Simple Signalling/Section Question - Single Track Line

[bécasse]

For shunting ahead into a single line (and it would be a common practice in at least one direction at just about every signalled station on a single line), the provision of a "shunt ahead" arm under the main arm on the starting signal was probably far more useful than providing an advanced starting signal and would have had a much lower cost. The main advantage of providing such a signal was that it authorised movements which would otherwise have required a hand signal from the signalman and at some stations the siting of the signal box might have made that difficult and/or time-consuming. Both starting and shunt ahead signals would share much the same interlocking, although they would obviously (except at some ER/NER locations) be locked to be mutually exclusive pulls, and any electric locking with the single line instrument (by no means universal especially historically) would differ. There were probably many, many more locations with a shunt ahead arm than an advanced starting signal, but even so most single line passing stations would have had neither.

Edited August 25, 2021 by bécasse

[Nick C]

1 hour ago, The Stationmaster said:

The methodology for shunting onto a single line has - I'm fairly sure, been changed anyway because it now has little utility on the national network and the heritage railways have in any case got to take account of various very different human factors.  However if they are still using (as some to my knowledge were not very many years back) BR era Block Regulations there is no logical reason at all to provided an Outer Home Signal on a single line with the possible exception of a terminus - but it wouldn't make any difference to the traditional procedure for occupying the single line for shunting purposes because it confers no additional protection at  all.   The only possible advantage of adding an Advanced Starting  Signal if the railway concerned is using the BR era Regulations is that it would serve as a reminder to a Driver etc making a shunt move not to go off into the section - but that is the only advantage it confers.  And in any case under ROGS a heritage etc railway has no legal obligation to provide such additional provided it has taken the necessary action and has the necessary data and paperwork from a Competent Person to support what it is providing.  Hence, for example, what I recommended, as a  Competent Person, to such a railway in terms of provision at a proposed new single line crossing loop is fully supported by all the necessary paperwork and covered by professional indemnity insurance.

I've not looked in detail at the BR block regulations to see what's changed, but one key thing to remember with heritage railways is that they tend to carry a lot more traffic than an equivalent single line would have done in BR days - so, for example, at a terminus you may well need to be able to accept the next train while the previous one is still running round - obviously you can't do this without the outer home and advanced starter. Most such stations were either not termini in BR days, or would generally only have had one train present.

 

Looking at a few diagrams on the SRS site, it seems that virtually all of the LSWR single-line termini had advanced starters, but few had outer homes - presumably this would have allowed run-rounds to take place without the need for blocking back.

[Michael Hodgson]

3 minutes ago, bécasse said:

For shunting ahead into a single line (and it would be a common practice in at least one direction at just about every signalled station on a single line), the provision of a "shunt ahead" arm under the main arm on the starting signal was probably far more useful than providing an advanced starting signal and would have had a much lower cost. The main advantage of providing such a signal was that it authorised movements which would otherwise have required a hand signal from the signalman and at some stations the siting of the signal box might have made that difficult and/or time-consuming. Both starting and shunt ahead signals would share much the same interlocking, although they would obviously (except at some ER/NER locations) be locked to be mutually exclusive pulls, and any electric locking with the single line instrument (by no means universal especially historically) would differ. There were probably many, many more locations with a shunt ahead arm than an advanced starting signal, but even so most single line passing stations would have had neither.

Shunt Ahead arms seem to me to have been more common on the Southern and its constituents than elsewhere, a hand signal being cheaper to implement.

 

Quite a number of safeguards such as locking signals with the block (Line Clear One Pull, proving signals to be On, sequential locking, Welwyn Control etc) were not required under the legislation requiring block working for passenger trains and not even available in the early days.  They were introduced as safety improvements following accidents, but they all had a cost.  So they were rolled out very gradually and priority was given to the lines most at risk (and of course "locking the stable door after the horse has bolted" in some cases).  This meant busy double track routes because where the service is infrequent, there is less chance that a mistake will result in a collision.

 

Single track lines, especially short dead-end branches, could be characterised as sleepy compared to the multiple track main lines.  Nothing would move for an hour or two, then it would suddenly become hectic for a short while (even more so at places where trains were crossing).  These lines would be low on the priority list for adding any optional safety enhancements.  By far the most important additional control on a single line was locking the starter with the instruments - because of the obvious risk that a mistake in that could result in a head-on collision.

