ML23296A083

Draft NRC Staff Evaluation of PNNL Methodology for Project Pele

ML23296A083
Person / Time
Site:	07109396
Issue date:	11/07/2023
From:	Bernie White Storage and Transportation Licensing Branch
To:
Bernie White, NMSS/DFM
References
Download: ML23296A083 (16)

Text

THIS NRC STAFF EVALUATION HAS BEEN PREPARED AND IS BEING RELEASED TO SUPPORT INTERACTIONS WITH THE ADVISORY COMMITTEE ON REACTOR SAFEGUARDS (ACRS). THIS PAPER HAS NOT BEEN SUBJECT TO NRC MANAGEMENT AND LEGAL REVIEWS AND APPROVALS, AND ITS CONTENTS SHOULD NOT BE INTERPRETED AS OFFICIAL AGENCY POSITIONS.

Methodology Evaluation Docket No. 71-9396 Project Pele Risk-Informed Methodology Development and Demonstration of a Risk Assessment Approach for Approval of a Transportation Package of a Transportable Nuclear Power Plant for Domestic Highway Shipment Summary By request dated February 20, 2023 (Agencywide Documents Access and Management System (ADAMS) Accession No. ML23066A202), as supplemented on September 18, 2023 (ML23268A328), on behalf of the Strategic Capabilities Office (SCO) within the U.S. Department of Defense, the Pacific Northwest National Laboratory (PNNL) requested that the U.S. Nuclear Regulatory Commission (NRC) review PNNLs document titled Development and Demonstration of a Risk Assessment Approach for Approval of a Transportation Package of a Transportable Nuclear Power Plant [TNPP] for Domestic Highway Shipment, hereafter referred to as the Methodology. PNNL developed the Methodology in the event that the Project Pele demonstration reactor cannot be shown to meet all the requirements in Title 10 of the Code of Federal Regulations (10 CFR) Part 71, Packaging and Transportation of Radioactive Material, for hypothetical accident conditions.

The demonstration transportable micro-reactor design consists of multiple modules including a reactor module and three other modules, an intermediate heat exchange module, a control module, and a power conversion system module, to support the micro-reactor. The four modules will be transported and contained in separate International Organization for Standardization (ISO)-compliant container express (CONEX1)-like boxes having dimensions of about 8 feet wide by 8 feet high by 20 ft long. The Methodology evaluated transport of the micro-reactor module and not the ancillary equipment needed to support the micro-reactor. The reactor module in its configuration contained in the CONEX-like box has been termed by PNNL as the TNPP and it is considered by the NRC staff as a transportable micro-reactor.

Because there is no specific guidance for the NRC staff to review a risk-informed methodology for an application for a transport package, the NRC staff reviewed the Methodology considering, where relevant, existing regulatory approaches and methods to evaluate the risk of transportation accident scenarios2 discussed in the following documents:

• Title 10 of the Code of Federal Regulations (10 CFR) Part 71, Packaging and Transportation of Radioactive Material

• Regulatory Guide 1.200, Acceptability of Probabilistic Risk Assessment Results for Risk-Informed Activities (ML20238B871) 1 The Project Pele modules will be transported in a custom-developed International Organization for Standardization container which resembles a CONEX box.

2 An accident scenario is one that leads to unexpected or unintended exposures. Exposures from normal conditions of transport are not considered to be unexpected or unintended exposures.

Enclosure 1

• Specific Safety Guide No. SSG-26 (Revision 1), Advisory Material for the IAEA

[International Atomic Energy Agency] Regulations for the Safe Transport of Radioactive Material (2018 Edition)

• Integrated safety analyses in NUREG-1520, Revision 2, Standard Review Plan for Fuel Cycle Facilities License Applications (ML15176A258)

• NUREG-2216, Standard Review Plan for Transportation Packages for Spent Fuel and Radioactive Material (ML20234A651)

Based on the statements and representations in the SCO request, as supplemented, the staff reviewed whether the Methodology provides a sufficient basis for incorporation in an application for package approval to risk-inform the demonstration that public health and safety is protected during the transportation of the micro-reactor for Project Pele. The NRC will review how the Methodology is applied in the application for Project Peles micro-reactor package to demonstrate compliance with the requirements in 10 CFR Part 71. As such, the NRC staff reviewed PNNLs proposed approach to determine the risk of transporting the microreactor.

The probabilistic risk assessment (PRA) results are not based on an actual route for transport or an engineering evaluation of the final design attributes of the Project Pele micro-reactor but uses engineering judgment of package damage and release of radioactive material. The NRC's endorsement of the Methodology applies only to the approach to perform the calculations in the Methodology and as discussed below, does not extend to the numerical assumptions and estimated results.

The Methodology includes a frequency/consequence plot, (F/C plot), and an approach to perform the risk calculations in a future application for package approval. The Methodology includes two F/C plots, one for the maximally exposed individual3 who is a member of the public and another for workers. Hereafter using the term workers includes individuals who are part of the protection program and could receive an occupational dose.

The Methodology provides an approach to perform the calculations to be used as a basis for incorporation in a future application for the approval of a transportation package. The Methodology also provides a logical and structured approach to identifying the safety or risk significance of a transportable micro-reactor and potential compensatory measures to be exercised by the shipper. While the F/C plots show the demarcation between acceptable and unacceptable based on accident frequency and the total effective dose equivalent (TEDE) for members of the public and workers, understanding both the sensitivity of the analysis results and what is credited for defense-in-depth (DID) will play a role in a future package approval.

