When Microsoft announced in September that it was forking over $16 billion to restart a nuclear reactor at Three Mile Island in Pennsylvania, the news seemed to kick off a resurgence for the technology.
In the weeks that followed, nuclear energy remained in the headlines: The Department of Energy’s Loan Programs Office closed a $1.52 billion loan guarantee to restart a nuclear generating station in Michigan. And, both Google and Amazon announced deals to support the development of small modular reactors: Google with Kairos Power and Amazon with Energy Northwest. Just this month, SMR startup Oklo signed a (nonbinding) deal with data center operator Switch to deploy 12 gigawatts of power in the next 10 years.
But the momentum actually started much earlier in the year. In late April, Vogtle’s Unit 4 finally came online. Then, in June, TerraPower, a company co-founded by Bill Gates, started construction in Wyoming on the Western Hemisphere’s first advanced nuclear facility. And in July, the Senate passed a law aimed at expediting permitting for nuclear projects, with overwhelming bipartisan support. (Talk of restarting Three Mile Island also dates back to July, though the news initially came without Microsoft’s name attached.)
2024, though, isn’t the first time excitement around nuclear power has resurged.
In 2005, President George W. Bush visited Sandia National Laboratory in Albuquerque, New Mexico, to sign the Energy Policy Act, which outlined major support for nuclear power, including establishing federal loan guarantees for new projects. “We will start building nuclear power plants again by the end of this decade,” Bush told the audience.
2007 saw the start of a wave of applications to develop new reactors, including for new reactors at the Vogtle Electric Generating Plant. But Bush’s 2010 goal was left mostly unrealized. By the mid 2010s, most applications had been withdrawn, and Vogtle’s additional reactors were the only ones to be built. So, considering the false starts of the last two decades, can 2024 really be considered the year of nuclear resurgence?
As Boundary Stone Partners’ co-founder Jeff Navin put it on an episode of the Political Climate podcast in July, “We have been here before as an industry. This is not the first time ‘nuclear renaissance’ has been talked about.”
However, he added, there are a few key differences this time around.
“I think we’ve gotten most of what we need from policymakers,” he added, pointing to a handful of nuclear-specific legislation passed in recent years. “We’ve gotten most of what we need in terms of public support, and now it’s on us to show that we can actually build and operate these things.”
Listen to Jeff Navin’s entire interview on Political Climate:
Perhaps even more important, Navin added, is the impact of electrification — and of course, the AI boom.
“Opposition to nuclear came out of the environmental movement at the time when the primary focus…was stopping the development of things and slowing things down so that communities and people could have a voice,” he said. “We’re now in a place where the primary need for the environment is to build lots of things as quickly as possible.”
The blitz of funding and partnership announcements in 2024, he added, “is about all of the new demand for clean energy from data centers and AI.” As utilities and hyperscalers attempt to divine what the power mix for the AI era will look like, it’s clear that nuclear is going to have to play a “key role,” Navin added. “And so as we see increased energy, electricity demand, we’re going to see increased support for nuclear.”
Challenges ahead
Whether or not 2024’s return of nuclear interest can endure is far from certain. If it does, though, and actual electrons are added to the grid, it will require huge amounts of cash.
All types of nuclear come with a high cost. Even restarting existing reactors (for which there are few viable options) isn’t simple. As MIT’s Jacopo Buongiorno noted in an interview with Latitude Media, “the investment necessary to bring back these reactors is not small: on the order of a couple of billion dollars per reactor.”
Challenges for first-of-a-kind SMRs are even steeper. Notably, when the Utah Associated Municipal Power System backed out of 600 a megawatt project with the SMR company NuScale, it was largely because the initial $3 billion cost of the project had multiplied to more than $9 billion by the time it was scrapped.
In its liftoff report on advanced nuclear, the Department of Energy acknowledged that the initial cost of projects will be “higher than is economically competitive.” To mitigate that challenge, the agency proposes a buyers consortium that could take the form of aggregated offtake by tech companies. That would allow buyers “to all be ‘one fifth’ of the orderbook versus ‘waiting to be fifth,’” DOE said.
And even the interest of hyperscalers themselves may pose yet another challenge. As Dominion Energy’s Stan Blackwell explained earlier this year, there’s a certain misalignment between tech industry expectations and utility realities.
“We have data center customers…call us up and say ‘hey can you put an SMR right next to my facility?’” Blackwell said, adding that those requests reflect a misunderstanding of what’s realistic. “You’re not ever going to get the permit to put that in a major metropolitan area.”
As DOE put it, nuclear still faces “a commercial stalemate” — and breaking out of it will require navigating regulatory, financial and logistical challenges that have stymied the industry for decades.
