In part 2 of Uncircling the Circle we looked at London Underground’s proposals for getting rid of the Circle Line in the form of a continuous loop and how they came to fruition. In doing so a document dating back to 2009 was used to explain the revised Circle Line proposals. What was rather skipped over was an arguably altogether more interesting diagram in that document with a vision of how services will run on the Sub-Surface Railway (SSR) by the end of 2018. In this article we look at that diagram and the proposed service.

Before we do so, however, we must first look at a very recent major change to the programme – the signalling.

Starting over on signalling

Given how dependent the plan is on the signalling system it was slightly worrying to read TfL’s Operational and Financial Performance Report and Investment Programme Report – First Quarter, 2013/14 which was published in September 2013.

In particular:

SUP Automatic Train Control (ATC): Progress to date on the new signalling system shows that there remain significant challenges to deliver the capacity uplift associated with this by 2018, within the overall programme funding. The programme team is engaged in discussion at a senior level with the signalling contractor to review the current status and expedite the technical solution and the delivery strategy in order to maintain the integrity of the required benefits and the overall completion date of 2018

Obviously London Underground’s new plans for services will only work with a suitably advanced and functional fit-for-purpose signalling system, so it was not surprising that any delays here would cause great concern, especially as memories of signalling problems on the Jubilee Line are still fresh.

A key sign that TfL were aware of the risk of non-delivery and looking to nip it in the bud came at the recent board meeting in December, where Mike Brown, gave an oral update on the situation. Although issues of commercial sensitivity were clearly part of the equation, that something serious was being discussed was clear from the fact that details were not made public.

There was, however, a statement in the latest Commissioner’s Report:

The first dynamic test run of an S7 train at the Old Dalby Test Track has taken place in manually protected mode. Discussions with Bombardier, including a director-level review, are continuing with regard to the automatic train control systems and the migration strategy to maintain the overall completion date of 2018.

One suspects that London Underground would have been much happier if the same established system that is currently working both the Jubilee Line and Northern Line (in part) was the chosen signalling system but London Underground has to abide by strict rules on competitive tendering. If events had continued to proceed as they were then it would not have been surprising if SSR resignalling became a topic for questions at a future GLA transport meeting.

Any delays to implementing the signalling on the SSR, however, also carried an additional concern – any delay to the SSR resignalling could impact on the Piccadilly Line upgrade which will be following the SSR upgrade. Currently there is common track regularly used by both lines as well as the ability to run the District and Piccadilly on the other lines’ track between Hammersmith and Acton Town should circumstances require it.

Taken all the above into account, eyebrows were thus raised in LR Towers by the sudden addition of a TfL Finance and Policy Committee Meeting to the 2013 calendar, right in the middle of Christmas party season and with no public access. Once again, commercial sensitivity was likely in play, but once again it suggested that a major change was coming. With TfL almost certainly dead-set against a change in the delivery timetable, that suggested a change of some kind in the contract between themselves and Bombardier was imminent.

That reality was finally confirmed today, 31st December, with TfL officially confirming that, in order to meet the current delivery timescale, Bombardier’s system is being dropped and the SSR signalling contract is being re-tendered:

London Underground (LU) and Bombardier Transportation today announced that the signalling contract for the Circle, District, Hammersmith & City and Metropolitan lines (Sub Surface lines) will be re-let by LU following discussions between the two companies.


Bombardier was awarded a contract in June 2011 following a competitive tender process, and has already undertaken a great deal of preparatory work. A new unified control centre has already been constructed, which means the project is ready to be advanced further. However due to the complex nature of the network and the nature of the work to be carried out over the next five years, LU and Bombardier have taken the decision that LU will re-let the contract and continue the works with another contractor in the New Year.

LU will this week seek expressions of interest in the Official Journal of the European Union (OJEU) and at this stage there is no change to the original delivery timetable of 2018. In the meantime Bombardier will continue to supply the S-Stock fleet…

That this is a big decision that is difficult to overstate. It’ll be a tough change to push through to the same delivery timetable, and only time will tell whether the correct decision has been made.

One thing, however, that deserves to be highlighted is that London Underground have been vindicated in their insistence that the live railway not to be used as a project test track, as was done on the Jubilee Line upgrade. Instead initial testing – testing that has contributed to this decision – was done at Old Dalby in Leicestershire. Had that not been the case then this decision would almost certainly not have been possible at all, and London may well have been doomed to repeat the mistakes of the Jubilee upgrade.

