COLUMN // ENERGY
FLOATING NUCLEAR TO THE RESCUE
Data centers are gobbling up energy faster than the U.S. can build new power plants and transmission lines. Could nuclear reactors built on floating platforms save the day?
Written by Michael E. Webber
OVER THE PAST 20 OR SO YEARS, electricity consumption in the United States has been flat. Advances in energy efficient appliances and lighting allowed Americans to get more energy services for the same amount of electricity.
That era of flat electric demand is over. Electricity consumption in the United States is growing quickly to supply power to electric vehicles, factories, heat pumps, and hyperscaled data centers. And there’s even more future demand built into economic assumptions.
Unfortunately, while we need clean dispatchable sources of power to match varying loads, there is a significant mismatch between the pace at which new electric loads come online (two years or less), new power plants are energized (two to 10 years), and transmission lines to connect the two can be built (six years to eternity).
In an ideal world, we could add clean, dispatchable power with nuclear reactors. Nuclear is clean, fuel-dense, reliable, and features low operational costs. But the most recently completed reactors have taken decades to build, and finding suitable sites introduces significant delays. And because of those built-in delays, so few reactors are under construction at any given time that each one is an expensive, bespoke project.
Enter floating nuclear. By taking a page from the U.S. military and from other energy systems, we can build these sorts of plants at scale and get them connected to major load centers more quickly than conventional, terrestrial options.
The global long-term trend of modern energy is to start onshore then move offshore. (I’ve written about that tendency in a previous column.) For instance, oil and gas production first started on land—up to the very edge of the shore, as in early 20th century Baku and Los Angeles—but the major oil companies developed technology to pursue large, economic reservoirs in places such as the Gulf of Mexico. The first modern wind turbines were erected on inland sites, but over the past decade or so, offshore wind farms have become common in windy, shallow seas off Europe and elsewhere.
Entrepreneurs have begun to take other energy technologies—such as solar, bioenergy, and even carbon sequestration—and develop offshore implementations.
One interesting wrinkle is that the trend isn’t just a move to offshore sites, but to buoyant, floating systems that are held in place by tethers rather than planted firmly on the seabed. Floating offshore oil platforms have been in use for decades, and now floating wind turbines and so-called floatovoltaics are becoming more common.
Developed for different purposes, nuclear reactors have been at sea since the 1950s. The U.S. Navy turned to nuclear propulsion to avoid the cost, bulky storage requirements, noise and fumes of diesel fuel. They pioneered nuclear aircraft carriers and submarines and according to the World Nuclear Association, today there are well over 150 nuclear ships worldwide.
It might be time to ramp that number up significantly. But rather than using nuclear power simply to propel ships, they can be designed with excess capacity to provide shore power to nearby cities and water line ports, factories, and ocean facilities.
There are some compelling reasons for this.
A floating nuclear station moored off the coast of a major urban area could feed electricity to those cities using just a few miles of subsea cables—as opposed to the hundreds of miles of overland cables needed to reach distant traditional power plants. These floating systems would also be surrounded by cooling water, which is handy for safety and to improve operational efficiency.
Unlike a terrestrial nuclear power plant, nuclear power ships wouldn’t have to remain in one place. Those same ships could provide port power, desalinated water, or other critical humanitarian services after hurricanes, wars, or other disasters. In fact, in the wake of Super Typhoon Haiya in 2013, the U.S. Navy sent the nuclear-powered USS George Washington to the Philippines for just that purpose.
Nuclear powered ships could also break with the painstakingly slow nature of terrestrial nuclear development. Nuclear-powered naval vessels require eight years or less to construct, and once the supply chains are set up, the construction periods can shorten significantly. Using standardized designs, centralized factories can fabricate multiples of the nuclear power plants before floating them to customers, all of which might help reduce costs and timelines compared to on-site construction of typical power plants.
A nuclear power plant on the barge Akademik Lomonosov has been providing power for a Siberian town since 2020. And there are multiple companies, such as Core Power and Saltfoss, working toward commercializing the concept.
In the face of rising electricity demand and new federal restrictions on siting wind and solar farms, we need new, reliable sources of power—and fast. Perhaps nuclear power ships can swoop in and save the day.
Michael E. Webber is the Sid Richardson Chair in Public Affairs, John J. McKetta Centennial Energy Chair in Engineering, and engineering academic director of the KBH Energy Center at the University of Texas at Austin.
© 2025 The American Society of Mechanical Engineers. All rights reserved.