ML25024A043

Public Enclosure - Terrapower, LLC - Summary Report on the Regulatory Audit of Topical Report, TP-LIC-RPT-0005, Radiological Release Consequences Methodology Topical Report, Revision 0

ML25024A043
Person / Time
Site:
Issue date:	01/23/2025
From:	NRC/NRR/DANU
To:
References
EPID L-2023-TOP-0055 TP-LIC-RPT-0005
Download: ML25024A043 (1)

Text

OFFICIAL USE ONLY - PROPRIETARY INFORMATION Enclosure OFFICIAL USE ONLY - PROPRIETARY INFORMATION TERRAPOWER, LLC. - AUDIT

SUMMARY

REPORT, TOPICAL REPORT TP-LIC-RPT-0005, RADIOLOGICAL RELEASE CONSEQUENCES METHODOLOGY, REVISION 0 Applicant:

TerraPower, LLC Applicant Address:

15800 Northup Way, Bellevue, WA 98008 Plant Name:

Natrium Project No.:

99902100

1.0 BACKGROUND

By letter dated November 6, 2023, TerraPower, LLC (TerraPower) submitted topical report (TR)

TP-LIC-RPT-0005, Radiological Release Consequences Methodology Topical Report, Revision 0 (Agencywide Documents Access and Management System (ADAMS) Accession No. ML23311A139) to the U.S. Nuclear Regulatory Commission (NRC) staff for review. The TR provides three evaluation models (EMs) that will be used to determine the consequences of radiological release source terms developed for the Natrium sodium fast reactor design according to an interfacing TR, NAT-9392, Radiological Source Term Methodology Report, Revision 0 (ML24261B944) also under review by the NRC staff. On December 18, 2023, the NRC staff found that the material presented in TP-LIC-RPT-0005 provides technical information in sufficient detail to enable the NRC staff to conduct a detailed technical review (ML23333A070).

TerraPower requested the NRC staffs approval of the EMs presented in this TR for use by applicants using the Natrium design. TerraPowers overall licensing approach for the Natrium design follows the Licensing Modernization Project (LMP) methodology described in Nuclear Energy Institute (NEI) 18-04, Revision 1, Risk-Informed Performance-Based Technology Inclusive Guidance for Non-Light Water Reactor Licensing Basis Development (ML19241A472). Regulatory Guide (RG) 1.233, Guidance for a Technology-Inclusive, Risk-Informed, and Performance-Based Methodology to Inform the Licensing Basis and Content of Applications for Licenses, Certifications, and Approvals for Non-Light Water Reactors, Revision0 (ML20091L698) endorses the LMP methodology described in NEI 18-04.

By letter dated April 24, 2024, the NRC staff provided its audit plan for the subject TR to TerraPower (ML24103A224). The audit was conducted virtually from May 7, 2024, through June 27, 2024, using TerraPowers electronic reading room (ERR). The NRC staff held an audit exit meeting with TerraPower on June 27, 2024. By letter dated July 26, 2024, TerraPower submitted a revision to the TR, renumbered as NAT-9391, Radiological Release Consequence Methodology Topical Report, Revision 0 (ML24208A181) to address the results of the audit, as summarized below.

2.0 AUDIT REGULATORY BASES The basis for the audit includes the following regulatory requirements:

Title 10 of the Code of Federal Regulations (10 CFR) 50.33(g)(2) with respect to radiological consequence analyses used to determine the size of the plume exposure

OFFICIAL USE ONLY - PROPRIETARY INFORMATION OFFICIAL USE ONLY - PROPRIETARY INFORMATION pathway emergency planning zone for applicants adopting 10 CFR 50.160, Emergency preparedness for small modular reactors, non-light-water reactors, and non-power production or utilization facilities.

