ML20113C423

AP600 Large-Break Loss-Of-Coolant Accident Phenomena Identificaton & Ranking Tabulation

ML20113C423
Person / Time
Site:	05200003
Issue date:	08/31/1995
From:	Boyack B LOS ALAMOS NATIONAL LABORATORY
To:
Shared Package
ML20113C412	List: ML20113C423 ML20113C438
References
LA-UR-95-2718, NUDOCS 9607010228
Download: ML20113C423 (119)

Text

{{#Wiki_filter:;._ _ _ _ _ _ _ _ _ l l 1 l August 1995 LA UR 95 2718 l l

Title:

AP600 LARGE-BREAK LOSS-OF-COOLANT ACCIDENT j PilENOMENA IDENTIFICATION AND RANKING l TABULATION i l Author: Brent Boyack Submitted to: Frank Odar US NRC I i I .,4,' } ~- 4

• =

n I l E: $z h } .I. I J l ~ 9607010229 960626 PDR ADOCK 052000 3 i

J i i i l i August 1995 L A-UR-95-2718 i J J

Title:

AP600 LARGE-BREAK LOSS-OF-COOLANT ACCIDENT j PIIENOMENA IDENTIFICATION AND RANKING l TABULATION j Author: Brent Boyack Submitted to: Frank Odar l US NRC i l l 2 1 1 1 j .. c w.- j 4 s l l 1 e i i i i i 9607010228 960626 PDR ADOCK 05200003 1 A PDR 4

i LA UR. 95o2718

Title:

AP600 LARGE BREAK LOSS OF-COOLANT ACCIDENT PHENOMENA IDENTIFICATION AND RANKING TABULATION i l Author (s): Brent Boyack i i Submitted to: Frank Odar US NRC m gj;=ss z#mEM_TQ2i' ' 5,;; _; g._ w+' Q[- =[ _ _. y# i

=

=

y-W si L== shal.1 ... = M MT/ M5.'T ~J $) %~ IMP 3 Los Alamos ~ % N alton AL L ABOR ATORY l Los Alamos Nationas Laboratory. en affirmative actiorvoqual oppoqunity employer. is operated by the University of California for the U S Department of Energy unoer contract W 7405 ENG 36 By acceptance of this airce the putmsher recognGes that the U S Govemmern retains a nonesclusive, royalty-free bcense to put*sh or reproduce the published form of this contnDution, or to allow others to ao so. for U S Govemment purposes The Los Alamos National LaDoratory requests fSat the pubbsner ceritify this adice as work performed unoer the ausaces of the U S Depaiment of Energy Form No 836 RS ET 26291041 e

I i CONTENTS ACRONYMS...................................................................................iv A CKNOWLEDG E ME NTS................................................................ v ABSTRACT......................................................................................1 EXECUTIVE S UMMAR Y..................................................................... 2 f 1.0 INTR ODUCTI ON..................................................................... 5 2.0 METHODOLOG Y.................................................................... 6 2.1 CSAU Methodology.......................................................6 2.2 PI RT Met hodology....................................................... 7 1 3.0 PLANT D E S C R I PT I O N............................................................. 10 i 3.1 AP600 D e s c ri p t 10 n............................................................ 10 3.2 Key AP600 Features - LB LOCA Respon se................................ I 2 3.3 Comparison to Current Generation Westinghouse PWRs...............12 4.0 LB LOCA SCENARIO DESCRIPTION...................................... 16 4 4.1 B l owd o wn Period........................................................... 34 4.2 R efi ll Peri od......................................................... 4 1 4.3 R e fl ood Period........................................................ 44 l 4.4 Long-Term Cooling Period................................................. 4 8 i 5.0 PIRT RE S ULTS....................................................................... 5 1 i 5.1 AP600 LBLOCA PIRT...................................................... 52 4 5.1.1 Blowdown Period.................................................. 5 2 1 5.1. 2 R e fill Period.......................................................... 54 5.1. 3 Reflood Period....................................................... 63 4 5.1.4 long. Term Coolin g Period......................................... 65 5.2 Comparison to PIRT for Westinghouse Four-Loop Plant............... 67 6.0

SUMMARY

............................................................................72 RE FE REN CE S................................................................................ 7 4 APPENDIX A - SYSTEMS, COMPONENTS, PROCESSES, 1 AND P H E N O M E N A............................................................... A-1 APPENDIX B - BRIEF BIOGRAPHIES OF PIRT PARTICIPANTS................B-1 4 ) ..e ee. m

FIGURES i 9 Code scaling, applicability and uncertainty (CSAU) evaluation methodology -13 1. AP600 passive safety systems 2_. 2a. 2b. AP600 passive containment cooling system 2................................. ..14 2c. AP600 passive safety injection 2-...._. _ _.._._.____.._.._......15 3. Vessel liquid volume fractions....................................................... 17 4. S y stem pre s sure s...................................................................... 17 5. Fuel maxim um claddmg temperatures............................................... 18 6. A ceumulator mass flows............................................................... 18 7. Core makeup tank mass flows........................................................ 19 8. Core make u p tan k liquid le veI s....................................................... 19 9. B reak mas s flows...................................................................... 20 10. B reak e xit voidin g..................................................................... 20 11. Cold-1eg m a s s fl o w s................................................................ 2 1 12. Cold -le g voi di n g..................................................................... 21 13. Hot-1eg 1 mass fl o w s........... ..............................................22 14. Hot-1eg 2 m a s s fl o w s.............................................................. 22 15. l_ cop 1 hot-le g voidin g................................................................ 2 3 i 16. Loop 2 hot-leg voidin g........................................................... 23 17. Pres suri zer mass flow.................................................................. 24 18. Heated core average vapor fraction................................................... 24 19. Core in le t mas s flows.................................................................. 25 20. Core o u tle t ma s s flo ws............................................................... 2 5 21. Fu el rod re acti vities.................................................................... 2 6 22. Stored energy distribution in fuel rod with time................................ 26 23. Average. power rod temperatures 0- 10 s........................................... 27 24. Average-power rod temperatures 20- 120 s....................................... 27 25. Maximum-power rod temperature s 0- 10 s........................................ 2 8 26. Maximum-power rod temperature s 20- 120 s........................................ 2 8 27. Downcomer ayera ge vapor frac tion................................................... 29 28. Total gu ide tu be flow................................................................... 2 9 29. Lower plen um average vapor fraction............................................... 30 30. Iower plen um temperature............................................................ 30 31. U pper head average vapor fraction................................................... 31 32. Upper plen um average vapor fraction................................................ 31 33. S team and feed water flows............................................................ 3 2 34. Total upper support plate drain hole flow............................................ 32 35. Downcomer to u pper h ead flow.................................................... 3 3 1 Source: B. E. Boyack. I. Catton. R. B. Duffey. P. Griffith. K. R. Katsma, G. S. Lellouche, S. Levy U. S. Rohatgi, G. W. Wilson. W. Wulff and N. Zuber. " Quantifying Reactor Safety Margins. Part 1: An Overview of the Code scaling. Applicability, and Uncertainty Evaluation Methodology, with Nuclear Engineering and Design. Volume 119,1 15 (1990). 2 Source: L. E. Hochreiter. L. E. Conway. S. V. Fanto, and L. K. Lau, " Integral Testing of the AP600 Passive Emergency Core Cooling Systems," Proceedings of the ARS 1994 International Topical Meeting on Advanced Reactors Safety held April 17 21,1994, Pittsburgh, Pennsylvania. Volume 2, p. 991 10N. ..u .a

f TABLES 1. S ummary of AP600 LB LOCA PIRT Re su lts..................................... 55 2. Comparative Tabulation of AP600 and Westinghouse Four-Imp PIRT R e s u l t s...................................................... 69 w f 54B

i ACRONYMS ADS Automatic Depressurization System AHP Analytical Hierarchical Process CL Cold leg CMT Core makeup tank CSAU Code scaling, applicability, and uncertainty DEGB Double-ended guillotine break DVI Direct vesselinjection ECC Emergency core coolant ECCS Emergency core cooling system HL Hot leg INEL Idaho National Engineering Laboratory i IRWST In-containment refueling water storage tank LBlDCA Large break loss-of-coolant accident LWR Light water reactor NPP Nuclear power plant NRC U. S. Nuclear Regulatory Commission PBL Pressure balance line PCCS Passive containment cooling system PCT Peak cladding temperature PIRT Phenomena identification and ranking tabulation PRHRS Passive residual heat removal system PRHRSHX Passive residual heat removal system heat exchanger PSF Passive safety feature PSIS Passive safety injection system PWR Pressurized water reactor RCP Reactor coolant pump S SIGNAL Safeguards signal SAR Safety analysis report SG Steam generator TPG Technical Program Group i 5V

ACKNOWLEDGMENTS The contributions of several different individuals and organizations are gratefully acknowledged. The NRC project manager for this activity was Frank Odar. His suppon of the Los Alamos AP600 confirmatory safety activities and his review of the draft document are appreciated. The PIRT activities of the Idaho National Engineer Laboratory (INEL) staff were monitored, the INEL PIRT documents were reviewed, and the information therein were used to check the completeness of the AP600 LBLOCA PIRT presented in this document. Jim Lime prepared the TRAC model for AP600 and performed the calculation and much of the analysts that is presented in Section 4 of this report. Finally, the effort of the AP600 PIRT team at I_os Alamos is appreciated. In addition to the author, the members of the PIRT team were: Steve Harmony, Thad Knight, Jim Lime, Frank Motley, Ralph Nelson, and Don Siebe. Finally, this report was edited by Gloria Mirabal. Her assistance is appreciated. Se ~ v

AP600 LARGE BREAK LOSS OF-COOLANT ACCIDENT PHENOMENA IDENTIFICATION AND RANKING TABULATION by Brent E. Boyack l ABSTRACT The Westinghouse-designed AP600 is a new-generation nuclear power plant that was designed to fulfill enhanced requirements for safety. Specifically, the reactor was designed to incorporate passive safety features that respond to an appropriately broad spectrum of operanonal and accident conditions. A formal applicanon for design certification of the AP600 has been submitted to the United States Nuclear Regulatory Commission (NRC). Historically, the large-break loss-of-coolant accident (LBLOCA) has been the focus of extensive safety assessment effons by the NRC. The primary safety criteria for evaluating safety system performance is the peak cladding temperature (PCT) limit. The NRC is using TRAC-PF1/ MOD 2 for AP600 LBLOCA confirmatory analyses and has undenaken to demonstrate the adequacy of TRAC-PFI/ MOD 2 for this application. The uncenainty in the calculated PCT is to be determined. An important initial effort is the preparation of a Phenomena Identification and Ranking Tabulation (PIRT) for the AP600 LBLOCA. Phenomena are first identified and then importance ranked relative to their influence on the PCT. This report documents the results of a PIRT activity for the AP600. The selected scenario is an 80% double-ended guillotine break LBLOCA. The AP600 designer, Westinghouse Electric Corporation, has identified this accident as the worst-case LBLOCA in the original AP600 Safety Analysis Report submitted to the NRC. The most important AP600 systems, components, and processes / phenomena occurring during each penod of a LBLOCA have been identified and tabulated. A detailed description of the AP600 LBLOCA PIRT results is presented in this repon. Results of an earlier PIRT for a Westinghouse four-loop plant and the AP600 PIRT are compared to provide insights about the adequacy of the AP600 PIRT. The most important outcome of the present assessment is that no new phenomena have been identified for the AP600 that were not present in the earlier Westinghouse four-loop plant PIRT. Thus, for the AP600 LBLOCA, there are no significant differences in the component modeling requirements that would be imposed on the accident analysts code compared with those identified in the earlier PIRT. H mee 1

EXECUTIVE

SUMMARY

ne Westinghouse-designed AP600 is a new-generation nuclear power plant (NPP) that has been designed to fulfill enhanced requirements for safety. Specifically, the reactor has been designed to incorporate passive safety features that respond to an appropriately broad spectrum of operational and accident conditions. A formal application for design certification of the AP600 has been submitted to the United f,tates Nuclear Regulatory Commission (NRC). Historically, the large-break loss-of-coolant-accident (LBLOCA) has been the focus 'I of extensive safety assessment effons by the NRC. Each reactor design must have the capability to withstand successfully the worst case LBLOCA while satisfying the acceptance criteria for emergency core cooling system (ECCS) performance for light-water-i cooled nuclear power plants found in the Code of Federal Regulations, Title 10, Section 50.46. One of tie basic safety criteria for evaluating ECCS performance focuses on a peak cladding temperature (PCT) limit of 1477 K (2200*F). For the AP600 NPP, the NRC is seeking to confirm the applicant's safety analyses, including the response of the AP600 to the worst-case LBLOCA. The NRC is using TRAC-PFIMOD2 for confirmatory analyses of AP600 LBLOCA. TRAC is a best-estimate code. In an earlier NRC effon, the adequacy of i TRAC-PFIMODI to light-water-reactor LBLOCA analyses was confirmed as pan of the pioneering effon to develop the Code Scaling, Applicability, and Uncenainty (CSAU) methodology. The NRC will demonstrate the adequacy of TRAC-PFIMOD2 for AP600 LBLOCA confirmatory analyses. An imponant effon in the code adequacy demonstration is the preparation of a Phenomena Identification and Ranking Tabulation (PIRT). De PIRT is NPP and scenario specific. Phenomena are first identified and then imponance-ranked relative to their influence on the primary safety criteria. For a LBLOCA, the primary safety criteria is typically the PCT. Once the PIRT is complete, the code to be used to analyze the transient is reviewed to ensure that its component models are adequate. A high degree of adequacy is required for those systems and components that most strongly impact the course of the transient as identified in the PIRT. This repon documents the results of a PIRT activity for the AP600 NPP. The selected scenario is an 80% double-ended guillotine break LBLOCA. Westinghouse has analyzed a spectrum of LBLOCAs and identified this accident as the worst-case LBLOCA in the original AP600 Safety Analysis Repon submitted to the NRC. Several resources were used in developing the PIRT. First, several LBLOCA calculations were carefully reviewt d. These included the TRAC-PFIMOD2 calculation summarized in Section 3 of this repon and a Westinghouse COBRA / TRAC calculation submitted with the A600 Safety Analysis Report. Second, the AP600 and earlier Westinghouse NPP designs were examined for similarities and differences. Insights from the considerable data base developed for the current commercial plants were then considered. Finally, a group of experienced staff at Los Alamos (See Appendix B) were gathered to express theirjudgment as to the relative imponance of AP600 systems, components, and processes / phenomena. The most important AP600 systems, components, and processes / phenomena occurring during each period of a LBLOCA are summarized below. A detailed presentation of the AP600 LBLOCA PIRT results is provided in Section 5. 2

Blowdown Period System l Component Process / Phenomena I Primary Coolant Break, RCPs Critical flow, degraded pump performance Reactor Core flow channels, fuel Flashing, stored energy release, gap rods conductance, post CHF heat transfer, reactivity insertion (void) Refill Period System Component Process / Phenomena PSIS Accumulators, valves Discharge Primary Coolant Break Critical flow Reactor Core flow channels, down-Flashing (l), bypass (2), multidimen-comer, fuel rods sional flow, post CHF heat transfed3), decay heat Reflood Period System Component Process / Phenomena PSIS Accumulators, Discharge Reactor Core flow channels, Entrainment(2), multidimensional downcomer, fuel rods flow, oscillations, level, post CHF heat transfed3), decay heat Long. Term Cooling Period System Component Process / Phenomena PCCS Concrete shield building, Buoyancy driven flow, mass ~ steelreactorcontainment transfer wall condensation and structure, spray tank valves, evaporation water tank PSIS CMT, IRWST, valves, Draining, gravity driven flow, sump resupply from containment Reactor Core flow channels, fuel Boiling, boiling, decay heat rods (1) Listed as interfactal heat and mass transfer in Table 1. (2) Listed as interfacial drag in Table 1. (3) Listed as transiuon and filtn boiling in Table 1. 3

Results of the earlier PIRT for a, Westinghouse four-loop plant and the AP600 PIRT were compared to provide insights about the adequacy of the AP600 PIRT. In addition, the comparison provided insights about the applicability of the original CSAU effort to the AP600. Although.there are differences between the designs, nocesses/ phenomena occurring in the two reactors are qualitatively similar during the slowdown, refill, and reflood periods of the LBLOCA. A comparative evaluation of the two PIRTs reveals many similarities. However, there are some differences in the imponance ranking of phenomena. Most of the differences between the AP600 and earlier Westinghouse four-loop plant PIRTs can be attributed to the design differences. He most significant AP600 design differences relate to the emergency core coolant delivery system, i elimination of the loop seal, provision for vonex suppression in the lower plenum, and , 1 modification of the upper plenum / upper plenum design (guide tubes and drain holes in the upper support plate). He impact of these design features on the existence and importance of processes / phenomena in the AP600 are detailed in Section 5 of this report. There are several additional design differences for which it has not been possible to assess the differential impact on processes and phenomena. These are core power density, core loading, and pump design. The core power density is lower in the AP600 than in the 'I Westinghouse four-loop design, and the core power distributions also differ. The reactor coolant pumps in the two designs are different. The impact of these differences has not been assessed. Effort will continue to asacss the impact of these differences. i The most important outcome of the present assessment is that no new phenomena have been identified for the AP600 during the blowdown, refill, and reflood phases that were not present in the PIRT for the earlier Westinghouse four-loop plant. Thus, for the AP600 LBLOCA, there are no significant differences in the component modeling i requirements that would be imposed on the accident analysis code compared with those identified in the earlier PIRT. ) 4

1 i l

1.0 INTRODUCTION

The Westinghouse AP600 is a new-generation nuclear power plant (NPP) that has been designed to fulfill enhanced requirements for safety.1 Specifically, the reactor has j been designed to incorporate passive safety features (PSFs) that respond to an appropriately broad spectrum of operational and accident conditions, thereby ensuring the level of safety needed to meet the requirements for advanced reactors specified in the Code of Federal Regulations.2 A formal application for design certification of the AP600 has 3 been submitted, and the United States Nuclear Regulatory Commission (NRC) is currently j reviewing the design to assess its safety. i Historically, the large break loss-of coolant accident (LBLOCA) has been the focus of extensive safety assessment efforts by the NRC. Each reactor design must have the i capability to successfully withs:and the worst case LBLOCA while satisfying the l acceptance criteria for emergency core cooling system (ECCS) performance for light-water-cooled nuclear power plants found in the Code of Federal Regulations, Title 10, Section i 50.46. One of the primary safety criteria for evaluating ECCS performance focuses on a l peak cladding temperature (PCT) limit of 1477 K (22000F). For the AP600 NPP, the NRC is seeking to confirm the applicant's safety analyses, including the response of the AP600 to the worst-case LBLOCA. Until 1989, only highly conservative methods were approved for evaluating the 1 i i response of a nuclear power plant to a LBLOCA. The conservative evaluation models did j not accurately simulate the actual :rrocesses and phenomena occurring during a postulated l LBLOCA, and this was perceivec as a serious shortcoming. To remedy this deficiency, the NRC developed and demonstrated the Code Scaling, Applicability, and Uncertainty l (CSAU) methodology.3 by quantifying the uncertainty associated with the best estimate calculation of key safety-related parameters for an accident in a NPP. The demonstration plant, scenario, and best-estimate code were a Westinghouse four-loop pressurized water reactor (PWR), a 100% double-ended guillotine break (DEGB) of a cold leg, and TRAC- ) PFl/ MODI,4 respectively. Subsequently, the Code of Federal Regulations was revised ) and the NRC issued guidance that applicants could license the ECCS of future plants using either the 10 CFR 50 Appendix K methodology, or they could use a best-estimate methodolo { quantified.gy, provided the uncertainty in key licensing parameters, e.g., the PCT l The NRC is currently using TRAC-PFl/ MOD 26 for confirmatory analyses of }- - AP600 LBLOCA. The NRC has undertaken the task to demonstrate the adequacy of j TRAC-PFl/ MOD 2 for AP600 LBLOCA confirmatory analyses. Code accuracy will be characterized for the LBLOCA application. d. l An essential, early step in the code adequacy demonstration is the completion of a i Phenomena Identification and Ranking Tabulation (PIRT) effort. The PIRT is plant and scenario specific. The primary objective of the present effort is to prepare a PIRT for the worst-case AP600 LBLOCA. The selected scenario is an 80% double-ended guillotine ~ break LBLOCA. Westinghouse has analyzed a spectrum of LBLOCAs and identified this l accident as the worst-case LBLOCA.7 A second objective is to compare the results of the i-Westinghouse four-loop LBLOCA PIRT with the AP600 LBLOCA PIRT to derive insigh:s about the adequacy of the AP600 PIRT. This report documents the results of the NRC-sponsored AP600 LBLOCA PIRT activity. [ i i

~

5 h

i -4 2.0 METHODOLOGY A brief explanation of the CSAU methodology and the role of the PIRT in that methodology follows. This discaision willlay a foundation for subsequerit discussions in this repon of similarities and differences between the LBLOCA PIRT prepared for a four-loop Westinghouse PWR and the AP600 LBLOCA PIRT documented in tlGs report. 2.1 CSAU Methodology The CSAU method is one approach by which the uncertainty in code-calculated results of key parameters may be evaluated. The CSAU method contains three eleinents: (1) Requirements and Code Capability, (2) Assessment and Ranging of Parameters, and (3) Sensitivity and Uncenainty Analysis (Fig.1). A complete description of the CSAU method and a demonstration of the method is found in Ref. 3. In the first element, Requirements and Code Capabilities, scenario modeling requirements are identified for a given plant and compared against code capabilities to l determine the code's applicability to the panicular scenario. In addition, an effon is made to identify potential code limitations. The modeling requirements are established by j identifying and ranking processes and phenomana important to the panicular scenario, and the result is a plant and scenario-specific PIRT. The need for such a screening process arises from the fact that the resources reqaired to quantify the uncenainty for every process and phenomenon occurring during the senected transient are too large. Further, the original CSAU demonstrated that relatively few LBLOCA processes and phenomena strongly influence uncenainty. In the second element, Assessment and Ranging of Parameters, the capability of the i code to calculate processes imponant to the scenario by comparing calculations against experimental data to determine code accuracy; to detennine scale-up capability; and to specify ranges of parameter variations needed for sensitivity studies, is determined. In addition, bounding analyses can be performed. ~ In the third element, Sensitivity and Uncenainty Analysis, the total uncenainty in a safety analysis is determined. The uncenainties arise from code limitations, scaling maccuracies embedded in the expenmental data (and therefore the code), and uncenainties associated with the state of the reactors at the initiation of a transient. The ultimate objective of the CSAU process is to provide a simple and direct statement of the calculated uncertainty in the primary safety criteria [e.g., the peak cladding temperature (PCT) for a LBLOCA) used as the basis for assessing safety and makmg licensing decisions relative to the revised ECCS rule. The first element, Requirements and Code Capabilities, contains six steps: (1) specify the plant scenario, (2) select the NPP, (3) identify and importance rank plant systems, components, processes /phenomene and document the result in a PIRT, (4) select a frozen code for analyzing the accider.t sequence in the selected NPP, (5) provide a complete set of documentation includiny, the theory manual; user's manual; and annotated code programmer's manual for the analpis code, and (6) determine code applicability. It is the third step: identify and imponance. rank plant systems, components, processes and phenomena and document the result in a PIRT, that is the subject of the remainder of this repon. That is, a LBLOCA PIRT has been prepared for the AP600. Similarities and { dissimilarities to the LBLOCA PIRT for Westinghouse four-loop PWR3 are compared and i discussed. ) -4 6 .s

The PIRT is a particularly imponant element in the CSAU process because it identifies what systems, components, and pmcesses/ phenomena most strongly influence the path and outcome of a given accident scenario in the selected NPP. Having this information, the remaining CSAU effons concentrate en quantifying uncenainties in modeling and calculating these key elements. It should be emphasized that the CSAU methodology is one approach to quantifying uncertainty. Other approaches to charactenzing accuracy may be defined. 2.2 PIRT Methodology The PIRT developed for the Westinghouse four-loop LBLOCA CSAU demonstration benefited fmm an extensive and pre existing LBLOCA data base consis:ing of experimental data and numerous code calculations. After identifying LBLOCA phenomena, two independent groups participated in importance rankmg the phenomena. Both relied on expens to evaluate the available information and rank the phenomena for importance. One ranking effort was completed by the Technical Program Group (TPG) developing the CSAU method and conducting the demonstration effon. The importance ranking of the phenomena was the result of extended group discussions and debate. The second ranking effon was conducted by staff at the Idaho National Engineering Laboratory (INEL). This effon also involved extended group discussions and debate but the debate was conducted within a formalized structure of pairwise ranking called the Analytical Hierarchical Process (AHP). The results of the two independent rankings were compared to confinn the findings. In general, the results of the two ranking efforts compared well, although there were modest differences in some of the rankings. By focusing on differences in phenomena ranked highly by one gmup or the other, some of the differences were subsequently reduced. For the AP600 PIRT reported in this document, the ranking effon more closely followed the approach used by the TPG in preparing its original demonstration PIRT. There were several reasons for this approach. First, although the AP600 is a new reactor design and the PSFs are a significant depanure from previous Westinghouse-design plants, the new or unique features of the PSFs have little impact on phenomena occurring during the LBLOCA until after the PCT is reached. The experience base (experimental and computational) is believed to remain largely applicable. Thus, the natural focus is toward the plant and scenario differences. Second, the experience with the two methods in the CSAU demonstration showed, at least for that effort, that the PIRT results were similar. Although the AP600 LBLOCA PIRT was expected to be similar to the LBLOCA PIRT for the Westinghouse four-loop plant, several explicit effons were made to create an independent PIRT assessment environment for the los Alamos staff members panicipating in the AP600 LBLOCA PIRT activity. First, a listing of AP600 systems, components, and processes / phenomena associated with a broad spectrum of accident initiators, including but not restricted to the LBLOCA, was prepared and distributed to the PIRT team in advance of the ranking sessions. This information was used to assess the relative imponance of systems, components, and processes / phenomena during the selected scenario. This information is largely reproduced in Appendix A of this repon. Second, a detailed scenario description was prepared and distributed to the PIRT team in advance of the ranking sessions. This information was a key resource used to understand LBLOCA processes and phenomena in the AP600 and to assess the relative importance of systems, components, and processes / phenomena during the selected scenario. This LBLOCA scenario description is provided in Section 4 of this report. Third, the PIRT information from the Westinghouse four-loop PWR LBLOCA PIRT was intentionally not distributed to the PIRT team during the ranking sessions, nor were results of the earlier PIRT described or presented in the ranking sessions. Although several members of the PIRT team had 7

