The Nuclear Regulatory Commission’s proposed Part 57, formally titled “Licensing Requirements for Microreactors and Other Reactors With Comparable Risk Profiles,” was published in the Federal Register on May 1, 2026, with the comment window closing June 15. The rule applies to reactors at or below 100 megawatts electric and builds a separate licensing pathway around the smaller safety case of factory-built units, rather than asking developers to retrofit themselves into procedures designed for 1,000 megawatt light-water plants.

The 6-to-12-month timeline from application to deployment is the line that has carried the news coverage. It is the right line to lead with, but it is not the most important part of the rule. Three structural provisions in the proposal change microreactor unit economics in ways the headline timeline does not capture, and those provisions decide which projects actually pencil between now and 2030.

Three structural changes, not one

Fleet approval of standard designs. Under the existing Part 50 and Part 52 frameworks, even identical units sited at different locations are reviewed substantially from scratch each time, because the licensing process was built around bespoke, site-tailored plants. Part 57 makes explicit what advanced reactor developers have been asking for since the early 2020s: a single design approval that supports deployment across multiple sites without rebuilding the technical case. The closest analog is the Federal Aviation Administration’s type certificate, where one approval covers an entire production run of an airframe.

This was already partially possible. NuScale’s Standard Design Approval for the uprated 77 MWe US460, issued by the NRC in May 2025, lets a future combined license application reference the approved design directly. Part 57 turns that pattern from a case-by-case exception into the default for microreactors. The economic effect is to move design recovery costs from a single-site amortization to a fleet amortization, which is the only way the manufacturing-line business model the microreactor sector has been promising investors actually works.

Manufacturing licenses. Part 57 contemplates licensing the production facility itself, with units shipped as completed assemblies and sited without each location rerunning the construction permit cycle. Combined with general-license provisions for limited construction activities ahead of full NRC sign-off, this collapses the longest-pole item in domestic nuclear schedules: the years spent on site-specific construction permitting. For developers building reactors small enough to ship by truck or rail, the relevant analog is closer to a packaged industrial heat exchanger than to a Vogtle-style multi-billion-dollar site project.

Autonomous operation and reduced on-site staffing. The proposed framework explicitly recognizes remote monitoring, remote operation, and reduced on-site staffing as acceptable operating modes. That is a material change. Operations and maintenance for conventional nuclear plants runs to staffing levels that simply do not work at the unit economics of a 5 to 20 megawatt microreactor sited at an industrial customer’s facility or military base. If the regulatory framework requires a Vogtle-style control room and shift complement for every microreactor unit, the per-megawatt fixed cost crushes the business case before the fuel is loaded. Part 57 does not eliminate operating staff, but it makes a centralized operations center supervising multiple deployed units the assumed default rather than the exotic exception.

The NRC’s own projection puts combined agency and industry savings at $3.76 billion to $11.84 billion over the implementation horizon, against a baseline that puts current advanced-reactor licensing in the five-plus-year range from docket to operation. Half of that compression in practice would still be a transformative shift.

What this changes for the deployment stack

The proposal does not turn microreactors into a near-term option for utility-scale generation. The 100 megawatt electric ceiling rules out, by design, the larger SMR designs that are oriented at grid-scale baseload (NuScale US460 at 462 MWe across six modules, Holtec SMR-300 at 300 MWe, TerraPower Natrium at 345 MWe). What Part 57 unlocks is a set of deployments that have been talked about for years and that have never penciled at the current licensing cost.

Behind-the-meter industrial heat. Dow Chemical’s Seadrift, Texas joint announcement with X-energy in 2023, targeting four Xe-100 modules at 320 megawatts thermal each, was the highest-profile industrial-heat pairing. Xe-100 sits at 80 megawatts electric per unit, inside the Part 57 envelope. Industrial process heat customers value firm low-carbon thermal output more than electricity per se, and many of them sit at facilities too small to support the staffing and licensing overhead of a conventional reactor. Part 57 makes the unit economics on those sites work, and the customer pool is real: chemicals, refining, cement, hydrogen.

Department of Defense installations. Project Pele, the DOD prototype mobile microreactor program run through the Strategic Capabilities Office and built by BWXT, was deliberately structured as a Department of Energy authorization at Idaho National Laboratory rather than an NRC license, because the Part 50 framework was a poor fit. The Eielson Air Force Base microreactor pilot announced earlier in the decade has been working through a separate authorization pathway. Part 57 gives DOD a civil-licensing route that fits the operational pattern: small unit, remote site, centrally supervised. That matters less for Pele itself than for the second and third wave of military microreactor deployments that DOD has signaled it wants but has not yet committed funding to scale.