10 CFR 50.34(a) with respect to the following minimum information to be included in a preliminary safety analysis report submitted as part of a construction permit application:

o 10 CFR 50.34(a)(1)(ii)(D)(1): An individual located at any point on the boundary of the exclusion area for any 2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> period following the onset of the postulated fission product release, would not receive a radiation dose in excess of 25 rem total effective dose equivalent (TEDE).

o 10 CFR 50.34(a)(1)(ii)(D)(2): An individual located at any point on the outer boundary of the low population zone, who is exposed to the radioactive cloud resulting from the postulated fission product release (during the entire period of its passage) would not receive a radiation dose in excess of 25 rem [TEDE].

o 10 CFR 50.34(a)(4) requires a preliminary analysis and evaluation of the design and performance of structures, systems, and components of the facility with the objective of assessing the risk to public health and safety resulting from operation of the facility and including determination of the margins of safety during normal operations and transient conditions anticipated during the life of the facility, and the adequacy of structures, systems, and components provided for the prevention of accidents and the mitigation of the consequences of accidents.

3.0 AUDIT PURPOSE AND OBJECTIVES The purpose of the audit is for the NRC staff to gain a more detailed understanding of TerraPowers EMs presented in the TR. A secondary purpose of the audit is to identify any information that will require docketing to support the NRC staffs safety evaluation.

4.0 SCOPE OF THE AUDIT AND AUDIT ACTIVITIES The audit followed the guidance in the Office of Nuclear Reactor Regulation Office Instruction LIC-111, Regulatory Audits, Revision 1 (ML19226A274).

Members of the audit team included the NRC staff listed below.

Roel Brusselmans, Project Manager, Audit Manager Michelle Hart, Senior Reactor Engineer, Audit Lead Kyle Clavier, Contracted Technical Contributor Keith Compton, Senior Reactor Scientist Stephanie Devlin-Gill, Senior Project Manager Zach Gran, Reactor Scientist (Severe Accident)

The participants from TerraPower for this audit were Patrick Donnelly, Gabrielle Schreier, Matthew Presson, Maria Pfeffer, Timothy Enfinger, Nick Kellenberger, Eric Pearson, and Shane Gardner.

OFFICIAL USE ONLY - PROPRIETARY INFORMATION OFFICIAL USE ONLY - PROPRIETARY INFORMATION In the audit exit meeting with TerraPower, the NRC staff summarized the audit purpose, activities, and high-level results. The NRC staff did not acquire any documents during the audit.

The NRC staff reviewed the following documents during the audit using the TerraPower ERR:

Global Nuclear Fuel, DBR-0073146, Design Basis Record Design Notes: Shine Dose Conversion Factors Supporting Design Basis Accident Radiological Consequence Analyses for the Natrium Design," Revision 0, July 2023 Global Nuclear Fuel, DBR-0073369, Design Basis Record Design Notes: Natrium Radiological Consequence Analyses - Licensing Basis Events (LBE) Isotope Sensitivity," Revision 0, August 2023 Global Nuclear Fuel, RRCAT01P, Software User Manual: Engineering Software Controlled Location, PLM 007N5941, Revision 1, August 2023 NAT-5881, Revision 1, Radiological Consequences LBE Analysis for Natrium

[Preliminary Safety Analysis Report/Construction Permit Application (PSAR/CPA)]

RRCAT01P, Software User Manual, Revision 1 5.0

SUMMARY

OF OBSERVATIONS As stated in the NRC staffs audit plan, the audit was focused on specific inquiries pertaining to the content of the TR. The NRC staff reviewed information through the TerraPower ERR and held discussions with TerraPower staff to understand and resolve questions. The table below replicates the transmitted audit questions and summarizes their resolution.

OFFICIAL USE ONLY - PROPRIETARY INFORMATION OFFICIAL USE ONLY - PROPRIETARY INFORMATION Number Question Resolution of Question 1

Regarding the radionuclide release characterization described in TR section 3.5.1, Isotope Sensitivity Method, provide clarification on the following topics:

a. How the isotope sensitivity analysis aggregates across exposure pathways and how it accounts for exposure duration.

b. With respect to the radionuclides which were screened out based on ((

)), what is meant by ((

))

for a radionuclide? Does this suggest that the MELCOR Accident Consequence Code System (MACCS) dose ((

))

files did not include the values, or that they are not available in general?

c. The TR does not provide a list of radionuclides that were screened out due to ((

)), so it is unclear if any of the radionuclides that would contribute significantly to individual dose were screened out (e.g., novel radionuclides ((

)) that are not included in FGR13GyEquiv_RevA.inp file in MACCS). How was it assessed that those radionuclides which were screened out would not contribute significantly to the dose result?