~ _ -. _ _ _ =_ . ~. _._ m previous casual exposures to the LBLOCA PIRT for the Westinghouse four-loop plant, the exposures took place several years ago. l The PIRT ranking process was as follows. An orientation session was conducted to familiarize members of the PIRT team with details of the current design. De response I of the AP600 to LOCA events having different break sizes was reviewed. With this background, the ranking effort was imtiated. As discussed in Section 4, the LBLOCA accident scenario is divided into four periods: (1) blowdown, (2) refill, (3) reflood, and (4) long-term cooling. The PIRT team considered one period at a time and ranked the importance of each system during that period of the transient. Component and u process / phenomena ranking followed as necessary. The importance of a system during the period was evaluated as "High" if it had a significant impact on the course of the transient, specifically ifit had a controlling impact on the ultimate PC.. Here, the emphasis on ultimate is meant to convey a focus on the PCT T for the total accident sequence rather than the PCT during each period. The imponance of the system was evaluated as " Medium" ifit had a moderate influence on the outcome. The importance of the system was evaluated as " Low" if it had a small influence on the outcome. Finally, the system was " Insignificant" if it had either no influence or an insignificant influence on the outcome. The PIRT team then considered the importance of each component. If the importance of the associated system was ranked "I.ow" or " Insignificant", no funher effon was made to rank the relative importance of either the components or associated processes / phenomena. De basis for this reduced ranking effo t comes from the original CSAU effort. In that effort it was demonstrated that only the most important processes / phenomena significantly affect code uncenainty. Ir. fact, phenomena of "high" imponance were further subdivided in the original CSAU ans only those phenomena in the highest of three subdivisions were carried forward for r.ncertainty quantification. For systems ranked as " Medium" during the period, only the most imponant components and phenomena were identified. However, the importance of components and processes / phenomena in a system of " Medium" importAnce was never assigned an unponance ranking higher than " Medium". Subsequent to the original ranking activity, additional information and insights were developed regarding both processes in the AP600 and the relative importance of various processes in commercial reactors. Proposed changes to the PIRT were prepared and the Justification for the changes documented. The PIRT team then reviewed the proposed changes, convened to discuss the proposed changes, and agreed upon the PIRT rankings appearing in this document. Brief biographies of the members of the team panicipating in the AP600 LBLOCA PIRT are presented in Appendix B. B 4 8 s-

1 i I Element 1 Specify SeleCe I freien 4 Requirements scenario and code code cacaoelities 4 se,eci NPP Prownde complete documentation } - Code manual User guise 5 Progammers Guios identify and rank Developmentas assessment phenomena (PIRT) 3 Mocol & correlations OE 4 Oeiermme cose 6 applicsoility i + Element 2 ,gj[gnt 7 j erlatett Assessment andrangin0of d parameters p,i,n, i nodaletation J for NPP Calculations i 4 Compare calculations Compare ce6culahons vs atta using NPP Set IET vs. (ETS using NPP _g ~ nodalization data base data base nodal:2ation Documeat Document j l I 4 Noding ves Chan I No Seas and NW'*'"e'm# C andesp n ent y uncertainty accuracy i O*g""'a' sies and to p, uncenaini, i \\ ein and oei. ne.tieci uncena nty of reactor anput 11 parameters and state Pefiorm NPP sensitivity 12 calculations y-Additional t g Comeane tHases 13 l margen H warranted by and ...1 limitatson in uncenamtses i este base. Element 3 i l coce. etc. j i ~~~~~~~~ Sensitmty and Total unconsenty uncenaenty analysis to calculate specific scenano 14 6n a ^ specific NPP .rig.1. Code scaling, applicability, and uncertainty (CSAU) evaluation methodology. 9 m _., _+- ____ - __ 4

3.0 PLANT DESCRIPTION The PIRT is plant specific. Herefore, a brief description of the AP600 system is pmvided as a necessary element in this documentation of the PIRT evaluation. In addition, a brief review of those systems and components having the greatest impact on the course of the LBLOCA during the blowdown, refill, and reflood periods of the transient culminating in the total core quench is provided. Finally, similarities and differences between the AP600 and the Westinghouse four-loop PWR and their impact on the LBLOCA are discussed as this information is an essential part of determimng the applicability of the earlier CSAU to AP600 3.1 AP600 Description The AP600 is an advanced passive 600 MWe reactor design being developed by Westinghouse in conjunction with the US Department of Energy Advanced Light Water Reactor Technology Program. The AP600 is a two-loop design. Each loop contains one hot leg (HL), one steam generator (SG), two reactor coolant pumps (RCP), and two cold legs (CL). A pressurizer is attached to one or the hot legs. The reactor coolant pumps are a canned-motor design and are attached directly to the steam generator. The loop seal is eliminated, an added safety feature in that core uncovery caused by the existence of water-1 filled loop seals is eliminated during a postulated small-break (SB) LOCA. The core is designed for a low-power-density and consists of 145 fuel assemblies with an active fuel length of 12 ft. The fuel assembly is a 17 x 17 array of fuel and control rods. The AP600 incorporates passive safety systems that rely only on redundant / fail-safe valving, gravity, natural circulation, and compressed gas. There are no pumps, diesels, or other active machinery in these safety systems. During plant shutdown, all the passive safety features will be tested to demonstrate system readiness, flow, and heat removal performance. These systems are shown in an isometric cutaway view of the AP600 reactor design in Fig. 2a, containment cutaway in Fig. 2b, and in a diagram in Fig. 2c. Two Passive Safety Injection System (PSIS) trains, each with an accumulator (ACC), a Core Makeup Tank (CMT), and an injection line from the In-containment Refueling Water Storage Tank (IRWST) and sump are connected directly to the reactor-vessel downcomer via a direct vessel injection (DVI) line. Depressurization of the primary system is an essential process that is required to ensure long-term cooling of the AP600. For example, the accumulators inject coolant into the reactor coolant system (RCS) only after the pnmary pressure has dropped to 700 psia. Coolant injection from large, safety-class water pools, specifically the IRWST and sump, can occur only after the reactor coolant system pressure decreases below the gravitational head of each pool. An Automatic Depressurization System (ADS) permits a controlled pressure reduction of the RCS. The ADS has four stages. Each of the first three stages consists of two trains providing redundant flow paths between the top of the pressurizer and the IRWST. The coolant discharged to the IRWST is condensed and accumulated for later injection into the RCS. He actuation signal for first stage ADS is a reduction in the coolant inventory of one CMT to 67% ofits initial value. ADS stage 2 is actuated 70 s after stage 1 actuation. ADS stage 3 is actuated 120 s after stage 2 actuation. The actuation signal for fourth-stage ADS is a reduction in the inventory of one CMT to 20% of its initial value and an interval of 120 s following third stage ADS actuation. The fourth stage ADS consists of two trains, one train connecting the top of the pressurizer hot leg (loop 1) and the containment, and the other train connecting the loop-2 hot leg and the containment. A direct discharge path to the containment is needed to ensure that the PCS pressure will equilibrate with the containment pressure so that the head-driven IRWST 10

l injection can proceed. The fractions of the total ADS discharge area for ADS stages 1-4 are 0.038,0.171,0.171, and 0.62, respectively. After the accumulators and CMTs are i depleted and the primary system has depressurized and approached the containment pressure, water injection is provided from the IRWST. This tank empties after several i days. Provisions are also made for recirculating coolant from a sump. IRWST and sump 3 recirculation may occur at the same time for some transients. The AP600 containment plays an essential role in the long-term cooling of the primary via the Passive Containment Cooling System (PCCS). Steam entering the containment, either through a break in the primary or through operation of the ADS, condenses on the inside of the steel containment shell. The condensate drains downward 1 and a large fraction is delivered via guners to either the IRWST or the sump. Heat transfer on the outside of the containment steel shell is by evaporation ofliquid, sprayed near the top of the steel reactor containment dome by the PCCS, and by convection to an air stream induced by buoyancy-driven flow (unforced). This air steam enters a high-elevation inlet, 4 flows downward to an elevation near the bottom of the cylindrical portion of the steel reactor containment structure, passes upward through the annular gap between the steel reactor containment structure and the concrete shield building, and is exhausted to the i atmosphere near the top of the concrete shield building. The PCCS spray inventory is eventually depleted. However, by the time the PCCS water supply is depleted, the decay heat has decreased sufficiently so that the buoyancy induced air flow through the air gap l between the steel containment structure and the concrete shield bsilding can remove the j core decay heat. t j Fct non-LOCA accidents, long term heat removal is provided by a Passive j Residual Heat Removal System (PRHRS) that removo core heat through natural circulation. The PRHRS receives water from the top of the. s t leg to which the pressurizer a is connected. The single PRHRS line connected to the hot leg divides into two lines, each feeding one of the two PRHRS heat exchangers residing in the IRWST. The two PRHRS j discharge lines then rejoin and connect to the outlet plenum of the steam generator outlet in the same loop. When functioning as the heat sink for the PRHRS heat exchangers, the i IRWST has sufficient water volume to remove decay heat for two hours before the inventory reaches saturation temperature. Isolation valves on the PRHRS lines open upon receipt of the safeguards (S) signal, and a buoyancy induced flow transports primary l coolant through the PRHRS. The PRHRS may also contribute to the removal of energy from the primary system during LOCA events. However, operation of the PRHRS is t interrupted when the PRHRS inlet void fraction becomes very high, degrading the energy transport to the PRHRS heat exchangers. Thus, the PRHRS is meffective for LBLOCAs, L has a limited interval of effectiveness for intermediate-break LOCAs, and has an extended l period of effectiveness for SBLOCAs. 4 Westinghouse has changed several features of the AP600 design since it was submitted for design certification. Although only the final design is of interest when assessing safety, it is important to assure that the design information used in performing i various safety assessments reflects either the current or final design. In performing the assessments presented in this document, we reflect the design features known to us as of November 15,1994. Since that time, additional design changes have been implemented. 4 The most significant for LBLOCA response has been an increase in the size of the pressurizer. For the PIRT reported in this document, the pressurizer component plays a i role of MEDIUM importance during blowdown and LOW importance otherwise. Thus, 2 the pressurizer design change may impact the PIRT ranking for the blowdown period. j J u j 11 I

l l 3.2 Key AP600 Features - LBLOCA Response The selected scenario is an 80% break in a single cold leg between the RCP and the - nozzle to which the cold-leg pressure balance line (PBL) is connected. For this PWR LBLOCA scenario, a large amount of primary coolant is discharged through the two-sided break, the core completely fills with vapor, the fuel rod cladding temperatures rapidly increase, emergency core coolant (ECC) is injected, and the core is quenched. This sequence of events occurs within the first five min following the break initiation. From break initiation through the time the entire core is quenched, the accumulators are the key and dominant components ensuring core cooling. It should be noted that in the l Westinghouse four-loop plant for which the original CSAU assessment was xrformed, the accumulators were also the key and dominant component ensuring core coo:ing during the interval from break initiation to core quench. Therefore, there are significant similarities in the responses of these two plants for the interval that culminates in total core quench. l The core is not completely filled with liquid at the time the last fuel rods are l quenched. The remainder of the scenario consists of refilling the core and providing long-term cooling while relying only on the passive safety features. The original CSAU assumed that core reheat did not occur once the core had been quenched. A similar assumption is made for the present AP600 PIRT effort. l 3.3 Comparison with Current Generation Westinghouse PWRs An important factor in assessing whether the earlier CSAU demonstration for a l I four-loop PWR is applicable to AP600 is the identification and evaluation of differences in key design features of the two reactors. We have already established that the prime focus of the review will consider those design features active during approximately the first five ndnutes following break initiation, the interval during which the ennre core is quenched. Several design differences do enter into the early phases of the LBLOCA scenarios for the two plants. For example, in the AP600, emergency cose coolant (ECC)is injected directly into the downcomer through the two DVI lines of the PSIS. Thus a cold-leg break does not directly result in the loss of ECC fmm either accumulator. In the earlier Westinghouse designs, each of the four accumulator injection lines are connected to a different cold leg. l Thus, the entire inventory of one accumulator is lost through the break and is unavailable I for core cooling. In the AP600, a turning vane directs the flow from each DVI line downward into the downcomer, thereby affecting the amount of ECC bypass. The average linear power in the AP600 core is approximately 70% of that in the four-loop Westinghouse PWR. One manifestation of the lower power density core is that the internal diameter of the AP600 is fully 90% of that in the four-loop Westinghouse PWR and the intemal vessel cross sectional area is about 82%. The two accumulators contain about the same volume of borated water as three accumulators in the Westinghouse four-loop plant. I 12 m

l __ 1

l ADS Velves (1/2 ADS Trains Shown)

Steam Generator [ PRHR HX IRWST O Pressurizer ( 45 b O[ Of sp.ra.r m v (O b8 i I v Wl suma s n i v / y can Screen V Reactor

• E' N Vessel Line A8A Fig. 2a.

AP600 passive safety systems. I e 6 13

- ~ - - - - ~' A Internal cordensation ard natural circulation transfers heat from the core to the steel containment. B The mntainment is cortinuously cooled by naturaldrculation of air between the containment vessel and surroundin0 shield building. C Initially, containment cooling is enhanced by gravity-fed water from tanks above the containment. e Q L / L L m r l ' i A v 1 l B i m 1 b 3 1 [i t 1 6 - 3

== EEE =

== == A ~ rg A ~ Q g [] 1 m a v A Fig.2b. AP600 passive containment cooling systems. 14

Break location for ACC B cut B pere:de CL DEBG ADS B DVI Line 8 y Pr.....r n CL1A Cold Leg Pressure Balance Lane B CL28 ~ a RCP1 RCP2B N Reactor SG1 l HL1 ye ggel HL 2 l SG 2 9 ^ '*8* RCP 2A RCP1B Cold Leg Presssre Balance Line A .J. .l. PRHRS DVI Line A Sos ger B T T ACC A CMTA Sparger A IRWST Sump I l lRWST Drain Lmes w Fig. 2c. AP600 passive safety injection. sep M e>w 15 i 1 I

4.0 LBLOCA SCENARIO DESCRIPTION l 'Ihree primary documents 89 have been used as source material for developing the 7 AP600-specific LBLOCA scenario description. Two sources.s contain LBLOCA analysis 7 results calculated with the.ECOBRA/ TRAC code.10 Reference 7 contains the LBLOCA analysis submitted with the AP600 Safety Analysis Report (SAR) and documents the worst case LBLOCA [80% double-ended guillotine break (DEGB) of a cold leg in the loop which does not contain the pressurizer]. Other LBLOCA cases (e.g.,100% DEGB and hot-leg LOCAs) are also presented in Ref. 7. The SAR analysis uses bounding assumptions concerning a number of initial and boundary conditions and parameters. Specifically, a bounding high power (102%), bounding p generator plugging values (10%), minimum system flow, mm, eaking factors, high stea imum accumulator delivery, minimum system pressure, maximum fuel pellet temperature, and minimum containment i i back pressure (14.7 psia) are chosen. Only safety-related systems operate during the accident. In addition, the hot assembly loca: ion is chosen such that there is no path through which fluid from the upper head can directly drain to the top of this assembly. Reference 8 contains additional plots for the 100% DEGB presented in Ref. 7. Reference 9 contains the results of an 80% cold leg DEGB LBLOCA calculation performed with the TRAC-PFl/ MOD 2 code.6 The calculation documented in Ref. 9 is based upon the AP600 design as of November 15,1994. There are no plans to conduct experiments to investigate the AP600's LBLOCA behavior. Reviewers have concludedll that the new passive safety systems that differentiate the AP600 from earlier Westinghouse commercial reactor designs have little impact on LBLOCA performance. As we have identified, compared, and ranked the systems, components, and processes / phenomena that tre active durmg a DEGB LBLOCA in the two reactors, we have reached the same conclusion. Therefore, we believe that insights from a previous study of LBLOCA phenomena in Westinghouse commercial reactors 3 are largely applicable. A related concept is that existing analysis codes can predict the thermal-hydraulic behavior of the AP600 during a LBLOCA event. The format and period definitions of the following scenario description of a LBLOCA in AP600 closely follow those of Ref. 3. The limiting LBLOCA has been identified as an 80% DEGB m a single cold-leg pipe between the primary coolant pump and the connecting point for the CMT PBL to the cold leg.7 Westinghouse has concluded that this hypothetical break will produce the maximum fuel rod cladding temperature. We have not found independent confinnation of this assertion in the literature. To facilitate analysis, the LBLOCA scenario is subdivided into four time periods that characterize events during the sequence. These time periods, termed blowdown, refill, reflood, and long term cooling are defined by the core and lower plenum liquid mass fraction behaviors; the first three periods are shown in Fig. 3. The response of the AP600 to a LBLOCA is characterized by the information presented in Figs. 3 through 35. The blowdown period is the result of a break in the coolant system through which the primary coolant is expelled. Early blowdown physical processes / phenomena include critical flow at the break, fluid flashmg and depressurizadon, redistribution of fuel rod 1 stored energy, and heating of the fuel rods due to degraded heat transfer. Later in the blowdown coolant reenters the core when the intact loop pump flows exceed the briefly exceed the break flows. Coolant also drains into the core dunng this period from the upper plenum. During blowdown, some components are affected more than others. In particular, the heat removal from the core results from the changing flow and heat transfer 16

1 I I I E 30 Too of upper penum -~.h Cold leg centertne ~jt ; j A, - - - M- ~ * - ",3 "- t - ~

• --Hot g coniertne

- DVI conene -- 7 i.:t' j [y.- a, - - l-3,, g,. -I,3f:,,,I,' g - , tj 1 - - 'fij ::::. { t-1.J"! o unoer core me. e e .e o 6--

l -- he o t-

+'d 5 i.^ - -20 ,a, f*--- Downcomer laguwllevel I."vyfel.,'f. ! ' C .. V ~ h e 5 g r i t -ees. $ $hh-.ee. N Top of lower penum w w l 0- --0 m RFJu RD1000 ,* ~ _y C 20 40 60 80 10 0 12 0 uo 16 0 TME (s) Fig. 3. Vesselliquid volume fractions. 17500000 p UPPER PLENW

15000000,

... PRE 55JR2ER 210 0 BRO 4N COLD LEG 28 - STEAM CENERATOR SECONDARY -- ACCUMULATOR 12500000_ W \\ k 67500000 ioso 7 f000000. s t N 2500000-- \\ .-sso X 0- --0 RUu IETLD00 2500000 O 20 40 60 80 10 0 12 0 wo 16 0 TNE (s) Fig. 4. System pressures. ~ 17

_ ~. >. - ~ -.,. -. l l 1300 ~ HOT ROO ~ # 1200 ......... A4 RACE ROO 8 1600 { 1% = 400 W 1000 g 1200 '1 d 800--. -. n00 300 700 - { 2 600_ _. 0 b" RC M Rgt000 I gg 0 20 40 60 80 10 0 12 0 MO 16 0 TME (s) Fig. 5. Fuel maximum cladding temperatures. 450 400. ACCUMULA'OR A 900 ......... ACCUMULA'OR B 350- - no 300 600 } J 230 4, 8 460 s g 200* i 15C 3 300 2 10C 50 50 0- ..o RE M RU1000 i I i 1 .g 0 20 40 60 80 10 0 12 0

1. 0 16 0 TME (s)

Fig. 6. Accumulator mass flows. 18 e

450 m i

400, AON@ ^

.' goo - - -- AOCUMULA'OR B 3So. . m 300 600 g. 460 300 10 0 15 0 So 0- --O e RCru NCR.000 0 20 40 60 50 10 0 12 0 uo 16 0 Tuc (s) Fig. 7. Core makeup tank mass flows. 6.32 20 3 9 CORE MAFLUP TAH A 6 30 ~-"~~~ CORE MAKEUP TAW B 20 6299 628 [

626, yg p g $

hd 624 tr e N -, 20.4200 5 622 f 620 ~ 6.18 g RErg REFLOOD I gg 202100 0 20 40 60 80 10 0 12 0 %Q 40 Tut (s) Fig. 8. Core makeup tank liquid levels. 19

.._m 22500 TOTAL BREA< ROW 20000 ......... PUW A SOE BREA< Row VESSEL SOE BREAK Flow 40000 i l 2000 ,,1 30000 1 12500 i 4 a **** d 20000 ..s V1 5 00 Q 3 .\\. 2 6000_-t. 10000 2500 "'. N A A/. V .n s g. 8 Rc 4 nort000 2500 O 20 40 60 80 10 0 12 0 too 40 TME (s) i Fig. 9. Break mass flows. 12 PuuP SCE ......... vtsstt ser ,, l' ' ~ 9 :;..

' :.., ;. -l If 6;* { ;!!':lj y T '.,:

! ":':;:p' ! j !'!l '. '- ': 3 , :1 08 l l' 2 o e

s i-

.j" Q 06 j .p .. j ! j,: !. : i l E l i 't

!?

4. 5. '

: 5 i.:

0.4 - ';

i 4

f,.

• l, - ;.

t e i: it?:! 02 ' s.

E

:

ji . ~..

• j?

r o J I _d_ ac s am000 I -02 O 20 40 60 80 to 12 0 wo 16 0 TME (s) Fig.10. Break exit voiding. 20

( WMO ~ 20000 7500 MACT COLD JG 1A \\ MACT CO2 JG 18 - MACT COLD 2G 2A 5000,-} . BR04N COLD EG 28, PUMP SDE -" e000 -- BR04N COLD MG 28. VESSQ. SOE 2500 O*-.-Steady. Stem Flow O' I 'l g -un b h } ) }T){'s f )W ~~ 't '.1 F a d = 1 h.Soco-./ - -**o{ / -f -7500 = -20000 -10000 g I -12500 REFLL RETLOCO -15000 C 20 40 60 80 10 0 120 14 0 16 0 TME (s) Fig.11. Cold leg mass flows. u MACT COLD EG 1A -- MACT COW LEG 18 0.e _ _ _. NTACT COLD LEG 2A 5 I h 06 E 04 02 0 RC M IETLOOD -02 0 N 40 60 80 00 12 0 wo 16 0 TNE (s) i Fig.12. Cold-leg voiding. 21 i

• 1 6000 12500 FLOW At VESSEL 5000 -*--- Siesdy-State Aow g

g 0000 4000 7500 3000 -M -_5000 ^ 2000 4 1000" N

• T N -

--O - - - ~ ^ 0- = -1000 .2500 -2000 -5000 -3000 d RETk RULC00 0 0 20 40 60 80 00 W W 16 0 TNE (s) Fig.13. 4 Hot-leg 1 mass flows. j e000. c500 FLDW AT VESSEL MD. FIDW AT STEAM GENERATOR 15000 6000 12500 5000 'e Steady State Row 10000 6 g WO h g m0 3000 2000 I E0 " ~ 5 4 .2. _g g. Ryu RETT 000 -e00 i C 20 40 60 80 to 12 0 16 0 TME (s) Fig.14. Hot-leg 2 mass flows. 22 -a

._. ~ l l 12 VOONG AT VESSEL 9 VOOAO AT STEAW CENERATOR f a rr i ;e ll . r-. : r, '- l 0.8 g !d y 0.6 E 1 i l l C2 0 e RE N acrL000 -02 0 20 40 60 80 10 0 12 0 uo 16 0 TME (s) Fig.15. Loop-l hot-leg voiding. i 1.2 voON AT vtSste 9 - - - VOONO A' STEAM CENERATOR l' t i O.8 2 Q -y 0.6 E f o' i 02 0 m RE N REFLOOD -02 0 20 40 60 80 10 0 12 0 40 16 0 l TME (s) Fig.16. Loop-2 hot-leg voiding. m 23 l 1 -.

't no 6000 2500 S000 2000 4000

P 300 E

.s 3000 2000 s s 500 C00 0- --0 RC M RER.000 i _i 0 20 40 60 80 00 12 0 wo 16 0 TME (s) Fig.17. Pressurizer mass flow. 1.2 ff I f l i p w c= L l 1 h 06 J 1 E 04 l 02 O d RE M RETT 000 -0J O 20 40 60 80 20 90 MO 16 0 i Tut (s) Fig.18. Heated core average vapor fraction. 24

1 0000 - -Sleedy-s:ste Fin 1 TotAt - 20000 6000 - LCI.O I VN - 16000 6000 u000

now 1

+ 20m. .. 000 { ~ j I p ] -2000- j -.-4000 4 000-i I -8000 F -6000 - f i e atru Mcramo j -exo 0 20 40 60 80 00 20 20 16 0 WE (s) Fig.19. Core inlet mass flows. 10000 l -s.or sc = rio. TOTAL ~ " ...... tou3 8000" "-18000 15000 S000 h 4000-- 9000 t2g - ex0 2000 I 3000 ( -J\\- Q __0 O_ d RETu REFLOOO -2000 0 20 40 60 80 20 12 0 20 16 0 mE (s) Fig. 20. Core outlet mass flows. o 25

0.075 REACTM*Y L RCD REAP ... ~... ~~ 'bRE REACTMTY 0050 - CCOLANT TEMMRA'URE REACTMTY -- VOO RACTON RCACTMTY p - ---- - - - o.e3 . M.. 0.000 / 5 L _.- _ ~ 'Juvvu4A/</^*W# + -00n -0m0 ~0UD ~ - 0.10 0 f RL M RETLOOS -0 23 O 20 40 60 80 10 0 00 16 0 16 0 TME (s) Fig. 21. Fuel rod reactivities. 1200 4 oT.0s U00_ 1500 1000- '*s.

• T=2s -.1350 E

E =T=3s -N'N'N U00 "s--_ .T.4 '00 '2 cc 10$0 .T=5s W 800 ~1........4-- ^ ^*' h

• iM

.T=6s - 900 h .,.,s 75o

• T=Us 600-

.r.as - 600 500 l 0.0 to 2.0 3.0 4.0 S.O 6.0 ROD RADUS (m) "I i Fig. 22. Stored energy distribution in fuel rod with time. 26

) I 850 ' 800

• T=0.0S
• 1 = 2.0 S y

f, 8 f,.. -+.....

• 7 10 S

( ,c.- .t "f=6.05 yon i

• t=805

.) w r } s' ' 7 - 10.0 5 650 ,...s...<,'s p a. 600 _ T'y sso 25 2s.b 26 2s.s 27 27.b 28 28.5 29 29.5 ELEVATION (m) Fig. 23. Average-power rod temperatures 0-10 s. 900 ,"L... . a 300 J* g

• f ' * *,

o T = 20.0 5 ^ V..

