In July, we wrote that Spain was mistaken in preparing to close one of its most reliable sources of clean power.

Spain just partially agreed.

On August 12, the Ministry for the Ecological Transition and the Demographic Challenge renewed the operating licenses for both Almaraz units through June 8, 2030, pushing Unit I’s closure back by roughly two and a half years and Unit II’s by about a year and a half. The order was signed by Vice President and Minister Sara Aagesen.

This is real progress, and it belongs to the people who fought for it. But it is not the end of the phaseout. Spain’s plan to shut down its entire nuclear fleet by 2035 remains intact.

What Changed, and What Didn’t

The extension followed a formal request submitted in October 2025 by Almaraz’s owners (Iberdrola, Endesa, and Naturgy) and a favorable report from Spain’s Nuclear Safety Council (CSN) in July.

The ministry’s own reasoning is worth reading closely. It described the move as a limited extension, adopted in light of uncertainty in international energy markets driven by conflict in the Middle East and the continuing war in Ukraine. It argued that keeping Almaraz online could help shield Spanish consumers and the electricity system from sudden spikes in fossil fuel prices.

That is, almost word for word, the argument advocates have been making for years. The government is now making it too.

An Insult to the Planet

In our July piece, we noted that the PNIEC never assumed a clean handoff from nuclear to renewables. It anticipated a larger near-term role for combined-cycle gas plants as “flexible backup.”

The government’s sensitivity analysis of this extension confirms it. Keeping Almaraz online reduces gas-fired generation from combined-cycle plants by roughly 7% in 2030 compared with the PNIEC scenario. The same analysis projects renewable generation about 1.4% lower. This is a point critics have raised, however, a 7% cut in gas burn against a 1.4% shift in renewable output should be an easy trade for anyone who is concerned about carbon emissions.

In other words, Spain’s own numbers show that keeping a nuclear plant open means burning less gas.

Is This a Stalling Tactic?

Here is a possibility worth taking seriously: the extension may be buying time towards a long term extension.

A temporary extension gives the government room to respond to an energy crisis without formally breaking with coalition partners whose platforms oppose nuclear power. It lets Madrid claim continuity on the 2035 phaseout while quietly relieving the pressure and allowing time for the continued growth of nuclear energy support. If political conditions shift (a new coalition arithmetic, another price shock, another blackout) the groundwork for a longer extension is already laid.

That reading is speculative. But if this is a stalling tactic toward a longer extension, sustained public pressure determines whether the next step gets taken and how far it goes.

Advocates cannot afford to relax. The extension is already being contested from the other side: the environmental organization Adenex has announced it intends to challenge the decision in court.

Phaseout Deadlines Hollow Out Plants Before the Date Arrives

Nuclear plants run on people. Reactor operators, welders, instrumentation technicians, and licensed engineers take years to train and can’t be hired off a shelf. The CSN’s own extension conditions require Almaraz to maintain adequate staffing. This is a demand the phaseout policy actively undercuts. No talented young Spaniard finishing an engineering degree builds a career in an industry their government has scheduled for demolition.

What Diablo Canyon Teaches About Not Letting Up

California’s Diablo Canyon Nuclear Plant is the clearest available lesson in what happens after a reprieve.

In 2016, PG&E planned to retire both units by 2025. Years of technical analysis, coalition-building, and on-the-ground organizing by groups including Generation Atomic, Mothers for Nuclear, Stand Up for Nuclear, Native Nuclear, and many others helped produce SB 846 in 2022, which authorized operation through 2030.

Advocates continued to push. In April 2026, the U.S. Nuclear Regulatory Commission renewed both units’ federal licenses, clearing them to run until 2044 and 2045.

Federal approval, though, isn’t sufficient. California law still caps operation at 2030, and going beyond that requires the state legislature to act. In August, Governor Newsom said he wouldn’t push for that extension in his remaining months, calling it a baton to pass to the next administration. Meanwhile 54 elected officials signed a letter urging the legislature to act this year, while a growing coalition continues to press the case.

The pattern is instructive. Every advance at Diablo Canyon came from advocates who treated the previous win as a starting point rather than a finish line. And every advance was followed by a new deadline that required the same work all over again. Almaraz is now at that stage.

The Fight Isn’t Over

For the advocates who have spent years fighting for Almaraz, this is not the time to stop.

The extension proves something important: Spain can change course. And it proves that the pressure worked.

If Almaraz continues to operate safely, reliably, and economically, Spain should be prepared to extend its operation again. The goal should be to recognize Almaraz for what it is: a valuable piece of clean energy infrastructure that should be allowed to operate for as long as it can safely and economically do so.

This is the beginning of the next chapter in the fight for Almaraz.

Want to help? Here’s how to keep the pressure on:

• Join Generation Atomic’s global community of nuclear advocates. Become an Atomic Ambassador or join one of our volunteer teams.

• Sign the petition: Sí a Almaraz, Sí al Futuro

• Follow the movement: Sí a Almaraz, Sí al futuro | Jovenes Nucleares

Subscribe now

If nuclear power didn’t cross your mind while celebrating Fourth of July weekend, you’re not alone. But while the rest of us were downing cans (or in my case, mojitos), the nuclear industry was racing towards a milestone even more important than the nation’s 250th anniversary: the criticality deadline for participants in the Department of Energy’s Reactor Pilot Program.

Last May, an executive order directed the DOE to bring at least three brand-new advanced reactors to criticality by July 4th, 2026, built and tested entirely outside the national laboratories that have historically done this kind of work. It’s worth noting that criticality isn’t the same as generating electricity – it just means that a nuclear core has successfully split atoms and achieved a self-sustaining chain reaction. It’s the moment a reactor stops being a very expensive arrangement of metal and graphite and starts being, technically, a reactor. Everything else, from electrons to revenue, comes later.

Designing, licensing, fueling, and starting up a nuclear reactor in under fourteen months is not something the American nuclear industry has done in the last fifty years. And when this announcement came out, most of the industry called it a pipe dream. The industry can’t even roll out a single reactor in five years, let alone multiple.

But miraculously, it did. Four times over.

Let’s talk about who crossed the line, who finished by the skin of their teeth, and who is still, by any measure, a long way from a working reactor.

Starting off strong is Valar Atomics. They didn’t actually achieve criticality first, but they’re easily the furthest along (at the time that this article is posted). Valar’s Ward 250 is a scaled-down version of what will be a 5-megawatt, helium-cooled high-temperature gas reactor using TRISO fuel. The reactor itself was built at Valar’s facility in Southern California, and in February, the Air Force flew it out to Utah aboard a trio of C-17 cargo planes. It reached zero-power criticality on June 18th at the company’s test site in Emery County, Utah, making Valar the first company in the program to achieve criticality outside a national laboratory altogether. By June 23rd, the company began power ascension, and on July 1st, they successfully powered a Nvidia Blackwell chip with their reactor. Valar and Nvidia used the same event to announce a feasibility study for a 30-megawatt, largely waterless AI data center on the Utah site, pairing Valar’s helium cooling loop with Nvidia’s closed-loop cooling tech. In theory, it would avoid millions of gallons of water that conventional data centers burn through annually. It’s part of the reason that Valar got the most applause, even though they weren’t first to achieve criticality.

That honor belongs to Antares, whose Mark-0 microreactor achieved zero-power criticality on June 4th at Idaho National Laboratory. It was the first privately developed, non-light-water reactor to go critical in the U.S. in more than four decades, which is a big deal for an industry that was all but declared dead after Fukushima. The reactor is sodium heat pipe-cooled and TRISO-fueled, bearing a closer resemblance to Cold War-era heat pipe research than the pressurized water designs that make up most of the existing American fleet. Antares’s plan is criticality in 2026, electricity in 2027, and power to military bases by 2028. We see it happening.

Third across the line was Deployable Energy. Their 1-megawatt gas-cooled microreactor is called Unity, which the company views as more of a “nuclear battery” than a power plant. The reactor was assembled in Houston, Texas, and driven to Idaho National Laboratory in… a Ford pickup truck. Unconventional, but a good indication that the technology is living up to its purported transportability. The company says it went from project kickoff to criticality in roughly 150 days, which is a timeline that outdoes even fellow participants.

Narrowly scraping the deadline is Austin-based Aalo Atomics. On June 25th, Energy Secretary Chris Wright appeared in person to authorize the startup of Aalo’s 10-megawatt Aalo-X, a sodium-cooled reactor fueled with low-enriched uranium. Days later, the company flew its roughly 200-person staff to its two-acre site outside Idaho National Laboratory, set up Airstream trailers, and livestreamed technicians loading uranium fuel bundles into the core. Their full-scale test core reached zero-power criticality at 12:20 a.m. Mountain time on July 4th, making them the largest reactor to hit the July 4th deadline.

While these successes were major, not everyone made it in time.

Radiant Industries’ Kaleidos reactor, for one, missed the July 4th cutoff, but still has an interesting summer ahead. Their design is a 1-megawatt, helium-cooled microreactor, built to fit inside a shipping container. The company received its safety approval from the DOE back in February, and on July 1st received its first batch of fuel at Idaho National Laboratory’s DOME facility. But they won’t declare success once hitting criticality, like the rest of the group – their short-term goal is 150 hours of full power, with zero operator intervention. Even in a crowd of overachievers, that’s ambitious. The company already has a deal to deliver reactors to Buckley Space Force Base, a 20-reactor agreement with a data center operator, and a factory under construction in Oak Ridge, Tennessee, which is targeted to crank out fifty reactors a year. Missing a symbolic date by a few weeks may end up mattering a lot less than what they’ll accomplish soon.

Close behind is Last Energy, whose design uses the same conventional pressurized water reactor technology that’s powered Navy submarines and most of the existing U.S. fleet for decades. Their current entry is a 5-megawatt test unit, with a 20-megawatt commercial version planned down the line. Since they’re not betting on exotic coolant or fuel geometry like the rest of the group, they may move through the regulatory process quickly. The company has finished building its pilot facility at Texas A&M’s RELLIS campus, installed reactor components, and submitted its final safety analysis to the DOE, with criticality expected sometime this summer (pending approval).

Next is Oklo, who you may know for dominating stock market headlines over the last year. This past February, they hit a market cap of $18 billion – a curious feat for a company that hasn’t generated a single electron. Today, they’re worth around $9 billion, and have more entries in the pilot program than anyone else. Their flagship is Aurora, a 75-megawatt liquid-metal-cooled, metal-fueled fast reactor that broke ground at Idaho National Laboratory back in September 2025. The DOE approved its preliminary safety analysis in June, but they haven’t released a firm criticality date. A second project, Atomic Alchemy, secured safety approval for an isotope production reactor in Lockhart, Texas, but has been stuck in authorization review since May. Their third project, Pluto, is a sodium-cooled fast test reactor that aims to use Cold War-era surplus plutonium, but seems to be in very early stages.

