In issue 180, (December 2019), Malcolm Dobell reported on recent experiences by train operators in introducing new trains, as presented to a recent IMechE’s Railway Division seminar.
This month, he considers the comments made by the Rail Delivery Group, infrastructure owner Network Rail, safety organisation RSSB and rail regulator the Office of Rail and Road (ORR). He also looks ahead to two new projects that could have an impact on travel in the coming years.
Mark Molyneux from the Rail Delivery Group (RDG) explored the reliability impact of so many new trains. He presented this against a background of a) generally improving fleet reliability, b) that about 15 per cent of delays are down to fleet issues and c) new trains may take a while to grow their reliability, but are, generally more reliable than the trains they replace.
With about half the national train fleet being replaced, the industry is rightly concerned that overall fleet performance will take a hit before recovering. In trying to understand whether this was inevitable, and what could be done to help, RDG has developed a guidance note, New Trains – A Good Practice Guide, which is currently out for consultation.
Typical issues highlighted include software function and validation (it is apparent that software developers and train engineers do not understand each other!), the need for clear pass/fail criteria for the demonstration of compatibility, and improvements to specifications and contracts so that trains are specified in terms of reliability, against realistic timescales and using industry guidance such as Key Train Requirements and Key Interface Requirements. All this is necessary as it is unacceptable to subject the railway’s customers to debugging in public!
The seminar turned to standards, infrastructure and system compatibility issues.
Rashid Wahidi from Network Rail introduced the chief engineer’s Vehicle Introduction Forum, a group which was unfamiliar to most of the IMechE audience.
It was set up in late 2017 with a number of objectives: to help develop good practice amongst those dealing with the unprecedented quantity of new vehicles being introduced over a short space of time, to assist new manufacturers entering the UK market without experience of the UK’s unique infrastructure constraints on compatibility, to guide new and established operators introducing vehicles for the first time under the current legislative regime, and to promote collaboration, shared learning, and the establishment of best practice to make the introduction process as smooth as possible.
All suppliers, lessors, lessees, key stakeholders and assessment bodies generally attend. The group has successfully produced a tracker covering all planned vehicle introductions – both new and cascade – so that possible network/resource constraints can be identified, developed alternative gauging and signal sighting assessment processes, initiated the development of a new method for pantograph encroachment assessment and kicked off a Key Interface Requirements document for people introducing vehicles to a route covering all areas to support a successful introduction. The latter will complement the industry’s Key Vehicles Requirements document.
Hugh O’Neil from RSSB described the Route Compatibility process outlined in Rail Industry Standard RIS-8270-RST: Route Level Assessment of Technical Compatibility between Vehicles and Infrastructure. To “put a vehicle into use”, three separate activities must be undertaken: a) a new vehicle must be “put into service” – a process to demonstrate that a rail vehicle is fundamentally safe as a rail vehicle and complies with the Technical Specifications for Interoperability, Notified National Technical Rules, b) it must be demonstrated that is it compatible with the infrastructure on which it runs – effectively demonstrating compliance with RIS 8270 – and c) it is integrated safely onto that infrastructure and all the other systems – technical and operational – used to manage residual risk.
The accountability for all this is firmly on the Duty Holder – the train operator (TOC) – although responsibility for a) and b) might be sub-contracted to the vehicle supplier. Hugh emphasised the accountability because, although Network Rail is necessarily heavily involved in the compatibility assessment, it is the TOC that must issue and sign the Statement of Compatibility. Stakeholders – Network Rail and, generally, other operators using that route – must be consulted, but none of them have any approval right.
When a Statement of Compatibility is issued, all documentation needed to define any limitations, restrictions or requirements on which the compatibility depends has to be updated, the data that describes asset characteristics relevant to compatibility must be maintained, updated and made freely available to relevant parties, and the infrastructure or vehicles must be maintained within the characteristics on which compatibility depends.
David Galloway from Network Rail carried on with the Safe Integration process. David defined Safe Integration as “the action to ensure the incorporation of an element (a new vehicle type, network project, subsystem, part, component, constituent, software, procedure, organisation) into a bigger system, does not create an unacceptable risk for the resulting system.” He said that Safe Integration of a rail vehicle with the local route or local characteristics is an essential part of meeting safety obligations under Railways and other Guided Transport Systems (Safety) Regulations 2006.
As an example, if a train with driver-controlled doors is being used on a route for the first time, each platform risk assessment will require to be reviewed and changes might be necessary properly to control the risks the assessments might identify.
