Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Concept development and testing of the UK’s first hydrogen-hybrid train (HydroFLEX)

Concept development and testing of the UK’s first hydrogen-hybrid train (HydroFLEX) Rail. Eng. Science (2021) 29(3):248–257 https://doi.org/10.1007/s40534-021-00256-9 Concept development and testing of the UK’s first hydrogen- hybrid train (HydroFLEX) 1 2 1 1 • • • • Charles Calvert Jeff Allan Peter Amor Stuart Hillmansen 1 1 Clive Roberts Paul Weston Received: 14 February 2021 / Revised: 29 July 2021 / Accepted: 18 August 2021 / Published online: 1 October 2021 The Author(s) 2021 Abstract In October 2018, Porterbrook and the University of 1 Introduction Birmingham announced the HydroFLEX project, to demon- strate a hydrogen-hybrid modified train at Rail Live 2019. The The need for countries to take action on climate change concept of modifying a Class 319 Electric Multiple Unit was was emphasised by the Paris Agreement of 2015 [1], which developed, with equipment including a fuel cell stack, trac- requires all parties to keep global temperature increase tion battery, 24 V control system and hydrogen storage ele- below 2 degrees centigrade, and to pursue an increase ments to be mounted inside one of the carriages. This was within 1.5 degrees centigrade, above pre-industrial levels. followed by procurement of a fuel cell stack, traction batter- Climate change impacts, notably the increased frequency ies, and control equipment, which was then installed inside and severity of extreme weather events, can affect the train, being fixed to the seat rails. One substantial change numerous aspects of human civilisation, and so mitigation from the concept was the provision of considerably more is essential [2]. hydrogen storage than the minimum necessary, providing the In order to achieve this, greenhouse gas emissions must train with more potential to be further modified to allow for be lowered, and the Great Britain’s railway is making pro- higher speed mainline testing. After the Rail Live exhibition gress; Ref. [3] notes that Great Britain’s rail emissions have where HydroFLEX was demonstrated, numerous modifica- been dropping in recent years, down to 36.6 g CO eper tions were performed to increase the reliability and power of passenger km in 2018–2019. This compares favourably with the HydroFLEX train, primarily concerned with modifying new cars, which, according to [4], produce an average of the base train logic, with the aim of a successful mainline test. 122.1 g of CO per km. Given an average car occupancy Supporting this effort was a multitude of documentation rate on 1.6 passengers per car in 2018 [5], 76.3 g of CO is concerning safety, operations, and approvals to gain emitted per car passenger km carbon emissions. However, approvals from the relevant approvals bodies. The project despite these data suggesting that trains produce half the demonstrated the feasibility of using hydrogen fuel cells as an amount of carbon emissions of cars further improvement are autonomous fuel for railway propulsion systems, which has required. A key driver of this is legislation and standards. the potential for full decarbonisation. This is the situation in the UK, a commitment to reduce greenhouse gas to net zero by 2050 was signed in 2019 [6]. Keywords Hydrogen  Fuel cell  Hybrid  Systems Most of the UK’s railway network relies on diesel- engineering powered rolling stock, which has the substantial problem of producing emissions at the point of use, including nitrous oxides and particulates, even exceeding EU limits in some & Charles Calvert enclosed stations such as Birmingham New Street [7]. One c.calvert.1@bham.ac.uk alternative is to use electrified railways; however, these can prove expensive, with a typical figure in 2010 prices being Birmingham Centre for Railway Research and Education, School of Engineering, University of Birmingham, £1 million per km per single line for mainline electrifica- Edgbaston, Birmingham B15 2TT, UK tion [8]. Jeff Vehicles Ltd, Beavercreek, USA 123 Concept development and testing of the UK’s first hydrogen-hybrid train (HydroFLEX) 249 Therefore, hydrogen propulsion has been presented as operation’’. As such, a systems engineering lead approach an alternative. Ref. [9] notes that the first hydrogen-pow- was required to bridge the traditional engineering disci- ered train appeared in 2002, a small mining locomotive plines involved in retrofitting an existing train with a brand powered by a 17-kW fuel cell. A more recent development new hydrogen-hybrid powertrain and to operate it on pas- has been the trial fleet of hydrogen-powered regional trains senger carrying demonstration services. in operation in Germany [10]. In Great Britain, however, Systems engineering follows stages of development there is a proliferation of restrictive loading gauges [11], from conception and concept development through engi- and there have been no previous mainline hydrogen-pow- neering development to a post-development stage. These ered train operations. However, the possibility of using stages are shown in Fig. 1, and this process was adopted in hydrogen has been considered for some time [12]. the development of HydroFLEX. To pioneer hydrogen trains in Britain, Porterbrook and the University of Birmingham announced the HydroFLEX train in October 2018 [13]. At the time of starting the 3 Concept development project press releases from the manufacturers of the iLint, Europe’s first hydrogen train stated that the train has a daily Concept development began in October 2018 and finished range of 1000 km [10]. This was taken to be an indication in December that year. to the project partners that hydrogen fuel cell technology The operational deficiency that the HydroFLEX concept had reached the level of maturity required to be imple- design aimed to address was a lack of zero-harmful-emis- mented on passenger trains and that an attempt should be sion options for powering self-propelled passenger trains in made to do so in the UK. the UK. This would be addressed by producing a demon- The objective of this piece of research was to ascertain strator hydrogen-hybrid passenger train using a surplus the suitability of hydrogen-hybrid trains on the British Class 319 Electric Multiple Unit (EMU). railway network through concept design, performance At workshops held at the beginning of the concept calculations, construction, and testing of a full-scale train. development phase, the following metrics were determined The approach taken was to retrofit an existing electric train from single train simulation of the intended route as the with the new power system, instead of designing and criteria for the functional specification, for an initial constructing an entirely new vehicle. demonstration passenger service at Rail Live 2019 exhi- bition, albeit at reduced speed (see Table 1). Extrapolating the data in Table 1, there is a minimum 2 Project systems engineering requirement for the hydrogen-hybrid system to provide 50.4 kWh of usable energy to meet an aim of running the The HydroFLEX project combined many different engi- train four times per day in passenger service. neering disciplines to achieve a singular objective, to make To achieve this, at a high level, our hydrogen-hybrid an electric train self-powered by adding a hydrogen-hybrid system concept consisted of the following elements: power system. • Hydrogen storage in high-pressure gas cylinders Electrical engineering was applied to ensure that enough • A hydrogen fuel cell stack to convert hydrogen and current was supplied at the correct voltage to the traction oxygen into electricity motors and auxiliary equipment to power the train. • A battery to store the output of the fuel cell in periods Mechanical engineering was required to ensure that all the of low demand and release energy in periods of high new equipment could be secured to the train in a safe demand manner, in accordance with the relevant load cases, and • A 24 V control system that the existing train could accept the additional load. Chemical engineering was required in the original design At the inception of the project, it was not clear what of the fuel cell and to determine how much hydrogen was approach to take with hydrogen storage. Removable and required to achieve the duty cycle. fixed options were available. However, after investigation In addition to the engineering disciplines involved, other it was decided that fixed cylinders on-board the train was factors such as logistics, management, and organisation the correct approach to take rather than utilising removable were involved in the project. If each of these concerns was hydrogen storage elements due to the determined lack of feasibility for inserting and removing removable storage addressed individually, the result would be a haphazard project that would not be able to meet the aim. Instead, a elements. The need for a bespoke 24 V control system was iden- different approach is needed. Kossiakoff [13] describes systems engineering as being tified early in the concept design phase as much equipment ‘‘focused on the system as a whole; it emphasizes its total requires this as a control voltage and energy to start the Rail. Eng. Science (2021) 29(3):248–257 250 C. Calvert et