Signalling - Limit of control

[paterson00]

I was told the other night that the limit of control of a signal is, up to the next signal and its associated overlap.

Is that it? Is it as straight cut as that or is there more to know about that?

[Colin]

In simple terms, the answer is essentially yes - at least as far as we are told on the operational side of the business.

I hope I'm not teaching you to suck eggs, but as you're probably aware, we mainly have absolute overlaps on LU....

So a given signal's limit of control (or signalling section) will be from the signal itself (which encompasses the overlap of the previous signal) through the berth track and onto it's own overlap past the next signal.

As I say, that is how it's taught on the operational side - those who actually work with it, such as Tom or DistrictSOM may have different ideas/perspective on the subject....

I have been intending for some time to reproduce a presentation I once had to do about overlaps onto here as it's a complicated thing to get your head round if it's a new subject - time and loss of the original presentation have got in the way but this is a timely reminder for me to get round to redrawing it & posting it up.

[londonstuff]

I'd appreciate that...it's quite a technical, but interesting, topic.

[Tom]

In a nutshell, that's it. All tracks and points from one signal to the end of the overlap of the following signal.

However, what you need to cause a signal to clear (the signal selection) is often different. The Signal Selection encompasses things like overrun protection, flank points, trackside equipment that has to be proved in a certain state, etc.. as well as the Limit of Control being unoccupied.

[paterson00]

I would also really appreciate that as I am a network rail trained chap who has never heard of (at least as far as I recall) an absolute overlap. Yours in hope....

[paterson00]

PS would I be correct in saying that an absolute overlap is one which the train WILL stop within as apposed to some of the overlaps on Network Rail where the trains speed will simply be reduced to minimise damage (on TPWS a train travelling over 70 mph will not be able to stop within the full overlap if stopped by TPWS)

[Tom]

You won't have heard of it mainly because it's a fairly outdated term now used only by the Operational Learning team.
With an absolute overlap, the signal has it's own replacing joint not far beyond, and a second joint at the end of the overlap. The so called 'ordinary overlap' (which only exists in about three or four places now) combines the berth and overlap into a single track circuit and the signal is only replaced once a train moves beyond the end of the overlap.

[paterson00]

Ok now I have a better undersatnding of it but would still really appreciate that upload.

"Overrun protection" by that I take it you mean a trainstop to protect an overrun of the signal?

[Tom]

No. I'm referring to replacing track circuits of conflicting signals which will cause your signal to return to danger in the event of a SPAD.

[Colin]

Jul 11, 2010 11:47:16 GMT @paterson00 said:

Ok now I have a better undersatnding of it but would still really appreciate that upload.

Noted

I'll try and get it done in the next couple of days, then Tom can rip it apart! ;D ;D

[Tom]

Be my guest
I've had to do it to those who should have known better in recent weeks so to offer constructive criticism to someone who actually has an interest would be a welcome change!

[Colin]

I was awaiting a check for accuracy from a trusted pair of eye's, but I also have about as much patience as an arsonist in a petrol station.....so here you are....

A few points to note before we start:

• The images do break our forum rules in terms of width, but it's a necessary evil in this case
  
• We are only interested in overlaps & signalling sections; this is not a full description of the whole signalling system
  
• The examples use an automatic signalling area for simplicity, but the same basic principles apply to all signals
  
• Take your time to understand the more complicated diagrams
  
• This is what operational staff are taught by LU's Operational Learning department, and is reproduced solely from my memory - signalling engineers may disagree with it!!

Fig 1 - we'll start with the basic area which shows three signals, a train and some insulated block joints. The insulated block joints are placed in the running rails (the rails that train wheels run on) such that each section of running rail is now separated from it's neighbour.


Fig 2 - as train 411 passes over the insulated block joint located at signal A661, the signal changes to red...


Fig 3 - then as the back of the train passes over the insulated block joint located at signal A663, signal A661 will clear to Green. With train 412 now approaching signal A661, we can see that this clearly isn't a safe situation.


