[Column Series] Key Issues to Address in Japan's Strategic Energy Plan (No.9) How Difficult is it to Expand Nuclear Power in the World?

in Japanese

[Special Contents] Key Issues to Address in Japan's Strategic Energy Plan

While discussing the next Strategic Energy Plan, the Japanese Government actively promotes the idea to maximize the use of nuclear power for decarbonization purposes. However, an analysis of international trends, based on data and facts, demonstrates that renewable energy (RE) is the decarbonized technology to be prioritized. New wind and solar projects are much cheaper than new reactors. Therefore, global electricity generation from RE is growing much faster than that from nuclear power. And in terms of electricity generation mix, RE largely prevails over nuclear in China, the United States, and Europe – the world’s three largest power systems, as well as in Japan. 

New Wind and Solar Three to Six Times Cheaper Than New Nuclear Power 

In December 2023, BloombgerNEF, the leading reference for energy economics, published its latest levelized cost of electricity (LCOE) analysis, for the second half of 2023 (2023-H2). It found that the unsubsidized global benchmark LCOEs (i.e., based on a central scenario using moderate assumptions) of onshore wind, solar photovoltaic (PV), and offshore wind were around three to six times lower than that of nuclear (Chart 1).        

Chart 1: Global Benchmark LCOE New Wind, Solar PV, and Nuclear 2023-H2

The outstanding cost competitiveness of these RE technologies is the result of economies of scale and technological innovation. 

Between 2000 and 2023, the world’s wind capacity (including on- & off-shore wind) increased from 17 gigawatts (GW) to 1,017 GW.¹ In the same period, the world’s solar capacity (including solar PV and concentrated solar power) increased from 1 GW to 1,419 GW, out of which +346 GW in 2023 alone. Thus, combined wind and solar capacity increased by more than 2,400 GW, which is enormous. 

Global Trends – Nuclear Power Stagnates and Is Three Times Smaller than RE 

Based on the Statistical Review of World Energy (73rd Edition) published in June 2024 by the Energy Institute, a not-for-profit organization headquartered in London, United Kingdom (this institute took over BP’s historical publication in 2023), while RE blossomed, nuclear progress stalled. 

The world’s peak for electricity generation from nuclear was reached in 2006 (2,803 Terawatt-hours (TWh)), nearly two decades ago (Chart 2). 

In 2023, electricity generation from RE became 3.3 times bigger than that from nuclear: 8,989 TWh against 2,738 TWh. 

Chart 2: World Electricity Generation from RE and Nuclear 2000-2023

This is because the growth of nuclear has historically been concentrated in three power systems: Europe, the United States, and Japan, where nuclear lost steam due to various reasons. 

These include the accident at the Fukushima Daiichi nuclear power plant in Japan in 2011, unreliability, and economic competition from cheaper alternatives. And the long-term decline in electricity generation from nuclear in Europe, Japan, and the United States has not been offset by increases in other countries (i.e., mainly China).        

China, the United States, Europe, and Japan – Renewable Energy Prevails Everywhere

According to the International Energy Agency, in the world’s major power systems in 2023, the share of RE ranged between 22% in the United States and 49% in Europe (Chart 3). In comparison, that of nuclear ranged between only 4% in China and 19% in Europe.

Chart 3: China, United States, Europe, Japan Electricity Generation Mix 2023

In these power systems the cost advantage of new wind and solar, with and without battery storage, over new nuclear is often evident (Chart 4). Only in China new nuclear is not completely outcompeted.     

Chart 4: Selected Countries LCOE New Wind, Solar PV, and Nuclear 2023-H2

The rest of this section separately provides specific information about the situations in China, the United States, Europe, and Japan.

■China

The Chinese nuclear power industry is currently the world’s most active and performant when it comes to new reactors. As of the beginning of September 2024, 56 reactors with a combined capacity of 54 GW were in operation in China.² All but three of these reactors were connected to the grid between February 2002 and April 2024. Moreover, China is constructing 28 reactors with a combined capacity of 30 GW.  

According to BloombergNEF, the unsubsidized benchmark LCOE of new nuclear in China is $62 per megawatt-hour (/MWh). This is remarkably affordable compared to global standards. This level of cost is achievable thanks to a combination of three domestic factors: low labor costs, a favorable financing environment – critical for capital intensive projects, and rapid construction times (six years on average).³

Despite China’s substantial and rather successful efforts in expanding nuclear in the past two decades, its reactor fleet is still smaller than that of France in terms of installed capacity (56 reactors with a combined capacity of 61 GW).⁴ This is somewhat striking insofar as China’s power system is, on an electricity generation basis, more than 18 times bigger than that of France.⁵

To keep putting things into perspective and moderate enthusiasm towards nuclear developments in China, it is also useful to compare these developments with those of wind and solar in this country. 

In China between 2020 and 2023, electricity generation from wind and solar grew approximately 5-6 times faster than nuclear (Chart 5).