A handful of others are further back still. Deep Fission broke ground in December, drilling a 6,000 foot-hole in Parsons, Kansas that will eventually hold the company’s Gravity reactor. Installing the reactor underground will reportedly eliminate the need for a traditional containment building, and will lessen the total land footprint. They cleared their initial safety agreement back in November, and are now targeting full-power operation sometime in early 2027. Natura Resources is continuing development of a molten salt research reactor at Abilene Christian University in Texas, with plenty of salt loop testing but no announced date for criticality. And Terrestrial Energy, a Canadian molten salt developer, has yet to offer a firm timeline. They have an agreement to build at Texas A&M, but appear to be using the program as more of a foothold into U.S. licensing than a race to win.

But back to the four companies that pulled this off. What does this milestone actually mean?

First, that they beat expectations. Fourteen months ago, this goal was considered extremely unrealistic, if not laughable. The DOE’s expedited licensing played the largest role in the accelerated timeline, but we also have to give credit to these companies for, well, busting ass.

While praise is well-deserved, it’s also true that criticality is the easy part (relatively speaking). What comes next is ramping to full power, running for months or years without incident, trading a DOE research authorization for an NRC commercial license, and finding a customer willing to bet (a lot of) money on a first-of-a-kind machine. This point is where almost every advanced reactor concept in American history has gotten stuck.

We could go a number of ways from here. The program could serve as intended and kickstart the commercialization of tiny-reactor technology, or it could result in a nice ‘ole defense contract, avoiding the civilian market entirely. It’s also possible that this so-called “innovation” never works once scaled, or that a company finds themselves amidst a scandal, goes bankrupt, and fizzles out. It’s happened before.

Regardless of what comes next, it’s easy to see that the American nuclear power industry is in better shape than it’s been in decades.

Check out the full video analysis on our YouTube, TikTok, or Instagram.

*Update August 19, 2026: Spain has agreed to extend Almaraz’s operation through June 2030. For our full breakdown and thoughts on the extension, read our newest article.

At a moment when energy security has never mattered more, Spain is preparing to close one of its most reliable sources of clean power.

The Almaraz Nuclear Power Plant has quietly powered Spain for over four decades. Construction began in the 1970s; Unit I entered commercial service in 1983, Unit II in 1984. Notably, 81% of the plant’s construction was carried out by Spanish companies, making it a point of national industrial pride from the start. Today, Almaraz supplies more than 7% of the electricity consumed in Spain. This is the equivalent of powering 4 million homes, around the clock, every day of the year, regardless of weather.

That outsized contribution isn’t unique to Almaraz. Nationally, nuclear power plants made up just 5.28% of Spain’s installed generation capacity in 2025, yet they produced 19.06% of the electricity generated that year. Few other sources deliver that much output from that small a footprint, a clear indicator of nuclear’s high capacity factor and its value as always-on baseload power.

Despite that record, Spain has scheduled Almaraz for closure in 2027 (Unit 1) and 2028 (Unit 2), part of a full nuclear phaseout across the country by 2035.

Why Spain Wants to Close It

Spain’s phaseout is often framed around a familiar set of critiques: long-term waste storage, the rising cost of maintaining “aging” reactors, and a public that grew more cautious after Fukushima. In reality, it’s more complicated.

The schedule traces back to 2019, when the government reached an agreement with the plants’ owners and ENRESA, the public company responsible for radioactive waste management and decommissioning, to phase out the country’s nuclear fleet. That agreement was approved as part of Spain’s National Energy and Climate Plan (PNIEC) in 2021, then ratified and submitted to the European Union in 2023.

Two forces shaped it. Politically, the governing coalition (in office then and, as of July 2026, still today) depends on the support of parties with strong environmental platforms, and many of Spain’s most influential environmental organizations have long opposed nuclear power in favor of a wind-and-solar transition. For many within the coalition, the phaseout reflects a principled commitment to a fully renewable future, even if the practical trade-offs are still being worked out. Economically, the picture looked very different in 2019: pre-pandemic electricity prices were low, and extending the plants’ life wasn’t seen as attractive. That changed dramatically after the pandemic, and even more so after Russia’s invasion of Ukraine. Energy prices spiked, energy security jumped back to the top of the agenda, and many of the assumptions behind the 2019 agreement started to look outdated.

Policy has to be judged by outcomes, and the evidence from across Europe raises serious questions about what happens when nuclear plants go offline and renewables alone are left to fill the gap.

A Legacy That Still Shapes the Debate

Spain’s caution toward nuclear runs deeper than Chernobyl or Fukushima. In the 1980s, the country adopted a nuclear moratorium that halted construction on ten reactor units already planned or underway. That decision was a turning point that triggered a long, gradual decline in nuclear development and public expertise.

The effects still linger. Public understanding of nuclear technology in Spain remains thin, and for decades the conversation has framed nuclear power around its risks while renewables became closely associated with progress and sustainability. In many ways, Spain has pursued a path like Germany’s Energiewende, aiming for a fully renewable electricity system by 2050.

So the debate over Almaraz isn’t really about nuclear power alone. It’s about how to balance climate goals, energy security, affordability, industrial competitiveness, and technological neutrality all at once. This debate has only sharpened since the war in Ukraine and the closure of the Strait of Hormuz, with more voices asking whether closing well-performing reactors is the right call.

A Gap That Has to Be Filled by Something

Renewable energy is meant to be the foundation of Spain’s clean energy future, and on paper there’s a roadmap: the PNIEC leans heavily on an ambitious expansion of renewable capacity to offset the phaseout. But even the most ambitious buildouts take time, and in practice, renewable deployment, grid connections, and system integration have all fallen well behind the plan’s own assumptions.

Tellingly, the PNIEC’s modeling doesn’t assume a clean handoff from nuclear to renewables. It explicitly anticipates a larger role for combined-cycle gas plants in the early years of the phaseout, using gas as “flexible backup” to keep the system stable while renewable capacity, storage, and grid flexibility scale up, with gas generation expected to rise before gradually declining again toward 2030 as the system matures.

In other words, even Spain’s own plan for replacing Almaraz counts on more fossil fuel generation in the near term. That matches the pattern across Europe: when nuclear capacity goes offline faster than renewables can replace it, gas fills the gap - the same gas whose price volatility hammered households after the war in Ukraine and the US bombing of Iran sent energy costs spiraling to record highs.

Nuclear plants like Almaraz provide “firm power”: electricity available on demand, regardless of conditions. France weathered the energy crisis better than most thanks to its large nuclear fleet, while Germany, which closed its last reactors in 2023, had to lean harder on coal and gas imports.

The lesson isn’t lost on Europe’s other policymakers. Belgium reversed its phaseout, the Netherlands is building new reactors, and the UK, France, and Finland are all investing in next-generation nuclear. Spain, meanwhile, is moving in the opposite direction, guided by a plan written in a world that no longer exists.

A Lifeline for Local Communities

For the towns surrounding Almaraz, this isn’t a policy debate. It’s a countdown.

The plant supports roughly 4,000 jobs and accounts for around 5% of the regional economy. Businesses, schools, and housing markets have grown around the assumption that the plant would remain. As one restaurant owner who serves hundreds of workers a day put it plainly: if Almaraz closes, so does most of his operation.

Energy transitions are often discussed in terms of carbon targets and megawatts. But behind those numbers are real communities who deserve more than a line on the calendar.

The Case for Changing Course

Extending Almaraz’s operational life would cost a fraction of building the equivalent power in new generation. The infrastructure already exists, the workers are already trained, and Spain’s own grid operator, Red Eléctrica, has warned that the phaseout poses real risks to supply security.

The concerns about waste and cost are real, but closing Almaraz doesn’t make the waste disappear, and Spain already has a plan for it. The 7th General Radioactive Waste Plan charts a path toward a permanent solution using a deep geological repository; shutting the plant only cuts off the electricity while that storage work continues. Meanwhile, greater fossil fuel dependence carries its own cost in both emissions and household energy bills.

Keeping Almaraz running doesn’t mean slowing the clean energy transition. It means completing it without leaving communities and households exposed.

Take Action

Sí a Almaraz, Sí al Futuro was founded in December 2024 by local mayors, residents, workers, and community organizations who refused to accept that their region’s future had an expiration date. In January 2025, more than 10,000 people took to the streets in one of the largest demonstrations the region has seen, and the movement has kept growing since, gathering signatures and pushing petitions now under consideration at the European Parliament.

That momentum builds on years of quieter groundwork. Jóvenes Nucleares began drawing attention to Almaraz as early as late 2023, when the topic had little public visibility, and in 2024 launched the Salvemos Almaraz initiative, now central to its outreach across Spain. Its work spans technical talks in schools and universities, introductory courses on nuclear science, informal sessions in social spaces like bars, visits to the Almaraz region and the plant itself, information stands, street outreach, digital content and videos, and even an appearance before the Spanish Congress of Deputies. These efforts are aimed at reaching people of every background and level of technical knowledge. The group has also carried Spain’s case into the broader Young Generation Network (YGN), building ties with counterpart groups across Europe.

This has been a sustained effort, years of early awareness-raising combined with continuous outreach, and there are several ways for you to get involved right now:

• Sign the petition: Sí a Almaraz, Sí al Futuro

• Learn more about the movement: sialmaraz.es

• Read the position paper and explore more ways to help: Jóvenes Nucleares’ Salvemos Almaraz page (available in several languages)

• Become a member of Jóvenes Nucleares to support its advocacy work and keep the campaign ready to act as the situation evolves: hazte socio

• Sign the Spanish Nuclear Society’s manifesto: ManifiestoNuclear-SNE.pdf

• Sign Alfredo García’s (”Operador Nuclear”) public petition: Change.org petition

All movements listed share regular updates and videos on the common social media channels.

Co-Author: Cristina Domingo Jiménez | Nuclear Engineer | Responsible of the Salvemos Almaraz initiative within Jóvenes Nucleares

Check out our video on this on YouTube, TikTok, or Instagram.

Subscribe now

From Nebraska to New York, Texas to Tennessee, and now potentially California, the list of states asking that question keeps growing fast enough that trackers can barely keep up. The latest count puts it at 31 states with an active, completed, or pending nuclear study, working group, or task force on the books, plus eight more weighing study-related bills behind them. The map below has the full state-by-state rundown.

