February 24th, 2021
Plans to build a prototype fusion power plant in the United States have come into tighter focus, as a new report lays out a rough timeline for building the multibillon-dollar plant and a strategy for developing its design. The United States should strive to start construction of the pilot by 2035 and to have it running by 2040, according to a report released this week by the National Academies of Sciences, Engineering, and Medicine (NASEM). To meet that tight schedule, the report calls for the U.S. Department of Energy (DOE) to help fund two to four teams that, in collaboration with private industry, would develop by 2028 different conceptual designs.
“It’s credible and doable,” says William Madia, vice president emeritus at Stanford University, who has often been critical of DOE’s fusion efforts. However, Edwin Lyman, director of nuclear power safety at the Union of Concerned Scientists, notes the report also points to numerous key technologies that are in a low state of technical readiness and questions whether they can be developed in time. “Reading between the lines, I didn’t feel that it gives you a lot of confidence that these time tables are realistic,” Lyman say.
The schedule suggested by the 91-page report, released on 17 February, does not reflect a bottom-up assessment of how long it would take to complete the R&D for a particular design. Rather, it presents a top-down evaluation of when fusion power needs to be feasible if utilities want to include it as they shift toward carbon-free energy sources by 2050. “The timeline really was determined from input from utilities,” says Brian Wirth, a nuclear engineer at the University of Tennessee, Knoxville, and one of the report’s 12 authors. Madia predicts that if fusion hasn’t shown itself feasible by 2035, it is “going to miss the train” and be left out of the future carbon-free energy mix.
Enthusiasm has been growing among U.S. fusion researchers to build a prototype power plant that would finally harness nuclear fusion, the process that powers the Sun, to generate electricity. The idea emerged in 2018 in another NASEM report that suggested the United States stick with its commitment to ITER, the gigantic international fusion experiment under construction in France, but also prepare to build a domestic pilot plant soon after. In December 2020, U.S. fusion researchers embraced the pilot plant in their new long-range plan.
Using intense magnetic fields, ITER will trap a plasma of deuterium and tritium—the heavy isotopes of hydrogen—in a doughnut-shaped vacuum chamber and heated to 150 million degrees Celsius. The experiment aims to show that fusion of the deuterium and tritium nuclei can produce more energy than generating the plasma consumes. The pilot plant would go a step further and produce useful amounts of electricity.
Supporters of the proposal say building a pilot plant would enable U.S. researchers to leap-frog competitors in Europe and Asia and reclaim the lead in fusion research. Moreover, they say, because of recent advances in magnets made out of high-temperature superconductors, 3D printing, and computer modeling, the plant could be much smaller and less expensive than ITER, which stands 30 meters tall and will cost in excess of $20 billion.
The pilot plant wouldn’t be overwhelmingly powerful. The new report envisions it producing a mere 50 megawatts of electrical power, a few percent of what a large gas-powered plant can generate. That’s just enough to see how such a plant would interact with the electrical grid, says David Roop, an engineer with DWR Associates LLC and an author of the new report. “This value of at least 50 megawatts allows you to test your ability to move power out onto the grid,” he says. The pilot plant should cost no more than $5 billion or $6 billion, the report says, which is the amount it estimates utilities would be willing to pay for the first commercial plant to follow, which would generate hundreds of megawatts of power.
Before any pilot plant can be built, several large technological challenges must be overcome, the report says. For example, if the plant works like ITER and fuses deuterium and tritium (there are a few other possibilities), then researchers must also develop a way to breed more tritium in a purpose-built “blanket” of material surrounding the reactor. Otherwise the plant could quickly consume the world’s entire supply of tritium, which comes only from certain nuclear reactors. If the blanket contains lithium, the neutrons will split some of those nuclei to form more tritium. But tritium is a highly regulated radioactive substance that’s difficult to handle, and Lyman questions whether such a closed system can be developed by 2035.
The debate over the timeline reflects tensions over the advent of new private fusion companies, which would presumably partner with DOE to develop the ideas they already have. Some startups have attracted the interest of high-tech billionaires. For example, Bill Gates is backing Commonwealth Fusion Systems, whose ideas focus on using magnet coils made of high-temperature superconductors, and Jeff Bezos is backing General Fusion, which uses mechanical pumps to squeeze and heat a plasma. Such private investment is a key reason Madia says he’s bullish on the pilot plant. “When guys like Bill Gates and Jeff Bezos get involved, they do interesting stuff,” he says.
But fusion startups are promising more than they can deliver, Lyman argues. He worries investors familiar with the high-tech economy, in which failure may not be fatal, may be taking the same high-risk attitude toward capital-intensive energy research, which is far less forgiving. “They think that they can just have an idea and get rich off the idea without actually having it fulfilled,” Lyman says. Developing a fusion reactor “is not something you can do in your garage, so you have to be cautious about that mindset.” Lyman also warns that fusion startups are asking the Nuclear Regulatory Commission to relax safety requirements in ways he says are not justified.
Still, the report’s call to form public-private teams to flesh out the various concepts for a pilot plant would be highly beneficial, says Steven Cowley, director of the Princeton Plasma Physics Laboratory, DOE’s fusion lab. “The U.S. needs to have its own version of designing a fusion reactor to see what it is and how close we are,” Cowley say. “I think there’s huge a value in that.”
Written by: Adrian Cho
Adrian Cho is a staff writer for Science.