Last month, a significant achievement took place on the Crossrail project, which is far more important than the headlines make it seem to be.
At a basic level, the Office of Rail Regulation approved the use of the new Class 345 trains that will be used on the Elizabeth line to carry passengers into the Heathrow tunnels.
The practical implication being that TfL Rail will in the next few weeks be able to run from Paddington to Heathrow direct, as they were supposed to start doing back in May 2018.
The delay has been caused by the bane and saviour of modern railways, the signalling system.
Modern trains are more like computers on wheels, which is great for running a highly intensive service and improving maintenance reliability but does also mean that an awful lot of what happens in a train is now controlled by a central brain – the Train Control Management System (TCMS).
This in turn talks to the signalling systems to direct where the train should go – with a human up front of course — but getting the two to work together is sometimes harder than expected. Crossrail even looked at adding the now obsolete GW-ATP signalling system used in the Heathrow tunnels to its trains to get around the problem but concluded it was not viable due to the age of the technology being used.
A lot of the problems are down to getting the modern train controls to talk to the different and overlapping radio signals used by railway signalling systems.
When you have lots of legacy signalling systems on the National Rail network, and the GW-ATP signalling system, one of the roles of the software is to filter away the noise to get just the information it needs. In the very old days had such a thing existed, it would have meant physically changing crystals in the radio kit. Today the hard work is done in software, so it’s a lot easier to update the software and test again to see if that works.
On top of that is all the testing to try and break the system, and ensure it doesn’t break, and that redundancy back-ups kick in as expected. And every time they update the software, they have to check that the tests they previously passed haven’t now failed.
It’s just taken a lot longer than expected to get the software to the point where it is safe for passenger use.
As it happened, that milestone was just about to be achieved as the lockdown took place, but being software, a remote upgrade was possible, and that was the final stage needed to get the regulator’s approval.
To a degree, this is a rather dry academic regulatory issue – nice to have, and it does move the Crossrail project on an extra step.
However, it’s a much bigger issue than it sounds. This is the next generation of train control systems, and the only place it’s been used before in the UK was for testing the technology on a line through Wales, the Cambrian line.
To deploy this new system on a part of the rail network that’s as complicated as the Heathrow to Paddington line is quite a substantial achievement.
Crossrail’s Technical Director, Colin Brown told me a couple of weeks back that due to the sudden lockdown, they hadn’t expected to get the regulatory approval, but they had managed to complete enough testing and validation for the approval for the latest software build, and now the trains are finally authorised for Heathrow Airport.
When the line is fully operational, a total of six Elizabeth line trains per hour will serve Heathrow Airport. Four will go to Terminals 2, 3 and 4 and two will go to Terminals 2, 3 and 5. This will also increase services to Ealing Broadway, Southall and Hayes & Harlington.
For all the headaches, the new signalling and train control systems offer a lot of benefits, from very tight adherence to the timetable, the ability to run the trains closer together to get the intense service being planned, operating the in-train air cooling and information screens, to controlling platform doors in stations.
To my mind though, the neatest trick is one that only the drivers will ever see.
Around half of westbound trains will terminate at Paddington, but to head back eastwards, once all the passengers are off, they carry on westwards to Westbourne Park, then return back to Paddington on the eastbound line.
Normally that means the train driver would drive to Westbourne Park, stop, walk through the train to the other end, then drive back. But with “auto-reverse”, as soon as the train leaves Paddington, the driver switches to automatic and starts walking through the train to the other end. By the time the train arrives at Westbourne Park sidings, the driver will be sitting in the drivers cab at the other end of the train ready to head back into Central London.
Now that’s really quite clever.
As you say, the auto-reverse is really quite clever.
Several services in the UK, do a normal reverse in various stations.
Could automation be developed, so that these services were speeding up?
You can also envisage shuttle services like the Greenford Branch, where the driver sits in either cab and controls the train with a sophisticated remote control.
The only question for me is whether the auto reverse functionality would still be an option during rush hour as the trains are rammed at that time and would take an age to walk through.
You never take a train into a siding or depot with passengers on board, except in the rarest of emergencies, and then the speed of turn around is the least of your problems.
I was on a GWR service to Bedwyn where the driver simply forgot to stop at a station. The train then stopped in between stations, where I assume a rather embarrassed driver discussed the situation with higher powers.
After a few minutes, we continued on with the option to get off at the next stop and wait for the return trip (20+ minutes) or the last stop (Bedwyn) and wait there instead (5 minutes or so).
This has happened in other unusual (winter weather) events as well. And in every case the driver had to clear the cars at the last station. Drive into the siding, walk back and then wait for signals to reboard passengers. It’s a muddle, so “auto parking” would be helpful for time
Which means those five rather tired Desiro’s units are surplus to requirements and join the increasing number of trains looking for a use elsewhere!
Had things gone to plan they would have gone much earlier.. I suppose electrifying branches to Windsor and Eton and Greenford would give them a new home .
Given the time it’s going to take to deliver Crossrail 2 then perhaps extending the Heathrow branch to link with SWR might be a quicker option !
Seeing as a train with 15% of capacity is now the max allowable under social distancing they need to be redeployed to the GE perhaps to ensure services are run with max capacity.
Those (thousands!) of us who live west of Reading and regularly use LHR really look forward to being able to change from GWR to Crossrail at Reading and then travel direct to LHR terminals instead of using the railair bus service with constant M4 traffic disruption.
This won’t change that – you’re confusing this with the unbuilt plans for a rail link to Terminal 5 from Reading.
Go to wrest Drayton and get service bus to airport. Much quicker than railair, for now.
Any update on the westbound Reading/LHR loop? Last I heard funding had gone away
I’d say that this approval is actually more importat than you first make it out to be – ‘dry academic regulatory issue – nice to have’
But then you say it’s a substantial achievement!
Good, clear description of the difference between ‘lights on sticks’ signalling and Digital Railway. But isn’t it also being used on Thameslink ‘core’ from London Bridge to St Pancras?
Not quite the same thing.
Tired Desiros to go to East Midlands Railway for driver training to Corby pending the release of trains from Greater Anglia?
This is indeed a moment to savour in the history of the provision of Crossrail 1.
You and others have covered with vigour the seemingly intractable mess that the engineers faced in joining the three at least different train control systems governing these tracks.
However, not wishing to pee on anyone’s chips, it appears that the auto reverse function means that for the sake of expedience, no-one is at the helm of the train to slam the brakes on if something is obstructing the line. Or are there cameras at either end linked to a human in a control room or coupled with radar/lidar to an HSE-rated computer?
Thameslink Core is ETCS (Not sure if level 2.5 or 3). Heathrow was also using ETCS. Didn’t the Heathrow link switch between AWS and ETCS? And of course, Thameslink switches between visual signalling with basic ATP, and ETCS. So what’s new. Well, for some reason, Crossrail has CBTC not ETCS. So the 345s have to cope with all four systems.
Nonetheless, a great and important achievement. Can now move to driver training, and a “soft launch” into passenger service.
West Ealing is a JUNCTION for Greeford and High Wycombe but not enough trains stop there.
Is there a reason why you felt the need to type junction in capital letters?
You’ll have to wait until crossrail fully ramps up the service and whether Chiltern et agreement to run into OOC from the Acton-Northolt Line