On electrified routes, longer, faster and more powerful trains are increasingly the norm and this is great, unless perhaps, you are the power engineer making sure that there is “enough electricity in the tank”. Alex Buchinger presented the challenges facing Network Rail in providing power for new trains. He outlined the modelling and assessment processes used to evaluate traction power needs and highlighted the many factors that have to be considered involving the electricity supply industry, the HV and/or overhead line equipment and, where relevant, the DC system. He also referenced system enhancements to support more demanding trains and the compatibility issues that sometimes have to be resolved such as harmonics induced into vulnerable circuits and resonance.
The most challenging message of all was the warning that some upgrades can take five years to organise after the requirement has been identified. This timescale may be longer than the time from the contract being let to the first train into service for some rolling stock procurements, so it begs the question “who can take the joined-up view when planning future power capacity requirements?”
Paul Hooper, head of interoperability and rail vehicles at the Office of Rail and Road (ORR), presented some of the problems and issues observed on trains that have been authorised into use. He explored nine questions:
- Why are TSI ‘compliant’ new trains revealing ‘hidden’ issues at or after authorisation?
- Are they really ‘hidden’ before authorisation? Should they have been identified earlier?
- If so, who should have identified them and at what point? During hazard ID? Design reviews?
- Is the problem actually with the Technical Specifications for Interoperability / Notified National Technical Rules conformance process or compatibility with legacy infrastructure? Is it the manufacturers’ or the operators’ problem, or shared?
- What is the role of the infrastructure manager?
- Are the issues inside or outside of authorisation process or straddle it through safe integration?
- Is Authorisation being used as a milestone to pay manufacturers? Is this part of the problem? An Authorised train which cannot be put into use is the same as an unauthorised train unless you are a manufacturer and want payment. In either case it cannot be used in service but could drive the wrong behaviours.
- What could be some other contributory factors giving rise to these unintended consequences?
- Should the Authorisation process / CSM be used to capture the residual issues and is it right that ORR feels like it is ‘refereeing’ the resolution of these issues?
Paul was conscious that past trains have had problems in service, but more recently, issues have occurred on fully TSI compliant trains that have been extremely serious; he cited the Brussels-Amsterdam high-speed FYRA trains that were eventually returned to their manufacturer is probably the most high-profile example.
Closer to home, Paul outlined other failures such as failures of the service brake on two different fleets where there was no safety integrity level defined for the software, and the issue with ghost images in the windscreen of another fleet. Are these genuine emergent properties or are there weaknesses in the TSIs or supporting Euronorms?
He turned to other issues that are ill defined in standards using stepping distances and the climbing and inter-car surfing risks caused by the umbilical cables between cars being set at various heights up the car ends as examples. Did these risks get identified early enough in the hazard identification/risk management process for appropriate attention during design process? Perhaps designers from other countries were more used to a better-behaved local population? He also mentioned the problem that the train fire safety standard BS6853 was superseded by EN45545 even though the requirements for seats in the latter had been shown to be inadequate.
He mused whether delayed infrastructure increases risk because vehicle production schedules get re-jigged, or people might be pressured to take short cuts, or that shortage of test facilities might lead to test not being sufficiently exhaustive, or the TSIs do not address all risks
Paul explained what the ORR can do. He said that when new trains are submitted for authorisation as ‘compliant’ and can theoretically pass the process, ORR use tools in the regulations such as conditions and limitations to manage non-compliances or, indeed, residual issues. He added that they use the safety assessment report produced by the independent assessment body and ORR decisions around safe integration and the Essential Requirement ‘safety’ to ensure that transferred risk has been properly accepted. He noted that ORR can only revoke authorisations before the vehicles are put into use but can and will use Health and Safety at Work and Railways and Other Guided Systems powers once trains are in service, and that Network Rail is becoming a concerned landlord over some new trains’ issues and application of the compatibility process.
Paul had asked other EU National Safety Authorities (NSA) for their experience. One NSA said it had retained additional national rules and had given its infrastructure manager a stronger permissioning role. He added that other NSAs are more intrusive in their authorisation assessment. However, apart from the FYRA project mentioned above, Paul said that Europe is not seeing the sorts of issues experienced in UK – why?
In attempting to answer this question, Paul observed that the UK has generally been a good European by adopting the TSIs and weeding out national rules. Could this perhaps have gone too far, taking our eye off the ball and letting safety-by-design lapse in some areas? Could it be that new trains are too sophisticated for the ‘standards’ regime to identify risks at the specification and design stage? Indeed, is the standards regime lagging behind new innovations? The ORR is being approached regularly on topics such as dual fuelling, hydrogen powered trains and very light rail, while the TSI for locomotives and passenger stock has few requirements for software.
