Text

Non-Proprietary - Independent Melcor Confirmatory Analysis for NuScale Small Modular Reactor
	ML19196A306
Person / Time
Site:	NuScale
Issue date:	04/30/2019
From:	Shawn Campbell, Hossein Esmaili, Jason Schaperow; NRC/NRO/DLSE/LB1, NRC/RES/DSA
To:	;
	Lauron C, 415-2736
References
	RES/FSCB 2019-01
	Download: ML19196A306 (74)
	v • d • e

RES/FSCB 2019-01 Independent MELCOR Confirmatory Analysis for NuScale Small Modular Reactor S. Campbell H. Esmaili J. Schaperow 1 April 2019 Fuel and Source Term Code Development Branch Division of Systems Analysis Office of Nuclear Regulatory Research United States Nuclear Regulatory Commission 1

PRA and Severe Accidents Branch, Division of Safety Systems and Risk Assessment, Office of New Reactors, U.S. NRC

FSCB-18-01 ACKNOWLEDGEMENTS The authors would like to acknowledge A. Hathaway (NRC/RES) and Z. Yuan, S. Choi, A. Krall, and M. Khatib-Rahbar (Energy Research, Inc.) for their effort to develop the MELCOR model of NuScale.

ii

FSCB-19-01 TABLE OF CONTENTS List of Tables................................................................................................................................ iv List of Figures ............................................................................................................................... v List of Acronyms .........................................................................................................................viii

1. Introduction ............................................................................................................................ 1 1.1. NuScale small modular reactor ...................................................................................... 1 1.2. Guidance for staff confirmatory analysis ........................................................................ 1 1.3. Outline of this report ....................................................................................................... 2

2. MELCOR Modeling Approach and Scenario Selection ......................................................... 2 2.1. Staffs MELCOR Modeling Approach ............................................................................. 2 2.2. Selection of Accident Scenarios ..................................................................................... 6

3. MELCOR Results and Comparison with the Applicants Results ........................................ 11 3.1. LEC-06T ....................................................................................................................... 11 3.2. LCC-05T ....................................................................................................................... 21 3.3. LCU-03T ....................................................................................................................... 29

4. Sensitivity Calculations to MELCOR Calculations ............................................................... 36 4.1. RVV Flow Area ............................................................................................................. 36 4.2. SG Nodalization ............................................................................................................ 38 4.3. Timing of ECCS Actuation ............................................................................................ 40 4.4. CVCS Break Location ................................................................................................... 42 4.5. Removal of the Bypass Flowpath Between Downcomer and Riser.............................. 45 LEC-06T .............................................................................................................................. 45 LCC-05T .............................................................................................................................. 48

5. Confirmatory Analysis for the NuScale Accident Source Term Topical Report ................... 51 5.1. Introduction ................................................................................................................... 51 5.2. Source Term from the RPV to Containment ................................................................. 51 5.3. In-containment Aerosol Removal Rate ......................................................................... 53 5.4. Containment Leak Rate ................................................................................................ 56 5.5. Release from the containment to the environment ....................................................... 59

6. References .......................................................................................................................... 60 A. Example Problem of Containment Leak Rate ...................................................................... 61 iii

FSCB-19-01 List of Tables Table 2.1 Comparison of select design features of the NuScale to a representative Combustion Engineering operating reactor ............................................................. 2 Table 2.2 Compilation of MELCOR calculations performed by NuScale for the FSAR ........... 7 Table 2.3 Scenarios Selected for Confirmatory Analysis ...................................................... 10 Table 3.1 A comparison of event timings in simulation LEC-06T for both the staff and applicants MELCOR calculations. ........................................................................ 14 Table 3.2 A comparison of event timings in simulation LCC-05T for both the staff and applicants MELCOR calculations. ........................................................................ 22 Table 3.3 Comparison of event timing for both NuScale and NRC staffs MELCOR calculations of LCU-03T ........................................................................................ 30 Table 5.1 Release fractions and time of the five representative accident scenarios from the applicants MELCOR calculations as compared to the Staffs MELCOR result.

DHRS was available in all scenarios until core damage. ...................................... 52 iv

FSCB-19-01 List of Figures Figure 2.1 Staffs MELCOR nodalization for NuScale model - Reactor Vessel ....................... 3 Figure 2.2 Staffs MELCOR nodalization for NuScale model - Containment Vessel ............... 4 Figure 2.3 Staffs MELCOR nodalization for NuScale model - Reactor bay water pool .......... 5 Figure 2.4 Staffs MELCOR nodalization for NuScale model - COR package ......................... 6 Figure 3.1 LEC-06T: Short-term pressure in the CNV, RPV and SGs. .................................. 15 Figure 3.2 LEC-06T: Long-term pressure in the CNV, RPV and SGs. ................................... 15 Figure 3.3 LEC-06T: Total core power. .................................................................................. 16 Figure 3.4 LEC-06T: Collapsed liquid level in the RPV, CNV and steam generators............. 16 Figure 3.5 LEC-06T: Integrated mass flowing through the ECCS valves. Note that RVV #2 and #3 curves are identical. .................................................................................. 17 Figure 3.6 LEC-06T: Mass balance of water in the CNV, RPV and steam generators. Note that the SG water mass is on the secondary vertical axis for viewing convenience.

.............................................................................................................................. 17 Figure 3.7 LEC-06T: Cladding temperatures of ring 1 of the core for each node (node 5 being the lowest and node 11 being the highest). ........................................................... 18 Figure 3.8 LEC-06T: Cladding temperatures of upper fuel elevation (5th highest of 6 axially stacked fuel nodes) for each ring and the peak cladding temperature of all fuel nodes..................................................................................................................... 18 Figure 3.9 LEC-06T: Hydrogen generated. ............................................................................ 19 Figure 3.10 LEC-06T: Fraction of radionuclide inventory released from the fuel. ................. 19 Figure 3.11 LEC-06T: Fraction of the radionuclide inventory airborne in containment. ........ 20 Figure 3.12 LCC-05T: Short-term pressure in the CNV, RPV and SGs................................ 23 Figure 3.13 LCC-05T: Long-term pressure in the CNV, RPV and SGs. ............................... 23 Figure 3.14 LCC-05T: Short-term swollen liquid level in the RPV, CNV and SGs. ............... 24 Figure 3.15 LCC-05T: Long-term swollen liquid level in the RPV, CNV and SGs. ............... 24 Figure 3.16 LCC-05T: Integrated mass flowing through the ECCS valves. Note that the three RVV curves are superimposed. ........................................................................... 25 Figure 3.17 LCC-05T: Integrated mass flowing through the CVCS break. ........................... 25 Figure 3.18 LCC-05T: Mass balance of water in the CNV, RPV and steam generators....... 26 Figure 3.19 LCC-05T: Cladding temperatures of upper fuel elevation (5th highest of 6 axially stacked fuel nodes) for each ring and the peak cladding temperature of all fuel nodes. .................................................................................................................. 26 Figure 3.20 LCC-05T: Hydrogen generated. ........................................................................ 27 Figure 3.21 LCC-05T: Fraction of the core inventory released from the fuel. ....................... 27 Figure 3.22 LCC-05T: Fraction of the core inventory airborne in containment. .................... 28 Figure 3.23 LCU-03T: Short-term pressure in the CNV, RPV and SGs................................ 31 Figure 3.24 LCU-03T: Long-term pressure in the CNV, RPV and SGs. ............................... 31 Figure 3.25 LCU-03T: Short-term swollen liquid level in the RPV, CNV and SGs. ............... 32 Figure 3.26 LCU-03T: Long-term swollen liquid level in the RPV, CNV and SGs. ............... 32 Figure 3.27 LCU-03T: Integrated mass flowing through the CVCS break. ........................... 33 Figure 3.28 LCU-03T: Mass balance of water in the CNV, RPV and SGs. The vendors SG mass does not include water in the feed and steam lines while the staffs does. 33 Figure 3.29 LCU-03T: Cladding temperatures of upper fuel elevations for the three rings and the peak cladding temperature. ........................................................................... 34 v

