Skip to Main Content
Skip Nav Destination

Investments in privately funded fusion ventures grow

13 October 2020

Five companies are betting they can achieve controlled thermonuclear fusion a decade or more before the $25 billion international ITER project.

General Fusion reactor.
General Fusion’s prototype reactor has pneumatically driven pistons that compress and heat a liquid metal that surrounds a plasma. Credit: General Fusion

The old adage is that fusion power is 30 years away and always will be. The timeline envisioned by the international collaboration for ITER—the colossal, $25 billion–plus reactor under construction in France—is only a little shorter than that. Attaining the first sustained fusion reaction, or burning plasma, is scheduled to occur in 2035 at the earliest. After that, an electricity-generating pilot plant will require about another decade to build.

That’s too long for the globe to wait for a clean new power source with a virtually unlimited supply of fuel, according to scientists and entrepreneurs working on alternative fusion power schemes. They say their machines could be ready to deliver electricity to the grid within a decade. Such promises have been made before, admits Michael Delage, chief technology officer of General Fusion in Vancouver, Canada, one of five fledgling private ventures that have attracted significant financing. “But I honestly believe that it is different this time. The underlying science and the ability to do modeling and simulation on how plasmas behave have really advanced a lot and out of the public eye.”

That doesn’t necessarily mean the companies are well on their way to success, says Dale Meade, former deputy director of the Princeton Plasma Physics Laboratory. He says codes have progressed incrementally over the last decade. “There have been improvements in the understanding, and we have been able to develop models that can replicate past experimental results, but have we improved our future predictability? We don’t know,” says Meade. “We need a burning plasma experiment to find out.”

In the coming years, the five companies will have to overcome engineering obstacles, surpass an energy-generating milestone that has eluded researchers for decades, and demonstrate that their reactors can serve as the basis for reliable, cost-effective power plants—while also attracting sufficient funding to move forward.

Financial infusion

Investments in private fusion ventures from billionaires, venture capital firms, philanthropists, and even oil and gas majors have grown in the past year. Measured by the amount of capital raised, TAE Technologies in Orange County, California, leads the pack at $750 million. That’s up by $150 million from early 2019, when named investors included the late former hedge fund manager Arthur Samberg, Charles Schwab, former Morgan Stanley CEO John Mack, venture capital firms New Enterprise Associates and Venrock, the UK’s Wellcome Trust, several sovereign funds, and Alphabet. CEO Michl Binderbauer says that list remains largely unchanged.

Illustration of TAE reactor.
A cross-section rendering of TAE Technologies’ Norman device. Plasmas are generated at each end and accelerated to merge at the center. Credit: TAE Technologies

Commonwealth Fusion Systems in Cambridge, Massachusetts, has raised more than $200 million, $84 million of it this year, mainly from venture capital firms, including the Bill Gates–backed Breakthrough Energy Ventures. Italian oil and gas producer Eni has contributed $50 million. CEO Bob Mumgaard says that the money should carry the 100-employee MIT spinoff through the end of 2021. Another 120 MIT employees are also working on the company’s test reactor.

In the UK, Tokamak Energy, started by former researchers at the Culham Centre for Fusion Energy, has attracted about £150 million ($193 million), including $26 million in R&D credits from the government. Investors include hedge fund managers and Legal and General, a UK financial services company. Under a $4 million cost-shared grant, Oak Ridge National Laboratory and Princeton Plasma Physics Laboratory will be working with the company on diagnostic techniques and plasma modeling, says David Kingham, Tokamak’s cofounder and executive vice chairman. It will cost about $1 billion to demonstrate the fusion process at grid scale, he says.

Jeff Bezos is the headline investor in General Fusion, which has raised $200 million in funding or commitments, according to Delage. About 80% of the company’s funding is from private sources; the Canadian government has provided the rest. Among its other publicly named investors are the Canadian oil and gas company Cenovus Energy, Singaporean investment firms Temasek and GIC, and the Malaysian sovereign wealth fund.

Zap Energy, based in Seattle, Washington, has been awarded $8 million in grants from the Department of Energy’s Advanced Research Projects Agency–Energy. The University of Washington spinoff has raised another $6.5 million from investors that include Chevron Technology Ventures and Chris Sacca’s Lowercarbon Capital. Benj Conway, Zap’s president, says the money raised “will take us to the precipice” of achieving break-even conditions and demonstrating gain—producing more energy from a fusion reaction than was required to initiate it.