Outer Home and/or Advance Starting signals on double track lines were often desirable to augment line capacity on busy routes and to give stations more opportunity to make shunts between trains or over junctions. Line capacity was often less of an issue on single track lines, because if you really can't handle the traffic you need double track.  One place on a single line you typically do want an outer home would be at a junction - so that you could accept a train from the single line when you couldn't set the route for it while there were other trains passing on the main line. 

 

Heavy traffic on single lines was perhaps more common through sparsely populated long routes such as the Scottish and Cambrian lines, the M&GN and S&DJR.  But even where line capacity is an issue, putting Advanced Starting signals or Outer Homes at the crossing loops didn't make much difference.  The key was pathing of services so that they crossed efficiently and re-arranging crossings in the case of late running.

 

[bécasse]

1 hour ago, Michael Hodgson said:

Shunt Ahead arms seem to me to have been more common on the Southern and its constituents than elsewhere, a hand signal being cheaper to implement.

 

You might be right about the Southern, I don't know, but the point I did try to get over was that they would normally only be provided where there was some problem in giving the relevant hand signal, usually because of the position of the signal box. In those days nobody, including the Southern, spent money unnecessarily.

[The Stationmaster]

3 hours ago, bécasse said:

For shunting ahead into a single line (and it would be a common practice in at least one direction at just about every signalled station on a single line), the provision of a "shunt ahead" arm under the main arm on the starting signal was probably far more useful than providing an advanced starting signal and would have had a much lower cost. The main advantage of providing such a signal was that it authorised movements which would otherwise have required a hand signal from the signalman and at some stations the siting of the signal box might have made that difficult and/or time-consuming. Both starting and shunt ahead signals would share much the same interlocking, although they would obviously (except at some ER/NER locations) be locked to be mutually exclusive pulls, and any electric locking with the single line instrument (by no means universal especially historically) would differ. There were probably many, many more locations with a shunt ahead arm than an advanced starting signal, but even so most single line passing stations would have had neither.

It is interesting to compare this with what the GWR were doing in the years prior to the Grouping and there the conclusion was reached that effectively there might be a choice between an Advanced Starting Signal and a Shunt Ahead arm below the Starting Signal depending on local circumstances but at many places the level of traffic justified neither.  A survey the previous year had found that on single lines Advanced Starting Signals at that time outnumbered Shunt Ahead arms but if the recommendations made in connection with the survey were followed the situation would be reversed.  

 

The final decision, in 1918, was that each location would in future be considered individually in order to reach the best solution for that place in relation to the level of traffic and so on.  Judging by signal box diagrams this continued to be the practice after then and there were of course even locations with a Shunt Ahead arm below the Advanced Starting Signal.   The position of signals in relations to the signal box seems to have had liittle bearing on what actually happened except, of course, when a Shunt Ahead arm was provided below an Advanced Starting Signal that had difficult sighting from/of the signal box.  But apart from that there were plenty of places where shunts onto the single line would have potentially started quitea way from the signal box so distance would seem to have had little impact in relation to other considerations.  Overall the GWR/WR was quite a prolific user of Shunt Ahead subsidiaries on both double and single lines but no doubt every example was considered on the same basic criteria.

[The Stationmaster]

2 hours ago, Michael Hodgson said:

Shunt Ahead arms seem to me to have been more common on the Southern and its constituents than elsewhere, a hand signal being cheaper to implement.

 

Quite a number of safeguards such as locking signals with the block (Line Clear One Pull, proving signals to be On, sequential locking, Welwyn Control etc) were not required under the legislation requiring block working for passenger trains and not even available in the early days.  They were introduced as safety improvements following accidents, but they all had a cost.  So they were rolled out very gradually and priority was given to the lines most at risk (and of course "locking the stable door after the horse has bolted" in some cases).  This meant busy double track routes because where the service is infrequent, there is less chance that a mistake will result in a collision.

 

Single track lines, especially short dead-end branches, could be characterised as sleepy compared to the multiple track main lines.  Nothing would move for an hour or two, then it would suddenly become hectic for a short while (even more so at places where trains were crossing).  These lines would be low on the priority list for adding any optional safety enhancements.  By far the most important additional control on a single line was locking the starter with the instruments - because of the obvious risk that a mistake in that could result in a head-on collision.

Outer Home and/or Advance Starting signals on double track lines were often desirable to augment line capacity on busy routes and to give stations more opportunity to make shunts between trains or over junctions. Line capacity was often less of an issue on single track lines, because if you really can't handle the traffic you need double track.  One place on a single line you typically do want an outer home would be at a junction - so that you could accept a train from the single line when you couldn't set the route for it while there were other trains passing on the main line. 

