Nuclear power in the U.S. is locked in a stalemate, according to a new report from the Department of Energy.

No matter how much renewable energy is deployed to decarbonize the grid, DOE is estimating that 500 to 750 GW of clean, firm power — including 200 GW of new nuclear — will be needed to reach net-zero emissions economy-wide by 2050.

“We’re going to need multiple reactor technologies to be successfully deployed at scale, from Generation 3 light water reactors to Generation 4 advanced reactors,” Kathryn Huff, who leads DOE’s Office of Nuclear Energy, told an online audience at Friday’s webinar on the recent Pathways to Commercial Liftoff: Advanced Nuclear report. Reactors of different sizes will also be needed, “from 1 MW, all the way up to gigawatt-plus reactors,” Huff said.

The report differentiates between Gen 3 and Gen 4 reactors based on the fuels they use and how they are cooled. Traditional, water-cooled Gen 3 reactors use low-enriched uranium, while Gen 4 reactors use high-assay, low-enriched uranium (HALEU) and alternative coolants such as molten salt.

“Each of these technologies has a different role to play in meeting our decarbonization goals,” Huff said. The first step to putting those 200 GW online will be “getting a committed order book of signed contracts for new reactors,” with five to 10 orders for each technology, ideally by 2025.

DOE is providing about $3.2 billion to help fund the construction of two new advanced, small modular reactors (SMRs), but the massive cost overruns and delays that have confounded Southern Co.’s Vogtle 3 and 4 reactors in Georgia have cast a long shadow over the industry, said Julie Kozeracki, a senior adviser at DOE’s Loan Program Office (LPO). (See Making the Case for Nuclear at NARUC.)

With Unit 3 just starting to produce power ― six years behind schedule and at more than twice its original $14 billion price tag ― “nuclear has a huge credibility problem to solve,” said Kozeracki, who helped author the report. “Everyone is staring at each other ― customers, suppliers, reactor designers. … Right now, every utility recognizes that they need new nuclear; they need clean, firm power. But they want to wait for someone else to go first, second, third, and order reactor No. 4 or reactor No. 5.

“But that’s not good enough,” she said. “Because if they all wait for the demo projects to be done, it’s going to be too late, and we’re going to miss the boat. … We need signed contracts, not press releases, not [memoranda of understanding] and not letters of intent, because you can’t finance a supply chain with MOUs.”

The report makes the case for nuclear as a carbon-free technology that checks a lot of boxes for grid reliability, energy security, economic development and equity, points underlined by speakers at Friday’s webinar.

“Between 2023 and 2050, about 200 electric GW of unabated coal assets are expected to retire,” Huff said. “Nuclear energy is uniquely positioned to replace those retiring assets with a similar electricity generation profile.”

“Deploying clean, firm power sources like nuclear will enable the increased deployment of renewable power,” LPO Director Jigar Shah said. “In addition to providing clean power, nuclear also uses land efficiently [and] has lower transmission requirements; so, they can site themselves on existing coal plant sites … and can leverage existing transmission infrastructure as fossil assets are retired.”

A 2022 DOE study identified close to 400 existing or retired coal plants that could be suitable for advanced nuclear development.

Huff also stressed the economic benefits of coal-to-nuclear transitions for communities affected by the closure of coal or other fossil fuel plants. “Nuclear is one of the few generation sources that can preserve the volume of high-paying jobs from those retiring coal plants,” she said. “A lot of the same people who maintain turbines and steam boilers and electricity around the plant can be rehired, and some of them don’t even need to be retrained to be leveraged into a nuclear power plant.”

The report notes that nuclear plants create about three times the number of jobs per gigawatt compared to wind projects and pay 50% more than wind or solar. Benefits for disadvantaged communities in general are also part of the picture.

“Access to reliable and resilient clean energy resources is not equitably distributed across the U.S.,” the report says. “Increasing grid reliability and resilience for underserved, overburdened communities can support improved health outcomes, public safety, economic security and overall quality of life.”

‘Megaproject Issues’

But how does nuclear compare with the low levelized cost of wind and solar, or even natural gas? It’s a question often asked by nuclear skeptics.