Variations on the tokamak

Commonwealth and Tokamak Energy are developing variants of the tokamak, a toroidal-shaped reactor that confines plasmas within magnetic fields. That design is the focus of most of the world’s government-supported fusion programs, including ITER (see the article by Richard Hawryluk and Hartmut Zohm, Physics Today, December 2019, page 34). Both companies will use magnets fashioned from high-temperature superconducting (HTS) tape to produce the strong magnetic fields needed to confine plasmas of more than 100 million kelvin in reactors that are a small fraction of ITER’s size. Commonwealth is working on a doughnut-shaped toroid, whereas Tokamak Energy is developing a spherical tokamak that resembles a cored apple (see Physics Today, August 2018, page 25).

Using 500 km of yttrium barium copper oxide HTS tape, Commonwealth plans to build by next June what Mumgaard says will be by far the world’s largest high-field HTS coil, producing an expected field of 20 tesla. It will be a full-scale version of the magnets to be used in Commonwealth’s test reactor, construction of which is scheduled to begin early next year. Commonwealth expects to demonstrate by 2025 a fusion reaction producing an energy gain, a feat that no one has yet accomplished. Mumgaard says the machine, known as SPARC, is expected to cost around $400 million.

A set of research papers published in a 29 September special issue of the Journal of Plasma Physics offers peer-reviewed support for Commonwealth’s plans. The papers were authored by 47 researchers from 12 institutions, including many from Commonwealth and MIT’s Plasma Science and Fusion Center, SPARC’s birthplace. The studies express confidence in the reactor’s plasma physics basis and performance predictions, says Martin Greenwald, deputy director of the MIT center who is a cofounder of Commonwealth and one of its leading scientists. SPARC should achieve its goal of an energy gain of at least 2, and possibly attain ITER’s expected gain of 10, the papers say. The nearest that any device has come to energy gain is the Joint European Torus (JET) in the UK, which got two-thirds of the way to the threshold. Commonwealth expects SPARC’s demonstration will attract investments from utility companies to finance a follow-on electricity-generating reactor.

But Meade says he is skeptical of Commonwealth’s cost and schedule estimates. He notes that Princeton’s Tokamak Fusion Test Reactor (TFTR), one of the two tokamaks in the world to have experimented with tritium, cost $1 billion in today’s dollars, and it took 10 years to ready for producing high-temperature plasmas in 1986. Given the additional development and cost of SPARC’s superconducting magnets, Meade expects SPARC to cost $1.5 billion to $2 billion and believes it will not be ready for tritium experiments until 2033, eight years past the company’s own timeline. But he says Commonwealth’s attempt is “still worth doing. I’ve argued all along that we needed another deuterium–tritium burning plasma experiment in parallel with ITER.”

Tokamak Energy has achieved a 2 T field in a plasma confined in its ST-40 spherical tokamak using conventional magnets. That’s double the top field strength attained by the biggest publicly funded spherical tokamak, Princeton’s National Spherical Torus Experiment Upgrade, says Kingham. The company expects to boost that performance to 3 T and 100 million K in ST-40 by the end of the year. HTS magnets will be incorporated in its next-generation tokamak, which is due for completion in 2025. That machine is expected to demonstrate energy gain, Kingham says, from a plasma with a 5 T field at 150 million K.

Unconventional approaches

General Fusion, TAE Technologies, and Zap Energy are pursuing more exotic approaches. General Fusion’s technology combines elements of magnetic and inertial fusion. A pulsed-power source generates a plasma that’s injected into a cavity surrounded by liquid lead–lithium. Before the plasma can decay, pneumatically driven pistons compress the liquid metal over 20 milliseconds, heating the plasma inside to fusion temperatures. Whereas laser-driven inertial fusion requires a compression ratio of about 40 to 1 for break-even conditions, General Fusion’s technology should work at a 7:1 ratio, says Delage. The nearly perfect symmetry required for laser-driven implosions will be greatly relaxed as well.

Scheduled for completion by 2025, the device is expected to achieve power-plant plasma temperatures and attain 10% of the conditions, known as the Lawson criteria, needed for break-even. From there, General Fusion plans to build a machine at power plant scale.

Zap Energy’s “sheared-flow stabilized Z-pinch” process should yield electricity-generating reactors that are “orders of magnitude cheaper” than fusion reactors requiring magnetic coils, says Conway. Reactors producing 200 MW thermal power would also be very small compared with tokamaks, says CTO Brian Nelson. Any number of those modules could be located with common ancillary facilities, such as a plant for handling and separating tritium. Some modules could be shut down when solar or wind power can meet grid requirements, or for maintenance. Steady-state tokamaks, by contrast, would need to be operated continuously.

