FUEL CELLS combine hydrogen with oxygen in the atmosphere to produce water and electricity, eliminating local pollution. But as things stand, electrolytic production and delivery of H2 as a fuel would consume around four times more power than the fuel cells would actually produce on a train. In comparison, the overall efficiency of electric traction from power station to the transformer output terminals on a train is about 80%.

Despite this formidable hurdle, over the last year there have been several articles in the international press about the possible use of fuel cells on trains. In December 2004, Yomiuri Shimbun reported that Japan's Railway Technical Research Institute plans to run a prototype train powered by a fuel cell by 2010. The US Army is working on a project to rebuild a shunting locomotive powered by fuel cells (p494), although the political emphasis in America is more on reducing dependence upon growing oil imports than halting climate change.

Earlier this year, the UK's Railway Forum, funded by the rail industry, issued a leaflet stating that 'fuel cells provide a very attractive alternative to further electrification and could also be a replacement in areas of the network where existing electrified infrastructure is likely to become obsolete.' At present, 31% of the British network is electrified - low by European standards.

Policy makers have high expectations of 'the hydrogen economy', especially fuel cells in transport. The International Energy Agency quotes Japan's Prime Minister Junichiro Koizumi as saying that 'the fuel cell is the key to opening the doors to a hydrogen economy. We will aim to achieve its practical use as a power source for vehicles and households within three years.' EU Commission Chair Romano Prodi has made a strong statement on the hydrogen economy, and in his 2003 State of the Union address President Bush committed $1·2bn to fuel cell cars1.

Environmental constraints

Setting cars and houses to one side, how far do technical realities support this enthusiasm for H2 as an environmental saviour where rail traction is concerned? Notwithstanding many predictions that oil will run out, untaxed diesel fuel continues to be available at prices similar in real terms to 30 years ago, and nobody seriously expects oil reserves to become exhausted in the next 30 years.

Prices are likely to rise as oil becomes harder to extract and there is greater demand from growing economies, especially China. This is unlikely to upset the balance between diesel and H2 fuel prices very much because, as we have seen in the last year, higher oil prices quickly push up coal and gas prices as electricity generators in particular switch fuels. It is difficult to argue that railways should adopt fuel cells merely to protect against a shortage of diesel.

In terms of local pollution, the UK experience is that over the last decade NOX emissions from transport have fallen by 34%, volatile organics compounds by 40%, CO and particulates by 42%, smoke by 50% and lead by over 90%. This happened despite a 10% increase in personal travel and a 15% increase in goods movements2.

Progressively more stringent European emission standards for new road vehicles will ensure that improvements continue, and it is anticipated that over the next 10 to 15 years emissions will fall below those in 1970 when road traffic was only 40% of today's volumes. Similar reductions are likely in other developed countries.

Much of the above reduction can be attributed to new designs of internal combustion engine for road vehicles, for example lean burn and stratified charge technologies, particulate traps and catalytic converters as well as the adoption of ultra-low sulphur diesel.

In general, rail has lagged behind road transport in adopting these measures, but if they are adopted it is likely that the diesel engine will remain environmentally acceptable. While air quality continues to cause concern, there is unlikely to be any reason for rail to adopt fuel cells to reduce local pollution other than in highly susceptible locations such as Californian conurbations. Local pollution is under control.

The same is not true of greenhouse gas emissions, and thus the principal environmental challenge for fuel cells is contributing to meeting limits on these gases, predominantly CO2. Transport is one of the major sources of CO2 in developed countries, and Fig 2 shows that CO2 emissions from transport continue to rise in the UK.

The UK government, in line with European Union recommendations, has set a target of a 60% reduction in CO2 emissions over the next 30 years. Could the use of fuel cells on main line railways contribute to this objective?

Inter-city case study

As a case study, consider the IC225 fleet used by UK operator GNER. Each train consists of a 25 kV locomotive hauling or pushing 10 coaches. Maximum output of the Class 91 locomotive is 5MW at the wheels, requiring 6·5MW peak output from fuel cells. A 225 km/h path from London King's Cross to Edinburgh, including hotel loads, would require around 25MWh total output from the fuel cells over a 4h period.

Fuel-burning trains must be able to operate for at least one full day without refuelling because going to a depot would significantly reduce productivity, as well as doubling movements in the congested approaches to the terminal station. Some IC225 trains make three one-way 600 km trips a day, a total of 15h in service. Two additional refuelling trips would add 4h to this schedule, requiring 25% more trains and 40% extra capacity in the station approaches.