Chart 5: China Change in Electricity Generation from Nuclear, Wind, and Solar 2023-2020

Considering the trade importance of the Chinese nuclear and RE industries by comparing export values is also enlightening. Since 2020, China has constructed only two nuclear reactors outside of its borders: Kanupp-2 & -3 (1,017 megawatts (MW) each) in Pakistan, worth around $10 billion (based on the construction cost).⁶ At the same time, the value of China’s solar PV exports were close to $159 billion.⁷

■United States

As of the beginning of September 2024, the United States was still the world leading country for nuclear power: 94 reactors with a combined capacity of 97 GW.⁸

In June 2024, Lazard, a leading financial advisory and asset management company, published a report dedicated to the cost of generating electricity in the United States.⁹
 It found that existing amortized nuclear reactors, whose costs are essentially fuel and operation & maintenance costs, can generate electricity at low cost: $32/MWh 

This achievement is made possible thanks to a stance towards safety which is not particularly stringent, and very high-capacity factors as reactors are operated in baseload mode (country-wide average of 92-93% over the last decade).¹⁰

Regarding safety more specifically, while the American approach aims at preserving the original safety level of reactors, the French approach aims at continuously upgrading it to the highest level (required by a law of 2006, but this practice was already implemented since the 1980s). Therefore, the American approach enables cost savings and higher reactor availability.        

Generating electricity at $32/MWh is remarkable but does, however, not prevent nuclear reactors to suffer from economic competition in the United States.  

Still according to Lazard, in this country unsubsidized new onshore wind and solar PV can generate electricity as low as $27/MWh and 29/MWh, respectively, and existing amortized combined-cycle gas turbines (CCGTs) can generate electricity at $30/MWh.¹¹

In the United States between 2013 and 2022, the competition from gas and RE led to the premature permanent shutdowns of 13 reactors (combined capacity of 10 GW).

Because nuclear is now recognized as a source of clean electricity by the American Government (which is controversial due to issues related to spent fuel & radioactive waste), existing reactors benefit from subsidies (i.e., up to $15/MWh) since January 1, 2024.¹²

As for new reactors, the American nuclear power industry is confronted with difficulties for advancing both large-scale reactors and small modular reactors (SMRs). 

The construction of the United States’ two latest large-scale reactors Vogtle-3 & -4 (1,117 MW each) which commercial operation started in July 2023 and April 2024, respectively, has been unsatisfying. Because of delays it took a little more than ten years, instead of four-five years as initially planned, to build this pair of reactors. Worse, Lazard’s estimates the LCOE of this project at $190/MWh, which is uncompetitive.¹³

It may be emphasized that the aforementioned construction period does not include all necessary preparatory work to start construction. The beginning of the Vogtle-3 & -4 project dates to the mid-2000s. It took nearly a decade to select the site of the new reactors, to certify the design of the reactors (i.e., Westinghouse’s AP-1000), to apply and obtain the construction permit and the operating license, and to reach financial agreements covering building expenses.

Furthermore, the United States’ most advanced small modular reactor (SMR) project, the Carbon Free Power Project, was terminated in November 2023.¹⁴ This project envisioned the construction of NuScale’s VOYGR-6 SMR (462 MW / 6 modules of 77 MW). The start of operation was planned for 2029. The cost of this project was last estimated at $120/MWh, which was too high to attract customers. 

■Europe

In Europe, electricity generation from nuclear power is in long-term decline. In this continent in 2023, electricity generation from nuclear reactors reached only 736 TWh, 34% less than its peak of 1,122 TWh in 2004.

This decrease is mainly due to Germany’s politically motivated nuclear power phase-out, widespread outages in France, numerous permanent shutdowns of old reactors in the United Kingdom, and the war in Ukraine.

The evolution of the European electricity mix, the perpetual quest for improved safety, and investments in lifespan extensions (so far, no nuclear reactor ever operated beyond 60 years) result in rising costs for nuclear power. For instance, in France the generating cost of existing nuclear reactors was previously estimated at approximately $45/MWh. In July 2023, the French Energy Regulatory Commission provided a new estimate of around $65/MWh.¹⁵

Despite these challenges, existing reactors in Europe, unlike in the United States, hold a clear cost advantage over fossil power plants. This is because in Europe the fuel costs of natural gas and thermal coal are rather high, and because relevant carbon pricing is implemented. For example, In Europe in 2024-H1, for typical CCGTs and coal power plants, the sum of the fuel and carbon costs was $85/MWh.   

This explains why electricity generation from fossil power collapses in Europe. This also explains why this year France, whose electricity is almost entirely generated from nuclear (63% in 2023) and RE (28%), exports record volumes of electricity to neighboring countries which rely more on fossil power.¹⁶

Though existing nuclear reactors demonstrate competitiveness in Europe, the same cannot be said about the new reactors based on Électricité de France (EDF)’s EPR design. 

New EPRs under construction such as those of Flamanville-3 (1,630 MW) in France and Hinkley Point C-1 & -2 (1,630 MW each) in the United Kingdom are plagued with significant cost overruns and long delays. 

