ML20212F842
ML20212F842 | |
Person / Time | |
---|---|
Site: | 05200003 |
Issue date: | 10/31/1997 |
From: | Andreychek T, Conway L, Woodcock J WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP. |
To: | |
Shared Package | |
ML20212F833 | List: |
References | |
WCAP-14408, WCAP-14408-R01, WCAP-14408-R1, NUDOCS 9711050105 | |
Download: ML20212F842 (704) | |
Text
{{#Wiki_filter:_. ._ __ _ . _ _ . - _ _ _ _ . _ _ _ _ . WESTINGHOUSE NON-PROPRIETARY CLASS 3 WCAP-14408 Revision 1 [h v WGOTHIC Application to AP600 1 4 J. Woodcock, T. S. Andreychek, L. Conway, T. Elicson, A. Forgie, J. A. Gresham, R. Haessler, T. O'Donnell, R. Ofstun, D. R Spencer, M. Sredzienski, M. Wills Nucleat Safety Analysis October 1997 J Westinghouse Electric Corporation Energy System Business Unit P.O. Box 355 Pittsburgh, PA 15230-0355 C 1997 Westinghouse Electric Corporation All Rights Reserved
-t * ' *m
- 8m nnn:1b-101497 9711050105 971023
{DR ADOCK 05200003
- m . -
d lii -
- MASTER TABLE OF CONTENTS -
1-2 4
. UST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . -. . . . . . . . . . ix LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xill
! LIST OF ACRONYMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxiv i ! 1 INTRODUCTION i . 1.1 OBJECITVE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 12 AP600 CONTAINMENT DBA REPORTS . . . . . . . . . . . . . . . . . . . . . . . . . 1-1 , 1.2.1 Accident Specification and Phenomena Identification and l Ranking Table Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1 , i 1.2.2 Scaling Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-3~ i 12.3 Heat and Mass Transfer Correlations Report . . . . . . . . . . . . . . . . . 1-6 . 12.4 WGOTHIC Code Description and Validation . . . . . . . . . . . . . . . . . 1-7 , 12.5 SS AR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7 4 1.3 APPUCATIONS REPORT CONTENT
SUMMARY
. . . . . . . . . . . . . . . . . 1-7 L 1.4 - USE OF LST AND VAUDATION RESUL'IS . . . . . . . . . . . . . . . . . . . . . . 1-9 L 1.4.1 LST Ma trix Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-10
- 1.42 Use of LST Separate Effects Data . . . . . . . . . . . . . . . . . . . . . . . . . 1-12 4
1.4.3 LST Confirmation of Phenomena . . . . . . . . . . , . . . . . . . . . . . . . 1-12 i 1.4.4 Code Comparison to LST as an Integral Test . . . . . . . . . . . . . . . . 1-12 1.5 INTERFACE WITH SSAR CALCULATIONS -. . . . . . . . . . . . . . . . . . . . . . 1 13
1.6 CONCLUSION
S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . - . . . . . . . . 1 13 REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-13 Lv i O 2 1.7 - TEST AND ANALYSIS PROCESS OVERVIEW AND HIGH AND MEDIUM RANKED CONTAINMENT PHENOMENA
2.1 INTRODUCTION
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1 2.2 ELEMENT 1 - AP600 PCS REQUIREMENTS AND CODE CAPABIUTIES . 2-1 2.3 ELEMENT 2 - ASSESS CODE VERSUS TESTS AND IMPORTANT l PROCESSES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3 L 2.4 ELEMENT 3 - ASSESS UNCERTAINTIES AND DEVELOP
, BOUNDING MODELS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 2.5 ELEMENT 4 - PERFORM DBA CALCULATIONS AND COMPARE TO SUCCESS CRITERIA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-8
2.6 CONCLUSION
S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-8
2.7 REFERENCES
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-8
! 3 OVERVIEW OF WGOTHIC
3.1 INTRODUCTION
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1 3.2 OVERVIEW OF THE CODE DEVELOPMENT AND VAUDATION . . . . 3-1 3.3 THE WGOTHIC CUME MODEL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 i 3.4 GENERAL CUME EQUATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-13 3.5 INTEGRATION OF THE WESTINGHOUSE CLIME MODEL -
INTO GOTHIC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-18
3.6 REFERENCES
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-20
. V l [ EGOTHIC Application to AP600 October 1997 oM006w-fm.non:1b 101497 Revision 1 d
, -- _.e,_.,--_.- ,, ,--_m. ,_ , . . - - - - - . - ., _ , - ,,,.w._,--,-..
iV MASTER TABLE OF CONTEN'13 (Cont.) 4 DESCRIPTION OF WGOTHIC EVALUATION MODEL
4.1 INTRODUCTION
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-1 4.2 AP600 WGOTHIC MODEL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-1 4.2.1 Reactor Cavity Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-4 4.2.2 Reactor Coolant Drain Tank (RCDT) Cavity Description . . . . . . . . 4-13 4.2.3 Southeast Accumulator Cavity . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-19 4.2.4 Northeast Accumulator Cavity . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-25 4.2.5 Eart Steam Generator Cavity (Upper and Lower) . . . . . . . . . . . . . 4-30 4.2.6 West Steam Generator Cavity (Upper and Lower) . . . . . . . . . . . . 4-39 4.2.7 North CMT Cavity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-48 4.2.8 Sou th CMT Cavity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-61 4.2.9 Chemical and Volume Control Cavity . . . . . . . . . . . . . . . . . . . . . 4-69 4.2.10 Refueling Room . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-75 4.2.11 Internal Refueling Water Storage Tank (IRWST) . . - . . . . . . . . . . . 4-82 4.2.12 Cylindrical Central Room . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-89 4.2.13 Above-Deck East Steam Generator Compartment . . . . . . . . . . . . . 4-95 4.2.14 Above-Deck West Steam Generator Compartment . . . . . . . . . . . 4-102 4.2.15 South Inner-Half Annulus Compartment . . . . . . . . . . . . . . . . . . 4-109 4.2.16 North Inner-Half Annulus Compartment . . . . . . . . . . . . . . . . . . 4-116 4.2.17 North Mid-Quarter Annulus Compartment . . . . . . . . . . . . . . . . 4 123 4.2.18 West Mid-Quarter Annulus Compartmem . . . . . . . . . . . . . . . . . 4-131 4.2.19 South Mid-Quarter Annulus Compartment . . . . . . . . . . . . . . . . . 4-140 4.2.20 East Mid. Quarter Annulus Compartment . . . . . . . . . . . . . . . . . . 4-147 4.2.21 North Outer Quarter Annulus Compartment . . . . . . . . . . . . . . . 4-155 4.2.22 West Outer Quarter Annulus Compartment . . . . . . . . . . . . . . . . 4-161 4.2.23 South Outer Quarter Annulus Compartment . . . . . . . . . . . . . . . 4-168 4.2.24 East Outer Quarter Annulus Compartment . . . . . . . . . . . . . . . 4-175 4.2.25 East Quarter Inner Dome Compartment . . . . . . . . . . . . . . . . . . . 4-182 4.2.26 North Quarter Inner Dome Compartment . . . . . , . . . . . . . . . . . 4-186 4.2.27 West Quarter Inner Dome Compartment . . . . . . . . . . . . . . . . . . 4-190 4.2.28 South Quar +er Inner Dome Compartment . . . . . . . . . . . . . . . . . . 4-192 4.2.29 East Quarter Outer Dome Compartment . . . . . . . . . . . . . . . . . . 4-199 4.2.30 North Quarter Outer Dome Compartment . . . . . . . . . . . . . . . . . 4-203 4.2.31 West Quarter Outer Dome Compartment . . . . . . . . . . . . . . . . . . 4-207 4.2.32 South Quarter Outer Dome Compartment . . . . . . . . . . . . . . . . . 4-211 4.2.33 Outside Containment Upflow Annulus Volumes . . . . . . . . . . . 4-215 4.2.34 Ou' . ? Containment Dowr low Annulus Volumes . . . . . . . . . . 4-220 4.2.35 PC imney Volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-224 4.3 CONDUCM T/PE DESCRIPTIONS . . . . . . . . . . . . . . . . . . . . . . . . . . 4-226 4.4 CLIME NODING DESCRIPTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ' 4-244 4.4.1 We t Stacks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-244 4.4.2 Dry S tacks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-260 4.4.3 Passive Cooling System Fiow . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-272 4.4.4 Condensate Film Stripping . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-274 9
LV. GOTHIC Application to AP600 october 1997 o:\3006w-fm.non:1b-101497 Revision 1
, v= 2 7 MASTER TABLE OF CONTENTS (Cont.) 4 4.5 _. INITIAL AND BOUNDARY CONDmONS . . . . . . . . . . . . . . . . . . . . . . 4-274
- i. --
4.5.1 Initial Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-274 - l~ 4.5.2 Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . _4-291 4.5.3 . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-294 L 4.6 AP600 CONTAINMENT EVALUATION MODEL
SUMMARY
. . . . . . . 4-297 l 4.7-
SUMMARY
OF EVALUATION MODEL TRANSIENT CALCULATIONS 4-319 i 4.7.1 LOCA Evaluation Model Base Case Results Summary . . . . . . . . 4-319 i j 4.7.2 MSLB Evaluation Model Base Case Results Summary . . . . . . . . 4-324
4.8 REFERENCES
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-353 i
, 5 INmAL AND BOUNDARY CONDmONS j
5.1 INTRODUCTION
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1
! 5.2 INmAL CONDmON SENSmVITY CASES . . . . . . . . . . . . . . . . . . . . . . 5-1 5.3 INmAL CONTAINMENT HUMIDITY . . . . . . . . . . . . . . . . . . . . . . . . . . 5-3 5.4 INmAL CONTAINMENT PRESSURE . . . . . . . . . . . . . . . . . . . . . . . . . . 5-4 i 5.5 INmAL CONTAINMENT TEMPERATURE . . . . . . . . . . . . . . . . . . . . . . . 9 t 5.6 AMBIENT HUMIDITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-12 ! 5.7 AMBIENT TEMPERATURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-15
- 5.8 SENSmVITY TO DROP MODELING ASSUMI'TIONS . . . . . . . . . . . . . . 5-20 l
5.9 CONCLUSION
S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-20 1-l 6 METEOROLOGICAL EFFECTS ON PCS PERFORMANCE
6.1 INTRODUCTION
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-1 6.2 WIND-INDUCED TURBULENCE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-1 6.2.1 Summary of Wind Tunnel Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-1 j 6.2.2 Tracking of a Wind-Driven Particle ................, ...... 6-3 6.2.3 Containment Time Constants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-3 6.2.4 Wind-Induced Oscillation Effect on Heat Transfer Coefficient . . . . . 6-5 6.2.5 WGOTHIC Evaluation Model Basis . . . . . . . . . . . . . . . . . . . . . . . . 6-5 6.3 RECIRCULATION OF CHIMNEY EFFLUENT , . . . . . . . . . . . . , . . . . . . . 7 6.3.1 Summary of Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-7 6.3.2 Evaluation of Effect of Recirculation . . . . . . . . . . . . . . . , , . . . . . . 6-8 6.3.3 - WGOTHIC Evaluation Model Basis . . . . . . . . . . . . . . . . . . . . . . . . 6-9
6.4 CONCLUSION
S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-10
6.5 REFERENCES
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-10 7 BASIS AND METHOD FOR CALCULATING THE PCS WATER EVAPORATION RATE FOR THE AP600 CONTAINMENT DBA EVALUATION MODEL
7.1 INTRODUCTION
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-1 7.2 - WATER APPLICATION AND DISTRIBUTION . . . . . . . . . . . . . . . . . . . . 7-3 .7.2.1 Containment Shell Surface Coating . . . . . .' . . . . . . . . . . . . . . . . . . 7-3 ~7.2.2 PCS Water Distribution Weir Description and Operation . . . . . . . . 7-5 7.2.3 PCS Water Distribution Testing Results . . . . . .-. . . . . . . . . . . . . . . 7 7.2.4 Delivered Water Flow Rate versus Time . . . . . . . . . . . . . . . . . . . . 7-10 7.3 AP600 WATER COVERAGE BASIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-13 7.3.1. Water Distribution Film Flow Rate, Pdist 7-14 7.3.2 Minimum Film Flow Rate, Fmin . . . . . . . . . . . . . . . . . . . . . . . . . . 7-16 WGOTHIC Application to AP600 - october 1997 - . o:\3006w-fm.non:1b.101497 - Revision 1
vi MASTER TABLE OF CONTEN'IS (Cont.) 7.4 EFFECT OF TWO-DIMENSIONAL (2 D) HEAT CONDUCTION THROUGH THE CONTAINMENT SHELL . . . . . . . . . . . . . . . . . . . . . . . 7-18 7.4.1 Geometry of the Wet and Dry Vertical Stripes on the Containment Outside Steel Surface . . . . . . . . . . . . . . . . . . . . . . . . 7-18 7.4.2 Inside and Outside Heat Transfer Boundary Conditions
