Beyond the Wire: Battery-Electric Multiple Unit (BEMU) Explained

The decision notice published by the UK Department for Transport in October 2017 cancelling the planned electrification of the Midland Main Line north of Kettering, the TransPennine route between Leeds and Stalybridge, and the Welsh Valleys routes was, in the words of the House of Commons Transport Select Committee’s subsequent report, “a significant setback to rail decarbonisation.” It also created an immediate operational problem: train operating companies had already ordered, or were committed to ordering, rolling stock on the assumption that electrification would proceed. Angel Trains, the rolling stock owner, and Hitachi had designed the Class 802 bi-mode specifically to operate on these routes using OCS electric power; without the electrification, the fleet would run in diesel mode on the very sections it had been bought to electrify. The DfT’s “solution” — announcing that new battery technology would bridge the gap — was initially received with scepticism by the rolling stock engineering community. In 2017, no purpose-built BEMU was in commercial service on a mainline UK railway. The battery technology that would be required — a 700–900 kWh pack capable of delivering traction current at 500–800 A for 50–80 km, surviving 12 years of daily charge-discharge cycles, and fitting under a train bodyshell designed to BS EN 15227 crash standards — did not yet exist as a validated railway product. Six years later, the picture had changed substantially. Stadler’s FLIRT Akku BEMU entered commercial service with several German regional operators from 2022 onwards; CAF’s Civity Battery variant was under test in the UK; Alstom’s Coradia Continental battery variant was operating in Germany. The trajectory was clear. But the honest assessment remains that every BEMU currently in service represents an engineering compromise between range, weight, capacity, and cost that the next generation of battery chemistry has not yet fully resolved — and that understanding where those compromises lie is essential to evaluating which routes BEMUs can realistically serve and which still require conventional electrification.

What Is a Battery-Electric Multiple Unit (BEMU)?

A Battery-Electric Multiple Unit (BEMU) is a self-propelled passenger railway vehicle that derives its traction power from a combination of a conventional overhead contact system (OCS) pantograph — used on electrified sections — and an onboard rechargeable battery pack of sufficient capacity to sustain full traction operation on non-electrified sections for a defined range. The battery pack is charged primarily from the OCS during electrified operation, and secondarily from regenerative braking energy recovered during all braking events regardless of mode.

The BEMU is distinct from: a conventional EMU (which has no meaningful battery traction range beyond emergency shore power supply); a bi-mode train (which combines electric traction with a diesel engine for off-wire operation, retaining direct combustion emissions); and a battery-only train (which has no pantograph and must be charged entirely from wayside infrastructure). The BEMU’s value proposition is specifically the replacement of diesel traction on partially electrified routes without requiring full electrification of the non-electrified sections — a decarbonisation strategy whose economics depend critically on the ratio of electrified to non-electrified route length on the specific lines served.

There is no single dedicated European standard for BEMU traction systems, but applicable standards include EN 50633 (Railway applications — Fixed installations — Electric traction — Rules for the application of power electronics equipment for rolling stock — specifically traction battery systems), IEC 62928 (Railway applications — Rolling stock — General requirements for onboard traction battery systems), and the TSI Rolling Stock energy subsystem requirements of Commission Regulation (EU) 2014/1302 as amended. In the UK, RIS-3703-RST governs rolling stock energy interface requirements including battery traction systems.

Battery Technology: Chemistry, Capacity, and the Weight-Range Trade-off

Cell Chemistry Comparison

The choice of battery cell chemistry for BEMU traction systems involves a multi-dimensional trade-off between energy density, power density, cycle life, operating temperature range, safety, and cost. Four chemistries have been deployed or seriously evaluated in railway applications:

System-Level Energy Calculation

The gap between cell-level energy density and system-level energy density is substantial and frequently understated in media coverage of BEMU capabilities. A battery system is not a collection of cells — it includes the battery management system (BMS) electronics, the thermal management system (liquid cooling circuits and heat exchangers), the structural housing (steel or aluminium crash-compliant enclosure to EN 45545 fire protection standards), fire suppression equipment (typically a gaseous suppression system per IEC 62933-5-1), and all high-voltage busbars, contactors, and fusing. These ancillary components typically add 30–50% to the mass of the bare cell pack:

The weight penalty of 6.6–10 tonnes per three-car unit represents a meaningful fraction of the axle load budget. Railway infrastructure imposes maximum axle loads (typically 18–22.5 tonnes on UK and European secondary routes) that limit the total mass available for all train systems. Adding 6–10 tonnes of battery equipment to a three-car unit with three powered bogies (six axles) consumes approximately 1.0–1.7 tonnes per axle of the axle load allowance — a significant constraint that forces BEMU designers to reduce structural mass or limit battery capacity relative to a pure-electric EMU of the same design.

