ERTMS rollout on ECML and GWML

coppercapped · 9 Jul 2019

Belperpete said:
Not sure what 80xs are, but trains running through the CrossRail core will work under ATO, as this is necessary to achieve the headways. At the Western end, they will switch out of ATO before they cross onto Network Rail territory - I can't recall whether it is intended that the driver will takeover manual control on the move, or while the train is stationary at Paddington "low level". I understand that at the Eastern end the trains will continue under ATO into Stratford station, where the driver will take manual control. I suspect that this will be the first place on NR that has ATO?

On the move, on the surface near Westbourne Park station where the Crossrail reversing sidings are situated. It is a requirement that CBTC is proved to be working before the trains (Class 345s) enter the tunnels. At the eastern end the tunnel portal is west of Stratford which is in the open so its easier to arrange the changeover when the train is at the station.

To add to Kevin Roche's post above, another reason for the adoption of CBTC for the tunnels was that it permitted automatic train operation from the deep level Paddington platforms to the reversing sidings at Westbourne Park allowing the driver to walk to the other end of the train while this is going on. The current flavours of ETCS cannot offer this feature.

RailUK Forums

kevin_roche · 9 Jul 2019

Belperpete said:
I can't recall whether it is intended that the driver will takeover manual control on the move

At the western end it is supposed to happen while on the move and it will now be a transition between CBTC and TPWS followed by another transition between TPWS and ETCS for those going to Heathrow.

Crossrail call it Plan B, The original plan A was to use ETCS between Paddington and Airport Junction ( Actually I suspect the real original plan was all the way to Reading, but we are now on plan C.)

Belperpete said:
By that time, CrossRail will (hopefully!) have been up and running for quite some time. Swapping to a new signalling system will require a lot of work, expense and disruption, not least because it will require extensive retesting. Unless ETCS L3 offers substantial advantages, I think it will be hard to justify either the cost or disruption.

I'm guessing that the hope is that ETCS L3 will be influenced by Siemens to such an extent that it is very similar if not the same. By using Euro Balises for positioning the only real question is if the radio link is the same or not. If substantially different I guess the upgrade will never happen or only happen when the wi-fi parts of the signalling gets to end of life.

LNW-GW Joint · 9 Jul 2019

kevin_roche said:
That is all quite disappointing.

Well, HS2 will be built with ETCS from the start.
That will extend to Lichfield/Crewe in the 2020s, and the northern WCML will also be resignalled in the next decade (Crewe, Warrington, Preston, Carlisle).
One hopes that will consider ETCS so that HS2 classic-compatible stock can take advantage of it.

edwin_m · 9 Jul 2019

Belperpete said:
Not sure what 80xs are, but trains running through the CrossRail core will work under ATO, as this is necessary to achieve the headways. At the Western end, they will switch out of ATO before they cross onto Network Rail territory - I can't recall whether it is intended that the driver will takeover manual control on the move, or while the train is stationary at Paddington "low level". I understand that at the Eastern end the trains will continue under ATO into Stratford station, where the driver will take manual control. I suspect that this will be the first place on NR that has ATO?

The Thameslink core has the first application anywhere of ATO over ETCS, which has been announced as working although I don't think it's being routinely used yet. I'm not sure whether Crossrail could have used that system (based on fixed block but with shorter blocks than the signal sections) or whether they needed moving block. Either way they put together a justification to convince the EU to make an exception to the normal presumption that the signalling on a new national network route should be ETCS-based.

kevin_roche · 9 Jul 2019

LNW-GW Joint said:
One hopes that will consider ETCS so that HS2 classic-compatible stock can take advantage of it.

The rules say all new lines have to be ETCS unless an exemption is granted. The crossrail exemption has the requirement that CBTC be replaced with ETCS Level 3 when it can be.

The new Digital Railway Deployment Plan from network rail will mean that any major replacement work will be ETCS too.