The Current timetable situation

With signalling considerations out of the way then, and taking TfL at their word that 2018 remains the delivery date, we can now consider what the intended plan is. To do so it would be helpful first to look at how the SSR was run prior to the uncircling of the Circle Line and also immediately afterwards. Very conveniently diagrams for these are provided in the document referred to in the first paragraph.

Current service

Diagram taken from “uncircling the circle” proposals. The figures in black are the trains per hour (tph) as was up to December 2009 and those in red the tph service proposed for after December 2009. It would appear this never actually took place in the form shown as the increase in Wimbledon service only happened when the Olympia shuttles were withdrawn.

The above diagram is very complex and difficult to follow. In as much as it can be comprehended at all, it shows a complex service pattern prior to uncircling the circle being replaced with an even more complex one afterwards.

The revised service (by and large the service we have today) has many unsatisfactory features.

  • Along the top of the circle you have 6tph Circle, 6tph Hammersmith & City (H&C) and 15tph Metropolitan. It is quite impossible to get any kind of consistent repeating pattern out of this and also means that some trains need minor retiming so that, on the Circle and H&C, 6tph does not mean you get a train every 10 minutes.
  • Along the bottom of the Circle you have a similar imbalance. This is more pronounced westbound where there are 9tph to Wimbledon, 13tph to Richmond/Ealing Broadway and 6tph Circle Line.
  • On the western extremity of the District you have 7tph to Richmond but 6tph to Ealing Broadway. The 7tph do not fit in very well with the 4tph with London Overground that share the same tracks.
  • There are 3tph peak hour terminators at Tower Hill. This would seem to be undesirable and would appear to be either because there is not enough rolling stock or because there simply aren’t the train paths to continue eastward or, more likely, both.
  • There is an imbalance on the Wimbledon branch. This to some extent is good because more trains go via Victoria where the greater demand is but on the other hand, again, it leads to an erratic service on the branch. The reason for this probably has more to do with the lack of C stock than fine tuning of the timetable to match customer demand.


  • The busy junctions at Earl’s Court are made more difficult to operate because they have to fit an additional 3tph to Olympia.

The rolling Stock Issue

There is another factor that is not present on the diagram – that three different types of rolling stock were in use at the time the diagram was created. Each would have its own speed profile and required dwell time at a particular station so there would not be any uniformity of performance when mixing the different stock on the same line. The rather underpowered (by today’s standards) Metropolitan A60 stock has now gone, replaced by the first batch of ‘S’ stock, but the issue of multiple types of stock will remain for a few more years yet.

The overall situation presented is clearly not going to be very satisfactory and was very much a case of “this is where we are, this is the rolling stock and the signalling we have got and this is the best plan we have to make the most of it”. It also has a feeling of “if we started afresh we wouldn’t end up with this” or to put it more colloquially “if I were you I wouldn’t start from here”.

If there is one good thing about all this, it is that all the rolling stock and the signalling were due for renewal, which gave London Underground a rare opportunity on the SSR to more or less start afresh (although obviously one could not ignore the more permanent infrastructure in terms of stations and routes that were already in place).

We will digress for a moment and take a look at how the Swiss would tackle this issue.


The Swiss have a word to describe the nature of their national rail timetable. It is called Taktfahrplan. As always with German compound nouns, a literal translation of the parts doesn’t really give you the full meaning. The word is a compound noun consisting of Takt (regular/beating/rhythmic) and Fahrplan meaning timetable. Actually Fahrplan itself is a compound noun of Fahr (travel/drive/distant) and Plan (map/schedule). Note that the German word Plan very often does not mean plan in the sense of a forward looking plan so is a bit of a linguist’s false friend. If one wants to emphasise something as going beyond what is normally expected then the Germans just stick in the prefix über. So perhaps that prefix would be justified for an intensive timetable. No wonder Mark Twain wrote an essay on the awful German language.

Anyway, to get very vaguely back on topic, the idea of the Swiss Taktfahrplan is to produce a timetable with lots of clockface (typically once or twice an hour) connections. This can produce the scenario of a station being empty for three-quarters of an hour and then trains arriving from all directions and disappearing again. There is no concept of rush hour. Alternatively and more relevant to us it could describe a network with a very regular repeating pattern of trains.

The Swiss are generally admired for their Taktfahrplan but there are exceptional factors which make it particularly appropriate to Switzerland – a small mountainous country with most of the population based in a relatively few reasonably sized towns and cities certainly helps. A vital aspect to der Taktfahrplan, which gets updated each year, is the planning that goes into it. One could refer to this as das Taktfahrplankonzept but the latest German spelling reforms frowns on people making up their own compound nouns because, as you can probably gather, it quickly starts to get very silly.

