When the Health and Safety Executive investigates the causes of the Southall rail crash, it will have to thoroughly probe the signalling system.
It dates back at least 30 years, and incompatibility between the old signalling system and a modern electronic safety system – Automatic Train Protection – under trial on the Bristol-London line, could have impeded the safety system’s functioning.
From early next century, at least the West Coast Main Line, which runs from London to Glasgow via some of Britain’s largest cities, should be safer. It is about to undergo a massive route modernisation to install a new signalling network.
‘The last time it had any tender loving care was in the 1960s and 1970s, and it shows,’ says Vincent Considine, project manager at Railtrack. ‘The route modernisation is going to take this venerable railway and bring it into the 20th century.’
Transmission Based Signalling (TBS) is expected to give trains of the future the signalling system they deserve.
It will be built using two new features: computers and radio transmission. The ageing infrastructure now in use is based on an unsophisticated combination of signal lights and track circuits.
In a track circuit, the train physically completes an electrical circuit as it passes, sending a message to a network management centre identifying the position of the train.
Each track circuit, placed 1.5km apart, triggers a set of lights, letting the train pass or telling it to stop.
‘Both the lights on sticks and track circuits are 19th century technology,’ says Considine. ‘Because of all the business needs to improve capacity on the line, and to improve its reliability and safety, there’s a big drive to modernise the way you signal the railway.’
The existing signalling system will not cope with the next generation of trains. Apart from constantly breaking down, it will not be able to handle Richard Branson’s tilting trains designed to negotiate the twisting tracks of the WCML at speeds of 125mph. At these speeds, drivers will not be able to check if a signal is red or green, or see round corners.
Current safety systems will also be inadequate. The automatic warning system only rings an alarm if the train goes through a red light, and will only trigger a stop 50m in front of the next signal – too late to prevent a crash.
ATP is modern enough – it was recommended by Sir Anthony Hidden after the Clapham rail disaster in 1988 – but it needs a modern infrastructure to support it. The problems of grafting an electronics-based system on to an electrical system of the 1960s has haunted railway engineers for years and will feature in the Southall crash investigation.
Track circuits also cost too much to maintain, and cannot take more than one train per 1.5km, a distance that does not provide a good business case for private rail operators.
There are other, human, problems. Electrical wire circuits are easily vandalised and traffic lights have been missed by the occasional sleepy driver.
TBS is expected to solve these problems.
Radio-based cab control will use vandal-proof radio waves to send signals directly to the driver’s cab to say how fast to travel and when to stop the train.
Using GSM, or some other digital radio network, high-speed trains will continually converse via the driver’s cab with the network management centre about the train’s position and speed.
As a back-up, small antennae will be positioned every few kilometres to check the speed and position of the train. ATP, which will be designed into the new system, will slow the train down if it goes too fast.
Benefits from this computerisation will include automatic access to information about track engineering, and last-minute hold-ups. The train driver will have a more immediate idea of delays and be able to communicate more accurately to train passengers.
The question engineers are trying to answer is where to put the intelligence. It could be on the train with the driver, the ‘smart’ train, or in the network management centre, the so-called ‘dumb’ train.’If we have dumb, and the trains from Europe are dumb trains, then they can work on our infrastructure,’ says Considine. ‘The alternative is the smart train, where the processing is local to the train, which means that our train wouldn’t mind what it drove on in Europe, but it would be more difficult if the Europeans had dumb trains.
‘Are we going to make it easier for the Europeans to come to us, or are we going to make it easier for us to go to Europe?’, he asks.
The financial motivation behind TBS is immense. Despite current perceptions, railway companies are keen to push more trains down the tracks to make the most of their licence and attract more customers.
The new rail operators have also pledged to provide newly refurbished rolling stock, and shareholders require a quick return on investment.
TBS will let network managers know exactly where a train is, and therefore allow them to use the track more effectively, running trains closer together. It will also allow trains to move in both directions – now impossible – which will save track space and time when taken up with repairs.
The business imperative has influenced the technology’s timescale. ‘Our present timescales are that we would like to complete development by 2001, such that we have installation complete by 2006. Clearly that doesn’t accord with other commercial drivers,’ says Considine.
‘There’s a lot of pressure to bring the operation date forward.’ Implementation will start as soon as development finishes.
Whoever manufactures the system for the WCML will receive a sizeable wedge of business. Railtrack expects the contract to build and install a TBS network for it to run to around £500m.
Two consortiums – GASL, comprising GEC-Alsthom Signalling, Siemens and Alcatel, and Transig, formed from Adtranz and Westinghouse Signals – are developing interoperable systems. One of these will win the contract for the WCML, and both will compete for refurbishments of further routes.
The systems have been developed from scratch. Railtrack has put in a ‘significant’ contribution.
TBS should have a noticeable effect on overall train effectiveness. Junctions, for instance, will be managed more efficiently. When two trains arrive simultaneously, neither will be forced to stop to let the other through: instead the system computer will synchronise their speeds to allow both through without delay.
With an integrally designed ATP, it might also make fatal accidents such as those at Clapham and Southall a distant memory.