FSCB-19-01 Figure 3.30 LCU-03T: Hydrogen generated. ....................................................................... 34 Figure 3.31 LCU-03T: Mass fraction released in-vessel. ..................................................... 35 Figure 4.1 LEC-06T: Absolute pressure in the CNV, RPV and steam generators (detailed). 36 Figure 4.2 LEC-06T: Integrated mass flowing through the ECCS RVVs for the staffs base and sensitivity cases and vendors case. .............................................................. 37 Figure 4.3 Collapsed liquid level in the RPV and CNV for the staffs base and sensitivity cases and vendors case. ...................................................................................... 37 Figure 4.4 LEC-06T: Comparison of short-term pressure in the RPV, CNV, and SGs........... 38 Figure 4.5 LEC-06T: Comparison of long-term pressure in the RPV, CNV, and SGs. ........... 39 Figure 4.6 LEC-06T: Comparison of collapsed water level in the RPV, CNV, and SGs......... 39 Figure 4.7 LEC-06T: Integrated mass flowing through the RVVs. Note that RVVs 2 and 3 are identical. ................................................................................................................ 40 Figure 4.8 LEC-06T: Collapsed liquid level in the RPV and CNV. ......................................... 41 Figure 4.9 LCC-05T: Short-term pressure in the CNV, RPV and SGs. .................................. 42 Figure 4.10 LCC-05T: Short-term collapsed liquid level in the RPV and CNV...................... 43 Figure 4.11 LCC-05T: Long-term collapsed liquid level in the RPV and CNV. ..................... 43 Figure 4.12 LCC-05T: Integrated mass flowing through the CVCS break. ........................... 44 Figure 4.13 LCC-05T: Integrated mass flowing through the ECCS valves. Note that the three RVV curves are identical. .................................................................................... 44 Figure 4.14 LEC-06T: Pressure in the RPV, CNV, and SGs. ............................................... 46 Figure 4.15 LEC-06T: Cladding temperatures of ring 1 of the core for each node (node 5 being the lowest and node 11 being the highest). ............................................... 46 Figure 4.16 LEC-06T: Fraction of radionuclide inventory released from the fuel. ................. 47 Figure 4.17 Pressure in SG 1, RPV and CNV for the staffs sensitivity case and the vendors calculation. .......................................................................................................... 48 Figure 4.18 Release fractions from the fuel for the staffs sensitivity case and the vendors calculation. .......................................................................................................... 49 Figure 4.19 Water level in the RPV and CNV for the staffs sensitivity case and the vendors calculation. .......................................................................................................... 49 Figure 4.20 Vapor temperatures in the RPV riser, SG 1, and Core region for the staffs sensitivity and base cases. .................................................................................. 50 Figure 5.1 Radionuclide inventory in containment as a fraction of total core inventory. ......... 52 Figure 5.2 Distribution of iodine in containment as a fraction of total core inventory. ............ 53 Figure 5.3 Containment removal rate and airborne aerosol concentration as reported in the Source Term Topical Report for NuScale. ............................................................ 55 Figure 5.4 Containment removal rate and airborne aerosol concentration as calculated from the staffs LEC-06T LOCA MELCOR simulation. .................................................. 55 Figure 5.5 Containment removal rate (averaged over 30 minute intervals) and airborne aerosol concentration as calculated from the staffs LEC-06T MELCOR simulation.

.............................................................................................................................. 56 Figure 5.6 Containment leak rate calculated from staffs MELCOR calculation versus the applicants assumed leak rate. .............................................................................. 57 Figure 5.7 Containment temperature and pressure. ............................................................... 58 Figure 5.8 Net mass density of materials in upper containment. ............................................ 58 Figure 5.9 Release fraction to the environment as calculated from the staffs MELCOR simulation. ............................................................................................................. 59 Figure A.1 Pressure history. ................................................................................................... 62 vi

FSCB-19-01 Figure A.2 Temperature history. ............................................................................................. 62 Figure A.3 Vapor density history. ............................................................................................ 63 Figure A.4 Leakage flow rate history. ..................................................................................... 63 vii

FSCB-19-01 List of Acronyms BAF Bottom of Active Fuel CDF Core Damage Frequency CE Combustion Engineering CNV Containment Vessel CVCS Chemical and Volume Control System DCA Design Certification Application DF Decontamination Factor DHRS Decay Heat Removal System EAB Exclusion Area Boundary ECCS Emergency Core Cooling System FSAR Final Safety Analysis Report IAB Inadvertent Actuation Block iPWR integral Pressurized Water Reactor LOCA Loss of Coolant Accident LPZ Low Population Zone MSIV Main Steam Isolation Valve NRC Nuclear Regulatory Commission PRA Probabilistic Risk Assessment PWR Pressurized Water Reactor PZR Pressurizer RCS Reactor Coolant System RPV Reactor Pressure Vessel RRV Reactor Recirculation Valve RSV Reactor Safety Valve RVV Reactor Vent Valve SA Severe Accident SG Steam Generator SGTR Steam Generator Tube Rupture SMR Small Modular Reactor TAF Top of Active Fuel viii

FSCB-19-01 Official Use Only - Proprietary Information Intentionally left blank ix

FSCB-19-01

1. Introduction This report documents the NRC staffs independent, confirmatory analysis of the severe accident simulations performed by NuScale Power, LLC. While the date of this report is April 2019, it was largely completed in the summer of 2018. The applicants data cited and referenced in this report do not reflect updates or changes since that time.

1.1. NuScale small modular reactor NuScale Power LLC submitted their Standard Design Certification Application (DCA) to the NRC for the NuScale Small Modular Reactor design in December 2016 (Reference 1). The NuScale Small Modular Reactor (SMR) is an integral Pressurized Water Reactor (iPWR) with a rated thermal power of 160 MW(t) and an electrical power output of 50 MW(e). The reactor coolant system (RCS) consists of a single, integrated reactor pressure vessel (RPV) that contains the reactor core, steam generators, and pressurizer. The steam generators (SG) consist of helical coils where, unlike large operating PWRs, primary side coolant flows on the outside of the tubes with steam being generated within the tubes. The containment consists of a steel vessel (referred to as the CNV) that is held at a vacuum during normal operations. There are no reactor coolant pumps used for forced flow through the core as the design depends upon buoyancy driven natural circulatory flow.

The NuScale SMR design includes an ECCS which provides protection for loss of coolant accidents. It consists of two Reactor Recirculation Valves (RRVs) toward the bottom of the RPV and three Reactor Vent Valves (RVVs) at the top of the RPV. Each valve includes an Inadvertent Actuation Block (IAB) which prevents the valve from opening until the differential pressure between the RPV and CNV is sufficiently low. Upon ECCS actuation, all five of these valves receive a signal to open. Once the IAB clears, the valves open and form a natural convection loop where steam escaping the RVVs condenses on the cooler CNV inner wall and re-enters the RPV through the RRVs. The Decay Heat Removal System (DHRS) is a heat exchanger that, when activated, passively removes decay heat via the steam generators into the reactor pool outside of containment.

1.2. Guidance for staff confirmatory analysis Section 19.0, Probabilistic Risk Assessment and Severe Accident Evaluation for New Reactors, of NUREG-0800, Standard Review Plan states the following:

For DC applications and COL applications not referencing the Level 2 PRA in the DC, the reviewer6 carries out an independent assessment of the plant response to selected severe accident scenarios using the latest version of the MELCOR computer code. The assessment should examine accident scenarios from the PRA, which are chosen based on a combination of frequency, consequence, and dominant risk. Some of these scenarios should be similar or identical to sequences analyzed by the applicant and reported in the PRA. The reviewer compares the results of corresponding sequences and release categories in the two studies. If the results of the assessment do not support and confirm the applicants simulation of the accident progression, analysis methodology, and interpretations of its analyses of the reactor, containment, and system 1

FSCB-19-01 response to severe accidents, the reviewer engages with the applicant to resolve the differences in results.

6 Support from an independent contractor or staff in the Office of Research may be necessary.

The staff followed the above Standard Review Plan 19.0 guidance in performing its independent confirmatory analysis. In addition, the staff used results from one of the scenarios to assist the staff in evaluating the NuScale Accident Source Term Methodology Topical Report (Reference 5).

1.3. Outline of this report Section 2 provides an overview of the NuScale design, the modeling approach used in the MELCOR input deck and a description of the process used for selecting the scenarios used for the staffs independent confirmatory analysis. Section 3 presents the staffs MELCOR results for each of the selected scenarios and includes comparisons to NuScales MELCOR results.