Fig 4 - we can make this safer by moving signals back so that there is a greater distance between trains before the signal behind clears.


Fig 5 - the area from the start of the signal section to the next signal is known as the berth track, and the area from the next signal to the end of the signalling section is the overlap. This is known as an ordinary overlap, with the overlap distance taking account of things like train speeds, gradients, etc.


Fig 6 - now let's show the signal sections for all of the signals on the diagram...


Fig 7 - It looks complicated now, so let's follow train 411 again and see what happens. As train 411 passes over the insulated block joint at the start of the A661's signalling section, signal A661 goes to Red.


Fig 8 - As train 411 continues along, passing signal A663, we can see that the next train (412) still has a Red aspect on A661 because train 411 is still in A661's signalling section.


Fig 9 - as train 411 passes over the insulated block joint into A663's signalling section, A663 will go Red. A661 will clear to Green as the rear of the train passes over the same insulated block joint. The overlap therefore maintains a safe braking distance between trains.


Fig 10 - we are relying on just one Red signal behind a train as it's protection; with the ordinary overlap system it's possible for a train to be in a given signal's overlap with a Green signal behind it. Have another look at figure 8.

So ideally we want two red signals behind any train. We can achieve this by placing an insulated block joint at each signal, known as a replacement block joint. This now makes the signalling sections larger - we'll look at A661's signalling section first....


Fig 11 - and now for all of the signals. Do take a moment to understand all the sections as it can be quite difficult to get your head around it when you first try and understand it.


Fig 12 - of course it'll be easier if we follow train 411 again. We'll start with train 411 passing over the replacement block joint - this will put signal A661 back to red.


Fig13 - train 411 continues along A661's signalling section, putting A663 to Red as it passes over A663's replacement block joint.


Fig 14 - notice now that train 411 is on A661's overlap section but now has two Red signals behind it? That's much safer and it's known as the absolute overlap system.


Fig 15 - as train 411 leaves A661's signalling section, A661 will clear to Green allowing train 412 to proceed.


Fig 16 - the same process happens all over again as train 411 passes over the replacement block joint at signal A665.


Short Overlaps

It's also worth mentioning short overlaps as I'm sure it would otherwise come up in further discussion. Station starting signals (the ones at the end of a platform) have a very short overlap on them because trains are moving much slower when they depart a platform. This is safe because the following train will be slowing down to stop at the platform.

The shorter overlap takes account of these slower speeds allowing more trains to pass through - if the overlaps on station starters took account of full line speeds with trains moving much slower, it would take much longer than necessary for the relevant signals to clear.

It is because of these short overlaps that trains must slow to 5mph when passing station starting signals if they are not stopping at the platform. There are of course exceptions to this where a "full speed overlap" exists at a station starting signal and trains will therefore be able to pass through a given platform at a higher speed in those cases.

[slugabed]

I think I get it,Colin...But,just a detail question:
Do station starting signals with a full speed overlap have an identifying marker of some sort?
Or is the driver expected to know them "by heart" as part of the route-learning procedure?
Clearly an experienced driver WILL know the ones on his route by heart...

[Colin]

No, they are not in anyway different.

This would be part of a drivers route knowledge and they would be expected to know it without any further reference.

[Deleted]

Let me see if I've understood

So the track consists of alternate "berth" and "overlap" sections (shown as being the same length in the diagrams but this is not necessarily so - only the "berth" sections have to be long enough contain a train).

Each berth section is unique to the signal behind it. Each overlap section is shared between two signals, the one behind it and the one behind that. Each signal controls a berth and two overlap sections. Only when all three of these sections are clear can the signal clear.

How am I doing?

[Colin]

Jul 13, 2010 21:19:29 GMT @innocentabroad said:

How am I doing?

Quite well

Jul 13, 2010 21:19:29 GMT @innocentabroad said:

So the track consists of alternate "berth" and "overlap" sections (shown as being the same length in the diagrams but this is not necessarily so - only the "berth" sections have to be long enough contain a train).

