With Crossrail well into the phase of building the railway infrastructure we look at progress in the tunnel section and what is being done to push forward ahead of schedule
Not just a real railway
A huge project like Crossrail can be looked at and broken down in many ways. Indeed Chairman Andrew Wolstenholme talks about it being two projects – the real railway and the virtual one. The real railway is obvious. The virtual railway is the the incredible amount of data being collected. This is modelling the railway in advance, and during, its construction so that all the data, diagrams etc. are up to date and correct. As a result Crossrail quite literally knows not only where every nut and bolt is, but also their history, type and how many other nuts and bolts share those properties.
The three construction phases of Crossrail
Another way of looking at Crossrail is three distinct construction phases. There is the tunnelling which, until about a year ago, was the primary activity. The second phase is fitting out of both the tunnels and the existing railway that will form part of Crossrail. The third, oft forgotten, phase is systems integration and testing. We are clearly into the second phase but, even more than most big construction projects, Crossrail is always thinking ahead and it is the third phase system integration that probably concerns people more.
Dealing with the known unknowns
A senior Network Rail manager, at the time working on Thameslink, put it this way. The construction phase is not such a problem. It may be daunting but it is all technology that is, or at least should be, predictable. In particular, timescales in the construction world should be very predictable and it is not often in the railway world that a hardware project overruns.
The predictability of hardware projects is in complete contrast to software projects (such as resignalling with ATO), which are notorious for either overrunning or not delivering what was specified. This is not to disparage the software side of things which, when it is successfully introduced, often produces greater benefits at a much lower cost and enables further improvements to be made by subsequent “tweaks”.
Systems integration is very different from construction and involves bring together and testing a myriad of processes (generally involving a lot of software) including such diverse elements as state-of-the-art signalling, ventilation management, radio reception, evacuation procedures and how to manage out-of-course events. In many cases, the technology may be almost cutting edge – mature enough to be reliable on its own but not necessarily tested in all scenarios with all combinations of the other elements of modern technology surrounding it. And that is before you add the human element. If construction is the known known then system integration is the known unknown.
Concentrate on the systems integration
The reason for spelling out the risk involved in various phases is to make it clear that any time that can be gained from the construction phase can be put to good use in the systems implementation phase. Of course in the old days the system implementation phase didn’t really exist. The physical construction was done, the staff were trained and HM Railway Inspectorate would be invited to come down and sign off the new railway. Public services would often start the next day.
As railways got more complicated, however, time would be allocated to make sure everything worked together. But all too often this time allocated was underestimated and then eaten into by late running construction. On Crossrail the current aim is, at worst, not to delay the systems integration phase and, ideally, to buy more time for it.
Scope creep bad, schedule optioneering good
It is often said that the important thing is to have a plan and to stick to it. It is also said that one should have an objective and stick to it. The two are very different things.
Once the exact form that Crossrail should take had been decided it was fairly clear it had an objective to build a railway from Maidenhead to Shenfield and Abbey Wood with the main tunnel section opening by December 2018. It is a project that is well known for resisting outside change that would alter this objective – “scope creep.”
That isn’t to say that Crossrail has been entirely immune from this. The main change has been to extend to Reading, but once Network Rail were committed to Great Western electrification this made a lot of sense and, in any case, the main reason for the project being cut back from its original objective of Reading to Maidenhead was to avoid it incurring costs that in the Railtrack era would have been assigned to it.
Another example of scope creep, largely forgotten, is level access at all stations. In this case it was something Crossrail itself wanted in order to make a better product – provided someone else found the money, of course. Crucially, this was the sort of add-on that could be provided without impacting seriously on other activities.
The importance of a Construction timetable
Construction is a very complicated activity. It relies on things being done in the right order and the right equipment, materials and appropriate people being available. Some items have long lead times. Others are scarce resources. On the positive side, with modern machinery and construction techniques things can progress very quickly if everything is planned properly. If one is lucky, one may even have an advantage of construction techniques and equipment improving in the lifetime of the project enabling time or money saving to be made. This is something that can be hoped for but not relied upon.
The secret is to have a well-planned (but not too ambitious) construction schedule so that everything can come together at the right time. Once this is in place one can determine an opening date.
Although an initial construction timetable is important, it should not be regarded as something that has to be adhered to. Nowadays, construction timetables tend to be very conservative. If you are ahead of schedule and a subsequent crucial task cannot be brought forward it thus does not really matter. On the other hand, if one is even slightly behind schedule the knock-on effects can turn a small delay into a large delay if critical plant cannot be rescheduled or specifically booked time slots cannot be taken advantage of.
