SUPPLEMENTARY REPORT TO AUTOMATIC TRAIN PROTECTION FOR THE RAILWAY NETWORK IN BRITAIN A STUDY by SIR DAVID DAVIES CBE FREng FRS President, The Royal Academy of Engineering SUMMARY This report represents a brief update on developments in automatic train protection following the full report on this topic by the same author published in February 2000. It includes additional comments on possible enhancements to AWS and the potential of the current ATP trials for future extension to a national system. The report also describes progress on the ATP trials and the fitment programme for TPWS including work on the TPWS+ and TPWS/E proposals. This leads to a discussion on the possible value of a systems authority and some brief comments on how the recommendations of the main report are being taken forward. August 2000 The Royal Academy of Engineering 29 Great Peter Street Westminster LONDON SW1P 3LW
Tel. 020 7222 2688
CONTENTS 1. Background 2. Additional work 2.1 Enhanced AWS 2.2 ATP based upon pilot trials 3. Progress on TPWS 3.1 Progress on fitment 3.2 Additional calculations 3.3 TPWS+ 3.4 TPWS/E 3.5 Limitations of TPWS 4. Progress on ATP Trials 5. Systems authority 6. Progress on ERTMS/ETCS 7. Funding for Research 8. Conclusions 9. References
1. Background This report represents a brief update on developments in Automatic Train Protection following the publication of my report (Reference 1) in February 2000. This supplement has been prepared at the request of the Department of Environment, Transport and the Regions (DETR). Although I do not envisage that this report will be published in the same way as the original document, it will be made available to relevant interested parties within the railway industry and to the joint Inquiry being chaired by Lord Cullen and Professor Uff. I also plan to make it available on the Web. The report covers some further developments that have taken place in the UK and the European continent including follow up to the recommendations made in the original report. It is perhaps relevant to record that since publication of the main report I have been asked by Railtrack with the support of the Deputy Prime Minister to take on the position of Chairman designate of the new company to be created shortly to oversee safety, namely Railway Safety. This follows the recommendations of the DETR Review of Railtrack Safety and Standards Directorate (Reference 2). Although it is unlikely that the new company will be formed until Autumn 2000 I have already started the process of discussions with various parts of the industry and this has greatly helped with the preparation of this report. Nevertheless, the views expressed are my own and should not be seen as those representing any part of the industry. 2. Additional Work 2.1 Enhanced AWS In my main report I indicated that towards the end of my studies some reports were made available suggesting that AWS could be enhanced to provide a train stop facility (reference 3) and that this might be a suitable alternative to TPWS. I indicated that, at the time that this would require further study and I have now had the opportunity to look at this in more detail. There are a number of ways of enhancing conventional AWS to provide additional features such as the train stop facility at a red signal. One approach that has received recent attention involves installing an additional permanent magnet shortly after the standard location of the conventional AWS magnet set, the additional magnet being suppressed except when the signal is at red. By means of a relatively small modification to the AWS receiver, it is possible for this to trigger a brake application associated with the red signal. Operationally this would function in a relatively similar way to TPWS but would have the major restriction that the location of the breaking point would have to be shortly after the existing location of the AWS magnet, allowing little or no flexibility for choice of braking position. On this basis and for a train with 12%g of breaking performance, a brake application so initiated should bring the train to rest safely within the overlap, providing the train was not travelling faster than about 55 to 60 miles per hour. As with TPWS, it is important to appreciate that the system would operate at all speeds but that the location for braking point could only ensure that the train was brought to rest within the overlap, provided the train speed was not greater than the above figure (depending on braking performance). Clearly the speed limitation is a serious disadvantage and is less than the corresponding
figure for TPWS. No figures are available for the installation cost of such a system but the trackside equipment would need power supplies plus access to the signal
