In the midst of such pessimism, it is comforting to find that technologies in the pipeline could cure our oil addiction.

The irritating thing about oil, apart from the pollution it causes, is that, in most other ways, it is an ideal fuel. Petrol, diesel and kerosene are all easy and safe to store and transport. They are also energy dense -- a small amount of fuel yields a lot of energy. These two properties make oil perfect to power vehicles of all kinds, and at the moment there is no practical alternative.

So when the oil does finally run out, what's going to happen to all the planes, trains and automobiles?

Many experts believe that the rising price of oil will drive a revolution in energy production and use, ushering in the so-called hydrogen economy. In this new world, energy will be harvested from renewable sources such as wind, solar and hydroelectricity, and used to split ordinary water into its constituents, oxygen and hydrogen. Later, the stored hydrogen will be recombined with oxygen from the air in a device called a fuel cell, which generates electricity and produces water as its only waste.

This technology is already available, though in its infancy. The hydrogen economy is a green dream, but without some significant technological developments it could easily turn nightmarish.

The problem is that hydrogen stored the conventional way -- compressed or liquefied -- is neither energy dense nor easily stored; not too much of a problem for supplying the national grid, but a serious challenge for powering a vehicle.

Carrying compressed or liquefied hydrogen also risks catastrophe in the event of an accident, since hydrogen is extremely inflammable. In fact, a naively designed hydrogen car would probably be even less useful than a "conventional" electric car powered by ordinary batteries.

So what engineers need is a way to store lots of hydrogen in a small space, safely. The alternatives to compression or liquefaction use another chemical to bind hydrogen molecules.

The best available binders at the moment are certain metals and alloys which absorb hydrogen to form a metal hydride, releasing some heat in the process. The hydrogen is later liberated by reheating the metal hydride. These tanks can only store about 2 per cent of their own weight in hydrogen, but, critically, they occupy less than a quarter of the space of the same weight of compressed hydrogen. Also, the gas will not escape if the tank is damaged.

Metal hydrides are the best storage option at the moment, but not good enough to take over from oil. Work is under way on other binders, the most promising of which are carbon nanotubes and nanofibres, which some predict could compactly store more than two-thirds of their own weight in hydrogen. If this speculation proves justified, fuel- cell cars of the future could run for thousands of kilometres between refills.

Not only are experimental hydrogen plants and vehicles operating around the world, but, as portable devices like cellphones and laptop computers become increasingly power-hungry, electronics manufacturers are looking to dump batteries and use tiny fuel cells here, too. All this activity should bring about the intense research needed to improve the efficiency of fuel-cell technology so that when the oil runs out we aren't left without a ride.

Hayden Walles is a science graduate and writer.