It is said that one of the problems of Channel Tunnel construction was that the British thought they were building a tunnel that just happened to contain a railway. The French, thinking more realistically, saw themselves as building a railway which just happened to need a tunnel. Luckily, this was a lesson that Crossrail took on board at a relatively early stage.
“Both Operations and Maintenance have been involved in the project almost since day one,” Explained Crossrail Systems Director Siv Bhamra to LR a little while back. “But that wasn’t part of the original plan.”
Andrew Wolstenholme, Crossrail’s Chief Executive since September 2011, later confirms that this was the case.
“When I arrived at Crossrail, I was told the post of Operations Director did not yet exist. That had to change. We lacked balance between the huge engineering and civils focus and the need to start planning at an early stage for the railway to come. That led to the appointment of Howard [Smith], who brought to Crossrail the knowledge and expertise that was missing.”
Andrew Wolstenholme, Crossrail Chief Executive
Howard Smith’s appointment a year later as Operations Director was a sound one. A hardened veteran of the launch of the London Overground, he is intimately familiar with the operational challenges of launching – and running – a new railway in London. Culturally, however, from an early stage it also meant the change in thinking that Bhamra described.
“There’s a real danger on big building projects like this,” warns Chris Sexton, Crossrail Technical Director, in discussing that culture “of making decisions that make sense from an engineering perspective but fail to account for how the railway will actually need to work.
“I hope we’ve avoided that. We’ve had the voices there to help us make decisions where the eventual operational benefits outweigh engineering judgement.”
It is the fact that the operational element has been embedded in Crossrail for some time which means that even as Crossrail are announcing the end of tunnelling we can already start to build up a picture of the railway those tunnels will contain. We will turn to this shortly, but first it is worth a final glance back at how the tunnelling process ended.
Here, what has been quite remarkable is that the tunnelling phase has proceeded relatively smoothly. There have been few changes of plan during the course of construction – most notably the extraction (or rather the lack of it) of the Tunnel Boring Machines (TBMs) from Farringdon – but this is the only major deviation of note.
That is not to say that, in engineering terms, there haven’t been other hitches. The TBMs stalled for a while at Whitechapel, for example. This could be regarded, as much as anything, as a consequence of the TBMs performing much better than expected. Combined with a confidence that they would make up time later though the work schedule could be re-arranged. Indeed the only major delay for tunnelling reasons was due to one of the TBMs hitting an unmarked test concrete pile in the Canary Wharf area. The delay was caused by the need to search for documentation to work out what it was and, most importantly, establish that there were no more of them.
As a final note here it is worth pointing out that some Sprayed Concrete Lining (SCL) work, primarily at Liverpool Street and Whitechapel, is still underway. Ultimately though the running tunnels are now complete and ready to receive the railway they will contain.
Building the railway
As was highlighted at the beginning of this article, it would be a mistake to assume that (just like plans for its operation) the construction of the railway itself hadn’t, in fact, already begun. We have covered before the sometimes-forgotten surface elements of Crossrail, and this work has presented plenty of opportunities for the laying of track and other activities. This is especially true on the surface section to Abbey Wood, where the scale of the Crossrail project has required the laying of over a mile of new double track between the Plumstead portal and Abbey Wood station.
This work has been complicated by the need to slew over the existing Network Rail double track railway. An excellent article in New Civil Engineer recently described some of the challenges involved. It is clearly much more complex than a visit to the area would lead you to believe. Much of that difficulty was down to unfavourable geology – alluvium with peat layers as well as potentially unstable embankments. It is highly telling though that despite the scale of work on this mile long worksite, which would normally warrant attention as a major piece of railway work in its own right, it is actually seen as a relatively minor element of the work involved in building Crossrail.
Elsewhere at the western end, electrification masts are starting to appear in great numbers and bridges have been raised or rebuilt. Two major tasks which have not had that much publicity are the rebuilding of Airport Junction and the work at Acton Yard, where a new dive-under has been constructed to avoid conflict with freight trains.