aspect information (as for TPWS) although the equipment costs might be slightly less. The trainborne equipment would comprise a relatively simple modification but would need all trains to be so modified. I think this is an interesting idea which would have been of significant value to the railways had it been considered and properly assessed quite some time ago (eg 20 years ago). Today, there is some concern about the overall reliability of AWS which is clearly based on old technology. Overall, I could not recommend this as a sensible way ahead at present though I believe it would have been a sensible development several years ago. A further issue is that AWS is now towards the end of its potential life with little life extension potential, whereas modern ATP systems must clearly be designed to be extended in the future to provide many additional capabilities. 2.2 ATP based upon the technology of the current pilot trials It has been pointed out to me that in my previous report, when comparing the options of TPWS with ETCS, I should also have covered a third option, namely that of extending one of the existing ATP trials systems to cover the whole of the country. This is a fair observation. However, my report was based on my argument that it was necessary to choose a single form of ATP for the future and that the choice should be ERTMS/ETCS. Nevertheless, the choice of one of the trialed ATP systems would have the advantage that such systems have already been the subject of extensive experimental trials (lasting nearly 10 years). Choosing one of these two systems for nationwide fitment would, therefore, have the advantage of avoiding the need for a full scale trial period for the ATP system. However, it is relevant to note that neither system has been trialed on an electrified part of the railway and some work would be necessary to confirm its capability in this environment. In contemplating such an action, it is also very important to bear in mind that neither equipment is compatible with the EU directive 96/48EC and it would, therefore, be necessary to replace such newly installed equipment with ETCS systems when any part of a high speed line was upgraded and/or resignalled. For this reason the choice of either the TBL system (as on the Great Western Main Line) or the SELCAB system (as on the Chiltern Line) for the whole of the UK would raise enormous operational and legal problems. In addition, these systems are currently regarded as rather out of date and one of them is no longer in production. Maintenance of old systems can present problems of availability of parts or heavy cost penalties. If such a system were chosen it would also condemn the UK to a complex mixture of ATP systems since ETCS is being fitted on the WCML and some degree of inter operation is currently planned. Switching between systems is very unattractive and can also lead to safety problems quite apart from the problem that such a choice would appear to breach an EU directive, which carries more force than UK law. This discussion also throws up the related question of whether either of the two existing pilot trials should be extended either in terms of track fitment or by fitting additional trains that spend part of their time running over the existing fitted track. Such decisions have to be considered in detail on a case by case basis. They are also complicated in instances where new franchise operations have not yet been resolved and where they affect the issues.
It is, of course, necessary to again bear in mind the fact that such fitments are not compatible with the longer term European system, ETCS, mandated by European directive, although there may be arguments for a short term benefit by some extension of fitment to track or trains. However, if we consider the example of Thames Trains which run on parts of the Great Western Main Line and would therefore benefit from some degree of ATP protection if they were so fitted, it turns out that such trains spend a very small proportion of their time running over ATP fitted track. The scale of benefit is, therefore, heavily limited and fitment would need to be set against alternative and perhaps more effective investments in train protection. Although it would probably be difficult to produce realistic comparative cost figures, it is fair to observe that the majority of the European railway industry is likely to move in favour of ETCS systems as the standard European solution and any alternative is likely to be costed at a premium. I have seen a copy of the supplementary statement to the Joint Inquiry by Mr R I Muttram which points out in paragraph 19 that there has already been an example of an extremely high quotation for the cost of fitting the Chiltern Line SEL system to additional trains. There could clearly be significant reluctance by the industry to fitting special systems which are not compatible with expected future plans. 