• T = 40.0 S b

,a',ga ',+ 700 ,9,. . T = 60.0 S ki 5 5'. = q s e., = T = 80.0 5 600 .. a, N.'.,

• T = 100.0 S e

al y sco -l/ .9, ' t = 120 0 5 \\ l . T = wo O s m 400 ...e ~ 300 25 2$3 26 26 S 27 27S 28 28.5 29 29 5 ELEVA90N (m) Fig. 24. Average-power rod temperatures 20-120 s. -er 27

i Soso 1000 / gbo oT=0.05 aT=2.05 goo g3o /jl..... - *.. 8

• T = 3.0 5 goo "T=6.0S j

h 73o

• T=8.05

' T = 10.0 S ,co

• ...........m., ~ s...,

abo ,,.a-j ~ ' 600 550 3on 26 2s.s 26 26 b 27 27.5 28 28.5 29 29 s ELEVAtON (m) Fig. 25. Maximum-power rod temperatures 0-10 s. 1300 ' 1200 o T = 20.0 5 + 110 0 ...., n.,y/,h 1'- e..... \\

• T = 40.0 S 1000

./

• T = 60 0 5

/ ', /,/ 900 } 'a,,. e ", - N = T = 80.0 S EW m*. W 830 j

s..

• T = 100.0 S w

\\ 700 ' T = 120.0 S 600 lI l .t., a T = 140.0 5 wo - ;)j l i l / l 00 300 tb 2b s 26 26.s 21 21h 28 28.5 29 29.5 ELEVA3ON (m) Fig. 26. Maximum-power rod temperatures 20-120 s. 28

09 O.s 0.7 0.6 0.5 l !d f f f1 04 0.3 f ( 1 )l1 l i t 0.i j 0 RCfra RCTLC00 0, 0 20 40 60 80 10 0 20 wo 16 0 TME (s) Fig. 27. Downcomer average vapor fraction. 200 350 0-J^ IM --O -_- 350 t 700 ^ 6 s 4% -1050 ~ 3 _ woo 2 -800- -_ - r750 ~ 2100 m RC M RETLOOD .nx 0 20 40 60 80 10 0 20 WO 16 0 TME (s) Fig. 28. Total guide tube flow. 29

'4 1 I-f 08 06 W E .3 .i 0.4 02 O ~ 2 uru urt000 1-o _o, 9 20 40 60 80 10 0 12 0 14 0 16 0 TwE (s) Fig. 29. Lower plenum average vapor fraction. W. 7og LCUO TEMPERATURE - - SATURATON TEMPERATURE 600 600 l th0 5 . m h ~ l< l-400 = 450 300 ,,.%~s~.....s.N=^-~~~.- _( uru ftf7L000 I I i g 0 20 40 60 80 10 0 12 0 ko 16 0 TME (s) Fig. 30. lower plenum temperature. 30 ..e !

I I f 0.8 hy 0e E i 02 5 1 0 t l M _, uru uncoo -02 O 20 40 60 80 10 0 12 0 20 16 0 TME (s) f i-Fig. 31. Upper head average vapor fraction. u j WP 4 0.8 It 5 Q 06 E oa 1 3 I 02 ~ 0 8 aru uno00 i 02 0 20 40 60 80 10 0 12 0 ko 16 0 TuE (s) Fig. 32. Upper plenum average vapor fraction. 31

e 800_ n50 EDwATR Flow 1 wwana rww 2

700, u gr g,

i A a 1 -I 1250 s00 2 I l 1000 400 -{,

• 3 g

- no 0 ~ 2 . wo 2 200 g. 250 , l 0-- l --O b RC L RfnD00 .me 0 20 40 60 80 to 12 0 wo 40 TME (s) Fig. 33. Steam and feedwater flows.

S00, gog 230 0-N

--O ~2b0 ~ 900 - 750 -2000 y - 9 00 -oso - ) -3000 -e00 -g -n30 ~ Rtru RErLD00 l 1 0 20 40 60 80 00 12 0 wo 40 TWE (s) Fig. 34. Total upper support plate drain hole flow. 32

I 1 - MC .~200 50 10 0 c.. 1. ..o = g 30- -mo .wo-. -. -2* 1 - -300 .w I AULL -400 200 atrLD00 O 20 40 60 80 10 0 12 0 MO 16 0 TME (s) Fig. 35. Downcomer to upper head flow. regimes in the core. The performance of the primary coolant pumps degrade as the coolant flashes. The steam generator heat transfer degrades after the steam-generator secondaries are isolated. The blowdown period ends when the intact-loop accumulator injection is initiated.3 During the refillperiod, the reactor system starts to recover as the PSIS components (CMTs and accumulators) start to inject coolant into the primary system. The important refill components and processes / phenomena concern the introduction of water into the reactor vessel downcomer and its subsequent distribution. Refill physical processes / phenomena are the operation of the PSIS, including interactions between the accumulators and CMTs, bypassing injected water through the downcomer to the broken cold leg, and penetration of safety injection water into the lower plenum. The refill period ends when the mixture level in the lower plenum appmaches the core inlet, and conditions are established for refloodmg the core with coolant. The refloodperiod begins once the lower plenum has refilled and the core liquid inventory enters a period of sustained recovery. The reflood process is highly oscillatory after the downcomer fills to the DVI line nozzle but the overall trend with increasing time is increasing core coolant inventory, i.e., a sustained recovery. Refilling of the core with coolant is well advanced by the end of the period. The reflood processes may be quite slow because much of the water is boiled and transported as steam and entrained droplets into the upper plenum and hot-leg piping. The reflood period ends when the entire core is quenched, that is, all fuel rod cladding temperatures are at or slightly above the coolant saturation temperature. we 33

.. g De long term cooling period continues after total core quench. At the time the fuel rod cladding is completely quenched, the core is only partially full. Accumulator discharge is still underway. After the accumulators empty, the CMTs resume draining their inventory into the primary. CMT draining leads to ADS actuation. IRWSTinjection is initiated when the primary pressure decreases to a level less than the static head in the IRWST. CMT and IRWST draining may occur simultaneously. Draining of the IRWST is expected to take several days, after which water from the sump is recirculated indefinitely. ADS stages 1-3 have an insignificant impact on the transient because the primary has largely depressunzed to containment conditions before they open. After the inventory in one of the CMTs drops to 20% of its initial value, fourth stage ADS opens a direct path for release of core-generated steam to the containment. For many accident scenarios, the depressurization process must be assisted by operation of the ADS. However, the LBLOCA has sufficient area to depressurize the primary, even in the absence of ADS actuation. The detailed scenario description that follows is largely based upon a TRAC-PFl/ MOD 2 calculation of an 80"; DEGB in a single cold-leg pipe between the primary coolant pump and the connecting point for the CMT PBL to the cold leg. Information 9 from other sources will be direct att ibuted to that source. For the TRAC calculation the containment back pressure was specified to be a constant one atmosphere. The scenario description is supported by figures displaying significant calculational results. In some cases, the figure legends identify components with alphabetic or numeric descriptors, for example, DVI line A or CMT B. Figure 2c, a schematic of the AP600 and its passive safety systems, shows the relationships between the plant components and the TRAC input model labels. With reference to Fig. 2c, loop 1 contains hot leg 1 (HL 1) and cold legs l A and IB (CL 1 A ar.d IB). The pressurizer is connected to HL 1. The PRHRS is connected between HL 1 and the outlet plenum of steam generator 1 (SG 1). One train of the founh i stage ADS attaches to the PRHRS inlet line which is in turn connected to HL 1. Loop 2 contains HL 2 and CL 2A and 2B. The second train of the fourth-stage ADS connects directly to HL 2. He PSIS has two trains which feed coolant directly to the downcomer ) through DVIlines A and B. Connected to DVIline A are CMT A, ACC A, and a IRWST drain line. Connected to DVIline B are CMT B, ACC B and a second IRWST drain line. 1 DVI A is connected to CL 2A by PBL A. DVI B is connected to CL 2B by PBL B. The postulated 80% DEGB occurs in CL 2B between the PBL B connection to CL 2B and RCP 2B. ) 4.1 Blowdown Period Overview The AP600 LBLOCA blowdown stans at the time the break opens, and ends when intact loop accumulator injection is initiated, a period of about 12 s. The initial barak mass flows are high, reflecting subcooled critical flows at the break planes. He mass flow from the vessel side of the break is much larger than that from the pump side because the pump and the steam generator piping produce a higher hydraulic resistance to flow. Because coolant flows from the vessel to the break through both the hot and cold legs, core flow rapidly stagnates; the flow at the core inlet reverses shortly after the break occurs. Early in the transient, flow from the intact loops is generally bypassed around the reactor vessel downcomer to the broken cold leg and out the break. As the primary coolant system rapidly depressurizes (Fig. 4), flashing occurs first in the highest temperature pans of the system, starting in the pressurizer and rapidly progressing through the upper plenum, hot legs, and then proceeding through the core and the steam generator and finally the lower plenum, downcomer, and cold legs. The d. 34 - 1

1ressurizer starts flashing earliest, but because of the rc6 stance to flow out of the surge ine, it takes almost 10 s to empty the pressurizer. De Nghest pressure in the primary is in the pressurizer until 7 s (Fig. 4). The flow from the prassurizer also contributes to the flow of upper plenum liquid into the top of the core. The steam volume addition rate from flashing reduces the primary coolant depressurization rate in the higher temperature portions of the primary system. Flashing begins in the core. Because of the core radial power profile, voiding in the core is non uniform. The core approaches a vapor-filled j condition within 2 s and the accompanying reactor kinetic effects produce a core power reduction. The reactor is tripped at about I s on either low primary pressure (presstuizer) or containment high pressure. De critical heat flux condition is reached in the core; heat transfer changes from the i nucleate to post-CHF heat transfer regimes, i.e., transition boiling, film boiling, and forced convection to vapor; and much of the core dries out (Fig. 3). The fuel rod cladding temperatures increase rapidly because of the degrading rod-to-fluid heat transfer (Fig. 5):, The pressure in the broken cold leg decreases rapidly and to a lower level than in the hotter parts of the primary system where the rate of pressure decrease is reduced by flashing (Fig. 4). The break flow regime changes from subcooled to saturated critical flow, voiding occurs in the break and increases the resistance to flow, and the break mass flow rate decreases rapidly. ) Processes / phenomena in the primary coolant system are tightly coupled during the early part of the blowdown period. The decreasing pump-side break flow rapidly affects core cooling processes / phenomena. RCP 2A and RCP 2B share a common plenum in i SG 2. Immediately following accident initiation, the flow through RCP 2B to the break is so high that flow delivered to RCP 2A is small. However, as the CL 2B break flow decreases, the flow delivered to RCP 2B recovers to a near-normal level. The increased flow through RCP 2A is delivered through CL 2A to the downcomer inlet annulus where it joins with the flows from CL 1 A and CL IB With the additional flow from CL 2A the coolant supplied through the intact cold legs (I A,1B, and 2A) exceeds the bypass flow to the vessel side of the break through CL 2B for a briefinterval. The excess coolant enters the still liquid-full downcomer and displaces liquid into the lower plenum, partially refilling the lower plenum, introducing some liquid into the core region, and restoring upward flow i through the core for a brief interval. Concurrently, core power decreases via void reactivity insertion and insertion of the shutdown rods. The release of stored heat from the core diminishes. The cladding temperature peaks (first peak) and begins to decrease. As the primary system pressure continues to decrease, flashing begins in all four cold legs and voiding appears m the pump-suction inlets. Pump performance degrades, and the intact cold leg mass flows rapidly diminish and approach zero. The interval of lower plenum refilling that began with recovery of the flow in CL 2A is terminated and the lower plenum resumes emptying. Flow enters the top of the core through the guide tubes and holes in the upper core plate (countercurrent flow). Sources of the liquid in the upper plenum include coolant present in the upper plenum from the time of accident initiation, coolant from the core, coolant entering the upper plenum from the pressurizer and HL 1, and coolant entering the upper plenum from the upper head. The core liquid volume fraction remains about the same for the remainder of the blowdown period, and the core a Cladding temperatures for each computational rod are searched and the maximum cladding temperature is located each time plotting data is stored. During the transient. different computational rods might be the site of the maximum temperature. For this AP600 LBLOCA calculation, the maximum cladding temperature was always on the same computational rod which resides in radial ring 1, cell 4 (see Ref. 9, Fig. 6). J 35 + -= y

'4 temperature decrease continues. Top-down cooling is provided by the liquid flow passing from the upper plenum te the top of the core. The fuel assemblies receive different amounts of water depending upon their location. Some of the design factors pmducing the non symmetrical top-down flows are location relative to the control rod guide tubes, location relative to the pressurizer hot leg (HL 1), and the core radial power profile. Thus, some portions of the core experience a top-down quench and others are cooled only a little from the top-down flow of coolant. Westinghouse calculations 7.8 predict more extensive blowdown core cooling following the first peak than predicted in the TRAC-PFI/ MOD 2 calculation.9 As described in the previous paragraph, TRAC predicts that the increased delivery of coolant to the downcomer and core following recovery of flow through CL 2A. 'Ihis coolant entering the core boils and the steam that is generated passes upward through the upper core support plate. The steam flow inhibits the downward flow of coolant from the upper plenum and guide tubes. As a consequence, top-down cooling of the core is limited. The Westinghouse calculations show an increase in the lower plenum liquid content at about the same time predicted by TRAC but this liquid does not enter the core. 'Ihus, there is no steam generation and related retardation of downward flow of liquid from the upper plenum into the core. Thus, the Westinghouse calculation shows extensive top down cooling for the 80% DEGB7 and a complete core quench for a 100% DEGB.8 It is not possible to declare which representation of blowdown phenomena is more correct at the present time. However, we do note that changes made to core heat transfer models as TRAC evolved from MODI to MOD 2 result in diminished heat transfer following the first peak in several LBLOCA calculations. Thus, changes to the TRAC heat transfer models play a substantial role in creating the phenomenological differences observed in the TRAC and Westinghouse calculations. Early in the blowdown period, the valves isolating the CMTs from the DVI lines are opened. A small flow is induced by the liquid head in the CMT and flows into the downcomer. However, all the flow from the CMT bypasses the core as it is entrained in the flow rising upward through the downcomer and carried into the bmken cold leg and out the break. System Resnonse Automatic Depressurization System: The ADS is not activated. Operation is based upon CMT level. The time at which the CMT level decreases to the trip level occurs well beyond the blowdown period. Passive Containment Cooling System Vapor and two-phase fluid pass through the break and exhaust into the containment. A constant containment back pressure of one atmosphere was specified for the TRAC calculation, In reality, the containment atmosphere pressure and temperature increase and a reactor trip signal is generated when a containment overpressure of 5 psi is reached. The increasing containment back pressure will have no effect on the primary because the break remains choked throughout this interval. Passive Heat Removal System: The S signal is generated on one of its initiators (e.g., high containment pressure or radiation or low pressurizer pressure). The S signal induces CMT actuation (opens the isolation valve). CMT actuation induces PRHRS actuation (opens the isolation valve). 36 l

/ j-Flashing and flow reversal occur in the PRHRS inlet piping immediately following break initiation, and two-phase flow passes into HL 1, rendering the PRHRS ineffective. Draining of the PRHRS inlet piping supplies an indetenninate amount of the 1 - liquid to HL 1 and thence the top of the core where it can participate in the top down quench of the core that occurs during this phase. There is a potential for draining of the PRHRS heat exchanger (PHRSHX) volume and outlet piping inventory into the SG 1 exit plenum. Passive Safety Injection System: l The primary system pressure remains above the 4.83 MPa (700 psi) accumulator set j point and there is no accumulator injection during this period (Fig. 6). CMT recirculation begins dun !g this period once the S signal induces CMT actuation (opens 4 the isolation valve). The opemng of the CMT isolation valve is followed by a small recirculation flow as shown in Fig. 7. However, the CMT remains filled with liquid (Fig. 8) because the CMT is replenished by liq uid delivered through the PBLs. The IRWST and sump do not drain during this period. Pnmary Coolant System: 1 Break flow. Liquid flows thmugh the break and into the containment during the first few seconds following the break. Figures 9 and 10 show the break mass flows and exit vapor fractions, respectively. For the remainder of the blowdown period, two-phase 4 critical flow with increasing vapor content passes through the bmak, and by the end of the blowdown period the liquid content in the break flow is small. The break flow peaks i carly, during the interval when liquid passes out of the break. The vessel side break l flow is significantly larger than the pump-side break flow. Flashing in the core i maintains a higher pressure on the vessel side. Also the steam generator tubes and operating pump are added resistances to flow out of the pump side of the break. The ) vessel-side break flow retains more liquid content longer than the pump side. 1 i RQL The RCPs operate normally until void appears in the pump suction shortly after 5 s. However, pump performance degrades rapidly thereafter, and the loop flows also j rapidly decrease. The RCPs continue to operate throughout this period. Cold legs. Because the RCPs continue to operate, the intact-loop cold leg flows (CL 1 A i and IB) remain near steady-state values until voiding occurs m the pump suctions at i about 5 s. 'Ihe flows then rapidly decrease. Figures 11 and 12 show the cold leg mass flows and voiding, respectively. The flow in CL 2A is reduced immediately following i accident initiation as the break in CL 2B consumes most of the flow arriving in the SG 2 i plenum that is common to both CL 2A and 2B. The flow in CL 2B is simply the break i flow. However, once void appears in the RCP-side of the break plane in CL 2B, the resistance to flow through the break increases and the break flow decreases. RCP 2A i _ then receives a nearly normal fraction of the flow delivered to the outlet plenum of SG 2 and the flow delivered to CL 2A approaches the flows in CL 1A and CL 2A at 5 s. l Voiding begins in the inlets of all RCPs shortly thereafter and the flows through the intact cold legs (I A,2A, and 2A) rapidly decrease. Processes / phenomena in the break, j cold legs, and vessel are tightly coupled as discussed in the blowdown period overview. l Hot lens. HL 1 (intact loop) and HL 2 (broken loop) display significantly different behavior immediately following accident initiation. The HL 1 flow bifurcates. Part of the flow continues in the normal flow direction through SG 1 (Fig.13). The remainder of the flow reverses and flows backward into the upper plenum. Flow in HL 2 continues in the normal flow direction (Fig.14). However, a briefinterval of high flow f-37 i'

results from the increased pressure difference induced by the break. Figures 15 and 16 show the hot-leg voiding. Voiding begins almost immediately in the hot legs because the temperatures are higher in this portion of the primat coolant system. Liquid ,l entering the upper plenum from HL 1 participates in top-down cooling of the core. i Pressurizer. The pressurizer empties a large fraction ofits inventory through the surge line and into HL 1 within the first 8 s following the break (Fig.17). The pressuriur inventory entering HL 1 splits, a portion continues in the normal flow direction through SG 1, and the remainder flows back into the upper plenum. The larger fraction of l pressurizer inventory flows into SG 1. The >ressunzer blowdown occurs while its pressure is higher than in the remainder of tle primary system (Fig. 4). Once the pressurizer pressure equalizes with the remainder of the pnmary system, its discharge rate slows. De pressurizer discharge is the primary source of makeup flow during this period. Steam nenerators. Voiding occurs in the hottest and highest portions of the steam generator primary much as that described for the hot legs. Similarly, the outlet flows track those of the corresponding cold legs. Reactor System: Control mds. Upon receipt of the S signal, the control rods insen over a period of about I s and are fully insened by a ) proximately I s following the S signal (i.e., essentially 2 s after the break). The initia shutdown of the reactor results from negative reactivity insened by core voiding (Fig. 21). However, the insened control rods maintain the reactor in a shutdown condition once the core is refilled. The control rods will not be discussed funher for the remaining periods. Lower nienum. The coolant inventory in the lower plenum decreases (Figs. 3 and 29) for the first 3 s after the accident. Dunng this interval, the break is supplied by both the flows through the intact cold legs (I A, IB, and 2A) and the vessel inventory. The lower plenum inventory recovers briefly once the flows through the intact cold legs can supply the break (see the cold-leg discussion for this period), but this partial refilling of the lower plenum lasts only until the RCPs void at 5 s, after which the flows through the intact cold legs decrease and approach zero. At the end of this period,50% of the liquid inventory of the lower plenum is lost. Figure 30 shows the lower plenum liquid and saturation temperatures. The lower plenum coolant reaches saturation at 5 s; it remains saturated until approximately 40 s, when the lower plenum is refilled with subcooled accumulator flow. Core renion. Flashing begins shonly after the break occurs as a result of primary system depressurization (Fig.18). The core inlet flow reverses (Fig.19) and passes downward to the lower plenum and u award to the broken loop through the downcomer. Vapor flow continues in the normal cirection throughout much of the core, resulting in countercurrent flow within the core and at the core exit. During this period, flow passes downward from the upper plenum into the upper portions of the core (Fig. 20). In addition, flow passes downward from the upper head and is delivered to the top of the core through the guide tubes. However, the deliver of coolant from the upper plenum to the core is interru )ted for about 2 s beginning at 2 s. This interrupts the top-down quench of the fue, rods (see fuel-rod discussion for this period). We note that the Westinghouse calculation does not show this interruption. The flow in the core has a multidimensional character and this is reflected in different rates of cladding heating and cooling throughout the core. Voiding in the core insens negative reactivity asnd is the initial mechanism for core power decrease before the control rods are insened (Fig. 21). 38

i t i A two-phase mixture continues in the core throughout this period, but the amount of liquid remaining in the core at the end of this period is small (Fig. 3). i Uoper elenum. The vessel upper plenum begins to void immediately following break i initiation and is almost fully voided by the end of the period (Fig. 32). Fluid exits the upper plenum through the broken loop hot leg and via top down draining of liquid j through the upper core support plate into the top of the core (Fig. 20). Countercurrent flow exists in the upper plenum. Vapor enters the upper plenum from the core region l and early in the period, the intact loop hot leg delivers water to the upper plenum. Uoner head The vessel upper head communicates with the upper plenum through the i guide tubes and drain holes in the u j the top of the downcomer annulus. pper head support plate. It also communicates with De guide tubes can also deliver coolant directly to the top of the core. After a delay of approximately 5 s, the upper head begins to void i and is almost fully voided by the end of this period (Fig. 31). De largest flow fium the i upper head is to the upper ? enum through the drain holes (Fig. 34), although this flow l does not begin for about 3 s following the accident. Further, the drain hole flow is exposed to the two-phase flow passing from the upper plenum to hot legs and an t d undetermined fraction is, therefore, entrained and carried into the hot legs. Upper head inventory is also delivered directly to the top of the core through the guide tubesb (Fig. 28). The guide tube flow begins immediately following accident initiation. The upper l head to downcomer flow is small (Fig. 35). 4 j Downcomer. The downcomer annulus remains either full or nearly full until about 7 s l (Fig. 27). Den the downcomer begins to rapidly ernpty and approximately 65% of the liquid inventory is lost by the end of the period. The voiding proceeds from the top of i the downcomer annulus downward under the influence of the cold-leg break. When the i voiding reaches the level of the broken cold leg nozzle, voiding contributes to a j continuation of the break flow decrease, which to this time has been dominated by the decrease of the primary system pressure. Early in the transient, flow from the intact cold 4 legs passes around the downcomer annulus and out the broken cold leg. In addition, j inventory from the other areas of the vessel supply the break until the flow has j dmunished to a level that can be supplied by flow through the intact cold legs. i Fuel rodt Heat transfer from the fuel rods to the coolant degrades as portions of the core void and the fuel cladding temperatures begin to rapidly increase. Essentially all the j stored energy in the fuel rods is redistributed from the center of the fuel rods to the outer periphery of the fuel rods and the cladding (Fig. 22). The cladding temperature increase slows markedly beginning at about 5 s, and the blowdown or first peak cladding temperature is reached at about 7 s (Fig. 5). The blowdown period fuel-rod temperature i increase is terminated by a combination of processes / phenomena occurring in the primary system. These processes / phenomena alter the core power to flow ratio. The decreasing core power and com? etion of the redistribution of core stored energy l ~ constitute the fuel rod power anc. energy component. The flow component is more complex. b in the TRAC model, the guidetube flow is delivered directly to the top of the core. However, the guidetube flow is distributed uniformly over all assemblies modeled in the cell. 'Ihe guidetube flow is delivered directly to a single assembly in AP600 In the AP600, the guidetubes have slots that communicate between the inner portion of the guidetube and the upper plenum. These are included in the TRAC model. These processes are specifically simulated in the ECOBRA/IRAC model.7 .39

