ETCS Roll-Out on ECML South Confirmed

[MarkyT]

It won't have too many noticeable changes, but there are some important aspects of ETCS that will change the service - SRTs and Junction margins will get rounded to 15 seconds rather than 30 seconds, The Braking Curve control replaces approach control signalling and cautionary aspects for a smoother approach and (I think) overlaps effectively become redundant, being replaced by the braking curve supervision. Which has implications for compact layouts such as at Peterborough.

It's a bit late to descope a resignalling scheme from ETCS to anything else once you're halfway through the project.

I've read that ETCS uses a concept familiar from previous forms of ATP, known as 'release speed' on final approach to the fixed limit of a movement authority, such as a signal or block marker board. Once below this speed, the onboard system does not supervise the speed in fine detail, however if the train then passes the limit of movement authority, then a 'train-stop' intervention takes place, which is calculated to safely bring the train to a stand within the overlap. Hence overlaps are required in ETCS schemes, although release speed values and overlap lengths can be varied to a greater extent than with conventional signalling to suit local conditions and layouts. I believe on simple plain track where block markers are provided every 0.5km or so, the standard overlap is one block length, to avoid any additional train detection sections just for overlap purposes.

== Doublepost prevention - post automatically merged: 30 Jun 2022 ==

ETCS could lead to changes in the permissible speed here and there where it is currently limited by signal sighting, but that's not going to be many nor very large increases. As others have said a few other things such as the shorter overlaps, better placed block sections and more timing granularity will help it trains to run 'smoother' and slightly closer together. But as others have pointed out there's probably not value for money in replacing hundreds of millions of pounds worth of points, track, platforms and OHLE.

ETCS might allow raised speeds on the slow lines in some places where geometry and track maintenance standards allow it. Currently, this is artificially limited by signal spacing with 4-aspect signals on the fasts and 3-aspect on the slows.

 

[zwk500]

I feel you have too little faith in the ability of the people who decide descoping!

It may be insane but the will find a way to do it.

I have no doubt they'll try, but a bit like when Cameron tried to change the Aircraft Carrier design halfway through, when somebody sits them down and explains it all they'll realise there's only 1 practical choice.

I've read that ETCS uses a concept familiar from previous forms of ATP, known as 'release speed' on final approach to the fixed limit of a movement authority, such as a signal or block marker board. Once below this speed, the onboard system does not supervise the speed in fine detail, however if the train then passes the limit of movement authority, then a 'train-stop' intervention takes place, which is calculated to safely bring the train to a stand within the overlap. Hence overlaps are required in ETCS schemes, although release speed values and overlap lengths can be varied to a greater extent than with conventional signalling to suit local conditions and layouts. I believe on simple plain track where block markers are provided every 0.5km or so, the standard overlap is one block length, to avoid any additional train detection sections just for overlap purposes.

My understanding is that ETCS calculates a Braking Curve for the train to stop safely at a given limit of authority (usually a block marker board). It includes a safety margin and compensates for gradient/train type etc. If the driver does not apply the brakes to keep the train within the braking curve, it will sound a warning and if there is not an adequate response from the driver it will intervene until the train is within the permitted curve. So the overlap is effectively taken over by the curve supervision. However, I haven't had any close involvement with ETCS schemes so there may still be a safety margin beyond each block marker (or some equivalent of double-reds) that still provides the safety an overlap does.

 

[Bald Rick]

Will this also mean speeds increasing to 225 km/h or 140 mph on the southern section of the ECML?

is any of the ECML South built for >125mph speeds like how the racetrack is but due to signalling was limited?
could we see any speed increases as part of this?

no.

 

[MarkyT]

Do they mean block signalling (as a system), which dates from the Victorian era, rather than PTC-style movement authorities?

In practical terms, movement authorities are still made up of locked routes in an interlocking, which in principle works in the same way as a mechanical lever frame. Talking of PTC, UP managed to equip their Big Boy with their version of the system. The cab is so big they managed to put a walkaround island in the middle with the electronics and interface screens. I believe the equipment takes power from the accompanying diesel loco that is always marshalled into the consist immediately behind the loco tender. Tornado has a steam powered generator on board I think.

The trouble is that the ECTS equipment will likely need to at least be able to implement a brake demand automatically, to pass the safety case. It is do-able, but may need some modification (such as an attachment point), which is probably do-able, however. Whether part of it will be removable is another matter.

