Two Different Reactors Have Now Won the Same Two Arguments With the NRC. Here's the Route for Everyone Behind Them.

Intended Audience: Advanced reactor developers who have a Regulatory Engagement Plan underway and a QAPD in place, and are now looking at the next step: proposing alternative Principal Design Criteria in place of the traditional General Design Criteria.

Executive Summary

Two advanced reactor developers, on two completely different reactor technologies, have each gotten the NRC to accept an alternative to PDC 16 (functional containment) and an alternative to PDC 26 (two diverse means of reactor shutdown). TerraPower's Natrium PDC topical report was approved April 10, 2024 (ADAMS ML24197A231). X-energy's Xe-100 PDC topical report was approved six months later, October 24, 2024 (ADAMS ML24284A012, topical report ML24319A155, Doc ID 004799-A). A sodium fast reactor and a TRISO-fueled HTGR have almost nothing in common at the barrier or mechanism level, and both cleared the same two arguments with the NRC, both grounded in the risk-informed, performance-based methodology of NEI 18-04. That's the real news here: this isn't one company's clever maneuver, it's a repeatable regulatory pattern with two independent proof points behind it. Neither company tried to make its reactor look like an LWR to satisfy the existing GDC. Both went back to the actual safety problem the GDC is trying to solve, limiting radionuclide release, maintaining redundant and diverse reactivity control, and built a solution shaped by their own technology instead of forcing their technology into an LWR-shaped answer. For a developer who has finished a QAPD and is looking at what comes next in a Regulatory Engagement Plan, proposing alternative PDCs in place of the GDC is that next step, and there are now two working examples of how to build the argument, on two different reactor technologies, both publicly documented in ADAMS.

Where Alternative PDCs Fit in Your Licensing Sequence

For a new entrant building out a Regulatory Engagement Plan, the sequence usually runs: establish the REP itself, stand up a QAPD that actually fits the size and maturity of the organization, and then start proposing the technical building blocks the rest of the licensing basis will rest on. Alternative Principal Design Criteria are one of the first substantial technical building blocks in that sequence, because the standard General Design Criteria in 10 CFR Part 50, Appendix A were written for light water reactors, and most non-LWR designs don't fit them without modification. NRC Regulatory Guide 1.232 gives applicants a defined path for proposing alternative PDCs, and I laid out the underlying SAFDL/SARRDL framework in an earlier post on this blog.

Proposing alternative PDCs early, well before a Construction Permit Application, means a developer gets NRC agreement on the shape of its safety arguments while the design still has room to move. Two different reactor technologies have now walked this exact path successfully. That's worth treating as a template, not just two isolated wins.

What a Principal Design Criteria Topical Report Actually Does For You

A PDC topical report isn't a license application. It's a pre-application document that asks the NRC to approve a set of design and safety criteria, adapted from the GDC, for a specific non-light-water reactor design. Getting one approved doesn't authorize construction or operation, but it does something useful: once it has an NRC safety evaluation attached, it can be incorporated by reference into a Construction Permit Application or Combined Operating License Application, not just by the company that submitted it, but by any applicant whose design fits within the approved envelope.

What a PDC topical report is not, usually, is sellable intellectual property. It's largely an extended, carefully worded argument with the regulator over what design and safety criteria are appropriate for a given reactor concept, and once it's approved, it's sitting in ADAMS as public record. Any developer with the technical background to follow the reasoning, the wording changes NRC required, the conditions attached, can read it and construct a comparable argument for their own design without paying anyone for it. What an approved PDC topical report can save another developer is the time and NRC engagement cycles it takes to work through that argument, not access to information they couldn't otherwise get.

Fuel qualification topical reports are a different animal, and this is where real sellable IP tends to show up in advanced reactor licensing. A fuel qualification case rests on proprietary data the developer actually had to generate: irradiation testing, material characterization, performance data under specific temperature and burnup conditions; work that costs real money and takes real time to produce, and that a competitor can't reconstruct just by reading ADAMS, because the underlying data isn't public even when the topical report's existence and general conclusions are. That's the kind of topical report a developer can actually license to other companies for real value. For investors evaluating a licensing-stage nuclear company, the topical report portfolio is worth asking about by type, not as one asset: a PDC topical report signals process discipline and NRC engagement maturity; a fuel qualification topical report signals genuine, defensible intellectual property.

