For six weeks in early 2006, engineers drilled nearly 3,000 holes into the 499-foot-tall cooling tower at the Trojan Nuclear Power Plant outside of Rainier, Oregon, and filled them with 2,800 pounds of dynamite. Early on May 21, a 2.5-second explosion echoed through the surrounding hills. Cheers erupted from the dozens of anti-nuclear activists who had come to watch the demolition from across the Columbia River.
The effort to topple Oregon’s nuclear industry had been years in the making. The Trojan plant, which came online in 1975, featured the largest reactor ever built. Twice in the 1970s, picketers blocked workers from entering the plant. Four years after it opened, the partial meltdown at the Three Mile Island facility in Pennsylvania spooked the nation, and Oregon, like many states, put a moratorium on new nuclear plants. By the time Trojan was blown up, the industry was over, at least locally. Or so it seemed.
In 2007, an engineer at Oregon State University named José Reyes began to resurrect it by imagining a reactor that would be “very, very different.” By shrinking and simplifying the standard nuclear reactor, Reyes believes he has created a technology that can generate power more safely at a fraction of the price. Last August, the Nuclear Regulatory Commission issued a final safety report for Reyes’ design, recommending its certification. Construction on the first reactor could begin as soon as 2025. That puts NuScale, the company Reyes co-founded, at the front of the race toward “advanced nuclear” power—a technology that advocates say will be essential to the transition away from fossil fuels.
Donald Trump’s Department of Energy was “all in” on advanced nuclear, as a press release put it, pouring hundreds of millions of dollars into research and development. President Joe Biden is a fan, too. As part of his plan to shift the United States to 100 percent clean energy by 2050, he has targeted further investment in small modular nuclear reactors like NuScale’s.
But are these investments worth the money—and the risks? New designs or not, nuclear plants face daunting issues of waste disposal, public opposition, and, most of all, staggering costs. We must ramp up our fight against climate change. But whether nuclear is a real part of the solution—or just a long-shot bid to keep a troubled industry alive—is a debate that will come to the fore in the short window we have to overhaul the nation’s energy portfolio.
Few issues divide us as cleanly as nuclear power. According to a 2019 Pew Research Center poll, 49 percent of Americans support opening new plants, while 49 percent are opposed.
The popular argument against nuclear power can be summed up in a few names: Chernobyl. Fukushima. Three Mile Island. Nuclear dread is palpable. Some formerly pro-nuclear countries, like Germany, began phasing out plants in the wake of the 2011 disaster in Japan. The dangers begin well before nuclear fuel arrives at a plant, and persist long afterward; the rods that fuel today’s plants remain radioactive for millennia after their use. How to ethically store this waste remains a Gordian knot nobody has figured out how to cut.
The argument in favor of nuclear power boils down to the urgent need to combat climate change. The United States’ 94 nuclear reactors provide about half of our carbon-free power. Like coal- and natural gas–fired plants, they supply a steady amount of energy, no matter the weather. Advocates also point out that in terms of deaths per unit of energy, nuclear is on par with renewable plants as one of the safest forms of energy available.
But if nuclear power is going to help us mitigate climate change, a lot more reactors need to come online, and soon. Eleven nuclear reactors in the United States have been retired since 2012, and eight more will be closed by 2025. (When nuclear plants are retired, utility companies tend to ramp up production at coal- or natural gas–fired plants, a step in the wrong direction for those concerned about lowering emissions.) Since 1970, the construction of the average US plant has wound up costing nearly three-and-a-half times more than the initial projections. Developers have broken ground on just four new reactor sites since Three Mile Island. Two were abandoned after $9 billion was sunk into construction; two others, in Georgia, are five years behind schedule. The public is focused on risks, but “nuclear power is not doing well around the world right now for one reason—economics,” says Allison Macfarlane, a former commissioner of the Nuclear Regulatory Commission.