In addition to development of the F/C plots, each step listed below is an integral part of the Methodology that includes defined relationships among accident scenarios, package evaluation, compensatory measures during shipment, uncertainty analysis, and assessments of DID. The 3

The definition of occupational dose from 10 CFR Part 20 states that it is a dose received by an individual in the course of employment in which the individual's assigned duties involve exposure to radiation or to radioactive material from licensed and unlicensed sources of radiation, whether in the possession of the licensee or other person. Occupational dose does not include doses received from background radiation, from any medical administration the individual has received, from exposure to individuals administered radioactive material and released under § 35.75, from voluntary participation in medical research programs, or as a member of the public.

applicant should perform the evaluations in an iterative fashion as they develop the design and package approval strategies.

A. Development of probabilistic risk assessment for package transport

1. Initiating Event and Accident Sequence Analysis,

i. Initiating events, ii. Accident sequence analysis (Evaluation of accident effects on the package),

2. Source Term Analysis (i.e., Level 2)/how much of what gets released,

i. Sub section: associated particle sizes,

3. Consequence Analysis (i.e., Level 3),

i. Dispersion of any material released to evaluate individual uptake, ii. Determination of TEDE,

4. Human Reliability Analysis,

5. Uncertainty Analysis (including sensitivity analyses), and B. Defense-in-Depth Consistent with the risk-informed, performance-based4 principles in the NRCs approach to regulation and decision-making, approval of a future application will be based on engineering analysis of the final design, risk insights (such as those obtained from the implementation of the Methodology), and engineering judgment. The NRC approval of a package application will depend on the application of each of these steps in the Methodology by the applicant and the assumptions used along with their justification. Therefore, as described below, these activities also define the Methodology as an important element to risk inform the basis that may be used by the applicant for transport package approval with exemptions of the Project Pele application.

The package application requesting exemptions would identify and provide the appropriate level of information needed to satisfy the regulatory requirements for exemption approval as stated in 10 CFR 71.12, Specific exemptions, to show that the transport of the package will not endanger life or property.

Chapter 1 Methodology Overview and Consistency with Commission Guidance The Methodology PNNL developed to establish a risk-informed framework would serve as one element to demonstrate that, after tests for hypothetical accident conditions are evaluated, potential exemptions from the dose rate and containment criteria in 10 CFR Part 71 could be granted while maintaining an acceptable level of safety. The Methodology, as developed by PNNL, serves as a tool to risk inform (1) an application for package approval by a vendor, (2) the design of the relative risk significance of transportable micro-reactor containment and shielding, and (3) the need for transportation compensatory measures. The Methodology includes establishing risk targets which are intended to be consistent with the level of risk from ionizing radiation to workers and the public from both operating nuclear power plants and fuel cycle facilities that is generally found to be acceptable by the NRC.

4 A risk-informed, performance-based regulation is an approach in which risk insights, engineering analysis and judgment (including the principle of defense-in-depth and the incorporation of safety margins), and performance history are used, to (1) focus attention on the most important activities, (2) establish objective criteria for evaluating performance, (3) develop measurable or calculable parameters for monitoring system and licensee performance, (4) provide flexibility to determine how to meet the established performance criteria in a way that will encourage and reward improved outcomes, and (5) focus on the results as the primary basis for regulatory decision-making.

The NRC staff notes that a peer review for the PRA that uses the Methodology in an application for package approval is useful and in certain applications in other licensing actions is an integral part of the NRCs risk-informed reviews. Peer reviews are part of the codes and standards for PRA development as endorsed by Regulatory Guide 1.200 and Regulatory Guide 1.247, TRIAL

- Acceptability of Probabilistic Risk Assessment Results for Non-Light Water Reactor Risk-Informed Activities. The NRC staff also notes that there is no standard for a PRA for transportation package approval or guidance for performing a peer review for this type of PRA, however general principles for conducting a peer review would be applicable. Given that this is a first-of-a-kind application, an independent review of the PRA would support an efficient review by the NRC staff.

Chapter 2 Comparison of the Proposed F/C Plot(s) to Existing Risk Criteria In section 4.0, Safety Goals and Risk Evaluation Guidelines, PNNL notes that risk evaluation guidelines do not exist for transportation of nuclear material as they do for nuclear power plants.

Section 4.1, NRC-Suggested Risk Evaluation Guidelines, discusses the NRC report, Risk-Informed Decision-making for Nuclear Material and Waste Applications (known as the RIDM report) (ML080720238), which proposed quantitative health guidelines (QHGs) to be used for determining the acceptability of risk. These QHGs are based on the quantitative health objectives (QHOs) from the 1986 NRC Safety Goal Policy statement (Volume 51 of the Federal Register (FR), page 30028 (51 FR 30028)) developed for nuclear power plants.

In section 4.2, Development of Risk Evaluation Guideline Surrogates for Safety Goals, PNNL proposes establishing surrogate measures of the QHGs by selecting likelihood-consequence pair limits based on existing guidance from the U.S. Department of Energy (DOE), the NRC, and the International Atomic Energy Agency (IAEA). Section 4.2.1, Risk Evaluation Guidelines Used by DOE for Nuclear Safety, discusses the risk evaluation guidelines used by DOE for nuclear safety. Section 4.2.2, NRC Performance Criteria for the Integrated Safety Analysis of Nuclear Fuel Cycle Facilities, discusses the performance criteria used by the NRC as delineated in subpart H of 10 CFR Part 70, Domestic Licensing of Special Nuclear Material, and NUREG-1520. Section 4.2.3, Risk References in the IAEA Q System, discusses the IAEA Q system described in Appendix I to IAEA SSG-26. Section 4.2.4, NRC-Endorsed Risk-Informed Methodology to Support the Licensing of Advanced Reactor Designs, discusses the risk-informed approach for selecting Licensing Basis Events described in NEI 18-04, Revision 1, Risk-Informed Performance-Based Technology Inclusive Guidance for Non-Light-Water Reactor Licensing Basis Development (ML19241A472), and endorsed by Regulatory Guide 1.233, Revision 0, Guidance for a Technology-Inclusive, Risk-Informed, and Performance-Based Methodology to Inform the Licensing Basis and Content of Applications for Licenses, Certification, and Approvals for Non-Light-Water Reactors (ML20091L698).