Belperpete · 9 Jul 2019

edwin_m said:
The Thameslink core has the first application anywhere of ATO over ETCS,

Ah, thanks for that. I knew that Thameslink was ETCS, but hadn't appreciated that it was ATO too.

coppercapped said:
another reason for the adoption of CBTC for the tunnels was that it permitted automatic train operation from the deep level Paddington platforms to the reversing sidings at Westbourne Park allowing the driver to walk to the other end of the train while this is going on.

Again, thanks for that info. I knew that the reversing moves are intended to be driven automatically - I hadn't appreciated that the driver wouldn't be in attendance while they are happening. I had assumed that, like with in-service passenger running, the automation was for headway reasons, not just to save a turn-round driver.

MarkyT · 9 Jul 2019

kevin_roche said:
The rules say all new lines have to be ETCS unless an exemption is granted. The crossrail exemption has the requirement that CBTC be replaced with ETCS Level 3 when it can be.

The new Digital Railway Deployment Plan from network rail will mean that any major replacement work will be ETCS too.

Or 'ETCS ready', which most modern signalling has been anyway for some decades by virtue of being processor based. The crossrail CBTC product is not designed to be ETCS ready. It is one of those bad old proprietary systems that ETCS was conceived to avoid main line rail authorities becoming locked into that are still common in the mass transit metro market, where going proprietary was the only realistic option when crossrail suppliers were selected. I suspect the cl. 345 trains have been engineered to emulate the required CBTC behaviour in their native European Vital Computer (EVC - the core on-board system required by ETCS) by means of the Specific Transmission Module (STM) method. This means they may not have any major Siemens components on board at all except possibly some plug-in antennae reader STMs for the particular type of track transponders used. These could notionally migrate to standard ETCS components in some future change with the full CBTC functionality replicated through a packet 44 Class B framework that would not be too dissimilar to that developed for the Thameslink core, also engineered by Siemens, and incorporating ATO, with sufficient stopping accuracy demonstrated now to allow platform edge doors to be used.

kevin_roche · 9 Jul 2019

MarkyT said:
Or 'ETCS ready', which most modern signalling has been anyway for some decades by virtue of being processor based. The crossrail CBTC product is not designed to be ETCS ready. It is one of those bad old proprietary systems that ETCS was conceived to avoid main line rail authorities becoming locked into that are still common in the mass transit metro market, where going proprietary was the only realistic option when crossrail suppliers were selected. I suspect the cl. 345 trains have been engineered to emulate the required CBTC behaviour in their native European Vital Computer (EVC - the core on-board system required by ETCS) by means of the Specific Transmission Module (STM) method. This means they may not have any major Siemens components on board at all except possibly some plug-in antennae reader STMs for the particular type of track transponders used. These could notionally migrate to standard ETCS components in some future change with the full CBTC functionality replicated through a packet 44 Class B framework that would not be too dissimilar to that developed for the Thameslink core, also engineered by Siemens, and incorporating ATO, with sufficient stopping accuracy demonstrated now to allow platform edge doors to be used.

The track transponders are allegedly the very same ones

The ETCS-compliant Trainguard Euro-balise is used for intermittent track-to -train communications.

See page 7 of the 2018 Trainguard MT brochure https://www.mobility.siemens.com/mo...on/train-control-systems/trainguard-mt-en.pdf

The radio system is described as:

Radio technology
Airlink operates in line with WiFi standards using either worldwide available free (ISM bands) or licensed frequency bands.
Lowest risk of interference is achieved by means of diversity measures such as front /rear on-board equipment with four antennas, intelligent use of the frequency spectrum and a repetition of telegrams when indicated. LTE radio technology is also supported. If LTE is used Airlink ensures the secured data transmission between onboard and wayside ATC units.

