New report details nuclear power’s demise

NUCLEAR POWER’S GLOBAL STAGNATION AND DECLINE

Authors: Assoc. Prof. Darrin Durant, Prof. Jim Falk and Dr. Jim Green

Date: 27 February 2025

EnergyScience Coalition 2025 Briefing Paper No 1

The EnergyScience Coali1on is an independent non-governmental organisa1on established as a collabora1on of concerned scien1sts, engineers and policy experts. www.energysicence.org.au

INTRODUCTION

Over the last 50 years nuclear power has been shaped within drama8cally over-es8mated claims for its future prospects, versus the reality of its actual performance. The long history of plumme8ng expecta8ons goes back as far as 1972 when 3450 gigawaGs (GW) of global capacity was predicted for the year 2000, dropping to the predic8on in 1978 of 728 GW. That was just before the industry suffered the crippling Three Mile Island Accident with declines in projec8ons con8nuing on towards the more sober reality of 371 GW of nuclear capacity opera8ng across the world in 2000. The current push in Australia to deploy nuclear reactors over the next several decades again contrasts an excessive op8mism for nuclear proponents against the con8nuing stagnant situa8on of nuclear programs worldwide.

The latest nuclear proposals are built on three specula8ons.

First, projected AI-related energy demand where ‒ as with nuclear power proponents in the 1970’s who projected huge demand for nuclear power that never eventuated ‒ there are already signs likely demand from this source is overstated: for example, the new leading AI entrant, DeepSeek, requires just 10% of the energy of compe8tors.

Second, specula8ve techno-op8mism that new technologies such as small modular reactors will resolve industry project management issues. Yet these small reactors are underwhelming where they do exist and unproven otherwise.

Third, prospec8ve wish-fulfilment, where dozens of nuclear ‘newcomer’ countries are offered as saviours, despite not having approvals and funding in place in a large majority of cases.

This paper presents evidence leading to the following conclusions:

The global nuclear power industry is stagna8ng rather than growing, so claims that ‘Australia is being le_ behind’ have no basis.

Where there is growth, it is primarily in authoritarian countries such as China and Russia.

Small modular reactors are offered as ‘magic bullets’ but they do not exist and available evidence suggests they would be uneconomic.

The number of countries opera8ng nuclear power reactors is the same as it was in the late 1990s.

•

• •

•

The number of poten8al nuclear power newcomer countries with approvals and funding in place, or construc8on underway, is just three and all those projects are funded heavily by the Russian state.

Renewables growth dwarfs nuclear growth, by a factor of 155 last year. The same applies to China: a ra8o of 100:1 last year.

Western democracies building new reactors have all experienced extraordinary cost overruns and construc8on blowouts.

The current state of nuclear power

So what is the state of nuclear power? The World Nuclear Industry Status Report presents these details regarding developments in 2024:

The number of opera8ng reactors fell from 413 to 411, which is 27 fewer than the peak of 438 reactors in 2002.

Global nuclear power capacity was stagnant at 371 GW. Indeed, there has been no growth over the past 20 years and very liGle growth in the decade before that (less than one percent per year).

Source, IAEA, Nuclear Power Reactors in the World 2024

Likewise, nuclear power genera8on (typically measured in terawaG-hours) has been stagnant for 20 years with very liGle growth in the decade before that.

Source: World Nuclear Associa=on.

As of 1 January 2025, nuclear power accounted for 9.15 percent of global electricity produc8on, down from 9.2 percent a year earlier and barely half of its peak of 17.5 percent in 1996. The drop in nuclear power’s share of global electricity genera8on is certain to con8nue. In one improbable scenario, the Interna8onal Atomic Energy Agency an8cipates 60 percent nuclear power growth by 2050 but says that the nuclear share of global electricity genera8on will s8ll fall in that scenario, by 2.3 percent.

As of 1 January 2025, the mean age of the global power reactor fleet was 32.1 years. In 1990, the mean age was just 11.3 years. As the rate of closure of ageing reactors increases, it will become increasingly difficult for the industry to maintain its long-term paGern of stagna8on let alone achieving any growth. The Interna8onal Atomic Energy Agency projects 325 GW of nuclear closures from 2018 to 2050 or 10 GW per year on average. Average construc8on starts and reactor startups over the past decade have fallen well short of 10 GW per year so a higher build rate will be required just to maintain the paGern of stagna8on.

Construc1on starts and numbers of reactors under construc1on

The picture of nuclear reactor starts and numbers under construc8on was reported to be similarly constrained:

There were nine power reactor construc8on starts in 2024: six in China; one in Pakistan implemented by Chinese companies; one domes8c reactor in Russia and one in Egypt being built by Russian agencies. In the decade from 2015-24, the average has been 6.5 construc8on starts per year. The historic one-year peak was 44 in 1976.

