untrained driver

 

[Ianno87]

Single lines operate perfectly safely all around the world. It was several other factors at play that day that specifically conspired to make the Single line at Cowden unsafe.

 

[DPQ]

Those pesky monkeys. If they hadn't have evolved into Humans...

 

[MarkyT]

Singling created the opportunity for conflicting movements, clearly not responsible for the actual accident taking place. To add my own semi serious suggestiin, might as well blame the foundryman casting the rail itself. Single track railways are a safe concept when managed safely.

For a lightly used route, singling may have been the only realistic way to keep it open economically. If it had closed instead, then clearly no accident could ever have occurred, although of course people driving to more distant stations for their journey could easily have generated more cumulative deaths than occurred in the incident. The Cowden inquiry gave extra impetus to completing the full cab secure radio scheme around the former NSE area, which had been underway for years. Tragically, the signaller could see clearly the impending collision many minutes beforehand but had no means of communication whatsoever to intervene.

 

[bramling]

Single lines operate perfectly safely all around the world. It was several other factors at play that day that specifically conspired to make the Single line at Cowden unsafe.

This is a can of worms firmly opened.

Firstly, the immediate causes:
1) Signal illumination was questionable
2) Was the AWS working?
3) Was the driver distracted?
4) Who was actually driving the train?

*None* of these questions were ever really definitively answered, which is quite unthinkable today, where there would be numerous pieces of evidence like on-train monitoring, CCTV. Also I seem to remember the signal head wasn't secured for evidence preservation after the accident either.

So before we even start to consider the layout of the railway as it was in 1994, there's plenty of other factors open to question. For balance, it's also worth adding that all these factors were fairly speculative, the only one which wasn't pretty circumstantial was (3).

Then there's the fact that the signal layout wasn't really ideal, with a station sited in between repeater and stop signals. Not in itself intrinsically unsafe, but hardly desirable. The layout also, IIRC, had a slightly longer than required distance between the two signals. OD58 is nicely visible from Ashurst platform normally, but on this day there was thick fog. Did anyone consider that possibility?

So was the layout safe? In the end the answer is no since the accident happened. My personal opinion is that, in the absence of TPWS, the layout would have been better with trap points, and likewise would have been better with the repeater signal in advance of Ashurst platform, not in the rear of it. Trap points aren't a panacea either as they would need to have been designed specifically not to derail an errant train into the path of another.

I'm not sure it boils down to cost cutting, but clearly the layout wasn't ideal.

I do wonder how much BR saved from the singling though, not least having introduced several junctions and sets of points which the line wouldn't have otherwise needed. I suspect Uckfield will probably be redoubled in my lifetime.

 

[D6130]

That is silly. Single Lines have as good a safety record as double track. You might as well blame the LBSCR because if they had not built the railway through Cowden, the accident would certainly not have ha

A better debate might be was it safe to abolish the use of physical tokens or staffs having to be carried by drivers on single line routes without a secure radio system being in place.

Agreed....but one of the other fundamental principles of safe single line railways the world over is the provision of trap points or derailer blocks at the exit from crossing loops and double line sections which, while they can lead to inconvenient low-speed derailments, absolutely prevent head-on collisions. The fact that these were not provided in the Uckfield line singling scheme was undoubtably a result of BR's penny-pinching in the 1980s - but that policy was imposed on them by the government of the day. The singling scheme in itself was designed to prevent the need for expenditure on major track renewals and not for operating convenience.

 

[Gloster]

I always understood that the savings were marginal: less track mileage to maintain but more use of the remaining track and more points and signalling. It was one of the lines that suffered because ignorant politicians (and others) demanded to see evidence of savings.

 

[D6130]

When I travelled on the line fairly regularly in the 1980s, most of the track South of Hurst Green was very old and worn jointed bull-head rail on rapidly rotting wooden sleepers, that urgently needed renewal. I believe that the major savings were made by only having to renew one track rather than two on much of the route.

 

[d9009alycidon]

Cost cutting and over rationalisation was also blamed for the Bellgrove and Newton crashes

 

[30907]

Agreed....but one of the other fundamental principles of safe single line railways the world over is the provision of trap points or derailer blocks at the exit from crossing loops and double line sections which, while they can lead to inconvenient low-speed derailments, absolutely prevent head-on collisions.

