A fusion research reactor offering performance similar to internationally backed ITER—but at one-eighth the volume and a fraction of the cost—could be built in as little as 10 years, researchers at MIT claim. Their design uses commercially available high-temperature superconductors and would produce magnetic field strengths double those of ITER.
The conceptual design for a tokamak fusion reactor proposed by MIT researchers. The design features rare-earth barium copper oxide high-temperature superconducting magnets for the principal magnetic coils (brown) and a “blanket” of molten salt (light blue) surrounding the vacuum vessel. Superconducting coils (green) shape and control the plasma (yellow); low-current copper control coils are shown in red. The central solenoid magnet (blue) is used to start up the plasma.
The conceptual design for a tokamak fusion reactor proposed by MIT researchers. The design features rare-earth barium copper oxide high-temperature superconducting magnets for the principal magnetic coils (brown) and a “blanket” of molten salt (light blue) surrounding the vacuum vessel. Superconducting coils (green) shape and control the plasma (yellow); low-current copper control coils are shown in red. The central solenoid magnet (blue) is used to start up the plasma.
Codeveloped by MIT’s Plasma Science and Fusion Center director Dennis Whyte and a dozen graduate students, the affordable, robust, compact (ARC) reactor is estimated to cost about $5 billion. Whyte and his team recently published the ARC design in Fusion Engineering and Design. No estimate for ITER’s cost exists (see Physics Today, May 2015, page 21), but extrapolating estimates of the US’s 9% share in the seven-party international project would put the total at about $40 billion.
With their high-temperature superconducting tapes made of rare-earth barium copper oxide (REBCO), the MIT-designed toroidal coils would produce a magnetic field of 23 T, nearly twice the 11.8 T to be generated by ITER’s niobium–tin low-temperature superconducting magnets. Although REBCO operates at liquid-nitrogen temperature, 77 K, the MIT designers would cool the material down to 20 K, which would substantially increase the critical current density of the tape-wound coils and therefore the magnetic fields they produce.
Coils made with the REBCO tapes have produced fields as high as 27 T in small-scale proof-of-principle experiments—50% higher than the highest fields achieved with Nb3Sn, says David Larbalestier, director of the Applied Superconductivity Center at the National High Magnetic Field Laboratory in Tallahassee, Florida.
The higher fields would allow a machine with ITER’s performance level to be built at the scale of the Joint European Torus in the UK and Japan’s JT-60, Whyte says. (JET’s plasma volume is about 100 m3, JT-60’s is 90 m3, and ITER’s is 840 m3). Indeed, were it possible to retrofit ITER with the high-temperature, high-field magnets, the reactor would produce gigawatts of power in a device that is designed to produce a maximum of 500 MW.
A small UK company, Tokamak Energy, has already built a small spherical tokamak using the REBCO material. “We know we can cool [the magnets] efficiently to temperatures of around 30–40 K where they superconduct very nicely,” says CEO David Kingham. But he admits that scaling up in size to a working fusion reactor would require “significant advances.” If that scaling works, a fusion reactor with REBCO coils producing significant amounts of fusion power might be built for as little as one-fifth the cost of the MIT design, he says. Such a device is unlikely to be commercially viable, however.
The private, 16-employee Tokamak Energy has raised $15 million from investors that include Oxford Instruments and its cofounder Martin Wood and the UK’s Rainbow Seed Fund. Tokamak Energy recently announced that it had sustained a plasma in its device for more than 24 hours, a duration that Kingham believes is a record. But the company has operated only with hydrogen so far, not the deuterium–tritium mix required for fusion.
A long history
Steven Cowley, head of the UK’s Culham Centre for Fusion Energy, home to JET, says high-field magnets in fusion research aren’t new. “There are two things we know about magnetic confinement: Things get better as [the reactor] gets bigger, and things get better if the field is stronger,” he says. MIT, in particular, has long been home to high-field advocates. The Alcator C-Mod fusion reactor at MIT, operating since 1993, is a high-field reactor, with copper magnets that have produced 24 T. But an ARC reactor with copper magnets would require 500 MW of electricity to run, compared to 1 MW with REBCO coils, according to Brandon Sorbom, the lead student on the MIT team. Copper magnets, moreover, would melt if operated for longer than a few seconds at a time.
