Part 53: A Viable Licensing Pathway for New Reactors

The Nuclear Regulatory Commission’s final 10 CFR Part 53 rule is the most significant modernization of reactor licensing in decades. Part 53 is designed differently from existing frameworks to be risk-informed and performance-based.

Part 53 provides a viable, technology-inclusive licensing pathway that reduces the need to force advanced reactor designs into regulatory constructs built for large light-water reactors. That mismatch has been a persistent source of delay, uncertainty, and regulatory friction.

The next test rests on whether it changes how reactors are licensed in practice—and whether developers choose to use it. For the nuclear energy industry, the key question is simple: is Part 53 usable now? Yes, but with constraints.

Part 53 provides a better pathway for advanced reactor licensing, but does not go far enough to create certainty for advanced reactor developers or ensure that benefits to society are considered in reactor licensing.

What Part 53 Is—and How It Works

Part 53 is an optional licensing pathway that does not replace 10 CFR Part 50 or Part 52—which were designed to serve as the licensing framework for conventional reactors. Developers can choose the framework that best fits their technology and deployment strategy.

Congress, through the Nuclear Energy Innovation and Modernization Act, directed the NRC to develop a technology-inclusive, risk-informed, performance-based framework—what is now Part 53.

The core shift is conceptual. Part 53 does not just change how safety is analyzed. It removes several assumptions embedded in earlier frameworks about how reactors must be designed, staffed, secured, and deployed. It does not prescribe specific accident scenarios in advance. Instead, it allows applicants to define risk based on their design. It introduces flexibility in operator requirements through generally licensed reactor operators and reduces prescriptive security assumptions by tying requirements to actual risk rather than fixed staffing models. It supports staggered submissions and common designs, which are necessary for repeated deployment. And it allows, in limited cases, siting decisions to reflect societal risk and benefit rather than relying solely on distance from population centers.

In practice, that means applicants must construct and defend a safety case. The safety case must identify what could go wrong, define the safety functions needed to protect the public, and show how the design and associated programs achieve those functions with sufficient margin. Applicants can use probabilistic risk assessment (PRA), other systematic risk evaluation methods, or combinations of both. The framework is deliberately flexible in how that analysis is performed.

The key point is that Part 53 regulates outcomes, rather than prescriptive design criteria.

A simple way to understand the difference is through an analogy. A prescriptive rule might require a vehicle to have four wheels. That works for cars, but it excludes motorcycles and misrepresents tractor-trailers. A performance-based rule instead asks whether the vehicle can be operated safely—whether it can steer, brake, and remain under control.

Traditional nuclear licensing frameworks often embed assumptions analogous to the “four wheels” requirement. They are based on how large light-water reactors are designed and operated. For technologies that do not share those characteristics, the result is not just added burden, but a structural mismatch that must be resolved through exemptions or reinterpretation.

What Changed from the Proposed Rule

The final rule includes several changes that materially improve usability.

First, the NRC moved away from a strictly PRA-only framework and explicitly allows broader systematic risk evaluations. This matters for technologies that do not have the data or operating history required to support full-scope PRA.

Second, the rule introduces more flexibility in operational programs, including the creation of generally licensed reactor operators. For smaller or more autonomous systems, this is a significant improvement. Operator requirements designed for large plants can become a primary driver of cost and complexity if they are not adjusted to match the plant’s operation characteristics.

Third, the rule improves the pathway for repeated deployment. It allows staggered applications and supports common designs across multiple sites. That is necessary if advanced reactors are expected to scale beyond demonstration projects.

Fourth, the final rule limits the requirement to use NRC-endorsed codes and standards to just safety-related structures, systems, and components. This reduces the need to reconcile non-safety-related components—often with already robust commercial standards—with rigid NRC requirements.

While these improvements can improve the efficiency of the licensing pathway, there are still questions about how Part 53 will work in practice. That will depend on how consistently the NRC treats prior applications, how it handles site-specific differences, and how much learning carries forward between projects.

Where Part 53 Still Falls Short of a Fully Risk-Informed, Performance-Based Framework

Part 53 moves toward the objective of a true risk-informed, performance-based framework, but does not fully realize it. Several issues stand out. Taken together, these issues position Part 53 as a hybrid. It is meaningfully more risk-informed and performance-based than previous frameworks, but it is not a complete implementation of that model.

1. No Clear Definition of “Safe Enough”

Part 53 allows applicants to use integrated risk metrics, but it does not define a single, transparent standard for what constitutes acceptable safety performance. The NRC has not codified Quantitative Health Objectives, but they remain referenced in the Final Rule. Newer concepts, such as comprehensive risk metrics, are introduced without a clear baseline.

The result is flexibility, but also uncertainty. Applicants are expected to define their own metrics, but most will default to established benchmarks to reduce regulatory risk. That creates an implicit standard without the clarity of an explicit one.

