Should the two track railway north of Preston be fitted with bidirectional signalling?

[Falcon1200]

moderator note-split from

4S38 has failed on Beattock Bank 29/4/2025

As indicated above it has failed just short of Harthope viaduct at 20.00 ( according to Tracksy) 1S81 and 82 very close behind. I'm no expert on this, but how would it be rescued?. Looking at Tracksy, if the Thunderbird at Carlisle was used it would have go along the up line from Lockerbie to...

www.railforums.co.uk

Anyone know why they can't run wrong road here? Guessing there will be good reason, but just wondering what it is really

For such a critical section of track, SimBids would be a minimum insurance policy....

There is no bi-directional signalling on any part of the WCML between Gretna Jc and Carstairs. And if I had a pound for every time I raised this issue (as NR Controller for the route for some years) Jeff Bezos would envy me. Depending on which section is involved, implementing Single Line Working requires up to 4 people on the ground, and this in a remote area with few staff based nearby.

The failed freight train had problems earlier at Carlisle and Beattock is notorious for failing freights so why on earth did NR route it ahead of all the cross border passenger trains

Perhaps the Freight Operator thought, and advised NR, that the problem had been resolved? It would not be the first time this situation has occurred.

Obviously delays occur on planes and roads too but being delayed 6 hours because of one broken down train will seem ridiculous to most people.

It is indeed absolutely ridiculous, as is the failure of BR, Railtrack and NR in turn to do anything about this crucial rail route. I have said before that there have been so many incidents where bi-directional signalling would have massively reduced delays that it would long since have paid for itself - Even if just on the Down line up Beattock bank.

 

[liam456]

There is no bi-directional signalling on any part of the WCML between Gretna Jc and Carstairs. And if I had a pound for every time I raised this issue (as NR Controller for the route for some years) Jeff Bezos would envy me. Depending on which section is involved, implementing Single Line Working requires up to 4 people on the ground, and this in a remote area with few staff based nearby.

indeed absolutely ridiculous, as is the failure of BR, Railtrack and NR in turn to do anything about this crucial rail route. I have said before that there have been so many incidents where bi-directional signalling would have massively reduced delays that it would long since have paid for itself - Even if just on the Down line up Beattock bank.

Fingers crossed this new Trilink plan to install ETCS on this very critical part of railway infrastructure will give us Bi-di signalling "for free" and fix this historical mistake once and for all!

 

[LNW-GW Joint]

I see the ill-fated 4S38 took 6h04m (364m) to climb Beattock bank yesterday.
Presumably it was a diesel loco which failed?

Realtime Trains | 4S38 1416 Seaforth C.T. Mdhc (Ews) to Mossend Euroterminal | 29/04/2025

Real-time train running information for 4S38 1416 departure from Seaforth C.T. Mdhc (Ews) to Mossend Euroterminal on 29/04/2025. From Realtime Trains, an independent source of train running info for Great Britain.

www.realtimetrains.co.uk

While I am aware of NR plans for ETCS installation between Crewe and Carlisle, I think I saw somewhere that it would not be deployed in Scotland.
But bi-di capability is sorely needed.

 

[zwk500]

Fingers crossed this new Trilink plan to install ETCS on this very critical part of railway infrastructure will give us Bi-di signalling "for free" and fix this historical mistake once and for all!

It's slightly misleading to refer to Bi-Di as coming 'for free' on ETCS (it obviously requires expansion of the interlocking data, the production and testing of which is a significant cost, and the requisite S&C to allow crossover moves) however it is much more cost-effective to do so than conventionally signalled lines as lineside you can just put in block markers where needed and don't need lamp proving and aspect sequence design for each crossover and so forth.

When ETCS is deployed on Preston-Carlisle and then eventually Carlisle-Carstairs I'd be astonished if Bi-Di wasn't included.

While I am aware of NR plans for ETCS installation between Crewe and Carlisle, I think I saw somewhere that it would not be deployed in Scotland.

The Scottish government certainly did talk about rolling back from ETCS deployment.

But bi-di capability is sorely needed.

The cost to come back and edit the interlocking now would be astronomical. Sadly, we are left with the legacy of the railway as was forseen when this section was designed, and local staff were much more readily available so unsignalled crossovers were less of an issue. However I wonder if some kind of bodge with Shunt/sub signals and SBSIs would cheap enough to justify for the flexibility during incidents until ETCS can be deployed on this section. IIRC the crossovers on this section are controlled from Ground Switch Panels, so are presumably motored and therefore include locking as standard?