32 states have launched nuclear studies or assessments, with 8 more study-related proposals under consideration (NARUC-NASEO, GAIN, Generation Atomic Policy Research)

What’s notable isn’t just the number of states; it’s which ones. These aren’t all the same kind of state with the same kind of politics or the same kind of grid, and a deep-red farm state and a deep-blue island chain can end up running through the same basic exercise for very different reasons. That kind of spread is usually a sign that something has shifted underneath the politics, not just within them.

Roll call

Minnesota approved a $500,000 study to look at whether new nuclear plants could be built in the state, with specific engagement planned for the Prairie Island Indian Community, which already lives next to the existing Prairie Island plant and its spent fuel storage site and has a direct stake in how the state handles nuclear going forward.

Michigan spent $250,000 on a study covering its existing nuclear resources and expertise, costs, technology options, integrated energy systems, and the policy questions that come with all of it.

Nebraska put $1 million toward screening potential sites across the state for advanced reactors, while Indiana teamed up with Purdue University to dig into small modular reactors, looking at cost, workforce, safety, community engagement, and the rest of the picture. Montana asked whether advanced reactors could step in as coal plants retire, and Florida studied the technical and economic case for advanced nuclear, including SMRs.

That’s just a sample. Arizona, Texas, Pennsylvania, and New Hampshire are all running their own version of the same study, and so are more than two dozen other states, spanning red, blue, and everything in between. The map up top has the full rundown if you want to check your own state. Different states are starting from different places and for different reasons, but the underlying approach tends to look the same: settle on a specific question worth answering, put a real budget behind it, and find out what the answer actually is before committing to anything larger.

Hawaii is a useful example of where that process can lead. The state studied the question and concluded that nuclear doesn’t make sense there right now, pointing to its small grid, cost, and other local barriers. That result is worth noting on its own, since it shows the process is capable of producing an honest no rather than simply clearing the way toward a yes that was decided in advance.

What Makes a Good Study?

A feasibility study isn’t supposed to be a foregone conclusion, and Hawaii’s result is a good example of why that matters. There’s a real difference between asking why nuclear is too expensive and asking what factors affect nuclear construction costs in the first place. One of those questions has already decided the answer before the research begins, while the other is actually willing to find out.

A good study works through that question honestly. It’s independent, transparent, and built around the specific state it’s studying rather than copied from somewhere else. It assesses nuclear in comparison to alternatives like gas, wind, solar, geothermal, storage, transmission upgrades, demand response, and geothermal in a real-world context. And it will work through cost ranges, timelines, risk, workforce needs, and what the local community actually thinks before drawing any conclusions. A bad study skips most of that work and treats nuclear as either magic or impossible from the start.

That distinction matters because of what happens next. A study that’s rigged in either direction rarely survives contact with the next legislative session, the next utility rate case, or the next community meeting, since the numbers eventually get checked against reality. A study that’s willing to land on no, the way Hawaii’s did, is the one that ends up with enough credibility to be trusted the next time it lands on yes somewhere else.

Even California!

California might be about to take its first real swing at a nuclear feasibility study in decades, and that’s notable given the state’s history. California has a long anti-nuclear track record, an existing legal moratorium on new construction, and a recent, very public fight over whether to extend the operations of Diablo Canyon, one of its largest sources of clean electricity. Now it’s weighing AB 2647, a bill that would direct the California Energy Commission to put together a full assessment of what role advanced nuclear could play in the state’s energy mix and critical infrastructure needs.

If that bill passes, it would be a genuinely significant shift for a state that essentially invented the modern anti-nuclear movement and, only a few years ago, came close to shutting down a source that provides roughly a sixth of its low-carbon power. A feasibility study wouldn’t undo any of that history, but it’s a meaningfully different conversation than whether to close Diablo Canyon, and that shift is worth paying attention to on its own.

Part of why a study is the version actually moving through the legislature is that it’s a much smaller political lift than the alternatives. AB 2647 is a good illustration of that: it actually started out as a bill to lift the moratorium outright, then got gutted halfway through the session and rewritten as a study bill instead. The version that’s left doesn’t ask anyone to vote for building a reactor, repeal the moratorium, or pick a winning technology. It just asks the Energy Commission to put real numbers on the table, which is a far easier thing to agree to than full-on legislation.

California isn’t out on its own here, either. It’s one of eight study-related proposals currently working their way through state legislatures, alongside places like New York, Washington, and Utah, each one stopping short of an actual project and asking for the numbers first.

Why are States Doing It?

The short version is that electricity demand is climbing, coal plants are retiring on schedule whether or not anything is ready to replace them, and grid reliability is getting harder to maintain on weather-dependent power alone.

And then there are the data centers. A growing list of tech companies with deep pockets have suddenly become very interested in firm, round-the-clock electricity, and they need it now, not whenever the next big transmission project clears permitting.

Put those pressures together, and a feasibility study starts to look less like bureaucratic caution and more like ordinary due diligence. States need to plan for rising demand without much certainty about what’s retiring or what’s going to take its place, and a data center customer willing to sign a long-term contract creates a different kind of urgency than a vague policy commitment ever would.

There’s also a simpler logic running underneath all of it. Jobs and tax revenue are an easier sell to a legislature once a study shows the economics actually work, and not many lawmakers want to be the one who voted to spend taxpayer money on a reactor before anyone checked whether it made financial sense in the first place.

None of this guarantees that any particular study leads to an actual reactor, and some of them, like Hawaii’s, will lead nowhere at all. That’s a perfectly fine outcome, since the point was never to guarantee a yes. The point was to get an honest answer.

The real win isn’t the study itself. It’s what happens afterward, when a state moves from nuclear curiosity to nuclear policy with real numbers attached instead of just enthusiasm. Thirty-one states already running a study, with eight more about to start one, is a very different starting point than the nuclear industry has been at in decades, and now we get to see which of them actually follow through.

Is your state on the list yet? Let us know in the comments!

Check out our video on this on YouTube, TikTok, or Instagram.

Subscribe now

When detractors criticise nuclear, they often talk about costs, claiming that nuclear plants are too expensive to build. And one of the first examples they point to is the Vogtle project in Georgia.

The original budget for Vogtle units 3 and 4 was around $14 billion. This ballooned over the course of the project to a final cost of approximately $32 billion. But while it’s true that this is a huge overrun, it has very little to do with nuclear technology and everything to do with how the project was executed.

Construction on Vogtle units 3 and 4 began in 2009, with Vogtle 3 finally starting up in 2023. But the important point is that these were the first new nuclear reactors to begin construction in the United States for more than 30 years. And that long gap mattered.

According to the US Department of Energy’s (DoE’s) Liftoff report, construction on Vogtle began before the plant’s design was fully complete. That decision triggered years of rework, schedule delays, and orders to alter the design. It’s a bit like starting on the foundations of a house before the blueprints are finished - when the plans change, any work that’s been completed previously has to be torn out and redone.

Starting again from scratch

At the same time, the US essentially had to rebuild a supply chain that largely gone dormant over the preceding three decades. There were only a limited number of vendors for specialised components, for example. Also, tens of thousands of workers had to be trained to sufficient standards to work in the nuclear industry, as the relevant expertise had moved on or retired. These are known as first-of-a-kind costs– things are more expensive when you’re building things for the first time.

But here’s the thing: they were incurred because the US nuclear industry essentially was building for the first time, having stopped building nuclear plants for so many years, not because they’re inherently required.

Another major challenge faced by Vogtle was in project management. Initially, the reactor designer (Westinghouse) was working with CB&I Stone and Webster (the constructors) to build Vogtle 3 and 4, and later acquired Stone & Webster. But this then meant Westinghouse was responsible for project management as well as design and construction - roles that require very different expertise.

This mismatch led to severe problems in the execution of the project, leading Westinghouse to file for bankruptcy protection in 2017. However, the project was then shifted to an owner-led model, with Southern Company taking over and bringing much more experienced construction oversight. This made for much more accurate cost estimates and improved project execution.

Learning curve

So here’s the good news: the learning effect is clearly visible in the data. Vogtle 4 was completed significantly faster and more than 20% cheaper than Vogtle 3, which is a direct reflection of the improvements in workforce productivity, supply chain performance, and construction sequencing.

What’s more, the DoE estimates that future reactors built with completed designs, standardized construction, and multi-unit sites could cost only nearly half as much as Vogtle, especially when paired with modern financing mechanisms.

So the lesson learned from Vogtle isn’t evidence that nuclear energy is unworkably expensive - it’s evidence that stopping and starting nuclear construction is expensive.

If we want clean, reliable energy at scale, the lesson isn’t to walk away from nuclear — it’s to keep building, so experience compounds instead of resetting, and first-time costs don’t get paid over and over again.

Check out our video on this on TikTok, Instagram, and YouTube.

Subscribe now

For decades, the nuclear industry operated under a simple assumption: bigger is better.

The idea made sense on paper. Larger reactors would generate more electricity, spreading costs across more megawatts and ultimately making power cheaper. But in practice, the opposite often happened. As reactors grew in size, they also grew in complexity, turning into one-off, bespoke megaprojects that were difficult to build, finance, and deliver on time.

We’ve seen this play out in real time. Projects like Vogtle in Georgia and Olkiluoto-3 in Finland arrived years behind schedule and billions of dollars over budget. Instead of benefiting from scale, they became cautionary tales of it.

To understand why, it helps to think about another energy technology: solar.

Imagine if every solar panel were the size of a football field and each one took a decade to build and required specialized infrastructure just to transport it. Solar wouldn’t be cheap. It would be one of the most expensive energy sources on Earth. But that’s not what we did. Instead, we built millions of small, standardized panels in factories. That shift toward repeatable manufacturing drove costs down dramatically.

According to the Right-Sizing Reactors Report, this is exactly what nuclear has been missing.

From Economies of Scale to Economies of Volume

The report describes a necessary shift from “economies of scale” to “economies of volume.” Here’s what that means: if a utility builds 7 gigawatts of large reactors, they might construct a handful of reactors, but each one is still treated like a custom project. That limits how much you can learn and improve from one build to the next. But if that same capacity is built using 300-megawatt small modular reactors, you end up constructing around 23 units. That’s 16 more chances to refine the process, to reduce costs, and to build better than the last.

This is how industries like wind, solar, and batteries got cheaper. Through repetition, iteration, and manufacturing at scale.

Fixing the Supply Chain Problem

Right-sizing nuclear doesn’t just improve learning,it also unlocks the supply chain.

Large reactors depend on massive components, like steel pressure vessels, that only a handful of specialized facilities in the world can produce. That creates bottlenecks, delays, and limited competition.

Smaller reactors, on the other hand, rely on more standardized components that can be manufactured by a wider range of suppliers. This introduces competition and flexibility into the system.

The report calls this “learning-by-searching”—the ability to find better, faster, and more affordable ways to build because you actually have options.