Section 4 gives the results of sensitivity calculations that were conducted by the staff. Finally, Section 5 provides results from one of the scenarios to assist the staff in evaluating the NuScale Accident Source Term Methodology Topical Report.

2

FSCB-19-01 Intentionally left blank 1

FSCB-19-01 a Charging line break at ft elevation with in2 hole area.

b Charging line break at ft elevation with in2 hole area.

c This is an earlier version of the code that no longer under active development CVCS line breaks are assumed to be on the injection side.

ER-P060-7085 is used for both the Chapter 19 and SAMDA analysis.

CDF values are taken from the following sources:

- the SAMDA values are from Table B-17 of the Environmental Report

- the ER-P060-4715 value is from Table 19.1-17 of the FSAR

- the ER-P060-4749/ER-P060-4857 and ER-P060-4750 values are from Figure 19.1-13 of the FSAR EE-P060-3637 is based on an older version of the NuScale design. Differences include number and size of ECCS valves and IAB setpoint.

EE-P060-5275 assumes DHRS deactivates when core damage starts.

8

FSCB-19-01 The staff reviewed the scenarios and their CDFs listed in Table 2.2 to select scenarios for confirmatory analysis. The available CDFs for Chapter 19 were the internal events CDF. The available CDFs for the Environmental Report were integrated CDFs for the range of initiators analyzed by the applicant. The Source Term Topical Report did not provide CDFs. Because of the availability of integrated CDFs from the Environmental Report and the overlap between the scenarios analyzed for the Environmental Report and the other two areas (Chapter 19 and Siting), the staff used the CDFs from the Environmental Report to select scenarios as discussed below.

The highest CDF scenario is the module drop accident. The module-drop analysis for both the Environmental Report and Chapter 19 is documented in Reference 7. Because of the importance of reactor pool scrubbing in determining the environmental release for this scenario, the staff did not see a need to perform MELCOR confirmatory calculations with the module on its side. The staff instead chose to focus its review on the applicants evaluation of reactor pool scrubbing. The staffs review of the applicants evaluation of reactor pool scrubbing is document in the staffs SER.

The next highest CDF scenario is a spurious RVV opening with partial ECCS actuation (remaining 2 RVVs open and no RRVs open) with DHRS unavailable. This scenario was also analyzed for Chapter 19 (LEC-06T in Table 2.2). A similar scenario was analyzed for Siting, but the Siting scenario included DHRS. The staff selected this scenario (without DHRS) for MELCOR confirmatory analysis.

The two next highest CDF scenarios are a) 3 RVVs open plus pressurizer heater port failure and b) 3 RVVs open plus stuck-open reactor safety valve. These two scenarios are similar to the 3 RVVs open scenario already being assessed in the staffs confirmatory analysis (see paragraph above), because these two scenarios have similar hole sizes and hole elevations to the 3 RVVs open scenario. Therefore, the staff did not select these two scenarios for confirmatory analysis.

The next highest CDF scenario is a CVCS line break inside containment with 3 RVVs open (RRVs fail to open). This scenario also was analyzed for Chapter 19 (LCC-05T in Table 2.2) and Siting. The staff selected this scenario for MELCOR confirmatory analysis because of the uniqueness of the CVCS break which is low in the RPV (i.e., a liquid break).

The next highest CDF scenario is the CVCS line break outside containment. This scenario also was analyzed in Chapter 19 (LCU-03T in Table 2.2) and has the largest source term. The staff selected this scenario for confirmatory analysis.

NuScale also analyzed a steam generator tube failure accident resulting in containment bypass.

However, this scenario was not chosen for confirmatory analysis, because its CDF is lower than the CVCS line break outside containment.

A summary of the scenarios selected for confirmatory analysis is given in Table 2.3 The staff applied its MELCOR model to each of the scenarios in Table 2.3. The staff compared the results of each its MELCOR simulations with the results for applicants MELCOR simulations. The detailed results provided by the applicant in References 9 and 10 were used for the comparison. The results of the comparisons are given in the Section 3.

9

FSCB-19-01 Table 2.3 Scenarios Selected for Confirmatory Analysis Scenario Environmental Chapter 19 Chapter 15 CDF (per Report year)1 3 RVVs open 2 2.4E-9 CVCS break plus 3 RVVs open2 2.2E-11 CVCS break outside containment 2 1.7E-11 1

The CDF values shown are for SAMDA (Environmental Report) and are for the range of hazards and represent the sum of the sequences associated with the release category.

2 The staffs confirmatory analysis assumes DHRS does not operate.

10

FSCB-19-01

3. MELCOR Results and Comparison with the Applicants Results This section presents the staffs independent MELCOR analysis with comparisons to the applicants results. Additional staff sensitivity MELCOR analysis are given in Section 4 of this report.

3.1. LEC-06T This scenario, referred to as LEC-06T in the applicants FSAR, is initiated by the spurious opening of a single RVV followed by partial actuation of the ECCS (other two RVVs open but both RRVs remain closed). The DHRS is assumed to be unavailable.

A summary of the key event timings is given in Table 3.1. Figure 3.1 through Figure 3.11 compare the simulation results of the staff and applicant. In the following discussion, the numbers cited are for the staffs MELCOR simulations unless otherwise stated.

Figure 3.1 shows the pressure in the RPV, CNV and SG in the first 15 minutes of the accident.

When the first RVV opens spuriously at time 0, the RPV and CNV pressures equalize after 2 minutes. Even though all three of the RVVs are open by 0.6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months , it takes 1.6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months and 2.6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months for the RPV pressure to decrease to 200 psi and 100 psi, respectively. After 2 minutes, flow from the RCS to containment is primarily steam that is generated in the core, enters containment and is condensed on the containment inner surface. Containment (along with RCS) pressure drops below atmospheric pressure at 7.9 hours1.041667e-4 days 0.0025 hours 1.488095e-5 weeks 3.4245e-6 months due to this condensation. As for the SGs, pressure initially spikes upon closure of the MSIVs and then drops with the corresponding drop in temperature in the SG.

Figure 3.2 shows the long-term pressure in the RPV, SGs (the two SGs are identical and therefore conditions within are identical as well), and CNV. Since the SG MSIVs close when the reactor trips, the SGs become isolated. When core damage and relocation occur, the temperature of the SGs slowly increases. Consequently, the pressure in the SGs also increases, rising to the SG relief valve setpoint around 23 hours2.662037e-4 days 0.00639 hours 3.80291e-5 weeks 8.7515e-6 months , passing additional water to containment. As for the RPV, it is open to containment via the 3 open RVVs so its pressure is the same as the CNV and lower than that of the SGs.

When the RPV water level falls below the bottom of the SGs at 1.4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months , the rate of cooling decreases since the SGs are hydraulically isolated. Consequently, the rate of depressurization in the SGs also slows. This phenomena is not as apparent in the applicants calculations where the SG temperature and pressure continue to decrease at a similar rate after the water level falls below the SGs. The difference in SG pressure between the staff and applicants calculations was discussed during an audit beginning March 6, 2018 (Reference 11). The applicant confirmed that no valves or systems (such as DHRS) had been left open. The applicant suggested it could be related to the staffs model not including heat transfer from the pipes connected to the SG (i.e., feed lines and steam lines) to the containment atmosphere whereas the applicant did include these structures.

The applicants RPV and CNV pressure is lower than what is predicted by the staff (about 30%

lower) in the first few minutes. Several explanations for this difference were posited and led to sensitivity calculations which are documented in Section 4. One explanation for the difference in pressure is that the RVV flow areas and form loss coefficients are different between the staff 11

FSCB-19-01 and applicants MELCOR input models. The staff requested as part of the 2017 PRA and severe accident audit (Reference 12) from the applicant the flow area of the RVVs and found that it had changed since the applicant had performed their MELCOR calculations. Staff used the new flow area in the confirmatory calculations and, as a result, the RVV flow area used in the staffs deck is smaller than that in the applicants deck (17.3 in2 versus in2). The sensitivity calculations documented in Section 4 showed that this difference in valve size had little impact on the early pressure difference and the integrated flow of water mass lost through the ECCS valves did not differ greatly from the base case, even in the long-term. Another sensitivity study documented in Section 4 brought to light another possible reason the early differences in RPV, CNV, and SG pressure. By increasing the number of nodes of the SGs both on the primary and secondary side, there was an increase in heat transfer from primary to secondary and the SG pressure slightly faster in the first few minutes of the transient, as shown in sensitivity calculations documented in Section 4. However, this too does not fully account for the differences in pressure.

