With a strong push from the Trump White House, the nuclear energy industry in the U.S. is, for the first time since Three Mile Island, bullish about its future. It’s about time, given that the average existing U.S. nuclear power plant was built based on 1980s technology.
A major reason for the virtual standstill in nuclear energy development in the U.S. was the Nuclear Regulatory Commission’s near-maniacal effort to reassure a skittish public that they would not issue permits to any nuclear power plant that had the potential for public harm.
The shot heard round the world signaling a change in U.S. nuclear energy policy was the summary firing of NRC Commissioner Christopher Hanson, whose divinity school background may have contributed to a perception he viewed his job as more a gatekeeper for regulatory control than a partner in building a U.S. nuclear future.
As Senate Environment and Public Works Committee chair Shelley Moore Capito (R, WV) said, “For decades, the NRC took too long, cost too much, and did not have a predictable and efficient process to approve new licenses or modernize outdated regulations.”
Newly installed NRC Chair David Wright has called the Trump directives not “just regulatory reform” but a “cultural transformation that positions the NRC to be a forward-leaning, risk-informed regulator for the future.” The agency’s internal culture is being reshaped into a more efficient and modern agency without sacrificing public safety, Wright said.
But it’s not just the NRC that is being transformed. Under presidents from Carter to Biden, nuclear was largely relegated to the closet as the primary focus was the media-driven “green energy” crusade. Wind and solar permits were issued without the cleanup requirements and prepayments mandatory for nuclear and fossil fuel facilities. Nuclear was deemed “dirty.”
The first Trump term was so mangled by political infighting (both intra-party and cross-party) that any real nuclear energy agenda lay buried among the lawsuits. In the interim, however, artificial intelligence made giant leaps and the demand for electric power for fast-growing data centers was exploding. Wind and solar cannot be relied upon by entities dependent upon 24/7/365 power – and nuclear is still viewed as the “cleaner” option vis-à-vis natural gas.
Even before Trump’s reelection, tech giants were busily signing nuclear energy deals to power their data centers. Last September the owner of the long-shuttered Three Mile Island Unit 1 nuclear power plant announced plans to restart operations in 2027, thanks to a 20-year power purchase agreement with Microsoft for a nearby AI data center.
Last October Amazon and Google both announced they would be investing in small modular reactors for AI data centers. Two months later Meta said it planned to follow suit. The amazing thing is the uncertainty that the SMR manufacturers will be able to deliver as quickly and as affordably as the tech giants demand. The simple reason? They have no track record yet. But energy demand is so high that waiting is not an option.
In the last few weeks, what was already a fast train picked up even more speed. On October 16 the U.S. Army unveiled its next-generation nuclear power Janus Program for the deployment of small modular reactors to support national defense installations and critical missions. Commercial microreactor manufacturers will partner with the Army’s Defense Innovation Unit with a goal of an operating reactor by September 30, 2028.
On October 26, Hyundai Engineering & Construction announced a basic design contract with Fermi America to construct four large nuclear reactors on a 8.1-square-mile property outside Amarillo, Texas. The Hyundai-designed AP1000 nuclear reactors will generate 4 GW for the HyperGrid complex, the world’s largest integrated energy and AI campus. The 11-GW project also includes 2 GW from small modular reactors, 4 GW from gas combined cycle plants, and 1 GW from solar and battery storage systems.
The integrated license application for the $500 billion project, the brainchild of former Energy Secretary Rick Perry and Fermi co-founder Toby Neugebauer, is currently under expedited review by the NRC. Meanwhile, Hyundai E&C is working on design tasks and preparations for the main construction phase, with finalization anticipated for an engineering, procurement, and construction (EPC) contract by spring 2026.
On October 28, Westinghouse Electric Co. joined Cameco Corporation and Brookfield Asset Management in a new strategic partnership with the U.S. government to accelerate the deployment of nuclear power. The government has committed to construction of at least $80 billion of new reactors using Westinghouse’s nuclear reactor technology to reinvigorate the U.S. nuclear power industrial base.
