In The Dangerous Sound of Silence we looked at the reasons why Crossrail’s delay was not spotted earlier. Possibly more importantly, we looked at why that delay was not fully communicated to those who needed to know about it once it became probable that the project would be significantly late and over-budget. What really matters now is to get Crossrail safely up and running as soon as realistically possible.

Enter Mark Wild

To that end, in November 2018, Mark Wild was appointed CEO of Crossrail. He gave a full and frank account of the task before him in public session to the Programmes and Investment Committee on 11th December 2018. He had also previously written to the chair of the GLA Transport Committee in a letter that has now been published prior to their meeting on 9th January 2019. Based on the information in both his briefing and his letter, we look at how the project is being put back on track.

Until recently Mark Wild was the Managing Director of London Underground. It is generally believed that the previous incumbent of this job was encouraged to move on by Mike Brown, the Commissioner, in order to make the post available to Mark Wild who has a wide range of experience in commissioning and running railways. If true, with extremely challenging projects such as Sub-Surface Resignalling on the criticial-to-deliver list, his hard-nosed prioritisation will probably turn out to be one of the best things Mike Brown has done during his role as Commissioner.

New broom needed

It must have been clear to all by Autumn 2018 that Crossrail really needed someone at the top who was fully dedicated to getting the new railway up and running (as opposed to constructed). Sir Terry Morgan would be a hard act to follow and he did formerly have a lot of experience as the man in charge of Tube Lines, which was a Public Private Partnership (PPP) that for a while looked after the infrastructure of the Jubilee, Northern and Piccadilly lines prior to becoming chairman of Crossrail. But, in that role, he didn’t have direct railway operating experience. In retrospect, it may be that Terry Morgan as Chairman and Andrew Wolstenholme as CEO was a successful double-act that was compromised by Wolstenholme’s departure.

With people losing confidence in Terry Morgan, rightly or wrongly, it was felt a new person was needed at the top. This was not least because it was perceived that it needed someone capable of radical and fresh thinking to look at what was happening. It could be argued that Mark Wild didn’t fully qualify as he was already on the Crossrail Board, although Wild says that he wasn’t involved with management decisions and didn’t know fully what was going on.

As with the Enron collapse in America in 2001, one wonders why the non-executive directors thought they were on the board if it wasn’t, at least in part, to know what was actually going on and prevent happening precisely what did happen. In the case of Crossrail, it seems to be generally believed, not least by the people appointed, that they were there to provide expertise and assistance to the project rather than scrutinise what others had done. In other words, they were there to feed information and advise to the executive board rather than to sanity check what the executive board told them.

Mark Wild formally took over as Chief Executive at Crossrail on 19th November 2018. If his previous association with the project is at least accepted, it is hard to think of a better, more-suited, more-qualified person to take over. Running London Underground isn’t easy but there are people around who can face the challenge and one can argue that the difficult decisions as to how to make London Underground ‘fit for the future’ have already been made. As part of the strategy to get Mark Wild into the Crossrail role, Nigel Holness, former Director of Network Operations, took over the role of MD of London Underground. In doing so he has delayed departure to a new job in Australia.

A new open (transparent) Crossrail?

Not surprisingly, in light of the allegations levelled against Sir Terry Morgan and the previous regime, Mark vowed that under his management he intends that Crossrail will be entirely open and transparent, subject to restrictions due to commercial sensitivity. He promised always to attend the Programmes and Investment Committee meetings to give an update (although that latter was swiftly amended to ‘or send my deputy’).

As Mark had only been in his role for around a month when he reported to the Programmes and Investment Committee on 11th December, he was still trying to assess what was needed and felt unable to commit to much. It was clear that he had already encountered various frustrations, one of which was the lack of an internal communications team – part of the winding down process. This meant he couldn’t even easily fully communicate with the organisation he was in charge of. He also commented on the near non-existence of the external communication team as a result of the same process, which would also need to be revived. We presume this means that at some stage the Crossrail website will come to life again.

A new management team

Something that was becoming abundantly clear was that Crossrail had been winding down its management team too early. Of course, this was done in the expectation that the job would be largely over by December 2018. In fact it was becoming obvious from Wild’s initial thoughts, that if ever there was a time that Crossrail needed a strong management team it was now.

