The great train robbery
This September Virgin Trains accepts the last of 53 nine-coach Pendolino tilting trains to run between London and Glasgow. The story goes that Britain invented the tilting train and should have been exporting it. So what went wrong? John Dwyer digs into what really happened
Potent myths surround the UK's on-off love affair with tilting trains: Britain developed tilters first but, when they turned out to be expensive and unreliable, sold the technology to the Italians, who are now selling it back to us. None of this is true. The real story is a more interesting brew of boardroom in-fighting, clashing cultures and politics.
Virgin's vision for the £1 billion Pendolino programme it launched in 2001 was to double rail travel on the West Coast Main Line (WCML) to 30 million passengers a year within five years. The programme foresaw trains running at full tilt 140mph along the WCML by May 2005.
But when escalating cost and delays to the WCML upgrade contributed to Railtrack's going into administration in October 2001, the upgrade, says a Virgin source, 'went into the melting pot'.
Virgin began a Pendolino service in July 2002, but Pendolino passengers did not arrive at Glasgow Central until January this year. No tilting was involved. Only 50 of the 401 miles of track and signalling from London and Glasgow are equipped for it.
By September the track and signals from London to Crewe will allow trains to tilt and reach 125mph. But until December 2005 trains from Crewe to Glasgow will still be restricted to 110mph. To reach their 140mph top speed, they must have in-cab signalling which, says a Virgin insider, 'does not yet exist'.
A similar tale attends Virgin's lower profile Class 221 SuperVoyager tilting diesel trains, that were introduced on the CrossCountry routes from October 2002.
Tilting trains are no gimmick. The railway rule is that every 1mph increase in average speed brings a one per cent increase in revenue. Corners make passengers uncomfortable at high speed. The French (TGV), German (ICE) and Japanese (Bullet Train) solution was to build new routes with long straight sections linked by steeply banked corners. But for one-tenth to one-twentieth the cost, you can put tilters on existing routes.
The WCML offers no such choice. It has too many curves and its northern section runs through valleys with no route for straighter tracks. BR's answer was the tilting advanced passenger train known as the APT.
In 1962 BR chief of research Sydney Jones recruited now-Professor Alan Wickens from the aircraft industry as director of advanced projects. When, in 1968, the APT project won 50:50 government funding, Wickens recruited a team, many from aviation, to carry it out.
Tilting trains were not new. Though United Aircraft/Sikorsky's passive-tilting Turbotrain went into service in 1968, US experiments go back to the 1930s, and France's TGV decision followed the failure of a tilt experiment in 1956.
But the project created friction between Wickens's group and BR's mechanical engineering department, responsible for acquiring new trains. Not only were Wickens and some of his colleagues not railwaymen, but they used computers and other non-railway working methods: "We adopted a systems engineering approach, familiar in the aircraft industry," says Wickens, "where the major design issues were identified and technical decisions made on the basis of calculation... [The] engineering culture on railways had been to design, build, deliver and sort out the problems in service."
He adds that, "There were many people on the railways who were anti-research, because it upset people, largely. New ideas you see. It's a very conservative industry."
The experimental train, APT-E, had its first run on 25 July 1971. Wickens says it proved BR could build a high speed tilting train. And though the APT-E programme was overspent, its £2.5 million was 'miniscule' for a government which spent £2 billion on Concorde between 1967-1976.
The cost of the prototype, APT-P, programme "cannot be more than £40 million, including depots and other operational investments, and it probably underspent its budget," says Wickens.
When the project went ahead in 1973, the specification for a large train with two power cars in the middle you could not walk from one end to the other was based on the needs of the WCML.
For Wickens, this was the first of BR's two grave mistakes. A slight cut in top speed would have simplified the design, turning APT from one train for one route into a range of trains which could speed up routes all over the country: "They could have had a train that looked very like the Pendolino."
Over-regulation also played a part. The TGVs have power cars at the end, says Wickens, because they run a high voltage line down the train. The UK's railway inspectorate would not allow this.
BR's second mistake was a 'mad' decision to build only four of the eight trains the government had offered 80 per cent funding for in 1973. In later public spending cuts, this number was cut to three. "That killed it," says Wickens. From then on, "the reduction in the size of the build programme gave a clear message about priorities."
In April 1973 BR transferred responsibility for both the prototype service trains and the APT design group to the mechanical engineering department. Though by the time of the transfer, there was "a very much better atmosphere at the senior levels," tension remained lower down between the APT group and their new department.
The culture at BR's manufacturing arm, British Rail Engineering (BREL), was also unsympathetic. BREL engineers were used to interpreting drawings 'quite liberally', says Wickens. This was not acceptable for a new project based on careful calculation, and the old hands resented new boys telling them how to build trains.
BREL did not stick to the build schedule. "They just didn't think it was important," says Wickens. "There wasn't the control over things that went on there that perhaps there should have been."
The first prototype, ATP-P, did proving trials in May 1979. The second ATP-P was completed in late 1979 and the third in 1980. BRB was pressing for a commercial service, so the APT's designers and builders still had problems to solve when the train started its inaugural return run to Glasgow on 7 December 1981.
Wickens agrees with many others that BR overdid the hospitality in Glasgow. This, not the tilt mechanism, caused some passengers to feel ill on the return journey, nicknaming APT the 'queasy rider'. Worse, snow and freezing weather caused the water-brakes to freeze up and, two days after its first run, the train stopped at Crewe.
The problems were minor and could have been sorted out, says Wickens, and none of them would have surfaced in public if the board had not rushed the APT into service prematurely. Though the train soldiered on in passenger service for another two years, as another project team member, Hugh Williams, told the BBC, "The train had received such bad publicity that I don't think the government could have given the substantial funding increase the project required."
Even from here, says Wickens, APT was far from a disaster. His dozen-long list of outcomes from the project include benefits to the high speed train, in service since 1974; dynamics research; a new train configuration which is still at the leading edge of world train design; aerodynamics work which contributed to Eurostar; longitudinally welded lightweight aluminium extrusions and much else. Though, as Wickens says, "We didn't export it, at all," he believes the APT, "had an enormous influence and stimulated a whole lot of work."
Wickens sums up: "In any major technical development the going often gets rough and the thing to do is to keep on going. The reason the Italians have been successful with tilting trains is they kept going. That's the most important lesson."