Shown in Figure 3.3, there is a small difference in total core power between the staff and the applicant. Figure 3.4 gives the predicted water levels with an associated image of the NuScale vessel for perspective. The RPV water level decreases rapidly with the first RVV opening since liquid water is passed initially (see Figure 3.5). From then on, the water level decreases steadily with steam passing through the three open RVVs to the containment where it condenses and partially fills the CNV. RPV water level reaches the top of active fuel (TAF) at 5.3 hours3.472222e-5 days 8.333333e-4 hours 4.960317e-6 weeks 1.1415e-6 months (this is the collapsed level, the swollen level reaches TAF at 6.7 hours8.101852e-5 days 0.00194 hours 1.157407e-5 weeks 2.6635e-6 months ). Figure 3.6 gives the overall mass balance of water (steam and liquid water) in the RPV, CNV, and SGs and demonstrates the exchange of water between the three components. The drop in water mass in the SGs around 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months in the staffs calculations corresponds to the steam generator relief valve opening on high secondary pressure and relieving into containment.

Following core uncovery, core heat up and degradation occurs (see Figure 3.7 and Figure 3.8).

Cladding oxidation begins at 8.7 hours8.101852e-5 days 0.00194 hours 1.157407e-5 weeks 2.6635e-6 months (Figure 3.9) and gap release in ring one of the core occurs at 8.8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months . In the staffs deck, the core is nodalized as three radial rings with 9, 14, and 14 assemblies respectively. In the applicants deck, the core is nodalized as four radial rings with , and assemblies respectively. In the staffs calculation, the two upper levels of ring 1collapse at 10 hours1.157407e-4 days 0.00278 hours 1.653439e-5 weeks 3.805e-6 months followed closely by the top level of ring 2 collapsing. It remains in this state until around 18 hours2.083333e-4 days 0.005 hours 2.97619e-5 weeks 6.849e-6 months when the remainder of the first ring collapses due to loss of strength as a result of being at a high temperature for an extended period of time. The remainder of ring 2 then heats to the point of collapse as well.

In the staffs calculations, up until the time of core damage, water in the RPV core region, lower plenum and downcomer is boiling off, going up through the pressurizer and out the open RVVs to containment. Around the time the core starts to heat up at 8.7 hours8.101852e-5 days 0.00194 hours 1.157407e-5 weeks 2.6635e-6 months , a flow pattern begins with steam flow going from the heating core, up the riser, turning around at the PZR baffle plate, down the downcomer, and back into the riser through a bypass flowpath at the 3 meter elevation. The flow rate appears to provide enough cooling to the riser and core that the collapse of the remainder of the lower nodes of ring one is delayed by several hours as can be seen in Figure 3.7 as well as the subsequent failure of ring 2 being delayed. During the audit beginning March 6, 2018 (Reference 11), the applicant stated that this bypass flowpath has been removed from the NuScale design and does not exist in the applicants MELCOR deck.

Hence, in the applicants analysis, flow from the downcomer to the core only occurs after the 12

FSCB-19-01 downcomer water level falls below the support plate. The applicants predicted SG pressure rise at hours is due to the water level going below the lower core support plate and this natural circulation beginning. Two sensitivity calculations are given in Section 4 for scenarios LEC-06T and LCC-05T with this bypass flowpath removed from the staffs deck. The removal had little impact on the LCC-05T calculation, but did have an impact on event times in the LEC-06T calculation.

Between 10.6 and 17.6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months , the bottom two nodes of ring one and the bottom four nodes (out of 6) of ring two are still upright. At 17.6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months , the RPV water level drops below the BAF. The core heats up due to core uncovery and ongoing oxidation of the zircaloy cladding. The remaining upright portion of ring 1 heats up over the following hour and collapses at 18.5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months with the remaining upright portion of ring 2 collapsing a few minutes later. With debris accumulating on the core support plate, it heats up leading to its failure in rings 1 and 2 fail at 19.5 and 20.7 hours8.101852e-5 days 0.00194 hours 1.157407e-5 weeks 2.6635e-6 months , respectively. At the end of the staffs simulation (as of 48 hours5.555556e-4 days 0.0133 hours 7.936508e-5 weeks 1.8264e-5 months ), ring 3 remains upright and rings 1 and 2 are a rubble bed sitting in the lower plenum and lower core region. Prior to a deck correction (water elevation in one of the RPV control volumes was corrected leading to additional water inventory in the RPV), this same scenario led to the core support plate of the outermost ring failing very late in the accident (47 hours5.439815e-4 days 0.0131 hours 7.771164e-5 weeks 1.78835e-5 months ) and ring 3 collapsing into rubble. Hence, there is some uncertainty associated with the timing and extent of core failure.

The release fraction of fission products from the fuel is shown in Figure 3.10 with the airborne fraction in containment given in Figure 3.11.

13

FSCB-19-01 Table 3.1 A comparison of event timings in simulation LEC-06T for both the staff and applicants MELCOR calculations.

LEC-06T Time (seconds)

Event Applicant Staff RVV LOCA 0 0 CNV Isolation signal 0 0 SCRAM 2 2 IAB cleared ** 11 Maximum CNV pressure 40 70 ECCS signal on low RPV level ** 2213 Partial ECCS actuation 1010 2213 ECCS signal on high CNV level 1010 4617 PZR heater isolation* 1432 0 Low low PZR level (20%) 2319 1380 RPV collapsed level below TAF 16200 (4.5hr) 19000 (5.3 hr)

High core outlet temperature (610F) 21973 (6.1hr) 30400 (8.4 hr)

Onset of cladding oxidation 24360 (6.8hr) 31259 (8.7 hr)

Gap release in ring 1 24539 (6.8hr) 31542 (8.8 hr)

Core damage (PCT>2200F) 25223 (7hr) 32600 (9.1 hr)

Gap release in ring 2 ** 32889 (9.1 hr)

Gap release in ring 3 ** 35171 (9.8 hr)

Maximum cladding temperature 29640 (8.2hr) 35560 (9.9 hr)

RPV collapsed level below BAF ** 63400 (17.6 hr)

Failure of core support plate (ring 1) 41785 (11.6hr) 70033 (19.5 hr)

Failure of core support plate (ring 2) ** 74559 (20.7 hr)

End of cladding oxidation 95400 (26.5hr) 104100 (28.9 hr)

End of simulation 259200 (72hr) 172800 (48 hr)

In the staffs MELCOR model, the PZR heater is not explicitly modeled. Instead a control function is used during the steady state to add enthalpy to the pressurizer. When the steady-state calculation is complete (at time 0), the enthalpy addition is assumed to end.

- Data not available.

14

FSCB-19-01 Figure 3.1 LEC-06T: Short-term pressure in the CNV, RPV and SGs.

Figure 3.2 LEC-06T: Long-term pressure in the CNV, RPV and SGs.

15

FSCB-19-01 Figure 3.3 LEC-06T: Total core power.

Figure 3.4 LEC-06T: Collapsed liquid level in the RPV, CNV and steam generators.

16

FSCB-19-01 Figure 3.5 LEC-06T: Integrated mass flowing through the ECCS valves. Note that the RVV #2 and

3 curves are identical.

Figure 3.6 LEC-06T: Mass balance of water in the CNV, RPV and steam generators. Note that the SG water mass is on the secondary vertical axis for viewing convenience.

17

FSCB-19-01 Figure 3.7 LEC-06T: Cladding temperatures of ring 1 of the core for each node (node 5 being the lowest and node 11 being the highest).

Figure 3.8 LEC-06T: Cladding temperatures of upper fuel elevation (5th highest of 6 axially stacked fuel nodes) for each ring and the peak cladding temperature of all fuel nodes.

18

FSCB-19-01 Figure 3.9 LEC-06T: Hydrogen generated.

Figure 3.10 LEC-06T: Fraction of radionuclide inventory released from the fuel.

19

FSCB-19-01 Figure 3.11 LEC-06T: Fraction of the radionuclide inventory airborne in containment.

