Decarbonising Railways: The True Cost of Replacing the Diesel Multiple-Unit

August 31, 2026

Print This Article Share it With Friends

by Antoine Belleguie (Rail Logistics Europe/Railway Gazette International) ... SNCF has set out its own vision in its PlaneTER programme, which aims to remove diesel traction from France’s regional network by 2050, converting DMUs to biofuel or replacing them with battery or hydrogen power. The operator says around 230 of its bi-mode Régiolis units could, in principle, be retrofitted. Yet a train working an intensively used suburban diagram under existing wires has little in common with one rostered for a lightly trafficked rural branch, and the economics that flow from those differences are stark.

Passenger rail is already heavily electrified; more than 85% of all passenger train movements around the world run on electricity, compared with little more than half of all freight trains. However, the residual diesel problem is concentrated precisely where it is hardest to solve: on regional and rural lines, worked by multiple-units that can often spend much of the day running under the catenary on shared trunk sections before peeling off onto a non-electrified branch.

Multiply that picture across France, Germany, and the rest of Europe, and the regional DMU is one of the most widespread yet addressable sources of avoidable emissions on the railway.

...

Before making any cost comparisons, though, it is worth setting out clearly what is currently available (Table I).

Conventional electrification by overhead catenary remains the most efficient route to zero operational emissions. Whether power is supplied at 25 kV 50 Hz, 1·5 kV DC or one of the other standard voltages, once it reaches the train, the electric drive converts 85 to 90% of the energy into motion. Running costs are low, but the capital cost of electrification can be high: typically €0·35m to €1·5m/km, and more if gauge clearance works at bridges and tunnels, or substation enhancements are needed.

One interesting development is partial, or discontinuous, electrification: wiring only the critical sections such as gradients and stations and bridging the gaps with bi-mode or battery units. SNCF Réseau is trialling this approach on the Marseille – Aix – Gardanne line, and several German schemes are proposing a similar strategy.

...

Hydrogen offers a credible replacement for diesel on longer non-electrified routes. Alstom’s Coradia iLint has been in commercial service in Germany since 2018, while the Siemens Mireo Plus H, Stadler Flirt H2, Alstom Coradia Stream H2, a hydrogen Régiolis and CAF’s Civia have followed. The catch is thermodynamic. A proton-exchange-membrane fuel cell reaches 55 to 60% efficiency at the stack in ideal conditions, but the net figure — once the compressor, thermal management and power electronics are included — falls to 45 to 50%, and that is before considering the losses incurred in producing, compressing and distributing the hydrogen.

Existing diesel power packs can be adapted to work as a hydrogen internal-combustion engine; the thermal efficiency of 35 to 45% is comparable to a modern diesel. Although these still generate nitrogen oxide emissions, these can be cut by 90% to 95% with catalytic after-treatment. This offers a transitional route that reuses much of the diesel powertrain. At the other extreme, liquid hydrogen may suit heavy long-distance haulage in sparsely electrified territories such as North America and Australia, but liquefaction at around -253°C consumes roughly 30% of the energy in the gas — 13 to 15 kWh for every kg — which for now confines its use to niche applications where volumetric density is decisive.

Biofuels are the quiet pragmatists which help to decarbonise the fleet that already exists. Pure first-generation biodiesel — B100, made from rapeseed methyl esters — cuts life-cycle greenhouse-gas emissions by 50 to 60% and can be used in diesel engines with only limited modification. However, it is more viscous than diesel and less stable in the cold.

SNCF Voyageurs has operated a fleet of Régiolis sets using B100 between Paris and Granville in Normandie since 2021. Hydrotreated vegetable oil goes further: a paraffinic fuel chemically close to fossil diesel, it is a genuine drop-in, with a cetane number above 70, significantly lower particulates and a 60 to 90% life-cycle carbon reduction depending on the feedstock. Rolls-Royce has cleared its MTU series 4000 engines to use HVO, while Stadler offers HVO compatibility across its WINK and Flirtino platforms.