This distinction catches a lot of small LWR developers off guard. Plenty of small modular reactor designs are marketed as built on "proven" LWR fuel technology, as if reusing an established fuel type means the licensing burden is already handled. It isn't. Fuel qualification is specific to a design's actual operating envelope, power density, burnup, temperature, core geometry, not just the fuel type. A developer using a well-established fuel in a new core design still has to generate or acquire fuel qualification data that actually covers their specific envelope. "Proven" describes the fuel. It doesn't describe the licensing basis.

Functional Containment: Two Technologies, One Accepted Pattern

Traditional LWRs already rely on more than one barrier. Fuel cladding, the reactor pressure vessel, and the containment structure together form a defense-in-depth stack, with the containment building acting as the last and largest of the three. The GDC's actual underlying problem isn't "build a large civil containment structure," it's "limit the release of radionuclides to the environment." Functional containment arguments take that problem at face value instead of assuming an LWR-shaped structure is the only acceptable answer, and the two approved examples show just how differently a correctly redefined solution can look depending on the technology.

TerraPower's Natrium argument (PDCs 13, 16, 81, and 82, approved in NATD-LICRPRT-0002, Revision 1) rests on physical characteristics specific to a sodium fast reactor: near-atmospheric operating pressure that prevents energetic releases, a large margin to sodium's boiling point that prevents coolant boiling, elimination of loss-of-coolant accident potential, single-phase liquid sodium coolant, and high-reliability fuel. NRC accepted this as consistent with the risk-informed, performance-based methodology of NEI 18-04 and the functional containment policy in SECY-18-0096, while explicitly declining to evaluate the specific functional containment barriers and performance criteria, reserving that for a future licensing action.

X-energy's Xe-100 argument (PDC 16) rests on something almost unrelated physically: TRISO particle coating layers standing in for fuel cladding, the graphite pebble adding an intermediate barrier, and the helium pressure boundary functioning the way a primary system boundary does, all shrunk to particle scale instead of relying on a large civil containment structure, because the source term itself is inherently limited by how TRISO particles behave at the temperatures the Xe-100 core can reach. One wording change during review shows how carefully this argument was built: NRC required X-energy to change "design conditions important to safety" to "design limit," a term that specifically means a value that cannot be exceeded.

In both cases, no specific numeric design limits were proposed or approved, it was too early in each design for those calculations. What NRC approved in both is the methodology: the approach each company's engineers will use to derive functional containment design limits once the underlying calculations are complete. Getting NRC agreement on methodology before the numbers exist is what narrows how much can be relitigated later. And it's worth being precise about what happens when those numbers do arrive: NRC performs its own independent calculations for the important safety parameters, using its own codes and its own engineers, to arrive at an answer independent of the applicant's; that's confirmatory analysis. What NRC does not do is audit the applicant's own calculation package line by line for arithmetic or procedural accuracy. That's a separate, internal check the applicant's QA program has to provide first, through the independent design review ASME NQA-1 requires before the calculation ever reaches NRC staff. Two independent checks, run by two different parties for two different purposes: NRC's confirmatory analysis verifies the answer is right, the applicant's NQA-1 independent review verifies the process that produced it was followed correctly. It's a clean, practical example of a QA program doing exactly what it's supposed to do.

Two Diverse Means of Shutdown: Two Technologies, One Accepted Pattern

The second criterion both companies won an alternative for is PDC 26, reactivity control: how the plant shuts itself down and stays shut down. NRC's General Design Criteria require two independent, diverse means of achieving and maintaining safe shutdown, so a single failure mode in one system can't leave the plant with no way to control reactivity. The GDC's underlying problem isn't "have a control rod system plus a boron injection system," it's "make sure no single failure mode can take away your ability to control reactivity." An LWR typically satisfies that problem with control rods plus boron injection, and neither Natrium nor the Xe-100 has boron injection, so both companies went back to the underlying problem and built an equivalent case from the ground up, each grounded in NEI 18-04's risk-informed, performance-based methodology, and they built two genuinely different solutions.

