Is the search for fusion energy, long dominated by doughnut-shaped devices called tokamaks, about to undergo a shape shift? Just as ITER, the world’s largest tokamak—and at tens of billions of dollars the most expensive—nears completion in the hills of southern France, a much smaller testbed with a twistier geometry will start throttling up to full power in Germany.
If the 16-meter-wide device, called a stellarator, can match or outperform similar-size tokamaks, it could cause fusion scientists to rethink the future of their field. Stellarators have several key advantages, including a natural ability to keep the roiling superhot gases they contain stable enough to fuse nuclei and release energy. Even more crucial for a future fusion power plant, they can theoretically just run and run, whereas tokamaks must stop periodically to reset their magnet coils.
In runs of a few seconds, the €1 billion German machine, dubbed Wendelstein 7-X (W7-X), is already getting “tokamak-like performance,” says plasma physicist David Gates, proving adept at preventing particles and heat from escaping the superhot gas. If W7-X can achieve long runs, “it will be clearly in the lead,” he says. “That’s where stellarators shine.” Theorist Josefine Proll of the Eindhoven University of Technology is equally enthusiastic: “All of a sudden, stellarators are back in the game.” The encouraging prospects are inspiring a clutch of startup companies, including one for which Gates is now leaving Princeton Plasma Physics Laboratory, to develop their own stellarators.
W7-X has been operating since 2015 at the Max Planck Institute for Plasma Physics (IPP) in Greifswald, Germany, but only at relatively low power levels and for short runs. Over the past 3 years, W7-X’s creators stripped it down and replaced all the interior walls and fittings with water-cooled versions, opening the way to much longer, hotter runs. At a W7-X board meeting last week, the team reported that the revamped plasma vessel has no leaks and is ready to go. It is expected to restart later this month, on its way to showing whether it can truly get plasma to conditions that, in a future device, would ignite fusion.
inside Wendelstein 7-X
Wendelstein 7-X’s twisting inner surface is now water cooled, enabling longer runs.IPP/JAN HOSAN
Both stellarators and tokamaks create magnetic cages for gas at more than 100 million degrees Celsius, so hot it would melt any metal container. Heating is provided by microwaves or high energy particle beams. The outlandish temperatures produce a plasma—a roiling mix of separated nuclei and electrons—and cause the nuclei to slam together with such force that they fuse, releasing energy. A fusion power plant would be fueled with a mix of the hydrogen isotopes deuterium and tritium, which react most readily. Research machines like W7-X that aren’t trying to generate energy avoid radioactive tritium and stick to safer, more plentiful hydrogen or deuterium.
To make their plasma-confining magnetic fields, tokamaks and stellarators employ electromagnetic coils looping around the vessel and through the central hole. But such a field is stronger nearer the hole than the outer edge, causing plasma to drift to the reactor’s wall.
Tokamaks tame the drift by making the plasma flow around the ring. That streaming generates another magnetic field, twisting the ionized gas like a candy cane and steadying it. Stellarators use weirdly shaped magnetic coils instead of streaming plasma to produce the twist. The tokamak scheme has long proved the more successful at holding plasma in place, but once plasma physicists had supercomputers powerful enough, they could tweak the complex geometries of stellarator magnets to improve confinement, a process called optimization.
W7-X is the first large, optimized stellarator and contains 50 bizarrely twisted superconducting coils, each weighing 6 tons. Its construction, begun in the mid-1990s, was tortuous, completed 10 years late and costing almost twice the €550 million originally budgeted.
Despite the wait, researchers haven’t been disappointed. “The machine worked immediately,” says W7-X director Thomas Klinger. “It’s a very easy-going machine. [It] just did what we told it to do.” This contrasts with tokamaks, which are prone to “instabilities”—the plasma bulging or wobbling in unpredictable ways—or more violent “disruptions,” often linked to interrupted plasma flow. Because stellarators don’t rely on plasma current, that “removes a whole branch” of instabilities, says IPP theorist Sophia Henneberg.
In early stellarators, the geometry of the magnetic field caused some slower moving particles to follow banana-shaped orbits until they collided with other particles and got knocked out of the plasma, leaching out energy. W7-X’s ability to suppress that effect means its “optimization worked as it was supposed to,” Gates says.
With this Achilles heel removed, W7-X mostly loses heat through other forms of turbulence—little eddies that push particles toward the wall. Simulating turbulence takes serious computing power, and theorists have only recently got a handle on it. W7-X’s upcoming campaign should validate the simulations and test ways to combat turbulence.
Wendelstein 7-X’s construction was long and complex (video produced in 2015).
The campaign should also showcase a stellarator’s ability to run continuously, in contrast to the pulsed operation of a tokamak. W7-X has already operated for runs of 100 seconds—long by tokamak standards—but at relatively low power. Not only were its components uncooled, but the device’s microwave and particle heating systems could only deliver 11.5 megawatts of power. The upgrade will boost the heating power by 60%. Running W7-X at high temperature, high plasma density, and for long runs will be the real test of stellarators’ potential for producing fusion power. An initial aim, Klinger says, is to get the ion temperature up to 50 million degrees Celsius for 100 seconds. That would put W7-X “among the leading machines in the world,” he says. Then, the team will push it for longer, up to 30 minutes. “We’ll go step by step, exploring uncharted territory,” he says.
W7-X’s achievements have prompted venture capitalists to back several startups developing commercial power-producing stellarators. First priority for the startups: Find a simpler way to make the magnets.
Princeton Stellarators, founded this year by Gates and colleagues, has secured $3 million and is aiming to build a demonstration reactor that will forgo the twisted magnet coils of W7-X. Instead, it will rely on a mosaic of about 1000 tiny square coils made of high-temperature superconductor (HTS) on the outside surface of the plasma vessel. By varying the magnetic field produced by each coil, operators will be able to change the shape of the applied field at will. “It takes complexity out of the coils and puts it in the control system,” Gates says. The firm hopes to initially develop a reactor that will fuse just cheap, abundant deuterium, to generate not power, but neutrons for manufacturing radioisotopes. If successful, the firm will then aim for a power-producing reactor.
Renaissance Fusion, based in Grenoble, France, has raised €16 million and plans to coat segments of the plasma vessel in a multilayered HTS, forming a uniform coating. Then, using a laser, engineers will burn off tracks within the superconductor to etch a twisting pattern of magnet coils. They aim to make a meter-long test segment over the next 2 years and a full prototype by 2027.
A third firm, Type One Energy in Madison, Wisconsin, received U.S. Department of Energy funding to develop HTS cables with enough bend to be used in stellarator magnets. The company would sculpt pieces of metal with computer-controlled etching machines, carving twisting channels into which the cable is wound to turn it into a coil. “Advanced manufacturing technology opens the door for the stellarator,” says co-founder David Anderson of the University of Wisconsin, Madison.
Anderson says the next phase of W7-X’s operation will accelerate the boom in stellarator efforts. “With half-hour discharges, you’re essentially steady-state,” he says. “This is a big deal.”