The countryside of Saint-Paul-lez-Durance in Provence is a serene terrain of thickly wooded hills. On chilly January mornings, the air becomes thick with mist and the sky glows red as the sun pokes up above the horizon at dawn. By mid-morning, that haze is usually gone, leaving behind a bright blue sky with only the faintest wisp of high-altitude cloud.
As picture-postcard scenes go, it is as rural and peaceful as it gets. But incongruously nestled among these hills and vineyards, the most sophisticated, expensive machine ever built is slowly taking shape at the local Cadarache nuclear facility. It is a scientific collaboration on a worldwide scale, meant to tackle one of the biggest challenges of the 21st century – with the human population growing every year, how do we continue to make ever more electricity past 2050 (the date that the EU has set for full decarbonisation of power generation) without destroying the environment? The scientists and engineers in Saint-Paul-lez-Durance think the solution is nuclear fusion – they want to recreate a star in a box on Earth.
Everything about the project, known as Iter (formerly known as the International Thermonuclear Experimental Reactor), is huge. The main fusion reactor will be built on a flattened area of concrete that has been blasted into the hills at Cadarache and stretches to 60 football pitches. Around 2.5m cubic metres of earth and rubble were excavated from what was originally a small valley that undulated by several hundred metres in parts. That concrete baseplate sits on dozens of pillars containing layers of rubber sandwiched between the mortar and cement – not only do these pillars raise the building above the height of the surrounding countryside (the height was calculated to be above the maximum height that water would flow past if the nearby dam broke), they also create a “seismic isolation pit” that will protect the building from earthquakes.
At the centre of the concrete box where the main building will go, you can already see a circle of steel bars that trace the shape of what will become the ring-shaped vacuum vessel, where the fusion reactions will take place. Ready to haul in the huge components over the coming years, four giant cranes are rooted into the site, one of them within the circle itself. When the main building containing the reactor is complete, it will rise 60 metres into the air and reach 10 metres below the ground.
When the million or so pieces that make up the Iter machine have been delivered to site and are finally bolted and welded together, the whole thing will weigh around 23,000 tonnes, three times the weight of the Eiffel tower. The entire reactor complex – including the foundations and buildings that will sit in the seismic isolation pit – will weigh 400,000 tonnes, more than the weight of the Empire State Building.
Visiting the Iter site, I meet Steven Cowley, who has been working on the theoretical physics of nuclear fusion for three decades and is now chief executive of the UK Atomic Energy Authority (UKAEA). The last time he saw the site, there was still mud at the bottom of the main pit. Standing over the recently finished concrete platform, he gestures to where the super-hot plasma will one day start burning and fusing atoms. “It’s not ordinary by any stretch of the imagination and when it’s working, you know, it will be one of the great wonders of the world.”
Cowley has been waiting for Iter his whole career. His commitment to it is not just driven by a desire to answer scientific questions that have occupied his mind for so many decades, though. “We don’t know where we are going to get our energy from in the second half of this century, and if we don’t get fusion working we are going to be really stuck,” he says. “We have to make [Iter] work. It’s not just because I work in it that I think that: it has to work and all this effort of thousands of people all the way round the world is to make sure that in 2100 you can flick a switch on the wall and have electricity.”
Nuclear fusion is different from the more familiar nuclear fission, which involves splitting heavy atoms of uranium to release energy and which is at the heart of all nuclear power stations. The promise of fusion, if scientists can get it to work, is huge – unlimited power without any carbon emissions and very little radioactive waste.
The process goes on at the core of every star and the idea that mimicking it could become a source of power on Earth has been around since the years after the second world war. But for many decades fusion has seemed out of reach, requiring materials and an understanding of the chaotic behaviour of hot plasmas that was beyond the technology of the time. However, decades of smaller experiments have led to Iter, the giant project in which fusion scientists have their best possible chance to finally show that this technology could work.
Iter has its roots in a summit between Ronald Reagan and Mikhail Gorbachev towards the end of the cold war, in 1985. They agreed on very little but, almost as an afterthought, they mentioned developing fusion as a new source of energy that could benefit all mankind. Europe and Japan joined the Americans and Russians on the tentative project soon after it was conceived and, today, it also includes China, India and South Korea – in total there are 35 countries involved.
