Train collides with van on level crossing at Teynham, Kent, 23/10

[Llanigraham]

Quick one, How long would it take for the train to stop?

How long is a piece of string?
Far to many variables to give an accurate figure.

 

[Chris M]

Also, it appears a second party was present at the crossing at the time of the incident and their actions should have left the van driver in no doubt as to whether it was safe to cross the line without contacting the signaller.

Without knowing what the actions of this second person were (and you give no hints) then it is not possible to say whether they left the van driver in "no doubt".
It is also entirely possible for a second party to leave the first party in "no doubt that it is safe to cross" when it is not actually safe to cross. The first scenario that comes to mind is that the second party had just finished crossing, and spoke to the first party saying words to the effect that "I spoke to the signaller who said it was safe to cross." which the van driver could very easily interpret as meaning that it was safe for them to cross also. Neither party is likely to be trained in safety critical communication, there could have been time for one party to cross the line when the second party spoke to the signaller but not when they completed their crossing (is there a requirement to let the signaller know when you have completed crossing if you are not crossing with animals and/or a slow vehicle?).

 

[Llanigraham]

There is now no requirement in the Rule Book for the person to report back to the signaller if they are driving a normal vehicle, however it can be at the discretion of the signaller if he considers there is a need.
At my Box if they were towing a trailer of any kind or in a tractor they were told to report back.

 

[2HAP]

RAIB Investigating.

https://www.gov.uk/government/news/collision-at-frognal-farm-user-worked-level-crossing

 

[bengley]

"On this occasion there is no evidence that the driver of the delivery van made a telephone call to the signaller. However, the signage provided at the crossing was not, in the RAIB’s view, clear."

The sign reads "STOP. Always telephone before crossing with vehicles or animals to find out if there is time to cross" It's a fairly large red retro-reflective sign placed immediately in front of the gates.

How, unless you are an idiot, is this not clear?

 

[SamYeager]

"On this occasion there is no evidence that the driver of the delivery van made a telephone call to the signaller. However, the signage provided at the crossing was not, in the RAIB’s view, clear."

The sign reads "STOP. Always telephone before crossing with vehicles or animals to find out if there is time to cross" It's a fairly large red retro-reflective sign placed immediately in front of the gates.

How, unless you are an idiot, is this not clear?

Given that the RAIB is not "an idiot" then there must be some reason why they feel the signage is unclear. Perhaps the signage associated with POGO might give the impression that there is no need to phone?

 

[2HAP]

Surely it's possible to link POGO to the track circuit so that its impossible to open gate if a train occupies the circuit?

 

[najaB]

Surely it's possible to link POGO to the track circuit so that its impossible to open gate if a train occupies the circuit?

And then you have to add a proving circuit...

 

[ComUtoR]

Surely it's possible to link POGO to the track circuit so that its impossible to open gate if a train occupies the circuit?

If there was a track circuit failure, how would you open the gate ?

 

[Llanigraham]

That also presumes there is a track circuit closely in advance of the crossing.

 

[2HAP]

If there was a track circuit failure, how would you open the gate ?

Electrical release from signalling centre?

Llanigraham - given the location of this one, it's a racing certainty there is a track circuit.

 

[GW43125]

Electrical release from signalling centre?

Llanigraham - given the location of this one, it's a racing certainty there is a track circuit.

Axle counter area I believe

 

[ComUtoR]

Electrical release from signalling centre?

If the track circuit shows 'Occupied', when clear then box wouldn't be able to release the gate. The track circuit shows occupied so failing safe would keep the gate locked. To allow them to override the locked gate will need them to confirm the track isn't actually occupied and the crossing is clear. They would need CCTV to remotely view the crossing.

The signaller can send a train into a section when showing 'occupied' so the gate would still need to be impossible to open even if there was a track circuit failure.

 

[Llanigraham]

Llanigraham - given the location of this one, it's a racing certainty there is a track circuit.

But not everywhere.
And as has been said if it is an axle counter area, then it certainly wouldn't work.

 

[nom de guerre]

Axle counter area I believe

Correct.

 

[nom de guerre]

Deleted.

 

[Tomnick]

Surely it's possible to link POGO to the track circuit so that its impossible to open gate if a train occupies the circuit?