In the five years from 2020‒24, there were a total of 40 reactor construc8on starts, of which 26 (62 percent) were in China, one in Pakistan carried out by Chinese companies, and the other 13 implemented by the Russian nuclear industry in Egypt, India, Turkiye, and at home. Russia also began building four reactors in China.

In the five years from 2020‒24, Chinese and Russian companies have been the only builders with reactor construc8on starts worldwide.

Source: World Nuclear Industry Status Report 2024 The World Nuclear Industry Status Report notes that:

As of 1 January 2025, there were 61 power reactors under construc8on, two more than a year earlier but eight less than in 2013 (and 173 fewer than the 1979 peak of 234).

Almost half of the 61 reactors are under construc8on in China (29), including four implemented by the Russian industry which is also building in Bangladesh (2), Egypt (4), India (4), Iran (1), Turkiye (4), and at home (6), thus a total of 25 reactors.

The only country besides Russia and China building abroad is France with two reactors under construc8on in the UK.

Almost all construc8ons (over 93 percent) are implemented either in nuclear weapon states or by en88es controlled by nuclear weapons states in other countries.

Of the 807 reactor construc8on starts since 1951, at least 93 reactors had been abandoned or suspended as of 1 July 2024, in 19 countries. That is 11.5 percent, or one in nine.

Small modular reactors

Given the actual experience of aGempts to develop and deploy small modular reactors it is not surprising that the op8mism of those pressing for renewed nuclear ac8vity has dwindled in rela8on to them. There were no small modular reactor (SMR) startups in 2024.

Indeed there has never been a single SMR startup unless you count so-called SMRs not built using factory ‘modular’ construc8on techniques, in which case there is one each in China and Russia. Cheap, quick mass-produc8on of SMRs is nowhere in sight. The so-called SMRs in China and Russia and China took 9 and 12 years to build, respec8vely, and in both countries planning plus construc8on took 20 years or more.

A few more so-called SMRs are under construc8on worldwide (in China, Russia and Argen8na) but none of these projects are using modular construc8on techniques. A 2024 report by the Australian Academy of Technological Sciences and Engineering notes that no small modular reactors exist in any OECD countries and the technology has not been proven technically or financially.

French u8lity EDF’s decision to suspend its Nuward SMR project was another setback in 2024 (following the previous abandonment of four other SMR projects: Flexblue, Antares, NP-300, ASTRID). So was the bankruptcy filing of US company Ultra Safe Nuclear ‒ the company’s assets were sold for a song (US$8.5 million) through an auc8on process conducted pursuant to Sec8on 363 of the US Bankruptcy Code.

Those failures followed the decision of US company NuScale to abandon its flagship project in Idaho in 2023 a_er cost es8mates rose to a staggering A$31 billion per GW, more than double the 2015 es8mate. The levelised cost es8mate rose to US$119 (A$189) per megawaG-hour (MWh). The Minerals Council of Australia states that SMRs won’t find a market in Australia unless they can produce power at a cost of A$60-80 / MWh. That is well under half the NuScale es8mate of A$189 / MWh.

A December 2024 CSIRO GenCost report provides these levelised cost es8mates for Australia:

Another (failed) nuclear renaissance?

Claims that 40‒50 countries are ac8vely considering or planning to introduce nuclear power, in addi8on to the 32 countries currently opera8ng reactors, do not withstand scru8ny. Current promo8onal claims lack substance and there is no credible reason to believe that the current claims of a nuclear ‘renaissance’ are likely to prove to have any more reality than the essen8ally never arriving ‘renaissance’ of the late 2000s.

Less than one third of the countries currently opera8ng reactors have ac8ve reactor construc8on programs: 10 out of 32.

As of 1 January 2025, reactors were under construc8on in just 13 countries, two less than a year earlier. Seven percent of the world’s countries are build reactors; 93 percent are not.

Of the 13 countries building reactors, only three are poten8al nuclear ‘newcomer’ countries building their first nuclear plant: Egypt, Bangladesh and Turkiye. In those three countries, the nuclear projects are led by Russian nuclear agencies with significant up-front funding from the Russian state.

The World Nuclear Associa8on (WNA) observes that apart from those three countries, no countries meet its criteria of ‘planned’ reactors, i.e. “approvals, funding or commitment in place, mostly expected to be in opera8on within the next 15 years”. The WNA lists six countries in its ‘proposed’ category, “specific programme or site proposals; 8ming very uncertain”. And the WNA lists 17 countries with “provisional plans, commitment pending or deferred”.

The number of poten8al newcomer countries with approvals and funding in place, or construc8on underway, is just three and all those projects are funded heavily by the Russian state.

