The nuclear reversal: why the energy crisis became an opportunity
For fifteen years, nuclear energy was the technology everyone respected but nobody invested in. Solar and wind were cheaper, policy favoured renewables, and every new reactor cost more and took longer to build. The story seemed settled.
Then two things happened at once: Russia cut off gas to Europe, and China started building reactors at scale.
Today, in April 2026, we can look back at the turning point: in 2024, the global nuclear fleet produced 2,667 terawatt-hours of electricity in a single year, breaking a record that had stood since 2006. Nearly seventy reactors are under construction worldwide. Well over 120 more are officially planned. The industry that seemed to be dying is now the fastest-growing power source on the grid. This is not a trend. This is a reversal.
The policy shift was real
Before 2020, nuclear faced a political headwind. Post-Fukushima anxiety had scarred public perception in developed nations. Natural gas looked like a practical bridge. Renewables were improving exponentially. Nuclear spent a decade defending itself against the assumption that it was obsolete.
The war in Ukraine changed the calculation overnight. Cutting Russian gas forced Europe to reconsider every energy source, including the ones they had spent a decade trying to phase out. Simultaneously, China and India were building nuclear plants because their electricity demand was exploding. They didn't have the luxury of waiting for renewables to "get cheaper." They needed power.
By 2024, the International Energy Agency officially said nuclear capacity would need to triple by 2050 to meet climate targets. 35 countries are exploring new nuclear programs. Japan restarted reactors. The UK approved new builds. Even Germany, which had controversially shut down its last three plants in 2023, began aggressively pivoting by 2026, with leadership publicly calling the phaseout a "huge mistake."
This wasn't ideology. This was math. You cannot decarbonise a grid at scale without baseload power. Renewables cannot provide that alone. The only options are nuclear or storage the size of entire nations. Policy caught up to physics.
The technology problem is being fought by going smaller
The traditional nuclear industry had a fatal flaw: it could only build massive plants. A 1,200-megawatt reactor required a decade of construction, $10 billion in capital, and a utility large enough to absorb the financial risk. Smaller grids couldn't play. Smaller companies couldn't finance. The business model was broken.
The theoretical solution was simple: stop building one giant reactor. Build smaller, modular ones.
NuScale Power's initial module design was certified by the U.S. Nuclear Regulatory Commission in 2023. It took three decades of development. However, the transition from paper to steel has been brutal. NuScale was forced to cancel its flagship project in Utah late in 2023 because projected construction costs spiraled to $9.3 billion — ironically matching the massive financial bets Small Modular Reactors were supposed to avoid.
The economic model is clearly not "solved" yet. But the industry is pushing forward because the theory of factory-built components is still the only way out of the cost trap. The goal remains: instead of waiting a decade for revenue on a massive build, a utility can deploy one module, watch it work, see the cash flow, and finance the next one.
TerraPower took the idea further. Their Natrium reactor, which broke ground in Wyoming in 2024, uses liquid sodium cooling and integrated thermal storage. It runs hotter, stores heat in molten salt for hours, and can ramp power up and down to follow grid demand. It generates electricity, yes. But it also produces high-temperature steam for hydrogen production, chemical processes, even desalination. For an industry that spent decades saying nuclear could only do one thing, Natrium says nuclear can do whatever heat can do.
The fuel problem was real. The infrastructure play is solving it
There was one thing standing between advanced reactors and deployment: fuel.
Advanced reactors need fuel enriched to 5 to 20 percent uranium-235. This is called HALEU. It's not weapons-grade. It's not illegal. But a few years ago, there was essentially zero commercial production of it in the United States. Without it, the entire advanced reactor ecosystem hits a wall.
The global enrichment supply chain is heavily concentrated, and Russia's supply became politically toxic. Western companies couldn't scale production in a matter of months. The bottleneck was severe.
But the market adjusted. Centrus Energy began shifting the tide, producing the first commercial HALEU in late 2023 in Ohio and delivering 900 kilograms of it to the Department of Energy by 2025.
To achieve the massive scale required, the U.S. government stepped in. In January 2026, the DOE issued identical $900 million contracts to Centrus, Orano, and a heavily-backed start-up named General Matter to rapidly expand production. General Matter's approach typifies the new, pragmatic industry mindset: repurposing existing infrastructure. By targeting abandoned nuclear facilities like Paducah in Kentucky and Hanford in Washington — sites with enrichment infrastructure already in place — they aim to scale production in months, not years.
This scales toward the 50 metric tons per year by 2035 that the entire advanced reactor fleet will need. This is not revolutionary. It is boring and practical and exactly what an industry infrastructure play looks like.
The geopolitics reversed
The old nuclear story was about developed countries building massive centralised plants. The new story is about Asia building massive decentralised fleets.
China has nearly 40 reactors under construction. India has 8 and wants dozens more. Russia, Turkey, Vietnam, Bangladesh — the construction pipeline is dominated by countries that need power and are willing to take nuclear seriously. The mature markets — the United States, France, the United Kingdom — are deploying advanced designs and life extensions, but trailing in raw volume.
This is a power shift. Nuclear expertise is moving east. The supply chains are being built in Asia. The standardized designs are Chinese. The industrial knowledge is accumulating in countries that are actually building at scale. By 2035, when the advanced reactor fleet starts coming online in the West, China will have been building standardized large reactors for a decade. The learning curves will be in China's favor.
This is not inevitable. It is just what happens when one region builds and another does not. Nuclear manufacturing will become a competitive advantage. The countries that learn to do it cheaply and fast will own the next-generation power systems.
What this means
The nuclear story flipped from "inevitable decline" to "necessary expansion" in about five years. The reason was not technological breakthroughs. NuScale took thirty years just to get a design certified. The reason was political — climate targets and geopolitics forced a recalculation of what "necessary" means.
Now the constraint is not policy. It is execution. Can the supply chains scale? Can the cost curves of SMRs actually be beaten down? Can financing flow? Can the workforce exist?
These are solvable problems. Boring problems. Not glamorous. But the glamorous part — proving that advanced reactors can work and getting the political green light — is mostly done. Now it is implementation. Implementation is what wins.
The nuclear industry moved from defending its existence to managing its expansion. That is worth paying attention to. Because energy infrastructure, once deployed, lasts for sixty years. The choices being made today are infrastructure that will still be running in the late 2080s.
China is making those choices. India is. Russia is building. The question is whether the West will.