Hoping to revive the moribund U.S. nuclear power industry, the Department of Energy (DOE) announced this week it will help build two radically new nuclear reactors within 7 years. Funded by DOE’s new Advanced Reactor Demonstration Program, the designs include exotic features such as cooling by sodium or helium instead of water in a bid to be safer and more economical than conventional power reactors.
DOE officials “were trying to do something new and push the technology forward but also to stay within that 7-year time frame,” says Ashley Finan, a nuclear engineer and director of the National Reactor Innovation Center at Idaho National Laboratory who was not involved in the choice. “I think these two [designs] were judged to be ready for demonstration.”
DOE will split the total cost of building each plant with private industry. Each project receives $80 million this year and could receive a total of between $400 million and $4 billion in funding over the next 5 to 7 years. The agency also intends to make additional, smaller awards this year for less mature ideas, Finan says. A committee of experts whose roster has not been published selected from several designs, she says.
The two winning designs deviate fundamentally from a conventional power reactor, which is essentially a boiler. Within the core of a nuclear reactor, atoms of uranium fuel split in a chain reaction, releasing energy and free-flying neutrons, which then split other uranium atoms. In a conventional power reactor, the energy heats highly pressurized “cooling” water that circulates through the core. Still under pressure, the cooling water flows to an external steam generator, where it boils water in a separate circuit, producing steam that drives turbines to generate electricity.
Instead of water, the 345 megawatt Natrium reactor from TerraPower, Inc., and GE Hitachi would use molten sodium metal as a coolant. Because sodium has a much higher boiling temperature than water, the coolant would not have to be pressurized, reducing the plant’s complexity and cost. The sodium would transfer its heat to molten salt, which could then flow directly to a steam generator or to a storage tank, to be held to generate steam and electricity later. In contrast to a conventional nuclear power plant, the Natrium plant could quickly ratchet up or down its total output even as its reactor continues to run steadily and efficiently. That could complement renewable sources such as wind and solar energy, which produce fluctuating power levels that need to be evened out.
In contrast, the Xe-100 design from X-Energy would use pressurized helium gas to cool its uranium-based fuel. That fuel would be packaged not in the conventional metal-clad rods, but in “pebbles”—spheres of graphite infused with countless ceramic kernels that contain the uranium. Like a giant gumball machine, the reactor would hold 220,000 pebbles, which would slowly descend through the core and, as their fuel was spent, would exit from a port at the bottom. Heated to 750°C, the helium would generate steam in a secondary circuit to produce electricity. In principle, the pebbles can’t melt, eliminating the risk of a meltdown. Each Xe-100 would generate 80 megawatts, and a plant would consist of four of the modular reactors.
Both plants should be simpler and cheaper than conventional nuclear power plants. Because Natrium sodium coolant is unpressurized, the reactor requires a smaller containment structure than a conventional reactor. The plant also “decouples” the reactor and the electricity generating portions of the facility, which sit on opposite sides of the storage tanks. Those features should allow engineers to reduce use of expensive reinforced concrete by 80%, says Tara Neider, a TerraPower engineer and project director for the Natrium design. “Natrium is all about making a nuclear plant simpler so it can be more efficient,” she says. Both companies say they have yet to choose sites for their reactors.
Both reactors would also depart from conventional designs in using a fuel that is more highly enriched in uranium-235, the fissile isotope that is key to generating a chain reaction. To minimize the risk that the fuel, fresh or spent, could be diverted to create a nuclear weapon, water-cooled power reactors run on fuel that it is 3.5% uranium-235. The Natrium and Xe-100 reactors would use fuel enriched to 20%, which would enable them to run longer on a batch of fuel and extract more energy from it. Such fuel isn’t currently produced in the United States, but current manufacturers could make it relatively easily, Finan says. The fuel would also be difficult to divert to weapons, she says, in part because it would require fewer refueling stops.
As in many things nuclear, what’s old is new: Since the birth of the nuclear age in the 1950s, engineers have built a handful of sodium-cooled reactors and even a couple of pebble-bed reactors. But the devil is in the design details, and both TerraPower and X-Energy aim to make reactors that are safe and can compete with cheaper forms of power. Ultimately, TerraPower hopes to market a Natrium plant for less than $1 billion, Neider says.
This year, DOE plans to make between two and five more awards totaling up to $30 million for advanced reactor designs that are even fresher, Finan says. “DOE wants to provide a pathway forward for these technologies that could well be game changers but are not quite ready for demonstration,” she says. Each “risk reduction” award winner would be eligible for a total of up to $400 million total over 7 years.
For U.S. nuclear engineers, the prospect of building new advanced reactors is thrilling, after more than 2 decades during which the United States commissioned just one new power reactor. “This is what we’ve been working for all along,” Neider says. “It’s exciting times.” But that bright future will depend on continued funding by Congress and support from what might be a new presidential administration. And, ultimately, developers have to prove that nuclear can compete economically with other forms of power.
*Correction, 19 October, 1:15 p.m.: This story has been updated to clarify that the "risk reduction" awards this year will total $30 million, instead of giving that amount to each winning project.