Braking issue on Caledonian Sleeper causes train to "run away" at Edinburgh

Discussion in 'UK Railway Discussion' started by 87015, 1 Aug 2019.

Thread Status:
Not open for further replies.
  1. HSTEd

    HSTEd Established Member

    Messages:
    9,853
    Joined:
    14 Jul 2011
    I'm simply suggesting, in this thread, that this train should be fitted with safety systems standard on almost all other passenger trains.
    Electronically controlled braking should be fitted as standard to all trains, yes.
    But the railway industry, especially the freight industry, is notorious for its refusal to adopt technological innovations - at least since privatisation.

    Fitting them would only cost $6,000-$8,000 per vehicle according to studies done in America. So the cost is negligible.
    Finding the total UK freight wagon fleet size is quite hard, but apparently 600 trains run per day, which implies something like 2000 wagons in use each day. So perhaps 20,000 wagons in total at a reasonable utilisation rate.
    That is something like $120m-$180m to fit the entire wagon fleet. Fitting the locomotives would be more expensive but there are so many less of them it almost doesn't matter.
    In reality fitting a very large fraction of the total freight train movements would require fitting a much smaller number of wagons, like the freightliner rakes that run up and down all day every day.

    There are various operational savings, and also the fact that it would completely eliminate run-arounds and other such freight operations because the same technology can be used for multiple working. So freight trains could top and tail or have non powered control vehicles.

    But it will never happen, because the freight industry can always rely on subsidies to keep it somewhat competitive.

    The cost of fitting this tiny fleet of 75 vehicles that operate with a special pool of locomotives with modern safety systems would have been negligible.
    These are not ancient tank wagons where you can claim an economic exception.
    This was brand new stock, and yet here we are.

    And relying on "competent people" as the basis of a safety case in my own industry got us Three Mile Island, Tokai-Mura and Chernobyl.
    It is only good fortune that there were not a lot of dead people that day.
     
    Last edited: 12 Aug 2019
  2. John Bishop

    John Bishop Member

    Messages:
    62
    Joined:
    15 Nov 2018
    Surely a running brake test would have flagged up the issue which should have been done on leaving Carstairs.
     
  3. a_c_skinner

    a_c_skinner Established Member

    Messages:
    1,162
    Joined:
    21 Jun 2013
    Is a running brake test not compulsory? I thought (for loco hauled trains) it was mandatory - based on a rule book on the www which seemed current.
     
  4. The BIgman1234

    The BIgman1234 Member

    Messages:
    35
    Joined:
    12 Aug 2019
    Not being a driver I can’t really say, but I do believe that they have to carry out one ‍♂️
     
  5. hwl

    hwl Established Member

    Messages:
    3,873
    Joined:
    5 Feb 2012
    Running brake test up hill with loco with a lots of regenerative/rheostatic braking and 6 axles?
     
  6. TimboM

    TimboM Established Member

    Messages:
    3,375
    Joined:
    12 Apr 2016
    Not necessarily. It's a climb out of Carstairs and the Class 92 has two rheo (dynamic) brakes each rated at 2.5 MW (the traction motors work in reverse and slow the loco down by converting the energy to heat and releasing it via resistors. Each brake stack gives off the energy of about 830 3-bar electric fires, so pretty hefty bits of kit).

    The loco's rheo brakes and a rising gradient together could quite feasibly have slowed the train sufficiently during a Running Brake Test leaving Carstairs to give the impression the train brakes were working. It's then possible that only when descending towards Edinburgh and needing to brake significantly on the approach to Waverley (which would be too much for the loco brakes alone) did it become apparent there was an issue.

    We'll need to wait months for the RAIB to conclude on what they believe actually happen, but in the meantime care should be taken to jump to conclusions that certain brake tests weren't done, as there's plenty of plausible explanations whereby both the brake continuity test at Carstairs and the running brake test after departing could've been done and the incident still occur.
     
