Welcome to nuclear history class. Today the professor is Yangbo Du and the student is you (and me). Let's get going~
"If you take primary energy production in France, go back to the early 70s, two-thirds was petroleum," Yangbo explains. "By the mid-1990s, petroleum was down to one-third. Almost all of that displacement was thanks to nuclear power."
The French success story reveals both nuclear's potential and its constraints. More importantly, it demonstrates why the industry's current approach to small modular reactors (SMRs) and data center partnerships might be missing the point entirely.
To be or not to Be(spoke)?
The difference between France's nuclear program and America's reads like a masterclass in industrial strategy versus chaos. While U.S. utilities pursued bespoke reactor designs—each plant essentially custom-built—France concentrated on a single pressurized water reactor design with three different capacities.
"What France did that the US did not do was concentrate on a single design of a nuclear reactor," Yangbo notes. "Different capacities, so not all plants had to be exactly the same size, but roughly three different sizes."
This standardization created the learning curve effects that economists predicted but rarely materialized elsewhere. French costs actually decreased through the program's early phases as workers, suppliers, and regulators all gained experience with the same technology. The U.S. experienced the opposite: each bespoke plant required unique engineering, custom components, and site-specific regulatory approval.
"You needed a whole host of factors in place to deploy nuclear successfully," Yangbo explains. "Government buy-in from the top level down, standard design so you can do repeat builds, and a supportive supplier ecosystem around the plants."
That ecosystem only remains viable with continuous construction. Stop-start cycles destroy the specialized knowledge base, force suppliers to exit the market, and require regulatory agencies to rebuild expertise from scratch.
China currently provides the clearest test of this theory. Building roughly one reactor monthly, they've maintained the supplier ecosystem and regulatory continuity that creates learning curve effects. Their costs remain "pretty manageable" compared to Western projects. Yet even China's program consistently falls short of government targets, and Yangbo identifies a concerning trend: "The fact that they're also trying to export a wide range of designs does not bode well going forward."
The implication challenges a core assumption in current nuclear strategy. Whether pursuing large reactors or SMRs, the path to cost competitiveness requires choosing one design and building it repeatedly—exactly what the fragmented, market-driven approach in most democracies struggles to achieve.
What Nuclear is Best For
The data center boom has created unexpected allies for nuclear developers among Big Tech companies willing to pay premium prices for clean, firm power. Meta's recent nuclear investments, Google's acquisition of a power generation company, and similar moves suggest nuclear's moment has arrived.
Yangbo sees this differently. "Nuclear is best matched with steady load 24-7," he explains. "If your load is going to vary basically hourly, nuclear is not at all a good fit for it."
Data centers might need high availability, but their actual power consumption varies significantly throughout the day. This mismatch between nuclear's strength (constant output) and data centers' reality (variable demand) suggests the current excitement might be misplaced.
The real opportunity lies in continental-scale grid integration, where nuclear's inflexibility becomes an asset. Yangbo describes the vision emerging across Europe: "Those nuclear plants in France would be a pretty good 10% of a future zero carbon European grid. So that would be your firm capacity that runs 24-7 no matter what."
This integration already shapes European energy strategy. France currently generates three-quarters of its electricity from nuclear—so much that during low-demand periods, nuclear plants must ramp down because there's nowhere to send the excess power. About a third of French nuclear plants now perform load-following operations they weren't designed for, leading to unplanned maintenance outages.
"But if you connect all of those energy generation and storage resources together with a continental-scale grid," Yangbo continues, "that's actually where you can see what is the most appropriate role of a nuclear reactor fleet."
The ideal vision integrates France's nuclear baseload with North Sea offshore wind, distributed solar across southern Europe, and pumped hydro storage in Scandinavia and the Alps. Each technology plays to its strengths while a continental grid balances supply and demand across time zones and weather systems.
The model works precisely because nuclear provides reliable output while other technologies handle variability.
Nuclear Power Plant Costs and Utility Challenges
The economic reality confronting nuclear expansion has shifted dramatically since the "nuclear renaissance" discussions of 20 years ago. When those conversations began, solar and wind carried significant cost premiums over fossil fuels. Today, utility-scale solar and wind with battery storage consistently undercut natural gas, coal, and new nuclear on levelized cost.
"The price of solar and wind has fallen by over 80% on the hardware, even 90% in some cases," Yangbo explains. "If you're really an investor looking at what is the best use of funds as far as energy is concerned, you'd go with what can be deployed very quickly and economically."
The modular advantage that makes renewables attractive extends beyond cost to risk management. Solar and wind farms generate revenue even when partially complete—a 10% finished project produces 10% of the intended power. Nuclear and fossil fuel plants produce nothing until 100% complete, creating massive capital risk for developers.
This fundamental difference explains why even China increasingly send its energy investments to solar, wind, and batteries despite maintaining its nuclear construction program.
SMRs promise to bridge this gap by bringing nuclear construction closer to manufacturing, but they face the same standardization challenge as large reactors. "For a nuclear plant, your reactor isn't always necessarily your biggest cost object," Yangbo notes. "What increases your costs even more significantly with a smaller reactor is the balance of plant."
The power generation equipment, security systems, and regulatory infrastructure don't scale linearly with reactor size. A small reactor still requires layers of defense around the perimeter, qualified operators, and emergency response systems. This suggests SMR deployment makes most sense at existing nuclear sites where infrastructure and expertise already exist.
The industry's current trajectory—dozens of SMR designs competing for market share—repeats the mistakes that undermined nuclear's first wave. Success requires the discipline to choose one design and build it repeatedly, exactly what market-driven systems resist.
Even the data center opportunity that's driving current investment comes with a timeline mismatch. Gas turbine wait times have stretched to four years due to data center demand, making seven-year nuclear development seem competitive. But data center technology evolves rapidly—today's facilities might be obsolete within three years as chip designs advance and efficiency improves.
"There is a very high likelihood that more and more companies will shift towards much more efficient computing models in response," Yangbo predicts. This could eliminate the power demand driving current nuclear investments before new plants come online.
The Continental Vision
Yangbo's perspective on nuclear stems from recognizing both its genuine value and its inherent limitations. Nuclear excels as firm, zero-carbon baseload power—the foundation upon which variable renewables can scale. It struggles as a flexible, market-responsive technology competing directly with increasingly cheap solar, wind, and storage.
The path forward requires abandoning fantasies about nuclear dominance and embracing a supporting role. "Nuclear is great given that you already have the installed base in place," Yangbo explains. "But going forward, you'd go with what can be deployed very quickly and economically."
This doesn't diminish nuclear's importance. The existing fleet provides decades of clean power while new continental-scale grids integrate diverse clean resources. France's nuclear plants become Europe's baseload backbone. Existing U.S. plants continue operating while regional transmission enables renewable scaling.
The key insight: nuclear's best path forward might require letting other technologies lead.
"Pretty much every single problem we're seeing right now has been solved by someone somewhere in the world already," Yangbo reflects. "It's really all about learning from what's already existed. Even some cases, what has been solved decades ago."
The lesson applies beyond nuclear to any capital-intensive clean technology: success requires systematic approaches that most market-driven democracies struggle to implement. Recognizing those constraints might be the first step toward working within them.
As the grid evolves toward continental scale and renewable dominance, nuclear's role becomes clearer: not the hero of the climate story, but an essential supporting character whose unique capabilities enable the real protagonists to shine.
To learn more about Yangbo Du's work in energy economics and policy, you can find him online through LinkedIn or a simple Google search—he notes he's likely the only person with his name maintaining an active online presence. Follow his insights as he continues focusing on immediate-term solutions for achieving 2030 climate targets.
Member discussion: