In February 2015, a battery-fitted Class 379 Electrostar undertook trials on the Harwich branch. This independently-powered EMU (IPEMU) was the first battery-powered train on the UK rail network for more than 50 years. During its trials, it ran with its pantograph retracted to give it a duty cycle of 30km on battery power and 50km powered by the overhead line equipment (OLE). At one stage, a range of 77km on battery power was achieved.
Our report on this trial in Issue 125 (March 2015) showed that its batteries weighed eight tonnes and had a capacity of 450kWh, half of which was required for heating and auxiliaries. This is an energy density of 56 Watts/kg.
Though this trial successfully demonstrated the potential of IPEMUs, it was to be over eight years before a production IPEMU entered service. This was Merseyrail’s Class 777/1 EMU which has a 55km battery range to enable it to operate on an unelectrified line to Headstone Lane. IPEMUs are now known as Battery EMUs (BEMUs) while tri-modes are trains that can operate on electric, battery, or diesel power.
In November 2024, Transport for Wales (TFW) introduced the Class 756 tri-mode unit. Next year, TfW will also introduce battery-powered Class 398 tram-train BEMUs on their Core Valley Lines which have discontinuous electrification. The Great Western Railways (GWR) Class 230 units are entirely battery powered and are to be introduced this year on their Greenford branch.
Thus, battery traction is now a mature technology that will increasingly be a part of future rolling stock orders with significant implications for future electrification. Hence a recent seminar run by the Institution of Mechanical Engineers (IMechE) titled ‘Rolling Stock 2025: Batteries and BEMUs’ was indeed timely.
Fast charging
The Class 230 battery unit was originally developed by Vivarail by re-engineering surplus London Underground D stock. After the company went into administration, GWR acquired battery trains and their fast-charging system. Julian Fletcher, GWR’s technology development manager for the trial on the West Ealing to Greenford branch described the programme to test this unit and its charging system in all conditions. These included: very cold (-18° C), hot, and wet weather; leaves, litter and oil on charging rails; charging system failure; crush loading; and grid failure. The trials proved the system was safe under all conceivable circumstances.
The charging system has a 400V three-phase 63A grid connection which trickle charges a 760V 430kWh fast charge battery bank, and has charging rails within the four foot at West Ealing station’s bay platform which are normally de-energised. When the unit is at the platform, it detects a balise which lowers the charging shoes and energises the charging rail. These charged the unit’s batteries in 3.5 minutes at the rate of 700kW.
Although the 8km round trip on the Greenford branch is not demanding in respect of battery range, the Class 230 travelled 113km on battery power on its journey from Long Marston testing ground to the Reading Train care depot. This shows its potential for use on branch lines much longer than the Greenford branch.
Julian also made the point that a re-engineered vehicle such as the Class 230 has half the embodied carbon of new stock.
Dublin DART+ BEMUs
Peter Smyth, Irish Rail’s chief mechanical engineer, explained how new BEMUs were being introduced to expand the Dublin Area Rapid Transit (DART) system which were ordered in 2021. Twenty-six five-car EMUs and 31 five-car BEMUs have been ordered which will enter service from 2026 onwards. These are 1500V DC OLE, 145km/h Alstom X’Trapolis units, and are fully articulated. They have 550 seats and are 82 metres long.
They offer level boarding with a low floor made possible as the traction battery, and its thermal management and power control systems are roof mounted. The water-cooled Lithium Nickel Manganese Cobalt Oxide (NMC) batteries have a capacity of 840kWh per five-car unit giving it a range of at least 80km. They have a minimum nine-year life to 75% capacity and have been sized after extensive simulations which included degraded scenarios.
The BEMUs will extend the DART service to Drogheda on the 38km unelectrified line from Malahide. Their worst-case duty cycle is a 105km round trip from Drogheda to Dublin Connolly station and back, of which 29km is on the DART’s electrified wires lines. For this reason, the BEMUs will be charged by a fast charger, although regenerative braking will charge batteries while under OLE.
The fast charger facilities have been installed at Drogheda on two platforms and one siding. It has 3.6MWh of energy storage to enable up to four trains of 2 x 5 cars to be charged simultaneously. It will take 20 minutes to charge a five-car BEMU via the two pantographs on its centre car with a charging current of 800A.
The introduction of these trains required a significant amount of infrastructure work including: balises to raise / lower pantographs, simulator building, trackside ETCS, a testing and commissioning facility, and significant depot alterations including a refurbishment of a spare shop with five-car lifting facility.
Peter also described the extensive battery fire protection precautions which included an underframe barrier providing 15-minute protection, thermal cameras, shock sensors for transport, and fitment and battery quarantine areas. While there may be an internal failure of a single cell, the resultant impact would be manageable. However, a fire and explosion from a multiple cell thermal runaway is assessed to be within the SIL2 safety target.
Dynamic testing of two BEMUs had started in February, and these are expected to enter service towards the end of next year.