The planning that goes into the Swiss Taktfahrplan generally has a ten year lead time and this is needed to ensure that the necessary locomotives, rolling stock, track layout, signalling and station capacity are available for the timetable when it is implemented. Being Switzerland, there also has to be liason with the many railway companies and the cantons which is probably a factor in why the process takes ten years. On the plus side the consistent political support for the railways means that planners can plan ahead with confidence.

The reason that Taktfahrplan would be so appropriate to describe the planned Sub-Surface Railway enhancements is that it really does have the two elements within it. The planning and thought gone into it is very much on a Swiss timescale and very comprehensive encompassing just about everything – track improvements, power upgrade, new trains, new signalling and longer platforms. It is a world away from what seems to be the usual British approach of “given that this is what we have got how can we tweak it to improve on it” and a feel that the whole thing was planned by a dysfunctional committee. The other Taktfahrplan element is the regularity of the trains and it is almost certainly the case that the current plan will result in a regularity of service never seen before on the SSR.

It is clear that many critical decisions concerning the SSR were made in the early part of the 21st century. It was to be an automatically run railway but with drivers at the front of the trains. The trains were to be of the same type but of two variations. Most would be seven carriages long with limited seats but for services north of Baker Street there would be eight carriages and a slightly different internal layout in order to provide a few more seats. These decisions would determine a lot of the decisions that followed such as power supply requirements but there was one more critical element that would influence other decisions and that was the timetable – or at least the timetable frequency. The exact timings could be decided later.

The important thing to grasp about the decision making process is that it is very much a case of “this is where we would really like to be in 2018” and make sure that all the infrastructure is upgraded as necessary to make that happen rather than simply a case of recognising a desire to increase capacity as much as reasonably practicable and devising a plan to achieve this. In other words it is clearly setting the agenda rather than just reacting to events and a general desire to increase capacity.

The Olympia Shuttle Issue

We are not generally privy to London Underground’s internal decision making but we get the strong suspicion that it was realised around 2009/2010 that running the Olympia service in addition to all the other trains planned to go through Earl’s Court was just not sustainable. If this was the case then it is fairly reasonable to assume that it was quickly realised that, if the change were to be made, it may as well be done sooner rather than later as this would enable the crowded Wimbledon service to be enhanced. In the same way that the continuous circle had to go, the Olympia service had to go – at least in peak hours.

Even if the Olympia service was sustainable from a timetable point of view one suspects that it would be very hard indeed to justify the new trains required to run this service in the peak hours when they cost around £10 million each. At weekends there is no issue – there will be trains that would otherwise be sitting in the depot doing nothing.

In a sense London Underground were lucky. Of prime consideration was the almost regular Overground service from Olympia of 4tph supplemented by one train an hour from Southern to West Brompton. This meant that the alternative service from Olympia is really little worse than the direct shuttle for most journeys. In the reverse direction it is not quite as good as one has to use the overbridge at West Brompton to change trains rather than just cross the platform, but it the circumstances it was considered to be an acceptable alternative.

West Brompton new link

Rather belatedly it seems, another opening is being created between platforms 2 and 3 at West Brompton to make changing from a southbound London Overground (or Southern) train to a northbound District Line train simpler.

A Rare Case of Political Support for Withdrawing a Train Service.

Another factor that made getting rid of the peak hour Olympia services less of a challenge than it might have been was was that it was “an easy sell”. By doing it prior to 2018 and reallocating the trains to the Wimbledon branch it was bound to be a popular move with many commuters who would back the plan. As it turned out, it would also go down well with a local MP and local Assembly Member. In 2011 TfL announced proposals to withdraw the Olympia service except at weekends. The withdrawn trains would be used to supplement the Wimbledon branch.

The proposal to withdraw Olympia trains is is probably the only occasion where one can find politicians urging their constituents to support London Underground’s proposal. Richard Tracey AM and Justine Greening MP were urging their constituents to write to London Underground supporting their plans. Of course, the losers would by and large be visitors to weekday exhibitions. They would typically come from outside London or even abroad and would be unaware of the consultation and, if they were, in the latter case possibly not have a good enough grasp of English to add their voice. This is not to say consultation procedures are wrong but that they are, inevitably, imperfect.

The 2018 Peak Period Plans

Planned service

The original proposal for the Sub Surface Railway 2018 peak hour frequencies. Since then it would appear that the intention not to run the Olympia shuttles in the peak period is permanent.

