Wrong side signal wiring fault

[martin_wynne]

This from RAIB today. Echoes of Clapham?

 

 https://www.gov.uk/government/news/wrong-side-signalling-failure-at-wingfield

 

Martin.

[Oldddudders]

25 minutes ago, martin_wynne said:

This from RAIB today. Echoes of Clapham?

Potentially, undoubtedly so. The same fatal false clear to the driver. Prompt reporting by the first driver and a suitable response from the signaller seem to have saved the day. Having once watched a track circuit go clear when I knew there was a light engine crossing it - at Fawkham Junction - the experience does give pause...... 

 

Clapham was caused by an overworked new-works tech not cutting off wire tails, which then shorted to give a false clear with catastrophic results in a busy area. We await detail of what went wrong here.

 

If I am doing wiring on the layout, I make a point of cutting off excess wire. No-one will die if I don't, of course, but one learns from others' ghastly experiences. 

[Bucoops]

Wasn't there another one recently too? Very concerning, but well done to both drivers and the signalman.

[Michael Hodgson]

Whilst I have my suspicions as the likely causes of this incident, I don't think speculation is helpful as it is currently under investigation.

[Bucoops]

31 minutes ago, Bucoops said:

Wasn't there another one recently too? Very concerning, but well done to both drivers and the signalman.

 

Ah, I was thinking of the same incident - an initial release made about it on 11th November.

[St. Simon]

31 minutes ago, Bucoops said:

Wasn't there another one recently too? Very concerning, but well done to both drivers and the signalman.

Hi,

 

There was a wrong side wiring failure a couple of years ago that stemmed from uncontrolled wiring revision back to a previous state as a result of cancelled works, whereby a track circuit was bypassed on the wiring.

 

For awareness, this has been already briefed out to Network Rail employees (and others I assume) and the appropriate actions have been put in place I believe.

 

Simon

[martin_wynne]

1 hour ago, Oldddudders said:

Prompt reporting by the first driver and a suitable response from the signaller seem to have saved the day.

 

Indeed. But which relied entirely on the in-cab radio system. Before that was installed, is/was there anything in the rules or signalling system to save the situation? There is no mention of putting down detonators. Is that still done?

 

Would it not be better for the first train in such a situation to proceed as far as possible beyond a faulty signal if the line is visibly clear, preferably beyond the next properly-working signal? 

 

Martin.

[Colin_McLeod]

I presume the driver of the first train would not know it was a faulty signal.  He got a red and stopped as quickly as he could and reported. 

 

Depending on what information he had, the signaller MIGHT be in a position to instruct the first train to proceed under caution and report again at the next signal.  (Maybe that's what you meant.)

 

[Michael Hodgson]

A radio call is a lot faster than putting down dets, and if a train is single manned, it takes a while to get to the back of a freight train before you can start walking a suitable distance to give rear end protection, and you can't go forward to protect the opposite line at the same time.

 

On track-circuited lines a track circuit clip was a faster means of protecting an adjacent line than dets/phone calls etc, but of course they don't work on lines Absolute Block lines that are not track circuited, nor on the increasingly common TCB-worked lines where axle counters are used rather than traditional circuits.

[martin_wynne]

7 minutes ago, Colin_McLeod said:

Maybe that's what you meant.

 

It was.

[Michael Hodgson]

1 minute ago, Colin_McLeod said:

I presume the driver of the first train would not know it was a faulty signal.  He got a red and stopped as quickly as he could and reported. 

 

Depending on what information he had, the signaller MIGHT be in a position to instruct the first train to proceed under caution and report again at the next signal.  (Maybe that's what you meant.)

 

Driver seeing a red when the preceding signal wasn't yellow must take the signal as indicating that the line has become obstructed in the meantime and that the signal has been put back to Danger in emergency.  It's technically a SPAD, but not his fault. 