.l for the Conduction Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-19
- 7.4.3 2 D Conduction (ANSYS) Model Description . . . . . . . . . . . . . . . . 7-20 7.4.4 Enhanced Evaporatica due to 2 D Conduction . . . . . . . . . . . . . . . 7-21 7.4.5 Insights from the PCS Large-Scale Testing . . . . . . . . . . . . . . . . . . 7-21 7.5 THE AP600 CONTAINMENT EVALUATION MODEL TREATMENT OF WATER COVERAGE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-31 7.5.1 PCS Film Coverage Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-31 7.5.2 EGOTHIC Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-36 7.6
SUMMARY
OF SUPPORTING TESTS AND SELECTED ANALYSIS . . . . 7-40 7.6.1 Westinghouse Wet Flat Plate Test . . . . . . . . . . . . . . . . . . . . . . . . . 7-40 7.6.2 Small-Scale Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-40 7.6.3 Large-Scale Tests (LSTs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-41 7.6.4 Estimated AP600 Range of Film Coverage Parameters . . . . . . . . . 7-50 7.6.5 AP600 Containment Shell Heatup Analysis . . . . . . . . . . . . . . . . . 7-5 2 7.7 AP600 CONTAINMENT DBA EVALUATION MODEL FILM COVERAGE SENSITIVITIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-55 7.7.1 Sensitivity of the Evaluation Model to the Input PCS Film Flow Ra te . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-55 7.7.2 Sensitivity to the Water Coverage Area . . . . . . . . . . . . . . . . . . . . 7-5 5 7.7.3 Conservatism in the Assumed Time Delay for Application o f the PCS Film . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-61
7.8 CONCLUSION
S AND
SUMMARY
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-64 7.9 NOMENCLATURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-67 7.10 REFE RENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-68 APPENDIX 7A PHYSICS OF LIQUID FILMS ON THE AP600 CONTAWMENT SHELL 8 AP600 CONTAINMENT PRESSURE SENSITIVITY DURING BLOWDOWN
8.1 INTRODUCTION
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-1 8.2 M ETH O D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-1 8.3 AN ALYSIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-2
8.4 CONCLUSION
S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-5 9 CIRCULATION AND STRATIFICATION WITHIN CONTAINMENT
9.1 INTRODUCTION
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-4 9.1.1 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-14 9.1.2 Lumped Patameter Biases and Capabilities . . . . . . . . . . . . . . . . . 9 -14 9.2 LARGE-SCALE TEST RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-17 9.2.1 LOCA Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-18 9.2.2 MSLB Configuration . . . . . . . . . . . .......................9-19 9.2.3 Method to Address Distortions in LST Stratification Data . . . . . . . 9-20 9.2.4 Application of Modeling Methods Developed for NUPEC M43 Lumped Parameter Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-21 WGOTHIC Application to AP600 october 1997 o:\3(X4w.fm.non:lo-101497 Revision 1
( , k
vil - MASTER TABLE OF CONTEN'IS (Cont.) j [O D 93 CIRCULATION AND STRATIFICATION ASSESSMENT FOR THE LOSS-OF-COOLANT ACCIDENT . . . . . . . . . , . . . . . . . . . . . . . . . . . . . . 9-26 !. 9.3.1 LOCA Break Scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-27 93.2 AP600 LOCA Evaluation Model ......,.................. 9-34 9.4 MAIN STEAMLINE BREAK (MSLB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 944 9.4.1 Break Locations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-44 4 9.4.2 AP600 MSLB Evaluation Model . . . . . . . . . . . . . . . . . . . . . . . . . . 9-45
- 9.4.3 MSLB Sensitivity Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-46
9.5 CONCLUSION
S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-48
9.6 REFERENCES
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-50 APPENDIX 9.A THERMAL AND CIRCULATION EFFECTS OF DROPS DURING A LOCA APPENDIX 9.B EFFECTS OF STRATIFICAITON ON HEAT SINK UTILIZATION APPENDIX 9.C ADDITIONAL INFORMATION ON AP600 CONTAINMENT CIRCULATION AND STRATIFICATION 10 NOMINAL INPUTS AND CORRELATIONS SENSITIVITIES
10.1 INTRODUCTION
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-1 10.2 SENSITIVITY STUDY RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-1 10.2.1 Sensitivity Study Cases for LOCA . . . . . . . . . . . . . . . . . . . . . . . . 10-5 10.2.2 Sensitivity Study Case for MSLB . . . . . . . . . . . . . . . . . . . . , . . . . 10-8 g 103 CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-8 g s
10.4 REFERENCES
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-11 a
11 TIMESTEP SEK3ITIVITY
11.1 INTRODUCTION
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 -1 11.2 M ETH ODOLOGY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-1 11.3 RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 -2 11.4
SUMMARY
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 -3
11.5 REFERENCES
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-6 12 SENSITIVITY TO CLIME NODING
12.1 INTRODUCTION
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-1 12.L1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-1 12.1.2 Problem Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 3 12.13 Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-3 12.1.4 Selected Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-4 12.1.5 Success Criteria ......................................12-4 12.2 MODEL DESCRII'TIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-5 - 12.2.1 Simple Annulus CIlme Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-5 12.2.2 AP600 Containment .Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-7 123 RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-8 123.1 Simple Annulus Clime Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-8 123.2 AP600 Containment Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-10
('-
}yGOTHIC Application to AP600 October 1997 o:\3006w-fm.non:Ib-101497 Revision 1
viil ~~ l l MASTER TABLE OF CONTENTS (Cont.) 12.4 SUMMA RY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 -12 l 12.4.1 Simple Annulus Clime Model Noding Study . . . . . . . . . . . . . . . 12-12 12.4.2 AP600 Containment Model Clime Noding Study . . . . . . . . . . . . 12-12
12.5 REFERENCES
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-12 O
O WGOTHIC Application to AP600 october 1997 o:\3006w-fm.non:1b.101497 Revision 1
;li LIST OF TABLES Table 21 Phenomena Identification and Ranking Table - Summar and Medium Ranked Phenomena - . . . . . . . . ................ . . . . . . .y of High 2-4 Table 3-1 - GOTHIC Validation Tests ; . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
35 Table 3-2 ' GOTHIC Phenomenological Models Validated by Test =. . . . . . . . . . . . . , . 3-6 Table 4-1 Central Cylindrical Room Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . 4-90 Table 4 2 Flow Paths for the Cylindrical Central Volumes . . . . . . . . . . . . . . . . . . 4-91 Table 4-3 Above-Deck East Steam Generator Compartment Parameters . . . . . . . . . 4 Table 4-4 Flow Paths for the Above-Deck East Steam Generator Volumes . . . . . . . . 4-%
. Table 4-5 Above-Deck West Steam Generator Compartment Parameters .. . . . . . . . 4-102 Table 4-6 Flow Paths for the Above-Deck West Steam Generator Volumes . . . . . . 4-103 ;
Table 4-7 South Inner Half Annulus Compartment Parameters . . . . . . . . . . . ._. . . 4-109 - Table 4-8 Flow Paths for the South Inner-Half Annulus Volumes . . . . . . . . . . . . . 4-110 Table 4-9 North Inner-Half Annulus Compartment Parameters . . . . . . . . . . . . . . . . ' 4-116 Table 4-10 Flow Paths for the North Inner-Half Annulus Volumes - . . . . . . . . . . . . . . 4-117 Table 4-11 North Mid-Quarter Annulus Compartment Parameters . . . . . . . . . . . . . 4-124 Table 4-12 Flow Paths for the North Mid-Quarter Annulus Volumes . . . , . . . . . . . 4-125 , Table 4-13 West Mid-Quarter Annulus Compartment Parameters . . . . . . . . . . . . . . 4-132 Table 4-14 Flow Paths for the West Mid-Quarter Annulus Volumes . . . . . . . . . . . . 4-133 Table 4-15 South Mid-Quarter Annulus Compartment Parameters . . . . . . . . . . . . . 4-140 Table 4-16 Flow Paths for the South Mid-Quarter Annulus Volumes . . . . . . . . . . . 4-146 Table 4-17 East Mid-Quarter Annulus Compartment Parameters . . . . . . . . . . . . . . 4-147 Table 4-18 Flow Paths for the East Mid-Quarter Annulus Volumes . . . . . . . . . . . . 4-148 - Table 4-19 North Outer Quarter Annulus Compartment Parameters . . . . . . . . . . . .- 4-154 Table 4-20 Flow Paths for the North Outer Quarter Annulus Volumes . . . . . . . . . . 4-156 O. Table 4-21 Table 4-22 West Outer Quarter Annulus Compartment Parameters . . . . . . . . . . . . 4-162 Flow Paths for the West Outer Quarter Annulus Volumer . . . . . . . . . . . 4-163 Table 4-23 South Outer Quarter Annulus Compartments Parameters . . . . . . . . . . . 4-169. Table 4-24 Flow Paths for the South Outer Quarter Annulus Volumes . . . . . . . . . . 4-170 Table 4-25 East Outer Quarter Annulus Compartment Parameters l . . . . . . . . . . . . . 4-176 Table 4-26 Flow Paths for the East Outer Quarter Annulus Volumes . . . . . . . . . . . 4-177 Table _4 _ East Quarter Inner Dome Compartment Parameters - . . . . . . . . . . . . . . . 4-182 Table 4-28 Flow Paths for the East Quarter Inner Dome Volumes . . . . . . . . . . . . . . 4-183
- Table 4 North Quarter Inner Dome Compartment Parameters . . . . . . . . . . . . . . 4-187 '
Table 4 Flow Paths for the North Quarter Inner Dome Volumes . . . . . . . . . . . . 4-187 Table 4-31 West Quarter Inner Dome Compartment Parameters . . . . . . . . . . . . . . . 4-190 Table 4-32 Flow Paths for the West Quarter Inner Dome Volumes , . . . . . . . . . . . . 4-193 Table 4-33 South Quarter Inner Dome Compartment Parameters . . . . . . . . . . . . . . 4-195 Table 4 34 ~ Flow Paths for the South Quarter Inner Dome Volumes . . . . . . . . . . . . 4-1% Table 4-35 East Quarter Outer Dome Compartment Parameters . . . . . . . . . . . . . . . 4-199 Table 4 Flow Paths for the East Quarter Outer Dome Volumes . . . . . . . . . . . . . 4-200 Table 4-37 North Quarter Outer Dome Compartment Parameters . _ , . . . . . . . . . . 4-203 Table 4-38' Flow Paths for the North Quarter Outer Dcme Volumes . . . . . . . . . . . . . ~4-204 Table 4 West Quarter Outer Dome Compartment Parameters . . . . . . . . . . . . . . 4-207. . Table 4-40 Flow Paths for the West Quarter Outer Dome Volumes ... . . . . . . . . . . . 4-208 Table 4-41 South Quarter Outer Dome Compartment Parameters . . . . . . . . . . . . . . 4-211 Table 4-42 Flow Paths for the South Quarter Outer Dome Volumes . . . . . . . . . . . . 4-212 O IGOTHIC Application to AP600 october 1997 o:\XJ6w fm.noruit>101497 Revision 1