 

Heavy traffic on single lines was perhaps more common through sparsely populated long routes such as the Scottish and Cambrian lines, the M&GN and S&DJR.  But even where line capacity is an issue, putting Advanced Starting signals or Outer Homes at the crossing loops didn't make much difference.  The key was pathing of services so that they crossed efficiently and re-arranging crossings in the case of late running.

 

The most effective way of improving things on single lines was  - additional block sections apart - to do what the GWR did on the Taunton - Barnstaple/ Minhead branches and put in trap points at the exit end of crossing loops and where double lines reduced to single.  This speeded up things and a saved time when trains were crossed at such locations because both trains could then be run in under clear signals instead of having to be stopped at the Home Signal then enter their respective loops one after the other. 

[Ian J.]

I have a feeling, when I can get diagrams of my S&P route drawn up, I'm going to need some help

 

Why would trap points make a difference to clear signals? Are there regulations allowing combined movement if they're provided?

 

As the S&P will be a heritage-style operation, I can do things differently to 'normal' practice if such a difference would be noted as significantly beneficial for operational and/or safety reasons. Also, while it's notionally in L&SWR territory (mid Dorset-ish), it doesn't have to stick rigidly to L&SWR practice.

[RailWest]

15 hours ago, Ian J. said:

 

 

Why would trap points make a difference to clear signals? Are there regulations allowing combined movement if they're provided?

 

 

Simple - at a loop with trap points at the exit ends, if an incoming train over-runs the Starting signal then (hopefully) it just ends up 'in the dirt'. If no trap-traps, then it might go head-on into the train waiting at the opposing Home (or, even worse, if trains could enter simultaneously, into the side of the carriages :-(  ).

Edited August 26, 2021 by RailWest

[RailWest]

2 hours ago, Michael Hodgson said:

Shunt Ahead arms seem to me to have been more common on the Southern and its constituents than elsewhere, a hand signal being cheaper to implement.

 

 

In the case of single-lines, with the exception of the S&DJR (who always managed to be just that little bit different from normal L&SWR practice), I would be hard pressed to think of more than one or two examples of SA arms on the Western Section. On the other hand, with the GWR it seemed be almost de rigeur to have one, usually under an Advanced Starting to boot :-)

[Michael Hodgson]

 

3 minutes ago, RailWest said:

Simple - at a loop with trap points at the exit ends, if an incoming train over-runs the Starting signal then (hopefully) it just up 'in the dirt'. If no trap-traps, then it might go head-on into the train waiting at the opposing Home (or, even worse, if trains could enter simultaneously, into the side of the carriages :-(  ).

Quite.  So the rule is (in the absence of traps) you stop both trains at their home signals.  When they're stationary you let one in, then the other.  Safer, but it causes delay.  With traps at both ends both can run straight in without the passengers asking why the **** have we stopped short of the platform.

 

Running both trains in at the same time is a common mistake on model railways.  

[The Stationmaster]

26 minutes ago, Ian J. said:

I have a feeling, when I can get diagrams of my S&P route drawn up, I'm going to need some help

 

Why would trap points make a difference to clear signals? Are there regulations allowing combined movement if they're provided?

 

As the S&P will be a heritage-style operation, I can do things differently to 'normal' practice if such a difference would be noted as significantly beneficial for operational and/or safety reasons. Also, while it's notionally in L&SWR territory (mid Dorset-ish), it doesn't have to stick rigidly to L&SWR practice.

Trap points work to separate the two approaching trains in order to eliminate teh risk of them over-running signals and colliding with each other.  If there are no trap points at a crossing loop exit end when two trains arrive at more or less the same time they both have to be stopped at their respective Home Signals. One of the trains is then allowed to proceed into its loop and the other train cannot move until the first has come to a stand in its loop.  That restriction doesn't apply if there are trap points at the exit end of each loop because. if a train does over-run it will simply run out of railway.

 

Now one important thing to note is that this is 'old railway' because by the 1980s the Railway Inspectorate were becoming increasingly concerned about passenger trains derailing on trap points so as layouts were renewed and remodelled trap points were removed from passenger lines including loop lines authorised to be used by passenger trains.

[Ian J.]

