We have looked at the New Tube for London (NTfL) before but, with very little officially announced, it has been hard to pin down the details. Thanks to the a presentation to the Rail and Underground Panel and a follow-up one to the Finance and Policy Committee though we can now, with a little more certainty, say something about it. In the past week or so there have also been a number of reports published that are relevant in some way to New Tube for London, and we can now look at these where relevant to provide some background as well.

The Prelude: Sub-surface Railway Resignalling

Amongst the numerous documents that have been published last week is the London Assembly Report of the Sub-Surface Railway (SSR) signalling fiasco. Whilst a lot of the cultural problems have already come to light and have been previously reported on (not least by ourselves back in 2013), there are some financial figures from that report that have to be interpreted very carefully.

The headline figure, inevitably picked up by the BBC is of £886m. Caution should be exercised, however, and one must look at the wording very carefully. For example, in the introduction to the London Assembly Report the Chairman, John Biggs, states:

And there is a staggering cost associated with this mismanagement, which leaves TfL with £886m less to spend on its capital programme than it thought it had.

This reflects the difference between the amount originally budgeted for the project some years ago and the expected final costs. But within that figure some of the cost is for extra features in the new specification. A lot of the difference can also be accounted for because Bombardier tended for the job with a price that simply was not achievable.

If a builder offers to build you a house for £100,000 but cannot stick to the contract and you find actually have to pay £1m to someone else get the job done, then it is hard to argue that you have lost £900,000. The truth is that unknown to you, from that outset, getting it built for £100,000 was never achievable. Put simply, in some sense TfL have not ‘lost’ or ‘wasted’ that money, because it was never going to happen.

Of course the above does mean that there has been a failure to budget properly, and this has had consequences for New Tube for London. The report does highlight, however, truly wasted money because of bad contract terms unfavourable to TfL. This meant money effectively being paid to get out of the contract and the realisation that the bulk of the work already done was largely useless and would have to be written off – a figure that is estimated as £67m of truly wasted expenditure. This is certainly disturbing, but the transport world in London is full of sub-optimal expenditure. The true cost of abandoning articulated buses because the Mayor did not like them has not been calculated and many would question whether the New Bus for London, especially when it has two crew members, offers true value.

Biggest of all the financial failures has to be Public Private Partnership (PPP), even though good things did come out of it. The Shaw report into the running of Network Rail refers to PPP and states that:

The combined cost to buy-out the shareholders and establish the three Infracos was over one billion pounds.

This would be on top of any additional cost of setting up PPP or the charge to London Underground during its years of existence of higher costs than would otherwise be necessary. It is hard to imagine that the final cost amounted to less than £3bn and figures for the overall real cost have been speculated to be as high as £10bn – depending on how you do your sums.

What about the passengers?

One very pertinent cost identified in the London Assembly report is the loss of income due to reduced number of passengers caused by lack of capacity. In the Executive Summary of the London Assembly Report it states that:

TfL estimates that there will be 11 million fewer journeys a year and that this will cost it £271m in lost fares income. The broader economy will also suffer, with TfL estimating damage in the hundreds of millions.

One could regard this as a case of “don’t build it and they won’t come”.

Of course a loss of fares income is not automatically the same as a loss of that sum of money, because other costs (such as staff and station upgrades) may have been also necessary to enable that income to be collected. Nevertheless it is a sobering thought that this substantial cost ought to be factored into the cost of delayed projects, and this does not bode well for delays to New Tube for London.

Sub-surface Resignalling: The real bad news is delay

It is not surprising that the London Assembly Budget and Performance Committee concentrated on budget and finance. We would argue though that the true financial loss was not great, however, and that the really big worry is the delay to Sub-surface Railway resignalling and the consequential delay to subsequent projects elsewhere. As we have said many times before, delays on SSR resignalling have knock-on effects on New Tube for London, and we are already seeing that replacement of the Northern line and Jubilee line fleet is unlikely to be possible before 2040 at the earliest as a result.

To be absolutely fair to the Budget and Performance Committee Report, they do not ignore this delay. In fact they rather overstate it. Again, according to the executive summary

The programme is now not expected to be completed until 2023 – five years late.

Like the financial cost of the abortive contract, they are not being entirely fair – for they are not comparing like with like. The 2023 end date refers to some off-peak enhancements which London Underground are being extremely vague about (although we think we know what they are). These were not included in the failed Bombardier contract, so a fairer interpretation is that the programme is four years late – still significant but not quite as bad as claimed.