The Regulatory Mismatch

But even the best technology won’t scale if the rules don’t allow it to.

Today’s regulatory framework was designed for large, custom-built projects—not factory-produced products. As a result, developers face high costs and repeated reviews that slow deployment. In some cases, Nuclear Regulatory Commission fees alone could consume up to 18% of a microreactor’s annual operating budget.

To address this, the Nuclear Innovation Alliance recommends three key reforms:

1. Standardize Reviews
   Once a design is approved, it shouldn’t need to be re-approved for every new site.

2. Scale Environmental Reviews
   Use generic impact statements for standardized technologies instead of starting from scratch each time.

3. Modernize Hearings
   Remove mandatory uncontested hearings that create bottlenecks, while still preserving pathways for legitimate public input.
   

The Future is Right-Sized

This isn’t about replacing large reactors with small ones. It’s about using the right tool for the job.

Large reactors will continue to play a critical role in powering big, stable grids. But small modular reactors can bring reliable energy to growing demands like AI data centers or help replace retiring coal plants. Microreactors can serve remote communities, industrial sites, and defense installations that currently rely on diesel.

The future of nuclear isn’t one-size-fits-all. It’s right-sized.

And it starts with a simple shift:

Stop building projects. Start building products.

Check out our video on this on YouTube, TikTok, or Instagram

Subscribe now

A paper published in Nature Communications this February claimed that U.S. counties closer to nuclear power plants have higher cancer mortality rates. The researchers, a doctoral student and his adviser at Harvard’s T.H. Chan School of Public Health, estimated roughly 115,000 excess cancer deaths over 19 years.

When this paper dropped, the media ran with it. But there’s a glaring problem: the study never actually measured radiation. To figure out how something like this gets published, I sat down with three experts to pull it apart: Philip Hult, a nuclear engineer here at Generation Atomic; DJ LeClear, a health physicist with nearly two decades of field experience; and Jim Smith, a professor of environmental science who has studied Chornobyl since 1990. Watch the full conversation here.

No Dosimetry, No Radiation Study

The study uses geographic distance as a stand-in for radiation exposure. Closer to a plant means more radiation, the authors assume.

Talk to anyone who actually measures this stuff for a living and the premise falls apart. DJ LeClear has spent years taking readings outside nuclear facilities, and he put it plainly: the doses are so low they’re barely detectable. The NRC puts the dose of living within 50 miles of a plant at about 0.01 millirem a year. You get the same dose from eating a banana.

We’ve had this conversation so many times at Generation Atomic that we built a tool for it. Every time there’s a scare story about tritiated water or routine emissions, the same confusion surfaces: people hear “radiation” and lose all sense of scale. Our Tritiated Water Dose Calculator lets you compare doses side by side. The gap between what a plant delivers to its neighbors and what would be required to produce the effects this study claims isn’t small. It’s cosmic.

The study ignores prevailing winds, where people spend their time, and dosimetry entirely. The paper itself acknowledges this. Without dose measurements, you don’t have a radiation study. You have a geography study wearing a lab coat.

Correlation Masquerading as Causation

Instead of tracking actual people and their actual radiation exposure, the authors just looked at county-level averages and distances. It’s like noting that counties near airports have higher rates of hearing loss, and concluding that jet exhaust damages eardrums. No, cities are just loud.

Nuclear plants sit near rivers, coastlines, and population centers. That’s engineering necessity, not coincidence. Counties near plants are also counties near industrial activity, urban density, and every health variable that comes with them. The authors tried to control for confounders. But as Smith noted, you can never fully untangle them in a study like this. And when the thing you’re measuring is orders of magnitude below background radiation, the signal is almost certainly something else.

If the logic held, as DJ LeClear pointed out, we’d see spiking cancer rates in Colorado, where natural background radiation runs roughly 100 millirem higher than the national average. We don’t.

The authors concede they haven’t proven causation. Then, in the same paper, they estimate 115,000 attributable deaths. Jim Smith was blunt: if you admit you haven’t proven a causal link, you have no business speculating on death tolls that will be stripped of context and weaponized.

The Ethics of Fear

Philip Hult called the paper’s timeline a “real ethical failure.” The national study dropped just three months after a Massachusetts-specific predecessor in Environmental Health. The original read as a grad student’s methodological exercise. The rapid expansion to national data, on the heels of media attention, looked less like science and more like strategy.

Smith, who has worked with communities around Chernobyl for over three decades, connected the dots to real harm. The World Health Organization concluded 20 years after Chernobyl that social and economic damage outweighed direct health effects. Radiophobia kills. It drives people from their homes, collapses local economies, and prevents rational decisions about energy. Studies like this feed that fear without scientific basis.

Think about what it would take for this paper to actually be right about radiation. Every single plant operator in the country would have to be falsifying emissions data. Every regulator would have to be complicit. Every independent researcher measuring the same thing would have to be in on it. Decades of coordinated fraud, not a single whistleblower.

That’s not a hypothesis. That’s a fever dream.

What Real Evidence Shows

We don’t have to guess about this. We have actual, massive studies—like the Million Person Study and INWORKS—that have tracked hundreds of thousands of nuclear workers across the globe for decades. These are people wearing physical radiation badges, giving us hard, undeniable data. And what does that mountain of real evidence show? At the tiny doses the public receives from a normally operating plant, the cancer risk is virtually zero. The Harvard paper asks us to ignore decades of hard data and just take their word for it.

The Real Cost

Nuclear energy is the largest source of carbon-free electricity in the United States. Every megawatt shuttered under the weight of unfounded fear gets replaced, overwhelmingly, by fossil fuels. The air pollution from those plants causes real, measurable, documented deaths.

A paper claiming 115,000 cancer deaths from nuclear proximity, without measuring a single radiation dose, isn’t just bad science. It’s a public health risk in its own right.

The Breakthrough Institute, the American Nuclear Society, and the Bulletin of the Atomic Scientists have all published detailed rebuttals. And if you want to hear the experts break it down themselves, watch the full video.

Subscribe now

You might have heard that the US nuclear industry is undergoing explosive growth thanks to favorable legislative conditions, renewed public interest, and - oh, the f*cking gigantic data centers that are popping up all over the place.

But there’s a major obstacle – although a lot of companies are proposing to build various designs of advanced reactor we don’t actually have the fuel to run a lot of them. Not because uranium is running out or we forgot how to make it, but because nobody will commit to producing it at scale.

The reason for this is that many advanced reactors need a special fuel called high-assay low-enriched uranium (HALEU). This isn’t non-existent: there’s already a pilot facility in Ohio making it, and earlier this year, the Department of Energy (DoE) announced a $2.7 billion funding boost to expand enrichment capacity, which is great news. But it’s a bit more complicated than that.

Buying uranium to enrich as HALEU and building the factory to create it represents a massive upfront cost, and right now, HALEU manufacturers don’t have customers ready to buy. Enrichment facilities are designed to run for more than 30 years with steady production, but most of the advanced reactor designs planning to use it are still on the drawing board, so it will be at least a few more years until the initial test reactors require actual fuel to load.

Developers can’t commit to buying HALEU when they don’t even know if their reactor design will get approved, bought, and built. And HALEU manufacturers can’t build the factory if there’s a chance they’ll have no customers. It’s a catch-22, and neither side can move.

Government needs to step in

That is, unless the government becomes the first customer. There are more than 50 advanced reactor projects in the pipeline, many of which will need fuel over the next two decades. If the government buys fuel upfront and stockpiles it, enrichment companies will have guaranteed contracts they can show to investors, and reactor developers will have fuel available when they need it. Once the market gets going, private companies can take over with their own long-term contracts – the government just needs to solve the coordination problem to get things started.

You might argue that this adds a middleman, but the fact is that this approach has worked before. The Strategic Petroleum Reserve, whereby the government buys and stockpiles oil to stabilize markets, works in the same way. Defense contracts mean the government commits to buying aircraft and ships, which lets manufacturers invest in production. The government isn’t replacing the market, but solving a temporary problem so the private markets can exist later on.

China and Russia also use this approach, and it’s the reason why they can build reactors really quickly. They build the reactor, supply fuel for its entire 50-year lifespan, handle the waste, and provide maintenance - all in house (well, all in-authoritarian-regime, but you get the idea). For countries making generational energy investments, government involvement makes all of the difference.

But the good news is, Congress figured this out four years ago. They created the HALEU Availability Program in 2020, specifically authorizing the government to purchase and stockpile enriched uranium. It’s been funded since 2022, but unfortunately the government still hasn’t bought a single gram of fuel yet.

Instead, the DoE is scrambling with workarounds, such as, for example, a pilot program allocating fuel one reactor at a time – which is hardly efficient. The reality is, we need five to ten times our current HALEU enrichment capacity. If the US can’t offer a complete and reliable nuclear ecosystem, other countries will.

Government-industry partnerships being decided right now will shape geopolitical alignment and energy outcomes for the next several decades. It’s time for the fuel supply chain to get its shit together.

Check out our video on this on YouTube, TikTok, or Instagram

Subscribe now

Microreactors sound great in theory. They’re being designed to be tiny, portable, and worlds cheaper than traditional large reactors. And although there are potentially great markets for them - such as college campuses, hospitals, data centers, and remote towns - they represent an expensive upfront investment, with significant regulatory challenges because they’d employ never-before licensed technology.

Some people are all in, but skeptics have raised some major doubts. One is that it’s uncertain whether anyone has enough money to front the full investment. Another is that it’s unclear whether anyone will want to buy the clean-energy equivalent of a diesel generator that’s hundreds of times the price. Finally, it may not be possible to license them under current regulatory pathways - you can’t get construction permits without a license, but you can’t get a license if your design is completely new and existing regulations don’t apply.

But there IS a clear customer, and it might just be the savior that the microreactor market needs: the US military. Remote military bases need a LOT of energy in a tiny, portable package, with no pre-built transmission or grid infrastructure. And the military doesn’t have to abide by traditional regulatory pathways – they can approve projects themselves.

In fact, they’ve been using microreactors to power their submarines for decades, with zero accidents to date. They also have an annual budget of more than $850 billion, meaning that they can make risky upfront investments if it means a great product in the end.

Marking the milestones

Enter Janus, a military program designed to break this gridlock by employing private sector companies to develop reactor designs. It uses milestone payments, which is the same model that got SpaceX off the ground. Companies don’t get paid for PowerPoint presentations or promises – only concrete milestones, like completing a design review, manufacturing a component, achieving criticality, or operating for 90 days. No delivery, no money.

In order to join the program, developers must design something that fits on a truck, can produce between 20 and 75 megawatts, can start without external power, and can pack up and leave within two years of shutdown.