Wild went on to praise the workers on the project. Some had worked for much of their working lives on the project and wanted to finish the job. He tried to allay Committee members’ concerns about morale and assured them that the workers continued to be positive about the project.

As the former MD of London Underground, Wild said he wouldn’t poach staff from existing LUL projects because he knew how disruptive and unfair that was. This too was slightly modified in the course of his verbal report, in this case to ‘without specific agreement to do so’. That clearly restricted the pool from which he could recruit but he assured members he had people lined up and was due to announce his senior team in the next few days – Friday 14th December to be precise.

It appears that no announcement of the new management team was made, certainly we cannot find one. However, the Crossrail website has clearly had some new life injected into it as a new Leadership Team page has appeared with links to details of individual members.

Where are we now?

Readers will not be surprised to learn that the signalling was giving problems. There was concern about ensuring trains successfully transitioned between one signalling system and another. Meanwhile, if we understand the situation correctly, there are significant problems just getting the signalling equipment on the trains to talk successfully to both the trackside-based communications equipment and the trains’ computer systems.

Transitioning

In his letter to Caroline Pidgeon, Chair of the London Assembly Transport Committee, Wild states:

The tests included: rolling stock transitioning between the different signalling systems at Westbourne Park (successfully going Eastbound into the central tunnels)

This is the sort of comment that must not be taken at face value. The clear implication is that the test was unsuccessful going westbound. It is a shame the letter isn’t truly more open and explains what happened when the train went westbound. Presumably the train must have returned westbound to get back to the depot. Did the testing reveal problems or has this simply not yet been tested?

Transitioning eastbound is easier than transitioning westbound at Westbourne Park. The more difficult part of transitioning is to successfully communicate with the signalling system the trains is transitioning to. It is somewhat easier to terminate communication with the signalling system the train is leaving.

To aggravate the problem transitioning westbound at Westbourne Park, the signalling system being transitioned to is TPWS+. This is a system based on the standard TPWS (Train Protection and Warning System) used on most of Network Rail, but it includes additional safety protection to bring it on par with ETCS when it comes to safety. One feature of TPWS (and TPWS+) is the intermittent (rather than continuous) nature of the system. This relies on fixed signals protecting discrete sections of track rather than utilising a ‘moving block’ to provide protection between trains. Stepping up from ‘fixed block’ to moving block should be relatively easy to achieve. Think of it, metaphorically, as lava dropping from the top of a lava lamp. The ‘moving block’ starts in a fixed position before moving with the train (protecting it in the rear). Doing it the other way around is probably very hard.

In the central section the signalling system is a bespoke version of Siemens’ Trainguard system, which has been updated with Crossrail in mind. So it should have been upgraded with facilities for transitioning to and from TPWS+ as part of the update. In the other direction (westbound) you are moving into TPWS+ which is an industry standard that one has to adhere to. You can’t simply ‘tweak’ TPWS+ so it works with Trainguard. Even if you could, that would require a derogation that needs ORR (Office of Road and Rail) approval. Such approval is not given lightly.

Part of the reason that there is the problem described above is that this was never the plan. The signalling system west of Paddington all the way to Heathrow was supposed to be ETCS (European Train Control System). It changed to TPWS+ because of the problems of introducing ETCS on the GWR in the necessary time frame. Somewhat ominously but understandably, it took ORR two years to approve the change because they needed to be satisfied that TPWS+ was as safe as ETCS on the lines affected.

More ominous still, the time allocated to the system integration phase of Crossrail was decided long before TPWS+ was substituted for ETCS and, to the best of our knowledge, the time allocated was not revised upward to take into account the known (or at least suspected) extra complexity. This does seem a puzzling oversight and one wonders, with the benefit of hindsight, whether it would have been better to stick to the original plan and use ETCS throughout on the Paddington – Heathrow section and accept any delays this may produce, rather than exert so much effort getting TPWS+ installed and working with Trainguard.