20

FSCB-19-01 3.2. LCC-05T This scenario, referred to as LCC-05T in the applicants FSAR, is initiated by a break in the CVCS line at elevation ft, inside of containment. Due to an oversight, the staffs MELCOR simulation has the break at 13 meters (42.7 ft.) which is near the pressurizer baffle plate. A sensitivity calculation is performed in Section 4 with the break location corrected to feet and shows little impact on the results of the simulation. Upon ECCS actuation and after the IAB allows ECCS valves to open, there is a partial failure (all three RVVs open and both RRVs remain closed). The DHRS is assumed to be unavailable.

A summary of the event timings is given in Table 3.2. Figure 3.12 through compare the simulation results of the staff and applicant. In the following discussion, the numbers cited are for the staffs MELCOR simulations unless otherwise stated.

Figure 3.12 gives the pressure in the RPV, CNV and SGs (the two SGs are identical and conditions within are also identical) in the first 30 minutes of the accident. As stated before, the initiating event is a break in the CVCS line into containment. The CVCS line break is a in2 pathway from a location above the top of the core in the riser (CV174 in Figure 2.1) to the space between the RPV and CNV (CV413 in Figure 2.2). As a result of the break, the RPV and SG pressures decrease. ECCS actuates and the RVVs open on low RPV level at 790 seconds. As steam condenses on the CNV wall, pressure in the CNV and RPV decreases. In the staffs calculations both here and in LEC-06T, the SG pressure follows RPV pressure until the water level falls below the bottom of the SGs (see Figure 3.13). This is not seen as significantly in the applicants calculation. This was discussed with the applicant during the March 2018 audit. (See discussion in the LEC-06T section above).

The long-term pressure is given in Figure 3.13. As mentioned in the LEC-06T discussion, the staffs MELCOR deck included a bypass flowpath that allows a recirculatory flow pattern to begin early on (around 2.6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months ) and increase the heat transfer to the SG tubes and cause a gradual increase in SG pressure. It also provides cooling to the area above the core. This bypass flowpath is not a part of the current NuScale design and was therefore not a part of the applicants MELCOR model. Hence, a flow from the downcomer to the core region only occurs in the applicants calculation when the RPV downcomer water level falls below the core support plate. In the applicants calculation, then, SG temperature and pressure rises more rapidly and at a later time (around hours). Section 4.5 describes a sensitivity calculation of the staffs LCC-05T calculation without the bypass flowpath included (see discussion in the LEC-06T section above). However, staff were still unable to replicate the sharp depressurization of the SG seen in the applicants calculations.

Water levels are given in Figure 3.14 and Figure 3.15. The water levels are similar between the staffs and applicants MELCOR simulations. Figure 3.18 shows that the mass of water in the RPV, CNV, and SGs for the staff and applicants simulations has good long term agreement regardless of the differences in the RVV modeling.

Following core uncovery, core heat up and degradation occurs (see Figure 3.19). Cladding oxidation begins at 3.4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months and gap release in ring one of the core occurs at 3.5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months . Recall that in the staffs deck, the core is nodalized as three radial rings with 9, 14, and 14 assemblies respectively while the licensees deck is nodalized as four radial rings with 21

FSCB-19-01 assemblies. In the staffs simulation, the upper levels of the ring 1 heat up and collapse at 4.4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months . The downward relocation of core debris into the water generates steam which temporarily cools the core. The remainder of ring 1 collapses 2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months later along with all of ring 2 after the water level has fallen below the BAF. Ring three is much slower to heat up since it has heat transfer to the heavy reflector and CNV vessel and collapses at 16.7 hours8.101852e-5 days 0.00194 hours 1.157407e-5 weeks 2.6635e-6 months .

There is more hydrogen generated in the applicants simulation as seen in Figure 3.20. In the applicants simulation, the two inner rings of the core remain upright for a greater period of time and the remaining two rings remain upright indefinitely. This allows for more interaction of steam with the zircaloy cladding and thereby hydrogen generation. In contrast, in the staffs simulation, rings one and two collapse before 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months and the outermost ring collapses around 16 hours1.851852e-4 days 0.00444 hours 2.645503e-5 weeks 6.088e-6 months .

Around 8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months , hydrogen generation slows and at 16 hours1.851852e-4 days 0.00444 hours 2.645503e-5 weeks 6.088e-6 months , it has ceased.

The release fraction of fission products from the fuel is shown in Figure 3.21 with the airborne fraction in containment given in Figure 3.22.

Table 3.2 A comparison of event timings in simulation LCC-05T for both the staff and applicants MELCOR calculations.

LCC-05T Timing (seconds)

Event Applicant Staff CVCS LOCA 0 0 SCRAM 3 3 PZR heater isolation* 40 3 Low low PZR level (20%) 90 240 ECCS signal on low RPV level ** 790 IAB cleared ** 900 Maximum CNV pressure 700 900 ECCS signal on high CNV level 681 915 RPV collapsed level below TAF 7860 (2.2hr) 7720 (2.2 hr)

High core outlet temperature (610F) 14625 (4.1hr) 10600 (2.9 hr)

Onset of cladding oxidation 16440 (4.6hr) 12179 (3.4 hr)

First gap release (ring 1) 16593 (4.6hr) 12424 (3.5 hr)

Core damage (PCT>2200F) 17190 (4.8hr) 13260 (3.7 hr)

Gap release in ring 2 ** 13835 (3.8 hr)

Gap release in ring 3 ** 15417 (4.3 hr)

Maximum cladding temperature 21120 (5.9hr) 16260 (4.5 hr)

RPV collapsed level below BAF ** 19644 (5.5 hr)

Failure of core support plate (ring 1) 33963 (9.4hr) 29146 (8.1 hr)

Failure of core support plate (ring 2) ** 37085 (10.3 hr)

Failure of core support plate (ring 3) ** 60206 (16.7 hr)

End of cladding oxidation 150000 (41.7hr) continues End of simulation 259200 (72hr) 172800 (48 hr)

In the staffs MELCOR model, the PZR heater is not directly modeled. Instead a control function is used during the steady state to add enthalpy to the pressurizer. When the reactor SCRAMs, the control function ceases.

- Data not available.

22

FSCB-19-01 Figure 3.12 LCC-05T: Short-term pressure in the CNV, RPV and SGs.

Figure 3.13 LCC-05T: Long-term pressure in the CNV, RPV and SGs.

23

FSCB-19-01 Figure 3.14 LCC-05T: Short-term swollen liquid level in the RPV, CNV and SGs.

Figure 3.15 LCC-05T: Long-term swollen liquid level in the RPV, CNV and SGs.

24

FSCB-19-01 Figure 3.16 LCC-05T: Integrated mass flowing through the ECCS valves. Note that the three RVV curves are superimposed.

Figure 3.17 LCC-05T: Integrated mass flowing through the CVCS break.

25

FSCB-19-01 Figure 3.18 LCC-05T: Mass balance of water in the CNV, RPV and steam generators.

Figure 3.19 LCC-05T: Cladding temperatures of upper fuel elevation (5th highest of 6 axially stacked fuel nodes) for each ring and the peak cladding temperature of all fuel nodes.

26

FSCB-19-01 Figure 3.20 LCC-05T: Hydrogen generated.

Figure 3.21 LCC-05T: Fraction of the core inventory released from the fuel.

27

FSCB-19-01 Figure 3.22 LCC-05T: Fraction of the core inventory airborne in containment.

28

FSCB-19-01 3.3. LCU-03T This scenario, referred to as LCU-03T in the applicants FSAR, is initiated by a break in the CVCS line at elevation ft., outside of containment (CV518 in Figure 2.3). ECCS is assumed not to actuate and the DHRS is assumed unavailable. Feedwater injection is assumed to end at the time of reactor SCRAM.

A summary of the event timings is given in Table 3.3. Figure 3.23 through Figure 3.31 compare the simulation results of the staff and applicant. In the following discussion, the numbers cited are for the staffs MELCOR simulations unless otherwise stated.

The staffs results reported in this section terminated a few hours after RPV lower head failure and corium relocation to the containment lower head. The cause of the termination was determined to be an issue with MELCORs new containment lower head model which becomes active following corium relocation to the containment lower head. By this time, in-vessel core-melt progression has ended and because this is a bypass event, the staff decided that additional effort on this confirmatory analysis was not cost beneficial. 2 Figure 3.23 gives the pressure in the RPV, CNV and SG in the first 30 minutes of the accident.