1a. TerraPower indicated that dose sensitivity, as defined in TR equation 3-1, does not include a term for exposure duration and that the dose sensitivity is computed assuming an equal exposure duration for all isotopes. TerraPower also indicated that a range of radionuclide decay times are considered in the core inventories. TerraPower stated that it will revise TR section 3.5.1.2, General Isotope Sensitivity Results, to clarify that the exposure pathways are equally weighted when sensitivity results are aggregated; specifically, TerraPower indicated it will state, ((

)) The TR revision, NAT-9391, contained the edit that TerraPower stated it would implement. TerraPower also clarified that TR equation 3-1 is repeatedly evaluated based on the time-dependent isotope inventory available.

1b. TerraPower stated that DCFs from Federal Guidance Report (FGR) 11, Limiting Values of Radionuclide Intake and Air Concentration and Dose Conversion Factors for Inhalation, Submersion, and Ingestion, and FGR 12, External Exposure to Radionuclides in Air, Water, and Soil, are used consistent with the TEDE and the methodology described in TR section 3.5.1. With regard to the radionuclides screened out, TerraPower stated that the TR will be revised to say: ((

)) The TR revision contained the edit that TerraPower stated it would implement.

OFFICIAL USE ONLY - PROPRIETARY INFORMATION OFFICIAL USE ONLY - PROPRIETARY INFORMATION Number Question Resolution of Question 1c. TerraPower stated that the methodology uses DCFs from FGRs 11 and 12, which the NRC staff stated in RG 1.183, "Alternative Radiological Source Terms for Evaluating Design Basis Accidents [(DBA)] at Nuclear Power Reactors, Revision 1 (ML23082A305) are acceptable for the calculation of TEDE. In addition, TerraPower stated that it ((

))

The NRC staff audited DBR-0073369, Design Basis Record Design Notes: Natrium Radiological Consequence Analyses - Licensing Basis Events (LBE)

Isotope Sensitivity, Revision 0, to confirm the analysis was consistent with the description of the isotope sensitivity method in the TR. The review of DBR-0073369 enhanced NRC staff understanding of the TR methodology, including the determination of the set of radionuclides applicable to any LBE source term for the Natrium reactor.

2 Regarding TR section 3.5.2, Adaptive Plume Algorithm, clarify whether the adaptive plume algorithm may result in plume segments significantly greater than one hour in duration. The NRC staff notes that the use of plume segments significantly longer than 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> may affect the choice of appropriate plume meander model used in MACCS. Does the LBE EM take this into consideration to ensure that the releases to environment are appropriately modeled?

TerraPower stated that the adaptive plume algorithm may result in plume segment durations that range from 1 minute up to 1 day, which are the limits of the allowed range in the MACCS code. TerraPower also stated that there are no limitations on release durations noted in either the MACCS user guide or in the Ramsdell and Fosmire plume meander model documentation, which is used in the TR LBE EM.

Furthermore, TerraPower communicated that there are

OFFICIAL USE ONLY - PROPRIETARY INFORMATION OFFICIAL USE ONLY - PROPRIETARY INFORMATION Number Question Resolution of Question no terms in the Ramsdell and Fosmire model that are based on plume duration.

The NRC staff and TerraPower discussed modeling considerations related to using relatively long plume durations, including the interaction with the building wake model and wind shifts that could affect the plume centerline concentration and subsequent dose estimates.

To further address NRC concerns related to impact of plume duration, TerraPower provided ERR access to NAT-5881, Revision 1, Radiological Consequences LBE Analysis for Natrium PSAR/CPA, which included a sensitivity study on the number of plumes modeled through the adaptive plume algorithm. The NRC staff observed a low sensitivity in the dose results to the number of plume segments in the analysis.

3 Regarding the LBE EM site characterization as described throughout TR section 3.6, MACCS Model Parameters, provide information to address the following topics:

a. Clarify the effect of uniform population distribution on population-weighted outputs. For example, a population concentrated close to the site would, all else equal, result in a higher population-weighted 10-mile risk than a uniform population over the 10-mile radius.

b. The analysis area (e.g., 10-mile zone, 50-mile zone) does not appear to be specified in the MACCS model; provide the analysis area used for each of the 3a. TerraPower stated that while it is expected that higher populations would be located further from the plant, making the uniform population model conservative, this assumption will be confirmed.