Refuelling in the station area every 15min would demand storage of sufficient H2 to bring the site into the requirements of the Seveso II Directive 96/82/EC on the control of major hazards. It is unlikely that a safety case could be made for introducing many tonnes of H2 into terminal stations along with thousands of people, high voltage electricity and hot equipment. So our H2 powered version of the IC225 'reference train' must have an onboard storage capacity of around 80MWh.

Onboard H2 storage options are liquid at -253°C and gas compressed at up to 700bar. Both are proven technologies, and storage in metal hydrides or carbon nanotubes is in the early stages of research. The US government has set a target for 2010 of achieving a usable energy/mass storage target for H2 of 2 kWh/kg, within a volume of 1·5 kWh/litre. Applying these targets to the 80MWh IC225 requirement gives an energy storage mass of 40 tonnes within a volume of 50m3, equivalent to a diesel fuel tank. Using the US 2010 target of $4/kWh output would price onboard fuel storage at $320000.

Feasibility of H2 train design

The development of hydrogen storage and fuel cell technologies will cost many billions of euros, and railways would of necessity have to adapt them from other users, most probably road vehicles.

Power equipment on the IC225, excluding motors and their associated converters common to both traction systems, has a mass of around 20 tonnes and occupies about 5m of the train length. How does this compare with our H2 version?

There are six types of fuel cell in use, and the most likely contender is the Proton Exchange Membrane (Fig 3). The PEM cell operates at low temperatures (less than 100°C) and responds well to fluctuating power demand. It has been demonstrated in ratings up to 250 kW and is being developed for road transport. In principle, one could arrange 60 of the PEM cells developed for road use, each giving 100 kW in series/parallel connection, to provide the power necessary for our reference train.

One benefit claimed for fuel cells is the ability to store regenerated energy for reuse. The maximum kinetic energy that has to be disposed of when braking an IC225 is around 1GJ, although normally most of this is dissipated by friction brakes as only four axles are motored.

It is easy to envisage the fuel cell of a road vehicle absorbing electrical energy and generating hydrogen and oxygen which are stored in the gas pathways to the cathode and anode. Trains have a very different duty cycle to cars, however, and the braking energy per kW of fuel cell rating for the reference train running at 200 km/h is 10 times greater than that for a 75 kW car. We can only say that the ability of a fuel cell to absorb typical inter-city braking energy is unproven.

So onboard H2 storage is feasible in our train, but would reduce the revenue space by about two-thirds of a vehicle if stored above floor. Using current designs for cars, a 6MW cell would have a volume of about 5m3 and a mass of 5 tonnes, comparable to a 25 kV transformer.

However, current PEM fuel cells have a coolant output temperature of 70°C, which is only 30°C above an ambient of 40°C. In comparison, the cooling group for a diesel engine runs with an input temperature of up to 120°C or 80°C above ambient. Thus the cooling group associated with a fuel cell powered train would be about 2·5 times the size and mass of that on a diesel train of similar power.

Impact on CO2 emissions

Emissions per passenger-km depend crucially on how electricity is generated and assumptions made about load factors. For example, in France, where most electricity is generated by nuclear power, the CO2 emissions attributable to trains are much lower than for cars even if the trains operate at speeds well in excess of 300 km/h. Most countries, however, depend on fossil fuels for power generation.

A direct comparison of fuel cells and electrification can assume that the electricity is generated in the same power station, and that the traction system does not affect the number of people who might wish to travel. The calculation then comes down to the relative efficiencies of the transmission chains.

Using figures for the hydrogen cycle presented by Professor Keith Ross3, the approximate efficiencies using current technology are in Table I.

These figures are indicative, and there will almost certainly be efficiency improvements in both fuel cells and hydrogen production. However, the table shows that there is a long way to go. At best, one can envisage the fuel cell route being only half as good as electric traction in terms of energy efficiency, and therefore CO2 emissions at whatever mix of power stations are supplying the national grid.

A major study by E4tech and others4 concluded that CO2 emissions are minimised by using renewable energy to supply the grid, and continuing to use oil products for transport, rather than using the renewable sources to generate hydrogen for transport. If so, there seems to be no environmental argument for the use of fuel cells for rail transport.


Huge sums of money are being poured into the development of fuel cells, and the technology is expected to progress rapidly. However, we have to question whether the fuel cell is an appropriate technology for the rail industry.

The US agenda is to decouple the personal mobility of the motorist from the vagaries of long-term supplies of oil. This does not apply to the rail industry as there is a well-proven technical solution to supplying energy from non-oil energy supplies: electrification.

From the point of view of inter-city train design, adopting fuel cells rather than overhead electrification seems likely to reduce the overall space available for passengers by about 10%.