X LIST OF TABLES (Cont.) Table 4-43 Outside Containment Upflow Annulus Volume Parameters . . . . . . . . . . 4-215 ! Table 4-44 Flow Paths for the Outside Containment Upflow Annulus . . . . . . . . . . 4-216 < Table 4-45 Outside Containment Downflow Annulus Volume Parameters . . . . . . . 4-220 Table 4-46 Flow Paths for the Outside Containment Downflow Annulus . . . . . . . . 4-221 Table 4-47 PCS Chimney Volume Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-224 Table 4-48 Flow Paths for the PCS Chimney . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-224 Table 4-49 Conductor Material Properties . . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . . 4-227 Table 4-50 Conductor 1 - 0.1" Thick Carbon Steel . . . . . . . . . . . . . . . . . . . . . . . . . . 4-227 Table 4-51 Conductor 2 - 0.03" Thick Carbon Steel . . . . . . . . . . . . . . . . . . . . . . . . . 4-229 Table 4-52 Conductor 3 - 0.06" Thick Carbon Steel . . . . . . . . . . . . . . . . . . . . . . . . . 4-229 Table 4-53 Conductor 4 - 0.105" Thick Carbon Steel . . . . . . . . . . . . . . . . . . . . . . . . 4-229 Table 4-54 Conductor 5 - 0.1007" Thick Carbon Steel . . . . . . . . . . . . . . . . . . . . . . . 4-229 Table 4-55 Conductor 6 - 0.1N4" Thick Carbon Steel . . . . . . . . . . . . . . . . . . . . . . . 4-230 Table 4-56 Conductor 7 - 0.158" Thick Carbon Steel . . . . . . . . . . . . . . . . . . . . . . . 4 230 Table 4-57 Conductor 8 - 0.132" Thick Carbon Steel ....................... 4-230 Table 4-58 Conductor 9 - 0.210" Thick Carbon Steel ....................... 4-230 Table 4-59 Conductor 10 - 0.2448" Thick Carbon Steel . . . . . . . . . . . . . . . . . . . . . . 4-231 Table 4-60 Conductor 11 - 0.252" Thick Stainless Steel ..................... 4-231 Table 4-61 Conductor 12 - 0.252" Thick Carbon Steel ......................,4-231 Table 4-62 Conductor 13 - 0.284" Thick Carbon Steel ...................... 4-231 Table 4-63 Conductor 14 - 0.404" Thick Carbon Steel ...................... 4-232 1 Table 4-64 Conductor 15 - 0.504" Thick Stainless Steel ..................... 4-232 Table 4-65 Conductor 16 - 0.854" Thick Carbon Steel ........ . . . . . . . . . . . . . 4-232 ! Table 4-66 Conductor 17 - 0.996" Thick Stainless Steel ..................... 4-232 Table 4-67 Table 4-68 Conductor 18 - 0.81" Thick Carbon Steel ....................... 4-233 Conductor 19 - 1.57" Thick Carbon Steel ....................... 4-233 h Table 4-69 Conductor 20 - 1.63" Thick Carbon Steel ....................... 4-233 Table 4-70 Conductor 21 - 1.754" Thick Carbon Steel ...................... 4-233 Table 4-71 Conductor ?.2 - 1.998" Thick Carbon Steel . . . . . . . . . . . . . . . . . . . . . . 4-234 Table 4-72 Conductor 23 - 3.014" Thick Carbon Steel ...................... 4-234 Table 4-73 Conductor 24 - 4.399" Thick Stainless Steel ..................... 4-234 Table 4-74 Conductor 25 - 4.874" Thick Carbon Steel ...................... 4-234 Table 4-75 Conductor 26 - 3.0" Thick Stainless Steel ....................... 4-235 Table 4-76 Conductor 27 - 25.04" Thick Concrete .......... . . . . . . . . . . . . . . 4-235 Table 4-77 Conductor 28 - 25.0" Thick Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . 4 -235 Table 4-78 Conductor 29 - 25.0" Thick Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . 4 -236 Table 4-79 Conducter 30 - 36.0" Thick Concrete ................... . . . . . . 4-236 Table 4-80 Conductor 31 - 25.0" Thick Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . 4-23 6 Table 4-81 Conductor 32 - 24.5" Thick Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . 4 -23 7 Table 4-82 Conductor 33 - 24.5" Thick Concrete .......................... 4-237 Table 4-83 Conductor 34 - 24.5" Thick Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . 4 -237 Table 4 84 Conductor 35 - 24.5" Thick Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . 4-23 8 Table 4-85 Conductor 36 - 24.0" Thick Concrete .......................... 4-238 Table 4-86 Conductor 37 - 49.0" Thick Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . 4 -238 l Table 4-87 Conductor 38 - 48.5" Thick Concrete .......................... 4-239 i e
}VGOTHIC Application to AP600 october 1997 o:\3006w.fm.non:lb-101497 Revtsion 1
xi LIST OF TABLES (Cont.)
' Table 4-88 Conductor 39 -'48.5" Thick Concrete .......................... 4-239 Table 4 Conductor 40 - 48.5" Thick Carbon Steel ....................... 4-239 Table 4-90 Conductor 41 - 49.0" Thick Concrete ..........................,4-240 Table 4 Conductor 42 - 48.0" Thick Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . 4-2 40 Table 4-92 Conductor 43 - 48.5" Thick Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . 4-2 40 Table 4-93 Conductor 44 - 48.5" Thick Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-241 Table 4-94 Conductor 45 - 48.0" Thick Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . 4-2 41 Table 4-95 Conductor 46 - 49.0" Thick Concrete ' . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-241 Table 4-% Conductor 47 - 49.0" Thick Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . 4-2 42 Table 4-97 Conductor 48 - 49.0" Thick Concrete .......................... 4-242 Table 4-98 Conductor 49 - 25.0" Thick Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . 4-2 42 Table 4-99 Conductor 50 - 48.5" Thick Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . 4-2 43 Table 4-100 Conductor 51 - 2.0" Thick Carbon Steel . . . . . . . . . . . . . . . . . . . . . . . . 4-2 4 3 Table 4-101 Conductor 52 - Dummy Baffle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-243 Table 4-102 Evaluation Model Function for the AP600 PCS Film Flow . . . . . . . . . . . 4-272 Table 4-103 Outside Containment Initial Conditions . . . . . . . . . . . . . . . . . . . . . . . . . 4-275 Table 4-104 Inside Containment Initial Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . 4 276 Table 4-105 LOCA Boundary Conditions Represented by Forcing Functions in the AP600 Evaluation Model . . . . . . . . . . . . . , . . . . . . . . . . . . . . . . . 4-277 Table 4-106 MSLB Boundary Conditions Represented by Forcing Functions in the AP600 Evaluation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-277 Table 4-107 Basis for Gwmetric Inputs ..................................4-294 Table 4-108 Basis for LOCA and MSLB Mass and Energy Inputs . . . . . . . . . . . . . . . 4-295 Table 4-109 Basis for PCS Inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-2%
O Table 4-110 Table 4-111 Summary of Containment Volumes and Heat Sinks Model Node Listing , 4-298 Master List of Flow Paths in AP600 Evaluation Model . . . . . . . . . . . . . . 4-300 Table 5-1 Initial Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1 Table 5-2 Initial Conditions Sensitivity Analysis Cases . . . . . . . . . . . . . . . . . . . . . . 5-2 Table 5-3 Summary of Pressure Results for LOCA Initial Condition Sensitivity Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-3 Table 5-4 Summary of Pressure Results for MSLB Initial Condition Sensitivity Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-3 Table 7-1 Ranges of the Film Coverage Parameters in the PCS Tests . . . . . . . . . . . . . 7-4 Table 7-2 Summary of the Phase 3 Water Distribution Test . . . . . . . . . . . . . . . . . . . . 7-9
' Table 7 PCS Time Sequence of Events (Based on 440 gpm flow rate) . . . . . . . . . . 7-13 Table 7 4 - Elevation D Heat Flux Comparison From PCS Large-Scale Tests RC048C and RC050C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-29 Table 7-5 Elevation E Heat Flux Comparison From PCS Large-Scale Tests RC048C and RC050C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-30 Table 7-6 Clime Wetted Perimeter and Basis for EGOTHIC Model . . . . . . . . . . . . 7-37 Table 7-7 Summary of STC Heated Flat Plate Tests . . . . . . . . . . . . . . . . . . . . . . . . . 7-42 Table 7-8 Summary of Small-Scale Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-44 Table 7-9' - - Summary of Large-Scale Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-48 Table 7-10 AP600 Shell Temperature and Outside Heat Flux . . . . . . . . . . . . . . . . . . 7-51 Table 7-11 Comparison of the Range of Film Coverage Parameters . . . . . . . . . . . . . . 7 Table 7-12 Transient Dry Shell Temperature Increase . . . . . . . . . . . . . . . . . . . . . . . . 7 Table 9-1 Circulation and Stratification Evaluation Summary . . . . . . . . . . . . . . . . . 9-5 WGOTHIC Application to AP600 October 1997 o \3006w-fm.non:Ib-101497 Revision 1
xil . - ~ . LIST OF TABLES (Cont.) Table 9 2 Flow Areas Connecting to North and South CMT Compartments (excluding Dead Ended Compartment Connections) . . . . . . . . . . . . . . . . 9 24 Table 10-1 Nominal inputs and Correlations Sensitivity Results . . . . . . . . . . . . . , , . 10-2 Table 11-1 Timestep Sensitivity Results ..................................112 Table 11-2 Timestep Limit Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-3 Table 12-1 Input Parameters for Annulus Clime Model Sensitivity Study . . . . . . . . . 12-6 Table 12-2 Predicted Heat Removal Rates for Various Clime Noding Schemes . . . . . 12-9 O I l I O l lyGOTHIC Application to AP600 october 1997 o:\3006w-fm.non:1b-101497 Revision 1 l
xill LIST OF FIGURES p) (v Figure 1-1 Relationship of AP600 Containment DBA Reports . . . . . . . . . . . . . . . . . . . 1-2 Figure 2-1 PCS Test and Analysis Process Overview . . . . . . . . . . . . . . . . . . . . . . . . . 2-2 l Figure 3-1 Summary of GOTHIC Historical Development ..................... 3-2 Figure 3-2 GOTHIC Modeling Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-3 Figure 3-3 Summary of MGOTHIC Historical Development . . . . . . . . . . . . . . . . . . . . 3-7 Figure 3-4 Westinghouse-GOTHIC Clime Wall Source Term Models . . . . . . . . . . . . 3-12 Figure 3-5 Clime Finite Difference Model Definitions . . . . . . . . . . . . . . . . . . . . . . . . 3-14 Figure 3-6 Clime Routines Flow Control Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-19 Figure 4-1 Cutaway View of the 100-foot Elevation of the AP600 Including Major Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-5 Figure 4-2 Cutaway View of the 100-foot Elevation of the AP600 Excluding Major Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-6 Figure 4-3 Rea ct or Ca vi ty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-7 Figure 4-4 Reactor Cavity Cross-Section - AA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-8 Figure 4-5 Reactor Cavity Cross-Section - BB . . . . . . . . . ....................4-9 Figure 4-6 Reactor Coolant Drain Tank Cavity . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-14 Figure 4-7 Reactor Coolant Drain Tank Cross-Section - AA . . . . . . . . . . . . . . . . . . . 4-15 Figure 4-8 Southeast Accumulator Cavity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-20 Figure 4-9 Southeast Accumulator Cavity Cross-Section - BB . . . . . . . . . . . . . . . . . . 4-21 Figure 4-10 q Northeast Accumulator Cavity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 26 Q Figure 4-11 Northeast Accumulator Cavity Cross-Section - BB . . . . . . . . . . . . . . . . . . 4-27 Figure 4-12 East Steam Generator Cavity ................................. 4-31 Figure 4-1? East Steam Generator Cavity Cross-Section - AA . . . . . . . . . . . . . , . . . . 4-32 Figure 4-14 East Steam Generator Cavity Cross-Section - BB . . . . .............. 4-33 Figure 4-15 West Steam Generator Cavity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-41 Figure 4-16 West Steam Generator Cavity Cross-Section - AA . . . . . . . . . . . . . . . . . . 4-42 Figure 4-17 West Steam Generator Cavity Cross-Section - BB . . . . . . . . . . . . . . . . , . 4-43 Figure 4-18 North CMT Cavity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-49 Figure 4-19 North CMT Cavity Cross-Section - AA . . . . . . . . . . . . . . . . . . . . . . . . . . 4-50 Figure 4-20 North CMT Cavity Cross-Section - BB . . . . . . . . . . . . . . . . . . . . . . . . . 4-51 Figure 4-21 Sou th CMT Cavity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-62 Figure 4-22 South CMT Cavity Cross-Section - AA . . . . . . . . . . . . . . . . . . . . . . . . . . 4-63 Figure 4-23 South CMT Cavity Cross-Section - BB . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-64 Figure 4-24 Chemical and Volume Control Cavity . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-71 Figure 4-25 Chemical and Volume Control Cavity Cross-Section - AA , . . . . . . . . . . . 4-72 Figure 4-26 Refueling Canal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-76 Figure 4-27 Refueling Canal Cross-Section - AA . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-77 Figure 4-28 Internal Refueling Water Storage Tank (IRWST) . . . . . . . . . . . . . . . . . . . 4-83 Figure 4-29 IRWST Cross-Section - AA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-84 p) ( v EGOTHIC Application to AP600 octreer 1997 o:\3006w-fm.non:lb-101497 Revision 1
