LEMONT, Ill. — Down a bright fluorescent hallway and through a secure door in one of the United States’ most storied federal research institutes is a room where only American citizens can go. It’s small and dimly lit, a rather modest workshop. Entering means surrendering your phone and camera; exiting requires first scanning your body for radiation. Much of the space is taken up by a blue, windowed box that’s about the size of three SUVs stacked on top of each other. Dozens of thick black gloves hang off the transparent walls.
Back in 2018, the year United Nations scientists warned that the window to avoid catastrophic global warming was quickly closing without dramatic cuts to fossil fuel use, Krista Hawthorne slid her hands into those gloves and, suddenly, everything made sense.
“Wow,” Hawthorne, 34, remembered thinking as she curled her fingers, Thanos-like, around a fist-sized chunk of ancient stardust metal. “I really get to do this.”
That single silvery sample of uranium contained roughly enough energy to power a single U.S. home for six centuries. But a traditional nuclear reactor could only tap into a fraction of that cosmic power before rendering the atomic fuel too contaminated with fission byproducts to be considered anything but dangerous waste.
Hawthorne, a young electrochemical engineer just starting her job at the Argonne National Laboratory here in the quiet Chicago suburbs, was determined to find a solution. If we just get to the 95% of energy left over in spent uranium fuel, it would be easier to stop burning so many planet-heating fossil fuels. And using more of that energy would shorten the period the final waste product needed to decay back to the radiation levels of uranium mined from the Earth.
“We’re looking at ways to recycle used nuclear fuel because it reduces the amount of waste that has to be disposed of,” Hawthorne said in a conference room outside her lab one morning last October. “It also decreases the amount of time the waste has to be isolated for, from about 300,000 years to about 300 years. And it provides — ”
A stoic scientist who chooses her words carefully, she corrected herself mid-sentence: “It would provide a sustainable source of fuel for advanced reactors.”
She may not need the conditional verb tense for long. Over the past year, the U.S. Department of Energy and a California-based nuclear reactor startup called Oklo have been jointly funding her experiments in a technique known as “pyroprocessing” in hopes of bringing fuel recycling to market. Last month, Oklo unveiled the proposal it submitted to the U.S. Nuclear Regulatory Commission to build the United States’ first commercial fuel recycling plant.
“There’s enough energy content in used fuel in the U.S. alone to power the country for 150 years, and that number is going to grow,” said Oklo chief executive Jacob DeWitte. “Each year, we produce enough waste to power the country for four years. It’s effectively an inexhaustible fuel. We’re taking a huge burden and liability and turning it into complete decarbonization that can power the country for centuries.”
Completing such a facility would be a milestone feat for the U.S. nuclear industry, which has not finished building a single new reactor ordered since the 1979 Three Mile Island accident, much less spawned a new sector. Sticking to the regulatory process that prevented any new nuclear from being built in decades, the U.S. Nuclear Regulatory Commission has been slow to approve any new next-generation fission technology. And the last time a company tried to build a nuclear fuel recycling plant, the White House at the time pulled the plug, fearing the facility might accelerate an apocalyptic arms race. Investors lost a fortune.
Yet many, like DeWitte, see the winds turning.
Countries that once vowed to eliminate nuclear energy are now looking to build new reactors as the limitations of renewables and natural gas to decarbonize big industrial economies become clear. There are few practical options other than turning to Russia. Moscow dominates the nuclear export market. Nearly half the 60 reactors under construction worldwide are Russian designs. And the only commercial supplier of the special fuel next-generation reactor companies like Oklo need is a state-owned Russian company.
Western sanctions against the Kremlin’s energy exports have spared the nuclear sector, in part because so few alternatives exist. Yet already companies reluctant to go into business with Russia since the Ukraine invasion are delaying projects as they struggle to get ahold of reactor fuel.
Last year, Congress earmarked billions in President Joe Biden’s landmark climate law for tax credits and subsidies to boost nuclear power and revive the domestic supply chain for the atomic energy industry.
The Department of Energy just recently picked three uranium mining companies to build up the federal government’s strategic reserve of the metal.
In October, Canadian mining behemoth Cameco bought the American nuclear pioneer Westinghouse as part of a joint venture with a clean-energy investor that promised to integrate one of the biggest uranium producers with the top U.S. designer of large, traditional reactors. X-energy — which, like Oklo, aims to sell small, next-generation reactors — met with regulators last month about licensing a facility near the Oak Ridge national laboratory to enrich fuel for its machines.