Lower blends of biofuel, from B10 to B30, and bio-derived compressed and liquefied natural gas complete the picture, though their rail use is largely outside Europe. LNG can cut carbon emissions by more than 80% when produced from biomethane, but none of these delivers zero operational emissions, and the upstream footprint depends heavily on feedstock traceability. Nevertheless, they offer the fastest and cheapest reductions currently available.

Mixing modes

...

The decision cannot be made on capital cost alone, and this becomes clear when considering ‘well to wheel’ energy efficiency — the share of primary energy that actually reaches the wheel once all of the upstream losses have been accounted for (Fig 1).

Fig 1. Well-to-wheel efficiency of the principal railway traction chains. The wide gap between electric and hydrogen options feeds straight through to the operating budget.

A 25 kV EMU converts three to five times as much of its source energy into motion compared to a hydrogen train. The chart includes the upstream losses incurred while generating and delivering the electricity, which is why the EMU sits at 75 to 85% rather than the 85 to 90% efficiency of electric drive alone. For a planner, that is the difference between a modest and a substantial energy cost in the operating budget, repeated every day for 30 years or longer. Efficiency, though, is only one input in a wider reckoning.

Putting a price on the choice

For decades, the rail industry has leaned on a single tool when comparing traction options: Total Cost of Ownership. This is indispensable, but on its own it can be misleading — because it only tells the story from the perspective of the operator’s balance sheet, and does not factor in the costs the train imposes on everyone else. The framework I developed during doctoral research conducted between Université Gustave Eiffel and Alstom, widens the lens to look at the Total Social Cost, which combines three terms: the operator’s TCO, the Total External Cost borne by society, and the net effect on public finances, FiPu (Fig 2).

...

On a lightly used branch, full electrification might cost several hundred euros for every tonne of CO₂ avoided, well above the line, while a battery or biofuel option slips beneath it. The technology has not changed; only the traffic volume. That single observation, that the best option is determined by the route, not the train type, is perhaps the most important conclusion of the whole exercise, yet easily lost in a procurement debate framed largely around vehicles.

Wider implications

This is relevant to both the operators and the authorities that specify their fleets. The instinct to standardise on a single technology across a network is understandable, as it simplifies maintenance, training and procurement, but it is rarely the cheapest path once the whole social cost is considered. The optimal answer must be shaped by the architecture of the network and the structure of the market it serves. A robust comparison demands holistic data: not just the cost of the train, but the energy chains, externality values, residual values and a realistic assessment of how much each route will be utilised over the coming decades.

...

A biofuel retrofit depends on having a secure, traceable supply of certified fuel. None of these is a reason to rule out an option, but each can quietly shift the balance of a comparison that looked settled on a spreadsheet.

...

Because so many of the benefits accrue to society rather than the operator, the business case often depends on well-targeted support — capital grants, green financing, or the adjustment of track access charges to reflect environmental performance, as is now being applied in Germany, Austria and Switzerland, for example.

There is a tension to resolve, however. The European green taxonomy, which increasingly governs access to sustainable finance, screens on the absence of direct operational emissions. That risks excluding the transitional options such as diesel-battery bi-modes or certified biofuels that can offer a cost-competitive abatement on lines where full decarbonisation may be years away. A framework built to reward environmental purity can end up penalising the most cost-effective option to lower emissions in the near term. Reconciling that purity with the economics of a credible transition is one of the more important debates that the rail sector currently faces. READ MORE

Tags

What You Can Do trains Switzerland Sustainability railroad particulates methane/biomethane market forces locomotives liquefaction Investing internal combustion engine (ICE) Hydrogen/Renewable Hydrogen hydrogen internal combustion engines HVO (Hydrotreated vegetable oil) GHG (Greenhouse Gas Emissions) Intensity Germany freight/cargo France European Union (EU) environmental policy Electric Car/Electric Vehicle (EV) efficiency drop-in biofuels/hydrocarbons diesel Compressed Natural Gas (CNG/R-CNG/bioCNG) Carbon Intensity (CI)/Carbon Footprint Canola/Rapeseed biodiesel bio-LNG (Liquified Natural Gas) B30 B100 B10 Austria

Comments are closed.