The surprising thing is that the report does mention what is possibly, to us, the most worrying aspect of all, yet it does not really spell out the harm done and almost understates the damage. In a falsely reassuring message we are told that:

The budget for the ATC element of the SSUP has been increased by 64 per cent from £1,382m to £2,268m. The majority of this additional expenditure will not become due until 2021-23, after the current business planning period, and hence TfL has not had to cancel or scale back any investment programmes already underway.

What this is really saying is that after 2021, beyond the current business planning period, TfL will have to cancel or scale back some investment programmes not already underway. This is primarily the very expensive New Tube for London. Cancelling is not an option, scaling back (in terms of specification) is probably not a realistic option either, so the inevitable consequence is that the critical New Tube for London programme will be delayed – which is what we have continually been seeing over the past few years.

Perhaps we could be as bold as to suggest the real failure of the SSR signalling contract was to believe in, and not scrutinise, a bid that was far too low in price for work that was not achievable and consequently forced a series of decisions that have led to the critical New Tube for London programme being delayed.

New Tube for London: The next big renewal unknown

With the future of Underground updates now fairly clearly mapped out until around 2023 – give or take some doubt as to what will actually be decided on for the Northern Line Upgrade 2 – it is time to regard the SSR resignalling fiasco as water under the bridge and to look beyond 2023, when we enter uncertain and less well charted territory. This leads us to look again at the huge upgrade to trains, signalling and just about everything else on four deep Tube lines happening under the New Tube for London banner.

A brave concept

It is hard to deny that the New Tube for London (NTfL) is a brave idea. The intention is to make best use of the existing tube infrastructure and create a new generation of trains that will last into the second half of this century. With the cost of new tube rolling stock already at around £2m for a single conventional carriage it is important to get the design right. The payback is that with the right infrastructure you can run far more trains than you can today. This applies to the Piccadilly line in particular which is currently struggling to run 24tph, yet 32-33tph should be possible. With NTfL you also have rolling stock that you can reasonably expect to last for 50 years and be updated as necessary to reflect the latest technology – as confirmed by the Commissioner, Mike Brown, in his latest Commissioner’s report in which he states:

The new trains will serve London for around 50 years and will be future-proofed to harness new advances in technology.

The project is extremely holistic with just about everything being being addressed – automation, the platform train interface, power upgrades, energy efficiency, the best train design possible in the space available, cooling, revised depots and even revised routeing where appropriate. It is a tall order and it is no wonder that until now the implementation date seemed to be slipping further into the future with every update.


According to the presentation slides, the New Tube for London project:

Encompasses 4 “deep tube” lines which were initially planned for upgrade under the PPP (2014 to 2020)

So, if things had gone according to the original plan, we would already be one third of the way through the NTfL upgrade. In retrospect it is hard to see how it ever would have been possible to upgrade in six years. The SSR upgrade – now unhelpfully called the four lines modernisation – is of comparable size, incorporates the same number of underground lines and is not now due to be complete until 2023. The idea of giving four Tube lines – even if one of them is the Waterloo & City – with a complete overhaul using the latest technology in just six years seems overly ambitious when viewed through modern glasses.

NTfL implementation timetable

The latest plan for implementation of NTfL

What we now have from a presentation to the Rail & Underground Panel is yet another set of completion dates for the implementation of rolling stock. What makes this one somewhat more plausible is that this time there appears to be a more realistic scenario for completely automatic operation (no staff of any kind). Furthermore, there also appears to be a commitment to this. Although previously (at least officially) no decision had been made, it was hard to believe that, in reality, there wasn’t an intention to move to completely automatic trains at some point. It was much more likely that the lack of a specific commitment was really an honest reflection of their being some doubt as to the best time to introduce what is more properly known as Unattended Train Operation within the lifetime of the project implementation.

Back to plan A

Years ago it appeared obvious to all that the best way to implement a programme of driverless trains was to start with the Waterloo & City line. This was rejected for various reasons – not least that upgrading the Piccadilly was the more urgent priority.

One of the sticking blocks seemed to be the “we will never order a train with a cab again” mentality. In other words, it was getting to the stage where it was unthinkable that the next generation would have a driver in the front cab for another forty years – at least on lines capable of full automation. A flaw in this plan was later identified. It was unrealistic to mix old manually driven stock and new driverless stock driverless on the same line, since the former prohibits platform edge doors under current rules and whilst the latter requires them. It was thus realised that you had to initially have some sort of cab – if only to get you through the transitional stage.

What seems to have now happened is that this coupling of the concept of new trains with that of driverless operation has been well and truly broken. The reason appears to be because it is more important in getting the new trains in service before the old ones become too much of a maintenance burden than it is to get them fully automated.