And Janus won’t just kickstart the process, it’ll also streamline it. Currently, there are more than 50 US microreactor companies that are wanting to manufacture all the components for their respective designs in-house - which is NOT how you build a strong supply chain.

The Janus program is designed to encourage consolidation, forcing companies to share suppliers and avoid reinventing the wheel. If it goes according to plan, manufacturing will become a well-oiled machine, regulatory approvals will happen more quickly, and costs will come down significantly for each new build.

Practice makes perfect

It took $8 billion for Boeing to build their first 787. Today, it costs 30 times less. Practice makes perfect, and modular nuclear is no different. If these reactors work, even if expensive at first, it will help create a domestic supply chain, give the Nuclear Regulatory Commission the data for commercial licensing, and prove that tiny nuclear is possible.

Project executives say that if reactors aren’t operational in 5-8 years, the program is a failure. Which sounds pretty ruthless, but it means we’ll know around 2030.

Are you betting on Janus? Let us know in the comments!

Check out our video on this on YouTube, TikTok, or Instagram

Subscribe now

The US has been trying to figure out what to do with nuclear waste for 44 years, and we’re not much closer to a solution than we were in 1982.

Not because it’s an impossible problem – Finland and Sweden are both building deep geological repositories right now.

And not because we don’t know where to put it – the Department of Energy (DoE) has been building partnerships and attracting community interest through its consent-based siting program for the past several years.

No, the real reason we can’t crack nuclear waste is because we keep putting the wrong people in charge.

The Nuclear Waste Policy Act of 1982 tasked the DoE with finding, developing, and operating a permanent repository for spent nuclear fuel. Yucca Mountain looked promising for a while, but Congress pulled funding in 2010, and it’s been sitting in limbo ever since.

In the meantime, nuclear plants across the country are storing their own waste on-site. It’s a solution that works fine for now, but it’s not a permanent one.

Enter NuCorp

Recently a bipartisan group of experts, including a former DoE official and former Nuclear Regulatory Commission (NRC) chair, published a report with a pretty straightforward solution: take this responsibility away from the DoE and give it to the people who have a real stake in solving the problem fast.

Their proposal is called NuCorp, an independent corporation led by nuclear reactor owners. These guys are already responsible for managing waste at their own plants, meaning that they understand the logistics, the safety requirements, and the timeline.

Any site selected under NuCorp would still require NRC licensing, as well as an independent advisory committee. This group of technical experts and representatives from relevant state, tribal, and local governments would oversee operations and report annually back to Congress and the public. The structure maintains the safety and oversight of a DoE job, it just puts implementation into more motivated – and experienced – hands.

Cash no problem

And here’s the money part that makes this whole thing possible. From 1983 to 2014, all US nuclear plant operators paid into something called the Nuclear Waste Fund, which was designed to finance a permanent disposal solution. By 2014, it had accumulated $40 billion.

Then, a federal court stepped in. The court basically said “you can’t keep collecting money for a program that doesn’t exist,” and ordered the DoE to stop charging the fee (which is a fair point). But even though operators stopped paying in 2014, the fund kept growing from interest alone, and now sits at around $52 billion.

The NuCorp proposal would use this money for its original purpose. Operators are currently paying to store waste at their plant sites, and would jump at the chance to engineer a real, permanent solution. Finland and Sweden are succeeding because they put the nuclear industry in the driver’s seat with strong regulatory oversight, and it’s time for the US to do the same.

Campus kings

Just last month, the DoE launched another initiative: Nuclear Lifecycle Innovation Campuses. The idea is to get states to volunteer to host regional hubs that handle everything from fuel fabrication and enrichment to reprocessing used fuel and storing the waste onsite. States like Wyoming, Tennessee, and Idaho are already competing for the chance, lured by the promise of up to $50 billion in private investment and thousands of jobs.

On paper, it sounds great – a whole-cycle solution with willing host communities. But it’s still the DoE running the show, and after four decades of false starts, that’s exactly the problem.

We can’t solve the waste debacle with another study or another decade of congressional back-and-forth, but by putting responsibility where it belongs. Forty-four years is long enough to wait.

Check out our video on this on YouTube, TikTok, or Instagram

Subscribe now

In January, US Energy Secretary Chris Wright ended the Department of Energy’s (DOE’s) longstanding use of ALARA to govern its standards for radiation exposure in the country’s nuclear energy industry. And the implications for the sector could be pretty seismic, to say the least.

But what is ALARA? It stands for ‘As Low As Reasonably Achievable’, and as its name suggests mandates a principle whereby nuclear operators must minimise radiation exposure as functionally possible to workers and the wider public. The concept is rooted in the ‘linear no threshold’ (LNT) model of radiation risk, which assumes that any exposure to radiation, no matter how small, has the potential to cause harm.

The problem is that the LNT model, which was adopted in the mid-20^th Century by scientific bodies and regulators, is based on assumptions that have pretty scant support from the actual scientific evidence (check out some of our other blogs and videos on the subject to find out why).

Be reasonable for a minute

So, bearing that in mind, what counts as ‘reasonably achievable’? Well, it depends on how you interpret that pretty ambiguous wording. And how you interpret it can cost billions of dollars. People working in commercial nuclear operations and DOE facilities have long pointed out how ALARA’s vague mandate has justified massive expenditures for ever-smaller reductions in radiation dose, even when those doses are already far below regulatory limits. In fact, a 2025 Idaho National Laboratory study found that 58% of monitored workers at commercial light-water reactors received no measurable radiation dose at all in 2022. Zero. And we’re still spending billions to reduce it further.

A good example of this is the Hanford site in Washington state. Hanford was the country’s main producer of plutonium for nuclear weapons during World War II and the Cold War - not a commercial nuclear energy site - but is widely described as one of the most costly clean-up sites in the US. Costs to clean up Hanford are projected to be in excess of $300 billion over the next 60 years.

However, recent surveys of radiation levels found that most areas of the site measured the same as natural background radiation. What’s more, any radiation hotspots were lower than the average level of background radiation in Finland. But, because these levels aren’t yet “as low as reasonably achievable”, the US is spending all that cash to remove every possible trace of radiation from the site.

Money, money, money

Of course, it’s important that we do clean up the site. But given the increasing evidence against the LNT model, not all of that vast expense is necessary. And just think what some of that money could be used for instead - expanding clean energy, subsidising healthcare, building affordable housing, help paying for childcare, education… the list goes on. When we pour resources into achieving negligible radiation dose reductions at legacy sites, we’re making a choice about priorities - and maybe that’s a choice that should be reconsidered.

Critics have said that abandoning ALARA will weaken worker protection, undermine public trust and blow a hole in regulatory coherence, as well as altering the nuclear industry’s established ‘safety first’ culture. But anyone who’s visited a nuclear plant knows that safety is not only the #1 priority for these plants, but also a massive source of pride for those who work there. So it’s hard to see how that would change.

ALARA asks us to continually reduce radiation doses below already conservative safety limits, even when we’re driving toward levels that have never been shown to cause health effects. So, it’s time to move on from disproven 1950s science and say: sayonara, ALARA.

Check out our video explainer on TikTok and Instagram!

Subscribe now

Last year was a great year for the nuclear industry. Nuclear power is now generating its highest-ever amount of electricity - around 2667 TWh in 2024, with the figure for 2025 likely to be similar. Advanced reactors took major steps towards commercialisation in the US and UK, and there were other big wins across the world - such as Belgium repealing its nuclear phase-out policy and the World Bank officially ending its long-standing ban on financing nuclear power projects.

But it’s clear that more needs to be done - after all, the fight to meet growing power demand and address climate change means we can’t wait. So here’s a list of the things we’d like to see happen in the nuclear industry across the globe in 2026.

1. Regulatory reform & streamlined licensing - especially in the US and UK

One clear theme of 2025 was interest around speeding up the process of getting reactors permitted and built, both in the US and elsewhere. And this is perhaps the most important and vital thing on our wishlist - looooooong licensing timelines scare off investors, developers and even utilities. Regulatory reform isn’t big and flashy, but without it nothing (or, at least, very little) else might happen.

What we’d like to see in 2026:

• The US Nuclear Regulatory Commission (NRC) adopting more predictable, risk-informed, performance-based licensing timelines, along with faster certification.

• In the UK, a similar continuation and acceleration of streamlined regulatory pathways - including clear timelines for final design assessments following acceptance.

So, for regulators and governments, we’d like to say: not so much “move fast and break things”, but “move a bit smarter and finish things”.

On the policy front, we need states to move from talking about nuclear to actually enabling it. 2025 showed us what works - and what doesn’t. The path forward requires clearing regulatory barriers, investing in workforce development, and creating bankable financial mechanisms.

Moratoriums must go. Illinois’ Governor just signed (SB0025) and removed its arbitrary 300 megawatt capacity cap. Meanwhile, Oregon, California, Minnesota, andRhode Island introduced moratorium repeal bills that all stalled in committee. Without repealing these restrictions, nuclear support becomes nothing more than rhetoric. You can’t develop what you’ve legally banned yourself from building.

Workforce programs need funding, not task forces. Michigan’s HB4125 provides grants to colleges and universities for nuclear degree programs, including scholarships for students who agree to work at Michigan facilities for three years. Colorado’s SB120and Texas’s workforce bills (SB1535) establish similar grant programs. Contrast this with Arkansas’s HB1073, which creates another task force. We don’t need more reports. We need training programs with actual funding.

Financial mechanisms matter. Texas’s HB14 created a three-tier funding structure with grants up to $200M, Indiana’s SB0423 established “SMR” cost recovery frameworks, and Tennessee’s HB1133 provides tax credits for nuclear facilities. These risk-sharing mechanisms recognize nuclear’s decades-long payoffs require significant upfront costs.

Nuclear policy isn’t stalled because of missing information or insufficient studies. It’s stalled because passing actual legislation requires political capital, and too many legislators aren’t willing to spend it. The question for 2026 is which states will follow through.

2. Real money flowing in, not just promises

Last year, as indeed many recent years have been, was full of big promises: net-zero pledges, nuclear “renaissances”, and advanced reactor optimism everywhere you looked. But what there wasn’t nearly enough of was capital actually flowing into projects.

For example, if we wanted to scale advanced reactors meaningfully by the 2030s, annual investment needs to jump from a few billion dollars currently to something closer to $25 billion per year by 2030, according to the International Energy Agency. That won’t happen via press releases alone.

So, our wishlist in this department:

• More private capital stepping in, not just government grants and cost-share programmes.

• New financing models - things like long-term power purchase agreements with data centers, industrial customers or hydrogen producers.