A classic example, much seen on Network Rail in recent months, is the problem of an external factor such as windy weather conditions – avoiding this is crucial if you are relying on a crane. Any plan that relies on windspeed being below a certain level on a particular day is to avoided when possible. Normally this is resolved by having back-up dates in case the work cannot be done on the first planned date – but Network Rail do not always have that luxury and sometimes one simply has to cancel the work and reschedule at a future, as yet unknown, date. Rescheduling is not a huge problem for a single independent task, such as like-for-like bridge replacement, unless there is an urgent necessity to get the work done. It is a different matter if there is a sequence of tasks that can only be carried out once the work involving the crane is completed.
Three distinct construction criteria
For much of Crossrail construction the engineers have the rare luxury of not having to fit in with a live running railway – something we will explore more in an upcoming feature. Indeed, another way of dividing the Crossrail railway construction phase is to look at the three distinct areas where the requirements are different.
The central bit and south eastern section
In the central unopened railway between Royal Oak (Paddington) and Abbey Wood, Crossrail has the luxury of being somewhat insulated from having to take into account the effect of activities on the day-to-day world outside. It is true that on the surface there has to be a level of co-ordination with others, but deep in the tunnels the project is largely free to progress as it sees fit. So, although there are exceptions such as agreed closure dates of tube stations or roads, the construction can largely go on at a pace that is determined by Crossrail and can be speeded up if the opportunity arises – or slowed down if necessary.
This independence was notable in the tunnelling phase when the tunnel boring machines (TBMs) made far better progress than planned for and, for the most part, bettered the most optimistic forecasts. Other work could be rescheduled and even resequenced to take advantage of this.
The eastern section
In contrast to the central tunnel section, east of Stratford Crossrail has to contend with a live railway but it does have the considerable advantage that the only passenger service that normally runs on it is already under Crossrail’s operational wing – even if advertised under a different name. The decision has been made here to go for lots of weekend closures and upgrade the railway whilst the opportunity is there. There are some fairly big projects going on such as the work at Shenfield which, amongst other things, involves an extra platform. Elsewhere there are new station footbridges and major station refurbishments which include platform extensions. There is also a comprehensive programme of railway renewal, including replacing the catenary.
There is never a good time to do the work involved on the eastern section of Crossrail, but it is far better to get this done before Crossrail begins in earnest. In a way it could be argued it has already been left later than it ideally should have been because TfL taking over the Shenfield – Liverpool Street service has already resulted in a considerable increase in the number of passengers, despite the accompanying weekend engineering works.
Clearly, east of Stratford, Crossrail construction is constrained by only having the railway closed at weekends though it is fairly free to use this time to its best advantage.
The western section
West of Old Oak, Crossrail has to work with a live 7-days a week railway. This, in some ways, is made worse by Great Western electrification also taking place. The sharing of tracks with the Great Western Railway (and freight) severely restricts both the time available and the opportunity to reschedule work once time slots are allocated – generally booked months, and sometimes years, in advance.
It is pertinent to note that the original plan was for Crossrail to open on the west side first and run new stock from 2017 onward from Maidenhead. The thought was that there was not much to be done, the platforms were long enough and the signalling was in place. How hard can a bit of electrification be? Eventually wiser counsel prevailed and it was realised that not only was the work on the western side not that simple but that, in fact, it would be easier to have the eastern side ready by 2017 even though more work needed to be done.
A manageable project
Part of Crossrail’s great success has been breaking the project down into manageable chunks, with each chunk let as a contract. Indeed Crossrail managers seem to refer to everything by its contract or C number – be it the Thames Tunnel or the fitting out of Tottenham Court Road station. In construction, as in many other walks of life including software, the trick is to get these manageable chunks to an optimum size – a true black art – and crucially to minimise (or at least manage) the interfaces between the different construction projects.
The lengths Crossrail has gone to so as to minimise interfaces and to prevent one contract affecting another is quite surprising. This can be see at Tottenham Court Road where the track work (one contract) is completely separated from the station work (a different contract) by a large temporary, but fairly substantial, wall. The structure for hanging the platform edge doors from the ceiling is almost complete but a small portion at each end has not been finished in order to achieve contract segregation by the required date. The only connection between the two contractual sites is a fire door for emergency evacuation.