I think it will have fairly limited interventions available, i.e cut power and apply service or emergency braking to a stop. The same valves in the braking system that TPWS/AWS uses currently for that might suffice, but an additional valve in the steam path from boiler to cylinders may be required to automatically cut power when required.

 

[DanNCL]

This will achieve absolutely nothing in performance improvements unless the DFT also cough up the money to either retrofit the 91s with ETCS or to replace them. They’ve not funded the former to my knowledge and things have gone quiet on the latter.

 

[Starmill]

This will achieve absolutely nothing in performance improvements unless the DFT also cough up the money to either retrofit the 91s with ETCS or to replace them. They’ve not funded the former to my knowledge and things have gone quiet on the latter.

They can simply withdraw them however and have LNER make do. Obviously that's certainly not a desirable outcome but it's entirely possible.

 

[DanNCL]

They can simply withdraw them however and have LNER make do. Obviously that's certainly not a desirable outcome but it's entirely possible.

Unlikely as LNER have pretty much returned to pre pandemic passenger numbers, which can’t be said for any of the TOCs that have had fleet reductions forced on them by the DFT. LNER is the one operation the DFT are least likely to cut back.

 

[zwk500]

This will achieve absolutely nothing in performance improvements unless the DFT also cough up the money to either retrofit the 91s with ETCS or to replace them. They’ve not funded the former to my knowledge and things have gone quiet on the latter.

If they're getting rid of lights on sticks, anything running on the Line will need ETCS. Which also raises the interesting question around Peterborough, as to whether all the freight will be fitted with ETCS or if they will have a hybrid Overlay system.

 

[TurboMan]

Seems a bit of a lost opportunity to go to ETCS Level 2 when it's 'signals away' and therefore could go to Level 3.

 

[DanNCL]

If they're getting rid of lights on sticks, anything running on the Line will need ETCS. Which also raises the interesting question around Peterborough, as to whether all the freight will be fitted with ETCS or if they will have a hybrid Overlay system.

There's a program to retrofit 66s with ETCS. I'm not sure about any other freight locos.

EMR's 170s will also need retrofitting unless there's a plan for TFW 158s to move to EMR for the Norwich route.

 

[MarkyT]

Seems a bit of a lost opportunity to go to ETCS Level 2 when it's 'signals away' and therefore could go to Level 3.

'Level 3' is increasingly being viewed as the means to remove much but not all of the fixed train detection while still retaining fixed, albeit 'virtual' blocks with markers. Only trains which can confirm their length and completeness on board can clear virtual fixed blocks. Not all ETCS-equipped trains can do this yet.

 

[Starmill]

Unlikely as LNER have pretty much returned to pre pandemic passenger numbers, which can’t be said for any of the TOCs that have had fleet reductions forced on them by the DFT. LNER is the one operation the DFT are least likely to cut back.

I don't think there's room for hubris!

 

[TurboMan]

'Level 3' is increasingly being viewed as the means to remove much but not all of the fixed train detection while still retaining fixed, albeit 'virtual' blocks with markers. Only trains which can confirm their length and completeness on board can clear virtual fixed blocks. Not all ETCS-equipped trains can do this yet.

Surely train length is essential for all trains that run under Levels 1 and 2? Otherwise how would the ceiling speed step up when the rear of the train clears an increase in linespeed? Or do you mean because the driver has to manually confirm train length at SoM which makes it open to human error?

 

[zwk500]

Surely train length is essential for all trains that run under Levels 1 and 2? Otherwise how would the ceiling speed step up when the rear of the train clears an increase in linespeed? Or do you mean because the driver has to manually confirm train length at SoM which makes it open to human error?

Under Level 2 there is still fixed train detection on the lineside that can communicate that the rear of the train is clear to the system. Under Level 3 the train would need to be able to prove itself complete.

 

[Sonik]

Under Level 2 there is still fixed train detection on the lineside that can communicate that the rear of the train is clear to the system. Under Level 3 the train would need to be able to prove itself complete.

I've often wondered about this. Easy for multiple units but the issue is freight, perhaps we will see a return of 'brake' vans, purely for detection purposes.

The issue is, Level 3 is obviously a desirable goal for NR due to the substantial reduction in infrastructure, but the costs for rear of train detection will fall to the FOC.