al. Fig. 1 Stages of system engineering, redrawn after [13] Table 1 Criteria for successful operation at Rail Live 2019 of the equipment involved for demonstrations. One of the first concept renders is shown in Fig. 2. Quantity Value Having decided to mount the power system within one Return run length 800 m of the carriages, the next task was to decide which carriage Return trips per day 4 to fit the new equipment in. As such, it was important to Traction energy for 2 9 400 m *6 kWh understand the base configuration of the donor train. Time for system on return journey 600 s The donated Class 319 train is made up of four car- Aux power 40 kW riages. In order these are titled as: driving trailer (A), Aux energy for 1 return journey 6.6 kWh pantograph motor open second (PMOS), trailer open sec- Total energy required for return journey (Traction ? Aux)12.6 kWh ond (lavatory), driving trailer (B) (see Fig. 3). Each of the driving trailer carriages feature a driving cab on the outer end. Class 319s that previously operated on the Thameslink network feature collector shoes on the outer bogies of the driving trailers to collect 750 V DC elec- various equipment could be stored in readily available and tricity, although these have been removed from some units inexpensive deep cycle leisure batteries. in service and were removed from the HydroFLEX donor It was decided at an early stage to abandon the idea of unit before the project was commenced. mounting equipment underneath the floor (as is common The PMOS carriage features a pantograph on the roof practice for British trains) [14]. This would require the for collecting 25 kV AC electricity from overhead contact design, manufacture, and installation of a bespoke under- wires and each of the wheelsets of the carriage is powered frame raft, which would require a greater budget and more by a traction motor. All other wheelsets on the train are time than was available. Further, storing hydrogen fuel in large quantities underneath the floor presents safety issues due to the buoyancy and flammability of the hydrogen gas. It is also clear that additional measures would have to be taken to comply with EC79 [15], if hydrogen were stored beneath the carriage floor. An alternative approach was selected, namely position- ing the equipment inside one of the carriages of the Class 319 EMU, utilising the seat rails to mount the equipment securely to the floor. Along with the overall benefit of the large volume that is available, fitting the equipment within a carriage gives protection from the elements, from dust and water ingress and from impacts that could be expected Fig. 2 Early concept render of HydroFLEX hydrogen-hybrid pow- if the equipment was located outside the vehicle. This ertrain located within the PMOS carriage. Note that the specific configuration also allowed for easy access to the equipment equipment is not modelled yet as configurations were being trialled at for maintenance and upgrades, an important feature for a this stage. The equipment from the nearest the viewpoint outwards is test train. Another advantage of this approach was visibility the battery, DC/DC converter, Ballard fuel cell and representative type 4 hydrogen tank Rail. Eng. Science (2021) 29(3):248–257 Concept development and testing of the UK’s first hydrogen-hybrid train (HydroFLEX) 251 unpowered. All the space between the bogies is occupied around the cabinets to comply with rail vehicle load cases with traction equipment. specified in regulations such as GM/RT 2100. The trailer open second (lavatory) carriage has auxiliary The overall system for supplying power to the train is equipment including a motor alternator set and a com- shown in Fig. 4. In this case, the DC-to-DC converter was pressor mounted between the bogies beneath the floor. All connected across the batteries, providing a parallel con- the space between the bogies is occupied by this nection to the traction system. equipment. Following some investigation into the practicability of From the four available carriages, it was decided that the using removable type 1 tanks, it was determined that this PMOS carriage was the most suitable for modification. The was not a viable option. As such, a search was made for a primary factor in this decision was that the traction workable alternative. equipment is located beneath the floor of the PMOS car- The search for hydrogen storage led to discussions with riage. This would allow for the new powertrain to be representatives of Swagelok, a supplier of EC79 rated electrically connected directly. If the powertrain was in any hydrogen piping and valves, revealed that a suitable sup- other carriage, it would necessitate the addition of a trac- plier of type 3 hydrogen storage tanks, Luxfer, operated in tion current return cable to the power equipment that would the UK 4 9 Luxfer W205N tanks were procured, with a have to span the length of the carriages and be connected total storage capacity of 20 kg of hydrogen. These were between carriages via a jumper connection. suitable for permanent attachment to the floor of the car- riage. These tanks are commonly used in hydrogen-pow- ered buses, and the project team benefited from this prior 4 Engineering development development to support this use case. While the total hydrogen capacity was substantially greater than what the 4.1 Hydrogen-hybrid power system concept required for a single journey, it was quickly rea- lised that additional duties during testing would be Much of the engineering development was constrained by required. These additional duties consisted of static testing, the limited time available. From the announcement of the idling time, and multiple-day operation between refuelling project in October 2018, to Rail Live 2019, when the train opportunities. The provision of greater range than initially was expected to be operated, there were just 9 months. In required also allowed for long distance testing following this time, equipment had to be procured, installed, tested, the initial low-speed testing around the Long Marston site. and finally operated. Therefore, the project team selected Static testing occurred both prior to the introduction to equipment which was readily available. service of the train and on a regular basis during the A Ballard FCVeloCity HD100, 100-kW fuel cell stack lifespan of the train, such as every morning before going was selected and procured along with its coolant and air into demonstration and testing service and after a pro- supply subsystems. This procurement also included a longed period of inactivity to confirm continued function of Danfoss DC-to-DC converter which could provide constant the system. DC power to the train from both the fuel cell stack and Idling time occurred between specific demonstration and battery. The fuel cell stack module was modified slightly in testing duties that occurred on the same day. Whilst the order to connect with this DC-to-DC converter. power produced by the hydrogen fuel cell during these It did not prove possible to order a single traction battery idling periods would not necessarily be wasted as it could to meet our requirements; instead, two 42-kWh Denchi be stored as energy in the batteries or used to keep auxiliary batteries was procured. Each battery consisted of 12 battery loads such as the motor alternator and the compressor slices with an additional control slice, located in a cabinet running. with equivalent dimensions to a 21 server rack. While this Multiple-day operation between refuelling opportunities did fit within the footprint of the vehicle, the centre of mass did happen on several occasion. This happened due to both of this assembly was relatively high, and the base relatively operational planning and as an unplanned reaction to other narrow; it was therefore decided to install an external frame trains being stabled on the siding where the refueller was located. By engaging with Luxfer it was confirmed that it would be possible for the tanks to be supplied within the tight deadlines of the project and that an assembly of four tanks could be supplied in a frame that met the GM/RT 2100 standards [16] for equipment to be mounted to rail vehi- cles. The solution would also include the refuelling and defueling nozzles, pressure gauges, high-pressure Fig. 3 Class 319 EMU train formation Rail. Eng. Science (2021) 29(3):248–257 252 C. Calvert et al. Precharge contactor DC Main contactor 1 Main contactor 2 DC Danfoss Ballard DC to DC fuel cell Denchi converter battery cabinets Fig. 4 Block diagram of the overall power system of HydroFLEX pipework, pressure reducer and systems integration, and The fuel cell stack was controlled by software running engineering support. on a laptop supplied by the University of Birmingham. Once the traction batteries were switched on, a switch 4.2 Control equipment could be operated in this software to start the other com- ponents. The software closed contactors between the DC- In order to start the hydrogen-hybrid power system com- to-DC converter, traction batteries, and fuel cell stack and ponents and operate them correctly when traction or aux- then started the fuel cell, enabled, and controlled the DC- iliary power is demanded, control equipment was to-DC converter to achieve the required power. The fuel necessary. The control system is also required to shut down cell software running on the University of Birmingham these components when required (including in emergen- laptop communicated with the control modules in the cies) and provide voltage, current, and temperature pro- traction batteries via CAN bus messages sent by the latter. In addition, the fuel cell had its own control system for its tection for the traction batteries. The control modules of traction batteries required a air, fuel, and coolant subsystems. 