Which track circuit? How long is it? Will the occupation of that track circuit give sufficient warning of an approaching train or will the controls need to extend back onto track circuits further in rear? I suspect that you’re proposing a system that merely supplements the existing process of phoning for permission, but I doubt it’d take users long to start relying on the gates being locked when a train’s coming. It gets more complicated if there’s a junction in rear, or even a signal. Would the crossing have to be taken on local control during a possession, as others affected by TC occupation do? Wrong direction movements?

It doesn’t take long for a suggestion that looks simple on paper to become very complicated!

But not everywhere.
And as has been said if it is an axle counter area, then it certainly wouldn't work.

Why? Track circuit occupation = axle counter section occupation, as far as the interlocking’s concerned (in normal operation) at least.

 

[Llanigraham]

Because you are making the presumption that there is an axle counter or track circuit immediately on the approach to this type of User Controlled Crossing, which is obviously not the case.

 

[Tomnick]

Because you are making the presumption that there is an axle counter or track circuit immediately on the approach to this type of User Controlled Crossing, which is obviously not the case.

Not at all, that’s already been covered and is just one of the reasons why it might not (and probably won’t) work. Not just ‘no track circuit’, but also ‘track circuit too short’ etc. I was just puzzled why it “certainly wouldn’t work” in, specifically, an axle counter area?

 

[najaB]

I was just puzzled why it “certainly wouldn’t work” in, specifically, an axle counter area?

I'd not say it 'wouldn't work' but it certainly makes things more complicated. A track circuit could operate the lock by way of a relay - the lock is kept shut by way of a spring and a solenoid moves it to the open position when the circuit isn't closed. That way it would fail safe. In an axle counter area, you either need a logic unit at the crossing or a feed back to the signalling centre.

 

[Tomnick]

I'd not say it 'wouldn't work' but it certainly makes things more complicated. A track circuit could operate the lock by way of a relay - the lock is kept shut by way of a spring and a solenoid moves it to the open position when the circuit isn't closed. That way it would fail safe. In an axle counter area, you either need a logic unit at the crossing or a feed back to the signalling centre.

I'm maybe getting a little bit out of my depth now, but a track circuit 'reports back' to the interlocking (whether it's a relay interlocking or SSI or whatever) whether it's 'occupied' or not. An axle counter section does similar, although it's presumably more complicated with conditional restoration and that. Either could be used in a circuit to determine whether the crossing should be able to be released or not, I'd have thought - but it'd have to be designed and checked as any other alteration to the interlocking would, and wouldn't necessarily involve just a single track circuit in each direction. On a single line in particular, it'd very quickly get complicated!

 

[najaB]

I'm maybe getting a little bit out of my depth now, but a track circuit 'reports back' to the interlocking (whether it's a relay interlocking or SSI or whatever) whether it's 'occupied' or not. An axle counter section does similar, although it's presumably more complicated with conditional restoration and that.

The difference is that a lock could be 'piggybacked' onto a track circuit at the purely electrical level without requiring any digital logic.

 

[ComUtoR]

If the lock on the gate gets broken, or is left open. Would that then be interlocked with the track circuit and then put the signal to red ? If the gate is open, is it then safe to assume that it is safe to cross and for how long ? Once that gate has been opened something must then prevent any train from passing.

 

[najaB]

If the lock on the gate gets broken, or is left open. Would that then be interlocked with the track circuit and then put the signal to red ? If the gate is open, is it then safe to assume that it is safe to cross and for how long ? Once that gate has been opened something must then prevent any train from passing.

Oh, I get that it's a bad idea, I was just explaining why it's a slightly less complicated bad idea in a track circuited area.

 

[2HAP]

OK, so it seems that you can't link the POGO to track circuits or axle counters.

Let's run this one past you then.

POGO electrically locked to signalling centre, which gives a release when a) telephoned by crossing user and b) safe for vehicle to cross.

 

[najaB]

OK, so it seems that you can't link the POGO to track circuits or axle counters.

You can, it just gets complicated really quickly.

Let's run this one past you then.
POGO electrically locked to signalling centre, which gives a release when a) telephoned by crossing user and b) safe for vehicle to cross.

That's basically how signaller controlled pedestrian crossings work (sorry, drawing a blank on the correct name).