 

[Trestrol]

Probably more to do with the age of signalling equipment in the Kings Cross and Peterborough PSB areas than anything else.

It is York to Edinburgh was re-signalled at electrification. So is twenty years newer than The South end which was not touched at the time. Kings Cross PSB needed rewiring as all the cabling was shot. No alterations were allowed for fear of causing a fault on wiring.

 

[TurboMan]

I've often wondered about this. Easy for multiple units but the issue is freight, perhaps we will see a return of 'brake' vans, purely for detection purposes.

The issue is, Level 3 is obviously a desirable goal for NR due to the substantial reduction in infrastructure, but the costs for rear of train detection will fall to the FOC.

That's true, I wasn't thinking about freight.

 

[Trestrol]

Well, apart from HS2 and the new bits of NPR, the northern WCML (Warrington, Preston, Carlisle) are likely to go ETCS fairly soon.
The signalling there is of a similar vintage (early 70s) to that being replaced on the ECML.

Don't hold your breath for these three. Carlise has been put back and as concequence is getting new signal heads. I think they are looking at 2034 for Carlisle. The problem with ECTS is its expensive and unless DaFT cough up for it NWR will never afford it. Make do and mend is the mantra into CP7. Plenty of schemes have been cancelled to save money some replacing signalling much older than Preston or Carlisle. Perth being a good example, dating from the 1950's.

 

[MarkyT]

Surely train length is essential for all trains that run under Levels 1 and 2? Otherwise how would the ceiling speed step up when the rear of the train clears an increase in linespeed? Or do you mean because the driver has to manually confirm train length at SoM which makes it open to human error?

The fixed track circuits or axle counters that are retained provide confirmation that a block is clear in levels 1 and 2. I remember early in the GWML ATP pilot scheme they fitted some class 47 locos that could be used on freight. The driver was responsible for inputting consist details, for braking purposes only in that case, but yes that system was open to human error. The requirement for virtual blocks is that the trains' on board systems must be able to automatically determine and report length and completeness. That's fairly easy for MU trains but is a major problem with current freight trains. One idea I have is that trains could be measured using conventional train detection tech at a specific location and then the value sent to the train and retained for use in subsequent block clearance events at sites with no trackside fixed train detection equipment.

== Doublepost prevention - post automatically merged: 30 Jun 2022 ==

Under Level 2 there is still fixed train detection on the lineside that can communicate that the rear of the train is clear to the system. Under Level 3 the train would need to be able to prove itself complete.

That's the easier side of the problem. Reliable determination of length in the first place is the harder part, particularly for variable-length non-MU consists like freight.

== Doublepost prevention - post automatically merged: 30 Jun 2022 ==

Don't hold your breath for these three. Carlise has been put back and as concequence is getting new signal heads. I think they are looking at 2034 for Carlisle. The problem with ECTS is its expensive and unless DaFT cough up for it NWR will never afford it. Make do and mend is the mantra into CP7. Plenty of schemes have been cancelled to save money some replacing signalling much older than Preston or Carlisle. Perth being a good example, dating from the 1950's.

Makes sense to change heads even for a short remaining lifespan as the new ones require nearly zero maintenance and are very reliable. The age of signalling equipment is not the only factor in system longevity. Some systems have deteriorated significantly after only 20 years, while the underlying major elements of many mechanical boxes have achieved a century or more in some cases. Signalling has always been an area where nearly every part can be renewed over time, often with improvements incorporated. A 'Triggers broom' or if you like your analogies more classical, A 'Ship of Theseus'.

 

[Mainsideman]

Is in cab signalling used on southeastern 395s on hs1?

 

[HSTEd]

The technology exists for freight trains to have ETCS-grade train completeness detection.
Just it's considered to be too expensive so we will never see the full potential of ETCS realised.

 

[Starmill]

Is in cab signalling used on southeastern 395s on hs1?

Yes, although as with the Cambrian this also needs to be updated to a newer version of ETCS.

 

[The Planner]

Don't hold your breath for these three. Carlise has been put back and as concequence is getting new signal heads. I think they are looking at 2034 for Carlisle. The problem with ECTS is its expensive and unless DaFT cough up for it NWR will never afford it. Make do and mend is the mantra into CP7. Plenty of schemes have been cancelled to save money some replacing signalling much older than Preston or Carlisle. Perth being a good example, dating from the 1950's.