As stated before, the initiating event is a break in the CVCS line outside of containment. Since this is a bypass accident, CNV pressure remains constant at 0 psia. Because it is a break of the CVCS injection line, water comes out of the break until the RPV water level decreases to the bottom of the CVCS line. As a result, Figure 3.23 shows that the RPV pressure is steady for the first 30 minutes of the accident. Figure 3.24 shows the longer-term RPV pressure reduction as a result of the CVCS line break outside containment.

The water level in the RPV, CNV and SGs is given in Figure 3.25 and Figure 3.26. RPV water is lost through the CVCS break. As is seen in Figure 3.27 the leak rate of the CVCS break is greater in the applicants analysis than in the staffs resulting in core uncovery occurring an hour sooner in the applicants analysis.

Figure 3.28 gives the water inventory of the RPV, CNV, and SGs. Since this is a bypass event, water that is lost from the RPV does not make its way into containment. The 1200 kg of water that appears in the CNV around 6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months is from the SG relief valve opening into containment.

Figure 3.29 provides the temperature of the uppermost elevation of the core for each ring. The temperature going to zero indicates the fuel in the cell is no longer intact (formation of debris).

By 10.2 hours2.314815e-5 days 5.555556e-4 hours 3.306878e-6 weeks 7.61e-7 months , all three rings of the lower core plate have failed resulting in downward relocation of the core into the lower plenum. Figure 3.30 gives the hydrogen generated by the zircaloy/steam reaction and Figure 3.31 gives the release fraction of volatile fission products from the fuel.

2 Following completion of this analysis, the code error was corrected an a rerun of this sequence with the new code version showed that there are some differences in the results following core damage, but well within uncertainties one would expect from a severe core damage accident.

29

FSCB-19-01 Table 3.3 Comparison of event timing for both NuScale and NRC staffs MELCOR calculations of LCU-03T LCU-03T Timing (seconds)

Event Applicant Staff CVCS LOCA 0 0 SCRAM on low PZR level (<35%) 69 69 Low PZR pressure 87 488 Low low PZR level (20%) 92 700 ECCS signal on low RPV level 1259 2269 IAB cleared ** 6380 (1.8 hr)

RPV collapsed level below TAF 4572 (1.3 hr) 7840 (2.2 hr)

Onset of cladding oxidation 8760 (2.4 hr) 13698 (3.8 hr)

Gap release in ring 1 8942 (2.5 hr) 14010 (3.9 hr)

Core damage (PCT<2200F) 9493 (2.6 hr) 15020 (4.2 hr)

Maximum cladding temperature 11640 (3.2 hr) 15220 (4.2 hr)

Gap release in ring 2 ** 15227 (4.2 hr)

Gap release in ring 3 ** 16662 (4.6 hr)

RPV collapsed level below BAF ** 16754 (4.7 hr)

Failure of core support plate (ring 1) 18110 (5.0 hr) 22887 (6.4 hr)

Failure of core support plate (ring 2) ** 31728 (8.8 hr)

Failure of core support plate (ring 3) ** 36748 (10.2 hr)

Relief valve opens allowing water to CNV 19800 (5.5 hr) 20240 (5.6 hr)

Failure of RPV lower head 42220 (11.7 hr) 57070 (15.9 hr)

End of cladding oxidation 63480 (17.6 hr) continues End of simulation 259200 (72 hr) *

Staffs MELCOR calculation terminated after failure of RPV lower head.

- Data not available.

30

FSCB-19-01 Figure 3.23 LCU-03T: Short-term pressure in the CNV, RPV and SGs.

Figure 3.24 LCU-03T: Long-term pressure in the CNV, RPV and SGs.

31

FSCB-19-01 Figure 3.25 LCU-03T: Short-term swollen liquid level in the RPV, CNV and SGs.

Figure 3.26 LCU-03T: Long-term swollen liquid level in the RPV, CNV and SGs.

32

FSCB-19-01 Figure 3.27 LCU-03T: Integrated mass flowing through the CVCS break.

Figure 3.28 LCU-03T: Mass balance of water in the CNV, RPV and SGs. The vendors SG mass does not include water in the feed and steam lines while the staffs does.

33

FSCB-19-01 Figure 3.29 LCU-03T: Cladding temperatures of upper fuel elevations for the three rings and the peak cladding temperature.

Figure 3.30 LCU-03T: Hydrogen generated.

34

FSCB-19-01 Figure 3.31 LCU-03T: Mass fraction released in-vessel.

35

FSCB-19-01 Figure 4.2 LEC-06T: Integrated mass flowing through the ECCS RVVs for the staffs base and sensitivity cases and vendors case.

Figure 4.3 LEC-06T: Collapsed liquid level in the RPV and CNV for the staffs base and sensitivity cases and vendors case.

37

FSCB-19-01 4.2. SG Nodalization This scenario is a sensitivity calculation based on LEC-06T.

The steam generator nodalization with four axial levels used in the staffs deck (see Figure 2.1) uses fewer volumes than the applicant. A consequence of the bottommost node being so large in the base model was that the heat transfer modeled in the steam generators was coarse and led to an increased overall RCS temperature in the steady-state. For this sensitivity, the topmost node and bottommost nodes of both SGs were split into three vertically stacked nodes and staff then ran LEC-06T with this modified nodalization.

Figure 4.4 shows that, with the finer nodalization, there was improved agreement between the staff and the applicants pressures in the first few minutes following the start of the accident.

In the longer-term, Figure 4.5 shows that the SG pressure in the sensitivity case is the same as the base case up to the time where the RPV water level falls below the TAF and the recirculation flow discussed in Section 3.1 begins. Beyond this time, the pressure difference in the base and sensitivity calculations may come from the difference in initial SG water inventory.

With the finer nodalization, the initial conditions in the SGs are slightly different for the sensitivity. The initial (swollen) water level in the SGs is higher (12.3m versus 10.5m).

The RPV and CNV water levels are given in Figure 4.6. The flow rate through the open RVVs is reduced and the rate of inventory loss is slowed.

Figure 4.4 LEC-06T: Comparison of short-term pressure in the RPV, CNV, and SGs.

38

FSCB-19-01 Figure 4.5 LEC-06T: Comparison of long-term pressure in the RPV, CNV, and SGs.

Figure 4.6 LEC-06T: Comparison of collapsed water level in the RPV, CNV, and SGs.

39

FSCB-19-01 4.3. Timing of ECCS Actuation This scenario is a sensitivity calculation based on LEC-06T.

There was a difference in timing for ECCS valve opening in the staff and applicants calculations. It stems from a change in the applicants ECCS actuation setpoint on CNV actuation level. The applicant changed their ECCS actuation setpoint after they completed their MELCOR analysis. Because of this change, the staffs calculations used a higher level for ECCS actuation than the applicants calculations and, as a result, the 2nd and 3rd RVVs opened at 1,010s versus the staffs 2,213s. A sensitivity calculation of LEC-06T was therefore performed to test the significance of this difference and whether the timing in ECCS actuation contributes to the core uncovering an hour later in the staffs calculations than in the applicants.

The deck was modified to force ECCS valves to open at 1,010s to match the applicants simulation.

This sensitivity resulted in a shift in the integrated flow through the RRVs. Figure 4.7 compares the integrated flow through the RVVs for the staffs two calculations to the applicants calculation. The flow through the first RVV (RVV1) is significantly less and the other two valves discharge more volume. In the end, however, the decreased flow through the first valve is compensated for by the increased flow through the other two. Figure 4.8 demonstrates that there is no impact on the staffs predictions of RPV level and the timing of core uncovery. In Figure 4.8, the RPV-staff line covers the RPV-sens line.

Figure 4.7 LEC-06T: Integrated mass flowing through the RVVs. Note that RVVs 2 and 3 are identical.

40

FSCB-19-01 Figure 4.8 LEC-06T: Collapsed liquid level in the RPV and CNV.

41

FSCB-19-01 Figure 4.10 LCC-05T: Short-term collapsed liquid level in the RPV and CNV.

Figure 4.11 LCC-05T: Long-term collapsed liquid level in the RPV and CNV.

43

FSCB-19-01 Figure 4.12 LCC-05T: Integrated mass flowing through the CVCS break.

Figure 4.13 LCC-05T: Integrated mass flowing through the ECCS valves. Note that the three RVV curves are identical.