It is not just in terms of construction, of course, that Crossrail’s operational work has begun. As of 31st May Howard Smith has technically been responsible for a running railway as the Crossrail concessionaire, MTR, officially took over all services between Liverpool Street and Shenfield on that date.
That those services are not currently branded as Crossrail is deliberate. Back in November we covered the fact that these services would be launched as “TfL Rail,” carefully avoiding the use of the Crossrail name until such time as new rolling stock and services were actually available.
Smith himself has since highlighted another reason for this decision. At the time the Crossrail concession, which follows a similar model to that used on the Overground, was let, neither he nor TfL knew who would get it. As a result ensuring an early takeover of Shenfield services (ideally under different branding) was seen as an important way for both parties to begin to build a long term relationship – so that when the “real” Crossrail came along there would be no disputes or misunderstandings as to working practice. Ultimately, of course, the contract was awarded to MTR – familiar to TfL through their status as co-owners of London Overground operator LOROL. Nonetheless, the principle was sound.
The Philosophy of Crossrail
Just what form that “real” Crossrail will take when it arrives is a far more philosophical question than one might think. Is it a mainline railway? A glorified tube line? A suburban metro system (akin to London Overground) or something else entirely? It has been cast in all of those images, by different people, at various stages of its planning, but until now determining just what Crossrail itself thinks it will be has been far trickier.
Luckily, as we move towards the operational railway, that is becoming easier to see. The result so far seems to be a railway that lies somewhere between mainline and tube line, as the operational approach Smith and his team are taking to a number of the issues below seems to show.
Few outside the railway industry likely realise just how far in advance timetable planning is carried out. Whilst final details and tweaks can be arranged in the months before implementation, the basic frequency and service needs to be agreed years in advance. This especially applies to the peak service of a new railway such as Crossrail, as this will largely determine the total number of trains required to be ordered.
As a result, in the simplest of terms, in order for Crossrail to operate trains through the central section and on to Network Rail tracks in 2019 it is actually necessary to agree on the service framework now.
To the second or to the half minute
That leads to the first of many interesting operating philosophy conflicts. The national railway in Britain has, almost since the start of proper train regulation, operated working timetables to the nearest half-minute. TfL on its automatically operated Underground lines has a working timetable that displays times to the nearest quarter minute, but which in many ways operates to the second. Each platform has a target maximum dwell time measured in seconds for example.
The issue of only operating to the nearest half minute may not be too much of an issue from the outset for Crossrail, but if the line ever goes above 24tph then this would appear to be a major area that Network Rail would need to address.
C-DAS to the rescue?
Driver Advisory Systems (DAS) are devices that advise the train driver of the best course of action to take as regards accelerating or braking, whilst leaving it to the driver to make the final decision. As such a DAS is not regarded as safety critical. An analogy would be a car driver with a sat-nav. The sat-nav can advise but cannot authorise a driver to go through a No Entry sign – the responsibility remains with the person behind the wheel. On the railways DAS stand in contrast to Automatic Train Operation (ATO) systems where the driver is relieved of the responsibility for many of the actions that the ATO system takes.
DAS are nothing new. In a very crude form the Southern Railway in the 1930s put white diamonds on posts next to the track to indicate to drivers that they should cease motoring and coast from that point onward. The modern electronic form of DAS is now quite common on the trains of many Train Operating Companies.
What is new, however, is a more intelligent system called C-DAS (for Connected DAS) which receives data from the outside world to give more intelligent advice to the driver. This could, for example, enable a driver to arrive at a junction at the optimum moment or take into account a late running train ahead to avoid applying power unnecessarily. If the technology is successful it is hoped that this would do a lot to overcome the issue of trying to mix and match Network Rail and TfL working. Trains on Network Rail track would approach the Crossrail portals obeying normal signals but at the same time the C-DAS system would be advising on speed so that they entered the Crossrail central section at precisely the right time.
