The Army Tapped Five Vendors for $2.2 Billion in Nuclear Microreactors. It Explicitly Expects Most to Miss the 2028 Deadline.
Project Janus aims to shield major power-projection bases from domestic grid failure using commercial microreactors, but officials warned that private capital must carry the bulk of the cost.
The Pentagon's largest military bases run on civilian electricity, backed up by storage yards of diesel fuel that would last days or weeks in a major conflict. On August 26, 2026, the U.S. Army and the Defense Innovation Unit moved to replace that vulnerability, selecting five companies under Project Janus to build, own, and operate commercial nuclear microreactors across five major power-projection installations.
The effort draws from a shared pool of up to $2.2 billion in fixed-price, milestone-based Other Transaction Authority agreements extending through fiscal year 2031. Modern power projection demands massive, continuous electricity for communications, radar arrays, counter-drone defenses, and directed-energy weapons. If the domestic electric grid fails under cyberattack or physical sabotage, diesel supply chains cannot guarantee sustained operations. Federal reporting on the awards framed Janus as an attempt to establish permanent, resilient base power through commercial partnerships rather than government-owned utility plants.
Five vendors, five military bases
The selections pair five distinct reactor concepts with five critical Army commands. Antares Nuclear, Inc. was tapped for Fort Bragg, North Carolina, home of the 82nd Airborne Division, where it will deploy its R1 sodium heat-pipe microreactors producing between 100 kWe and 1 MWe. BWXT Advanced Technologies, a business unit of BWX Technologies, was selected for Fort Campbell, Kentucky, to deploy a 20-MWe version of its BANR high-temperature gas-cooled reactor using uranium oxycarbide TRISO fuel. General Atomics Electromagnetic Systems was assigned to Fort Hood, Texas, to advance its GA-TES liquid-metal-cooled reactor, which scales from 5 MWe to 20 MWe using encapsulated uranium-zirconium hydride fuel.
Radiant Industries, Inc. secured up to $750 million to deliver 15 factory-built, 1-MWe Kaleidos helium-cooled TRISO microreactors by 2030, beginning with an initial three-unit deployment at Fort Benning, Georgia. Westinghouse Government Services, a unit of Westinghouse Electric Company, will field its eVinci heat-pipe microreactor at Fort Drum, New York. Technical analysis of the program notes that the five vendors collectively plan to deploy more than 20 commercial microreactors across the military footprint.
A COTS-style commercial risk model
Rather than purchasing reactors outright, the Army is adapting the contracting model NASA used for commercial spaceflight under the Commercial Orbital Transportation Services program. The $2.2 billion allocation acts as a pool of milestone subsidies covering development, construction, and the initial year of operation. After the first year, installations will transition to commercial power purchase agreements. The vendors will own, build, and operate the reactors themselves.
Crucially, the Army expects private investment to fund the majority of total program costs. The agreements tie disbursements strictly to demonstrable hardware and testing milestones. If a contractor falters or falls behind schedule, the Army holds total flexibility to cancel future milestones with zero termination fees and redirect those funds to competing vendors or new entrants.
The 2028 mandate and cost reality
The program responds to May 2025 presidential executive orders that designated the Army as executive agent for military nuclear energy, setting a firm deadline to have at least one Army-regulated reactor operating on a domestic base by September 30, 2028. Army energy leadership was blunt about the odds. Dr. Jeff Waksman, principal deputy assistant secretary of the Army for installations, energy and environment, told reporters that the service does not expect all five companies to turn on a reactor in 2028, adding that they definitely will not. The multi-vendor portfolio is designed so that even if several companies slip, at least one reactor reaches sustained operation on time.
Cost visibility remains another open question. Waksman acknowledged that nobody knows what commercial microreactors will actually cost until multiple units are manufactured and deployed. Early units will carry high first-of-a-kind expenses, and the Army does not expect microreactors to compete with wholesale natural gas prices, viewing the premium as an acceptable price for operational energy resilience.
Defense licensing and fuel supply bottlenecks
Project Janus bypasses the Nuclear Regulatory Commission by utilizing defense licensing authorities under the Atomic Energy Act. The newly established Army Reactor Regulatory Office will oversee safety authorizations, working to align standards so vendors have a straight line toward eventual civilian NRC licensing. Environmental reviews under the National Environmental Policy Act remain mandatory, and spent fuel must be removed from military installations within two years of reactor shutdown, with the Department of Energy taking title to the waste.
The largest external constraint facing the program is fuel availability. Four of the five designs rely on High-Assay Low-Enriched Uranium, or HALEU, enriched up to 19.75 percent and packaged into TRISO particles. Domestic HALEU enrichment capacity is severely constrained, requiring close coordination with the Department of Energy and the National Nuclear Security Administration to allocate stockpiles for initial prototypes. Industry coverage emphasized that while the Army is funding initial momentum, long-term commercialization will depend on whether fuel supply chains and private capital markets can sustain the industry once government subsidies step back.