 

The signaller would not know the line was clear, in most cases he could only know that whatever obstruction may or may not exist, it was not the previous train.  Whilst he should know the signal should have showing a proceed aspect, he would not know the previous signal had been showing green.  So the most likely cause of such a SPAD report by a driver would be that the driver had missed a Yellow (whatever he might say and genuinely believe) and that he had unexpectedly met a red caused by a track circuit failure .  Most "signalling failures" as announced at stations are actuallly track circuit failures, causing signals to "fail" safe to red, but that's exactly what they're supposed to do.  He's got to report the problem up through channels.

 

 

[phil-b259]

4 minutes ago, Michael Hodgson said:

So the most likely cause of such a SPAD report by a driver would be that the driver had missed a Yellow (whatever he might say and genuinely believe) and that he had unexpectedly met a red caused by a track circuit failure .

 

Actually the nature of most track circuit failures means that drivers can easily get a green onto a red situation - the failure having occurred after passing the green signal but before they reach the subsequent signal - which has reverted to red because of the failure.

[martin_wynne]

4 minutes ago, Michael Hodgson said:

he could only know that whatever obstruction may or may not exist, it was not the previous train.

 

Which doesn't prevent the first train from proceeding slowly as far as the line is visibly clear.

 

The signaller obviously did know that something was amiss, hence the call to the following train.

 

 

[The Stationmaster]

3 hours ago, martin_wynne said:

 

Indeed. But which relied entirely on the in-cab radio system. Before that was installed, is/was there anything in the rules or signalling system to save the situation? There is no mention of putting down detonators. Is that still done?

 

Would it not be better for the first train in such a situation to proceed as far as possible beyond a faulty signal if the line is visibly clear, preferably beyond the next properly-working signal? 

 

Martin.

Signal capable of showing a stop aspect so it would have had an SPT (or have been within walking distance of the controlling signal box).  Driver-to-shore radio (apart from the E button function) is no different from any other means of communication as it is exactly that - a means pf communication.  The E button is in many respects a modern, and far quicker/potentially more effective, equivalent of detonator and hand signal protection.

 

But in this case all the Driver of the first train knew was that he approached a signal at danger when he had insufficient distance to halt his train; the Driver didn't know why the signal was at danger and wouldn't know until he contacted the controlling signal box. 

 

It is at this stage pointless to discuss what happened next because we don't know the full details (and we won't for some time) apart from the fact that Driver of the second train approached a signal that was incorrectly showing a proceed aspect when the section in advance of it which it was supposed to protect was apparently occupied by another train.  Beyond that there are various possible reasons why things went wrong and the purpose of the investigation is to establish what happened, why it happened, and ensure that a recurrence is avoided by  issuing either amended procedures or whatever other means necessary to properly enforce observance of existing procedures.

[TheSignalEngineer]

I have no detailed knowledge and won't speculate, but I have feared for some time that the industry is teetering on the edge of another Clapham. That weekend I had been in charge of a section of a major commissioning. I spent Friday night supervising a changeover interfacing new wiring to existing equipment. It didn't go well so after breakfast and a shower I returned to site for a conference with the design team to explain what was going wrong and agree the modifications necessary. I rearranged the Saturday night work programme diverting some experienced staff to a scratch team to do the modifications. 

Back to the hotel for a meal, I changed into my outdoor gear and returned to site to lead the modification team myself. Fortunately all went well and we completed the outstanding work by breakfast on Sunday so a few hours sleep was in order.

Returning in the evening I took up my booked Sunday night work checking sighting and operation of equipment from the test train before signing in my part of the job at the start of Monday's rush hour. At breakfast I got a message to say that there had been a serious incident at Clapham. When I got back to the site office there were details coming through and I had a tip as to what the cause had been so rounded up some volunteers to go through the locations we had been in to check we hadn't done the same thing. I finally arrived home at Monday teatime, lay down and promptly fell asleep where I remained for 17 hours. 

Such working was common then. I had been away from home for 82 hours and had about 10 hours sleep.

My fear is that everything goes round in circles and managers without personal experience of attending a crash scene are too quick to destroy the things that were put in place by those who had to pick up the pieces in the past.