Signalling and the rule book
From the passenger point of view Crossrail may be seen as trying to be all things to all men – a conventional railway outside the centre of London and a high frequency underground service in the centre. From the train driver’s perspective things will be somewhat different and, whatever the appearances are to the public, it is important that the drivers see this as one railway with one set of signalling and one rule book they have to follow. It is true that on the Bakerloo Line and the District Line that Underground drivers effectively have to change from an Underground signalling system and rule book to a Network Rail one (and, of course, Chiltern drivers have to do this the other way round) but this is generally regarded as highly unsatisfactory and best avoided if possible.
Despite initial appearances, from the drivers and signallers perspective, Crossrail is very much a railway that is part of National Rail. It will be run using the Network Rail rulebook which will no doubt have to be supplemented to cater for such issues as Platform Edge Doors and evacuation procedures. The signalling will be supervised from either Didcot (west of Paddington) or the new Network Rail Regional Operating Centre at Romford and both of these locations will have Network Rail signallers. The drivers radios will operative using GSM-R (the National Railway system), even in the tunnels, and not on Tetra network which is the radio network London Underground currently uses.
Stick to the timetable or run an even interval service?
Closely associated with network supervision is network regulation of trains. This brings up the issue of whether one sticks to the timetable. It is here that we are are probably going to see the greatest problems with Crossrail. On Underground lines with a frequent service the objective is largely to maintain even intervals rather than adhere too strictly to the timetable. This can be a problem if sharing tracks on Network Rail where there are other services, but in practice this hasn’t been too a great issue with an absolutely maximum of 4tph of other services to consider.
This need to stick to the timetable will be a bit of an issue for the line out to Shenfield, with the need to accommodate some freight crossing the line on the level and a lot of empty coaching stock movements to Ilford depot. It will be much more of an issue on the Great Western Line out towards Reading where slots are scarce and the railway has to accommodate fast, semi-fast and stopping services.
A further issue is that it would appear to be inconceivable that Crossrail trains, at least those west of Paddington, will not be shown on the National Rail Timetable. It would also be somewhat strange if not all the journey was shown. So it seems that Crossrail, rather like London Overground, will probably put great emphasis on running trains to the timetable and limit out of course working in order to plug any gaps in the service. London Overground themselves are currently finding out how hard it can sometimes be to balance these two competing approaches on their newly-acquired West Anglia services. No doubt they will learn their lessons swiftly, and it should be hoped that Crossrail are watching efforts there closely.
How frequent an off-peak service?
How Crossrail will approach off-peak services is another operational issue that remains of great interest. Indeed this is a subject we have looked at in the past. As a rule of thumb, London Underground nowadays tend to run an off-peak service, even on Sundays of at least 75% of the peak service with the philosophy of “turn up and go” prominent. They have also tried, as hard as is possible, to have a very simple service pattern. There are very few short workings and the interval between trains is generally consistent.
West of Paddington, Crossrail is largely constricted by having what slots are available and these, in principle, were established before the Crossrail Act was passed. On the Shenfield branch it was agreed at the time of the Crossrail Act passing that they could have 6 trains per hour (tph) off-peak to Shenfield which is basically the same as shown now in the TfL Rail timetable.
The problem is, to maintain an even interval service with 6tph to Shenfield will also require 6tph on the Abbey Wood branch. This then gives a combined total of only 12tph through the central section. This would be woefully inadequate by London Underground standards – in essence a train every five minutes – and would even be considerably worse than Thameslink from December 2018 onwards.
It is encouraging to learn from Smith now that, despite all the documentation stating otherwise, the draft agreement with Network Rail for the future frequency of Crossrail off-peak services to Shenfield is actually 8tph. This is a critical improvement because this means 16tph off-peak through the central section which, though still not quite up to London Underground standards, is considerably better than 12tph. Even with this improvement, Canary Wharf will only have an off-peak Crossrail train every 7½ minutes and one wonders whether this will be considered acceptable by those working there.
We must issue a word of slight caution here as regard to off-peak services. Whilst Crossrail and Network Rail are working on the planning assumption that there will be 8tph off-peak to Shenfield, we presume at even intervals, the agreement with ORR (Office of Road and Rail – formerly Office of Rail Regulation) is still only for 6tph so, theoretically, at some point in the future, Crossrail could find themselves limited to 6tph on that branch. However, once 8tph is established, it is very hard to conceive of any circumstances in which this could happen. Just to establish certainty and pre-empt any danger of that happening, at some point a revised Track Access Option application will go to the ORR to ensure that the 8tph cannot be taken away.
