By Martin Redfern
BBC radio science unit
The US space agency (Nasa) is progressing with ambitious plans to explore the Solar System using nuclear power.
Nuclear power would allow missions to orbit - not merely fly by
Their hope, eventually, is to use electricity generated by nuclear power to propel a space probe and power its instruments on a voyage to the icy moons of Jupiter, satellites that just possibly might harbour life beneath their ice.
Before then, nuclear technology could be proved with a less ambitious mission, perhaps a nuclear-powered probe to the Moon.
As long ago as 1907, just two years after Einstein discovered his famous equation E=mc2 which hinted at the vast power locked within the atom, Robert Goddard, who was himself to go on to pioneer chemical rockets, wrote: "The navigation of interplanetary space depends for its solution on the problem of atomic disintegration."
Once the power of the atom bomb had been demonstrated and the Cold War set in in the 1950s, all sorts of amazing proposals were developed for nuclear power in space.
Among them was project Orion, a plan to launch and propel spacecraft weighing thousands of tonnes and carrying dozens of passengers by detonating nuclear bombs behind a pusher plate.
Orion was cancelled when a nuclear test ban treaty came into force but another project, Nerva, to use a nuclear reactor to produce a rocket jet was the front runner for a possible human mission to Mars after the Apollo moon landings.
That, too, was scrapped and it was left to the Russians to launch several nuclear reactors into space to power spy satellites during the 1970s.
Dreams of nuclear power in space did not die with the collapse of the Soviet Union and restrictions on the Nasa budget.
Project Orion envisaged "pulsed" nuclear propulsion
Many space scientists agree that nuclear power is the only viable way of exploring the outer Solar System.
Using chemical rockets to move between planets and their moons is not really practical because of the fuel mass a spacecraft would have to carry with it, and relying on solar power to drive instruments is problematic because of the distance from the Sun.
Travel out to Mars and there's only half as much solar energy as reaches the Earth.
At Saturn, it is a hundredth of the power we are familiar with. Arrays of solar cells to produce really useful power there would be impossibly big.
So far, probes to the outer planets, such as the Cassini craft now orbiting the ringed planet, have used Radioisotope Thermal Generators (RTGs) - solid state electrical generators powered by the heat of radioactive decay. But that power is limited.
The RTGs on Cassini would not produce enough to run a hair dryer (three units produce about 700 watts).
Scientists would love more electrical power for their instruments - but there is another use for which nuclear electric power could make all the difference: the ion engine.
By using electrical energy to ionise atoms and accelerate them in a jet, it is possible to propel them at more than 10 times the speed a chemical rocket jet can manage.
That means that you need less than a tenth of the propellant to travel a certain distance.
Already the European Space Agency has used an ion engine to take its Smart 1 craft to the Moon and an ion engine for a planned mission to Mercury is now under test at the UK's QinetiQ's labs in Hampshire. Both use solar power, but for deep space that is not enough.
CURRENT ION PROPULSION
Uses electrical power provided by solar panels to accelerate a propellant to high velocity
Smart 1 uses the propellant xenon, a colourless gas
Electrons trapped inside a chamber by a magnetic field collide with xenon gas, creating xenon ions and more electrons
The resulting ion beam pushes the space craft forward
The thrust produced is the same as the pressure exerted by a sheet of paper held in the palm of a hand
Over long periods, it can make a spacecraft travel faster
In a Prometheus model, the electricity would be generated from a nuclear power pack
Project Prometheus proposes using a nuclear reactor not much bigger than a dustbin, linked to a turbine or other generator to provide perhaps 250 kW of power.
Mention that to space scientists and their eyes glaze over with dreams of the instruments they could run and the multiple destinations they could visit.
An ion drive might be slow - it produces a thrust little more than a human breath - but it can keep it up for years on end and have enough puff left over to flit between multiple destinations.
Spare power, for example, could run radar to look for oceans beneath the ice of Jupiter's moon Europa. It could also power a communications system that would replace a trickle of data with a broadband flood of pictures.
Critics such as Bruce Gagnon of the Global Network Against Nuclear Power In Space talk of the risks of making, launching and using nuclear reactors.
The first project manager of Prometheus, Alan Newhouse, counters: "A reactor would be launched shut down and never having operated, so there would be only a very small amount of radioactivity involved. So, as a potential danger to Earth, it's not there."
But the engineering challenges are immense. No one has ever made a nuclear reactor that could run for many years without human intervention.
Reactors on Earth are all near convenient cooling systems. In space, they would need large areas of radiator, perhaps not as large as solar arrays but still substantial.
Reactors for Prometheus are being developed by a US naval laboratory that makes reactors for submarines.
It is possible they may also learn from Russian reactors designed for space which were sold to the US after the break-up of the Soviet Union and of which little has been heard since. The US National Academy of Sciences has declared a mission to Europa to be the highest priority for space science in the next decade and the first Prometheus mission was due to be Jimo, the Jupiter Icy Moons Orbiter.
But, explains Prometheus project manager John Casani, "people are saying that the Jimo mission is so important, so high profile, maybe we should take a baby step before we take that giant step.
"So, we are looking at a mission that would be less technically challenging. But Jimo is the crown jewel of the suite of missions we've been looking at."
The Nasa budget for 2006 includes funding to develop a test mission for Prometheus, possibly to the Moon. That might be ready in 10 years' time and, says John Casani, the hope would be to launch a new mission every two years thereafter.
They might include Jimo and then a mission to Saturn's moon Titan, glimpsed by the Huygens probe earlier this year.
There might be a mission to Neptune's moon Triton, even further from the Sun, and the power of Prometheus might be used to support unmanned rovers and possibly manned missions to the Moon and Mars.
John Casani says: "My own favourite mission would be a kind of nuclear tug boat to an asteroid."
Sooner or later, many astronomers believe that an asteroid will be spotted on a collision course with the Earth.
One day, another big one will hit the Earth
Given several years' notice, a nuclear-powered ion-drive rocket could use its gentle thrust to push the whole asteroid into a safe orbit without the need or risk posed by blowing the giant rock apart.
There will doubtless be protests and opposition to the use of nuclear power in space. Its value for deep space probes is undeniable, but critics fear that that could open a back door to nuclear power in space for the military as well.
Leo Enright investigates Project Prometheus at 21PM GMT on BBC Radio 4 on Wednesday 9 March. The Radio 4 website will retain a recording of the programme after transmission.