Based on information from the above sources, Section 4.2.5.4, Selection of the Likelihood-Consequence Pair Limits for the Surrogate Risk Evaluation Guidelines, proposes likelihood-consequence pair limits with an objective for the proposed limits to be more conservative than the RIDM QHGs. These limits are converted to health effects and compared to the applicable RIDM QHGs in Table 4.6, Comparison of Selected Dose-Consequence Limit Surrogates to the Limiting QHGs.

Finally, Section 4.3, Proposed Surrogate Risk Evaluation Guidelines Established to Meet the Safety Goal QHOs, proposes risk evaluation guidelines against which bounding representative accidents (BRAs) will be compared for acceptability. PNNL suggests that the BRAs will be conservative enough to be compared to the QHGs separately, without summation. The proposed guidelines for the public and workers are shown in Table 4.7, Proposed Radiological Risk Evaluation Guidelines. Additionally, Figure 4.7, Proposed Offsite Public Risk Evaluation Guidelines Chart for Transport of a TNPP Package, shows the proposed guidelines for the public compared to the Frequency-Consequence Line from NEI-18-04. The guidelines for workers are shown in Figure 4.8, Proposed Worker Risk Evaluation Guidelines Chart for Transport of a TNPP Package.

NRC Review The Methodology uses multiple references to inform the development of risk criteria to be used to judge the acceptability of transporting a transportable micro-reactor package and specifically ties the proposed guidelines to the QHGs defined in the RIDM report. Several of these references, including the RIDM report, were developed or approved by the NRC and they provide a reasonable basis for developing and evaluating the proposed risk evaluation guidelines.

Section 4.2.5.4 of the Methodology describes the comparison of the proposed risk evaluation guidelines to the QHGs, which is also summarized in table 4.6. There are limitations to this comparison. The QHGs are intended to be calculated as the risk to an average individual within the relevant population at significant risk from the licensed activity. Conversely, the Methodology suggests calculating the risk to the maximally exposed individual for each BRA. In Section 4.3 the Methodology suggests that the BRAs will be calculated in a manner that is sufficiently conservative that aggregation is unnecessary and would be overly conservative. The potential non-conservatism of not aggregating the BRAs can be limited by not overly subdividing into an excessive number of BRAs. The sub-division performed for the methodology description was reasonable.

The Methodologys comparison to the QHGs is conservative in that it suggests calculating the risk to the maximally exposed individual, rather than averaging the risk over a population. The difference in the risk to the maximally exposed individual and the average individual among some population will depend on many factors including: 1) the area/population averaged over,

2) the amount of dispersion, and 3) protective actions limiting maximum dose (and risk) more than average dose. As noted in Section 4.3, in the case of a transportation package, the population is spread out along a long route. The difference between the maximally exposed individual at 25 meters and the average among a population spread over the entire route would be quite large. The difference may not be as large for workers, depending on how the population is defined. Considering these differences, for the transportation of a transportable micro-reactor package the staff considers that the proposed risk evaluation guidelines are bounded by the QHGs.

The risk criteria in NRC endorsed guidance for fuel cycle facilities (NUREG-1520) and advanced reactor designs (NEI-18-04 endorsed by Regulatory Guide 1.233) both consider sequences separately and doses at the site boundary rather than doses averaged across a population. For the public, the risk evaluation guidelines proposed by the Methodology are less than these criteria, except for a small portion of the proposed guidelines exceeding the NEI 1804 risk line as can be seen in Figure 4.7. For workers, the risk evaluation guidelines proposed by the Methodology are about an order of magnitude lower to the likelihoods identified in NUREG-1520, to the extent that they overlap.

The NRC reviewed PNNLs proposed risk evaluation guidelines and finds them acceptable to be used as an objective means of comparing the likelihood and consequences of different scenarios in relation to the frequency and consequence targets as illustrated by Figures 4.7 and 4.8, because the guidelines are consistent with NRC risk-informed approaches.

Chapter 3 Elements of the Probabilistic Risk Assessment 3.1 Initiating events and accident sequence analysis The Methodology describes the process for identifying hazardous conditions and grouping them into accident scenarios for further analysis.

3.1.1 Initiating events To identify initiating events, the Methodology first defines the safety functions of the TNPP.

Similar to other transportation packages, these are identified in Section 5.2, Identification of TNPP Package Safety Functions, as containment of radiological material, radiation shielding, and maintaining criticality safety. As discussed in Section 5.3.1, Approach to Development of Accidents Scenarios, PNNL expected that the accident scenarios would not be complex, and thus uses hazard analysis to explore a broad range of possibilities rather than complicated system interactions. The results are compared to NUREG-2125, Spent Fuel Transportation Risk Assessment (ML14031A323), to review the comprehensiveness of the process.

The details of the hazard identification are discussed in Section 5.3.2, Identification and Assessment of TNPP Transportation Hazardous Conditions. Subject matter experts filled out hazardous conditions worksheets to generate a catalogue of hazardous conditions (grouped by hazard category and including descriptions) as well as preliminary estimations of frequency, consequence, risk, etc. for each identified hazardous condition. The result of this process is documented in Appendix B, Evaluation of TNPP Package Transportation Hazardous Conditions, of the PNNL report. An applicant for package approval should similarly use a systematic approach to identify initiating events, including a literature review, and document the results of the analysis in the application.

