London Underground's Jubilee Line Extension is in a difficult situation. It is the victim at best of over-inflated productivity claims by its signalling contractor; at worst of an over-optimistic desire to push the realms of known technology.
'It's a hellish position,' says project director Hugh Doherty explaining why the new line will now not open until spring 1999 - more than a year later than originally planned.
The revolutionary moving block signalling technology - vital to London Underground if it is going to cram more trains down its tunnels to meet the future capacity demands on the network - is now abandoned in the short term. This will not now appear on the Jubilee Line until well into the new millennium.
Westinghouse, which has the contract to design and install the signalling, has let him down. The fact is: trains can run through unfinished stations but only if the signalling system is up and running.
'You cannot change horses late in the day on a sophisticated thing like a signalling system,' says Doherty, as he discusses the hard decisions that had to be made when Westinghouse confirmed that it was going to miss its deadlines. 'Once you've contracted a company to supply a system, you are with that company.'
'It was not until February this year that Westinghouse finally admitted that it could not do it, so you are left with the consequences,' he says. 'It is unfortunate, it is regrettable, but it was a fact of life. All we could do was work with them to deliver what could be the best system that would allow the railway to work.'
This meant putting in place short term contingencies to deal with immediate problems while working towards longer term solutions.
Reverting to a traditional fixed block system, where trains run at fixed distances apart, was the only short term option sure of success.
Doherty had initially hoped that Westinghouse could deliver such a system in time to allow a partial opening in September 1998 with 17 trains an hour between Stratford and Waterloo. But even this is was abandoned last February in favour of opening the entire line between Stratford and Stanmore in spring 1999 (NCE 5 February).
'The priority is to deliver the fixed block system because of the ongoing delays with the moving block. We will bring in the moving block whenever Westinghouse has reached a stage with development to allow that to happen,' he adds.
This move to a traditional fixed block system was driven by JLE when, Doherty claims, his team realised that moving block would not be ready, despite Westinghouse's assurances. JLE then brought in specialist BTR to review the work to see if it could be done. 'But finally they admitted that they could not even deliver that,' he said with some dismay.
Doherty refuses to accept that the signalling problems could have been foreseen. 'If we could have started from day one knowing what we know now, we would have gone for a fixed block system and brought in the moving block when it was ready,' he admits.
London Underground really wanted moving block to underpin its future Tube-wide signalling system and JLE would have pioneered the technology. Unfortunately, Westinghouse underestimated the amount of development work it had to do to make it work.
'It is just unfortunate that my project got caught up in it,' he says. 'We recognised the risk of the moving block system and knew that we had to keep a check on progress - and we did. Development did not move on in the way that it should have and we perhaps introduced mitigation rather late in the day.'
Doherty has his doubts about whether or not LU really appreciated the risk it was taking with Westinghouse and moving block technology. 'I haven't seen the kind of formal risk analysis that one might do now,' he admits. 'But I know the type of risk analysis that I would do now.'
He says that by nature engineers are ambitious and optimistic, and this was certainly the case with moving block. But Doherty defends this 'can- do' attitude as a vital factor in the struggle to get things done. 'You have to be a bit careful with your enthusiasm for risk assessment as you might double the cost and make the project non-viable.'
The irony of his situation is that, as a civil engineer, he has overseen some of the most technically challenging pieces of work ever attempted. The civils work has had problems - it still has more than a few - but within a few months of the scheduled completion of the 43 month civils programme, construction work is being pulled home leaving the signalling the critical element.
Doherty doubts that a closer relationship between JLE and its signalling supplier would have produced a better result. But it is clear some lessons from JLE have been taken on board by the teams working on the signalling work Railtrack's £2.5bn West Coast Main Line upgrade. Here, a seamless project team drawn from the client and preferred supplier GEC Alsthom has been set up from the outset to see the project through (see News).
Doherty is reluctant to blame the procurement process for the problems with the signals. 'This was not a contractual issue. I don't see that other arrangements would necessarily have helped us. More fundamental is how much you rely on a system that has to be developed during the life of a project,' he says. 'I think I would rather have a reliance on known technology at the outset, rather than rely on a development succeeding during the lifetime of the project.'
Certainly the episode will make LU more cautious about bringing untried technology into its projects, 'and rightly so,' he says. LU perhaps did not truly appreciate the engineering consequences of its desire for such advanced technology.
Compare this signalling approach to that taken by engineers when faced with the possibility of using sprayed concrete for tunnel linings - another relatively untried, undeveloped technology. This was only accepted as a solution after contractors had proved to JLE that it was safe.
'Perhaps the demonstration (of the moving block system) should have been at the outset of the project,' says Doherty. 'But we knew that this couldn't take place.' The logical decision therefore, he says, would have been to build this project with known technology and bring in moving block on the next one.
'With hindsight, people might say that is what should have been done. But there was a wish to have that technology much sooner,' he adds, pointing out that JLE called for a system that gave a capacity rather than any specific system. 'How the contractor achieved that capacity was up to him,' says Doherty.
Westinghouse has now taken on more people and JLE has put its own people into the Westinghouse team to ensure revised deadlines are met. But, says Doherty: 'If the contractor is struggling, there is no easy solution that you can implement quickly.'
Westinghouse has had a fall guy in chief executive John Mills, who lost his job in February. Doherty refuses to be drawn on the extent to which the JLE problems led to Mills' departure, but says simply: 'He was directly responsible for the work'.
Doherty sticks doggedly to the no blame culture he instituted on JLE, but he points out that this does not mean changes are not made when needed. 'You have got to have the skills around you. We have made changes in our own team and insisted on changes in the contractors team. That is not blame culture, it is trying to bring to bear the most appropriate skills to solve whatever problem you are faced with.'