Rail 626: Network Rail shows weakness of high speed rail business case

Network Rail’s analysis of the business case for a north south high speed line looks like a massive own goal. It suggests yet again that its directors have little political nous and have made the classic mistake of commissioning research without knowing the answer they wanted. Not only does the analysis suggest that the case for building the line is at best tenuous, the organisation has not endeared itself to the government through the timing and nature of the publication. The report flatters to deceive, begging more questions than it answers and, if anything, further muddying the waters.

 First, though, there is the rather odd timing and nature of the report. HS2, the government company set up by Lord Adonis to examine the feasibility of a high speed line is due to hand over its report to him by the end of the year, with publication expected very early in 2010. Therefore there was some consternation in government circles about what Network Rail was up to and why the company felt it timely to release this report at this time. Certainly in Whitehall it was felt that this was an attempt by Network Rail to park its tanks on the government’s lawn. There was concern that Network Rail was adding complexity to the situation, and while at the launch, Iain Coucher, the company’s chief executive, assured the audience that there had been cooperation with the HS2 company, there was confusion about the precise nature of the status of the two reports.

 In fairness, Network Rail research was launched last autumn before Lord Adonis had firmed up his support for high speed rail and set up the HS2 team to examine the feasibility, but nevertheless Network Rail, not for the first time, does not seem to have considered the political sensibilities before making high profile public interventions.

Certainly there was no attempt by Network Rail to slip out this report without fanfare. Quite the opposite. The report was issued in August during a quiet news day in order to maximise impact and interest, and was carefully (though badly) stage managed There were weeks of prevarication over the date of the launch (and even then NR managed to forget to invite Rail‘s editor) with several postponements, apparently due to technical difficulties with the presentation which incidentally was appalling. The computer generated images showed a robot type figure having a yellow train built around him whizzing through a landscape reminiscent of the post apocalyptic vision of Conor McCarthy’s The Road, a vision rather like those car ads without people or indeed any human frame of reference. It was charmless and artless, the work of a poorly briefed PR agency commissioned by people who have no understanding that the railway is about serving passengers, not developing technology.

 Then there was the secrecy. I was inundated with calls from both TV and radio stations before the event, and rang Network Rail for a briefing on the expected contents. I was treated as if I were asking for intimate details of Iain Coucher’s personal life. Of course, I was not seeking the whole story but simply enough to be able to talk intelligently and knowledgeably about the research. But no. I was merely steered towards the fact that it would focus on the West Coast line – not exactly an earth-shattering surprise, that one – and told all would be revealed at the launch, which was held in a swanky office in a new development behind Bishopsgate, as soulless as the CGI in the DVD.

 Yet, at slow news times, as any competent press officer knows, there is a lot of media interest prior to an event, but it tends to dissipate afterwards. So there was plenty of rather sketchy coverage in advance and little afterwards, suggesting Network Rail’s press strategy could do with some rethinking.

 Perhaps the lack of detailed analysis was fortunate because the report itself does not stand up well to scrutiny. Indeed, it contains an argument which undermines much of the case for a high speed line since the most astonishing aspect of the Network Rail document, completely missed in the national press, is that it effectively argues that a high speed rail route which does not go the whole way between London and Scotland is not really worthwhile within the terms of its analysis.

 The analysis is based on determining a Benefit Cost ratio for a variety of options for a high speed line, and the results overall are very poor. Several options, notably those that do not continue north of Manchester, have terrible Benefit Cost ratios, all under one. In other words, spending say £20bn on building the line would reap societal benefits of less than that, as low as £12bn in one case.

 This type of analysis, as regular readers will know, is based on an assessment of the benefits and costs of a project. Most of the benefits are notional. There is, of course, the fare box which is included in the benefits column (interestingly a third of the revenue would be displaced from the existing railway), but as that would not bring in enough to pay for the project, the ‘benefits’ figure is topped up by giving a monetary value to the small time savings made by people using the line. I have never been a great fan of this type of analysis, since the values attributed to time savings are arbitrary and, of course, wholly theoretical, but it does at least provide a basis on which to compare projects. And it is obvious that a high speed line comes up badly compared with many other big schemes..

 Aware of the flaws in this methodology, the Department for Transport generally likes a Benefit Cost ratio of at least 2 before it will consider funding a major scheme. Between 1.5 and 2 is considered okay and anything less than that is poor value for money.

