We’d all like to live in the world of fairytales and gumdrops where all our power comes from the wind and the sun. But when you talk to people who know what they’re talking about, one thing becomes clear: Wind and solar will be a big part of our energy future, but they won’t be everything. Those are renewable power sources, which fit under the larger umbrella of low-carbon energy, but the sun doesn’t shine all the time and the wind doesn’t always blow. A bunch of companies are working on new battery technologies to store energy from wind and solar for the downtimes, but we need other slices in our energy pie to crank out power 24/7. These are called “firm” or “baseload” power sources, and we’ve covered a couple—like hydro and geothermal—that will hopefully replace some of what we get from oil and gas.
But there’s one source of clean, constant energy that already supplies 20% of the electricity in the United States: nuclear power. In the real world of the future, a big slice of the pie is radioactive.
“No one who's a real engineer could even imagine anything beyond about 80% wind and solar. I think it's far less than 80%,” says Paul Dabbar, a distinguished visiting fellow at the Center on Global Energy Policy at Columbia University. He’s a former nuclear submarine officer who went on to become undersecretary for science and innovation at the Department of Energy. Somewhere between 20 and 50% of our energy pie needs to be constantly and reliably available, which wind and solar can’t supply without battery technologies we don’t currently have. Even then, there’s some skepticism among utilities and other folks whose job it is to keep the lights on. Geothermal and hydro are helping around the edges, but the issue is scale. “You're either going to run gas plants with carbon capture,” added Dabbar, referencing still-unproven tech for capturing the carbon emissions from natural gas plants before it floats up to the atmosphere, “or you're going to need nuclear.”
Some in the environmental movement oppose nuclear, and Germany, for one, was on the path to shuttering its plants for a while. But that’s not realistic now, and it isn’t realistic when you consider our future electricity demands. “If we're really going to displace all this fossil power and electrify the transportation sector to some high degree, and maybe electrify other sectors, that's just going to increase the amount of electricity we use by quite a lot,” says Dr. Matt Bowen, a research scholar at the Center on Global Energy Policy. “All the scenarios that the Princeton Net Zero America folks looked at, I think our electricity generation always at least doubles after 2050. We're going to shut down all the nuclear power plants and still somehow double our generation in the next twenty-something years? It just seems highly implausible to me.”
There’s no low-carbon transition where power stays reliable and affordable without nuclear. Once they’re built, these plants are cheap: the fuel is cheaper than gas and coal, you’ve got some maintenance (which, by the way, you’ve also got on wind and solar), and a plant can operate for 80 years. And again, it's all based on mature technology that’s already powering our homes and businesses. The possibilities for fusion—the process of combining atoms to create energy that saw a breakthrough in December—are a whole ‘nother conversation.
For now, we’re talking fission: splitting atoms. We’ll probably need to build more of these power plants, which isn’t exactly easy. First of all, there’s public apprehension. We all have some image in our minds of nuclear disaster, whether that’s Cherynobyl or Fukushima or Three Mile Island here in the States. But in the scheme of things, deaths connected to nuclear power are very rare, particularly compared to oil and coal.
“The Chernobyl reactor was a horrible design that was literally illegal in the United States. It was unsafe,” Dabbar says. “As the temperature went up in the reactor, the power went up. That fundamental issue is very bad. Even the Russians themselves don't build it anymore. In Japan, it was a bad location. Don't put it in certain places that have potential problems, like a tsunami. If you look at Three Mile Island, no one died, nothing actually got out. The reactor got wiped out. It melted down inside, but it didn't hurt anybody. Obviously, that was not positive, and the things that allowed that to happen have been fixed. It was primarily about instrumentation and training.” That last one was in 1979 in Pennsylvania, which still gets 36% of its power from nuclear today. It sits behind seven states, including South Carolina (56%) and Illinois (54%). The kind of siting issues that Fukushima exposed will always be a problem, particularly as the same climate we’re trying to salvage is serving up more extreme weather events, but they’re a problem for any energy project. In Texas in 2021, it was the gas plants that froze.
Another sticking point for the public is waste disposal. Dabbar ran the largest disposal program in the world while at the Department of Energy, and he’ll tell you it isn’t a technological problem. “There has been mostly a political block on the topic, not science, environmental, or engineering limitations. The ability to consolidate, concentrate and put the nuclear waste into a safe form, in a safe location, is completely doable. That's completely not a challenge.” The challenge is finding that safe location. There was agreement on a site at Nevada’s Yucca Mountain, but waves of politicking have stalled that plan. Dabbar backs the idea of a central storage site, either “long-term geological storage that could last millennia” or “interim storage, which could be 100 or 200 years.” We already ship nuclear material around the U.S. by rail and truck and store it in a secure location. We know how to do this because we do it all the time when nuclear-powered submarines and aircraft carriers need to refuel.
(Fun fact: Dabbar says newer designs for these reactors allow some Navy ships to go 40-plus years without refueling. Even with the older submarines, it was about 20 years.)