Not stopping at …
TfL’s philosophy when it comes to scheduled omitting of station stops is to avoid it when possible. Apart from operating constraints that this causes, it tends to confuse the occasional traveller who wants a simple, easy to understand, message. Nowadays fast trains on London Underground are limited to a few peak-hour services on the Metropolitan Line. Until very recently all London Overground services called at all stations. With the introduction of services to Liverpool Street this is no longer quite true but, in all probability, having services sometimes skipping stations is a situation that was forced on London Overground rather than something they wanted.
Whilst, in general, Crossrail will consist of all-stations services, there will be exceptions west of Paddington. Acton Main Line, Hanwell and West Ealing will see trains not stopping there throughout the day. These, for the most part, will be trains serving Heathrow replacing the Heathrow Connect service. What is more curious is that it looks like there will be 2tph off-peak that will run from London all stations to West Drayton, then fast to Slough, then fast to Maidenhead. This appears to be making best use of limited capacity by extending a suburban metro service to Slough to provide a better service to that town. It will also provide a much better service for the tourist market to Windsor. The continuation fast to Maidenhead is merely in order to terminate at the first opportunity.
When it comes to platforms it is apparent that here too there is a clash of ideas. With Platform Edge Doors already decided for sub-surface stations it is clear that in the central section the platforms are going to be Underground-style step-free straight platforms with no discernible gap. West of Paddington and at Stratford, and stations east thereof, one would like to think Network Rail standards apply. The problem is that currently they don’t.
The Network Rail standard for platform dimensions is very much an aspiration. Given that nearly all platforms on the system already existed (and therefore had grandfather rights) when the standard was introduced, one can see why a compliant platform is very much the exception.
There are two problems that Crossrail inherits with platforms, both of which are very different. On the Great Western Railway the loading gauge has historically been generous and the desire to minimise the step up into the train not seen as a priority. As a result just about all of the platforms are too low and too far from the track.
By contrast on the line out to Shenfield the loading gauge at stations tended to be narrower than the average. Furthermore, the extremities of the platforms are often higher than they should be. This is because frequently there are bridges at the end of the platform. To save money when overhead electrification was installed a dual voltage system of 25kV was installed with the voltage going down to 6.25kV at bridges with limited clearance. When it was decided to eliminate the 6.25kV sections and have a universal overhead line voltage of 25kV they solved the problem of clearance by lowering the track – which had the effect of making the platforms too high. So on this part of Crossrail the platforms are too close and too high.
It would clearly not be acceptable for Crossrail to have a greater gap than necessary on the Network Rail owned stations. This means that as well as many platform extensions for the longer trains, there will be a lot of work done by Network Rail on Crossrail’s behalf to ensure that the platforms are compliant. One could look at a map of the Crossrail route and take into account that Reading station would already be compliant and draw the conclusion that around 26 stations would be potentially affected and this would amount to around 52 platforms. Unfortunately it is not as simple as this, as on both the Great Western and the Great Eastern there are times (typically Sunday morning) when the line must be run as a two track railway instead of a four track railway to enable maintenance to be carried out on the other two tracks. This means that, at some stations, all four platforms at each station, ideally, should be fully compliant with the Network Rail standard for platforms.
On the final network there will be at least 112 platforms then that a Crossrail train could potentially call at, as this number will be fitted up for cab CCTV. There are also 61 existing platforms which will need extending. For a railway with only 40 stations, 10 of which will have completely new platforms, and Selective Door Operation (SDO) implemented at various locations, this is an awful lot of platforms to extend.
Mind the doors
SDO is necessary because some platforms on Crossrail will not be long enough to accommodate a full length Crossrail train and cannot be extended. Even something as simple as this brings a clash of different practices. London Underground and London Overground select individual pairs of doors for exclusion. The Train Operating Companies and Network Rail appear to have a policy of doing this by carriage. Of course, if you have walk-through trains it is arguable how sensible it is to do it by carriage. It remains to be seen which of these approaches Crossrail will take.