(- { xiv LIST OF FIGURES (Cont.) Figure 4-30 South Inner-Half Annulus Compartment . . . . . . . . . . . . . . . . . . . . . . . . . 4-92
, Figure 4-31 Cylindrical Central Room Cross-Section - BB . . . . . . . . . . . . . . . . . . . . . 4 93 Figure 4-32 Cylindrical Central Room Cross-Section - AA . . . . . . . . . . . . . . . . . . . . . 4-94 ;
Figure 4-33 Above-Deck East Steam Generator Compartment . . . . . . . . . . . . . . . . . . 4-97 Figure 4-34 Above-Deck East Steam Generator Room Cross-Section - BB . . . . . . . . . . 4-98 Figure 4-35 Above-Deck East Steam Generator Room Cross-Section - AA . . . . . . . . 4-99 Figure 4-36 Above-Deck West Steam Generator Compartment . . . . . . . . . . . . . . . . . 4-104 Figure 4-37 Above-Deck West Steam Generator Compartment Cross-Section - BC . . 4-105 Figui.: 4-38 Above-Deck West Steam Generator Compartment Cross-Section - AA . 4-106 Figure 4-39 South Inner-Half Annulus Compartment . . . . . . . . . . . . . . . . . . . . . . . . 4-111 Figure 4-40 South Inner-Half Annulus Cross-Section - BB . . . . . . . . . . . . . . . . . . . 4-112 Figure 4-41 South Inner-Half Annulus Cross-Section - AA . . . . . . . . . . . . . . . . . . . 4-113 Figure 4-42 North Inner-Half Annulus Compartment . . . . . . . . . . . . . . . . . . . . . . . . 4-118 Figure 4-43 North Inner-Half Annulus Cross-Section - BB . . . . . . . . . . . . . . . . . . . . 4-119 Figure 4-44 North Inner Half Annulus Cross-Section - AA . . . . . . . . . . . . . . . . . . 4-120 Figure 4-45 North Mid-Quarter Annulus Compartment . . . . . . . . . . . . . . . . . . . . . , 4-126 Figure 4-46 North Mid-Quarter Annulus Cross-Section - BB . . . . . . . . . . . . . . . . . . 4-127 Figure 4-47 North Mid-Quarter Annulus Cross-Section - AA . . . . . . . . . . . . . . . . . 4-128 Figure 4-48 West Mid-Quarter Annulus Compartment . . . . . . . . . . . . . . . . . . . . . . . 4-134 Figure 4-49 West Mid-Quarter Annulus Cross-Section - BB . . . . . . . . . . . . . . . . . . . 4-135 Figure 4-50 West Mid Quarter Annulus Cross-Section - AA . . . . . . . . . . . . . . . . . . 4-136 Figure 4-51 South Mid-Quarter Annulus Compartment . . . . . . . . . . . . . . . . . . . . . . 4-141 Figure 4-52 South Mid-Quarter Annulus Cross-Section - BB . . . . . . . . . . . . . . . . . . 4-142 l Figure 4-53 South Mid-Quarter Annulus Cross-Section - AA . . . . . . . . . . . . . . . . . . 4-143 Figure 4-54 East Mid-Quarter Annulus Compartment . . . . . . . . . . . . . . . . . . . . . . . 4 - 14 9 Figure 4-55 East Mid-Quarter Annulus Cross-Section - BB . . . . . . . . . . . . . . . . . . . . 4-150 Figure 4-56 East Mid-Quarter Annulus Cross-Section - AA . . . . . . . . . . . . . . . . . . . 4-151 Figure 4-57 North Outer Quarter Annulus Cross-Section . . . . . . . . . . . . . . . . . . . . . 4-157 Figure 4-58 North Outer Quarter Annulus Cross-Section - BB . . . . . . . . . . . . . . . . . 4-158 Figure 4-59 North Oute. Quarter Annulus Cross-Section - AA . . . . . . . . . . . . . . . . 4-159 Figure 4-60 West Outer Quarter Annulus Compartment . . . . . . . . . . . . . . . . . . . . . 4 - 164 l Figure 4-61 West Outer Quarter Annulus Cross-Section - BB . . . . . . . . . . . . . . . . . . 4-165 l Figure 4-62 West Outer Quarter Annulus Cross-Section - AA . . . . . . . . . . . . . . . . . 4-166 ( Figure 4-63 South Outer Quarter Annulus Compartment . . . . . . . . . . . . . . . . . . . . . 4-171 Figure 4-64 South Outer Quarter Annulus Cross-Section - BB . . . . . . . . . . . . . . . . . 4-172 Figure 4-65 South Outer Quarter Annulus Cross-Section - AA . . . . . . . . . . . . . . . . 4-173 Figure 4-66 East Outer Quarter Annulus Compartment . . . . . . . . . . . . . . . . . . . . . . 4-178 l l 9 EGOTHIC Application to AP(IJO october 1997 c:\3006w4m.non:1b-101497 Revision 1
i
-xv LIST OF FIGURES (Cont.)
Figure 4 East Outer Quarter Annulus Cross-Section - bB . . . . . . . . . . , , . , , . . . 4-179-Figure 4-68 East Outer Quarter Annulus Cross-Section - AA . . . . . . , . . . . . . . . . . . 4-180 - Figure 4-69 East Quarter Inner Dome Cross-Section - BB , . . . . . . . . . . , . . . . . . . . 4-184 Figure 4-70 Eaat Quarter Inner Dome Cross-Section - AA . . . . . . . . . . . . . . . . . . . . 4185 Figure 4 North Quarter Inner Dome Cross-Section - BB . . . . . . . . . . . . . . . . . . . 4-188 Figure 4-72 North Quarter Inner Dome Cross-Section - AA . . . . . . . . . . . . . . . . . . . 4189 Figure 4-73 West Quarter Inner Dome Cross-Section - BB . . . . . . . . . . . . . . . . . . . . 4-191 Figure 4-74 West Quarter Inner Dome Cross-Section - AA ................... 4-194 Figure 4-75 South Quarter Inner Dome Cross Section - BB . . . . . . . . . . . . . . . . . . . 4-197 Figure 4-76 South Quarter Inner Dome Cross Section - AA . . . . . . . . . . . . . . . . . . . 4-198 Figure 4-77 East Quarter Outer Dome Cross-Section - BB . . . . . . . . . . . . . . . . . . . . 4-201 Figure 4-78 East Quarter Outer Dame Cross-Section - AA . . . . . . . . , . . . . . . . . . . . 4-202 Figure 4 79 North Quarter Outer Dome Cross-Section - BB . . . . . . . . . . . . . . . . . . 4-205 Figure 4-80 North Quarter Outer Dome Cross-Section - AA . . . . . . . . . . . . . . . . . . 4-206 Figure 4-81 West Quarter Outer Dome Cross-Section - BB . . . . . . . . . . . . . . . . . . . . 4-209 Figure 4-82 West Quarter Outer Dome Cross-Section - AA . . . . . . . . . . . . . . . . . . 4-210 Figure 4-83 South Quarter Outer Dome Cross-Section - BB . . . . . . . . . . . . . . . . . . 4-213 Figure 4-84 South Quarter Outer Dome Cross Section - AA . . . . . . . . . . . . . . . . . . 4-214 Figure 4-85 Outside Containment Upflow Annulus Cross-Section - BB . . . . . . , . . 4-217 Figure 4-86 Outside Containment Downflow Annulus Cross-Sectior. - BB . . . . . . . . 4-222 Figure 4-87 PCS Chimney Cross-Section - BB - . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-225 Figure 4-88 WGOTHIC Conductor Noding Example . . . . . . . . . . . . . . . . . . . . . . . . 4-228 Figure 4-89 Wet / Dry Clime Modeling Diagram . . . . . . . . . , , . . . . . . . . . . . . . . . . 4-245 Figure 4-90 Containment , Baffle, Shield Building Modeling with Climes . . . . . . . . . 4-246 Figure 4-91 Wet Stack Clime Numbering Diagram . . . . . . . . . . . . . . , , . . . . . . . . . . 4-247 Figure 4-92 Dry Stack Clime Numbering Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . 4-261 Figure 4-93 AP600 Evaluation Model LOCA Function 3: PCS Evaporation Limited Fihn Flow Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-273 Figure 4-94 AP600 Evaluation Model LOCA Function 2: IRWST Drain Rate .......................,........................... 4-278-Figure 4-95 AP600 Evaluation Model LOCA Function 4: Liquid Drop
- Mot el Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-279 Figure 4-% AP600 Evaluation Model LOCA Function 5: Liquid Mass Break Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . 4-280 Figure 4-97 . AP600 Evaluation Model LOCA Function 6: Liquid Break Enthalpy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-281 Figure 4-98 AP600 Evaluation Model LOCA Function 7: Steam Mass Break Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-282 WGOTHIC Application to AP600 october 1997 o:\3006w fm.non:1M01497 Revision 1
xvi I LIST OF FIGURES (Cont.) Figure 4-99 AP600 Evaluation Model LOCA Function 8: Steam Break Enthalpy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-283 g Figure 4-100 AP600 Evaluation Model LOCA Functim 11: Thermal Conductor Heat Transfer Coefficient Control . . . . . . . . . . . . . . , . . . . . . 4-284 Figure 4-101 AP600 Evaluation Model LOCA Function 12: Steam Mass ADS Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-285 Figure 4102 AP600 Evaluation Model LOCA Function 13: PCS Annular Sump Drain Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-286 Figure 4-104 AP600 Evaluation Model MSLB Function 12: Steam Break Enthalpy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-288 Figure 4-105 AP600 Evalustion MSLB Function 13: Short-Term MSLB Steam Mass Break Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-289 Figure 4-106 AP600 Evaluation Model MSLB Function 13: Long-Term MSLB Steam Mass Break Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-290 Figure 4-107 AP600 WGOTHIC Evaluation Model Section B-B . . . . . . . . . . . . . . . . . . 4-310 Figure 4-108 AP600 EGOTHIC Evaluation Model Section A-A . . . . . . . . . . . . . . . . 4-311 Figure 4-109 AP600 EGOTHIC Evaluation Model 135'-3" Elevation . . . . . . . . . . . . . . 4-312 Figure 4-110 AP600 WGOTHIC Evaluation Model 148'-0" Elevation . . . . . . . . . . . . . . 4-313 Figure 4-111 AP600 EGOTHIC Evaluation Model 170'-0" Elevation . . . . . . . . . . . . . . 4-314 Figure 4-112 AP600 EGOTHIC Evaluation Model 189'-6" Elevation . . . . . . . . . . . . . . 4-315 Figure 4-113 AP600 EGOTHIC Evaluation Model 209'-0" Elevation . . . . . . . . . . . . . . 4-316 Figure 4-114 AP600 EGOTHIC Evaluation Model 224*-9" Elevation . . . . . . . . . . . . . . 4-317 Figure 4-115 AP600 WGOTHIC Evaluation Model 240'-6" Elevation . . . . . . . . . . . . . . 4-318 Figure 4-116 AP600 WGOTHIC Evaluation Model 254'-4" Elevatien . . . . . . . . . . . . . . 4-328 Figure 4-117 Below Deck Compartments and Flowpaths . . . . . . . . . . . . . . . . . . . . . . 4-329 Figure 4118a LOCA Break Pools Below the CMT Room Floor (107'-2") . . . . . . . . . . . . 4-330 Figure 4-118b LOCA Break Pools in the Accumulator and CVS Cavities . . . . . . . . . . . 4-331 Figure 4-119 LOCA Below-Deck Water Levels Above 107'-2" . . . . . . . . . . . . . . . . . . . 4-332 Figure 4-120 LOCA IRWST Water Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-333 Figure 4-121a LOCA Pressure History: Short-Term . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-334 Figure 4-121b LOCA Pressure History: Long-Term . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-335 Figure 4-122 LOCA Steam Concentration for East Quarter Compa rtments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-336 Figure 4-123 LOCA Steam Concentration for West Quarter Compartments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-337 Figure 4-124 LOCA Temperature Distribution for East Quarter Compartments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-338 Figure 4-125 LOCA Temperature Distribution for West Quz er Compartments . . . . . . . . . . . . . . . . . ...... ................... 4-339 O
}VGOTHIC Application to AP600 october 1997 o:\3006w fm.nortib-101497 Revision 1
xvil LIST OF FIGURES (Cont.) 4 : Figure 4-126a Temperature Distribution for Central Cylindrical Regions . . . . . . . . . . . 4-340-
- Figure 4-126b Temperature Distribution for the Dome Regions . . . . . . . . . . . . . . . . . . 4-341 . ' Figure 4-127 Air Temperature Distribution in the PCS Annulus - LOCA Event . . . . . 4-342 Figure 4-128 Air Flow Through the PCS Annulus - LOCA Event . . . . . . . . . . . . . . . . 4-343 l Figure 4-129a LOCA Gas Circulation Pattern in the Upper East Steam 2 Generator Compartment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 344 - Figure 4-129b LOCA Gas Circulation Pattem in the East Outer g Quarter Compartments , . -. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-345 !
