Feature

# The new shape of fusion

See allHide authors and affiliations

Science  22 May 2015:
Vol. 348, Issue 6237, pp. 854-856
DOI: 10.1126/science.348.6237.854

ITER, the international fusion reactor being built in France, will stand 10 stories tall, weigh three times as much as the Eiffel Tower, and cost its seven international partners $18 billion or more. The result of decades of planning, ITER will not produce fusion energy until 2027 at the earliest. And it will be decades before an ITER-like plant pumps electricity into the grid. Surely there is a quicker and cheaper route to fusion energy. Fusion enthusiasts have a slew of schemes for achieving the starlike temperatures or crushing pressures needed to get hydrogen nuclei to come together in an energy-spawning union. Some are mainstream, such as lasers, some unorthodox (Science, 25 July 2014, p. 370). Yet the doughnut-shaped vessels called tokamaks, designed to cage a superheated plasma using magnetic fields, remain the leading fusion strategy and are the basis of ITER. Even among tokamaks, however, a nimbler alternative has emerged: a spherical tokamak. Imagine the doughnut shape of a conventional tokamak plumped up into a shape more like a cored apple. That simple change, say the idea's advocates, could open the way to a fusion power plant that would match ITER's promise, without the massive scale. “The aim is to make tokamaks smaller, cheaper, and faster—to reduce the eventual cost of electricity,” says Ian Chapman, head of tokamak science at the Culham Centre for Fusion Energy in Abingdon, U.K. Culham is one of two labs about to give these portly tokamaks a major test. The world's two front-rank machines—the National Spherical Torus Experiment (NSTX) at the Princeton Plasma Physics Laboratory in New Jersey and the Mega Amp Spherical Tokamak (MAST) in Culham—are both being upgraded with stronger magnets and more powerful heating systems. Soon they will switch on and heat hydrogen to temperatures much closer to those needed for generating fusion energy. If they perform well, then the next major tokamak to be built—a machine that would run in parallel with ITER and test technology for commercial reactors—will likely be a spherical tokamak. A small company spun off from Culham is even making a long-shot bet that it can have a spherical tokamak reactor capable of generating more energy than it consumes—one of ITER's goals—up and running within the decade. If it succeeds, spherical tokamaks could change the shape of fusion's future. “It's going to be exciting,” says Howard Wilson, director of the York Plasma Institute at the University of York in the United Kingdom. “Spherical tokamaks are the new kids on the block. But there are still important questions we're trying to get to the bottom of.” TOKAMAKS ARE AN INGENIOUS WAY to cage one of the most unruly substances humans have ever grappled with: plasma hot enough to sustain fusion. To get nuclei to slam together and fuse, fusion reactors must reach temperatures 10 times hotter than the core of the sun, about 150 million degrees Celsius. The result is a tenuous ionized gas that would vaporize any material it touches—and yet must be held in place long enough for fusion to generate useful amounts of energy. Tokamaks attempt this seemingly impossible task using magnets, which can hold and manipulate plasma because it is made of charged particles. A complex set of electromagnets encircle the doughnut-shaped vessel, some horizontal and some vertical, while one tightly wound coil of wire, called a solenoid, runs down the doughnut hole. Their combined magnetic field squeezes the plasma toward the center of the tube and drives it around the ring while also twisting in a slow corkscrew motion. But plasma is not easy to master. Confining it is like trying to squeeze a balloon with your hands: It likes to bulge out between your fingers. The hotter a plasma gets, the more the magnetically confined gas bulges and wriggles and tries to escape. Much of the past 60 years of fusion research has focused on how to control plasma. Generating and maintaining enough heat for fusion has been another challenge. Friction generated as the plasma surges around the tokamak supplies some of the heat, but modern tokamaks also beam in microwaves and high-energy particles. As fast as the heat is supplied, it bleeds away, as the hottest, fastest moving particles in the turbulent plasma swirl away from the hot core toward the cooler edge. “Any confinement system is going to be slightly leaky and will