TerraPower's approach uses two different control rod assembly designs combined with two distinct insertion mechanisms: a gravity-driven scram latch release and a motor-driven insertion function. TerraPower supported the argument with NEI 18-04 methodology, using probabilistic risk assessment and failure mode analysis to demonstrate adequate independence and diversity between the two. NRC found the approach "reasonable in meeting the underlying intent of PDC 26," while stating explicitly that "the NRC staff is unable to make a final determination regarding whether TerraPower has demonstrated conformance of the Natrium design with PDC 26 at this time," reserving that determination for a future licensing action.

X-energy split the same requirement into three distinct criteria by operational state rather than by mechanism: PDC-RFDC 26 for movable poison systems used during design basis events, PDC-CDC 26 for diverse means of reactivity insertion during anticipated operational occurrences, and PDC-OCDC 26 for reactivity control during normal operations. NRC found this structure "addresses the underlying intent of MHTGR-DC 26" and confirmed it satisfies the two-diverse-means requirement across all three operational states, conditioned on X-energy demonstrating proper NEI 18-04 methodology when the criteria are actually applied in a future licensing review.

Neither company got a final, unconditional determination on PDC 26. Both got the NRC to accept that their proposed approach reasonably meets the underlying intent of the requirement, with the actual conformance demonstration reserved for a real application. Two different reactivity control schemes, two different ways of structuring the argument, the same regulatory outcome.

The Route This Opens for Developers Proposing Alternative PDCs

Two working examples, on two unrelated reactor technologies, is enough to treat this as a route rather than a coincidence. A few things are worth knowing before walking it.

First, both approvals are design-specific. A developer can't cite either topical report and expect the NRC to accept it wholesale for a materially different design. What's transferable is the argument structure, multiple credited barriers for functional containment, a clearly reasoned case for independence and diversity for shutdown, not the specific technical content behind either company's version.

Second, the sequencing is the real lesson, and now there are two data points showing it works regardless of the underlying technology. Getting the methodology locked in through a topical report, before a real application forces every number to be final, lets a design keep maturing on schedule instead of having its safety case frozen prematurely by an early filing.

Third, budget real engineering and licensing time for NEI 18-04 methodology if your case leans on it the way both of these did. Both companies' PDC 26 approvals are conditioned on demonstrating that methodology, specifically the PRA and failure mode analysis it requires, correctly at application time. That's a substantive technical demonstration with its own schedule, not a paperwork step.

What Small Teams Should Do Now

Read the safety evaluations, not just the topical reports, for both approved PDC sets. The specific wording NRC required, and the specific language NRC used to describe what it was and wasn't determining, tell you more about what the agency will actually accept than the original submissions do.

Map your own design against both approved structures before assuming you need to build an argument from zero. Even where the specific technical content doesn't transfer, as it largely doesn't between a sodium fast reactor and a TRISO-fueled HTGR, the underlying approach to building the case often does.

Talk to the NRC early about which parts of your safety case you intend to finalize at the alternative-PDC stage versus reserve for the application stage, and be explicit that you're establishing methodology, not final numbers. Both companies made that split explicit and got NRC to agree to it in writing. That's a negotiation worth having on purpose, not something to discover by accident partway through review.

Frequently Asked Questions

Q: Has more than one advanced reactor developer gotten the NRC to accept a functional containment argument?

A: Yes. TerraPower's Natrium PDC topical report was approved April 10, 2024 (ADAMS ML24197A231), and X-energy's Xe-100 PDC topical report was approved October 24, 2024 (ADAMS ML24284A012). The two arguments rest on entirely different physical characteristics, sodium fast reactor thermal-hydraulics for Natrium, TRISO particle fuel behavior for the Xe-100, which is what makes the pattern significant rather than technology-specific.

Q: Has more than one developer gotten NRC to accept an alternative to PDC 26 (two diverse means of shutdown)?