31 minutes ago, The Stationmaster said:

Trap points work to separate the two approaching trains in order to eliminate teh risk of them over-running signals and colliding with each other.  If there are no trap points at a crossing loop exit end when two trains arrive at more or less the same time they both have to be stopped at their respective Home Signals. One of the trains is then allowed to proceed into its loop and the other train cannot move until the first has come to a stand in its loop.  That restriction doesn't apply if there are trap points at the exit end of each loop because. if a train does over-run it will simply run out of railway.

 

Now one important thing to note is that this is 'old railway' because by the 1980s the Railway Inspectorate were becoming increasingly concerned about passenger trains derailing on trap points so as layouts were renewed and remodelled trap points were removed from passenger lines including loop lines authorised to be used by passenger trains.

 

So would such an arrangement be considered unsuitable for a heritage line today? I can't remember any time when I've visited a heritage line and two trains have moved simultaneously into a crossing station with loop from either end (can't say I noticed trap points or not), so I can only presume that this trap point arrangement is no longer considered acceptable.

[Michael Hodgson]

10 minutes ago, Ian J. said:

 

So would such an arrangement be considered unsuitable for a heritage line today? I can't remember any time when I've visited a heritage line and two trains have moved simultaneously into a crossing station with loop from either end (can't say I noticed trap points or not), so I can only presume that this trap point arrangement is no longer considered acceptable.

There is one such station on the West Somerset line, but it's an anomaly because it is allowed to do this under "grandfather rights" as the station happens to be the summit of a hill and there is less risk of a driver being unable to stop.  The general rule on preserved lines is still let one in first and only let the second one in once the first has stopped.

[RailWest]

13 minutes ago, Michael Hodgson said:

There is one such station on the West Somerset line, but it's an anomaly because it is allowed to do this under "grandfather rights" as the station happens to be the summit of a hill and there is less risk of a driver being unable to stop.  The general rule on preserved lines is still let one in first and only let the second one in once the first has stopped.

Actually, there are two now :-)

 

The layout at Crowcombe Heathfield (on a summit) with traps at each end replicates what was there in GWR and BR days until the line closed.

 

The layout at Williton (much more on the level-ish) used to be similar until modified in the 1960s by BR. It has now been reinstated with traps at both ends again. The ability to admit trains simultaneously certainly helps to cut down on possible delays.

[Michael Hodgson]

Weybourne on the North Norfolk effectively has a trap at the Sheringham end, in that the platform line continues into a yard.

However there is no equivalent at the Holt end.

 

Likewise Thuxton on the Mid Norfolk has a siding at the southern end of the loop, but not at the opposite end.

 

I'm not sure of the current layout at Horsted Keynes - that may well allow trains to enter from both ends.

[Nick C]

2 hours ago, Michael Hodgson said:

Weybourne on the North Norfolk effectively has a trap at the Sheringham end, in that the platform line continues into a yard.

However there is no equivalent at the Holt end.

 

Likewise Thuxton on the Mid Norfolk has a siding at the southern end of the loop, but not at the opposite end.

 

I'm not sure of the current layout at Horsted Keynes - that may well allow trains to enter from both ends.

https://www.bluebell-railway.co.uk/bluebell/pti/lgepics/diag-1.html

 

It looks from that, that both 3 and 4 roads are trapped by the crossovers, so presumably would be possible. I don't know whether it is done though. 

[KingEdwardII]

Amazing that a "nowheresville" like Horsted Keynes had at one point a 5 platform station.

 

OK, it was a junction station, but was there really the traffic to justify such a large facility?

 

From what I have read, originally the line through Horsted Keynes was double track, with only the southbound line to Lewes being single track, which may influence the discussion on this thread.

 

Even today the small mainline station at Ford (right next to Gaugemasters' headquarters!) manages with only 2 platforms despite being a significant double track junction station with a very busy timetable...

 

Yours, Mike.

 

 

[St Enodoc]

9 hours ago, Nick C said:

https://www.bluebell-railway.co.uk/bluebell/pti/lgepics/diag-1.html

 

It looks from that, that both 3 and 4 roads are trapped by the crossovers, so presumably would be possible. I don't know whether it is done though. 

The full diagram may make it clearer:

 

https://www.derekhayward.co.uk/BluebellRailway-1/Photographic-Tours/Horsted-Keynes-Signal-Box/i-NRPVmLs/A.

[RailWest]

One important thing to remember, which does not seem to have been mentioned so far, is that - regardless of whether traps are provided or not, and whatever 'the rules' may say - it depends also upon what the interlocking does - or does not - allow. In a 'typical' simple passing-loop with no traps, then the opposing Home signals would be interlocked so that it would be impossible physically to pull them both 'off' at the same time, thereby enforcing the relevant rule about 'only one train at a time'. 