MarkyT · 9 Jul 2019

kevin_roche said:
The track transponders are allegedly the very same ones

So presumably uses the standard ETCS balise reader module and antenna on board, just in a different mode to when in standard mainline ETCS territory.
That doesn't surprise me. Replacing any custom hardware in the old Matra system with more standardised components. The radio system is probably where ETCS itself will go in the future. The original GSM-R is very out of date now and alternatives are being looked into. Finland is implementing ETCS with a Tetra-based radio system for example, perhaps because the lower frequency used gives longer range, permitting large geographic coverage with a smaller number of transmitters. Probably rather important to control costs in a large sparsely populated country.

The Planner · 9 Jul 2019

LNW-GW Joint said:
Well, HS2 will be built with ETCS from the start.
That will extend to Lichfield/Crewe in the 2020s, and the northern WCML will also be resignalled in the next decade (Crewe, Warrington, Preston, Carlisle).
One hopes that will consider ETCS so that HS2 classic-compatible stock can take advantage of it.

Crewe will be, the others may well be CP8. ETCS overlay is more likely as I doubt Crewe will be ETCS from the start.

kevin_roche · 9 Jul 2019

MarkyT said:
Finland is implementing ETCS with a Tetra-based radio system for example, perhaps because the lower frequency used gives longer range, permitting large geographic coverage with a smaller number of transmitters. Probably rather important to control costs in a large sparsely populated country.

Yes I forgot that Tetra was also being considered. In London one of the issues is the capacity of the system and the need to limit coverage to as small an area as possible in order to avoid interference. With GSM-R only having 2 channels that is awkward to achieve.

RichardGore · 9 Jul 2019

Belperpete said:
By that time, CrossRail will (hopefully!) have been up and running for quite some time. Swapping to a new signalling system will require a lot of work, expense and disruption, not least because it will require extensive retesting. Unless ETCS L3 offers substantial advantages, I think it will be hard to justify either the cost or disruption.

It does not need to be justified because the decision to switch has already been taken—from my understanding that was a condition on using the CBTC system under a derogation from the European Commission.

What that means if we have left the Union is anybody’s guess, but assuming that either we remain members at that point, or we are bound under a future agreement, or even under international law, it is legally required that we upgrade the signalling in the core to L3.

RichardGore · 9 Jul 2019

Belperpete said:
I suspect that this will be the first place on NR that has ATO?

The Thameslink core has ATO, and that’s NR territory.

coppercapped · 9 Jul 2019

RichardGore said:
It does not need to be justified because the decision to switch has already been taken—from my understanding that was a condition on using the CBTC system under a derogation from the European Commission.

What that means if we have left the Union is anybody’s guess, but assuming that either we remain members at that point, or we are bound under a future agreement, or even under international law, it is legally required that we upgrade the signalling in the core to L3.

This is 'angels dancing on a pin' territory...

Is changing the signalling in the core from Siemens' CBTC to L3 an 'upgrade'? Will L3 permit more trains to be run? If it doesn't, it is not an upgrade.

If I remember what I read correctly, TfL demonstrated that as the Crossrail tunnels were never going to be used by any trains other than its own, the Commission's argument that the line be made 'interoperable' was not made. Also the CBTC permitted modes of operation that ETCS could not offer, certainly at the time that plans were being made. As a result Crossrail was granted a derogation and to keep everybody happy it was agreed that when the time comes to replace the signalling Crossrail would adopt ETCS.

In practice this is meaningless. The CBTC signalling will still be working in 40 years time - and who knows what ETCS, CBTC or any one of other possibilities will be available or even if the EU's remit will still run to the UK.

edwin_m · 9 Jul 2019

It's probably in Crossrail's interest to upgrade to Level 3 at the time the 345s are replaced, or even if they need to enlarge that fleet (assuming Level 3 is available by that time). Given that the trains will need ERTMS anyway for Heathrow and probably for the rest of the GWML and maybe elsewhere, not fitting CBTC removes some cost and a lot of complexity.

RichardGore · 9 Jul 2019

coppercapped said:
This is 'angels dancing on a pin' territory...

Thank you, I’m very flattered to be compared with Thomas Aquinas.

coppercapped said:
Is changing the signalling in the core from Siemens' CBTC to L3 an 'upgrade'?

