Return to Archive -by date - by topic - Archive 2002.
British Rail could push signalling innovation: Railtrack thought it could do the same – big mistake
There is a tendency to regard Computer Based Interlockings (CBI) as a different technology to our home grown Solid State Interlocking (SSI), conceived by BR Research and commercialised in a joint venture with what is now Alstom and Westinghouse.
In fact, SSI pioneered the CBI concept and the perception is all about terminology not technology. SSI is a type of CBI.
Hence the difference between Mk 2 SSI and the third and fourth generation CBIs on the market, essentially comes down to processing power and capacity plus computer aided scheme design – the ‘front end' which is what determines signalling prices .
In fact, signal interlockings not being a high intensity continuous process, blinding speed is not that essential in a CBI processor. So, while CBI will offer more of everything, SSI remains a competent technology, selling round the world.
Railtrack's mistake was to dismiss this competence while simultaneously overestimating both the readiness of the latest generation CBIs and the work needed to introduce them to the UK rail network with its safety case-based acceptance process.
And that crashing noise you can hear is chickens coming home to roost in a terminal velocity dive through the hen house roof.
Railtrack awarded three CBI pilot installations to Adtranz (now Bombardier) for Horsham, , Ansaldo (Manchester South) and Siemens ( Dorset Coast ). Subsequently Bombardier's Ebilock CBI was selected for Thameslink 2000, of Which more anon.
In addition to these pilot schemes, Westinghouse is getting on with acceptance of its new Westlock CBI. Of which more anon too.
All three pilot schemes have found life hard. One problem is that acceptance has to satisfy two criteria. First, vital safety software has to be validated. This is the easier bit, since in most cases cross acceptance of the original validation for other European railways is sufficient.
However, it is the application safety case which is causing everyone to sweat. This involves demonstrating that the software can run British signalling principles safely.
Current status of the three pilots is as follows.
Siemens is still on target to commission the Dorset Coast scheme in October this year. But being straight up guys they told me in July that the timescale was ‘tight'.
And, yes, it is defining and clarifying the requirements for signalling rules that is an on-going problem. Developing the control tables, in particular, is proving ‘difficult. SSI has some of the UK specific requirements designed into the processor's software, so replication is not easy if you are not SSI-smart.
Bombardier is due to commission Horsham in November, but informed sources report similar pressures to those on the Dorset Coast scheme.
Finally, while one hesitates to machinegun the survivors of a torpedoed project, Ansaldo has fallen even further behind on Manchester South. Commissioning already deferred from 2001 to 2003 has now been put back to 2004 at the earliest.
Explaining the latest delay, James Martin, Railtrack¹s West Coast Route Modernisation project director spun in vain. ‘This is the first time the new Ansaldo signalling technology has been used in this country and the safety approval process is taking longer than anticipated, even though the system is working perfectly well in other parts of Europe'. He added ‘We remain totally committed to this important project', Reversion to SSI, considered and rejected over a year ago, must again be an option.
Manchester South was bad enough, but the real blow to Railtrack's CBI aspirations came at the end of June when the three SSI suppliers, Alstom, WS Atkins and Westinghouse, were summoned to a meeting with Railtrack and invited to bid for the Thameslink 2000 resignalling.
This was a shocker, because Adtranz Signal, as was, had been nominated as preferred bidder for the design and development contract for the Thameslink Associated Signalling Works (TASW) in June 2000. Horsham was a vital precursor to this mega-scheme which would provide 24 train paths an hour each way through the central section between London Bridge and Kings Cross.
Resignalling would use the Ebilock CBI and EBICOS train management system. With its extra processing power Ebilock should have required a smaller number of interlocking modules.
One estimate was that four Ebilock CBIs could provide the same capacity as the 13 SSI modules needed for the scheme. Informed Sources suggest that as the scheme design developed, the ratio was approaching parity.
An expected benefit was the ability to switch control between the existing interlocking and the new Ebilock during commissioning. As part of the development work for the UK , Ebilock is designed to communicate with standard Railtrack trackside equipment. Stageworks would have been hugely simplified.
But even though Horsham was making progress, informed sources suggest that adapting the Ebilock/EBICOS package to Thameslink 2000 threatened to delay the opening of the project to 2012.
Reversion to SSI should not materially affect performance. When Thameslink 2000 was out to tender, Westinghouse was asked to put in an SSI based solution as a reality check. At the time Westinghouse reckoned it was a viable solution.
So back to square one, but with a build-up in throughput starting at a modest 16 trains an hour each way. Mind you, given Railtrack's claims for a shortage of signalling design resources, isn't it a bit perverse to get all the SSI suppliers beavering away rather than reviving the SSI based scheme worked up by Westinghouse?
And speaking of Westinghouse, their fourth generation CBI is coming along nicely. The vital safety platform comes from Triconex, an American sister company within the Invensys group.
But the clever bit is that Westinghouse is porting the SSI data across onto this platform which runs two out of three redundancy, just like SSI, and will also be able to use existing SSI comms and trackside modules.
According to Westinghouse sources, the development programme at Chippenham and in California has been accelerated by six months. Currently the Infrastructure Safety Review Panel has approved the Westlock acceptance Safety Plan and the company expects to start acceptance proper in March next year. Assuming success, look for Leamington Spa, where SSI was piloted in 1985, to see the first installation – but don't ask when.