The government says this partnership will facilitate the growth and future of the U.S. nuclear power industry and the supporting supply chain. The entire project, which will deploy two-unit Westinghouse AP1000 reactors, is expected to create more than 100,000 construction jobs and support or sustain 45,000 manufacturing and engineering jobs across 43 states.
These are only a sampling of the active and planned contracts for nuclear power plant construction that have sprung out of the unplowed ground with the change in philosophy at the NRC and the White House. All systems are brightly lit green – but obstacles remain in the road.
Even with greatly shortened licensing timeframes, it will take time to complete site designs, obtain permits and licenses, and begin delivering much needed electricity to tech giants and other customers. Yet the biggest problem may be finding enough nuclear fuel at affordable prices to meet the mushrooming demand.
One option, says Curio CEO Ed McGinnis, is recognizing that spent nuclear fuel (including that from nuclear weapons) can safely be turned into fresh usable nuclear fuel and valuable rare metals and materials (like rhodium, palladium, krypton-85, and americium-241).
McGinnis, once a top nuclear energy official with the Department of Energy, says that only about 4% of the energy value has been used after five years of operation, and that fissioning uranium generates a plethora of other highly valuable isotopes that can be used for medical purposes, space exploration, and industrial processes.
Curio’s dry electrochemical and pyro-processing system separatees isotopes and fission products and also separates metallic elements like uranium and plutonium – each of which can be used as reactor fuels – enough, says McGinnis, to produce up to a third of the U.S. nuclear uranium feedstock a year from one facility. To boot, reprocessing dramatically shortens the half-life.
U.S. production of uranium oxide in 2024 jumped exponentially to 677,000 pounds from just 50,000 pounds in 2023, and exploration and development activities in 2023 were the highest in a decade. On a down note, anti-nuclear activists have been waging a campaign to shut down the White Mesa Mill in Utah that processes uranium ore – and in the U.S. today only about 5% of nuclear fuel has been processed domestically.
The nuclear fuel conundrum is but one of the obstacles in the path of the massive U.S. nuclear power industry growth that is also a vital component of the growth of AI data centers and other emerging electricity-hungry technologies that are shaping our future. But all systems are go – and that is the giant step that had to be taken first.
Duggan Flanakin is a senior policy analyst at the Committee For A Constructive Tomorrow who writes on a wide variety of public policy issues.
This article was originally published by RealClearEnergy and made available via RealClearWire.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Changing the approval procedure is essential. As is, the Green Blob can abuse process for over a decade.
Agreed. But the principal problem is the NRC. As long as it is rewarding for the NRC staff to obstruct or lengthen approval processes, this will continue. At NRC, there are no penalties, either institutional or personal, for simply saying “no”.
Worse, the malign influence of Barack Obama cannot be overlooked here. His administration put in place a number of regulatory obstacles to prevent cost-effective plant financing and construction..
But probably the worst element of government process in managing nuclear power licencing and construction was the appointment of an ideological antinuke as Chair of the NRC. Greg Jaczko did his best to ensure that NRC would be in no position to approve of anything nuclear. At the time of the Fukushima earthquake in 2011, Jaczko did his best to spread antinuclear hysteria across the United States.
Let’s face it: Jaczko brought no skills or knowledge whatsoever to USNRC. He was appointed simply because he was a politically influential antinuke. It is thus entirely fitting that his nuclear career ended with him mired in sexual harrassment charges by NRC staff.
The current chair is David Wright, Trump’s appointee. He’ in favor of nuclear development and streamling approval processes. Are you saying Jaczko’s influence still overshadows current NRC chairman?
AI Overview:
The damage Jaczko did lives on. He is a major part of the reason that there was no follow-on from Vogtle 3 and 4. That government can appoint such a clown for such a decisive role means that capital takes flight from any long term investment.The uncertainty created is permanent.
Jaczko was a political hack — worked for Harry Reid.
“To boot, reprocessing dramatically shortens the half-life.”
What does that mean?
Also: “and 1 GW from solar and battery storage systems.”
What’s the point?
Also, how does, “two-unit Westinghouse AP1000 reactors,” square with, “Hyundai-designed AP1000 nuclear reactors“?
But anyway, good news. I’m looking forward to seeing how the SM nuclear reactors perform. And what about the thorium option? Still viable?
