In a small Wyoming town that spent a century running on coal, the federal government just gave the green light to build something the U.S. nuclear industry hasn’t seen in a generation: a brand-new, advanced reactor.
The U.S. Nuclear Regulatory Commission issued a construction permit in early 2026 to TerraPower subsidiary US SFR Owner, LLC, authorizing construction of Kemmerer Power Station Unit 1, a sodium-cooled fast reactor paired with a molten salt energy storage system. The permit, designated CPAR-1 under Commission Order CLI-26-5, is documented in the NRC’s public records and represents the first construction permit the agency has granted for a new commercial reactor design in decades.
The approval landed roughly nine months ahead of the timeline federal officials had projected, according to the Department of Energy’s Office of Nuclear Energy, which described the review process as ahead of schedule and under budget. And the timing coincides with reports from Bloomberg and other outlets that Meta Platforms has agreed to purchase power from multiple future Natrium reactors to feed its growing fleet of artificial intelligence data centers, a deal that, if confirmed, could fundamentally change how advanced nuclear projects get financed in the United States.
A fast-tracked review for a first-of-its-kind reactor
TerraPower’s subsidiary filed its construction permit application with the NRC on March 28, 2024, setting the agency’s review clock in motion. From that point, the NRC ran two parallel tracks: a safety evaluation and an environmental review. Both had to be completed before any permit could be issued.
The NRC completed its final safety evaluation by late 2025, according to the agency’s own safety review update, with the DOE separately stating the milestone was reached in December 2025. The permit followed shortly after, meaning the entire process from application to authorization took less than two years.
That pace is striking by historical standards. U.S. nuclear licensing has traditionally stretched across many years, sometimes exceeding a decade from initial application to full authorization. The last time the NRC issued a construction permit for a genuinely new reactor design was in the 1970s. (The agency granted combined licenses for the AP1000 reactors at Plant Vogtle in Georgia in 2012, but those followed a different regulatory pathway.) The compressed Kemmerer timeline suggests the NRC’s processes can move faster when an applicant and the agency are aligned on expectations, though the precise original schedule has not been published in NRC documents. The “nine months early” figure is an informed approximation drawn from DOE characterizations and industry reporting, not an official NRC performance metric.
Why the Natrium design is different
The reactor planned for Kemmerer is not a scaled-down version of the pressurized water reactors that dominate the existing U.S. fleet. It uses liquid sodium as a coolant instead of water, which allows the reactor to operate at higher temperatures and lower pressures. That combination offers potential safety advantages: sodium-cooled reactors do not require the massive, high-pressure containment vessels that define conventional plants, and the physics of the design allow the reactor to shut itself down passively if cooling is lost.
The more commercially distinctive feature is the integrated molten salt energy storage system. The reactor itself is designed to produce 345 megawatts of electricity at its baseload rating, but the storage bank can boost the plant’s output to 500 megawatts during periods of peak demand. In practice, the reactor runs at a steady level while operators charge or discharge the thermal storage system to modulate how much electricity reaches the grid.
That flexibility is the core of TerraPower’s pitch to utilities and corporate buyers. Traditional nuclear plants are built to produce constant output with limited ability to ramp up or down, which makes them excellent baseload generators but poor partners for grids increasingly shaped by variable wind and solar power. A Natrium plant, at least on paper, could fill the role now occupied by natural gas peaker plants: absorbing excess renewable generation during sunny or windy hours and then dispatching stored energy when demand spikes.
The Meta deal: what’s reported and what’s missing
The reported Meta agreement has generated as much attention as the permit itself, but the two rest on very different evidentiary foundations. As of June 2026, neither Meta Platforms nor TerraPower has published a binding contract, regulatory filing, or detailed joint announcement spelling out the terms of any power purchase arrangement. No corresponding disclosure appears in NRC dockets or in the securities filings that would typically accompany a large, long-term off-take agreement.
What exists are reports from credible financial outlets indicating that Meta intends to buy power from several Natrium reactors to supply its AI data centers. The broad strokes are consistent with a pattern that has emerged across the tech industry: Google has partnered with Kairos Power on small modular reactors, Amazon has invested in X-energy, and Microsoft signed a deal to restart a reactor at Three Mile Island. Each of these companies faces the same underlying pressure. Training and running large AI models requires enormous, reliable electricity supplies, and corporate climate pledges demand that the power be low-carbon.
But the specific contours of the Meta-TerraPower arrangement remain unverified. It is not known how many reactors would be covered, what capacity each would commit to Meta’s operations, or how power deliveries would be structured relative to the plants’ broader grid obligations. Pricing terms have not been disclosed. Nor is it clear whether Meta would be purchasing physical electricity, clean energy credits, or some combination to support its public sustainability commitments.
The financial implications could be significant. The Kemmerer project is backed by substantial federal cost-sharing through the Department of Energy’s Advanced Reactor Demonstration Program, which has committed roughly $2 billion in matching funds to accelerate first-of-a-kind deployments. A robust corporate off-take contract could lower TerraPower’s cost of capital by providing predictable revenue, potentially allowing the company to attract private financing for additional reactors beyond the initial demonstration unit. But without transparent financial disclosures, the interaction between federal support and private commitments remains unclear.
What Kemmerer stands to gain
The reactor will be built near the site of PacifiCorp’s Naughton coal plant, which has been winding down operations as the utility shifts away from fossil fuels. For Kemmerer, a town of roughly 2,600 people in Lincoln County, the Natrium project represents a direct economic bridge from coal to advanced nuclear. TerraPower has said the plant will create approximately 250 permanent jobs and up to 1,600 construction jobs during the build-out, numbers that carry real weight in a community where the coal plant was a primary employer.
Construction is expected to begin in 2026 now that the permit is in hand, with a target operational date around 2030. That timeline is ambitious for a first-of-its-kind facility, and delays are common in nuclear construction. But the accelerated permitting process has at least removed one of the longest-lead obstacles.
What this means for the advanced nuclear industry
TerraPower’s permit is a concrete milestone, but it does not by itself prove that advanced nuclear power is ready for commercial scale in the United States. The reactor still has to be built on time and on budget, a challenge that has tripped up nuclear projects around the world, most recently the Vogtle expansion in Georgia, which came in years late and billions over its original estimate.
What the permit does demonstrate is that a first-of-a-kind advanced reactor design can navigate the existing U.S. regulatory framework faster than many in the industry expected. If the reported Meta deal materializes with binding terms, it would add a second proof point: that private capital from outside the traditional utility sector is willing to underwrite nuclear construction based on long-term power demand rather than regulated rate recovery.
Together, those two developments would represent a meaningful shift in how nuclear projects get built and paid for in this country. Separately, the permit is solid ground and the Meta deal is still an open question. Readers following this story should watch for a formal contract announcement, any corresponding NRC or SEC filings, and TerraPower’s own updates on construction progress in Kemmerer. Those will determine whether the Natrium program becomes a template for the next generation of American nuclear power or remains, for now, a promising but singular experiment.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