Power Architecture: How Energy Flows Through a BEMU

The electrical architecture of a BEMU integrates three energy flows — OCS input, battery storage, and traction output — through a DC intermediate bus that connects all three. The design of this bus and its power electronics determines the efficiency, flexibility, and degraded-mode capability of the system.

The DC Intermediate Bus

In a conventional 25 kV AC EMU, the pantograph feeds a main transformer that steps the 25 kV AC down to a lower AC voltage (typically 900–1,800 V AC), which is then rectified to a DC intermediate bus (typically 1,500–3,000 V DC) feeding the traction inverters. In a BEMU, the battery pack is connected to this same DC intermediate bus through a bidirectional DC-DC converter. The converter steps the battery voltage (typically 400–900 V DC for a lithium traction pack, depending on the series/parallel cell configuration) up to the intermediate bus voltage, allowing the battery to inject current into the bus for traction, and steps the intermediate bus voltage down to the battery voltage for charging. The bidirectional converter enables four operating modes:

Battery Management System (BMS)

The Battery Management System is the safety-critical computer that monitors every cell in the battery pack (or cell group, in large packs) and controls charge and discharge to prevent damage, thermal runaway, and over-discharge. Key BMS functions in a railway context include: State of Charge (SoC) estimation — the calculation of remaining usable energy as a percentage of full capacity; State of Health (SoH) estimation — the determination of how much the battery’s total capacity has degraded from its new condition; cell balancing — redistributing charge among cells in a pack to prevent individual cells reaching full charge while others remain partially charged; thermal management control — adjusting cooling circuit pump speed and heater operation to maintain all cells within the operational temperature window (typically 15–35 °C for NMC cells); and fault isolation — disconnecting individual modules or the entire pack from the DC bus if any parameter exceeds safety limits. The BMS of a modern BEMU traction pack is certified to IEC 62133 (battery safety) and IEC 61508 SIL 2 (functional safety of the disconnect function), reflecting that pack isolation is a safety-critical action whose failure could result in thermal runaway propagation to the vehicle structure.

Range Analysis: What Determines How Far a BEMU Can Go

The non-electrified range of a BEMU is not a fixed number — it is a function of at least six independent variables that interact in ways that make simple “range” claims in promotional material frequently misleading. Understanding the true determinants of range is essential to route feasibility assessment.

Key Range Variables

• Gradient profile: A 3-car BEMU climbing a sustained 1 in 60 gradient (16.7‰) at 100 km/h consumes approximately 2.5–3.5× more traction energy per kilometre than the same train on level track. Many UK rural routes that BEMUs are intended to serve — Inverness–Wick, Shrewsbury–Aberystwyth, the Cambrian Coast — include sustained grades that significantly reduce the level-terrain range figure.
• Battery state of health (SoH): A battery pack warranted for 2,000 charge-discharge cycles to 80% of original capacity will deliver noticeably less range after 5–6 years of service. A route designed for a 90 km non-electrified section at fleet introduction may become operationally marginal (requiring slower speeds or additional charging stops) by year 8 of service if the SoH declines faster than projected.
• Ambient temperature: NMC lithium-ion cells deliver approximately 85–90% of rated energy at 0 °C and 75–80% at −10 °C compared to their 20 °C benchmark capacity. This temperature dependence directly reduces effective range in winter operations and is a critical design parameter for BEMU operations in northern Scotland, Scandinavia, or alpine regions.
• Passenger loading: A fully loaded train (seated capacity + standing) has higher mass than an empty train, increasing rolling resistance and gradient energy consumption by 5–15%. BEMU range figures quoted by manufacturers typically assume average load factors; peak-load winter services on scenic routes (where tourist traffic coincides with worst temperature conditions) represent the worst-case range scenario.
• HVAC demand: Cabin heating in winter consumes 15–25 kW per car on modern EMU platforms — a continuous auxiliary load that runs whether or not the train is moving. At temperatures below 0 °C, HVAC can consume 10–15% of total energy per journey, directly reducing battery range by the same proportion.
• Regeneration opportunity: A route with frequent stops (station spacing <5 km) on level or downhill grades provides significant regeneration opportunity. A long-distance rural route with sparse stations and uphill gradients provides very little. Regeneration recovery on typical BEMU routes ranges from 8% to 32% of gross traction energy.

