Purple Reign – How Crossrail Will Be Run


There was a time when your humble Editor worked in the Civil Service. From this period of personal history two clear memories stick out. Firstly that there was once a genuine publication titled Guidance on Guidance, a document of which Sir Humphrey would likely have been proud. Second was the lesson that the more innocuous a committee’s name, the more interesting and important the topics it discussed were likely to be.

This second lesson certainly applies to the TfL Finance & Policy Committee. It was a rare emergency meeting of this body in December that provided the first public hint that major changes were coming to the Sub-Surface Signalling contract and the Committee’s papers, which TfL generally make available on their website, always provide a welcome glance into the current status and future thinking on key projects.

This is particularly true for the upcoming committee meeting on the 17th July, the papers of which have now been made available online. Traditionally we would cover all of the topics to be found therein in one post, but this month sees a raft of key projects come before the committee – Crossrail Franchising, Future Ticketing, The Northern Line Extension and the aforementioned SSL Signalling contract.

We begin in this post with one of the most interesting – the supporting paper that provides a rather fascinating insight into just how Crossrail will be run.

How it unfolds

In March 2013 Rail for London (RfL), TfL’s surface rail arm, issued a tender notice in the OJEU seeking an operator for Crossrail’s services. The papers finally provide some insight into what that operator will be… well… operating, and how.

Crossrail train services will be opened in a staged manner:

  • May 2015. Stage 0: Liverpool Street – Shenfield services will be taken over and operated by the existing Class 315 rolling stock.
  • May 2017. Stage 1: The Class 345 rolling stock will be gradually introduced into passenger service.
  • May 2018. Stage 2: The Crossrail concessionaire will begin operating 4tph between Paddington and Heathrow Terminal 4, replacing Heathrow Connect and part of the current Great Western inner suburban service.
  • December 2018. Stage 3: The formal introduction of Crossrail services on the Central Operating Section from Paddington – Abbey Wood.
  • May 2019. Stage 4: The formal introduction of Shenfield – Paddington services through the Central Operating Section.
  • December 2019. Stage 5: Full operation with through services.

Track and Trains

As with the Overground, Crossrail will run over Network Rail rails, in addition to the Central Operating Section (of which TfL will both retain ownership and duty of maintenance). This means track access fees. Technically the operator will pay these, but in reality they will be passed on as-is to TfL – nicely removing that element of operator risk from the contract. Just how track access works is actually detailed in a second document, and is worthy of discussion in its own right, which we’ll do in a future article.

Ahead of the arrival of the new rolling stock, the concessionaire will operate 44 class 315s leased from Evershot Rail. In a piece of joined up thinking (of which Mwmbwls would be proud), these will be maintained at Ilford Depot by the concessionaire of TfL’s new Greater Anglia Franchise. Operation (and obviously maintenance) of these will be phased out as the new Bombardier 345s arrive and are commissioned into service, along with their new depot facilities.

Learning from the Overground

Whilst the above is all important, what will be of most interest to travellers and industry watchers alike will be how Crossrail will be operated. On that note, it will come as a surprise to no-one that the Crossrail Concession will be closely modelled on that which TfL have used successfully on the Overground.

We have looked at just what this is and why this works so well in London before, but in short RfL will set the service level they require and the operator will be paid to deliver it, regardless of what is collected via the farebox, with deductions made for failing to meet performance targets. This accomplishes two goals.

Firstly, it removes revenue risk completely from the contract. The operator gets to calculate far more accurately how much money it is going to cost to run, and thus a big element of risk that they would previously have passed on in cost terms to TfL is removed.

Secondly, this keeps the operator focussed on delivering services that are tailored to meeting the needs of commuters and general travellers, rather than optimising towards higher fare payers. The Concession will also be based on a deduction-only model – i.e. the assumed baseline for delivery is 100% with all failures to meet that deducted from the payment made, rather than a model that sets a lower target (e.g. “99.4% of trains must be on time”) and then rewards for going above that. All deductions (and indeed payments in general) will be adjusted against RPI.

The available papers rather nicely lay out what performance measures the operator will be judged on, providing an interesting insight into current franchise thinking within TfL. Broadly speaking, the measures are broken down into three categories – Headway, Delay and Capacity.

