• Our booking engine at tickets.railforums.co.uk (powered by TrainSplit) helps support the running of the forum with every ticket purchase! Find out more and ask any questions/give us feedback in this thread!

‘Digital Signalling’ to be introduced on the ECML

Status
Not open for further replies.

MarkyT

Established Member
Joined
20 May 2012
Messages
6,232
Location
Torbay
Well I would have preferred we deploy TVM or LZB in the 80s and 90s, but better later than never!
Both proprietary systems that would have locked the industry into particular suppliers for perpetuity with only a ruinously expensive and disruptive complete refit of both infrastructure and traction available to switch away again. The global rail industry has wanted an interoperable system for nearly a century. ETCS is that system and while it may not be perfect as it stands, it is open, extensible, flexible and can evolve.
 
Sponsor Post - registered members do not see these adverts; click here to register, or click here to log in
R

RailUK Forums

D365

Veteran Member
Joined
29 Jun 2012
Messages
11,397
Unless one of those became the EWR interchange station for the ECML, in which case some fast line trains might want to call and a speed limit might be needed through the platform.

EWR is planning for a Tempsford/south of St Neots station.
 

HSTEd

Veteran Member
Joined
14 Jul 2011
Messages
16,635
Both proprietary systems that would have locked the industry into particular suppliers for perpetuity with only a ruinously expensive and disruptive complete refit of both infrastructure and traction available to switch away again.
If the entire British railway market for signalling equipment is up for grabs, people would have developed compatible solutions.

EDIT:
Thales, Alcatel and Siemens all manufacture LZB derived equipment, and there are likely others, so its hardly a single supplier situation.

The global rail industry has wanted an interoperable system for nearly a century. ETCS is that system and while it may not be perfect as it stands, it is open, extensible, flexible and can evolve.
And has cost us 30-40 years of wasted time.
 
Last edited:

MarkyT

Established Member
Joined
20 May 2012
Messages
6,232
Location
Torbay
EWR is planning for a Tempsford/south of St Neots station.
I wonder if the proposed site moved sufficiently north it might replace St. Neots station. In any case, a 125mph speed limit for fast line platforms somewhere in this area might destroy any case for 140mph, such that any case exists at all in the first place.
 

MarkyT

Established Member
Joined
20 May 2012
Messages
6,232
Location
Torbay
If the entire British railway market for signalling equipment is up for grabs, people would have developed compatible solutions.

And has cost us 30-40 years of wasted time.
Both these systems are predominantly high speed focussed, not for general use, for which France developed their KVB eventually after not having anything beyond the crocodile (AWS equivalent) until the 1990s, and Germany used their 'Indusi' PZB which was admittedly invented in the 1930s, and is something like a more advanced TPWS which is actually failsafe, but was only rolled out on a fairly limited proportion of the network until fairly recently (nothing in the fromer DDR I believe fro example) LZB and TVM also have complete conventional signalling infrastructure under the hood with fixed train detection, point machines, interlockings etc. so they would not have saved anything on any signalling renewal work in the UK over that period. TPWS or something very similar could plausibly have been rolled out maybe a decade earlier, but the UK was experimenting with proprietary systems under the BR ATP pilot scheme and was not actually significantly behind everywhere in Europe despite what the press have implied. Privatisation undoubtedly took eyes off balls however. UK was behind the best, no doubt. Sweden, for example, had a full supervision ATP overlay beginning in the 1970s and complete throughout most of its network by the 1980s. This is now being transitioned to be emulated using ETCS equipment.
 

HSTEd

Veteran Member
Joined
14 Jul 2011
Messages
16,635
Both these systems are predominantly high speed focussed, not for general use, for which France developed their KVB eventually after not having anything beyond the crocodile (AWS equivalent) until the 1990s, and Germany used their 'Indusi' PZB which was admittedly invented in the 1930s, and is something like a more advanced TPWS which is actually failsafe, but was only rolled out on a fairly limited proportion of the network until fairly recently (nothing in the fromer DDR I believe fro example) LZB and TVM also have complete conventional signalling infrastructure under the hood with fixed train detection, point machines, interlockings etc. so they would not have saved anything on any signalling renewal work in the UK over that period.