The Charging Distance Calculation

BEMUs in Service: Global Deployments

The Denmark Case Study: Banedanmark Skjern–Holstebro

The Banedanmark Skjern–Holstebro line, which entered BEMU operation in March 2025, represents the clearest proof-of-concept for terminus-charging BEMU operations on a long non-electrified route. The 67 km line runs entirely without OCS between Skjern — where it connects to the electrified main line — and Holstebro. Stadler FLIRT Akku units are charged during the 5-minute terminus dwell at Holstebro via an OCS stub installed specifically at the terminal platform, and recharged from the main line during the approach run into Skjern junction. The operating data from the first months of service showed average energy consumption of 9.1 kWh/km in winter (November–February), rising to 7.8 kWh/km in summer — a 14% seasonal variation driven primarily by HVAC load. Battery range at 20% reserve was confirmed at 74 km in winter conditions and 87 km in summer, with terminus charging recovering 22–28% of battery capacity per 5-minute stop. No service cancellations due to battery range issues were recorded in the first six months of operation.

BEMU vs. Diesel (DMU) vs. Bi-Mode vs. EMU: Full Technical Comparison

Frequently Asked Questions

1. Why is a BEMU’s battery range significantly shorter than an electric vehicle’s, given that both use lithium-ion batteries?

The comparison is frequently made but is misleading because the energy demand profiles and mass scales are so different. A battery electric car (BEV) such as a Tesla Model Y weighs approximately 2,000 kg and consumes roughly 0.15–0.20 kWh per kilometre at motorway speeds. Its 75 kWh battery provides approximately 375–500 km of range. A three-car BEMU weighs approximately 150,000 kg — 75 times heavier — and consumes approximately 8–12 kWh per kilometre on level terrain. Scaling a BEV’s range to BEMU mass would require: 75× the energy consumption = 75 × 0.175 = 13.1 kWh/km (broadly consistent with actual BEMU figures); and 75× the battery capacity to achieve the same range = 75 × 75 kWh = 5,625 kWh of battery. At current system-level energy density of 150 Wh/kg, a 5,625 kWh battery would weigh 37,500 kg — nearly as much as the entire traction equipment of the BEMU itself, and approximately four times what the axle load budget can accommodate. The fundamental constraint is that trains are very heavy relative to their power-to-weight ratio, rolling resistance (though low per tonne) accumulates over long distances, and the battery mass fraction of the total vehicle is constrained by structural and infrastructure limits that do not apply to road vehicles. Battery energy density would need to approximately triple (to ~450 Wh/kg system level, approaching projected solid-state battery performance) before a BEMU could offer 300+ km range on a comparable axle load to today’s designs.

2. What happens to a BEMU’s service if the battery state of charge drops below the safety reserve during a journey — is there an emergency procedure?

The battery management system on a BEMU includes a hierarchical response to low state of charge (SoC) that is integrated into the train management system and signalled to the driver via the driver machine interface (DMI). The response stages are typically: a caution threshold at approximately 20% SoC, which triggers an advisory message to the driver and may activate a speed reduction to reduce energy consumption; a warning threshold at approximately 10% SoC, which triggers a mandatory speed reduction (commonly to 50% of line maximum) and a request for operational support from the control centre; and a minimum reserve threshold at approximately 5% SoC, which activates the battery isolation sequence — the traction pack is taken offline to protect the cells from deep discharge damage, and the train coasts to a stop using service brake. Unlike a diesel DMU running out of fuel, a BEMU that reaches minimum reserve does not physically strand in the same way — it comes to a controlled stop using the service brake while retaining auxiliary power (safety bus, lighting, door operation) from a smaller dedicated auxiliary battery, maintaining passenger safety systems throughout. The emergency procedure from this state requires either: a rescue by a traction-powered unit that can push the BEMU to the nearest OCS section; or a wayside recovery vehicle with a portable fast-charging capability — a scenario that has driven several BEMU operators (including Norsk Tog in Norway) to procure mobile battery charging trailers deployable from road access points adjacent to non-electrified sections.

3. How does a BEMU perform on routes with long tunnels — where reduced speed limits apply and the train may spend 20–30 minutes underground without regeneration opportunities?