There is also another door issue – whether or not all the doors will be opened by the driver on arrival at a station or whether it will be left to the passenger to press a button to open the door. This may seem like a minor issue, but it one that will potentially have an impact on the passenger journey. In the central sections, for example, where many may see Crossrail through the window of the Underground (particularly as interval times decrease) there will be an expection of automatically opening doors. Yet one can also imagine passenger complaints about waiting on a cold train at Abbey Wood on a winter’s morning, or the near pointless opening of all the doors at lightly-used Iver station.
Easily overlooked is the issue of CCTV for the driver to open the doors. The London Underground practice is to have cameras located on the platform and have the display in the drivers cab with the signal being send via a leaky feeder cable. This is generally regarded as operationally the ideal set up but technically more complex and expensive. The current London Overground practice is to have the cameras on the outside of the trains.
Cameras on the outside of the trains would clearly be unsuitable for platforms with Platform Edge Doors. We also mentioned the desire not to have drivers having to adopt different procedures on different parts of the railway. From this it follows that the decision has been made to go for platform cameras linked via a leaky feeder cable to the driver’s cab throughout.
It is on issues like this that Network Rail is forced to adopt new practices. Until now CCTV images of the platform provided by leaky feeder signal to the train has been the preserve of London Underground. One wonders whether, having established the principle on Crossrail, London Overground will look at this again and push for CCTV cameras on the platform in future.
Another area where Crossrail is going to rewrite the rule book, rather than rely on precedent, is on the subject of evacuation procedures. One innovation will make evacuation less urgent and the other will make it easier to carry out.
Crossrail will be the first deep running line in Britain with air conditioned trains. This means that the interior of the trains should not be so hot in the first place and if the power remains on the trains will remain cool almost indefinitely – whether or not the train is in a tunnel section. In the main tunnel section there will be a rigid overhead bar rather than a catenary wire so one would hope that the number of times the power has to be deliberately turned off is substantially reduced with the elimination of possible dewirement.
Even if the power is off for any reason, the air conditioning will be able to keep functioning for a considerable time by using the on-board batteries. This should reduce the urgency of evacuation although there will still be the problem of no toilet on board. Having the option of keeping the passengers aboard the train is generally helpful because sometimes the quickest overall solution is to have passengers remain on the train if the duration of the delay can be fairly reliably established. Once you evacuate a train, getting things moving again generally takes a long time.
The other change to evacuation procedures is that passengers will be able to use the walkway in the tunnels. This should make evacuation procedures considerably quicker and there should be no need to turn off the power to the rigid overhead bar that supplies electricity to the train in order to facilitate the evacuation. This is a major advantage, as one of the reasons that evacuation is so problematic on the Underground is the need to be absolutely certain there is no possibility of the power coming back on once passengers are walking down the track.
There is plenty more information to come. Indeed one major area we have skipped over in this article is that of signalling – arguably one of the most problematic areas of Crossrail – similarly rolling stock. We have also not mentioned fare structure, the policy towards fare enforcement and refund policy. These topics, as well as a detailed look at the purpose of the implementation plan and its different stages, we will look at next.
For now, however, it seems safe to say that Crossrail have earned the right to celebrate what is a huge milestone – the end of tunnelling. They should also be commended for recognising that it is simply that – the end of one stage and the start of the next, one for which they appear to have successfully prepared for some time. In that light it is tempting to end this article with the Churchill quote that gave this piece its name. Instead, however, it feels more appropriate to leave the last word, in effect, to Crossrail themselves.
“If we don’t deliver this railway to the best of our ability. If we don’t make it as good as it can be. Then we will only have ourselves to blame.”
Chris Sexton, Crossrail Technical Director
It’s a very valid point, the pressure for which Crossrail’s operational team, and those involved in completing and fitting out its stations and tunnels, will now increasingly bear the weight of in the years to come.
Want to know when part 2 comes out? sign up to our newsletter on the homepage