Building a second sun

Building a second sun: Take $10 billion, add coconuts




THE balmy south of France has always been a magnet for sun worshippers. So it is perhaps fitting that here, not far from the Côte d'Azur, an international team of researchers is building a machine to recreate the sun. It will take tens of thousands of tonnes of steel and concrete, plus a whole host of more unusual materials: beryllium, niobium, titanium and tungsten; frigid liquid nitrogen and helium. Oh, and a supply of burnt coconuts.

This eclectic mix of ingredients will be turned into ITER, the International Thermonuclear Experimental Reactor - the next big thing in nuclear fusion research. When completed in 2018, the reactor will fuse together two heavy isotopes of hydrogen to release vast quantities of energy. In theory, the result will be clean electricity galore with no carbon emissions and far less radioactive waste than today's nuclear fission reactors leave behind.
So why we are not already flooding our electricity grids with fusion energy? While the concept of nuclear fusion is simple, the practicalities are anything but. That's because the nuclei themselves are reluctant participants: each
carries a positive electrical charge and these repel one another, so forcing two nuclei together is almost impossible. Only at stupendously high temperatures do the nuclei acquire enough energy to overcome their mutual aversion, smash into one another, and fuse.



It is much the same picture in the sun. There, heat is generated from the fusion of hydrogen nuclei. But the fuel barely smoulders even at 15 million kelvin, the temperature of the sun's core. It is consumed so slowly that the supply lasts for billions of years.
At a fusion power plant, the fuel needs to be burned on human, not cosmological, timescales. The heavier isotopes deuterium and tritium are a little easier to burn than ordinary hydrogen, but even so, to get a good blaze going inside ITER the temperature will have to be racked up to a hellish 150 million kelvin. That brings a mountain of engineering problems. Not least is how to contain a plasma of electrons and atomic nuclei that is 10 times as hot as the sun's core.
Even the most hardy of construction materials cannot withstand temperatures of more than a few thousand kelvin. So the solution is to weave a cage for the plasma from magnetic fields.
ITER follows the design of several smaller experimental reactors where physicists have already achieved the temperatures required for fusion. The nuclear fuel is held inside a ring-shaped reactor called a tokamak.
Magnets outside the ITER ring combine to generate a spiralling field that holds the superhot plasma in place. To make its magnetic cage, ITER will use superconducting coils of niobium-alloy wire weighing a total of 10,000 tonnes and cooled by a supply of liquid helium.
Trapped in its cage, the fusion fuel is simultaneously cooked in three different ways. While electrical circuits force a current through the plasma, it is blasted with microwaves and bombarded by high-energy atoms generated by small particle accelerators dotted around the ring. Even under this triple-pronged attack, so far no tokamak has yielded much fusion energy. ITER should do better by firing up a much bigger, denser ring of plasma. A lot of power will have to be pumped in to start the plasma sizzling, but if all goes to plan, 10 times as much will emerge.
All that power poses a threat to ITER because the magnetic cage is not impregnable. The violet-hot plasma will radiate X-rays, a trickle of charged particles will always escape, and the fusion reaction will create high-energy neutrons, which are electrically neutral and can't be contained by magnetism. So despite the magnetic cage, ITER's plasma will blast the surrounding walls with several megawatts of heat per square metre, far more than in previous tokamaks or conventional nuclear fission reactors.
The solution is simple: use water to carry the heat away. "Of course this is exactly what we want from fusion in the end - to extract the heat," says Mario Merola, head of the ITER division responsible for internal components of the reactor.
Once again, it's the practicalities that are the problem. The main reactor wall, known as the blanket, will be made from 440 stainless steel blocks nearly half a metre thick and riddled with high-pressure water pipes. This steel blanket should absorb most of the neutrons, which will heat the blanket from within. Near the inner wall, the water pipes can be no more than 2.5 centimetres apart, otherwise the steel between them would become dangerously warm and soft.
For the innermost surface facing the plasma, steel is no good. Incoming plasma particles would chip iron atoms out of the steel and back into the chamber, where they would pollute the fuel and damp down the fusion reactions. So the ITER team has chosen to face the wall with tiles made of beryllium. While beryllium is toxic to humans, it is quite palatable to the plasma because it is such a light element, close in atomic weight to deuterium and tritium. So although some beryllium will get blasted off the walls, it won't quench the reactor's fire.