The Headway Measure

Headway is the temporal gap between a train and the one preceeding it. On Crossrail, every train which passes through the Central Operating Section between Paddington and Whitechapel will be subject to the Headway Measure. If that train exceeds by 40% or more the timetabled headway between services then the operator will be fined.

40% may initially sound like a generous measure, but it is worth remembering that for trains on the central section, particularly during peak hours, that timetabled headway is going to be very small – often around 2.5mins. This means that every train which is a minute late through the central section will be subject to a fine.

The baseline fine is currently set at £150 per minute, rounded to the nearest second, per train. When the headway increase is the result of an identified delay, that figure will be modified by whose fault the delay is – if it is the fault of the operator then they’ll pay 110% of that baseline figure. If it’s someone else’s fault (e.g. Network Rail) then the operator only pays 10% of the baseline figure.

Handily, TfL include some examples of how this might play out in reality:

a peak service operating 6 minutes headway at a recording point due to a passenger action where timetabled headway is 2.5 minutes – in this example, the delay is attributed to the Operator who incurs a headway deduction of £577.5; that is, 3.5 minutes over the timetabled headway of 2.5 minutes multiplied by £150 per minute plus ten per cent.


three peak services operating 3.5 minutes headway at a recording point due to a fault with the train doors where timetabled
headway is 2.5 minutes – in this example, delay is attributed to the RSP but the Operator incurs a headway deduction of £45; that is 3 lots of 1 minute over the timetabled headway of 2.5 minutes multiplied by £150 per minute multiplied by 10 per cent.

Including the Headway Measure makes sense when you consider what one of TfL’s biggest concerns with Crossrail is likely to be, which is that Crossrail is effectively a service that has to mix both metro and regional services. TfL are clearly looking to ensure that the metro element – close headways through the central section – is taken very seriously indeed. This is necessary to minimise the risk of overcrowding on platforms and to cater for the fact that in the Central Section it effectively needs to be “like the Underground” as that’s how people interacting with it there will treat it.

In this context, headway is really important as passengers don’t judge the Underground’s performance on how strictly it runs to its timetable (although one naturally exists in the background), they judge it on how long they have to wait until the next train.

All the above is obviously good news for commuters, for it means that any excess minutes (or, in the peak, just the first excess 31 seconds as minutes are rounded to the nearest one) they are left standing on the platform are costing Crossrail’s operator money. Proactive and flexible thinking about the service pattern every hour of every day is thus not only encouraged, it is effectively (thanks to the deduction-only contract) mandatory.

The Delay Measure

Whilst the Headway Measure is good news for commuters, it does naturally raise the question as to what steps will be taken to ensure that services aren’t overly optimised towards performing well in the Central Operating Section to the detriment of services on the more regional sections of the line.

To help mitigate the risk of disruption and delay in this context is the Delay Measure. This mandates that the operator will be charged £125 per minute (again indexed at RPI) based on lateness at destination of 3 minutes or more, measured to the nearest second at the terminating point for each service (i.e. any single point to point journey). As with the Headway Measure, this baseline figure will be modified based on whose fault the delay is – 110% If it’s the operators, 10% if it isn’t.

It may seem harsh to punish the operator at all for delays that are outside of its control, but as TfL explain:

This is driven by experience in similar arrangements in TfL which suggest that while the Operator is not directly in control of certain events which impact performance, it does have the ability to significantly influence the impact of them on the railway.

In other words it’s the response to the delay that matters – as that’s ultimately what affects passengers, not the cause. You thus punish for every delay because that way you incentivise a swift return to as good a service as possible, as quickly as possible.

Again, TfL include a couple of handy examples:

an incident caused by the Operator results in a service arriving 3 minutes late at destination – the Operator would incur a delay deduction of £412.5 in this example; that is, £125 per minute multiplied by 3 minutes lateness plus ten per cent.


an incident caused by the RSP results in 3 services arriving 3 minutes late at destination – the Operator would incur a delay deduction of £112.50 in this example; that is, £125 per minute multiplied by 9 minutes lateness multiplied by 10 per cent. The RSP would incur deductions under its own performance regime.

The Capacity Measure

The final major category of performance measures against which the Crossrail operator will be judged is the Capacity Measure. This is intended to, as much as possible, prevent the operator from resorting to running services “fast” (i.e. not stopping at advertised stations) in order to quickly and easily avoid falling foul of headway or delay issues.