Figure 4-130a LOCA Gas Circulation Pattem in the Upper. West Steam Generator Compartment . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . 4-346 Figure 4130b LOCA Gas Circulation Pattern in the West Outer Quarter Compartments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-347 Figure 4-131 LOCA Gas Circulation Pattern Across the Operating Deck . . . . . . . . . . 4-348 Figure 4-132 LOCA Steam Pressure Ratios Between the CMT Room Floor and the Operating Deck . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-349
- Figure 4-133 LOCA Steam Pressure Ratios at the Operating Deck Inside the 50.975' Radius . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-350 Figure 4-134 MSLB Temperature Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-351 Figure 4-135 MSLB Pressure History . . . . . . . . . . . . . . . , . . . . . . . . . . . .. . . . . . . . . 4-352 Figure 5-1 Case 1 - Initial Containment Humidity Sensitivity ~- LOCA . . . . . . . . . . . 5-5 Figure 5-2 Case 8 - Initial Containment Humidity Sensitivity - MSLB . . . . . . . . . . .. . 5-6 Figure 5-3 Case 2 - Initial Containment Pressure Sensitivity - LOCA . . . . . . . . . . . . 5-7 Figure 5-4 Case 9 - Initial Containment Pressure Sensitivity - MSLB . . . . . . . . . . , . . . 5-8 Figure 5-5 Case 3 - Initial Containment Temperature Sensitivity - LOCA . . . . . . . . . 5-10 Figure 5-6 Case 10 - Initial Containment Temperature Sensitivity - MSLB . . . . . . . . 5 Figure 5-7 Case 4 - Ambient Humidity Sensitivity at 115'F Ambient Temperature - LOCA . . . . . . . . . . . . . . . . . _ . . . . . . . . . . . . . . . . . . . . . . 5-13 Figure 5-8 Case 11 - Ambient Humidity Sensitivity at 115'F Ambient Tempera ture - MSLB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-14 - Figure 5 Case 5 Ambient Temperature Sensitivity - LOCA . . . . . .. . . . . . . . . . . . 5-16 Figure 5-10 Case 12 - Ambient Temperature Sensitivity - MSLB . . . . . . . . . . . . . . . .. . 5 - Figure 5-11 Case 7 - Film Temperature Sensitivity - LOCA . . . . . . . . . . . . . . . . . . . . 5-18 4 Figure 5-12 = Case 14 - Film Temperature Sensitivity - MSLB . . . . . . . . . . . . . . . . . . . 5-19 Figure 5-13 : Blowdown Drop Fraction Sensitivity.- LOCA . . . . . . . . . , . . . . . . . . . . . 5-21 Figure 6 Particle Path Through AP600 PCS With and Without Wind . . . . . . . . . . . 6-4 Figure 6-2 1D Containment Shell Model Inside Temperature Results . . . . . . . . . . . . . 6-6 Figure 7-1 Illustration of PCS Water Distribution Weir Assembly . . . . . . . . . . . . . . . 7 Figure 7-2 Conservative Estimate of the Gravity-Driven PCS Flow Rate . . . . . . . . . . 7-11 > Figure 7-3 Weir Outflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-12 WGOTHIC Application to AP600.- October 1797 c:\3006w-fm.non:1b-101497 Revision 1
xylii LIST OF FIGURES (Cont.) F.gure 7-4 Comparison of Water Distribution Model to Phase 3 Water ' Distribution Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-15 Figure 7-5 Determmation of Gamma-Min from LST, SST, and Flat Plate Data . . . . . 7-17 Figure 7-6 Normalized Water Evaporation Rate (2-D/1D Conduction) versus Overall Containment Wetted Fraction . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-22 Figure 7-7 Containment Steel Shell Temperature Gradients ('F) with 2-D Heat Conduction; 20 psig,25% Wetteo .............................. 7-23 2 Figure 7-8 Containment Steel Shell Thermal Flux Gradients (Brufnt-ft ) in Y-Direction; 20 psig, 25% Wetted . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-24 I Figure 7-9 Containment Steel Shell Total Thermal Flux (Bru/hr-ft 2); 20 psig, I 25% We t t ed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-25 Figure 7-10 Thermal Flux in Y-Direction on Sutside Surface of Containment Wall 2 [ B tu /hr-r t ] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-26 Figure 7-11 Thermal Flux in Y-direction on Inside Surface of Containment 2 Wall [ B ru / hr-ft ] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-27 Figure 7-12 Large-Scale Test Water Coverage Pattern . . . . . . . . . . . . . . . . . . . . . . . . . 7-46 Figure 7-13 Sensitivity to the Input PCS Film Flow Rate ...................... 7-56 Figure 7-14 Comparison of Peak Containment Pressure as Function of PCS Coverage Area ............................................7-58 Figure 7-15 PCS Runoff Flow Rates as a Function of Coverage Area . . . . . . . . . . . . . 7-59 Figure 7-16 Comparison of Evaporation Model Peak Pressure with 100% and 50% Constant Coverage Models ...............................7-60 Figure 7-17 Comparison of Evaporation Model PCS Runoff Flow Rate with l 100% and 50% Constant Coverage Models ............,.,........7-62 Figure 7-18 Difference in the Integrated Energy Transferred to the Top of the Dome (with PCS Film Applied at 35 and 337 Seconds) . . . . . . . . . . . . . . 7-63 Figure 8-1 Comparison of Single-Node Model with EM Pressure Curve . . . . . . . . . . . 8-3 Figu-e 8-2 Comparisca of Response with No Heat Sinks to EM Response . . . . . . . . . 8-4 Figure 9-1 Measured Steam Concentrations for LST . . . . . . . . . . . . . . . . . . . . . . . . . 9-52 Figure 9-2 LST with Diffuser Under Steam Generator - Steam Flow 0.11-0.17 lb/sec - Internal Fluid Temperature - Group 1 .. ........ 9-53 Figure 9-3 LST with Diffuser Under Steam Generator - Steam Flow 0.11-0.17 lb/sec Saturation Temperature - Group 1 . . . . . . . . . . . . . . . . 9-54 Figure 9-4 LST with Diffuser Under Steam Generator - Steam Flow 0.11-0.17 lb/sec - Internal Steam Pressure Ratio - Group 1. . . . . . . . . . . . 9-55 Figure 9-5 LST with Diffuscr Under Steam Generator - Steam Flow 0.11-0.17 lb/sec - Internal Fluid Temperature - Group 2 . . . . . . . . . . . . . 9-56 cigure 9-6 LST with Diffuser Under Steam Generator - Steam Flow 0.11-0.17 lb/sec - Saturation Temperature - Group 2 . . . . . . . . . . . . . . . . 9-57 O lyGOTHIC Application to AP600 October 1997 a:\3006w-fm.non:Ib-101497 Revtsion 1
xh LISF OF FIGURES (Cont.) O riS< 9-7: 'st -i'a oi'< u a < s' - c * < - s' - rie-: 0.11-0.17 lb/see - Internal Steam Pressure Ratio - Group 2 . . . 1. ...... 9-58 Figure 9-8 . LST with Diffuser Under Stehm Generator - Steam Flow 0.27-0.36 lb/sec - Intemal Fluid Temperature . . . . . . . . . . . . . . . . . . . . . . 9-59 Figure 9-9 : LST wtth Diffuser Under Steam Generator - Steam Flow 0.27-0.36 lb/sec - Saturation Temperature . . . . . . . . . . . . . . . , . . . . . . . 940 Figure 9-10 LST with Diffuser Under Steam Generator - Steam Flow 0.27-0.36 lb/sec - Internal Steam Pressure Ratio . . . . . . . . . . . . . . . . . . . . 941 Figure 9-11 LST with Diffuser Under Steam Generator - Steam Flow 0.49-0.62 lb/sec - Internal Fluid Temperature . Group 1 . . . . . . . . . . . . . 942 - Figure 9-12 LST with Diffuser Under Steam Generator - Steam clow 0.49-0.26 lb/sec - Saturation Temperature - Group 1 . . . . . . . . . . . . . . . . . 943
- Figure 9-13 LST with Diffuser Under Steam Generator - Steam Flow L 0,49-0.62 lb/sec - Internal Steam Pressure Ratio - Group 1. . . . . . . . . . . . 944 .
