Urgent Safety Message from RAIB

[martin_wynne]

 

 https://www.gov.uk/government/publications/urgent-safety-advice-022023-overspeeding-through-spital-junction/

 

 

Edited May 30, 2023 by martin_wynne
shorten title

[The Stationmaster]

I often used to wonder about the change to the Signalling Principles which allowed a splitting colour light to be sited up to half a mile in rear of the point of divergence, ex zpevcially when slow speed turnouts are involved.  In this case it looks on the face of it like poor design for such a slow diverging speed.  

 

But I also wonder about the sighting of the signal (or the competence of the Drivers involved) if two Drivers failed to notice the Ji illuminated).   The interesting thing is that while Grand Central has been around for quite a while Lumo is a fairly recent arrival on the scene and their Driver standards management surely needs to be very carefully looked at in the light of this incident involving their train?  The same of course might be said of Grand Central but they certainly used to be a competently managed concern.

[Wickham Green too]

1 hour ago, The Stationmaster said:

... Grand Central has been around for quite a while Lumo is a fairly recent arrival on the scene and their Driver standards management surely needs to be very carefully looked at ...

Chances are the drivers concerned have been around longer than Lumo has, at least !

[keefer]

It's all a bit ambiguous as to why the incidents happened.

As Mike said is it a sighting issue or driver comprehension (signal/route knowledge) issue?

Even explaining the two cases doesn't really give a clue - what is the official definition of 'to observe' a signal? Is it simply to 'see' it or is it to 'see and obey'? (Although the latter may be covered by 'and react to').

The way it's written to me suggest the drivers didn't know what a JI is and just assumed they'd continue straight on like they always do. (which obviously may not be the case - it's just the way it sounds to me).

@The Stationmaster how do you find the wording of these more modern documents, compared to the older DoT Official Reports etc.? I get the feeling they're trying to make them easier to understand by using more 'everyday' language, which can make them less 'precise'. Whereas the older reports may have been slightly more technical but were unambiguous as everything was simply and clearly presented and defined.

What kind of trains were involved? i believe Lumo use the class 8xx units, so they are relatively rigidly-coupled which could mitigate coaches tipping over. Mk4 sets may also do better than normal LH-stock as they have a Tightlock-style buckeye arrangement within rakes (prong and socket each side, which prevents rotation about the coupler axis).

Rather concerning anyway, in that it involves high-speed trains at a major station on the ECML (as opposed to, say, a single-line branch off a secondary route).

 

Edited May 26, 2023 by keefer

[Flying Pig]

17 minutes ago, keefer said:

The way it's written to me suggest the drivers didn't know what a JI is and just assumed they'd continue straight on like they always do. (which obviously may not be the case - it's just the way it sounds to me).

 

It says "did not observe and react to" which suggests that drivers with an expectation of remaining on the Up Fast may see what they expect - i.e. they do not perceive the lit junction indicator - and drive according to what they believe they have seen.

[Edwin_m]

9 hours ago, Flying Pig said:

 

It says "did not observe and react to" which suggests that drivers with an expectation of remaining on the Up Fast may see what they expect - i.e. they do not perceive the lit junction indicator - and drive according to what they believe they have seen.

This Safety Advice is issued because their investigation so far suggests there is a significant risk of another incident at this location (unsurprising as there have been two already) which warrants a warning being put out.  The investigation continues, and the final report may shed more light on whether it was failure to observe or failure to react, but for now they will deliberately have left this vague. 

[phil-b259]

A simple (ish) technical solution is one (or more) TPWS overspead sensors between the signal and the junction which are energised when the route indicator is lit.

 

Moving the signal closer to the point of divergence, although superficially sounding easy is likely to be more technically complicated.

[Mallard60022]

I remember the media reporting this (near Peterborough I believe). Social media 'onboard' output created sensational reports of course, BUT it did sound pretty frightening.   

P

[The Stationmaster]

2 hours ago, Edwin_m said:

This Safety Advice is issued because their investigation so far suggests there is a significant risk of another incident at this location (unsurprising as there have been two already) which warrants a warning being put out.  The investigation continues, and the final report may shed more light on whether it was failure to observe or failure to react, but for now they will deliberately have left this vague. 

It raises, or should raise, a whole host of issues.  