Likewise, Oklo wants to produce fuel for its own reactors. But if its proposed facility comes to fruition, the company aims to recycle its own waste in a repeating cycle that could, in the long term, make reprocessing cheaper than mining, refining and enriching fresh fuel. It would also establish a U.S. competitor to Russia’s closed-loop fuel services that appeal to countries that want atomic energy but have no appetite to store waste.
“As a country looks at buying a nuclear system, they go and see a country like Russia offering to design, build and even operate a nuclear plant and then, at the end of life, they take that waste back and manage it,” DeWitte said. “That’s an attractive offer when there’s only one on the table, especially when we in the U.S. aren’t offering anything close to that.”
How Nuclear Bombs Blew Up Uranium Recycling
Back when U.S. government scientists first started tinkering with techniques to recycle nuclear fuel, uranium seemed to be in short supply.
It was the 1940s, during the Manhattan Project to build the first atom bomb. Researchers at the time, including the legendary physician Enrico Fermi, believed uranium was far more scarce in the Earth’s crust than it later proved to be. Hoping to shore up supplies for a nuclear arsenal, and for what had been initially planned as 1,000 U.S. civilian power reactors, Fermi and his team at the University of Chicago began probing at ways to reuse spent fission fuel.
By the 1960s, two methods showed promise. One approach involved roasting spent fuel in special kilns to get at useful fuel. But pyroprocessing only worked for fuel that could be used in so-called fast-neutron reactors, a completely different design that replaced water with liquid sodium as a coolant, making it possible to sustain faster reactions that burned up more byproduct contaminants and actually generate more useful plutonium than fission consumed.
Another technique, called “aqueous reprocessing,” involved chopping used nuclear fuel rods into small pieces and using acid to separate useful fuel such as uranium and plutonium from other metals and fission products.
Targeting self-sustaining nuclear power industries, France, Japan, Russia, and the United Kingdom all opened aqueous recycling plants. In the early 1970s, the chemical conglomerate Allied Corp. broke ground on what was to be the U.S.’s first commercial reprocessing plant in Barnwell, South Carolina.
As has so often been the case with civilian nuclear technology, the prospect of world-ending war cast a long shadow. The initial goal of developing recycling was to guarantee the U.S. military’s supply of some of the world’s most dangerous radioactive materials. There’s a thin line between reprocessing fuel and enriching weapons-grade ingredients.
Even as the U.S. and the Soviet Union stockpiled more and more bombs, the world had agreed in 1968 to the Treaty on the Non-Proliferation of Nuclear Weapons. It was supposed to prevent any more countries from getting the bomb. In 1974, however, India tested its first atomic explosion, becoming the first nation since the treaty was signed to join the nuclear club.
Hoping to curb further proliferation, then-President Jimmy Carter, who had trained in the Navy’s nuclear submarine program, indefinitely prohibited all commercial nuclear fuel reprocessing in 1977. The Barnwell plant, a $1.3 billion investment in today’s dollars, went bust before it even opened.
Two years later, the Three Mile Island accident heralded the end of the civilian nuclear buildout in the U.S., with roughly a tenth of the reactors that were first envisioned.
Fresh uranium fuel was cheap and abundant enough to meet those reactors’ needs. And aqueous reprocessing — having really been designed as a technique to make plutonium for bombs — could only recycle fuel once; any more than that was too complicated and costly. Even after former President Ronald Reagan lifted the ban on nuclear recycling in 1981, no investor wanted to build a first-of-its-kind facility for which the market demand was as dubious as the long-term government support.
What little incentive power plant owners had to spend money on recycling uranium dwindled further in 1982, when Congress passed the Nuclear Waste Policy Act, requiring the federal government to take possession of and permanently dispose of all spent fuel from civilian reactors. The law put a tax on nuclear electricity, raising money for a final disposal site. Rather than recycle uranium, the U.S. plan was to bury the unusable fuel.
Five years after the legislation passed, the U.S. chose Yucca Mountain in Nevada as its first and only location for a permanent waste dump.
“There’s enough energy content in used fuel in the U.S. alone to power the country for 150 years, and that number is going to grow.”
Compared to the volumes of electricity reactors produce, there isn’t currently that much spent fuel. All the fuel waste generated in the U.S. since the 1950s could fit in a single Walmart Supercenter. Instead, however, it’s spread out among 75 locations in 35 states, most of them on-site at nuclear power plants.
