Level crossing stupidity...

[royaloak]

Not all controlled signals are TPWS fitted, if I would guess that less than half are fitted, it isnt as widespread as people think.

 

When I was running in and out of Waterloo it was said that in the morning peak going into Waterloo if you were running on greens you were late, double yellows was where we needed to be. 

[The Stationmaster]

As a signalman I have had to sit in meetings where management want to remove our box copies of the route emergency plans, which are the only things that we have that show ALL the crossings and signal and points with their 6 figure grid references and how to get to them. Why did they want to remove them? Well they are uncontrolled documents and *may* be out of date.

When I said how do I call the emergency services to a incident they backed down, so they are for reference purposes only.

And before you ask, no accommodation and footpath crossings are not shown on the box diagrams, nor is the yardage to the signals.....

 

Andy G

 

I could never understand how some other Railways/Regions could get away without showing signal yardages on 'box diagrams - it was standard practice on the GWR/Wr going right back into the late 19th century.   Interesting comment regarding crossings - when I wrote our Area's emergency plan back in the 1970s (when we first told to prepare one) I put in OS map references for all our occupation etc and footpath crossing having first plotted them and mileposts as accurately as I could onto 1" OS maps and the emergency services I spoke at that time all said they wanted map references rather than anything.  We had about 120 miles of railway, including branches, on our patch and I'd defy anybody to find their way to a particular spot by road working solely from milepost mileage.

 

Removing such information from signalboxes strikes me as totally daft as even if they are in the Sectional Appendix as the milepost mileage is still irrelevant when it comes to outside bodies (and to be honest doesn't make much sense to some on the inside either).

[uax6]

But then you also get into the silly issue if you happen to deal with the ambulance service first, as they want postcodes Not 6 figure gridrefs! Try having a guess at a postcode for a crossing in the middle of nowhere!

What I find more strange is that the EP is not common in boxes round here. We have it over the desk and it is the singular most useful document in the entire box...

 

Andy G

[phil-b259]

But then you also get into the silly issue if you happen to deal with the ambulance service first, as they want postcodes Not 6 figure gridrefs! Try having a guess at a postcode for a crossing in the middle of nowhere!

What I find more strange is that the EP is not common in boxes round here. We have it over the desk and it is the singular most useful document in the entire box...

 

Andy G

 

That is more due to the march of technology than anything else allied with the inability (or them being too afraid they will end up in front of the guys in the funny wigs) for someone to update the rules.

 

As you say all the 'normal' emergency services are geared up to use Postcodes in their computer systems that locate the nearest responder etc and for said responder to get to site quickly.

 

Grid references are only of use if you are using OS based mapping - which is ideal when we are talking about hill walkers say who need rescuing from the fells and are thus a long way from habitation.

[Colin_McLeod]

Why not give post codes for level crossings? They may not get any mail but if that's the system the emergency services use then it could save lives by helping them to get there quicker.

[Titan]

Why not give post codes for level crossings? They may not get any mail but if that's the system the emergency services use then it could save lives by helping them to get there quicker.

 

Great for urban areas, but I suspect that in the countryside where occupation crossings usually are, one postcode could cover several miles and a number of different crossings, you would be lucky to get the right road!

[Colin_McLeod]

To spell it out: I meant that each and every level crossing of all types be allocated it's own unique post code which would be fed into the emergency services system and tells the emergency services on their geographical database where each level crossing is. The idea could extend to cover other non railway locations that might need the emergency services.

[JDW]

That's actually quite a good idea, the "post"code wouldn't even necessarily need to be a local one, so long as the right people know it and can use it, nor even the usual format - besides the usual AB12 3CD type, the system can seemingly cope with other formats like FIQQ 1ZZ (Falkland Islands), so there are plenty of "spare" options, but of course, it would need lots of different people to work together...

[uax6]

But of course 6 figure grid refs do all this and so much more accurately too.

 

Andy G

[Reorte]

But of course 6 figure grid refs do all this and so much more accurately too.

Postcodes would probably tie in with existing systems much more easily, just need a database update.

 

If you give the emergency services a grid ref would they ask you for something else instead?

Edited February 1, 2017 by Reorte

[martin_wynne]

Interesting reference to the drawbacks of axle counters in this RAIB report today:

 

 https://www.gov.uk/government/news/serious-signalling-irregularity-at-cardiff-east-junction

 

Martin.

[The Stationmaster]

Interesting reference to the drawbacks of axle counters in this RAIB report today:

 

 https://www.gov.uk/government/news/serious-signalling-irregularity-at-cardiff-east-junction

 

Martin.

 

Just one of several drawbacks which they have, regrettably I fear it might only be a matter of time before another emerges in the 'wrong' way.  But in normal working they do seem to be far more reliable than track circuits and they are probably cheaper overall in a modern installation.  However generally I would think that treadles are still probably the best system for level crossings.