- Figure 9-14 LST with Diffuser Under Steam Generator - Steam Flow
[ 0.49-0.62 lb/sec - Internal Fluid Temperature - Group 2 ............. 945 Figure 9-15 LST with Diffuser Under Steam Generator - Steam Flow 0.49-0.62 lb/sec - Saturation Temperature - Group 2 . . . . . . . . . . . . . . . 946 '.' Figure 9-16 LST with Diffuser Under Steam Generator - Steam Flow O.49-0.62 lb/sec - Internal Steam Pressure Ratio - Group 2 . . . . . . . . . . . 947 i Figure 9-17 LST with Diffuser Under Steam Generator - Steam Flow l 0.76 0.84 lb/sec - Internal Fluid Temperature . . . . . . . . . . . . . . . . . . . . . . 948 h Figure 9-18 LST with Diffuser Under Steam Generator - Steam Flow
- .0.76 0.84 lb/sec - Saturation Temperature . . . . . . . . . . . ._- . . . . . . . . . . 949 Figure 9-19 LST with Diffuser Under S'eam Generator - Steam Flow O.76-0.84 lb/sec - Internal Steam Pressure Ratio . . . . . . . . . . . . . . . . . . . . 9-70
[ Figure 9-20L LST with Diffuser Under Steam Generator - Steam Flow - l 1.10-1.20 lb/sec - Internal Fluid Temperature . . . . . . . . . . . . . . ....... 9-71 l Figure 9-21
~ ~
LST with Diffuser Under Steam Generator - Steam Flow I ! 1.10 1.20 lb/sec'- Saturation Temperature - . . . . . c. . ............... 9-72 Figure 9 LST with Ddfuser Under Steam Generator - Steam Flow ! 1.10-1.20 lb/sec'- Intemal Steam Pressure Ratio . . . . , , . . . . . . . . . . . . . . 9-73 i Figure 9 LST with Diffuser Under Steam Generator - Steam Flow - l 1.54-1.68 lb/sec - Intemal Fluid Temperature . . . . . . . . . . . . . . . . . . . . . . 9-74
- Figure 9-24 LST with Diffuser Under Sterm Generator - Steam Flow r - 1.54-1.68 lb/sec - Saturation Temperature ' . . . . . . . . . . . . . . . . . . . . . . . 9-75
[ Figure 9-25 . LST with Diffuser Under Steam Generatcr - Steam Flow 1.54-1.68 lb/sec - Internal Steam Pressure Ratio . . . . . . . . . . . . . . . . . . . . 9-76 3 Figure 9-26 LST with Diffuser Up 6 Feet - Steam Flow 0.76 & 1.68 lb/sec - Internal Fluid Temperature . . . . . . . . . . . . . . . . . . . . . . . . . 9-77
- ( ,
f ( ' WGOTHIC Application to AP600 octobw 19a7 l o:\3006w-fm.non:1b-101497 Revision 1
xx LIST OF FIGURES (Cont.) Figure 9 27 LST with Diffuser Up 6 Feet - Steam Flow 0.76 & 1.68 lb/sec - Saturation Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-78 Figure 9-28 LST with Diffuser Up 6 Feet - Steam Flow 0.76 & 1.68 lb/see - Internal Steam Pressure Ratio . . . . . . . . . . . . . . . . . . . . . . . 9-79 Figure 9-29 LST with Steam Injection: 3 Inch Pipe - Steam Flow 0.76 - 0.95 lb/sec - Intemal Fluid Temperature . . . . . . . . . . . . . . . . . . . . 9 80 Figure 9-30 151 with Steam injection: 3 Inch Pipe - Steam Flow 0.76 - 0.95 lb/sec - Saturation Temperature . . . . . . . . . . . . . . . . . . . . . . . 9-81 Figure 9 31 LST with Steam Injection: 3 Inch Pipe - Steam Flow 0.76 - 0.95 lb/sec - Intemal Steam Pressure Ratio . . . . . . . . . . . . . . . . , , . . %82 Figure 9-32 LST with Steam Injection: 3 Inch Pipe - Steam Flow 1.25 - 1.31 lb/sec - Internal Fluid Temperature ............. . . . . . 9-83 Figure 9-33 LST with Steam Injection: 3 Inch Pipe - Steam Flow 1.25 - 1.31 lb/sec - Saturation Temperature . . . . . . . . . . . . . . . . . . . . . . . 9-84 Figure 9 34 LST with Steam Injection: 3 Inch Pipe - Steam Flow 1.25 *.31 lb/sec - Intemal Steam Pressure Ratio . . . . . . . . . . . . . . . . . . . 9-85 Figure 9-35 CMT Compartment Layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-86 Figure 9-36 Simplified AP600 Containment Diagram . . . . . . . . . . . . . . . . . . . . .... 9-87 Figure 9-37 EGOTHIC Calculated LOCA Blowdown Steam Pressure Ratio for Jet Momentum Dissipated in SG East Co npartment . . . . . . . . . . . . . . . . . . . 9-88 Figure 9-38 EGOTHIC Calculated AP600 Containment Pressure - Sensitivity to Loss Coefficients for LOCA Jet Momentum Dissipated in SG East Compartment ............................................. 9-89 Tigure 9-39 WGOTHIC Calculated Flow Pattern - Sensitivity to Loss Coefficients for LOCA Jet Momentum Dissipated in SG East Compartment at 20 Seconds ............................................... 9-90 Figure 9-40 EGOTHIC Calculated Flow Pattem - Sensitivity to ' oss Coefficients for LOCA Jet Momentum Dissipated in SG East Comp, at 1000 Seconds . . . . 9-91 Figure 9-41 ECOTHIC Calculated Flow Pattem - Sensitivity to Loss Coefficients for LOCA Jet Momentum Dissipated in SG East Comp. at 1550 Seconds . . . . 9-92 Figure 9-42 EGOTHIC Calculated Flow Pattem - Sensitivity to Loss Coefficients for LOCA Jet Momentum Dissipated in SG East Comp. at 80050 Seconds . . . 9-93 Figure 9-43 EGOTHIC Calculated AP600 Containment Heat Removal Rates - LOCA Jet Momentum Dissipated in SG East Compartment . . . . . . . . . . . 9-94 Figure 9-44 EGOTHIC Calculated AP600 Containment Steam Pressure Ratio for LOCA Jet Momentum Dissipated in SG East Compartment . . . . . . . . . . . 9-95 Figure 9-45 EGOTHIC Calculated AP600 Cont. Pressure - Sensitivity to Heat Transfer Coefficient for Study of Undissipated Jet Effects During a LOCA . 9-96 Figure 9-46 EGOTHIC Calculated AP600 Ccntainment Pressure-LOCA Jet Momentum Dissipated in SG East Compartnent . . . . . . . . . . . . . . . . . . 9-97 3
}yGOTHIC Application to AP600 october 19e o \3006w.fm.non:lb 101497 Revi-km 1
________._.._._.-_._m.._.__._______.__ _ y 4 7 4 LIST OF FIGURES (Cont.) , !. Figure 9-46A EGOTHIC Calculated AP600 Containment Below Deck Compartment . , ] Pressure for LOCA Jet Momentum Dissipated in SG East Compartment . 9-98 U Figure 9-47 WGOTHIC Calculated Flow Pattem - LOCA Jet Momentum Dissipated .
]
4' in SG East Compartment at 20 Seconds . . . . . . . . . . . . . . . . . . . . . . . . . . 9-99 - Figure 9-48 EGOTHIC Calculated Flow Pattem - LOCA Jet Momentum Dissipated , ' in SG East Compartment at 1000 Seconds . . . . . . . . . . . . . . . . . . . . . . . 9100 ' Figure 9-49 EGOTHIC Calculated Flow Pattem - LOCA Jet Momentum Dissipated [ in SG East Compartment at 1500 Seconds . . . . . . . . . . . . . . . . . . . . . . . 9-101 i Figure 9-50 EGOTHIC Calculated Flow Pattem - LOCA Jet Momentum Dissipated l in SG East Compartment at 8000 Seconds . . . . . . . . . . . . . . . . . . . . . . . 9-102
- Figure 9-51 Details of WGOTHIC Flow Paths to Above-Deck Region from CMT,
[ Refueling Canal, and IRWST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-103 i Figure 9-52 EGOTHIC Calculated AP600 Containment Pressure - LOCA Plume l Rising into CMT Room . . . . . . . . . . . . . . . . . . . . . . _. . . . _ . . . . . . . . . . _ 9-104 j Figure 9-53 WGOTHIC Calculated Flow Pattem - LOCA Plume Rising into l', CMT Room at 1000 Seconds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-105
- j. Figure 9-54 EGOTHIC Calculated Flow Pattem - LOCA Plume Rising into 1 CMT Room at 1400 Seconds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-106 I Figure 9-55 WGOTHIC Calculated AP600 Containment Heat Removal kates -
LOCA Plume Rising into CMT Room . . . . . . . . . . . . . . . . . . . . . . . . . . 9-107
- Figure 9-56 HGOTHIC Calculated AP600 Containment Pressure - LOCA Plume Rising into CMT Room and SG Compartments . . . . . . . . . . . . . . . . . . . 9-108
!. Figure 9-57 EGOTHIC Calculated Flow Pattem - LOCA Plume Rising into [' CMT Room and SG Compartments at 1000 Seconds . . . . . . . . . . . . . . . 9-109 Figure 9-58 WGOTHIC Calculated Flow Pattem - LOCA Plume Rising into
- CMT Room and SG Compartments at 1500 Seconds . . . . . . . . . . . . . . . 9-110
- Figure 9-59 EGOTHIC Calculated AP600 Containment Heat Removal Rates - ,
LOCA Plum 3 Rising into CMT Room and SG Compartments . . . . . . . . 9-111 Figure 9-60 EGOTHIC Calculated AP600 Containment Steam Pressure Ratio for MSLB Above-Deck . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-112 Figure 9-61 EGOTHIC Calculated AP600 Containment Pressure - MSLE Above Operating Deck . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-113 Figure 9-62 EGOTHIC Calculated AP600 Containment Pressure - MSLB in CMT Room . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-114 Figure 10-1 LOCA Sensitivities to Nominal Inputs and Correlations . . . . . . . . . . . . . 10-3 Figure 10-2 MSLB Sensitivity to Nominal inputs and Correlations . . . . . . . . . . . . . . . 10-4 Figure 10 3 Comparison of Evaluation Model and Nominal Mass Release . . . . . . . . . 10-9 Figure 10-4 Comparison of Evaluation Model and Nominal Energy Release . . . . . . . 10-10 Figure 11-1 AP600 Containment LOCA Transient Pressure Prediction with Full, One-Half, and One-Quarter Timesteps . . . . . . . . . . . . . . . . . . . 11-4 Figure 11-2 AP600 Timestep Comparison, Percentage Difference in Predicted Pressures Between Successive Code Versions . . . . . . . . . . . . 115 Figure 12-1
- Figure 12-2 Simplified _ Westinghouse-GOTHIC Line Diagram of Annulus Clime ClimeModel Wall. .Source . . . . . . , ,Term . . . . . .Models12-14 . . . , . . .