 

Edited December 21, 2022 by TheSignalEngineer

[The Stationmaster]

17 minutes ago, TheSignalEngineer said:

I have no detailed knowledge and won't speculate, but I have feared for some time that the industry is teetering on the edge of another Clapham. That weekend I had been in charge of a section of a major commissioning. I spent Friday night supervising a changeover interfacing new wiring to existing equipment. It didn't go well so after breakfast and a shower I returned to site for a conference with the design team to explain what was going wrong and agree the modifications necessary. I rearranged the Saturday night work programme diverting some experienced staff to a scratch team to do the modifications. 

Back to the hotel for a meal, I changed into my outdoor gear and returned to site to lead the modification team myself. Fortunately all went well and we completed the outstanding work by breakfast on Sunday so a few hours sleep was in order.

Returning in the evening I took up my booked Sunday night work checking sighting and operation of equipment from the test train before signing in my part of the job at the sart of Monday's rush hour. At breakfast I got a message to say that there had been a serious incident at Clapham. When I got back to the site office there were details coming through and I had a tip as to what the cause had been so rounded up some volunteers to go through the locations we had been in to check we hadn't done the same thing. I finally arrived home at Monday teatime, lay down and promptly fell asleep where I remained for 17 hours. 

Such working was common then. I had been away from home for 82 hours and had about 10 hours sleep.

My fear is that everything goes round in circles and managers without personal experience of attending a crash scene are too quick to destroy the things that were put in place by those who had to pick up the pieces in the past.

 

Oddly also relating to Clapham I often wonder to what extent some TOCs (and NR) apply Hidden 18 monitoring when we hear stories about the extent to which they are relying on traincrew - Drivers in particular- working Rest Days and/or where voluntary Sunday working has been introduced.  the pressure on rostering staff to cover turns in order to keep trains running will inevitably be considerable as will the pressure be on the 'willing' (or 'grabbing') hands to get in some extra days outside their roster.  The basic roster will always be the first line of compliance when respecting Hidden 18 and once you start losing that idea by relying on Rest Day working and using volunteer coverage for Sunday turns you are getting ever closer to forgetting what matters.  Generally BR also applied the recommendation to rest intervals and number of days permitted to be worked continuously by regarding 'hours' as meaning total periods of hours worked rather than just specific days etc.

 

[Bucoops]

The NHS monitors and restricts bank shifts. Even for non-critical positions (my Wife is a Ward Clerk and does quite a few extra shifts but she regularly hits the hard limit on hours). Not so much at the time of the report, but with modern computerised systems it should be simple to flag excessive hours.

[phil-b259]

20 minutes ago, The Stationmaster said:

Oddly also relating to Clapham I often wonder to what extent some TOCs (and NR) apply Hidden 18 monitoring when we hear stories about the extent to which they are relying on traincrew - Drivers in particular- working Rest Days and/or where voluntary Sunday working has been introduced.  the pressure on rostering staff to cover turns in order to keep trains running will inevitably be considerable as will the pressure be on the 'willing' (or 'grabbing') hands to get in some extra days outside their roster.  The basic roster will always be the first line of compliance when respecting Hidden 18 and once you start losing that idea by relying on Rest Day working and using volunteer coverage for Sunday turns you are getting ever closer to forgetting what matters.  Generally BR also applied the recommendation to rest intervals and number of days permitted to be worked continuously by regarding 'hours' as meaning total periods of hours worked rather than just specific days etc.

 

 

Down our way the company go beyond Hidden and put a maximum limit of 60 hours per week (Hidden put in place a ceiling of 80) but in other respects follows the 'Hidden rules' as regards minimum rest periods etc

 

However this is less about managing fatigue and more about saving money -  with the result that shifts are left uncovered and / or teams short staffed, thus putting more pressure on the remaining team members.

 

Tired and demoralised staff are far more likely to make mistakes or give in to pressure when up against it. Its all very well the suits saying they have all these wonderful procedures - they are not the ones out in the freezing cold/ pouring rail / baking sun with everyone and their mother ringing up wanting to know when things will be done.

 

Its all very well NR producing lots of fancy e-learnings / powerpoint presentations but that does **** all to address the real issues.