The number of poten8al newcomer countries with approvals and funding in place, or construc8on underway, without substan8al Russian state funding, is zero. That is the underwhelming reality underlying highly exaggerated claims about 40-50 countries pursuing nuclear power.

Over the past five years, there have been no reactor construc8on starts worldwide other than Russian and Chinese projects at home and abroad. Apart from China’s construc8on project in Pakistan, and France’s twin-reactor project in the UK, only Russia is building reactors abroad (in China, Bangladesh, Egypt, India, Iran and Turkiye). But Russia’s capacity to fund such projects is diminishing, and the poli8cal fallout over Russia’s invasion of Ukraine puts further constraints on poten8al future projects abroad (as well as compromising and complica8ng exis8ng nuclear export projects).

There is no evidence of a forthcoming wave of nuclear newcomer countries in the coming years and decades. At most in prospect is a trickle. That has been the historical paGern with just seven newcomer countries building and opera8ng reactors over the past 40 years and just three this century: the Czech Republic (1985), Mexico (1990), China (1991) Romania (1996), Iran (2011), the UAE (2020) and Belarus (2020).

The number of countries opera8ng power reactors in 1996–1997 reached 32. Since then, nuclear newcomer countries have been matched by completed phase-outs and thus the number is stuck at 32. That is the underwhelming reality underlying exaggerated claims about 40-50 countries pursuing nuclear power.

Four countries have phased out nuclear power: Italy (1990), Kazakhstan (1999), Lithuania (2009) and Germany (2023). Five countries have begun reactor construc8on projects but abandoned those projects and never completed any power reactors: Austria, Cuba, Philippines, Poland and North Korea.

Countries now planning to phase-out nuclear power include Spain, Switzerland and Taiwan. An ‘organic’ nuclear phase out is underway in many other countries: exis8ng reactors are ageing and the prospects for new reactors are slim or nil.

It is doubvul whether the number of nuclear newcomer countries over the next 20‒30 years will match the number of countries comple8ng phase-outs.

Nuclear growth dwarfed by renewables

In striking contrast to nuclear power’s net gain of 4.3 GW in 2024, the Interna8onal Energy Agency’s (IEA) October 2024 ‘Renewables 2024’ report es8mates 666 GW of global renewable capacity addi8ons in 2024. Based on the IEA’s es8mate, renewables capacity growth was 155 8mes greater than that of nuclear power.

The IEA expects global renewable capacity to increase by more than 5,520 GW from 2024-2030 and the Agency expects renewables to jump from 30 percent of global electricity genera8on in 2023 to 46 percent in 2030.

Source: Interna=onal Energy Agency
In 2025, renewable electricity genera8on is set to overtake coal-fired genera8on.

In 2026, wind and solar power genera8on are each poised to surpass power genera8on from nuclear.

In 2029, electricity genera8on from solar PV is set to surpass hydropower, becoming the largest renewable power source globally, with wind-based genera8on expected to surpass hydropower in 2030.

A BloombergNEF analysis finds that renewable energy investments reached $US728 billion (A$1.16 trillion) in 2024, up 8 percent on the previous year, whereas nuclear investment was flat at US$34.2 billion (A$54.4 billion). Thus renewable investments were 21 8mes greater than nuclear investments.

In striking contrast to massive cost overruns with nuclear projects, renewable costs have fallen sharply. Lazard investment firm data shows that u8lity-scale solar and onshore wind became cheaper than nuclear power from 2010‒2015. From 2009‒2024, the cost of u8lity- scale solar fell 83 percent; the cost of onshore wind fell 63 percent; while nuclear costs increased 49 percent.

The IEA states: •

Sources: World Nuclear Industry Status Report 2024, Lazard 2024. China to the rescue?

In the 20 years from 2005 to 2024, there were 109 power reactor startups and 108 permanent closures worldwide. Of these, 51 startups were in China with no closures. Outside China, there has been a net decline of 50 reactors over the same period with 58 startups and 108 closures.

Thus worldwide nuclear power can only be said to be stagna8ng because of growth in China, outside of which there has been significant decline.

In China, nuclear growth is dwarfed by the growth of renewables. In 2024 there were three reactor startups with a combined capacity of 3.5 GW. That is 100 8mes less than the 356 GW of solar (277 GW) and wind (79 GW) capacity installed in China in 2024.

In 2020, China aimed to install at least 1,200 GW of solar and wind capacity by 2030. The target was surpassed in 2024. The Interna8onal Energy Agency expects China to install 3,207 GW of new renewable electricity capacity from 2024‒30. That expected growth in China alone is 8.5 8mes greater than current worldwide nuclear power capacity of 377 GW. And it is 40 8mes greater than the 80 GW of installed (50 GW) and under-construc8on (30 GW) nuclear capacity in China.