TfW’s BEMUs
TfW Head of Fleet Readiness Rowan Philips explained how the Core Valley Lines out of Cardiff have been transformed with discontinuous electrification and bespoke rolling stock providing a metro-style operation with level boarding. This also required signalling, station, and depot upgrades as well as a new depot for tram-trains.
The discontinuous electrification is 170 single track kilometres (stk) of OLE with 30 catenary free sections (CFS), which generally start at low bridges, and 60 permanently earthed sections (PES). Twenty-four kilometres of the OLE is permanently earthed and CFS have avoided the need to electrify 81stk. A balise before the CFS lowers the pantograph. Should this fail to operate, the OLE is raised to 6.3 metres above track level after 10 seconds running time before the low bridge to raise the pantograph above its maximum normal height causing it to drop automatically.
Two types of battery trains operate under this discontinuous electrification. There are 24 tri-mode Class 756 units which have three or four passenger vehicles with a short ‘Power Pack’ vehicle near the centre of the unit. This contains a 480kW diesel generator set and three LTO battery modules that can supply up to 1,300kW. These trains entered service in January on services between Rhymney and Barry / Bridgend. Battery capacity on the three and four-car trains is respectively 559kWh and 447kWh.
A frequent service on the Teherbert, Aberdare, and Methyr routes is to be provided by 36 Stadler Class 398 CityLink units. These three-car tram-train BEMUs have 138kWh LTO batteries that supply up to 600kW. These are currently under test and are expected to enter service early next year.
The Class 756 units are maintained at Cardiff Canton which has been upgraded with, for example, high level access to a battery store and the provision of OLE for charging. The Class 398’s are maintained at a new depot at Taff Wells.
Their energy storage system (ESS) which comprises the traction batteries and battery management system is supplied and supported by ABB. Rowan advised that its maintenance is largely condition based visual inspections for which training and competency requirements were minimal. Rowan advised that the six-month operational experience with the Class 756 had shown the reliability of the ESS to be very good.
Nevertheless, the risk from defective batteries must be addressed as required by the Dangerous Substances and Explosive Atmospheres Regulations. Mitigation to address this includes training to identify critically defective cells and quarantine processes.
As the units are tested and introduced into service, battery charge is closely monitored and compared against that predicted from modelling
Hitachi’s big battery
The biggest battery considered by the seminar was that fitted to a TransPennine Class 802 bi-mode unit which was the subject of a presentation by Hitachi’s Christopher Dautel. The Class 802 is a five-car unit which, in addition to its 25kV OLE power supply, has three cars powered by 700kW Rolls Royce MTU stage IIIB compliant generator units (GU). They operate on the 220km route between Newcastle and Liverpool of which a 90km unelectrified section has steep gradients, long tunnels, covered stations, and big urban areas.
This trial commenced in May 2024 with a Class 802 on which one GU had been replaced by a 575kWh battery and a roof-mounted cooling unit. This reduced the weight of the car concerned by 0.5%. The battery supply was configured so that the train could:
- Run on battery only in operation and when stabled.
- Use battery in combination with other two GUs.
- Recharge from OLE and recover braking energy with potential to charge from GUs.
The preliminary report of this trial showed that:
- Had all three GUs been replaced by batteries, CO2 emissions between Newcastle and Liverpool would be 835kg which is 38% of current emissions.
- Between 0 to 72 km/h the battery provided the same acceleration as the GU.
- The battery had a range of at least 70km at speeds up to 140km/hr.
- Fuel savings in fast mode and eco mode were respectively 35% and 57%.
- It took 50 minutes to charge the battery from the OLE with a current limit of 50A.
- The battery could power train auxiliaries for five hours.
- If all three GUs were replaced by batteries, a return run of 135 km off wire was feasible e.g. York to Scarborough.
Batteries and power
Luke Nolson, traction sales manager for ABB explained the issues associated with battery traction and the capabilities of different cell technologies. He described ABB’s extensive experience of providing batteries and other traction equipment including those on TfW’s battery trains. On electric vehicles, traction batteries offer catenary-free operations and reduced peak power requirement. For diesel-electric vehicles they offer engine downsizing, reduced maintenance and fuel consumption, as well as providing an acceleration boost.
Luke described how ABB had undertaken tests to determine aging and power capability of the following cell technologies: NMC (Nickel-manganese-cobalt), LFP (lithium ferro-phosphate), LTO (lithium-titanium-oxide). The results of these tests are shown in Table 1.
This shows that although they have a low energy density, LTO cells have significant advantages. For example, their excellent charging performance can collect more regenerated energy.
The required battery capacity is determined by total required energy, traction and auxiliaries load, safety factors, cell degradation, and charging rate. Power restrictions are necessary to maximise battery life. As shown in the diagram, an LTO battery has a higher state of charge range without restrictions than a conventional battery.
Luke stressed that battery monitoring to maintain the battery’s state of health and maximise its life is essential. As an example, battery temperature needs to be monitored against permitted number of hours for each temperature range.