[uax6]

Postcodes would probably tie in with existing systems much more easily, just need a database update.

 

If you give the emergency services a grid ref would they ask you for something else instead?

The fire and police (in experience) are happy with 6FGR, but the ambulance service want postcodes. But interestingly I've been told that they can change to a 6FGR on their system, but it appears that they may not be well versed in doing so.

 

Andy G

[ejstubbs]

Interesting reference to the drawbacks of axle counters in this RAIB report today:

 

 https://www.gov.uk/government/news/serious-signalling-irregularity-at-cardiff-east-junction

 

Martin.

 

Why does it say: "The new signalling system uses axle counters for train detection, and in this situation the system would not have identified that the train was in the wrong place"?  The summary of the incident doesn't seem to suggest a belief that the axle counters were actually faulty.

 

Is it because the axle counters only detect a train passing one specific location, whereas a track circuit covers a complete stretch of track?

 

Looks like good work by the driver to have spotted the problem before anything serious happened, though.  Is it usual for drivers to check that points have been set correctly?  (Depending on the speed the train is doing, presumably - I assume it's hardly practical at 125mph.)

Edited February 1, 2017 by ejstubbs

[uax6]

Track ccts detect the train throughout their length. Axle counters count the train in at the beginning of the section, and then again at the end of the section. In this case as the route wasn't set (and the signalling system appears to not know that the points where there) it would have counted the train in, but not have noticed that the train had wondered off somewhere else. It would have just shown the train being somewhere on the expected route until it counted it out again.

 

To say track ccts would have noticed this train is a tad difficult, because presumably the works would have removed the track cct before the end of the possession, so would not have been there either.

 

From a personal point of view, I prefer track ccts to axle counters, as, with good maintenance they are far more reliable than axle counters, and far less suseptable to being upset by P-way wielding shovels near them.... The railway likes them because they are cheap.

 

Andy G

[Coryton]

The fire and police (in experience) are happy with 6FGR, but the ambulance service want postcodes. But interestingly I've been told that they can change to a 6FGR on their system, but it appears that they may not be well versed in doing so.

 

Andy G

 

I would have thought in a world where most people carry a GPS receiver around with them, grid references would be more useful than ever to the emergency services.

 

(And I believe that the widespread availability of cheap GPS chips on phones was driven by a US government mandate for phones to be able to report their position when making emergency calls...gets around the chicken-and-egg situation that new technology is expensive so most people don't want to pay for it so the market is small so the new technology is expensive...)

[avonside1563]

There's also this option for location identification now, specifically designed for places which don't have an address identifying them. http://what3words.com/

[Coryton]

Why does it say: "The new signalling system uses axle counters for train detection, and in this situation the system would not have identified that the train was in the wrong place"?  The summary of the incident doesn't seem to suggest a belief that the axle counters were actually faulty.

 

Is it because the axle counters only detect a train passing one specific location, whereas a track circuit covers a complete stretch of track?

 

With the usual proviso that I have no professional involvement in the railways, if I've read the report correctly the train would have been diverted onto a different line and thus into a different block. As it would have entered the block from an unexpected direction (since the points were supposed to be locked out of use) there wouldn't have been axle counters to detect it entering the block, and it would have presumably been invisible to the signalling system until it left the block. I don't know what happens when more train leaves a block than was detected entering it, but presumably it is flagged in some suitable fashion.

 

On the other hand, a track circuit detects the presence of a train no matter where it comes from, whether it be conventionally into the block, via a surprise set of points or dropped neatly onto the track by helicopter.

 

To say track ccts would have noticed this train is a tad difficult, because presumably the works would have removed the track cct before the end of the possession, so would not have been there either.

 

I don't follow that - the crossover itself which had been taken out of use might have had any track circuit removed, but why would the line that the train would have been diverted onto not have had a fully functioning track circuit?

 

In thinking this through it's occurred to me that axle counters can potentially do something that I don't think a track circuit can, which is to detect if a train enters an already-occupied section. This sounds like quite a useful advantage, assuming there is some method of sending a stop signal to a train other than via signals.

 

Edited to add: I like the way the "news story" on the government web site is illustrated by a "Library image of a set of points"

Edited February 1, 2017 by Coryton

[Reorte]

Another person with little knowledge sticking his oar in here, but if the other line would've been on a different track circuit, and is now done by axle counters instead, presumably there are axle counters for both entering and leaving (can't see how else it would work) so the train would be detected as being on that line, even if it wasn't supposed to be, and even if it would still show as being in the one it had left. And either no axle counters or a track circuit being removed for a line that wasn't supposed to be used anyway would not show up either way. So the situation might block more than track circuits would've done but is still safe as far as knowing what's occupied and what isn't (better to have it showing a block as occupied when it isn't than vice-versa).