 

There might be a simple and obvious problem with signal sighting - is the JI obscured by lineside junge for instance? Apart from the impact of the Signalling Principle (I don't know if it has been revised since the half mile relaxation was added) there is the whole issue of signal sequences and aspects as well as the reasons the signalling was designed in taht way - as Phil implies  the siting of this signal will be down to a lot more than this particular divergence.  There is a least one similar situation on the Berks & Hants l Line installed under teh Reading panel Extension schme which took advantage of the relaxation to allow 3 aspect signalling to be used.

 

But there is another - in my view far more serious matter - and that is the failure of two Drivers to properly observe and react to the indication given by thr signal.  If there is no sighting issue then the obvious question which needs to be asked here is whether or not these are the first two instances of this happening or is there a longer term history of such errors at a signal which seemingly has been there for quite a time.   The answers to that question could be critical with some very important implications starting with the number of times passenger trains get a JI there (has it increased due to the number of extra trains on the route?), or is it a training and monitoring of Drivers shortcoming on the part of the operators.

 

It will take some time to get through all that.

 

Coming back to Keefer's question I think we do see more use of 'simpler' language in todays' Reports although as they are aimed at people within the industry that shouldn't really be necessary.  But Drivers have been required to 'observe' signals for umpteen years and we're now in the third century in which that requirement has applied - and 'observe' in that context means 'look out for and comply with'.  

 

The big change iin Reports has been to move from establishing the mechanics of what happened to addressing in far greater detail the background reasons which led to any shortcomings.  Thus organisation and process are now more clearly addressed (although they always have been addressed) with far more specific references to individual documents etc.  However just because a dcument exists it doesn't mean that everybody knows every detail of it but they should understand, and apply, its meaning.  

 

Regrettably the involvement of the legal trade actively participating in public Inquiries about things they clearly don't undersrand might have a bit of responsibiluty there.  For example some downright ridiculous things were said, or questions asked, by barristers at both the Southall collision and Ladbroke Grove iInquiries - to no useful effect whatsoever in establishing the causes of these incidents or how they might be avoided in future.  I'm definitely not comparing RAIB with that sort of thing but the more active involvement of the legal trade might well bear some responsibility for the need to 'dumb down technical matters.  And in any case it is of course very much the situation in the wider media (and they read RAIB Reports of course, and the spout things from them but in their own version).

[phil-b259]

2 minutes ago, The Stationmaster said:

 

 

But there is another - in my view far more serious matter - and that is the failure of two Drivers to properly observe and react to the indication given by thr signal.  If there is no sighting issue then the obvious question which needs to be asked here is whether or not these are the first two instances of this happening or is there a longer term history of such errors at a signal which seemingly has been there for quite a time.   The answers to that question could be critical with some very important implications starting with the number of times passenger trains get a JI there (has it increased due to the number of extra trains on the route?), or is it a training and monitoring of Drivers shortcoming on the part of the operators.

 

 

It might be noteworthy that ordinarily Grand Central and Lumo don't stop at Peterborough and as such might we not be looking at a case of the mind seeing what it wants to see which then means their brain says they are remaining on the fast line.

 

 

[Michael Hodgson]

8 minutes ago, The Stationmaster said:

Coming back to Keefer's question I think we do see more use of 'simpler' language in todays' Reports although as they are aimed at people within the industry that shouldn't really be necessary. 

 

I agree that it's the industry that has to take notice of from these reports and act on their contents, and they should all understand railway-speak, but the old blue accident reports were always formally addressed to a politician who would be worried about being asked a Question in the House, and the Inspectorate were responding  to an instruction from him.  So reports should always have been worded in such a way that the layman or lobby correspondent can understand. 

 

This has been stylistically formalised by starting with a summary or management overview, and a glossary of ralway jargon is now standard as an appendix.  On the whole I think this is positive, although once you seen a few of these, you really don't need the boilerplate about not seeking to apportion blame etc.  No doubt all that guff is there because of the involvement of the ignorant lawyers you referred to.

[62613]

22 minutes ago, phil-b259 said:

 

It might be noteworthy that ordinarily Grand Central and Lumo don't stop at Peterborough and as such might we not be looking at a case of the mind seeing what it wants to see which then means their brain says they are remaining on the fast line.

