Now, at last, one company is working to make buried reactors a reality.
Deep Fission proposes digging boreholes 30 inches in diameter and about a mile deep to house each of its 15-megawatt reactors. And it’s making progress. In August, the Department of Energy selected Deep Fission as one of the 10 companies enrolled in the agency’s new reactor pilot program, meant to help next-generation startups split their first atoms by July. In September, the company announced a $30 million reverse merger deal with a blank check firm to make its stock market debut on the lesser-known exchange OTCQB. Last month, Deep Fission chose an industrial park in a rural stretch of southeastern Kansas as the site of its first power plant.
Based in Berkeley, California, the one-time hub of the West Coast’s fading anti-nuclear movement, the company says its design is meant to save money on above-ground infrastructure by letting geology do the work to add “layers of natural containment” to “enhance safety.” By eliminating much of that expensive concrete and steel dome that encases the reactor on the surface, the startup estimates “that our approach removes up to 80% of the construction cost, one of the biggest barriers for nuclear, and enables operation within six months of breaking ground.”
“The primary benefit of placing a reactor a mile deep is cost and speed,” Chloe Frader, Deep Fission’s vice president of strategic affairs, told me. “By using the natural pressure and containment of the Earth, we eliminate the need for the massive, above-ground structures that make traditional nuclear expensive and slow to build.”
“Nuclear power is already the safest energy source in the world. Period,” she said. “Our underground design doesn’t exist because nuclear is unsafe, it exists because we can make something that is already extremely safe even safer, simpler, and more affordable.”
But gaining government recognition, going public, and picking a location for a first power plant may prove the easy part. Convincing others in the industry that its concept is a radical plan to cut construction costs rather than allay the public’s often-outsize fear of a meltdown has turned out to be difficult, to say nothing of what actually building its reactors will entail.
Despite the company’s recent progress, I struggled to find anyone who didn’t have a financial stake in Deep Fission willing to make the case for its buried reactors.
Deep Fission is “solving a problem that doesn't actually exist,” Seth Grae, the chief executive of the nuclear fuel company Lightbridge, told me. In the nearly seven decades since fission started producing commercial electrons on the U.S. grid, no confirmed death has ever come from radiation at a nuclear power station.
“You’re trying to solve a political problem that has literally never hurt anyone in the entire history of our country since this industry started,” he said. “You’re also making your reactors more expensive. In nuclear, as in a lot of other projects, when you build tall or dig deep or lift big and heavy, those steps make the projects much more expensive.”
Frader told me that subterranean rock structures would serve “as natural containment, which also enhances safety.” That’s true to some extent. Making use of existing formations “could simplify surface infrastructure and streamline construction,” Leslie Dewan, a nuclear engineer who previously led a next-generation small modular reactor startup, told IEEE Spectrum.
If everything pans out, that could justify Deep Fission’s estimate that its levelized cost of electricity — not the most dependable metric, but one frequently used by solar and wind advocates — would be between $50 and $70 per megawatt-hour, lower than other SMR developers’ projections. But that’s only if a lot of things go right.
“A design that relies on the surrounding geology for safety and containment needs to demonstrate a deep understanding of subsurface behavior, including the stability of the rock formations, groundwater movement, heat transfer, and long-term site stability,” Dewan said. “There are also operational considerations around monitoring, access, and decommissioning. But none of these are necessarily showstoppers: They’re all areas that can be addressed through rigorous engineering and thoughtful planning.”
As anyone in the geothermal industry can tell you, digging a borehole costs a lot of money. Drilling equipment comes at a high price. Underground geology complicates a route going down one mile straight. And not every hole that’s started ends up panning out, meaning the process must be repeated over and over again.
For Deep Fission, drilling lots of holes is part of the process. Given the size of its reactor, to reach a gigawatt — the output of one of Westinghouse’s flagship AP1000s, the only new type of commercial reactor successfully built from scratch in the U.S. this century — Deep Fission would need to build 67 of its own microreactors. That’s a lot of digging, considering that the diameters of the company’s boreholes are on average nearly three times wider than those drilled for harvesting natural gas or geothermal.
