SSR resignalling

[bramling]

LUL are renowned for chopping and changing requirements and most signalling contractors had their fingers burnt with abortive tender costs. It has taken some time, money and light contractural touch to rebuild these relationships. Perhaps they now see LUL as a soft touch particularly regarding claims/ variations

The other issue is that LU is not a particularly well informed client.

So much expertise was lost from LU as a result of the PPP years, and they have never really recovered. To make things worse TFL is a rather more political organisation than LT ever was.

It’s rather sad to see how low LU has sunk when one looks back at some of the incredible innovation and engineering they presided over more or less throughout their existence last century.

== Doublepost prevention - post automatically merged: Friday at 23:31 ==

That's where the contract needs to be more watertight with fines for non compliance.

“Okay, we understand you want to keep your old signalling on the SSR, and we wish you the best of luck with the life-expired cabling in various areas, perhaps Westinghouse will allow you to have another go with DTG-R”

Unfortunately, when you have a flagship project which is decades late and do nothing isn’t really an option, you don’t really get to dictate the terms of the deal too heavily.

 

[boiledbeans2]

LUL are renowned for chopping and changing requirements and most signalling contractors had their fingers burnt with abortive tender costs. It has taken some time, money and light contractural touch to rebuild these relationships. Perhaps they now see LUL as a soft touch particularly regarding claims/ variations

Now, the suppliers laugh at the incompetence of LU/TfL. The smart suppliers underbid during tender. They get awarded the contract. As GFE mentions, LU are renowned for chopping and changing requirements. The smart suppliers then bring their lawyers in and calculate how much to charge LU/TfL for varying the contract.

 

[Recessio]

Forgive me if this is a stupid question, or if it has been asked before: but why go for a proprietary system rather than something like the interoperable standards of ETCS ATO?

 

[Bald Rick]

Forgive me if this is a stupid question, or if it has been asked before: but why go for a proprietary system rather than something like the interoperable standards of ETCS ATO?

Two reasons:

ATO over ETCS wasn‘t a thing when this was specified. Thameslink has the first application of it worldwide, and it was (and remains) bespoke and non-standard. ATO over ETCS being implemented now is to standard.

ETCS does not (or at least did not) have the ‘cycle time’ to accommodate very high frequency metros. Cycle time is the time in seconds between each communication between the train and the Radio Block Centre (RBC). Although the difference between ETCS and CBTC systems is only a few seconds, this makes a difference for capacity.

 

[bramling]

Forgive me if this is a stupid question, or if it has been asked before: but why go for a proprietary system rather than something like the interoperable standards of ETCS ATO?

I don’t think ETCS was sufficiently mature then, and certainly wasn’t able to support ATO in any proven way. In any case, both Infracos wanted relatively proven systems in order to minimise the level of risk.

There would also be the issue that TFL at least had some prior connection with the Westinghouse and Thales systems, the former being at least outwardly similar to the Central Line, and the latter having been on the DLR.

What is possibly more interesting is whether ETCS could have been used for the central section of Crossrail, given its originally planned opening time was similar to the Thameslink Programme.

 

[100andthirty]

It was London Underground (i.e., not an Infraco) that specified and contracted the sub-surface signalling after the DTG-r system was abandoned as a result of Metronet going into administration. The first attempt was with Bombardier, which was abandoned and then placed with Thales (now Hitachi) for the Seltrac CBTC system but with WiFi comms instead of loops. Usually, metro CBTC system are implemented on comparatively simple lines. Even the relative complexity of the Northern line is simple compared with the sub-surface network. I doubt that either party knew what problems would be encountered wne setting off on this contract. Just one example - look at the length of the overrun areas between the train stopping positions and the buffers at Stanmore and Stratford on the Jubilee. Then compare with the space available at Hammersmith H&C. Clearly a solution was found but it was neither quick nor easy. The same sort of isses applied all over the network.

 

[Bald Rick]

What is possibly more interesting is whether ETCS could have been used for the central section of Crossrail, given its originally planned opening time was similar to the Thameslink Programme.

ETCs with the bespoke ATO overlay was certainly an option, but as you say, TfL wanted proven kit.

There was an unofficial bet between Thameslink and Crossrail as to who would get a train operating on ATO in passenger servcie first, which Thameslink won.

 

[Roast Veg]

How far from the now-standardised ATO-on-ETCS is Thameslink?

 

[notverydeep]

Forgive me if this is a stupid question, or if it has been asked before: but why go for a proprietary system rather than something like the interoperable standards of ETCS ATO?

ETCS is Europe’s railways way of admitting that Metro signalling principles are superior…

 

[bramling]

It was London Underground (i.e., not an Infraco) that specified and contracted the sub-surface signalling after the DTG-r system was abandoned as a result of Metronet going into administration. The first attempt was with Bombardier, which was abandoned and then placed with Thales (now Hitachi) for the Seltrac CBTC system but with WiFi comms instead of loops. Usually, metro CBTC system are implemented on comparatively simple lines. Even the relative complexity of the Northern line is simple compared with the sub-surface network. I doubt that either party knew what problems would be encountered wne setting off on this contract. Just one example - look at the length of the overrun areas between the train stopping positions and the buffers at Stanmore and Stratford on the Jubilee. Then compare with the space available at Hammersmith H&C. Clearly a solution was found but it was neither quick nor easy. The same sort of isses applied all over the network.

I always felt the Victoria Line resignalling seemed to be heavily engineering-led, which seems to have led to a good outcome.

Ironically the Jubilee was probably more complex than the a Northern. Interface with conventional signalling in the Neasden/Wembley area, as well as interfaces with two fully signalled depots. They also had the challenge of migrating an existing control centre, which the Northern didn’t. For sure the Northern has a few fiddly layouts, but overall the complexity was actually quite low. I think to some extend this lulled LU into a false sense of security when it came to SSR, and by this time a lot of expertise had gone (or been purged when LU took over the Infracos).

All of LU has seen quite a decline over the last three decades in terms of the quality of how the operation is run. Hence why the service we see nowadays is quite lumpy, and incidents seem to take a lot longer to resolve - despite the influx of modern technology. This is what happens when an organisation fails to value expertise. The mind boggles as to why any organisation would spend over half a century developing expertise, and then see that as a threat.

 

[Bald Rick]

How far from the now-standardised ATO-on-ETCS is Thameslink?

Far enough that
a) it won‘t be extended, and,
b) it won’t be used elsewhere

 

[Roast Veg]

So much the same situation as TASS then. Pity.

 

[Recessio]

Thanks all for the very useful explanation!