One suspects that deferring, for the most part, the automation proposal to the end of the implementation project makes implementation much less vulnerable to problems. A lot of people will immediately see this in terms of union difficulties, but it is probably the technical challenge to get a fully automated system working in a single track deep Tube tunnel that has really focused thinking. It could be argued that, whilst the rest of the scheme is very involved and complex, it really isn’t pushing at the boundaries of technology and is more a case of integrating already established railway components and technology. In contrast, with modern automated driving and supporting signalling, we are not yet at the stage where the risks associated with this are acceptably low enough to be confident it can be implemented on the Underground without major problems.

The consequence of this revised thinking appears to be the eminently sensible option of getting the trains in service on the Piccadilly line and then the Waterloo & City line. The latter will be automated, but with only two stations and no outdoor sections (and hence no issue with wet rails) this should not be a major technical challenge.

There was much early enthusiasm for starting off with the Waterloo & City line but, amongst the issues that make is less attractive as a proposition, is the difficulty of getting the trains back to the surface should early major problems develop. At least, when being the first Underground line with driverless trains, all problems should be rectifiable in the Waterloo depot and there should be no need to bring the trains up to the surface again.

After completion of implementation of the new stock on the Waterloo & City line, the trains will be introduced onto the Bakerloo line – which was never intended to be run without a driver in a cab. Finally, they will be introduced on the Central line and only then will the issue of running automated trains on a major Tube line (complete with outdoor sections) be tackled. The Piccadilly line – the line that was originally to have been first – will now follow. It is thus the repositioning of driverless trains on the Piccadilly from the possible front of the queue (never confirmed) to the back of the queue that really is the big change.

The major consequence of this revised plan is that there should be no further delay getting the trains in service on the Piccadilly line and Bakerloo line. These two stocks are already over forty years old and both are going to need more major refurbishment – the Bakerloo line in particular – just to get them to last until the New Tube for London comes along.

The price to be paid is employing drivers on the Piccadilly line for a lot longer than intended. This will probably cost London Underground around £30m per year for around 15 years, but this will be partially offset by earlier-than-otherwise-achievable automation on the Central line. The cost, though huge, is relatively small in relation to the entire New Tube for London plan. It also greatly reduces the element of risk as London Underground gets the best part of a decade to draw on the experience of automation on the Waterloo & City line before introducing it elsewhere.

A further potential benefit of introducing automation on the Central line before the Piccadilly is that the Central offers far more opportunities for introducing temporary shuttles, possibly off-peak only, for trying out or progressively implementing driverless operation. Another advantage is that the issue of joint running with the Metropolitan between Rayners Lane and Uxbridge does not need to be addressed for a further few years.

Signalling Progress

One of the very positive signs of New Tube for London progressing forward is seeing these papers presented as a matter of urgency and the reason behind them. The “Invitation to Negotiate” (ITN) has already been issued for the trains and now it is said that the signalling is on the critical path, and that there is a requirement to issue an ITN for the signalling.

Whilst the list of signalling option providers available for the SSR was somewhat small, the potentially massive and prestigious project to resignal the deep Tube lines should attract all the main players. Thales, with its TBTC system on the Northern and Jubilee lines and being installed on the sub-surface lines, is probably the safe but uninspiring bet. Siemens, no longer tied down by Crossrail or ERTMS development (on Thameslink in particular) and with, hopefully by then, an extremely good track record, will almost certainly become another contender. One would imagine that Bombardier would have a lot of reputational damage to overcome and would be very much an outsider. The joker in the pack could well be Hitachi, who have spent years trying to understand the UK rail market and are only now beginning to enjoy some successes. Their determination to be a major world, and especially European, player in both the train and the signalling market may well see resources made available that few would have expected a few years ago.

More costly than Crossrail?

Lord Adonis (and others) have in recent months talked about South London and the need to inject money “the equivalent of a Crossrail” to improve and upgrade existing services. This is talked about as if it were a revolutionary statement. In fact New Tube for London has an estimated final cost of £16.5bn. With Crossrail billed on the BBC as the £15bn railway it is actually cheaper than New Tube for London from a certain point of view, although these figures appear to be the expected outturn costs separated by a few years, so it is perhaps an an overly crude cost comparison.

A further reason why NTfL could be regarded as very much more expensive than Crossrail is that there are all sorts of funding options for Crossrail and, more especially, Crossrail 2. New Tube for London is essentially an extreme case of asset renewal and it is unlikely to attract funds from developers, oversite developments, section 106 grants or other forms of revenue associated with new projects.

So, at the other extreme, one could argue that a lot of the NtfL programme is simply about replacement with modern best practice and shouldn’t really be considered as investment at all. The programme is more than just replacement though and the increase in capacity, on the Piccadilly line in particular, really will come from a substantial infrastructure and rolling stock investment programme that arguably deserves to be placed in the same category as Crossrail or Thameslink.