Nuclear was not always so troubled. In the 1950s, some environmentalists advocated “atoms not dams,” preferring nuclear plants to hydropower projects that destroyed wild landscapes. Until Three Mile Island, public support was strong. Dozens of plants came online. In the 1970s, Reyes, seeing an industry full of promise, decided to pursue a degree in nuclear engineering.
After a decade of working at the Nuclear Regulatory Commission, in 1987 Reyes went to teach at Oregon State University, where he built a model of the Trojan plant’s reactor, scaling down its pipes and chambers to examine what might happen if its cooling pumps failed. After he presented his findings at a conference, Westinghouse—then one of the biggest builders of nuclear reactors—tapped Reyes to analyze its technology. Under his lead, Oregon State eventually became what he calls “the Consumer Products Safety Commission for the nuclear industry.”
Reyes was determined to simplify reactors. The conventional model is a giant tangle of tubes, any of which might fail. In a plant’s control room, operators monitor hundreds of data sources at once; Reyes sees this oversupply of information as a needless source of stress. So he designed a smaller, more streamlined reactor. The university, which still hosts NuScale’s test facilities, helped him commercialize his idea.
When I visited NuScale before the pandemic hit, most of the company’s 400 employees worked out of a stucco building on the outskirts of Corvallis. Reyes’ office looks out toward the Oregon Coast Range. Now in his mid-60s, he is soft-spoken but energetic. On his gray blazer, he wore a pin with the company’s logo: an N composed of atomlike dots.
A nuclear reactor, he explained, is like a giant teakettle—it’s just a mechanism for boiling a whole lot of water. The heat comes from metal fuel rods, filled with uranium pellets, which are immersed in water inside the reactor core. The fuel is hit with neutrons, causing the uranium atoms to split, one after another, in a chain reaction that releases tremendous heat. The water does double duty: It carries the heat into a steam generator, where another batch of water is boiled and sent into turbines that produce electricity. The water also functions as a coolant, preventing the fuel rods from growing so hot that they melt. The most famous nuclear disasters have all been caused, at least in part, by failed cooling systems: Water could not be pumped back into the core quickly enough to protect the fuel.
Reyes aims to reduce the likelihood of such a catastrophe by placing the reactor core and steam generator inside a much smaller containment vessel, reducing the amount of exterior piping and thereby the number of places where coolant can escape. Instead of relying on pumps and motors to push water through the reactor, as is typical, Reyes’ system is passive and less dependent on human intervention. For instance, the NuScale reactor features valves that flip open automatically if power fails, sending water back into the reactor core to cool it.
Reyes also shrank the reactor—by a lot. All of a NuScale reactor’s components will be housed in a shell that is just 76 feet tall—twice the height of a typical telephone pole—and 15 feet wide. More than 100 of these shells could be packed inside a conventional nuclear reactor containment building. A typical NuScale plant will consist of 4 to 12 reactors in their shells, all surrounded by a large pool of water, one more layer of protection. Each reactor will generate up to 77 megawatts of energy, which means a “12 pack” will produce nearly as much power as a traditional 1,000-megawatt reactor. Their small size, though, means these new nuclear plants can replace old coal plants on the same site, plugging right into the existing grid—an enormous cost savings.
This reimagined nuclear plant, Reyes says, will supplement renewable power. Right now, the United States turns to wind and solar power for less than 10 percent of its energy. While the price of renewables has been dropping, relying on wind and solar for all of our energy needs would depend on battery storage and other technologies that aren’t yet ready. Until they are, backup facilities would be required to ensure power never drops because of a cloudy or windless day. The question is what path we want to follow to get to zero emissions: Do we build a huge number of solar and wind-powered plants, enough to ensure power is always coming from somewhere? Or build fewer such plants and supplement them with a baseload supply of steady electricity from a handful of nuclear facilities. A recent study by a consortium of utilities found that this second approach could drop the price of supplying carbon-free energy to the Pacific Northwest by $8 billion a year—though only if NuScale can keep its costs as low as it has promised.