TerraPower stated it would update TR section 3.6.4, Plant Area, to clarify that the population will only be modeled as uniform after it is confirmed that this is conservatively representative for the plant location considered. TerraPower also stated that it would revise the TR to specify that the population distribution for the plant location may also be used in the methodology.

The TR revision contained the edit that TerraPower stated it would implement.

3b. TerraPower stated that it would revise TR section 3.6.4 to add additional information to specify that the boundary of the computational model will be larger than

OFFICIAL USE ONLY - PROPRIETARY INFORMATION OFFICIAL USE ONLY - PROPRIETARY INFORMATION Number Question Resolution of Question consequence metrics under consideration.

the analysis area to prevent model boundary conditions from influencing analysis results. The TR revision contained the edit that TerraPower stated it would implement.

4 Regarding the LBE EM meteorological data described in TR section 3.6.1, Data File Specifications, provide information to address the following topics:

a. Clarify whether the TR would apply only before site selection or whether it could be used after site selection, and whether this would affect the source of meteorological data.

b. Clarify how the generic meteorological data from Electric Power Research Institute report 3002003129, Advanced Nuclear Technology: Advanced Light Water Reactor Utility Requirements Document (URD), would be shown to be appropriate for a specific site.

c. Describe whether the URD generic meteorological data provides sufficient meteorological information to successfully complete a MACCS code calculation with respect to the atmospheric dispersion model (e.g., at least one full annual cycle of hourly meteorological data that are representative of long-term meteorological conditions of the site).

Clarify how the quality and completeness of meteorological data is assessed for the URD data. The NRC staff notes that RG 1.23, Meteorological Monitoring Programs for Nuclear Power Plants, 4a. TerraPower stated that the TR methodology may be applied either before or after site selection, including use of the generic meteorological data from the URD, considering it is shown to be conservative in place of site-specific data.

4b. TerraPower stated that the details of the comparison of the URD data to the site will be deferred to the analysis which applies the TR methodology.

TerraPower also clarified that the TR methodology will be implemented with site-specific weather data separately from the URD data and the calculated dose results compared to determine the bounding conditions.

TerraPower stated it would revise the TR to clarify the intent of this comparison. The TR revision contained the edit that TerraPower stated it would implement.

4c. TerraPower described the URD generic data as including 8,760 hourly meteorological data points, as well as seasonal morning and afternoon mixing height, and is intended to be representative of an 80th percentile bounding site. TerraPower stated that the data was reviewed by the NRC in Annex B of NUREG-1242, Volume 2, Part 1, NRC Review of Electric Power Research Institute's Advanced Light Water Reactor Utility Requirements Document (ML100430013) and that it determined that the data provides sufficient information to support state-of-practice analysis (e.g., see ML14100A187, section 19.1.4.3.1).

OFFICIAL USE ONLY - PROPRIETARY INFORMATION OFFICIAL USE ONLY - PROPRIETARY INFORMATION Number Question Resolution of Question Revision 1 (ML070350028), provides one acceptable approach.

5 Regarding the LBE EM atmospheric transport and diffusion, as described in TR sections 3.6.2, Atmospheric Dispersion, and 3.6.3, Plume Deposition, provide information to address the following topics:

a. Clarify how uncertainty in dispersion coefficients can be assessed if a lookup table is used.

b. Clarify whether a user of this methodology is directed to use their own plume buoyancy or the given default constant value for a low sensible heat release rate.

c. Clarify the technical basis (e.g., primary reference) for the resuspension parameters.

5a. TerraPower stated that the diffusion coefficients specified in the LBE EM are the Eimutis and Konicek (E&K) parameterization of the Pasquill-Gifford diffusion curves, as typically used in dispersion models acceptable to the NRC (such as NRCs codes XOQDOQ and PAVAN). TerraPower clarified that the E&K parameterization is input to the MACCS code via a lookup table because the functional form of the parameterization includes a non-zero constant term that conflicts with the power law model in MACCS, as noted in appendix D of NUREG/CR-7270, Technical Bases for Consequence Analyses Using [MACCS],

(ML22294A091). TerraPower further stated that it determined that use of the E&K parameterization without explicit additional uncertainty assessment was appropriate since it is consistent with implementation in the codes stated above, as well as technical guidance.