There are no fixed lengths as such - the diagrams were only done that way to help make it easier to understand.

The berth track would be at least a trains length but can be much much more; the overlap behind it covers the safe braking distance of a following train and can therefore vary quite a lot in distance required depending on local geography, line speed, etc. Ditto the next overlap, which is covering the next safe braking distance to the next berth track.

It's because of those two reasons that platform areas are always berth tracks (hence the railway term "fully berthed in the platform").

Jul 13, 2010 21:19:29 GMT @innocentabroad said:

Each berth section is unique to the signal behind it.

Keeping it simple, the berth track is unique to each signal section, yes.

Jul 13, 2010 21:19:29 GMT @innocentabroad said:

Each overlap section is shared between two signals, the one behind it and the one behind that.

The overlap is on two signalling sections, but I wouldn't say it's shared between the two in the way I'm reading you as saying it.

Each signal section uses a different electrical frequency so although two signals may read over the same section of track, they cannot see each other and are therefore independent - all they're both looking for is the presence of a train.

Jul 13, 2010 21:19:29 GMT @innocentabroad said:

Each signal controls a berth and two overlap sections. Only when all three of these sections are clear can the signal clear.

Again in simple terms, yes.

[paterson00]

I really appreciate you taking the time to do that Colin and my next question is, presumably the berth track and the overlap could all be one track circuit? I'm sure that that is common practice on Network rail but correct me if I am wrong.

[tubeprune]

Jul 14, 2010 4:15:03 GMT @paterson00 said:

I really appreciate you taking the time to do that Colin and my next question is, presumably the berth track and the overlap could all be one track circuit? I'm sure that that is common practice on Network rail but correct me if I am wrong.

I think you've missed the point here. Colin was trying to explain that it is usual practice on LU to have separate overlap and berth tracks, then you don't get the NR scene of a train passing a signal and it remaining green until the front reaches the start of the section of track controlling it 183 metres ahead. On LU, the signal is replaced to danger as it passes it (or very close to it).
[correction]
So, each signal is responding to 3 separate track circuits: Two form the section being protected and one the overlap of the next signal. On LU automatic sections they are usually numbered/lettered - e.g 611a, 611b (for the first signal, say A611) and 613a for the overlap of A613. All three circuits must be up (complete) for the signal to clear.[/correction]

[paterson00]

Jul 14, 2010 6:43:24 GMT tubeprune said:

Jul 14, 2010 4:15:03 GMT @paterson00 said:

I really appreciate you taking the time to do that Colin and my next question is, presumably the berth track and the overlap could all be one track circuit? I'm sure that that is common practice on Network rail but correct me if I am wrong.

I think you've missed the point here. Colin was trying to explain that it is usual practice on LU to have separate overlap and berth tracks, then you don't get the NR scene of a train passing a signal and it remaining green until the front reaches the start of the section of track controlling it 183 metres ahead. On LU, the signal is replaced to danger as it passes it (or very close to it).

So, each signal is responding to 3 separate track circuits: Two form the section being protected and one the overlap. On LU automatic sections they are usually numbered/lettered - in Colin's example 611a, 611b and 611c. All three circuits must be up (complete) for the signal to clear.

I'm glad that you highlighted that. The answer is in the original explanation isn't it? So my next question is, if I am correct in saying that Network Rail common practice is to have on track circuit as the berth and overlap track, why is this done?

Presumably it is only laid out this way in rural area's where it isn't necessary to have a high trains per hour figure unlike the Underground and that the Network Rail signalling would become more like an Underground set-up as demonstrated as the lines become nearer to the city centres, London etc?

[Colin]

NR (Network Rail) has to have one set of standardised rules for the whole of the UK. This encompasses anything from a slow single passenger car in rural areas and single locomotives through to 12 car express passenger trains and long freight trains.

Having a set up similar to LU's just wouldn't work with this type of mixed traffic railway.

