Select Committee on Transport Written Evidence

Memorandum by Messrs R M Bell, D A Marshall and P G Rayner (FOR 126)



  The purpose of this paper is to put before part 2 of the inquiry a constructive proposal for a better way of managing operations and safety on the national railway network.

  The objective is to propose the creation within the industry of a "controlling group"—referred to here as the Railway Operations & Safety Executive (ROSE)—because the authors (see below) consider that the most fundamental issue facing the industry is a lack of clear responsibility for the safe running of the network.

  Much attention has been given to achieving safety through regulation and processes. But we feel strongly that the industry should be structured to regulate itself taking account of the interdependencies and inter-relationships that are necessary—both between organisations and within organisations comprising the industry's complex structure—to ensure the industry can function safely and efficiently, and can provide a reliable and dependable service to its customers.

  This paper considers what the authors consider is a structure to ensure safety through clearly defined responsibilities for railway "Operations"—defined as everything concerned with any movements taking place on the railway network.

  It then looks at how the proposed structure can fit into the existing industry framework as simply as possible. We hope that by concentrating safety responsibility in the heart of the industry, the task of the regulatory authorities can be made less onerous.

  The paper does not suggest fundamental changes in the role of HMRI, (s)SRA, or the Office of the Rail Regulator, nor does it suggest any further external controls of the industry.


  Two of the authors of this paper have had a long career in railway operations and engineering and in recent years have together investigated the background, causation and implications of the Watford, Southall and Ladbroke Grove accidents. Out of those investigations, and through the production of other papers concerning Train Protection Systems and Cost Benefit Analysis, it became clear that major problems existed in the Industry. In addressing these issues, Alan Marshall—a colleague manager for over 15 years, has joined Roy Bell and Peter Rayner, he has had a long career in public affairs, and also contributed to the review of Cost Benefit Analysis (and was recently made Corporate Communications Director by the International Council of the Chartered Institute of Transport).


  It is significant that safety arrangements in the railway industry are now having to be imposed on a plethora of private companies in a manner not applied since the halcyon days of railway development in the 19th Century. . . when major accidents and serious concerns about safety led to significant legislation being enacted by Parliament. Since then—and especially in the 70 years following the formation of four privately owned mainline railway groups in 1923—total responsibility for railway infrastructure, operations and safety was placed with the "vertically-integrated" railway service provider(s). From 1948 until 1994 safety on the railways was overseen by British Railways—monitored by Her Majesty's Railway Inspectorate—through Regional organisations, not dissimilar to the geographical boundaries of 1923-47 "Big Four" railway companies—and even today Railtrack production arrangements (eg, train planning, weekly notices, etc) still reflect the "Big Four" geographic boundaries.

  It is a simple fact that the present arrangements, stemming from the 1993 Railways Act, tend to hinder or prolong timescales. There is a large number of parties involved, each depending on occurrences, and having vested interests:

    —  Railway industry companies (various, approximately 100 in total).

    —  Her Majesty's Railway Inspectorate (HMRI, part of the HSE).

    —  Police—Home Constabularies.

    —  British Transport Police.

    —  Crown Prosecution Service (Procurator Fiscal in Scotland).

    —  Coroners.

    —  Emergency services.

    —  Insurance companies.

    —  Lawyers of various sorts.

    —  Injured parties.

    —  Equipment manufacturers and suppliers.

    —  Equipment installers.

    —  Equipment maintainers.

  There is a good management adage that says "You cannot control what you cannot measure"—and this sums up the Safety Procedures of the present railway industry structure.


  The best and maybe the only way to run a railway efficiently, punctually and therefore safely is to create a vertical chain of command and for everyone in that chain to know who the overall boss is. The key players in a railway are the signallers and the drivers: to have them trained, certified and managed by several different organisations is a recipe for inefficient (and therefore unsafe) operation. They need, and arguably should have, the same person in charge of them and feel part of the same organisation. The Railway Control Organisation on a day-to-day basis needs to be geographically based and cover all trains—Express, Local, Freight, Postal and Suburban networks—for all interact, and all in some way affect each other.

  Using an orchestral analogy, it is the overall music that needs conducting—not the playing of separate instruments.

  The railway is a dangerous place at the interface between train and track and to inculcate and achieve a strong safety culture there is a need for strong leadership with sufficient powers to run the total operation.

  With power comes responsibility. So Train Control should be part of the Railway Operating & Safety Executive—which would function as a sort of railway equivalent to the Civil Aviation Authority—while Railtrack would continue to develop and maintain the infrastructure and manage its contractors . . . just as BAA Ltd operates its Airports, and Air Traffic Control manages the safe operation of aircraft movements—while the CAA oversees the allocation of "slots" (the equivalent of rail industry timetable "paths") and safety standards.

  To set up an industry-wide operations and safety organisation, the country needs to be divided into geographically logical, route-based pieces, not more than six in number. The leaders of these "Pieces" should be responsible for the actions of all train movements, responsible for all operating staff irrespective of the organisations by whom they are employed.