5b. TerraPower stated that the methodology directs to user to model plume buoyancy as a function of sensible heat (i.e., select the Power Model Plume Rise option),

while the default constant value is provided only as an alternative if the sensible heat for a given source term is not available. TerraPower stated it would revise the TR to add clarity.

5c. TerraPower stated that the relevant references will be added to TR section 3.6.3.

6 Regarding the LBE EM modeling of protective actions, as described in TR section 3.6.5, Protective Actions, provide information to address the following topics:

6a. TerraPower stated that it would revise the TR to include consideration of long-term protective actions in the evaluation of the individual latent cancer risk by

OFFICIAL USE ONLY - PROPRIETARY INFORMATION OFFICIAL USE ONLY - PROPRIETARY INFORMATION Number Question Resolution of Question

a. Clarify how the individual latent cancer risk will be computed without consideration of long-term protective actions.

b. Clarify how dose reduction factors associated with occupancy of structures or vehicles are developed and applied.

including the use of the MACCS CHRONC module, including description of the durations of the intermediate and long-term phases and assumptions related to protective actions such as dose criteria for protective actions, and modeling of groundshine and weathering. TerraPower stated that inputs are taken from guidance in NUREG/CR-7270 and are consistent with Environmental Protection Agencys Protective Action Guides (PAGs). TR section 3.4, Computational Model, will be revised to state that CHRONC module will be used, but not the food model option for ingestion.

6b. TerraPower stated that the dose reduction factors will be described and their basis documented in the analysis supporting a license application, not in the TR methodology. The analysis will be available for audit during the review of the application referencing this TR.

TerraPower stated that the values would generally be consistent with NUREG/CR-7270. The NRC staff noted that when assigning dose reduction factors, there would be a difference in the values when modeling protective actions.

7 Regarding the LBE EM dosimetry modeling and use of MACCS, provide information to address the following topics:

a. Clarify how the MACCS parameter L-ICRP60ED relates to the TEDE. TR section 3.6.1 states that DCFs from the FGR13GyEquiv_RevA.inp file supplied with MACCS are used which reflects total effective dose (TED) based on Federal Guidance Report No. 13: Cancer Risk Coefficients for Environmental Exposure 7a. TerraPower stated that it would revise the TR to indicate that the methodology will use the newer DCF file that calculates TEDE.

7b. TerraPower stated that the TR will be revised to add the specification to set ENDEMP to 30 days. No additional time will be added, considering the discussion in the MACCS user manual which implies that the time for transit of radionuclides is already accounted for in that the time starts once the first plume segment arrives at the dose location. The TR revision

OFFICIAL USE ONLY - PROPRIETARY INFORMATION OFFICIAL USE ONLY - PROPRIETARY INFORMATION Number Question Resolution of Question to Radionuclides, (FGR-13). The NRC staff notes that the MACCS DCF file has recently been updated so both TED and TEDE, as well as FGR-13 dose coefficients, can be used from the same DCF file.

b. Clarify how MACCS will be configured to estimate a 30-day dose (e.g., Will MACCS parameter ENDEMP be set to 30 days? Will additional time be added to ENDEMP to allow for transit of radionuclides?)

c. Clarify the exposure duration for MACCS parameter PWILCF (related to individual latent cancer risk) and how that relates to typical practice for comparison to the quantitative health objectives.

contained the edit that TerraPower stated it would implement.

7c. TerraPower stated that the term PWILCF, understood to represent the exposure duration used to calculate the population-weighted risk of latent cancer fatality, will include long-term consequences.

Consideration of long-term consequences over a 50-year long-term phase is consistent with evaluations in State-of-the-Art Reactor Consequence Analysis (SOARCA)1 and will be accomplished through use of the MACCS CHRONC module.

8 The NRC staff notes that TR section 3.6.6, Dosimetry, states dose is computed with the [linear no threshold (LNT)] model. Though dose is an input to the LNT model, the LNT model applies to health effects and not dose. Clarify whether the statement dose is computed with the LNT model in the TR could be phrased as, cancer fatality risk is computed with the LNT model.

TerraPower agreed with the proposed clarification and stated it would revise the TR accordingly. The TR revision contained the clarification that TerraPower stated it would implement.