Can you give an example? It is sometimes done but doesn't seem to be routine practice on NR or heritage lines.

 

[edwin_m]

Agreed....but one of the other fundamental principles of safe single line railways the world over is the provision of trap points or derailer blocks at the exit from crossing loops and double line sections which, while they can lead to inconvenient low-speed derailments, absolutely prevent head-on collisions. The fact that these were not provided in the Uckfield line singling scheme was undoubtably a result of BR's penny-pinching in the 1980s - but that policy was imposed on them by the government of the day. The singling scheme in itself was designed to prevent the need for expenditure on major track renewals and not for operating convenience.

Trap points generally aren't provided on passenger lines, I believe because the risk of derailing when there is no other hazard is usually considered to outweigh the reduced risk of collision. Even at the start of a single line they are rare - I can only think of Dorchester - there must be others but surely not many.

 

[Taunton]

That is silly. Single Lines have as good a safety record as double track.

This was not the case with the various 1970s-80s excessive singling projects - where I include single-lead junctions and single-lead crossovers as well, the same concept. I can readily think of half a dozen or more head-on collisions which happened on them within a short number of years of them being introduced - the Glasgow area seemed to be particularly prone to them. Nowadays they not only don't get done, but quite a number have been stripped out.

 

[Ianno87]

This was not the case with the various 1970s-80s excessive singling projects - where I include single-lead junctions and single-lead crossovers as well, the same concept. I can readily think of half a dozen or more head-on collisions which happened on them within a short number of years of them being introduced - the Glasgow area seemed to be particularly prone to them. Nowadays they not only don't get done, but quite a number have been stripped out.

Although that is in part due to lack of other mitigations to manage the risk. E.g. the risk of a single lead junction (and single lines) is considerably reduced by TPWS etc.

 

[The Ham]

Those pesky monkeys. If they hadn't have evolved into Humans...

I blame the amphibians leaving the sea!

 

[Ianno87]

More seriously, any movement of a rail vehicle involves some level of risk. It's how that risk is managed and mitigated that is key.

 

[tpfx89]

Out of potential interest people might have, just wanted to point out that I covered this one in episode 8 of my podcast at the start of this year.

Episode 8 - Cowden - Signals to Danger

2 Passenger trains collide head-on a rural single line. Five… Read more

signalstodanger.com

 

[D6130]

Can you give an example? It is sometimes done but doesn't seem to be routine practice on NR or heritage lines.

Probably the classic example, which is much photographed and includes a long sand drag due to the steep falling gradient, is at Goathland on the NYMR. Other than that, I'm sure that I can recall from the days of my mis-spent youth seeing trap points at the exits from several of the crossing loops on single lines in the Scottish Highlands. (Perhaps @Cheshire Scot could confirm or deny?). In my early railway career on the Southern Region, I seem to remember that the Botley-Fareham section - singled in the early 1970s - was protected by trap points at both ends.

 

[30907]

Probably the classic example, which is much photographed and includes a long sand drag due to the steep falling gradient, is at Goathland on the NYMR.

Thanks. That one had occurred to me, and the 1 in 49 will be a factor (it also helps that the line was historically double track!).
Another IIRC was Royal Albert Bridge Junction (on the Plymouth side).

And (a bit OT) trap points on running loops were very common in the days of unfitted freights - I still recall the sight of an N 2-6-0 that had overrun the end of the old Up Local at Shortlands Jn c.1958.

In my early railway career on the Southern Region, I seem to remember that the Botley-Fareham section - singled in the early 1970s - was protected by trap points at both ends.

It isn't now according to satellite views.

Back OT - I agree the (minimal) signalling at Cowden was a factor.

 

[Cheshire Scot]

Probably the classic example, which is much photographed and includes a long sand drag due to the steep falling gradient, is at Goathland on the NYMR. Other than that, I'm sure that I can recall from the days of my mis-spent youth seeing trap points at the exits from several of the crossing loops on single lines in the Scottish Highlands. (Perhaps @Cheshire Scot could confirm or deny?). In my early railway career on the Southern Region, I seem to remember that the Botley-Fareham section - singled in the early 1970s - was protected by trap points at both ends.