A: Yes. TerraPower proposed two different control rod assembly designs with two distinct insertion mechanisms, supported by PRA and failure mode analysis; NRC found this reasonable but reserved final conformance determination for a future licensing action. X-energy proposed a three-part criterion split by operational state, supported by NEI 18-04 methodology; NRC found this addressed the underlying intent of the standard criterion, with the same kind of future determination reserved. Different mechanisms, same regulatory pattern.

Q: Did either topical report include the specific numeric safety limits behind its functional containment or shutdown argument?

A: No, and neither was supposed to. No specific design limits were proposed at this stage in either case, it was too early in both designs for those calculations. A PDC topical report establishes the methodology for deriving those limits later, so there's less room for the NRC to push back once the actual numbers are calculated and submitted in a real application.

Q: Does the NRC independently check an applicant's engineering calculations for accuracy?

A: For the important safety parameters, yes, through confirmatory analysis using NRC's own codes and its own engineers to independently arrive at an answer rather than simply accepting the applicant's number. What NRC does not do is audit the applicant's own calculation package for arithmetic or procedural accuracy, that's the job of the independent design review required under the applicant's QA program (10 CFR 50, Appendix B and ASME NQA-1). The two checks serve different purposes: NRC's confirmatory analysis verifies the answer is right, and the applicant's NQA-1 independent review verifies the process that produced it was followed correctly.

Q: Where does proposing alternative PDCs fit in a new developer's licensing sequence?

A: Typically after the Regulatory Engagement Plan is underway and a QAPD fitted to the organization's actual size and maturity is in place. Alternative PDCs are one of the first substantial technical building blocks in a non-LWR licensing basis, since the standard GDC in 10 CFR Part 50, Appendix A were written for light water reactors and rarely apply without modification.

Q: Can other developers rely on an approved PDC topical report from Natrium or X-energy?

A: Only for the parts of the argument that genuinely match their own design, the argument structure is more transferable than the specific technical content. A developer with a materially different core design, coolant, or reactivity control scheme would still need its own topical report or a clearly justified basis for referencing an existing one. Unlike fuel qualification data, a PDC topical report isn't really proprietary once approved, it's public in ADAMS, so the value to a third party is in the time saved constructing their unique argument, not in licensing something otherwise unavailable.

Q: Is a PDC topical report the same kind of asset as a fuel qualification topical report?

A: No. A PDC topical report is mostly an argument, a negotiated set of design and safety criteria, and it becomes public record once approved, so any developer can read and adapt the reasoning without paying for it. A fuel qualification topical report rests on proprietary data the developer generated, irradiation testing, material characterization, performance data under specific conditions, that a competitor can't reconstruct just by reading ADAMS. That data is genuinely sellable IP in a way an approved PDC argument generally isn't.

Q: What is NEI 18-04, and why does it matter here?

A: NEI 18-04 is the risk-informed, performance-based licensing methodology developed under the Licensing Modernization Project, built around probabilistic risk assessment and failure mode analysis. Both TerraPower's Natrium and X-energy's Xe-100 grounded their PDC 26 arguments in NEI 18-04 methodology, applied to two very different reactivity control schemes.

Q: What's the actual strategic lesson in these two approvals, beyond "it can be done"?

A: Neither company tried to make its design look like an LWR to fit the existing GDC. Both went back to the safety problem the GDC is actually trying to solve, limiting radionuclide release for PDC 16, ensuring no single failure mode removes the ability to control reactivity for PDC 26, and proposed a solution shaped by their own technology's physics instead of forcing their reactor into an LWR-shaped answer. That's the transferable move for any non-LWR developer: identify the underlying safety problem before assuming the GDC's LWR-specific solution is the only acceptable one.

Sarah Gibboney, P.E., is the Founder and Principal Licensing Engineer of Gibboney Nuclear, PLLC. She has contributed to the licensing of eight reactor designs, two of which are operating and selling power commercially, and co-authored Construction Permit Applications for both Advanced Reactor Demonstration Program awardees, TerraPower's Natrium and X-energy's Xe-100.

Related reading: The Functional Containment Licensing Argument · What the Natrium Construction Permit Actually Means, and What Still Has to Happen