Signalling Problems from start - Independent

[metrailway]

The Independent reports that Crossrail has major issues with its signalling system:

Crossrail faces the prospect of signal trouble from day one
Engineers say the interface with the national rail network is causing problems and could even fail when new line opens.

Crossrail, the South-East's £15bn new rail line, is on course for a humiliating signalling failure when it opens in 2018.

The project chairman, Terry Morgan, confirmed that a "mitigation plan" was being put in place as it grows increasingly likely that signalling on the 73-mile Berkshire-to-Essex link won't work properly. This would mean a reduction of train services from the start.

Crossrail uses a complicated mix of traditional metro signalling and state-of-the-art systems to link trains on the 26 miles of new tracks beneath London to existing major cross-country rail lines.

Engineers and rail experts are struggling to create "interfaces", meaning the trains are unable to smoothly switch from the relatively conventional Metro system to a higher-tech standard once they have left central London. Sources say with only three years to go these problems might not be resolved in time.

...

A source who has recently worked on Crossrail said: "The most likely scenario is that Crossrail will open with a significantly reduced service." A second source close to the project said: "In the extreme you wouldn't be able to run any trains. Signalling technology in this industry is something that has challenged everybody for a very, very long time."

[domh245]

Engineers and rail experts are struggling to create "interfaces", meaning the trains are unable to smoothly switch ...

(my emphasis)

I don't see what is wrong with the transition being at the first station outside of the core, in a fashion that is similar to what is planned for thameslink, unless the problem is with getting the ATO system to talk with the signalling system. In either case, the engineers on Xrail can learn from the Thameslink bunch, who will be facing a similar problem.

Having read a bit further into the article, this little gem pops up

looking to introduce a highly sophisticated signalling system across the railways, which will link with Crossrail.

This incorporates technology similar to GPS

Satellites don't work underground do they?

The article isn't very clear, so I want to try and find more about this in something like Rail, or the rail engineer - as at least then we might get more detailed information, it could be anything from reading that!

>ERTMS isn't very good/isn't developed enough yet
>They can't get their ATO system working
>They can't agree on how the trains will swap between systems
>They haven't got the signalling systems to talk together

&c.

[will]

Hope it doesn't go the same way the JLE did when they tried to integrate different systems. I always thought it was daft using 2 or 3 different systems on a single line moving from ATO to normal signalling!

[snoggle]

To be fair to Crossrail they have long recognised that the signalling system and integration with the NR network is a massive project risk. It is one of the reasons why the connections to NR and especially to the west are up to a year later than opening the tunnel core section. The headline could have been written several years ago and been broadly truthful given what the Crossrail Project Director said about signalling integration risks.

I am not particularly surprised to see headlines emerging on this theme especially as NR are considering changes to their wider philosophy on new systems. I do think there's the inevitable "doom and gloom" element to the headlines given Terry Morgan has said Crossrail are developing a mitigation plan which is entirely sensible given the risks.

[domh245]

I've heard that the problem is apparently between getting the ERTMS and the CBTC system that is going to be installed on the core to talk to each other. At this stage, would it be too late to scrap the CBTC system for the core and use a similar ERTMS based ATO system as per thameslink, because then there will only need to be 1 interface point where Xrail comes up onto the GEML, and any lessons that can be learnt from this project can be transferred to thameslink and vice-versa.

[snoggle]

This issue came up at the Transport Committee this week when Crossrail made their annual appearance in front of the committee.

Terry Morgan was perfectly clear that there are no issues with Crossrail's signalling system - it's on programme and proceeding as expected. There are also agreed interfaces with Network Rail for both eastern and western interface points. The issue about triggering mitigation was purely in recognition of the fact that Network Rail's policies about Digital Railway may bring about change to the interface to the GWML. It is entirely prudent to ensure those changes were understood and steps taken to ensure trains will be able to run.

Obviously none of this is a cast iron guarantee of perfection with the signalling but I got no sense of concern or panic from watching the webcast of the meeting.