杏吧原创

Editorial: 鈥�Nuclear transmutation is a cool idea for a hot problem鈥�

AS THE locomotive and its long train of white cylinders grinds to halt, a small figure of a man atop one of the canisters raises his arms in a victory salute. He has brought to a standstill the infamous CASTOR train.

Similar protests accompany each of the train鈥檚 1000-kilometre journeys from the Normandy coast of France to the plains of northern Germany. The most recent was in November 2011. Besides peaceful demonstrations, overhead wires have been brought down, rail lines blocked and signalling installations sabotaged.

CASTOR stands for 鈥淐ask for Storage and Transport of Radioactive Material鈥�. In each of the white cylinders is high-level nuclear waste generated years ago by German reactors. It is returning from the French reprocessing plant at , where unreacted uranium and plutonium have been extracted to be recycled into fuel. The rest, including isotopes that will stay dangerously radioactive for many thousands of years, is heading for an 鈥渋nterim storage facility鈥�, a disused salt mine at Gorleben on the banks of the river Elbe.

The protestors argue that transporting such hazardous material in this way is too risky, as is keeping it in temporary storage with no vision for its future. They might have a point.

Is there another option? Possibly, if the work of the scientists and engineers who are beginning to test ways of 鈥渢ransmuting鈥� highly troublesome radioactive materials into other, less intractable elements lives up to its promise. It sounds a little like the ancient dream of turning lead into gold. But whereas the alchemists of old had the mythical philosopher鈥檚 stone, the transmuters of today are putting their faith in particle accelerators. Will they have any more success?

Highly radioactive nuclear waste is the not-so-secret filthy secret of the nuclear industry. Germany isn鈥檛 alone with the problem: global stockpiles grow by more than 10,000 tonnes each year. In the US, nearly 65,000 tonnes or about 26,000 cubic metres of spent fuel sits in interim facilities at 75 sites across the country while political wrangling about its ultimate fate continues. That contentious legacy is expected to double by 2050 鈥� and is dwarfed by high-level waste left over from America鈥檚 nuclear weapons programme (see 鈥淭oxic legacy鈥�).

Nuclear waste owes its existence to the activities of neutrons, the subatomic particles found in the nucleus of every element except hydrogen. Uranium-235 atoms, the active ingredient in nuclear fuel, have 143 neutrons, too many to be completely stable. Over time, a lump of uranium-235 slowly undergoes its own transmutation: individual atoms split in two in the process called fission, shedding a couple of neutrons and a lot of energy as they do so. If one of these neutrons strikes a neighbouring uranium-235 atom, it too can fall apart, producing more neutrons, and so on. The result is a self-sustaining, energy-giving chain reaction.

Intractable nasties

The problems start when fission doesn鈥檛 happen. About 95 per cent of spent nuclear fuel is still in the form of uranium 鈥� primarily uranium-238, a non-fissionable but radioactive isotope that dominates mined ore. Uranium can be extracted and reprocessed into fuel at facilities like La Hague, but this is an expensive business. Freshly mined uranium is much cheaper, so most countries leave the spent fuel as it is.

And reprocessed or not, spent fuel contains other, less tractable, nasties. Occasionally when a uranium atom is hit by a neutron within a reactor, it simply absorbs it. That can happen to any one atom several times, and a series of heavier elements results, including plutonium, americium and neptunium. These 鈥渉eavy actinides鈥� are the real sting in the nuclear tail. Their typical half-lives of thousands of years mean they will remain dangerously radioactive for tens or even hundreds of thousands of years, long after most of the other components of nuclear waste have decayed away (see 鈥淣o quick fix鈥�).

But if neutrons produce this high-level waste, they also provide a way to clean it up. If a neutron strikes a heavy actinide atom hard enough, sometimes that atom can fission too. The resulting lighter elements are, as a rule, much less of a problem. 鈥淭he products of fission are many and varied,鈥� says , a nuclear engineer at the University of Cambridge. 鈥淎lmost all are radioactive but their half-lives are, in general, orders of magnitude shorter.鈥� The waste consists for the most part of isotopes such as krypton-85 or caesium-137, with half-lives of 11 and 30 years, respectively. They need to be stored for a few hundred years before their activity is low enough to be safe 鈥� still a problem, but not a burden on countless future generations. This is the promise of transmutation.