48 V supply. This was supplied by two DC-to-DC con- verters using the 110 V DC train supply. The two con- 4.3 Installation verters produce 24 V each and are connected in series to produce the required voltage. Each battery was started The company dg8 was responsible for designing the using software running on a laptop supplied by Denchi. arrangements for mounting the new equipment to the train, This software also showed the operator the individual cell and for ensuring these arrangements were compliant with voltages and any alarms (including overvoltage, under- rail vehicle legislation. They designed the rafts for the fuel voltage, overcurrent, over-temperature, and under-temper- cell and its various subsystems (including DC-to-DC con- ature). These alarms automatically trigger a shutdown of verter), the external battery frames, and the 24 V control the relevant battery if required. When the fuel cell stack system box reinforcement. was connected, various traction battery outputs and charge The final design layout of equipment is shown as a CAD paths were enabled to allow the traction battery voltages to image in Fig. 5. This is largely representative of the layout balance. that was implemented. The only detail difference between The laptop control was supported by an auxiliary control this CAD image and the true layout is the slightly smaller battery box containing 6 9 12 V batteries, arranged as size of the true hydrogen storage. three series pairs of batteries connected in parallel pro- This equipment was installed between March and May ducing the required 24 V supply. The batteries were 2019, primarily using the door openings. These openings charged from the three-phase train supply and could supply proved wide enough for all rafts to be moved in and the 24 V System for over an hour, which is much longer installed without difficulty. than is necessary to complete the start-up process and thus establish the three-phase train supply. Rail. Eng. Science (2021) 29(3):248–257 Concept development and testing of the UK’s first hydrogen-hybrid train (HydroFLEX) 253 4.4 Rail Live power supplies. As such, the harsh changes in demanded load created by the HydroFLEX train were routinely trip- Following functional testing, the engineering development ping the protections at their original setpoints. achieved its objective of supplying a working train for use Further, the train control was modified, with new train at Rail Live 2019, where numerous low-speed demonstra- control logic equipment being implemented (see Fig. 6). tion runs were performed with the public on-board. All This allowed engineers in the PMOS to control the train were successful. logic, rather than it being automatically controlled from the Class 319 power controller. When notch 2 was selected on the controller, for example, this would illuminate a button 5 Post-development on the logic equipment, allowing engineers to prevent the train from drawing more power if the fuel cell stack and Although the HydroFLEX train had been intended to debut traction batteries were not ready. This equipment also at Rail Live 2019 as a demonstrator, it was felt that suffi- cient potential existed in the design to carry out additional testing, including on the UK mainline railway. This was part-funded by Innovate UK FOAK 3 funding, from the UK’s Department for Transport. 5.1 Power The first issue to be solved beyond Rail Live was that of power. In operation, the traction batteries proved more vulnerable to peaks in load than had been anticipated. This was due to the traction load fluctuating more rapidly than the design load of the battery, which was intended for steady charging and discharging rates when used as an uninterrupted power supply system. While it was in theory possible to operate the train in notch 2 (the Class 319’s power controller has four power notches, notch 4 being the most powerful and notch 1 being the least) this state would often cause one or both of the traction batteries’ protective systems to cut out the battery for battery protection. To attempt to mitigate the vulnerability of the batteries to the power transients demanded by the Class 319 traction system the slice modules were returned to the manufacturer for modifications to the control hardware and software. The modifications included changes that allowed for charging and discharging at a higher rate than before the modifica- tion. The conservative thresholds that the battery protection system was originally set at were originally imposed due to Fig. 6 Train control logic equipment, showing the exterior (above), the control software being intended for use in uninterrupted and interior wiring (below) Fig. 5 CAD image of the design of the inside of the PMOS carriage. From the left to right the visible components are the 24 V control box, hydrogen tanks, fuel cell stack and DC-DC raft, and the protective structural frame for the traction batteries Rail. Eng. Science (2021) 29(3):248–257 254 C. Calvert et al. AsBo/Safety compliances Ops. plans Test plan Safety plan System definition HAZID & HAZOP Hazard record Vehicle maintenance Evidence instructions (VMI) Assessment plan Assessment record Assessment record Safety assessment report Technical compliance Build compliance Verification plan Compliance matrix Attestation certification (mods. only) Assessment record Route compatibility NR submission NRSC RIS-8270-RST Compliance, arguments LoNo ROG Ricardo SoC pack RIS standards list & evidence (ROG) network rail ORR authorisation Initial meeting Decision on TSI statement Review Final meeting Fig. 7 Overall schematic of the approvals process, showing the four categories of the process allowed engineers to monitor other traction system activity control of passengers, etc. This concerns public safety such as the Wheel Slide Protection (WSP) system activity. while using the HydroFLEX train. • System definition. This document defines the system An additional mode of operation, beyond what had originally been envisaged, was also implemented. In nor- such that safety issues with our system could be identified and mitigated. mal operation, the Class 319 EMU engages a field divert at notch 4. This is because as the speed of the train rises, the • HAZID & HAZOP. This consists of two phases: back-EMF produced by the traction motors also increases, – HAZID or Hazard Identification. This is defined as leading to a limitation in power. To allow for higher speed ‘‘A systematic review of the possible causes and operation on the UK mainline railway with the limited consequences of hazardous events’’ and is largely power available from the fuel cell and traction batteries, the concerned with the effects of hazardous events [18] train control was further modified, with the field divert – HAZOP or Hazard & Operability. HAZOP is a being activated when notch 3 was selected. systems level review, to identify any further hazards that may occur either when the system is operating 5.2 Mainline documentation, test and approvals as intended or any likely deviations from expected operations [19]. It was necessary to seek formal authorisation to take the train onto the mainline via an approvals process. • Hazard record. This is a centralised repository for all To standardise procedures, an engineering operations hazards identified, which was used as a reference for manual was collated. This was just one of many documents the operations plant, test plans, and the vehicle necessary for approvals purposes. An overview of the maintenance instructions (VMI). approvals process is shown in Fig. 7. The technical compliance consisted of verifying that the The following are the key elements of the assessment new equipment (noted as ‘‘mods.’’ or modifications in body (AsBo)/safety compliances. Fig. 7) was compliant with the relevant technical standards • Safety plan. A safety plan was produced by Porterbrook (including GM/RT 2100)[17] to be permitted to run on the covering evacuation, procedures, dispatch procedures, mainline railway. Rail. Eng. Science (2021) 29(3):248–257 Concept development and testing of the UK’s first hydrogen-hybrid train (HydroFLEX) 255 Fig. 8 GPS data for a mainline test run to Evesham. The colour coding of the GPS track shows the speed, and the plot on the right shows speed readings; although one reading indicates 91.5 km/h, this is spurious and was not corroborated by any other sensor. Altitude data are shown in the lower plot against time from system start-up in hours and minutes. The thick red lines added to the elevation plot bracket the time when the train was in operation Route compatibility is essential; while it may be theo- 5.3 Mainline running retically possible for the train to operate on the rail network in Great Britain, in practice it