Amongst other things, with this wall in place at Tottenham Court Road, the tracklayers can plan their work including ventilation management plans using temporary ventilation (especially vital once the diesel power trains are in the tunnels) without having to consider the impact on, or liaise with, the team at Tottenham Court Road station.
As we shall shortly see, this separation of contracts does have its downside.
Current state in the middle
Today the overall project is progressing according to schedule, more-or-less. Well established techniques for laying track learnt from the Channel Tunnel, French TGV routes and HS1 mean that a sophisticated production line system is progressing.
In recent months Crossrail has been proudly showing off its concreting train, its multi-purpose gantries and its drilling rig.
Crossrail video explaining tracklaying. Accompanying text here.
The concreting train is purpose built and can concrete approximately 180 metres a night. The general plan is that concreting is done at night with maintenance on the plant and any site preparation necessary done during the day. Only if noise for residents above became a significant issue would this be switched around, but this has not been necessary so far.
To all intents and purposes a production line phase of fitting out, where one large production factory on rail wheels simply follows another and the railway gets built, is now in effect on Crossrail. Again, this is something we will talk more about at a later date and whilst that is a simple overview, the reality is much more complicated than that. In general, rail-mounted machinery will return to its base on the surface each day or night for maintenance and restocking.
With around 21 months to go until January 2018, when Crossrail aims to be in an integration and testing phase, they are clearly going to have their work cut out if they are to have the 42 kilometres of tunnel and various short surface sections ready in time. In other words, they are progressing well but they need to – because the schedule does not allow for otherwise.
Crossrail can’t just rush and get everything done at once because they are constrained. This is basically down to three issues: sequencing the tasks, access considerations and restrictions brought about by noise issues. We consider each of these in turn.
Things have to be done in a particular sequence. The concrete base has already been laid through the tunnels. The next stage is generally to lay temporary jointed rails. This enables the concrete to be poured for the track. Once the concrete is poured then permanent track can be laid. With track in place one can bring along other “rolling factories” to do their work in a very precise, ordered way.
The above is a simplification but, generally, the idea is to get rails in the concrete so that one has a precise alignment and other plant can work. Of course it is not as simple as that. One needs plant to lay the rails in the first place. In the case of Crossrail, these are on pneumatic tyres. The four multi-purpose gantries, in particular, can only run on pneumatic tyres. On these, the wheels can be adjusted heightwise so on one side the wheels are at track level and on the other they climb up a ramp to be at platform level. Furthermore the whole frame can “squeeze itself in” so that it fits within the Network Rail loading gauge and can travel on the live railway upstairs.
The hole story
Another complication is that by sequencing activity one has, at the start, nothing happening in certain places and plant lying idle elsewhere. So, although the preference is for the drilling rig to run on rails it also has the capability to run on tyres and, of course, there is no reason why the necessary holes for rigid catenary, cable brackets and more cannot be drilled in advance of the rails being in place. This is especially helpful when one considers how many of the tasks subsequent to rail laying require the holes to be in place before they can commence. It would be possible to drill the holes by hand but that would be less accurate and the marking out in advance of drilling would be very time consuming.
As a digression, it is interesting to note that Crossrail did not have the holes already pre-drillled or cast in advance, but chose to wait until the tunnels had actually been built. The actual reason is unknown but anyone aware of the history of the construction of the Spitfire fighter plane will be aware how crucial it is to get holes in exactly the right place. One of the lesser known stories of World War II is how Lord Nuffield (of Morris cars fame) nearly lost us the war by trying to build Spitfire planes like cars. Amongst other problems, he tried to build the planes with the holes for the rivets pre-drilled into the assembly parts, only for them to not line up when the plane was assembled.
The second of the major constraints that Crossrail faces in the tunnels is access for rail vehicles. The Pudding Mill Portal is not, to the best of our knowledge, being used for access. This leaves the Royal Oak Portal west of Paddington and the Plumstead Portal south of the Thames. This means that it will be a long time before rails get laid at somewhere around Farringdon. As well as that, you then have the problem of getting stuff out. If other plant is following on behind then it is not simply a matter of reversing. There is a crossover tunnel at Holborn, currently without rails, but even if it had rails laid, not all the plant would be capable of using it.
Noise and consequent track considerations
Crossrail claim that in most normal circumstances noise will not be a problem. The sleepers will generally have their own rubber sock that should remove a lot of the vibrations. The idea of modifying the structure around the track to reduce vibrations is nothing new and it has been successfully used in a different way by Network Rail for the Thameslink service under St Pancras Hotel. On Crossrail the use is much more extensive and the only tunnel not to be fitted with any kind of noise dampening track structure is the Connaught tunnel under the Royal Docks.