Whilst TVM is predominantly high speed focussed, there is nothing that requires it inherently be so, as the installation in the Channel Tunnel proves.
(It doesn't particularly care what each speed code means, that is decided based on the network code in the datagram, which potentially allows code 'juggling' to allow far more than six codes)

LZB apparently allows for arbitrary speed limits, although finding English language documentation is something of a challenge!

Also the complete absence of all signal gantries is likely to have saved some money/space, and potentially avoided at least notorious fatal accidents, like Ladbroke Grove. Can't get confused which signal is for you, or have sunlight interfere with signal reading, if the signal aspect is on your desk!
Worth noting that ETCS will, for the forseable future, largely due to freight operator intransigence, require a complete signal interlocking (with train detection) system be provided anyway.

Also as to early deployments not being practical because of tech, in the US they had full blown cab signalling with train protection in the 20s! (PRR pulse code system and it's many many derivatives)
 
Last edited:

edwin_m

Veteran Member
Joined
21 Apr 2013
Messages
24,793
Location
Nottingham
The government press release says it'll go as far as Stoke Tunnel, unless that's a 'phase 2' part. Would seem like an odd place to stop!
I seem to recall that one end of the tunnel is the fringe between Peterborough and Doncaster signaling control areas, so it may just be replacing Peterborough for the time being.
 

edwin_m

Veteran Member
Joined
21 Apr 2013
Messages
24,793
Location
Nottingham
Privatisation undoubtedly took eyes off balls however. UK was behind the best, no doubt.
I was on the committee that led to the development of TPWS very much in parallel with the privatization process, motivated by the need to adopt something that was easier and cheaper for wider application than the proprietary and mutually incompatible BR-ATP systems. Once that got under way the main delay was waiting for government funding (during which Ladbroke Grove happened, which it would have prevented).
 

NotATrainspott

Established Member
Joined
2 Feb 2013
Messages
3,223
In principle digital signaling allows higher speeds. However any attempt to increase the speed will affect the fastest trains most, and probably won't affect the slowest trains at all. The increase in difference between fastest and slowest journey times would exacerbate the capacity problems the ECML already has, and outweigh the fairly small capacity benefit provided by the digital system itself. Therefore I don't see it happening.

If the capability is there, it could be used to help timetable recovery. Going from 125 to 140mph won't make a vast difference to most journey times but it might help a train arrive in the right slot at a junction if it's a few minutes late.
 

edwin_m

Veteran Member
Joined
21 Apr 2013
Messages
24,793
Location
Nottingham
If the capability is there, it could be used to help timetable recovery. Going from 125 to 140mph won't make a vast difference to most journey times but it might help a train arrive in the right slot at a junction if it's a few minutes late.
On a busy mixed-traffic railway this only works if nobody has let a slower train out in front of the delayed fast train, and if they haven't that any slow trains thus delayed don't cause knock-on delays.
 

MarkyT

Established Member
Joined
20 May 2012
Messages
6,232
Location
Torbay
Whilst TVM is predominantly high speed focussed, there is nothing that requires it inherently be so, as the installation in the Channel Tunnel proves.
(It doesn't particularly care what each speed code means, that is decided based on the network code in the datagram, which potentially allows code 'juggling' to allow far more than six codes)

LZB apparently allows for arbitrary speed limits, although finding English language documentation is something of a challenge!

Also the complete absence of all signal gantries is likely to have saved some money/space, and potentially avoided at least notorious fatal accidents, like Ladbroke Grove. Can't get confused which signal is for you, or have sunlight interfere with signal reading, if the signal aspect is on your desk!
Worth noting that ETCS will, for the forseable future, largely due to freight operator intransigence, require a complete signal interlocking (with train detection) system be provided anyway.