is necessary to ensure there Following the approvals and testing process, HydroFLEX are no barriers to it operating on a particular route. This was approved for mainline running. Three mainline tests were conducted in September 2020, between Long Mar- began with a search of the rail industry standards (RIS), followed by a Letter of No Objection (LoNo) from Rail ston, the site at which the train had been tested and oper- ated, to Evesham. Operations Group (ROG), who are licensed to operate trains on the mainland UK network. Ricardo, as the AsBo, All test runs were successful with only minor issues encountered. The GPS data from one of these runs are provided a Statement of Compliance (SoC), and following the submission to Network Rail, a Network Rail Safety shown in Fig. 8, with performance and electrical data shown in Fig. 9. During mainline testing, performance data Certificate (NRSC) was awarded. The Office of Rail & Road (ORR) took some interest in were also harvested from the traction system, allowing for the project, and a review was carried out; however, it was tuning of the system for later runs. These data will also not considered to be necessary to issue a Technical Stan- inform the design of subsequent types of hydrogen-hybrid dards for Interoperability (TSI) statement. train. Figure 9 shows the main system data plotted as a Numerous tests had to be carried out to support the function of time for the test runs. The whole test cycle approvals effort, including takes a considerable amount of time because the train needed to be prepared and ready, and to meet certain • Electro-magnetic compatibility (EMC) testing. This timing points. This means that the duty cycle comprises of was carried July 2020, using both lineside and on-board extended periods of idling where the train is held before equipment to monitor electromagnetic emissions from given the authority to advance to the next section. the train. The green line in Fig. 9 shows the train speed. The • Annual exam of the equipment on board the train. initial movements are in the test site and are characterised • Fitness to run (FTR) exams. These were carried out by several short and slow movements. The train then enters every day before a mainline test, with sign off by a the mainline and travels on a low-speed branch to join the competent inspector required before operations live railway. Once given permission to move onto the high- speed section, the train then was permitted to travel at its highest speed to Evesham station. After a stop, the train moved beyond the station and returned on the adjacent track for the return journey. Rail. Eng. Science (2021) 29(3):248–257 256 C. Calvert et al. Fig. 9 Performance data plot for the same run from Long Marston to Evesham, showing electrical data for the base Class 319 (top), speed data (middle) and data from the current monitoring device (bottom) A few other data values are included on the plot. These from the HydroFLEX project will continue to inform the include the system DC bus voltage, the traction motor design of follow-on classes of hydrogen train. current, the DC link bus current, the available power, and Acknowledgements This project required the support of a large the CMD gain which moderates the power demanded by number of stakeholders. It is impossible to give thanks and list all the train to within the capability of the propulsion system. contributors individually, but the authors would like to thank the The propulsion system needed to be dynamic enough to following organisations: Porterbrook Leasing Company Ltd., Dr Jeff cope with rapid demand changes from the driver, and Allan of Jeff Vehicles Ltd., Chrysalis Rail Services Ltd., Ballard Power Systems Inc., Denchi Power Ltd., Swagelok, Rail Alliance, considerable effort has gone into the design to cope with Rail Operations Group, Luxfer, dg8 Design & Engineering Ltd. The changes in notch power settings at different speeds. authors would like to acknowledge the generous support of Porter- brook Leasing Company Limited and the funding supplied by Inno- vate UK. Without this support, the HydroFLEX project would not have been possible. 6 Conclusion Open Access This article is licensed under a Creative Commons The HydroFLEX team showed the possibility of operating Attribution 4.0 International License, which permits use, sharing, hydrogen-hybrid trains were on the British railway net- adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the work. This was achieved by developing a concept design, source, provide a link to the Creative Commons licence, and indicate consisting of new equipment mounted inside the PMOS if changes were made. The images or other third party material in this carriage, and progressing this design through detailed article are included in the article’s Creative Commons licence, unless design and installation to testing. The resulting Hydro- indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended FLEX train was successfully demonstrated at Rail Live. use is not permitted by statutory regulation or exceeds the permitted Beyond this its potential was developed through modi- use, you will need to obtain permission directly from the copyright fications to the design, and the train was approved to run on holder. To view a copy of this licence, visit http://creativecommons. the mainline railway, with numerous documents produced org/licenses/by/4.0/. to support this. The learning experience and data gathered Rail. Eng. Science (2021) 29(3):248–257 Concept development and testing of the UK’s first hydrogen-hybrid train (HydroFLEX) 257 10. Alstom (2018). World premiere: Alstom’s hydrogen trains enter References passenger service in Lower Saxony. Available at: https://www. alstom.com/press-releases-news/2018/9/world-premiere-alstoms- 1. United Nations (2015). Paris Agreement. https://unfccc.int/files/ hydrogen-trains-enter-passenger-service-lower. Accessed 10 Nov meetings/paris_nov_2015/application/pdf/paris_agreement_ english_.pdf. Accessed 10 Nov 2020 11. RSSB (2015). Requirements for the application of standard 2. Ye B, Jiang J, Liu J, Zheng Y, Zhou N (2020) Research on vehicle gauges. https://www.rssb.co.uk/standards-catalogue/ quantitative assessment of climate change risk at an urban scale: CatalogueItem/GERT8073-Iss-3. Accessed 10 Nov 2020 Review of recent progress and outlook of future direction. Renew 12. Hillmansen S (2003) The application of fuel cell technology to Sustain Energy Rev 135:110415 rail transport operations. Proc Inst Mech Eng Part F J Rail Rapid 3. ORR (2019). Rail emissions: 2018–19 annual statistical release. Transit 217(4):291–298 https://dataportal.orr.gov.uk/media/1550/rail-emissions-2018-19. 13. Porterbrook (2019). Meet the FLEX family. Available at: https:// pdf. Accessed 10 Nov 2020 www.porterbrook.co.uk/innovation/case-studies/the-flex-family 4. UK Government (2015). New car carbon dioxide emissions. Accessed 10 Nov 2020 https://www.gov.uk/government/publications/new-car-carbon- 14. Kossiakoff A, Sweet WN, Seymour SJ, Biemer SM (2011) Sys- dioxide-emissions#:*:text=The%20average%20carbon% tems engineering principles and practice, 2nd ed. Wiley 20dioxide%20emissions,of%20carbon%20dioxide%20per% 15. Rolling Stock Strategy Steering Group (2018). Long term pas- 20kilometre. Accessed 10 Nov 2020 senger rolling stock strategy for the rail industry, 6th ed. https:// 5. Statista. 2020. Average car and van occupancy England www.raildeliverygroup.com/files/Publications/2018-03_long_ 2002–2018 Statistic | Statista. https://www.statista.com/statistics/ term_passenger_rolling_stock_strategy_6th_ed.pdf. Accessed 20 314719/average-car-and-van-occupancy-in-england/. Accessed Oct 2020 13 June 2021. 16. European Parliament And Council (2009) Regulation (EC) No 6. UK Government, (2019). UK becomes first major economy to 79/2009 of the European Parliament and of the Council of 14 pass net zero emissions law. https://www.gov.uk/government/ January 2009 on type-approval of hydrogen-powered motor news/uk-becomes-first-major-economy-to-pass-net-zero- vehicles, and amending Directive 2007/46/EC. https://eur-lex. emissions-law. Accessed 10 Nov 2020 europa.eu/eli/reg/2009/79/oj. Accessed 04 Feb 2021 7. Hickman AL, Baker CJ, Delgado-Saborit JM, Thornes JE (2018) 17. RSSB (2020). Railway group standard GMRT2100: rail vehicle Evaluation of air quality at the Birmingham New Street Railway structures and passive safety (Issue 6). Station. Proc Inst Mech Eng Part F J Rail Rapid Transit 18. Crawley F. (2020) A guide to hazard identification methods, 2nd 232(6):1864–1878 ed. Elsevier, Oxford 8. Keen P, Phillpotts R (2010) Low cost electrification for branch 19. Crawley F, Tyler B (2015) HAZOP: guide to best practice, 3rd. lines. https://assets.publishing.service.gov.uk/government/ Elsevier, Oxford uploads/system/uploads/attachment_data/file/3872/low-cost- electrification-report.pdf. Accessed 10 Nov 2020 9. Hoffrichter H (2013) Hydrogen as an energy carrier for railway traction. Dissertation, University of Birmingham, Birmingham. Rail. Eng. Science (2021) 29(3):248–257 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Railway Engineering Science Springer Journals

Concept development and testing of the UK’s first hydrogen-hybrid train (HydroFLEX)

Download PDF
10 pages

Loading next page...
