Although global nuclear generating capacity grew 2.9% to a record 2,663 terawatt-hours (TWh) in 2024, the industry won’t likely be able to sustain that growth in the face of limited investment, aging power plants, continuing project delays, and overwhelming competition from cleaner, more affordable renewable energy, a leading industry analyst concluded this week.
The nuclear industry struggled for real relevance in 2024, with just seven new reactors brought online and four shut down, while solar alone added hundreds of gigawatts of new capacity, Mycle Schneider writes [pdf], in the latest edition of his annual World Nuclear Industry Status Report (WNISR). Energy storage “passed a trigger point,” there are first signs of “a revolution behind the meter,” and low-income countries are starting to leapfrog to renewables.
“Risks around aging fleets, sluggish construction, accelerating system disruption from renewable energy, and China-centred development are expected to impact growth and lead to declines in regional electricity production shares,” Reuters states, citing the report. “To keep global nuclear output steady through 2030 the world would need 44 additional startups beyond those already scheduled, lifting annual startups to roughly 2½ times the past decade’s pace.” [Assuming, against a great deal of history, that a nuclear project scheduled today could begin generating power in just five years—Ed.]
That won’t be easy, with more than one-third of the world’s 63 nuclear construction projects behind schedule, 14 of them reporting increased delays. Meanwhile, “competition from cheaper non-hydro renewables and battery storage is expected to have a broad impact, as investment in renewables was 21 times that of nuclear last year, while added capacity was more than 100 times net nuclear additions,” Reuters says. “Battery costs are also falling, down about 40% in 2024, while nuclear plant costs continue to rise.”
Most of the nuclear growth is in China, the WNISR data show, where capacity grew by 3.5 gigawatts in 2024 while solar added 278 GW. “Between 2005 and 2024, there were 104 startups and 101 closures,” the report states. “Of these, 51 startups and none of the closures were in China. Thus, outside China, there has been a net decline by 48 units over the same period.”
The report warns that it will be increasingly difficult for nuclear to break into a global energy system dominated by renewables. “New energy technologies disrupt markets and systems,” it explains. “Photovoltaics directly produces electricity from solar radiation in harmless, nanometer-thin semiconductor junctions, allowing for ongoing steep cost reductions and performance increases. This is complemented by similar advances in power electronics and batteries. Together, these new technologies are evolving towards a highly flexible, fully electrified energy system with a decentralized control logic, outcompeting traditional centralized fossil and nuclear systems.”
Nuclear energy, it adds, “increasingly has difficulties to survive in this context.”
“The continuous disconnect between public narrative and industrial reality remains very worrying,” Schneider told PV Magazine.
“It is startling to see the breathtaking speed of the transformation of the world’s power systems,” he added. “While the nuclear industry proudly announces—correctly—a new record of nuclear power generation, the event is insignificant in the system context.” The 14 TWh of output the industry added this year broke an 18-year record, but it was only the equivalent of a single large reactor, “or what UK offshore wind turbines generated in the fourth quarter of 2024.”
Outside China, the report states, “nuclear generation in 2024 remained 363 TWh below the 2006 level, an almost 14% plunge.”
Despite the incremental growth in nuclear output worldwide, the industry is still mired at 9% of the global electricity mix, its lowest value in four decades, down from 17.5% in 1996. That share “is likely to erode further…unless project delivery and economics improve markedly,” Reuters writes. The world’s nuclear fleet is also aging—around two-thirds of reactors have been in operation for more than 31 years, and their average age increased slightly from 32 to 32.4 years between mid-2024 and mid-2025, another measure of the small number of new projects going online.
And so far, very few of those projects are the small modular nuclear reactors (SMRs) that have captured the imagination of politicians and pundits in Canada, the United States, the United Kingdom, and beyond. China is the only country with two designs in operation or built, with limited operational data available. Elsewhere, they “remain largely aspirational, as despite rising public and private funding, no Western SMR construction has begun,” Reuters says.
In September 2020, with nuclear already losing ground to affordable renewables, Schneider tagged SMRs as “PowerPoint reactors, not detailed engineering, and it’s not the first time. They’ve been doing this for decades.”
“Nobody, not even industry, pretends they can produce anything before 2030. That’s the earliest,” he told The Energy Mix at the time. So “it’s already very simple—it’s much too late, and we don’t know if it’ll work or what it’ll cost.”