Utah Associated Municipal Power Systems, a state-owned agency that sells electricity across six Western states, is betting that it can. The agency, which aims to offer its members the choice of fully carbon-free power, sees NuScale as the best available option for undergirding its existing wind and solar plants. In 2015, UAMPS announced a plan to build 12 NuScale reactors at the federally run Idaho National Laboratory. NuScale projected total construction costs at $3 billion—nearly a third less than the most recently completed US reactor, which came online in 2016 at a cost of $4.7 billion (though it will supply more power). And the next plant should cost even less, since NuScale’s small reactors will be built on an assembly line, rather than on-site. But the price will drop only if more customers buy them. “Taxes are more popular than nuclear power,” jokes Doug Hunter, the CEO of UAMPS.
To change that perception, Hunter and his team have spent the last few years visiting towns and utility companies that buy power from UAMPS, explaining the potential role of nuclear power and the safety of NuScale’s design. His persistence paid off. By 2020, the majority had signed on to the NuScale project—though only as long as they had plenty of chances to back out if the project went south.
The day after I visited NuScale, I drove north to Rainier to tour what’s left of the Trojan nuclear plant. Power lines overhead disappeared into mist-shrouded hills, and geese squawked alongside a 29-acre lake in a quiet, wooded park. Portland General Electric, the plant’s owner, secures the 34 casks that hold the plant’s old fuel behind a chain-link fence. Otherwise, there is no marker indicating the site’s history. The cooling tower, once visible for miles, has been gone for more than a decade.
Lloyd Marbet, one of the activists who helped ensure the tower’s demolition, moved to Oregon in 1969 as a young anti-war protester. He soon became an anti-nuclear organizer, too, and in 1980 was involved in a campaign to ban construction of more nuclear plants in Oregon until there was a permanent waste disposal facility (though Marbet had hoped to go further and ban the technology entirely). A ballot measure he championed narrowly passed, and the law remains in effect.
I met up with Marbet over lunch at an organic cafe and then trailed his orange, electric BMW to the office of the Oregon Conservancy Foundation, the latest nonprofit he directs. Built in a shipping container and sheathed in aluminum insulation, the office felt like the cluttered, homey cockpit of a spaceship that might be able to manage the Kessel Run in less than 12 parsecs.
Marbet admires the science of nuclear reactors; when he toured Trojan in the 1970s, he was impressed by its engineering. But he insists there is no perfect mousetrap. Even with new technology, we will need to mine uranium—a process that has leached radioactive waste into waterways—and find somewhere to put the spent fuel. (The current practice, which persists at Trojan and will be employed at NuScale’s plants, is to hold waste on-site. This is intended to be a temporary measure, but every attempt to find a permanent disposal site has been stalled by geological constraints and local opposition.) Marbet believes we need to transition away from coal and gas immediately. But he worries that nuclear is too expensive, and a new round of investment might pull money away from more effective, and cleaner, solutions.
In the years after Oregon’s ban on new plants, Marbet and his allies wrote three more ballot measures aimed at shuttering Trojan. Their campaigns highlighted the plant’s safety and design flaws, and featured sealed court transcripts that an anonymous source had delivered to Marbet in 1989, wrapped in a paper bag. Those memos showed that Trojan’s builders had hired a wildly underqualified engineer to design some essential buildings—mistakes caught only after the facility was complete.
Despite such evidence, all three ballot measures were defeated. Then, in 1993, just months after the last campaign, Portland General Electric discovered leaking steam tubes inside the plant. They decided to close it forever—less than halfway through its operational life. “I had people come up to me on the street and say, ‘Jeez, you know, we thought you were full of it, but now we see that the concerns that you were raising were legitimate,’” Marbet told me.