 

Of course the real answer is to actually employ enough people in the first place - yet NR seem hell bent on slashing maintenance activities to the bare minimum and redeploying staff instead!

Edited December 21, 2022 by phil-b259

[TheSignalEngineer]

1 hour ago, phil-b259 said:

Of course the real answer is to actually employ enough people in the first place.....

They also seem hell-bent on replacing established permanent staff with agency people. That makes it more difficult to track hours worked and rest periods. I caught people coming off one job, sleeping in the van for a couple of hours then working for someone else. They were on the books of more than one agency with no communication between them in those days. I don't know if the current systems are foolproof, but in the early days of privatisation we demonstrated that the Railtrack ones were not good.

In my BR days the only agency staff in my depot were an HGV driver who was purely for delivering material / collecting scrap and a clerk/typist who covered for leave and sickness in the office.

Later as a Project Engineer although I my on-site qualifications I only used the staff safety ones enough to keep my ticket valid. I gave up that part in my last role but still took the test to keep my knowledge up for managment purposes as there were times I needed to review associated documents and write technical material for tender submissions. 

If I had to go on-track it was usually in a group with an agency PICOW or the like. I was appalled at the lack of knowledge, both in Rules and railway geography,  show by some of the staff who turned up and blackballed two firms who failed audits by our Safety Engineer following my reports. 

[Michael Hodgson]

On 21/12/2022 at 13:09, martin_wynne said:

 

Which doesn't prevent the first train from proceeding slowly as far as the line is visibly clear.

 

The signaller obviously did know that something was amiss, hence the call to the following train.

 

 

A broken rail is one possible cause of such a fault causing a traditional track circuit to give a false occupied status.  You wouldn't necessarily expect to see that from the cab, even in good daylight. 

 

The simple track circuit is a fail safe design; it puts a low voltage battery feed across the rails at one end of the section being monitored with a relay at the opposite end.  An axle in the section offers a path of much lower resistance than the relay coil, causing the relay to drop out.  Discontinuity of supply from battery failure or through a broken rail can also cause the relay to drop.  Modern track circuits tends to be rather more complex.

[St. Simon]

2 minutes ago, Michael Hodgson said:

A broken rail is one possible cause of such a fault causing a traditional track circuit to give a false occupied status.  You wouldn't necessarily expect to see that from the cab, even in good daylight. 

 

The simple track circuit is a fail safe design; it puts a low voltage battery feed across the rails at one end of the section being monitored with a relay at the opposite end.  An axle in the section offers a path of much lower resistance than the relay coil, causing the relay to drop out.  Discontinuity of supply from battery failure or through a broken rail can also cause the relay to drop.  Modern track circuits tends to be rather more complex.

 

Hi,

 

Off Topic I know, but I think that is a myth, a broken rail can cause a track circuit failure in the right conditions, but it won't definitely cause a Track Circuit Fault. It is just as common, if not more common, that broken rails are detected via patrols or reports rather than a Track Circuit Failure. The reality is that it needs to be a clean 'wide' break in dry conditions to cause a track relay to drop. I think the 'rail break' argument taken out of any Track Circuit conversation

 

Equally, it's not true that a Track Circuit is Fail Safe. Not all trains operate Track Circuits reliably, On Track Plant being one such type, and there are conditions (where the vehicle axle does not present the least resistive circuit) that allow the relay to be picked whilst a train is on the circuit. So, it is only fail safe in most of the operating conditions that experiences, not totally fail safe.

 

Simon

[Michael Hodgson]

Yes, and you also can't rely on on track circuits to detect rubber tyred vehicles.

[Colin_McLeod]

No traction tyres then. 😁

[The Stationmaster]

And you can't rely on axle counters to detect anything unless it physically  passes over one of the counters.

[St. Simon]

8 hours ago, The Stationmaster said:

And you can't rely on axle counters to detect anything unless it physically  passes over one of the counters.

 

Quite right!

 

The more you delve into signalling, the more you realise that very few bits of it are truly 'fail safe', instead the key is that it is the system as a whole that is fail safe.

 

Simon

Edited December 24, 2022 by St. Simon