Friday 21 December 2012

The real West Coast scandal

The West Coast Main Line: it seems disruption is for life, not just for Christmas. Photo: Network Rail
It wasn’t quite a Mayan prophecy, but no doubt a few pundits were left somewhat disappointed when the rail franchising Armageddon of December 9 failed to materialise. Despite the many column inches devoted to the undoubtedly serious flaws in the Inter-City West Coast refranchising process, there was never any genuine risk that services on a network carrying 30 million passengers per annum would simply cease because of a foul-up of the functionaries.

And so it proved. The trains are still running, and Virgin Rail Group has a short-term deal to continue to operate ICWC services for a further 23 months. Passengers could be forgiven for breathing a sigh of relief as they prepare for the big Christmas getaway...

Or not. Because (and do please forgive the London broadsheet hacks for failing to notice this), ICWC trains do (effectively) stop running on December 23 for five days over the Festive period. Sure, there will be a limited rail service linking London with Nuneaton via a circuitous route along the Chiltern main line, but no direct services will run between Euston, Manchester, Liverpool and stations to Glasgow; ironically Chiltern Railways services between London Marylebone and Birmingham Moor Street may also be affected by the need to share tracks with Virgin.

The reason for the hiatus is, of course, engineering works – all four West Coast Main Line tracks through Bletchley will be closed for resignalling. Infrastructure manager Network Rail’s desire to perform work at Christmas is perhaps understandable, as there is less commuter traffic generally and no trains at all on December 25-26. But the Bletchley blockade lasts much longer, seriously hindering leisure travel at one of the busiest times of year, and it isn’t a ‘bog standard’ track possession, it is one of a panoply of work packages deferred from the infamous West Coast Route Modernisation Programme undertaken between 1998 and 2009.

As regular readers of this blog will already be well aware, the project was effectively halted prematurely to ensure the outturn cost was kept below the symbolic £10bn mark. Nobody now disputes contractor Bechtel’s assessment that the true capital cost of WCRM was in the region of £13bn. Note that this does not include the wider economic disbenefits caused by a decade of disruption; this sum has never been meaningfully assessed. My own view is that the value of leisure travel in a service-dominated economy is seriously underestimated by transport economists.

But on the West Coast Main Line, it seems disruption is for life, not just for Christmas. This blog has remarked many times that the institutional unreliability of Europe’s busiest mixed-use railway utterly undermines the claims by opponents of High Speed 2 that more services could be accommodated on it. Now these concerns have been confirmed in a landmark report authored by Virgin Rail Group COO Chris Gibb and issued on behalf of the operator and Network Rail.

Mr Gibb outlines a series of interrelated challenges which impact on Network Rail’s ability to achieve an acceptable level of reliability on the busiest London – Rugby core. Among the specifics highlighted are the growing incidence of trespass and suicide, where the report points to Réseau Ferré de France data showing that fewer than 3% of suicides on the French network occur on Lignes à Grande Vitesse, where trespass risk can be modelled at the design stage; an unintended benefit of having fewer platform faces of course.

The difficulty in gaining access to the WCML for maintenance is also addressed:

'The section between Watford and Euston is some of the most difficult to maintain and ageing infrastructure, passing through an urban area which limits access to and alongside the railway whilst influencing the railway with earthworks issues, trespass and other ‘neighbour’ issues. The current possession arrangements are barely enough to hold the infrastructure in its current condition, which in turn is not good enough to sustain good performance.'

An examination of overhead power supply equipment revealed a yet more damning indictment of the ‘incremental upgrading’ approach:

'It appears that the West Coast Route Modernisation project team were more focussed on within budget/on time delivery of the project, than the medium/long term component performance, and this approach has clearly cost NR and the industry dearly in terms of poor performance.'

You get the idea.

It is abundantly clear that many millions of pounds will be needed just to keep the southern WCML in a fit state to handle its current workload, let alone the several extra 125 mile/h services per hour envisaged by the 51m Group under its alternative to HS2.

There will inevitably be those who, come what may, will insist that interminable spending on route upgrading is always the right policy. But global best practice suggests otherwise: it is generally under-appreciated that two of the world’s most commercially-successful high speed rail projects, Japan’s pioneering Tokaido Shinkansen between Tokyo and Osaka, and the Paris – Lyon line in France, were both authorised after the respective governments concluded that upgrading the conventional routes in each case would be too costly for the accrued benefits[1].

