ML25344A109
| ML25344A109 | |
| Person / Time | |
|---|---|
| Site: | 99902069 |
| Issue date: | 01/05/2026 |
| From: | NRC/NRR/DANU |
| To: | Kairos Power |
| References | |
| KP-TR-020 | |
| Download: ML25344A109 (0) | |
Text
OFFICAL USE ONLY - PROPRIETARY INFORMATION
SUMMARY
REPORT FOR THE REGULATORY AUDIT OF KAIROS POWER, LLC - SAFETY ANALYSIS METHODOLOGY FOR THE KAIROS POWER FLUORIDE SALT-COOLED HIGH-TEMPERATURE TEST REACTOR TOPICAL REPORT OCTOBER 2024 - JULY 2025
1.0 BACKGROUND
AND PURPOSE Kairos Power LLC (Kairos) began pre-application discussions with the U.S. Nuclear Regulatory Commission (NRC) staff on its proposed Kairos Power Fluoride-Salt-Cooled, High-Temperature Reactor (KP-FHR) in October 2018. In December of 2020, Kairos announced that the Hermes test reactor (henceforth known as Hermes 1 for clarity) would be used to support future development of the KP-FHR. On December 14, 2023, the NRC staff issued a construction permit (CP) to Kairos for the Hermes 1 test reactor (Agencywide Documents Access and Management System (ADAMS) Accession No. ML23338A260). On November 21, 2024, the NRC staff issued CPs to Kairos for the Hermes 2 testing facility (ML24324A020).
By letter dated June 4, 2024, Kairos submitted KP-TR-020-P, Safety Analysis Methodology for the Kairos Power Fluoride Salt-Cooled High-Temperature Test Reactor (ML24156A162), to support future KP-FHR test reactor operating license (OL) applications. On July 3, 2024, the NRC staff determined that the topical report (TR) presented sufficient information to begin a detailed technical review (ML24184A052). The NRC staff issued an audit plan on September 26, 2024 (ML24269A202). The purpose of the audit was for the NRC staff to gain a better understanding of KP-TR-020-P by providing an opportunity to review and discuss material used to support statements or conclusions in the TR.
2.0 AUDIT REGULATORY BASES The bases for the audit are the regulations in Title 10 of the Code of Federal Regulations (10 CFR) pertinent to the NRC staffs review of a future Hermes 1 operating license (OL) application, particularly 10 CFR Part 50, Domestic Licensing of Production and Utilization Facilities, sections 50.34, Contents of applications; technical information, 50.40, Common standards, and 50.57, Issuance of operating license.
3.0 AUDIT OBJECTIVES The primary objective of the audit was to enable a more effective and efficient review of KP-TR-020-P through the NRC staffs review of supporting documentation and discussion with Kairos. A secondary purpose of the audit was to identify any information that will require docketing to support the NRC staffs safety evaluation.
4.0 AUDIT ACTIVITIES The audit was conducted from October 2024 through July 2025, via the Kairos electronic reading room (ERR). The NRC staff conducted the audit in accordance with the Office of
OFFICAL USE ONLY - PROPRIETARY INFORMATION 2
Nuclear Reactor Regulation (NRR) Office Instruction LIC-111, Revision 1 Regulatory Audits (ML19226A274).
Members of the audit team were selected based on their detailed knowledge of the subject and are listed below:
Alex Siwy - Senior Nuclear Engineer Alexander Chereskin - Materials Engineer Andrew Bielen - Senior Reactor Systems Engineer (Neutronics)
Ben Adams - General Engineer Brian Bettes - Project Manager Cayetano Santos Jr. - Senior Project Manager Michelle Hart - Senior Reactor Engineer Pravin Sawant - Senior Nuclear Engineer, Audit Lead Tarek Zaki - Reactor Systems Engineer The NRC staff reviewed the following documents via the ERR:
Kairos Power, ((
)), Hermes Base Model Documentation for KP-SAM, dated August 28, 2024.
Kairos Power, ((
)), Treatment of In-Vessel Argon 41 Inventories and Releases, dated May 23, 2024.
Kairos Power, ((
)), Analysis Methodology for Salt Spill Events for the Hermes Test Reactor, dated October 8, 2024.
Kairos Power, ((
)), KP-FHR Materials Handbook, dated April 29, 2020.
Kairos Power, ((
)), Primary Thermodynamic Data for Base-Case MST Modeling, dated May 29, 2024.
Kairos Power, ((
)), Grouping structures for modeling radionuclide evaporative release from Flibe, dated March 12, 2025.
Kairos Power, Core design inputs to KP-TR-020.
Journal of the Electrochemical Society, Effect of Metal Chloride Impurities on Equilibrium Potential of Fe/FeCl2 in Eutectic LiCl-KCl, dated July 25, 2023.
Kairos Power, ((
)), Hot Pebble Factor Methodology for KP-SAM Transient Analysis, dated August 29, 2024.
Kairos Power, ((
)), Figure of Merit and bounding values for solubility and the dilute solution limit for impurities in Flibe, dated February 2, 2025.