 

[alxndr]

Another possibility could perhaps be an overlay system, such as used by EBIgates, which would allow more flexibility and avoid needing to redesign the signalling circuitry in that area. These work by additional axle counters that aren't in anyway linked into the signalling system or interlocking. In EBIgate useage these feed into a PLC (Programmable Logic Controller) which determines whether there is a train approaching either shows a red or green light to the user. In a failure situation no light shows, there is a board instructing any users to telephone the signaller, and an alert is sent so that techs are aware. There are varients which operate barriers too.

It would probably be necessary to install lights as well to enable a failure situation to become apparent and users to be directed to contact the signaller, but it could surely be possible to wire a gate solenoid into this to prevent the gates being opened with a train approaching. A manual method of opening the gates could be possible in failure situations. There is the potential for this to be abused, of course, but equally there's currently no guarentee that a user is going to pick up the phone and contact the signaller, especially if that means leaving a vehicle to do so.

There would be expense involved, but I'd hazard a guess that it would be cheaper than reworking track circuits/axle counters and interlockings to provide suitable detection.

 

[eastdyke]

The RAIB Rail Accident Report has been published today:
https://assets.publishing.service.g...a/file/735874/R122018_180823_Frognal_Farm.pdf
The report makes reference to previous and subsequent incidents at POGO crossings and has quite a bit to say on the overall understanding and use of such installations.

Summary
On 23 October 2017, a passenger train collided with a parcel delivery van at Frognal
Farm user worked level crossing, near Teynham, in Kent. The train was travelling at
89 mph (143 km/h). It did not derail, and no-one on the train was hurt, but the train
was damaged by the impact. The van was severely damaged and the van driver
suffered serious injuries.

The van driver was delivering a parcel to a property on the far side of the crossing.
He initially went to an incorrect address, where he was given directions which
involved going over the crossing to reach the correct address. The level crossing
was equipped with power-operated gates, controlled by a button at the side of
the approach road, and a telephone which vehicle drivers were required to use to
contact the signaller to obtain permission to cross. Being unfamiliar with user worked
crossings, the van driver did not notice the telephone and pressed the button to
operate the gates. They opened, and so he returned to his van to drive across the
crossing, believing it was safe to cross.

There were multiple signs associated with the crossing which were placed in a way
that meant they did not stand out to the van driver. The van driver had been told that he needed to press a green button to open the gates at the crossing, and he was
focused on locating this button. The fact that the gate opened when the button was
pressed, coupled with the van driver’s previous experience of other types of level
crossing, may have reinforced his view that it was safe to cross.

The RAIB has found that an underlying cause of the accident was that the system
where authorised users are responsible for briefing visitors about the safe way to use
private crossings, is unreasonable in present-day circumstances.

The RAIB has made four recommendations, the first directed to Network Rail, the
Department for Transport and the Office of Rail and Road to improve the signage at
private crossings and review the concept of authorised users. The second is directed
to the Department for Transport and the Office of Rail and Road to change the law
covering the signage at private crossings. The third is to Network Rail, to improve the safety of private crossings equipped with power operated gate opening equipment.
The last is also directed to Network Rail, to review the way in which it collects
and maintains data about regular users of private crossings, so that it can better
communicate important information about crossing safety.

Edited to replace the link to the RAIB report, original link had stopped working.

 

[Muzer]

Fantastic report, some particular quotes I think really needed to be said:

The RAIB has found that an underlying cause of the accident was that the system
where authorised users are responsible for briefing visitors about the safe way to use
private crossings, is unreasonable in present-day circumstances.

The van
driver’s goal was to locate the green button to operate the crossing, not to look for
and comprehend an array of signage.

It is no longer reasonable (if it ever was) to expect authorised users to provide information to everyone who may approach a private level crossing about how to use it safely.

 

[Bald Rick]

Fantastic report, some particular quotes I think really needed to be said:

I haven't read the report in detail. However the quotes here are a good example of how safety on the railway is improved through learning; the RAIB have now decided that something that has been deemed reasonable for the last (nearly) two centuries is now not so.

Resolving this will, of course, cost money. And quite a lot of it. It rather puts ORR in a bit of a spot.

Expect the closure programme of user worked crossings to step up a notch.