[Deleted]

The berth track would be at least a trains length but can be much much more

I can see that it could be from a signal engineer's standpoint but if the aim is to maximise line capaicty why would it be? AIUI safety is provided by the overlapping.

Each signal section uses a different electrical frequency so although two signals may read over the same section of track, they cannot see each other and are therefore independent - all they're both looking for is the presence of a train.

You could've said so first off .

I suppose you would also argue that the track immediately beyond a convergence - such as the e/b platform at Leytonstone or the w/b immediately beyond our old friend Praed St junction - isn't "shared" either. Or perhaps their is some nuance I'm missing...

[norbitonflyer]

Jul 14, 2010 21:12:05 GMT @innocentabroad said:

The berth track would be at least a trains length but can be much much more

I can see that it could be from a signal engineer's standpoint but if the aim is to maximise line capaicty why would it be? AIUI safety is provided by the overlapping.

I've been reading the Hidden report into the Clapham Junction disaster.

AIUI the overlap (usually 200 yards) is there simply to ensure a train is well clear of a signal before another train is allowed into the section approaching taht signal. This is to provide a safety margin in the event of an overshoot.

The spacing of the signals is determined by the braking curves of the trains working the line - in 3-aspect territory the maximum distance required that any train permitted to use the line will need to stop, in 4-aspect territory it is half that distance (I am simplifying slightly, as sighting distances come into it). At Clapham Junction, the spacings are between 700 and 800 yards, including the overlap.

Originally, a "berth" track circuit was a short length of track usually located at the point a train would stand when held at a red signal, and only needed to be long enough to ensure that a train standing there would be detected - it could be much shorter than a train. The signalling was interlocked such that once a train had been admitted to the section, and the signal protecting that section put to danger behind it, the signal could not be cleared again until the berth circuit at the next signal had been occupied and then cleared again, indicating that the train had passed through the section.

Nowadays we have continuous track circuits, and the train should be detected at all times.

What happened at Clapham was that a false feed prevented the berth track circuit detecting a train, so that the signal it controlled only went to red when a train entered the overlap circuit beyond the next signal. A driver, seeing the signal go to red 30 yards ahead of him in these circumstances, stopped at the next signal to report the SPAD, but was of course then standing on the faulty berth circuit so that it was lost to the signalling system and another train entered the section and ran into the first. Indeed, a third train also entered the section. The power had gone off by then and the driver, believing that his train had lost power, was trying to coast the remaining mile to Clapham Junction so that he could detrain the passengers, but when the driver saw the rear of the second train he was able to stop in time. It was remarked that, even with three trains now in the section, the signal was still showing a proceed aspect.

[Tom]

Jul 14, 2010 17:42:32 GMT @paterson00 said:

I'm glad that you highlighted that. The answer is in the original explanation isn't it? So my next question is, if I am correct in saying that Network Rail common practice is to have on track circuit as the berth and overlap track, why is this done?

More than likely to save having lots of 183m track circuits that serve little purpose beyond replacing signals to danger.

I have also heard that it was so a guard would never see the signal replace. How that works (or doesn't work) in controlled areas was never really answered....

[Tom]

Jul 14, 2010 22:30:14 GMT norbitonflyer said:

AIUI the overlap (usually 200 yards) is there simply to ensure a train is well clear of a signal before another train is allowed into the section approaching taht signal. This is to provide a safety margin in the event of an overshoot.

Nice in theory but 200yds (183m, now reduced to 180m) prior to a collision occuring will have negligible effect for any train at a reasonable speed where no attempt has been made to stop before passing a signal at danger.

Don't forget that LU calculate their overlaps individually as there is a lot less variation in train characteristics.

[mrfs42]

Jul 14, 2010 22:30:14 GMT norbitonflyer said:

I've been reading the Hidden report into the Clapham Junction disaster.

I've just looked at the affected circuit diagram - in extract.

Without wishing to comment any further on the working practices that had evolved - this has nothing whatsoever to do with LU signalling!