Audit meetings were held on the following dates (meetings were virtual unless otherwise noted):
October 9, 2024 - Entrance meeting October 21, 2024 November 7, 2024 November 12, 2024 November 26, 2024 December 10, 2024 January 13. 2025 January 21, 2025 February 4, 2025
OFFICAL USE ONLY - PROPRIETARY INFORMATION 3
February 5, 2025 February 18, 2025 February 26, 2025 February 27, 2025 March 18, 2025 April 1, 2025 April 8, 2025 April 23, 2025 July 21, 2025 - Exit meeting The NRC staff transmitted audit questions to Kairos on the following dates:
September 26, 2024 (ML24270A019 Public) (ML24270A021 Non Public)
December 12, 2024 (ML24347A062 Public) (ML24347A025 Non Public)
January 10, 2025 (ML25010A115 Public) (ML25010A210 Non Public) 5.0
SUMMARY
OF AUDIT OUTCOME The NRC staffs audit focused on the review of supporting documents associated with the topics identified in the audit plan and subsequently transmitted questions. The NRC staff reviewed information through the ERR and held discussions with Kairos staff to understand and resolve questions. In many cases, Kairos updated KP-TR-020-P to resolve items discussed in the audit. Gaining access to underlying documentation and engaging in audit discussions about various aspects of the TR facilitated the NRC staffs understanding and aided in assessing the proposed methodology. In addition, during an audit call on July 21, 2025, the NRC staff shared with Kairos draft limitations and conditions that were under consideration as the NRC staff prepared the draft safety evaluation (SE).
The table below reproduces the transmitted audit questions and summarizes resolutions. As noted in the table, Kairos updated the TR based on the audit discussions. These updates are included in Revision 1 of the TR, which was submitted in letter dated July 29, 2025 (ML25210A576). The equation numbers in the audit questions were updated to reflect the numbering in Revision 1 of the TR.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 4
Question Number Audit Question Resolution GEN-1 The descriptions of test reactor design features (TR section 1.1.3, KP-FHR Structures, Systems, and Components), postulated events (TR section 3.2, Postulated Event Descriptions), phenomena identification and ranking tables (PIRTs) (TR section 3.5, PIRT Summary), and KP-SAM input model (TR section 4.3, KP-SAM Base Model) are specific to the Hermes 1 design. The NRC staff acknowledges that the significant portions of these methodology elements presented in the TR are also applicable to the Hermes 2 design. However, as identified in the safety evaluation of the Hermes 2 CP application (ML24200A114), there are many unique design and safety features that need to be considered for the safety analysis of the Hermes 2 design. Without the Hermes 2 design specific information, the NRC staffs review and approval of the methodology as presented in the TR will be limited to only the Hermes 1 design. Any extension of the safety analysis methodology in the TR to the Hermes 2 design would require review and approval of the design specific information. Please confirm that the scope of the methodology presented in the TR is limited to only the Hermes 1 design.
Kairos stated during audit discussions that the scope of the TR methodology is limited to the design features described in TR section 1.1.3. In addition, Kairos stated that deviations from the design features described in TR section 1.1.3 would be justified in a future licensing submittal, as required by Limitation 5 in TR section 6.2.
GEN-2 Please describe the findings of any commercial grade dedication (CGD) studies or early gap evaluations performed for the selection of the System Analysis Module (SAM) code developed by Argonne National Laboratory. Please discuss changes implemented in the SAM code to develop KP-SAM. Please discuss the process for the implementation and verification of these updates to the SAM code. State the specific version of the KP-SAM code that is being used.
Kairos updated TR section 4.1.4 to state that KP-SAM version ((
)) was used for this TR, which was based on the SAM code. Kairos clarified that the SAM code was evaluated through CGD in accordance with the Kairos Power software quality assurance program. In addition, Kairos stated during audit discussions that the KP-SAM theory manual lists the changes made to the SAM code and the specific SAM code is listed in the CGD documentation.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 5
GEN-3
((
))
((
))
GEN-4 SAM relies on multiphysics object-oriented simulation environment (MOOSE) for some of the thermal fluids modules and fluid properties it uses, but once SAM is compiled these modules and properties would have already been imported from MOOSE and would be included within the SAM executable. However, the TR states that the fluid properties are hard coded in KP-SAM. Does KP-SAM need to import any information from MOOSE for code compilation or is it independent? Also, does SAM need to interact with any of the MOOSE modules (e.g., tensor mechanics module) during execution?
Kairos stated during audit discussions that KP-SAM relies on MOOSE for compiling; however, material properties are internal to KP-SAM. Kairos clarified that the term hard coded means that the material properties are found within KP-SAM and are not imported from MOOSE during code execution. In addition, Kairos stated that KP-SAM is not coupled to any other MOOSE-based modules.
GEN-5 The reactor core is modeled using KP-SAM 2D porous media and the coolant loops are modeled using KP-SAM 1D. Please describe how 1D and 2D domains are connected while ensuring consistency, continuity, and preserving energy and momentum.
Please explain if the "domain overlapping" technique developed by the Nuclear Energy Advanced Modeling and Simulation (NEAMS) or a similar method is used to achieve the domain overlapping and any validation approaches used to validate the method.
Kairos clarified that it is using a 1D approach and updated section 4.3 to provide additional description of the base model that includes nodalization and modeling for the converging and diverging sections of the reactor, bypass model, upper plenum model, etc.
The 1D approach does not involve domain overlapping.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 6
GEN-6 Section 1.2, Regulatory Background, of the TR states that [t]he development of this safety analysis methodology is consistent with the applicable portions of the evaluation model development and assessment process (EMDAP) described in Regulatory Guide [RG] 1.203, Transient and Accident Analysis Methods [ML053500170].
Please provide information on plans for addressing each step in the RG 1.203 process in the ERR.
During audit discussions, Kairos presented slides to describe how the steps in RG 1.203 are addressed in the TR. In addition, Kairos updated TR section 1.2 to clarify that the EMDAP is only applied to the KP-SAM methodology.
