Signal failure around Clapham Junction area affecting South West Trains [04/08]

[infobleep]

South West Trains aren't having much luck lately. Another signalling fault, this time at Clapham Junction or between their and Wimbeldon. Again happened towards the end of the morning peak. Also signalling problems around Fareham this morning.

Whether these will cause problems for the rest of the day, due to trains and crew being out of place, only time will tell.

I was on the 8.20 from Guildford which was diverted to the fast lines after Surbiton and missed out all intermediate stops. Passengers were advised to board or stay on board and then catch a train out of Waterloo. Never heard that type of advice before so must not be many if any trains are going north of Surbiton on the slow lines.

 

[infobleep]

Problem now fixed according to Twitter. It's a track circuit failure on the London bound slow line. South West Trains notice: [http://m.journeycheck.com/southwest...54&webMsg=false&engWorksPeriod=&serviceTime=]

Cancellations to services between Wimbledon and Clapham Junction*

Owing to signalling problems between Wimbledon and Clapham Junction fewer trains are able to run on London bound suburban lines.*

Impact:*
Train services running to and from these stations may be cancelled, delayed by up to 20 mins or revised at short notice. Disruption is expected until 11:00 04/08.*

Customer Advice:*
A signalling problem (track circuit failure) on one of our main lines into London Waterloo through Clapham Junction is affecting trains this morning.*
.*
Our engineers will be on site at 09:00 to look into the problem affecting our use of the line. Trains that use platform 10 at Clapham Junction will be running very slowly through this platform as we are using a method of talking past signals (two) to move trains through this area.*
.*
We need to limit the number of trains through this area into London Waterloo, the following train alterations will commence as a result of the incident:*
.*
London Waterloo to Dorking trains will be cancelled in both directions - The 0954 London Waterloo to Dorking will be the first to run from London Waterloo.*
London Waterloo to Chessington Trains will be cancelled in both directions - The 1016 London Waterloo to Chessington South train will be the first to run from London Waterloo.*
London Waterloo to Hampton Court Trains will be cancelled in both directions - The 1006 London Waterloo to Hampton Court will be the first to run from London Waterloo.*
.*
Guildford to London Waterloo trains will call additionally at Berrylands and run fast from Berrylands to Waterloo.*
Woking to London Waterloo trains will be run fast from Surbiton to Waterloo.*
Shepperton to London Waterloo trains will be run via Richmond.*
.*
Alternative travel arrangements:*
.*
Customers with South West Trains can use London Underground, London Buses via any reasonable route.*
First Capital Connect trains from Wimbledon via any reasonable route.*
Trams London from Wimbledon via any reasonable route.*
First Great Western trains via any reasonable route.*
Southern Trains via Dorking from Clapham Junction via any reasonable route.*
We are sorry for the disruption to your train journey this morning.*

Message Received :04/08/2014 09:41

A note about the problems hasn't appeared on the Surbiton online departure board. It may not do now as they have fixed the problem. It's just residual delays.

 

[talldave]

What exactly is a signal failure? Broken wire? Bad connection? Blown fuse?

And what's the MTBF of "a signal" - that has the potential to fail?

It always sounds like a generic "another part of our badly maintained infrastructure has just died" excuse.

Surely if they were serviced/maintained/overhauled more frequently than the MTBF, then most failures wouldn't occur would they?

Simplistic, I know, but if airlines had as many failures as Network Rail, we'd have planes falling out of the sky on a daily basis.

 

[Lockwood]

To extend that example a bit... Track circuit failures aren't a train problem. So, in an airline example, it would be equivalent of the sky not working?

 

[hassaanhc]

To extend that example a bit... Track circuit failures aren't a train problem. So, in an airline example, it would be equivalent of the sky not working?

Probably the radar system failing

In today's case it was a Track Circuit failure, which detects whether a train is in that section or not. If it fails then it should show the section as occupied even when it isn't, so there is no danger of trains colliding. However you sometimes get a "wrong-side" failure, where the system has failed to detect that a train is there. These are incredibly dangerous, it was a type of wrong-side failure that lead to the Clapham 1988 disaster.

 

[Tomnick]

Of course, there's no single failure mode for a track circuit (or any other item whose failure might be described as 'a signal failure'). Anything from flooding, through the failure of a block joint to provide insulation between adjacent track circuits, to something conductive touching both rails (such as a signal wire - been there!) can cause a track circuit to show occupied when clear. Whether the current maintenance regime is sufficient (especially in the face of continued pressure to run the whole job as cheaply as possible), I'll leave others to decide - but I do know that it's well-structured and thorough, to the point (yes, they fail too) that they've now got equipment to monitor some bits of kit to help identify something that's showing early signs of premature expiration.

 

[infobleep]

I once saw a Panorama documentary on health and safety and monitoring that companies are complying in the industry. It didn't paint a good picture.