Yes.

Will L3 permit more trains to be run? If it doesn't, it is not an upgrade.

This is, if you’ll forgive me, a rather narrow view of what constitutes an upgrade. Replacing semaphore arms with colour-light signals doesn’t in itself increase capacity, but I think most people would agree that it’s not unreasonable to call that an upgrade. The additional OSS loops installed under TPWS+ are a safety improvement but under normal conditions make no changes to capacity at all, that would still be reasonable to call an upgrade. Installing a cab signalling system to allows trains to run faster, even if not more of them, would be an upgrade.

Moving to a standardised system with more options for rolling stop is an upgrade. Moving away from being beholden to Siemens’ patents and monopoly power is an upgrade. Moving to standardised components with which more engineers are familiar is an upgrade. Moving to a system with open standards and more options for suppliers is an upgrade. Moving to a system which does not require a separate driver training from the rest of the network is an upgrade.

Also the CBTC permitted modes of operation that ETCS could not offer, certainly at the time that plans were being made.

Hence the derogation. This does not apply to future signalling options.

In practice this is meaningless. The CBTC signalling will still be working in 40 years time - and who knows what ETCS, CBTC or any one of other possibilities will be available or even if the EU's remit will still run to the UK.

The EU’s remit presumably would still exist since it provided investment. I would hazard that a 40 year lifespan is perhaps an optimistic assessment given that it seems the signalling isn’t even reliably working right now. Regardless, 40 years is not a terribly long time in the railway.

Bald Rick · 9 Jul 2019

edwin_m said:
The Thameslink core has the first application anywhere of ATO over ETCS, which has been announced as working although I don't think it's being routinely used yet. I'm not sure whether Crossrail could have used that system (based on fixed block but with shorter blocks than the signal sections) or whether they needed moving block. Either way they put together a justification to convince the EU to make an exception to the normal presumption that the signalling on a new national network route should be ETCS-based.

Yes it’s all live on Thameslink, but not routinely used, pending driver training.

Crossrail took the decision not to go from ETCS L2 and ATO as at the time of the decision it was not a proven system in service*, albeit Thameslink were going down that route. Let’s just say the Thameslink signalling boys have been looking rather smug recently, as their system works and the Crossrail system .....

*another reason was the perceived need for auto-reverse at Paddington (as mentioned by @coppercapped) which ETCS/ATO doesn’t (yet) provide. Finally the ‘cycle time’ - the frequency that messages are sent from the block centre to the trains - is about 1 second for CBTC but can be up to 5 sec for ETCS, and in high frequency applications every second counts.

Nevertheless, had Crossrail chosen ETCS+ATO, there would be a much reduced problem with signalling /systems testing.

MarkyT · 9 Jul 2019

It's not an open access interoperable railway realistically. That's not to say another TOC (or whatever they become post Williams/Brexit) couldn't run through there if there was a really compelling case to do so, but they'd have to use carefully designed trains custom made to suit the platform edge door spacing and match the performance profile of the 345s so as not to reduce capacity, and they'd have to run the same stopping pattern for similar reasons. It's interesting that the Trainguard MT is now using some standard ETCS hardware as that suggest a migration path to a set of features implemeted within the ETCS ecosystem. I think that's how signalling will go generally. The hardware will be interoperable but software changes could allow different functionalities to be engineered into custom systems for the particular client. ETCS allows for this already with class B systems, which are generally legacy national protection systems, but can also be special implementations like WCML TASS, Thameslink ATO and even the totally new ZBS system implemented on the Berlin S-Bahn which is a limited supervision direct replacement for their previous mechanical train stop system using Eurobalises. https://en.wikipedia.org/wiki/Zugbeeinflussungssystem_S-Bahn_Berlin

MarkyT · 9 Jul 2019

Bald Rick said:
...Finally the ‘cycle time’ - the frequency that messages are sent from the block centre to the trains - is about 1 second for CBTC but can be up to 5 sec for ETCS, and in high frequency applications every second counts.