The company isn’t just distinguished by its unique approach. Deep Fission has a sister company, Deep Isolation, that proposes burying spent nuclear fuel in boreholes. In April, the two startups officially partnered in a deal that “enables Deep Fission to offer an end-to-end solution that includes both energy generation and long-term waste management.”
In theory, that combination could offer the company a greater social license among environmental skeptics who take issue with the waste generated from a nuclear plant.
In 1982, Congress passed a landmark law making the federal government responsible for the disposal of all spent fuel and high-level radioactive waste in the country. The plan centered on building a giant repository to permanently entomb the material where it could remain undisturbed for thousands of years. The law designated Yucca Mountain, a rural site in southwestern Nevada near the California border, as the exclusive location for the debut repository.
Construction took years to start. After initial work got underway during the Bush administration, Obama took office and promptly slashed all funding for the effort, which was opposed by then-Senate Majority Leader Harry Reid of Nevada; the nonpartisan Government Accountability Office clocked the move as a purely political decision. Regardless of the motivation, the cancellation threw the U.S. waste disposal strategy into limbo because the law requires the federal government to complete Yucca Mountain before moving on to other potential storage sites. Until that law changes, the U.S. effort to find a permanent solution to nuclear waste remains in limbo, with virtually all the spent fuel accumulated over the years kept in intermediate storage vessels on site at power plants.
Finland finished work on the world’s first such repository in 2024. Sweden and Canada are considering similar facilities. But in the U.S., the industry is moving beyond seeing its spent fuel as waste, as more companies look to start up a recycling industry akin to those in Russia, Japan, and France to reprocess old uranium into new pellets for new reactors. President Donald Trump has backed the effort. The energy still stored in nuclear waste just in this country is sufficient to power the U.S. for more than a century.
Even if Americans want an answer to the nuclear waste problem, there isn’t much evidence to suggest they want to see the material stored near their homes. New Mexico, for example, passed a law barring construction of an intermediate storage site in 2023. Texas attempted to do the same, but the Supreme Court found the state’s legislation to be in violation of the federal jurisdiction over waste.
While Deep Fission’s reactors would be “so far removed from the biosphere” that the company seems to think the NRC will just “hand out licenses and the public won’t worry,” said Nick Touran, a veteran engineer whose consultancy, What Is Nuclear, catalogs reactor designs and documents from the industry’s history.
“The assumption that it’ll be easy and cheap to site and license this kind of facility is going to be found to be mistaken,” he told me.
The problem with nuclear power isn’t the technology, Brett Rampal, a nuclear expert at the consultancy Veriten, told me. “Nuclear has not been suffering from a technological issue. The technology works great. People do amazing things with it, from curing cancer to all kinds of almost magical energy production,” he told me. “What we need is business models and deployment models.”
Digging a 30-inch borehole a mile deep would be expensive enough, but Rampal also pointed out that lining those shafts with nuclear-grade steel and equipping them with cables would likely pencil out to a higher price than building an AP1000 — but with one one-hundredth of the power output.
Deep Fission insists that isn’t the case, and that the natural geology “removes the need for complex, costly pressure vessels and large engineered structures” on the surface.
“We still use steel and engineered components where necessary, but the total material requirements are a fraction of those used in a traditional large-scale plant,” Frader said.
Ultimately, burying reactors is about quieting concerns that should be debunked head on, Emmet Penney, a historian of the industry and a senior fellow at the Foundation for American Innovation, a right-leaning think tank that advocates building more reactors in the U.S., told me.
“Investors need to wake up and realize that nuclear is one of the safest power sources on the planet,” Penney said. “Otherwise, goofy companies will continue to snow them with slick slide decks about solving non-issues.”
The three shades of gray in the center of the continent shows the path of smoke from the Canadian fires.MyRadar
Information under the smoke tab on the AirNow app.AirNow
The FireSmoke Canada website tracks PM2.5 with a handy animation but does not offer information about individual fires like MyRadar and AirNow do.FireSmoke Canada
A chart showing average monthly spot prices for Brent crude oil throughout 2025.EIA
A chart showing average annual spot prices for Brent crude oil throughout from 2020 to 2025.EIA