lose particles,” Wilson says. Studies of tokamaks of different sizes and configurations have always pointed to the same message: To contain a plasma and keep it hot, bigger is better. In a bigger volume, hot particles have to travel farther to escape. Today's biggest tokamak, the 8-meter-wide Joint European Torus (JET) at Culham, set a record for fusion energy in 1997, generating 16 megawatts for a few seconds. (That was still slightly less than the heating power pumped into the plasma.) For most of the fusion community, ITER is the logical next step. It is expected to be the first machine to achieve energy gain—more fusion energy out than heating power in. In the 1980s, a team at Oak Ridge National Laboratory in Tennessee explored how a simple shape change could affect tokamak performance. They focused on the aspect ratio—the radius of the whole tokamak compared to the radius of the vacuum tube. (A Hula-Hoop has a very high aspect ratio, a bagel a lower one.) Their calculations suggested that making the aspect ratio very low, so that the tokamak was essentially a sphere with narrow hole through the middle, could have many advantages. Near a spherical tokamak's central hole, the Oak Ridge researchers predicted, particles would enjoy unusual stability. Instead of corkscrewing lazily around the tube as in a conventional tokamak, the magnetic field lines wind tightly around the central column, holding particles there for extended periods before they return to the outside surface. The D-shaped cross section of the plasma would also help suppress turbulence, improving energy confinement. And they reckoned that the new shape would use magnetic fields more efficiently—achieving more plasma pressure for a given magnetic pressure, a ratio known as beta. Higher beta means more bang for your magnetic buck. “The general idea of spherical tokamaks was to produce electricity on a smaller scale, and more cheaply,” Culham's Chapman says. But such a design posed a practical problem. The narrow central hole in a spherical tokamak didn't leave enough room for the equipment that needs to fit there: part of each vertical magnet plus the central solenoid. In 1984, Martin Peng of Oak Ridge came up with an elegant, space-saving solution: replace the multitude of vertical ring magnets with C-shaped rings that share a single conductor down the center of the reactor (see graphic). U.S. fusion funding was in short supply at that time, so Oak Ridge could not build a spherical machine to test Peng's design. A few labs overseas converted some small devices designed for other purposes into spherical tokamaks, but the first true example was built at the Culham lab in 1990. “It was put together on a shoestring with parts from other machines,” Chapman says. Known as the Small Tight Aspect Ratio Tokamak (START), the device soon achieved a beta of 40%, more than three times that of any conventional tokamak. It also bested traditional machines in terms of stability. “It smashed the world record at the time,” Chapman says. “People got more interested.” Other labs rushed to build small spherical tokamaks, some in countries not known for their fusion research, including Australia, Brazil, Egypt, Kazakhstan, Pakistan, and Turkey. The next question, Chapman says, was “can we build a bigger machine and get similar performance?” Princeton and Culham's machines were meant to answer that question. Completed in 1999, NSTX and MAST both hold plasmas about 3 meters across, roughly three times bigger than START's but a third the size of JET's. The performance of the pair showed that START wasn't a one-off: again they achieved a beta of about 40%, reduced instabilities, and good confinement. Now, both machines are moving to the next stage: more heating power to make a hotter plasma and stronger magnets to hold it in place. MAST is now in pieces, the empty vacuum vessel looking like a giant tin can adorned with portholes, while its €30 million worth of new magnets, pumps, power supplies, and heating systems are prepared. At Princeton, technicians are putting the finishing touches to a similar$94 million upgrade of NSTX's magnets and neutral beam heating. Like most experimental tokamaks, the two machines are not aiming to produce lots of energy, just learning how to control and confine plasma under fusionlike conditions. “It's a big step,” Chapman says. “NSTX-U will have really high injected power in a small plasma volume. Can you control that plasma? This is a necessary step before you could make a spherical tokamak power plant.”