Only 11 countries hosted nuclear construction projects in mid-2025—two fewer than in mid-2024, and five fewer than in mid-2023—and companies controlled by the Chinese and Russian governments accounted for 44 of 45 of the world’s reactor construction starts between January 2020 and mid-2025, the WNISR states. China accounted for seven of the nine construction starts in 2024. Russia “continues its key role” as a supplier of uranium fuel for its own reactors, and as a provider of parts and service to western nuclear companies, while the Zaporizhzhia site in Ukraine, Europe’s largest, is a “constant cause of concern” due to the continuing occupation by Russian troops.
The nuclear industry may also be running out of the uranium fuel it relies on to run its reactors. Earlier this month, the World Nuclear Association warned that supplies will be tight if the industry ramps up, with existing mines expected to halve their output between 2030 and 2040. The “significant gap” between demand and supply projections would threaten an awaited nuclear power “renaissance,” the Financial Times wrote (although this week’s analysis would appear to solve that problem).
The WNISR also reports that:
• The total of 408 nuclear reactors in operation in mid-2025 was unchanged from the previous year, down 30 from the industry’s peak in 2002.
• Russia dominates the international market for nuclear reactor construction projects, with 20 under way in seven countries. While multiple countries have announced plans to build their first nuclear reactors, only Bangladesh, Egypt, and Türkiye have started construction, all with Russian designs.
• In France, the Flamanville nuclear plant finally went online, 12 years late and at a cost of US$25.6 billion, “a staggering sixfold increase over the original $4.3 billion estimate.”
• At the site of the 2011 Fukushima nuclear disaster in Japan, “onsite and offsite challenges remain overwhelming, with an initial removal of fuel debris amounting to around a billionth of the total,” the WNISR states. “A focus on food safety monitoring finds an opaque system that makes it challenging for the government to convince observers it has control over the situation.”
• Of the 218 nuclear reactors that have been shut down worldwide, only 23 have been fully decontaminated, and only nine of been released by regulators and declared safe.
Pierre Arnold 
Mitchell Beer 
Delores Broten 
Nicholas Geary 
did you mean only 23 have been fully DE-contaminated?
Thanks, yes, and correcting that immediately!
And this – Nikiforuk at his best!
https://thetyee.ca/Analysis/2025/09/22/New-Nuclear-Fever-Debunked/
Mycle Schneider is a leading nuclear industry analyst? That’s like saying Donald Trump is a leading expert on windmills.
Well, heh, that’s kind of extreme!
But Nicholas, I find this so interesting, tbh! I’ve been following Schneider’s work for some years, and I’ve noticed that whenever anyone doesn’t like his conclusions they attack him ad hominem, rather than going after the substance of his work. We don’t have the resources to fact check every data point in a ±580-page report, no news organization does — whether it comes from an analyst who’s paid to stand by their numbers, or a trade association that is meant to combine data with a decent or indecent share of spin. But I appreciate about the World Nuclear Industry Status Report is that it *is* a compendium of industry data and analysis is verifiable by anyone with doubts.
So, not asking you to do an unreasonable amount of homework, either, but this is a genuine question — can you point to any specific data points in the WNISR that you think are inaccurate or misleading? Throw us a substantive question, rather than a personal attack on the author, and we’ll run it down one way or the other.
Is it now an ad-hominem slur to characterize Schneider as the anti-nuke he is? That’s new. And yes, he has some data in his report, but half-truths, eliminated context, and cherry-picked facts ladled with opinion and spin does not constitute analysis.
Take this tidbit from his reporting on the Fukushima water release program: “The water-treatment system, Advanced Liquid Processing Systems (ALPS), continues to underperform, with only 34 percent of the 1.18 million m³ of the water it had processed satisfying regulatory standards as of March 2025—little changed from the previous year. The rest needs to be repurified.”
The 34% figure is accurate. Characterizing the system as underperfoming is not. The first priority Tepco had was bringing down the radiation levels in the vicinity of the storage tanks to reduce worker exposure, and for that, they did a rapid treatment which removed the bulk of the contaminants and brought proximity radiation down to workable levels. For some of the water, they also did a trial run of high purification at slower processing speed, to confirm the system could take the water to release-standard purity. But trying for high purity in a tank system which had frequent transfers from tank to tank was impractical and wasteful, and years ago they decided they would set up a dedicated ALPS line just for the release operation. So the ALPS line dealing with water coming into the tanks isn’t underperforming, because they are not using that system to bring the water to release standard, and the release operation ALPS line is not underperfoming because every release batch has had lower contamination than the regulatory standard–which he conveniently doesn’t mention.
In his section “Power Plants Need to Be Large” he writes: “Due to several physical laws, thermal and mechanical devices are intrinsically more efficient, e.g., for power generation, when they are larger. These basic characteristics as well as the constant search for economies of scale have led to ever increasing average sizes of nuclear power plants. They also make the search for an economical small reactor rather illusionary.”