Kozeracki says it doesn’t matter “because of the value it’s providing for a resilient, decarbonized grid. As a clean, firm resource, nuclear doesn’t need to compete with solar by itself or with natural gas by itself. It needs to compete with solar; with really long-duration energy storage, or natural gas with carbon capture,” technologies that have yet to be proven at scale, she said.

The bigger challenge ahead is getting the orders and then completing projects “reasonably” on time and on budget, a measure the report defines as plus or minus 20%.

The report tackles the cost and time overruns at Vogtle, which Kozeracki said “were not nuclear-specific boondoggles. … They are general megaproject issues that you see with any megaproject, from building bridges to Olympic stadiums.

“The design just wasn’t complete enough before construction began, which created a cycle of rework,” she said. “There wasn’t a detailed-enough integrated project schedule or fast-enough turnaround on” quality assurance.

Vogtle’s workforce of 9,000 at peak also created “diseconomies of scale,” the result of trying to manage “a city’s worth of people,” Kozeracki said.

One solution for bringing down costs is a “consortium approach,” said David Crane, director of the Office of Clean Energy Demonstrations, which is overseeing the advanced nuclear demo projects. As described in the report, the strategy would allow a group of companies, such as utilities, to “enter a cost-sharing agreement for the construction of multiple reactors, likely of the same design. This pooled demand would allow for sharing risk across multiple owners and could smooth the cost curve from the first reactor to the last.”

This approach also relies on a pipeline of five to 10 projects for different types of reactors, Crane said, so that “first of a kind” does not become “one of a kind.”

“If we could get an order book going for a new wave of nuclear reactors by 2025, then I think we’ll be on our way,” Crane said. “If we don’t start until [2035] … it’s virtually impossible. Given the lead time that’s associated with nuclear, we need to be moving now.”

Crane also said “Gen 3-plus” reactors ― light-water SMRs ― could be an important first step for the industry because of the “synergies” that SMRs have with advanced reactors. Kozeracki agreed, saying that light-water SMRs are a proven technology; the Navy has been using them to power nuclear submarines since the 1950s.

“SMRs may be a bit of a ‘gateway drug’ to get us back into the habit of building new nuclear in the U.S. at scale again,” she said.

Kozeracki pointed to the example of South Korea, which has built out a successful nuclear industry by “picking one design and sticking to it and building it over and over again,” she said. The country recently set a new goal for nuclear to generate more than a third of its power by 2036, up for about 27% today, and to sell 10 reactors on the international market by 2030, according to World Nuclear News.

Supply Chains

By comparison, the U.S. has 92 reactors with a capacity of about 95 GW, a fleet that generates 20% of the nation’s power and 50% of its carbon-free power. If new reactors start coming online by 2030, with a solid supply chain, the report says, the industry could reach a steady state of growth, about 13 GW per year through 2050.

But, as Crane said, even a five-year delay on deployments to 2035 could have serious impacts, requiring an annual growth rate of 20 GW per year, which could result in an overbuilt supply chain.

The challenge here is that the U.S. nuclear supply chain is adequate for keeping the existing fleet fueled but not primed for expansion or advanced technologies. For example, the U.S. has the capacity to mine and mill ― the first steps in processing nuclear fuel ― about 2,000 metric tons of uranium per year. Getting to 200 GW will require hitting 50,000 MT per year.

Other steps in the process are equally lagging, and the country has no commercial capacity at all to produce the HALEU needed for advanced reactors. Previously, the industry depended on a single processing facility in Russia for HALEU, but because of the country’s invasion of Ukraine, companies have had to quickly look for other sources.

The difference between the low-enriched uranium used in existing reactors and HALEU is each fuel’s level of the U-235 isotope needed to sustain a nuclear chain reaction. For low-enriched uranium it is around 5%, but for HALEU, it can be up to 20%.

TerraPower, the Bill Gates-funded company that is developing one of the DOE-funded advanced reactors, announced in December a two-year delay on project completion, from 2028 to 2030, because it has not been able to procure the HALEU it needed. The company’s Natrium project is to be located in Kemmerer, Wyo., near a coal-fired plant scheduled to close in 2025.