1.1-1 Anti-Siphon features play a critical role in limiting the loss of coolant from the reactor vessel in the event of a primary heat transport system (PHTS) break. The description of this important function and the methodology for modeling its performance are not provided in the TR. Please discuss the design and the modeling approach for the cold and hot leg anti-siphon design features. Please identify the phenomena from the PIRT tables (tables 3-1, Thermal Fluids PIRT Summary (High and Medium Importance Phenomena), through 3-3, Source Term PIRT Summary (High and Medium Importance Phenomena)) that address anti-siphon features.
Kairos updated TR section 1.1.3 to describe the anti-siphon design features in the PHTS. In addition, Kairos updated TR table 3-1 to list phenomena that are addressed by the anti-siphon features. Kairos added TR section 4.3.10 to describe how the anti-siphon features are modeled in the base model. The conservative bias in the event-specific salt spill modeling of the anti-siphon activation level was also added to TR section 5.5.1.
2.2-1 In section 2.2.1.1, Quantification of MAR [material at risk for release] Sources, of the TR, the provided methodology describes activation of argon (Ar) and a calculation of Ar-41 inventory based on an evaluation for a given region. Discuss what is meant by a region and how are they determined? How many regions are there? What is the sensitivity of the calculation to the region definition and use of region-averaged values?
Kairos provided documentation on the ERR describing how inventories of Ar-41 are calculated and released. During the audit, Kairos presented a summary of the ERR documentation for the NRC staff.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 7
2.2-2 What is the basis for TR equation 2.35 regarding the natural convection mass transfer relationship between LiF-BeF2 (Flibe) and cover gas?
Kairos stated in audit discussions that the relationship is based on an empirical correlation of the heat transfer between heated plate flow and gas in an open space. Kairos also stated that tests are being conducted per KP-TR-012-P-A, "KP-FHR Mechanistic Source Term Methodology [MST] Topical Report,"
revision 3 (ML22088A228) to validate the correlation.
2.2-3 Discuss an example of the Ar-41 solubility-limited pore release model (TR sections 2.2.1.1, Argon-41, and 2.2.4.2, Argon-41 Release) calculation to give Ar-41 release fractions from the reflector and fuel graphite pores.
This question was addressed through the resolution of audit question 2.2-1.
2.2-4 It is unclear from TR section 2.2.6, Gas Space, if the control room radiological habitability analyses will model post-accident isolation of the control room, unfiltered in leakage, and ventilation system operation. In addition, how will any potential radiation shine dose contribution be evaluated for the control room? See discussion in NRC RG 1.183, Alternative Radiological Source Terms for Evaluating Design Basis Accidents at Nuclear Power Reactors, Revision 1, section 4.2, Control Room Does Consequences, for further guidance.
Kairos updated TR section 2.2.6 to describe how the control room post-accident isolation, unfiltered in leakage, and ventilation system operation is modeled.
Kairos stated during audit discussions that the model is conservative because analysis does not credit any filtering or shielding for the control room.
2.2-5 The discussion of control room breathing rates and occupancy factors in TR section 2.2.6 refers to TR reference 21, which is NRC RG 1.194, Atmospheric relative Concentrations for Control Room Radiological Habitability Assessments at Nuclear Power Plants (ML031530505). RG 1.194 does not provide guidance on the control room habitability analysis assumptions on breathing rates and occupancy factors. The staff notes that TR reference 33, RG 1.183, does give such guidance.
Discuss why RG 1.194 is referenced instead of RG 1.183 in TR section 2.2.6.
Kairos confirmed in audit discussions that this was a typographical error and updated the TR section 2.2.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 8
2.2-6 Section 2.2.1, "Quantification of MAR [material at risk for release] Sources," states that a new intermediate volatility noble metal (IVNM) radionuclide transport grouping is created to regroup some noble metal elements for transport.
Section 3, "The Revised Base Case Grouping Structure," of ((
)), "Primary Thermodynamic Data for Base-Case Mechanistic Source Term Modeling: Selection, Justification, and Uncertainty Quantification," revision 2, states that the IVNM radionuclide transport grouping is
((
)). Additionally, please describe how noble metal elements were determined to be in the IVNM grouping instead of the low or high volatility noble metal groupings.
Kairos updated TR section 2.2.3.2, to clarify that the noble metal groupings are determined by comparing the equilibrium vapor pressure of each element.
Additionally, Kairos stated the IVNM grouping still uses a representative transport species that has a higher vapor pressure than the rest of the species included in the IVNM group. This maintains conservatism because all of the elements in the IVNM group will be transported at a greater rate than if vapor pressures for individual species were used.
The NRC staff confirmed the basis for the IVNM grouping in internal report ((
)).
Specifically, the internal report stated that the new representative species (In) has a higher vapor pressure than all other species in the IVNM grouping, which is consistent with the method to group the low and high volatility noble metals in KP-TR-012-P-A.
While the new grouping is less conservative than the previous groups (i.e. only low and high volatility noble metals), it is still conservative due to the bounding nature of the selected transport species.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 9
2.2-7 Section 2.2.3.2, "Evaporative Release," states that the elements which fall into multiple transport groups based on the Flibe redox potential are now placed in other groups based on the vapor pressure of these elements. Please describe why the latter grouping structure remains conservative given that solubility of certain elements (e.g., Niobium) in the salt is likely dependent on the redox potential.
Additionally, please confirm that the mass transfer coefficient for evaporation from Flibe in equation 2-35 will be validated experimentally.
