Blog: Lessons from NuScale's Design Certification Process

In August 2020, NuScale Power reached a significant milestone, becoming one of the first developers to receive a design approval for its small modular reactor (SMR) from the Nuclear Regulatory Commission (NRC) staff. NuScale’s design is a simplified and smaller version of pressurized light-water reactors (PWR) that the NRC has regulated for decades. That the licensing process has so far required 5-years and a 12,000-page Design Certification Application (DCA) which together highlight its complexity. If licensing a reactor that is based on common PWR is this challenging, there will be a further disconnect between the NRC’s existing licensing framework and the innovative designs behind a new generation of advanced nuclear reactors. In a recent comment to the NRC, we analyzed NuScale’s DCA and made recommendations for consideration in future applications.

One of the main inefficiencies in the DCA process comes from the prescriptive nature of the NRC’s existing licensing framework, which is based on experiences with large, light water reactors (LLWR). NuScale required 17 exemptions, with some additional topics left for individual licensees attempting to build NuScale’s SMR to address. To streamline this process, we recommend the NRC do away with prescriptive standards for licensing that are irrelevant to new reactor designs. As an example, 10CFR part 50 “Domestic Licensing of Production and Utilization Facilities”, provides control room staffing requirements based on a set of assumptions that are applicable to LLWRs. This would require NuScale licensees to seek an exemption if they want to use the number of operators NuScale actually recommends. As an alternative, designers and licensees should be allowed to demonstrate how safety objectives are being met on a design-by-design basis, without having to seek exemptions.

Another inefficiency we identified in NuScale’s DCA process relates to unresolved technical issues. Specifically, the NRC identified three issues that were “unresolved open items,” including NuScale’s shielding wall design, containment leakage from the combustible gas monitoring system, and stability of the steam generator in the presence of density wave oscillations. NuScale’s DCA was still approved with these issues outstanding, but the burden of addressing them will fall on NuScale licensees. To help simplify this process, we recommend that the NRC provide a generic version of these outstanding issues that are not specific to the NuScale design so they can be resolved through multiple mechanisms with the help of the wider research community. The NRC could further streamline this process by identifying its outstanding considerations for these design issues and providing guidance on how to resolve them.

The NRC has tried to alleviate some of this regulatory burden. Starting in 2017, with guidance from the Nuclear Energy Institute (NEI), the agency began the process of streamlining licensing for non-light water reactor technologies. Unfortunately, despite this effort, the “Licensing Modernization Project” LMP has proven to not perfectly match the licensing strategy for most developers.

Despite this setback, advanced nuclear experts and developers have expressed confidence that there are still other pathways to licensing advanced reactors by 2030. For some, like Kairos Power, that has meant crafting their own versions of LMP. For others, like Oklo Inc., it’s meant formulating a wholly original way of going through the licensing process. But if NuScale’s experience is any indication, this bespoke approach won’t be easy or efficient.

There have been other complications following LMP’s approval. While industry advocates like NEI only intended for LMP to modernize the existing licensing process, their commitment to LMP gave the NRC the mistaken impression that it should also serve as a functional template for the new Part 53 framework that is currently being developed. Unraveling that misperception has taken a lot of effort from the Breakthrough Institute and many other organizations.

Success in the absence of a new licensing framework — the creation of which Congress mandated in 2019 — will require cooperation among developers and from the NRC. For developers, this means knowledge sharing their successes and failures with one another as they move forward with their respective DCAs. For the NRC, this means accommodating these custom approaches to the LMP process. It also means doing away with overly prescriptive requirements that are irrelevant to new reactor designs. With the lessons of each success and failure, industry and the NRC can hopefully chip away at the barriers to arrive at a more efficient licensing process.