 
/lp/springer-journals/concept-development-and-testing-of-the-uk-s-first-hydrogen-hybrid-qhYtS0RLz3

References (20)

Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2021
eISSN
2662-4753
DOI
10.1007/s40534-021-00256-9
Publisher site
See Article on Publisher Site

Abstract

Rail. Eng. Science (2021) 29(3):248–257 https://doi.org/10.1007/s40534-021-00256-9 Concept development and testing of the UK’s first hydrogen- hybrid train (HydroFLEX) 1 2 1 1 • • • • Charles Calvert Jeff Allan Peter Amor Stuart Hillmansen 1 1 Clive Roberts Paul Weston Received: 14 February 2021 / Revised: 29 July 2021 / Accepted: 18 August 2021 / Published online: 1 October 2021 The Author(s) 2021 Abstract In October 2018, Porterbrook and the University of 1 Introduction Birmingham announced the HydroFLEX project, to demon- strate a hydrogen-hybrid modified train at Rail Live 2019. The The need for countries to take action on climate change concept of modifying a Class 319 Electric Multiple Unit was was emphasised by the Paris Agreement of 2015 [1], which developed, with equipment including a fuel cell stack, trac- requires all parties to keep global temperature increase tion battery, 24 V control system and hydrogen storage ele- below 2 degrees centigrade, and to pursue an increase ments to be mounted inside one of the carriages. This was within 1.5 degrees centigrade, above pre-industrial levels. followed by procurement of a fuel cell stack, traction batter- Climate change impacts, notably the increased frequency ies, and control equipment, which was then installed inside and severity of extreme weather events, can affect the train, being fixed to the seat rails. One substantial change numerous aspects of human civilisation, and so mitigation from the concept was the provision of considerably more is essential [2]. hydrogen storage than the minimum necessary, providing the In order to achieve this, greenhouse gas emissions must train with more potential to be further modified to allow for be lowered, and the Great Britain’s railway is making pro- higher speed mainline testing. After the Rail Live exhibition gress; Ref. [3] notes that Great Britain’s rail emissions have where HydroFLEX was demonstrated, numerous modifica- been dropping in recent years, down to 36.6 g CO eper tions were performed to increase the reliability and power of passenger km in 2018–2019. This compares favourably with the HydroFLEX train, primarily concerned with modifying new cars, which, according to [4], produce an average of the base train logic, with the aim of a successful mainline test. 122.1 g of CO per km. Given an average car occupancy Supporting this effort was a multitude of documentation rate on 1.6 passengers per car in 2018 [5], 76.3 g of CO is concerning safety, operations, and approvals to gain emitted per car passenger km carbon emissions. However, approvals from the relevant approvals bodies. The project despite these data suggesting that trains produce half the demonstrated the feasibility of using hydrogen fuel cells as an amount of carbon emissions of cars further improvement are autonomous fuel for railway propulsion systems, which has required. A key driver of this is legislation and standards. the potential for full decarbonisation. This is the situation in the UK, a commitment to reduce greenhouse gas to net zero by 2050 was signed in 2019 [6]. Keywords Hydrogen  Fuel cell  Hybrid  Systems Most of the UK’s railway network relies on diesel- engineering powered rolling stock, which has the substantial problem of producing emissions at the point of use, including nitrous oxides and particulates, even exceeding EU limits in some & Charles Calvert enclosed stations such as Birmingham New Street [7]. One c.calvert.1@bham.ac.uk alternative is to use electrified railways; however, these can prove expensive, with a typical figure in 2010 prices being Birmingham Centre for Railway Research and Education, School of Engineering, University of Birmingham, £1 million per km per single line for mainline electrifica- Edgbaston, Birmingham B15 2TT, UK tion [8]. Jeff Vehicles Ltd, Beavercreek, USA 123 Concept development and testing of the UK’s first hydrogen-hybrid train (HydroFLEX) 249 Therefore, hydrogen propulsion has been presented as operation’’. As such, a systems engineering lead approach an alternative. Ref. [9] notes that the first hydrogen-pow- was required to bridge the traditional engineering disci- ered train appeared in 2002, a small mining locomotive plines involved in retrofitting an existing train with a brand powered by a 17-kW fuel cell. A more recent development new hydrogen-hybrid powertrain and to operate it on pas- has been the trial fleet of hydrogen-powered regional trains senger carrying demonstration services. in operation in Germany [10]. In Great Britain, however, Systems engineering follows stages of development there is a proliferation of restrictive loading gauges [11], from conception and concept development through engi- and there have been no previous mainline hydrogen-pow- neering development to a post-development stage. These ered train operations. However, the possibility of using stages are shown in Fig. 1, and this process was adopted in hydrogen has been considered for some time [12]. the development of HydroFLEX. To pioneer hydrogen trains in Britain, Porterbrook and the University of Birmingham announced the HydroFLEX train in October 2018 [13]. At the time of starting the 3 Concept development project press releases from the manufacturers of the iLint, Europe’s first hydrogen train stated that the train has a daily Concept development began in October 2018 and finished range of 1000 km [10]. This was taken to be an indication in December that year. to the project partners that hydrogen fuel cell technology The operational deficiency that the HydroFLEX concept had reached the level of maturity required to be imple- design aimed to address was a lack of zero-harmful-emis- mented on passenger trains and that an attempt should be sion options for powering self-propelled passenger trains in made to do so in the UK. the UK. This would be addressed by producing a demon- The objective of this piece of research was to ascertain strator hydrogen-hybrid passenger train using a surplus the suitability of hydrogen-hybrid trains on the British Class 319 Electric Multiple Unit (EMU). railway network through concept design, performance At workshops held at the beginning of the concept calculations, construction, and testing of a full-scale train. development phase, the following metrics were determined The approach taken was to retrofit an existing electric train from single train simulation of the intended route as the with the new power system, instead of designing and criteria for the functional specification, for an initial constructing an entirely new vehicle. demonstration passenger service at Rail Live 2019 exhi- bition, albeit at reduced speed (see Table 1). Extrapolating the data in Table 1, there is a minimum 2 Project systems engineering requirement for the hydrogen-hybrid system to provide 50.4 kWh of usable energy to meet an aim of running the The HydroFLEX project combined many different engi- train four times per day in passenger service. neering disciplines to achieve a singular objective, to make To achieve this, at a high level, our hydrogen-hybrid an electric train self-powered by adding a hydrogen-hybrid system concept consisted of the following elements: power system. • Hydrogen storage in high-pressure gas cylinders Electrical engineering was applied to ensure that enough • A hydrogen fuel cell stack to convert hydrogen and current was supplied at the correct voltage to the traction oxygen into electricity motors and auxiliary equipment to power the train. • A battery to store the output of the fuel cell in periods Mechanical engineering was required to ensure that all the of low demand and release energy in periods of high new equipment could be secured to the train in a safe demand manner, in accordance with the relevant load cases, and • A 24 V control system that the existing train could accept the additional load. Chemical engineering was required in the original design At the inception of the project, it was not clear what of the fuel cell and to determine how much hydrogen was approach to take with hydrogen storage. Removable and required to achieve the duty cycle. fixed options were available. However, after investigation In addition to the engineering disciplines involved, other it was decided that fixed cylinders on-board the train was factors such as logistics, management, and organisation the correct approach to take rather than utilising removable were involved in the project. If each of these concerns was hydrogen storage elements due to the determined lack of feasibility for inserting and removing removable storage addressed individually, the result would be a haphazard project that would not be able to meet the aim. Instead, a elements. The need for a bespoke 24 V control system was iden- different approach is needed. Kossiakoff [13] describes systems engineering as being tified early in the concept design phase as much equipment ‘‘focused on the system as a whole; it emphasizes its total requires this as a control voltage and energy to start the