Kairos stated during audit discussions that the change in methodology from KP-TR-012-P-A is minor. KP-TR-012-P-A used release fraction as opposed to vapor pressure used in this TR. Kairos clarified that the change was made to switch the burden of proof to before the analysis as opposed to after the analysis. In addition, Kairos confirmed that equation 2-35 in the TR will be validated experimentally.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 10 2.2-8 Section 2.2.3.3, "Confirmation of Solubility/Dilute Solution," states that salt soluble fluorides are evaluated to ensure mole fractions for non-native impurities are less than a limit of 6235 [6325, correction of error] parts per million (ppm) and therefore ensures solute-solute interactions are not measurable, and the impurities are dilute and soluble. The reference cited as justification for this limit appears to study a different molten salt composition. Given the different salt composition, please explain how this study can be used to derive a limit to ensure a dilute solution for impurities in Flibe.
Kairos updated TR section 2.2.3.3 to include explanation on the solubility limit of salt soluble fluorides. Specifically, Kairos updated TR section 2.2.3.3 to state that fluoride and chloride salts are chemically similar and that the solute-solute interactions are independent of the solvent, therefore the use of data from chloride salt experiments is applicable to fluoride salts.
The NRC staff noted that while this data may be useful to inform the decision on a purity limit to maintain a dilute solution assumption, the stricter limit on non-native salt-soluble fluorides is stated in TR section 2.2.3.3. This limit would apply to species that could form compounds such as cesium iodide, which may cause non-ideal vaporization behavior. Kairos noted that the salt soluble fluoride (SSF) limit is based on plutonium trifluoride solubility data, which tends to be less soluble than other SSFs. The NRC staff noted that it is not making a determination in its SE on whether chloride and fluoride salts are similar enough to use the data interchangeably.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 11 2.2-9 Please provide the following clarifications on how the vapor pressure correlation for cesium fluoride (CsF) was determined:
o How were the terms in equation 2.45 of the TR derived from the cited data in the NIST-JANAF database?
o How was the uncertainty on the correlation developed?
o Confirm the experimental work described in section 4.3.1.6 of KP-TR-012-P-A, which we understand is outside the scope of this TR, will be used to verify whether the assumption of an activity coefficient of 1.0 for the CsF vapor pressure is conservative.
Kairos stated in audit discussions that the terms in equation 2.45 were determined through fitting the equation to the data provided in the NIST-JANAF database.
Kairos stated that the uncertainty is the maximum variance provided by the NIST-JANAF database applied to all data points.
Kairos confirmed that the experimental work to be done as described in KP-TR-012-P-A will be used to verify the assumption of an activity coefficient of 1.0 for CsF.
3.2-1 In section 3.2 of the TR, the methodology addresses seven postulated event groups. It states that the methodology for other postulated event groups will be included in a future licensing submittal. Discuss potential event categories left out of the current methodology.
Kairos stated during audit discussions that there are no foreseen additional event groups to be included in a future licensing submittal as confirmed by Limitation 1 in TR section 6.2
OFFICAL USE ONLY - PROPRIETARY INFORMATION 12 3.2-2 Section 3.2.6, Pebble Handling and Storage Malfunction, of the TR states that the mechanical damage and loss of pebble handling and storage system (PHSS) cooling scenarios that could result in a release of radioactive MAR are assumed to be mitigated or precluded by design. These precluded scenarios are not discussed in the Hermes 1 CP application (ML23151A743); please explain the discrepancy.
Kairos stated during audit discussions that the loss of the PHSS cooling is not within scope of the TR.
Kairos clarified that for Hermes 1, the description of how loss of the PHSS cooling is mitigated or precluded by design will be provided in a future licensing submittal. The NRC staff communicated that a Limitation and Condition may be needed regarding the modeling of PHSS malfunction event.
Kairos also stated that the current methodology is not applicable to mechanical damage and loss of PHSS cooling scenarios as it assumes that these are mitigated by design. The only event discussed under PHSS malfunction is a break in transfer line. TR section 5.6.2 provides discussion of the method for calculating dose from non-maximum hypothetical accident (MHA) release pathways. This dose is added to the dose of the bounding primary system event. No need is identified for a separate KP-SAM calculation for this event. No unique phenomena are identified in PHSS for modeling using KP-SAM.
3.2-3 Section 3.2.6 of the TR indicates that a break in the PHSS is detected by the reactor protection system (RPS) through the low pressure in the PHSS actuation signal. Is this a safety-related signal? Is the low pressure signal used to detect other PHSS malfunctions? If not, how are other PHSS malfunctions detected?
Kairos stated in audit discussions that the signal is a safety-related signal. In addition, Kairos stated that other PHSS malfunctions are not within the scope of the TR.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 13 3.2-4 Clarify the role of KP-SAM in analyzing the following event categories:
o PHSS malfunction o Radioactive release from subsystems and components o Salt spill Kairos stated during audit discussions that KP-SAM is not used for evaluating a PHSS malfunction.
Kairos stated during audit discussions that KP-SAM is not used to evaluate radioactive release from subsystems and components.
Kairos updated the TR to add section 4.3.10 which describes how a salt spill is modeled. Kairos also updated TR sections 3.3 and 3.4 to clarify the figures of merit (FOMs) and PIRT phenomena for the non-intact postulated events that clarified that KP-SAM is used to analyze a salt spill event and added discussion on modeling of salt spill event in KP-SAM.
Kairos clarified that for the salt spill and PHSS malfunction event the dose calculated for the bounding primary system event is added to the releases from non-MHA pathways. KP-SAM is not needed for the calculation of releases from subsystems of component.
3.2-5 Clarify the scope of application of the KP-SAM and KP-BISON codes for each of the event categories identified in section 3.2 of the TR. If multiple codes are used in the methodology, please clarify the scope and interfaces for the different codes and models.