The steel and beryllium plates will also be battered by mechanical forces generated by the interaction of the electric currents and magnetic fields passing through them. Each 4-tonne plate will experience forces of up to the weight of 100 tonnes, so they will have to be firmly locked in place - and sturdily built, even though they are punctured with holes for the pipes. "The design of the blanket modules is one of the most technically challenging parts of the whole machine," says Merola.
A different kind of armour plate is needed around the bottom of the chamber. Here, a device called the divertor is used to keep the plasma pure. The main by-product of the fusion reaction is helium nuclei, which would eventually build up and stifle the nuclear fire. The divertor's job is to skim off the outermost layer of plasma, which can then be cooled and siphoned off to have the helium "ash" and other impurities removed. The surface of the divertor will get hot enough to melt beryllium, so it will be covered in tungsten and carbon fibre, both materials with melting points above 3000 kelvin.
Armour plates
Wearing this water-cooled suit of armour, ITER should be able to resist the steady heat given off by the ring of plasma, but that isn't all it has to contend with. The plasma inside a tokamak turns out to be similar to the sun in more ways than one: like its big brother this doughnut-shaped mini-star can suddenly and violently erupt in an event called an edge localisation mode, or ELM. In a fraction of a millisecond, the surface of the plasma ring balloons outwards, sending out an explosive burst of particles. "It looks like a solar flare," says ITER researcher Alberto Loarte.
"The problem is that the release is very intense and localised, so local power densities are very high - many gigawatts per square metre," Loarte explains. That is more than a million times the power density of sunlight hitting Earth. Even though an ELM is very brief, this flash of energy is still enough to vaporise a thin layer of beryllium, tungsten or carbon. As ELMs are expected at a rate of a few per second, the vital armour would soon be destroyed.
However, the ITER team has a plan to fight these turbulent fires: throw ice cubes at them. The technique was developed during the 1990s at a reactor called ASDEX in Garching near Munich in Germany. Like other tokamaks, ASDEX has to keep its plasma topped up with fuel, and to do this it is equipped with a gas-powered peashooter that fires pellets of frozen deuterium into the plasma chamber.
The ASDEX team discovered that the sudden puff of gas released when a pellet enters the plasma can trigger an ELM-like outburst. So it is possible to cut ELMs down to size by deciding where and when to aim your pellets. "You can decide to trigger these events frequently, making very many small ELMs," says Loarte. These smaller flares should be less damaging.
But this may not prove a perfect method of control: the ELMs might still eventually eat through ITER's inner armour, and a further line of defence could be needed. In 2006, at the DIII-D tokamak operated by General Atomics in San Diego, California, physicists discovered that they could prevent ELMs from happening altogether by using an array of small magnetic coils inside the reactor. Tucked in just behind the protective blanket wall, these coils produce small magnetic fields that perturb the surface of the plasma and somehow prevent the ELMs bursting out. "The theoretical basis for this is not yet clearly understood," says Loarte. Both methods have been developed further at the JET fusion reactor near Oxford, UK, but the ultimate test will be at ITER to find out if they can control ELMs without releasing too much plasma and shutting down the fusion reaction.
A more insidious hazard is posed by neutrons. As well as heating up the fabric of the reactor, high-energy neutrons from the fusion core will damage the crystal structure of anything they hit, potentially making a strong metal brittle and weak. The bombardment of neutrons generated by ITER will be more intense than anything ever seen on Earth - so will the reactor crumble away? Merola is confident that it won't. The blanket will be made from austenitic stainless steel - a type of steel also used in domestic cutlery - which has a resilient crystal structure that keeps its strength even if many atoms are knocked out of place. "Austenitic steel is very forgiving," says Merola.
Fusion remains a controversial goal, not least because of the expense of the research still required. ITER alone will cost more than $10 billion. Sceptics also like to point out that ever since the idea was first touted in the 1950s, fusion's promise of clean power has receded endlessly into the future.
The ITER team are now hoping to drag it closer to the present. If they can successfully hold a slice of the sun at the reactor's heart, we might finally be on the verge of getting usable energy out of this electric dream.

2002 was an excellent year for coconuts. It's strange how you need this very natural material to make a fusion device

Totally tropical tokamak
Is your fusion reactor getting clogged with garbage? Try using some coconut. It's not the most obvious ingredient for such a high-tech device, but it will play a vital part in the ITER reactor in France.
The coconuts will be used to generate 10,000 cubic metres of nothing - the vacuum essential to ITER's operation. Some of this vacuum, in the central chamber, separates the plasma from the surrounding solid walls and allows fusion to proceed unhindered by air molecules. A whole lot more nothing is needed to fill the vacuum jackets that insulate ITER's supercooled magnets.
The vacuum pumps that will suck air out of ITER will also need to hoover up the waste helium from the fusion reaction, along with other debris created when hot plasma smashes into the reactor wall. "This can only be done with very large cryogenic pumps," says Christian Day of the Karlsruhe Institute of Technology in Germany, who is building ITER's vacuum pumps.
A cryopump works by capturing gas molecules on a cold surface. The pumps in ITER will have two stages, chilled to 80 and 5 kelvin respectively. While most gases will freeze on contact at these temperatures, the helium and hydrogen isotopes emerging from ITER are less easily caught. They can only be captured by adsorption, a process that involves atoms of the gas sticking loosely to a solid surface - and the greater the surface area, the better. "We wanted a material that behaves like a sponge, with lots of internal surfaces," says Day.
He and his team have spent 20 years searching for the ideal adsorber. Their quest was wide-ranging, taking in sintered metals and porous minerals called zeolites, but eventually they found that the material with the greatest capacity to adsorb gases was charcoal. They tried charcoal made from industrial polymers, from different varieties of wood, from peat, and from textiles such as wool and cotton, and more.
Once source stood out. "We found that coconut-shell charcoal is the best," Day says. "It is somehow strange that you need this very natural material to make a fusion device."
To make coconut charcoal, you first char the shells in a low-oxygen atmosphere, then wash them in acid and bathe them in steam to clean out a network of fine pores. The walls of these pores form a huge internal surface - about 1200 square metres per gram of charcoal. "We found the decisive point is pore size distribution," says Day. "We need pores of all types, because different pore sizes work best for different gases."
The provenance of the coconut matters too. The best fusion coconut turned out to come from one particular Indonesian island, Day found, and the quality of the resulting charcoal even depends on the year. "2002 was an excellent year for coconuts," he says.
So the team bought up most of the vintage 2002 Indonesian coconut-shell charcoal. "We have a few tonnes in a garage in Karlsruhe," Day says. That will be enough to supply ITER's cryopumps a few times over, so there should be a surplus - perhaps destined for the first commercial fusion power stations some decades from now. Unless 2035 is an even better year for coconuts.