"4 Immediately following break initiation, the vessel side flow is larger than can be supplied by coolant entering the downcomer annulus from the intact cold legs. Thus, the required additional flow is supplied from vessel inventory. When core inventory depletion by flow processes is combined with coolant flashing in the core, a large fraction of the core becomes vapor filled. At approximately 3 s, the vessel-side break flow can be completely supplied by the flows entering the downcomer annulus from the intact cold legs. This is caused by the combination of flow recovery in CL 2A (Fig. I1) and the decrease in break flow (see blowdown period overview and cold leg discussions for this period). The coolant delivered to the downcomer annulus in excess of that needed to supply the break is delivered to the lower plenum and core. The lic uid inventories of both the lower plenum and the core increase (Fig. 3). The coo. ant en,tering the core boils and the resultant steam flow cools the core. The flow of liquid from the upper plenum into the upper portions of the core and the associated heat transfer processes reduce the claddmg temperatures in the upper portions of the core. He cladding temperatures are decreasing at the end of the blowdown period. Fuel-rod cladding ternperatures versus position along the fuel rod are shown as a function of time in Figs. 23 thmugh 26. The temperature data is for the least-cooled rod i in the TRAC calculation,i.e., the rod that experiences the ultimate PCT. A detailed view of the progression of rod heatup, cooling, and reflood can be obtained through careful study of such data. Figures 23 and 24 are for an average-power rod and Figs. 25 and 26 are for a maximum-power fuel rod. This fuel rod experiences only a limited amount of top-down cooling. Other fuel rods experience much more cooling. For example, approximately the upper 1-1/2 m (5 ft) of a number of rods are quenched during the blowdown period (See Ref. 9, Appendices E, F). This quenching arises from liquid flows entering the top of the core through the upper core support plate (Fig. 20). Guide tubes. The vessel guide tubes pmvide a flow path whereby fluid from the upper head proceeds directly to the top of the core (Fig. 28). Although some of 6e holes in the upper core plate are covered by the guide tubes, others are not. This distinction is lost in the TRAC input model which divides the vessel into 16 sectors, each of which represent in some average way fuel elements that are both directly under guide tubes and fuel elements that are removed from guide tube locations. Significant guide tube flow delivery to the top of the core occurs only during the blowdown period. This flow begins immediately as the pressure in the upper plenum and core decrease under the influence of the break. The steady-state fluic. temperature in the upper head is between the hot-and cold-leg temperatures, but closer to the cold leg temperature. Thus, flashing in the upper head is delayed for several seconds after the start of the accident but sustains guide tube flows as the primary pressure decreases. The guide tubes play no significant role dudng either blowdown or the remainder of the transient and will not be discussed further. Vessel structures. The vessel structures include the core support plates, core barrel, downcomer walls, etc. These structures also contain stored heat that interacts with the coolant. With the exception of the downcomer, vessel structures appear to play an insignificant role in this or the remaining pedods of the LBLOCA and they will not be discussed further. Bvnass flows. There are several small flows that bypass the core and are unavailable for core cooling. These include rod thimble flow, core bypass flow, reflector cooling flow, and reflector cavity flow. During steady state, these flows are about 7.5% of the total loop flow. The fluid in these regions flash during the depressurization transient. The bypass flows play no significant role during either blowdown or the remainder of the transient and will not be discussed further. e I 40 - 4

l ^ ~ Steam Generator System (Secondary Side): The S signal initiates feedwater valve closure, and closure is completed during this 4 period. In addition, the containment overpressure signal leads to closure of the main steam isolation valve (MSIV) during this period. Thus, the steam generator is isolated during this period and both feedwater and steam flows decrease to zero (Fig. 33). The SG pressure is slightly higher than the primary pressure at the end of this period but secondary to-primary heat transfer is insignificant because the tubes on the pnmary side of the SG are highly voided. 4.2 Refill Period Overview i. 'Ihe refill period begins at the time accumulator injection flow is initiated, and ends when sufficient water from the PSIS has entered the lower plenum to nearly or fully refill 1 the lower plenum. The end of the refill period also corresponds to the start of the Mood period. Identification of the transition from the refill period to tne reflood period is somewhat uncenain, as the lower plenum is not completely refilled and the collapsed liquid level is fluctuating when the sustained core inventory increase begins. Refill begins at about 12 s and ends at about 43 s (Fig. 3). Early in the refill period, the CMTs supply water to the DVI line. However, as the primary system pressure continues to decrease, ECC flows from the accumulators increase and eventually terminate the CMT flows. Accumulator injection is initiated at the start of i i the refill period when the primary pressure decreases to 4.83 MPa (700 psi) (Fig. 4). The coolant flows to the reactor vessel downcomer through the DVI lines. Early in the period ) the injected liquid does not descend to the lower plenum. Rather, ECC entering the downcomer through the DVI lines is entrained by core generated steam passing upward through the downcomer, carried into the broken cold leg, and exhausted into the containment through the break. Thus, the liquid inventories of the core, lower plenum, and downcomer continue to decrease during the early part of the refill period. In this steam-filled environment, the core begins to reheat (Fig. 5). ECC bypass continues only as long as core-generated steam passes through the downcomer and exhausts to the containment through the break. Steam can be generated only so long as coolant is entering the core. No coolant enters the core from the lower plenum during the early part of the refill period (Fig.19). More important is the cessation of coolant flow from the upper plenum into the top of the core at approximately 20 s (Fig. 20). Shortly thereafter, all remaining liquid in the core turns to steam and by 23 s the core is steam filled (Fig. 3). Concurrent with the termination of top-down liquid flow at 20 s, core-generated steam production decreases. The downcomer begins to refill at 20 s and a second later the lower plenum begins to refill. At 43 s, the lower plenum first fills to the lower core support plate (Fig. 3). There is a small accumulation of water in the core beginning at 37 s but the core is still heating at the end of the period (Fig. 5). System Response 4 Automatic Depressurization System: j The ADS is not activated during this period (See Blowdown Period System ~ Description). i' - 41 -m-, -e----

l .<1 Passive Containment Cooling System The containment atmosphere pressure and temperature continue to increase during this .I period. The pressure peaks at about 50 psia (Ref. 8) and remains between 45 and 50 psia for an extended period. The system pressure appears to remain high eno, ugh to i keep the break choked thmughout period. We again note that a constant containment back pressure of one atmosphere was specified for the TRAC calculation. Precise statements regarding the time at which the break planes unchoke are not possible because i the calculation was not performed with a coupled model of the primary system and the containment. Passive Heat Removal System: i 'I The flow entering the PRHRS piping from the hot leg is highly voided at the start of the refill period and fully voided by the end of the period (Fig.15). Vapor entering the i PRHRSHX condenses, but the energy transferred to the ITWST heat sink is minor. The PRHRS is not designed to operate during accidents in which highly-voided conditions exist in the primary system. Passive Safety injection System: Accumulator flow injection begins at about 12 s (Fig. 6) when the system pressure drops below the 4.83 MPa (700 psi) accumulator set point (Fig. 4). The accumulator flow rapidly terminates the CMT flow that began during the blowdown period. Neither the IRWST nor the sump drain during this period. Pnmary Coolant System: 1 Break flow. He pump-side break flow is essentially all va)or (Fig.10), and critical j flow conditions persist during this period. The vessel side reak flow entrains some i liquid flow from the downcomer but it too is highly voided. The decrease of the primary pressure (Fig. 4) and the highly voided nature of the flows from each side of the break contribute to a diminishing break mass flow (Fig. 9). At the start of the refill period, the total break flow is approximately 5700 kg/s. At the end of the refill period, the break flow is near zero. However, the steam flow passing from the core to the pump side break is sufficiently large to influence the upper plenum pressure. His effect becomes more important during the reflood transient. RQL The reactor coolant pumps trip at about 16 s. This has very little effect on the mass flow because the flow through the pumps is highly voided at the start of the refill period, and pump performance is highly degraded. Cold legs. The cold legs progress from a highly voided to a near fully voided condition during this period (Fig.12). The connection to the cold leg PBL voids, and a vapor path is o>ened to the top of the CMT, which initiates CMT draining as previously discussec Hot legs. The broken loop HL 2 is essentially fully voided from the beginning of the refill period (Fig.16). The intact-loop HL 1 is highly but not fully voided at the beginning of the period (Fig.15). There is a flow from this hot leg back into the upper alenum between 10 and 20 s. This flow is from the pressurizer and backflow from the . oop steam generator. Both hot legs are fully voided by the end of the refill period. 1 42 m

Pressurizer. After the pressurizer pressure equalizes with the primary pressure near the end of the blowdown period, the rate at which the pressurizer inventory is discharged into the loop-l hot leg slows (Fig.17). The pressurizer discharges the remainder of its inventory into the loop-l hot leg by the end of the refill period. Steam renerators. The steam generator tube bundles void at about 16 s and remain voided. As the primary system pressure decreases, secondary-to-primary reverse heat is possible. However, because the primary side tube bundles are vapor filled, the secondary-to-primary heat transfer is insignificant. Some mass flows back into HL 1 from the steam generator between 10 and 20 s. Reactor System: Lower Plenum. The key process during this period is refilling of the lower plenum to set the stage for core reflood. At the start of the refill period, the lower plenum inventory decreases (Figs. 3,29), and it continues to decrease until approximately 22 s. During the interval between 12 and 22 s, the break flow rapidly decreases as discussed previously. Steam generation in the core has diminished because the core approaches a fully voided condition and there is little flow of either liquid or vapor at the bottom of the core, which can pass into the downcomer and retard the flows (Fig.19). Consequently, the ECC flows injected through the DVI lines begin to refill the downcomer and lower plenum at about 22 s. The liquid level in the lower plenum first reaches the bottom of the core at approximately 43 s, ending the refill period and markmg the beginning of the reflood period. Flow oscillations occur during the lower plenum refilling process. Core Reelon. The core is at decay power levels and is maintained in a shutdown condition by the control rod reactivity (Fig. 21). The core is about 90% voided at the start of the refill period and fully voided by 20 s. Between 12 and 20 s, the downward flow of liquid into the top of the core from the upper plenum is decreasing and approaches zero (Fig. 20). After 20 s, lacking the introduction of additional liquid into the core, the vapor residing in the core is heated and moves upward, but the vapor mass flow is very small and produces little cooling of the core. Uoper Plenum. The upper plenum is about 90% voided at the start of this period, becomes fully voided by 25 s, and remains fully voided throughout the remainder of the refill period (Fig 3). The upper plenum supplies essentially all the liquid flow into the core between 12.3 and 20 s. The coolant flow from the upper plenum into the core supplies the liquid that sustains steam generation in the core and ECC bypass. Unper Head. The upper head is about 90% voided at the start of this period, becomes fully voided shortly thereafter, and remains fully voided throughout the remainder of the refill period (Fig. 3). The liquid in the upper head at the beginning of the period enters the upper plenum through the upper support plate drain holes (Fig. 34) and the downcomer through the downcomer-upper head flow path (Fig. 35). There is no flow into the upper plenum through the guide tubes after the upper head liquid level drops below the top of the guide tubes near the end of the blowdown period. Downcomer. See the description of lower plenum processes and phenomena for the refill period. Fuel Rods. At the start of the refill period, the fuel rods cool as a result of liquid introduced into the core late in the blowdown period. The temporary restoration of liquid flow into the core during the blowdown period occurred when the intact-loop cold leg flows were sufficient to supply the rapidly decreasing vessel side break flow ~ 43

However, subsequent voiding in the cold legs at the end of the blowdown period terminates the core flow recovery and the core becomes vapor filled during the refill )eriod as previously discussed. In this vapor filled environment, the cladding-to-coolant 1 eat transfer degrades and the fuel rods begin to reheat (Fig. 5). As previously discussed, the flows refilling the lower plenum are oscillatory. Thus, some liquid begins to appear at the core inlet before the end of the refill period (Fig.19). He core heats up (Fig. 5) and continues to heat to the end of the period. The flow of liquid downward from the upper plenurn into the core between 12.3 and 20 s removes some of the core decay heat, especially at the top of the core. i Steam Generator System (Secondary Side): Secondary to-primary heat transfer is initiated during this period, but the steam-generator pnmary is vapor filled and the reverse heat transfer is limited. 4.3 Reflood Period Overview Reflood begins after the lower plenum refills and the core begins to refill (Fig. 3). Reflood is completed when the entire core is quenched. Initially, core reflood is quite ra sid because (1) the downcomer head is increasmg as it refills, (2) the downcomer heac is initially resisted by only a steam-filled cose, and (3) core-generated steam flows through the core, coolant loops, and break are small so the pressure resistitig the downcomer head is small. For this analysis, the ieflood period begins about 43 s after accident initiation. Because of the high fuel rod temperatures at the beginning of reflood, the entire spectrum of thermal regimes, starting with single phase liquid and progressing upward through the core with nucleate boiling; transition boiling; film boiling; and single-phase steam flow, are encountered. However, by the end of this period, fuel temperatures have peaked and the entire core is quenched. Because of droplet carryover from the core and subsequent deentrainment at the upper core plate and grid spacers, top quenching and local quenching occur in addition to bottom quenching. Higher vapor velocities and liquid entrainment occur in the central region of the core where higher-powered fuel rods are located. The entrained liquid has a cooling effect on the fuel rod regions above. De upper portions of the core remain cooler because significant cooling occurred during the previous period as liquid from the upper plenum entered and cooled the upper portions of the core. Some of the entrained liquid is deentrained at the upper core plate. The remainder is carried into the upper plenum, where it is deentrained, fomung a two phase pool, or carried over into the hot legs. Liquid from the pool can reenter the low-power regions of the core through the upper support plate because of the lower vapor velocities in those regions. A three-dimensional flow pattern results: in the core, flow is from the low-to high-power regions, while in the upper plenum the flow passes from the high-to low-power regions. Liquid from the upper plenum two-phase pool may be further entrained and carried over into the hot legs. In traversing the upper plenum, this liquid may be further deentrained on upper plenum internal structures. As the bottom quench progresses upward through the core, more liquid is carried over to the upper plenum pool. Conditions exist (steam passing through the upper plenum pool) by which liquid can be entrained and carried into the hot legs. Droplets carried through the hot legs to the steam generators will evaporate by reverse steam generator heat 44 a

i I j l 1~ 1 transfer, causing a pressure increase in the steam generator bundle. This phenomenon is 1 called " steam binding." There is some liquid accumulation in the steam-generator inlet late in this period, which will alter the loop pressure losses somewhat. For much of the reflood period, manometer oscillations between liquid in the downcomer and the core lead to an oscillatory reflood rate. The downcomer hquid level i (head) is the driving potential for liquid to enter the core. Liquid passing through the i downcomer and entering the core from the lower plenum boils; the resultant increased steam flow from the core to the pump-side break causes the upper plenum and core l yressures to increase. The increased pressure in the core alters the core-downcomer force nlar ce and reduces the flow from the downcomer. With reduced core flow, there is less steam generation, the core side pressure decreases and agair alters the core downcomer 4 j force balance causing an increased downcomer to core flow in this part of the cycle. l Associated *;dtn the oscillations is increased liquid entrainment and carryover that accelerate ) j core quenching. i After the accumulators have injected their inventory into the primary coolant 4 system, the CMTs resume draining. Quenching of the entire core is estimated to occur before the accumulators empty. Neither the present calculation nor that of Ref. 8 were extended to the time the accumulators empty. System Resnonse l Automatic Depressurization System: I The ADS is not activated during this period (See Blowdown Period System Description). j Passive Containment Cooling System t The cooling processes of the PCCS become more important as the transient proceeds. i The primary functions of the containment are several. First, the containment is to i provide a barrier to fission product release. To do so, the integrity of the containment structures must be preserved. Thus, the containment pressure increase must be limited. l Second, the decay heat of the core must be transferred to the ultimate heat sink, the j atmosphere. Safety systems are provided to cool the containment, so the containment pressure increase will be limited and the containment integrity will be retained for the i long-term cooling period. The break planes unchoke during the refill period. Precise statements regarding the time j at which the break planes unchoke are not possible because the calculation was not performed with a coupled model of the primary system and the containment. The containment and primary coolant system are coupled during this period and the coupling may have a significant impact on core cooling processes. The PCCS is not modeled in the TRAC calculation. The following qualitative description of PCCS operation is i provided for completeness. 3 Early in the transient, energy transfer to the atmesphere through the steel containment structure is enhanced by evaporation of liquid deposited on the outside of the steel containment structure near the top of the structure. Buoyancy induced air flow through the air gap between the steel concrete structure and the concrete shield building also cools the structure. By the time the PCCS water supply is depleted during the long-term ~ i cooling period, the decay heat had decreased sufficiently so that the buoyancy induced 45 1

air flow through the air gap between the steel containment structure and the concrete shield building can remove the decay heat of the core. Passive Heat Removal System: The PRHRS mass inventory is depleted either from flashing or from draining, and this system does not contribute to energy removal from the primary. Passive Safety Injection System: The sole source of emergency coolant injection during this period is via accumulator injection (Fig. 6), and this circumstance condnues until the accumulator empties during the long-term cooling period. At the start of the reflood period, about 12% of the accumulator inventory has been discharged into the primary. At 140 s following break initiation, slightly more than 70% of the accumulator inventory has been discharged into the primary and the accompanying cooldown of the core is well underway. The accumulator pressure is high enough to keep the CMT from injecting (Fig. 7). Typically, analyses of the LBLOCA transient are terminated before CMT injection resumes because the key safety acceptance parameter, the PCT, has been reached and the com is cooling. Primary Coolant System: l Break flow. The impact of the break on primary system behavior, major during the l blowdown and early refill periods, becomes less important throughout the refill period. The pump-side break continues to be of moderate importance during the reflood period because it affects the core pressure which in turn affects the core reflood rate. Neither the pump-side nor vessel-side break flows are critical; the break flows are small (Fig. 9). The pump-side flow is vapor until the refill process is well advanced (Fig.10). The vessel-side break flow is nearly so, but significant amounts of liquid appear at the vessel-side break plane after 60 s, as the oscillating downcomer liquid level approaches the cold-leg nozzle connections to the vessel. RCP1 The reactor coolant pumps have coasted down and are inactive. 1 Cold legs. The intact cold legs are essentially voided and there is no flow through them during the reflood period (Figs.12 and 11, respectively). After 60 s, the oscillating downcomer liquid level reaches the cold-leg-nozzle connections to the vessel and significant amounts of liquid begin to move through the broken-loop cold leg to the break. Hot lens. The hot legs are vapor filled during the early part of the reflood period (Figs. l 15-16). As the upper plenum begins to accumulate liquid later in the transient (Fig. 3), small amounts of liquid are entrained and carried into the hot legs. Reverse heat transfer from the secondary side of the steam generators vaporizes the liquid carried into the j steam generators. The vapor generation in the steam generators participates in a process / phenomena called " steam binding." The resultant pressure increases propagate toward the core. The increased core pressure offsets some of the downcomer liquid head and the core reflood rate decreases, i Pressurizer. The pressurizer is voided. l Steam generators. See the previous discussion of reflood phenomena in the hot legs. During the latter phases of the reflood period, liquid is carried to the steam generator 46 i g ..y i-- +- ,_m .m

inlets (Figs.15 and 16). Some of the liquid is deentrained there, and that which is canied into the steam generator tubes is vaporized. Reactor Sysuu: Lower Plenum. The reflood period begins when lower plenum refilling is completed. Because of the oscillatory filling process, the actual time assigned to the stan of this period is somewhat uncenain. However, by 45 s, the lower plenum is essentially full (Fig. 29) and lower plenum liquid is subcooled (Fig. 30). Core Rerion. He core inlet flow is highly oscillatory from a time shonly before the end of the refill period and continuing until the core is completely reflooded (Fig.19). Two reasonably distinct reflood rates are calculated (Fig. 3). The initial reflooding rate is about 2.5 cm/s (1 in/s) and continues for about 25 s after reflooding begins. The ultimate PCT is attained during this interval, shonly before 60 s. The several core-downcomer level oscillations that occur dudeg this interval have a low frequency. This interval ends when the downcomer fills to the level of the DVI line nozzles (Fig. 3). The reflood rate subsequently slows because the downcomer level has reached the DVI nozzles and the downcomer head is no Icnger increasing with the refilling process, partially because liquid discharge through the vessel side cold leg break has resumed (Fig. 11). The frequency of the oscillations increase, beginning at this time, approximately 68 s. The calculated reflood rate beginning at 68 s is about 0.8 cm/s (0.3 in/s) ne flow oscillations are considerably damped in the confined flow passages of the core [ compare the core outlet mass flows (Fig. 20) to the core inlet mass flow (Fig. 19)]. He liquid fraction, although oscillatory, steadily increases (Fig. 3) as the core vapor fraction decreases (Fig.18). The manometric oscillations between liquid in the downcomer and the core produce the oscillatory reflood rate. Liquid entering the core i from the lower plenum boils and the increased pressure in the core reduces the flow from the downcomer as previously discussed. Upoer Plehun The upper plenum begins the reflood period in a fully voided condition (Fig. 32). However, as the reflood period continues, the core liquid level increases, accompanied by significant steam generation that carries liquid into the upper plenum (Fig. 20). A small amount ofliquid accumulates in the upper plenum (Fig. 3), but there are several competing processes that prevent a large accumulation of liquid. These processes include entramment and transpon of liquid into the hot legs, as previously discussed, and the flow ofliquid to regions of the upper core support plate above the lower powered regions of the core. As the steam generation in these regions is small, some of the upper plenum liquid flows downward into the top of the core. Upoer Head. De upper head remains fully voided during this period (Fig. 31). Downcomer. The downcomer continues to refill from accumulator flow injection through the DVI line (Fig. 3), until the level of the DVI nozzles is reached at about 68 s. Several slow oscillations occur while the downcomer is filling, but more rapid and continuous flow oscillations arise after the downcomer fills to the level of the DVI nozzles. The downcomer does not continue refilling once the liquid reaches the level of the DVI nozzles. Carryover of liquid into the cold legs and out the break also contributes to limiting funher increases in the downcomer liquid level (Fig. I1). The ECC delivered to the lower plenum is subcooled (Fig. 30). The delivery of subcooled liquid to the lower plenum suggests that condensation in the downcomer is limited. However, this code-calculated result should be revisited because the uncenainty associated with modeling condensation in complex geometries is large. The post-CHF 47

i heat transfer processes that actually occur in the core are dependent upon the amount of subcooling so the amount of condensation occurring in the downcomer is significant. Fuel Rodt At the stan of the reflood period, fuel rod temperatures in much of the core are increasing (Fig. 5). However, as the core begins to refill, the heat transfer environment of the fuel rods change. The primary direction of the quench is from the Fottom up, as shown in Figs. 24 and 26, for the average-power and maximum powcr rods respectively. The upper 1-1/2 m (about 5 ft) of much of the core were c uenched during the refill period. There is a minor reheat of the upper core during the ear.y part of the reflood period, but the cladding temperatures rema n well below the peak cladding temperatures lower in the core. Coolant advancing upward from the bottom of the core is the primary mechanism for core cooling during reflood. The lower sections of the fuel rods are cooled by liquid convection and nucleate boiling, higher elevations of the j fuel rods by transition and film boiling (>ost-CHF heat transfer), and even higher elevations of some fuel rods receiving top-c own flow by nucleate boiling. The quench front advances upward in the core and eventually the entire core is quenched as the core refills. Steam Generator System (Secondary Side): See the discussion of hot-leg phenomena during the reflood period, l i 4.' 4 Long-Term Cooling Period Overvieg The TRAC-PFl/ MOD 2 calculation presented in this document does not cover the long-term cooling period, nor do the Westinghouse-prepared calculations documented in i Refs. 7-8. Thus, the descriptive information provided for the long-term cooling period is r based upon information presented in Ref.12, discussions of long-term cooling processes j and phenomena at several recent NRC-sponsored meetings (Ref.13), and information appeanng in quick-look reports documenting tests in the ROSA facility (Ref.14). The long-term cooling period is entered once the entire core is quenched. 3 However, the core is only partially refilled at this time, and accumulator injection is approaching completion. After the accumulators have emptied, CMT injection resumes. In addition, if the primary pressure is sufficiently low, IRWST injection may start. Simultaneous CMT and IRWST injection have been observed in a ROSA intermediate-break loss-of-coolant-accident experiment.14 When the flow out of at least one CMT into the primary reduces the CMT liquid volume to 67%, the ADS is actuated and three stages of valves at the top of the pressurizer, and one valve stage connected to a hot leg, open in sequence. However, the ADS operation does not strongly impact the course of the LBLOCA transient because primary depressurization is well advanced when ADS stages 1-4 actuate. The break alone is sufficient to depressurize the primary to the containment pressure (Fig. 4). The AP600 long-term cooling period behavior is similar for many transients and accidents because the planned end states of events such as a LBLOCA, SBLOCA, main steam line break, and steam generator tube rupture are similar.12 Specifically, the primary is depressurized, liquid enters the primary from the IRWST and sump, there is boiling in i the core, and the primary has a direct path for releasing mass and energy to the 1 48 \\

containment (through the founh stage ADS valve). Some differences in processes and phenomena occurring during the late phase of various transients and accident are expected, but eese differences are not expected to have a major impact on the late-period course of the sequences. Draining of the IRWST is expected to take several days, after which water from the sump is recirculated indefinitely. Water from the IRWST passes to the vessel downcomer through drain lines that connect to the DVI lines. The injected water is heated in the core, a portion evaporates and a two-phase liquid level is established in the core, and the steam generated in the core enters the containment through the break'and the founh stage of the ADS. The passive containment cooling system plays an imponant role during the long-term cooling period by condensing steam released through fourth-stage ADS and returning the liquid to the IRWST and sump. System Resoonse Automatic Depressurization System: The ADS actuates late in the reflood period. The ADS is actuated when the volumetric liquid level in one of the CMTs decreases to 67% of its fell value. The initial ADS actuation is to open the first stage valves between the pn:ssurizer and the IRWST. The second and third stages open at timed intervals to provide additional flow area between t'ae same two locations. Continuous draining of the CMTs does not begin until draining of the accumulators approaches completion, i.e., on 'he order of 5 min, following LBLOCA initiation. The driving potential for releasing primary mass and energy through ADS stages 1-3 and into the IRWSTis negligible when ADS stages 1-3 are actuated. The primary pressure is already low and any flow must overcome the sparger submergence head. The fourth stage ADS opens a path between the hot leg and the containment. This path will panicipate in the continuing release of primary mass and energy to the containment already occurring through the cold-leg break. Passive Containment Cooling System See the system response description for the reflood period. Passive Heat Removal System: The PRHRS is essentially voided and, therefore, does not contribute to the transport of energy from the primary to the IRWST. Passive Safety Injection System: The accumulators and CMTs empty during this period. Coolant is provided over a period of several days via IRWST flow injection. There is some evidence (e.g., the ROSA test of a 200% DVI line break, Ref.14) that the intervals of CMT and IRWST injection may overlap. The IRWST injects its inventory through drain lines that connect to the DVI lines. Throughout the interval following the LBLOCA initiation, liquid collects in the sump region. The rate ofliquid accumulation is speculative at present. However, lines connect the sump to the IRWST drain lines which, in tum, connect to the DVI lines. Sump injection is by the same process as IRWST injection, via the ~ differences in head between the coolant source and the primary. As the IRWST drains, the IRWST driving head will diminish. At some point, the IRWST and sump heads 49 e,y

-4 approach equality, and injection frcm both will continue until the IRWST completely drains. Primary Coolant System: I Brak flow. Opening of the fourth stage ADS will provide a more direct path for release of core-generated vapor to the containment. The pump side break flow, already extremely small, will further diminish. The AP600 is designed such that in the very late phases of an accident, after the accumulators, CMTs, and IRWST have drained, the . evel of liquid surrounding the vessel will cover the cold leg piping. Liquid will drain into the broken cold leg and terminate shy iwtner break flow. RCP.L The pump rotors are either stationary or freewheeling depending upon the velocity of the vapor flows through the loops. Cold lens. The cold legs will eventually fill with IRWST water. Hot lens. The hot legs will stratify with core-generated vapor passing through the top of the pipe and passing to the contamment through the fourth stage ADS. Pressurizer. The pressurizer remains voided. ) Steam neaerators. The steam generator tube bundle remains voided. Reactor System: Lower Plenun The lower plenum is full. Core Recion. Boiling continues in the cost for an extended period. Eventually, decay power decreases sufficiently so that core cooling is maintained by the buoyancy induced, single-phase flow of subcooled liquid through the core. 4 Unner Plenum. The upper plenum will refill to the hot-leg elevation. There is a potential for condensation events as subcooled water comes into contact with vapor in the upper plenum. Unoer Head. The upper head remains voided during this period. There is a potential for condensation events as subcooled water comes into contact with vapor in the upper head. Downcomer. The downcomer will refill to the cold leg elevation Fuel Rods. The fuel rods are completely immersed in water, but boiling continues until the decay power reduces to a sufficiently low level that core cooling is maintained by buoyancy mduced, single-phase, subcooled liquid flow through the core. Steam Generator System (Secondary Side): Secondary-to-primary heat transfer has diminished to zero as the primary and secondary reach equilibrium. I 50 a