3.1.2 Accident sequence analysis (Evaluation of accident effects on the package)

The Methodology includes an evaluation of accident sequence analysis and the potential for fuel release and dispersion and dose consequences to a worker or bystander. The evaluations in the Methodology do not include engineering analysis to determine damage to the packaging or contents. As discussed in Section 5.3.3.1, Delineation of Accident Scenarios from the Identified Hazardous Conditions, the Methodology initially assigns six accident consequence groups based on expected (not calculated) contributors to the material at risk (MAR). The NRC reviewed the approach PNNL used to determine the MAR; however, the NRC did not review the numerical assumptions and estimated results for the accident consequence groupings. The NRC expects that an applicant using the Methodology would either provide an analysis of the expected MAR for each accident or justify the consequence groupings.

Consistent with the Methodology, for each accident sequence, the applicant would utilize appropriate initial and boundary conditions and evaluate damage to the package based on the accident conditions. These conditions could include the impact energy in the event due to truck speed and any other impacting traffic, the most damaging orientation of the package (unless a specific orientation is the only one possible), and thermal energy input in the case of a fire. The accident sequence evaluation should include a structural evaluation of the package either as it would be prepared for shipment, or in a damaged state if any of the normal conditions of transport affect the package performance during an accident. The results of these structural and thermal evaluations should include damage to the package, including amount of degraded shielding, leak paths for radioactive material, and the amount of damaged fuel, if any, to determine the amount of radioactive material released for the source term analysis.

3.2 Source Term Analysis 3.2.1 Radionuclide Inventory The Methodology includes determining the source term potentially available for release during an accident and the remaining radioactive material to determine TEDE to the maximally exposed individual. The source term and remaining radioactive material can be determined by assessing:

• the radionuclide inventory of the package,

• the MAR, and

• the quantity of material in various components of the package that could be released in an accident as it migrates out of the tri-structural isotropic (TRISO) fuel towards the containment boundary.

The source term analysis by PNNL calculates the radionuclide inventory based on reactor operations. In Section 5.1 of the Methodology, PNNL provides its basis for estimating the radionuclide inventory, which includes the operational characteristics and operating history of the reactor and the cooling time prior to shipment. In appendix A, PNNL provided its estimates of the radionuclide inventory from 30 days to 2 years after shutdown after 3 years of operation.

In Section 5.1.3, Estimated Radiological Inventory, PNNL performed a two-phase screening of radionuclides to identify the radionuclides with non-negligible contributions to dose rates in the event of potential accidents. The first phase screened out all radionuclides whose inventory is less than a millicurie.

For the second phase, PNNL utilized the A2value5 to screen out small quantities of radionuclides with low dose potential. PNNL stated that radionuclides without a corresponding A2 value in Table A-1 in Appendix A to 10 CFR Part 71 were included in the list of radionuclides screened in, if its activity was greater than or equal to 0.001 (0.1%) of the appropriate A2 value from 10 CFR 71, Appendix A Table A-3. This means that the activity of the radionuclide would be greater than 5.4x10-4 Curies for beta or gamma emitting radionuclides, 2.4x10-6 Curies for pure alpha emitters and neutron emitters. Based on its screening analysis of radionuclides at 90-day cooling time, PNNL stated that more than 99.99999 percent of the radionuclide inventory would be included in the consequence analysis.

3.2.2 Material Available for Release In Section 5.1.4.1, Development of Estimates of Material at Risk for Different Accidents, PNNL states the MAR consists of two types. The first is radioactive material that migrates out of the 5

As defined in 10 CFR 71.4, the A2 value means the maximum activity of radioactive material, other than special form material, LSA, and SCO material, permitted in a Type A package. The radionuclides with smaller A2 values have a larger hazard, because by the Q System radionuclides are normalized based on radiological hazard to a human receptor. The derivation of the A2 values in Appendix A to 10 CFR Part 71 is determined using the Q System as discussed in Appendix I to SSG-26.

TRISO fuel during normal reactor operations, anticipated operational occurrences, and potential transportation accidents. The second type of MAR is activation products in the reactor components. Radioactive material released from the TRISO due to a transportation accident will contribute as accident-related MAR. In Section 5.1.4.1, PNNL names the primary (radioactive material released from TRISO) and secondary (activated components) contributors to the MAR, but states that the secondary contributors would be negligible. To determine the MAR, PNNL develops release fractions from the TRISO fuel based on TRISO fabrication errors, and fuel failure in an accident and attenuation factors as material migrates from the TRISO kernel through the fuel compact.

As described in Section 7.1, Source Term Methodology for Transportation Accident Scenarios, PNNL used the following equation from the DOE Handbook DOE-HDBK-3010-94, (Reaffirmed 2013), Airborne Release Fractions/Rates and Respirable Fractions for Nonreactor Nuclear Facilities, to determine the airborne release of the source term:

Source Term = MAR x DR x ARF x RF x LPF Where: MAR = the material at risk, DR = damage ratio, i.e., fraction of MAR affected by stresses due to the accident ARF = airborne release fraction RF = respirable fraction LPF = leak path factor.

PNNL developed the MAR for three regions within the transportable micro-reactor package to account for radioactive material that migrates out of the TRISO and its reduction due to plating out within the package. The three regions include the fuel assemblies (TRISO compacts in inside graphite sleeves), the core (fuel assemblies in coolant channels, moderator blocks and control rods), and the reactor pressure boundary (reactor vessel and primary cooling system).