  The responsibility for the safety of the infrastructure must be clearly defined: it cannot be spread among a number of contractors or sub-contractors. There are two possible ways we think this can be achieved. Firstly the responsibility could be carried by the Railway Operating & Safety Executive. It would need to employ engineers of experience in the respective technical disciplines to ensure all the maintenance contractors were performing to the required safety Standards. Alternatively, resources within Railtrack could be strengthened to perform this role, which would need to be much deeper than the present contract management and the current audit procedures.

  The conflict between the need to grant track possessions for work to be carried out against the need to deliver a punctual and reliable service to customers would have to be the responsibility of the Railway Operating & Safety Executive, which would necessarily have to have overall authority for the national timetable.

  However this is achieved, the engineers responsible for infrastructure safety—whether within Railtrack or the Railway Operating & Safety Executive—would need to have the power to close lines or apply emergency speed restrictions if needed for safety reasons. A properly maintained railway infrastructure would mean that this would be a rare event, and normally all engineering work would be pre-planned under the leadership of the Railway Operating & Safety Executive.


  There are at present too many boundaries between the different companies and agencies in the railway industry. Despite the interdependencies and inter-relationships within the industry these boundaries are often not coterminous, and are often unclear, yet it is these boundaries that are really essential and need to be made clear.

  We believe the present muddle needs addressing and that there is an essential need to reinstate a professional level of operational and safety management across the railway industry.

  Railtrack should do what its existing resources have been developed to undertake—ie, to act a procurement and asset management agency—while leaving a new focused organisation to look after industry-wide railway operational and safety issues.

  The new organisation should function as a Railway Operating & Safety Executive, also undertaking detailed regular safety auditing for which the industry should be responsible (and for which HMRI is not resourced, anyway)—a body monitoring and advising the whole railway industry. HMRI would then be able to continue its sample audit and independent inquiry roles on behalf of the HSE.

  Similarly, this new Railway Operating & Safety Executive should be responsible for accepting and certifying new equipment and systems which are then ratified by HMRI on behalf of the HSE, and setting the training standards and providing certification for those involved in safety-critical roles.

  The accident/incident inquiry process also requires to be modified. As a result of the structure brought in since the rail industry was prepared for privatisation, prosecution of offending parties has been seen to take precedence over the public inquiry (or publication of internal inquiry reports and recommendations), and until the inquiry is concluded coroner's inquests cannot proceed where fatalities have occurred. Yet, without an inquest verdict, victims' families can be left without compensation, or can only then begin arguing for compensation. It is significant that the manslaughter charge arising from the Watford train crash in August 1996 was not tried until March 1998. Only after the jury's "not guilty' verdict did HMRI publish the findings and recommendations of its inquiry, while almost two years after the accident, the inquest into the one fatality at Watford had still not been held. In the case of Southall, Professor Uffs inquiry was delayed for two years by the manslaughter charges, etc, and the issues relating to train protection have only just been considered at the Joint Inquiry three years after the Southall crash occurred.

  Compare these circumstances with those following the tragic accident at Escheder, Germany, in 1998—after which the first report was placed before the German parliament within 15 days.

  A major, delaying problem in Britain now is that the body responsible for industry inquiries (Railtrack) is inevitably in difficulty in the conduct of these inquiries for, in most instances, Railtrack starts as one of the "suspects" in any police investigation and, therefore, must be inhibited in the way it can conduct an inquiry, including accessing witnesses, etc. The proposed Railway Operations & Safety Executive, however, would not suffer from such difficulties.


  We do not believe any further upheaval of HMRI is sensible or required. We do believe, however, that HMRI should become more active in monitoring safety trends—thus enabling meaningful statistics to be provided year on year. To facilitate this we recommend that any reporting process (eg, RIDDOR—the Reporting of Injuries, Diseases and Dangerous Occurrences Regulations 1995) should include agreed key safety indicators. A list would need to be agreed with HMRI. Examples of possible Key Safety Performance Indicators (KSPIs) are:


        Broken rails

        Signals passed at danger (SPADs)

        Main/running line derailments


        Level crossing incidents

        Trackside near misses (of personnel)

        Signalling irregularities

        Wrong-side signalling failures

        Passenger vehicle and freight wagon/container door incidents

        Suspected suicides

        Dangerous trespass


        Signal box inspections

        Signal sighting committees

        Driver inspections (drivers ridden with)

        Safety Case compliance


        Competency Assessments

        Rolling stock depot audits

        Ultrasonic rail inspections

  Once the KSPIs are determined, train companies and Railtrack should be required to report the necessary data (including nil returns). Currently, the "by exception" method of reporting assumes everything is okay unless otherwise advised and could result in data being missed out of the system. It is essential that a pro-active system be adopted, quickly. In particular, "near misses" should be investigated in much greater depth, as events leading up to Ladbroke Grove clearly demonstrated.