Also as to early deployments not being practical because of tech, in the US they had full blown cab signalling with train protection in the 20s! (PRR pulse code system and it's many many derivatives)
It's pointless arguing about 'might have beens'. BR commenced trials of two different manufacturers' systems but very quickly realised the proprietary system route was not the correct way ahead and both were much much more expensive than TPWS. We are where we are now and ETCS is the only show in town. Interlocking is ALWAYS required, even in the case of non-fixed train detection techniques. Interlocking is conceptual though but functionally must be proven safe enough, although could notionally be distributed and embedded in trackside devices, even to some extent processed onboard the train instead of being centrally located as tends to be the case today. On main lines, even if self length and completeness detection is widely available, fixed virtual blocks mediated by trackside interlockings are most likely to be retained. Freight companies are not intransigent, they is simply attempting to remain in business without having to fit complex intercommunicating electronic systems on every wagon that could potentially bankrupt themselves or destroy reliability for all traffic. I'll give you signal gantries, but practically, migration strategies may have required them to be kept in the medium term anyway. TVM doesn't work below 60kph (I think?) hence why St Pancras is conventionally signalled with KVB protection. Stratford International is a fast layout with TVM throughout on the main lines and platfrom loops, with KVB going into the Temple Mills depot area. Eurotunnel is predominantly TVM but if they have any junction areas to sidings etc that are below 60kph, the standard French solution would be to drop into KVB. DB uses PZB in slower areas, LZB out on higher speed main lines. I think all of these systems will eventually either be replaced by ETCS or, more interestingly perhaps, be migrated to an ERTMS emulation, using standard ETCS hardware and protocols on track and train, but retaining the legacy functionality as limited or full supervision class B systems. This is the new evolutionary approach becoming more widely used on the continent. Use full L2 on new lines and very high capacity areas, create a new hybrid as required elsewhere that nevertheless retaining interoperability for the latest traction through the limited supervision mode, or with a special Class B legacy emulation.
 

HSTEd

Veteran Member
Joined
14 Jul 2011
Messages
16,635
Freight companies are not intransigent, they is simply attempting to remain in business without having to fit complex intercommunicating electronic systems on every wagon that could potentially bankrupt themselves or destroy reliability for all traffic.

Instead we expend vast sums of public money to allow them to avoid spending that money.
They should be required to bear the full commercial cost of all trackside train detection infrastructure, for as long as they do not adopt ECP brakes, and the full cost of all signal masts, colour light signalling equipment, TPWS equipment, AWS equipment and a bunch of other things whilst they do not have ETCS equipment installed.
If they go out of business, so be it.

Perhaps a carrot can be provided in terms of grant for fitting the required equipment, but US experience indicates the costs are not particularly high. Supposedly costs have been estimated at only $5,600 per wagon

Even ten thousand wagons would cost on order of $60m. How big is the UK wagon fleet?
I'll give you signal gantries, but practically, migration strategies may have required them to be kept in the medium term anyway. TVM doesn't work below 60kph (I think?) hence why St Pancras is conventionally signalled with KVB protection.
TVM in the Channel Tunnel apparently includes a 30km/h speed code.
I think the reason KVB was employed at St Pancras was predominantly that only six speeds were available because of the decision to only utilise a single route code, and given that international trains would have to have KVB anyway, there was little reason not to have it fitted at St Pancras. The Line was built essentially as if it was in France after all.

EDIT:

Found the study the $5600 USD comes from, its here - AU$8000 per wagon, which is about ~5500USD at present exchange rates.

EDIT #2:
We're going off topic though - so I will say this:

I'm well aware that ETCS is now the only game in town, but there simply should be no more conventional resignalling projects in the pipeline at all - it should be ETCS L2 or ETCS L2/L3 Hybrid only from now on.
If the freight operators aren't prepared to operate in that environment then they should get out of the business.
 
Last edited:

Ianno87

Veteran Member
Joined
3 May 2015
Messages
15,215
If the capability is there, it could be used to help timetable recovery. Going from 125 to 140mph won't make a vast difference to most journey times but it might help a train arrive in the right slot at a junction if it's a few minutes late.

Better spending less money (e.g. dwell time initiatives) to make the trains not routinely a "few minutes late" in the first place.
 

Taunton

Established Member
Joined
1 Aug 2013
Messages
10,019
Going from 125 to 140mph won't make a vast difference to most journey times but it might help a train arrive in the right slot at a junction if it's a few minutes late.
Gerry Fiennes, in "I Tried to Run a Railway" had, as ever, this one nailed half a century ago. An express was being regularly delayed by a local, the proposal was put to him to add time to the express schedule to allow for this. He said No. "The solution is not to slow down the express, the solution is to make the local run on time. Do it".

