ML21337A138

M211209: Slides - Briefing on 10 CFR Part 53 Licensing and Regulation of Advanced Nuclear Reactors - E Lyman Slides

ML21337A138
Person / Time
Issue date:	12/03/2021
From:	NRC/OCM
To:
Shared Package
ML21301A174	List: ML21337A124 ML21337A136 ML21337A137 ML21337A138 ML21337A139 ML21337A140 ML21337A141 ML21337A142 ML21343A023 ML21349A911
References
M211209
Download: ML21337A138 (11)

Text

UCS Perspectives on Part 53 Dr. Edwin Lyman Director of Nuclear Power Safety Climate and Energy Program Union of Concerned Scientists December 9, 2021

The foundation of a new reactor licensing rule

• A new reactor licensing rule should

- Clearly provide for levels of safety, security, and environmental protection at least equivalent to the operating fleet

• Or preferably, greater levels than the operating fleet

- Employ risk-informed provisions only to the extent that state-of-the-art risk assessment methods allow

- Maintain robust defense-in-depth

- Contain clear requirements for prototype testing

- Reject provisions that could result in disproportionate outcomes to disadvantaged communities 2

The Part 53 process is broken

• Although UCS appreciates the greater degree of public engagement for developing the draft rule language required in SRM-SECY-0020-0032, the intermittent and iterative process has been ineffective and inefficient, and the high degree of industry involvement threatens to compromise the NRCs regulatory independence

- What was intended to be a flexible, technology-inclusive approach has split into at least three alternatives

• Following the next release of draft language, UCS recommends that a more conventional notice-and-comment rulemaking process be restored, and that the NRC consider developing a regulatory basis document 3

Part 53 should be scrapped

• Development of a technology-inclusive rule does not require starting from scratch and rewriting the entire rule book

• The population of new reactor designs that the NRC is likely to receive for review for many years to come is not actually that diverse

- Coolants: liquid-metal, gas, molten salt

- Fuels: metal, TRISO, molten salt

- Size: very small (<20 MWe), small (<300 MWe); medium (<600 MWe)

• The accident spectrum is fairly well-characterized for many of these design categories 4

Alternative approach

• Rather than a new part, NEIMAs intent can be addressed through a series of appendices to Parts 50/52

- Coolant-specific, size-specific, and fuel specific

• Each appendix specifies

- All regulations in Parts 50/52 that are not applicable

- Design-specific alternatives for inapplicable requirements that correspond to a safety or security function (e.g. 50.46)

• Anticipated operational occurrences, design-basis accidents, and severe accidents determined through a structured process such as the Licensing Modernization Project, with sufficient error bars on frequency estimates to adequately encompass uncertainties

• Should be complemented with a requirement to conduct systematic searches for hazards, initiating events, and accident scenarios, as recommended by ACRS 5

Core safety approach

• To ensure a clear correspondence with requirements for the operating fleet, fundamental defense-in-depth principles for AOOs and DBAs should be maintained on an equivalent level, including safety-related SSCs and the single-failure criterion

• Severe accident risks should not be greater than those for the operating fleet (more on this later) 6

Alternative approach

• Part 53 will duplicateor may even be inconsistent withother rulemakings taking place for new reactors

- Limited-scope security

- EP requirements

• The risk-informed licensing approach should focus first and foremost on the design and construction of the plant

- Requirements covered by other parts of the regulations should remain where they are

- risk-informed voluntary alternatives should be available once designs have been approved (or demonstrated)

• A structured approach is needed for application of analytical safety margins to operating flexibilities to ensure sufficient layers of defense-in-depth are maintained 7

Severe accidents and the use of PRA

• There is no credible way to risk-inform licensing without a PRA other than providing defense-in-depth measures with unquantified risk benefitsi.e. a deterministic approach

• There is no plausible way to define a maximum credible accident without a PRA (maximum credible implies a likelihood threshold) 8

Role of Quantitative Health Objectives in licensing

• The QHOs are not adequate metrics for incorporation as regulatory requirements

- Magnitude is too high

• latent fatality QHO corresponds to CDFs from 5x10-4 for an open containment (NUREG-1860) to 4x10-2 for a frequency-weighted average risk (2018 EPRI Margins study), compared to an average of 5x10-5 for the operating fleet

- Do not include societal risks (land contamination)

- Are based on population-averaged radiological risks that are insensitive to the disproportionate effects of ionizing radiation on disadvantaged populations such as Black people 9

Conclusion

• The 9-month extension in the Part 53 schedule should provide some room for the NRC to reconsider the direction of the rulemaking

• UCS recommends a more modest approach that preserves the foundational principles underlying the current rules, while providing for design-specific 10

Acronyms

• AOO: Anticipated Operational Occurrence

• CDF: Core Damage Frequency

• DBA: Design-Basis Accident

• PRA: Probabilistic Risk Assessment

• TRISO: Tristructural Isotropic

• QHOs: Quantitative Health Objectives

• SSCs: Structures, Systems, and Components

• UCS: Union of Concerned Scientists 11