The Army is moving ahead with microreactors at five military installations, aiming to prove that small nuclear plants can keep critical missions powered when the grid fails. The project is framed as military resilience, but it also gives Washington a high stakes proving ground for the next phase of U.S. nuclear power.
WASHINGTON | Filed at 7:42 p.m. Eastern time
The U.S. Army is preparing to put nuclear reactors back on American military bases.
Not the towering cooling stack image many Americans associate with nuclear power. Not the massive civilian plants that take decades to license, finance and build. The Army’s new bet is smaller, faster and more strategic: nuclear microreactors that can generate electricity for critical missions even if the commercial grid goes dark.
On Wednesday, the Army announced that it will award up to $2.2 billion over five years to five companies to own, build and operate microreactors at five military installations. The selected pairings are Antares Nuclear at Fort Bragg in North Carolina, BWXT Advanced Technologies at Fort Campbell in Kentucky, General Atomics Electromagnetic Systems at Fort Hood in Texas, Radiant Industries at Fort Benning in Georgia and Westinghouse Government Services at Fort Drum in New York.
The program is called Janus, a name taken from the Roman god of transitions. It is an unusually fitting choice. The Army is trying to move nuclear energy from the world of promising prototypes into the world of real power delivery. If the project works, it could reshape how the military thinks about base resilience, battlefield logistics and remote energy. It could also help jump start a commercial microreactor industry that has promised much but has not yet placed a working reactor on the U.S. commercial grid.
If it fails, it will become another reminder that advanced nuclear power is easier to celebrate in policy speeches than to deliver safely, affordably and on schedule. That is the tension at the center of the Army’s announcement.
The military case is direct. American bases depend heavily on the civilian electric grid. That grid is vulnerable to storms, cyberattacks, physical sabotage, heat waves, fuel disruptions and war related pressure. Military operations increasingly require reliable electricity for command centers, communications, radars, computer systems, drone operations, batteries, housing, training ranges, weapons support, intelligence systems and data heavy operations. Diesel generators can help, but they depend on fuel supply chains that may be difficult to maintain in a crisis.
Army leaders are now arguing that nuclear microreactors can provide a different kind of backup: continuous power that can run for years without refueling.
Army Secretary Dan Driscoll framed the awards as a step toward reliable baseload power directly on installations, saying the service needs energy resilience to project combat power without depending on vulnerable external grids. Jeff Waksman, the Army’s principal deputy assistant secretary for installations, energy and environment, put the issue even more plainly in a briefing with reporters. The domestic grid, he said, is potentially at risk in conflict, and the Army cannot assume it will always be able to move fossil fuels where they are needed. That makes nuclear energy attractive to military planners.
But the project is not only about keeping lights on during outages. It is part of a much broader U.S. nuclear revival pushed by the Trump administration, which signed executive orders in 2025 to speed advanced reactor deployment, expand the nuclear industrial base and use the military as an early customer. One order directed the Secretary of Defense, acting through the Secretary of the Army, to begin operating an Army regulated nuclear reactor at a domestic military installation no later than Sept. 30, 2028. The Army’s Janus awards are the clearest sign yet that deadline is becoming a procurement reality.
The service does not expect every company to succeed. Officials openly acknowledge that some designs may fall behind, fail technical milestones or prove too difficult to field by 2028. That is why the Army selected five vendors. It is spreading risk across competing approaches, with payment tied to technical milestones rather than simply writing checks upfront. The Army expects more than 20 total microreactors to be built and operated across Defense Department installations if the program scales.
That target is ambitious
Microreactors are generally described as small, factory built nuclear plants that can be transported and installed closer to the customer than conventional reactors. The Army said each unit could provide between 1 megawatt and 20 megawatts of power, depending on the company’s design. Some installations may receive multiple reactors. Antares and Radiant are planning three pack deployments, according to Army officials. Fort Drum’s public announcement said the systems would generate under 20 megawatts of electricity and occupy less than five acres.
Those numbers matter because major Army posts consume enormous amounts of power. A microreactor will not power an entire large installation by itself. Army officials have made that clear. Bases will remain connected to the commercial grid, and the reactors are intended to support critical infrastructure rather than replace every outside electricity source. That makes Janus a resilience project, not a full energy independence project.