To make matters even worse, the change of plan now means you also have another transitioning area that originally was not going to exist. This will need to occur north-east of the Heathrow tunnels on the Airport Branch. In hindsight it is clear how much Crossrail were relying on Network Rail being able to introduce ETCS on the Great Western Main Line (GWML) in good time prior to Crossrail opening – as was the plan. It also shows how apparently unrelated issues such as problems on Network Rail’s GWML modernisation project (some of which were in Wales) have affected Crossrail. As has often been pointed out – Crossrail is good for the rest of the country. Unfortunately sometimes the rest of the country can be bad for Crossrail.

The Train Control Management System

A surprise in Mark’s letter is the reason he gives for the Crossrail Class 345 trains not having yet reached Heathrow in service. He states:

The delay in operating the new trains to Heathrow is because of issues in developing the Train Control Management System (TCMS) compatible with the European Train Control System (ETCS) that operates in the tunnels to the airport.

This again raises far more questions than it answers. For starters, to the uninitiated, it is far from obvious what TCMS is. Note that ETCS is the European Train Control System. So you might think that TCMS is to do with the signalling system. Earlier in Mark Wild’s letter there is a reference to the on board Train Control Management Systems (TCMS). One would certainly be forgiven for thinking this is the on-board element of the ETCS signalling system.

In fact TCMS, as far as Bombardier is concerned, covers just about every aspect of on-board train control from heating & ventilation to passenger display screens and door opening. Whereas some aspects of it, such as giving passengers the ability to send and receive emails, are not safety critical, other parts of it are. So it has to work – at least for the safety-critical bits of it.

A brochure from Bombardier describes their offering of TCMS and shows how it will become the bedrock of future builds of trains supplied by them. It is clearly a system that once it is up and running should provide great benefits but, development has been fraught with difficulty – as anyone who has followed the saga of the Class 710 on the Gospel Oak – Barking line will be well aware.

It is therefore far from obvious why the TCMS is delaying introduction of the Class 345 trains in the Heathrow tunnel, whereas the trains can run between Paddington and Hayes & Harlington and are also reported to work with ETCS on the test track. The reason seems to be down to the fact that the TCMS is responsible for filtering out unwanted frequencies as part of the on-board signalling function. It needs to do this in the Heathrow tunnels to avoid unwanted signals from the GW-ATP signalling system installed in there. Older readers brought up on superheterodyne radios need to be aware that this filtering functionality is now generally done with software, so the issue has become a software problem rather than a hardware problem.

No timelines

Currently it would be impossible to say how long it would take before Crossrail would be in a position where trains could successfully operate on routes with multiple signalling systems. Let’s not forget that they need this just to get the trains from the depot at Old Oak Common to Paddington. This was not helped by there not being a reliable signalling system on the trains (something that is not Crossrail’s direct responsibility as the Bombardier trains were procured by TfL and not Crossrail), which made it difficult to carry out meaningful testing at an early stage of the testing programme. Indeed, after the fourteenth attempt in Autumn 2018 they had only just succeeded in getting the software on the trains working sufficiently well to conduct proper tests. This goes a long way to back up Terry Morgan’s claim that a lot of the problems were down to not having a single working train with which to test the signalling, despite some oddly aggressive briefing from the Mayor and others to the contrary.

The problem with software on the trains has highlighted various points not described in detail by Mark Wild:

  • Even stage 1 of Crossrail implementation Shenfield – Liverpool Street (which included a service consisting of only Class 345 at weekends) hasn’t been fully implemented. In fairness, probably all that is stopping this happening is the need to loan 7-car trains to run the Paddington – Hayes & Harlington shuttle service.
  • Given that stage 2 (Paddington High Level to Heathrow) has only been implemented by using existing class 360 trains and not the new 9-car class 345 trains, and with no class 345s yet able to use the Heathrow tunnels because of issues with legacy signalling still present there, Crossrail has yet to truly achieve any of its staged openings.
  • If it is only now that software is working in basic mode, then in reality the testing programme is about a year late. This should have been happening soon after the transformer at Pudding Mill Lane was due to go live.
  • Arguably most damaging to timescales is just how interaction between various parts of the project does not seem to have been taken fully into account and timescales have not been updated to allow for changed circumstances. If the trains can’t reliably be used for testing then other more complex signalling functions and the power cannot be fully tested – let alone the interaction between trains and stations, such as stopping precision and platform edge doors.

Those responsible for the train delays may see it as something that can be caught up on later, but the reality is that it impinges on other testing factors so the full significance of the delays are not necessarily taken into account. It should not be underplayed just how much the trains matter.