However in the rail industry I do think they are serious about health and safety these days, hence all this additional equipment to monitor for problems.

In terms of trains, they now seem to be running normally again. The 10.56 from Surbiton to Waterloo was put on the fast lines but was able to stop at Clapham Junction so well done. A signaller or a member of staff at South West Trains must have spotted a gap it could use to stop at Clapham Junction.

 

[hassaanhc]

I once saw a Panorama documentary on health and safety and monitoring that companies are complying in the industry. It didn't paint a good picture.

However in the rail industry I do think they are serious about health and safety these days, hence all this additional equipment to monitor for problems.

In terms of trains, they now seem to be running normally again. The 10.56 from Surbiton to Waterloo was put on the fast lines but was able to stop at Clapham Junction so well done. A signaller or a member of staff at South West Trains must have spotted a gap it could use to stop at Clapham Junction.

It helps that at Clapham Junction there are two platforms for the London bound fast lines

 

[swt_passenger]

What exactly is a signal failure?

When they say 'signal failure' in a public facing statement you should really read it as a 'signalling system failure' - where 'signalling system' can include almost anything whatsoever to do with the signals themselves, the controlling signalling centres, all the points and the train detection systems, and all the associated power supplies and communication systems.

Their typical customer isn't interested in the details though...

 

[carriageline]

Well I have heard customers moan that signal failures are too vague, if you turn around to them and say W1721A points no reverse detection due to faulty clutch drive, they will probably look at you like you have just explained brain surgery to them

 

[infobleep]

Well I have heard customers moan that signal failures are too vague, if you turn around to them and say W1721A points no reverse detection due to faulty clutch drive, they will probably look at you like you have just explained brain surgery to them

Well I think it's great that SWT said it was a track circuit failure. Yes I don't immediately have a clue what they are on about but I can look it up online and learn something. That's assuming I'm not stuck at a railway station or on a train with no O2 mobile reception of course. Lol!

 

[hassaanhc]

Well I have heard customers moan that signal failures are too vague, if you turn around to them and say W1721A points no reverse detection due to faulty clutch drive, they will probably look at you like you have just explained brain surgery to them

I'd prefer they straight up said Points Failure in that case rather than just Signal Failure

 

[infobleep]

It helps that at Clapham Junction there are two platforms for the London bound fast lines

Does two platforms allow them to run services closer together?

 

[AM9]

Does two platforms allow them to run services closer together?

Usually yes because the dwell time can overlap the approach of a succeeding and departure of a previous train.

 

[carriageline]

I'd prefer they straight up said Points Failure in that case rather than just Signal Failure

Fair enough, but customers don't necessarily know what points are. Plus, I doubt most of them care what the actual failure is, just how long it will delay them

 

[Yabbadabba]

I think just keep it simple such as, signal failure, points failure or signalling system failure, rather than "a TDM indication failure combined with a remote interlocking failure due to the national grid causing a power spike in the power supply to the signalling system" as in what happened at South Croydon relay room 19th June.

 

[swt_passenger]

Usually yes because the dwell time can overlap the approach of a succeeding and departure of a previous train.

That's correct in the general case, but IIRC the platform geometry at Clapham Jn doesn't lend itself to alternate use, firstly because P7 (the loop platform) has a much lower speed entry compared to P8, and secondly (and there's probably a better and much more detailed explanation) there's a 'reduced overlap' or something beyond the P7 starter signal which also prevents alternate use in the usual manner. What I think it means in practice is that you cannot have an arrival in P7 while a departure is taking place from P8.

 

[Tomnick]

Usually yes because the dwell time can overlap the approach of a succeeding and departure of a previous train.

I'm not sure whether that's the case here though - I think the outlet from the 'loop' platform falls within the overlap of the signal at the end of the UMF platform, meaning that a second train can't approach on the latter until the first has passed clear of the connections from the former. I don't know whether the opposite applies though, or whether the overlap can swing onto the ladder through the yard towards the Windsor side.

 

[Yabbadabba]

That's correct in the general case, but IIRC the platform geometry at Clapham Jn doesn't lend itself to alternate use, firstly because P7 (the loop platform) has a much lower speed entry compared to P8, and secondly (and there's probably a better and much more detailed explanation) there's a 'reduced overlap' or something beyond the P7 starter signal which also prevents alternate use in the usual manner. What I think it means in practice is that you cannot have an arrival in P7 while a departure is taking place from P8.

Looking at the panel diagram for Wimbledon panel 1 for Clapham Junction, there is a reduced overlap for platform 8 so you can depart platform 7 and arrive in platform 8 at the same time, though it will be an approached controlled signal into platform 8. But it doesn't look like you can depart from platform 8 and arrive in platform 7 at the same time as the overlap for platform 7 looks like a full overlap which swings back onto the Up Main through the very points that the departing train from platform 8 is using.