 

[68000]

Bi-directional signalling will not happen on that portion of the WCML until the interlockings are replaced as they are 1970s era geographical relay interlockings and there is little chance NR or the supply chain now has the capability of doing such design and testing work. The interlockings are probably due for renewal within the next 10 years so should be looked at then. One of the problems will be the need to bring the signalling up to modern standards (eg signal sighting) which generally means more invasive intervention than a like for like replacement and will cost more.

 

[Dr Hoo]

While I am aware of NR plans for ETCS installation between Crewe and Carlisle, I think I saw somewhere that it would not be deployed in Scotland.
But bi-di capability is sorely needed.

Yes, I’m sure that Transport Scotland had decided not to have it in Scotland.

 

[Zomboid]

You'd have to say that it needs to be provided "as soon as sensibly achievable". If there's a medium term plan for ETCS then that will provide the opportunity to get it sorted out.

It's probably not practical to go out and fit it any time soon as a standalone thing, though.

 

[The Planner]

Its pretty nailed on for when Carlisle and Preston are done.

 

[liam456]

Is that because the signalling that is inside WSSC is new enough that Transport Scotland don't want the expense of modification/renewal before it is actually life-expired but for the signalling in NW&C it all needs ripping out anyway. Is there a significant extra cost in terms of specifying ETCS vs ETCS + Bi-di signalling?

Thanks for the insight @zwk500. Certainly some crossovers exist along the route but even with whatever CBI is used for the Scottish signalling, the cost there to add bi-di would be high as well?

 

[MarkyT]

Fingers crossed this new Trilink plan to install ETCS on this very critical part of railway infrastructure will give us Bi-di signalling "for free" and fix this historical mistake once and for all!

It wouldn't be 'free', but it would likely be cheaper an upgrade than bi-di in an equivalent conventional signal scheme. Sometimes, the additional wrong direction signals need significant structures in places for sighting reasons, especially on curvy lines through hilly regions. ETCS block markers don't need continuous long-range sighting on approach in the same way and they are smaller so can sometimes fit between tracks where signals wouldn't. There's still more complex interlocking in implementing bi-di than a unidirectional scheme in ETCS, nonetheless. As others have said, it's impractical and a waste of effort to attempt to implement bi-di in the existing relay interlockings, especially poor value since they are likely due for renewal in the medium term. Bi-di might be an aspiration that could help tip the balance in favour of earlier renewal however, assuming other parts of the ETCS puzzle are in place.

 

[zwk500]

Is that because the signalling that is inside WSSC is new enough that Transport Scotland don't want the expense of modification/renewal before it is actually life-expired but for the signalling in NW&C it all needs ripping out anyway.

I think Transport Scotland were concerned with the generally high cost of ETCS schemes rather than any specific technical details of the equipment invovled.

Is there a significant extra cost in terms of specifying ETCS vs ETCS + Bi-di signalling?

Extra cost yes, don't know how significant it is as that would depend on a range of factors. Anything from extra crossovers to needing to review all the wrong-direction stopping points would need to be considered.

Thanks for the insight @zwk500. Certainly some crossovers exist along the route but even with whatever CBI is used for the Scottish signalling, the cost there to add bi-di would be high as well?

AIUI from the Sectional Appendix and signal diagrams it looks like the crossovers are now directly controlled from WSSC, but there are no signalled routes across the crossovers. There's also a Local Instruction that the crossovers are only to be used during Single Line Working or in a possession.
The cost to add bi-di to this stretch would be extremely substantial. The new signals would need to be positioned correctly for sighting and the risk approaching each crossover. You'd need additional routes from existing signals to the new lines (including new route indicators, also requiring sighting), as well as the new signals. You'll probably need a distant signal for the wrong direction to allow a reasonable linespeed (assuming you've not gone for full-capacity bi-di due to cost). As the signals would be on the 'wrong' side of the line for drivers they may need mounting on gantries or cantilevers to provide correct sighting. And that's before we get to the work required for all the risk assessments and TPWS fitment, etc...

As I mentioned there may be a bodge to fit shunt signals to give a signalled route across the crossover and then a SBSI to allow the signaller to authorise wrong-direction working at low speed without needing a Pilot on-scene or anything like that. However this would likely not be worth the effort for the limited benefits proceeding at caution for that long would bring. It would also likely not be fully interlocked, and thus may not be agreed to by the ORR/RSSB on safety grounds if there is a possibility the signaller could signal a right-direction move when a wrong-direction move was still taking place.

 

[GRALISTAIR]

Its pretty nailed on for when Carlisle and Preston are done.

Thank you for that information. I trust your replies.