All good questions. I hope we get answers.
The first of four BWRX-300 SMRs are under construction right now at the Darlington nuclear power station in Ontario.The expected first power is to be in 2028. The provinces of Alberta, Saskatchewan and New Brunswick are all part of the four-province coalition to develop these reactors. The design was completed and fully certified by the regulator the Canadian Nuclear Safety Commission. Construction began in May of this year.
Site preparation began in 2022.
As for thorium, all CANDU reactors can use thorium fuel. This fuel use was demonstrated in the late 1970s.
Of the 26 CANDU nuclear reactors in operation today* around the globe (Canada, Argentina, China, India, Pakistan, Romania, and South Korea), none use thorium fuel. There must be an straightforward reason for this.
*Note: three previously active CANDU reactors have been shutdown and Canada has decommissioned eight. This accounting does not include the CANDU-derived nuclear reactors that were designed and built by India. Only India operates a CANDU system that is not based on the CANDU 6 design.
— https://en.wikipedia.org/wiki/CANDU_reactor
“There must be an straightforward reason for this.”
Really? In a world that believes CO2 is poison and that sunlight & unicorn farts can power a modern economy you want a ‘straight forward reason’ for why an industry that has been demonized at every turn isn’t actively using Thorium in a CANDU reactor…
I read of a demonstration project in China that showed that Thorium can be used as a fuel in a CANDU reactor (though my memory is that it was a ‘combined fuel’ with Uranium). That it isn’t commercially available or widely used is a question of regulatory constraints and likely market dynamics (e.g. its probably still cheaper just to use Uranium). As well when used as a fuel some of the Thorium is turned in to U233 which can be used to make those things that go ‘boom’ REALLY loud…nobody likes loud noises.
The original posters point is still true, but there’s no rush to convert to the use of Thorium.
Yes, really!
Most of us WUWT readers have heard all the advantages cited by proponents of using thorium to fuel a fission nuclear reactor, such as
— Thorium is three to four times more abundant in the Earth’s crust than is uranium, with attendant claims that it can provide a rapid, long-term energy solution and some estimates that US alone has enough thorium deposits to power the country for over 1,000 years.
— Thorium-fueled fission reactors would produce significantly less high-level, long-lived radioactive waste compared to conventional uranium-fueled fission reactors. The waste produced from using thorium would be less toxic and its radioactivity would drop to safer levels within a few hundred years, as opposed to the tens of thousands of years for uranium-originated nuclear power plant waste.
— The U-233 produced in a thorium fuel cycle is contaminated with U-232, which emits strong gamma radiation. This makes the material difficult to handle and easy to detect, presenting an intrinsic barrier to the illicit use of the fissile “waste” material for nuclear weapons production.
If the above were all true, what’s not to like about thorium as a nuclear fuel???
So, I suspect the straightforward answer is most probably related to the old adage “follow the money” . . . that is, thorium-fueled nuclear power plants will not be significantly cheaper than current uranium-fueled nuclear power plants on a life-cycle-levelized cost per delivered kWh.
So, I suspect the straightforward answer is most probably related to the old adage “follow the money”
And? How does this negate the statement that Thorium can be used in a CANDU reactor? And to be clear I noted “e.g. its probably still cheaper just to use Uranium”…that it is still more cost effective to use Uranium does not negate ANY of the statements you ascribe to proponents of the use of Thorium, or that Thorium has been demonstrated to be usable in a CANDU style nuclear reactor. When it becomes time that humanity needs to use Thorium we will.
Push comes to shove, there’s no rush, but first the nuclear industry needs to recover from the unfathomably inaccurate negative campaign against it.
Hmmmm . . . the last nuclear power plant reactor to begin operating in the US was the Vogtle Unit 4 in Georgia, which started commercial operation on April 29, 2024. This was the second new reactor to come online at the Vogtle Electric Generating Plant, following Unit 3 in July 2023.
