Can we trust hydrogen again?

Despite a safety record unmatched in transport history, with 50,000 passengers carried without a fatality, the use of hydrogen in air passenger transport came to a spectacular end on 6 May 1937 in the half-minute it took for the airship Hindenburg to become a pile of smoking debris.

Hydrogen was first used as a means of getting airborne more than 200 years ago, with the first manned flight of a hydrogen balloon taking place only 10 days after the Montgolfiers' first flight.

Now, nearly 60 years after the Hindenburg disaster, hydrogen is making a comeback to civil aviation - but as a fuel rather than for lighter-than- air flight.

Engineers from Daimler-Benz Aerospace, with German and Russian research partners, are carrying out work that will lead to the "Cryoplane", a commercially viable aircraft powered by liquid hydrogen (LH2), based on the familiar Airbus.

As an aviation fuel, hydrogen actually has a lot to commend it. Nuclear power is too dangerous, electrical propulsion requires heavyweight equipment, and natural alcohols have a comparatively low energy/ mass ratio. Hydrogen, on the other hand, can be generated from water, and when liquefied and stored cryogenically below its boiling point of -253C has nearly three times the energy/ mass ratio of kerosene - the principal component of jet fuel.

Back in 1956, the US NACA-Lewis Flight Propulsion Laboratory modified a B57 bomber for flight using liquid hydrogen in a programme believed to have been driven by the need for extremely long range, high-altitude flights by reconnaissance aircraft. That job has since been taken over by satellites. The availability of cheap fossil fuels and a lack of appreciation for its environmental benefits put the work on a back burner.

The 1990s are different. Economical, accessible reserves of crude oil are expected to run out between 2020 and 2040. Though conventional transport fuels can be produced from other fossil sources such as tar, shale and coal, this involves higher production costs that will make alternative technologies commercially attractive. Kerosene's price will escalate; that of liquid hydrogen will fall, relatively. And although only 3 per cent of world carbon dioxide emissions come from aviation, the Rio agreement to use taxation and regulation to reduce such emissions will affect it as much as other industries.

"Burning kerosene produces both water and a lot of greenhouse carbon dioxide - gas that can stay around in the atmosphere for 100 years," says Dr Heinz Klug of Daimler-Benz. "The burning of LH2 produces no carbon dioxide, but a lot of water vapour."

That is also a greenhouse gas, whose effects vary with altitude. Near the ground, water vapour dissipates within days, but in the stratosphere it can persist for six months. Jet vapour trails composed of ice crystals also contribute to the greenhouse effect; the water from burning LH2 might produce more of these than kerosene. However, ice particles from LH2 exhaust will be bigger, supposedly reducing their effect. Dr Klug says: "A test aircraft will prove the point, but we can avoid the effect - at a small fuel consumption penalty - by reducing slightly the cruising altitude of LH2-powered airliners."

Liquid hydrogen combustion emits none of the secondary products associated with kerosene except nitrogen oxides, but research suggests that emission of even these by-products can be reduced to one-third - and perhaps as much as one-tenth - of those from today's jet engines.

So liquid hydrogen may be an effective fuel. But anyone who has seen the film footage of the Hindenburg disaster will certainly question its safety.

During a fuel spillage, LH2, unlike kerosene, vaporises rapidly and rises away from the spillage site, burning upwards - avoiding the horrendous "fire carpets" that typify kerosene fires. Nor does it contaminate soil as kerosene does. Even if it does catch fire, leaking hydrogen burns rapidly but tends not to explode; an explosive mixture can be formed only in a confined space. Paradoxically, the Hindenburg proves the point: it burned, but didn't explode.

Surprisingly, perhaps, almost no heat is given off by hydrogen while burning, so the aluminium fuselage of a modern airliner should be enough to withstand a fuel fire and protect the passengers within.

Those safety benefits aside, the commercial use of LH2 in aviation will place tremendous demands on technology for production, storage and distribution. Liquid hydrogen boils at -253C, only 20C above absolute zero, requiring effective insulation. It requires four times the tank capacity of kerosene, posing design problems for fuel storage both on the ground and in the air.

In a conventional aircraft, fuel is stored in the wings. This keeps the fuel near the centre of gravity of the aircraft, and avoids large changes in balance as the fuel is used up. But liquid hydrogen fuel needs to be kept cold, so requires more insulation than can fit into wing space, and larger tanks to hold the greater volume needed.

Daimler-Benz's solution is to carry the fuel "piggy-back" fashion above the main cabin - giving the proposed Cryoplane an unusual but characteristic appearance.

The potential scale of the project to move to the use of LH2 is enormous. The first stage - converting domestic European wide-bodied aircraft - would involve some 500 aircraft and 70 airports. It also requires a quantum leap in technology to increase LH2 production from the present 20 tons per day to 6,000 tons. With current technology, that would use the electrical output of 10 large power stations, but experiments are being carried out as part of the "Euro-Quebec Hydro Hydrogen Pilot Project", jointly funded by the European Union and Quebec Province, to use Quebec's surplus hydroelectricity to produce liquid hydrogen, which would then be shipped to Europe in insulated tanks.

But when will this happen? A Tupolev-155 aircraft has been operating since 1988 as a flying test-bed with a starboard engine that can use kerosene, natural gas or hydrogen. Product cycles within aeronautics are long: it can take 50 years from developing a new aircraft to phasing out the last of the series, and a radical development such as a new fuel could further prolong that process.

But assuming that development targets are met, Daimler-Benz anticipates that the first series-produced LH2 Cryoplane will be a small, regional aircraft, and is working on a demonstrator programme. The intention is to modify a 30-seat Dornier Do-328 for the year 2000, with a series version entering service in 2005. Whether it will shake off those newsreel images of the Hindenburg is another matter.