[The Stationmaster]

42 minutes ago, RailWest said:

One important thing to remember, which does not seem to have been mentioned so far, is that - regardless of whether traps are provided or not, and whatever 'the rules' may say - it depends also upon what the interlocking does - or does not - allow. In a 'typical' simple passing-loop with no traps, then the opposing Home signals would be interlocked so that it would be impossible physically to pull them both 'off' at the same time, thereby enforcing the relevant rule about 'only one train at a time'. 

The interlocking with traps is also interesting as the Home Signal also requires the trap point to be bolted by its Facing Point Lock. (although it can be in either position however if the trap is set for exit obviously the Home in the opposite direction is locked at danger).

 

It is also interesting to hear that some heritage etc lines are putting traps back but presumably they are (I hope) properly risk assessing the change before implementing it as they are now legally required to do when making changes which alter potential risks.  it could get a bit embarrassing if a member of the Inspectorate turned up and asked them why they had installeda trap point    

 

The statistics collected by BR and the previous incarnation of the Railway Inspectorate were quite revealing as they showed that the instances of a passenger train derailing where there was a trap point were considerably greater than the number of collisions where there wasn't one.  Although back in those days the consequences of something happening weren't given as much attention as they would be nowadays in a full risk assessment.  However a trap point, ideally with a proper run-off or sand drag does provide flank protection so there are arguments both for and against hence the need for proper risk assessment.

[Nick C]

2 hours ago, KingEdwardII said:

Amazing that a "nowheresville" like Horsted Keynes had at one point a 5 platform station.

 

OK, it was a junction station, but was there really the traffic to justify such a large facility?

 

From what I have read, originally the line through Horsted Keynes was double track, with only the southbound line to Lewes being single track, which may influence the discussion on this thread.

 

Even today the small mainline station at Ford (right next to Gaugemasters' headquarters!) manages with only 2 platforms despite being a significant double track junction station with a very busy timetable...

 

Yours, Mike.

 

 

Bear in mind it's only really 4 platforms, as 4/5 are both sides of the same track. The line from East Grinstead via HK to Haywards Heath was indeed double track, so I would assume the idea was to provide separate up and down platforms for both main and branch. It's also quite possible, that it was built that size in expectation of future traffic, which then never materialised. Were it still on the national network, the village would probably now be a commuter town just like Uckfield, Oxted etc...

[Ian J.]

OK, I've worked out a general route diagram for S&P, and thought about where passing places would be on the line. I was going to use AnyRail to come up with some station diagrams so I could work out signalling arrangements, however it seems to be somewhat lacking in a decent array of signal types in both semaphore and colour light for U.K. outline.

 

Can anyone make a suggestion of a software program that I could use for creating general track layout and signalling diagrams with that has a good selection of U.K. outline signals (both semaphore and colour light)?

[KingEdwardII]

On 28/08/2021 at 21:54, Ian J. said:

Can anyone make a suggestion of a software program that I could use for creating general track layout and signalling diagrams

I use XTrackCad and I have created a suitable set of semaphore signal icons for it that I've used for my own layout. It is fairly easy to add new signal arrangements to the list of icons, although a bit laborious. Once you've created the icons in the list, you can add them to your track diagram very easily.

 

Some folk find XTrackCad somewhat hard to use, although it works for me. You might like to try out the basic program to create track layouts first before looking at adding signals.

 

Yours,  Mike.

[Ian J.]

I investigated further how to create objects in AnyRail and I've been able to cobble together a Home and Distant and white/red and black/yellow discs. I've kept the original drawn items in ungrouped form in a separate file so I will be able to reuse them to put together variations. It's a bit fiddly to use for object drawing, but I worked out how to get round a problem or two and the objects look OK.

 

So, onto my next question: what are the 'rules' for where ground signals go? The only one I've ascertained so far is the black/yellow for exits to a 'main' line where there is another viable route between the signal and the 'main' line exit, meaning the black/yellow can be passed at 'on'.

 

Here's a diagram I drew up yesterday for Cold Holt, the first station on the S&P, without any ground signals:

 

[SZ]

55 minutes ago, Ian J. said:

So, onto my next question: what are the 'rules' for where ground signals go?

 

I think a most basic rule is to only signal movements that are expected to happen.