Before the project opening date was pushed back in August to next Autumn, today was the first working day Crossrail would have been operating on the core Paddington to Abbey Wood route. There’s been a lot of condemnation of the late announcement of the delay (and also the requirement to be bailed out to the tune of £1bn!) – but in truth, for such a big project to be delayed is not that unusual, which I’ll go into more about later.
The most outwardly obvious reason why the project may have been delayed is that some stations aren’t ready – now, it’s possible that progress on these has been deliberately slowed after the announcement, but I can’t quite see how some stations, such as Bond St, would have opened on time.
However, this isn’t the main reason for the delay – the primary reason is that the testing programme is several months behind schedule, and there’s one major factor in this – signalling.
Crossrail’s trains have had to be set up to work with three, entirely different, signalling systems. The section west of Paddington will ultimately use ECTS (essentially, a newly developed in-cab signalling system), the ‘Core’ section will use a form of ATO (Automatic Train Operation), and the Eastern section running out of Liverpool Street (oh, and for now, the Paddington – Heathrow section too) uses AWS/TPWS, which is the system currently in use on a majority of mainline railways in Britain.
Heathrow – Paddington
Let’s look at the Paddington to Heathrow section first. Even before the December launch, the Class 345 trains Crossrail will ultimately use should have been operating between Paddington and Heathrow under the TfL Rail brand (these will then be rebranded to Crossrail when the core section through London opens) – however they’re currently being curtailed at Hayes and Harlington. Why? You’ve guessed it, signalling!
Basically, the conventional signalling used by the Heathrow Express units in the Heathrow tunnels is having issues interfacing with the shiny new Crossrail units. ECTS has been installed for the use of Crossrail, however the units are having issues with interference between the conventional signalling and the new ECTS setup. As such, trains are being forced to terminate at Hayes & Harlington, with only the old Class 360 units TfL inherited from Heathrow Connect being able to continue on to the airport.
Core (Paddington to Abbey Wood)
However, ECTS isn’t the only signalling system being installed for Crossrail. The new section built under Central London uses ATO – essentially, due to the frequency of services (and also the use of platform-edge doors, so trains must stop in the exact position to line the doors up with these!), trains will be automatically driven in the central section, using a similar setup to that of the Thameslink route in their core section under the capital. This, too, is having issues – an issue with some of the electrical transformers being used on this central section has delayed testing, and there are still issues related to this which I understand are yet to be resolved. This basically means that the central section is inoperable at present.
Liverpool St to Shenfield
This is the only section where TfL Rail’s Class 345 are running the full intended route, though 50% of services should be branching off and continuing through the central core section. This wasn’t scheduled to happen until at least the end of next year anyway, but will now likely be delayed further due to the issues on the other parts of the network. As such, it’s likely we’re not going to have the full network running until at least the end of 2020.
That’s all great, but other major signalling projects have completed on time, haven’t they?
Actually, delays on signalling projects are more common than you might think. Modern railway signalling is complex, and Crossrail is far from the first victim of delays in completion:
Eurostar launched its new service to Amsterdam earlier this year, more-or-less on time, however the previous major project of theirs, introducing the Class 374 units, was anything but simple, as it has to cope with 4 different signalling systems (that of the UK, France, Belgium & the Netherlands). Whilst the delivery, and subsequent passenger launch, of the first unit wasn’t too badly delayed, it couldn’t run to Brussels for several months afterwards whilst glitches were ironed out with the train’s Belgian signalling system implementation. Additionally, the train isn’t compatible with traditional UK signalling, so until Ashford International was upgraded with signalling similar to the ECTS/KVB systems used elsewhere on its route earlier this year, only older trains could call there.
Whilst not on the same scale disruption-wise, Thameslink is still being manually-driven through the central Core section under London and the May 2020 deadline for ATO to be in force is likely to be pushed back – though ATO is virtually ready to go subject to final tests, timetabling issues, widely reported back in May, mean that the implementation is delayed and it’ll likely be 2020 before Thameslink run the full peak 24 trains per hour through this vital cross-London link.
So, when will Crossrail fully launch? Will we even see it in Autumn 2019?
With an announcement today hinting at a possible further delay, we may not even see Crossrail run a train in public service next year – sources are hinting that the cross-London route won’t open until possibly May 2020. However, I’d personally predict we’ll see a service this time next year, even if it’s a limited one. Nothing material has changed (at least, not that’s been put in the public domain), since the Autumn 2019 date was announced, so a significant further delay would come as a surprise.
(Updated 11th December to clarify ATO not being the primary reason for the Thameslink delay)