Figure 12-3 Noding Diagram, 8 Clime Node Model . . . . . . . . . . . . . . . . . . . . . . . . . 12-15 WGOTHIC Application to AP600 october 1997 c:\3006w fm.non:1b-101497 Revision 1
xxii LIST OF FIGURES (Cont.) Figure 12-4 Comparison of Transient IIcat Transfer Rates; Case 1 . , , . . . . . . . . . . . 12-16 Figure 12-5 Comparison of Transient Heat Transfer Rates; Case 2 . . . . . . . . . . . . . . 12-17 Figure 12-6 Comparison of Transient Heat Transfer Rates; Case 3 . . . . . . . . . . . . . . 12-18 Figure 12-7 Comparison of Transient Heat Ttansfer Rates; Case 4 . . . . . . . . . . . . . . 12-19 Figure 12-8 Comparison of Transient Heat Transfer Rate Differences; Car.e 1. . . . . . 12-20 Figure 12-9 Comparison of Transient Heat Transfer Rate Differences; Case 2 . . . . . . 12-21 Figure 12-10 Comparison of Transient Heat Transfer Rate Differences; Case 3 . . . . . . 12-22 Figure 12-11 Comparison of Transient Heat Transfer Rate Differences; Case 4 . . . . . . 12-23 Figure 12-12 Comparison of Heat Flux Profiles; Time t = 2000 Seconds; Case 1. . . . . 12 24 Figure 12-13 Comparison of Heat Flux Profiles; Time t = 2000 Seconds; Case 2 . . . . . 12-25 Figure 12-14 Comparison of Heat Flux Profiles; Time t = 2000 Seconds; Case 3 . . . . . 12-26 Figure 1215 Comparison of Heat Flux Profiles; Time t = 2000 Seconds; Case 4 . . . . . 12-27 Figure 12-16 Comparison of Film Temperature Profiles; Time t = 2000 Seconds; C a se l . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-2 8 Figure 12-17 Comparison of Film Temperature Profiles; Time t = 2000 seconds; C a se 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-2 9 Figure 12-18 Comparison of Film Temperature Profiles; Time t = 2000 seconds; C a se 3 . . . . . . . . . . . . . . . . . . . , , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 -30 Figure 12-19 Comparison of Film Temperature Profiles; Time t = 2000 seconds; C a se 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-31 Figure 12-20 Comparison of Air Temperature Profiles; Time t = 2000 seconds; C a se l . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-3 2 Figure 12-21 Comparison of Air Temperature Profiles; Time t = 2000 seconds; Ca se 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 -33 Figure 12-22 Comparison of Air Temperature Profiles; Time t = 2000 seconds; Case 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-34 Figure 12-23 Comparison of Air Temperature Profiles; Time t = 2000 seconds; Case 4 . . . . . . . . . . . . . . . . . . . .............................. 12-35 Figure 12-24 AP600 Cor.tainment Model Clime Noding Pattern . . . . . . . . . . . . . . . . . 12-36 Figure 12-25 AP600 Containment Model Double Vertical Clime Noding Pattern . . . . 12-37 Figure 12-26 AP600 Containment Model Double Stack Clime Noding Pattern . . . . . . 12-38 Figure 12-27 AP600 Containment Model Double hiesh Point Clime Noding Pattern . 12-39 Figure 12-28 Pressure History, AP600 Containment Model; Double Clime . . . . . . . . . 12-40 Figure 12-29 Wet Heat Flux vs. Clime; AP600 Containment Model, Base Case . . . . . . 12-41 Figure 12-30 Wet Heat Flux vs. Clime; AP600 Containment Model, Double Clime . . . 12-42 Figure 12-31 Dry Heat Flux vs. C!ime; AP600 Containment Model, Base Case . . . . . . . '2-43 Figure 12-32 Dry Heat Flux vs. Clime; AP600 Containment Model, Double Clime . . . 12-44 Figure 12-33 Film Temperature vs. Clime; AP600 Containment Model, Base Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-4 5 Figure 12-34 Film Temperature vs. Clime; AP600 Containment Model, Double Clime ............................................12-46 Figure 12-35 Dry Surface Temperature vs. Clime; AP600 Containment Model, B a se Ca se . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-47 Figure 12-36 Dry Surface Temperature vs. Clime; AP600 Containment Model, Double Clime ........ ...................................12-48 Figure 12-37 Annulus Pressure vs. Clime; AP600 Containment Model, Base Ca se . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-49 O
}%OTHIC Application to AP600 october 1997 o:\3006w.fm.non:1b 101497 Revision 1
xxili LIST OF FIGURES (Cont.) Figure 12-38 Annulus Pressure vs. Clime; AP600 Containment Model,' . Double Clime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-50
- Figure 12 39 Air Temperature vs. Clime; AP600 Containment Model, Base Case . . . . 12 i Figure 1240 . Air Temperature vs. Clime; AP600 Containment Model, Double Clime , ,12-52:
i: Figure 1241 . Air Density vs. Clime; AP600 Containment Model, Base Case . . . . . . . . .;12-53 Figure 1242 Air Density vs. Clime; AP600 Containment Model, Double Clime . . . ._. 12-54 ; i-l Figure 12 43 - Air Velocity vs. Clime; AP600 Containment Model, Base Case . . . . . . . . .- 12-55 [ Figure 1244- Air Velocity vs. Clime; AP600 Containment Model, Double Clime . . . . . 12-56
- Figure 1245 Heat Rejection History Comparison, AP600 Containment Model;
! Double Clime Sensitivity Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .' 12 57 Figure 12-46 Integrated Heat Rejection Comparison, AP600 Containment Model; Double Clime Sensitivity Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-58 , Figure 12 Pressure History Comparison, AP600 Containment Model; i _ Double Stack Sensitivity Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17-59
- Figure 12-48 Heat Rejection History Comparison, AP600 Containment Model;
, Double Stack Sensitivity Case . . . . . :. . . . . . . . . . . . . . . . . . . . . . . . . . . 12-60 Figure 12-49 Integrated Heat Rejection Comparison, AP600 Containment Model; Double Stack Sensitivity Case . . . . . . . . . . . . . . . . '. . . . . . . . . . . . . . . . 12-61 i- Figure 12-50 Pressure History Comparison, AP600 Containment Model; Double Mesh Sensitivity Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-62 F 4 Figure 12-51 Heat Rejection History Comparison, AP600 Containment Model; Double Mesh Sensitivity . Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-63 Figure 12-52 I .tegrated Heat Rejection Comparison, AP600 Containment Model; l_ . Double Mesh Sensitivity Case . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . . 12-64 i l' i-i t-i k i-4 4 i i l 4 1 4 i I t LO WGOTHIC Application to AP600 h 1"7 j Tuoo6w-tm.non:1b 101497 ##I .i
xdv LIST OF ACRONYMS AND ABBREVIATIONS O CFD Computational Fluid Dynamics CMT Core Makeup Tank CVS Chemical and Volume Control System DBA Design Basis Accident DBE Desigr. Basis Event DECLG Double-Ended Cold Leg Guillotine EM Evaluation Model HDR Heissdampfreaktor HVAC Heating, Ventilation, and Air Conditioning IRWST Internal Refueling Water Storage Tank LOCA Loss-of-Coolant Accident LST Large-Scale Tests M&E Mass & Energy MIT Massachusetts Institute of Technology MSLB Main Steamline Break NAI Numerical Applications,Inc. NUPEC Nuclear Power Engineering Corporation PCS Passive Containment Cooling System PIRT RCS Phenomena Identification and Ranking Table Reactor Coolant System h SG Steam Generator - SRP Standard Review Plan SSAR Standard Safety Analysis Report STC Science & Technology Center TS Technical Specifications UWO University of Western Ontario
'dGOTHIC Westinghouse-GOTHIC
. O WGOTHIC Application to AP600 October 1997 o:\3006w-fm.nortib-101497 Revision 1
. . ~ . . - ~ . . - . . - . . - - . . . . ,
i
=
WESTINGHOUSE NON-PROPRIETARY CLASS 3 i Section 1 o Introduction i e i h I i i i. t i ?- 4 . I l-i 4 l
- t.
J. i t l t i ? ?
- l. -
f t-t. i. i-i 4. a , 1
} +
k
- i i
T f c:\3006w.1.nortib-101497 i-i- 4- . - _ . _ _ . _ . _ . - - _ _ . _ _
. .lii TABLE OF CONTENTS en (d uST OF FIG URES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ill 1.0 -- INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 1.1 OBJECTIVE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 1.2 AP600 CONTAINMENT DBA REPORTS . . . . . . . . . . . . . . . . . . . . . . . . . 1-1 1.2.1 Accident Specification and Phenomena Identification and Ranking Table Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1
- 1.2.2 Scaling Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-3 1.2.3 Heat and Mass Transfer Correlations Report . . . . . . . . . . . . . . . . . 1-6
- 1.2.4 }yGOTHIC Code Description and Validation . . . . . . . . . . . . . . . . . 1-7 1.2.5 SSAR............................................... i-7 i
1.3 ' APPLICATIONS REPORT CONTENT
SUMMARY
. . . . . . . . . . . . . . . . . 1-7 1.4 USE OF LST AND VALIDATION RESULTS . . . . . . . . . . . . . . . . . . . . . . . 9 -
j 1.4.1 LST Matrix Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 10 1.4.2 Use of LST Separate Effects Data . . . . . . . . . . . . . . . . . . . . . . . . 1-12 1.4.3 LST Confirmation of Phenomena . . . . . . . . . . . . . . . . . . . . . . . . . 1-12
\ - 1.4.4 Code Comparison to LST as an Integral Test . . . . . . . . . . . . . . . . 1-12 1.5 INTERFACE WITH SSAR CALCULATIONS . . . . . . . . . . . . . . . . . . . . . . 1-13 l 1
1.6 CONCLUSION
S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-13 1 1 i 1.7 REFE RENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-13 ' LIST OF FIGURES Figure 1-1 Relationship of AP600 Containment DBA Reports . . . . . . . . . . . . . . . . . . . 12
\,_/
Introduction October 1997 o:\3006w-1.non:1b-101497 Revision 1 i
1-1
1.0 INTRODUCTION
1.1 OBJECITVE The computer code used for the AP600 containment pressure design basis accident (DBA) analysis is EGOTHIC. EGOTHIC is used to calevi *e a consewative AP600 pressure transient response and to specify temperatures for equipment qualification. 'Ihe containment DBA analysis makes use of the lumped parameter approach which is based on 30 years of nuclear industry experience. The industry experience has identified lumped parameter limitations and biases that are due primarily to the oversimplification of the momentum formulation. Limitations and biases have been identified based on international tests at different scales (Section 9). Biases and conservatism are applied to models for important phenomena in the w JGOTHIC Evaluation Model to develop a bounding methodology, so that AP600 containment pressure is conservatively estimated. This report describes specific modeling and defines methods used to develop conservative input for the EGOTHIC code to create a bounding Evaluation Model. Using design parameters _ sp?cified in the AP600 SSAR, the hcensing basis AP600 Evaluation Model is used to calculate the design basis pressures and temperatures reported in the SSAR. 1.2 AP600 CONTAINMENT DBA REPORTS b G As shown in Figure 1-1, this report fits into the framework of licensing documentation which _ defines the containment DBA methods. Following is a brief summary of the purposes of containment DBA reports. 1.2.1 Accident Specification and Phenomena identification and Ranking Table Report WCAP-14812 (Ref.1.1) describes the containment and passive containment cooling system (PCS), . defines DBA accidents, identifies success criteria, and ranks the importance of phenomena that must be considered. A cross-reference to relevant tests and test data reports is also included. A systematic process has been followed to identify and rank phenomena, including input and review by members of industry, academia, and regulatory authorities. Introduction October 1997 o:\3006w-1.noruttr101497 Revision 1
1-2 O WCAP-14812 Accident Specification and PIRT (includes Overview of Bounding Methodology) N
~
1 Test Reports WCAP-14845 WCAP-14326 WCAP-14382 (Incorporatedby Scaling Analysis Heat & Mass WGOTHiC Code Reference in Higher Transfer Description & LevsIReports) Correlations Validation e O WCAP-14407 WGOTHIC Application Methodology o SSAR Design inputs DBA Calculations Figure 11 Relationship of AP600 Containment DBA Reports i introduction October 1997 o;\3006w-1.non 1h101497 Revision 1
13 As a convenient vehicle for defining the Evaluation Model approach, the following information is provided in WCAP 14812 for each phenomenon:
- PIRT ranking
- Basis for PIRT ranking
- Test results Sct.llng results Sensitivity studies Expert review
- How phenomena are implemented in the Evaluation Model
- Justification of Evaluation Model treatment of phenomenon Test experience Modeling guidance Sensitivity studies
- E aluation Model treatment of uncertainty 1.2.2 Scaling Report O The application of scaling to a specific methodology is related to the type of analysis being performed and the regulatory needs to be satisfied. The regulations require supporting documentation for the use and sufficiency of the database to develop bounding models for the full-scale AP600 containment pressure transient. The objectives for the scaling of the AP600 pressure transient and the approximately 1/8 geometric scale test vessel, called the Large-Scale Test (LST), are derived from regulations and regulatory guides, WCAP 14845 (Ref.1.2) descritws how nealing has been used to derive the appropriate nondimensional parameters and their Ah4 ranges to examine phenomena for bottom up model validation. Separate eficcts tests (SETS) are identified and the test parameter ranges compared to AP600 ranges to show sufficiency of the test database for application to containment DBAs. Scaling is used to identify distortions in the LST that are then addressed in the bounding methodology.