Paul concluded by wondering whether the changes to regulations over the last decade have led to unintended consequences, whether current UK specifications are robust enough, whether operator involvement at the design stage is being stifled, and whether manufacturers are missing some skills to thoroughly review designs. He also observed that some changes to TSIs in areas of crashworthiness have, in turn, led to bigger inter-vehicle gaps, slimmer longer vehicles and increased stepping distances with legacy infrastructure.
This led to Paul’s final question with probably a deliberate double meaning: “What should be done to manage gaps that have appeared which lead to unintended consequences?”
Your writer would observe the old adage that “standards are for the guidance of wise men (people) and the observance of fools”. Determining the goals – the purpose and function of a train – and then using the standards to help achieve the goals is much more challenging than simply assessing that standards have been complied with. It requires an experienced project team, that is knowledgeable about how the trains will be used and about rolling stock engineering, that knows the right questions to ask and that engages with the train’s designers. This is a different proposition from only buying consultancy services that assess compliance with specifications, TSIs and NNTRs.
And now for something completely different
There followed two presentations on projects that are at quite different stages of development. The first was from Helen Simpson from Porterbrook Leasing about the prototype HydroFLEX unit (issue 175, June 2019) which is a partnership between Porterbrook and the University of Birmingham. One Class 319 unit was converted to hydrogen and battery power in just nine months and demonstrated at Rail Live in June 2019.
Helen’s presentation started with how low-carbon energy sources are developing in UK and, particularly, how wind farms generate more electricity than is needed overnight when demand is low. Essentially, this surplus energy could be stored if suitable storage systems were available, and one such means is to create hydrogen, which leads to comparatively low cost.
The UK’s hydrogen creation and distribution network is small, but growing. Helen focussed on safety and approvals for the train, especially the next stage for main-line running. She said that many of the issues she expected to be problems have proved not to be. For example, there are well-developed standards for the use and transport of hydrogen that can be adopted or adapted for rail, making the assurance process a little easier.
Helen said that the team had learned a great deal about the practical performance of battery and fuel-cell powered trains, especially how best to deal with peak demand on actual and simulated journeys. She added that she had become quite expert in some areas, such as the risk of escaping hydrogen on a train powered from the OLE where electrical sparks are an inevitability. In essence, Helen said that the issue to manage is having hydrogen escape into confined spaces – once in the atmosphere it disperses to a very low dilution incredibly quickly.
That said, there is still much to do though the hazard identification and risk assessment process to identify criteria against which the acceptability of the train for main line running can be judged. InnovateUK funding has now been granted for a main line trial around February 2020.
DLR comes of age
Finally, there was a presentation from Phil Shrapnell and Andrew Dunsby from TfL Engineering about the new trains for the Docklands Light Railway (DLR) that have recently been ordered from CAF to replace the current B90/92/2000 fleet which are near the end of their 25-year design life. DLR is growing, and Phil said that 43 trains have been ordered. This includes ten extra trains for growth with an option, which he believed might be exercised imminently, for a further 14 trains for further growth.
Current DLR vehicles are two-section, three-bogie articulated cars run in pairs or threesomes. It is probably a sign that Docklands has come of age as a fully-fledged metro that the new trains will be five-car trains with two bogies on each car. The trains will be just under 88 metres long and will have wide, open gangways between cars (nominally just over 17 metres long each), which will lead to long inter-car gangways as DLR has minimum curve radii of 40 metres including reverse curves on crossovers.
Other features include resilient wheels with inside frame bogies, liquid flange lubrication, three motor cars, two doors per side on end cars and three doors/side on intermediate cars, ADO for a few short platforms, air conditioning, interior CCTV system, comprehensive passenger information system and VDU screens for adverts, LED lighting adjustable for time of day, USB sockets, and remote condition monitoring.
An unusual feature is the requirement for obstacle detection, useful on a railway where the staff member on board is rarely at the front of the train looking out, although Phil acknowledged that making such a system compatible with some of DLR’s sharp curves will be a challenge. Some of the trains will also be fitted with unattended track monitoring equipment.
Phil talked about some of the challenges in addition to the previously mentioned 40-metre-radius curves. A specific issue is designing the interfaces for Passenger Service Agents. These include controls at each doorway – enabling control, door close and door reopen controls as well as a fixed microphone, and the emergency driving position. Making these suitable for 5th percentile female to 95th percentile males is a requirement. Phil concluded with the programme dates of mid 2020 for the final design review, testing to start in 2022 including 20,000 km testing on a Spanish test track that will be equipped with Thales Seltrac for the purpose. And finally, delivery of the first train to DLR by December 2022.
This was a really stimulating seminar, and illustrates that scale of the challenge that the UK rail industry has over the next two to three years.
Congratulations are due to the IMechE’s organising committee, led by Martin Elliott, chief technical officer of Ricardo Rail, for organising the event, and to the unusually candid speakers.