I was on the committee that led to the development of TPWS very much in parallel with the privatization process, motivated by the need to adopt something that was easier and cheaper for wider application than the proprietary and mutually incompatible BR-ATP systems. Once that got under way the main delay was waiting for government funding (during which Ladbroke Grove happened, which it would have prevented).
It's a real irony that the worst two accidents of that era which would have been preventable by an enhanced protection system, Southall and Ladbroke Grove, both happened on the one main line on which ATP hardware had already been installed but the two train operators involved had both put together a financially-driven justification why they should not use it on their trains.
 
Last edited:
Joined
21 Feb 2011
Messages
194
Location
Doncaster
The government press release says it'll go as far as Stoke Tunnel, unless that's a 'phase 2' part. Would seem like an odd place to stop!
Odd until you realise that Stoke Tunnel is the area where control of the ECML swaps between Peterborough PSB and Doncaster PSB...
 

MarkyT

Established Member
Joined
20 May 2012
Messages
6,232
Location
Torbay
It's a real irony that the worst two accidents of that era which would have been preventable by an enhanced protection system, Southall and Ladbroke Grove, both happened on the one main line on which ATP hardware had already been installed but the two train operators involved had both put together a financially-driven justification why they should not use it on their trains.
It was very tragic indeed, but by privatisation, the proprietary GW-ATP pilot technology was already acknowledged to be a dead-end. The only other major user was Belgian state railways who employed a much less sophisticated variant, TBL1, with only simple distant warning and train stop protection, while BR implemented full speed supervision ATP functionality using the equipment. In 2006, Belgian infrastructure authority Infrabel started to replace TBL1 in its entirety and expand its coverage with a Eurobalise-based equivalent, TBL1+. This is now complete so the old technology is extinct in Belgium. The only other application was on the East Rail Line in Hong Kong where it is due to be phased out in 2020. GWML will remain the world's last example of this technology, originally developed by the Belgian ACEC conglomerate, now defunct and with its rail business assets transferred to Alstom who have clearly rationised their product portfolio. The other BR pilot scheme, on the Chilterns, uses a kind of intermittent fixed block SelTrac, with similar, but in this case non-continuous transmission loops placed between the rails. That was supplied by Alcatel who sold their signalling business to French conglomerate Thales. They have been very successful in selling SelTrac in the metro market.
Transmission balise-locomotive (Dutch: Transmissie Baken Locomotief) (TBL) (English: "Beacon-to-Locomotive Transmission") is a train protection system used in Belgium and on Hong Kong's East Rail Line.
...
TBL1+ began to be installed on the Belgian rail network in 2006, with full conversion to the system expected by 2012. The system is a development of TBL1, but uses Eurobalises, providing compatibility for later reconfiguration to an ETCS system. An additional safety measure is the addition of an additional balise 300m up-line, which forces an automatic stop on trains travelling at more than 40 km/h ahead of a red signal. The roll-out of TBL1+ was accelerated after the Halle train collision on 15 February 2010, which killed 19 and injured 171 people. Similar to the Swiss EuroSignum approach, it simply converts the aspects from TBL1 into a digital transmission packet so that it is compatible with ETCS, while not interfering with ETCS functionality, by using the national datagram (packet number 44) that is defined in the ERTMS / ETCS specification requirements specification (subset 026).
 

Irascible

Established Member
Joined
21 Apr 2020
Messages
1,988
Location
Dyfneint
IIRC the TVM codes transmitted to the train don't inherently mean anything, but are indexes into a table on the train itself to get the speed/colour? so it's not inherently limited to fast trains, it's just you wouldn't want to mix them usually. It interested me enough ( signalling does in general ) to study how it works, but that was quite a few years ago now.
 

High Dyke

Established Member
Joined
1 Jan 2013
Messages
4,267
Location
Yellabelly Country
Kings Cross & Peterborough Box Areas , Doncaster Box Area from Stoke Tunnel ?
Correct. Doncaster take over at the south end of Stoke Tunnel.