The Methodology includes diffusion of radioactive material out of the TRISO particle during normal operation and releases from the TRISO due to manufacturing defects. PNNL used the fabrication and failure parameters along with the attenuation factors to develop the TRISO release fractions. PNNL then used the release fractions to calculate the MAR that would be available for release from the package in an accident to contribute to inhalation/ingestion dose and the remaining material to determine the direct dose.

The Methodology contains an estimate of the fraction of material that is respirable based on a particle size with an aerodynamic equivalent diameter of 10 microns or less. For each accident, the Methodology estimates the damage ratio based on the energy in the accident, the physical phenomena that causes the release, and the physical and chemical form of the MAR. While the Methodology provides the estimate of damaged fuel and airborne releases for accident generated stresses for each of the three regions in Tables 7.2 and 7.3, Damage Ratios for the Bounding Represented Accidents, and Combined Airborne Release Fractions and Respirable Fractions for Represented Accidents, respectively, the estimates are not based on structural and thermal evaluations performed on the package and therefore these numerical assumptions were not reviewed by the NRC. The NRC expects that the applicant would either develop these release estimates, along with the leak path factors, based on structural and thermal evaluations on the package due to an accident or otherwise justify its assumed factors.

In Section 7.2, Description of the Source Term for Each Bounding Representative Accident, PNNL estimates the source term (overall release fraction for the MAR) for each accident using the values it provided earlier. For the same reasons discussed related to the development of the factors developed in Section 7.1, the NRC did not review the overall numerical assumptions and estimated results for the source term.

NRC Review Screening radionuclides that contribute less than a millicurie may be appropriate because the smallest A2 value listed in Appendix A to 10 CFR Part 71 is 2.4 millicuries for Actinium-227, any quantity less than a millicurie would be authorized for shipment in a Type A package. Type A packages are self-certified under the Department of Transportation (DOT) regulations and are not required to withstand accidents. In the Q System, for evaluation of a Type A quantity, the entire contents are assumed to be released in an accident and the effective or committed effective dose to an individual would be less than 5 rem after an accident, which for 1 millicurie would be less than 2 rem for a radioisotope with an A2 greater than 2.4 Curies. The NRC staff accepted screening of radionuclides based on A2 values in past applications for package approval when calculating external dose rates from photons and neutrons. However, if the TEDE for an accident is close to a dose threshold, the applicant for package approval should perform a sensitivity analysis on all radionuclides eliminated to show that their contribution is negligible.

The NRC reviewed PNNLs process for determining the quantity of MAR and finds it acceptable.

The process developed by PNNL provides an approach that an applicant for package approval could use to develop the quantity of MAR available for release from the package based on the inventory of material that migrates out of the TRISO core and containment boundary. However, the applicant for package approval should justify its assumptions for radioactive material that is not included in the dose analyses, such as secondary contributors from irradiated reactor components. Additionally, the release fraction approach used in the methodology is consistent with other documented approaches used in estimating releases from containerized spent fuel such as in NUREG-1864 (Dry Cask Storage Risk) and NUREG-2125 (Transportation Package Risk).

3.3 Consequence Analysis PNNL proposed basing the Methodology on the Q System, as described in Appendix I to SSG-26, to determine the dose to an individual. The Q System is used by the IAEA to determine the maximum quantity limit for material in a Type A package and establish leakage rate limits for Type B and Type C packages based on radiation dose consequences for human receptors. The Methodology utilized the same exposure pathways as the Q System because these exposure pathways were agreed upon by the IAEA Member States to be the dominant pathways for the public and workers that could be exposed to radioactive material released during a transportation accident. The Q System exposure pathways include external photon dose; external beta dose; inhalation dose; skin contamination and ingestion dose; and dose from submersion in a cloud.

In the Methodology, consistent with SSG-26, PNNL did not include the ingestion dose and the dose due to submersion in a cloud. SSG-26 does not include the ingestion for a transportation accident because as stated in item I.45 of SSR-26, appendix A, the inhalation pathway will normally be limiting for internal contamination and, therefore, explicit consideration of the ingestion pathway is unnecessary. It is likely that after an accident that contaminates land, either there would be remediation of the land to levels or unrestricted access, or the land would be cordoned off. The submersion doses calculated in Appendix I to SSR-26 are for rapid depressurization indoors, therefore, PNNL does not include the submersion dose because the accident is assumed to take place outside and release from the package would be a puff release.

3.3.1 External Photon Dose For direct photon dose to a worker, PNNL included both the material released that is too large to be inhaled and therefore deposited on the ground, and the remaining radioactive material left in the damaged package. For the external photon dose, PNNL states that it increased the source term in Section 7.2, Description of the Source Term for Each Bounding Representative Accident, by a factor of 100 to account for total material released rather than just the respirable material. PNNL used dose coefficients from SSG-26, which for material released from the package does not account for dispersion and assumes the receptor is 1 meter away from any released material. For external dose to a member of the public, PNNL assumed that the person is located 25 meters from the package compared to 1 m for the worker. This assumption is based on emergency responders isolating the area around the release for 25 meters in all directions, according to the 2020 Emergency Response Guidebook. While the Emergency Response Guidebook does include an isolation distance of at least 25 meters, this is after emergency response personnel arrive on the scene. The applicant using the Methodology should justify the distance to a member of the public as it is possible that a member of the public receives the maximum direct photon dose prior to emergency response personnel cordoning off the area. The Methodology calculates external dose due to unreleased material for an individual standing 1 meter from the package with potentially degraded shielding.