Kairos stated during audit discussions that KP-SAM provides input into KP-BISON for the following events: increase in heat removal, decrease in heat removal, loss of forced circulation, and reactivity-initiated events, and salt spills. Kairos updated TR sections 3.3.1 and 4.2.4 to clarify the use of KP-BISON.
3.3-1 Section 3.3.1.5, Peak Structural Graphite Temperature, of the TR states that the acceptance criterion for the peak structural graphite temperature FOM is set at 950 degrees Celsius (°C). Does this account for the effect of temperature on temperature-dependent graphite properties?
Kairos stated during audit discussions that the acceptance criterion for the peak structural graphite FOM accounts for temperature-dependent properties such as thermal expansion and thermal conductivity.
Kairos also stated that any necessary graphite qualification activities would be conducted outside the scope of this TR. The FOM in this case is an analytical assumption and does not necessarily reflect the required qualification temperature.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 14 3.3-2 Section 3.3.2, Dose Figures of Merit, of the TR presents different FOMs for the postulated events where the integrity of a Flibe-containing component is compromised (e.g., salt spill event) as compared to the surrogate FOMs for the rest of the postulated events in TR section 3.3.1, Surrogate Figures of Merit. However, in the Hermes 1 CP application, the surrogate FOMs from TR section 3.3.1 were also used as FOMs for the salt spill and PHSS break postulated events. Discuss the differences between the two approaches.
Kairos revised the discussion in TR section 3.3 on the FOMs and clarified each FOM selected for the methodology, acceptance criteria, and how it is calculated.
3.3-3 With respect to the dose FOM, how does the acceptance criterion account for principal design criteria (PDC) 19, Control Room? Could there be a potential event (e.g., release from a subsystem or component) for which the release results in an exclusion area boundary (EAB) 30-day total effective dose equivalent (TEDE) bounded by the MHA, but the control room dose is not bounded by the MHA?
Kairos stated during audit discussions that given how the control room operator receptor location is modeled as in the outside environment (i.e., without modeling the receptor inside a structure or ventilation envelope), it did not identify a postulated event that would have a control room dose greater than the MHA control room dose. Given the relative difference between the control room and offsite atmospheric dispersion values are the same as for the MHA and considering the non-MHA accident releases would have offsite doses less than the MHA, the non-MHA control room doses would be also be bounded by the MHA.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 15 3.5-1 Please confirm that the phenomena identified in the thermal fluids PIRT (TR table 3-1) are applicable to each event category identified in TR section 3.2. ((
)) Discuss how this PIRT was extended to other event categories in TR section 3.2. Similarly, confirm that the reactivity-initiated PIRT addresses all the phenomena needed to model the neutronics in each event category identified in TR section 3.2. Please discuss any considerations given to the event category-specific PIRT.
Kairos updated TR section 3.5 and table 3-2 to clarify that the phenomena in the thermal fluid PIRT are applicable to all postulated events except for the radioactive release from subsystem and component event group. Some phenomena are specific to certain postulated events, and these are clearly identified in the PIRT. The updated TR also clarifies that the reactivity initiated PIRT addresses the reactivity initiated and increase in heat removal events.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 16 3.5-2 TR tables 3-1 through 3-3 identify important phenomena that need to be modeled by the methodology to calculate the required FOMs.
Please clarify the following for the phenomena listed below: 1) the specific code models and correlations used; 2) the basis for selecting the models and correlations; and 3) their applicability, maturity or validation status.
o ((
o o
o o
o o
Kairos updated TR section 4.2.2 to identify the code models and correlations for the phenomena listed, the basis for their selection, and the applicability range.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 17 o
o
))
3.5-3
((
))
Kairos stated during audit discussions that KP-SAM does not have the ability to vary the bed porosity as a function of location, a single value is used. Moreover, Kairos indicated it did not perform sensitivities on bed porosity as part of the safety analysis methodology.
Kairos stated that non-homogenous effects due to asymmetrical flow distribution would be homogenized quickly and that during normal operations, there are no 3D effects that would require 3D bed porosity resolution near the wall.
Kairos updated TR section 4.2.4 to describe the hot pebble factor methodology developed to account for the impact of non-uniform flow (and porosity) and power distribution.
3.5-4
((
))
Kairos stated during audit discussions that neglecting radiative heat transfer to the adjacent structures increases maximum coolant temperature. This maximum coolant temperature is used to determine the temperature of surrounding components which is a conservative approach.
Kairos added TR section 4.2.4.3 to clarify how neglecting radiative heat transfer from the pebble bed to adjacent structures results in a conservative prediction of peak reflector temperature and peak vessel temperature FOMs.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 18 3.5-5
((
))
Kairos addressed this question, in part, through audit discussions on question 3.5-4. Kairos stated that the peak FLiBe temperature will be used for peak vessel temperature.
Kairos added TR section 4.2.4.2 to clarify how neglecting emissivity of graphite structures and other vessel internal structures results in a conservative prediction of peak reflector temperature and peak vessel temperature FOMs.
3.5-6
((
))
Kairos stated during audit discussions that bypass flow is not directly controlled once a transient is initiated.
Kairos updated TR section 4.3 and added section 4.3.8 to describe the modeling approach for bypass in KP-SAM base model. Kairos stated that sensitivity analysis will be performed to evaluate the impact of bypass flow fraction on the methodology FOMs.
3.5-7
((
))
Kairos updated TR table 3-1 to clarify how the methodology addresses heating by gamma and neutron radiation, which is detailed in TR sections 4.2.4.2 and 4.2.4.3. In addition, Kairos added TR table 4-4 to summarize the modeling limitations of the methodology, which includes 3D distribution of power, flow, and temperature.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 19 3.5-8 o ((
o o
o o
))
((
))
OFFICAL USE ONLY - PROPRIETARY INFORMATION 20 3.5-9
((
))
Kairos stated in audit discussions that section 15.7 of the KP-SAM user manual describes the treatment of material properties. As described in the response to audit question Gen-4, Kairos clarified that hard coded means that it is specified in KP-SAM and not a user input.
