Nobody said building a fusion power plant would be easy. Physicists and engineers have been working for decades to crack the problem. But over the last year or so, fusion startup Zap Energy took a deeper look at its pathway to a working power plant and decided that it would be quicker to build a fission power plant first.
Wait, what?
“Fission and fusion are two sides of the same coin,” Zap’s new CEO, Zabrina Johal, told TechCrunch. “They have so many challenges that are congruent with each other.”
Zap is among the better-funded fusion startups, having raised more than $300 million, so this partial pivot holds some shock value, no matter how many synergies exist between fission and fusion.
It starts to make more sense against the backdrop of rising energy demand from AI data centers, which is expected to nearly triple by 2030. Tech companies want electricity today, and one of the challenges facing every fusion startup is that grid-ready power plants won’t be ready for several more years — likely a decade or more.
“There is not enough power and energy in the world to build all the data centers that are needed,” Johal said. “It just meant we need to pull this in faster; we need to get something that’s relevant to the grid today.”
Two ways to split an atom
Fission is commercially viable in a way that fusion is not. Fusion is the practice of fusing two light atoms like hydrogen, which also releases energy. One experiment has been able to produce more energy than the fusion reaction needed to ignite, but it wasn’t anywhere close to what a power plant would need to generate. Fission splits heavy atoms like uranium to produce power, and we’ve been doing that since the 1950s.
Techcrunch event
San Francisco, CA
|
October 13-15, 2026
Despite decades of experience, building fission reactors cost-effectively remains a significant challenge. Fission startups building small modular reactors (SMR) are counting on mass manufacturing to help bring costs down, though that theory has yet to be proven. Benefits from scaling production can take around a decade to materialize.
Johal said that Zap expects to start generating revenue from the new fission business within a year. “Our business model is not dependent upon generating electrons,” she said. Revenue could come from federal programs from the Department of Defense and the Department of Energy, but it could also include “milestone payments” and reserved production capacity from companies that need massive amounts of electricity, she said.
Milestone payments could be an intriguing model for Zap and other energy startups to follow.
It’s similar in concept to how ASML extracted money from Intel, TSMC, and Samsung to develop extreme ultraviolet lithography (EUV). The semiconductor manufacturers effectively paid a premium for ASML shares, underwriting R&D in the technology and reserving capacity once EUV machines entered production.
But there’s a fundamental difference between what Zap is attempting and what ASML pulled off. When ASML ginned up its “Customer Co-Investment Program for Innovation,” it was clear the Dutch company was the only show in town — everyone else had given up on EUV. In the energy world, tech companies have a range of different technologies and suppliers to pick from. They’ll want to see something extra special in Zap’s fission proposal before they pony up.
On that front, potential buyers can already start assessing Zap’s plans. The startup’s fission reactor will be based on the 4S, a molten salt-cooled design that was jointly developed by Toshiba and Japan’s power industry research institute. Ultimately, it was never built, but Johal said the design comes with “no intellectual property entanglement.”
Johal expects there will be enough demand in the 2030s that Zap will find plenty of customers, despite being years behind other fission startups. “There will not be enough reactors in the near term,” she said.
Follow the money
For Zap’s fission gambit to pay off, one of two things needs to occur: It’ll have to bring in revenue or bring in new investment.
Given Johal’s comments on government funding and milestone payments from large energy users, revenue is the obvious play. The cost of developing one reactor concept is eye-wateringly high. The cost of developing a second may not be double, but it’s almost certainly not free. The more cash the better.
Zap isn’t the only fusion company to pursue side businesses to bring in revenue. Commonwealth Fusion Systems and Tokamak Energy are selling its high-temperature superconducting magnets to other fusion companies and experiments, while others like TAE and Shine Technologies are in nuclear medicine.
Some of those revenue opportunities are more aligned with building a fusion power plant than others. Zap argues that its fission plan will help it move faster on everything but the fusion reactor itself, including things like materials testing and power systems. The company also argues that it can gain experience in regulatory domains, though Johal said this is more about building relationships with regulators than navigating specific rules. The Nuclear Regulatory Commission, a cautious government agency, has provided fusion companies with a separate set of guidelines. For all their similarities, fusion and fission are still very different technologies.
Or maybe Zap won’t need new revenue if it can attract a new class of investors. If Zap can tap into enthusiasm for fission startups, maybe it can find an exit for existing investors sooner. For example, X-energy, which has yet to build a power plant, went public last week in an upsized IPO that brought the company $1 billion.
Much of this assumes that Zap will be able to show progress on connecting a small modular reactor (SMR) to the grid in the early 2030s.
Zap’s arguments that adding fission to its plate will help it reach commercial fusion power sooner are compelling, but time may prove me wrong. Still, it’s hard to square those ambitions with the challenges — and costs — of building a second reactor based on a very different technology. There are enough similarities to prevent this from being a one-eighty, but it’s far enough from Zap’s previous path that it will need to tread carefully to ensure it doesn’t turn into a permanent detour.
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.
