While I usually do not engage in on-line nuclear discussion, given today is the anniversary of the worst civilian nuclear disaster ever happened, I think I would give my two cents of thoughts on this matter.

Nuclear is simply too slow, costly, and risky, while VRE is fast, cheap, and low risk

When arguing for the growth of nuclear sector in energy system, people often use current share or current growth rate of renewable energy to talk about decades into the future. This overlooks the exponential curve we are witnessing currently in renewable and their complementary flexible technologies. For example, this year IEA forecasted that by 2030, electricity generation from solar PV would grow by 118% compared with 2025. This is impressive on its own, but it is still an underestimation, if you consider the yoy growth rate of variable renewable energy capacity (22.2% according to IRENA), which would result in an 170% growth in the same time frame.

Of course, variable renewable energy sources are also cheaper and less risky (in terms of cost overrun), and the learning curve is much better for renewables than for nuclear. Nuclear also infamously requires longer construction time as time passes.

Arguing that the fastest and cheapest available energy sources cannot replace fossil fuel sources fast enough, so that we should divert our resources to deploy much slower and much expensive energy sources is self-contradictory.

Nuclear competitiveness is not unfairly hindered politically

There is also the argument that nuclear competitiveness is hindered politically when compared with renewables. Given the risk profiles and flexibility of project development, I would say the opposite is true. Suppose for a moment, a nation decided to build 10 GW of solar PV on ten of thousands of rooftops and thousands of solar farms in a given time frame. Most of them would be built on time, but let us just say 10% of them did not deliver on time. That would still mean 9 GW of low carbon and renewable electricity capacity. While this risk characteristic is good both financial and climate wise, it also means that any sudden change in the political landscape can result in cancellation of the expansion plans immediately with little opportunity costs of incomplete projects.

In an alternative timeline, suppose the same nation decided to build one single 1 GW nuclear reactor. Let us be generous to nuclear and just assume that once operational, this reactor will produce the same amount of electricity annually with the aforementioned 10 GW of solar PV. Given the bad record of the industry, it is almost certain that the project would not be completed on time. This means 0 GW of low carbon electricity capacity when the planned time frame is reached (looking at the currently missing 7% of electricity promised when Hinkley Point C began construction…).

Financial and climate wise, this is a very risky project, but it also means that once committed to the project, it becomes a project “too big to fall” and also a politically “stable” option, since the opportunity cost of abadoning the project becomes too high. The true essence of politics, after all, is to create irreversible path dependencies so that your successors have no choice but to follow what you have planned. Regardless of the moral hazard that might accompany, this sets nuclear (political wise) in a more favorable position than renewable energy. The fact that renewable energy has become the preferred technology of new energy generation investment regardless simply means that despite this political advantage, nuclear is still too costly and risky as an investment project.

The necessary fast and large scale adoption of new technology in the market for nuclear does not exist

New nuclear is often painted as some leap in technology. The need of technology innovation for a successful nuclear renaissance is true, but the plausibility of fast and large scale adoption of the new technologies is yet to be proven.

As of February 2026, the gap between actual operational and announced goals of SMRs remained large. Even the “realistic forecast” of 80 GW of SMRs by 2040 seems far-fetched.

Meanwhile, as news of record-breaking growth of renewables probably have become normalized, the fast and large scale adoption of technological innovation involved is often overlooked. 5 years ago, monocrystal PV, mainly PERC at that time, just completed its domination in the solar PV market share. In the past 5 years Top-con has squeezed the share PERC into oblivion, and HJT are currently on the rise. In a decade, conversion efficiency of commercially available solar PV has increased by about 5 p.p., or by 20%, all when the installation speed experiences an exponential growth. Such fast and large scale deployment and adoption of new technologies has rarely been seen in the history of energy generation, and the trend is no where ending.

In the early 2020s, PERC replaced traditional monocrystal PV, only to soon be replaced by n-type PV technologies.

There is no need for investment in baseload generators in a VRE-dominated energy system

Finally, the necessity of firm baseload generators is often brought up as an argument for new nuclear. Ignoring the fact that solar plus battery are becoming cheaper than new conventional power plants and around-the-clock solar + BESS systems already exist, there is neither technical nor economical case for the investment of baseload generators in an energy system dominated by renewable energy, especially the ones where incumbent hydro and nuclear power plants already exist.

For example, here is a recent study on European energy system in 2045 investigating under what scenarios will the investment of a baseload power plant become economically justified in the system. Here is one paragraph of their conclusion:

System-level modeling for Europe shows that the question is not whether new baseload plants are essential for a secure, net-zero grid—they are not. The defining question is whether they can become economical in a system dominated by low-cost renewables. The evidence suggests their competitiveness is questionable, requiring costs that current projects consistently fail to achieve.

What happens if we decide to build those new nuclear power plants aggressively anyway? Overall, it matters little to how the energy system will operate (see Figure 2 and 3 of the aforementioned study).

Economic dispatch of the European energy system by 2045, with or without new baseload generators. The impacts of new baseload generators on the system is minuscule compared with VRE.

Just look at China, with the largest nuclear construction program on the planet, nuclear share in electricity has nonetheless remained steady at no more than 5% in the last few years. These investments are simply too slow and too expensive to make a difference in the general picture of the energy landscape, presenting a misguided distraction at best, and an ill-intended campaign for slowing down renewable investment at worst.

Conclusion

In the end, I think on-line nuclear discussion is usually toxic and unproductive. Facts simply do not care about people’s feelings. Electricity generation from both solar and wind alone will surpass that from nuclear this year, and in a decade or so solar will probably become the largest source of electricity generation. The burden of proof lies in nuclear people to demonstrate the necessity of their beloved new technology (be it SMR, thorium reactor, fusion reactor, or whatever they believe to be the silver bullet to solve all of nuclear industry’s problems) in this renewable-dominated future, and so far they are not doing it well.