It was in four-aspect territory without trainstops. Where's the similarity please?

Jul 14, 2010 22:30:14 GMT norbitonflyer said:

Nowadays we have continuous track circuits, and the train should be detected at all times.

Do we? Everywhere on NR or just in London?

Every day is a learning day, as is said.....

Jul 14, 2010 21:12:05 GMT @innocentabroad said:

Or perhaps their is some nuance I'm missing...

Er... Flank protection? I'm sure it's been hinted at before on the Forum.... Probably belongs in another thread too! ;D

[mrfs42]

Jul 14, 2010 23:38:07 GMT Tom said:

Jul 14, 2010 17:42:32 GMT @paterson00 said:

I'm glad that you highlighted that. The answer is in the original explanation isn't it? So my next question is, if I am correct in saying that Network Rail common practice is to have on track circuit as the berth and overlap track, why is this done?

More than likely to save having lots of 183m track circuits that serve little purpose beyond replacing signals to danger.

I have also heard that it was so a guard would never see the signal replace. How that works (or doesn't work) in controlled areas was never really answered....

Why don't you raise the question 'elsewhere[1]' - I'm sure it has more to do with the 'guessing' of the position of the guard in a push-pull set than meets the eye.. (this also really belongs in another thread and probably another forum....)

[1] quite a few of them look over the wall here....

[paterson00]

Jul 14, 2010 22:30:14 GMT norbitonflyer said:

Jul 14, 2010 21:12:05 GMT @innocentabroad said:

The berth track would be at least a trains length but can be much much more

I can see that it could be from a signal engineer's standpoint but if the aim is to maximise line capaicty why would it be? AIUI safety is provided by the overlapping.

AIUI the overlap (usually 200 yards) is there simply to ensure a train is well clear of a signal before another train is allowed into the section approaching taht signal. This is to provide a safety margin in the event of an overshoot.

What does AIUI mean? Please excuse my lack of what may be basic knowledge.

[tubeprune]

AIUI= As I Understand It

The NR overlap was regarded as a nominal distance provided for drivers overrunning in foggy or wet weather. OS Nock in his "Railway Signalling" AC Black & IRSE, 1980, says that statistics in the 1970s showed that overruns were seldom more than 300ft (90m appprox.). In 1978, it was decided that a fixed overlap would be provided for all new designs, set at 600ft - double the statistical norm. It could be reduced for routes with less than 60mph line speed.

There is a simple rule for anyone trying to understand the complexities of railway signalling. Treat it like eating an elephant - only take small bites.

[Colin]

Jul 15, 2010 7:55:59 GMT tubeprune said:

There is a simple rule for anyone trying to understand the complexities of railway signalling. Treat it like eating an elephant - only take small bites.

That is absolutely true - and as I said at the start of my post with the diagrams, it was not a full description of the signalling system.

This thread is about a given signal's limit of control, and the overlap diagrams were a way of explaining that - anything more is really a thread drift and so a new thread should be started so that different aspects (pardon the pun) of LU's signalling system can be explained properly.

[paterson00]

Jul 15, 2010 7:55:59 GMT tubeprune said:

AIUI= As I Understand It

The NR overlap was regarded as a nominal distance provided for drivers overrunning in foggy or wet weather. OS Nock in his "Railway Signalling" AC Black & IRSE, 1980, says that statistics in the 1970s showed that overruns were seldom more than 300ft (90m appprox.). In 1978, it was decided that a fixed overlap would be provided for all new designs, set at 600ft - double the statistical norm. It could be reduced for routes with less than 60mph line speed.

There is a simple rule for anyone trying to understand the complexities of railway signalling. Treat it like eating an elephant - only take small bites.

Thank you very much for that, I had been wondering where the standard overlap measurement had come from as the same measurement is used for placing the AWS's

I take it the book you are referring to is the green book? I am going to ask another question in a new thread about recommended books to read to understand railway signalling and then take a bite at a time.