Might be even later than that, Warrington and Preston will likely get done first. It will only be one big conversion per control period.

 

[MarkyT]

The technology exists for freight trains to have ETCS-grade train completeness detection.
Just it's considered to be too expensive so we will never see the full potential of ETCS realised.

No, it just means that for the foreseeable future, fixed train detection will have to be retained on lines used by those trains that can't report length and completeness. Moving block is of no significant capacity or performance benefit over fairly short fixed level 2 blocks on a mainline rail network. Fixed train detection can then be removed progressively by the virtual block technique where it is practical. That's a pragmatic strategy that allows for future freights with the required tech.

 

[HSTEd]

No, it just means that for the foreseeable future, fixed train detection will have to be retained on lines used by those trains that can't report length and completeness. Moving block is of no significant capacity or performance benefit over fairly short fixed level 2 blocks on a mainline rail network. Fixed train detection can then be removed progressively by the virtual block technique where it is practical. That's a pragmatic strategy that allows for future freights with the required tech.

But unless the freight operators are compelled to fit train completeness technology, they never will.
Which means enormous amounts of, expensive to install and expensive to maintain, fixed train detection equipment will be retained forever. The commissioning of all that extraneous equipment will slow the conversion to ETCS even further.

Given that basically everyone on this forum will be dead before a substantial fraction of the UK rail system is ETCS, "doing it later" basically means "never".

 

[MarkyT]

Is in cab signalling used on southeastern 395s on hs1?

Yes, although as with the Cambrian this also needs to be updated to a newer version of ETCS.

The French coded track circuit system, 'TVM430', is used on HS1, rather than ETCS. In lower speed station areas around St Pancras and Ashford International, the digital balise-based train protection system 'KVB' (also French in origin) is used, overlaid on conventional signalling. The balises (track beacons) look very similar to the ETCS ones and I believe a standard ETCS on-board balise reader can interrogate them if the vehicle has the correct emulation software.
TVM430 - Transmission Voie-Machine (track to train transmission) 430 is the current version, an improvement on the original TVM300. I think the figure suffix indicates the maximum speed in kmh the system can support.
KVB - Contrôle de Vitesse par Balises (Speed control by beacons).

 

[brad465]

Will this also mean speeds increasing to 225 km/h or 140 mph on the southern section of the ECML?

is any of the ECML South built for >125mph speeds like how the racetrack is but due to signalling was limited?
could we see any speed increases as part of this?

I would say I don't see any point considering HS2 is coming along, but the way things are going the speed and capacity benefits won't benefit the east coast given the scaling back.

The commonly quoted figure I see is 140mph over 125mph saves 3 seconds a mile, which I don't see any benefit to over this planned rollout, although maybe more beneficial if/when the whole route is done to at least Darlington. I also suspect everytime a points renewal is due they'd have to swap for swing nose crossings like the type seen on HS1, for as long as it takes to do every affected point set.

 

[snowball]

Are level crossings allowed above 125 mph?

 

[mmh]

It's also got nothing to do with "levelling up" because it won't really make any difference to the service operated, which is what the customer sees.

If it's not going to make any difference to the service the passenger experiences, what is the actual point of it?

 

[MarkyT]

But unless the freight operators are compelled to fit train completeness technology, they never will.
Which means enormous amounts of, expensive to install and expensive to maintain, fixed train detection equipment will be retained forever. The commissioning of all that extraneous equipment will slow the conversion to ETCS even further.

Given that basically everyone on this forum will be dead before a substantial fraction of the UK rail system is ETCS, "doing it later" basically means "never".

The tech isn't really 'off the shelf' yet. There's on-going development in Europe for an automatic coupling and brake line connection standard for freight vehicles. I expect better completeness and length determination will be a side benefit. In the meantime I disagree with you that the train detection is 'extraneous equipment' it is the required infrastructure for continuing to run the freights that run today. Front loading a project to force all freight companies to renew or modify all their equipment at their expense with a risky cutting edge technology is likely to force a lot of freight off the railway. Modern processor-based systems are fairly easy to alter too so if there's a sensible saving to be made by subsequently removing much of the fixed train detection, once all trains used on the line section are suitable, then it can and no doubt will be done.