 

[MarkyT]

As I mentioned there may be a bodge to fit shunt signals to give a signalled route across the crossover and then a SBSI to allow the signaller to authorise wrong-direction working at low speed without needing a Pilot on-scene or anything like that. However this would likely not be worth the effort for the limited benefits proceeding at caution for that long would bring. It would also likely not be fully interlocked, and thus may not be agreed to by the ORR/RSSB on safety grounds if there is a possibility the signaller could signal a right-direction move when a wrong-direction move was still taking place.

I agree that solutions without full interlocking would be difficult to justify

Any cheap solutions for providing such interlocking can often start to get more difficult and expensive than anticipated, especially when old installations have wiring and cabling condition concerns. While such installations can be perfectly safe for many more years undisturbed in their original configurations, they can be put at risk when running in new wires and particularly when pulling out old wires and equipment. There is also the spare capacity of trackside cabling to consider for the additional circuits required. Such capacity is often limited in simple plain line 'auto section' installations.

The original 1980s SIMBIDS concept was developed as a cheap(er) way to provide limited reverse direction capability in areas having traditional relay-based colour light signalling, primarily to avoid the need for pilot working during planned engineering work and emergencies. The system as applied on the GWML (and I think parts of ECML), had AWS functionality deliberately excluded in the reverse direction, with no magnets for the new signals themselves and, perhaps more importantly, no suppression of the permanent magnets for the normal direction signals. Hence a reversibly signalled train would first pass a special SIMBIDS 'no AWS' board, then get a series of warnings for every right direction signal reverse passed en route through the section until reaching the crossover back to the normal line where another special board announced recommencement of AWS. That's considered highly undesirable now as the repeatedly cancelled warnings with no subsequent actions risk conditioning drivers into reflexively doing so elsewhere.

The technical reason for this was to avoid having to retrofit the expensive and heavy suppressed permanent inductors at every right direction signal and create new control circuits for each such installation based on the newly provided directional route locking between adjacent interlockings. While the handful of circuits needed for the new end-to-end directional locking could usually be accomodated in existing cabling, or multiplexed via vital reed transmission in a trunk cable, significantly more cable cores than existed typically through sections of continuous multi-aspect headway signalling would have been required for full AWS suppression, leading to widespread new cabling and housings to terminate them, likely approaching the cost of a more comprehensive trackside renewal which could more easily have had fully featured bi-di built in from new.

The GWML form of SIMBIDS cannot be replicated in new schemes today, mainly because of the AWS compromises.

 

[Helvellyn]

The original 1980s SIMBIDS concept was developed as a cheap(er) way to provide limited reverse direction capability in areas having traditional relay-based colour light signalling, primarily to avoid the need for pilot working during planned engineering work and emergencies. The system as applied on the GWML (and I think parts of ECML), had AWS functionality deliberately excluded in the reverse direction, with no magnets for the new signals themselves and, perhaps more importantly, no suppression of the permanent magnets for the normal direction signals. Hence a reversibly signalled train would first pass a special SIMBIDS 'no AWS' board, then get a series of warnings for every right direction signal reverse passed en route through the section until reaching the crossover back to the normal line where another special board announced recommencement of AWS. That's considered highly undesirable now as the repeatedly cancelled warnings with no subsequent actions risk conditioning drivers into reflexively doing so elsewhere.

Not sure if it's the same system but SIMBIDs was installed on the SWML as well. I want to say Worting Junction (South of Basingstoke) to Winchester (possibly Shawford), and Woking to Guildford.

 

[brad465]

Its pretty nailed on for when Carlisle and Preston are done.

Good to hear, is the proposed time for doing this disclosable yet?

 

[zwk500]

Good to hear, is the proposed time for doing this disclosable yet?

There is a map with layers by Control period and year here: https://www.networkrail.co.uk/runni...gy/digital-railway-long-term-deployment-plan/

https://www.networkrail.co.uk/wp-content/uploads/2019/06/Digital-Long-Term-Deployment-Plan.pdf

To view the layers you will probably need to download it and open in Adobe Viewer (or similar) rather than open it directly in a web browser.

Preston - Carnforth is shown as CP8 Y1, 2030. Carnforth - Upperby is shown as CP8 Y3, 2032, Carlisle Station area as CP8 Y4, 2033. AIUI the map is from 2019 so may have slipped a bit with COVID. The map also only covers England and Wales so Gretna Jn - Carstairs is not shown.

 

[D6130]

Not sure if it's the same system but SIMBIDs was installed on the SWML as well. I want to say Worting Junction (South of Basingstoke) to Winchester (possibly Shawford), and Woking to Guildford.