The upgraded machines will each have a different emphasis. NSTX-U, with the greater heating power, will focus on controlling instabilities and improving confinement when it restarts this summer. “If we can get reasonable beta values, [NSTXU] will reach plasma [properties] similar to conventional tokamaks,” says NSTX chief Masayuki Ono. MAST-Upgrade, due to fire up in 2017, will address a different problem: capturing the fusion energy that would build up in a full-scale plant.

Fusion reactions generate most of their energy in the form of high-energy neutrons, which, being neutral, are immune to magnetic fields and can shoot straight out of the reactor. In a future power plant, a neutron-absorbing material will capture them, converting their energy to heat that will drive a steam turbine and generate electricity. But 20% of the reaction energy heats the plasma directly and must somehow be tapped. Modern tokamaks remove heat by shaping the magnetic field into a kind of exhaust pipe, called a divertor, which siphons off some of the outermost layer of plasma and pipes it away. But fusion heat will build up even faster in a spherical tokamak because of its compact size. MAST-Upgrade has a flexible magnet system so that researchers can try out various divertor designs, looking for one that can cope with the heat.

Researchers know from experience that when a tokamak steps up in size or power, plasma can start misbehaving in new ways. “We need MAST and NSTX to make sure there are no surprises at low aspect ratio,” says Dennis Whyte, director of the Plasma Science and Fusion Center at the Massachusetts Institute of Technology in Cambridge. Once NSTX and MAST have shown what they are capable of, Wilson says, “we can pin down what a [power-producing] spherical tokamak will look like. If confinement is good, we can make a very compact machine, around MAST size.”

### VIDEO

Fusion scientists explain the new shape at http://scim.ag/fusionshape.

BUT GENERATING ELECTRICITY isn't the only potential goal. The fusion community will soon have to build a reactor to test how components for a future power plant would hold up under years of bombardment by high-energy neutrons. That's the goal of a proposed machine known in Europe as the Component Test Facility (CTF), which could run stably around the clock, generating as much heat from fusion as it consumes. A CTF is “absolutely necessary,” Chapman says. “It's very important to test materials to make reactors out of.” The design of CTF hasn't been settled, but spherical tokamak proponents argue their design offers an efficient route to such a testbed—one that “would be relatively compact and cheap to build and run,” Ono says.

With ITER construction consuming much of the world's fusion budget, that promise won't be tested anytime soon. But one company hopes to go from a standing start to a small power-producing spherical tokamak in a decade. In 2009, a couple of researchers from Culham created a spinoff company—Tokamak Solutions—to build small spherical tokamaks as neutron sources for research. Later, one of the company's suppliers showed them a new multilayered conducting tape, made with the high-temperature superconductor yttrium-barium-copper-oxide, that promised a major performance boost.

Lacking electrical resistance, superconductors can be wound into electromagnets that produce much stronger fields than conventional copper magnets. ITER will use low-temperature superconductors for its magnets, but they require massive and expensive cooling. High-temperature materials are cheaper to use but were thought to be unable to withstand the strong magnetic fields around a tokamak—until the new superconducting tape came along. The company changed direction, was renamed Tokamak Energy, and is now testing a first-generation superconducting spherical tokamak no taller than a person.

Superconductors allow a tokamak to confine a plasma for longer. Whereas NSTX and MAST can run for only a few seconds, the team at Tokamak Energy this year ran their machine—albeit at low temperature and pressure—for more than 15 minutes. In the coming months, they will attempt a 24-hour pulse—smashing the tokamak record of slightly over 5 hours.

Next year, the company will put together a slightly larger machine able to produce twice the magnetic field of NSTX-U. The next step—investors permitting—will be a machine slightly smaller than Princeton's but with three times the magnetic field. Company CEO David Kingham thinks that will be enough to beat ITER to the prize: a net gain of energy. “We want to get fusion gain in 5 years. That's the challenge,” he says.

“It's a high-risk approach,” Wilson says. “They're buying their lottery ticket. If they win, it'll be great. If they don't, they'll likely disappear. Even if it doesn't work, we'll learn from it; it will accelerate the fusion program.”

It's a spirit familiar to everyone trying to reshape the future of fusion.

View Abstract