The TR does not address the treatment of uncertainties in material properties and geometry in KP-BISON for the calculation FOMs for this methodology. As the methodology for calculation of the peak tri-structural isotropic (TRISO) silicon carbide (SiC) layer temperature and fuel failure fraction FOMs is not presented in this TR, the NRC staff communicated that a Limitation and Condition may be needed regarding the use of KP-BISON.
3.5-10
((
))
Kairos clarified during audit discussions that the 3D temperature distribution affects the temperature distribution among different pebbles across the core.
In addition, Kairos stated during audit discussions that the HPF methodology accounts for uncertainties in 3D temperature and flow distributions in the core.
Kairos updated the TR to include section 4.2.4 to clarify how the bounding HPF accounts for uncertainties.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 21 3.5-11
((
))
((
))
3.5-12
((
))
((
))
3.5-13
((
))
Kairos added TR sections 4.3.8 which provides more detailed information on modeling of bypass flow.
Kairos also updated TR section 5.1 to clarify the approach for bypass flow sensitivity calculations to determine conservative treatment for modeling of bypass for different event categories.
3.5-14
((
))
((
))
OFFICAL USE ONLY - PROPRIETARY INFORMATION 22 3.5-15
((
o o
))
Kairos updated TR section 5 to clarify the parameters that will be monitored for the determination of limiting events using sensitivity calculations which involve biasing coolant thermophysical properties among several other biases. The final outcome of these sensitivity calculations and the final values selected for the coolant thermophysical properties be reviewed as a part of an OL application.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 23 3.5-16
((
))
Kairos stated during audit discussions that neglecting the loss of heat to reactor building during a heat up event is a conservative assumption. In addition, Kairos updated TR section 4.3.3 to clarify that adiabatic boundary conditions were being used in the methodology.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 24 3.5-17 Table 3-1 of the TR states that ((
o
))
o Please discuss how the DHRS performance is modeled for these calculations and for the sample calculations presented in appendix A.
Kairos updated the executive summary and appendix A of the TR to clarify that the sample calculations in appendix A are meant to provide a generic system response for selected transients and do not represent final design information. Kairos updated TR section 4.3.5 to clarify that DHRS is modeled as a simple heat flux boundary condition as part of the methodology. Kairos also added text to appendix A section 1 to describe the DHRS modeling approach for that example calculations presented in the appendix.
((
))
OFFICAL USE ONLY - PROPRIETARY INFORMATION 25 3.5-18
((
))
((
))
3.5-19
((
o o
o
))
((
))
OFFICAL USE ONLY - PROPRIETARY INFORMATION 26 3.5-20
((
o
((
))
OFFICAL USE ONLY - PROPRIETARY INFORMATION 27 o
o o
))
OFFICAL USE ONLY - PROPRIETARY INFORMATION 28 4.1-1 Section 4.1.3, Numerical Methods, of the TR states that "continuous finite element methods formulation for the spatial discretization" is used in SAM. Further, the TR states that the detailed discretization is managed in MOOSE and the user can control numerical method orders.
o Please discuss any limitations or assumptions for the use of this specific numerical method and discretization, as well as the spatial stabilization method.
o Please discuss justifications for the selected numerical method order.
o Please describe the approach or guidelines established to control or manage the numerical errors.
Kairos stated in the audit discussions that the spatial stabilization method used is the default provided in KP-SAM and discretization is consistent with the guidance in KP-SAM user guide. Kairos also noted that mesh sensitivity studies were performed to confirm the independence of the simulation solutions from spatial discretization. Kairos updated TR section 4.3.1 consistent with the audit discussion.
Kairos noted that the KPSAM Users Guide (Reference 28) provides guidance on nodalization to manage the numerical errors. The NRC staff notes that the planned validation against the integrated effects test data discussed in TR section 4.2.3 will also provide additional justification for the selected nodalization.
4.2-1 Table 4-1, Parameter Ranges to Address Heat Transfer in KP-FHR Test Reactor Pebble Bed, of the TR provides ranges for Reynolds number and Prandtl numbers covered in the Pebble Bed Heat Transfer SET. If available, please discuss the simulant fluid, pebble sizes, and bed porosities covered in the experiment. Please confirm that SET instrumentation will be adequate to calculate contributions to the heat transfer from the near-wall heat transfer and pebble-to-fluid heat transfer mechanisms.
Kairos stated during audit discussions that the ranges provided in the TR are intended to provide context for how the SET would be used to confirm the correlations used but not provide specific test design information. Kairos stated that specific experiment information would be provided in a future licensing submittal as identified by Limitation 3 in TR section 6.2.
Kairos also confirmed that the SET would resolve both the heat transfer from the near-wall and pebble-to-fluid heat transfer. Kairos updated TR section 4.2.2.3 to clarify this.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 29 4.2-2 Impact of asymmetric flow, temperature and power (neutron flux) distributions in different regions of the reactor is accounted for by using several factors such as HPF, peak vessel temperature factor, peak temperature factor, and peak reflector temperature.
The following phenomena are addressed using this approach:
o Volumetric heating of structures and coolant by gamma and neutron radiation o Core flow 3D effects; o Conjugate heat transfer from fuel to coolant and the resulting 3D fuel temperature distribution; o Conjugate heat transfer from graphite structures to coolant and the resulting 3D temperature distribution; and o Conjugate heat transfer from coolant and cavity to reactor vessel structure and the resulting 3D temperature distribution.