4 i 5.0 PIRT RESULTS 'Ihis section summarizes the results of the AP600 LBLOCA PIRT. For each phase of the LBLOCA, i.e., Blowdown, Refill, Reflood, and Long-Term Cooling, a hierarchy of rankmg evaluations is presented. The hierarchical rankmg logically proceeds in order from 4 i systems, to components, and finally to processes / phenomena. A slightly different process was followed for both the original PIRT demonstration 3 and the recent INEL AP600 PIRT efforts for a small-break loss-of-coolant, main steam line break, and steam generator tube rupture scenarios.t2 In contrast to these earlier PIRT evaluations, the importance rankings for the present AP600 LBLOCA PIRT are explicitly performed at each of three levels: i systems, components, and processes / phenomena. - One of four importance ranks is assigned. For example, the importance of the i system is evaluated as "HIGH" if the perfonnance of the system controls the course of the transient during the period, specifically ifit has a controlling impact on the ultimate PCT. Here, the emphasis on ultimate is meant to convey that the focus is on the PCT for the total j accident sequence. Of course, the ultimate PCT is the result of a series of coupled processes occurnng in the systems and components of the AP600, and these linkages are considered by the PIRT team in the ranking process. The imponance of the system is i evaluated as " MEDIUM" if it has a moderate influence on the outcome of the sequence l-resulting in the ultimate PCT. The imponance of the system is evaluated as "IDW' if it has j a small influence on the outcome. Finally, if a system has either no influence or an insignificant influence, it is rated as " INSIGNIFICANT." For those systems ranked " LOW" or " INSIGNIFICANT", no funher effon is made to rank the relative imponance of either components or processes / phenomena. For systems receiving an importance ranking 4 i of either "H/GH " or " MEDIUM", ranking proceeds to the next level, and each component is ranked according to its impact on the course of the transient, specifically its impact on the 3 ultimate PCT. The highest possible rank for a component in a " MEDIUM"-ranked system is " MEDIUM." A similar logic is applied to the ranking of processes / phenomena. The 4 j highest potential ranking of processes / phenomena is limited to the highest ranking of the }; associated component. Within any system, component, and process, a variety of " basic phenomena" occur. In Ref.15, the following are listed as basic phenomena: evaporation due to depressurization, evaporation due to heat input, condensation due to pressurization, condensation due to heat removal, interfacial friction in venical flow, interfacial friction in horizontal flow, wall to fluid friction, pressure drops at geometric discontinuities, and pressure wavc propagation. The concept of " basic phenomena" has been used in the PIRT process descrited in this document. Specifically, the term basic phenomena is introduced as a general entry in the list of processes / phenomena for a given component. No ranking is provided for basic phenomena. However, it is understood that each component must, at a minimum, have a model for each of the listed basic phenomena should the phenomena occur in the scenario. Thus, if a component exists in the facility and enters into the scenario, the basic phenomena occurring in that component must be modeled, even if the component rank is " LOW." In some cases, some of the basic phenomena listed are very imponant, e.g., evaporation due to depressurization (flashing) in the core flow channels of the reactor vessel during the blowdown and refill periods. For such cases, that element of the listed basic phenomena is individually ranked in the process / phenomena list. In addition to the general thermal. hydraulic expenise of the individual members of the PIRT team, several resource materials were available to the PIRT members both in 51-

advance of, and during, the ranking sessions. A detailed tabulation of AP600 systems, components, functions, and processes / phenomena was prepared. Descriptions of AP600 systems and components were also prepared. PIRT insights were developed and documented based upon the documentation reviewed. This supporting information is presented in Appendix A of this document. While preparing an early version of Appendix A, an attempt was made to identify and clearly differentiate between processes and phenomem,. However, this proved difficult to accomplish and seemed to irquire a more extensive effort than appropriate for this PIRT activity. Therefore, the process entry was replaced with a statement of system or process function and processes and phenomena were considered as a combined category (process / phenomena). Finally, a detailed LBLOCA scenario description was provided (Section 4). Much of the description is based upon a code-calculated LBLOCA sequences. The PIRT team acknowledges the potential pitfalls of excessive reliance on code-calculated results in a PIRT activity. However, the insights gained from a thoughtful consideration of code-calculated scenados are thought to outweigh the potential for makmg enuneous rankmg conclusions. 5.1 AP600 LBLOCA PIRT The LBLOCA scenario is described in Section 4.0. The results of the PIRT team evaluation are presented in this section. Brief biographies of the individuals participating in the PIRT sessions are presented in Appendix B. 5,1.1 Blowdev'n Period The blowdown period covers the interval from break initiation until the stan of accumulator injection, an interval of approximately 12 s. Only the primary coolant system and reactor system have a controlling impact (HIGH) during the blowdown period. The primary coolant system, as defined within this document, consists of the steam-generator primary side, hot-leg piping, cold-leg piping, pressurizer, RCPs, and the i break. Within the primary coolant system, The impact of RCP component operation is judged to be HIGH. Until they void, the RCPs continue to force fluid through the intact loops. These flows supply the break, and when the break flow diminishes, briefly provide sufficient flow to supply not only the break but deliver flow to the downcomer, lower plenum, and core. This coolant panicipates in the termination of the blowdown period cladding temperature increase and clearly limits the ultimate PCT. A brief time later, when the pump inlets void, the cold leg flows rapidly decrease, the break flow again, exceeds the intact loop flows, the core inventory again begins to decrease, and the stage is set for the resumption of cladding heatup during the refill period. The most imponant pump phenomenon is degraded performance caused by voiding in the pump inlet (HIGH). The most imponant primary coolant system feature is the break component (HIGH), which is the cause of all the phenomena leading to core heatup. The most important phenomenon occurring in the break itself is critical flow (HIGH). However, two additional factors influencing the break flow are the inlet voiding, which arises from flashing in the cold leg and downcomer, and the decreasing primary pressure. The impacts of the cold-leg components are judged to be MEDIUM. The cold legs are the conduit for supplying coolant to one side of the break and to the core via the downcomer and lower plenum. There is a briefinterval before the cold-legs void when the cold-leg flows exceed the break flow and supply flow to the core via the downcomer and lower plenum. This flow participates in the termination of the blowdown cladding heatup. The impacts of the pressurizer and surge line are also judged as MEDIUM, As the pressurizer inventory is injected into the attached hot leg, its inventory splits. Part passes through the steam generator to the cold leg where it supplies the break and contnbutes to the mass balance that eventually terminates the cladding heatup during the blowdown interval. The remainder 52

!^ i l passes through the hot leg into the upper plenum where it contributes to the quenc,hing of i fuel rod cladding in the upper portions of the core. The steam generator, primary-j i side component impact is judged to be LOW because the steam-generator secondary is l j isolated early, and this action limits the primary to-secondary heat transfer. The impacts of the hot leg components are judged to be LOW. The inveltory of the intact hot leg reverses and is delivered to the upper plenum where it participates in top-down quench of a j portion of the com. However, the impact on the ultimate PCTis believed to be small. The reactor system, as defined in this document, consis.s of the vessel-related i components, specifically the control rods, core flow channels, downcomer, fuel rods, guide tubes, lower plenum, structures, reflector / bypass, upper head, and upper plenum. "wo of the vessel components in the reactor system have a controlling (H/GH) influence l on the course of the transient. These two components, the fuel sud and core flow l channel components, are closely coupled. The fuel rod component provides the energy source and is affected by the heat transfer processes which lead to cladding heatup. I During this period, both stored energy redistribution and decay heat are important, but j stored energy redistribution (HIGH)is dominant process. Another highly ranked phenomena occurring in the fuel rod is rod structure heat transfer, which moves the energy within the fuel rod to the cladding surface. The gap conductance largely (HIGH) i determines the energy transport rate to the cladding surface. The blowdown period has two subintervals. The early intervalis characterized by a j rapid depressurization in the core coolant channel component, a sharp reduction in core liquid inventory, and a region of stagnation as coolant passes out both the top and l bottom of the core under the influence of the two sides of the break. The key channe) flow l process occurring during the early interval is flashing, an interfacial mass transfer process j which produces a nearly voided core (HIGH). Heat transfer from the fuel rod component to the coolant is small during this early interval and the fuel rod cladding j heatup is at a nearly adiabatic rate. The later interval is characterized by flow reversals at both the core inlet and outlet. The interaction between RCP and break flows resulting in j the core inlet flow reversal has been previously described. In addition, coolant flows countercurrent to the steam flow from the upper plenum into the top of the core - (MEDIUM). The key interfacial heat and mass transfer processes occurring during the latter interval are flashing and boiling, the latter cooling the lower and mid-portions of the i core and generating steam. The steam flows upward and interrupts the downward flow of coolant from the upper plenum, thereby decreasing the potential for top-down coolinga, The key (HIGH) convective heat transfer phenomena occurring at the fuel rod surface during the latter part of the period is post critical heat flux (CHF) heat transfer associated j with transition and film boiling. i j The upper plenum component is the source of coolant flowing into the upper portions of the core. With quenching of the uppe j temperatums are experienced in the mid-pom,r portions of the core, the highest clad on of the core, an area most heavily influenced by the cooling processes associated with the brief period of core inlet flow. Thus this component is of MEDIUM importance; it is the source of the top down rewet that is ranked j as having medium importance. The impact of the guide tube component is also 2 i' a There are significant differences in the amount and impact of the top-down cooling predicted by TRAC 9 i and COBRA / TRAC.7 The PIRT Ranking Team concluded that MEDIUM was the appropriate rank for the upper plenum, upper head, and guide tube components. This ranking assigns slightly more weight to the TRAC-calculated perspective.1lowever, the rankings of these components and the associated top-down quench phenomena should be reviewed periodically. i; - 53 l lj _

MEDIUM. The guide tubes deliver upper plenum fluid to the top of the core, providing coolant for the top-down quench. Similarly, the im)act of the upper head component is MEDIUM, as it is the source of the water passing t trough the guide tubes and the upper support plate drain holes. The lower plenum component is also ranked as MEDIUM because it is the linking component between the phenomena in the downcomer component ranked MEDIUM and core flow channel components. Both components deliver flow, but the phenomena occurring in these components are basic! The impact of the control rod component during this period is LOW. The negative reactivity inserted by voiding in the core is sufficient to shut down the core (HIGH). It is assumed that the control rods insert during this period and keep the core shutdown during the long term ~ cooling period when the core refills. De impact of the reflector / bypass component is LOW. The impact of the vessel structural component is Low because the release of stored heat to the primary is slow. One system is ranked as having a small impact (LOW) during this period. The steam generator system (secondary side) system has a small impact on the primary system. The feedwater and main steam lines are isolated early in the period. Although some primary system energy is transferred to the secondary, the impact on the key primary system processes is small. } A number of systems are ranked as having either no impact or an insignificant impact (INSIGNIFICANT) during this period. The ADS system is' inactive. The PCCS has an insignificant impact on the primary system during this period. Critical flow conditions exist at the two break planes, so there is no feedback from the changing containment conditions to the primary system. The PRHRS has an insignificant impact on the course of the transient. Opening of the PRHRS isolation valve is initiated upon receipt of the S signal about I s after break initiation. However, the normal buoyancy-induced circulation does not initiate because the PRHRS inlet piping inventory blows down into the j hot leg under the influence of the depressurization transient. Some coolant does pass from the PRHRS into the hot leg, thereafter passing to the upper plenum where it participates in t i the top-down cooling of the core. The PSIS and associated valves and piping have an j insignificant impact on the course of the transient during the blowdown period. The j l accumulator component is inactive according to the period definition. There is a small i CMT component injection into the DVI lines and thence into the downcomer. However, this coolant is entrained in the upward flow of coolant through the downcomer and carried out the vessel side of the break. The IRWST component and sump component are j j inactive. i j A tabulation of the system, component, and phenomena rankings for this phase of j the LBLOCA is presented in Table 1. i j 5.1.2 Refill Period l Re Refill period covers the interval from the time ECC injection is initiated until the lower plenum is refilled, the interval between approximately 12 and 45 s. i The primary coolant system continues to have a controlling influence (HIGH) on the course of the transient during this period. The break component continues to i have a controlling (HIGH) impact on the course of the transient but its relative importance j diminishes toward the end of the period as the mass and energy flow through the break i decrease and inventory loss through the break diminishes. The most imponant phenomenon occurring in the break itself continues to be critical flow (HIGH). The i decreasing primary pressure continues to be an important factor causing the break flow to q decrease during this period, the inlets to the vessel-and pump-side breaks having 54 1 v m yr v D

l 2 h b Table 1. Summary of AP600 LRLOCA PIRT Results l System Period Compnnent Period l Process /Phennmenon l Period l 1 l2 l3 l4 lI l2 l3 l4 l lI l2 l3 l4 l ADS lI lI lI lM Valve stages I,2,3 L Cntical flow / flow L Valve stage 4 M Basic phenomena Piping L Batic phenomena Sparger L Condensation Pressure fluctuations Flow oscillations Basic phenomena PCCS lI lI lM l11 Concrete shield building 11 Buoyancy-driven flow fl Air inlet Basic phenomeaa Air exhaust d Air path Piping L Basic phenomena Stect Reactor Containment M H Inside Structure Buoyancy-driven flow L M Pressurization M M IIcat transfer M M Mass transfer - wallcondensation M 11 Draining L M Noncondensible gas M M Dru wall conduction L M Outside IIcat transfer - air flow L H Mass transfer-tvaporauon L 11 Draining L M Noncondensible gas L L Basic phenomena Valves-spray tank H Basic phenomena Water Tank (PCCS) H Basic phenomena

Table 1. Summary of AP600 LilLOCA PIRT Resulls (continued) l System l Period l Component l Period l Process / Phenomenon l Period l lI l2 l3 l4 l lI l2 l3 l4 l lI l2 l3 l4 l PRIIRS lI lI ll ll PRIIRilX Primary side Buoyancy-driven flow Iligh-point trapping IIcat transfer Noncondensible gas Dru-wall heat transfer Scoondary side Buoyancy-driven flow Ileat transfer to IRWST pool Basic phenomena Piping Phase separation y Basic phenomena Valves Basic phenomena PSIS lI lIl lIl lIl Accumulator I 11 11 L Noncondensible gas injection I M Discharge 11 11 Basic phenomena CMT I L I 11 Buoyancy-driven flow / Recirculation L Draining 11 Refilling L Thermal stratification M Flashing L L Condensation L Noncondensible gas L Flow termination via accumulator L injection Ambient heat loss L Basic phenomena y n e

Table 1. Summary of AP600 LHLOCA PIRT Results (continued) l System l Period l Component l Period l Process / Phenomenon l Period l lI l2 l3 l4 l lI l2 l3 14 l l1 l2 l3 l4 l IRWST I I I 11 Ambient heat loss I PRilRS operation I Pool heat transfer Pool water level Thermal statification/ mixing / multidimensionality Evaporation ADS operation I Condensation Thermal statification/ mixing / multidimensionality Evaporation Pool heat transfer Pressure oscillations O ECC operation Gravity-driven flow / Draining 11 Oscillations L Resupply from containment H Basic phenomena Piping I M L M Flashing in IRWST discharge hnes 1 Basic i e..uwe.& t Valves - associated with I 11 L 11 Basic phenomena specific components Sump 1 I I 11 ECC operation Gravity-driven flow / Draining 11 Oscillations L Resupply from containment II Ambient heat loss L Basic phenomena l l )

a Table 1. Summary of AP600 LBLOCA PIRT Results (continued) l System l Period l Compnnent l Period l Process /Phennmenon l Period l lI l2 l3 l4 l lI 12 13 l4 l El l2 l3 l] Primary Coolant System lIl lH IM lL Break 11 11 M I Critical flow / flow H H M Flashing L L L Basic phenomena Cold leg piping M L L I Flow asymmetries M Stored energy release L Flashing M Phase separation /stratifcation M Condensation L Basic phenomena flot leg piping L L M I Flow and related phenomenon M Entrainment/Deentrainment Countercurrent flow (CCF) (e.g., at sacam generator inlet plenum) ta Flow asymmetries L Stored energy release I Flashing (void generation) 1 Phase separation /stratifcation L Basic itc.a,..aia tw. W. M L I I Flashing (void generation) M Flow and related phenomena Draining / blowdown M Refill L Phase separaten/stratifcanon L Entrainment/Deentrainment L CCF L Stored energy release L Basic phenomena Surge line M L I I CCF L Basic phenomena Reactor coolant pumps 11 L L I Degraded performana H Coasadown L + Basic phenomena Steam Generator Primary L L M I Reverse heat transfer (e.g., steam M Side binding) Stored energy release I Basic rii..a.r.cre.

__....._ ___..._....~._.- Table 1. Summary of AP600 LDLOCA PIRT Results (continued) l System l Period l Component l Period l Process / Phenomenon l Period l ll l2 l3 ]4 l lI l2 l3 l4 l lI l2 l3 l4 l Reactor System lil lli Ill lli Vessel - conuoi rods L L L L Reactivity change Vessel-core flow chanacis Il II 11 11 Flow Reversal / stagnation M L L L Top Down / CCF M M M L Multidimensionality M M 11 L Oscillations L L 11 M level L L 11 L Interfacial heat and mass transfer 11 11 11 L Inscrfacial drag M M 11 L Basic iteri.c.e Vessel - downcomer M 11 11 M Flow g Intcrfacial drag / Bypass M 11 L L CCF L M L L Multidimensionality M II L L ECC mixing L L L L Head / level / oscillations L M 11 M Flashing M L L L Stored energy release - hot wall L M M L Condensation L M L L Noncondensible gas effects L L L L Basic i44..c w

Table I. Summary of AP600 LBLOCA PIRT Results (continued) l System l Period l Component __l Period l Process / Phenomenon l Period l lI l2 l3 l4 l lI l2 l3 l4 l lI [2 l3 l4 l Vessel-fuel rtxts 11 Il 11 11 Rod structure heat transfer Conduction M M M M Gap conductance 11 M M M Stored energy relear II L L L Rod convective heat transfer l'orted convection to liquid L L L L Nucleate boiling L L M L Transition boiling 11 11 l{ L Film boiling 11 11 11 L Forced convection to vapor L 11 L L j Reactivity [ S Void 11 L L L l Moderator temperature L L L L 8 Fuel temperature (Doppler) L L L L Baron L L L L Decay heat M 11 11 }{ Rasic phenomena Vessel-guide tubes M i i I Flow Draining Flow regime Flow resistance Basic phenomena Vessel-lower plenum M M L L Flow transient Emptying /sweepout M L Refill L M Mixing L M Multidimensionality L L Oscillations L L Stored energy release L L Basic phenomena Vessel - structures L L L L Stored encrey releae Vessel-:eth: -a-y L L L L Flashing I Basic phenomena

s s i i Table 1. Summary of AP600 LBLOCA PIRT Results (continued) f i l System l Period l Component l Period l Process / Phenomenon l Period l [ lI l2 l3 l4 l lI l2 l3 l4 l lI l2 l3 l4 l Vessel-upper head M I I I Stosed energy release Draining Guide tubes M Dovmcomer L Drain holes M I Flashing M Basic -iau,ss i Vessel-upper plenum M M M L Flow transient CCF M M M i Entramment/deentramment M L M ? Phase separation / stratification L L L Multidimensional M L M Flow distributen (hot legs, core) M L M e Flashing (void generation) L L L Stored energy release L L L ~ Basic phenomena l Steam Generator System L I I L (Secondary Side) Main feedwater line Basic i a-.a6 4 Main steam line SG blowdown line Stanup feedwaterline Other piping Valves Basic phenomena MS Isolaten i MS Safety [ MS PORV MFW Isolati<m i I F L I i

approached a fully voided state during the blowdown phase. Within the primary coolant system, the hot leg components completely void midway in the period and their impact is LOW. He impact of the steam-generator primary side component continues to be LOW. The steam-generator-secondary pressures have decreased below the primary pressure, but the hot legs voiding limits the secondary to-primary heat transfer. De cold-leg components completely void early in the period and their impact is LOW as the flow delivered to the downcomer is small at the start of the period and essentially zero at the end of the period. The pressurizer component injects its remaining small fraction ofinitial inventory through the surge line early in the period, and the impact of these two components is LOW. The performance of the RCP components degraded during the blowdown period. Because the pumps coast down under highly voided conditions and they deliver little flow to the downcomer, their impact is LOW. The reactor system continues to have a controlling (#/GH) influence on the course of the transient during this period. The fuel rod component and the core flow s; channel component continue to play a dominant (#/GH) and coupled role in the course j of the transient. As essentially all the stored energy is removed during the blowdown period, decay heat is the dommant (HIGH) energy source. Fuel rod convective heat transfer occurs at the interface between the fuel rod component and the fluid in the core flow channel. As was the case for the blowdown period, the refill period has two subintervals. The first is a continuation of top-down flow and associated cladding cooling i at the top of the core. Once the upper plenum empties and this top-down flow terminates, s; the second subinterval is characterized by a completely vapor filled core and associated resumption of cladding heatup. The key (H/GH) convecuve heat transfer phenomenon occurring during the first subinterval is post CHF heat transfer (transition and film boiling) to the fuel rod arising from the top-down flow of liquid from the u pper plenum into the upper ponions of the core. The key (HIGH) fuel-rod convective heat transfer process during the second subinterval is convection to vapor. During the refill period,'the processes / phenomena occurring in the downcomer component have a controlling (H/GH) influence on the course of the transient. Emergency coolant enters the downcomer via the DVIline. During the first subinterval, core-generated steam passes upward through the downcomer. The injected coolant is entrained and convected via interfacial drag (H/GH) to the broken loop. The process by which the emergency coolant bypasses the core and passes out the vessel side break is multidimensional (HIGH). The second subinterval is initiated after the upper plenum empties. Lacking any coolant flow into the core, steam generation ceases, the upward flow of steam through the downcomer rapidly decreases, emergency coolant is no longer entrained and convected to the break, and the coolant injected through the DVI line into the downcomer flows downward to the lower plenum. Several of the components of the vessel have a moderate (MEDIUM) influence on the course of the transient during the refill period. The processes / phenomena occurring in the vessel upper and lower plenum components influence the course of the transient. The processeshhenomena occurring in the upper plenum and their impact on the course of the transient c uring this period have been discussed previously. The lower plenum provides a conduit for the ECC fluid to reach the core, and it must be refilled following the blowdown period to replace the liquid inventory lost during that period and the first subinterval of the refill period. Refilling of the lower plenum marks the end of the refill period. Downcomer refilling proceeds at a slower rate; the downcomer liquid fraction is approximately 60% at the end of the period. t ' The impact of the control rod component during this period is LOW as the core is shut down and in an unchanging neutronic state. With no flow through the guide tube J 62

l - - -.~. -. _. - i 4 1 4 component, the impact is INS /GN/F/ CANT. The upper head component is nearly j voided at the stan of the period, and its impact on course of the transient is negligible (INSIGNIFICANT). The impact of the renector/ bypass component is LOW. The impact of the vessel structural component is LOW because the release of stored heat to the primary is low in the nearly or completely voided parts of the vessel. 3 The PSIS transitions from no impact during the blowdown period to a controlling } impact on the course of the transient during the reflood phase. At the end of the refill j_ period, the PSIS has operated to refill the lower plenum and create the conditions for i refilling and cooling the core. Therefore, the importance of this system during the refill period is HIGH. The accumulator component is the primary source of ECC during 4 i this period and, therefore, its importance is HIGH as is the imponance of the associated valves. As a conduit for the PSIS flows, the piping component is of moderate (MEDIUM) imponance. There is a small CMT component injection into the DVI lines early in the period, but this coolant bypasses the core and is lost through the break. i Therefore, the impact of this component on the course of the accident is LOW. The IRWST component and sump component are inactive and thus INSIGNIFICANT. i Several other systems are ranked as having an insignificant impact l (INSIGN/F/ CANT) during this period. The ADS system is inactive. The PCCS has an 1 insignificant impact on the primary system during this period as critical flow conditions continue at the two break planes and therefore there is no feedback from the containment to } the primary. There is no flow through the PRHRS and thus no transport of primary { system energy to the IRWST. The steam generator system (secondary side) i feedwater and main steam lines are isolated. The primary pressure falls below the secondary pressure, but the steam generator primary is voided and thus the secondary to i j primary heat transfer is negligible. / A tabulation of the system, component, and phenomena rankings for this phase of l l the LBLOCA is presented in Table 1. j 5.1.3 Renood Period 4 The reflood period covers the interval from lower plenum refill to the time the entire core is quenched. Although the entire core is quenched at the end of this period, boiling j dominates the core heat transfer processes and the core is not yet refilled. i The reactor system continues to have a controlling (HIGH) influence on the j course of the transient during this period. The fuel rod component and the core flow i channel component continue to play a dominant (HIGH) and coupled role in the course i of the transient. Decay heat generanon is the energy source for the transient and is, j-therefore, a dominant (HIGH) process. At the end of the previous period, the ECC refilled the lower plenum. During the reflood period, the ECC advances into, cools, and 1 quenches the core. In concert with downcomer liquid level and processes that produce i pressure changes in the upper plenum, the core liquid level plays an important role (HIGH) j m establishing the core reflood rate. For the larger pan of the refill transient, the processes j in the core, downcomer, and flow path to the break through the upper plenum compete, j causing the advancing core liquid level to oscillate. The oscillatory mteraction between the core and downcomer liquid levels impacts the core cooling processes (HIGH). Ponions of i the fuel rod below the quench front are in nucleate boiling, and the cladding temperature is slightly above the coolant saturation temperature. Above the quench front, the PCT a t continues to increase during the early ponion of the reflood period. The post-CHF heat j transfer above the quench front consists of transition and film boiling (H/GH). Large )- amounts of vapor are generated below and in the transition boiling region above the quench i. i 63 i. l i 1-.

