Fusion’s Missing Pieces

Scientific American, June 2012 Volume 306, Number 6

Geneva was cold and gray when air force one touched down in November 1985. President Ronald Reagan had come to meet Mikhail Gorbachev, the newly appointed leader of the Soviet Union. Reagan was convinced that the risk of catastrophic nuclear war was high, and he wanted to reduce the two superpowers’ swollen arsenals. Gorbachev also recognized that the arms race was strangling the Soviet economy.

Yet the tête-à-tête quickly degenerated. Reagan lectured Gorbachev on the history of Soviet aggression. Gorbachev attacked Reagan’s Strategic Defense Initiative, an ambitious plan to knock incoming nuclear weapons out of the sky. Negotiations nearly broke down. At five in the morning, the two sides agreed to a joint statement with no firm commitments. At the bottom—almost as a footnote—Reagan and Gorbachev inserted
a gauzy pledge to develop a new source of energy “for the benefit of all mankind.”

That note set in motion a project that has evolved into arguably the most ambitious scientific undertaking of the 21st century—a mash-up of complex experimental technologies that will, if all goes well, underpin the final solution to humanity’s
energy crisis.

ITER (formerly the International Thermonuclear Experimental Reactor) will attempt to reproduce the sun’s power hereon earth. It will generate around 500 megawatts of power, 10 times the energy needed to run it, using little more than hydrogen, the most abundant element in the universe. The project will illustrate a proof of principle for a technology that could lead to a nearly unlimited supply of energy for the power-hungry world. Politicians from seven participating members, including the U.S. and Russia, have enthusiastically enlisted their nations in the effort.

Yet like the summit that birthed it, ITER (pronounced “eater”) has not lived up to expectations. Cost estimates have doubled and doubled again as engineering problems find bureaucratically expedient solutions. For instance, rather than pooling resources, the seven partners are producing bits and pieces in their home countries, then assembling them at ITER’s building site in the south of France. The process is akin to ordering nuts, bolts and brackets from a catalogue, then trying to build a 747 in your backyard. Progress is glacial. Less than a year ago ITER was a 56-foot-deep hole in the ground, which has only recently been filled with nearly four million cubic feet of concrete. The start date has slipped from 2016 to 2018 to late 2020. The first real energy-producing experiments will not come before 2026—two decades after the start of construction.


The ITER fusion reactor promises to be a landmark step on the road toward unlimited clean energy. Once running, the machine will produce 10 times the amount of energy needed to power it. Yet for all its promise, the ITER project is in trouble. Billions of dollars over budget and years behind schedule, the reactor will not start power-production experiments until 2026 at the earliest. The complex reasons behind the troubles include unforeseen engineering difficulties and the baroque bureaucratic squabbles of a global partnership of seven major stakeholders. Critics contend that ITER has become a pie-in-the-sky boondoggle whose only purpose is to suck money away from productive clean-energy research projects like wind and solar energy.

And ITER is just the beginning of this putative new source of energy. Even if it is successful, another generation of test reactors will have to follow it, and only after these have run their course will local municipalities begin to build fusion plants to supply the grid. ITER is but one step in a project that will continue for decades, if not centuries.

Supporters argue that ITER is the only hope, in the long term, of meeting the world’s unquenchable demand for power. But even they have been forced to recalibrate their utopian expectations. The project now seems to be propelled by institutional inertia—it is easier for individual governments to stay the course rather than be the lone pariah who pulls out early. Critics, meanwhile, have more ammunition with each delay and cost overrun. ITER, they say, is a colossal waste of money at a time when funding is desperately needed in other areas of energy research. Both sides agree: when the project is finally completed, it had better work.


in theory, fusion is the perfect energy source. It depends on the one thing in physics that everyone has heard of: energy equals mass times the speed of light, squared (E = mc2). Because the speed of light is so great, E = mc2 means that a very small amount of mass can generate an enormous quantity of energy.

All nuclear reactions exploit this basic law of the universe. In the case of ordinary nuclear power plants, heavy uranium nuclei split apart to create lighter elements. During this fission, a tiny fraction of the uranium’s mass turns directly into energy. Fusion is the same, except backwards. When light nuclei such as hydrogen come together, they create helium ions that weigh slightly less than their parents. Per unit mass, fusion fuel can release around three times the energy of uranium fission. Even more important, hydrogen is far more abundant than uranium, and fusion’s helium waste products are not radioactive.

“Fusion is seductive,” says Gyung-Su Lee, a South Korean scientist who has devoted years to ITER negotiations. “It’s like people searching for ways to make gold in the Middle Ages. It’s the holy grail of energy research.”