I suspect this response time issue will improve with replacement radio systems, as the original GSM-R is obsolete and there are so many alternatives available today.

Bald Rick said:
Nevertheless, had Crossrail chosen ETCS+ATO, there would be a much reduced problem with signalling /systems testing.

Paddington was an awfully complex and difficult layout to signal with the original SSI technology in the early 1990s. Designers ran into no end of timing problems with messages passing between the individual interlocking areas, but they persevered and thankfully solved all the issues eventually. The project really was on the edge of the new technology's capabilities at the time. The similar size and complexity of Waterloo around the same time was one reason why a relay based solution was selected there instead. Later processor based interlockings as used today don't have the same problems as each processor system can cover a much larger area. I suspect many of the problems they're having at Paddington today are in tying up the two varieties of trackside equipment and control functions, rather than the trains' ability to work with either systems. The transition is simply from CBTC to TPWS, just as it is at the GEML interface, but the GWML layout is far more complex and incorporates the reversing sidings of course.

HSTEd · 10 Jul 2019

Replacements for GSM-R won't be arriving any time soon.

It took a long time to get this far and tearing up the standard for a handful of marginal gains is not really going to get us much.

Bald Rick · 10 Jul 2019

HSTEd said:
Replacements for GSM-R won't be arriving any time soon.

It took a long time to get this far and tearing up the standard for a handful of marginal gains is not really going to get us much.

That’s funny, because I swear I’ve seen the programme for swapping to GPRS, and it’s happening soon. Siemens have the contract I think.

HSTEd · 10 Jul 2019

Bald Rick said:
That’s funny, because I swear I’ve seen the programme for swapping to GPRS, and it’s happening soon. Siemens have the contract I think.

I'm not sure I can characterise the adoption of GPRS, which is built upon the traditional GSM standard, as a replacement for GSM-R.

Bald Rick · 10 Jul 2019

HSTEd said:
I'm not sure I can characterise the adoption of GPRS, which is built upon the traditional GSM standard, as a replacement for GSM-R.

I’m no expert in this, but I understand that GPRS has much better data capacity and is what is necessary to enable the upgrades to ETCS cycle time.

HSTEd · 10 Jul 2019

Bald Rick said:
I’m no expert in this, but I understand that GPRS has much better data capacity and is what is necessary to enable the upgrades to ETCS cycle time.

My understanding is that GPRS is primarily an adaption of GSM, rather than a new technology like 3G or above.
Which means I would characterise GPRS...-R? As an adaption of GSM-R rather than a replacement.

Much of the same hardware is in use for example.

Belperpete · 10 Jul 2019

RichardGore said:
It does not need to be justified because the decision to switch has already been taken

I accept that a decision has been made that at some indeterminate time in the future a switch will be made. However, when it comes to the crunch, I cannot see any elected mayor justifying spending a considerable amount of ratepayers' money, and causing a considerable amount of disruption to the people who he needs to re-elect him, for something that will give no benefit, purely because of some vague promise that was made several decades ago. There will always be something more pressing to spend the money on, and some excuse to defer the upgrade ("we are not yet confident that L3 is sufficiently proven or suitable for use on CrossRail").

coppercapped · 11 Jul 2019

RichardGore said:
Thank you, I’m very flattered to be compared with Thomas Aquinas.

Yes.

This is, if you’ll forgive me, a rather narrow view of what constitutes an upgrade. Replacing semaphore arms with colour-light signals doesn’t in itself increase capacity, but I think most people would agree that it’s not unreasonable to call that an upgrade. The additional OSS loops installed under TPWS+ are a safety improvement but under normal conditions make no changes to capacity at all, that would still be reasonable to call an upgrade. Installing a cab signalling system to allows trains to run faster, even if not more of them, would be an upgrade.