• More creative thinking around nuclear as infrastructure: bonds, regulated asset bases, and public-private partnerships that actually reach final investment decisions.

3. A real comeback in large reactor construction

Yes, yes, we know that common-or-garden large reactors might not be as glamorous and sexy as advanced ones, but let’s face it: they still do the heavy lifting when it comes to decarbonising power systems at scale. China understands this, and so does India. France is re-learning it. The US and UK… well…

So with that in mind, we demand the following:

• At least one new large-reactor project in the US move decisively past the planning stage - with a clear design choice, financing structure and timeline.
  

• In Europe, a bit less “hmmm, maybe nuclear” and a bit more “nuclear, yes please!”. Poland and the Czech Republic are showing the way here.
  

• Japan leaning back into nuclear in an even bigger way that they’re currently planning.
  

• More standardization - fewer bespoke designs, more repeat builds (think Liberty ships, not Faberge eggs).
  

• Clear accountability between utilities, EPC firms and vendors.
  

• Early, honest project cost estimates instead of optimistic numbers that fall apart mid-build.

Big reactors aren’t dinosaurs. They’re the fastest way to add truckloads of firm, clean power - if countries are willing to commit. Just ask the United Arab Emirates.

4. Out with the new, in with the old (wait, what?)

It’s always the new reactors and projects that grab the headlines. But again, just like the advanced nuclear/traditional large reactor dichotomy we pointed out above, sometimes the less glamorous aspects are the most important. A huge part of nuclear’s value in fighting climate change comes from the plants that already exist, and many aging reactors just need a bit of investment to remain reliable for years.

In 2026, we’d love to see:

• More license extensions and life-extension investments in the US and Europe, plus restarts where possible.

• Clear plans to resolve long-standing holdups - whether that’s delayed projects, political roadblocks, or infrastructure challenges.

• More uprated power outputs from existing reactors - if we can get more out of them, it’s the easiest way to add clean power.
  
  The US nuclear uprates slated for 2026 are telling an optimistic story. Constellation’s Byron and Braidwood plants will be squeezing out an extra 135 MW, not because of some heroic technical breakthrough, but because the Inflation Reduction Act’s 45Y clean power tax credits are rewarding reliability and zero-carbon output. Vistra and Energy Harbor are going to be pumping out more power at Perry and Davis-Besse, creating a whopping 433 MW increase in output, driven in no small part by companies like Meta realising that data centres do not, in fact, run on vibes and solar press releases. Southern Company is adding 112 MW across Vogtle 1 and 2 and Hatch 1 and 2 to meet increases in demand, while Energy Northwest’s Columbia Generating Station is targeting a 186 MW boost to help regional decarbonisation goals. Different owners, different reasons, but the same underlying message: America’s existing nuclear fleet is being souped up, uprated and brought back into the conversation as critical infrastructure, not legacy baggage.

• RESTARTS - yes, we really can’t say this enough. It’s about more than just getting extra clean power - it’s about sending a powerful signal that tried-and-trusted nuclear plants are too valuable to be left on the sidelines.
  
  In the US, those that are closest include Palisades, Crane Clean Energy Center (the new name for Three Mile Island), and, come 2028, Duane Arnold. Elsewhere in the world, in Europe Belgium’s Tihange 3 and Doel 4 are now back online, and we’re willing to bet you a bratwurst that Germany will change its stance and consider reopening at least one of its shuttered plants. Elsewhere, we’re hopeful that Japan will restart Hamaoka but they’ll definitely be sparking up Kashiwazaki-Kariwa again.

This isn’t a nostalgia trip for old reactors (as much as we love ‘em), it’s about recognising that keeping existing nuclear power plants online is one of the fastest, cheapest ways to cut emissions right now.

5. Supply chains (my baby’s got me wrapped up in supply chains…)

Another unglamorous aspect of the industry, but again, one that’s absolutely vital. A lot could be improved in nuclear supply chains, both in terms of materials and human resources. For example, large forgings, reactor pressure vessels, steam generators, high-integrity valves, nuclear-grade concrete and instrumentation all depend on a surprisingly small number of suppliers worldwide. Plus, any planned expansion of the nuclear fleet entails an expansion of the workforce. Therefore, we need to see:

• Investment in domestic and allied nuclear manufacturing, especially in the US and Europe. This might involve a commitment to fewer one-off components and more standardised parts across projects, as well as long-term procurement deals to give certainty to suppliers and operators alike.

• More apprenticeships in construction and engineering trades, as well as clear pipelines from universities into nuclear jobs - including construction and operations staff.

• Concrete progress on creating new Western mining/milling, conversion, and enrichment capacity.

And finally…

6. Let’s get it together across the world

The energy transition can’t succeed as a patchwork of national efforts; the nuclear industry needs to act globally too. China’s rapid build-out shows what scaling can look like, India is displaying huge ambition, and Japan is showing how trust can be rebuilt. Meanwhile, Europe’s policy debates are revealing what happens when energy security collides with ideology. There’s something to learn from all of it - if we’re willing to collaborate.

This wishlist isn’t about wishful thinking; it’s about momentum reaching escape velocity.

All the necessary ingredients are there: public support is improving, climate pressure is rising, electricity demand is exploding, and nuclear is increasingly seen as a solution rather than a problem. What the nuclear industry and governments need to deliver in 2026 is execution.

If we get faster licensing, smarter financing, real advanced reactor builds, stronger existing fleets and deeper global cooperation, then 2026 won’t just be another year of talk - it’ll be the year nuclear energy truly shifted gears.

And honestly? It’s about time.

What would you like to see happen for nuclear in the coming year? Let us know in the comments!

Subscribe now

Workers in the nuclear industry are classified as radiation professionals, which doesn’t come as a surprise to many people. Radiation workers are subject to government oversight, strict monitoring, and annual dose limits. Less known, however, is another class of radiation professionals hiding in plane sight: airline pilots and flight attendants.

The average commercial pilot receives between 2 and 6 millisieverts of radiation annually – roughly equivalent to 20 to 60 chest X-rays. Nuclear plant workers average just 1.5 millisieverts per year. Pilots routinely receive up to four times more radiation than people who work directly with nuclear reactors, but from where?

Not engines or electronics, but from outer space.

High-energy particles from distant stars and our sun constantly bombard Earth. The atmosphere filters out most of this cosmic radiation before it reaches ground level, which is why surface dwellers receive relatively modest doses. At 35,000 feet, though, planes are above much of that protection, and cosmic rays that would normally be absorbed by the atmosphere reach cabin interiors directly.

There’s also ultraviolet radiation to consider. Flying for an hour at 30,000 feet delivers UVA exposure comparable to a 20-minute tanning bed session, though exposure varies by route. Flights over polar regions receive higher doses than equatorial routes because Earth’s magnetic field provides less shielding near the poles.

Aircraft offer little meaningful protection against cosmic radiation. A single cross-country flight exposes passengers to approximately 0.04 millisieverts, which for occasional travelers, is negligible – about 1.3% of the background radiation we get from natural sources each year (3 millisieverts). For pilots making 150 flights per year, though, this exposure can accumulate. Flight attendants often receive even higher doses because they work longer hours and fly more frequently. International aviation authorities require airlines to monitor crew radiation exposure. Personnel can be reassigned to lower-altitude routes or have schedules adjusted if they approach the 20 millisievert annual limit, and pregnant crew members may be moved to ground duties to minimize fetal exposure.

Nuclear power plants are engineered specifically to minimize radiation exposure through multiple layers of protection. Thick concrete containment structures, steel reactor vessels, water pools, and physical distance all work together to reduce doses. Workers wear dosimeters that continuously track cumulative exposure, and facilities enforce strict time limits for working near radiation sources. Aircraft have no comparable shielding. The fuselage (airplane body) offers little meaningful protection against cosmic radiation, which means pilots and passengers are simply exposed to it for the entire duration of the flight.

For airline workers, the International Commission on Radiological Protection (ICRP) recommends a limit of 20 millisieverts per year, averaged over five years, with no single year exceeding 50 millisieverts. The U.S. Nuclear Regulatory Commission (NRC) limits nuclear workers to 50 millisieverts annually, though average exposure falls 33 times below this amount.

What It Means

Both aviation and nuclear power are highly regulated industries with excellent safety records. Annual radiation doses for pilots and nuclear workers fall well below levels associated with any measurable health risks, and the human body continuously repairs the kind of low-level radiation damage that occurs from these exposures.

A widespread misconception is that all radiation, in any quantity, is dangerous. This assumption neglects to recognize that low levels of radiation exist across our everyday lives, in everything from routine medical procedures and potassium-rich foods to soil, fly ash, and airplanes.

Nuclear workers receive low doses because the industry has tight administrative controls. Pilots receive marginally higher doses because cosmic radiation at 35,000 feet is unavoidable and fuselages don’t block it. Not all radiation is dangerous, but not all radiation is where you’d expect.

Makes you think twice about who the real radiation professionals are.

Small Modular Reactors (SMRs) have attracted attention to nuclear energy that hasn’t been seen in decades. The phrase “SMR” now appears in landmark federal legislation, bills to lift state nuclear construction bans, and deployment plans for data centers, military bases, and disaster relief zones. It all sounds like progress, but is it possible that this phrase has accidentally restricted development?

The International Atomic Energy Agency (IAEA) defines SMRs as advanced nuclear reactors with power output up to 300 megawatts per module, roughly one-quarter the capacity of traditional large reactors. They’re called modular because components can be factory-built and transported to sites rather than constructed on-site like traditional plants, promising faster deployment and lower upfront costs.

The phrase “advanced nuclear” refers to a much broader range of technologies. Small modular reactors are certainly included under the advanced nuclear umbrella, but so are innovative ideas like floating nuclear plants, microreactors to replace diesel generators, high-temperature reactors providing industrial process heat, and fast reactors that can consume nuclear waste as fuel. Most importantly, advanced reactors don’t have to be small. They can be just as large as traditional light water reactors, or even larger. The confusion between “advanced nuclear” and “SMRs” was mostly semantic until it became legal language – then it got expensive.

Illinois banned new nuclear construction in 1987, but in 2023, legislation to lift the moratorium passed both chambers of the state legislature. The bill would have allowed new advanced reactor construction, opening pathways for next-generation nuclear development. That is, until Governor J.B. Pritzker vetoed it.

The bill that Pritzker signed limited reactors to 300 megawatts. He was concerned that “advanced reactors” represented a category so broad that it could make way for large reactors like the AP1000 to be built in the state. SMRs were the hot new technology said to revolutionize energy markets, and bigger reactors were seen as politically scary. The entire purpose of lifting the nuclear construction moratorium was to enable new advanced reactor deployment and position Illinois as a leader in next-generation nuclear energy. Instead, the fixation on SMRs as a magic category excluded the technologies most likely to succeed in the near term.