Although it is impossible to estimate accurately the costs of a new technology when fundamental development has only just started, all indications are that the life-cycle costs per passenger of a fuel cell train will be significantly higher than for a conventional electrified railway.

From the environmental point of view, using fuel cells for inter-city traction looks like a disaster. It would greatly increase the primary energy consumption and hence CO2 production, compared with electrification, and hence would reduce one of the justifications for public investment in the rail network.

Table I. Comparative efficiency from power station to train drive

Hydrogen power

Electrolysis 50%
Storage 75%
Fuel cell 65%
Total 24%

Electric traction

Grid 94%
25 kV distribution 92%
Train transformer 93%
Total 80%


1. Moving to a hydrogen economy: dreams and realities. International Energy Agency, OECD document IEA/SLT(2003)5, January 2003

2. Transport Statistics Great Britain, 2003

3. Ross K. Moving towards hydrogen. Paper at the Will the lights go out? conference, Lancaster University,December 2004

4. A strategic framework for hydrogen energy in the UK. Report by E4tech, Eoin Lees Energy and elementenergy. December 2004

  • CAPTION: Fuel cells are unlikely to prove as efficient in railway applications as main line electrification, argues Roger Kemp, using the power requirements of a GNER IC225 high speed trainset as a case study
  • Fig 1. Transport's share of UK energy use has risen steadily over the past 25 years
  • Fig 2. CO2 emissions from transport in the UK have been increasing since 1990 as other sources decline Source: Defra
  • CAPTION: Fig 3. Basic operating principles of a Proton Exchange Membrane fuel cell as being developed by Johnson Matthey in the USA
  • CAPTION: Fuel cell research in the transport sector is being led by the automotive industry, with 'zero-emission' buses in several European cities. InSeptember 2004 Transport for London, First Group, BP and DaimlerChrysler put three fuel cell buses into service on route RV1 in a trial due to run until December 2005

Hydrogen offers no alternative to main line electrification

Interest in fuel cells to power trains has been growing, with Japan's Railway Technical Research Institute planning a prototype by 2010. Political leaders talk of the hydrogen economy eliminating local pollution and mitigating climate change, and prototype buses and cars already exist. Within the rail industry, it has been seriously suggested that hydrogen could avoid costs associated with electrification, but Professor Roger Kemp argues that exploiting rail's unique ability to use electric traction directly meets environmental objectives more effectively.

L'hydrogène n'offre pas d'alternative à l'électrification des grandes lignes

L'intérêt pour les piles à combustible afin de fournir l'énergie nécessaire aux trains va croissant, avec l'institut ferroviaire de recherche technique au Japon qui prévoit un prototype pour d'ici 2010. Les leaders politiques parlent d'une économie basée sur l'hydrogène, éliminant la pollution locale et atténuant le changement climatique, puisque des autobus et des voitures prototypes fonctionnent déjà. Au sein de l'industrie ferroviaire, il a été suggéré sérieusement que l'hydrogène pourrait éviter les co–ts associés à l'électrification, mais le Professeur Roger Kemp soutient que tirer parti de la capacité unique du rail à utiliser directement la traction électrique permet d'atteindre les objectifs environnementaux plus efficacement

Wasserstoff ist keine Alternative zu Elektrifizierung

Das Interesse in Brennstoffzellen als Antrieb für Züge wächst laufend, und das Japanische Eisenbahn-Forschungsinstitut plant einen Prototypen bis 2010. Politiker sagen, dass die Wasserstoff-Wirtschaft lokale Luftverschmutzung eliminiert und den Klimawandel bremst. Prototypen von Bussen und Automobilen existieren bereits. Innerhalb der Bahnindustrie wurde ernsthaft vorgeschlagen, dass Wasserstoff die mit einer Elektrifikation verbunden Kosten vermeiden könne, aber Professor Roger Kemp stellt fest, dass die für die Bahn einmalige Möglichkeit, elektrische Traktion direkt zu nutzen, die Umwelts-Ziele wirksamer erreicht

El hidrógeno no ofrece ninguna alternativa a la electrificación de las líneas principales

El interés en las pilas de combustible para los trenes ha ido creciendo, con el plan del Instituto Ferroviario de Investigación Técnica en Japón de construir un prototipo para el año 2010. Los líderes políticos dicen que la economía del hidrógeno eliminaría la polución local y mitigaría el cambio clim tico, y ya existen prototipos de autobuses y coches. En la industria ferroviaria, se ha sugerido seriamente que el hidrógeno podría evitar los costes relacionados con la electrificación, sin embargo, el profesor Roger Kemp sostiene que explotar la capacidad única del ferrocarril para utilizar la tracción eléctrica de forma directa cumple los objetivos medioambientales de forma m s eficaz