Details trickle out

The latest papers, the most detailed of which is the one presented before the Finance & Policy Committee do not really tell us anything new about NTfL, but we do see some items confirmed in writing that were already known.


For the first time, we have written confirmation that NtfL will include uprating the voltage on the included lines from a nominal 630V to 750V. This will help reduce transmission losses and also mean more power can be applied to the motors. As transmission losses are inversely proportional to the square of the voltage and are significant on the Underground, this will have a beneficial effect that is higher than a 19% increase in voltage would suggest. In simple terms this one improvement produces three benefits – more power available to traction motors, a reduced energy bill due to less electricity being wasted in transmissions and reduced heat output.

Piccadilly line to Ealing Broadway

Part of the NtfL package involves the Piccadilly line taking over the District line to Ealing Broadway. We covered this in considerable detail in Upgrading the Piccadilly. TfL have been very strangely reluctant to confirm this, but many of their documents allude to it without explaining the exact intentions. The briefing for the Rail & Underground is no exception and the diagram refers to the “West London service change” whilst the text gives no indications as to what that is. It was unfortunate that the Rail & Underground panel decided, quite unnecessarily, to cover the entire NTfL topic under “part 2” arrangements which meant that the public were excluded, so even non-confidential issues were hidden from public view. They were probably entitled to do this, but it was hardly in the spirit of open democratic accountable governance. It meant that yet again any details of this proposal failed to come out into the public domain.

To actually see what the West London service change is, you actually have to look at Lord Adonis’s report Transport for a World City published by the National Infrastructure Commission (NIC) which feels obliged to mention it when it states:

Piccadilly line – A peak service level of 33-36 tph, with air-cooled, walk-through Underground trains, by 2025, over the current line geography, and possibly the Ealing Broadway branch currently served by the District line.

Unfortunately some of the dates and other details elsewhere are inconsistent with the briefings for various panels and the TfL board which suggests that, when it comes to NTfL, the NIC report cannot be relied on entirely for up-to-date exact detail.

Overcrowding Diagram

Forecast AM peak overcrowding 2031 – most severely affected Underground lines

What is also not apparent is why it is thought necessary to transfer this section of the Underground to the Piccadilly line. This was explained in a previous article, but it has since emerged that part of the rationale is to enable more District line trains to serve inner stations on the District line’s Wimbledon branch. The Adonis report has highlighted this on an overcrowding diagram. The overcrowding on this section of the District line is not expected to be resolved by Crossrail 2.

The curious omission of the Bakerloo line to Lewisham

The various panel and board reports on NTfL also fail to refer to the Bakerloo line extension to Lewisham. This is almost like reading about plans for the Northern line without having a mention of the extension to Battersea Power Station – but the Northern line extension is being built now whereas the Bakerloo line extension is currently an aspiration.

This omission is strange, as surely it must feature in the plans and be a major factor in constraining implementation options – as any option that does not provide new trains for the Bakerloo line and resignal the line prior to 2030 would suggest that the NTfL team do not sincerely believe that the Bakerloo line extension will happen. This despite it being very much the policy of the current Mayor and almost certainly that of any future one as well.

Stranger still is the Finance & Policy Panel report that states:

the order of the Central and Bakerloo modernisations has been reversed. With commitment of additional investment on existing Central line assets to secure their continued safety and reliability, the Bakerloo line modernisation has been prioritised ahead of the Central line to accelerate the replacement of the oldest trains on the network which are operating on the Bakerloo line. It is intended that the relative order of these two lines in the NTfL delivery sequence be kept under review, informed by emerging asset condition and available funding.

No mention whatsoever about prioritising the Bakerloo line ahead of the Central line so that the Bakerloo line upgrade is complete in time for the opening of the Lewisham extension.

Northern line

Finally, we keep mentioning this but, if the NTfL programme is serious and as future-proofed as it is supposed to be, it must surely follow that part or all of the Northern line should logically follow in from the current proposed end of the programme (around 2037) in time to replace at least some of the Northern line trains by 2040. There is probably an understandable desire not to formally include this in the NTfL programme – not least because it will considerably increase costs – but it surely must make sense to make some provision for this – if only to make sure that the contract for the trains includes options to supply further trains at a later date for another line, should the client wish to do so.

Lining Up the lines in a line

What us becoming clearer is how budgetary constraints are forcing a sequential upgrade of individual underground lines. The cost of any delay is going to have ramifications for years to come, with the first possible opportunity for a breather and a bit of slack sometime in the 2040s. The London Assembly is right to be concerned with money wasted on failed contracts, but the consequential delay may turn out to be an even bigger concern for future generations.

Written by Pedantic of Purley