Return to Archive -by date - by topic - 2001 Archive.

INFORMED SOURCES May 2001

Resource issues at heart of ETCS implementation

Applique ETCS could be a case of new wine in old bottles.

Three types of equipment are covered by the recommendations in The joint inquiry into train protection systems by Professor Uff and Lord Cullen.

First of all there are the two British Rail Automatic Train Protection (BR-ATP) pilot schemes which provide partial coverage on Great Western and Chiltern.

While the SEL equipment on the Chiltern route is obsolete, Alstom has actively promoted and costed the infill of its TBL system on Great Western. Platforms 13 and 14 at Paddington are not fitted and there is only partial ATP installation between Reading and Newbury. There is also a gap between Bath and Bristol Temple Meads.

Filling these gaps would involved fitting track mounted loops at 46 signals. Note that this would not cover the relief lines, which would add between 20 and 60 signals. Alstom's proposals also include the line through Bedwyn. Total cost for the three components is £7million.

Alstom has also costed a second scheme, extending protection to Severn Tunnel Junction plus the lines from Swindon to Kemble and Reading to Oxford , at £5.5million

Despite these low costs, the Inquiry saw no point in fitting the relief lines and other parts of the Great Western zone, given the ‘limited future' for BR-ATP. Thus the recommendation is only that gaps should be infilled to give full continuous coverage on the main line between Paddington and Bristol Temple Meads.

What – no Turbos?

What surprised me was the decision not to go for retro-fit of BR-ATP on the Thames Trains Class 165/166 diesel multiple unit fleets nor to equip the relief lines with ATP.

According to the Report, retro-fitment to other traction on the GWML is likely to generate problems not previously encountered on Great Western and Chiltern Lines. Anyway, TPWS protection is likely to be available before BR-ATP.

These considerations may apply to most traction, but the Thames and Chiltern Class 165/166 fleets are to the same design. Presumably, provision was made for the installation of BR-ATP on the Chiltern units, simplifying a retro-fit programme on Thames .

So, while the Report is highly critical when it comes to the limitations of TPWS compared with ATP, it is happy to rely on it on the route where the accidents which triggered the inquiry happened on the type of train involved in the worst of them.

As back up there is the belief that Thames DMUs will be fitted with ETCS ‘within a timescale not incompatible with the timescale likely to be required to achieve the full operation of retrofitted BR-ATP'. These ETCS-fitted trains could read BR-ATP track loops with the addition of a Specific Transmission Module or STM.

So Recommendation 24 says that consideration should be given to fitting Thames Trains with ETCS plus an STM to allow the use of BR-ATP track loops, with compliance by 1 January 2003 .

Now given that ETCS is still at the pilot stage, that no one has yet produced an STM and that the combination of retrofitted train, plus STM plus the interface with the track equipment would have to be safety validated, getting it running in 20 months time does seem a trifle optimistic.

It gets even messier on Chiltern, which wants its version of BR-ATP extended from Aynho Junction to Snow Hill. According to the report, this 48 mile section could be fitted with BR-ATP – although you might have difficulty in getting the manufacturer to supply obsolete kit.

Alternatively, says the report, you might use ETCS Eurobalises to extend protection. Chiltern's new Class 168 DMUs, which could be retrofitted with ETCS would then read BR-ATP to Aynho Junction through an STM and then use ETCS for the rest of then trip to Birmingham .

Even Uff and Cullen have doubts over the ability of the existing BR-ATP to read Eurobalises through a notional reverse STM. In which case, the Class 165 fleet would be limited to the BR-ATP equipped track south of Aynho junction.

GWML resignalling

But then I realised that what these recommendations on ATP for the GWML overlook is that the interlockings at Reading , Swindon , Slough and Bristol are already overdue for replacement. Which is where you realise what the lost years of privatisation have done to the railway.

Remember how under the malign influence of consultants and bean counters Railtrack's Project Destiny said that patch and mend was more cost effective than renewal? Remember how Gerald Corbett said that re-signalling costs were so high that they didn't make a business case.?

Remember those reports in this column about the ensuing hiatus in signalling orders. Followed by reports of the signalling companies laying off design staff? And believe it or not, even with TPWS inflating from a £33million project in 1996 to £450million and rising today, the signalling majors are still worried about diminishing future work loads.

So while the extension of ATP seems like a reasonable policy in isolation, the urgent need to resignal the GWML, means that you could end up with clockwork BR-ATP interfacing with whatever new interlockings Railtrack adopts and being read through STMs.

This highlights the mess train protection has got into. In a rational world, BR-ATP on GWML would soldier on only to 2006, say, when in a competent railway we would expect to have digital interlockings talking to cab signalling through Eurobalise or even digital radio (GSM-R) on completion of the GWML resignalling.

Thus signalling reality matches exactly the timescale for ETCS fitment on the GWML recommended by the Report, which suggests that the GWML, rather than the ECML should follow the WCML in the ETCS programme.

Lest anyone is tempted to ask how you manage the changeover from ATP to ETCS while keeping the trains running safely or what's involved in fitting ETCS to an 1975 vintage IC125 power car, or whether a reverse STM woukld be simpler I shall claim protection under the Railway Privatisation (It wasn't my idea) 1992 Act, and try and get back to you.

TPWS cope creep

On TPWS there is much agonising over whether it would be better to drop the scheme and put all the resources into bringing forward ETCS. Overall, while the Report does not accept that any decision about TPWS is inevitable, its cancellation would not advance the date for the introduction of ETCS. How true, since ETCS is still at the testing stage.