If all it takes is neutrons to kick-start things, it is tempting to think transmutation might happen of its own accord: there are plenty of neutrons flying about in a nuclear reactor, after all. But neutrons released by fissioning uranium-235 are slowed in a conventional reactor until they have an energy of 0.025 electronvolts. That is just right for splitting more uranium, but eight orders of magnitude below the few megaelectronvolts needed to split a heavy actinide. Previous efforts to build 鈥渇ast-neutron鈥� reactors capable of eating up actinides have stalled on grounds of cost and safety.

This is where particle accelerators come in. By flinging their products 鈥� high-speed protons 鈥� into a lead target, a stream of high-energy neutrons can be ejected. Channel these into a reactor containing spent nuclear fuel and they can break down the heavy actinides.

That鈥檚 the theory, at least, and it isn鈥檛 new: particle physicist and Nobel laureate was pushing the idea of such accelerator-driven systems (ADS) 20 years ago. But the technology has lagged behind. Reliably producing neutrons with energies high enough to blow apart heavy actinides is a problem. Today鈥檚 high-power proton accelerators produce particles in bursts, and are not designed with reliability as their primary goal. They tend to 鈥渢rip鈥� or lose their beam regularly: no great problem in a research accelerator, but not a good idea for keeping a transmuting reactor continuously and stably supplied with neutrons. 鈥淟osing the beam in an ADS means that your reactor will shut down and start to cool,鈥� says Hamid A茂t Abderrahim of the , in Mol. This could potentially cause dangerous stresses in the materials containing the reactor.

A茂t Abderrahim heads up a project to show that such problems can be overcome. MYRRHA 鈥� the 鈥� will be the first large-scale test of the ADS concept. Funded by the European Union, construction is due to start in 2015 and the reactor should be operational by 2023. In the meantime, since January this year, the researchers have had Guinevere to play with 鈥� the world鈥檚 first demonstration accelerator-driven lead-core reactor. 鈥淕uinevere is a baby MYRRHA operating at a smaller power,鈥� says A茂t Abderrahim. His team will use it to test the principle and show that they can measure and control neutron levels in the reactor core.

Back to the future

In a bid to get around the tripping problem, MYRRHA will not steer protons around a circle, as is the case with a machine such as the Large Hadron Collider at CERN near Geneva, Switzerland. Instead, it will accelerate them in a straight line. This is a simpler and more reliable design, but it means the accelerator must be much bigger, and more expensive, for the particles to reach the necessary speed: the accelerator will be several hundred metres long, compared with a radius of 10 to 20 metres for an equivalent circular machine.

That is not going to win over those who think nuclear power is already a hideously expensive white elephant. But MYRRHA has also been designed to test out an idea first floated by Rubbia in the 1990s: that accelerator-driven systems might themselves be used to generate electricity. To do that, you need to mix the actinide waste with a fissile fuel. Using uranium-235 would be self-defeating 鈥� that would just produce more of the heavy actinides you are trying to get rid of. But the picture changes when you consider an alternative fuel: thorium.

Thorium is a nuclear fuel with a long list of potential advantages over uranium. It is three to four times more abundant, and all of it can be used as a fuel, whereas natural uranium deposits contain only 0.7 per cent uranium-235. Thorium was extensively used in early prototype fission reactors. When nuclear power really took off in the 1970s, however, new deposits of uranium were being discovered all the time, and thorium reactors suffered from a further disadvantage: unlike uranium reactors, they don鈥檛 generate much plutonium, the raw material for nuclear bombs. 鈥淚 am convinced that uranium won out because there is no military application of thorium,鈥� says , a particle physicist at the University of Huddersfield, UK, who researches thorium as a fuel. 鈥淣uclear power and nuclear weapons were developed hand in glove.鈥�

Thorium is not itself fissile. Its atoms first soak up neutrons to form uranium-233, which is fissile and falls apart in a burst of energy when the next neutron hits. Sustaining this two-step process requires more neutrons than are generated, so an outside neutron source is needed 鈥� exactly what an accelerator would supply. And although running an accelerator requires an awful lot of electricity, that energy demand would be dwarfed by the reactor鈥檚 output: you would only need a 20 megawatt accelerator to generate 600 MW, says Parks, and potentially one even smaller than that. The set-up has another benefit too: the fission reaction can be switched on and off at the flick of a switch. 鈥淭he chain reaction would die out if it wasn鈥檛 for the accelerator producing neutrons, which means that a Chernobyl-style accident is impossible,鈥� says Parks.