The inside of MIT’s Alcator C-Mod high-field tokamak. The copper-magnet device has produced what researchers there say is a world record for plasma pressure of 3 atmospheres at thermonuclear temperatures. Internal components are protected from neutron damage by tiles of high-melt-temperature metallic armor (center). The louvered components on the outside walls are RF heating antennas.
The inside of MIT’s Alcator C-Mod high-field tokamak. The copper-magnet device has produced what researchers there say is a world record for plasma pressure of 3 atmospheres at thermonuclear temperatures. Internal components are protected from neutron damage by tiles of high-melt-temperature metallic armor (center). The louvered components on the outside walls are RF heating antennas.
A high-field reactor design known as Ignitor was first proposed in 1977 by MIT physicist Bruno Coppi as a machine that would achieve ignition, the point at which the fusion reaction becomes self-sustaining. It’s been redesigned several times over the years. But despite a pledge of €78 million ($87 million) from the Italian government in 2011 and in-kind contributions from Russia, Ignitor still awaits construction at a facility near Moscow. Coppi blames the delay on an intransigent Italian bureaucracy.
Coppi is hopeful that the new MIT concept will encourage people “to wake up to the fact that the high-field approach is the most promising, because we’ve known that for many years.”
An earlier high-field reactor concept was proposed in the 1970s by the late Robert Bussard, who was with the fusion program at the Atomic Energy Commission. Bussard, with funding from publisher Bob Guccione, worked to develop a high-field tokamak that used disposable copper magnets. That research ended when the funding ran out.
Crushing the Mariana Trench
Higher fields would permit a much smaller reactor to produce a given amount of power—a doubling of the field will produce 16 times the fusion power density. However, a major challenge is to build coils strong enough to withstand the tremendous forces the fields generate. “What everybody in the world is trying to do is make superconducting magnets that can do that,” says Cowley. The European Union has a research program, he notes, as does Brookhaven National Laboratory, which is pursuing the high-temperature materials for accelerator applications.
Cowley reckons that the pressures on the magnets in the MIT design would be two and a half times the crushing 1000 atmospheres at the bottom of the Mariana Trench, the deepest point in the ocean. “I think it is doubtful we will get to 23 T in 5 years. We might be able to go 10–20% above what ITER is doing in 5 years, and in 10 years maybe another 10–20%,” he says. He notes that ITER’s 11.8-T magnets, built to withstand 550 atmospheres, are pushing the current technological envelope.
But Sorbom says that the ARC reactor “is very structurally sound” and notes that modeling and finite element analysis indicated a 50% safety margin in the yield stress of the materials. “You put in more structure. You have to be clever about the shape of the coil to minimize the local stress. Typically you add in some superstructure to the coil,” adds Whyte.
Coppi also insists that Ignitor could withstand the forces, and he says that the new MIT design borrows structural components from his project.
Norbert Holtkamp, deputy director of SLAC, says that large magnets made with high-temperature superconductors are at least a decade away. That’s how long it took for first-generation niobium–titanium materials in Fermilab’s Tevatron and the Nb3Sn superconducting magnets in CERN’s Large Hadron Collider upgrade and planned for use in ITER, he notes. “It’s clear that for a lot of applications besides fusion, high-field magnets will make great advances in science, whether for high-energy physics, medical devices, or research devices,” says Holtkamp, a former principal deputy director general of ITER. “It’s a good path for development. It’s just a steep one and a difficult one.”
Another novel feature in the ARC design is the use of circulating molten salt in place of metal or other solid material for the reactor’s “blanket,” or inner wall. Materials are a major problem for reactors, since they tend to get brittle with a steady bombardment of 14-MeV fusion neutrons. Not only would the mixture of molten fluorine, lithium, and beryllium carry off neutron-generated heat to produce steam for electric turbines, the lithium-rich liquid also would breed a sufficient amount of tritium to sustain fusion, says Whyte. “Having a robust tritium breeding ratio is one of the most challenging engineering problems for a commercial reactor,” he notes.
The toroidal coils of the ARC reactor would have built-in joints to ease disassembly of the reactor. Says Sorbom, “You can lift the top off the machine and take the inside out to do maintenance.”