 

 

That seems to be fairly common in accidents, particularly derailments; "I wasn't usually routed that way", or "I didn't expect that signal to be at danger" comes up in quite a few accident reports. These remind me a bit of the Bourne End crash, just after WW2, where a train was due to be re - routed over a double junction, due to p.w. work in Watford Tunnels; although in that case, there seems to have been some confusion about what a particular signal indication meant, and whether the driver actually saw the signal because of the lighting conditions.

Edited May 27, 2023 by 62613

[The Stationmaster]

On 27/05/2023 at 13:28, 62613 said:

That seems to be fairly common in accidents, particularly derailments; "I wasn't usually routed that way", or "I didn't expect that signal to be at danger" comes up in quite a few accident reports. These remind me a bit of the Bourne End crash, just after WW2, where a train was due to be re - routed over a double junction, due to p.w. work in Watford Tunnels; although in that case, there seems to have been some confusion about what a particular signal indication meant, and whether the driver actually saw the signal because of the lighting conditions.

But it all  comes back to road knowledge on this sort of situation and that takes us back y traininhg, management systems, and supervision.  All of which take us back, once more, to the 'lessons learned' from the Ladbroke Grove collision which came down to being caused by a Driver who having passed a signal at dangerdidn't even realise which line his train was on and that it was one where he wasn't meant to be.

[Robert Shrives]

One drawback to reduced manning levels and changes in operations is the loss of a "secondman" in the cab - the years spent in understudy mode in a cab surely helped in road awareness and experiences that no process driven management lead training school can ever hope to match.  I suspect we will never have the luxury for second staff up front and for a  zillion valid reasons I am sure they would not be needed or desired within the industry.  

In the lulls between incidents in my work area talk often spins on to the "do you remember when xxx when wrong, some younger colleagues hear about larger accidents and bigger near misses with historical and geographical references that they did not know happened. A recent rail break on the Crosscity, within a fishplate was another Hither Green but was spotted first and quite a few had never heard of it. 

 

I await report with interest and hope it will feature in the active discussions within the industry.

Robert         

[Grovenor]

On 27/05/2023 at 13:07, phil-b259 said:

 

It might be noteworthy that ordinarily Grand Central and Lumo don't stop at Peterborough and as such might we not be looking at a case of the mind seeing what it wants to see which then means their brain says they are remaining on the fast line.

Which is, of course the very reason we have timed approach release on diverging routes, to ensure that speed is reduced. The obvious loophol;e that then occurs when the signal is so far back that the train can easily accelerate to an unsafe speed between signal and turnout has been known since the 1960s at least and can be avoided at the design stagewith a bit of effort, so those who agreed on the relaxation of design standards that the SM references should take some of the responsibility for allowing a reduction in the safety provided by the design.

 

Of course there are now mitigations possible that were not available earlier through use of TPWs, and evehtually ETCS, as Phil mentioned.

Edited May 29, 2023 by Grovenor

[Edwin_m]

3 minutes ago, Grovenor said:

Which is, of course the very reason we have timed approach release on diverging routes, to ensure that speed is reduced. The obvious loophol;e that then occurs when the signal is so far back that the train can easily accelerate to an unsafe speed between signal and turnout has been known since the 1960s at least and can be avoided at the design stagewith a bit of effort, so those who agreed on the relaxation of design standards that the SM references should take some of the responsibility for allowing a reduction in the safety provided by the design.

At the time of the Peterborough re-signalling, the usual train that might be accelerating after passing those signals would have been a Deltic plus 8.  Occasionally that Deltic might have been running light, in which case I guess its acceleration on maximum power would be phenomenal (and there was a similar incident with a 90 at Bletchley a few years back).  But these days most passenger trains approaching that signal are 80x units in electric mode, whose acceleration will leave the Deltic plus 8 in the dust.   

[Michael Hodgson]

5 minutes ago, Edwin_m said:

Occasionally that Deltic might have been running light, in which case I guess its acceleration on maximum power would be phenomenal (and there was a similar incident with a 90 at Bletchley a few years back). 

Odd though it sounds, locos have a lower speed restriction when runing light than when they've got a train behind them.

[Bucoops]

2 minutes ago, Michael Hodgson said:

Odd though it sounds, locos have a lower speed restriction when runing light than when they've got a train behind them.

 

Brake force presumably?