Please discuss how these peaking factors are applied and biased. Please discuss the methodology for validating these factors. Please explain what, if any, consideration was given to the use of 3D calculations to validate 1D or 2D modeling approaches.
Kairos updated TR section 4.2.4 to clarify the HPF methodology and the methodologies for calculation of peak reflector temperature and peak vessel temperature surrogates. Kairos updated the TR to replace the term temperature factor with temperature surrogate and clarified that there are only peak vessel and peak reflector temperature surrogates. The updated discussion clarified how these methodologies address key KP-SAM modeling limitations of 3D flow, power, and temperature distribution. The discussion provided further information on the process for implementation of these methodologies which involves performing supporting calculations using the core design methodology and CFD model.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 30 4.2-3
((
))
((
))
4.3-1 In the KP-SAM base model, does each core node have both fuel and moderator pebble heat structures? How many types of fuel pebble heat structures are defined (e.g., average, hot)? How is power deposited or distributed within these heat structures?
Kairos stated during the audit discussion that each axial node has separate structures for fuel and moderator pebbles. Kairos updated TR section 4.3.2 to clarify that the power is calculated using point kinetics model and described the approach for specification of axial power shape. There is no radial power distribution due to the use of a 1D model. The impact of radial power distribution is accounted for by the HPF methodology.
4.3-2 Please explain how the radial flow from the core to reflector is represented in the KP-SAM base model.
Kairos added TR section 4.3.8 to describe the modeling of reflector bypass which accounts for flow through the reflector gaps and channels. Kairos stated that radial and azimuthal flows and temperature gradients within the core are not directly calculated by their 1D KP-SAM core model. Kairos added TR section 4.2.4 which describes the HPF methodology and methodologies for peak vessel and reflector temperatures. As described in the TR, these methodologies account for radial and azimuthal flow distribution effects and are informed by the CFD calculation.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 31 5-1 Please discuss the process for identification of the limiting event for the different event categories in section 5, Event-Specific Methods. Please discuss the sensitivity calculations that will be performed, and the FOMs used for the identification of the limiting event.
Kairos updated TR section 5 for different event specific methods to clarify the approach for sensitivity calculations and identified the parameters or FOMs monitored to determine the limiting event.
5-2 Please explain how the safety analysis methodology accounts for undesirable impacts due to availability of the non-safety related plant control system and operator actions. Please discuss specific examples.
Kairos stated during audit discussions that the sensitivity studies will consider operator actions prior to the initiation of the event that could create the most challenging initial conditions. Kairos also confirmed that the activation of RPS or the shutdown system would prevent operator actions by design.
5-3 Please describe the core design inputs and reactor physics parameters (reactivity coefficients and other neutronic data) for the core model to calculate reactivity response and the power distribution. In addition, please explain how the conservative biases/uncertainties are applied to these parameters in safety analysis calculations?
Kairos updated TR section 4.3.2 to clarify the use of a point kinetics model to calculate the power response in KP-SAM. Kairos added table 5-1 which describes the core design analysis inputs for the KP-SAM model. TR section 5 describes the treatment of biases for the reactor initial power and reactivity parameters for the event specific methods. The NRC staff also reviewed additional base model documentation in the ERR to confirm the input used to model the power boundary condition.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 32 5.1-1 Section 5.1.1, Bias Application, describes conservative biases for different parameters used in the analysis of an increase in heat removal event. In general, it appears that the parameters are biased to maximize the heat-up.
o Please justify why maximizing the heat-up would be the bounding event for the reactivity response.
o Please explain how the direction of bias is determined for different parameters considering the multiple FOMs.
Specifically, please identify if there is any FOM for which the required bias direction for conservative analysis is different than the biasing prescribed in section 5.1.1 to maximize the heat-up.
Kairos clarified during audit discussions that maximizing the heat up maximizes the power of the reactor for theincrease in heat removal event, which would reduce the margin to fuel failure. Furthermore, Kairos also updated TR section 5 for different event specific methods to clarify the approach for selecting bias direction and preforming sensitivity calculations and identified the parameters or FOMs monitored to determine the limiting event.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 33 5.1-2 Section 5.1.2, Sensitivities, describes sensitivity studies proposed for the analysis of an increase in heat removal event. The material property sensitivity analysis includes only the properties of the coolant.
o Please discuss how the impact of uncertainties in properties of other materials is accounted for in the methodology. For example, heat capacities of reflector, fuel and moderator pebbles, reactor vessel, and other internal structures which could significantly affect the total stored energy in the model.
o Please discuss how the uncertainties in total mass and heat transfer surface areas of different structures in reactor vessel (e.g., reflector, fuel and moderator pebbles) and primary system are accounted for in the methodology.
Kairos updated TR section 4.3.6 to clarify that the material properties in the base model are not biased except for the thermal conductivity of ET-10 graphite.
Kairos stated that the justification for the use of unbiased material properties will be provided in a future licensing submittal as described in sensitivity analysis approach in TR section 5. Kairos also clarified that heat capacities of Flibe and graphite are large and small geometrical uncertainties (mass and surface area) would not impact the overall thermal mass of the system.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 34 5.4-1 Section 5.4.1, "Bias Application," states that [a]
transient radial peaking factor is not applied because the withdrawal of all control elements results in a symmetric reactivity insertion. Transient radial and pebble peaking factors are applied for the single element withdrawal events to account for axial and radial changes in the power shape during the reactivity-initiated event. A hot pebble factor, which conservatively treats the power shape and associated uncertainties, is applied to calculate the highest temperature fuel pebble. Please discuss how three-dimensional power distribution or asymmetry in power response is accounted for in a single element withdrawal event. In addition, discuss the differences between the transient radial peaking factor, transient pebble peaking factor, and hot pebble factor, and their physical importance for the methodology.