,1 front. While movin various forms, e.g., g upward, this vapor entrains liquid via interfacial drag (HIGH) in slugs and droplets. The post CHF heat transfer between the two-phase coolant and fuel rod cladding includes both wall to liquid and wall to vapor process. In addition, interfacial heat and mass transfer processes above the quench front are imponant (H/GH) and contribute to precursor cooling. Eventually, sufficient cooling occurs in ponions of the core undergoing film cooling, and the PCT begins to decrease. Several multidimensional processes occur during reflood, which have been observed in large scale facilitics such as the Slab Core Test Facility, Cylindrical Core Test 4 Facility, and Primarkreislaufe Facility. Multidimensional effects in the cme enhance heat transfer by supplying more water to the high-power bundles below the quench front and more entrained water in the liquid deficient region. The multidimensional character of the reflood process is judged to be important (#1GH). The downcomer component plays an important (H/GH) role in the core reflood process. Imponant phenomena (#lGH) are related to the liquid level in the downcomer and related manometer-type oscillations that are coupled to core heat transfer processes because these affect the reflood rate and the PCT attained during the LBLOCA. The vessel upper plenum component plays a significant but moderate role (MEDIUM) during the refill period. First, the resistance between the core and break is one J component in determining the reflood rate. During refill, the path to the vessel-side break is blocked by coolant while the path to the pump-side break is open. The path pressure drop is due mainly to wall friction in the steam-generator tubes and in the pumps. There is also an acceleration pressure drop or steam binding (MEDIUM) in the steam generator component (MEDIUM) due to evaporating drops and superheating the vapor. Entrainment/deentraiment processes (MEDIUM) in the upper plenum and the hot leg component (MEDIUM) play imponant roles in the production of steam in the steam generator component. A second process of MEDIUM importance in the upper plenum component is countercurrent flow. Liquid is entrained in the core and carried into the upper plenum. A portion is entrained and carried through the hot legs to the steam generators as 4 previously described. A ponion of the liquid remaining in the upper plenum flows into the core over the lower powered assemblies near the outer periphery of the core. The associated multidimensional and countercurrent flows are of moderate (MEDIUM) unponance. The vessel lower plenum component has little (LOW) influence because its basic role is to provide a conduit (basic phenomena) for the ECC fluid that refloods the core. The remaining components (vessel upper head, guide tubes, structures, and reflector bypass) in the reactor system continue to have either no impact or a insignificant impact on the course of the transient (see the discussion for the Refill Period). The PSIS dominates the course of the accident by supplying the coolant that cools and quenches the core. Therefore, the importance of this system is HIGH. The accumulator component is the primary source of ECC during this period and, therefore, its importance is HIGH. The associated valve and piping components are rated as LOW because they are in continuous operation following their actuation during the initial period. The CMT component does not inject coolant into the primary until the accumulators empty. Thus the impact of this component on the course of the accident is INS /GNIFICANT. The IRWST component and sump component are inactive and thus are INSIGNIFICANT. M _s

i i 4 i The primary cooling system is of moderate importance (MEDIUM) during the refill period. The flow, entrainment/deentrainment and steam binding processes / phenomena that occur in the upper plenum, hot legs, and steam generators influence the rate at which steam exits the pnmary coolant system through the break as previously described. Entrained liquid and associated processes, e.g., steam binding, increase the flow resistance and elevate the pressure in the upper plenum. In response to the elevated pressure in the upper plenum, the core reflood rate, which is driven by the hydrostatic head in the downcomer, decreases. I Many systems are ranked as having either no impact or an insignificant impact (INSIGNIFICANT) during this period. The ADS system is inactive. The PCCS has an i insignificant impact on the pnmary system, but the break does unchoke during this period.b There is no flow through the PRHRS and therefore no transport of primary i system energy to the IRWST. The steam generator system (secondary side) feedwater and main steam lines are isolated. The primary pressure falls below the secondary pressure but the steam generator primary is voided; therefore, the secondary to primary heat transfer is negligible except when entrained liquid en! cts the steam generator j through the hot legs as previously described. l' A tabulation of the system, component, and phenomena rankings for this phase of the LBLOCA is presented in Table 1. 2 f 5.1.4 Long Term Cooling Period The long-term cooling period covers the interval from the time the entire core is quenched until a stable long-term cooling mode consisting of either IRWST or sump injection is established. To reach this final state, the CMTs inject their inventory, ADS stages 1 through 4 actuate, and the primary and containment pressures equilibrate sufficiently so that IRWST head-driven injection can initiate. 'Ihe long-term cooling period encompasses the final two periods employed by the INEL for the recent AP600 small break loss-of-coolant accident PIRT 12 The rankings from Ref.12 are considered for applicability to the LBLOCA, revised as appropriate, and are provided here in a format consistent with the remainder of this section.c Once a quasi steady state has been established following CMT and accumulator injection, multiple systems interact to provide long-term cooling. Liquid coolant enters the primary system via head-driven flows from the IRWST and/or sump. The coolant proceeds to the core where boiling generates steam that passes through the upper plenum to the pressurizer hot leg where it is released to the containment. Within the PCCS (steel i b The TRAC analysis used by the PIRT Ranking Team assumed a constant containment back pressure. Therefore,importance rankings for the PCCS should be revisited as analyses which couple the primary system and containment system performance become available. 1 c The key measurement standard used %.he PIRT Ranking Team in assessing imponance was wheder the system, component, and process / phenomena had a significant impact on the course of the transient, specifically if it had a controlling impact on the ultimate PCT. For the LBLOCA scenario of this report, the ultimate PCT occurs during the reflood phase. Logically, therefore, no system, component, and process / phenomena during the long-term cooling phase should be ranked as having HIGH importance. The PIRT Ranking Team took a different approach by considering what systems and components have a significant impact relative to maintaining core cooling. Thus, a different key measurement standard has been used during the long-term cooling period. 65

reactor containment structure, concrete shield building, and spray tank subsystems) processes and phenomena occur by which the steam released to the steel reactor containment structure through the 4th-stage ADS is condensed and returned to either the IRWSTor sump via gutters. j The reactor system continues to have a controlling (#1GH) influence on the j course of the transient during this period. The fuel rod component and the core flow channel component continue to play a dominant (HIGH) and coupled role in the course of the transient. Decay heat is the energy source for the transient and is, therefore, a 4 dominant (#1GH) phenomenon. Fuel roc convective heat transfer occurs at the interface between the fuel rod component and the core flow channel. *Ihe fuel-rod convective heat l transfer is by nucleate boiling,.which keeps the cladding temperature close to the coolant saturation temperature. Because the core is already quenched, phenomena in the l downcomer component are of diminished importance (MEDIUM) when evaluated a relative to PCT Core-downcomer manometer oscillations continue until the core is refilled, but the impact on the course of the final period of the LBLOCA is moderate (MEDIUM). Phenomena in the lower plenum component are also of diminished l. imponance (LOW). The PSIS dominates the course of the accident by supplying the coolant that i ensures that the core remains cooled and quenched. Therefore, the imponance of this l system is HIGH. The CMT component mjects its coolant into the pnmary while the final primary pressure reduction occurs. Draining (HIGH) of the CMTs is necessary to open the founh stage ADS providing a direct path for vapor to pass into the containment, panicularly after the cold-leg break is covered by liquid. Thus the impact of this component and its associated valves on the course of the accident are HIGH. The l IRWST component and sump component activate after the CMTs drain and provide the inventory for long-term cooling; therefore, these components and the accompanying valves are of HIGH imponance. Resupply of coolant from the containment is also { imponant (H/GH). The accumulator component either emptied during the reflood period or early in the long-term cooling period; therefore, its imponance is LOW. . The PCCS is an imponant system (HIGH) during the long-term cooling phase. i The Steel Reactor Containment Structure, concrete shield building, PCCS water tank, and associated valves are all imponant (HIGH) as together they provide the i path through which decay heat is rejected to the atmosphere. After the founh stage ADS F has actuated and the CMTs have completed draining, the primary pressure decreases to the a level at which the IRWST pool head is sufficient to permit injection of the IRWST j inventory into the primary. Subsequently, the IRWST will drain and the sump will e activate. For long-term cooling, an adequate supply of cooling water must be returned to the IRWST and/or sump. Primary coolant is heated in the core and released as vapor to the j containment, initially through the break, but most predominantly through the fourth stage ADS after it opens. The vapor condenses on the inner surface of the steel reactor i i containment structure (HIGH) and the condensed phase (liquid) then drains downward j (MEDIUM) where it is collected by gutters and returned to the IRWST. The latent heat of l the vapor as well as the heat released upon condensation must be rejected to the ultimate ' heat source. Otherwise, a breach of euontainment could occur because of pressurization i (MEDIUM). Vaporized coolant would be lost through the containment breach and a subsequent core uncovery and cladding heatup would occur were makeup water not provided. Thus, the heat transfer caused by the air flow in the inner space between the i concrete shield building and the steel reactor containment structure, and the evaporation of water sprayed on the outside of the steel reactor containment structure, are important j (HIGH). The movement of spray water on the outside of the containment (draining) is of 1 4 + - - - .~v

4 i i ~ l troderate imponance (MEDIUM) because the draining process determmes the coolant flow rate and surface coverage. 1 Several systems continue to have either a LOW orINSIGNIF/ CANT impact during this penod (see discussion for the reflood period). A tabulation of the system, component, and phenomena rankings for this phare of the LBLOCA is presented in Table 1. 5.2 _ Comparison to PIRT for Westinghouse Four-Loop Plant The AP600 LBLOCA PIRT piesented in this document and the earlier LBLOCA PIRT for the Westinghouse four-loop plant 3 conducted as pan of the CSAU demonstration cffon have areas of both similarity and dissimilarity. Most of the differences are clearly related to the design features of the two plants. For example, the AP600 has a separate piping system (DVI lines) through which emergency core coolant is injected directly into the vessel downcomer. In the Westinghouse four-loop plant, the accumulator inventory is injected into the cold legs. Given such design differences, some of the processes and i phenomena also differ. As a consequence of some differences in the system, component and process / phenomena importance arises. In this section, the design differences are identified and the manner in which the design differences affect the relative importance of systems, components, and processes / phenomena in the two designs is discussed. Additional design differences exist for which it has not been possible to identify specific impacts on the processes and phenomena occurring in the two designs. These design i differences will also be listed. Other differences appear to arise from two sources. First, the process followed for the AP600 LBLOCA PIRT effort was somewhat different. For example, the AP600 PIRT team followed a hierarchical ranking approach within each accident period. Following this approach the importance of AP600 systems was first ranked. The importance of components withm each system was then ranked. Finally, the processes / phenomena occurring within the components during the period were ranked. If a system or component had either an insignificant or no impact on the course of the sequence, processes / phenomena were not ranked for that component during that period. This approach, with its initial focus on systems and components, undoubtedly led to a different perspective by the AP600 PIRT team. A related outcome of the differing perspectives was j the occasional absence of a one-to-one correspondence between the process / phenomena l lists for the two PIRTs. For example, the earlier PIRT for the Westinghouse four loop plant lumped all heat transfer phenomena occurring during the refill period into the descriptor "reflood heat transfer." The AP600 PIRT team did not include reflood heat transfer as a process / phenomena; rather, the team retained the individual reflood heat i transfer processes, specifically forced convection to a liquid, nucleate boiling, transition boiling, film boiling, and, forced convection to a vapor. Phenomena at the interfaces between these heat transfer regimes were not explicitly identified. Examples are the CHF condition at departure from subcooled nucleate boiling (DNB), saturated DNB, and dryout. Second, the PIRT team members were intentionally not exposed to the earlier PIRT results for the Westinghouse four-loop plant and, therefore, ranked the phenomena based upon their background and experience. Thus, differences of opinion regarding the relative importance of phenomena can be expected. The author did review the Westinghouse four-loop PIRT results with the objective of resolving differences, where possible. All such changes were reviewed with the members of the PIRT team for acceptability. 67

{ i A comparison of the present AP600 PIRT results and those from the PIRT for the Westinghouse four-loop plant 3 is given in Table 2. The table is a modified listing of the a summary rankings (Table 1) from Ref. 3. With few exceptions, only those phenomena from the AP600 LBLOCA PIRT (Table 1) appearing in the Westinghouse four loop plant ,l PIRT are listed in Table 2. Further, only the AHP ranking is presented for the earlier PIRT as more phenomena were covered by the AHP process. A numerical ranking with 4 values between 1 (low imponance) and 9 (high importance) was used in the AHP.d For this comparison, the AHP PIRT results for the Westinghouse four-loop plant have been recast so that numerical values of 1-3 are ranked as LOW, values of 4-6 are ranked as MEDIUM, and values of 7-9 are ranked as HIGH. Differences between two adjacent j ranks are not deemed significant. However, differences of two ranks (i.e., HIGH and j IDW ranks) are deemed significant and are discussed in the following paragraphs. 4 l The largest number of rank differences arise from differing design features, and the associated differences in processes and phenomena and their imponance. These are discussed below. The earlier Westinghouse four-loop PIRT ranked steam binding as an important (HIGH) phenomenon. With this perspective, associated phenomena in other j parts of the primary system were also perceived as being important.

Thus, entrainment/deentrainment phenomena in the upper plenum and hot legs were also ranked as imponant (HIGH) because they controlled the amount of liquid delivered to the steam l

generator. The AP600 LBLOCA PIRT team concluded that steam binding, while still 3 imponant, is of MEDIUM imponance. This was due, in part, to the small amounts of j liquid that were predicted to be entrained and carried to the steam generators from the upper plenum. Because there is a significant degree of uncertainty in entrainment/deentrainment ,f modeling, this conclusion should be revisited should additional information become available. 3 i Another design difference of imponance is related to the delivery of emergency cor: coolant to the core. In the earlier Westinghouse four loop design, the accumulators i delivered their inventory directly to the cold legs which, in turn, delivered the coolant to the 4 vessel downcomer. In the AP600, the emergency core coolant is delivered to DVI lines, which are completely separated from the primary loops. Because the ECCS is a key i system (HIGH), ranking differences at both the component and process / phenomena level arise. For example, in the earlier PIRT, condensation caused by the accumulator injection into a partially voided cold leg is imponant (HIGH). However, the AP600 cold leg component is of only moderate importance (MEDIUM) during the blowdown phase, and of small or insignificant importance (LOW or INSIGNIFICANT) during the remaining phases of the LBLOCA. In the AP600, condensation phenomena arising from ECCS mjection are shifted to the downcomer component. One consequence of the AP600 loop and ECCS design is that loop processes and phenomena are less important than in the earlier Westinghouse four-loop plant, and this difference shows in the imponance ranking of phenomena in the loops. Thus, phenomena arising during the reflood phase in the hot legs are assessed as important (HIGH) in the Westinghouse four-loop design but unimponant (MEDIUM ) m the AP600 design. d The following perspective is providing regarding consideration of highly-ranked phenomena in the CSAU process, specifically as implemented in the original CSAU demonstration.3 Only those phenomena having a rank of 9 were carried forward into the uncenainty quantification. In the current study, processes / phenomena ranked HIGH would have a comparable AHP ranking of 7. 8, or 9. Thus, further refinement of the AP600 LBLOCA PIRT rankings for processes / phenomena ranked HIGH may be necessary should uncenainty quantification proceed. 68

~ Table 2 Comparative Tabulation of AP600 and Westinghouse Four Loop PIRT Results Blowdown Refill Reflood System / Component AP600 W4L AP600 W4L AP600 W4L See notes at end of tablefor H. M L, NR depnitions Stored Energy H H L L L L Oxidation NR NR L NR H Decay heat M L H L H H Gas (gap) conductance H L M L M M Core DNB M* M L* L L* L Post-CHF H M H H H M Rewet M* H M* M M* L Reflood heat transfer H* H Nucleate boiling L M L L M L 1-phase vapor nat cire NR M NR M 3-D flow M L M L H H Void generation / distribution H* M H* M H* H Entrainment/deentrainment M* L M* L H* M Flow reversal / stagnation M L L L L L Radiation heat transfer NR L Upper head Guide tube flow M NR L NR L NR Drain hole flow M NR L NR L NR Upper plenum Entrainment/deentrainment M L L L M H Phase separation L L L L L L CCF drain / feedback M L M L M M 2-phase convection NR L NR L NR M 3-D flow M NR L NR M NR Hot leg Entramment/deentramment L L L L M H Flow reversal L L L L L* Void distribution L L L L M* M 2-phase convection L L L L M* L Pressurizer Early quench NR H Critical flow in surge line L* H Flashing M H L L L L Steam generator Steam binding L L L M H AP, form losses L L L L L L 69

~t Table 2 (continued) Comparative Tabulation of AP600 and Westinghouse Four Loop PIRT Results 4 Blowdown Refill Reflood System / Component AP600 W4L AP600 W4L AP600 W4L l Pump 2-phase performance H H L M L AP, form losses L* L L* L M* H y Cold leg / accumulator i Condensation L L L H L M noncondensible gases NA NA L NA H t HPI mixing NA NA L NA L Discharge (accumulator) NR H NR H NR 3 Downcomer Entminment/deentramment M* L H* H L* L Condensation L M H L L Countercurrent, slug noeg L L H H L L t' Hot wall L M M M L s 2-phase convection-bypass M L M L L L Saturated nucleate boiling NR L NR L NR L g 3-D effect M L H H L L Flashing M L L L Liquid level oscillations L M L H H Lower plenum Sweep out M L L M L M Hot wall L L L H L M Multi D effects L L L L. L H Break Critical flow / flow H H H H M L Flashing L L L L L L Containment pressure L L L M M L L*P 2-phase AP M* H M* H M* M Oscillations NR H NR H Flow split (asymmetries) M H L H L L Nas Not explicitly ranked. The ranking presented is inferred from related process / phenomenon ranks. 1. AHP ranks 7 through 9 of the original CSAU study 3 were translated to "H/GH" or "H" in the ranking rationale for this study. 2. AHP ranks 4 through 6 of the original CSAU study 3 were translated to "MED/UM" or "M" in the ranking rationale for this study. 3. AHP ranks 1 through 3 of the original CSAU study 3 were translated to " LOW" or "L" in the ranking rationale for this study. 4. Several processes or phenomena in the original CSAU study 3 were "NOT RANKED" or "NR" for this study. Other processes are "NOT APPLICABLE" or "NA" in AP600. A process or phenomena that was identified in the original CSAU study but not ranked is denoted with a " ". The same mark is used if it was unranked in the AP600 PIRT. .i 70 J

-~. I b A third design change affecting imponance assessments relates to the lower plenum design. A vonex suppression plate has been incorporated in the AP600 design to specifically [ reduce multidimensional behavior. Tests have demonstrated the effectiveness of the plate. Thus, multidimensional effects in the lower plenum during the reflood period are assessed 1 as being important (H/GH) in the earlier Westinghouse four-loop design but unimportant ) i (LOW)in the AP600 design. 1 \\ Additional design differences exist in the region of the upper head and upper j_ plenum. Drain holes are provided in the AP600's upper suppon p, ate so that the upper head will not function like a pressurizer. Upper head coolant liquid inventory participates 1 in top-down cooling, which is ranked of MEDIUM importance during the blowdown I period. Both designs had guide tubes but the AP600's guide tubes deliver more coolant from the upper head directly to the top of the core than the guide tubes in the earlier Westinghouse design. The guide tubes are ranked of MEDIUM importance during the l blowdown period. In addition, flow fmm the upper head to the upper plenum through the drain holes is approximately twice the flow delivered through the guide tubes. De original i i PIRT did not deal with either components or processes in the same detail. Rather, the j focus was on the upper plenum and the processes by which coolant was delivered to the l Core. There are several additional design differences for which it has not been possible to j assess the differential impact on processes and phenomena in the two designs. The core j power density is lower in the AP600 than in the Westinghouse four-loop design. However, the relative impact of the lower power density has not been assessed. With j respect to the LBLOCA PCT, the combination of power density, core power distribution, l heat transfer regimes, and flow rate and distribution are all imponant and strongly coupled. i i With the information available, it was not possible for the PIRT Ranking Team to isolate l and identify the effect of core power density alone on the LBLOCA PCT. Efforts will j continue to assess the impact of these differences. The core loadings are different in the 3 two designs. The impact of the differences has not been assessed. Efforts will continue to i assess the impact of these differences. The RCPs in the two designs are different. For l cach design, phenomena occurring in the pumps during the blowdown priod were l assessed as having an imponant (HIGH) impact on PCT. During the blowcown period, the process of interest is pump performance degradation caused by voiding in the pump inlet. Pump performance is highly sensitive to the presence of void, and pump l performance degrades markedly in the presence of void. Sensitivity studies would be i panicularly helpful in displaying the sensitivity of the PCT to uncenainties in the pump performance characterization. We do note that the imponance of the RCPs is ranked as HIGH for the Westinghouse four-loop and LOW for the AP600 during the reflood period. This difference is associated with other design differences, specifically the l configuration of the ECC delivery piping as discussed pn:viously. l The remaining differences in ranking relate more directly to differing assessments j of importance by the ranking teams and to definitions used by the ranking teams. The L assignment of priority to decay heat is an example of a difference in assessing imponance. ne AP600 PIRT team assigned a HIGH ranking to decay heat in the fuel rods for the refill j and reflood periods. Decay heat was ranked LOW during the refill period for the Westinghouse four-loop plant. The AP600 PIRT team concluded that because all stored energy is removed from the fuel rods during blowdown, decay heat must be ranked as high because it is the dominant energy source affecting both cladding heatup and the core i heat transfer environment. Some of the significant ranking differences relating to core heat transfer processes appear to be related to definitions. For the Westinghouse four loop a plant, a phenomenon called "reflood heat transfer" was defined. The AP600 PIRT team l did not use this term. Rather, reflood heat transfer processes were made explicit by listing e i s 71 1 4 I i

1 u ql and ranking the fuel-rod convective heat transfer processes. Although some of the specific ) heat transfer processes during reflood were more imponant than others, the perspective that. ni reflood heat transfer is important (H/GH) is consistent. The most imponant outcome of the present assessment is that no new phenomena have been identified for the AP600 during the blowdown, refill, and reflood phases that l were not present in the PIRT for the earlier Westinghouse four-loop plant. i The passive safety features such as the core makeup tanks, In-containment Refueling Water Storage Tanks, and Passive Containment Cooling System do not play a l significant role in the outcome of the transient until the core is entirely quenched. During the blowdown, refill, and reflood periods of a LBLOCA, the two plants have many more i areas of similarity in systems, components, and processes / phenomena than areas of i dissimilarity. l 'The PIRT displays the imponance of the plant systems, components, and processes and phenomena during the LBLOCA scenario. Within the CSAU method, the PIRT is then i used as pan of the code assessment activity to ensure that the code's system and component models, fundamental models, and carrelations are applicable and adequate. A high degree of applicability and adequacy is required for those systems, components, and processes i and phenomena that most strongly impact the course of the transient as identified in the '3 { PIRT. Lesser standards apply for those components that have little impact on the course of i the transient. For the blowdown, refill, and reflood periods of a LBLOCA, there are no significant differences between the earlier PIRT for the Westinghouse four-loop plant and l the AP600 PIRT. There are no new phenomena modeling requirements that would be j required of the accident analysis code. The plant differences may enter into code 4 j uncertainty quantification of the LBLOCA PCI'in a significant way. j 6.0

SUMMARY

3 A Phenomena Identification and Ranking Tabulation, or PIRT, has been completed 'I for the AP600 response to an LBLOCA. Several significant supporting activities were - undertaken to lay the foundation for preparing the AP600 LBLOCA PIRT. A detailed i description of the AP600 LBLOCA scenario was prepared and documented. This ii j information appears in Section 4 of this repon. A detailed tabulation of AP600 systems, components, and processes / phenomena information was prepared. Considerable effon was expended in reviewing source documentation and summarizing this information. This information appears as Appendix A of this repon. Using the scenario description and the l information regarding AP600 systems, components, processes, phenomena, and the expenise of the PIRT team members, the AP600 LBLOCA PIRT was developed. This information appears in Section 5 of this repon. Also, a comparative analysis was F i completed to assess the similarities and differences between the LBLOCA PIRTs for the J AP600 and the Westinghouse four loop plant for which a PIRT was prepared as pan of the CSAU demonstration effon. i 1 Although there are differences between the designs, processes and phenomena - occurring in the two reactors are qualitatively similar during the blowdown, refill, and reflood periods of the LBLOCA. A comparative evaluation of the two PIRTs reveals many similarities. However, there are some differences in the imponance ranking of phenomena. Most of the differences between the AP600 and earlier Westinghouse four-loop plant i PIRTs can be attributed to the design differences. Perhaps the most significant design j . difference is associated with the ECC system configuration and operation. In the AP600, l-the two accumulators each inject their inventory into direct vessel injection lines that terminate in the vessel downcomer. The direct vesselinjection lines are separate from the 4 4 4 72 I

~ cold legs. In contrast, each of the four accumulators in the Westinghouse four-loop plant inject directly into the cold leg. Thus, the break directly affects ECCS injection m, the i broken loop, and ECCS injection also affects the phenomena in the intact cold legs. Neither of these circumstances pertain to the AP600. Additional design differences that contribute to differences in the two PIRTs are detailed in the body of this report. There are, however, a smaller number of differences that are not linked to design features. Several potential reasons for such differences can be identified. First, the AP600 t PIRT ranked systems and components by accident period, as well as the phenomena occurring within the systems or compon6nts during the period. This approach, with its initial focus on systems and components, toay have affected the perspecuves of the PIRT j team members. Second, the PIRT team members were intentionally not exposed to the earlier PIRT results for the Westinghouse four-loop plant and, therefore, ranked the phenorr.ena based upon their backgrour.d and experience. Thus, differences of opinion n garding the relative importance of phenomena can be expected. An effort has been made to rewhe these differences. The AP600 PIRT ranks the importance of the plant systems, components, and processes / phenomena during an 80% DEGB LBLOCA scenario. Within the CSAU method, the code to be used to analyze the transient is then reviewed to ensure that its component models are adequate. A high degree of adequacy is required for those systems and components that most strongly impact the course of the transient as identified in the PIRT. Lesser standards apply for those components that have little impact on the course of the transient. For the blowdown, refill, and reflood periods of a LBLOCA, there are no significant differences in the component modeling requirements that would be imposed on the accident analysis code compared with those identified in the CSAU PIRT. The passive safety features such as the core makeup tanks, In-containment Refueling Water Storage Tanks, and Passive Containment Cooling System do not play a significant role in the outcome of the transient until the core is entirely quenched. The most important outcome of the present assessment is that no new phenomena have been identified for the AP600 during the blowdown, refill, and reflood phases that were not present in the PIRT for the earlier Westinghouse four loop plant. 73

i REFERENCES 1. Advanced Light Water Reactor, Utility Requirements Document, Volume 1, ALWR 8 Policy and Summary of Top-Tier Requirements, Electric Power Research Institute (1990). 1 2. Code of Federal Regulations, Volume 10, Part 50.47(b)(2)(i)(A) l 3. B. Boyack, R. Duffey, P. Griffith, G. Lellouche, S. Levy, U. Rohatgi, G. l Wilson, W. Wulff, and N. Zuber, " Quantifying Reactor Safety Margins: Application of Code Scaling, Applicability, and Uncertainty Evaluation Methodology to a Large Break Loss-of-Coolant Accident," EG&G Idaho, Inc. l report NUREG/CR-5249, also EGG-2552 (October 1989). i 4. Safety Code Development Group, " TRAC-PF1/ MODI: An Advanced Best-i Estimate Computer Program for Pressurized Water Reactor Thermal-Hydraulic Analysis," Los Alamos National Laboratory report NUREG/CR-3858 (July 1986). l 5. "Best Estimate Calculations of Emergency Core Cooling System Performance," US Nuclear Regulatory Commission Regulatory Guide 1.157 (May 1989).