Ignoring what happens north of Baker Street for the moment, the solution produced from 2018 onwards is almost an example of perfection with balanced frequent regular services and just about every part of the sub-surface railway used to maximum capacity. The peak hour service is based on 8tph and multiples thereof and trains should arrive at any location in a consistent predetermined sequence.

Off peak and Weekends

Unlike in Switzerland, in London it is generally not realistic to run the same frequency throughout the day Monday-Friday. In the off-peak the same basic pattern of services will be maintained but based on 6tph or multiples thereof. Weekend times would be the same except that the Olympia service will run. Off-peak the District trains that terminate at Barking in the peaks will also probably be cut back to Tower Hill to reflect the disparity between traffic in the centre and in the inner eastern suburbs where 12tph off-peak should be adequate. Terminating at Tower Hill will not unduly affect many travellers but will cut down on unnecessary train mileage and simplify working through the critical Minories and Aldgate East junctions.

The proposed SSR 2018 scenario will have no doubt have been run through computer models many, many times to show that it is workable – even when perturbations are applied. It is also a fairly safe assumption that such a timetable can only work if there is intelligent junction control to optimise passage through the critical junctions.

The Metropolitan Line has to be Different

We earlier ignored the Metropolitan Line north of Baker Street. This will not quite fit into the 8/16/32/tph pattern. For starters, the way services run are more complicated than they strictly need to be. The obvious thing to do is ensure that the 16tph from Uxbridge all go to Aldgate and other Metropolitan Line trains terminate at Baker Street. The need to continue to provide some service from Amersham and Chesham to the City probably kills any chance of simplifying the service pattern in this way.

The Metropolitan service interval does not quite fit into the 8tph (or multiples thereof) at regular consistent spacing except for south of Baker Street where it has to in order to fit in with the rest of the SSR. The Watford Junction service is only planned to be 6tph at 10 minute intervals both peak and off-peak. Clearly 8tph would fit better with the regular 8tph peak SSR ethos but one has to be realistic when the demand is not expected to be there. The Amersham/Chesham service does not appear to fit into the regular 8tph pattern either but in a way it does with 4tph to Amersham, 2tph to Chesham and 2tph that would go to Chesham if the track was double-tracked and the demand was there.

There is also the issue of just how many Metropolitan Line trains can be terminated at Baker St during the peaks. The plan requires 6tph per terminating platform which in itself is probably quite a challenge with the terminating platforms on the outside of the through ones. It would probably be a step too far to attempt 8tph per terminating platform here and, because of the tight curves, it would be impractical to redesign the track layout to make the middle platforms the terminating ones.

Will it work?

Come December 2018, or maybe some time after that, we should see if the grand plan works. We will also see just how the decision to re-let the signalling contract pans out. Presumably before then we will see some partial upgrades and these should give us (and London Underground) a big clue as to whether it all looks like working. The problem for London Underground is that there would not appear to be any obvious fallback position that wouldn’t introduce horrible complications. Ultimately one could drop down to 28tph but that would not really offer a better service than today.

One suspects that a great effort will be put in to making sure the plan does work and the proposed frequencies can be achieved. The decision on getting to grips with the signalling is indicative of that. If there are problems one suspects that, like the signalling issue, they will be robustly focused on in order to make things work. As the project comes nearer to fruition this would apply especially to any local difficulties that can be quickly resolved with special measures – such a restricting access to a particular station at certain times if necessary or reorganising pedestrian routes so that people can exit the station quicker.

Can even more be achieved?

One suspects that the plan is about as far as one can go with frequencies given that many junctions are going to have to handle 32tph. As the level of traffic goes up it will probably be a case of working harder just to maintain the current situation. If more investment is made after 2018 it will probably be on stations in order to keep passenger movement flowing rather than more trains, longer trains or further enhancements of the signalling system. Victoria (District & Circle Lines) is one station that is known to be on the list of stations where it is hoped that a major capacity upgrade scheme can be implemented in future.

One would like to think that after 2018 we will have stability and a long period of little disruption or change. This would probably be the case were it not for the upgrade of the Piccadilly Line in 2021-24 which, amongst other things, shares tracks with the District Line. Although the Piccadilly Line is a deep tube line and not a subsurface line, it currently has to be treated in some ways as part of the sub-surface railway. So for our next look at an Underground line we will look at proposals for the Piccadilly Line and in passing see how these proposals affect the sub-surface railway – and the District Line in particular.

Written by Pedantic of Purley