Kairos updated TR section 5.4.1 to replace the term transient with event-specific. Kairos added TR section 4.2.4 which describes the HPF methodology and the methodologies for peak vessel and reflector temperatures. These methodologies account for the impact of non-uniform power and flow distributions for different event categories, including the single rod withdrawal event.
5.5-1 Section 5.5.1, Bias Application, mentions that certain Flibe properties, including surface tension, are specified for salt spill events. Please describe how surface tension is used in the salt spill analysis and whether it has a significant impact on radionuclide release (e.g., via bubble formation and popping).
Kairos stated during audit discussions that surface tension is used to determine aerosolization as described in KP-TR-012-P-A, however, it is not used to determine radionuclide release.
5.5-2 Section 5.5.2, Sensitivities, states that the limiting salt spill event is conservatively modeled as a combination of the most limiting break conditions rather than a single break scenario. Please describe how this is achieved.
Kairos stated during audit discussions that the salt spill event has multiple pathways not used in the MHA. Kairos evaluated how the size of the break effected the release of radionuclides. Kairos stated that the salt spill event sums up the various break possibilities rather than selecting a specific break.
Kairos stated that that the assessment uses the most limiting conditions as related to the offsite dose results.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 35 5.6-1 Discuss how the number of TRISO particles with intact silicon carbide layers that are dislodged from the oxidized fuel pebbles is determined.
Kairos stated during audit discussions that the analysis takes a marginally conservative non-physical assumption to account for uncertainty in the modeling of the release mechanisms from the TRISO. The assumption is that for the pebbles exposed to oxygen in the accident, all TRISO particles with an intact SiC layer (based on steady-state conditions) in those pebbles fall into the Flibe and are dissolved to mix their radionuclides into the Flibe for subsequent release.
5.6-2 Discuss how the TR methodology is implemented and the choices made by the analyst in the example analysis in appendix A.4, Pebble Handling and Storage System Malfunction (Transfer Line Break),
including the modeling of damaged fuel pebbles and graphite dust.
During the audit, Kairos discussed the example analysis in appendix A.4. Kairos clarified that the example analysis does not model graphite dust and that a methodology for determining the quantity of graphite dust will be included in a future licensing submittal.
Kaiross walkthrough of the example analysis aided in NRC staff understanding of the methodology for the PHSS malfunction.
5.6-3 Please describe how the temperature distribution of pebbles in the pebble extraction line and the pebble insertion line is determined in section 5.6, Pebble Handling and Storage System Malfunction.
Kairos updated TR section 5.6.1 to clarify that all pebbles in the extraction line and insertion line are assumed to be at nominal temperature greater than 400 °C and are subject to oxidation.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 36 5.7-1 Section 5.7, Radioactive Release from Subsystem or Component, states that the limiting event in this group is expected to be a seismic event, which results in failure of the subsystems and components not designed to withstand the design basis earthquake.
o Please clarify if the radioactive MAR for the seismic event is only located in the subsystems and components that fail (i.e., are not designed to withstand the design basis earthquake), or if there are releases from other systems or components that communicate with the failed subsystems and components, potentially with a delayed or protracted release.
o Section 5.7 states that the evaluation model for this event group is composed of establishing MAR inventories, combining releases from multiple subsystem or component failures, and the gas space transport model. Please explain in detail how the gas space transport model will be used.
Kairos stated during audit discussions that external events are outside the scope of this TR. However, Kairos confirmed that the methodology described in section 5.7 of the TR will be used to evaluate release from systems not designed to survive the event.
Kairos stated that if there is a gas space release then the methodology described in t KP-TR-012-P-A will be used.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 37 6.1-1
((
))
((
))
A-1 Please provide the analytical limits or reactor protection system setpoints used for the example calculations in appendix A, Example Calculations of the TR.
Kairos stated during audit discussions that section 1.1.3 of the TR describes setpoints for the RPS and that section 3.2 of the TR describes the event specific basis for the setpoints. Furthermore, Kairos updated appendix A to clarify that the calculations in the appendix do not represent final design information but are provided for illustration purposes. Kairos is not requesting review and approval appendix A.
OFFICAL USE ONLY - PROPRIETARY INFORMATION 38 6.0 EXIT BRIEFING The NRC staff conducted an audit closeout meeting on July 21, 2025. At the closeout meeting, the NRC staff reiterated the purpose of the audit and discussed their activities.
7.0 ADDITIONAL INFORMATION RESULTING FROM AUDIT As a result of the audit, Kairos revised the subject TR to address questions discussed during the audit.
8.0 OPEN ITEMS AND PROPOSED CLOSURE PATHS Not applicable. There were no deviations from the audit plan.
Package: ML25197A042 Draft proprietary version: ML25210A101 Final Proprietary version: ML25344A108 Final Nonproprietary version: ML25344A109 Transmittal email for Final Report: ML25344A111 OFFICE NRR/DANU/UAL1:PM NRR/DANU/UAL1:LA NRR/DANU/UTB1:LEAD NAME BBettes DGreene PSawant DATE 7/29/2025 8/11/2025 12/10/2025 OFFICE NRR/DANU/UTB1:BC NRR/DANU/UAL1:BC NRR/DANU/UAL1:PM NAME CdeMessieres JBorromeo BBettes DATE 12/10/2025 12/10/2025 12/10/2025