Closer to home, we can hardly dismiss the south end of the WCML as an obstreperous one-off. As this week’s multiple overhead power failures at Hitchin illustrated, the East Coast Main Line from London King’s Cross is similarly fragile despite a series of infrastructure upgrades over the past three decades, whilst readers of industry newsletter Rail Business Intelligence will already be aware of insiders’ mounting concerns about the costs and benefits of the London – Cardiff route modernisation, now priced at around £7bn.

The interrelated concerns about capacity, reliability and service patterns on the UK’s principal rail axes will not go away, even as memories of ‘Franchisegate’ fade. Over more than four decades, the alternatives to a new line to link the capital with our most important provincial centres — 14 of which would be served by trains using HS2 under current plans — have been tried repeatedly. Thanks to the Gibb report, we now have empirical proof of the limitations of ‘patch and mend’.

The real West Coast scandal is about infrastructure, not franchising.
-------------------------------------------------------------------------------------------------
1. For more details on the early years of high speed rail development, see Hughes M., Rail 300: The World High Speed Train Race, David & Charles, Newton Abbot 1988

13 comments:

  1. Technical failures, and the need for maintenance, are not restricted to legacy networks - as Eurostar / HS1 / LGV Nord attest.

    Remember this, for example?

    Is the best way of boosting Britain's transport resilience to build new high-speed rail lines? I don't see any evidence.

    The country seems to have largely lost its ability to design or engineer anything properly. Without competent people, the future looks bleak.

    ReplyDelete
  2. Many thanks Nick for this timely reminder of the limitations posed by a patch and mend strategy.

    Yes, of course those hostile to HS2 (largely driven by their own self-interest agenda?) can highlight isolated incidents of breakdown on new infrastructure but this viewpoint deliberately sidesteps the main issue, which is the frequency of maintenance required on existing lines, in contrast to new infrastructure.

    Let's face it, the existing WCML is a legacy from the Victorian era of pioneer railway building - long may it continue to serve but after nearly two centuries, it is fraying around the edges (an understatement if ever there was one!) - given a choice between patch and mend and build anew, the ordeal undergone during the WCML upgrade debacle provides a salutory lesson in how NOT to implement step change improvements in the UK's rail network, given a choice.

    With High Speed Rail the UK has a choice - I say let's go for the bold, innovative approach rather than the existing, proven to fail, patch and mend strategy!

    ReplyDelete
    Replies
    1. The vast majority of the railway mileage in western Europe is a legacy from the Victorian era.

      Most expenditure in the last hundred years has been on replacement of worn-out / obsolete assets, not new-build lines. Or as Mr Padav would call it, "patch and mend".

      In Britain, mismanagement of large rail projects is the norm (see Intercity Express Programme, Edinburgh Trams, West Coast Refranchising, etc). The resource waste tends to get bigger with the size of the project.

      Compared to HS2, the cost of adding capacity on the existing West Coast and Chiltern routes is chicken feed.

      Mr Padav's claim that the WCML modernisation "effectively threw the best part of £9bn of taxpayers money down the pan for very little improvement in terms of capacity, reliability and speed" has no factual basis.

      According to Bechtel, "Capacity on the West Coast Main Line was increased by 60,000 seats per day, with more than 1,000 new train services per week. The modernization also reduced travel times along the lines significantly, cutting 43 minutes off the trip between London and Manchester, and nearly an hour off the journey between London and Glasgow."

      Recasting the Birmingham - Euston frequency from two to three trains per hour, increased Inter City West Coast seats between those points by 50% at a stroke. And Pendolino lengthening (to 11 carriages) increased ICWC second class seating by another 50%.

      Delete
    2. "Compared to HS2, the cost of adding capacity on the existing West Coast and Chiltern routes is chicken feed."

      Yes, I seem to recall scrutinising your proposals for doing this on the Chiltern Line. As I remember, it involved disregarding the £8bn price tag envisaged by Atkins (in Scenario C), and instead claiming that only two small sections of four-tracking were needed, with no arguments offered as to why the rest of the improvements were unnecessary. (There was an argument as to why four-tracking Banbury-Oxford was unnecessary, which I agree with, but that wasn't in the list of improvements.)

      For pity's sake, it was exactly the rose-tinted vision of WCML upgrade being a quick, cheap and easy improvement (albeit with different motives) that got us into this mess in the first place. That is why I am extremely suspicious of anyone claiming quick and easy solutions now, especially ones coming from anonymous bloggers.

      Delete
    3. Atkins' (anonymous) creators of the High Speed 2 Strategic Alternatives stated that their March 2010 Report identified "additional interventions required to improve the rail offering, as an alternative to constructing a High Speed Rail Line*."