In the meantime, the Cold War ended. In 1993, then-President Bill Clinton signed a deal with Russia to buy any reactor fuel made from disassembled atomic weapons, flooding the U.S. market with uranium too cheap for many U.S. makers of fresh fuel to compete. In 1994, the U.S. shut down its experimental sodium-cooled fast reactor.
That same year, engineers began tunneling into Yucca Mountain.
The engineering challenge proved more surmountable than the political one. Nevadans, especially Indigenous tribes who lived near the proposed facility, feared the canisters of spent waste destined for Yucca Mountain might corrode and contaminate their land and water. Shortly after taking office, former President Barack Obama fulfilled a campaign pledge to cut off funding to the Yucca Mountain project. Since that location had been the only one legally designated as the first depository, the government could not start work on another spot, effectively ending the U.S. effort to build a permanent nuclear repository.
Why Fuel Recycling Is Making A (Possible) Comeback
Growing up in the shadow of the nuclear industry in New Mexico, DeWitte knew about radioactive waste and became “super captivated” by the concept of recycling, figuring that the benefits of a closed-loop fuel supply chain would be too obvious not to realize them in his lifetime.
But as he got older and went to college for nuclear engineering, he was struck by how little progress had been made. He wanted to know why. For his senior undergraduate project, he worked on a reactor capable of running on waste.
Reviewing the history and policies that charted the nuclear power industry’s course over the past 70 years, DeWitte identified what Chris Gadomski, the lead nuclear analyst at the energy consultancy BloombergNEF, called the “chicken and egg situation”: Would newly built reactors create demand for recycling services? Or would the establishment of a recycling plant offer a compelling enough solution to the radioactive waste problem that the public would once again support building new reactors?
“There is plenty of uranium in North America and Australia,” Gadomski said in an email when asked about the viability of nuclear fuel recycling. “What is wrong with more mining and milling?”
To many opponents of nuclear energy, the answer is: Mining and milling take heavy environmental tolls, including on the rural communities — many of them Indigenous — that abut such sites.
“The uranium mills were all torn down in the mid-1990s because there’s no growth here, and no expectation of growth,” said Paul Robinson, a uranium supply chain expert and research director at the nonprofit Southwest Research and Information Center. “Nuclear power is not an economical way to boil water, so I don’t think there’s a chance of any full-scale reactors being proposed anywhere.”
He added: “If there really is a climate crisis, then we’re going to need to be less energy-intensive rather than more energy-intensive with our approach.”
What Comes Next
Nuclear reactors harness the energy released when the nucleus of an unstable atom splits in two, starting a chain reaction that gives off tremendous heat that can be used to boil water and spin electrical turbines. In nature, most uranium is uranium-238, which has an even number of neutrons.
The majority of nuclear reactors run on uranium-235, which, with three fewer neutrons, has an unstable nucleus that makes the isotope good fodder to sustain a fission reaction. But uranium-235 makes up just 0.7% of uranium mined from the Earth.
Through a process called enrichment, which involves converting the metal to its gaseous form and refining the material in centrifuges with thousands of rapidly spinning vertical tubes, the concentration of uranium-235 increases to between 3% and 5%, according to the World Nuclear Association. Anything enriched beyond that would be unusable for a traditional light water reactor.
But sodium-cooled fast reactors modeled on the experimental machine U.S. scientists used from the 1960s to the 1990s can use fuel enriched closer to 20%.
“It’s a poor analogy, but you can think of it as needing premium gasoline for a high-performance car,” said Roger Blomquist, a principal nuclear engineer at Argonne and former Navy submarine officer. “A regular conventional reactor can’t recycle fuel more than once because there’s too much garbage in the recycled fuel material. A fast neutron reactor doesn’t care; it uses that garbage to make electricity.”
While the U.S. buys three times more uranium overall from Russia than it produces at home, the country is entirely dependent on Moscow for that specific kind of fuel.
If next-generation fast reactors like those Oklo and the Bill Gates-backed TerraPower designed are to hit the market in the next decade, as those companies promise, customers will want fuel from somewhere that won’t hold energy exports hostage.
That’s still a big “if.” The few countries building new nuclear plants — the United Arab Emirates, India, and China, for example — are going for large, traditional light water reactors to maximize electricity production.
In the U.S., where most utilities are subject to the whims of the market, the cost and time it takes to build a big reactor has proven too much for investors who expect financial success on a tri-monthly basis. The only two reactors under construction are the 1,100-megawatt units at the Plant Vogtle facility in Georgia, which, at more than $31 billion, are years delayed and roughly double the original budget.