3. Progress on TPWS 3.1 Fitment programme I have attended a number of meetings of the TPWS Industrial Liaison Group (last meeting on 11 August) and my overall conclusion is that the programme of fitment remains a tight timetable but there is good confidence that the timescale will be met. The programme has made significant allowance for new problems emerging in the early stages of the project and various problems have indeed emerged. One of the most significant limitations on the rate of fitment relates to the lack of fully trained design and technical staff to undertake the surveying and design work for each signal location. At the time of the last Industry Liaison Meeting it was reported that 1,200 signals have been fully surveyed, 400 signals have been fully fitted with the equipment and 40 signals fully commissioned. (There is naturally a delay on commissioning the equipment in areas where there is no operation of TPWS fitted trains.) I am,therefore, overall broadly content with the progress and it looks like meeting the accelerated timetable agreed late last year. 3.2 TPWS+ A significant factor in my report on ATP was recommending the development of TPWS+ as a way of extending the effectiveness of TPWS to an operating speed of 100 miles per hour. It is important to emphasise that TPWS works at all speeds but that the geometry may preclude the train, under emergency braking, coming safely to rest within the overlap distance. To achieve this for 100 mph requires an additional pair of TPWS loops in an advanced location on the track and, to ensure that the system still works adequately for lower speed trains, requires the original loops to be in place for 70 mph operations. TPWS+ therefore requires two successive speed traps to operate, as discussed in my report. Mention was made in my report of the possibility of further extending this to three speed traps. This is certainly a possibility but I do not believe that anyone is currently working on three speed traps. When work has been undertaken and assessed on a two speed trap
TPWS+ we will have a better idea on how drivers react to multiple speed traps and could at that stage decide whether it could be extended further. Overall, I still have doubts but I do believe it is well worth while developing a two speed trap TPWS+ for speeds operating up to 100 miles per hour. Since my report the TPWS Liaison Group has encouraged further work on the concept and some simple tests have been made and have confirmed that the equipment will operate at speeds up to 125 mph. Plans are now well advanced for fitting 6 signal locations on the ECML for a trial lasting about 6 months. It is expected that the trials of this will start late in 2000 and should enable the operational performance of the system to be assessed. This should enable the industry to evaluate whether TPWS+ loops could be put in place beyond the end of 2002 but before the end of 2003. If this tight timetable were completed it would ensure that TPWS+ would operate satisfactorily up to 100 mph by the time that all trains were fitted. 3.3 TPWS/E In my report I also commented on the possible value of developing the TPWS/E system which would be based on more modern technology than TPWS. In particular, it is a technology which would be fully compatible with ETCS and would therefore provide a suitable migration route to full ATP in a cost effective way which would also speed up the transition to full ATP. Since completing my report a prototype TPWS/E System has been demonstrated by Siemens in Switzerland. I attended this demonstration together with the representatives of Railtrack, Safety and Standards Directorate, and train operating companies. There has since also been a technical presentation by Siemens to the TPWS Industry Liaison Group indicating possible technical routes ahead. This has led to an industry wide workshop on the concept and a report recommending an early technical trial in the UK. Funding for this stage is currently being sought. I remain convinced that this could well be an effective way ahead. 3.4 Limitations on TPWS There are, of course, many examples of accidents which neither TPWS or TPWS/E could prevent. Quite apart from the problem of vehicles travelling at high speeds (ie. 75mph for TPWS and 100 mph for TPWS+) or vehicles that are not equipped with adequate braking (ie. 12%g of braking rate) there is a theoretical possibility that a driver braking to bring the train speed profile below that of the speed trap could then accelerate immediately afterwards. It is difficult to handle this unusual and indeed suicidal situation (after all, the driver is in a vulnerable position in the front of the train). It is important to point out that even in situations in which the operation of TPWS does not succeed in bringing a train completely to rest within the overlap, in most cases, the speed of the train in any collision will be dramatically reduced so that the severity of an accident will be substantially reduced. Despite these limitations the analysis in my report indicated that more lives could be saved by adopting the TPWS solution and migrating to full ATP later. Nothing new has emerged to change that analysis.