      The Atkins proposals focused on speeding up existing lines, as an alternative to building an all-new line. But if capacity, not speed, is the issue, then neither HS2 nor Atkins' Scenario C Banbury by-pass (etc) would be necessary. Lyon to Paris takes 2 hours by TGV, but I don't see the French government prioritising getting that down to 45 minutes.

      The West Coast Route Modernisation 'completed' in 2008 was a scope-limited modernisation, not an upgrade. How much it would cost to 'add capacity' on a particular railway, depends on the amount of capacity being added. In the 1950s, locomotive-hauled 11-carriage trains operated on the Chiltern Main Line. A reminder that *modernisation* is not synonymous with *upgrade*; in the 1960s and 1970s, Chiltern was modernised by shortening trains, partial line singling, and removal of fast tracks at stations (i.e., downgraded).

      All railways need to be modernised from time to time. In France, Germany, and Italy, such work tends to be less disruptive. They had no equivalent of the Beeching cuts, and have more diversionary routes available.

      Delete
    4. "The Atkins proposals focused on speeding up existing lines, as an alternative to building an all-new line."

      Sadly, no. The exact terms of reference for the Chiltern Line improvements are: "Euston – Birmingham services diverted to the Chiltern Paddington – Birmingham route;" and "This package includes infrastructure enhancements to enable the Chiltern route to maintain existing
      London – West Midlands journey times, safeguarding existing benefits for passengers." The focus is on capacity. Journey time improvements are just a bonus (and not a terribly impressive bonus either: 73 minutes Ldn-Bhm instead of 84).

      Even if you can demonstrate the Banbury by-pass is unnecessary if you just focus on capacity and ignore speed, there's still masses of other expensive improvements in the list. Proving one improvement might not be necessary is not proof that you solve the WCML capacity problems for "chicken feed". You will have to either produce a service pattern showing this is possible with the rail improvements you propose, or point to a similar railway elsewhere in the world that offers a similar level of service (or, at the very least, point to a rail expert arguing this who people might take seriously). Otherwise it's just an empty sales pitch.

      "In France, Germany, and Italy, such work tends to be less disruptive. They had no equivalent of the Beeching cuts, and have more diversionary routes available."

      True, but you're missing one important aspect: France, Germany and Italy all have high-speed lines adding to the total number of railways available. Notably, these lines duplicate existing routes. Yes, duplication is a good thing. If one line has to close, you can send passengers down the other (unless both lines are close to full already, in which case you need to consider triplication). We don't know if they would cope so well without their HS networks. My suspicion is that lack of Beeching cuts alone wouldn't have had the same effect.

      I'm open to suggestions for re-opening lines axed by Beeching as a solution the WCML capacity problems (the freight line idea was a reasonable one), but the current anti-HS2 logic seems to be re-opening lines that help very little or not at all. Suffice to say that re-opening the Oxford-Cambridge line won't help if you're travelling between Rugby and London. Line closed at Watford Junction, can't commute to London from Milton Keynes? No problem! Hop on a train to Cambridge instead! I think not.

      Delete
    5. The February 2011 Strategic Alternatives report stated that

      "Atkins was appointed by the Department for Transport (DfT) in August 2009, to consider road and rail improvement alternatives to the High Speed Rail proposition, being developed by HS2 Ltd. The main objective of HS2 Ltd at the time was to consider the case for new high speed services between London and the West Midlands."

      From the beginning, very high speed has taken priority over everything else, capacity included.

      The Atkins Scenario C proposal would not "maintain" the existing London - Birmingham rail travel time; it would reduce it to 73 minutes. The marginal costs of that time reduction are enormous, and there is no capacity case for a Saunderton – Seer Green tunnel, or Banbury by-pass.

      Scenario C reserved two Chiltern paths per hour per direction for freight, but there is no particular reason why Southampton goods trains couldn't be routed via the Cotswold Line.

      Chiltern operated with 12-vehicle formations in the 1950s, and does not require £8 billion of spend to restore such capability.

      With the construction of a building a Chalton - Ridgmont or similar connection, the Varsity Line could have been quite useful for Milton Keynes commuters. The obvious difficulty is that St Pancras has been largely given over to HS1.

      Delete
    6. The problem with your theories is that is relies on two major assumptions:

      1) Journey time and (train) capacity improvements are separable; and
      2) If you do just the train capacity improvements, the cost will only be a fraction of the Scenario C amount.

      The reality is that often speed improvements and capacity improvements are the same thing. You can look on four-tracking as a way of preventing fast services getting stuck behind stopping services, with a bonus of improving capacity. Or you look on four-tracking as way of preventing local services having to skip stops to make way for faster services (as is happening on the Chiltern Line now), with added capacity as a bonus. But in practice, what's the difference?