3.3.2 External Beta Dose In Section 7.3.3, External Dose Due to Beta Radiation, PNNL states that the external beta dose in the Methodology, like SSG-26, is based on skin contamination due to material released from the package for a person 1 meter away from the accident. The Methodologys dose coefficients include a range of shielding factors that are dependent on beta energy and based on an absorber thickness of approximately 150 mg/cm2. PNNL did not include explicit calculations of annihilation radiation as it is expected to be a small contribution to potential doses. PNNL included the 0.51 MeV gamma rays in the photon energy per disintegration in the derivation of the photon dose coefficients for the radionuclides. The Methodology includes conversion electrons as monoenergetic beta particles. Similar to the external gamma dose, the dose coefficients do not account for dispersion. In addition, PNNL states that it increased the source term in Section 7.2, Description of the Source Term for Each Bounding Representative Accident, by a factor of 100 to account for total material released rather than just the respirable material.

3.3.3 Inhalation Dose As discussed in Section 7.3.4, Inhalation Dose, the Methodology utilized the estimated airborne source term and a human uptake value of 1x10-3 based on the discussion in item I.36 in appendix I to SSG-26 about determination of a receptor within 10 meters of an outdoor release. SSG-26 estimates that human uptake at 100 meters is increased by a factor of approximately 30 at a 1-meter distance rather than 100 meters from the package. PNNL used this information to develop a scaling function to determine the uptake factor at 25 meters for the public. SSG-26 then converts the uptake to dose to a human receptor based on the effective dose coefficient for inhalation as provided in table II.2 of appendix II in SSG-26, Dose and Dose Rate Coefficients of Radionuclides.

3.3.4 Skin Contamination Dose PNNL proposed calculating the dose due to skin contamination using the source term from Section 7.2 in the Methodology along with the equivalent skin dose rate per unit activity per unit area of the skin found in table II.2 of appendix II in SSG-26. While the Methodology does calculate the dose to the skin, the Methodology presumes that workers handling radioactive material after an accident would wear appropriate protective clothing. Therefore, the Methodology calculates the skin dose in the radiation dose consequences, however PNNL neither included it in the risk results nor used it for comparison with the risk evaluation guidelines.

3.3.5 Neutron Dose The Q System does include neutron dose due to californium-252, however there are addition radioisotopes in spent fuel that decay due to spontaneous fission, such as some plutonium, uranium, and curium radioisotopes. PNNL estimated the dose due to these isotopes and stated that for these other isotopes the dose is dominated by Curium-242 and Curium-244, consistent with studies completed on light-water reactor spent fuel. Based on its assessment, PNNL estimated that the dose due to neutrons is less than one half percent of the dose due to photons. In addition, based on literature review, PNNL estimates that the dose due to alpha reactions causing neutron emission is less than 10 percent of the dose due to spontaneous fission. The NRC staffs experience with evaluating spent fuel packages for light-water reactor fuel is that areas with significant amounts of dense materials such as steel and lead, which may have small amounts of material to stop neutrons, can have a neutron dose rate contribution greater than or equal to the gamma dose rate.

3.3.6 Exposure Pathways Not Addressed by the Q System PNNL proposed not to include pathways excluded in the Q System for exposure to radioactive material. The exposure pathways used in the Q System were judged by the IAEA Member States to be the dominant pathways for the public and worker exposure to radiation resulting from a transportation accident involving radioactive materials. While other exposure pathways (e.g., resuspension, skyshine, drinking water ingestion, etc.) could be evaluated using SSG-26, PNNL stated that they are not expected to be significant exposure pathways for transportation accidents involving irradiated fuel as they would be expected to be mitigated by emergency response to a transportation accident.

3.3.7 Radionuclides without Dose Coefficients in IAEA SSG-26 Based on its screening analyses, PNNL determined that there are six radionuclides for which SSG-26 does not include an effective dose coefficient, Barium-136, Tritium, Yttrium-89m, Promethium 146, Antimony-127, and Terbium-161. Since Barium-136 and Yttrium-89m are decay products of decay of Cesium-136 and Strontium-89, respectfully, they are included in the analysis. As discussed in item I.51 of Appendix I to SSG-26, progeny with a half-life less than 10 days are assumed to be in secular equilibrium with the longer-lived parent; however, the progenys contribution to each Q value calculation is summed with that of the parent. The radionuclides Tritium, Pm-146, Sb-127, and Tb-161, were deemed by PNNL to have a negligible source term and dose contributor.

NRC Review The NRC staff reviewed the proposed consequence calculations in the Methodology but did not review the numerical assumptions and estimated results for the dose calculations because, as discussed above, the effects of an accident on the package were not explicitly evaluated. The NRC staff determined that an applicant using the Methodology for package approval would likely determine all appropriate doses to an individual, whether a worker involved in the shipment or a member of the public, and that the equations used to calculate dose are representative or bounding. However, the staff notes that exclusion of submersion dose is based on the accident occurring outdoors and a puff release from the package. If an accident were to happen that is a prolonged release or take place in a tunnel, then neglecting the submersion dose should be justified.

The NRC notes that the Q System is used by the IAEA to determine A16 and A2 values, along with release limits for Type B packages containing normal form material. The A1 and A2 values in the Q System are based on limiting the dose to 5 rem to an individual, assuming the package releases all of its contents. Similarly, the maximum quantity of material that may be released from a Type B package after the tests for hypothetical accident conditions is also limited to an A2, which may be a small fraction of the material in the package. These limits set by the Q System presume a limited amount of material is available for release based on the maximum dose from the material released by a Type A or Type B package. Since utilizing the Methodology could mean that the package releases more than the allowed limit in 10 CFR 71.51(a)(2) after the tests for hypothetical accident conditions, one cannot assume a priori that exposure pathways that are neglected in the Q System (e.g., resuspension, skyshine, ingestion, etc.) do not provide a significant dose to a member of the public. Therefore, an application for package approval should justify any dose that is not included in the calculation of the TEDE.