44

FSCB-19-01 4.5. Removal of the Bypass Flowpath Between Downcomer and Riser As a result of discussions during an audit beginning March 6, 2018 (Reference 11), the staff found that a m2 area bypass flowpath going from the downcomer (CV110 in Figure 2.1) to the area inside the riser just above the core (CV in Figure 2.1) was eliminated from the NuScale design in 2015. Therefore, the staff ran two sensitivity calculations below (one for LEC-06T and one for LCC-05T) to assess the impact of its removal.

LEC-06T As described in Section 3.1, during a severe accident the bypass flowpath allows a small amount of vapor flow from the RPV downcomer to the riser creating a recirculatory flow of gas in the RPV before the lower core plate is uncovered. The case with bypass flow predicts lower core plate failure for rings 1 and 2 at 18 hours2.083333e-4 days 0.005 hours 2.97619e-5 weeks 6.849e-6 months , while the case without bypass flow predicts lower core plate failure at 13 hours1.50463e-4 days 0.00361 hours 2.149471e-5 weeks 4.9465e-6 months see Figure 3.7 and Figure 4.15). Comparing Figure 3.7 with Figure 4.15 demonstrates this extended time to failure of the bottommost nodes of ring one in the base case.

While the fuel collapses in ring one collapses at 13 hours1.50463e-4 days 0.00361 hours 2.149471e-5 weeks 4.9465e-6 months , the core support plate for ring one fails at 15 hours1.736111e-4 days 0.00417 hours 2.480159e-5 weeks 5.7075e-6 months which accelerates the loss of water inventory. The water level in the downcomer region falls below the lower core plate around 16.5 hours5.787037e-5 days 0.00139 hours 8.267196e-6 weeks 1.9025e-6 months and flow from the downcomer, through the rubble bed in the lower plenum, through the core region and up the hot leg riser begins. When this pattern is established, hot vapor flows over the SG tubes and the water within the tubes heats up. Pressure in the SGs increases more quickly than in the staffs base case (see Figure 3.2), though still not as quickly as in the applicants case as shown in Figure 4.14. Also, the sensitivity case has an increased flow of steam over the upright third ring of the core. The lower core plate of ring 3 fails at 23.4 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months which is in contrast to the original calculation where the lower core plate of ring 3 does not fail. As a result, there is a higher release fraction of fission products in this case with ring 3 heating up as a rubble bed.

In summary, core degradation progresses more quickly in the sensitivity calculation. Failure of the core support plate occurs at 15, 17 and 23 hours2.662037e-4 days 0.00639 hours 3.80291e-5 weeks 8.7515e-6 months for rings 1, 2, and 3, respectively. With the core support plate under all three rings failing in this case, the rubble bed in the lower plenum is larger and heats up more and the release fraction from the fuel is greater (as seen in Figure 4.16).

45

FSCB-19-01 Figure 4.14 LEC-06T: Pressure in the RPV, CNV, and SGs.

Figure 4.15 LEC-06T: Cladding temperatures of ring 1 of the core for each node (node 5 being the lowest and node 11 being the highest).

46

FSCB-19-01 Figure 4.18 Release fractions from the fuel for the staffs sensitivity case and the vendors calculation.

Figure 4.19 Water level in the RPV and CNV for the staffs sensitivity case and the vendors calculation.

49

FSCB-19-01 Figure 4.20 Vapor temperatures in the RPV riser, SG 1, and Core region for the staffs sensitivity and base cases.

50

FSCB-19-01

5. Confirmatory Analysis for the NuScale Accident Source Term Topical Report 5.1. Introduction The NuScale Accident Source Term Methodology Topical Report (Reference 5) provides a method for estimating the EAB, LPZ and control room doses. It also includes a sample analysis of EAB, LPZ and control room doses which is referenced in the FSAR. The topical report covers the source term from the RPV to the containment, the in-containment aerosol removal rate, and containment leak rate. The staffs independent confirmatory analysis performed in support of its review of the topical report is described below. Section 5.2 assesses the source term from the RPV to the containment. Section 5.3 assesses the in-containment aerosol removal rate. Section 5.4 assesses containment leak rate. Section 5.5 assesses the release from the containment to the environment.

The results of the MELCOR base case calculation in Section 3.1 for RVV LOCA (LEC-06T) is used for the staffs independent assessment of the topical report. LEC-06T is initiated by the spurious opening of a single RVV followed by partial actuation of the ECCS (the remaining two RVVs open and both RRVs remain closed). The DHRS is assumed unavailable in this scenario.

The timing of key events for LEC-06T is compared between the applicants and staffs MELCOR calculations in Table 3.1.

5.2. Source Term from the RPV to Containment In the topical report, the release fraction of fission products to containment for the representative source term is determined by taking the median release from four MELCOR calculations based on an earlier version of the NuScale design. The applicant subsequently performed five MELCOR calculations based on the final version of the NuScale design (as described in FSAR Revision 0 and underlying documentation) to confirm the analysis. Table 5.1 shows the results of the five MELCOR calculations based on the final version of the design and also the results of the staffs MELCOR results for LEC-06T.

The topical report states that the release durations into containment are estimated as the time when 90% of the volatile fission products that reach the containment by the end of the accident have entered the containment. The topical report does not state which fission product group (e.g., iodine, cesium) is used for this metric. In the staffs MELCOR simulation, 90% of the iodine enters containment by 40.6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months while 90% of the cesium enters containment by 9.8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months for a duration of either 31.8 or 1.0 hours0 days 0 hours 0 weeks 0 months . For the staffs comparison, the former is chosen since iodine is typically the dominant contributor to EAB, LPZ, and control room doses.

The second through sixth columns of Table 5.1 give the applicants cumulative release fractions into containment for the five MELCOR calculations based on the final version of the NuScale design. The ninth column of Table 5.1 gives the staffs cumulative release fraction into the containment of noble gases, iodine, and cesium at the end of 48 hours5.555556e-4 days 0.0133 hours 7.936508e-5 weeks 1.8264e-5 months . The fraction of core inventory of volatile fission products in containment is given in Figure 5.1 . The curve for noble gases is not flat but reaches a relative peak value of 0.55 before decreasing again due to flow back into the RPV. Similar behavior is not seen for iodine and cesium, because these are in aerosol form and deposit in the containment so cannot flow back into the RPV through the open RVVs. Table 5.1 lists this peak value for the release fraction of noble gases into containment.

51

FSCB-19-01 Figure 5.2 gives the release fraction of iodine that is released from the core and present in the RPV and containment.

Table 5.1 Release fractions and time of the five representative accident scenarios from the applicants MELCOR calculations (Reference 13) as compared to the Staffs MELCOR result. DHRS was available in all scenarios until core damage.

Case Case Case Case Case Median Reg. NRC Staff 1 2 3 4 5 Guide (LEC-06T) 1.183 CVCS CVCS CVCS RRVs RVVs Repres- PWR RVVs break, break, break, open open entative DBA open no RRVs RVVs Source LOCA ECCS open open Term1 Start of gap 17.6 3.8 8.1 6.2 21.3 3.8 30 8.8 (hrs) sec Release 12.0 1.0 9.0 1.3 14.0 1.0 1.8 31.8 duration (hrs)

Releas NG 0.39 0.19 0.41 0.19 0.48 0.39 1.0 0.55 e I 0.21 0.035 0.16 0.019 0.14 0.14 0.4 0.18 fraction Cs 0.25 0.059 0.22 0.031 0.20 0.20 0.3 0.29 1

The representative source term per the topical report methodology uses the shortest time for the start of release and the release duration. It then uses the median release fractions.

Figure 5.1 LEC-06T: Radionuclide inventory in containment as a fraction of total core inventory.

52

FSCB-19-01 Figure 5.2 LEC-06T: Distribution of iodine in containment as a fraction of total core inventory.

5.3. In-containment Aerosol Removal Rate The NuScale accident source term methodology topical report describes a number of removal mechanisms and uses the STARNAUA code (with thermal hydraulic input from the applicants MELCOR simulations) to calculate the gravitational settling removal rate of airborne aerosols in the containment.

Figure 5.3 shows the airborne aerosol concentration and removal rate that is provided in the applicants Accident Source Term Topical Report for an example source term calculation (Figure 5-13 of the topical report). Staff identified two issues with this figure. First, the airborne concentration is reported in g/cm3 which is unrealistically high. Second, the applicants aerosol removal rate increases linearly as the aerosol concentration decreases. As the airborne concentration decreases, the likelihood of agglomeration should also decreases and the deposition rate should slow.