The difference matters for public understanding. These reactors are not being sold as miniature versions of civilian nuclear plants that will run everything on a base. They are being positioned as hardened, localized power sources for essential operations. In a crisis, they could keep command functions, communications, medical systems, data centers, security systems and mission critical infrastructure operating while other parts of a base draw from the grid or backup generators. Supporters say that is exactly the kind of use case microreactors need.
Commercial nuclear developers have struggled to prove that small reactors can compete economically with gas, renewables, batteries and large scale nuclear plants. Military bases, especially remote or strategically sensitive ones, may value resilience enough to pay more per kilowatt hour. Reuters reported that the Army currently pays about 40 cents per kilowatt hour to power Arctic installations and hopes microreactors could eventually cut those costs. Radiant told Reuters that its containerized reactors could deliver power between 20 and 30 cents per kilowatt hour by avoiding transmission fees.
Even if those numbers hold, microreactors will not be cheap. Waksman acknowledged that nobody thinks microreactors will be less expensive than conventional nuclear plants. The Army’s goal is to use Janus to help drive costs down by giving companies real projects, real customers and real deployment pressure. That is why the program is being watched outside the Pentagon.
Data centers are consuming massive amounts of electricity as artificial intelligence and cloud computing expand. Industrial sites are searching for reliable clean power. Remote communities, mining operations and islands are studying alternatives to diesel. If the Army can prove that microreactors can be built, regulated, operated and maintained safely, the commercial market could follow. If the Army cannot make the model work with federal support and mission urgency, private customers may hesitate. Janus is therefore both a defense program and a market signal.
The regulatory structure will be one of the most closely scrutinized parts. The reactors will be regulated by the Army rather than the U.S. Nuclear Regulatory Commission, which oversees commercial nuclear reactors. Army officials say they are trying to align their processes with NRC expectations so that companies will not need major design changes if they later seek commercial licensing. That alignment is crucial. A military shortcut that cannot translate into civilian approval would limit the program’s broader value. Critics will focus on safety, waste, cost and transparency.
Small reactors may be physically smaller than conventional plants, but they still involve nuclear fuel, radioactive materials, emergency planning, security protocols and long term accountability. Communities near military bases will want to know where reactors will be placed, how accidents would be handled, who responds to emergencies, what evacuation planning exists, how spent fuel is managed and what happens if a contractor fails financially or technically.
Army officials say the reactors are designed to shut down safely in case of failure and that no long term spent fuel storage will occur on installations. The Army is working with the Energy Department on arrangements to remove radioactive waste, and Fort Drum’s public release said used fuel and activated materials would remain on site only for a brief federally mandated cooldown period before transport to Department of Energy or national laboratory facilities for long term storage or recycling. Those assurances are important. They will need details.
Nuclear projects succeed not only through engineering, but through trust. Military bases often have surrounding communities that depend on them economically but also expect to be informed about risks. A reactor at Fort Drum is not only an Army matter. It affects the North Country in New York. A reactor at Fort Bragg is not only a procurement milestone. It affects North Carolina communities that live beside one of the largest military installations in the country. The Army will have to communicate early, clearly and repeatedly.
The politics of nuclear power have changed in recent years. Concerns over climate change, energy security, grid reliability and data center demand have made nuclear power more attractive to some policymakers who once focused primarily on risk. At the same time, long construction timelines, cost overruns, waste storage debates and accident fears still shape public opinion. Microreactors enter that debate with a promise: smaller systems, factory manufacturing, faster deployment, passive safety and reliable carbon free power. Promise is not proof.
The United States has seen advanced nuclear enthusiasm before. Many designs remain precommercial. Fuel supply chains are not fully mature. High assay low enriched uranium, a fuel needed by many advanced reactors, has been an area of national concern because domestic supply remains limited. Skilled nuclear labor is also a constraint. Manufacturing capacity, quality assurance, security rules and transport regulations all add complexity. The Army is attempting to compress that complexity into a 2028 operational goal.
That timeline is aggressive, which is both the point and the risk. Supporters argue that the nuclear industry needs a demanding customer to move faster. Critics argue that nuclear projects are not the place for deadline driven optimism. The program will test whether military urgency can accelerate nuclear deployment without weakening safety culture.