• l

.e \\ 6. J. W. Spore et al., " TRAC-PF1/ MOD 2 Code Manual - Theory Manual," Los Alamos National Laboratory report NUREG/CR-5673, also LA-12031-M (to be ,j issued). 7. AP600 S AR Chapter 6, Rev. 0 (Effective June 26,1992). 8. " Additional Information in Support of Westinghouse Response to RAI 952.44 - l 952.46," Attachment to Westinghouse Letter NTD 94-4114, Proprietary (June, 1994). l 9. J. F. Lime and B. E. Boyack, " Updated TRAC Analysis of an 80% Double-Ended Cold Leg Break for the AP600 Design," Los Alamos National Laboratory document LA-UR 95-1594 (1995). 10. S. I. Dederer, L. E. Hochreiter, W. R. Schwartz, D. L. Stucker, C. K. Tsai, and M. Y. Young, " Westinghouse Large Break Best Estimate Methodology, Volume 1 Model Description and Validation, Volume 2, Revision 2, Application to Two-Loop l PWRs Equipped with Upper Plenum Injection," Westinghouse, Inc. report WCAP-l 10924-P-A, Proprietary (December 1988). l l 11. J. M. Taylor, " Integral System Testing Requirements for Westinghouse's AP600 Plant," US NRC document SECY-92-030 (January 1992). 12. C. D. Fletcher, G. E. Wilson, C. B. Davis, and T. J. Boucher, " Interim Phenomena Identification and Ranking Tables for AP600 Small Break Loss-of-l Coolant Accident, Main Steam Line Break, and Steam Generator Tube Rupture Scenarios," Idaho National Engineering Laboratory report INEL 94/0061 (November 1994). 13. B. E. Boyack, " Letter Report for T/H Consultants Meeting of June 13-15, 1995," Los Alamos National Laboratory Group TSA-12 letter TSA-12-95 212 to F. Eltawila, USNRC (June 19,1995). 74 l

14. R. Shaw, T. Yonomoto, and Y. Kukita, " Quick Look Report for ROSA /AP600 Experiment AP-DV-01," DRAFT for REVEW (December 1991). 15. N. Aksan, F. D'Auria, H. Glaeser, R. Pochard, C. Richards, and A. Sj6 berg, " Overview on CSNI Separate Effects Tests Validation Matrix," Paper presented at the 5th CAMP Meeting, October 19-21,1994, Idaho Falls, Idaho. l l i O 75

r" d APPENDIX A AP600 SYSTEMS, COMPONENTS, PROCESSES, AND PHENOMENA 4 Information conceming AP600 systems, components, processes and phenomena are organized and tabulated in this Appendix. Numerous sources 10 were examined and t summarized. The tabulation was organized in a logical hierarchy, proceeding in order through AP600 systems, components, system or component functions, process / phenom-3 ena descriptions, and other related factors that should be considered when evaluating the operation or importance of a given system or component. In addition, pertinent PIRT-significant insights are entered as system, component, instrumentation, and PIRT comments. These insights relate to system and component operations, the processes and phenomena occurring in AP600, and other information germane to importance assessments. The information in Appendix A was made available to the PIRT team members in advance of the phenomena ranking sessions. This is a nonproprietary version of Apxndix A. Reference 10 is a Westinghouse l proprietary document, the code qualification c ocument, generally referred to as the CQD j for the Westinghouse safety analysis code WCOBRA/ TRAC. It contains extensive sensitivity studies performed with WCOBRAffRAC for three currently operating a Westinghouse-designed nuclear power plants: Indian Point 2, Sequoyah, and North Anna. Sensitivity studies were performed related to plant physical condition (e.g., dimensions, flow resistances, fuel conditions, etc.), reactor power parameters (e.g., bumup, moderator temperature coefficient, and power history), reactor fluid concutions (e. system pressure, accumulator volume), accident boundary conditions (e.g.,g., loop flow break location, coraainment pressure, and pump operation), and models (e.g., break, pump, ECC bypass). A nonproprietary version of Appendix A has been prepared but it is not issued with this nonproprietary document. A-1

Systems, Components, Processes, Phenomena Tabulation AP600 PIRT Support l System l Component l Function l Process / Phenomenon l Other Factors l Automatic Contmiled system Phenomena described at System Depressurization System depressurization component level inlet / outlet conditions and envimnmental conditions Valve stages 1,2,3,4 Contmiled flow release Critical flow Component (mass and energy) inlet / outlet and Basic phenomena 9 and Note I environmental conditions Layout and design features instrumentation 2 y Valve status (1-3) Measurement / status A3 Valve status (4) Measurement / status Piping Flow delivery Basic phenomena 9 and Note 1 Component inlet / outlet and envimnmental conditions Layout and design features Sparger Deliver and distribute Condensation Component ADV flow Pressure fluctuations inlet / outlet and Flow oscillations envimnmental conditions Basic phenomena 9 and Note 1 Layout and design features l E 1---.-

Syst2ms, Compon=ts, Processes, Phen::mena Tr.bulstion (continzd) AP600 PIRT Support l System l Component l Function l Process / Phenomenon l Other Factors l System comment:1 The purpose of the ADS is to depressurize the primary coolant system sufficiently so that borated water can drain by gravity from the IRWST into the primary cooiant system. This injection is only possible if the static head created by fluid available from water in the IRWST is greater than the differential pressure created by fluid escaping from the primary coolant system into the containment. He ADS is composed of four stages. Stages I,2, and 3 pass fluid from the top of the pressurizer through a network of valves and discharge lines to spargers where the fluid is injected below the IRWST pool surface. Stage 4 passes fluid from the hot leg directly into the containment atmosphere. The valves are actuated in sequence following indication of a low inventory signal in either of the core makeup tanks. 6 System comment: De ADS is not activated until well past the time of PCT in LBLOCA events. hhkeur/lztdown SysMm D System comment:1 The makeup / letdown system adds coolant to and removes it from the primary coolant system at virtually identical low rates during normal operation. An S signal results in isolation of the letdown flow, and start-up of a second centrifugal makeup pump; thus, the system function changes to injection only during most accident sequences. The makeup / letdown system is nonsafety grade and therefore it is not proper to take credit for its performance in a licensing basis calculation. Normal Residual IIcat Removal System (NRIIRS) System comment: His is a low-pressure pumped system. It is a nonsafety-class system. Westinghouse intends that the planned mode of systems operation is that the NRHRS will be actuated after the 3rd stage ADS has operated but before the 4th stage ADS operates. He pumped coolant will prevent the CMTlevel from dropping to 20% and prevent the operation of the 4th stage ADS. Because the NRIIRS is nonsafety grade, it is not proper to take credit for its performance in a licensing basis calculation.

1 Systems, Components, Processes, Phenomena Tabulation (continued) AP600 PIRT Support l System l Component l Function l Process / Phenomenon l Other Factors l Passive Containment Contain, collect, cool, Phenomena described at System Cooling System and distribute discharges component level inlet / outlet and from the primary cooling environmental system conditions { i Concrete shield building Induce and direct air Buoyancy-driven flow Component l Airinlet flow that transmits lleat transfer inlet / outlet and Air exhaust energy to the ultimate Conduction environmental i Air path heat sink Convection conditions Radiation l Mass transfer Layout and i design features Basic phenomena 9 and Note I i Component comment:1 Heat transfer on the outside of the containment steel shell is by evaporation of i liquid deposited near the top of the steel reactor containment dome by the PCCS, and by convection to an E air stream induced by buoyancy-driven flow (unforced). 'Ihis air steam enters a high-clevation inlet, i flows downward to an elevation near the bottom of the cylindrical portion of the steel reactor containment i structure, passes upward through the annular gap between the steel reactor containment structure and the I concrete shield building, and is exhausted to the atmosphere near the top of the concrete shield building. l Instrurnentation2 l Containment i Pressure Measurement / variable Radiation Measurement / variable Containment spray tank level Measurement / variable I l Containment spray l tank valve status Measurement / status PCC water tank flow Measurement / variable PCC water tank level Measurement / variable PCC water tank valve status Measurement / status 1 L

Systems, Compon2nts, Processes, Phen men 3 TchulItion (continnd) AP600 PIRT Support l System l Component l FInction l Process / Phenomenon l Other Factors l instrumentation comment:6 An S signal occurs due to high containment pressure at about I s following the LBLOCA initiator. Piping Flow delivery Basic phenomena 9 and Note 1 Component inlet / outlet and environmental conditions Layout and design features i w

Systems, Components, Processes, Phenomena Tabulation (continued) AP600 PIRT Support l System l Component l Function l Process / Phenomenon l Other Factors ] Steel reactor Transferenergy from the Inside Component Containment structure containment atmosphere, Buoyancy-driven flow inlet / outlet and across the steel contain-Pressurization 6 environmental ment shell, to the outside Ileat transfer conditions air flow Conduction Convection Layout and Return ECC to the Radiation design features primary system Mass transfer Evaporation (spray) Limit radioactive ' Condensation (wall) releases to the Draining environment Noncondensible gas Thru wall conduction Outside Heat transfer Conduction Convection Radiation Mass transfer-evaporation Draining Noncondensible gas Basic phenomena 9 and Note 1 Component comment:1 The safety-grade containment heat removal mechanism involves transferring heat from the containment atmosphere, across the steel containment shell, to the atmospheric heat sink. On the inside surface of the containment structure, heat transfer is by condensation of steam. Condensate flows downward to gutters, which direct the condensate to the IRWST. Movement of the containment atmosphere is by buoyancy driven flow. Noncondensible gas has the potential to degrade condensation heat transfer on the inside structure surface. PIRT comment:3 A higher containment pressure was found to significantly improve the vessel mass inventory and subsequent reflood transient. i l r

Systems, Components, Processes, Phenomena Tabulation (continued) AP600 PIRT Support l System l Component l Function _l Process / Phenomenon l Other Factors l Valves Controlled flow release Basic phenomena 9 and Note 1 Component Spray tank inlet / outlet and Water tank environmental conditions Layout and design features Spray tank Discharge fluid Basic phenomena 9 and Note I Component (Containment) inventory inlet / outlet and environmental conditions Layout and design features >b Water tank (PCCS) Discharge fluid Basic phenomena 9 and Note I Component inventory inlet / outlet and envimnmental conditions Layout and design features Passive Residual Heat Energy removal from Phenomena described at System Removal System the primary system component level inlet / outlet (PRIIRS) conditions and environmental conditions

Systems, Components, Processes, Phenomena Tabulation (continued) AP600 PIRT Support l System l Component l Function l Process / Phenomenon l Other Factors l Ileatexchanger Primary to secondary Primary side Component (PRilRilX) heat exchange Buoyancy-driven flow inlet / outlet and liigh-point trapping environmental lleat transfer conditions Noncondensible gas Thm-wall heat transfer Layoutand Secondary side design features Buoyancy-driven flow Ileat transfer to IRWST pool Basic phenomena 9 and Note 1 Instrumentation 2 IIX flow Measurement / variable Flow IlX outlet T Measurement / variable Temperature Valve status Measurement / status Open/ shut oo Piping Flow delivery Phase separation at connections Component to hot leg inlet / outlet and environmental Basic phenomena 9 and Note I conditions Layout and design features Valves Controlled flow release Basic phenomena 9 and Nose 1 Component inlet / outlet and environmental conditions Layout and design features A .~

i Syshms, Ccmpon:nts, Processes, Phen mx:n Tchulition (conti2ued) AP600 PIRT Support l System l Component l Function l Process / Phenomenon l Other Factors l System comment:1 The function of the PRIIRS is a safety-grade decay removal; the PRIIR system consists of two heat exchanges in parallel. De heat exchanger primary side is connected to the AP600 primary and the secondary side is immersed in the IRSWT. Primary-side flow may be either forted or naturally circulating; however, in most accident sequences the primary pumps will be tripped. De buoyancy-driven flow is influenced by processes and phenomena in the hot legs (see Primary Coolant System - Ilot Leg Piping). Noncondensible gases trapped in the inverted high-point trap may affect flow through the heat exchanger primary. Ileat exchange with the IRWST pool heat sink is strongly influenced by the IRWST pool conditions (see Passive Safety injection System - IRWST). Passive Safety Injection Inject water fmm Phenomena described at System System (PSIS) multiple sourtes into the component level inlet / outlet primary to cool the core conditions and under transient and environmental accident conditions conditions lo Accumulator Borated waterinjection Noncondensible gas injection Component into the primary when inlet / outlet and the primary pressure Basic phenomena 9 and Note I environmental falls below a preset conditions value (currently 700 psia) 12yout and design features Component comment:1 AP600 contains two accumulators, each connected via a medium-diameter line to a DVI line. The accumulator is not an active component during normal operation. For transients in which the primary coolant pressure falls below the accumulator gas charge pressure, borated water is injected from the accumulator into the primary coolant system. If the depressurization transient continues for sufficient time that the entire liquid inventory is discharged, nitrogen is injected into the primary coolant system and has the potential to affect both processes and phenomena for the remainder of the sequence.

Systems, Components, Processes, Phenomena Tabulation (continued) AP600 PIRT Support l System l Component l Function l Process / Phenomenon l Other Factors l PIRT comment:6 Plant design is such that no active failure of a safety-related component significantly impacts LBLOCA performance. Minimizing the accumulator delivery bounds the uncertainty in system performance pertinent to the LBLOCA event. PIRT comment:7 A primary difference between the AP600 design and a current generation plant is the long duration accumulator injection that provides ample core inventory makeup for final quenching. Core Makeup Tank Borated water injection Buoyancy-driven Component into the primary flow / Recirculation inlet / outlet and following loss of Draining environmental primary system coolant Refilling conditions inventory and Thermal stratification independent of the Flashing Laymt and primary system pressure Condensation 6 design features Noncondensible gas Flow termination via Ei accumulator injection 5,6 Ambient heat loss Basic phenomena 9 and Noe I l a o n

I g 1-i I l I Systrms, Components, Processes, Phenomena Tabulation (continued) AP600 PIRT Support l System l Component l Function l Process / Phenomenon l Of her Factors l i Component comment:1 ne two CMTs are completely filled with borated water at containment temperature during normal operation. The top of each CMT connects to the cold-leg piping and the bottom discharge line delivers flow to the reactor vessel downcomer via the passive safety injection system line. His discharge line contains the normally closed CMT actuation valve, which opens upon I receipt of the S signal.6 A buoyancy-driven flow arises shortly thereafter, even while the CMTs remain { liquid full. Following initiation of the accumulator flow, the CMT flow diminishes and the CMT level does not approach the first stage ADS valve actuation point until after the accumulator tanks empty.6 ne ( eventual outcome of either loss-of-coolant accidents or transients that activate the ADS is that voiding arises in the cold legs at the junction with the pipes leading to the CMTs. A vapor path is created and draining of the CMTs is initiated. The accumlators empty long after the blowdown portion of the LBLOCA is complete. De CMTs do not drain to the ADS trip level until later; therefore, actuation of the ADS does not occur until long after the AP600 PCT occurs.6 PIRT comment:7 A RELAPS parametric calculation of a LBLOCA was performed in which one emergency core coolant system train (Accumulator and CMT feeding a DVI line) was assumed to be y inactive. For the base LBLOCA transient, RELAP5 calculated an early core heatup (3 to 7 s) that was ~ terminated by guide tube inventory draining into the top of the core and a second core heatup (20 to 55 s) that was terminated by accumulator injection. The following was reported for the parametric case. The early heatup was predicted to be more severe as measured by a higher peak cladding temperature. The difference was attributed to reduced CMT injection in the first few seconds of the transient. The downcomer condensation due to liquid from a single CMT was significantly less compared with the base case; the pressure in the downcomer remained higher, and the dynamic head in the core did not decrease as quickly. His caused less driving head for liquid flow from the guide tubes into the core, and the cooling flow did not penetrate below the core midplane. The differences seen in cladding temperature responses during the late heatup were directly related to the core inventory difference, i.e., less fluid injected with one less accumulator and CMT. Instrumentation 2 CMTlevel Measurement / variable Ixvel CMTvalve status Measurement / status Open/ shut IRWSTlevel Measurement / variable Level Accum level Measurement / variable 12 vel Accum valve status Measurement / status Open/ shut

Systems, Components, Processes, Phenomena Tabulation (continued) AP600 PIRT Support l System l Component l Function l Process / Phenomenon l Other Factors l Instrumentation comment:6%e first stage ADS is actuated when the CMTlevel decreases to 67%. He second and third stage ADS actuations occur at prespecified timed intervals following first stage ADS actuation. He fourth stage ADS is actuated when the CMT level decreases to 20% concurrent with other requirements. IRWST Late time and large Ambient heat loss Component volume source of ECC inlet / outlet and injection into the PRHRS operation environmental primary. IIcat sink for Pool heat transfer conditions core-generated energy. Pool water level Dermal statification/ mixing /_ Layout and multi-dimensionality design features Evaporation ADS operation y Condensation J Dermal statification/ mixing / multidimensionality Evaporation Pool heat transfer Pressure oscillations 4 ECC operation Gravity-driven flow / oscillations Resupply from containment Basic phenomena 9 and Note i l E ~ 4 a m

Systems, Ccmpo :nts, Processes, Ph n::mena Tchul tion (continu:d) AP600 PIRT Support i System l Component l Function l Process / Phenomenon l Other Factors l Component comment:1 ne IRWST is a very large irregularly shaped compartment located around about one-half of the AP600 containment periphery. The functions of the IRWST are to: pmvide the heat sink for the PRIIRS (two PR11R heat exchangers are immersed in the IRWST), provide the discharge sink for ADS stages 1-3 (two ADS spargers are located within the IRWST), and provide the source of safety-grade injection following primary system blowdown (via the ADS or otherwise). The IRWSTis nearly filled with water at the containment temperature during normal operation; the normal water level is above the upper horizontal tube bundle section of the PRIIRHX. However, the pool level can vary, decreasing as water is boiled off or evaporated and increasing as a result of ADS sparger discharge or condensate return from the containment interior. Pool level and temperature distribution determine the static head available for injecting coolant into the reactor vessel. Hey also determine the heat transfer coefficients and sink temperatures applicable on the outside of the PRilRilX tubes. Energy exchanges between the pool and the IRWST walls (steel-lined concrete) may be significant for some transients. Ambient heat loss from the IRWST walls and pool free surface can be a significant fraction of the IRWST heat load. Steam is injected into the IRWST pool through small holes in the spargers from the ADS. The sparging process is generally expected to result in complete condensation of the steam flow. >a Piping Flow delivery Flashing in IRWSTdischarge Component lines inlet / outlet and i environmental Basic phenomena 9 and Note 1 conditions Layout and design features Component comment:1 The flow resistances of the IRWST discharge piping and fittings can be a significant factor in determining the IRWST injection flow rate. Component comment:1 The " cold-leg" PBLs are actually the CMT inlet lines. Voiding at the cold-leg PBLs during CMT recirculation is significant because accumulation of a large void at the top of the CMT inlet line and creation of a vapor path to the top of the CMT is the mechanism by which recirculation is terminated and the CMTs are drained. l l

Systems, Components, Processes, Phenomena Tabulation (continued) AP600 PIRT Support l System l Component l Function l Process / Phenomenon l Other Factors ] Valves Basic phenomena 9 and Nme I Component Accumulator check Control flow inlet / outlet and environmental CMT motor driven Control flow conditions Layout and design features Sump Late time and continuous ECC operation Component source of ECC injection Gravity-driven flow / Draining inlet / outlet and into the primary. Ileat Resupply from containment environmental sink for core-generated Ambient heat loss conditions energy. Basic phenomena 9 and Nde t layout and design features Component comment:1 The sump is a compartment within the containment into which liquid may collect. g Two medium-diameter injection lines, each containing a check valve, connect the bottom of the sump to the passive safety injection system lines. The function of the sump is to collect effluent from a pipe break so that it is available for injection into the primary coolant system; this is the only manner in which water enters the sump. During normal operation, the sump is empty. Sump activation is expected to occur during the late stages of an accident (perhaps days) when the IRWST water level falls below that in the sump. PIRT comment:3 While in some cases variations in emergency core cooling system flow rates and temperatures did not h.ve a significant effect on the peak cladding temperatures, it was clear from the calculated results that the vessel mass inventory during reflood was strongly affected by the safety injection flow rate and temperature, and the accumulator temperature. In addition ECC temperature affected the mixture level in the downcomer which in turn affected how much emergency core coolant was entrained from the vessel during reflood. In general, higher ECC flow and/or lower temperature resulted in increased vessel mass

• inventory and improved core cooling during reflood.

g

i i 1 i 1 i i Systems, Components, Processes, Phenomena Tabulation (continued) AP600 PIRT Support l System l Component l Function l Process / Phenomenon l Other Factors l Primary Coolant System Circulation of primary Phenomena described at System l coolant with the component level inlet / outlet attendant transport of conditions and energy from the reactor environmental core to the steam conditions generators. Break Discharge of mass and Critical flow 3,4,5,6 Component energy from the primary Flashing inlet / outlet and 4 into the containment environmental (accidental). Basic phenomena 9 and Na I conditions (e.g. contamment d pressure ) Layout 6 (size, y location, area, and orientation) Component comment:I De break component refers to a rupture in the pressure boundary of the primary coolant system that is assumed to open very quickly, if not instantaneously. He break cannot be isolated. De geometry of the break is not specified. Fluid in the primary coolant system is accelerated out the break because the pressure in the containment is lower than in the primary coolant system. The fluid can be single-phase liquid or steam, two-phase liquid and steam, with or without noncondensible gas. For the LBIDCA, a spectrum of break locations and sizes is analyzed to identify the worst-case LBLOCA. Currently, Westinghouse claims the worst case is an 80% cold-leg break in a loop connected to a CMT, not a loop connected to the pressurizer. De break is located in loop B1 (Westinghouse designation) between the primary coolant pump exit and thejunction of the cold-leg PBL to the CMT. Component comment:6De break was modeled to occur in one of the cold legs in the loop containing the CMTs. Past sensitivity studies with WCOBRA/ FRAC have demonstrated that locating the cold-leg break in the loop that does not contain the pressurimer is conservative.

Systems, Components, Processes, Phenomena Tabulation (continued) AP600 PIRT Support l System l Component l Function l Process / Phenomenon l Other Factors ] PIRT comment:6 Break size is treated as a parameter with fixed location. He Co = 0.8 case results in a higher PCF that the Co = 1.0 case. The later case experiences larger downflows through the guidetubes and open hole / support column flow paths, apparently because the larger break size causes a faster depressurization, higher flashing rate, and results in the larger flows from the upper head to the top of the core (~50% larger). These flows, in turn, limit the cladding temperature rises more than for the Co = 0.8 case. A second study takes the break at the cold-lee solit (break area is one-half that of a 1.0 DECLG). The system depressurizes much more slowly than is the case with the merged line cold-leg break. In addition, flow at the top of the core remains upward everywhere through the first 10 s of the transient. Because flow at the bottom of the core reverses, there is a stagnation point at about the 4-foot level. When the break is in the hot leg, the core flow continues in the same direction at a reduced mass flow rate because of core flashing. Because there is no core flow reversal, the hot assembly fuel rods do not dry out and begin to heat. Cold. leg piping Flow delivery Flow asymmetries Component Stored energy release inlet / outlet and y Flashing environmental J. Phase separation / stratification conditions 3 Condensation,6 Layout and Basic phenomena 9 and Noe I design features Component comment:1 De cold legs are four large-diameter pipes that carry primary coolant from the discharges of the coolant pumps to the reactor vessel downcomer during normal operation. The diameters, lengths, and arrangement of the four cold legs are virtually identical. However, there are several asymmetries related to connecting piping. In loop 2 (the nonpressurizer loop), the medium-diameter CMT PBLs are connected to the tops of both cold legs. Voiding in the cold legs affects recirculation flow phenomena and draining of the CMTs. The two cold legs on loop I cach have small diameter offtake lines for the pressurizer spray. I i

i i I i i Systems, Ccmponnts, Processes, Phen menc Tcbulztion (centinued) AP600 PIRT Support l System l Component l Function l Process / Phenomenon 1 Other Factors l Component comment:1 Recirculation through the CMT PBL starts when the isolation valve is opened upon receipt of the S signal. Recirculation stops when the associated cold leg saturates and voids. Steam enters the CMT PBL, collects at the top of the PBL, and blocks the return of flow to the CMT. Creation of a vapor path to the top of the CMT is the process fmin which CMT draining arises and, therefore, is the process by which the CMT inventory replaces coolant lost from the primary. PIRT comment:3 Variations in relative resistance of the path from the core to the break on the loop side, and from the core to the break on the vessel side, were found to affect the PCT significantly for all plants. De relative resistance was affected, in turn, by the following: (a) break discharge coefficient,(b) break type, (c) break location, (d) vessel inlet nozzle resistance on the broken loop, and (e) pump resistance on the broken loop. Ilot-leg piping Flow delivery s plow and related phenomenon Component Entrainment/Deentrainment4 inlet / outlet and CCF (e.g., at steam generator envimnmental y; inlet plenum) condmons Flow asymmetries w Stored energy release Layout and Flashing (void generation)4 design features Phase separation /stratificationi Basic phenomena 9 aruf Note 1

Systems, Components, Processes, Phenomena Tabulation (continued) AP600 PIRT Support l System l Component l Function [ Process / Phenomenon l Other Factors l Component comment:1 The hot legs are two large-diameter pipes that carry primary coolant from the reactor vessel upper plenum to the steam generator inlet plena during normal operation. The bot-leg piping runs are horizontal as they leave the reactor; each hot leg has a single medium-angle, large-radius bend that turns the pipe upward for several feet to meet the steam generator inlet plenum. He two hot-leg piping runs are virtually identical. Dere are several asymmetries between the hot legs that are related to connections to other piping. He loop 1 piping connects at a right angle to the large diameter surge line with the connection nozzle located on the top of the hot leg, about midway between the hot-leg bend and the steam generator inlet plenum. Loop 1 also connects to the PRIIRS inlet piping. He medium-diameter nozzle is located on the top of the horizontal hot leg. Each hot leg also connects to one branch of the fourth stage ADS; these connections are on the tops of the horizontal hot legs, near the bend. 1 Component comment: Ileat transfer mechanisms associated with stored energy release and heat loss to ambient are discussed. Piping layout features that can induce asymmetries are discussed. Flow regime effects are discussed. " Phase separation at the hot-leg / pressurizer surge-line tee controls the state of the fluid passed to the pressurizer, and this affects ADS stages I,2, and 3 performance. Phase separation at the hot-leg / ADS stage-4 tees controls the state of the fluid passed out ADS stage 4 to the containment. g Phase separation at the hot leg /PRIIR inlet line tee controls the state of the fluid entering the PRllR system. Instrumentation 2 Pressurizer pressure Measurement / variable Pressure Pressurizer level Measurement / variable ifyeg Ilot-leg T change Measurement / variable Temperature Cold-leg T change Measurement /vanable

pio, llot-leg flow change Measumment/ variable Cold-leg flow change Measurement / variable Instrumentation comment:6 He reactor trip signal occurs within the first second of the LBLOCA based upon the measumment of pressurizer pressure.

