Just as Commonwealth Fusion Systems, a company spun off from the Massachusetts Institute of Technology in 2018, was announcing that its mission to build a compact fusion power plant (the SPARC) based on the ARC tokamak design was nearing completion, a team of engineers from ETH Zurich (the Swiss Federal Institute of Technology) were throwing cold water on the idea that fusion power plants (FPPs) can be cost-competitive in future net-zero energy systems.
Tobias Schmidt, founding director of the Albert Einstein School of Public Policy at ETH Zurich, lead author Lingzi Tang, and two others published their findings in Nature Energy on March 23. They concluded that the large unit size, extraordinary complexity, and intermediate need for customization of FPPs are empirically linked to experience rates (cost reductions for additional units) of 2% to 8% rather than the industry’s estimates of 8% to 20% – and that, paired with expected high initial capital costs, spells financial doom for the fledgling industry.
Schmidt told Techxplore.com that, after hearing promises from “some actors in the fusion space” of extremely low levelized costs for FPPs, his team applied to fusion an ETH framework that analyzes why some technologies learn with higher experience rates than others. The team compared magnetic fusion (which prompts fusion by confining hot plasma using powerful magnetic fiels) and inertial fusion (which works by compressing fuel using lasers).
Casey Crownheart, writing for MIT Technology Review, said the ETH team asked fusion experts to evaluate FPPs on size, complexity, and customization to predict an experience rate. Tang told Crownheart that “there was almost unanimous agreement that fusion is incredibly complex,” and that reaching the experience rate of 23% for solar modules, 20% for lithium-ion batteries, or even 12% for onshore wind power was highly unlikely.
Tang believes it would take a lot of deployment and quite a long time for the price of building an FPP to drop significantly – and thus that “questions should be raised about current investment levels in fusion,” especially investments by the federal government. “If you’re talking about decarbonization of the energy system,” said Tang, “is this really the best use of public money?”
Not coincidentally, CFS is building its prototype FPP – the SPARC (smallest possible affordable, robust, compact) tokamak – in collaboration with MIT at its Devers, Massachusetts, facility, with completion now scheduled for 2027. CFS CEO Bob Mumgaard says it will be the world’s first commercially relevant fusion energy machine to produce more energy from fusion than it needs to power the process – or net energy generation (Q>1).
Mumgaard went on the record with Reuters’ Tim Gardner just days after the ETH Zurich report was published and spoke extensively about the progress his company is making toward its goal of building a larger FPP in Virginia starting in this decade. Mumgaard noted that the ETH Zurich authors are unaffiliated with fusion (Schmidt even said they would not be doing further studies on fusion costs). These academics did not contact him or anyone who is building “anything.”
“When you actually look at … the speed at which we’re able to build our things, the ability that we’ve actually shown to decrease the cost of every component that goes into it, those are following well-known industrial trends that are not from the 1950s or 1960s but are from today. And so we remain very bullish that power bills in Virginia in the 2030s will include fusion.”
Mumgaard says the scaling of costs for FPPs depends not on “the right geographies” or “the right contracts” or even the various flows of big systems but rather on how quickly you can manufacture plants and install them. France switched to nuclear energy in just two decades, and it is even possible that fusion can go significantly faster – possibly as rapidly as what we are seeing in the digital world.
A CFS spokesperson added that, while “we appreciate that researchers are looking into the economics of fusion power plants, the [ETH Zurich study] makes some assumptions that we believe are flawed – notably examining learning rates from the top down instead of at the component level and overestimating customization costs. We track our economics very carefully … and we fully expect to be competitive with other energy sources.”
Mumgaard, who was just appointed as the first representative of the fusion industry to the President’s Council of Advisors on Science and Technology (PCAST), is ebullient about fusion’s – and his company’s – future, much of which depends on the success of the SPARC reactor. The appointment emphasizes the realization that fusion has moved from basic science to applied science to demonstrations that we understand how the science works.
The planned commercial fusion power plant (the ARC), to be sited in Chesterfield County, Virginia, is a partnership with Dominion Energy. CFS anticipates first deliveries of up to 400 megawatts of electricity in the 2030s – with Google and ENI as its first customers. But Mumgaard says CFS is planning to build multiple ARC units in rapid fire if capital is available and the markets are right.
In so doing, says Mumgaard, you learn as you go from one plant to another and eventually get to the point where you believe you have it all sorted out – the technology works, the supply chain exists, the design is refined enough to build units in parallel. The scaling is enhanced because the tritium fuel used in tokamak reactors is regenerative.
Mumgaard says all you need to begin endless production of tritium is “a little starter, sort of like a sourdough starter.” SPARC’s tritium came from a fission power plant but can make its own. Any tritium that leaks out we recapture and reuse; it’s well understood how to contain tritium and well written into the regulations.
While CFS has raised billions in private funding, Mumgaard joined the Fusion Industry Association in calling for a $10 billion federal infusion [denied in 2026] to expedite the deployment of fusion power plants in the U.S. The argument was that, while a decade ago the U.S. was shutting down its facilities and sending its experts to help China bring its own facilities online, things have changed and now it is urgent that the U.S. catch up and surpass the Chinese.
Mumgaard envisioned federal dollars going into three pots – research at the national science laboratories and universities to build the foundations for a future fusion industry; helping the first commercial plants get up and running; and translational research that helps move laboratory innovations into the supply chain and into power plants.
If Mumgaard – and CFS’ investors – are right, the ARC is generating electricity at competitive prices ten years from now and a new industry could begin rapid growth. Success would also bring in new investors – and new customers to help CFS (and perhaps other fusion companies) to turn a profit.
Once fusion energy achieves commercial success, Mumgaard envisions a high demand, especially in countries like Japan, South Korea, and Singapore that are attracted to an energy source that uses so little land and natural resources, ends dependency on LNG delivered by tankers, and avoids the threats of nuclear weapons being cobbled together from plutonium.
If that happens, China’s massive public sector investment in fusion might not matter.
This piece originally appeared at RealClearEnergy.com and has been republished here with permission.