 

[swt_passenger]

Looking at the panel diagram for Wimbledon panel 1 for Clapham Junction, there is a reduced overlap for platform 8 so you can depart platform 7 and arrive in platform 8 at the same time, though it will be an approached controlled signal into platform 8. But it doesn't look like you can depart from platform 8 and arrive in platform 7 at the same time as the overlap for platform 7 looks like a full overlap which swings back onto the Up Main through the very points that the departing train from platform 8 is using.

Thanks, that sounds a better explanation than I had in the old memory banks. But I suppose it's reasonably correct to say that neither platform route is optimised for stopping services.

 

[455driver]

Plus the real reason which is that trains are not allowed to call at Platform 8 (except in an emergency) due to the massive step down because of the cant through the platform.

Re overlap from P7 effecting a departure from P8, I dont think this is the case because the points (and so the overlap) can be set into Clapham yard. I am not a signalling expert so there is probably some reason this isnt possible but I am not aware of it.

 

[Yabbadabba]

Plus the real reason which is that trains are not allowed to call at Platform 8 (except in an emergency) due to the massive step down because of the cant through the platform.

Re overlap from P7 effecting a departure from P8, I dont think this is the case because the points (and so the overlap) can be set into Clapham yard. I am not a signalling expert so there is probably some reason this isnt possible but I am not aware of it.

In general signalling standards change with time and it will depends what standards were in use at the time when the re-signalling was done. At my location Horsham platform 4 is the only place I know where the full overlap swings into a siding, which the best of a bad option and the approaching signals are all approached controlled and Gatwick Airport platform 6 now has a full overlap into the run off again with approach controlled signals, but this was after the re-signalling as before you could only get the full overlap when it could swing the points back onto the down fast line. All other locations at my ASC require either a reduced overlap with an approach controlled signal or a full overlap that requires the route away from the siding.

 

[infobleep]

Plus the real reason which is that trains are not allowed to call at Platform 8 (except in an emergency) due to the massive step down because of the cant through the platform.

Re overlap from P7 effecting a departure from P8, I dont think this is the case because the points (and so the overlap) can be set into Clapham yard. I am not a signalling expert so there is probably some reason this isnt possible but I am not aware of it.

I've only got out at platform 8 twice but don't remember it before no any worse than platform 17.

As for giving out information, they can just say a signalling failure, as they did in this case and then on their Web Site, give out further information as to what exactly it is. This is exactly what SWT did.

 

[Darren R]

Apologies for taking this off at a bit of a tangent, but it's made me wonder something.

When a track circuit failure occurs, just how do the engineers start to trace the source of it? As has been mentioned, there can be several reasons for such a failure, added to which some track circuit sections are quite lengthy - several miles in fact (although I'm guessing this isn't the case at Clapham Junction! )

Is it not sometimes like looking for the proverbial needle, or is there some way of knowing what to look for and where?

 

[Railsigns]

As has been mentioned, there can be several reasons for such a failure, added to which some track circuit sections are quite lengthy - several miles in fact

Track circuits are never several miles long. What might appear to the signalman as such will in reality comprise a number of separate track circuits indicated as one, or alternatively it could be a long axle counter section.

 

[moggie]

They ought to just announce that the TR defaulted to safe state operation while the cause is under investigation by rail technicians.

Normals would be none the wiser but it sounds much more positive that the term failure which in fact is a misnomer as the equipment has operated to its failsafe state as intended.

 

[infobleep]

They ought to just announce that the TR defaulted to safe state operation while the cause is under investigation by rail technicians.

Normals would be none the wiser but it sounds much more positive that the term failure which in fact is a misnomer as the equipment has operated to its failsafe state as intended.

Are there instances where track circuit will go to their failsafe mode but nothing has failed in terms of components or software?

 

[carriageline]

Not completely sure if I understand, but yes there are. For example, flooding, broken rails (you may constitute that as a failed component), coke cans bridging the gap etc. I have never heard of a track circuit showing occupied without reason though, that's not to say it hasn't happened!

What's worse is when a track circuit stays down when you expect it to be up (occupied, if I have them the right way!!) , 99% of the time that's just an indications fault

 

[moggie]

Are there instances where track circuit will go to their failsafe mode but nothing has failed in terms of components or software?

Indeed but it can vary with the style of Track Circuit equipment but temporary low ballast resistance due to a combination of wet weather and poor insulation between rail and sleeper and equipment set up can often cause intermittent faults to show, but will work quite happily when it dries out to damp rather than wet conditions.

As another mentioned a build up of contamination between adjacent rails at a block joint can also have the same effect yet the Track Circuit equipment itself is fault free.

There will be numerous scenarios which do not result from equipment faults per se'. Axle Counters remove many of these......but introduce their own bag of nails. But at least the signaller can press a 'button' or two and clear them again - yes really! <tin hat on>