For a performance-based framework, that lack of transparency limits predictability.

2. Defense in Depth Remains Partly Deterministic

A fully risk-informed framework would treat defense in depth as a function of uncertainty and risk insights. Part 53 moves in that direction, but retains deterministic elements, including limits on reliance on single features.

Those constraints reflect legacy assumptions about reactor design. For systems with strong inherent safety characteristics, they can force additional layers that do not materially change risk, but do add complexity.

Defense in depth is therefore only partially integrated into the safety case. It remains, in part, a separate requirement.

3. Flexibility That Shifts Burden to Implementation

Allowing systematic risk evaluations instead of requiring a PRA improves accessibility. But it also shifts the burden from meeting a defined standard to defining one. The rule permits multiple approaches, but does not fully define how they will be evaluated. That shifts importance to guidance, staff interpretation, and early precedents.

For developers, this is a familiar tradeoff: flexibility in theory, uncertainty in practice.

4. Transferability Without a Framework

NRC declined to codify a formal pathway for transitioning from Part 50 or 52 into Part 53. More significantly, it explicitly stated that Part 53 applicants may not directly reference approvals issued under those frameworks. That is more restrictive than simply declining to define a transfer process. It prevents applicants from leveraging prior licensing decisions directly, even where the underlying safety case may be comparable.

At the same time, it does not fully exclude transition. Technical information from Parts 50 and 52 can be reused in Part 53. Certain methodologies are aligned across all three rules. Additional regulatory guidance from the NRC is expected in the near future. But the framework still lacks a clear, low-friction path between licensing regimes.

The issue is clear when compared to other rulemakings. The NRC has proposed a pathway allowing applicants to directly reference prior Department of Energy or Department of War reactor authorizations in licensing reviews. That demonstrates that direct-reference models are feasible when the agency chooses to implement them. The result is an intermediate state: transferability is possible, but not structured, creating even more uncertainty.

Part 53 and Benefits to Society

The ADVANCE Act required the NRC to explicitly recognize the benefits of nuclear energy to society. That requirement is reflected in the agency’s updated mission statement. But in Part 53, societal benefits are not broadly integrated into the rule itself.

The NRC declined to incorporate explicit societal-benefit language into the purpose of the rule or to create a formal mechanism for weighing those benefits in licensing decisions. Instead, the concept appears in a narrower form.

The clearest example is reactor siting.

Part 53 allows applicants, under certain conditions, to justify siting a reactor in more populated areas by comparing societal risks and societal benefits. This is a departure from earlier frameworks, which deterministically treated remoteness as the default objective.

This change reflects a more realistic question: not just whether a plant can be sited safely, but whether siting it in a particular location creates public value—through emissions reduction, grid support, industrial use, or reliability.

However, the framework stops there.

Societal benefits are not treated as a general decision variable. They are not systematically integrated into the broader licensing framework, and they are largely excluded from formal cost-benefit analysis in the rulemaking. As a result, Part 53 acknowledges societal benefit, but does not fully operationalize it.

The Case for Industry: Enabling Innovation

For advanced reactor developers, Part 53 addresses a structural problem that has existed for decades: the NRC trying to fit all reactor licensing into the large light-water reactor licensing box.

Legacy frameworks require non-light-water designs to be translated into regulatory constructs that were never built for them. That translation proved inefficient and introduced uncertainty, because the closer a design is forced into an ill-fitting framework, the more interpretation is required.

Part 53 reduces that mismatch. It allows developers to start from their design and build a licensing case that reflects how the system actually behaves, rather than how a conventional light-water reactor behaves.

But the rule does not reduce the burden of proof.

Applicants still need to perform rigorous analyses, define safety functions, demonstrate performance, and establish operational programs. The difference is that the burden shifts from demonstrating compliance with predefined rules to demonstrating that the overall safety case is coherent and sufficient. That is a meaningful change, but not necessarily an easier one. Part 53 is better aligned with advanced reactor designs, but that does not make it any less safe of a licensing pathway.

Part 53 is most likely to make sense for developers early in the licensing process, particularly those with designs that do not align well with light-water-reactor assumptions. It is also attractive for developers willing to construct a full safety case using modern analytical approaches. To encourage use, the ADVANCE Act provides an award to cover the cost of licensing for the first applicant through the Part 53 process.

It may be less attractive for projects already committed to Parts 50 or 52, or for developers that prioritize predictability and precedent over flexibility.

As with most regulatory choices, the decision is not about what is theoretically better. It is about which pathway reduces risk for a specific project.

Conclusion

Part 53 is a substantive step forward. It provides a licensing pathway that is better aligned with advanced reactor technologies and modern risk analysis.

The framework reflects a transition away from purely deterministic regulation, but not fully into a risk-informed, performance-based system. The success of Part 53 will not be determined by the text of the rule. It will be determined by whether developers use it and whether it reduces, rather than redistributes, regulatory friction.