Moving to a standardised system with more options for rolling stop is an upgrade. Moving away from being beholden to Siemens’ patents and monopoly power is an upgrade. Moving to standardised components with which more engineers are familiar is an upgrade. Moving to a system with open standards and more options for suppliers is an upgrade. Moving to a system which does not require a separate driver training from the rest of the network is an upgrade.

Hence the derogation. This does not apply to future signalling options.

The EU’s remit presumably would still exist since it provided investment. I would hazard that a 40 year lifespan is perhaps an optimistic assessment given that it seems the signalling isn’t even reliably working right now. Regardless, 40 years is not a terribly long time in the railway.

Ah! Semantics! Don't cha luv 'em!

In principle I agree with you that many of the points you make are indeed 'upgrades' in the sense of the technical development. I was considering an upgrade in this context to be an improvement in the service offered to passengers - and in this sense replacing semaphores with coloured lights with no other changes will not improve the service, although it may well have other benefits. In the same way changing CBTC to ETCS may have other benefits but the change will, of itself, not offer a faster or more frequent service to the passengers.

In any event, as others have posted, the technological change may be moot - as the technologies converge the differences may be marginal. This only underlines the point I was trying to make, and with a better example given by Belperpete - that a commitment made now to changeover from one system to another at some undefined point in the future is probably not to be taken literally.

kevin_roche · 11 Jul 2019

coppercapped said:
In any event, as others have posted, the technological change may be moot - as the technologies converge the differences may be marginal. This only underlines the point I was trying to make, and with a better example given by Belperpete - that a commitment made now to changeover from one system to another at some undefined point in the future is probably not to be taken literally.

In case there was any confusion I have no great expectations. Just stating that the commitment had been agreed. As we all know when it come to ETCS precious few cast iron commitments appear to be as solid a few weeks after they have been agreed.

kevin_roche · 12 Jul 2019

There are references to the ETCS installation in the Heathrow Tunnels in the Jacobs second period report on the Crossrail implementation plan. See Page 17.

Unfortunately what appear to be the most interesting parts have been redacted.

Infrastructure

As described in our last report,the testing of the train has identified 3 faults with N (redacted) ETCS. These are: inadequate (redacted) The train position issue is expected to be fixed in June 2019, but the (redacted) issues are proving difficult to fix, with the current date of (redacted) at risk.

This needs to be resolved before passenger service can begin. Lack of (redacted) at the start of driver training would require training runs to be repeated.

The position concerning the ETCS wayside receiving its ‘Putting into Use’ certificate from the W&W ETCS groupremains unchanged. This has been outstand(redacted) since November 2018 and may not occur until the 39 tests are completed, currently scheduled for August2019. CRL is aware of the issue.

Chris125 · 13 Jul 2019

kevin_roche said:
Peter Gracey of Bechtel told me at presentation to the IET Solent Section on 27th March that there was no problem with interference between ETCS and GW-ATP. I have no idea what he meant by that, he would not explain further.

I'm sure I read a comment somewhere else that suggested the same thing - that what we'd all been led to believe wasn't, at least now, the real issue and IIRC suggested it was actually the 345s at fault. Intriguing if true .

kevin_roche · 13 Jul 2019

Chris125 said:
I'm sure I read a comment somewhere else that suggested the same thing - that what we'd all been led to believe wasn't, at least now, the real issue and IIRC suggested it was actually the 345s at fault. Intriguing if true.

I have read elswhere, unfortunately I don't remember where, that the issues were with the software defined receiver on the 345. No evidence of that however so it could have been pure speculation.

ERTMS rollout on ECML and GWML

Established Member

RailUK Forums

Member

Veteran Member

Veteran Member

Member

Established Member

Established Member

Member

Established Member

Veteran Member

Member

Member

Member

Established Member

Veteran Member

Member

Veteran Member

Established Member

Established Member

Veteran Member

Veteran Member

Veteran Member

Veteran Member

Veteran Member

Established Member

Established Member

Member

Member

Established Member

Member