Zap Energy prototype.
Zap Energy’s device is designed to extend the lifetime of lightning-like Z-pinch plasma. Credit: Zap Energy

The Z-pinch has been pursued for decades, notably at Sandia National Laboratories’ Z machine. But the lightning-like plasmas at Z last only a few nanoseconds before being ripped apart by instabilities, says Nelson. Zap’s sheared-flow technology extends the lifetime to 20 microseconds by causing the current to flow between electrodes at different velocities across the radii of a cylindrical plasma. Zap power reactors, which the company says could be available this decade—Conway declines to be more specific—would be pulsed at a few hertz. A circulating liquid lead–lithium blanket would carry the heat to steam turbines while also breeding tritium and protecting reactor components from neutrons.

Over three years Zap has quadrupled the electric current through the Z-pinch plasma and is now less than a factor of two from the level needed for breakeven, says Conway. He and Nelson say their goal is to reach break-even in three years.

In TAE’s scheme, plasmas are created at opposite ends of a linear device and accelerated to collide, forming a larger, hotter plasma. Beam injectors supply angular momentum to stabilize the football-shaped plasma. (See Physics Today, February 2019, page 28.) TAE is finishing up experiments on a test device known as Norman and ordering components for its next reactor, to be named Copernicus. Like its competition, TAE aims to fuse the hydrogen isotopes deuterium and tritium, although the company ultimately plans to combine protons and boron-11 in a reaction that would not produce any neutrons. “We are within a factor of two or so of where we need to be for D–T conditions,” says Binderbauer. “I’m convinced we will be able to do that with Copernicus.” A third planned reactor, DaVinci, a “nearly commercial-scale power plant,” is to begin operating in the latter part of the decade, he adds.

TAE won’t be using tritium in its new device. Plasmas will be composed of ordinary hydrogen, and experimental results will be extrapolated to determine how D–T will perform. Binderbauer says the regulatory burden of experimenting with tritium would delay R&D and add unnecessary cost.

To date, Princeton’s TFTR and the UK’s JET are the only non-weapons facilities to have run fusion experiments with tritium. But the heavy hydrogen isotope is used in biomedical applications, says Commonwealth’s Mumgaard, and the licensing process is comparable to that required for a cancer treatment center. A Wisconsin company, SHINE Medical Technologies, is anticipating Nuclear Regulatory Commission approval for its plan to use tritium to produce the medical isotope molybdenum-99. The licensing process will take two years from the company’s application submission.

Common challenges

All D–T fusion power approaches require materials that can withstand constant bombardment from the high-energy neutrons produced in the reaction. Several of the companies are counting on liquid metal blankets surrounding the plasma to absorb neutrons, breed the large quantities of tritium that will be needed for sustained fusion burns, and transfer the heat from the reaction chamber to generate steam. Zap and General Fusion have selected liquid lead–lithium.

Commonwealth proposes molten salt, similar to what is used to transfer heat in concentrated solar arrays, for its thermal and electrical conductivity properties. A liner between the plasma and molten salt would need to be replaced periodically. “We think it’s exciting that we’re at the stage that we are thinking about things like [blankets],” says Mumgaard, noting that reports from the National Academies of Sciences, Engineering, and Medicine and the DOE’s Fusion Energy Sciences Advisory Committee have both recommended more R&D on blankets and materials.

Other issues that are far from being resolved are removing the alpha particles that build up in the plasma and could choke off fusion, and making a liquid metal flow through the reactor. One of the challenges facing Commonwealth and Tokamak Energy is maintaining their superconducting magnets at 20 K in compact devices next to plasmas of 100 million K or more while preventing the energetic fusion neutrons from heating the superconducting coils.

Crucially, the companies will have to prove they can increase the duration and energy gain of the fusion reaction. It’s a big step to go from experiments with tritium lasting a few seconds to breeding the large quantities of tritium in the reactor that can sustain a fusion reaction for months or years. And Meade says the small energy gains of 1–2 times the input energy that the five companies are aiming for pales in comparison to the 30× gain that will be needed for a power plant to be commercially attractive.

Although convinced they will beat ITER to the punch, developers say the international effort should continue. “The economics of extrapolating ITER-scale tokamaks to the power scale looks pretty challenging, but that doesn’t mean the tokamak is an invalid concept,” says Delage. “I don’t think it’s the time to be taking any of the options off the table.”

ITER also serves as a political unifying force. “The first commercial reactor doesn’t mean that fusion research should stop,” says Kingham. “There will be lots of scope for further improvements in subsystems, diagnostics, and plasma control. It’s not realistic for us to do it all ourselves.”

Editor’s note, 14 October: The article has been updated to correct the amount of money raised, number of employees, and amount of high-temperature superconducting tape at Commonwealth Fusion Systems.

Close Modal

or Create an Account

Close Modal
Close Modal