Meanwhile, outside the US, here is a listing of just some of the most recent nuclear power plants to begin commercial operations:
— in France, Flamanville 3, began feeding power to its grid in December 2024,
— in United Arab Emirates, Unit 4 at the Barakah Nuclear Power Plant began feeding power to its grid in September 2024; the preceding reactor, Unit 3, began feeding power to the same grid in February 2023,
— in India, the following new reactors began feeding power to various grids: Rajasthan Unit 7 in April 2025, Kakrapar Unit 4 in March 2024, and Kakrapar Unit 3 in June 2023
— in China, the following new reactors began feeding power to various grids: Zhangzhou-1 in January 2025, Fuqing Unit 7 in December 2022,
and Shidaowan (world’s first high-temperature gas-cooled reactor) in December 2023,
— in Belarus, Unit 2 (Astravets 2) began feeding power to its grid in November 2023,
— in Slovakia, the Mochovce 3 rector was operational in October 2023, but it is unclear if it is currently feeding electrical power into a grid,
— in Finland, Olkiluoto 3 began commercial operation in May 2023.
So, If I’ve done the math correctly, there have been 14 new nuclear reactors in power plants that have become operational (at least 13 of which are in commercial operation feeding electricity into grids) in just the last three years.
The nuclear industry needs “a recovery from a negative campaign”? No . . . if one has the money and the time and the right politics, one can currently play in the nuclear power plant game, even in the US.
Also this: since the nuclear industry also supplies reactors to the US Navy, the last I heard there was no issue whatsoever (other than total vessel cost) with the Navy building new submarines and aircraft carriers powered by nuclear reactors.
Hmmm . . . as the saying goes: that was then, this is now. Here is what the
GE Vernova-Hitachi website (https://www.gevernova.com/nuclear/carbon-free-power/bwrx-300-small-modular-reactor ) currently states, with my bold emphasis added:
“GE Vernova Hitachi’s (GVH) BWRX-300 is not a concept, a plan, or a PowerPoint. It’s now under construction. Its configuration, based on proven technologies, has been selected for the leading project in the Western world with construction underway in Canada on the first BWRX-300. Construction of the first unit is estimated to be completed by the end of 2029 and in commercial operation by the end of 2030.“
Hmmmm . . . first power in 2028, or maybe in 2030 . . . has it already suffered a two-year schedule slip in just the three years since construction began?
I should query Google’s AI bot to see how it and its cousins like waiting to be fed. /sarc
All “AP1000”-based designs (including China’s CAP1400) are derivatives of Westinghouse technology. In the case of China, they purchased the intellectual property from Westinghouse (a requirement imposed on Westinghouse to be allowed to build a few units in China — the only units that Westinghouse will ever build there) and then stripped out any technology that Westinghouse still had ownership of and replaced it with “Engineered in China” parts.
This type of thing is not uncommon in Asia. For example, South Korea’s current nuclear infrastructure is based on technology originally developed by American company Combustion Engineering.
Still viable, but nobody is currently working on a serious Thorium-based design. It is difficult enough to get HALEU (High-Assay Low Enriched Uranium) fuel these days, and that is based on the traditional uranium-fuel infrastructure, which has been around for over 70 years.
“To boot, reprocessing dramatically shortens the half-life.”
What does that mean?
Reprocessing removes the relatively long lived radioactive constituents (Uranium, Plutonium etc.) for reuse in a power plant leaving only the relatively short lived constituents. Thus the half-life of the actual waste (stuff that can’t be used in a power plant or otherwise) is significantly less than the original depleted fuel.
I’ve been pointing out the huge lost opportunity of this type of ‘recycling’ since the ’80s and try not to even call it ‘waste’. Of course if the cost of reprocessing is higher than just digging up raw material & processing it then at any given time its not worth it. At the very least though it can be safely stored until its cost effective to recycle.
Also: “and 1 GW from solar and battery storage systems.”
What’s the point?
Hell if I know but how about we go with ‘green washing’…
Also, how does, “two-unit Westinghouse AP1000 reactors,” square with, “Hyundai-designed AP1000 nuclear reactors“?
Quick searching shows that ‘AP’ stands for ‘Advanced Passive’, so two companies designing a similar ‘Advanced Passive’ nuclear reactor supplying 1000 MW (1 GW) of power (or at least that’s what I think the 1000 stands for). Nothing odd that 2 companies are competing to deploy similar technology, that’s a good thing.