To compete against imported solar panels and subsidized natural gas, the nuclear industry is hoping that smaller, lower-power reactors, known as small modular reactors or SMRs, can be produced in bulk, bringing down costs through economies of scale.
“We need building nuclear reactors to be less like building airports and more like building airplanes,” said Alex Trembath, the deputy director at the Breakthrough Institute, a pro-nuclear environmental think tank in California.
First, the NRC will need to give the green light to build the first SMRs. After years of stalled efforts, the agency last month gave the Portland, Oregon-based startup NuScale’s reactor design final certification, making it the first SMR and only the seventh reactor design ever approved for use in the U.S.
But NuScale’s design is essentially just a scaled-down version of a traditional light water reactor. Companies such as GE Hitachi Nuclear Energy and Holtec International are pursuing regulatory approval for similar types of reactors. Because the model hews so closely to existing reactors, industry analysts have generally forecast those designs to hit the market sooner than the kinds of advanced machines Oklo aims to produce.
Instead, Oklo’s design borrows from the fast reactors the U.S. produced in the 1960s but never fully commercialized.
This has been a challenge. The NRC rejected Oklo’s licensing application last year on the grounds that it failed to provide enough information for the agency to assess the design safety. The Breakthrough Institute in turn blamed the agency’s “higher than necessary hurdles” for advanced reactors.
The application process alone took NuScale six years, 12,000 pages, and more than half a billion dollars. In a lengthy 2021 report, the Nuclear Innovation Alliance, a think tank that works closely with nuclear energy companies, called on Congress to reform licensing fees and provide more federal financing for new reactors.
The NRC said it could not provide an estimate of how long it would take to assess Oklo’s bid for a recycling facility since the company had only submitted “a licensing project plan” so far.
“The staff would have to check any future application for completeness before we could set a review schedule,” said NRC spokesperson Scott Burnell. “The X-energy fuel facility project is also in its early stages so I don’t have a schedule available there.”
Asked to weigh in on the viability of fuel recycling generally, Burnell deferred to the Department of Energy: “The NRC’s only role is determining whether an applicant meets the relevant safety and environmental requirements.”
Biden’s Inflation Reduction Act, his landmark climate and infrastructure spending package, delivered on the money side, earmarking hundreds of millions of dollars for tax credits and funding for next-generation reactors. It also included $700 million to spur a domestic supply chain for the kind of advanced uranium fuel on which companies like Oklo would depend.
Already, Oklo has won four Department of Energy awards to support its efforts to commercialize fuel recycling. Totaling more than $17 million, the cost-share awards allowed Oklo to split the cost of research at Argonne with the federal agency.
DeWitte said he’s hopeful that the NRC will ultimately approve both his company’s designs for reactors and fuel recycling.
“Historically, it’s been dominated by utilities who don’t have a mandate or the bandwidth to take a different approach to doing things in a more modern way,” he said of the licensing process. “What I’m excited about, just generally speaking, for advanced reactors is there’s a recognition the NRC has that this is different, but also a reality that this has to happen.”
While there’s no direct analog for the NRC to approve a fuel recycling plant, the agency has greenlighted fresh fuel manufacturing facilities.
“If you think about fuel fabrication, you’re handling radioactive materials and doing chemical operations on them,” DeWitte said. “That’s the same thing we’re doing.”
While most reactor companies sell their machines to utilities, Oklo plans to own and operate its own plants, meaning that its recycled fuel is “about helping us supply our own growth,” he said.
For now, the company has fuel it acquired years ago from the federal government. By contrast, its billionaire-backed rival TerraPower delayed its plans to build its first sodium-cooled SMRs because it could not obtain advanced reactor fuel from Russia.
In that sense, DeWitte said he saw TerraPower less as a competitor than a potential customer. Eventually, he expects fuel recycling and sales to represent 40% of Oklo’s business, with reactors making up the other 60%.
“With the reactors being bigger than the recycling side, ideally we could provide for what we do but we may also need to buy fuel from others, and vice versa,” DeWitte said. “I don’t think of it as competitive, but as complementary.”
There are other companies competing to enter the nuclear recycling market. But with the U.S. generating roughly 2,000 metric tons of spent fuel per year, DeWitte said there’s more than enough waste to go around.
“Everyone serves different niches, and there’s so much material out there,” he said. “We’re just excited to make a big dent in that.”
CORRECTION: This story was updated to clarify that X-Energy’s design is not a fast reactor.