At the beginning of September 2024, it was expected that Flamanville-3 would be connected to the grid before the end of autumn 2024, after a construction time of seventeen years, out of which more than twelve years of delay.¹⁷The cost estimate for this reactor increased from $5.4 billion to $18.7 billion.¹⁸

In January 2024, the construction time of Hinkley Point C was revised to ten-thirteen years, including three-six years of delay, with start of operations in 2029-2031.¹⁹ The cost estimate for this pair of reactors increased from $34.4 billion to $54.4-61.4 billion.  

These failures raise serious questions about the successful realization of plans for new reactors.

France intends to build 6-14 new large-scale reactors by 2050. The site of Penly, which already hosts two operational reactors (1,330 MW each), should welcome a first pair of new EPR2 (1,670 MW each) by 2035-2037.²⁰ This means that in France over the next decade at least, except for Flamanville-3, all new decarbonized electricity will come from RE only.

The EPR2 should be a simplified EPR benefiting from the feedback of Flamanville-3. However, there are already some concerns that the first EPR2 could again be a first-of-a-kind reactor. This would penalize the EPR2 from being constructed faster and at lower cost.

And the United Kingdom pursues the goal of covering a quarter of its 2050 electricity demand with nuclear (against 13% in 2023).²¹According to the country’s “Civil Nuclear: Roadmap to 2050” published in January 2024, this would require deploying up to 24 GW (against 6 GW today).²² It is envisioned that securing investment decisions is needed to deliver 3-7 GW every five years from 2030 to 2044. 

This objective will not be achieved without the participation of several foreign companies. Finding the right partners will be difficult because international reactor vendors are scarce: mainly a few American, Chinese, French, Russian, and South Korean companies. 

Due to Russia’s invasion of Ukraine, and tense diplomatic relationships with China, this leaves only American, French and South Korean companies as credible options. French EDF already has a well-established presence in the United Kingdom. In May 2024, the Financial Times reported that South Korean Korea Electric Power Corporation (KEPCO) held early-stage discussions with British officials about a new nuclear power plant at Wylfa.²³ Both EDF and KEPCO are strained financially making governmental support a necessity.  

Sizewell C power plant, a replica of Hinkley Point C (i.e., based on the same EPR design), is the most likely project to move forward next. In May 2024, Sizewell C has been granted a nuclear site license.²⁴ The final investment decision on this project could be made in 2025. Once reaching this stage, more information on the cost and/or schedule should be communicated.         

Finally, important progress regarding the back-end activities of the nuclear power industry (i.e., reactor decommissioning, and disposal of spent fuel & radioactive waste) taking place in Europe also deserves to be reported in this section. 

At the end of August 2024, the company Posiva announced that the trial run of final disposal started in Olkiluoto, Finland.²⁵  During this trial run that should last several months, final disposal will be tested without actual spent fuel. The purpose of the trial run is to verify safe final disposal before the start of the actual final disposal operation (the exact date of which is not indicated). No country ever came so close to implementing such a solution.

Other countries with advanced plans for deep geological repositories include France and Sweden. In France in January 2023, the application for creation authorization of the Cigéo project was submitted to the French nuclear safety authorities. Construction could take place from 2027, and commissioning is scheduled for 2035-2040.²⁶ In Sweden in January 2022, the Government approved the construction of the Forsmark repository.²⁷ Construction could start in the 2020s, and the repository could be ready to start operations about ten years later. ²⁸

Neither the United States nor Japan made similar tangible progress. 

In 2022, the European Union Taxonomy (i.e., a classification system for sustainable activities) required that Member States pursuing nuclear power should have in place a detailed plan to have in operation a deep geological repository by 2050.

■Japan

Japan is the most chaotic and uncertain nuclear power market. More than thirteen years after the Fukushima-Daiichi accident, nearly two-thirds of the country’s 33 existing reactors (combined capacity of 33 GW) are still not operational (Chart 6). This is because restarting nuclear reactors is a costly and time-consuming process, sometimes faced with public opposition.

Chart 6: Japan Current Status of Nuclear Power Plants, as of August 13, 2024

According to the International Energy Agency, the 12 reactors (combined capacity of 12 GW) which were operational in fiscal year 2023 (FY, from April 1, 2023, to March 31, 2024) accounted for only 8% of Japan’s total annual electricity generation.²⁹ This is well below the country’s FY 2030 target of 20-22% which is probably out of reach. 

Another crucial issue in Japan is the lack of official up-to-date data on the costs of nuclear, both for existing and new reactors. Even BloombergNEF specialized in LCOE estimates, does not provide any data on the costs of nuclear in Japan. Without transparency, it is impossible to organize constructive debates which should lead to objectively deciding the fate of the country’s power sector.    

A new working group on the cost of generating electricity is currently organized by the Japanese Government for the next Strategic Energy Plan. The analysis conducted should be based on the latest data and refer to the other leading countries such as the United States, France, and the United Kingdom. It would then find that onshore wind and solar PV outcompete new nuclear reactors and compete with the existing amortized nuclear reactors.

The findings presented in this column are an invitation to the Japanese Government to responsibly set a realistic target for nuclear power in FY 2040. This is because an excessively ambitious target for nuclear risks undermining prospects for RE – the most efficient decarbonization solution on the supply side – and jeopardizing the country’s objective to reach carbon neutrality by 2050.