.....and also on the Brighton main line between Balcombe Tunnel Junction and Preston Park, IIRC.

 

[zwk500]

.....and also on the Brighton main line between Balcombe Tunnel Junction and Preston Park, IIRC.

SIMBIDs is also present on the C2C line, might be on the GEML (the sectional appendix just marks 'reversible signalling' and doesn't specify SIMBIDs or not, and the GEML reversible speeds do get above 50mph for wrong-direction which I thought was a SIMBIDS limit), and possibly the ECML on the section controlled by Morpeth box (it's definitely bi-di, not sure if it's specifically SIMBIDS).

 

[MarkyT]

SIMBIDs is also present on the C2C line, might be on the GEML (the sectional appendix just marks 'reversible signalling' and doesn't specify SIMBIDs or not, and the GEML reversible speeds do get above 50mph for wrong-direction which I thought was a SIMBIDS limit), and possibly the ECML on the section controlled by Morpeth box (it's definitely bi-di, not sure if it's specifically SIMBIDS).

The GWML implementations had special instructions in the sectional appendix that explained the AWS limitations. The reverse direction was also not used routinely without advance notice of work in operating publications, except in certain defined emergency scenarios. In later years, the instructions didn't even refer to SIMBIDS explicitly by name, just the characteristics of the sections concerned and any restrictions. I recall some southern region examples referred to as SIMBIDS were actually engineered to full bidirectional standards with complete AWS provision etc. SSI signalled areas like C2C are unlikely to be 'true' SIMBIDS as the AWS suppression issue is much easier to solve with distributed I/O processor-based systems.

 

[yorksrob]

The section does seem pretty prone to flooding - although I don't know how often the other track in these cases could be used.

The section of double track on the Brighton main line has this arrangement, so it has been done.

 

[zwk500]

The GWML implementations had special instructions in the sectional appendix that explained the AWS limitations. The reverse direction was also not used routinely without advance notice of work in operating publications, except in certain defined emergency scenarios. In later years, the instructions didn't even refer to SIMBIDS explicitly by name, just the characteristics of the sections concerned and any restrictions. I recall some southern region examples referred to as SIMBIDS were actually engineered to full bidirectional standards with complete AWS provision etc. SSI signalled areas like C2C are unlikely to be 'true' SIMBIDS as the AWS suppression issue is much easier to solve with distributed I/O processor-based systems.

Indeed, it's all very confusing with the mixed up references.

 

[Harpo]

SIMBIDs is also present on the C2C line,

….and came originally with the limitation that it could only be used pre-planned. (Not sure why.)

At least one of the Anglia bi-di sections also had a stipulation that both entry and exit signals had to be cleared to start a movement.

I also recall that the associated safe-cess footpaths in Anglia were very expensive, with one senior person calling them ‘motorways’ such was their cost.

All stuff that makes justification very didficult.

 

[MarkyT]

The section of double track on the Brighton main line has this arrangement, so it has been done.

Indeed, it's all very confusing with the mixed up references.

The confusion may have arisen because most reversibly signalled double track lines have many fewer signal blocks in the reverse direction than the normal direction. A five mile section between crossovers might have five separate blocks in the normal direction, while only having one in the reverse. Some thought that reduced reverse capacity was the main factor defining SIMBIDS, but the configuration predated the SIMBIDS concept, was already provided in many places, and is often perfectly adequate for emergency and planned diversionary use as it assumes services much be drastically reduced to be able to run alternately on a single track over part of the route instead of the usual double.

On the Brighton main line, signalling was like this for many years, though standard principles were followed in full, with full AWS for instance, so there were no particular restrictions on its use, except the practical impossibility of pathing a wrong direction move during normal weekday traffic. It is such a busy line with so many branches however that special weekend timetables can be devised with 'flights' (convoys) of trains closely following each other in the same direction through a single line section, before the direction is switched, say every half hour, and a convoy in the opposite direction can proceed. It's not possible to do this with limited capacity in the reverse direction however, so some years ago the line was upgraded to provide the same signal density and following headway capacity in the reverse direction as the normal direction. This is still unusual in the UK.

 

[zwk500]

The confusion may have arisen because most reversibly signalled double track lines have many fewer signal blocks in the reverse direction than the normal direction.

Indeed

It's not possible to do this with limited capacity in the reverse direction however, so some years ago the line was upgraded to provide the same signal density and following headway capacity in the reverse direction as the normal direction. This is still unusual in the UK.