That misses (or deliberately omits) the potential savings from changing the nature of the technology. The reactor vessel Kairos is building, for example, is only rated for around 75 MW electric, which is about 1/16 the scale of an average large high-pressure nuclear reactor vessel. But the Kairos vessel runs at low pressure, so it doesn’t need to be made in a single huge forging, and can be rolled, stamped, and welded with common and cheap production techniques, reducing the vessel cost by a factor of around a thousand–or 50 to 70 times cheaper than the big vessels per MW of capacity. Plus, the Kairos plant won’t need that massive steam containment dome. Or the expensive interleaved, multi-redundant safety systems designed to keep the reactor pressurized and the core under water at all times. And it can use smaller, cheaper off-the-shelf steam turbines instead of the enormous custom-built units the big plants use. etc.
Section heading: “Controlling a Potentially Explosive Nuclear Chain Reaction” Did that section explain how nuclear power reactors could potentially produce a nuclear explosion? No it did not. It just talked vaguely about the risks of nuclear bomb explosions and nuclear reactor accidents. (Number of times this report mentions risk: more than 100. Number of times this report mentions Triso: zero) And the reason he doesn’t describe how a nuclear power reactor accident can produce a nuclear explosion is that there is no such scenario. And any actual nuclear expert would know that. So either he’s not the nuclear expert he styles himself to be, or he is including innuendo he knows to be false because it sounds scary. Either way, doing nuclear power expert cosplay and putting out a FUD piece dressed up to look like an industry report is simply dishonest.
I didn’t read this report. I’ve read some of his previous reports, so I know the schtick. The above examples were gleaned from a few minutes of skimming. I expect a comprehensive critique from an actual nuclear expert would be voluminous. The good news for Schneider is that the old-tech version of nuclear he’s long been railing against is in decline. And with new kinds of nuclear in development, orders for more old-tech will probably fall off while many potential customers wait to see how the demo reactors perform. And the new kinds won’t reach the commercial stage for roughly another five years, so he’ll be able to keep doing this report in the same manner for that long, at least. But for actual analysis of what’s good and what’s bad about Gen 4 nuclear (he completely misses the proliferation hazard of the China thorium reactor, btw) we’ll have to look for that information elsewhere.
No, that’s not ad hominem. Comparing Schneider’s expertise on the nuclear industry to Donald Trump’s on windmills? Much more so. Unlike Schneider, I don’t think Trump is capable of publishing multiple hundred pages (or multiple hundred words) in an exhaustive annual review with more than 2,500 references.
You seem to think it’s automatically disqualifying for someone to assess or comment on the nuclear industry if their expertise on the subject leads them to be deeply skeptical of the technology and its prospects. That’s amusing, but worrying. A really mature industry would welcome that kind of informed critique — and BTW, I would set exactly the same expectation for the renewables.
“You seem to think it’s automatically disqualifying for someone to assess or comment on the nuclear industry if their expertise on the subject leads them to be deeply skeptical of the technology and its prospects.”
I think it is disqualifying to try to mislead through hyperbolic editorializing, cherry-picking, context-stripping, and placing exaggerated emphasis on dangers which have small risk or effect–to the neglect of far greater risks and harms which these could displace. Schneider’s expertise isn’t in the field of nuclear power. It’s in the field of anti-nuclear propaganda.
“A really mature industry would welcome that kind of informed critique”
Regarding “mature industry” a lot of his FUD was directed at new companies trying to develop new kinds of nuclear. They haven’t even gotten into production yet. Regarding “informed critique”, you asked if I could point to any specifics in the WNISR that I thought were inaccurate or misleading. I provided examples and explained how they were misleading. And your response was to just ignore those and to simply reassert that this was an “informed critique”. And why should the traditional nuclear industry welcome or even countenance anything from an opposing industry which has been hell-bent on its destruction, with a long track record of dishonesty?
I do actually feel the public dialog over nuclear power would benefit from rational skepticism and informed criticism. There are many nuclear pathways ahead of us. Some of them have potential to be of great benefit to the planet. Some of them are overhyped and unrealistic. Some of them are actually fraught with danger. And with dozens of designs in development simultaneously, the public doesn’t know what to make of the flood of information, claims, and rhetoric coming from many sources, and there is a felt need for honest, impartial, expert analysis. But it looks like the leading anti-nuke individuals and organizations are not in a good position to fill that role, as they long-ago opted for the fearmongering-for-dollars business model. So if anti-nuke is now seen as a disreputable label, they did that to themselves.