Below, the staff used the results of its MELCOR simulation for LEC-06T to assess the aerosol removal rates that the applicant calculated with STARNAUA. MELCOR models removal of airborne fission products through Brownian diffusion, gravitational settling, thermophoresis, and diffusiophoresis. While the applicant only modeled gravitational settling/agglomeration, the staff included all available aerosol removal mechanisms in MELCOR. Below, the staffs MELCOR prediction of the concentration of airborne versus deposited aerosols is used to estimate a removal rate for containment to be compared to the applicants.

The concentration of airborne fission products in containment is given by the equation 53

FSCB-19-01

= () () +

where,

() is the concentration of airborne aerosols in the containment at time (g/m3)

() is the removal rate of aerosols from the air (1/hr)

() is the source of aerosols into the containment as a function of time (g/m3).

A temporal removal rate can be estimated using the MELCOR results at discrete times (, +

1).

1 +1 +1

+1 +1 All values on the right hand side can be inferred from MELCOR output for a given scenario for all aerosol and vapor mass. The source mass, , is inferred by performing a mass balance on the total aerosol/vapor mass present in containment (on heat structures, in pool, and airborne for all CNV control volumes).

Figure 5.4 and Figure 5.5 show the results of the staffs MELCOR calculations and the derived removal rate. The temporal removal rate in Figure 5.4 is averaged over 30 minute intervals in Figure 5.5 to smooth the oscillations due to sensitivity to the discrete output time intervals and local containment thermal-hydraulic conditions.

In the above equations, the release to the environment from design-basis leakage was neglected. The design-basis leak rate (0.2 percent of containment volume per day or 1E-4/hr) is sufficiently small in comparison with the deposition rate (2/hr) that it can be neglected with no loss of accuracy in estimating the deposition rate.

One difference between the staffs MELCOR model and the applicants STARNAUA model is that the staffs calculations do not include structural aerosol whereas the applicants does. As a result, RAI 9224 requested the applicant provide the basis for inclusion of structural aerosol.

54

FSCB-19-01 Figure 5.3 Containment removal rate and airborne aerosol concentration as reported in the Source Term Topical Report for NuScale.

Figure 5.4 LEC-06T: Containment removal rate and airborne aerosol concentration.

55

FSCB-19-01 Figure 5.5 LEC-06T: Containment removal rate (averaged over 30 minute intervals) and airborne aerosol concentration.

5.4. Containment Leak Rate The NuScale Accident Source Term Methodology topical report assumes that the containment leak rate is the design-basis leak rate of 0.2%/day and half that after 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months . In the staffs MELCOR model, the leak rate is directly modeled by a flow path from the upper containment atmosphere to the environment. The flow area of this path is chosen so that 0.2%/day leakage occurs when the containment is pressurized with air to the design pressure (1000 psia) and at a temperature of 70°F. These are the test conditions for the containment leak rate testing per Chapter 6 of the NuScale FSAR.

A comparison of the design basis leak rate and the staffs prediction for LEC-06T is given in Figure 5.6. The containment pressure and temperature for the staffs calculation is given in Figure 5.7. The leak rate calculated by the staff starts higher than the design basis leakage and then more than doubles in rate following core damage. The reason for this is multifaceted. The flow rate through the leakage path is dependent upon containment pressure, containment temperature, and material density which are all changing as the accident progresses. The elevated flow rate of 0.3%/day in the first few hours of the simulation is due to the high temperature of the containment atmosphere as compared to the test conditions. Figure 5.7 shows that the temperature is much greater than the test temperature of 70°F leading to a greater flow rate. As steam condenses, the CNV pressure falls and between 7.9 and 9.6 hours6.944444e-5 days 0.00167 hours 9.920635e-6 weeks 2.283e-6 months , the pressure in the CNV drops below the atmospheric pressure (Figure 5.7) and reverse flow through the leak path occurs (Figure 5.6). When core damage occurs and hydrogen is 56

FSCB-19-01 generated through the zircaloy oxidation, the containment pressure and temperature increase.

The containment leak rate increases rapidly as well. Even though the pressure and temperature after 10 hours1.157407e-4 days 0.00278 hours 1.653439e-5 weeks 3.805e-6 months is less than the initial value, the leak rate is more than twice as much. This is because of the much smaller density of the containment atmosphere at this time (see Figure 5.8) since it is composed mostly of hydrogen (molar concentration >0.8). See Appendix A for a demonstration calculation on the impact of vapor density on leak rate.

Figure 5.6 Containment leak rate calculated from staffs MELCOR calculation LEC-06T versus the applicants assumed design leak rate.

57

FSCB-19-01 Official Use Only - Proprietary Information Figure 5.7 LEC-06T: Containment temperature and pressure.

Figure 5.8 LEC-06T: Net mass density of materials in upper containment.

58

FSCB-19-01 5.5. Release from the containment to the environment Figure 5.9 presents the staffs results for the integrated fission product release from the containment into the reactor building for its MELCOR calculation for LEC-06T. Because the staffs MELCOR model assumes the containment leakage occurs at the top of the containment vessel, there is no scrubbing in the reactor pool. The results in Figure 5.9 can be used to confirm the applicants analysis in its source term topical report (Reference 5) which is currently being updated and the applicants analysis in its Environmental Report (Reference 4).

Figure 5.9 LEC-06T: Release fraction to the environment.

59

FSCB-19-01

6. References

1. NuScale Final Safety Analysis Report, Revision 1, March 2018

2. Calvert Cliffs, Units 1 & 2, Updated Final Safety Analysis Report, Revision 48, September 14, 2015 (ADAMS No. ML15279A455)

3. Updated MELCOR Calculation Notebook: NuScale Integral Pressurized Water Reactor, ERI/NRC 13-205, Revision 7, December 2017.

4. Applicant's Environmental Report - Standard Design Certification, Rev 0, December 2016

5. Topical Report, TR-0915-17565, Accident Source Term Methodology, Revision 1, Non-Proprietary Version, April 8, 2016 (ADAMS Accession No. ML16099A394)

6. Severe Accident Selection Methodology, Rev 0, NuScale report ER-P020-4896-R0

7. Dropped Module Consequence Analysis, Rev 1, NuScale report ER-P060-7085-R1

8. Probabilistic Risk Assessment Large Release Frequency Definition, Rev 0, NuScale report ER-P000-7004-R0

9. NuScale Power, LLC, Supplemental Information in Support of NRC Audit of Final Safety Analysis Report, Tier 2, Chapter 19 Probabilistic Risk Assessment and Severe Accident Evaluation, letter from NuScale to NRC, August 3, 2017

10. NuScale Power, LLC, Supplemental Information in Support of the NRC Audit of Final Safety Analysis Report, Section 19.2 Severe Accident Evaluation, letter from NuScale to NRC, April 4, 2018

11. U.S. Nuclear Regulatory Commission Audit of the Probabilistic Risk Assessment and Severe Accident Evaluation of the NuScale Power, LLC Design, September 13, 2018 (ADAMS No. ML18254A340)

12. Nuclear Regulatory Commission Audit of the Probabilistic Risk Assessment and Severe Accident Evaluation for the NuScale Design, November 3, 2017, ADAMS No.

ML17305A024

13. Accident Radiological Analyses Sensitivity to Severe Accident Analysis, Rev 0, NuScale report ER-0000-5278-R0 60

FSCB-19-01 A. Example Problem of Containment Leak Rate In order to demonstrate how the leak rate is affected by the conditions inside the containment, a simplied MELCOR model was constructed with two control volumes connecting to the environment. The flow paths have the same characteristics as the leak flow path in the NuScale input model. The control volumes are time independent with initial conditions shown in the figure below. CV400 represents the leak test conditions, while CV500 tries to mimic the conditions inside the NuScale containment following a severe accident. For simplicity, the material inside CV500 is assumed to be filled with hydrogen.

The code is run for 100 seconds and the results are show in the following figures. The resulting leakge rates are mostly consistent with the behavior in the NuScale analysis with FL499 calibrated to produce 0.2% leak rate and how the leak rate changes in response to changes inside the containment.

61

FSCB-19-01 Figure A.1 Pressure history.

Figure A.2 Temperature history.

62

FSCB-19-01 Figure A.3 Vapor density history.

Figure A.4 Leakage flow rate history.

63

ML19196A306

Text

Navigation menu

Search