[Edwin_m]

3 minutes ago, Michael Hodgson said:

Odd though it sounds, locos have a lower speed restriction when runing light than when they've got a train behind them.

Indeed, though they will still have more acceleration.  No such restriction if a driver of an 80x unit applies full power after passing a route indicator.  

Edited May 29, 2023 by Edwin_m

[Michael Hodgson]

9 minutes ago, Edwin_m said:

At the time of the Peterborough re-signalling, the usual train that might be accelerating after passing those signals would have been a Deltic plus 8. 

It is a requirement that the braking curve is such that trains can stop if they encounter adverse signals, the signal spacing would have been based on all the types of train running at the time.  So when the Deltics were phased out in favour of faster stock, S&T would have had to re-evaluate the suitability of the then existing signalling to establish whether higher speeds could be approved. 

 

Between Peterborough and Stoke Tunnel they introduced the experimental flashing green signsl which effectively turned the system to five-aspect for 140 mph running of class 91s.  Flashing green was an example of speed signalling - it meant reduce speed to 125; other slower trains could treat it as a clear signal.  The experiment was not followed up, but I believe the flashing greens are still there.  The view was taken that above 125, lineside signalling wasn;t good enough, and in-cab signalling is required instead.

[Michael Hodgson]

9 minutes ago, Bucoops said:

 

Brake force presumably?

I believe so

[njee20]

76mph through a 30mph crossing?! Bloody hell! Astonished it stayed on. That’s normally the preserve of models!

[MarshLane]

On 26/05/2023 at 18:41, The Stationmaster said:

But I also wonder about the sighting of the signal (or the competence of the Drivers involved) if two Drivers failed to notice the Ji illuminated).   The interesting thing is that while Grand Central has been around for quite a while Lumo is a fairly recent arrival on the scene and their Driver standards management surely needs to be very carefully looked at in the light of this incident involving their train?  The same of course might be said of Grand Central but they certainly used to be a competently managed concern.

 

I cannot verify this @The Stationmaster but I believe the Lumo driver was a former LNER driver who transferred.  A good source told me that was the case, in which case it was likely he had traversed the route thousands of times.

 

The problem here with both Lumo and Grand Central, where it does not affect XC, Greater Anglia or LNER, is that neither are booked to stop at Peterborough, so it is an out of the ordinary move, and one where the awareness and attention of the driver comes in play.  XC always calls, LNER drivers are use to calling and Greater Anglia start from here.  But Lumo and GC are not booked to use the platform roads, and on both occasions I believe the issue came when the signallers diverted them onto the Up Slow to allow a faster service to pass.

 

Also the fact that the incidents are 12 months apart does (in my personal view) point more to a driver attention aspect than it does training or competency.

 

Edited May 29, 2023 by MarshLane

[Edwin_m]

9 hours ago, Michael Hodgson said:

It is a requirement that the braking curve is such that trains can stop if they encounter adverse signals, the signal spacing would have been based on all the types of train running at the time.  So when the Deltics were phased out in favour of faster stock, S&T would have had to re-evaluate the suitability of the then existing signalling to establish whether higher speeds could be approved. 

The issue that may be significant here relates to acceleration not braking.  One possible cause of such an overspeed is passing the signal at the normal speed for a diverging train, then accelerating so the junction itself is passed at a much higher speed than permitted.  This is much more likely with a light engine or a modern high-powered EMU than with a traditional locomotive and coaches.  

 

There was indeed a re-evaluation of signalling when HSTs were introduced, which were the first to have the higher braking rate which is now universal for new passenger stock and mandatory for anything exceeding 100mph.  One consequence was the introduction of flashing aspects, to replace approach control at certain junctions where the HST might be able to brake later to comply with the signals but end up too fast at the junction itself.  

[DY444]

Translation: Urgent Safety Message - Drivers need appropriate route knowledge before driving trains through a given location and to pay attention to signals when they drive through that location.  

 

When has it ever been anything other than totally understood by everyone that you regulate your speed to that of the divergence irrespective of how far out the signal is?  The only times you don't are if you're not paying proper attention or you don't know the road.  The end.

 

All this waffle about how far out the signal is etc looks like smoke and mirrors.  It seems to me that the drivers concerned in these incidents didn't know the road properly and shouldn't have been driving over it.  I suspect both being OAO may not be a complete co-incidence.

Edited May 30, 2023 by DY444