9 Regarding the LBE EM treatment of uncertainty and conditional consequence quantification, provide information to address the following topics:

9a. TerraPower stated that the TR methodology does not provide specific sources of parameter distributions.

Which parameters are uncertain, the treatment of that uncertainty, and the distributions are all items that will be addressed and documented in the analyses which apply the TR methodology. TerraPower stated it would 1 Additional information on the NRC SOARCA project is available on the NRC public website at https://www.nrc.gov/about-nrc/regulatory/research/soar/overview.html

OFFICIAL USE ONLY - PROPRIETARY INFORMATION OFFICIAL USE ONLY - PROPRIETARY INFORMATION Number Question Resolution of Question

a. Clarify the source of parameter distributions to be used in sensitivity and uncertainty analyses.

b. Clarify how model uncertainty will be addressed.

c. Clarify how presentation of meteorological variability (used to estimate the mean, 5th, and 95th percentiles) will be combined with the estimation of weather uncertainty.

revise the TR to recommend that NUREG/CR-7270 be used as a basis for parameter distributions. The TR revision contained the edit that TerraPower stated it would implement.

9b. TerraPower stated that the TR methodology does not address model uncertainty and that it would be addressed in the analysis which applies the TR methodology. TerraPower further stated that all aspects of the computational model were constructed to appropriately represent all phenomena considered in the LBE EM.

9c. TerraPower described that the meteorological variability combines with the estimation of weather uncertainty. The variability of weather within a given year is captured by reporting the mean, 5th, and 95th percentile results. The uncertainty in the weather would be addressed by considering the most limiting year of site-specific meteorological data. TerraPower stated it would revise item 3.7 in TR table 3-1, Listing of Licensing basis Event Evaluation Model Inputs, to add this restriction. TerraPower further clarified that if a meteorological data file based on the URD is used instead of site-specific data, it must be shown to be conservatively bounding of the limiting year of data specific to the plant location.

10 The NRC staff requests the documentation related to the RRCAT computer code be made available for audit (e.g., user manual and other documents that provide insight into the calculations that are done by this computer code). This information will assist the NRC staff in better understanding how the RRCAT computer code follows the applicable consequence analysis TerraPower provided access to the RRCAT user manual in the ERR. TerraPower stated that the RRCAT computer code follows applicable consequence analysis models contained in RG 1.183 and noted that the code was developed prior to the issuance of RG 1.183, Revision 1. The NRC staffs audit of the documentation provided clarification of the code

OFFICIAL USE ONLY - PROPRIETARY INFORMATION OFFICIAL USE ONLY - PROPRIETARY INFORMATION Number Question Resolution of Question models contained in RG 1.183 and the RADTRAD consequence analysis code.

capability and required user inputs and aided in understanding of the TR methodology and the DBA EM.

11 The NRC staff requests that example calculations, or other resources, be made available for audit to assist the NRC staff in understanding how the shine dose conversion factor (SDCF) calculation is being implemented in this EM. The supporting ((

))

code input and output files and supporting technical report information would assist the NRC staff understand how the proposed SDCF is being implemented.

TerraPower provided access to DBR-0073146, Revision 0, Design Basis Record Design Notes: Shine Dose Conversion Factors Supporting Design Basis Accident Radiological Consequence Analyses for the Natrium Design. TerraPower stated that this analysis was performed without crediting the shielding from radiation provided by the control room structure, therefore, it results in overly conservative SDCFs and is not expected to be used in design and licensing.

TerraPower also noted that the analysis includes significantly more detail than presented in TR section 5.5, Shine Dose. The NRC staffs audit of the analysis aided in understanding of the intent of the development of SDCFs and calculation of the SDCF values. The NRC staff observed that the analysis used reasonable assumptions and methods and was consistent with the description in TR section 5.5.

OFFICIAL USE ONLY - PROPRIETARY INFORMATION OFFICIAL USE ONLY - PROPRIETARY INFORMATION 6.0 REQUESTS FOR ADDITIONAL INFORMATION RESULTING FROM AUDIT As a result of the audit, the NRC staff did not identify any requests for additional information related to this TR. However, TerraPower revised the TR to address audit items.

7.0 OPEN ITEMS AND PROPOSED CLOSURE PATHS There are no open items resulting from this audit.