g

Syst:ms, Cemponents, Processes, Phen 2meno Tcbul: tion (contixu:d) AP600 PIRT Support l Systene l Component l Function l Process / Phenomenon l Other Factors ] l Pressurizer 3 Primary system pressure Flashing (void generation)4 Component control. Flow and related phenomena.5 inlet / outlet and l 4 environmental Flow path between Draining / blowdown.6 c nditions Refill Primary and IRWST for discharge of mass and sWohibh Entrainment/Deentrainment Layout and energy through ADS stages 1-3. Countercurrent flow (CCF) design features Stored energy release Basic phenomena 9 and Note 1 Component comment:I A large thick-walled cylindrical tank, with spherical upper and lower heads, that is elevated above the hot legs. During normal operation, the pressurizer contains well-separated water and steam. Primary system pressure control is attained by the combined actions of pressurizer spray and heater power. 1 s Component comment: The pressurizer behavior generally controls the depressurization rate of the primary coolant system during the initial phases of most accidents. The pressurizer is the only region in 'o the primary coolant system containing saturated water during normal operation; therefore, declining t pressurizer level and pressure results in expansion of the pressurizer vapor space and flashing of the saturated water to steam. Countercurrent flow limiting (CCFL) phenomena may occur at the pressurizer / surge line interface during ADS stages I,2, and 3 operation. PIRT comment:6 A minimum pressure (system) is conservative [ infer that higher PCTs are obtained]. Surge line Flow path CCF Component inlet / outlet and Basic phenomena 9 and Noec I environmental conditions Layout and design features

Systems, Components, Processes, Phenomena Tabulation (continued) AP600 PIRT Support l System l Component l Function l Process / Phenomenon l Other Factors l Component comment:1 The pressurizer surge line is a large-diameter pipe from the bottom of the pressurizer to the loop-l hot leg. Flow from the loop-l leg, through the surge line, pressurizer, and ADS valves is the primary pathway for the initial system depressurization that is integral to the AP600 passive safety concept. Reactor coolant pumps Degraded performance 3,4,6 Component Coastdown inlet / outlet and environmental Basic phenomena 9 and Note 1,4 conditions Layout and design features (e.g., heat, moment of inertia, etc.) Q Component comment:1 The mechanical energy stored in each reactor coolant pump flywheel is considerable and coastdown of the pump impeller, shaft and motor takes several minutes following interruption of the pumps power supply. During this interval, pumping action, albeit ever reducing, is maintained. The flow resistance of a stalled rotor is a significant fraction of the total loop resistance during buoyancy-driven flow. If the buoyancy-driven flow is sufficiently high to cause rotation, the pump resistance is reduced considerably. PIRTcomment:3 Westinghouse sensitivity studies have found that variations in relative resistance of the path from the core to the break on the loop side, and from the core to the break on the vessel side were found to affect the PCT significantly for all plants. PIRT comment:3 On the intact loops, the most significant parameter was found to be whether the RCS pump was powered. It was found that for most plants, maintaining power to the pumps resulted in a higher PCI', because it inhibited downward flow through the core as the system depressurized. PIRT comment:6 Minimum steady-state flow values are chosen in the Westinghouse "superbounded" approach. 5 e 4

1 i Systrms, Ccmponcats, Processes, Phenomexa Tabuintion (continu d) AP600 PIRT Support l System l Component l Function l Process / Phenomenon l Other Factors l Steam Generator Primary-to-secondary Reverse heat transfer Component Primary Side heat transport (e.g., steam binding)4 inlet / outlet and Stored energy release environmental conditions Basic phenomena 9 and Note I,4 Layout and design features Fouling Component comment:I Steam generator primary-to-secondary heat transfer is controlled by the fluid state and finw conditions on the primary and secondary sides of the tubes. Initially the pressure and temperature on the primary side are higher than on the secondary. During primary depressurization transients, secondary-to-primary heat transfer can occur. IIcat transfer in either direction is adversely affected by voiding at the surfaces through which energy is being transferred. b PIRT comment:6 Bounding steam generator tube plugging values are chosen in the Westinghouse "superbounded" approach. Reactor System Generation, transport, Phenomena described at System and control of energy component level inlet / outlet produced via nuclear conditions and fission environmental conditions Vessel-control rods Rod insertion Reactivity change Worth Insertion rate

Systems, Components, Processes, Phenomena Tabulation (continued) AP600 PIRT Support l System l Component l Function l Process / Phenomenon l Other Factors ] Vessel -core flow Receive and transport Flows Component t channels energy generated by Reversal / stagnation4 inlet / outlet and fission m the fuel rods. 4 envimnmental Top down condinons 3 Entramment/Deentrainment.4 5 CCF,6 gg 3 Multidimensionality.4.6 design features Level and oscillations Void generation / distribution 4 Location relative 3 6 Flashing 6 to drain paths Boiling Assembly design Basic phenomena 9 and Nok: I Component comment:1 The reactor core is an arrangement of more than 100 fuel assemblies, each Y composed of more than 200 vertical, small-diameter fuel rods. The core heated length is 12 ft. Each O fuel assembly also contains more than 20 guide tubes and one instrument tube. Coolant normally flows upward through the core. The flow areas through the bottom core plate, top core plate, grid spacers are smaller than the open flow area within the fuel rod bundles. PIRT comment:3 During blowdown, the most significant core power distribution parameters is the maximum average fuel temperature. This establishes the first PCT, and generally affects later peaks. The initial fuel temperature is affected primarily by the time in cycle and by the peak linear heat rate. 1 l i L i I f l g 2

t s Syst::ms, Ccmpomnts, Processes, Phturmens Tr.bulition (continnd) AP600 PIRT Support l l System l Component l Function l Process / Phenomenon l Other Factors l PIRT comment:3,6 The assembly receives different amounts of water directly from the upper head and upper plenum depending upon its location relative to the control rod guide tubes. There are a variety of upper plenum configurations above the assemblies. Each configuration must be evaluated to assess its impact on assembly flow during blowdown. While liquid remains in the upper head above the top of the guide tubes, the guide tubes are the preferred path for draining ofliquid into the upper plenum. Once the upper head begins to flash, liquid drains directly down the guide tubes, and a fraction is able to penetrate into the core against the continuous steam upflow.6 Flow of liquid into the core open hole / support column locations is also significant but its impact is delayed (by 5-10 s) relative to the guide tube flows. The flow in the open hole and guide tube assemblies is sufficient to quench the fuel m the average assemblics (Rods 3 and 4, respectively in the Westinghouse identification scheme) at the 6 -10 ft levels. Because the hot assembly can receive draining liquid from the upper head only indirealy, liquid r downflow into this assembly is delayed. Nevertheless, after some time (~10 s given), liquid bu;lds up in j the global region above the upper core plate and begins to flow through the plate at the hot assembly location. There is a significant flow of entrained liquid droplets into the hot assembly. 'Ihe liquid flow is sufficient to quench the rot rod and hot assembly near the top of the core but not at lower levels. [ Uf PIRT comment:3 It was observed that the peak cladding temperature location during reflood was typically high in the core. Consequently, the core power distribution, as well as the hot assembly i power, affected the reflood peak cladding temperature. t PIRT comment:6 Bounding high core power, peaking factors, fuel pellet temperature, and axial power shape are chosen in the Westinghouse "superbounded" approach. PIRT comment:7 Lower core thermal power in the AP600 design results in lower temperatures during [ the blowdown phase compared to current plants, h

Systems, Components, Processes, Phenomena Tabulation (continued) AP600 PIRT Support l System l Component l Function l Process / Phenomenon l Other Factors l J Vessel-downcomer Flow path Flow Component Two-phase convection inlet / outlet and (bypass)4 environmental CCF4 conditions 3 Entrainment/Deentrainment.4 Layoutand 3 Multidimensionality.4.6 g; 4 ECC mixing IIcad/ level / oscillations 4 Flashing 4 Stored energy release 6 3 Condensation.4.6.7 Noncondensible gas effects 4 y Basic phenomena 9 and Note I L" Component comment:1 The Passive Safety injection System terminates at two large-diameter Direct Vessel Injection Nozzles on the downcomer. Baffles exist so that the injected flow is directed downwards. There may be nonuniform fluid temperature distribution in the downcomer during accident sequences because of the potential for different temperature fluids entering the downcomer from the cold legs and the Direct Vessel Injection Nozzles. The static head created by the liquid level standing in the downcomer is the driving force to push fluid through the core during phases with IRWST injection. PIRT comment:1 The flow of steam from the upper head, through the bypass, into the upper regions of the downcomer can result in condensation. The condensation of steam on cold walls and liquid within the downcomer affects downcomer level and temperature. L ~

l Systems, Components, Processes, Phenomena Tabulation (continued) AP600 PIRT Support t l System l Component l Function l Process / Phenomenon l Other Factors l PIRT comment:3 For the AP600 design, a PIRT indicates that bypass during blowdown is an important l process. PIRT comment:6 De primary pressure decreases and eventually the accumulator begins to inject coolant into the primary system. Initially, some of this injected liquid is bypassed to the break. Later, accumulator liquid begins to flow into the lower plenum. Vessel-fuel rods Generation of energy via Rod structure heat transfer Component nuclear fission while Conduction inlet / outlet and 4 environmental retaining radioactive Gap conductance materials within a barrier Stored energy release 3,4 conditions to fission product release Rod convective heat transfer and transport to the Nucleate boiling and departure Fuel pin design coolant interface. from nuclear boiling ? (DNB)4 Decay hear.6 s Time in cycle 3 ewet,4,6 Post critical heat flux (CilF) Powerlevel6 heat transfer 4 6 3 6 i Ilot assembly d Noncondensible gas effects 6 Core average Reacavity Void 6 Peripheral 6 Moderator temperature Fuel temperature (Doppler) Boron Decay heat Basic phenomena 9 and f*xe 1

Systams, Compo:sts, Processes, Phtecmens Tabulstion (continued) AP600 PIRT Support l System l Component l Function l Process / Phenomenon l Other Factors l Component comment:1 The core heated length is 12 ft. Each fuel rod is loaded with uranium dioxide fuel pellets that are axially restrained with springs. Dere is a small radial gap between the outer radius of the pellet and the inner radius of the tube; this gap is filled with an inert noncondensible gas. PIRT comment:1 During normal operation "the fuel pellet temperatures are very high (thousands F). The heat stored within the fuel rods, therefore, is very high, and its release into the coolant can be a significant" process. PIRT comment:6 Bounding high core power, peaking factors, fuel pellet temperature, and axial power shape are chosen in the Westinghouse "superbounded" approach. Vessel-guide tubes Channel for control rod Flow Component movement Draining inlet / outlet and Flow regime environmental Flow resistance conditions h Basic phenomena 9 and Nac 1 Layout and design features PIRT comment:3 When considering the hot assembly location, assemblies receive different amounts of water directly from the upper head depending upon the location relative to the control rod guide tubes. FIRT comment:7 A total core rewet was predicted in a RELAP5 calculation due to cooling from liquid supplied by the guide tubes and the upper head. Vessel-instrumentation Core neutron flux Measurement / variable Neutron flux Control rod position Measurement / variable Position Core exit T Measurement / variable Temperature 1 ~ ~ ~ ~ T -.- L ~ - - - ~

V Systems, Components, Processes, Ph nomens Tabulation (continued) AP600 PIRT Support l System l Component l Function l Process / Phenomenon [ Other Factors l Vessel-lower plenum Flow path Flow transient 5 Component Emptying /sweepout.4 inlet / outlet and 3 Refill environmental Mixing condmons 3 Multidimensionality,4,6 Oscillations Layoutand Stored energy released design features Basic phenomena Component comment:1 The AP600 lower plenum arrangement differs significantly from those of existing plants. It is relatively open, with little resistance to flow, and contains a vortex suppression plate to stabilize the flow before it enters the core. Experiments have been conducted at the University of Tennessee. Vessel - structures (e.g., Structural support Stored energy release 6 Layout and a upper and lower core design features support plates, etc.) Vessel-reflector / bypass Flow path Flashing Component inlet / outlet and Basic phenomena 9 and Note i environmental conditions Layout and design features 1 Component comment: A small fraction of the total core flow is bypassed around the core fuel assemblies and is used to cool ancillary regions within the reactor vessel.

Syst ms, Compomis, Processes, Phin mena Teb:l; tion (contiz=d) AP600 PIRT Support l Sytem l Component l Function l Process / Phenomenon l Other Factors l Vessel - upper head Cool structures in the Stored energy release Component upper head Draining inlet / outlet and Guidetubes environmental Downcomer conditions Upper plenum Flashing Layout and design features Basic phenomena 9 and Note 1 Component comment:1 During normal operation, flow entering the upper head comes upward from the core through the guide tubes and upward through the bypes from the upper annulus of the downcomer. Flow leaves the upper head downward through a support plate into the upper plenum. The guide tube flow rate is higher than that of the bypass path. Thus, during normal operation, the upper-head fluid temperature is below, but near the core outlet temperature. Component comment:1 The upper head contains hot fluid and is elevated high in the reactor vessel. g During depressurization events, fluid in the upper head is among the first to flash, resulting in high void fr. actions. Some fluid may be displaced to other components in the primary coolant system by the flashing process. The release of stored energy in the upper-head region tends to sustain a voided upper head and produces a pressurizer-like effect. Vessel-upper plenum Route coolant at the core Flow transients Component exit to the two hot legs. CCF4 inlet / outlet and Entrainment/deentrainment4 cn mental c Phase separation / stratification 4 Multidimensional Layout and Flow d,stnbution i design features (hot legs, core) Flashing (void generation) Stored energy release Basic phenomena 9 and Note I g J

Systems, Components, Processes, Phenomena Tabulttion (continued) AP600 PIRT Support l System l Component l Function l Process / Phenomenon l Other Factors l Component comment:1 He upper plenum contains hot Guid and is elevated high in the reactor vessel. During depressurization events, fluid in the upper plenum is among the first to flash, resulting in high void fractions. Some fluid may be displaced to other components in the primary coolant system by the flashing process. Given the complexity of the internal structures in the upper plenum, entrainment and deentrainment phenomena can be important both to the pressure drop through the upper plenum and the condition of the fluid passed to each of the hot legs. Because the upper plenum is a connecting volume for flow paths, asymmetries may result fmm interactions within the upper plenum. Steam Generator System Receive and transport Phenomena described at System (Secondary Side) core-generated energy to component level inlet / outlet the turbines conditions and environmental conditions ? Instrumentation 2 SG secondary level Measurement / variable ifyeg Turbine throttle valve Measurement / status Position Main feedwater line Flow delivery Basic pNnomena9 and Noe 1 Component Main steam line inlet / outlet and SG blowdown line environmental Startup feedwater line conditions Other piping Layout and design features

r Systems, Components, Processes, Phenomena Tabulation (continued) AP600 PIRT Support l System l Component l Function l Process / Phenomenon [ Other Factors l Valves Contmiled flow release Basic phenomena 9 and Note 1 Component MS Isolation inlet / outlet and MS Safety envimnmental MSPORV conditions MFW Isolation Layout and design features i ?8 r

l i Notes: l 1. Reference 9 recognizes that there are many phenomena that can be characterized as " basic phenomena." All or some of these basic phenomena arise in essentially all thermal-hydraulic pmcesses. Therefore, rather than provide a repetitive listing of these phenomena, the term " Basic phenomena" is used. He basic phenomena listed in Ref. 9 are as follows: 1. Evaporation due to depressurization 2. Evaporation due to heat input 3. Condensation due to pressurizatien 4. Condensation due to heat removal 5. Interfacial friction in vertical flow 6. Interfacial friction in horizontal flow 7. Wall to fluid friction 8. Pn ssure drops at geometric discontinuities 9. Pressure wave propagation f 5 t 1. .-m.- ..-2 m u

c ? References l t 1. C. D. Fletcher, G. E. Wilson, C. B. Davis, and T. J. Boucher, " Interim Phenomena identification and Ranking Tables for f AP600 Small Break Imss-of-Coolant Accident, Main Steam Line Hreak, and Steam Generator Tube Rupture Scenarios," i Idaho National Engineering Laboratory DRAIT report INEL-94/0061 (October 1994). 2. Westinghouse AP600 Plant Description Report, US Department of Energy neport DE-AC03-86SF16038 (January 1989). Note: Contains Westinghouse Proprietary Class 2 data. 3. S. M. Bajorek, M. Y. Young, and K. Ohhwa, "Best Estimate LOCA Code Development and Assessment in Support of Operating and Advanced Plants," ARS '94 International Topical Meeting on Advanced Reactors Safety, Pittsburgh, Pennsylvania (April 17-21,1994). 4. B. Boyack, R. Duffey, P. Griffith, G. Lellouche, S. Levy, U. Rohangi, G. Wilson, W. Wulff, and N. Zuber, " Quantifying Reactor Safety Margins: Application of Code Scaling, Applicability, and Uncertainty Evaluation Methodology to a Large-Break Loss-of-Coolant Accident," EG&G Idaho, Inc. report NUREG/CR-5249, also EGG-2552 (October 1989). 5. J. F. Lime and B. E. Boyack," TRAC Analysis of an 80% Double-Ended Gold-leg Break for the AP600 Design," Ims g Alamos National Laboratory document LA-UR-94-2503 (1994). 6. AP600 SAR Chapter 6, Rev. 0 (Effective June 26,1992). 7. " Additional Information in Support of Westinghouse Response to RAI 952.44 - 952.46 " Attachment to Westinghouse Ixtter NTD-94-4114 (June 1994). Note: Westinghouse Proprietary Class 2. 8. J. E. Fisher, "Large Break LOCA Calculations for the AP600 Design," Proceedings of the Fifth International Topical Meeting on Reactor'Ihermal Hydraulics - NURE'HIS, Salt Lake City, Utah (September 21-24,1992). i 9. N. Aksan, F. D'Auria, H. Glaeser, R. Pochard, C. Richards, and A. Sjoberg, " Overview of CSNI Separate Effects Tests Validation Matrix," Proceedings of the International Conference on 'New Trends in Nuclear System "Ihermohydraulics, Volume 1, Pisa, Italy (May 30-June 2,1994). 10. S. M. Bajorek, S -K Chow, L. E. Hochreiter, S. B. Nguyen, M. E. Nissley, K. Ohkawa, and M. Y. Young, " Code Qualification Document for Best Estimate LOCA Analysis: Assessment of Uncenainty," Westinghouse Electric Corpor./. ion document WCAP-12954-P, Volume IV, Westinghouse Proprietary Class 2 (June 1993). g

__ _ _. _.... _.. = - _ -. _. _ _ i Definitions 1. Process: a series of occurrences, progressing fri m one state to another 2. Pknomenon: an occunence within nature following the laws of physics, an observable event 3. Condition: Factors affecting the process or phenomenon f >b i w l-4 i h L

.. - ~. f APPENDIX B 4' BRIEF BIOGRAPHIES AP600 LBLOCA PIRT TEAM MEMBERS 1 B. E. Bovack Brent E. Boyack is the leader of the Terrestrial Reactor Technology Team within the Reactor Design and Analysis Group at Los Alamos National L.aboratory. He has published extensively on light water reactor accident behavior, thermal-hydraulic code assessment, code uncertainty quantification, gas-cooled thermal and fast reactor fhild dynamics and heat i transfer, and computer model development. He was Chairman of the MELCOR Peer Review Committee. He has participated in NRC review groups focusing on the quantification of uncertainty in the NRC-developed thermal-hydraulics code TRAC-PFl/ MODI, quaufication of uncertainty in the NRC-developed thermal-hydraulic code RELAP5/ MOD 3, the Severe Accident Scaling Methodology development, and the Expert Panel review of a proposed AP600 Integral Test Facility. He directed the Los Alamos participation in the multinational 2D/3D program, which generated LBLOCA data for j TRAC assessment. Before accepting employment with Los Alamos National Laboratory, i he was a Senior Staff Member at General Atomic Company, where he held technical and management positions working in the High Temperature Gas-Cooled Reactor and the Gas Cooled Fast Reactor programs. He was also employed by the Boeing Company where he applied numerical methods to the solution of engineering analysis problems. He obtained his B.S. and M.S. degrees in Mechanical Engineering from Brigham Young University. He obtained his Ph.D. degree in Mechanical Engineering from Arizona State University and is a registered Professional Engineer in California. l 1 S. C. Harmony Stephen C. Harmony is a technical staff member in the Reactor Design and Analysis Group 1 in the Technology and Safety Assessment Division of Los Alamos National Laboratory. { He has 10 years of experience in the simulation of nuclear reactor thermal hydraulics. In the 2D/3D program he modeled and analyzed LBLOCA experiments conducted in the Slab Core Testing Facility. He modeled and analyzed thermal hydraulic transients in the PIUS j plant, a design with many passive-safety features. As part of a safety review of the Savannah River heavy-water production reactors, he developed models that simulated the unique features of the coolant circulation pumps, and participated in a quality assurance assessment of the TRAC model of Savannah River. Prior to joining the Reactor Design and Analysis Group, he supervised the production of explosive composites in a pilot-scale facility at Los Alamos National laboratory, and designed and operated bench-scale separation facilities, both at LANL and in his previous position with Celanese Chemical 4 Company. Mr. Harmony has B. S. and M. S. degrees in Chemical Engineering from Texas Tech University. He is a member of AIChE. i T. D. Knieht 1 Thad D. Knight is a Technical Staff Member with the Advanced Reactor Safety Group at i Los Alamos National Laboratory. He has twenty years of experience in reactor and nuclear safety at the national laboratories and with the Department of Energy (and its predecessor agencies). The Department of Energy and the Nuclear Regulatory Commission have funded his work, which has primarily involved the analysis of light-water power reactors and experiments relating to the performance of these reactors; also, he has worked to 4 j B-1

j 2 i 4 i review and to improve the safety of the heavy water production reactors at the Savannah l River Site. At Los Alamos National Laboratory, he has worked primarily in thermal-hydraulic research related to the assessment, application, and development of the TRAC

• l PWR code. The assessment and application activities have involved working closely with t

j-the LOFT, Semiscale, and MIST experiment facilities. He had similar experience at the Idaho National Engineering Laboratory with the RELAP4 series of codes. His most recent j activities at Los Alamos relate to safety analysis of nuclear materials storage. He received B.A., M.M.E, and Ph.D. degrees in Mechanical Engineering from Rice University. i _L F. Lime -1 i l i James F. Lime is a water-reactor-accident-analysis specialist in the Advanced Reactors Safety Group at Los Alamos National Laboratory. He is currently the principal analyst for y the AP600 thermal-hydraulic analysis using TRAC-P. He also recently participated in the analysis of the advanced PIUS mactor. Prior assignments have included thermal-hydraulic modeling and accident analysis of DOE new and current production reactors and of commercial PWR plants. The DOE reactors include the New Production Heavy Water i Reactor, the Savannah River K and L Reactors, and the Washington Hanford N Reactor. ) Commercial reactor analyses include Bellefonte, Ginna, Davis Besse, H. B. Robinson, i Oconee, and Zion. He has published extensively in technical reports and conference i papers. Prior to employment with Los Alamos, he was employed at the Idaho National Engineering Laboratory where he was involved in the development of a best-estimate l advanced containment analysis code (BEACON). Previous work experience was in the aerospace industry with the Marquardt Corporation and with RockwellInternational. He i obtained his B. S. degree in Mechanical Engineering from the University of California at Berkeley. He has conducted post-graduate work at the University of California and the i University of Idaho. i l F. E. Motlev i ( Frank E. Motley has over 28 years of experience in nuclear-reactor related thermal-hydraulic research and development, and reactor systems accident analyses. Currently he F 3 i is principal investigator for Russian Reactor Safety Analyses Laboratory-Directed Research i and Development program, which involves the application of TRAC and NESTLE to j RBMK reactors. Prior to that assignment he was principal Investigator for the thermal-hydraulic safety analyses for the Savannah River Restart Program. He was principal Investigator for the ROSA IV program (an international cooperative reactor safety research 1 - program including Japan and France). He provided technical leadership, analysis and review of tests, preparation and review of reports, and analyses of TMI accident until the j time of core disassembly. As section leader for the 2D/3D program (an international i cooperative reactor safety reseamh program including Germany and Japan), he provided i technical leadership, analysis of review of tests, preparation and review of reports, and 4 PWR accident calculations. Prior to his employment at Los Alamos, he worked for 14 years at the Nuclear Energy Division of Westinghouse Electric Co. At the time he left l Westinghouse he was a fellow engineer in thermal and hydraulic design responsible for i developing work scope and costs for thermal-hydraulic tests, design of test apparatus and j procedures, developing data analysis and correlations, and licensing new reactor design . correlations. Mr. Motley obtained a B. S. in Mechanical Engineering and M. S. in Nuclear l Engineering from the University of Kansas. He is a member of ASME, ANS, and Sigma i Xi. 1 i R. A. Nelson i l B-2 2

Ralph A. Nelson is a technical staff member in the Advanced Reactor Safety Group in the Technology and Safety Assessment Division of Los Alamos National Laboratory. He has over 20 years of experience in the investigation of fundamental thermal-hydraulic phenomena and associated model development. He has published extensively on boiling heat transfer. His 1992 paper at the National Heat Transfer Conference was awarded the best paper prize. He is an Associate Technical Editor for the ASME Journal of Heat Transfer. He is a member of the ASME K-13 Nucleonics Heat Transfer Committee and served as chairman for 3 years. He is a member of the ASME Heat Transfer Division Executive Committee and the K-2 Long Range Directions and Issues Committee. Prior to coming to Los Alamos, he was a Scientific Specialist at the Idaho National Engineering Laboratory. He obtained his B.S., M.S., and Ph.D. degrees from North Carolina State University. D. A. Siebe l Donald A. Siebe is a technical staff member in the Reactor Design and Analysis Group in the Technology and Safety Assessment Division of Los Alamos National Laboratory. He has eight years experience in the simulation and analysis of reactor thermal-hydraulic systems and related experimental facilities. His work includes performing TRAC-PF1 code assessments against experiments from the MIST facility and from the 2D/3D program test facilities. He was the lead Los Alamos participant in the final analysis and documentation phase of the 2D/3D program. In addition, he has worked on the thermal analysis and safety of reactor-like accelerator targets, and has performed simulations of gas-cooled reactors. His publications include 20 technical reports,9 conference papers and,2 refereed journal papers. He obtained his B. S. and M. S. degrees in Mechanical Engineering from the University of Idaho, and his Ph.D., also in mechanical engineering, from Arizona State University. i B-3}}