The only advanced nuclear technology that’s achieved commercial operation in the United States so far is Westinghouse’s AP1000. These reactors have passive safety systems, similar to many Gen III+ designs, and produce over 1,000 megawatts of capacity. Four units are operating at Plant Vogtle in Georgia right now.

TerraPower, another advanced reactor developer that is currently constructing non-nuclear infrastructure in Wyoming, aims for 345 megawatts. Their Natrium reactor integrates molten salt energy storage to provide dispatchable output, and is backed by Bill Gates’s billions, paving the way for commercial success within a matter of years. These technologies are advanced by all definitions, but don’t fit into the neat package that earns the SMR stamp of approval.

Any reactor with feasible design improvements should be developable under advanced nuclear legislation. Fixation on “SMR” as the category that defines innovation creates arbitrary cutoffs that omit viable technologies, and correcting these subtle exclusions starts with us—stakeholders, advocates, and industry professionals who want progress over particularity.

Check out the video explainer on our TikTok and Instagram!

A technical competition is underway to determine which company will be first to navigate regulatory approval, secure financing, and commercially operate an electricity-generating advanced nuclear reactor in the United States. We’re calling it the Advanced Reactor Race™, and most U.S. based companies are inventing new designs from scratch. Kairos Power is perfecting a design that worked in the 1960s but never achieved commercial deployment.

All U.S. molten salt reactors trace their lineage to the Molten Salt Reactor Experiment (MSRE) that operated at Oak Ridge National Laboratory from 1965 to 1969. The experiment proved that molten salt coolant works, that fluoride salts can transfer massive amounts of heat, and that the concept is scientifically sound. What it didn’t prove is that the technology could operate economically at commercial scale.

Kairos, founded in 2016, is attempting to finish what Oak Ridge started. Their reactor is called Hermes, named after the Greek god known for speed. Whether this naming choice actually accelerates construction timelines is still up in the air.

Hermes is a fluoride salt-cooled high-temperature reactor using a molten salt mixture called Flibe — a combination of lithium fluoride and beryllium fluoride. Flibe can transfer enormous amounts of heat at low pressure, which eliminates the thick pressure vessel walls required by water-cooled reactors and reduces certain accident scenarios. Unlike the original MSRE, which dissolved fuel directly into the liquid salt, Hermes uses solid TRISO fuel pebbles. Uranium atoms split inside the ceramic-coated pebbles, releasing heat that the liquid Flibe carries away through natural circulation—no pumps required.

In 2023, Hermes earned a construction permit from the NRC, making it the first non-light water reactor design permitted in the United States in over 50 years. It was a regulatory milestone, and construction began at Oak Ridge in May 2025. The initial version is a test reactor designed to prove the technology works. It will only produce heat, not electricity, which simplifies regulatory requirements and lets Kairos demonstrate that Flibe cooling performs as predicted in operational conditions.

But Kairos isn’t just building one reactor at a time — they immediately applied for construction permits for Hermes 2, an electricity-producing version using identical nuclear technology but adding steam turbines. The NRC reasoned that if Hermes 1 was safe to build, Hermes 2, which uses the same reactor technology, should be safe too. Both permits were approved. Hermes 1 is now targeted for operation in 2027, while Hermes 2 is scheduled for 2030.

And Kairos isn’t just building reactors — they’re also working to mass-manufacture their coolant. They’ve constructed a molten salt production facility in New Mexico and recently operated the world’s largest Flibe system at their Engineering Test Unit , which became  the largest molten salt transfer since the 1960s experiments. Manufacturing Flibe at commercial scale hasn’t been done before, as the beryllium fluoride component is particularly challenging to produce in quantity. Kairos is developing this capability in tandem with reactor construction, which adds technical risk but addresses a fundamental supply chain gap.

In October 2024, Google announced a power purchase agreement with Kairos for 500 megawatts of nuclear capacity to be delivered between 2030 and 2035. The deal, structured through Tennessee Valley Authority (TVA), is the first corporate agreement to purchase nuclear energy from multiple small modular reactors.

This commitment has encouraged other buyers to invest in advanced nuclear too. While traditional utilities have been hesitant to commit to unproven reactor designs, tech companies facing exponential data center growth and carbon-free energy requirements are willing to take the risk. The agreement gives Kairos financial certainty to continue construction and provides evidence of actual customer demand for advanced reactor output.

The Assessment

If Hermes 1 achieves criticality in 2027 as planned and demonstrates stable operation, Kairos will validate five decades of molten salt research and potentially open pathways for other advanced designs. If unexpected materials or operational challenges emerge, they’ll discover why commercializing MSRE was so difficult in the first place. Either way, they’re actually building the thing, which puts them ahead of most of the competition.

Check out our video explainers on TikTok and Instagram!

Advanced reactor developers are facing a regulatory paradox: the Nuclear Regulatory Commission (NRC) won’t license new designs without operational data, but obtaining that data requires building a reactor, which requires a license. While many companies are stuck in this catch-22, Austin-based Aalo Atomics, a startup founded in 2022, found a way around it.

MARVEL (Microreactor Applications Research Validation and EvaLuation) is a 100-kilowatt research microreactor at Idaho National Laboratory that kicked off operations in 2020. The federal program was designed to test sodium-cooled fast reactor technology and generate operational data that could inform future commercial development.

MARVEL’s project lead, Yasir Arafat, left INL in 2023 to co-found Aalo Atomics – a commercial-scale version of the MARVEL reactor that Arafat helped design. The Aalo Pod is a 50-megawatt sodium-cooled reactor consisting of five clustered 10-megawatt Aalo-1 reactors, amounting to 100 times more power than its parent design. Their target market is artificial intelligence data centers, which face explosive power demands and are willing to pay premium prices for reliable, carbon-free electricity.

In the Aalo pod, liquid sodium metal flows through the reactor core, absorbing heat from nuclear fuel. Sodium is a highly efficient heat conductor that operates at low pressure, posing a strategic advantage over traditional water or gas coolants. This eliminates the thick pressure vessels required by water-cooled designs, but brings along a host of other operational challenges, including the fact that it burns on contact with air and explodes on contact with water. Russia has struggled through sodium fire incidents for decades between its BN-600 and BN-800 commercial reactor models, providing the rest of the world with a detailed roadmap of what not to do when your coolant is chemically incompatible with both air and water.

The Regulatory Strategy

The company is constructing Aalo-X, their experimental reactor, directly adjacent to MARVEL at Idaho National Laboratory. Research reactors don’t require the same NRC licensing process as commercial reactors, as they receive rigorous government supervision during operation. By comparing data between the government’s research reactor and their scaled-up version, they can demonstrate to the NRC that their methodology for increasing reactor size is technologically viable. In this arrangement, the government gets valuable research data, and Aalo gets the operational results they need for licensing approval. It’s a clever strategy, but also means a lengthy period until commercialization comes around.

Aalo’s approach has attracted serious backing. The Department of Energy selected them for the Nuclear Reactor Pilot Program, which challenges participating companies to achieve criticality by July 4, 2026. They’ve secured $100 million in Series B funding, and have signed agreements with URENCO that makes them the first advanced reactor developer to receive fuel delivery in 2026. Fuel commitment matters significantly, as it’s a typical bottleneck for advanced reactors using non-standard enrichments. This arrangement gives them a slight advantage.

The Assessment

The real question is: Will Aalo be first to commercial operation, per their claims? Likely no, but they aren’t in bad shape either. They have a clear technical strategy, federal support, substantial funding, and secured fuel supply. They’re building actual hardware at Idaho National Laboratory, not just presenting slides. And they’re not alone in struggling toward aggressive deadlines, as almost every participant in the Nuclear Reactor Pilot Program will likely miss the July 2026 target.

Building on MARVEL’s foundation instead of starting from square one gives Aalo better odds than most competitors attempting similar timelines, and it’s likely that their plans will eventually see success.

Be sure to check out this video explainer on our TikTok and Instagram!

Most companies in the advanced reactor race are trying to scale down gigawatt-sized nuclear plants into something commercially viable. Radiant Nuclear, founded in 2020 by former SpaceX engineers, is taking the opposite approach: build something so small it fits on a single truck, then factory-generate dozens of them per year.

Coming Back to Earth

Radiant’s founders were originally working on energy systems for Mars colonization. In space, you need to generate power with no existing infrastructure, no water, and no ability to call for backup. But there are plenty of places on Earth like this too — remote military bases, disaster zones, mining operations, and anywhere diesel generators currently burn through millions of gallons of fuel annually. And tiny reactors are a hot sell in the terrestrial market right now, so the pair set out for Earth instead.

Their design is called Kaleidos, a one megawatt microreactor that uses TRISO fuel and helium cooling. The entire system fits in a shipping container, requires no site excavation, and can theoretically be dropped off and plugged in within 24 hours. It’s designed to replace diesel generators, which means it needs to be truly portable — not “portable with six months of site preparation and a small army of specialists.”

The reactor uses air cooling via forced convection (aka fans), and can run for several years before refueling. It’s perfect for remote areas, and unlike competitors, doesn’t require a dedicated water supply.

But here’s where Radiant’s plan gets (very) ambitious. They don’t want to build reactors one at a time, proving each one works before building the next. They want to build a factory that produces 50 microreactors per year.

In October 2025, the company announced it will build its first manufacturing facility, called R-50, at a former Manhattan Project site in Oak Ridge, Tennessee. It will require a $280 million investment and create 175 new jobs, with a long-term goal of producing dozens of reactors per year, though no company has yet produced nuclear reactors at this scale. It assumes you can standardize, automate, and quality-control nuclear manufacturing the way Boeing builds airplanes or Tesla builds cars. This might work, but it also might be spectacularly complicated.

The U.S. Department of Energy selected Radiant for the Advanced Reactor Pilot Program, which requires participating companies to achieve criticality by July 4, 2026. That’s building and starting up a reactor in less than a year, which is ambitious, to say the least. Radiant will conduct initial testing at Idaho National Laboratory’s DOME (Device for Observing Modular Energy) facility beginning Spring 2026.

The Department of Defense awarded Radiant a contract in August 2025 to deliver one unit to a military installation by August 2028. Each Kaleidos displaces approximately 1.8 million gallons of diesel fuel annually, which is a 110-fold improvement in energy density compared to conventional generators. But the convenience comes with a lofty price tag. The military might be able to front the cost of flexible power, but does the public want micro technology this expensive? Only time will tell.

The Assessment

Radiant won’t be first to achieve criticality, but they do have actual hardware under construction, with federal support for testing, paying customers lined up, and a manufacturing facility breaking ground. This is more progress than many competitors can claim.