Just to follow up on the ‘shortened half-life’ idea. This is a HUGE missed opportunity of the nuclear industry. The environmentalist scream bloody murder about the ‘highly radioactive waste that lives for millennia/thousands of years’. This is just outright FALSE. The long lived isotopes are NOT ‘highly radioactive’, its the short lived isotopes that are (e.g. they decay faster giving off their radiation more quickly).
The long lived isotopes are LOW level radioactive materials and can be ‘recycled and reused’.
Ultimately the ‘highly radioactive’ materials in the spent fuel will decay in 1 human life-time. The long lived material which isn’t highly radioactive can be reprocessed and recycled. If the nuclear industry wasn’t scared of its own shadow they’d mount a marketing campaign to correct the record & demonstrate how grand they are by ‘recycling’.
BTW, even some of the short lived highly radioactive material such as Iodine-131 can be used in medical procedures so maybe they could reprocess the spent fuel to capture this element & remove it from the final ‘waste’…perhaps its too costly but its doable in theory.
Westinghouse developed the original AP1000. They sold the base technology to companies in China and South Korea when they ran into financial difficulties. China and South Korea have changed the design enough to claim that it is not a copy and are selling the design to other countries.
While I see a decent future for SMRs, I don’t see them being deployed fast enough to provide power for the initial large AI requirements. Re-commissioning a moth-balled plant may be possible, but in the US, the first SMR isn’t scheduled to go live until 2030. And there’s a lot of potential obstacles between now and then. First, the typical delays and missteps of deploying the first of anything. But you can bet environmentalists will try every possible lawsuit to stall deployment, or at least increase the costs enough through law-fare that future deployments are deemed uneconomical.
Then there are the judges. Our judicial system has been compromised by ideological jurists who rule as they wish, not as the law dictates. Yes, they are overturned eventually, with both a time and cost penalty charged to the defendants,
But the big fear will be the 2028 elections. If Democrats are voted back in power, there may be no SMRs at all deployed. All of this may push many new AI data centers to build NG power plant facilities ASAP.
The nuclear renaissance is good news, but I am kind of worried that it is being driven by a frenzied AI schedule instead of a deliberate sifting of options to converge on a solution or solutions that hit the sweet spot between cost, safety, utility and deliverability.
More and more the whole AI thing is looking like a bubble to me and if someone ends up fielding a crap nuclear solution that has major flaws just because it can meet schedule it could set the whole nuclear industry back another 50 years. That would be a real shame.
There are a lot of interesting technology choices out there, but many of them are nascent technology/designs that need to be well proven in a deliberate manner.
Not so. In Ontario, new nuclear construction at Darlington, and still to come at Bruce, is being driven by electrical load growth entirely separate from any new applications such as AI.Natural load growth is the single factor driving all new nuclear planning and construction in Alberta, Saskatchewan and New Brunswick.
More good news.
We continue to build nuclear here in Ontario because it is practical and long term, unlike wind and solar, which take up too much land. We were saddled with wind and solar in some parts of the Province, even when the locals did not want those installations, but that fervour seems to have abated for the moment.
The question for newly installed NRC Chair David Wright, as he positions the NRC to be a “forward-leaning, risk-informed regulator” is “Which of the following deserves the most attention:”
— proven, but updated, PWR fission reactor designs and builds?
— unproven but highly marketed SMR fission reactor designs and builds?
— unproven but greatly discussed molten salt reactor (MSR) designs and builds?
— yet-to-be-demonstrated-at-commercial-scale fusion reactors, said to be only 20 years away, again?
So many choices, so many AI-directed dollars waiting to be spent!
When the EPA gets rid of the ‘Linear no threshold’ rule for radiation safety I’ll start to believe nuclear power will have a renaissance.
The EPA should have done this at the same time as issuing its reversal on the CO2 ‘Endangerment Finding’.
LNT is as farcical as CO2 being a pollutant, there is 0 scientific evidence for it and plenty of scientific evidence that small amounts of radiation are vital to human survival (not to mention evolution!)…