These days, he’s watching the industry creep back. A Republican state senator named Brian Boquist has proposed a bill three times that would permit city or county voters to exempt themselves from the 1980 law, allowing a nuclear facility to be built within their borders. (The bill has failed twice; the latest version is with the senate committee.) Boquist does not seem particularly committed to fighting climate change: He and other members of the Republican minority refused to show up to vote on a cap-and-trade bill in early 2020, causing the Senate to fall short of a quorum. (When Gov. Kate Brown threatened to retrieve legislators using state troopers, Boquist said to “send bachelors and come heavily armed.”)
In 2017, as the legislature debated Boquist’s first pro-nuclear bill, Marbet testified that NuScale was making “an end run around [voters] in their quest for corporate profit.” He also noted the company’s ties to the Fluor Corporation. The Texas-based multinational engineering firm that has been NuScale’s majority owner since 2011 has invested $9.9 million in campaign contributions over the past 30 years, with nearly two-thirds going toward Republican candidates. (Fluor is currently under investigation by the Securities and Exchange Commission due to allegedly sloppy accounting practices.)
Marbet admits his view of the industry is jaundiced, but his experiences make him skeptical of NuScale and its claims. He worries, too, that if small reactors take off, operators will revert to old habits, cutting corners to make a buck. He points to a draft rule approved last year by the Nuclear Regulatory Commission, over the objections of FEMA, that would reduce the size of the emergency planning zone around nuclear plants: Rather than a 10-mile-wide circle, a plant would only need an evacuation plan for the space within its fence lines. NRC commissioner Jeff Baran opposed the change, noting it is based on assumptions about small reactors, like NuScale’s, that remain on the drawing board, and might open the door to weakening safety standards for existing plants.
Old-line environmental groups like Greenpeace and the Sierra Club remain staunchly opposed to nuclear power, but politicians have been more open to it. President Barack Obama was an outspoken proponent of nuclear’s potential. For 2020, the Senate Appropriations Committee unanimously agreed to spend more than President Trump requested on nuclear research, and the Senate is currently considering a bipartisan bill that will streamline the permitting process and establish a national uranium reserve.
Now, as part of his $2 trillion climate plan, Biden is calling for a federal research agency that would pursue carbon-free energy sources, including small reactors. Biden’s was the first Democratic Party platform in 48 years that explicitly supported an expansion of nuclear energy. His pick to lead the Department of Energy—which devotes the majority of its budget to nuclear projects—is former Michigan Gov. Jennifer Granholm, who has little experience in the field. Gina McCarthy, the former EPA administrator who is Biden’s chief domestic climate coordinator, has said that nuclear could play a key role in baseload power supply but indicated that waste disposal issues ought to be resolved before the technology is widely adopted.
NuScale is not the only player: Currently, seven companies are in discussions with the NRC about advanced reactor designs. Last March, a Silicon Valley startup called Oklo applied for a license to build and operate a plant that uses liquid metal as a coolant. This might allow some old fuel rods to be reused—though some scientists worry it could create new and more dangerous kinds of waste. If it meets its ambitious timeline, Oklo will beat NuScale to market. But Oklo’s “microreactor” will only generate 1.5 megawatts, enough to power 1,000 homes, whereas NuScale’s first plant will produce hundreds of times more power. The company is targeting small, remote communities in regions like northern Alaska. Ahmed Abdulla, a researcher working with the University of California, San Diego’s Deep Decarbonization Initiative, calls this a “niche” technology, unlikely to contribute widely to the national grid. He doubts that any of the current designs are likely to be widely adopted in the next few decades—NuScale’s included. The economics, he says, “render nuclear virtually irrelevant in the US.”
A major hurdle for any advanced nuclear product is the regulatory process. NuScale spent more than $500 million developing its licensing application. The path to approval has consumed 12 years already, and it’s not over yet. In the months after my visit to NuScale, the Nuclear Regulatory Commission noted “several potentially risk-significant” questions that remain unanswered about the company’s reactor design, especially about its new version of a steam generator. Nonetheless, the NRC granted its initial approval of the design at the end of the summer; now NuScale awaits official, final certification by the commissioners, which is expected sometime this year. But further analysis of the generators will be required before a license is granted to actually build a plant.