  7. a_c_skinner

    a_c_skinner Established Member

    Messages:
    1,162
    Joined:
    21 Jun 2013
    As it was the same locomotive that couldn't stop the train in Waverley it would clearly have exposed the fault. Looking at the old logs of ECML depatures from KGX they didn't do them uphill in the tunnels but soon afterwards. It isn't uphill all the way to Edinboro. Anyhow how to do it is secondary to my question which is are they still supposed to be done. The last thing I read about RBTs suggested they were not being done with sufficient assiduity.
     
  8. TimboM

    TimboM Established Member

    Messages:
    3,375
    Joined:
    12 Apr 2016
    No, it wouldn't clearly have exposed the fault. A 300 tonne "swinger" on the back of a loco when travelling uphill has a significant slowing/drag effect as soon as the loco stops providing power and the loco's own (rheo) brakes are applied. It's the opposite going downhill with the momentum of 300 tonne of unbraked stock rendering the locomotive brakes largely ineffective.

    In broad terms, the line from Carstairs to Edinburgh climbs for the first half, then descends into Edinburgh - so it is uphill for quite some way after leaving Carstairs and beyond the distance you'd want to wait before performing a Running Brake Test.
     
  9. The BIgman1234

    The BIgman1234 Member

    Messages:
    35
    Joined:
    12 Aug 2019
    Personally I hope it’s a mechanical/electrical fault rather than human as I wouldn’t like to anyone loose their job .
    The new sleeper train has been an absolute disaster since they started running it . The delay it has caused with trying to spilt and join it has been diabolical. I’d ban them of the railway till they got all the problems sorted !!!!
     
  10. Bletchleyite

    Bletchleyite Veteran Member

    Messages:
    40,530
    Joined:
    20 Oct 2014
    Location:
    Up and down the south WCML (mostly)
    Locomotive-hauled coaching stock, sorry. (As distinct from multiple units)
     
  11. Surreytraveller

    Surreytraveller Established Member

    Messages:
    1,273
    Joined:
    21 Oct 2009
    Well, there's currently multiple unit stock coupled up in multiple where drivers cannot walk through. Are all stations to Glasgow / Edinburgh platform staff despatched? Not sure - but could they run DOO as far as GLC/EDB then have guards? But I suppose with LHCS the driver has no control over the doors?
     
  12. theironroad

    theironroad Established Member

    Messages:
    2,267
    Joined:
    21 Nov 2014
    Location:
    London
    Certainly on older slam door stock the rule book required that a brake continuity test was required after any attaching or detaching.

    If brakes cocks were operated after a brake contiuity test, then that brake test is now void and a new brake continuity best would be required.

    I don't sign Carstairs to Edinburgh Waverly, but I'm sure there must be an opportunity gradient wise to do a proper running brake test that would have set some metaphoric alarm bells ringing in the drivers head.
     
  13. a_c_skinner

    a_c_skinner Established Member

    Messages:
    1,162
    Joined:
    21 Jun 2013
    So there should be a running brake test at a suitable location and carried out in a suitable way and (one of) the questions becomes why did this not reveal this very serious problem?
     
  14. BRX

    BRX Established Member

    Messages:
    1,945
    Joined:
    20 Oct 2008
    Doing a brake test at a location where it's not possible to have an unambiguous result seems a bit pointless. I remember this being discussed in the RAIB report on the Carrbridge runaway/derailment. There was the complication of winter weather there but I think there was some stuff about it not being possible to be clear about the results on a climbing gradient.
     
  15. Justapunter

    Justapunter Member

    Messages:
    273
    Joined:
    21 Jun 2018
    All this Antediluvian running brake test stuff. Sounds like stuff Stephenson and Gooch deliberated over. These are super expensive super complicated electronically controlled 21 century (after the rebuilds) trains. And we are talking about points of nonsense like not following the manual, whoops, oh well....

    At the cost of this stuff, and the potential for disaster, the incremental cost of an electronic check on the brakes, given no doubt all the canbus wiring etc, is the square root of nothing. And there are repeated attempts to gloss over a fundamental and sort of obvious safety system. I am more concerned about 400 tons plus stopping than moving....