Lee is a fierce believer in fusion’s power. In 1980 he arrived as a graduate student at the University of Chicago to study quantum field theory, one of the toughest corners of physics. But America changed Lee’s thinking. “In the U.S., money is everything,” he says, and quantum field theory offers only intellectual riches. He began to look for something more practical to study and eventually settled on fusion. “It’s very difficult, scientifically and also in engineering,” he notes. Yet if it worked, the payoff would be huge: energy would be widely available and cheap; fossil fuels would become irrelevant. The world would be transformed.

Scientists such as Lee have been seduced by fusion for half a century. Many before him have promised its impending arrival. Although some of those researchers were charlatans, the vast majority of them turned out to be plain wrong. Fusion is tough, and nature breaks promises.

Here is the core challenge: because hydrogen ions repel one another, scientists must slam them together to make them fuse. ITER’s strategy is to heat the hydrogen inside a magnetic cage. The particular type of magnetic cage it employs is called a tokamak— a metal doughnut circled by loops of coil that generate magnetic fields. These magnetic cuffs squeeze a charged plasma of hydrogen ions as it warms to hundreds of millions of degrees—temperatures no solid material can withstand.

In the 1970s tokamaks looked so promising that some researchers predicted they could build fusion electricity plants by the mid-1990s. The only challenge was scaling research reactors up to sufficient size—in general, the bigger the tokamak, the hotter the plasma can get, and the more efficient fusion becomes. Then problems arose. Plasma conducts electricity and so can suffer from self-generated currents that make it buck and writhe. Violent turbulence snaps the plasma out of its cage, firing it toward the machine’s wall. As the temperature rises, the tokamak grows to give the plasma space, and the magnetic fields need to be stronger to hold it. Extra room and stronger magnetic fields require higher electric current in the doughnut’s copper coils. And higher current requires more power. Put simply: the larger and more powerful a machine becomes, the more energy it consumes trying to hold everything together.

This feedback meant that conventional tokamaks would never produce more energy than they consumed. Lee and others knew of only one solution: superconductors—special materials that, at very low temperatures, can carry extremely high current with no resistance. If a tokamak’s magnets were superconducting, they could be pumped up with current and left to run indefinitely. It would solve the energy problem but would not be cheap. Superconductors are exotic, expensive materials. And to work, they need to be constantly cooled with liquid helium to just four kelvins above absolute zero.

Even in 1985 it was clear that neither Russia nor America could build a tokamak large enough to produce net energy. When ITER officially began, it was as a joint project among the U.S., the Soviet Union, Japan and Europe. The design was enormous and used the latest technology of the time. In addition to superconductors, ITER incorporated advanced accelerators to fire neutral beams of atoms into the core to heat it, along with sophisticated antennas that would act something like a microwave for plasmas. Rather than using plain hydrogen for fuel, ITER would use deuterium and tritium, two hydrogen isotopes that fuse at lower temperatures and pressures. Deuterium is relatively common—a drop of ocean water contains many trillions of deuterium atoms— but tritium is rare, radioactive and pricey. The original construction costs were estimated at $5 billion, but by the mid-1990s a more thorough accounting of the machine’s complexities had doubled the price. In 1998, in large part because of the expense, the U.S. left the project.

Shortly thereafter, a small team desperate to keep the project alive hastily redesigned it at half the size and half the cost. Unfortunately, because of “the limited time to finish the design, some things were forgotten,” admits Gunther Janeschitz, a senior scientist with ITER and a member of the original redesign team. The member states fought over all the big bits of the machine, but some of the little things feedthroughs, connections—never got assigned. “There were holes between two of the components, and none of the procurement packages really described it,” he says.

These gaps are the scourge of ITER because the machine is not really being manufactured by the ITER organization itself. Established nations such as Russia and Japan want their investment in ITER to go to scientists in their state-run laboratories, whereas newcomers such as India and China want to give their burgeoning industry a chance to learn advanced new technologies. Therefore, member states contribute fully built units to the enterprise (along with a small financial contribution to the central organization). Superconducting cables for its magnets will arrive from Hitachi in Japan, but they will also be supplied by Western Superconducting Technologies Company in China and the Efremov Scientific Research Institute of Electrophysical Apparatus in Russia. The machine’s giant vacuum vessel will be constructed in Europe, India, Korea and Russia; the heating systems will come by way of Europe, Japan, India and the U.S., which rejoined the project in 2003. The central ITER organization must take these parts, figure out what is missing, then cobble everything together into the most sophisticated experiment ever built.

The challenge becomes clear at a medieval château overlooking the Durance River on the
other side of a two-lane highway from ITER’s temporary headquarters. Here ITER’s members gather inside a purpose-built meeting room crammed with flat screens and microphones. The partners have no interest in letting a reporter in on the negotiations, but during a coffee break, Lee tells me a minor crisis is unfolding behind closed doors. “The Indians think a pipe should end here, and others think it should end there,” he says, gesturing to opposite ends of the room with a small chocolate tart from the pastry table. “The obvious solution is to meet in the middle, but this is not technically possible. So we hand it on to the DG.”

