Nuclear energy is entering its biggest expansion in decades. After years of stagnation, global nuclear capacity is projected to grow between 50% and 150% by 2050, driven by climate targets, energy security concerns, and a new generation of reactor designs that look very different from the massive plants built in the 20th century. The question is no longer whether nuclear has a future, but how large that future will be.
How Much Nuclear Capacity Is Expected by 2050
The International Atomic Energy Agency raised its nuclear power projections for the fifth consecutive year in 2025. In the conservative scenario, global nuclear generating capacity rises 50% from current levels to 561 gigawatts by 2050. The optimistic scenario nearly triples that growth, reaching 992 gigawatts. For context, the world currently operates roughly 370 gigawatts of nuclear capacity, so even the low estimate represents a substantial buildout.
The International Energy Agency’s net-zero modeling puts nuclear’s role in concrete terms: reaching global climate goals requires nuclear power to exceed 5,000 terawatt-hours of annual generation by 2045. That’s roughly double what nuclear plants produce today. Meeting that target means not just extending the life of existing reactors but building hundreds of new ones across multiple continents.
Small Modular Reactors Are Closest to Market
The most tangible shift in nuclear’s near future is the small modular reactor, or SMR. These are factory-built units typically producing 300 megawatts or less, compared to the 1,000+ megawatt output of conventional plants. Their smaller size means lower upfront costs, shorter construction timelines, and the ability to site them in locations where a full-scale plant wouldn’t be practical.
Several SMR projects in the United States are actively moving forward. The Tennessee Valley Authority is advancing deployment of a 300-megawatt reactor designed by GE Vernova Hitachi at the Clinch River site in Tennessee, with plans to accelerate additional units with other utility partners. Holtec International plans to deploy two 300-megawatt SMRs at the Palisades site in Michigan, a former nuclear plant location that already has grid connections and community familiarity with nuclear operations.
These first deployments matter because they’ll establish whether SMRs can actually be built on time and on budget. The nuclear industry’s biggest credibility problem has been cost overruns and delays at conventional plants. If SMRs deliver on their promise of modular, predictable construction, they could unlock orders from utilities that have avoided nuclear for decades.
Micro-Reactors for Remote and Industrial Use
Even smaller than SMRs, micro-reactors are designed to go where the grid doesn’t reach. Westinghouse’s eVinci, one of the leading designs, is a 15-megawatt transportable unit that functions essentially as a nuclear battery. It runs for more than eight years before refueling, uses passive heat pipes instead of pumps and coolant loops, and can be shipped to a site rather than built on one.
The intended applications read like a list of places that currently depend on diesel generators: remote mining operations, military installations, disaster relief zones, island communities, and Arctic settlements. Beyond electricity, micro-reactors can produce industrial heat for hydrogen generation, district heating, and manufacturing processes that require high temperatures. This is a genuinely new market for nuclear energy, one where the competition isn’t solar or wind but fossil fuels burned in isolated locations with no clean alternative.
Advanced Reactor Designs and Their Fuel Challenge
Beyond SMRs, a broader class of next-generation reactors is in development. These include designs that use molten salt instead of water as a coolant, high-temperature gas reactors that can withstand extreme heat, and reactors with fuel encased in tiny ceramic shells called TRISO particles that are physically incapable of melting down. The U.S. Department of Energy is funding pilot fuel production lines for several of these concepts, including Terrestrial Energy’s molten salt fuel fabrication process and Valar Atomics’ TRISO fuel line for its high-temperature reactor.
Most of these advanced designs share one bottleneck: they need a specialized fuel called high-assay low-enriched uranium, or HALEU. This fuel is enriched to just under 20%, compared to the 3-5% enrichment used in today’s reactors. As of mid-2025, there is exactly one facility in the Western world licensed to produce it: the Centrus Energy plant in Piketon, Ohio, which reached the 900-kilogram production mark and received a contract extension to produce another 900 kilograms over the following year. One metric ton is a start, but scaling up a commercial fuel supply chain for dozens of advanced reactors will require significantly more capacity.