4. Progress on ATP Trials Both the Chiltern Line Trials and the Great Western Line Trials have moved to successful operation. The Chiltern Line ATP System was accepted by HSE for incorporation into normal operational service in January 2000. Application has been made to the HMRI to accept the Great Western System for operational service and the response is awaited. Substantial effort has been put in by both train operating companies to achieve this performance. It is, nevertheless, significant to recognise that this has happened about 10 years after the start of the trials. However, many lessons have been learned and many of these can be carried across to trials of ETCS. This is particularly important since Railtrack have now set aside an experimental site for trials of ETCS prior to its first installation on the West Coast Main Line. The documentation of problems associated with the two sets of ATP trials have been set out in other references (eg. reference 4) but the overall message again is that it is far easier to install such systems in new rolling stock than add it to existing stock particularly where the wiring is old and needs significant modification. In this respect I am very pleased to note that the S&SD are introducing regulations to ensure all new stock will be fitted with appropriate location spaces for ATP to ensure that they are ATP ready as they come into service. It would be helpful to extend this to specifying necessary wiring for ETCS, but this is more difficult because different companies in Europe produce different versions of ETCS that have different wiring requirements (though they must all be compatible with each other for interoperation). The issues of possible extensions to the trials was covered briefly in section 2.2. This topic also relates to attempts to make the trials compatible with future ETCS and this will be discussed in section 6. 5. Systems Authority My report made reference to the problems created for any Project Manager or Project Team needing to have overall responsibility to implement a major system such as ATP or TPWS on the current railway system. This was on the grounds that no individual or project team could hold the necessary authority to implement such a project across the fragmented railway industry. To do so would require all related railway companies to grant the project and financial authority to such a body (without even knowing the extent of the financial commitment). Since my report (and that of Professor Uff on the Southall Inquiry, reference 4) there have been many discussions on the possible need for and form of a system authority for such projects. This led to a one day workshop on the topic (looking primarily at the issue for TPWS) earlier this year. There are still significant differences of opinion across the industry on the need for such an authority and how it could be implemented and I have set out below my own observations on the topic. By way of example, consider the issues involved in the creation of a systems authority for a railway administration where all parts of the industry are under common ownership (eg. such as British Railways). It would be possible for the Chairman of the Board to authorise the creation of such a body on the basis that it could consult broadly on the design of a new system and be given financial authority across the whole area of the railway administration. It would presumably
take decisions based upon the best solution for safety or efficiency of the system in such a way that would minimise overall cost but would not have to concern itself that decisions would sometimes increase costings in one area of the organisation at the expense of others. However, when we consider the same situation for the current fragmented industry there are many options that arise with a new development where difficult decisions have to be made (even on issues of detail). Alternative solutions to problems can result in heavy financial costs falling on one part of the industry (for example, a train operator or train leasing company) as compared to costs falling upon Railtrack or upon a freight operator. In the absence of a single overarching authority it is very difficult to decide a suitable and fair basis for how such costs would fall. To some extent this is not very different from the problems facing S&SD (or the new company Railway Safety) who have to choose the parameters of a Railway Group Standard which can cause costs to fall on different parts of the industry. Some have argued that it may be possible to introduce legal requirements to enforce the detailed introduction of safety systems but even if this is true it is difficult to see how alternative (often detailed) ways of meeting the costs could be agreed without subsequent agreement of all the various parties. I am, therefore, not convinced that it is a practical projection to impose such a systems authority on a fragmented industry. It can only operate by consensus and it is not possible to legislate for consensus. Despite the above I have been impressed with the way things have worked out within the TPWS Industrial Liaison Committee. This has effectively worked as the systems authority for TPWS. As a group it does not have a single authority to commit funding but provided all the individual members of the groups (representing all the relevant industries) can come to a common conclusion then the system implementation can, and indeed has, progressed well. It is, of course, possible that this has happened only because a major problem has not arisen. It is not evident to me that there is a feasible way of resolving some problems without some enforced authority across the industry. It has also been pointed out that the whole process would be aided significantly if there was a proportion of central funding (perhaps derived by a levy on the whole industry) which could be used to take some decisions forward (at least by doing studies). I therefore believe that the future for systems authority is still in question and requires additional work but it is interesting to report on the success of a particular way of achieving it with TPWS. Currently the Office of the Rail Regulator is working with S&SD to examine the way forward. This will probably involve a Railway Group Standard on Systems Authority. The same will probably apply to forms of Design Authority representing a body that would monitor and control the continuing integrity of a system through its working life. On the basis of the above, I suspect that the solution must depend to some extent on collaboration and agreement between the parties and I will be watching the progress with interest. 6. Progress on ERTMS/ETCS European collaboration on the Europe wide ETCS system has progressed well and international agreement has now been reached for the functional specification for ETCS. This has been reported in the technical Press (reference 5) and a CD Rom has been issued with the full specification. There is also a complex programme for demonstration of concepts and operational trials planned for ETCS in parts of Europe including France, Germany, Italy, Spain and the UK.