Remote and off-grid industrial sites. Mining operations in Alaska, Canada, and Australia have been quoting microreactor offtake interest for half a decade with no actual procurement, because the per-site licensing burden has always killed the business case. The Westinghouse eVinci heat-pipe microreactor at roughly 5 megawatts electric, Last Energy’s 20-megawatt PWR module, and BWXT’s BANR design all target this market. Fleet approval plus manufacturing licenses plus reduced staffing requirements is the package those projects need.

Data center co-located microreactors. This one is more contested. Hyperscalers have publicly preferred larger units (Constellation’s Three Mile Island restart for Microsoft, Talen’s Susquehanna arrangement with Amazon, Vistra negotiations with Comanche Peak) because the load per facility justifies hundreds of megawatts. But the second wave of data center sites under development, particularly the secondary-market campuses being announced in Pennsylvania, Indiana, and Wyoming, sit at load levels (50 to 200 megawatts) where a small cluster of microreactors plausibly competes with grid co-location under the FERC PJM tariff framework being drafted this year.

Who is positioned to actually file

Part 57 only matters to the degree that developers actually use the new pathway rather than continuing under Part 50 and Part 52 out of habit. The proposed rule does not force a transition, and the existing pathways remain available. Several developers are clearly positioned:

  • Oklo. The Aurora compact fast reactor design, originally 1.5 megawatts electric and now in scaled-up configurations, was the test case for advanced reactor licensing under the old framework. The NRC denied Oklo’s first combined license application in January 2022 on completeness grounds. The company has been preparing a refiled application that, on its current schedule, would arrive after the Part 57 final rule takes effect. Oklo also signed a fuel-recycling agreement with the Idaho National Laboratory and has data-center offtake conversations on record.
  • X-energy. Xe-100 at 80 megawatts electric sits inside the Part 57 envelope. The DOE Advanced Reactor Demonstration Program award and the Dow Seadrift commercial pairing put X-energy further along the customer-engagement curve than most peers. TRISO fuel manufacturing through X-energy’s TRISO-X subsidiary is also in progress, which addresses part of the HALEU question for the company’s own deployments but does not solve it for the sector.
  • Kairos Power. The Hermes 1 demonstration reactor under construction at Oak Ridge, Tennessee, is licensed under the old framework, but Kairos has signaled that successor commercial units would use whatever framework is best fit by the time they file. Hermes designs sit comfortably under 100 megawatts electric and use FLiBe molten salt cooling.
  • Westinghouse. The eVinci microreactor at approximately 5 megawatts electric is the smallest mainstream commercial design and the most natural fit for the autonomous-operation provisions of Part 57. Westinghouse has not announced a Part 57 filing intention, but if the rule survives the comment period substantially intact, eVinci is the unit best engineered to make full use of it.
  • BWXT. Beyond Project Pele, BWXT’s BANR (BWXT Advanced Nuclear Reactor) commercial offering targets the same off-grid industrial market. BWXT is also the dominant US producer of HALEU through downblending of high-enriched uranium, which gives the company a structural advantage in fuel supply that no other developer matches.
  • Last Energy. A US-headquartered startup focused on a 20 megawatt PWR microreactor design, with European customer interest leading the company’s pipeline. The smaller-PWR strategy is intentionally conservative on technology risk, betting on Part 57 (and equivalent EU frameworks) to clear the regulatory hurdle.

The names not on this list matter too. NuScale’s US460 at 77 megawatts electric per module is technically within the Part 57 ceiling but the company’s commercial model is built around six-module 462 megawatt plants, which are not. Holtec’s SMR-300 and TerraPower’s Natrium are above the threshold. Those developers will continue working under Part 50 and Part 52, and the gap between their licensing experience and the Part 57 cohort will become a meaningful competitive variable.

Where the chokepoint moves

Part 57, if adopted substantially as proposed, takes licensing off the critical path. The chokepoint then moves to three places that are already known but are now more binding.

HALEU fuel supply. Most advanced reactor designs require high-assay low-enriched uranium fuel, between 5 and 19.75 percent U-235. The current domestic HALEU supply chain is small. Centrus Energy’s Piketon, Ohio facility has been operating at demonstration scale. DOE’s surplus plutonium designation for SMR fuel, advanced under the rule announced May 27, 2026, expands the inputs available. Recent expansion efforts at Centrus, plus DOE allocation programs, are scaling supply. Even so, the projected HALEU demand from a microreactor sector deploying at the Part 57 timelines is materially above any current production schedule. If licensing compresses to 12 months and fuel availability remains a 24-to-36-month constraint, the binding constraint simply migrates.