Rail. Eng. Science (2021) 29(3):248–257 250 C. Calvert et al. Fig. 1 Stages of system engineering, redrawn after [13] Table 1 Criteria for successful operation at Rail Live 2019 of the equipment involved for demonstrations. One of the first concept renders is shown in Fig. 2. Quantity Value Having decided to mount the power system within one Return run length 800 m of the carriages, the next task was to decide which carriage Return trips per day 4 to fit the new equipment in. As such, it was important to Traction energy for 2 9 400 m *6 kWh understand the base configuration of the donor train. Time for system on return journey 600 s The donated Class 319 train is made up of four car- Aux power 40 kW riages. In order these are titled as: driving trailer (A), Aux energy for 1 return journey 6.6 kWh pantograph motor open second (PMOS), trailer open sec- Total energy required for return journey (Traction ? Aux)12.6 kWh ond (lavatory), driving trailer (B) (see Fig. 3). Each of the driving trailer carriages feature a driving cab on the outer end. Class 319s that previously operated on the Thameslink network feature collector shoes on the outer bogies of the driving trailers to collect 750 V DC elec- various equipment could be stored in readily available and tricity, although these have been removed from some units inexpensive deep cycle leisure batteries. in service and were removed from the HydroFLEX donor It was decided at an early stage to abandon the idea of unit before the project was commenced. mounting equipment underneath the floor (as is common The PMOS carriage features a pantograph on the roof practice for British trains) [14]. This would require the for collecting 25 kV AC electricity from overhead contact design, manufacture, and installation of a bespoke under- wires and each of the wheelsets of the carriage is powered frame raft, which would require a greater budget and more by a traction motor. All other wheelsets on the train are time than was available. Further, storing hydrogen fuel in large quantities underneath the floor presents safety issues due to the buoyancy and flammability of the hydrogen gas. It is also clear that additional measures would have to be taken to comply with EC79 [15], if hydrogen were stored beneath the carriage floor. An alternative approach was selected, namely position- ing the equipment inside one of the carriages of the Class 319 EMU, utilising the seat rails to mount the equipment securely to the floor. Along with the overall benefit of the large volume that is available, fitting the equipment within a carriage gives protection from the elements, from dust and water ingress and from impacts that could be expected Fig. 2 Early concept render of HydroFLEX hydrogen-hybrid pow- if the equipment was located outside the vehicle. This ertrain located within the PMOS carriage. Note that the specific configuration also allowed for easy access to the equipment equipment is not modelled yet as configurations were being trialled at for maintenance and upgrades, an important feature for a this stage. The equipment from the nearest the viewpoint outwards is test train. Another advantage of this approach was visibility the battery, DC/DC converter, Ballard fuel cell and representative type 4 hydrogen tank Rail. Eng. Science (2021) 29(3):248–257 Concept development and testing of the UK’s first hydrogen-hybrid train (HydroFLEX) 251 unpowered. All the space between the bogies is occupied around the cabinets to comply with rail vehicle load cases with traction equipment. specified in regulations such as GM/RT 2100. The trailer open second (lavatory) carriage has auxiliary The overall system for supplying power to the train is equipment including a motor alternator set and a com- shown in Fig. 4. In this case, the DC-to-DC converter was pressor mounted between the bogies beneath the floor. All connected across the batteries, providing a parallel con- the space between the bogies is occupied by this nection to the traction system. equipment. Following some investigation into the practicability of From the four available carriages, it was decided that the using removable type 1 tanks, it was determined that this PMOS carriage was the most suitable for modification. The was not a viable option. As such, a search was made for a primary factor in this decision was that the traction workable alternative. equipment is located beneath the floor of the PMOS car- The search for hydrogen storage led to discussions with riage. This would allow for the new powertrain to be representatives of Swagelok, a supplier of EC79 rated electrically connected directly. If the powertrain was in any hydrogen piping and valves, revealed that a suitable sup- other carriage, it would necessitate the addition of a trac- plier of type 3 hydrogen storage tanks, Luxfer, operated in tion current return cable to the power equipment that would the UK 4 9 Luxfer W205N tanks were procured, with a have to span the length of the carriages and be connected total storage capacity of 20 kg of hydrogen. These were between carriages via a jumper connection. suitable for permanent attachment to the floor of the car- riage. These tanks are commonly used in hydrogen-pow- ered buses, and the project team benefited from this prior 4 Engineering development development to support this use case. While the total hydrogen capacity was substantially greater than what the 4.1 Hydrogen-hybrid power system concept required for a single journey, it was quickly rea- lised that additional duties during testing would be Much of the engineering development was constrained by required. These additional duties consisted of static testing, the limited time available. From the announcement of the idling time, and multiple-day operation between refuelling project in October 2018, to Rail Live 2019, when the train opportunities. The provision of greater range than initially was expected to be operated, there were just 9 months. In required also allowed for long distance testing following this time, equipment had to be procured, installed, tested, the initial low-speed testing around the Long Marston site. and finally operated. Therefore, the project team selected Static testing occurred both prior to the introduction to equipment which was readily available. service of the train and on a regular basis during the A Ballard FCVeloCity HD100, 100-kW fuel cell stack lifespan of the train, such as every morning before going was selected and procured along with its coolant and air into demonstration and testing service and after a pro- supply subsystems. This procurement also included a longed period of inactivity to confirm continued function of Danfoss DC-to-DC converter which could provide constant the system. DC power to the train from both the fuel cell stack and Idling time occurred between specific demonstration and battery. The fuel cell stack module was modified slightly in testing duties that occurred on the same day. Whilst the order to connect with this DC-to-DC converter. power produced by the hydrogen fuel cell during these It did not prove possible to order a single traction battery idling periods would not necessarily be wasted as it could to meet our requirements; instead, two 42-kWh Denchi be stored as energy in the batteries or used to keep auxiliary batteries was procured. Each battery consisted of 12 battery loads such as the motor alternator and the compressor slices with an additional control slice, located in a cabinet running. with equivalent dimensions to a 21 server rack. While this Multiple-day operation between refuelling opportunities did fit within the footprint of the vehicle, the centre of mass did happen on several occasion. This happened due to both of this assembly was relatively high, and the base relatively operational planning and as an unplanned reaction to other narrow; it was therefore decided to install an external frame trains being stabled on the siding where the refueller was located. By engaging with Luxfer it was confirmed that it would be possible for the tanks to be supplied within the tight deadlines of the project and that an assembly of four tanks could be supplied in a frame that met the GM/RT 2100 standards [16] for equipment to be mounted to rail vehi- cles. The solution would also include the refuelling and defueling nozzles, pressure gauges, high-pressure Fig. 3 Class 319 EMU train formation Rail. Eng. Science (2021) 29(3):248–257 252 C. Calvert et al. Precharge contactor DC Main contactor 1 Main contactor 2 DC Danfoss Ballard DC to DC fuel cell Denchi converter battery cabinets Fig. 4 Block diagram of the overall power system of HydroFLEX pipework, pressure reducer and systems integration, and The fuel cell stack was controlled by software running engineering support. on a laptop supplied by the University of Birmingham. Once the traction batteries were switched on, a switch 4.2 Control equipment could be operated in this software to start the other com- ponents. The software closed contactors between the DC- In order to start the hydrogen-hybrid power system com- to-DC converter, traction batteries, and fuel cell stack and ponents and operate them correctly when traction or aux- then started the fuel cell, enabled, and controlled the DC- iliary power is demanded, control equipment was to-DC converter to achieve the required power. The fuel necessary. The control system is also required to shut down cell software running on the University of Birmingham these components when required (including in emergen- laptop communicated with the control modules in the cies) and provide voltage, current, and temperature pro- traction batteries via CAN bus messages sent by the latter. In addition, the fuel cell had its own control system for its tection for the traction batteries. The control modules of traction batteries required a air, fuel, and coolant subsystems. 