Interestingly, it seems that stations have been assigned specific financial values – effectively an “importance” rating calculated on the basis of how often they are stopped at in the timetable and their passenger volumes. Sadly the paper doesn’t include a breakdown of what those values are other than a few specific examples (see below), but hopefully a detailed list will appear at a later date.

Those financial values again act as a baseline penalty – missing a stop once results in the corresponding fine value. Missing it with the subsequent timetabled service as well results in that subsequent miss being fined at five times the station value.

TfL’s summary of how the Capacity Measure works is quite succinct, so it seems simplest to repeat it here:

The capacity deduction the Operator incurs will be the aggregate of the station values for each missed station stop versus scheduled station stops on each single timetabled journey, therefore ensuring that the Operator is always incentivised to maximise the number of scheduled station stops. The value of the capacity deduction that the Operator may be charged for a full cancellation ranges between £10,000 and £36,000 depending on the scheduled station stops to be made on any timetabled journey.

Again, they also handily provide examples:

a westbound service scheduled to arrive at Tottenham Court Road at 08.00 but misses Forest Gate before 7.45 – the Operator would incur a capacity deduction of £1,260; that is a station value of £36 or Forest Gate multiplied by a station stop factor relating to first station missed on the service of 35.


two westbound services scheduled to arrive at Tottenham Court Road at 09.30 and 09.35 both miss Stratford – the Operator would incur a capacity deduction of £2,450 for the first service; that is, £70 station value for Stratford multiplied by a station stop factor relating to first station missed on the service of 35. The Operator would also incur a £12,250 deduction for the second service; that is £70 station value for Stratford multiplied by a station stop factor relating to first station missed on the service of 35, multiplied by the successive station stop factor of 5

Other performance measures

Whilst headway, delay and capacity form the primary measures against which performance will be judged, the papers suggest that there will be others. Given the emphasis placed on station environment and staffing on the Overground, it can be assumed that these relate more specifically to judging those. TfL will own outright the four new Crossrail stations on the Central Section and these will be leased back to the operator, who will also take over 24 more stations under lease from Network Rail. As now seems to be standard practice for TfL concessionaires, on taking over those stations the operator will be required to conduct a full deep-clean and freshen them up visually/rebrand – as well as man them on an ongoing basis.

Drawing Conclusions

The combination of headway, delay and capacity measures – whilst not unique – does seem to represent a good approach to seeing Crossrail managed in a way that accommodates the challenges that are going to be inherent in running a line that has to combine both metro and regional services.

That Crossrail would lift heavily from the operating practices of the London Overground was always something of a no-brainer. That card was marked when Howard Smith was appointed as Director of Operations. As we noted at the time:

The key words in the [announcement of his appointment] are, of course, “Crossrail operating concession.” Smith has been instrumental in making London Overground the success it is, both from a passenger and a franchise perspective.

Indeed his name will be familiar to anyone who has followed our coverage of the Overground over the years, as he has consistently proven to be a figure who gets the importance of both urban and suburban rail, and has been a vocal proponent of taking the concession model forward. He’s also continually demonstrated a thorough knowledge of the practicalities and possibilities of rail work both north and south of the river. In that regard, Crossrail’s gain is very much TfL’s loss.

What is interesting here though is that Smith and his team have clearly realised the importance not only of capturing the successes of the existing model, but of trying to develop it out to cater for the unique challenges Crossrail presents as well. The triple approach of headway, delay and capacity could be an interesting solution to the issue of managing a railway of such occasionally competing demands, and one suspects that an awful lot of modelling has taken place behind the scenes to try and find a punitive balance between them that will give the operator room to breathe, whilst also keeping them on their toes.

If it works – and given the track record of both TfL and Smith himself there’s reasonable grounds for optimism – then it will be interesting to see whether this approach gets used elsewhere. For there is no denying that one of the big questions still to be answered when it comes to railway franchising is how you incentivise operators to balance the needs of both commuters and distance travellers. If the Crossrail concession is thus successful, then it could have some major ramifications indeed.

In our next piece on the Finance and Policy Committee, we’ll take a look at the Northern Line extension to Battersea

Written by John Bull