In the UK the West Coast Main Line will be the first major installation of ETCS, which will be implemented at Level One and Level Two. Before such a major installation can be taken forward it is, of course, necessary to have some experimental trials and a special test track is being set aside and equipped for this purpose by Railtrack. It is clear that the railway administrations in Europe are strongly behind ETCS in terms of heavy political support. There are, nevertheless, some pockets of doubt where there is a major investment in current national ATP systems which still have a long life to run. It is important to appreciate that there is no single technical specification for ETCS but a functional specification with specifications for interoperability. Thus different countries with different signalling contractors will develop their own systems which have to be approved to be compatible with the international standard and hence capable of inter-running on the various national infrastructures. Looking at installations on the West Coast Main Line it will be important to ensure that the track infrastructure could accept subsequent future inter running from other trains although this is assumed to be constrained for many years to Eurostar plus appropriately fitted freight locomotives. It is considered unlikely that the trains running on the West Coast Main Line, although equipped with ETCS, would wish to inter-run to other countries in the foreseeable future. However, it is possible that some of them will wish to run over larger areas of the railway including the GW Main Line. In this case some interface unit between ETCS and the GW version of ATP may be needed. Such interfaces are permitted within the overall ETCS specifications. There is still considerable uncertainty about the time for evolution to Level Three Systems. Although a number of experimental demonstrations are planned based on level 3 systems, they are not expected to include significant numbers of trains fitted. It will be necessary to get relevant experience of the extent of high performance radio coverage needed for Level 3 type operations before taking things much further. In my opinion, the timescale for a full operational trial is probably about 10 years away. I mentioned in my report that developments in navigation systems, such as GPS satellite navigation, could also have applications in this area. I am pleased to hear of a European project called Project Apollo aimed at examining the relevance of modern developments in navigation to the next stage of ETCS. Again, this will depend on trials of radio coverage and also satellite communication coverage but the potential benefits come in terms of substantially reduced trainborne costs and far simpler trainborne equipment. 7. Funding for Research In my report I suggested that it was important to develop an enhanced programme for research to support both safety systems and ways to enhance the overall operational performance of the railway and that either the Strategic Rail Authority or Railtrack might well be the suitable organisations to take this forward. I am pleased to report that the bid for funding the new Railway Safety company includes a significant allocation of funding for safety related research. I also note that the funding for Railway Safety, which has been included in the strategy announced by the Rail Regulator for the 5 year period from 1 April 2001, includes this research funding. This programme includes a variety of safety related research but does not include work aimed at increasing capacity or performance such as Level-3 type systems. It is assumed that this will fall to the Strategic Rail Authority although I have no indication yet about their interest in taking this forward.
8. Conclusions Overall, I am relatively satisfied with the progress that has been made in the six months following the publication of my report. I recognise that the Government felt unable to give a formal reaction to my report because this could appear to anticipate the outcome of the joint inquiry to be started in the Autumn of 2000. Despite that, I am pleased that Railtrack, HSE, the Safety and Standards Directorate and several other parts of the railway industry have moved on their own initiative to take many of the recommendations forward. I look forward to hearing the reaction of the Joint Inquiry to the same issues. References 1. Automatic Train Protection for the Railway Network in Britain A Study (D E N Davies) published by The Royal Academy of Engineering, February 2000. 2. Railtrack s Safety and Standards Directorate Review of Main Functions By DETR. February 2000. 3. Proposals for an enhanced version of existing AWS Equipment Howells Railway Products Ltd. Private Report. 4. Inquiry into the Southall Accident by Professor J Uff Published by HSE 2000. 5. Railways Agree ERTMS Specifications J Claudo Traveso IRJ May 2000. P 29-52.