Its design is centred on heating a cloud of hydrogen gas to 10 times hotter than the core of the sun, some 150m degrees celsius, inside a ring-shaped container called a tokamak, which has superconducting magnets fixed around it like hoops fitted on a circular curtain rail. These magnets create an overlapping set of fields that keep the electrically charged gas inside from touching the sides of the tokamak and therefore losing energy.
Building a working tokamak is not straightforward. “The plasma is a bit like a lump of jelly and you are holding it with a magnetic field which is a bit like knitting wool – and imagine holding a lump of jelly with a few pieces of knitting,” says Cowley. The magnets have to be strong and Iter’s design uses superconducting magnets that only work at -269C.
Since the earliest designs, several generations of tokamak-based nuclear fusion reactors have proved that it is possible to build and run the technology at increasingly large sizes. The biggest of these is the Joint European Torus (Jet), based at Culham in Oxfordshire and run by the UKAEA. In the early 1990s, experiments there showed it was possible to fuse hydrogen and then release the resulting energy in a controlled way.
But it took more energy to fuse atoms at Jet than the scientists got back out at the end – which is useless if you want to use the technology to build a power plant. Iter’s primary goal is to fix that problem by creating what they call a “burning” plasma, something that keeps going without the need for external heating, in the same way that a log fire keeps burning after it has initially been set alight by a match. Its design is a scaled-up version of Jet and the scientists here want to produce 500 megawatts of power, 10 times its predicted input.
But scientific challenges are not the only complexities with a mega-project such as Iter. With so many countries involved, so much money and so many engineering contracts, the path to laying even the first building block of this experimental reactor has been far from smooth.
The seven partners agreed on Cadarache in 2004 and they signed an agreement two years later, which costed the project at an estimated €5bn to build and a similar amount to run for its 20-year lifetime. The agreement stated that, as hosts for the project, Europe pays 45% of the total cost while the remaining partners split the bill for the rest between them. Countries do not pay funds directly to Iter but rather provide the equivalent value in parts and services to the reactor project. The ratios are important – they were to remain in place even if the cost rose. Which it did: after a design review in 2008 that incorporated several advances in fusion science into the basic design and also took into account the increased cost of steel and concrete, the construction budget rose to €15bn.
When the Iter agreement was signed in 2006, the reactor was supposed to begin operations in 2016. With the subsequent redesign and construction delays, the current timetable does not involve a switch-on until 2020 and there will not be a working plasma in the tokamak before around 2022. The all-important fusion reactions are not likely to occur before 2027, more than 20 years after building started.
Iter’s director general, the Japanese plasma physicist Osamu Motojima, has been in charge since 2010 and is now in the final months of his tenure. His team came under criticism in 2013 in an assessment carried out by independent consultants, who said the project’s management was inflexible and top-heavy. Motojima says that managing a project to develop a radically new technology with so many political partners was a new experience for the world, and required a “new standard of collaborative culture”. The cost increases, he added, were mainly the result of inflation from the original estimate in 2001 and also the increase in cost of basic building materials. “In general, cost is increasing but that is, I believe, within acceptable levels for stakeholders and the public.” Iter’s governing council has since accepted the thrust of the findings of the assessment report and promised change in how the project is organised, in a bid to keep it on track.
Some of the higher costs are perhaps inevitable when you are building new technologies from scratch. In Iter’s case, this is exacerbated by the partners’ wish to build the components of the reactor in facilities all over the world.
The task of making the million or so parts of Iter has been distributed among the seven partners. Iter will need 100,000km of superconducting wire for its magnets, for example – enough to wrap around the equator twice – which will be made in China, Japan, Russia, South Korea and the US. The magnets are being built in pieces in France and also China. The vacuum chamber, which will contain the hot hydrogen gas, will come in pieces from South Korea and also parts of Europe.