    It also takes the human point of failure out of the equation.... so whom ever has made a mistake doesn’t find themselves able to do it again. And they shouldn’t have been put in that position in the first place....
     
  16. Put Kettle On

    Put Kettle On Member

    Messages:
    20
    Joined:
    2 Apr 2019
    Location:
    Here & there , but mainly there .
    Climb out of Carstairs for 2 1/2 to 3 miles , then there is a descent before levelling off & then a climb up to Cobbinshaw before another descent towards Midcalder Jcn . Before Midcalder there is a 70 MPH restriction , from Midcalder Jcn basically level for a bit before a final drop down towards Curriehill & into Edinburgh.
    Approaching Slateford , towards Haymarket Jcn , the train would need to be reducing speed, for the 40 through Haymarket East, continues at 40 through Haymarket tunnels before a 20 MPH through Princes Street Gardens & in to Waverley.
    Why did the phone call to Signaller only take place whist between Haymarket & Waverley, surely it would be evident something was amiss before there .
    It is reputed to have still doing around 40 MPH going through Waverley, so at what speed did it cross Haymarket East & travel through tunnels & gardens .
     
  17. hexagon789

    hexagon789 Established Member

    Messages:
    5,811
    Joined:
    2 Sep 2016
    Location:
    Glasgow
    Perhaps an edit to make an RBT using purely the air brakes compulsory to check their effectiveness?
     
  18. theironroad

    theironroad Established Member

    Messages:
    2,267
    Joined:
    21 Nov 2014
    Location:
    London
    It's not points of nonsense, it's about the driver who is driving that particular formation knowing the effectiveness of the braking capability available to them for that journey.

    It allows the driver to get a feel for how good the brake is, as different units, vehicles and formations may be reacting differently. It also allows the driver to test the performance of the brakes and stopping distances required especially in low rail adhesion conditions where braking generally takes longer.

    Buy hey, it's all nonsense really.
     
  19. CheshireCrusty

    CheshireCrusty Member

    Messages:
    55
    Joined:
    6 Feb 2014
    I offer my congratulations and thanks to all those that stopped this becoming a very tragic incident, but naturally, as a potential passenger I was very concerned about this incident and hope the RIAB will look at all aspects.

    The questions that initially occurred to me were:
    1. Did the driver have any means of communicating the problem to the TM/Guard? I appreciate there were likely to be other urgent actions required of the driver, but did he not have a readily accessible method of contacting the TM to discuss the required action? (As a passenger I can recognise the "bell code" for driver calling TM on Virgin's DMUs and EMUs).
    2. What did the TM have to do to apply the carriage brakes when aware of the problem - for example did he have to go to a specific carriage (I keep thinking of the "Guards" Brake Van on old LHCS) and how does that relate to other duties required as the train approached a stop (Haymarket) or its terminus (Waverley)?
    3. What did the driver have to do to alert the signaller of the runaway? Again being old, I seem to remember there was a steam whistle for this but that doesn't work with centralised signal control rooms - I assume there is now a panic button or similar to provide instant communication?
    4. I assume the CS is routinely pathed into Platform 11 a bidirectional through platform, but did the signaller change the path through Haymarket or Waverley to avoid crossovers etc.?
    5. Is it a standard operating procedure to provide wherever possible (irrespective of any signal shown) a clear path ahead for a train approaching a through platform, even though that train may be scheduled to stop or terminate at the platform?
    6. I understand the train eventually stopped at Abbeyhill - was this by default or design? I ask this question because it seemed a good choice to stop - on an upgradient out of Waverley, there is a crossover presumably providing bi-directional signalling etc.
    7. Finally if the train was going at 40mph through Waverley at what speed was it going through Haymarket and how did that relate to the line speed?
    Presumably there will be far more intelligent and appropriate questions asked during the investigation, but as a concerned passenger these were the questions that immediately sprung into my mind.
     