 The analysis found that there is not sufficient demand between London and Birmingham, and even between London and both Birmingham and Manchester to justify a high speed line. The reasoning is that there would only be maybe four trains per hour to each of Manchester and Birmingham, whereas the line would have a capacity of, say, 16 to 20 and therefore would be greatly underused. In other words (and I rarely use italics, but this is worth stressing), the line currently being examined by HS2 to Birmingham is not worth building!

 It is only by having trains running from Scotland that there would be sufficient usage to justify a line, just as the Kent domestic trains were needed in order to ensure there was a case for HS1. But since HS2 would be built in stages, with the last one, to Scotland, unlikely to be finished realistically before 2030, this means its viability is questionable. There will always be the fear that circumstances – global warming, huge fuel price increases, energy shortages, deep recession, reduction in travel demand – might prevent the construction of the line north of Manchester, therefore leaving Britain with a truncated high speed line which could not even pass the rather low threshold of the benefit cost analysis used by the Department.

 But OK, let’s accept that the decision for going ahead with a full line to Scotland, the option chosen by Network Rail seems complicated and expensive because it involves a series of branch lines running of a main spine. The report considers a number of options and assesses the benefit cost ratio for each one, suggesting that the strongest case is for the full Monty to both Glasgow and Edinburgh with branch lines diverging off to Birmingham and Manchester/Liverpool. In other words, the option with the most railway and the most expensive to build since it would have the largest amount of track. In other words, the solution that an engineering company – remember Network Rail used to have a mission statement extolling its ‘engineering excellence’.

 Then there is the problem of Heathrow. Network Rail suggests, quite rightly, that running trains all through Heathrow (from the unknown station called London Central) is not worthwhile and suggests a spur. But this adds a whole host of complexities. As the line is a spine with branch lines, all trains would be point to point, and it is unlikely that there would be a case for frequent trains from Heathrow to Birmingham or Manchester, negating the whole idea. Frankly, I have always thought Heathrow is a dangerous red herring for this project and this analysis seems to back that up.

  While it would be churlish to call Network Rail’s effort to try to establish a business case as guesswork, it is not far off it. The number of variables and imponderables, known unknowns and unknown unknowns in Donald Rumsfeld’s memorable phrase, is so huge that to boil down to precise numbers for the benefit cost ration of each option. It is clear, though, that the environmental case, like the economic one, is feeble.

 It is instructive to note that a recent analysis by Halcrow into a potential high speed line in Brazil between Campinas, Rio de Janeiro and Sao Paulo suggest that it would cost double the previous estimate ($20bn rather than $10bn) and carry fewer passengers than previously predicted. Indeed research by Bent Flyvbjerg, the world’s expert on the cost of megaprojects, has shown that there is a consistent pattern of underestimating costs and overestimating passenger numbers. Either would weaken the business case such that it might not even reach the weak 1.5 threshold, let alone the 2 which most major projects are expected to require before being given the go ahead.

 Therefore, while it is welcome we are at last having a debate over a north south high speed it is clear that its supporters have an uphill task in proving the case for it. Certainly, for Iain Coucher to argue in the press release that ‘the line has a sound business case that will pay for itself’ is pure nonsense since most of those benefits are notional. To suggest that a government would give the go ahead to a scheme with such a feeble benefit cost ratio, with the added complexity that only a complete line gives a positive ratio at all, is fanciful. The report barely advances the debate by a few inches, and the HS2 company will have to come up with a far more convincing case to cement the cross party support for the line that Lord Adonis craves.

  • Andrew Campbell

    The NR proposals appear to be for a line which is top-drawer from end to end. If their headline costs are correct, then surely a good few billion can be chopped off by dropping some of the gold-plating. WCML capacity problems are at their worst at the southern end of the route, so the question is how far north does the line need to go to get a decent journey time from Glasgow/Edinburgh to London? One short enough to encoruage significantly more travellers to take the train.

    If you look at the French system, after Paris-Lyon, the next step in the early 1990s was TGV Atlantique, which runs from outside Paris to Tours with a branch to Le Mans. High-speed trains run on from that on the conventional lines to the likes of Bordeaux and Nantes. Extending the high-speed line from Tours to Bordeaux is now on the drawing board, but it’ll be another decade or so before it’s fully open. But the TGV service from Paris (and north and east of Paris) to Bordeaux and beyond works, and passengers get the benefit of having part of their journey on high-speed line, just as in the old days on Eurostar through Kent. Of course it’s true that the old Paris-Bordeaux journey was slower than current WCML trips, so the added benefit of the Paris-Tours line was greater than would be the case with a London-Crewe/Preston high-speed line.