Patrick White, project manager for the nonprofit Nuclear Innovation Alliance, said the fuel supply chain for existing reactors is “robust,” but the industry would need clear demand signals before it will be ready to invest in expansion for new light-water SMRs.

“It’s a little bit of a chicken-and-egg problem,” White said in an interview with NetZero Insider. “Uranium enrichment companies and some of the other players in the supply chain [need] a clear line of sight on what their future commercial demand is going to be [so] they know that these major capital investments are worthwhile.”

He estimated that bringing new production online would take three to five years, a time frame that could support Crane and Kozeracki’s vision for an initial buildout of Gen 3+ SMRs.

For HALEU, the challenge is less the enrichment process itself, which is similar to low-enriched fuel, but making sure “your facility is designed and licensed to produce the higher enrichment,” along with some assurance of future demand, White said.

One possibility would be for the government — in this case, DOE — to be an initial off-taker of HALEU in order to guarantee production and sales to help bring companies into the market, he said.

Other recommendations in the “Liftoff” report include low-cost federal loans to suppliers to help them build capacity for future projects, as well as public-private collaboration to create a “HALEU bank,” a stockpile to meet the needs of the demonstration projects.

DOE is actively exploring other options as well. One example is the $200 million that the department provided to X-energy to build a HALEU facility to produce fuel for its XE-100 reactor, the second project in its advanced reactor program. The XE-100 uses a specialized kind of HALEU, which X-energy will produce at a facility in Tennessee. The scheduled online date is 2025.

The department also announced in November a $150 million cost-shared award to American Centrifuge Operating to install the necessary equipment at one of its plants in Ohio to produce HALEU.

Will Utilities Take the Plunge?

Storage of spent nuclear fuel is another major obstacle. According to the report, most reactors in operation are storing their spent fuel on site while the federal government tries to find interim and permanent storage locations. Previous efforts to build a spent fuel storage facility at Yucca Mountain in Nevada were abandoned after strong opposition from the state, as well as environmental and tribal groups.

According to the report, “New legislation would be required to build a federal consolidated interim storage facility or allow development of geologic repositories for permanent disposal at sites other than Yucca Mountain.” DOE is advocating for a new “consent-based siting process” to get local buy-in before attempting to build either interim or permanent storage.

DOE defines consent-based siting as “an approach to siting facilities that focuses on the needs and concerns of people and communities. Communities participate in the siting process by working carefully through a series of phases and steps with the department (as the implementing organization). Each step and phase helps a community determine whether and how hosting a facility to manage spent nuclear fuel is aligned to the community’s goals.”

White says the industry “has a clear understanding of how to keep [spent fuel] in a safe and stable state” for on-site storage at reactors currently in operation. After spent fuel has been cooled for a year or more in cooling pools, it is stored in “dry casks,” metal or concrete-encased containers.

Similar best practices should be used for newer SMRs and advanced reactors, White said.

But both interim and long-term waste storage remain open questions. “I don’t think this is something that’s technically impossible,” White said. “But I think a lot of it is making sure that we’re incorporating both the geology, the nuclear science and the social science, and [making] sure we come up with politically feasible ideas that aren’t necessarily overburdening or unfairly putting the responsibility for managing the waste on any single community.”

Like Huff and Crane, however, White stressed the importance of building a strong order book of five to 10 projects. “The challenge is we get stuck in this process of doing one-off reactors, and we don’t necessarily get the signal that we need to build out an effective supply chain,” he said.

Some utilities are planning for those first, second and third projects. In Washington state, PacifiCorp has partnered with TerraPower on the Natrium demonstration project and recently released an integrated resource plan that included two additional Natrium reactors.

The Tennessee Valley Authority is also moving forward with plans for a GE-Hitachi SMR at its Clinch River site near Oakridge. Speaking at the National Association of Regulatory Utility Commissioners’ Winter Policy Summit in February, CEO Jeff Lyash predicted TVA could build up to 20 nuclear plants by 2050.

“I have no interest in building one reactor,” Lyash said. “In order for us to be successful, TVA needs something on the order of 20 reactors over that period of time. So, if you can’t see your way to reaching nth-of-a-kind costs, supply chain, workforce, project execution for a portfolio of reactors, I don’t see the point in building one.”