Nuclear Waste Finally Has a Solution
The question “what do you do with the waste?” has dogged nuclear energy for its entire existence. Finland is providing the first real answer. On Olkiluoto Island, a facility called ONKALO has been carved 430 meters into ancient bedrock, creating a network of tunnels that will eventually stretch 50 kilometers. It is the world’s first deep geological repository designed to permanently store spent nuclear fuel.
The concept is straightforward: seal spent fuel in copper and steel canisters, embed them in clay, and place them deep in stable rock formations that haven’t shifted in nearly two billion years. Finland plans to store 6,500 tons of spent fuel in approximately 3,250 canisters, with the repository designed to remain safe for at least 100,000 years. Construction of the underground facility began in 2004, the government granted a construction license in 2015, and Posiva, the company operating it, is now in the process of obtaining its final operating license.
Finland’s success matters far beyond its borders. Sweden is pursuing a similar repository, and countries like Canada and France are developing their own deep geological storage programs. The existence of a working model dismantles one of the longest-standing arguments against nuclear power.
Uranium Supply and How Long It Lasts
The IAEA’s most recent assessment identified roughly eight million tons of known uranium resources globally as of 2023. At current consumption rates, those reserves could be depleted by around 2080, giving the world roughly six decades of fuel at today’s usage levels. That timeline sounds tight, but it comes with important caveats.
Exploration tends to follow demand. When uranium prices rise, mining companies invest in finding new deposits, and historically, identified reserves have grown over time. Advanced reactor designs are also significantly more fuel-efficient than current plants, extracting more energy from the same amount of uranium. Some next-generation concepts can even use spent fuel from existing reactors as feedstock, which would effectively turn today’s nuclear waste into tomorrow’s fuel supply. Seawater extraction, while currently expensive, contains enough dissolved uranium to power reactors for thousands of years if the technology becomes economical.
Where Fusion Fits In
Nuclear fusion, the process that powers the sun, remains the long-term wildcard. The ITER project in southern France is the world’s largest fusion experiment, designed to demonstrate that a fusion reaction can produce more energy than it consumes. ITER’s timeline has slipped repeatedly since its original schedule, which targeted first plasma by December 2025 and full deuterium-tritium operations by 2035. The project has since announced further delays and cost increases, and a revised schedule is expected.
Dozens of private fusion companies are pursuing smaller, faster approaches. Several have achieved significant plasma milestones, and a few claim they’ll have demonstration plants running in the early 2030s. Even optimistic projections, though, don’t put commercial fusion electricity on the grid before the 2040s at the earliest. Fusion will not solve near-term climate or energy challenges. Its role, if it works, is as a virtually limitless energy source for the second half of the century and beyond.
What’s Driving the Momentum
Three forces are converging to push nuclear forward in ways that weren’t true even five years ago. First, the electricity demand from data centers, artificial intelligence, and electric vehicles is growing so fast that renewables alone can’t keep pace, especially when the grid needs power around the clock regardless of weather. Several major tech companies have signed direct power agreements with nuclear developers for exactly this reason.
Second, energy security has become a top political priority. Countries that relied heavily on imported natural gas learned painful lessons from supply disruptions, and nuclear offers a domestic or allied-nation fuel source with years of fuel stored on-site. Third, the climate math is unforgiving. Reaching net-zero emissions by mid-century while simultaneously growing global electricity demand requires every large-scale clean energy source available. Nuclear is the only proven technology that produces massive amounts of carbon-free electricity on demand, day and night, in any climate.
The obstacles are real: high construction costs, long permitting timelines, public skepticism in some regions, and the need to build a workforce and supply chain that has atrophied in countries that stopped building reactors. But the direction of policy, investment, and technology development all point the same way. Nuclear energy’s future is larger than its present, likely by a wide margin.