Stations

Wild also confirmed that various stations were problematic. Surprisingly, Woolwich is progressing well and, by now, should already be handed over to Crossrail. The two real problem stations, less surprisingly, are Bond St and Paddington. One of the things he is keen to do is liaise with MTR Crossrail, the future operators, and establish just which bits of the stations need to be complete to enable them to be opened.

Wild didn’t explain in detail what could be done to open a station without completing all the work, but it isn’t hard to guess. For double-ended stations it may be possible, subject to safety approval, to initially open just one entrance. It is also possible that not all the escalators need to be in use on day one. Temporary arrangements could also be agreed with staff for alternative mess-room arrangements until such time as the proper facilities are ready.

Whilst admitting that there were lots of issues still to be sorted out, Wild saw an initial objective was to get the contractors out of the stations. As long as they were still present it became next to impossible to do meaningful testing within the station. Furthermore, other issues such as having working fire alarms, could not be relied upon so long as the contractors were present.

It was not explained how the issue of completing the stations would be addressed. Would the original contractor be expected to finish the job later or would the contract be terminated and fresh contracts issued to ‘finish the job’ – presumably during engineering hours?

Finishing the tunnels

It came as a bit of a shock to learn that there was still work to be done in the tunnels. True, this appeared to only be retrieval of redundant kit, but it was a surprise this wasn’t already done. Wild believed this could be left as something non-urgent, but at the same time he emphasised that it is something that really ought to be sorted out eventually. There seems no doubt that Elizabeth line will open before it is fully ‘finished’ with work still to be done in many areas.

Intriguingly, (and not mentioned by Wild himself) LR sources have suggested that a lot of the problems with the trains not talking to the track-based equipment are down to problems with the ‘leaky feeder’ cable. Apparently, it is not leaky enough. This would explain redundant kit being present in a new tunnel if the already-installed leaky feeders were supplemented with a replacement and the original cable left in place. It would also help explain Siemens placing adverts in late November 2018 for GSM-R testers – long after that fitting of cables was due to be complete and operational.

The outer stations

Something we are determined not to lose sight of at LR is the fate of the existing outer stations on Crossrail which have been problematic. This appears to be partly due to budgetary constraints and partly because the work is overseen by Network Rail, so Crossrail has little to do with it. Indeed this could be seen as an area where more direct involvement by a TfL insider would achieve better results than formerly where Crossrail and Network Rail both reported to TfL but didn’t really seem to communicate directly with each other on this issue.

For once, it would seem to be better just to repeat what was written by Wild in his letter.

Network Rail’s programme of works for the western stations will see enabling works continuing at four stations (Ealing Broadway, West Ealing, Southall and Acton Main Line), over the Christmas period. These are expected to complete by early February, allowing the main works contractors to mobilise to site in January. Ilford and Romford station improvements, also being carried out by Network Rail, are fully funded. At Ilford, enabling works are also due to be undertaken this Christmas to provide a vital increase in capacity and step-free access, with completion likely to be in early/mid-2020. These improvements are not critical to the opening of the Elizabeth line and Ilford station will continue to operate as this work is undertaken.

Maybe next year, maybe not

On the all important subject of when the Elizabeth line will open, Wild would not be drawn at his briefing given on the 11th December. He explained that he didn’t want to give a date he couldn’t commit to and he couldn’t at this stage even give a spread of dates that was known to be achievable. This was partly down to the unknown factor of signalling, but also down to the fact that a lot more investigative work needs to take place before he feels in a position to assess the possibilities. He expressed grave doubt that it would open in 2019 but didn’t exclude the possibility.

System Testing

The problem, it appears, is the complexity of systems testing. In the words of Mark Wild himself, the new Crossrail stations are the most technically advanced stations on the planet. People talk about the digital city as something in the future but, when it comes to Crossrail stations, it is the here and now.

This is not a new revelation. In Crossrail lectures, most notably (but not only) by Terry Morgan, this aspect has been emphasised with the presenter seeking to justify how long this phase was. With the benefit of hindsight, perhaps two years rather than one year should have been allowed. There is the obvious danger of allocating time required for integration testing based on experience of similar recent projects but, by the time the latest project comes to fruition, things have got even more complex.