The NRC staff also notes that in the Methodology, dose consequences for some scenarios with a frequency less than 5x10-7 per year were not included, however, the package application should provide an analysis of all doses to a worker and member of the public or justify their exclusion.

The NRC staff found that PNNLs document, Development and Demonstration of a Risk Assessment Approach for Approval of a Transportation Package of a Transportable Nuclear Power Plant for Domestic Highway Shipment, outlines a methodology that if followed, with appropriate justifications as discussed above, could provide an approach to demonstrate compliance with the regulations in 10 CFR 71.12. Based on the structural and thermal analyses performed during the accident sequence analysis, as discussed in Section 3.1.2, the Methodology is an acceptable approach to determine that a package application containing exemptions to the tests or criteria for hypothetical accident conditions will not endanger life or property.

Chapter 4 Human Reliability Analysis Unlike operating reactors, there are no active systems associated with the transportable micro-reactor that require a human action to mitigate an accident. In the Methodology, PNNL considers two possible human errors and calculates the probability for each of these two human errors, during reactor disassembly for transport and transportation package preparation. The 6

As defined in 10 CFR 71.4, the A1 value means the maximum activity of special form radioactive material permitted in a Type A package. Special form material is also defined in 10 CFR 71.4.

only other human error would be driver error which would be included in accident rates.

Following a quality assurance program during fabrication and package preparation should reduce the potential for human error. The estimates provided by PNNL appear to provide a conservative estimate of the frequency of human errors that may affect the package during an accident. The application for package approval should describe the human actions that could affect the package performance during transport and determine whether they are sufficiently limited such that reasonable estimates of the human actions do not require rigorous human reliability analysis methods.

Chapter 5 Uncertainty Analysis (including sensitivity analyses)

In Section 10.1, Role of Uncertainty in PRA, the Methodology references relevant discussions of uncertainty in NRC guidance documents such as Regulatory Guide 1.174, Revision 3, An Approach for Using Probabilistic Risk Assessment in Risk-Informed Decisions on Plant-Specific Changes to the Licensing Basis (ML17317A256), Regulatory Guide 1.200, NUREG-1855, Revision 1, Guidance on the Treatment of Uncertainties Associated with PRAs in Risk-Informed Decisionmaking (ML17062A466), and the RIDM report as well as requirements found in national PRA consensus standards. The Methodology addresses model uncertainty and the variability of very large truck (semitrailer) crash data.

PNNL analyzes modeling uncertainty using sensitivity studies which are primarily discussed in Section 9, Sensitivity Study on PRA Modeling Inputs. The first step considers lists of assumptions found in the PRA and determines whether they could be a candidate for a sensitivity study. This step considered assumptions on the hazardous condition evaluation, the accident likelihood determination, the consequences determination, and the compensatory measures to reduce or mitigate risk. In the second step the candidate sensitivities are further screened for feasibility, as some may be difficult to study because of a lack of data. The results of the sensitivity studies are compared to the risk evaluation guidelines, particularly whether the risk conclusions are sensitive to the assumption. Finally, insights from the sensitivity studies are discussed in section 9.3, Insights Gained from the Sensitivity Studies.

The NRC reviewed the steps taken to define, perform, and analyze the sensitivity studies and determined that the process used by PNNL is sufficient to identify, characterize, and understand the impacts of the key sources of model uncertainty.

For parametric uncertainties, PNNL notes that for reactor PRAs, guidance recommends using mean values, which necessitates the development of probability distributions. PNNL states that the development of probability distributions would be difficult for a transportable micro-reactor PRA as it is a first-of-a-kind and the data may not be available to develop the probability distributions for some likelihoods, and it would be burdensome for developing consequences. In Section 10.2.2, Evaluation of the Impact of the Variability in the Very Large Truck Crash Data, PNNL estimates the variability of very large truck crash rates by considering subsets of the available data. While development of full parametric uncertainty distributions is not necessary for an application, the NRC would expect an applicant using the Methodology to examine other key parameters to establish high and low estimates, unless the risk results (likelihood and consequences) are demonstratively bounding.

Chapter 6 Defense-in-Depth Section 11, Defense-in-Depth and Safety Margin Concerns, articulates a DID approach for the transportation of packages containing a transportable micro-reactor based on (1) the robustness of the TRISO fuel and package containment, (2) support of safety functions during transport that do not rely on active systems, (3) the transportable micro-reactor package transportation risk is quantified and shown to be low, (4) sensitivity studies show that most sources of uncertainty in PRA modeling assumptions in the Methodology and inputs do not impact the conclusions about risk from accidents, and (5) compensatory measures will be administered but were not credited in the transportable micro-reactor package PRA in the Methodology to reduce risk to the worker and the public and uncertainty about risk through preventive and mitigative actions and features.

Key aspects of the DID approach in Section 11 associated with the demonstration design are:

a. Passive elements of the design provide safety functions (e.g., not reliant on alternating current (AC) power or operator intervention).

• TRISO fuel limits release and is designed to be tolerant of elevated temperatures that can occur in nuclear power plant accidents

• The reactor coolant boundary acts as a containment boundary including isolation devices for disconnected piping

• The reactor module will absorb much of the energy of an impact crash protecting the reactor coolant boundary from more significant damage

• The CONEX-like box provides another barrier for release of radionuclides

• Rod locking mechanisms to prevent reactivity insertion

• Radiation shielding is provided by the reactor and supplemented by transport shielding in the walls of the CONEX-like box

b. The PRA demonstrates the risk of the safety systems failing is low, and verification of the performance of the safety features and components will be provided in the application for package approval (e.g., release of radiological material from a transportation accident involving significant mechanical forces is very small).

c. Administrative transport controls, not credited in the PRA in the Methodology, will be applied to limit potential accidents and help assure the reliability of the safety systems.

