How Four Advanced US Microreactors Achieved Criticality in a Policy-Driven Summer Sprint

For decades, the American nuclear industry has been synonymous with staggering scale and paralyzing delays. Mega-projects required billions of dollars and years of construction, cementing a reputation that nuclear power is too slow and expensive to meet urgent energy demands. The sheer size of legacy reactors meant that every installation was a bespoke engineering challenge, highly vulnerable to supply chain disruptions and regulatory hurdles.[7]

But a quiet paradigm shift has just materialized at the opposite end of the size spectrum. Over the summer of 2026, four advanced microreactors achieved "criticality"—the point at which a nuclear chain reaction becomes self-sustaining. This milestone ends a 40-year drought for novel, non-light-water reactor designs in the United States, signaling a pivot from massive construction projects to nimble, factory-built hardware.[1][7]

The rapid succession of these tests was not coincidental. It was driven by the Department of Energy's Reactor Pilot Program, initiated by a May 2025 executive order that mandated at least three advanced test reactors reach criticality by July 4, 2026. By setting a hard deadline, the policy forced regulatory and engineering frameworks to align, compressing timelines that historically stretched over decades into a matter of months.[1]

The first to cross the line was Antares Nuclear, whose Mark-0 demonstrator went critical on June 4 at the Idaho National Laboratory. Valar Atomics followed two weeks later with its Ward 250 unit at the Utah San Rafael Energy Lab, and the Los Alamos-designed ZiaCore system achieved the same status in Nevada by August. Each test proved that the underlying physics of these miniaturized systems functioned exactly as modeled.[1][7]

In reactor physics, criticality is the fundamental threshold of operation. It occurs when each fission event releases enough neutrons to trigger exactly one additional fission, creating a stable, ongoing chain reaction without growing out of control. Achieving this state is the definitive proof that a reactor core can sustain the nuclear processes required to generate continuous heat.[2]

These summer demonstrations were "zero-power" criticality tests. They operated at high temperatures to validate the core design and safety parameters, but they did not yet convert that thermal energy into electricity. Generating actual electrons for the grid is the next phase of the program, with developers aiming to produce usable power by 2027.[1]

Unlike traditional gigawatt-scale plants that rely on vast volumes of water for cooling, these microreactors utilize entirely different architectures. They are designed to produce between 1 and 50 megawatts of power and are small enough to fit inside standard shipping containers. This modularity allows them to be fully assembled in a factory, shipped via truck or rail, and deployed in a matter of days.[2][3]

The safety profile of these systems relies heavily on Tri-structural Isotropic, or TRISO, fuel. Instead of traditional metallic fuel rods that can melt under extreme heat, TRISO encapsulates tiny uranium kernels within multiple layers of porous carbon, pyrolytic carbon, and silicon carbide.[4]

These ceramic layers act as individual, microscopic containment vessels. They are capable of withstanding temperatures up to 1,600 degrees Celsius without melting or releasing radioactive byproducts. This inherent structural integrity eliminates the need for the massive concrete containment domes that define legacy nuclear plants, drastically reducing both the physical footprint and the construction cost.[4]

Heat removal is also fundamentally altered in these advanced designs. Systems like the Westinghouse eVinci and the Antares Mark-0 utilize passive heat pipes, while others employ inert helium gas or liquid sodium. These coolants operate efficiently at high temperatures without the high-pressure environments required by traditional water-cooled reactors.[1][5]

Because these coolants do not boil off or require active mechanical pumping in the same way water does, they significantly reduce the risk of pressure-related accidents. If the system loses power or operators are incapacitated, the reactors are designed to automatically shut down and cool themselves using natural physical processes, establishing a "walk-away safe" operational standard.[5][6]

Proving the physics of these reactors unlocks new infrastructure possibilities. The U.S. military is a primary early adopter, seeking to deploy microreactors to secure off-grid power for defense installations under the Army's Janus Program. A reactor that can run for a decade without refueling offers unparalleled strategic independence from vulnerable civilian power grids.[1][3]

Beyond defense, the commercial sector is eyeing microreactors to power the energy-intensive data centers driving the artificial intelligence boom. They also offer a viable solution for providing district heating, water desalination, and reliable electricity to remote mining operations and isolated communities that currently rely on expensive, carbon-heavy diesel generators.[3][6]

While the physics have been validated, the economic viability of the technology remains untested. Critics note that achieving criticality in a laboratory setting does not guarantee that mass-producing these units will be cost-effective. The high cost per megawatt-hour of small reactors may struggle to compete with increasingly cheap renewable energy paired with grid-scale battery storage.[7]

Nevertheless, the successful summer sprint establishes a replicable licensing and testing pathway. By moving from theoretical blueprints to operational hardware in less than a year, the nuclear sector has demonstrated a newfound agility. If the manufacturing economics can be solved, these microreactors lay the groundwork for a highly resilient, decentralized zero-carbon energy grid.[1][7]