      This is the case with both of your examples. A Banbury By-Pass both improves journey times and eliminates a constraint on capacity with a occupied platforms and a flat junction. The Saunderton - Seer Green tunnel I presume is on the list because there's no other conceivable way of four-tracking through High Wycombe.

      As I have said several times before, you are welcome to come up with your own proposals stating what improvements will be made, how much you think it will cost, and how it will be possible to run the services. So far, you haven't. Arguing that 12-car trains ran in the 1950s is useless - it is train capacity, not seat capacity, that is the problem if you're adding another 3tph to a line. (And, frankly, if you are still confusing passenger capacity and track capacity this way, no-one's going to believe you're some sort of rail expert.)

      Finally, on the Calton-Ridgmont proposal, that's actually not a bad idea for a line. A Luton-MK service would be a good thing in its own right, and I imagine you could do this by extending some Luton-terminating Thameslink services to Milton Keynes. With the MK situation getting desperate pre-2026 with or without HS2, we need to consider every option going.

      But it's quite wrong to suggest it's HS1 standing in the way of this, however much you and other antis might want to scapegoat the line for every problem on the railways. With local services now running through the station on a ultra-high-frequency line, the smaller number of MML platforms isn't that much of an issue (and no issue at all if you're extending existing Thameslink services). The biggest constraint at the moment is track capacity on the way to London, not how big the station is once you've got there.

      As for a WCML substitute when the line is closed, don't get too excited. You might be able to divert some trains down the line, but with the four-track WCML barely coping with its current load there's no chance of pouring all of this on to the four-track MML on top of its regular services. There's no getting round the ECML, MML, WCML and Chiltern Line are all busy enough as it is. If you really want to solve the problem through reverse Beeching, you'll need to find some lines into London to re-open.

      Delete
  3. "Arguing that 12-car trains ran in the 1950s is useless - it is train capacity, not seat capacity, that is the problem if you're adding another 3tph to a line."

    Presumably, you mean track capacity. It does not follow that moving West Midlands intercity traffic to the Chiltern Main Line would add three trains per hour to it. There is already one short-length fast passenger train each hour between London and Birmingham on that route. Atkins Scenario C reserves 2 paths per hour for goods, but there is no good reason to route Southampton freight via Chiltern.

    There's no getting around the fact that trains on the Chiltern line are much shorter than on the West Coast, Midland, or East Coast Main Lines, making their passenger capacity much lower. Path utilisation is not particularly good either, but that's true of many lines north of London.

    As previously mentioned, the Chiltern cut-off was built with separate through and platform tracks at stations. At High Wycombe, capacity was taken out by British Rail, and the remaining tracks re-aligned to slightly increase linespeed.

    So, capacity went down, and speed went up. The reality on the Chiltern Main Line is that speed improvements and capacity improvements are not the same thing. In Atkins' proposed Scenario C, most of the spend is targeted on journey time, not capacity.

    Chiltern doesn't have country stopping services; with the exception of Kings Sutton, village stations closed in the 1960s. At Banbury, there are crossovers, just as there are on other lines, but it's not a junction.

    For efficient path utilisation, the London Chiltern commuter trains would need to run at intercity speed (up to ~200 km/h) on shared track sections.

    ReplyDelete
  4. "Arguing that 12-car trains ran in the 1950s is useless - it is train capacity, not seat capacity, that is the problem if you're adding another 3tph to a line."

    Presumably, you mean track capacity. It does not follow that moving West Midlands intercity traffic to the Chiltern Main Line would add three trains per hour to it. There is already one short-length fast passenger train each hour between London and Birmingham on that route. Atkins Scenario C reserves 2 paths per hour for goods, but there is no good reason to route Southampton freight via Chiltern.

    There's no getting around the fact that trains on the Chiltern line are much shorter than on the West Coast, Midland, or East Coast Main Lines, making their passenger capacity much lower. Path utilisation is not particularly good either, but that's true of many lines north of London.

    As previously mentioned, the Chiltern cut-off was built with separate through and platform tracks at stations. At High Wycombe, capacity was taken out by British Rail, and the remaining tracks re-aligned to slightly increase linespeed.

    So, capacity went down, and speed went up. The reality on the Chiltern Main Line is that speed improvements and capacity improvements are not the same thing. In Atkins' proposed Scenario C, most of the spend is targeted on journey time, not capacity.

    Chiltern doesn't have country stopping services; with the exception of Kings Sutton, village stations closed in the 1960s. At Banbury, there are crossovers, just as there are on other lines, but it's not a junction.

    For efficient path utilisation, the London Chiltern commuter trains would need to run at intercity speed (up to ~200 km/h) on shared track sections.