A decade ago, NuScale suggested it might have a plant in operation by 2018. Now construction won’t begin until 2025 at the earliest. The plant at Idaho National Laboratory won’t be fully operational until 2030. Factoring in interest and other costs not included in NuScale’s $3 billion estimate, UAMPS expects a total 40-year lifetime cost of $6 billion for the plant. Some critics see this as the same old story: grand, early promises—a “dog and pony show,” as Marbet calls NuScale’s PR—followed by cost overruns and delays. Reyes intentionally used materials familiar to regulators, so as to speed along the process. But other advanced reactor designs, which use new kinds of fuel and coolant, may face an even slower and more expensive journey.
Recently, nine towns—more than a quarter of the subscribed members—pulled out of UAMPS’s project after changing their minds about their energy needs or worrying that it was becoming a financial sinkhole. (Meanwhile, one new town signed on.) The plant’s economics depend on running near full capacity, which will only happen if utilities outside of UAMPS also buy some of its power. The Department of Energy says it will chip in nearly $1.4 billion over the next nine years, which should help bring down the cost of the plant’s energy. But the projected price—$55 per megawatt-hour—is still above the current costs for solar and wind projects. And the federal money will require annual congressional approval. It’s possible that other new ideas might pop up, competing for limited dollars.
Biden’s climate plan hinges on a massive expenditure on research. What his administration will have to quickly decide, though, is how to divvy that pot. Allison Macfarlane, the former NRC commissioner, told me other industries deserve far more of our resources and attention than nuclear. Batteries, in particular, could steady out the uneven flow of renewables. They may even work better, since nuclear plants are difficult to power up or down in response to changing conditions. Once a pie-in-the-sky idea, battery storage now offers costs at least “in the ballpark” of nuclear, says Stan Kaplan, a former US Energy Information Administration analyst. Prices have dropped 70 percent in the past few years and are projected to drop another 45 percent before NuScale’s plant comes online. California—which also has a moratorium on nuclear builds—is rapidly expanding its storage capacity. Within 10 years, the niche that NuScale is aiming for might already be filled.
But nuclear proponents say we still need “an enormous amount of hedges,” as Josh Freed, vice president of the think tank Third Way, put it to me. For nuclear to persist as a hedge, it all but requires government assistance, given the enormous upfront costs of R&D. Another challenge is vetting which projects have real promise. “You have all these reactor vendors pitching their wares, and making all sorts of outrageous and false claims,” says Edwin Lyman, the director of nuclear power safety with the Union of Concerned Scientists. These claims have also been the basis of lowering safety standards, which offers a large indirect subsidy for operators. There needs to be a stronger peer-review process, he says, to make sure the government is only sponsoring truly worthwhile projects.
A recent study from Princeton found that even without nuclear power, the relative cost of a decarbonized energy system in 2050 could be about the same as in 2015, which at the time was a historic low. The study found nuclear could reduce costs even further—if it becomes as cheap as its advocates hope. But Abdulla, the UC San Diego researcher, has calculated that in order to make advanced reactors accessible within the next few decades—even relatively simple reactors, like NuScale’s—the government would need to provide hundreds of billions of dollars in subsidies and substantially simplify the regulatory process. Abdulla believes nuclear energy should have been “an arrow in our quiver.” But given the economics, he says, “I fear the arrow has broken.”
That makes for a strange moment. No existing clean energy source is quite so scalable as nuclear, which is why engineers are bullish. And if money were no object—if we could snap our fingers and scatter reactors across the landscape—we would take a huge step toward curbing emissions. But if Abdulla’s numbers are right, the nuclear dream looks dead on arrival. For many, that feels like relief: We can shake off our nuclear dread. But it should be bittersweet, too. It’s a door closed, after all, and until we know another has opened, the changing climate gives us plenty else to fear.
Technical illustration: Remie Geoffroi
This article has been updated.