Washington, June 21, 2026, 18:04 EDT
Key takeaways
- Valar Atomics’ Ward 250 hit zero-power criticality, the second advanced U.S. reactor in 14 days to do so.
- The Army and Defense Innovation Unit are putting up more than $2 billion over five years to drive commercial work on military microreactors.
- The next hurdles: a third criticality by July 4, then electricity-producing tests set for 2027, and plans for a military reactor in operation by September 2028.
Two private advanced reactor projects hit zero-power criticality in under 14 days, a step that shows the military’s microreactor push is getting results. The Department of Energy said Valar Atomics’ Ward 250 reactor reached the threshold in Utah on June 18, just two weeks after Antares Nuclear did the same with its Mark-0 reactor at Idaho National Laboratory on June 4. With this, the military’s pilot program is now two out of three on its goal for July 4. Washington’s bigger bet is commercial—Army and Defense Innovation Unit have put up more than $2 billion over five years to bring these first-of-a-kind reactors to market for bases, private industry and far-off sites.
Ward 250 is the first DOE-authorized reactor to be built outside a national lab. Four months ago, its parts and unfueled core were shipped nearly 700 miles from California to Utah on a military C-17, testing if a power system like this can move through the defense logistics network already in place. Valar says the reactor could one day deliver up to 5 megawatts. “Nine months ago, this was an empty site,” CEO Isaiah Taylor said after the criticality test. The Department of Energy’s Energy.gov
Information gain: The DOE program is now 67% of the way to its criticality goal, with 13 days left as of June 21. That doesn’t mean it’s commercially ready, but two separate companies have done fueled tests at two sites. Timing is key here, since the administration has told the military to start running a domestic advanced reactor by September 30, 2028.
The spending points to how big the planned push is from demo to real-world use. The Army named nine potential sites for Janus Program reactors. If the $2 billion minimum pledge is split evenly, that’s more than $222 million per site. That’s not an official Army budget for each site, but gives a sense of how much cash might go to cover first-unit costs and unknowns.
Janus is set up with commercial companies running the reactors, with milestone payments and contracts based on NASA’s Commercial Orbital Transportation Services model. The military would act as the main buyer instead of building all the reactors itself. Suppliers would get key experience, certified parts and the chance for follow-on orders, seen as steps from high-cost one-offs to standard builds. Kenneth Luongo at The National Interest likened it to the Navy’s 1950s push on pressurized-water reactor tech.
Pentagon officials say they’re taking more than one path. Project Pele, a mobile 1-to-5-megawatt reactor prototype, is designed to fit into four 20-foot containers, move by truck or C-17, and use TRISO particle fuel that can handle high heat. Formal testing could start as early as 2027. Janus targets commercial reactors of about 1 to 20 megawatts for Army bases in the U.S. “Next, we need a microreactor to generate electrons,” Army nuclear chief Jeff Waksman said after the Antares test. The Department of Energy’s Energy.gov
Air Force is building up a separate customer base with its power-purchase deal at Eielson Air Force Base, where the government buys electricity and heat but doesn’t own the plant. Another Defense Department program matched Antares with Joint Base San Antonio, Radiant Industries with Buckley Space Force Base, and Westinghouse with Malmstrom Air Force Base for possible deployment. The approach gives the military several developer options and lets it see different designs in action at real sites, instead of relying on one company.
Debate is shifting from climate talks to energy as critical infrastructure. In a June 20 commentary for OilPrice, supply-chain consultant James Durso said dependable power for AI, chip fabs, weapons manufacturing and command systems should be handled like strategic assets. “Energy dominance is no longer simply economic policy,” Durso said. He pointed to behind-the-meter reactors—set up next to where the electricity is used—as a way to keep some operations running without full dependence on long transmission lines. Oilprice
The debate is spreading outside the U.S. Gulf Daily News ran an analysis looking at whether SMRs could become a new energy edge for the Gulf. In that region, the pitch isn’t just electricity—these small reactors might provide heat for desalination and industrial use along with steady electricity, a use the International Atomic Energy Agency has pointed to before. So the export opportunity is bundled energy-and-water infrastructure, not just a shrunken nuclear power plant.
The military’s strongest impact may end up being simple demand, not advances in reactor technology. Civil nuclear buyers want fixed prices, a license, and working plants they can trust before they sign up. Defense sites look for black-start capability, grid independence, and lower diesel use. In practice, the Army appears to be buying the track record and production runs that private customers will later expect—a demand-side play for nuclear. That reading comes from the Janus contract setup and the NRC’s push on faster microreactor licensing.
There’s a big catch. Zero-power criticality isn’t the same as selling power, the Mark-0 doesn’t have power-conversion gear, and licensing, fuel, cost, security, and waste are all still up in the air for commercial rollout. Edwin Lyman, nuclear-safety director at the Union of Concerned Scientists, told AP the Antares test had “absolutely no bearing” on commercial viability or safety. Backers point to TRISO fuel, passive design, staged oversight from the feds. The next hurdles are a third reactor by July 4, power tests in 2027, and a military unit up and running by September 30, 2028. AP News