    Last edited: 13 Aug 2019
  20. edwin_m

    edwin_m Veteran Member

    Messages:
    15,504
    Joined:
    21 Apr 2013
    Location:
    Nottingham
    A series of short blasts on the horn is used if a train is in an emergency situation, but when doing that a driver only has one hand free to do anything else. I've no idea if these trains have any sort of intercom from the loco.
    It would be an emergency call to the signaller on the train radio. There is also a button to stop all nearby trains in an emergency, but the better course of action might be not to use this, as it might stop trains across junctions or locking the routes across junctions ahead of them, thereby making it more difficult for the signaller to find a route through the station. If these trains kept moving the routes reserved for them over junctions would release as they passed. I haven't seen any mention of the emergency button being used, so I tend to think it wasn't.
    The signalling generally provides an "overlap" distance beyond a danger signal as a train approaches, and an overrun into the overlap won't result in a collision, but it's normally no longer than 225m and it may be less in a busy station. The overrun here was nearly three times that.
     
  21. ac6000cw

    ac6000cw Established Member

    Messages:
    1,799
    Joined:
    10 May 2014
    Location:
    Cambridge, UK
    Given that we are 19 years into the 21st century (and more than 100 years since radio communications were invented), I would expect by now that guards/conductors - in situations where there isn't any other form of modern intercom system available - to be equipped with some sort of portable GSM-R handset so they can communicate directly with the driver and signaller in safety-critical situations.

    Is this the case, or are we still running a steam-age railway in some respects?
     
  22. GB

    GB Established Member

    Messages:
    5,385
    Joined:
    16 Nov 2008
    Location:
    Somewhere
    The yellow "Urgent" button on the GSMR would have been the one to use...it connects straight to the signaller without bringing all the other services to a stand like the red emergency button would have ....however I believe a normal GSMR call was made....why this was we will have to see in the report. Class 92s have no direct intercoms with the rest of the train, standard back to back radios are in use like they are on charter services, whether they were actually in use/working will again have to wait for the report.
     
  23. O L Leigh

    O L Leigh Established Member

    Messages:
    4,087
    Joined:
    20 Jan 2006
    The two-pipe air brake may have it's origins in antiquity, but the fact that we still use it today is testament to the fact that it works. If it didn't work or introduced an unacceptable level of risk that couldn't be mitigated by other means it would have been replaced by now. The fact that it hasn't should be telling you something. As a result, what is the point of having to go away and invent a system to replace it.

    You may not like it, but the success of every critical operation in organisations from NASA down relies on someone doing their job properly, whether it's launching a satellite, undergoing surgery or brake-testing a train. Someone has to design a system, build and test it rigorously and ensure that it's been correctly deployed. You cannot simply automate every process that you don't like because automation itself introduces risks that require mitigation. Automation fosters an over-reliance on the efficacy of the automated system causing people to simply stop checking.

    Besides, as I've said repeatedly, this incident will have been caused by a process failure. There's no fault that an automatic system could have detected that wouldn't have already been found if this train had been properly brake-tested.

    It's also worth pointing out that the requirement to carry out a running brake test applies to all stock irrespective of it's age or braking system.
     
  24. ac6000cw

    ac6000cw Established Member

    Messages:
    1,799
    Joined:
    10 May 2014
    Location:
    Cambridge, UK
    Re. remote brake line pressure monitoring at the rear of trains - in North America (and I assume other countries running long freight trains) railways have been using radio-linked 'end-of-train' devices to do this for nearly 40 years (the devices look like modern flashing red tail lamps, but also have a connection to the brake pipe). Some types can also be commanded to vent the brake pipe in an emergency, in the event that it has become blocked by e.g. ice

    They allow brake continuity testing/monitoring from the loco cab whenever you like...as well as being a train-handling aid e.g. so the driver knows when a brake release has reached the end of the train.
     