When the Crossrail bill was going through Parliament there were numerous objections concerning noise but clearly, in general, Parliament was satisfied that everything reasonable had been done to counter it, and even the residents of Mayfair had to be content with this.
Exceptions were made though in a couple of cases. These were because the vibration would interfere with recording studios located in Soho and also the Barbican Concert Hall. Parliament felt that these objections required more than the standard mitigating measures and specified that what is generally known as floating slab track was to be used in these locations.
Not the complete answer
One has to be careful with extolling the praises of floating track and other noise reducing track construction techniques too much. Crossrail can proudly show off their floating track construction deep below the streets of London, but the Baldrick-like flaw in the cunning plan soon becomes apparent as one can clearly hear and feel the vibration present from an existing Tube line – presumably the Central line. The technique is probably completely impractical for retrofitting on existing lines. The space available would likely be a problem but the length of closures required to install it would probably make it utterly impractical on their own. In a similar manner one does wonder about the benefit of quieter track in the vicinity of London City Airport.
It’s all done with springs
Floating slab track is track laid on a slab of concrete that is supported on very strong springs.
The technology is not new and has been used before on various railways including HS1. However, for various reasons, it is expensive. It requires a lot of manpower to install. To get the necessary weight of supported concrete track you also either have to make the tunnels even bigger – which as you are using TBMs means that, in practice, they have to bigger along the whole length of the route – or you have to use an especially dense form of concrete. The especially dense form of concrete brings its own separate issues as you cannot use standard concreting plant and machinery with it.
Now having a relatively small site where you have to put floating track is not too bad. You simply work from the west to access sites to the west of the floating track and work from the east to access sites to the east of the floating track. Meanwhile work advances on the floating track ready for the final fitting out when the production line finally arrives on site.
The problem for Crossrail is that there are two sections of floating track. These effectively isolate the section in between. A further complication is that you potentially cannot get the heavy concrete to it, which limits preparation to setting out the reinforced steel bars (rebars) by manual labour – a massive job in itself.
Getting the concrete in
Ultimately the Crossrail construction programme could not be accelerated further because of the restriction that the concrete had to be delivered from one of the extremities of the tunnels.
The issue of concrete delivery was recognised during the phase when an initial concrete base was put into the tunnels. This concreting was not so critical and did not require a sophisticated concreting machine to deliver it. Not for the first time, the Kingsway tram subway provided an opportunity for Crossrail and from there they simply drilled a hole into the the top of one of the running tunnels to provide a flexible chute for concrete delivery. This could then be delivered to wherever it was needed.
Unfortunately the Kingsway subway solution wasn’t really viable as a way of providing heavy concrete for the section between Tottenham Court Road and Bond Street station. The extra weight would have produced problems and the concrete really needs to be continuously pumped through rigid pipes.
It turned out that what one would have thought was an ideal location, Tottenham Court Road station, was not really suitable either. Although the site may look big there is a lot of concurrent activity going on and the thing would have been a logistical nightmare.
In contrast Bond Street (we presume this means the Hanover Square site) provided more possibilities. The problem here, as elsewhere, was that this operation simply had not been envisaged and, though it sounded simple, there would be all sorts of issues. It is now that the isolation of the different contracts was working against a beneficial change and it took six months of negotiating to sort out all the issues.
More fortuitously the Fisher Street shaft, just to the east of the Holborn Kingsway subway dive-down entrance, was ideal for pumping concrete. The problem was that it wasn’t especially close to where it was needed and as, far as is known, no-one had pumped heavy concrete over the distance necessary. As it was, new plant would have to be imported to the UK and there would probably be no future use for it in this country. The heavy concrete is more typically used in Sweden for basements as the water table is quite high in Sweden so the walls and floor need to be especially heavy to prevent the entire structure from floating. However, when building walls and basement floors in Sweden you generally have easy access to the site.
Ianvisits has already reported on the visit Crossrail made possible to see what is going on under London. You can find it here. Amongst other things the photos are probably better than you will see on this site. Geofftech has done his usual excellent video for Londonist which can be found here. No doubt New Civil Engineer and all the other usual suspects rounded up by Crossrail for a visit will also provide excellent interesting accounts of what is going on. RailEngineer has reported in the past on the details of the tracklaying process. Here we just remind you that it is some of the less exciting advances, such as working out how to deliver a load of concrete, that can actually provide a significant benefit to a project like Crossrail.