鈥淩unning an accelerator requires a lot of electricity, but that would be dwarfed by the reactor鈥檚 output鈥�

The crucial point, though, is that thorium atoms have a smaller number of neutrons 鈥� 142 in thorium-232 against 146 in uranium-238 鈥� and that makes a huge difference to the waste it produces. A thorium atom has to capture more neutrons to make the troublesome heavy actinides, so the reactor makes less of them. 鈥淚t manages its own waste while it鈥檚 operating, but it鈥檚 got the capacity to deal with more than its own waste,鈥� says Parks. 鈥淪o you have a device that is simultaneously generating power, exploiting an available resource and getting rid of problem waste.鈥�

鈥淚t鈥檚 a rational idea,鈥� says A茂t Abderrahim. 鈥淭o go to producing energy, for me, makes sense.鈥� With MYRRHA, his team will test how well the technology works in practice.

Barlow and his colleagues in the , meanwhile, are working to reduce the size and therefore cost of any future facility. They are developing 鈥渇ixed-field alternating gradient鈥� machines that spin particles in a twisting motion as they are accelerated, counteracting the particles鈥� normally increasingly irregular behaviour as they approach very high energies. Last year, the team built a at the Daresbury Laboratory in Cheshire, UK. The next step, to build an ADS-spec device, will take five years to complete, says Barlow. Parks has high hopes. 鈥淎 few more years of accelerator development might bring the missing piece in the jigsaw,鈥� he says.

The latest work from Parks鈥檚 own group, though, suggests an intriguing twist: an accelerator might not be necessary after all. His student has calculated that if thorium and heavier actinide waste are mixed in the right proportions, enough neutrons should be generated from the decaying waste to drive thorium fission, maintaining energy generation and safety without the need for an external neutron source ().

There has never been any obstacle to that possibility. 鈥淚t鈥檚 more that nobody had thought of doing it,鈥� says Parks. He and Lindley came upon the idea while considering how to combine an accelerator with existing nuclear reactors. 鈥淚 started off with a very complicated-looking reactor, and essentially kept simplifying it,鈥� says Lindley. 鈥淚 found that each iteration kept working, in some cases better than the original complicated design, until I was left with a normal reactor.鈥�

Barlow still needs some convincing. He points out that the gradual conversion of thorium to uranium-233 within the fuel will change neutron production and consumption patterns over time. 鈥淚 think you need a tap labelled 鈥榥eutrons鈥� that you can adjust to keep the reactor viable,鈥� he says. 鈥淚f it turns out that you can run thorium-fuelled power stations without an accelerator, I鈥檒l be very pleased 鈥� and very surprised.鈥�

Even if any variant of a transmutation reactor can be shown to work at a reasonable cost, it will still have big hurdles to jump. A conventional nuclear reactor core is kept cool by pumping water through it, but collisions with water molecules slow neutrons down. To keep neutrons whizzing about with the energies needed to fission heavy actinides, a reactor needs heavier coolant molecules that neutrons ping off while retaining most of their energy. For MYRRHA, that coolant will be liquid lead: a nasty, corrosive material that is difficult to contain within a reactor core.

Establishing any energy-generating technology using thorium will also take time. The small amounts of the element currently produced as a by-product of mining valuable rare earth elements are enough to keep research reactors ticking over, but not to supply an industry. A whole new infrastructure would be needed, from mining to refining. That is not fundamentally difficult, says Parks 鈥� but it won鈥檛 come for free.

But then, nor will any solution to the problem of high-level nuclear waste. Imagine if, in a decade or so, the CASTOR trains begin to roll across Europe carrying their toxic cargo to a facility where it can be zapped out of existence. Would anyone be protesting?