The race to turn nuclear fusion into a commercial energy source is attracting billions of dollars, but the technology’s biggest challenge may no longer be proving that fusion works — it is proving that it can produce reliable, affordable electricity at scale.
That is the central tension highlighted in a Financial Times analysis of the rapidly expanding private fusion industry, where technology companies, billionaires and energy groups are backing increasingly ambitious plans to commercialise the reaction that powers the Sun.
In late August, politicians and investors gathered in the New Mexico heat to break ground on a fusion test facility being built by Pacific Fusion in Albuquerque. Eric Schmidt, the former Google chief executive and an investor in the company, said that if fusion eventually generated most of America’s electricity, “it solves all the problems” created by the energy demands of AI.
The financial backing is already substantial. Fusion start-ups attracted a record $3.8bn in private funding during the first eight months of this year, surpassing the previous annual record of $3.3bn. The number of private fusion companies has also more than doubled over five years to more than 50, according to the Financial Times analysis, citing the Fusion Industry Association.
Yet no fusion experiment has produced more electricity than the facility itself consumes, and no fusion power plant has demonstrated reliable commercial operation. Alain Bécoulet, chief scientist at the international ITER project, says connecting a fusion plant to the grid “will never happen” within the next five to 10 years.
The scientific milestone achieved at the US National Ignition Facility in 2022 illustrates the gap. The experiment generated 3.15MJ of fusion energy, but the laser system consumed about 422MJ of electricity.
The Financial Times also points to engineering obstacles. Fusion reactors must withstand extreme heat and radiation, while producing their own tritium fuel remains a challenge. Mohamed Abdou of UCLA estimates that first-generation plants could suffer component failures frequently, with replacements taking “many months”.
Economics present another uncertainty. Proposed plants would consume substantial amounts of electricity to create and sustain plasma, potentially leaving a much smaller share available for sale to the grid. General Fusion estimates lifetime costs of $64-$73 per MWh, while Commonwealth Fusion Systems has a guideline target of about $50/MWh.
As fusion companies seek increasingly large sums from governments, pension funds, sovereign wealth funds and public markets, questions are also emerging over whether investors can properly assess their claims. One investor says many newcomers are still on an “early learning curve”.
The industry’s long-standing credibility problem remains. Fusion has been described as “three decades away and always will be”.
Princeton’s Steven Cowley says researchers have managed to “put the Sun in a bottle”, but adds a crucial qualification: “We haven’t put it in a commercial bottle.”
By Sabina Mammadli