This has been in the offing for some years now. And knowing how slow the railway goes it will probably another ten years before it even gets done. Kings Cross PSB is still partially operating, with York ROC covering the computer game signalling parts along the area. And Peterborough PSB is still going strong.
 

MarkyT

Established Member
Joined
20 May 2012
Messages
6,232
Location
Torbay
IIRC the TVM codes transmitted to the train don't inherently mean anything, but are indexes into a table on the train itself to get the speed/colour? so it's not inherently limited to fast trains, it's just you wouldn't want to mix them usually. It interested me enough ( signalling does in general ) to study how it works, but that was quite a few years ago now.
Makes sense, but everywhere a particular train goes has to use codes meaning the same thing! It's a bit of a dead-end system though with codes superimposed on track circuits for movement authority. Digital balises and radio are the way forward, whether ETCS or not. Interesting that China still likes track circuit codes which they're formalised their own ETCS derived modern digital hierachy. I think that enthusiasm originated from American/Japanese practice. The US is where track circuit speed codes started of course with the Pennsylvania RR pulse code systems of nearly a century ago!
The Chinese Train Control System (CTCS, simpl.chinese: 中国列车控制系统) is a train control system used on railway lines in People's Republic of China. CTCS is fundamentally identical to the European Train Control System (ETCS).
It has two subsystems: ground subsystem and onboard subsystem. The ground subsystem may based on balise, track circuit, radio communication network (GSM-R), and Radio Block Center (RBC). The onboard subsystem includes onboard computer and communication module.
CTCS has several levels:
CTCS-0: Track Circuit + Cab Signalling + LKJ2000, [2] - CTCS-0 equates to no CTCS installed.
CTCS-1: Track Circuit + Cab Signalling + LKJ2000 + Balise,[2]
CTCS-2: Track Circuit + Balise + ATP, the track circuit is used both for block occupation detection and movement authorization, its architecture is similar to TVM-300.
CTCS-3D: Track Circuit + Balise + ATP, CTCS-3D is equivalent to the European ETCS Level-1
CTCS-3: Balise + GSM-R + ATP, using CTCS-2 as the backup system, [2] CTCS-3 is equivalent to the European ETCS Level-2 + CTCS-2.[3]
CTCS-4: Balise + GSM-R + ATP, moving block
Levels 2, 3, and 4 are back-compatible with lower levels.
 

Irascible

Established Member
Joined
21 Apr 2020
Messages
1,988
Location
Dyfneint
Makes sense, but everywhere a particular train goes has to use codes meaning the same thing! It's a bit of a dead-end system though with codes superimposed on track circuits for movement authority. Digital balises and radio are the way forward, whether ETCS or not.

Well more to the point, the tables on every train have to be set up with some knowledge of the performance of every other train on any line it's going to use - or presumably you could use different lookup ( speed ) tables for different routes for the same train ( as far as I remember the codes themselves don't ever change, they're a complicated set of aspects if you like - I will have to go reaquaint myself I think ), but that does not sound sufficiently robust unless there's some *extremely* strong interlocking. It'd be theoretically doable, issue the train with a token that locks in a particular table, but it seems a messy kludge. ETCS is a better solution, no doubt there - so it should be.
 
Last edited:
Joined
6 Jan 2018
Messages
109
Location
Carluke
I detest this “Digital Railway” nonsense conjured up by a prior NR CEO. The railways have been digital since time immemorial.

Semaphores were “mostly” On or Off (1 or 0). Block bells are a sequence of On or Off to relay a message. Telegram systems were a series of On or Off. Relay based interlockings are an outcome based on a sequence of On or Off. Track Circuits are Clear or Occupied. Distant transmission systems in use since the 50s to communicate with assets many many miles from the signalling centres were some of the earliest examples of digital transmission over Time Division Multiplex. SSI is a properly digital system continually sending and receiving a string of 1 or 0 to control remote assets and receive indications. I can trace faults by observing a stream of bits (which, in a nod to history, we call a Telegram), on a system designed in the late 80s.

The Railway was using “digital” long before many of these visionaries were glints in their grandparents eyes.