There’s a thornier political problem on the other side of the process: acquiring the fuel. Russia currently dominates key markets servicing the nuclear fuel supply chain. After uranium is mined, it goes through a number of different processing stages. Two of them are, individually, “conversion” and “enrichment.” Russia has big influence on these two markets, but the U.S. and the rest of the West have been working to disentangle their supply chains since Russia’s invasion of Ukraine. “The United States should be able to largely get away from Russia this year,” says Bowen, who published a paper on the topic in May. We’ll always need to deal with the fact that we don’t actually mine much uranium domestically, but we can get more control over the nuclear equivalent of oil refining.
OK, with all the (spent) fuel and (geo)politics talk out of the way, we come to the real obstacle to nuclear proliferation in the U.S.: It’s too damn expensive to build a nuclear power plant. The infamous recent example is the expansion of Plant Vogtle, a Georgia project in which two new reactors could come online this year after six years of delays and $20 billion in cost overruns. The $33 billion tab is more than double the original cost projections. Dr. Magdalena Klemun is a professor at the Hong Kong University of Science and Technology and previously studied the economics of nuclear power at MIT. She says American issues with delivering nuclear power plants on time and on budget are tied to increasingly stringent safety requirements—no bad thing—and declining construction productivity that’s hit across industries, but has hit nuclear particularly hard.
“Any onsite construction project is always prone to issues with the management of the construction team, issues with the supply chain, the on-time delivery of different materials and components,” Klemun says. But with nuclear, “every single screw that you place requires documents before that plan where exactly it is to be placed, and then documents after that confirm that everything has been done exactly right.” These projects suffer from ballooning “soft costs” on the worksite. More than that, because we essentially stopped building nuclear power plants for decades, we might have lost a step.
Dabbar’s view is slightly different. “‘On time, on budget’ is something that has not been a part of the culture of the nuclear industry,” he says. “To a reasonable degree, many people in the industry just assumed everything was going to be significantly over schedule and over budget. It's a little bit like you're building a home and your contractor gives you a price, but they just immediately think that it will never work and you're going to have to pay double. Well congratulations, it's almost certainly going to be double.”
This has understandably freaked out local utilities and scared away the kind of capital investment that could get these projects off the ground more often. “If you're a power company CEO and you're looking at various types of power plants to go build, a lot of executives [who backed nuclear plants] have been fired, and companies go into bankruptcy,” Dabbar says. And it’s not new: “If you go back to all the other nuclear construction cycles in the U.S., there's one bankruptcy after another. This isn't just Georgia and South Carolina. There was Public Service in New Hampshire and El Paso Electric, and Long Island Lighting went into bankruptcy and disappeared.” Large nuclear projects often feature “very, very poor controls over the construction, budget, and schedule.”
At the Department of Energy’s National Reactor Innovation Center, director Ashley Finan says work is underway on construction technologies that will bring down costs. “We've built prototypes of steel bricks, which some people say is a Lego-type approach to concrete and steel for nuclear.” She adds they’re also looking at how a best practice from the tunneling industry, “vertical-shaft boring,” could bring down cost and time. But other large civil engineering projects have some of these same problems, Dabbar says, and he believes there's a solution that harkens back to his days on a nuclear submarine: build them smaller.
Instead of a giant 1,200-megawatt reactor, you could have four 300-megawatt reactors on the same site. With Small Modular Reactors (SMRs) we have the advantage of building much of them in factories, with standardized processes that tamp down on the cost overruns and delays. “This is where all the ex-Navy people have come from for years,” he says. “The Navy takes Small Modular Reactors—better known as just a nuclear reactor for a submarine or an aircraft carrier—built at a central location, primarily in Virginia near Roanoke. Then they put them on a semi truck, and they ship them off to the shipyard and they weld it in. A very significant portion of the whole reactor is made at a factory someplace and shipped. Navy reactors are delivered on time and on budget into submarines and aircraft carriers all the time. It's actually a very well-honed machine.”
What if instead of a big, hulking, site-specific nuclear power plant design, we had a standardized design for a reactor that’s a quarter or a tenth the size that we could manufacture in a factory and ship to a site, where it could be installed in a process closer to what we have for nuclear subs and aircraft carriers? If we need more power than one reactor can provide, we could stack several on a single site. In one model, we could plug these things into former coal plant sites and use the transmission infrastructure already in place to connect them to the grid quickly. There are also plans for “microreactors” that are even smaller than the SMRs. They could power a single large facility, like a hospital, or a remote community on their own. This might freak you out, but keep in mind that many of these new advanced reactor designs have inherent safety features that allow them to power down and eliminate danger without even much human intervention.
Now just imagine we’re cranking these out in factories and shipping them wherever they’re needed around the country. In fact, you don’t need to imagine: on Tuesday, the U.S. Nuclear Regulatory Commission approved an SMR model from NuScale. An individual reactor will produce 50 megawatts on its own, but it can be stacked in groups of four, six, or twelve. The assembly line changed everything for a reason. We need one for nuclear power plants, because we need nuclear power.