-
Our booking engine at tickets.railforums.co.uk (powered by TrainSplit) helps support the running of the forum with every ticket purchase! Find out more and ask any questions/give us feedback in this thread!
Level crossing incident near Norwich new RAIB investigation
- Thread starter notverydeep
- Start date
- Status
Sponsor Post - registered members do not see these adverts; click here to register, or click here to log in
R
RailUK Forums
edwin_m
Veteran Member
A track circuit (working properly) will positively confirm that there is no train within its extent, so can be used to clear a signal for example. All a treadle does is detect when a wheel is standing over it, and if there is none it can give no information about whether there is a train in the section, and a treadle at the end of the section can't prove that the whole train has passed over it. So a treadle can't be used to clear a signal.
However a track circuit may fail to detect a train when the rail is contaminated. Usually this is just for a short time when the first few wheels of the train are within the track circuit and the rest of the train is still on the previous one, so it doesn't normally cause the train to disappear from the signaling system. But if that track circuit is a strike-in for a level crossing it shortens the time between the warning to road traffic and the arrival of the train. The treadle overcomes this because it's virtually guaranteed to detect the first wheel.
More modern installations use axle counters instead of track circuits, which are a bit like treadles but count the number of wheels passing in each direction so by comparing the counts it is possible to prove that the section between them is clear. I presume, but I don't know for certain, that level crossings in axle counter areas don't have treadles as well.
The crossing in this case appears to be using a predictor, which is a more sophisticated piece of kit able to detect the position and speed of the approaching train and trigger the level crossing to give a moreorless constant warning time for fast and slow trains.
The crossing in this case appears to be using a predictor, which is a more sophisticated piece of kit able to detect the position and speed of the approaching train and trigger the level crossing to give a moreorless constant warning time for fast and slow trains.
eman_resu
Member
Australia - NSW and Victoria
https://www.bbc.co.uk/news/av/uk-en...train-misses-crashing-into-car-by-025-seconds
Video from train published on BBC website.
Video from train published on BBC website.
Wilts Wanderer
Established Member
- Joined
- 21 Nov 2016
- Messages
- 2,484
I came here to comment on that. It reinforces the idea that the unit momentarily went missing from the POV of activating the track circuits.
But the system shouldn't fail wrong side like that. Reported in other threads that NR are fitting treadles where this non standard predictor system is in use.
Seems the HAZOP at design stage was not thorough enough. "What happens if the track circuit detection fails before the train has passed?" "The barriers will raise." Hmm, better design that behaviour out...
Whatever happened to the concept of "failsafe"?
I'd have thought a system that relied on track circuits to close barriers should keep the barriers down once a train is detected (even if it subsequently "disappears"), only to be raised by triggering a sensor of some description after the crossing.
I'd have thought a system that relied on track circuits to close barriers should keep the barriers down once a train is detected (even if it subsequently "disappears"), only to be raised by triggering a sensor of some description after the crossing.
I suspect this will be the conclusion of the investigation. I am rather surprised that such a thing could happen if there was any sort of fault.
Grumpy Git
On Moderation
As an electrical engineer who designs and installs fail-safe control systems I find this remarkable!
The vast majority of the country’s signalling system relies on trains operating track circuits. Yes, there are good arguments for treadles for level crossings as well, but what about everything else? Axle counters are used in some areas but they are not everywhere, the track circuit is still the basis for train detection across the majority of the network and they are not backed up with treadles. Those level crossings have worked well with a variety of rolling stock for around 19 years, it’s only with the introduction of the new trains that a problem has occurred, what are the implications across the wider network? What would result if a train “disappears” from a signallers panel? I know, I’ll let you give it some consideration.
Agreed. I'm amzed that this possibility was allowed to take place.
Maybe the track circuits are fine, and it's this new stock that somehow doesn't reliably activate those circuits, but surely NR must have some stake here in terms of asessing the suitability of new stock for the PW and signalling? In post #14, it was pointed-out that these Stadler units appear to have Track Circuit Actuators fitted, in which case it seems strange that there should have been any possibility of a unit "disappearing" in the manner that this one seems to have done.
thing is, we have track circuit failures on a regular basis, all that happens is the signal protecting the area where the failure is reverts to danger. Then a track inspection is done to asses what the issue is. All nice and safe.
when one fails on a level crossing that has no signals protecting it, where’s the failsafe?
brick60000
Member
- Joined
- 18 Apr 2013
- Messages
- 442
That video makes for some harrowing footage and as others have said, you almost have to watch it a few times to believe it’s actually real. Thank goodness the stars aligned that evening and nobody was hurt! Certainly a wake up call I guess, and a stark reminder that things can still go wrong even with all the safety in the world.
In particular when the fail-safe is ineffective or absent.
Unfortunately, it seems to me that many AHB crossings don't have very good sightlines from the road, presumably as the NR people who write the regulations don't think they need them.
Last edited:
Grumpy Git
On Moderation
Fail-safe means just that, it can't be "ineffective" by definition.
It is The Proposer of Change, as defined in the Common Safety Method for Risk Assessment (CSM-RA) that has the responsibility for risk assessing any change. In this case it is the TOC who is proposing the change so it’s up to them to involve the other stakeholders - Network Rail, Manufacturer etc. - and work with them to mitigate any risks.
It’s interesting that these trains need TC actuators when most don’t, I wonder what the mitigation is for TC actuator failure?
Thanks - I'm not employed in the industry, I'm afraid; just an enthusiast.
I meant if a fail-safe system does not work properly.
I don’t think there was a TC failure. It looks like the train failed to keep the TC shunted so it cleared. If that happens anywhere other than at a level crossing then you could find clear signals when there is a train on the track, not very clever!
If a TC at a level crossing fails then the lights will flash and the barriers will lower.
Visibility of oncoming trains themselves by road users is not a consideration for this type of crossing. It is assumed the warning systems will always work. Opening up good sightlines on the typical rural roads that have AHBCs could be rather difficult and might involve removal of old crossing keepers' houses for example that are likely to be in private ownership now.
Grumpy Git
On Moderation
A "fail-safe" system MUST be just that, (i.e. fail-safe), if it "does not work properly" (your words) its failed, so when it fails, it goes to a SAFE state.
If it fails and the result is that it is not SAFE, then it is NOT fail-safe!!!
On the face of it it seems likely a new heavier 3 or 4 car train would be more likely to be assured of shunting the track circuits than the often shorter sprinter trains that have plyed the line until now. However, something I came across on the ex LSWR West of England route to Exeter may be relevant. Where a single class of unit generally provides the entire service, with modern suspension every single passing axle of the 159s followed the exact same path on the railhead, leaving a very narrow bright metal strip with the rest of the railhead deeply rusted. The problem with that is that if a different but equally or even more modern short train or track machine traversed the route on rare occasions then it was highly likely it would disappear as it's wheel-rail contact point could be different and might be riding on the poorly conductive rust. On the WOE that is no longer such a concern as the new signalling east of Pinhoe uses axle counters. Through Exeter Central it was never such a concern, as a variety of classes operate the denser service.
Indeed, that was the idea I was thinking of. While attempting to check, as was mentioned by "Taunton" might be useful, in practice, it's impractical.
I think we're getting into semantics here. Something can be designed to be failsafe but, due to errors, it might not actually be able to fulfil this criterion; for want of another way of putting it, it has failed to be failsafe!
Interesting thought - is this well-nown within the industry?
- Status