Getting all of this to the site in Cadarache is a huge logistical challenge. The giant magnets and vacuum chambers will begin arriving this month along a specially designated 104km road from the nearest Mediterranean port, known as the Iter Itinerary. The pieces will be enormous – the 18 D-shaped coils of wire that will make the main magnets each weigh 360 tonnes, approximately the same as a fully loaded jumbo jet. The heaviest component will weigh 900 tonnes including its transport vehicle and some will be more than 30m long; the tallest will rise four storeys from the road.
Part of this distribution of labour is to do with bringing jobs, money and prestige into hi-tech industries within the partner nations. But also, these countries will need their own expertise in building and running fusion power plants if Iter is successful and this form of power takes off.
Iter’s job is to show that fusion can be achieved and controlled for sustained periods in the tokamak design. It will not be a power station itself: that is the job of the next generation of fusion reactors, which will be built by countries individually with knowledge gained from the Iter experiments and which are collectively known as “DEMO” projects. China has already started planning the precursor to its DEMO project, a test device called China Fusion Engineering Test Reactor. Construction could start by 2020 and the test plant could be in operation by the mid-2030s, ready to move on to building the first fusion power plants a decade or so after that. Plans to build DEMO power plants are at an earlier stage in Japan and South Korea. There are many hurdles to get over between Iter and a commercial power plant – engineers need to come up with new materials for the walls of the vacuum vessels and the shielding around the plasma, for example. The high-grade steel being used in Iter is good enough to deal with the plasma and the radiation from the relatively small amounts of power (500MW) that will be produced there for a few minutes at a time, but commercial power plants will need much hardier materials in order to deal with the result of a plasma producing three or four times that much power, day in, day out. If all the technical and design refinements in successive experimental reactors go to plan, it is expected that the very first fusion power plants could be producing electricity for the grid by 2045-2050.
It has always been a dark joke about nuclear fusion that a commercial power station is 30 years away and that it always will be. But Iter, despite its delays and cost overruns, might finally tip the balance. It’s hard not to be hopeful that, because of the experiment under way in Cadarache, a commercial fusion reactor really is only 30 years off.
Indeed, if Iter gets its plasma working and scientists can extract more energy from the burn than they had to put in, it will change the world, says Cowley. “When this machine works I will be here, you know. The end of my career is going to be watching this machine do a fusion burn,” he says. “There are probably, over history, a handful of historic moments where in a flash the future changed. In a flash the future will change with this machine… What this is going to show is that man can make a star.”
Fusion facts: how to re-create a star on Earth
At the centre of our sun, the nuclei of hydrogen atoms (which are bare protons) are being jostled around under unimaginable pressures. Once in a while, two of them will overcome their mutual repulsion and fuse to form a nucleus of helium, releasing a little energy in the process.
It is not a straightforward process, however. Such is the repulsive force between protons that it takes millions of years to fuse two of them, even in the intense conditions within a star. Fusing bare hydrogen is therefore out of the question if you want a power station to mimic the process on Earth, so physicists tend to use two of its isotopes instead, deuterium and tritium, which are heavier versions of hydrogen that contain one and two extra neutrons in their nuclei respectively. Deuterium is abundant in seawater – around one in every 6,000 molecules contains it – and tritium can be made by the fusion reactor itself. By heating the deuterium-tritium mixture to hundreds of millions of degrees celsius inside a ring-shaped vessel, the two elements fuse to form helium, energy and fast-moving neutrons. The neutrons will be absorbed by shielding around the reactor vessel that contains lithium, and this interaction will create more tritium.
There is a virtually limitless source of fuel in the world’s oceans to feed future nuclear fusion reactors. And though there are some radioactive waste products that come from the process, they all have short half-lives and will become inert within a few hundred years, as opposed to the thousands of years for which waste from fission reactors stays dangerous.
Further, fusion power plants cannot go critical, produce runaway reactions or a meltdown in the way people sometimes worry about with nuclear fission. Like an internal combustion engine, fusion reactors will only burn the fuel put into them. And you only need a very small amount of fuel for each fusion burn in a large power station – tenths of a gram of the mixture will fill up the reactor – since the deuterium-tritium mixture is a million times more energy-dense than petrol.