Long tunnel operation is one of the more challenging scenarios for BEMU energy management, and several proposed BEMU deployments in tunnel-intensive regions (Norway’s Bergensbanen, the proposed Scottish Borders BEMU services) require specific analysis. In a tunnel, the usual operating constraints — ventilation limits, fire safety, emergency egress requirements — impose maximum speeds typically 20–40 km/h below open-line limits, which reduces traction energy consumption proportionally (power roughly proportional to v², so 80 km/h in tunnel vs 100 km/h open gives approximately 36% energy saving). However, tunnels eliminate all opportunities for regenerative braking recovery from gradient descent that would otherwise be available on approaches to mountain tunnel portals — the kinetic energy that would be recovered descending the approach ramp is consumed ascending to the tunnel portal from the far side. More critically, some long railway tunnels operate under safety regulations that prohibit battery-electric traction for a specific segment of tunnel length due to fire safety concerns about lithium-ion battery thermal runaway events in enclosed spaces. EN 45545-2 (fire protection of railway vehicles) and ERA Technical Opinion ERA-2022-INF-06 address battery fire safety in tunnels; the current consensus permits BEMU operation in tunnels up to approximately 5 km long if the battery pack meets the fire suppression and thermal runaway containment requirements of IEC 62933-5-1, but requires specific safety case approval for tunnels above 5 km. The Seikan Tunnel (53.9 km, Japan) and Channel Tunnel (50.5 km) would require battery systems well beyond current technology capability to maintain thermal runaway containment for the duration of a worst-case tunnel fire event, meaning that BEMU operation through these specific structures would not be approvable under current safety frameworks.

4. Can a BEMU be retrofitted from an existing EMU design, or does BEMU capability require a ground-up new design?

Both approaches have been pursued in practice, and each has advantages. Purpose-designed BEMUs — such as the Stadler FLIRT Akku, Siemens Mireo Plus B, and Alstom Coradia Continental Battery — are designed from the outset with underfloor battery pack mounting points, appropriate structural reinforcement for the added mass, and traction architecture optimised for dual-source power management. The advantage is an integrated design with optimal mass distribution, thermal management routing, and BMS integration. Retrofit conversions — adapting an existing EMU bodyshell to accommodate battery packs — have been pursued for several fleets, most notably Eversholt Rail and Porterbrook’s “Project DECarbonisation” proof-of-concept on a Class 379 (originally designed as a 25 kV AC EMU). The Class 379 conversion demonstrated technical feasibility but highlighted the constraints of retrofit: the underfloor space available after the original traction equipment is retained for OCS operation is limited, existing structural cross-sections are not designed for the lateral and vertical loads of large battery enclosures, and the cooling system routing for a large battery pack requires substantial modification of the existing under-floor layout. The class 379 conversion achieved a battery capacity of approximately 200 kWh — sufficient for 15–20 km of battery range — which, while useful for station area operations and OCS-free sections within Network Rail controlled infrastructure, falls well short of the 60–100 km range that operational BEMU deployment requires. The practical conclusion is that retrofit conversions of existing EMU designs are viable for creating vehicles with limited battery range (10–30 km) suitable for short OCS gaps, but purpose-designed BEMUs are necessary for routes where off-wire operation exceeds 30–40 km.

5. What is the end-of-life disposal and recycling challenge for BEMU traction battery packs, and how does the industry address it?

Railway traction battery packs reaching end-of-first-life (typically at 80% of original capacity after 8–12 years and 2,000–4,000 charge cycles, depending on chemistry) present a significant material recovery challenge. A single three-car BEMU traction pack contains approximately 6–10 tonnes of battery system mass, of which the electrochemically active materials include: 400–700 kg of lithium (in the electrolyte and cathode); 600–1,200 kg of nickel (NMC cathode); 200–400 kg of cobalt (NMC cathode); 300–600 kg of manganese (NMC cathode); and 500–800 kg of copper (current collectors and cabling). At 2024 commodity prices, the recoverable metal value of a single BEMU pack is approximately £35,000–65,000 — sufficient to make dedicated pyrometallurgical or hydrometallurgical recycling commercially viable if collection logistics can be organised. Several European manufacturers (Umicore for NMC chemistry; Retriev Technologies in North America for LFP) have established contractual take-back programmes with rolling stock owners: the rolling stock operating company (ROSCO) is contractually committed at purchase to return end-of-life battery packs to the manufacturer’s designated recycler, and receives a small residual value credit against the cost of replacement packs. Alternatively, “second-life” repurposing — using end-of-first-life BEMU packs (at 80% capacity) as stationary energy storage for depot demand management or grid balancing — has been trialled by DB Energie and Porterbrook, with results suggesting that BEMU packs retain adequate capacity and cycle life for 5–8 years of stationary storage application before final recycling. EU Battery Regulation 2023/1542, which entered force in stages from 2024, imposes mandatory recycled content targets for railway traction batteries and requires manufacturers to provide a digital battery passport documenting cell chemistry, origin, and state of health — making the full supply chain for BEMU traction batteries substantially more transparent and accountable than was previously the case.