Mass-producing several microreactors per year, let alone 50, has never been done. Nuclear manufacturing has always wanted to be industrial and repeatable, but has never gotten there yet — at least in the United States. If Radiant can pull it off, they’ll change the game. If they don’t, they’ll have built a very expensive factory in Tennessee.

The race to build the first commercial advanced nuclear reactor features startups, established manufacturers, and…. a defense contractor that's been building submarine reactors for 70 years. BWX Technologies (BWXT) has commercial reactor designs ready whenever they feel like it, which apparently isn’t yet.

They’ve manufactured nuclear reactors since the USS Nautilus, the world’s first nuclear-powered submarine, in the 1950s. Today, their reactors power every Ohio, Virginia, Seawolf, and Los Angeles-class submarine in the U.S. Navy, plus every nuclear aircraft carrier. They have $2.61 billion in Navy contracts and more under construction.

BWXT knows how to navigate the complex intersection of nuclear regulation, manufacturing, and operation. They’ve been doing it for seven decades. But if they build microreactors all the time, why haven’t they come out with a commerical version?

Well, it exists — just not high on the priority list. BWXT’s entry in the advanced reactor race is called BANR (BWXT Advanced Nuclear Reactor), a 50 megawatt thermal high-temperature gas reactor designed to provide industrial process heat and electricity for mining, oil and gas production, municipalities, data centers, and campuses. It uses TRISO fuel and helium coolant. The entire unit is modular and factory-built, designed for transport by rail, ship, or truck.

As of October 2025, they haven’t filed any permit applications with the NRC, and they have no announced timeline for commercial deployment. This isn’t because the technology isn’t ready or because they lack capability. It’s because commercial nuclear markets are uncertain, regulatory pathways for novel designs are time-consuming, and building submarine reactors is more profitable. BWXT can build BANR whenever market conditions justify the investment. Until then, they’re content to wait.

Now, BWXT is building a microreactor, just not for commercial customers. Project Pele is a 1.5 megawatt demonstration microreactor for the Department of Defense that fits into four 20-foot shipping containers. It’s expected to achieve criticality by 2028 and will power military bases for up to three years without refueling. The military doesn’t care about market disruption or venture capital — they care about reliable power in locations where diesel supply chains are vulnerable or expensive.

Aside from reactors, BWXT is building critical infrastructure for the entire advanced nuclear sector. In July 2025, they completed installation of advanced TRISO fuel manufacturing equipment in Virginia. This facility is now producing commercial quantities of TRISO fuel, which several other advanced reactor companies will use in their designs.

The company was selected in December 2020 for the Department of Energy’s Advanced Reactor Demonstration Program, receiving an initial $30 million grant. They’ve been conducting feasibility studies since then, including a two-year contract with the Wyoming Energy Authority completed in 2025 to evaluate microreactor deployment in a former coal town called Kemmerer.

The Assessment

BWXT isn’t trying to be first, because they don’t need to be. They’re the only company in North America manufacturing both naval reactor components and commercial nuclear fuel. They have seven decades of operational experience, established relationships with regulators, and government contracts that provide stable revenue while everyone else burns through funding.

If they one day decide that commercial reactor deployment makes sense, they can file permit applications and get approved quickly. The NRC knows them and trusts their manufacturing quality, which is far more than others can say. If commercial markets aren’t the success they’re slated to be, BWXT may never enter at all.

Be sure to check out this video explainer on our TikTok and Instagram!

There’s a new tech war brewing in the United States, and it’s fueled by billions of dollars, national energy security, and the massive egos of companies convinced that they can reinvent nuclear power. It’s the race to deploy the first game-changing advanced reactor.

GE Hitachi’s plan is to make existing technology smaller and remove the pumps. They’ve manufactured boiling water reactors since the 1960s, and have built 67 licensed reactors across 10 countries. They understand water-cooled reactor physics because they’ve been implementing it commercially for six decades. Their reactors are operating in Japan, Taiwan, Mexico, Spain, Sweden, and the United States. This experience matters when trying to convince regulators and utilities that your new design will work.

BWRX-300

The BWRX-300 is essentially a smaller, simplified version of GE Hitachi’s Economic Simplified Boiling Water Reactor (ESBWR), incorporating components from their Advanced Boiling Water Reactors operating in Japan. It produces 300 megawatts of electrical output—roughly one-third the capacity of their larger designs.

The key innovation is eliminating active cooling pumps, instead utilizing natural convection. Heated coolant rises through the core, transfers heat to steam generators, becomes denser as it cools, and sinks back down. This passive safety approach eliminates pump-related failures. There are no motors to lose power, no mechanical seals to leak, and no control systems to malfunction.

The reactor’s parent design already has NRC approval, and fuel supply chains are well-established. All that’s left is to get building.

Construction Underway

The Canadian Nuclear Safety Commission issued a construction license in April 2025 for Ontario Power Generation to build a BWRX-300 at the Darlington site in Ontario. Approval from the province followed in May 2025, making it the first advanced reactor design to receive construction approval in North America.

While the Darlington site will eventually host four BWRX-300 units, Ontario Power Generation is only building one to start. They want to prove the technology before committing to the full deployment, and criticality is targeted for 2030. Each reactor will generate enough electricity to power 300,000 homes, adding even more clean capacity to a famously low-carbon grid.

Luckily, this technology is slated for the U.S. too. Tennessee Valley Authority (TVA) submitted construction permit applications to the NRC in April and May 2025 for a BWRX-300 at Oak Ridge, Tennessee. They also applied for $800 million in Department of Energy funding through the Generation III+ SMR program. If received, reactor deployment could be as early as 2033.

Beyond North America, Poland, Sweden, Estonia, the United Kingdom, and Hungary have all expressed deployment interest. In July 2025, a letter of intent was signed to deploy up to ten BWRX-300 reactors in Hungary.

When utilities across multiple countries independently select the same reactor design, it means something good. Utility companies are conservative, risk-averse organizations that answer to regulators, ratepayers, and investors. They don’t make buying decisions based on marketing hype — their pick is what they believe will operate reliably and economically.

The Assessment

GE Hitachi’s design lacks the dramatic, “brand new” appeal of sodium-cooled fast reactors or molten salt systems, but it’s likely to turn on before many competitors.

They took technology that utilities understand, regulators trust, and operators know how to run, and shrunk it for contemporary electricity markets that value “tiny” over “turns on.”

So buckle up — here’s a round-up (in no particular order) of the five silliest things that nuclear activists have come out with over the past 12 months.

1. Nuclear costs will last generations

Our first stop is in Somerset, UK, where EDF is building the mighty Hinkley C nuclear power plant. Sadly, the plant is behind schedule and over budget — partly through delays caused by Covid and other factors. But even with this much ammunition in their favour, anti nuclear protesters still seem to feel the need to resort to the occasional ‘pork pie’ to bash the project.

“Hinkley C will be paid for by our children and grandchildren,” said the spokesperson of the imaginatively named Stop Hinkley protest group recently. But this is inaccurate: cost overruns are real, but energy contracts, government guarantees and the complex way electricity markets and levies work mean bill impacts aren’t as simple as the slogan implies. It’s a political framing rather than a neutral technical statement.

2. Nuclear is a ‘trojan horse’ for fossil fuels

Next up, we head Down Under — where billionaire renewables enthusiast Simon Holmes a Court has been peddling his uniquely Australian take on anti nuclear whackaloonery.

His latest take is to claim that the Australian coalition government’s policy to introduce nuclear in the country is “a Trojan horse for extending fossil fuels”. Um, how, exactly, if Australia is looking to replace its coal plants with nuclear?? If he thinks those coal plants will be replaced with renewables any time soon, he’s wrong — unless he wants to see a renewables-heavy grid without baseload that will likely behave like Spain’s.

3. Discarding LNT will create ‘sacrificial victims’

Not to be outdone, US pressure group Beyond Nuclear have been drinking similar Kool-Aid, claiming that abandoning the Linear No Threshold (LNT) theory of radiation exposure will result in ‘sacrificial victims’. Cue a massive Gish Gallop of debunked tropes, most of them citation free (or linking to non peer-reviewed work by other anti nuclear pressure groups), but bonus bonkers points for linking to at least one debunked study and another one that’s nearly 70 years old. The reality is that LNT was never based on sound science in the first place — check out ‘The LNT Report’ for the full story.

4. We can’t trust nuclear companies

You can always rely on the Bulletin of the Atomic Scientists for some daft pronouncements. Clearly in need of some alternative fodder to push to the media instead of their incredibly tiresome and decades-old ‘Doomsday clock’ gimmick(sometimes you wish the apocalypse would actually happen just so that you don’t have to listen to this tired old schtick anymore), a certain Matt Smith (thankfully not the former Doctor Who star) is dissing Holtec because… actually, I’m not entirely sure. Because the company wants to expand? Because it wants to build SMRs? The whole article is really just an disingenuous attempt to sling mud at a company in the hope that some of it will stick. And none of it does.

5. Nuclear will distract us from the ‘real’ solutions

Climate Action Network — one of Europe’s largest environmental pressure groups — put out a daffy report earlier this year in which they claimed small modular reactors (SMRs) would be a ‘distraction’ from “more realistic, lower-cost solutions such as renewables, storage, and demand side solutions”. Ah, ‘demand-side solutions’ — presumably, in their heads, meaning improved energy efficiency; but which in reality will more likely mean higher prices, rolling brownouts and the like (also known as rationing, but they’d never put it that way).

6. ‘New nuclear a backward step’ says politician in country that’s had nuclear power since ::checks notes:: 1959

Oh, alright then — just because it’s Christmas, here’s a bonus one — from Scottish Green politician Mark Ruskell. In what must be an attempt on the world record for Most Inaccuracies In A Single Paragraph, he said earlier this year that: “New Nuclear energy would be a massive backwards step. It is a toxic and costly distraction from the clean, green renewable investment that we badly need to make. It is unsafe, unreliable and will not do anything to lower the obscene bills that households across our country are being hit with.” How sad that he’s chosen to ignore Scotland’s proud history of nuclear generation — one that began in the Fifties and which has not only provided low-cost, low-carbon energy but also jobs for life that the locals have loved.

So there we have it — for my money, the silliest things that have emanated from anti nuclear voices over the past 12 months. A resounding ‘bah humbug’ to the lot of them, though to be fair, in a way we’re glad they’re there — partly because we never tire of using the opportunity to rebut the pronouncements to anti nuclear activists, groups and ‘experts’ as a way to educate people and do some science outreach work. But partly also because they’re also always just so hilariously wrong.

Merry Christmas!

Subscribe now