Until 2010 the DG, or director general, was a soporific Japanese diplomat named Kaname
Ikeda. As these kinds of problems mounted, Ikeda resigned under pressure from ITER’s council and was replaced by Osamu Motojima, a veteran Japanese fusion researcher whose quiet nature belies what insiders describe as a tough and sometimes autocratic personality. Motojima and his deputies, veterans of the U.S. and European programs, sit down to work out a deal with the Indians in a converted stable next to the conference room. While the team haggles, Harry Tuinder, at the time ITER’s chief legal adviser (he has since left the organization for the European Commission), sits in the courtyard and lights a cigarette. I ask him if it would not make more sense if Motojima had the authority to force every nation to contribute the parts he needs. “That would basically be degrading all the relationships you try to reinforce,” Tuinder says, leaning back in his chair. At the end of the day, it is members’ willing participation, not the power of ITER’s director general, that will make the project come together.


as negotiations drag on, ITER’s costs have doubled yet again to an estimated $20 billion, although the piecemeal way in which it is being built means that the actual cost may never be known. Its completion date has slipped by another couple of years.The soaring price and lengthening delays have been fueling opposition to the giant tokamak, particularly in Europe, which is supplying around 45 percent of its construction costs.

“If we really want to put money to save the climate and have energy independence, then obviously this experiment is nonsense,” says Michel Raquet, energy adviser to the Green Party of the European Parliament. The European Union is currently working on a budget that will accommodate the estimated €2.7 billion that ITER requires to complete construction by 2020. The Greens, ITER’s chief opponents in Europe, fear that the money
will come at the expense of such renewables as wind and solar.

In the U.S., which will pay just 9 percent of the cost, opposition is more muted. “It’s not threatening—it’s just a waste of money,” says Thomas Cochran, a nuclear campaigner with the Natural Resources Defense Council. Cochran asserts that he would rather devote his energy to fighting other nuclear research programs that generate long-term waste or spread nuclear weapons technology. The U.S. Congress seems similarly indifferent about the program. “All I can say is that there’s no move to kill it,” says Stephen Dean, president of Fusion Power Associates, which advocates for the development of fusion energy. But that may change. The budget President Barack Obama presented this year funds a steep rise in ITER costs by slashing spending on domestic fusion research. Even then, the $150 million ITER will receive is 25 percent less than the U.S.’s scheduled contribution.

Other nations are also encountering trouble with their commitments to ITER. India has struggled to hand out contracts, and last March’s massive earthquake off the coast of Japan damaged key facilities there. “Every country has its own reasons for delays,” says Vladimir Vlasenkov, a member of the Russian delegation. Russia, he hastens to add, is on track.

ITER will prove whether fusion is achievable. It will not prove whether it is commercially viable. There is good reason to think it might not be. For starters, the radiation from fusion is very intense and will damage ordinary material such as steel. A power plant will have to incorporate some as yet undeveloped materials that can withstand years of bombardment from the plasma—otherwise the reactor will be constantly down for servicing. Then there is the problem of tritium fuel, which must be made on-site, probably by using the reactor’s own radiation.

Arguably the greatest obstacle to building a reactor based on ITER is the machine’s incredible complexity. All the specialized heating systems and custom built parts are fine in an experiment, but a power plant will need to be simpler, says Steve Cowley, CEO of the U.K.’s Atomic Energy Authority. “You can’t imagine producing power day in and day out on a machine that’s all bells and whistles,” he says. Another generation of expensive demonstration reactors must be built before fusion can come onto the grid. Given ITER’s lumbering development, none of these will be up and running before the middle of the century.

Despite these setbacks and the uncertain future of fusion energy as a whole, it is difficult to find anyone familiar with ITER who thinks the machine will not get built. Peer pressure is one reason: “The French are in it and won’t back out because the U.S. is in it and won’t back out,” Cochran says. Political visibility for the countries involved—and substantial penalties for pulling out early—also serves to keep the project moving, Tuinder observes.

Those legitimate, if cynical, reasons for staying the course aside, many scientists genuinely feel that fusion is the only hope to meet the world’s energy demands. “I was scared about the future energy of the world—I didn’t know where it would come from,” says Raymond Orbach, chief scientist at the Department of Energy at the time the U.S. rejoined the project. “It’s CO2-free, it’s essentially unlimited, it has no environmental impact—come up with an alternative.” Most fusion scientists think a climate crisis is inevitable anyway. Further down the line, after humanity has learned its lesson, “we’d better have a set of technologies ready,” Cowley admonishes. It is going to work, this line of thinking goes, because it must.