48 V supply. This was supplied by two DC-to-DC con- verters using the 110 V DC train supply. The two con- 4.3 Installation verters produce 24 V each and are connected in series to produce the required voltage. Each battery was started The company dg8 was responsible for designing the using software running on a laptop supplied by Denchi. arrangements for mounting the new equipment to the train, This software also showed the operator the individual cell and for ensuring these arrangements were compliant with voltages and any alarms (including overvoltage, under- rail vehicle legislation. They designed the rafts for the fuel voltage, overcurrent, over-temperature, and under-temper- cell and its various subsystems (including DC-to-DC con- ature). These alarms automatically trigger a shutdown of verter), the external battery frames, and the 24 V control the relevant battery if required. When the fuel cell stack system box reinforcement. was connected, various traction battery outputs and charge The final design layout of equipment is shown as a CAD paths were enabled to allow the traction battery voltages to image in Fig. 5. This is largely representative of the layout balance. that was implemented. The only detail difference between The laptop control was supported by an auxiliary control this CAD image and the true layout is the slightly smaller battery box containing 6 9 12 V batteries, arranged as size of the true hydrogen storage. three series pairs of batteries connected in parallel pro- This equipment was installed between March and May ducing the required 24 V supply. The batteries were 2019, primarily using the door openings. These openings charged from the three-phase train supply and could supply proved wide enough for all rafts to be moved in and the 24 V System for over an hour, which is much longer installed without difficulty. than is necessary to complete the start-up process and thus establish the three-phase train supply. Rail. Eng. Science (2021) 29(3):248–257 Concept development and testing of the UK’s first hydrogen-hybrid train (HydroFLEX) 253 4.4 Rail Live power supplies. As such, the harsh changes in demanded load created by the HydroFLEX train were routinely trip- Following functional testing, the engineering development ping the protections at their original setpoints. achieved its objective of supplying a working train for use Further, the train control was modified, with new train at Rail Live 2019, where numerous low-speed demonstra- control logic equipment being implemented (see Fig. 6). tion runs were performed with the public on-board. All This allowed engineers in the PMOS to control the train were successful. logic, rather than it being automatically controlled from the Class 319 power controller. When notch 2 was selected on the controller, for example, this would illuminate a button 5 Post-development on the logic equipment, allowing engineers to prevent the train from drawing more power if the fuel cell stack and Although the HydroFLEX train had been intended to debut traction batteries were not ready. This equipment also at Rail Live 2019 as a demonstrator, it was felt that suffi- cient potential existed in the design to carry out additional testing, including on the UK mainline railway. This was part-funded by Innovate UK FOAK 3 funding, from the UK’s Department for Transport. 5.1 Power The first issue to be solved beyond Rail Live was that of power. In operation, the traction batteries proved more vulnerable to peaks in load than had been anticipated. This was due to the traction load fluctuating more rapidly than the design load of the battery, which was intended for steady charging and discharging rates when used as an uninterrupted power supply system. While it was in theory possible to operate the train in notch 2 (the Class 319’s power controller has four power notches, notch 4 being the most powerful and notch 1 being the least) this state would often cause one or both of the traction batteries’ protective systems to cut out the battery for battery protection. To attempt to mitigate the vulnerability of the batteries to the power transients demanded by the Class 319 traction system the slice modules were returned to the manufacturer for modifications to the control hardware and software. The modifications included changes that allowed for charging and discharging at a higher rate than before the modifica- tion. The conservative thresholds that the battery protection system was originally set at were originally imposed due to Fig. 6 Train control logic equipment, showing the exterior (above), the control software being intended for use in uninterrupted and interior wiring (below) Fig. 5 CAD image of the design of the inside of the PMOS carriage. From the left to right the visible components are the 24 V control box, hydrogen tanks, fuel cell stack and DC-DC raft, and the protective structural frame for the traction batteries Rail. Eng. Science (2021) 29(3):248–257 254 C. Calvert et al. AsBo/Safety compliances Ops. plans Test plan Safety plan System definition HAZID & HAZOP Hazard record Vehicle maintenance Evidence instructions (VMI) Assessment plan Assessment record Assessment record Safety assessment report Technical compliance Build compliance Verification plan Compliance matrix Attestation certification (mods. only) Assessment record Route compatibility NR submission NRSC RIS-8270-RST Compliance, arguments LoNo ROG Ricardo SoC pack RIS standards list & evidence (ROG) network rail ORR authorisation Initial meeting Decision on TSI statement Review Final meeting Fig. 7 Overall schematic of the approvals process, showing the four categories of the process allowed engineers to monitor other traction system activity control of passengers, etc. This concerns public safety such as the Wheel Slide Protection (WSP) system activity. while using the HydroFLEX train. • System definition. This document defines the system An additional mode of operation, beyond what had originally been envisaged, was also implemented. In nor- such that safety issues with our system could be identified and mitigated. mal operation, the Class 319 EMU engages a field divert at notch 4. This is because as the speed of the train rises, the • HAZID & HAZOP. This consists of two phases: back-EMF produced by the traction motors also increases, – HAZID or Hazard Identification. This is defined as leading to a limitation in power. To allow for higher speed ‘‘A systematic review of the possible causes and operation on the UK mainline railway with the limited consequences of hazardous events’’ and is largely power available from the fuel cell and traction batteries, the concerned with the effects of hazardous events [18] train control was further modified, with the field divert – HAZOP or Hazard & Operability. HAZOP is a being activated when notch 3 was selected. systems level review, to identify any further hazards that may occur either when the system is operating 5.2 Mainline documentation, test and approvals as intended or any likely deviations from expected operations [19]. It was necessary to seek formal authorisation to take the train onto the mainline via an approvals process. • Hazard record. This is a centralised repository for all To standardise procedures, an engineering operations hazards identified, which was used as a reference for manual was collated. This was just one of many documents the operations plant, test plans, and the vehicle necessary for approvals purposes. An overview of the maintenance instructions (VMI). approvals process is shown in Fig. 7. The technical compliance consisted of verifying that the The following are the key elements of the assessment new equipment (noted as ‘‘mods.’’ or modifications in body (AsBo)/safety compliances. Fig. 7) was compliant with the relevant technical standards • Safety plan. A safety plan was produced by Porterbrook (including GM/RT 2100)[17] to be permitted to run on the covering evacuation, procedures, dispatch procedures, mainline railway. Rail. Eng. Science (2021) 29(3):248–257 Concept development and testing of the UK’s first hydrogen-hybrid train (HydroFLEX) 255 Fig. 8 GPS data for a mainline test run to Evesham. The colour coding of the GPS track shows the speed, and the plot on the right shows speed readings; although one reading indicates 91.5 km/h, this is spurious and was not corroborated by any other sensor. Altitude data are shown in the lower plot against time from system start-up in hours and minutes. The thick red lines added to the elevation plot bracket the time when the train was in operation Route compatibility is essential; while it may be theo- 5.3 Mainline running retically possible for the train to operate on the