Commercial nuclear reactors today largely share a common blueprint: atoms of a radioactive material like uranium split and release neutrons, which sustain a chain reaction that gives off heat. That heat is typically used to turn water into steam, which spins a turbine to generate electricity. Most existing plants use the same fuel and coolant, rely on water, and are built at a massive scale, with each facility effectively custom designed for its specific site. While these giants have supplied power grids around the world for decades and have recently seen renewed interest as concerns about climate change and energy security rise, they are expensive and slow to construct, which limits how quickly nuclear power can expand.
A new generation of nuclear power technologies aims to reinvent both the physical form and operating logic of reactors, with advocates arguing that such innovations could revitalize the industry and help displace fossil fuels without emitting greenhouse gases. Demand for electricity is climbing as hotter temperatures and growing economies bring more air conditioners online, industry pursues modernization to cut climate pollution, and the boom in artificial intelligence drives construction of more power-hungry data centers. Against this backdrop, nuclear can play a larger role only if the next wave of plants proves to be safe, reliable, affordable, and quick to deploy. One major concept is the small modular reactor, or SMR, which is designed to be factory-built and assembled on site in a more standardized fashion than today’s bespoke plants.
Small modular reactors work like traditional gigawatt-scale reactors but at a fraction of the size and power output, with reactor cores that can be just two meters tall, making them easier to install and scale by adding multiple units at a site. This smaller footprint could unlock new applications, from supplying power to military bases and remote locations to providing industrial heat in sectors such as chemical manufacturing, with projects under way from companies including BWXT and X-energy. Early SMR deployments include operational plants in China and Russia and the Linglong One demonstration project in China, which is being built alongside two large existing reactors and is expected to come online by the end of the year, while in the US, Kairos Power has regulatory approval to build its Hermes 2 demonstration reactor, which should be operating by 2030. A key unknown is how much standardization will actually cut costs, since even modular reactors must be adapted for site-specific conditions like earthquakes, floods, and hurricanes, which can add significant planning and construction expenses.
Reactor developers are also rethinking fuel. The critical parameter is the proportion of uranium-235, the isotope that sustains a chain reaction, which makes up about 0.7% of naturally occurring uranium and must be enriched for reactor use. Weapons material is enriched to uranium-235 concentrations over 90%, while existing reactors typically run on fuel enriched between 3% and 5% uranium-235. New designs may turn to high-assay low-enriched uranium, or HALEU, which ranges from 5% to 20% uranium-235 and can sustain a chain reaction for much longer before refueling, especially as enrichment levels rise. HALEU also enables alternative fuel architectures such as tri-structural isotropic, or TRISO, fuel, which uses uranium kernels less than a millimeter across that are coated in layers of carbon and ceramic and then embedded in graphite pellets, creating a built-in containment system that can withstand neutron irradiation, corrosion, oxidation, and temperatures over 3,200 °F (1,800 °C) while allowing heat to flow out to the coolant.
Cooling systems are undergoing similar reimagining. In most current reactors, water serves as a coolant and must be kept under very high pressures so it remains liquid as it circulates, which requires thick, reinforced containment structures and careful monitoring to prevent leaks that could lead to a meltdown. New reactor concepts use alternative coolants such as gas, liquid metal, or molten salt that can operate at much higher temperatures, with loops reaching upwards of 500 °C compared with a maximum of around 300 °C in water-cooled designs, improving the efficiency of steam generation and heat transfer. Because metal and salt coolants remain liquid at high temperatures at pressures closer to one atmosphere, they can eliminate the need for the high-pressure systems that water requires, potentially simplifying some safety and engineering challenges even as they introduce others, such as corrosion risks with molten salt in the presence of oxygen or the need to prevent sodium metal from contacting water. Companies like Kairos Power are testing these ideas with molten-salt-cooled reactors, including a 50-megawatt unit slated to come online in 2030 that will sell power to the Tennessee Valley Authority. Across all these experiments in size, fuel, and cooling, the long-term test will be whether such reactors can not only generate electricity effectively but also operate safely and economically for decades.