    ReplyDelete
  5. "Arguing that 12-car trains ran in the 1950s is useless - it is train capacity, not seat capacity, that is the problem if you're adding another 3tph to a line."

    Presumably, you mean track capacity. It does not follow that moving West Midlands intercity traffic to the Chiltern Main Line would add three trains per hour to it. There is already one short-length fast passenger train each hour between London and Birmingham on that route. Atkins Scenario C reserves 2 paths per hour for goods, but there is no good reason to route Southampton freight via Chiltern.

    There's no getting around the fact that trains on the Chiltern line are much shorter than on the West Coast, Midland, or East Coast Main Lines, making their passenger capacity much lower. Path utilisation is not particularly good either, but that's true of many lines north of London.

    As previously mentioned, the Chiltern cut-off was built with separate through and platform tracks at stations. At High Wycombe, capacity was taken out by British Rail, and the remaining tracks re-aligned to slightly increase linespeed.

    So, capacity went down, and speed went up. The reality on the Chiltern Main Line is that speed improvements and capacity improvements are not the same thing. In Atkins' proposed Scenario C, most of the spend is targeted on journey time, not capacity.

    Chiltern doesn't have country stopping services; with the exception of Kings Sutton, village stations closed in the 1960s. At Banbury, there are crossovers, just as there are on other lines, but it's not a junction.

    For efficient path utilisation, the London Chiltern commuter trains would need to run at intercity speed (up to ~200 km/h) on shared track sections.

    ReplyDelete
    Replies
    1. Okay, so if I've understood this correctly, you are arguing:

      * You can replace 3tph Ldn-Bhm via WCML with 2tph Ldn-Bhm via Chiltern with the existing fast service serving as the third tph; and

      * By diverting the two Southampton-bound freight paths per hour, we're all square.

      Big flaw with the second point: diverting these freight trains only frees up capacity where we don't need it, between Banbury and Leamington (where no four-tracking is proposed). It is between Marylebone and Bicester where there is a current capacity shortage, which already has relatively little freight on it.

      You might be able to downscale the improvements if you're only adding two fast services into the line instead of the four envisaged in Scenario C. But as people are already complaining about cuts to local services to make way for the existing Chiltern Main Line service, that's a dubious proposition.

      It's all very well citing one example at High Wycombe where a change allegedly reduced capacity but increased speed, but to claim this applies to all rail improvements is highly ropey. But I'm not interested in theories. I want a firm proposal of what rail improvements will deliver which capacity improvements and how, and you still haven't given one.

      The only proposal I have seen which addresses capacity between London and Bicester is the idea of intercity-speed commuter services, and there's all sorts of problems with that. For a start, even after Evergreen 3 the maximum speed is 100mph, so you're not going to get a speed of 125mph/200kmph without even more upgrading. (Even ECML services rarely go over 110mph.) And when the local stations are 4-6 minutes apart (more like 2-4 minutes apart), by the time you factor is accelerating and decelerating it's doubtful whether there'll be any time worth saving at all.

      Again:

      * What improvements do you think are needed (in particular four-tracking)?
      * How much do you think it will cost?
      * What service pattern would you propose to run, including both mainline and local services?

      Repeatedly pointing to spare capacity on the least congested stretch of the route is not good enough.

      Delete
    2. As previously mentioned, the Bicester cut-off was built with separate through and platform tracks at stations. At High Wycombe, capacity was taken out by British Rail, and the remaining tracks re-aligned to increase linespeed.

      So, capacity went down, and speed went up, showing that speed and capacity improvements are not necessarily the same thing. In Atkins' proposed Scenario C, most of the spend is targeted on journey time, not capacity.

      "Arguing that 12-car trains ran in the 1950s is useless - it is train capacity, not seat capacity, that is the problem if you're adding another 3tph to a line."

      It does not follow that moving West Midlands intercity traffic to the Chiltern Main Line would add three trains per hour to it. There is already one short-length fast passenger train per hour on that route. Atkins Scenario C reserves 2 paths per hour for goods, but there is no particular reason to route Southampton freight via Chiltern, rather than on the Cotswold Line.

      There's no getting around the fact that Chiltern line trains are much shorter than on the West Coast, Midland, or East Coast Main Lines, making their passenger capacity much lower.

      Chiltern doesn't have country stopping services; with the exception of Kings Sutton, village stations closed in the 1960s. At Banbury, there are crossovers, just as there are on other lines, but it is not a junction.

      For efficient path utilisation, London Chiltern commuter trains would need to run at up to ~200 km/h on shared track sections.

      Delete