Project financing. Microreactor unit economics depend on factory-line cost compression that only materializes at production volumes the sector has not yet reached. The first 10 to 20 units shipped by any vendor will be expensive. Customer offtake contracts for those units have to absorb that cost or developer balance sheets do. Lender appetite for first-of-a-kind nuclear paper, even under a friendlier licensing regime, is constrained by the still-fresh Vogtle and CFPP cancellation experiences. Part 57 helps the financing case by removing licensing risk, but it does not pay the first unit’s cost premium.

Customer offtake structure. Most of the announced microreactor deals are framework agreements, not binding offtake. Dow Seadrift, the Eielson Air Force Base pilot, the various mining-sector letters of intent: the conversion rate from announcement to executed power purchase agreement has been low. Customer organizations capable of underwriting first-mover risk on factory-built nuclear are a narrow group, and the contracts are non-trivial to draft. The licensing compression Part 57 offers makes the customer side more attractive but does not by itself produce the contracts.

What to watch through the rest of 2026

The comment window closes June 15. The substantive engagement to monitor will come from three quarters.

Industry will press for specific clarifications: how broadly fleet approval extends across design variations, how manufacturing license inspections are structured, how the autonomous-operation provisions interact with cybersecurity and physical security requirements. State regulators (the public utility commissions in jurisdictions hosting prospective microreactor sites) will push back on aspects of the framework that could limit their authority over distribution-level interconnection and siting. Public-interest groups and the Union of Concerned Scientists will challenge the autonomous-operation provisions and the reduced-staffing assumptions in particular.

Final rule publication, if the agency holds to a reasonable timeline, would arrive in late 2026 or early 2027. The first filings under the new pathway would land shortly after. Oklo’s refiled application, if structured to use Part 57, would be the highest-profile early test. The first manufacturing license application, whoever submits it, would set the precedent for the production-line approval pattern the entire sector is counting on.

Risks to the read

The proposed rule is not the final rule. Substantial provisions could be modified in response to comments, in either direction. The autonomous-operation provisions in particular have a credible chance of being tightened during final rulemaking, which would partly undermine the operating-cost case.

A change in NRC commissioner composition could slow or refocus the rulemaking. The current commission has signaled clear support for the framework. A different composition in 2027 or beyond could reopen settled questions.

Industry uptake could be slower than the framework supports. Developers with existing Part 50 or Part 52 filings have sunk costs in the old pathway and may continue under it rather than restart under Part 57. The first wave of Part 57 filings will tell whether the sector actually exercises the new option.

State-level barriers could swamp the federal compression. Several states (California, New York, New Jersey at various points) have additional procedural or political obstacles to nuclear deployment that the NRC framework does not address. Part 57 helps where the constraint is federal licensing, which is most jurisdictions but not all.

The frame

The headline 6-to-12-month timeline is real, and it matters, but it is also the kind of number that policy announcements always produce and that practical experience usually compresses by less than projected. The more durable read on Part 57 is structural. The rule treats microreactors as manufactured products rather than as one-off plants, and aligns the regulatory framework with the factory-line business model the sector has been describing for a decade. That alignment is the part that survives whatever practical slippage occurs in actual licensing timelines.

The investment thesis adjusts. Companies whose engineering, fuel supply, and customer pipeline can absorb a Part 57 pathway are positioned to scale much earlier in the decade than the consensus analyst forecasts suggest. Companies whose plans implicitly require Part 50 timelines and bespoke staffing will look comparatively expensive. The next 18 months, between the close of comments in June and the first credible Part 57 filings in late 2026 or early 2027, are when that gap becomes legible.

Sources

  • Nuclear Regulatory Commission, “Licensing Requirements for Microreactors and Other Reactors With Comparable Risk Profiles,” proposed rule, published in the Federal Register May 1, 2026, comment period closing June 15, 2026.
  • American Nuclear Society, “NRC introduces microreactor regulatory framework,” April 27, 2026, ans.org/news/2026-04-27/article-7981/nrc-introduces-microreactor-regulatory-framework.
  • NuScale Power, press release on US460 Standard Design Approval, May 29, 2025.
  • Department of Energy, Advanced Reactor Demonstration Program awards to X-energy and TerraPower, 2020 and subsequent.
  • Department of Defense, Strategic Capabilities Office, Project Pele documentation and BWXT prime contract materials.
  • Dow Chemical and X-energy, joint announcement on Seadrift, Texas Xe-100 deployment, 2023.
  • Centrus Energy, HALEU production updates from the American Centrifuge Plant, Piketon, Ohio.
  • US Department of Energy, surplus plutonium SMR fuel program announcement, May 27, 2026.
thesisnuclearmicroreactorsnrcpermittingsmrhaleu