There is also a strategic question about contractor ownership. The Army said the reactors will be contractor owned and operated. That could reduce some burdens on the service and let specialized companies handle operation. It also raises accountability questions. If a reactor sits on a military base but belongs to a private company, who is responsible for cost overruns, outages, security failures, workforce training, fuel management and decommissioning? Contracts can answer those questions legally. Public trust requires those answers to be understandable. The base selections themselves reveal the Army’s priorities.
Fort Bragg is a major power projection and special operations hub. Fort Campbell supports highly deployable forces. Fort Hood is one of the Army’s largest armored force centers. Fort Benning is a central training and maneuver institution. Fort Drum, home of the 10th Mountain Division, carries a cold weather and rapid deployment identity that fits the resilience narrative. These are not symbolic locations alone. They are installations where energy interruptions could affect missions with national security implications.
The program’s long range logic extends even farther. Army officials have spoken about possible future use at remote installations in Alaska or on islands. Those are places where diesel logistics are expensive, exposed and strategically fragile. If microreactors work on domestic bases, the military may eventually look at more expeditionary or remote applications. That would be a profound shift in military energy planning. It would also intensify debate.
Moving nuclear technology closer to operational military settings raises questions about battlefield vulnerability, enemy targeting, transport security and international perception. A fixed reactor on a domestic base is one challenge. A more mobile or remote reactor concept is another. Project Pele, the Pentagon’s earlier mobile microreactor effort, helped demonstrate feasibility, but Janus now appears focused on installations and sustained power rather than brief demonstration. That difference makes Janus more practical and more consequential. The Army is not trying to win a science fair. It is trying to buy power.
For the nuclear industry, that is the opportunity. For taxpayers, that is the accountability test. Up to $2.2 billion in federal awards, plus expected private capital, will flow into companies that still must prove their designs can operate reliably. The Army says it can pull future milestones if companies fail to perform and that there are no termination fees. That gives the government leverage, but it does not eliminate risk. Failed prototypes can still consume money, time and political capital.
The broader national question is whether this is the right way to ramp up U.S. nuclear power.
One argument says yes. The military has historically helped mature technologies that later transformed civilian markets. The internet, GPS, jet engines and commercial space all benefited in different ways from defense linked demand. If advanced nuclear needs a first customer with high reliability needs and tolerance for early costs, the Army may be that customer.
The counterargument says nuclear power should not be rushed through military channels to avoid the slower scrutiny applied to civilian energy infrastructure. Communities, environmental groups and nuclear safety advocates will want assurance that Army regulation does not become weaker regulation. The credibility of the whole effort may depend on whether the Army can show that faster does not mean looser. That is the story now unfolding.
Five bases. Five companies. A 2028 deadline. A promise of resilient power. A test of America’s nuclear ambitions at the same time AI, data centers, electrification and national security are straining the grid.
The Army is calling Janus a transition. That is accurate. It is a transition from diesel dependence toward localized nuclear backup. It is a transition from paper reactor designs toward operating machines. It is a transition from nuclear power as climate and energy policy toward nuclear power as military infrastructure. It may also become a transition in how Americans think about nuclear energy itself.
If the reactors operate safely, cut reliance on vulnerable fuel supply chains and prove useful at military bases, the program could accelerate confidence in microreactors for civilian uses. If they stumble, the setback will feed old doubts about nuclear cost, complexity and risk.
For now, the Army is making the first big move. The service is betting that the next era of American energy will not be built only around power plants far away from the mission. It will be built closer to the places that cannot afford to lose electricity at all.
That is the wager behind the $2.2 billion announcement. Not just that microreactors can work. That the future of national security may depend on them.
Reporting and sourcing transparency note: This article is based on current public reporting from the Associated Press, Reuters, Defense News, Breaking Defense, Axios, official Army materials, DVIDS, White House executive order materials and related Energy Department and Army microreactor program reporting.
Energy, defense and public safety information note: This article is for news and public information only. Nuclear licensing, environmental review, contract milestones, reactor schedules, fuel arrangements and base selections may change as the Janus Program develops.