Lessons for Australia

Alongside the risk of Fukushima-scale disasters, the weapons prolifera8on risks, the risk of aGacks on nuclear plants (and the reality of aGacks on nuclear plants in Ukraine), and the intractable nuclear waste legacy, the reality is that nuclear power just can’t compete economically.

The industry’s greatest problem at the moment is a recogni8on of this by investors, resul8ng in a capital strike. Even with generous government/taxpayer subsidies, it is becoming

difficult or impossible to fund new reactors ‒ especially outside the sphere of China and Russia’s projects at home and abroad.

Who would bet tens of billions of dollars on nuclear power projects when the recent history in countries with vast exper8se and experience has been disastrous:

In France, the latest cost es8mate for the only recent reactor construc8on project, at Flamanville, increased seven-fold from €3.3 billion to €23.7 billion (A$39.4 billion) for just one reactor. Construc8on took 17 years. No reactors are currently under construc8on in France.

In the US, one project in South Carolina, comprising two Wes8nghouse AP1000 reactors, was abandoned in 2017 a_er A$14.3 billion was wasted on it. Wes8nghouse declared bankruptcy and almost forced its parent company Toshiba to do the same. All that remains is the nukegate scandal: an avalanche of legal ac8on including criminal cases.

The only other reactor construc8on project in the US ‒ the twin-reactor Vogtle project in the state of Georgia ‒ reached comple8on at a cost 12 8mes higher than early es8mates. The final cost was at least US$17 billion (A$27 billion) per reactor. Comple8on was 6‒7 years behind schedule. No power reactors are currently under construc8on in the US.

In the UK, the 3.2 GW Hinkley Point twin-reactor project was meant to be complete in 2017 but construc8on didn’t even begin un8l 2018 and the es8mated comple8on date has been pushed back to 2030-31. The latest cost es8mate ‒ £23 billion (A$46.3 billion) per reactor ‒ is 11.5 8mes higher than early es8mates. The UK Na8onal Audit Office es8mates that taxpayer subsidies for the Hinkley Point project could amount to £30 billion (A$60.4 billion).

No other reactors are under construc8on in the UK. The last power reactor startup in the UK was 30 years ago, in 1995, since when there have been 24 permanent reactor shut- downs.

The es8mated cost of the planned 3.2 GW twin-reactor Sizewell C project in the UK has jumped to nearly £40 billion (A$80.5 billion) or A$40.2 billion per reactor, twice the cost es8mate in 2020. Securing funding to allow construc8on to begin is proving to be difficult and protracted despite a new ‘Regulated Asset Base’ funding model which foists the enormous risk of enormous cost overruns onto taxpayers and electricity ratepayers.

Those three countries ‒ France, the US and the UK ‒ have vast nuclear exper8se and experience. They all enjoy synergies between civil and military nuclear programs. All of the above-men8oned construc8on projects were (or are) on exis8ng nuclear sites. All projects were (or are) long-delayed and tens of billions of dollars over-budget.

Claims that poten8al nuclear ‘newcomer’ countries such as Australia, without any of those advantages, could build reactors quickly and cheaply strain credulity. Whether renaissance or perhaps baroque, pro-nuclear claims seem based more on the emo8on and passion than the prac8cal reali8es of technical, economic and poli8cal ra8onality.

About the authors:

Dr. Darrin Durant is Associate Professor in Science and Technology Studies at the University of Melbourne. He has published widely on environmental controversies including climate poli=cs and debates about nuclear power and nuclear waste disposal, as well as on the rela=on between experts and ci=zens in democra=c decision-making, disinforma=on and democracy, and climate denialism. His most recent book is Experts and the Will of the People: Society, Populism and Science (Palgrave, 2020). Of relevance to nuclear maWers is

•

•

•

•

•

•

Nuclear Waste Management in Canada: Cri=cal Issues, Cri=cal Perspec=ves (UBC Press, 2009), and he is currently comple=ng a book on Australian debates about nuclear power 1998-2025.

Professor Jim Falk is a Professorial Fellow in the School of Geography, Earth and Atmospheric Sciences at the University of Melbourne; Senior Fellow in Melbourne Climate Futures; and Emeritus Professor at the University of Wollongong. Over some fi_y years he has been the author of many books and papers dealing with climate change and energy issues with a con=nuing focus on issues associated with nuclear technology.

Dr. Jim Green is the na=onal nuclear campaigner with Friends of the Earth Australia, a member of the Nuclear Consul=ng Group, and a former editor of the ‘Nuclear Monitor’ newsleWer produced by the World Informa=on Service on Energy.