rail network in Great Britain, in practice it is necessary to ensure there Following the approvals and testing process, HydroFLEX are no barriers to it operating on a particular route. This was approved for mainline running. Three mainline tests were conducted in September 2020, between Long Mar- began with a search of the rail industry standards (RIS), followed by a Letter of No Objection (LoNo) from Rail ston, the site at which the train had been tested and oper- ated, to Evesham. Operations Group (ROG), who are licensed to operate trains on the mainland UK network. Ricardo, as the AsBo, All test runs were successful with only minor issues encountered. The GPS data from one of these runs are provided a Statement of Compliance (SoC), and following the submission to Network Rail, a Network Rail Safety shown in Fig. 8, with performance and electrical data shown in Fig. 9. During mainline testing, performance data Certificate (NRSC) was awarded. The Office of Rail & Road (ORR) took some interest in were also harvested from the traction system, allowing for the project, and a review was carried out; however, it was tuning of the system for later runs. These data will also not considered to be necessary to issue a Technical Stan- inform the design of subsequent types of hydrogen-hybrid dards for Interoperability (TSI) statement. train. Figure 9 shows the main system data plotted as a Numerous tests had to be carried out to support the function of time for the test runs. The whole test cycle approvals effort, including takes a considerable amount of time because the train needed to be prepared and ready, and to meet certain • Electro-magnetic compatibility (EMC) testing. This timing points. This means that the duty cycle comprises of was carried July 2020, using both lineside and on-board extended periods of idling where the train is held before equipment to monitor electromagnetic emissions from given the authority to advance to the next section. the train. The green line in Fig. 9 shows the train speed. The • Annual exam of the equipment on board the train. initial movements are in the test site and are characterised • Fitness to run (FTR) exams. These were carried out by several short and slow movements. The train then enters every day before a mainline test, with sign off by a the mainline and travels on a low-speed branch to join the competent inspector required before operations live railway. Once given permission to move onto the high- speed section, the train then was permitted to travel at its highest speed to Evesham station. After a stop, the train moved beyond the station and returned on the adjacent track for the return journey. Rail. Eng. Science (2021) 29(3):248–257 256 C. Calvert et al. Fig. 9 Performance data plot for the same run from Long Marston to Evesham, showing electrical data for the base Class 319 (top), speed data (middle) and data from the current monitoring device (bottom) A few other data values are included on the plot. These from the HydroFLEX project will continue to inform the include the system DC bus voltage, the traction motor design of follow-on classes of hydrogen train. current, the DC link bus current, the available power, and Acknowledgements This project required the support of a large the CMD gain which moderates the power demanded by number of stakeholders. It is impossible to give thanks and list all the train to within the capability of the propulsion system. contributors individually, but the authors would like to thank the The propulsion system needed to be dynamic enough to following organisations: Porterbrook Leasing Company Ltd., Dr Jeff cope with rapid demand changes from the driver, and Allan of Jeff Vehicles Ltd., Chrysalis Rail Services Ltd., Ballard Power Systems Inc., Denchi Power Ltd., Swagelok, Rail Alliance, considerable effort has gone into the design to cope with Rail Operations Group, Luxfer, dg8 Design & Engineering Ltd. The changes in notch power settings at different speeds. authors would like to acknowledge the generous support of Porter- brook Leasing Company Limited and the funding supplied by Inno- vate UK. Without this support, the HydroFLEX project would not have been possible. 6 Conclusion Open Access This article is licensed under a Creative Commons The HydroFLEX team showed the possibility of operating Attribution 4.0 International License, which permits use, sharing, hydrogen-hybrid trains were on the British railway net- adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the work. This was achieved by developing a concept design, source, provide a link to the Creative Commons licence, and indicate consisting of new equipment mounted inside the PMOS if changes were made. The images or other third party material in this carriage, and progressing this design through detailed article are included in the article’s Creative Commons licence, unless design and installation to testing. The resulting Hydro- indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended FLEX train was successfully demonstrated at Rail Live. use is not permitted by statutory regulation or exceeds the permitted Beyond this its potential was developed through modi- use, you will need to obtain permission directly from the copyright fications to the design, and the train was approved to run on holder. To view a copy of this licence, visit http://creativecommons. the mainline railway, with numerous documents produced org/licenses/by/4.0/. to support this. The learning experience and data gathered Rail. Eng. Science (2021) 29(3):248–257 Concept development and testing of the UK’s first hydrogen-hybrid train (HydroFLEX) 257 10. Alstom (2018). World premiere: Alstom’s hydrogen trains enter References passenger service in Lower Saxony. Available at: https://www. alstom.com/press-releases-news/2018/9/world-premiere-alstoms- 1. United Nations (2015). Paris Agreement. https://unfccc.int/files/ hydrogen-trains-enter-passenger-service-lower. Accessed 10 Nov meetings/paris_nov_2015/application/pdf/paris_agreement_ english_.pdf. Accessed 10 Nov 2020 11. RSSB (2015). Requirements for the application of standard 2. Ye B, Jiang J, Liu J, Zheng Y, Zhou N (2020) Research on vehicle gauges. https://www.rssb.co.uk/standards-catalogue/ quantitative assessment of climate change risk at an urban scale: CatalogueItem/GERT8073-Iss-3. Accessed 10 Nov 2020 Review of recent progress and outlook of future direction. Renew 12. Hillmansen S (2003) The application of fuel cell technology to Sustain Energy Rev 135:110415 rail transport operations. Proc Inst Mech Eng Part F J Rail Rapid 3. ORR (2019). Rail emissions: 2018–19 annual statistical release. Transit 217(4):291–298 https://dataportal.orr.gov.uk/media/1550/rail-emissions-2018-19. 13. Porterbrook (2019). Meet the FLEX family. Available at: https:// pdf. Accessed 10 Nov 2020 www.porterbrook.co.uk/innovation/case-studies/the-flex-family 4. UK Government (2015). New car carbon dioxide emissions. Accessed 10 Nov 2020 https://www.gov.uk/government/publications/new-car-carbon- 14. Kossiakoff A, Sweet WN, Seymour SJ, Biemer SM (2011) Sys- dioxide-emissions#:*:text=The%20average%20carbon% tems engineering principles and practice, 2nd ed. Wiley 20dioxide%20emissions,of%20carbon%20dioxide%20per% 15. Rolling Stock Strategy Steering Group (2018). Long term pas- 20kilometre. Accessed 10 Nov 2020 senger rolling stock strategy for the rail industry, 6th ed. https:// 5. Statista. 2020. Average car and van occupancy England www.raildeliverygroup.com/files/Publications/2018-03_long_ 2002–2018 Statistic | Statista. https://www.statista.com/statistics/ term_passenger_rolling_stock_strategy_6th_ed.pdf. Accessed 20 314719/average-car-and-van-occupancy-in-england/. Accessed Oct 2020 13 June 2021. 16. European Parliament And Council (2009) Regulation (EC) No 6. UK Government, (2019). UK becomes first major economy to 79/2009 of the European Parliament and of the Council of 14 pass net zero emissions law. https://www.gov.uk/government/ January 2009 on type-approval of hydrogen-powered motor news/uk-becomes-first-major-economy-to-pass-net-zero- vehicles, and amending Directive 2007/46/EC. https://eur-lex. emissions-law. Accessed 10 Nov 2020 europa.eu/eli/reg/2009/79/oj. Accessed 04 Feb 2021 7. Hickman AL, Baker CJ, Delgado-Saborit JM, Thornes JE (2018) 17. RSSB (2020). Railway group standard GMRT2100: rail vehicle Evaluation of air quality at the Birmingham New Street Railway structures and passive safety (Issue 6). Station. Proc Inst Mech Eng Part F J Rail Rapid Transit 18. Crawley F. (2020) A guide to hazard identification methods, 2nd 232(6):1864–1878 ed. Elsevier, Oxford 8. Keen P, Phillpotts R (2010) Low cost electrification for branch 19. Crawley F, Tyler B (2015) HAZOP: guide to best practice, 3rd. lines. https://assets.publishing.service.gov.uk/government/ Elsevier, Oxford uploads/system/uploads/attachment_data/file/3872/low-cost- electrification-report.pdf. Accessed 10 Nov 2020 9. Hoffrichter H (2013) Hydrogen as an energy carrier for railway traction. Dissertation, University of Birmingham, Birmingham. Rail. Eng. Science (2021) 29(3):248–257

Journal

Railway Engineering Science – Springer Journals

Published: Sep 1, 2021

Keywords: Hydrogen; Fuel cell; Hybrid; Systems engineering

There are no references for this article.