  25. GB

    GB Established Member

    Messages:
    5,385
    Joined:
    16 Nov 2008
    Location:
    Somewhere
    ...but if the train is fully charged up and then the aircocks between vehicles operated, the EOTD will still show the correct pressures for that part of the train....particularly if the cocks have been closed not long before departure.
     
  26. Sleeperwaking

    Sleeperwaking Member

    Messages:
    91
    Joined:
    20 Oct 2018
    I think ac6000cw is referring to a system that is always on, rather than something that is only checked before setting off. I.e. you could have something that constantly compares the coach brake pipe pressure to the loco brake pipe pressure, and flag an issue if the two differ by X% / some tolerance. That would probably identify isolation of the coach brake pipe during a running brake test / first service brake application, which could be flagged to the driver by an alarm in the cab (potentially also the guard). The driver could then decide what action to take.
     
  27. ac6000cw

    ac6000cw Established Member

    Messages:
    1,799
    Joined:
    10 May 2014
    Location:
    Cambridge, UK
    Correct (my understanding is that is how standard EOTD's operate - the loco cab unit requests the pressure info fairly frequently from the EOT unit). It means that when any train air brake application is made, the driver can check that it actually reaches the rear of the train - and how long it takes to get there. With a train as short as a passenger train, the propagation delay (of the pressure reduction) ought to be quite fast, so a 'running brake continuity test' shouldn't take very long in theory.
     
  28. GB

    GB Established Member

    Messages:
    5,385
    Joined:
    16 Nov 2008
    Location:
    Somewhere
    The whole of the brake pipe is charged fully to 5 bar. If you split the train pipe in two a few minutes before departure by closing the cocks between two vehicles, the 1st half of the train will still be at 5 bar and the second part will still be at 5 bar. It takes some time for air to leak out, particularly on new stock.

    Certainly if you were then to apply the brake when stationary you would see a difference (the brake pipe in the first half being commanded to apply while the second part is still at full release) but if the train is on the move by the time its is in a suitable position to do a running brake test it would be too late.

    If the cocks were closed before the train had built up its air or the train was being held on the train brake and not just the loco brake then yes you would see a difference on the EOTD between the back and front, but the train wouldn't have gotten very far as the brakes on the second part of the train would still be on...or at least partially on.
     
  29. edwin_m

    edwin_m Veteran Member

    Messages:
    15,504
    Joined:
    21 Apr 2013
    Location:
    Nottingham
    I think the end of train device is more about the issues that arise specifically with very long American freight trains. If power is applied before the brakes have released all the way to the back of the train then a broken coupling or maybe even a derailment could result. It may also have a role in avoiding the "running out of air" issues that can arise on single-pipe systems and were discussed a few pages back as a possible cause of this incident. I'm not sure how it would help in brake continuity testing, because someone would still have to go to the other end of the train to attach it, and it would only take a few seconds to test the brake while there. There's also the risk of setting up the wrong channel on the radio or whatever they use, so the EOT being monitored might be somewhere other than the end of the same train.
     
  30. Justapunter

    Justapunter Member

    Messages:
    273
    Joined:
    21 Jun 2018
    It was nonsense. Because it didn’t work. The driver getting a feel didn’t work. The stopping distances were clearly wrong. And it was lucky it stopped without hitting something (and seemingly possibly only with assistance from non driving crew). An electronic assessment and failsafe would have prevented the mistake.

    The problem is not an automated process. It’s a lack of electronic check on the manual processes. And it is not acceptable when it can be engineered out. Some rail staff could easily be looking at years in prison and living with the guilt and shame, it just embarrassment. Yet people persist with “well, men are trusted. But they do mistakes. So we won’t. Bother with reducing/removing the risk.”

    We have the technology. Why on earth would you rely on humans alone where you don’t have to. The microprocessors can do the test, or prohibit movement until the brakes are definitely sorted. But, again, it seems railway knows best. Or rather it’s a bit like my exmother in law. She knew everything. All the time. Including how to tell NASA how to fly. And no doubt how to tell CS drivers how to do brake tests.....
     
Thread Status:
Not open for further replies.

Share This Page