• escort vehicle in the front of and behind the truck carrying the transportable micro-reactor package

• route selection that avoids bodies of water and other measures, as necessary, such as inspecting bridges over bodies of water, closing bridges to other traffic, scheduling shipment to avoid high winds while on bridges

• ship at night to avoid high traffic

• fire detection and suppression installed on transport vehicle

d. Recovery plans will be in place in the event of transportation incidents and accidents.

• Transportation personnel trained in emergency response procedures (e.g., setup of a safety perimeter)

• PRA sensitivity studies used to assess worker and public safety (distance from accident location and duration of exposure) to enhance recovery response NRC Review The Commission has articulated its approach to applying the principles of DID for ensuring safe use of radioactive material in both reactor and materials applications. Key aspects of this DID approach from Staff Requirements - SECY-98-144 - White Paper on Risk-Informed and Performance-Based Regulation, dated March 1, 1999 (ML003753601) are:

• Defense-in-depth is an element of the NRC's Safety Philosophy that employs successive compensatory measures to prevent accidents or mitigate damage if a malfunction, accident, or naturally caused event occurs at a nuclear facility. The defense-in-depth philosophy ensures that safety will not be wholly dependent on any single element of the design, construction, maintenance, or operation of a nuclear facility. The net effect of incorporating defense-in-depth into design, construction, maintenance, and operation is that the facility or system in question tends to be more tolerant of failures and external challenges.

• The concept of defense-in-depth has always been and will continue to be a fundamental tenet of regulatory practice in the nuclear field, particularly regarding nuclear facilities.

Risk insights can make the elements of defense-in-depth clearer by quantifying them to the extent practicable. Although the uncertainties associated with the importance of some elements of defense may be substantial, the fact that these elements and uncertainties have been quantified can aid in determining how much defense makes regulatory sense. Decisions on the adequacy of or the necessity for elements of defense should reflect risk insights gained through identification of the individual performance of each defense system in relation to overall performance.

NUREG-2150, A Proposed Risk Management Regulatory Framework, dated April 2012 (ML12109A277) provided additional perspectives regarding the application of DID in transportation:

While the term defense-in-depth is not explicitly used, the current regulatory approach for approving and inspecting radioactive shipping packages follows the risk-informed and performance-based defense-in-depth approach in a general sense. For example, the safety requirements for different types of shipping packages become more stringent with the quantity (radioactivity), or hazard, contained. (page 4.8-3)

NUREG-2150 also made the following finding regarding the risk assessments and DID:

Finding 2.2: Risk assessments provide valuable and realistic insights into potential exposure scenarios. In combination with other technical analyses, risk assessments can inform decisions about appropriate defense-in-depth measures. (page 2-6)

The NRC used NUREG-1520 in part to assist the evaluation of DID approach for transportation of transportable micro-reactor package. The consideration for applying DID in the context of accident sequences that could result in undesired consequences at a fuel cycle facility are similar to the undesired consequences in transporting a transportable micro-reactor package (i.e., material releases or loss of containment and criticality). NUREG-1520 generally discusses safety at a fuel cycle facility in terms of items relied on for safety (IROFS) and the integrated safety analysis (ISA), whereas the approach for a transportation package of a transportable micro-reactor uses terms of safety features and components and PRA in conveying the same concepts. NUREG-1520 provides the following concepts that are directly relatable to the transportable micro-reactor risk framework in the Methodology:

An integrated safety analysis (ISA) identifies potential accident sequences in the facilitys operations, designates items relied on for safety (IROFS) to either prevent such accidents or mitigate their consequences to an acceptable level, and describes management measures to provide reasonable assurance of the availability and reliability of IROFS. (page 3-1)

Defense-in-depth is the degree to which multiple IROFS or systems of IROFS must fail before the undesired consequences (e.g., criticality, chemical release) can result. IROFS that provide for defense-in-depth may be either independent or dependent, although IROFS should be independent whenever practical because of the possibility that the reliability of any single IROFS may not be as great as anticipated. This will make the results of the risk evaluation more tolerant of error. (page 3-B-7)

The NRC staff has determined that the DID provided in the Methodology is appropriate because:

1. the approach describes the multiple layers of defense that is not reliant on a single feature or component,

2. the passive safety features and components provide a reasonable assurance of the availability and reliability,

3. the PRA has estimated the risks to be very low,

4. administrative transport controls, not credited in the PRA, will be applied to limit potential accidents and help to further assure the reliability of the safety systems,

5. recovery plans will be in place in the event of transportation incidents and accidents as a final safety protection.

Taken together, these protections are consistent with the Commissions statements regarding defense-in-depth and statements in NUREG-1520 regarding defense-in-depth with respect to fuel cycle facilities.

Chapter 7 Conclusion The NRC staff found that the Methodology proposed in Development and Demonstration of a Risk Assessment Approach for Approval of a Transportation Package of a Transportable Nuclear Power Plant [TNPP] for Domestic Highway Shipment, outlines an approach that if followed, with appropriate justifications, as discussed above, could be used as the basis to demonstrate that the transportation package will not endanger life or property nor the common defense and security, and thus could be granted an exemption from certain portions of the regulations under 71.12. This conclusion is based on the structural and thermal analyses performed during the accident sequence analysis, as discussed in Section 3.1.2 and the adequate calculation of all dose pathways that contribute to the TEDE for a member of the public and a radiation worker.