There are many benefits to ETCS and ERTMS, but most of them are lost on a mixed use Railway, carrying traffic with maximum speeds that vary from 60 to 125mph. This, is just a sound bite, to appease those in the Fens, West Anglia and the surrounding areas that feel left out by HS2.
 

JN114

Established Member
Joined
28 Jun 2005
Messages
3,346
They’re saying ECML is to be first Mainline with “Digital Railway”, but here on Western route we already have the traffic management part of ERTMS/Digital Railway, and on top of the recently commissioned installation of ETCS L2 overlay on the Heathrow branch, they’re currently installing and testing ETCS on the GWML between Paddington and Slough. Not to mention the existing installation (without Traffic management?) on Thameslink and Cambrian...
 

cle

Established Member
Joined
17 Nov 2010
Messages
4,008
Is there any potential to go above 140mph in this system, where journey times might actually be impacted? Curious as to the limits, and if that is system, stock, stations, track spacing.... anything else?

North of York would especially be useful, as future HS2 services could also use it. But appreciate the main raceway is below there :)
 

Bald Rick

Veteran Member
Joined
28 Sep 2010
Messages
29,070
Is there any potential to go above 140mph in this system, where journey times might actually be impacted? Curious as to the limits, and if that is system, stock, stations, track spacing.... anything else?

North of York would especially be useful, as future HS2 services could also use it. But appreciate the main raceway is below there :)

It’s literally everything except alignment.
 

4F89

Member
Joined
17 Aug 2018
Messages
860
Instead we expend vast sums of public money to allow them to avoid spending that money.
They should be required to bear the full commercial cost of all trackside train detection infrastructure, for as long as they do not adopt ECP brakes, and the full cost of all signal masts, colour light signalling equipment, TPWS equipment, AWS equipment and a bunch of other things whilst they do not have ETCS equipment installed.
If they go out of business, so be it.

Perhaps a carrot can be provided in terms of grant for fitting the required equipment, but US experience indicates the costs are not particularly high. Supposedly costs have been estimated at only $5,600 per wagon

Even ten thousand wagons would cost on order of $60m. How big is the UK wagon fleet?

TVM in the Channel Tunnel apparently includes a 30km/h speed code.
I think the reason KVB was employed at St Pancras was predominantly that only six speeds were available because of the decision to only utilise a single route code, and given that international trains would have to have KVB anyway, there was little reason not to have it fitted at St Pancras. The Line was built essentially as if it was in France after all.

EDIT:

Found the study the $5600 USD comes from, its here - AU$8000 per wagon, which is about ~5500USD at present exchange rates.

EDIT #2:
We're going off topic though - so I will say this:

I'm well aware that ETCS is now the only game in town, but there simply should be no more conventional resignalling projects in the pipeline at all - it should be ETCS L2 or ETCS L2/L3 Hybrid only from now on.
If the freight operators aren't prepared to operate in that environment then they should get out of the business.
How to repair/maintain the pway without freight operators? Idiotic.
 

HSTEd

Veteran Member
Joined
14 Jul 2011
Messages
16,635
How to repair/maintain the pway without freight operators? Idiotic.

Using departmental rolling stock?

The value of purchasing additional such stock to make up for any shortfall is drastically exceeded by the costs associated with the status quo continuing functionally forever.
 

Tom Quinne

On Moderation
Joined
8 Jul 2017
Messages
2,225
I’m sure Network Rail will foot the bill for a National fitment of all cabs when the system goes live, just like TPWS and GSMR.

I wonder if that is factored into the £1.2bn “investment”?
 

Tom Quinne

On Moderation
Joined
8 Jul 2017
Messages
2,225
Before we “invest” in digital signalling (DfT description) maybe we should actually invest in higher speed turnouts for loops, and main to relief crossings to allow slower trains to get out the way of faster ones...faster?

Reassess the need for approach control signalling, where 2000t freight trains are brought down to a crawl only for the signal to clear into a loop or relief line?

Trust professional drivers to driver their trains professionally maybe?

But we all know treating Railway professionals like kids is a money making scheme which bleeds money out of the industry.
 
Status
Not open for further replies.

Top