    No government is going to come up with the cash to build London to Scotland in one go, so any plan is going to have to present real benefits for London-Scotland journeys without a full-length line. Perhaps the Scottish government could fund the stretch from Carstairs to Carlisle to speed things up (that stretch does suffer from sharing passenger services with freight) but I can’t see where they’d get the money from, never mind the costs for getting through Cumbria.

    Of course rail’s disadvantages on the London-Scotland routes are not so much about journey time, but about cost, comfort, and the legacy of weekend alterations, delays and replacement buses over many years. Reducing costs, particularly for walk-up fares, and improving service reliability may not be sexy but it’s a lot cheaper.

  • “Of course rail’s disadvantages on the London-Scotland routes are not so much about journey time, but about cost, comfort, and the legacy of weekend alterations, delays and replacement buses over many years.”

    We get the government—and, therefore, the infrastructure—we deserve. We’re extremely willing to whine and moan about the status quo, but heaven help you if you present us with the bill for fixing it to our desired standards.

    “Reducing costs, particularly for walk-up fares, and improving service reliability may not be sexy but it’s a lot cheaper.”

    Except the only way to achieve that on many routes is, er, by building new lines! If you’re going to build brand new railways, you might as well make the effort to do it right.

    Pick any two of “Good”, “Cheap” and “Fast”. I recommend against picking “Fast” unless there’s an urgent need for the project to be completed in five years as opposed to, say, ten or fifteen.

    The problem of cost can be mitigated by making it a rolling programme, rather than taking the usual election-term-friendly Big Bang approach of trying to get it all built yesterday. Split the new line into phases, sliced up according to cost. Build the “quick win” elements first, so the pressure to complete the entire project as quickly as possible is reduced. Include passive provision for future requirements, such as four-tracking, but don’t bother building it until you actually need it. (And make sure it’s possible to add the new bits later without closing the stuff you’ve just built first!)

    Government needs to follow the example of manufacturing industries and use a “just-in-time” approach to national infrastructure construction, rather than buying everything up-front every time.

    This means planning for the long term, but that’s what the civil service is for: set up a “Department for National Infrastructure” which exists solely to manage and improve all publicly-owned infrastructure. Give it a fixed, annual budget, and let them get the hell on with it without interfering every time a new minister is hired.

    (Oh, and give them BT’s phone network while you’re at it. And roads. And every other important piece of national infrastructure for which no logical or artificial “competition” can be provided. This offers some useful synergies, such as upgrading public services and constructing decent, accessible ducting for same at the same time as the road above them is being renewed or resurfaced. But that’s a rant for another website!)

  • John Fleming

    Surely the worst part of the own goal is that Network Rail wants to build a Shinkansen and not a TGV/ICE. In other words, they propose a completely new system, with different train profile, new tunnels, new bridges, new stations, the lot, rather than something that can be evolutionary, which was the essence of the success of TGV/ICE. Please could we return to the real world, the affordable world, and for pity’s sake, get on with it.

  • Pingback: LondonUnlocked.org | Eleven cities campaign for high-speed rail network()

  • The Thin Controller

    And don’t get anyone from Scotland started on the Edinburgh Trams scheme! That is showing all the traits of a megaproject – a complete (and possibly deliberate) under-estimate of the project costs, to which will eventually be added a complete (and possibly deliberate) over-estimate of the passenger numbers.

    Scale that headache up by a factor of thirty and we are looking at high-speed rail.

  • “Surely the worst part of the own goal is that Network Rail wants to build a Shinkansen and not a TGV/ICE.”

    There’s nothing wrong with that approach. The UK’s geography and population spread bears more resemblance to Japan and Italy than France or Germany. We have clusters of urban sprawl, rather than a more even spread.

    New lines make sense given that the existing lines we have were built in an age when building a tunnel under a river only a couple of hundred metres wide was considered a “mega-project”. Virgin’s execrable “Pendolino” trains are small and cramped, thanks to the UK’s tiny loading gauge, and even after the massive WCML upgrade, they’re still limited to 125mph.