That has a familiar ring to it, at locations approached on a significant rising gradient, although in some cases - e.g. Tyndrum Upper, such protection was provided by the siding entry points. Siding exit points were all trap protected.

 

[edwin_m]

several of the crossing loops on single lines in the Scottish Highlands.

Here's one doing its job: RAIB_Carrbridge2010.pdf (railwaysarchive.co.uk)

At 16:04 hours on 4 January 2010, train 4N47, the late running 13:14 hrs Inverness to Mossend Yard1 service, derailed on exiting the run-out located at Carrbridge station, Badenoch and Strathspey, Scotland (see figure 1), having previously passed over trap points located within the station’s passing loop.
...
Once over the summit the line has a steep descent (1 in 60 maximum slope) towards Carrbridge where the gradient flattens out on the approach to and through the station
...
Because of the risks posed by the steep descent which leads into the station from Slochd summit ... the passing loop is fitted with trap points (116 points) which can be set either to allow trains to enter into the loop or on to a 27 metre long run-out

The cause was brake failure due to ice. The wording strongly implies that the traps and run-out were only provided because of the gradient.

That has a familiar ring to it, at locations approached on a significant rising gradient, although in some cases - e.g. Tyndrum Upper, such protection was provided by the siding entry points. Siding exit points were all trap protected.

Because of the hazard of unbraked vehicles rolling towards the main line (as seen recently in Wakefield), sidings will always be provided with trap points, unless there is a switch to a headshunt or similar that serves the same purpose. Either will be worked and detected with the main line turnout, so they always trap when a train isn't signalled in or out and never when one is, and in power signalling they will have an auto-restore or a reminder function so they get restored to normal (=trapping) setting when the move is complete.

 

[norbitonflyer]

And (a bit OT) trap points on running loops were very common in the days of unfitted freights - I still recall the sight of an N 2-6-0 that had overrun the end of the old Up Local at Shortlands Jn c.1958.

Trap points prevented the Norton Fitzwarren disaster of 1940 from being even worse than it was. (Train leaving Taunton ran through trap points at the end if the four track section because the driver thought he was on the adjacent track and was reading the wrong signals, which were clear for an overtaking train. Serious as the derailment was (27 deaths), the trap points prevented an even worse high speed sidelong collision with the other, faster, train, which had come alongside a few seconds before the first train reached the trap points

 

[edwin_m]

Trap points prevented the Norton Fitzwarren disaster of 1940 from being even worse than it was. (Train leaving Taunton ran through trap points at the end if the four track section because the driver thought he was on the adjacent track and was reading the wrong signals, which were clear for an overtaking train. Serious as the derailment was (27 deaths), the trap points prevented an even worse high speed sidelong collision with the other, faster, train, which had come alongside a few seconds before the first train reached the trap points

Don't have time to check right now but I think the train at Norton Fitzwarren was spread across all the tracks when the engine ploughed into the dirt in the traps, and the last van of the passing train was hit by flying ballast as it got clear just in time. So if the other train had been a few seconds later it would presumably have been a much bigger disaster - traps or no traps.

 

[Taunton]

The "trap" points at Norton Fitzwarren in 1940 were little more than a carriage length beyond the crossover to the main line, and a bufferstop. The layout was relatively new, the 4-tracking having been installed in 1933. There had been previous comment that the overrun provision was insufficient, the buffers being hard up against a stream which ran under the line, for freight trains were commonly brought to a stand there waiting for a path westwards. The subsequent big sand drag was installed after the accident. I've written here before about all the other measures then put in place there to prevent a recurrence.

The damage to the overtaking newspaper train was to several of the rear vehicles, broken windows etc, such that the guard pulled it up at the next box west, Victory Crossing. As the derailed King locomotive ended up sideways across all tracks, it wasn't even a few seconds that separated them.

 

[whoosh]

This is a can of worms firmly opened.

Firstly, the immediate causes:
1) Signal illumination was questionable
2) Was the AWS working?
3) Was the driver distracted?
4) Who was actually driving the train?

*None* of these questions were ever really definitively answered, which is quite unthinkable today, where there would be numerous pieces of evidence like on-train monitoring, CCTV. Also I seem to remember the signal head wasn't secured for evidence preservation after the accident either.

Covered in dead flies I believe. In foggy weather. But the dead fly issue affecting the illumination seemed to be glossed over.