[Grovenor]

In this case there was no axle counter in the crossover between the parallel tracks as that crossover was supposed to be secured out of use pending removal and hence was not part of the signalling design. Hence a train wrongly going through the crossover would not have been seen in the axle counter block covering the new section that it entered.

Had the design used track circuits then the train wrongly taking the crossover would have been detected by the track circuit when it reached the wrong line.

The problem in this case was not a design issue but a failure of the project team to secure the points concerned before handing the lines back to traffic, and that is what the RAIB will be investigating.

Regards

[Coryton]

In this case there was no axle counter in the crossover between the parallel tracks as that crossover was supposed to be secured out of use pending removal and hence was not part of the signalling design. Hence a train wrongly going through the crossover would not have been seen in the axle counter block covering the new section that it entered.

Had the design used track circuits then the train wrongly taking the crossover would have been detected by the track circuit when it reached the wrong line.

The problem in this case was not a design issue but a failure of the project team to secure the points concerned before handing the lines back to traffic, and that is what the RAIB will be investigating.

Regards

 

The use of axle counters doesn't (in itself) change the likelihood of an out-of-use crossover being inadvertently used. It does change the potential impact. So overall risk goes up.

[royaloak]

Another person with little knowledge sticking his oar in here, but if the other line would've been on a different track circuit, and is now done by axle counters instead, presumably there are axle counters for both entering and leaving (can't see how else it would work) so the train would be detected as being on that line, even if it wasn't supposed to be, and even if it would still show as being in the one it had left. And either no axle counters or a track circuit being removed for a line that wasn't supposed to be used anyway would not show up either way. So the situation might block more than track circuits would've done but is still safe as far as knowing what's occupied and what isn't (better to have it showing a block as occupied when it isn't than vice-versa).

If the train had gone over the set of points onto the 'wrong' line it would only be detected when it got to a set of axle counter heads (which can be several miles away), until it reaches them the system will assume that line is clear even though the train was actually on that line but the track the train is supposed to be on would show occupied even though the train was no longer on it.

 

A track circuit would pick the train up the instant it got onto the wrong' line.

 

Anyhow well done to the driver for spotting the problem and stopping short and turning what could have been a very serious incident into nothing more than a oops incident.

Edited February 1, 2017 by royaloak

[royaloak]

In thinking this through it's occurred to me that axle counters can potentially do something that I don't think a track circuit can, which is to detect if two trains attempt to enter a section simultaneously. This sounds like quite a useful advantage, assuming there is some method of sending a stop signal to a train other than via signals.

They will if both trains go over a set of axle counter heads as they enter the section, in this case that wouldnt have happened because the set of points which were set wrong didnt have any associated Axle counter heads, the train would have crossed onto the other track without being detected meaning there is no reason another train coudnt be signalled into the section even though it was already occupied.

 

A track circuit will 'find' the train as soon as its wheels break the circuit, ie as soon as it came off the set of points, which would show in the signal box as an occupied track circuit and that would prevent a route being set which involved that track circuit, any route already set would be cancelled with the associated signals reverting back to danger instantly.

 

In this day and age with reversible lines its a good job the driver was on the ball, at some locations (Bristol Temple Meads for one) you can be routed into a platform but actually go 2 or 3 different ways, you dont know which way you are going until you get to the points and then you have to work out where your last set of points are to get you where you want to go actually are, you get no help from the signal or the route indicator bar the platform number.

Edited February 1, 2017 by royaloak

[ejstubbs]

it's occurred to me that axle counters can potentially do something that I don't think a track circuit can, which is to detect if two trains attempt to enter a section simultaneously.

 

Do axle counters know which way the train is travelling as it counts the axles?  I assume they must do, otherwise two identical trains (eg two 3-car multiple units) entering at opposite ends of a section at different times might look to the signalling system like one train traversing the section, and it would think the section was unoccupied when in fact there would be two trains in it   (which of course shouldn't happen anyway, but if it did I imagine it would be useful to know about it rather than blithely allow yet another train to run into the section and pile in to the wreckage of the first two...hmm, cheery thoughts for a Wednesday afternoon)

[Coryton]

They will if both trains go over a set of axle counter heads as they enter the section, in this case that wouldnt have happened because the set of points which were set wrong didnt have any associated Axle counter heads, the train would have crossed onto the other track without being detected meaning there is no reason another train coudnt be signalled into the section even though it was already occupied.

 

Indeed.

 

And - if the train had gone over the points and carried on without hitting something, the signalling system would have seen a set of axles leaving the block that never entered it. I presume at this point signals would automatically get set to red until someone worked out what had just happened?