    The WCML upgrade makes it clear that we’re at the stage where the old infrastructure is just not good enough any more. Our Victorian ancestors did the best they could—and 150+ years is pretty good going—but it’s time we stopped relying on that cop-out “Dunkirk Spirit”: Time to stop muddling through with patch-and-mend when it’s clear that brand new infrastructure is urgently needed.

    A new network of high-speed lines will give us a system with enough resilience to make it feasible to close and completely rebuild (or convert) some of the older lines to make them fit for purpose once again. Who knows: maybe we could even bite the bullet and gauge-enhance the entire classic network to match the same loading gauge as the rest of the continent. (No more tiny, cramped Pendolinos!)

  • @The Thin Controller:

    The Edinburgh Tram scheme is hitting a problem I’ve been predicting for some time: it’s a hell of a lot more expensive to build these things when you don’t have a plentiful supply of disused railways to build them on, as Croydon, Manchester and Nottingham did. (Even the Tyne and Wear Metro uses old alignments for the bulk of its route.)

    This is why I don’t see traditional trams returning to central London any time soon. It would bring the city’s roads to a grinding halt during construction and the trams would still have to share road space with existing traffic. Boris’ bizarre hatred of the Citaro “bendy” buses certainly isn’t going to help!

    Some form of suspended rail / monorail network would be a lot cheaper and quicker to build, while causing far less disruption. This is the key selling point of these alternatives. (Many Asian nations, including Japan, have invested heavily in these technologies for years. Japan has six monorail metros already in use.)

    These suspended technologies have the added advantage of inherent grade separation: the roads don’t lose any capacity at all, so you get a new ‘flying tram’, but don’t have to give up any road space for them. Traditional trams offer, at best, an “either / or” solution: either a tram is using the space, or a car, bus or truck is doing so. This does nothing to help congestion.

  • James Strachan

    What we are seeing is a power struggle between the Department for Transport and Network Rail.

    DFT have set up this tame company “HS2” to plan for a high speed rail line.

    Network Rail want to assert a right not just to maintain the network but to plan new developments and to control usage of the entire network.

    So they must get out some semblance of a plan before “HS2” can submit their plan.

    As Tom Lehrer put it, “I published first”.

    The industry should do better than this.

  • “What we are seeing is a power struggle between the Department for Transport and Network Rail.”

    The UK government system tends to use a “musical chairs” approach to appointing ministers which seems deliberately designed to stop them from achieving anything useful.

    There are two kinds of politician, defined by their reason for going into politics: they’re either in it for their own personal gain, or (on rare occasions), they genuinely want to improve things.

    The former might be raging egotists, merely corrupt or just plain greedy. This generally results in either empire-building behaviour, or (if they’re still quite early in their career) the “safe pair of hands” approach of lying low and doing bugger all of any merit while they wait for the music to start up again.

    The latter aren’t rare because so many people are greedy, or power corrupts: they’re rare because these people have to first genuinely care, and understand, the particular area they’re dealing with. That musical chairs system means many ministers spend a lot of time dealing with affairs of no genuine interest to them personally. Lord Adonis is an aberration, not the norm. He’s what happens when the system accidentally places the right man in the right chair!

    Empire-building is normal in any large corporate (or corporate-like) system. This is human nature, laced with a liberal dose of NIH Syndrome. (I.e. “If we didn’t think of it then, by definition, it’s a terrible idea!”) We’re seeing much the same thing that happened back when the unlamented ORR was throwing its own weight around to get a bigger slice of the pie.

    The rail industry was privatised at a time when nobody thought rail was going anywhere. Its fragmentation was deliberate. Its lack of a central “Voice of Rail” was also deliberate—Thatcher famously detested rail (and public transport in general), and the Major administration was no better. That *anything* got done in the face of this, let alone a project like Thameslink, is a testament to the people in charge of British Rail.

    The mere fact that BR was still managing to achieve such projects resulted in the Tories going out of their way to ensure rail would never have a single “voice” ever again. They wanted rail to wither away and *die*. To that end, shattering the industry into a bunch of disparate, bickering fiefdoms was the perfect solution.

    Quite why everyone seems to be so surprised at the result escapes me. “Divide and conquer!” Thatcher applied exactly the same technique to the mining and other heavy industries. It’s not new.

    What is surprising is that New Labour clearly didn’t give a damn about rail either.

    (Oh, sure, they pay lip-service *now*, but they know damned well they won’t be in power after the next general election, so they’re not going to have to worry about where all the money will have to come from. Making promises is easy. *Keeping* them is hard.)