Dr Dave Hewings, Network Rail’s head of engineering and asset management (electrification & fixed plant) for Wales and Western, explained how traction batteries can make the best use of the electrification asset. However, this requires the electrification system design philosophy to take full account of battery electric train operation. For example, design should not assume continuous electrification and take account of continually improving battery energy densities and the need for dynamic battery charging.
He felt that, for freight, an option might be for the locomotive to haul a wagon with traction batteries. Yet the energy stored in the 5,700-litre fuel tank of a Class 66 locomotive is equivalent to batteries weighing over 800 tonnes which would require 27 x 20ft shipping containers.
He noted that while energy demand for electric trains has always been calculated, rather than assumed, battery train range is often assumed rather than calculated. One issue is that traction system analysis pays little attention to stationary trains as auxiliary system demand is considered to be insignificant. In this respect having a figure for maximum auxiliary demand not suitable for battery performance analysis. He felt that, although air temperature can be used to analyse the range of heating and cooling demand, probabilistic modelling is needed to avoid worst-case design limits.
Research
Professor Stuart Hillmansen of the University of Birmingham’s Centre for Railway Research and Education noted that not that long ago battery traction and discontinuous electrification were only being simulated and that now they were a reality. This question now is what can be achieved with these new(ish) technologies. Answering this question requires an understanding of the physics of railway traction in which the acceleration of a train is net-force divided by the train’s mass and this force is the train’s tractive effort minus resistance and gravitational force due to gradient.
Stuart noted that resistance is proportionate to speed squared and that the power required by a train is proportional to the cube of its speed. Hence, a 125mph train requires almost twice the power of a 100mph train. As a result, he felt that BEMUs were suited to suburban services with an off-wire range of around one hour or 80 km, although he expected technology would improve this range.
He noted that charging trains from the OLE offered various opportunities including the use of trains as mobile energy storage devices and limiting maximum pantograph power draw, and that this highlighted the need for an energy management system.
This and other related research related to BEMUs was described in a presentation by RSSB’s Richard Turner which described the research reports shown in Table 2.
Traction strategy
What the introduction of BEMUs means for future rail decarbonisation plans was considered by Jonathan Williams, Network Rail’s lead strategic planner, who advised that, consistent with available funding, Network Rail has the objective of promoting the effective use and development of rail capacity. He considered that the establishment of Great British Railways (GBR) provides greater opportunities to integrate track and train to develop this strategy.
In addition, the commitment to a net-zero railway by 2050 had to be delivered. The Traction Decarbonisation Network Strategy (TDNS) that was published in 2020 had recommended this required electrification of 90% of the network. Currently 46.5% stk of the network is electrified. However, as technology had advanced since then, Network Rail is refreshing TDNS on the basis that BEMUs:
- Have a worst-case range of 50 miles running at 100mph.
- Have a re-charge time of 20 minutes which is best delivered by dynamic charging.
- Run under wires for at least 25% of their journey.
Battery use for freight trains is still assumed to be limited to ‘last mile’ operation. Although BEMUs can decarbonise railway with much less electrification, Jonathan acknowledged there is still much to learn about deploying BEMUs at scale.
David Clarke of the Railway Industry Association (RIA) also agreed that the TDNS’s electrification proposals were unrealistic. As reported in Issue 208 (May-June 2024) RIA’s report ‘Delivering a lower cost, higher performing, net zero railway by 2050’ is a plan of thirds which are: existing electrification; additional electrification; and BEMUs. Hence, RIA proposes that the amount of electrification required is two thirds of the network.
The RIA report considers that electrification is needed for freight and high-speed passenger services and shows that it has significant benefits in addition to decarbonisation. David considered that, with a positive business case, an electric railway is “a better railway however you define it”. However, the problem is affordability and lack of confidence in industry to deliver. He felt that this was unfortunate as much has been done to reduce the cost of electrification, for example voltage control clearances which dramatically reduce the number of bridge reconstructions.
Yet, there is still no committed funding for a rail decarbonisation strategy in England and Wales with no agreement on what should be electrified. David explained how this presented significant problems for the supply chain and for freight operators who must consider how to replace their locomotive fleets.
He explained how RIA is producing a report showing that electrification is affordable. This will propose a commercial approach to minimise benefits and make the programme business case for implementation. It will also propose an efficient rolling programme of electrification and a robust rolling stock plan to maximise early benefits by deploying BEMUs.
After hearing most of the day’s presentations, delegates could be forgiven for concluding that, with the introduction of BEMUs, there is no further requirement for electrification.
Yet, as the RIA report makes clear, the network traction strategy needs to consider freight and 100 mph+ passenger trains as well as suburban passenger trains. In short, a whole-system approach must be taken. Whole-life costs must also be considered which require an understanding of battery costs over a vehicle’s life as well as the costs of discontinuous electrification. As well as CFSs and PRSs, this includes the cost of HV cables feeding all wired sections. The seminar did not mention such costs.
Hence, although the IMechE’s Railway Division is to be congratulated for organising an informative seminar, various questions were left unanswered.
Image credit: Irish Rail