This appears to only be the case between Balcombe Tunnel Junction and Copyhold Jn. Looking at the latest signal maps I've got (Jan 24), Haywards Heath to Keymer junction and Keymer Junction to Preston Park has only 1 Signal section for the wrong direction, but with YY/Y aspect sequence from Distant/Repeater signals on approach. Copyhold Jn to Haywards Heath is 4-track, with the Slow Lines Bi-Di and the Fast Lines not (although both fast line platforms are able to turn trains round to/from the south). The OTT Map seems to be out of date.

Lines with full-capacity BiDi in the UK include the Trent Valley (slightly different as it's a 4-track route with the bi-di intended for 2-track running), the MML between Kettering and Leicester and the West London Line/Willesden reliefs from Clapham Junction to Wembley Central.

 

[Zomboid]

Bicester to Oxford is bi-di, but I'm a bit surprised to learn that Bicester to Bletchley isn't.

 

[MarkyT]

Lines with full-capacity BiDi in the UK include the Trent Valley (slightly different as it's a 4-track route with the bi-di intended for 2-track running),

The 4 track section of GWML between Wantage Rd and Challow, installed in the 1990s for HSTs to overtake new imported coal traffic between Avonmouth and Didcot, is a bit like that. The middle Mains are BIDI with the outer Reliefs UNIDI. It's still reduced capacity for the reverse direction though I think. The BIDI was there before the extra tracks. It was provided in the mid 1970s before HST introduction.

 

[Oxfordblues]

One very minor but useful scheme on the WCML between Lancaster and Morecambe South Jn would allow Morecambe-Lancaster trains to traverse the Down Main without crossing-over and crossing-back, leaving the Up Line uninterrupted.

 

[zwk500]

Bicester to Oxford is bi-di,

Only Kennington South Jn to Oxford Parkway/Wolvercot North Junction is equal capacity in both directions though. Didcot North Jn - Kennington South Jn and Oxford Parkway - Bicester is reduced capacity of only 1 signal between crossovers for wrong-direction movements (but I think 'full' signalling so not SIMBIDS).

but I'm a bit surprised to learn that Bicester to Bletchley isn't.

I assume as costs escalated it got descoped quite quickly, to await ETCS.

The 4 track section of GWML between Wantage Rd and Challow, installed in the 1990s for HSTs to overtake new imported coal traffic between Avonmouth and Didcot, is a bit like that. The middle Mains are BIDI with the outer Reliefs UNIDI. It's still reduced capacity for the reverse direction though I think. The BIDI was there before the extra tracks. It was provided in the mid 1970s before HST introduction.

I think most of the GWML between Didcot and the Severn Tunnel/Bristol Temple Meads is reduced capacity in the wrong direction (although some sections might get 2 trains in the wrong direction rather than just 1). However it is at least usable without prior notice, I think, so presumably has full protection of the wrong-direction signals.

 

[hexagon789]

moderator note-split from

4S38 has failed on Beattock Bank 29/4/2025

As indicated above it has failed just short of Harthope viaduct at 20.00 ( according to Tracksy) 1S81 and 82 very close behind. I'm no expert on this, but how would it be rescued?. Looking at Tracksy, if the Thunderbird at Carlisle was used it would have go along the up line from Lockerbie to...

www.railforums.co.uk

There is no bi-directional signalling on any part of the WCML between Gretna Jc and Carstairs. And if I had a pound for every time I raised this issue (as NR Controller for the route for some years) Jeff Bezos would envy me. Depending on which section is involved, implementing Single Line Working requires up to 4 people on the ground, and this in a remote area with few staff based nearby.



Perhaps the Freight Operator thought, and advised NR, that the problem had been resolved? It would not be the first time this situation has occurred.



It is indeed absolutely ridiculous, as is the failure of BR, Railtrack and NR in turn to do anything about this crucial rail route. I have said before that there have been so many incidents where bi-directional signalling would have massively reduced delays that it would long since have paid for itself - Even if just on the Down line up Beattock bank.

Another Freight has failed on Beattock Bank - 6S50.

Surely with so many incidents recently, questions need to be asked of both Network Rail and freight operators in terms of resilience, disruption, rescue contingencies and so on...

Its getting to be a habit and it's not as though they can blame poor adhesion at the moment.

 

[Iskra]

Another Freight has failed on Beattock Bank - 6S50.

Surely with so many incidents recently, questions need to be asked of both Network Rail and freight operators in terms of resilience, disruption, rescue contingencies and so on...

Its getting to be a habit and it's not as though they can blame poor adhesion at the moment.

Could the current heat have something to do with it?