What is notable here is that in 2013 Terry Morgan gave a prestigious lecture to a packed Institute of Engineering and Technology (IET). He highlighted the need for systems testing and explained the complexity of it all. There were many questions from a highly qualified audience but no-one in the audience sought to ask a question on whether the year allowed for all the system testing was really enough.

To summarise, Wild said that system testing is “costly in terms of time but not in terms of money”. Obviously delays also have a financial cost because they mean that the opening – and the revenue it brings – is delayed, but the message here seems to be that it doesn’t make any kind of sense to try and short-circuit the testing phase – even when short of money.

Looking forward

The program for the future seems to be

14th January 2019: Full dynamic testing of trains would begin. Testing would continue 5 days a week. It cannot be more than that because there is maintenance to do and staff cannot work extra hours over prolonged periods. This is expected to continue for six to twelve months.

End of January 2019: Mark Wild would be in a position to outline priorities and assess the likelihood of various opening dates.

Early February 2019: Certain members of the TfL board and scrutiny committees will get a private briefing on the current situation to be followed by a detailed scrutiny by board members.

Sometime in January – March 2019: Farringdon station gets handed over.

Late 2019: The stations with a larger amount of outstanding work are handed over.

From the above it seems like Sunday 15 December 2019 is the earliest possible realistic date for Crossrail to open in central London with a date in 2020 looking more likely.

Issues not mentioned

No light was shed on a couple of consequential issues as they were not mentioned.

In the Summer of 2019 Network Rail is scheduled to reconfigure platforms 16-18 at Liverpool Street. This is to provide two further 240m platforms to replace the three short platforms and (together with platform 15) will provide for three full-length platforms for future suburban services.

This platform work at Liverpool St has been planned for years and a lot of follow-on work, such as the reconfiguring of Bow Junction, is dependent on this. Nothing was said about how this would be accommodated or whether the reconfiguration would be delayed. Our latest information is these works may have to continue as scheduled, otherwise other Network Rail plans are dislocated, so there may be new operating pressures.

Even if TfL had the final say, they would still have a difficult decision to make. TfL must be keen to introduce longer class 345 Crossrail trains on TfL Rail, so as to handle passenger numbers better, but they cannot really afford to have Liverpool Street out of action as a terminus. We understand that temporary use of Liverpool Street low level (i.e. Crossrail) is being considered for the duration of the works, with empty trains running on west in order to reverse. The expectation is that this will require trains to be turned at Fisher Street (Holborn) crossover.

Reversing at Fisher Street should be possible, even during the peak period, if the tunnel is handed over solely to run the TfL Rail service to Liverpool Street. This would mean that other testing involving train movements could not take place and, presumably, risk delaying the project even further. Alternatively it might be possible to send the out-of-service trains all the way to Westbourne Park to reverse.

The other issue is what to do about services to Reading in December 2019. TfL’s revised business plan states quite categorically that TfL want to run trains from Paddington to Reading in order to capture the revenue. The problem here is lack of platform availability at Paddington (high level) in peak hours – and off-peak is not much easier. The low level (Crossrail) platforms might be available by then but, presumably, this would entail being able to use the tracks between Westbourne Park and Paddington as two bi-directional lines.

LR understands that talks are underway with the DfT at a high level as to how to provide a Crossrail service from Paddington to Reading in December 2019. Assuming this happens, we can then envisage that there maybe further pressure to run these trains as through Elizabeth line trains as soon as possible to free up platforms at Paddington.

Late but in a better place

2019 is looking like it will be a bumper year for advancement of rail schemes in London – as schemes that should have been completed in 2018 are finally opened. Unfortunately it looks like we cannot say with any confidence whether Crossrail will be on that list. What we can say is that at least the right person for this phase of the job appears to have been appointed. A sense of realism has been introduced and there does seem to be the making of a plan to ensure that Crossrail is opened as soon as practically possible.

Thanks to Jonathan Roberts for supplying his detailed notes taken at the Programmes and Investment Committee meeting and for assistance with this article and to ngh for detailed information on various issues. Thanks also to ngh for considerable technical input clarifying and correcting an original draft.

Written by Pedantic of Purley