While the concept of TPWS as a relatively cheap and quickly available stop gap in place of nation wide ATP ‘remain proper and viable aims', the Report's authors are left with ‘considerable reservations' about its effectiveness. ‘Its limitations have become more apparent as its cost has escalated' they comment, adding, ‘We doubt that it would have been adopted had these factors been known more reliably at the outset'.

In particularly, they are concerned that its effectiveness is least where the risk is greatest – namely high speed trains with braking described as ‘falling well below the optimum', which I take to mean 9%g compared to the 12%g of modern stock with Enhanced Emergency Brake.

Despite these misgivings, the report accepts that TPWS ‘offers a material degree of reduction in SPAD risk'. The main outstanding issue on bog-standard TPWS (called TPWS-A by the report) is which signals should be fitted. Some junction signals don't need it, some plain line signals do and all multi-SPAD signals should be fitted. The report calls for urgent studies to ensure that resources go to the signals most in need of protection. Hell's teeth, what has Railtrack been doing since 1996?

Given the concern over the limitations of TPWS-A, the report considered two enhancements. TPWS+ adds one or more additional Over Speed Sensor (OSS) loops, colloquially the ‘speed trap', further in rear of the signal set to a higher speed than the current 74mile/h. However, the further back the additional OSS , the greater the possible mismatch between braking curve, train speed and positioning of the sensor.

For example, at 100mile/h a modern multiple unit could brake safely well after an IC125 or a Class 47 with Mk 2 stock. Setting the OSS for the older train would cause unnecessary TPWS intervention with the multiple unit.

Older readers will recall that one of the problems with BR-ATP was handling the wide variety of speed and braking rates on trains A&M without affecting capacity. That was why BR signal engineers thought they needed a their own ATP rather than buying Ebicab, as was, off the shelf.

Then there is TPWS-E, promoted by Siemens. This uses Eurobalises on the track in place of loops and could also talk to ETCS fitted trains. While this ‘smart' TPWS offers upgrade advantages and greater flexibility in terms of train types, the Report decides that it would take 2 ½ years, including tests before it would be available.

Readers might compare this realism about a system which is nearly ready for testing with the optimistic time scale for a BR-ATP-reading STM on Thames Trains.

Anyway, while there are no recommendations on the implementation of TPWS-E, the Report says that the rail industry should explore the project further. Both Railtrack and the HMRI see potential on lightly trafficked lines where ETCS is not appropriate.

ETCS mandatory

But the meat of the report is ETCS. The fitment of ETCS to the TEN lines covered by Directive 96/48EC, plus the current draft Directive extending interoperability to conventional lines, will be incorporated in UK legislation through Regulations. The Joint Inquiry, says that these Regulations should be in absolute terms and not dependent on reasonable practicality. And to a set timescale

Given the concern over another catastrophic accident happening, the Report says that the HSE should establish a programme of consultation and drawing up of Regulations for the fitment of ETCS with the objective of Regulations being in force within three years (my italics).

Three years to draw up Regulations when the Report castigates the railways a number of times for their tardiness? But, as the Health & Safety Executive points out, development of the new Regulations will involve the DETR, the railway industry, rail workers representatives, the travelling public, the Strategic Rail Authority, the Rail Regulator ‘and other stakeholders'. The three years also reflects ‘further work on technical feasibility by industry'. The HSE also points to the Recommendations in the report, (below), covering ‘the trial and testing work that needs to be done'.

Reading the Report you get the impression that the ETCS work at Old Dalby (see last month) is a generic pan-industry development programme rather than the sharp end of Alstom Signalling's commercial contract for the WCML. I'm not sure whether Virgin, Alstom and Railtrack would welcome a new System Authority ‘directing' their programme.

Pilots

Requirement and objectives of the Regulations are set out in the box) but ETCS pilot schemes are the subject of three recommendations. These are:

Fitment of ETCS Levels 1 or 2 between Aynho Junction and Snow Hill by 1 January 2003 ;

Fitment of ETCS plus an STM to Thames Trains by the same date;

And the selective fitment of GSM-R radio fixed equipment to selected routes in advance of ETCS fitment to trains, by 1 January 2004 .

As you can see from the box trackside ETCS, is also required to be in service on the East Coast Main Line

by 2005 or 2006. Note that this is the current timescale for the commissioning of Level 2 ETCS under Phase 2 of the West Coast Route Modernisation. And our continental cousins, with their well known lack of irony, refer to the WCRM installation as one of the ETCS ‘pilot schemes'.

Which brings us back to the people in the Passengers Group who in a press conference (immediately after the Report's release and in the same room at the HSE offices) hailed the Recommendations as a vindication of their demands for action. I do not believe that the signalling resources exist to install ETCS on three main lines simultaneously over the next five years on top of on-going resignalling work and TPWS installation.

Uff and Cullen say as much when they highlight two concerns. The first is the availability of adequate resources within the UK and European railway industry. The second is the achievable timescales for fitting the major items of train protection under consideration. ‘The fitment of new train protection systems may be dictated by resources rather than finance or management decisions', they warn, adding, ‘this will apply particularly to the fitment of ETCS Level 1 and 2'.

As a starting point for the Consultations on the Joint Inquiry Report someone has to produce a sensible programme for Railtrack's global signalling requirement. And then take the flak for dashing people's hopes. -

Continues.........