  By adopting a pro-active approach, when a rising trend in unsafety becomes apparent, the Railways Operating & Safety Executive (ROSE) will take the lead in reversing the trend. Only in very rare cases will HMRI have to use Improvement Order regulations. Hence, with an issue like SPADS we would in the future expect ROSE to achieve quick results to prevent tragedies occurring, such as at Watford, Southall and Ladbroke Grove.

  We would also expect ROSE to be able to direct other parts of the industry on a wider basis, including a collective group of interdependent and inter-related organisations—such as a Railtrack Zone, one or more TOCs and the Railtrack's infrastructure contractors to ensure that remedial safety measures are undertaken promptly—for example, to overcome an adverse trend in the incidence of broken rails. We would not expect such a problem to require use of the Improvement Notice process.

  Finally, it is necessary to ensure HMRI is adequately staffed to carry out what it does well—which is monitoring the industry and conducting important inquiries, also responding to trends of a disturbing nature, such as vandalism and road driver abuse of level crossings.

  Of course, to be adequately staffed, HMRI must be adequately funded by the Government. As all parties to the industry stand to gain from improving safety we would suggest one means of funding a Railway Operations & Safety Executive (ROSE) would be through a levy on all holders of Railway Safety Cases, based perhaps on their turnover in the rail industry.


  As a first step to the setting up of a geographical organisation with comprehensive control over train movements, the creation of a Control Centre with overall day-to-day authority will start to bring clarity back to the interdependencies of the system.

  The Railway Operating & Safety Executive, with its geographic territories, will need to have a competent Inspectorate to monitor drivers and signallers alike, thus overcoming the inadequate training and monitoring that showed up so badly at both Southall and Ladbroke Grove in relation to driver standards, also with regard to the Slough Signalling Centre.

  As to the Executive's Train Control function, its empowerment should include the ability to alter services, introduce stops and act in the common good to maximise customer (passenger and freight) satisfaction.

  To set up such a Control is feasible. By way of illustration Roy Bell and Peter Rayner, with John Edser, produced a paper to the Institution of Railway Signal Engineers (IRSE) at the "International" Conference on Railway Control, "Aspect 91" (read by P Rayner). The Paper was titled "Railway Operational Control—the IECC and the next 10 years".

  Clearly the interface on stations and on trains between retail/customer service staff and staff with additional Operating/Safety responsibilities is an area requiring careful consideration. The advantages of staff remaining with the different companies for administration, salary, uniform and other conditions is accepted; to transfer all staff with operational/safety responsibilities to the ROSE would be a step back in time and could restrict innovative ideas and not assist safety. Rather, the ROSE should set the training standards of such staff and ensure their on-going competence and compliance with laid-down operating and safety procedures.

  The arrangements in place with the airlines are an example for, provided cabin crew are trained for example in evacuation procedures (see Maidenhead Fire railway inquiry), there is no need for them all to be with the same employer.

  Although safety improvement might be achieved by putting all train drivers under one organisation, the benefits might be long term and, in any event, any such proposals would need careful risk assessment. But it should be noted that the present company-based working arrangements can lead, albeit unintentionally, to boredom and breed complacency which experts in Human Factors might argue to be safety issues, and the ROSE would need to address such issues. However, to suggest a transfer of drivers and conductors to a central authority is not recommended. Nonetheless, all train crew when operating services must obey the rules of the Railway Operating & Safety Executive whilst wearing the uniforms of their respective employers.

  The proposal is not dissimilar to what happens in many retail stores, which have many companies trading within them in franchised areas, but the stores' overall owners control these companies' operations. This is analogous to the way we see the Railway Operations & Safety Executive functioning, with accountabilities to the Secretary of State through the Strategic Rail Authority and the HSE. It should also have a Management Board comprising representation from within the rail industry, but also with strong non-executive membership, too—an extension of the principle already adopted for Rail Safety Ltd.

  The funding required for ROSE should be drawn from industry wide owners who would pay for it as they pay now for the Police. Indeed signallers were always called "Bobbies" as indeed they started off as Policemen.

  If this structure is considered workable, it could fit into the present industry structure without creating significant changes to the Safety Regulators, or to the (s)SRA or the Office of the Rail Regulator, and the new operating and safety organisation would, hopefully, quickly earn the confidence of the public, the industry and the HSE (which could go back to the monitoring and supervisory roles that it previously excelled at). The new executive, of course, would ensure consistency and comparability of the required re-active and pro-active statistics.

  Railtrack would continue to own and maintain the infrastructure, employing accredited and competent contractors. Train operating companies would continue to own or lease their rolling stock and to be responsible for its maintenance. Signalling and train planning staff (Timetable compilation), together with Rail Safety Ltd., would transfer from Railtrack to the Railway Operations & Safety Executive, which would then be able to provide the industry with a skilled and dedicated expertise in operations and safety.

R M Bell MBE

D A Marshall

P G Rayner

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2005
Prepared 11 May 2005