The following shows how ob}ectives are met le ung in support of AP600 containment DBA metheds. O Introduction Cktober 1997 c:\3006w 1.non.lb-101497 Revison 1 j l
- - - _ _ _ _ _ 1
1-4 The scaling analysis in WCAP-14845 (Ref.1.2) satisfies the three stated objectives for AP600 containment pressure scaling. The conclusions of the scaling analysis are: g
- 1. Support Development of Bounding Methodology The scaling analysis confirmed the identification in the PIRT (Ref.1.1) of high ranked phenomena. The high ranked phenomena inside containment are the break source, gas compliance, and condensation on the shell and heat sinks. The high ranked phenomena outside containment, are evaporation of the external liquid film and the PCS natural circulation flow rate. In addition, the scaling analysis confirmed the PIRT ranking of lower order phenomena including convectio., and radiation heat transfer, liquid film conductance, and liquid film energy transport.
The high ranked phenomena and the parameters that most strongly affect them are the ones that must be bounded in the Evaluation Madel. Specifically the phenomena and how they are bounded for the Evaluation Model are:
- Dreak source mass and energy releases are calculated using assumptions that makes them conservatively high
- Gas compliance is made conservatively low by lower bounding the volume e Condensation rate test data are lower bounded and the biased correlation is used O
- ShcIl and heat sink heat capacity is calculated using m!ain.um volume and heat capacity
- The effects of circulation and stratification were examined and ranged and the combinations that produced the highest transient pressures were used to define the Evaluation Model
- Evaporation rate test data are lower bounded and the biased correlation is used
- The minimum external water flow rate is applied
- The effects of external water application and stability were examined and ranged and a conservative coverage model was developed
- The PCS air flow path maximum flow resistance is used The net effect is an Evaluation Model that bounds all the dominant processes so as to produce the maximum predicted pressure respcnse.
O Introduction Cktoter 1997 oA3006w 1.nortit*101497 Reviskm 1
15
- 2. Specify Individual Model Constitutive Relations.
ApproprtMe constitutive relations, models, and/or bounding methods are identified tor each of the dominant phenomena and parameters in objective 1 above. Condensation and evaporation are modeled using conventional free and forced convection mass transfer relationships, characterized by Reynalds, Grashof, and Schmidt numbers. The range of these dimensionless variables ne essary to cover AP600 operation are defined and SETS are identified and used to validate the selected mass transfer correlations. The range of dimensionless variables in the data are shown to encompass the expected range of operation in AP600. Bounding models were developed for film stability and for circulation and stratification. The models and modeling approach used in the Evaluation Model for film stability and circulation and stratification are discussed in Sections 7 and 9, respectively. The PCS air flow path resistante is bounded using measurements from a geometrically scaled test, with calculations to extend the results to operation at 1.5 times higher Reynolds number air flow, and with an additional 30 percent bias for added conservatism. The volume compliance and heat sink heat capacities are modeled using conservatively low
] volume calculations and reasonable solid material density and heat capacity values.
The external delivered waNr flow rate was conservativdy calculated using hydraulle models and assumptions. The resulting external source flow history is reduced by the film stability and runoff model, and only the flow that is expected to etaporate, consistent with film stability and evaporation rate,is applied to the Evaluation Model.
- 3. Investigate Use of LST to Validate Elements of the Bounding Evaluation Model The LST was used as a wurce of separate effects data to validate external dry heat transfer, condensation mass transfer, film stability, and stratification models as discussed under 1 and 2-above. Component vel distortions in the LST were addressed using local measurements of temperature, steam concentration, and velxities from the LST, and by supplementing the LST r
data with data from other sources where the range of LST parameters was insufficient to cover AP600 operation. The scaling analysis for the LST as an integral test shows the three primary drivers, or dominant system-level phenomena, for the pressure transient are the break source, the gas volume, and the heat sink surface area dependent condensation rate. The scaling analysis shows the 7 integrated system response is well scaled for small values of dimensionless time and for longer ( values of dimensionless time, and shows a distortion for intermediate times. The magnitude of introduction October 1997 c:\3006w.1.rortit>101497 Revison 1
i l 16 the distortion for intermediate time (between blowdown and quasi-steady,long term operation) is too large to scale the LST pressure transient results directly to the AP600 in this range, However, the distortion is small enough that the LST data can be used to examine known g lumped parametar modeling biases in the bounding Evaluation hiodel. Furthermore, testing the Evaluation hiodel against a range of LST cases, with step char.ges in boundary conditions (source flow rate, external fan, extemal water on or off) covers the range of system level dimensionless groups for AP600 operation through the refill and peak pressure time phases between the blowdown and quasi-steady, long term operation. The boundary condition variations result in a range of relative values of condensation rate and steam pressure sofficient to span the range for AP600. The lumped parameter Evaluation Model does not resolve intemal velocity and concentration fields due to its simplified momentum model and large lumped volumes. Comparisons between preliminary versions of the Evaluation Model and the system level LST response showed that pressure was reasonably well predicted, with a modest conservative margin. Examination of intemal processes clearly identified the existence of competing intemal effects in which the excessive vek> cities predicted by the lumped parameter model overpredicted the velocity component of mass transfer, while overmixing underpredicted the steam concentration component of mass transfer. Consequently, these competing effects in AP600 predictions are addressed. The effect of overpredicted velocities was resolved by using only free convection for intemal heat and mass transfer, thereby climinating velocity from the condensation correlation. The overmixing issue vas resolved by examining and blasing the effects of circulation and stratification in the Evaluation Model as discussed in Section 9. h 1.2.3 Heat and Mass Transfer Correlations Report WCAP-14326 (Ref.1.3) documents the analytical and experimental bases for heat and mass transfer correlations associated with: . Condensation mass transfer
- Evaporation mass transfer
. Convective heat transfer + Liquid film thermal resistance For modeling convenience, an explicit representation of the liquid film thermal resistance is modeled, with condensation or evaporation occurring at the film surface. This is in contrast to the more traditional approach c mbining mass transfer and liquid film resistance and then using the solid surface temperama. The explicit representation allows clearer treatment of elements of uncertainty in mass transfer and liquid film over the AP600 range of conditions. O Introduction Cktober 1997 oMxbw.1.rmit=101497 Revision 1
K?MkW&*k
.-.y p 17 1.2.4 EGOTillC Code Description and Validation O WCAP 14382 (Ref.1.4) documents the implementation of " climes" subroutines in the GOTHIC code. Climes are used to represent heat and mass transfer on the containment shell, shield building, and baffle. The report shows comparisons to the LST using both lumped parameter and distributed parameter models, identifies lumped parameter biases and competing effects based on the LST calculations, and describes the derivation of noding guidan.e for the AP600 Evaluation Model. ECOTHIC verification and validation has been completed using calculations of separate effects tests (Reference 1.4, Section 4 and Reference 1.3, Sections 3.1 and 3.3).
1.2.5 SSAR The methodology in the WGOTHIC Application Report is used, along with design input specified in Section 1.5, to perform the licensing basis DBA containment calculations reported in the AP600 SSAR, Chapter 6.2. 1.3 APPLICATIONS REPORT CONTENT
SUMMARY
The Introduction outlines the containment DBA analysis approach, summarizes the use of the LST, and shows how the Evaluation Model methods are incorporated in the containment DBA analysis reported in SSAR 6.2 for long term loss-of coolant accidents (LOCA) and main steamline breaks (MSLB). Subsequent sections document elements of the methodology, as follows. Section 2 contains a summary of high and medium ranked phenomena. This section describes the process used to develop the bounding containment Evaluation Model. Each step or element in the process is briefly described, and those phenomena that were determined to be of high or medium rank are presented, with a summary of how those phenomena are addressed by the EGOTHIC Evaluation Model. Section 3 presents an overview of the Westinghouse-GOTHIC code package. The EGOTHIC features, development history, and validation programs are briefly described. The models and features that were added by Westinghouse to adapt GOTHIC to model the AP600 PCS are also described. Section 4 presents the EGOTHIC design basis Evaluation Model of the AP600 using design inputs specified in this report. In this section the code inputs are described for the AP600 model. The code inputs include free volumes, elevations, heat sink characteristics, and boundary conditions. Graphics are included which aid in visualizing both the AP600 layout and the EGOTHIC model of the AP600. The methodology defined in this section is used for the licensing basis SSAR Evaluation Model. O introduction Octotwr 1997 c:\3006w 1.non:It>101497 Revision 1
18 Lection 5 contains a number of sensitivity cases varying the initial conditions assumed for the design basis ana fses. These include sensitivities on initial containment humidity, initial containment pressure, initial containment temperature, outside humidity, outside temperature, g and boundary condition drop assumptions. Except as noted specifically for a sensitivity study, all sensitivities in this report are based on the base case Evaluation Model described in Section 4. This section provides the basis for choosing the conservative initial conditions assumed for the DDA analyses. Section 6 describes the effects of meteorological changes on the performance of the PCS. In this section the effects of PCS effluent entrainment into the PCS inlet are studied. In addition to recirculation, the effects of wind on PCS performance are identified. The results of these studies show that wind effects are beneficial to containment cooling since they augment the natural draft velocity that develops during PCS operation. The effluent recirculation due to inversions or strong winds is shown to have a negligible effect on PCS performance and containment pressure response. Section 7 supplies the methodology for calculating the PCS applied water flow rate input for the AP600 containment Evaluation Model. Based on conservatively bounding liquid film test data from various tests, the coverage and evaporation rate are conservatively calculated, and only the amount of water which evaporates is applied to the Evaluation Model. Thus, there is a conservative bound on the amount of evaporative cooling credited in the Evaluation Model. The implementation of evaporation limited flow applied in the Evaluation Model also conservatively underpredicts subcooled liquid film heat removal from containment. The basis for the delay g time in the application of the film as well as the coverage areas and other parameters are presented. Sensitivities to coverage area and other parameters are presented which demor. strate the conservatism in the method used to determine coverage for the AP600. Section 8 presents the sensitivity of the AP600 blowdown pressurization transient. The PCS model that uses climes is compared to a single volume model of the AP600 created based on l Standard Review Plan (SRP) methodologies. The single volume model uses lyGOTHIC conductors to model the containment shellinstead of the clime model and uses the Uchida heat transfer correlation instead of the Westinghouse-developed clime heat and mass transfer correlations. The results of this comparison show that there is very little difference between the two models for the blowdown phase of the transient. A sensitivity to heat sinks during blowdown is also presented. l Section 9 addresses circulation and stratification within the AP600 containment. Circulation and l stratification can be affected by break location, orientation, and type, in addition to noding assumptions. The effects of circulation and stratification inside the AP600 are assessed for an MSLB and the various time phases (i.e., blowdown, refull, peak pressure, and long-term) of a LOCA. Based on these results, biases have been incorporated into the Evaluation Model as described in Sections 4 and 9. h introduc9on oaober 1997 oMKbw.1.non:1b-101497 Reviuon 1
19 Section 10 describes the conservatism contained in some Evaluation Model assumptions made for the design basis LOCA accident analysis that are intended to maximize the peak pressure. in this section, the conservatism in the heat and mass transfer biases, the initial conditions for inside and outside of containment, PCS water temperature, material properties, steel concrete gap, external annulus loss coefficient, dead ended compartment modeling biars, and the LOCA mass and energy releases are described in a step wise fashion. The final result is a quantification of the conservatism contained in the Evaluation Model in the above parameter. Based on these sensitivities, there is approximately 13 psi of margin in the design basis analysis second peak pressure as compared to the nominal case second peak. Since the nominal case maximum pressure occurs during blowdown, there is approximately 11.5 psi of margin in the maximum calculated pressure between the design basis case and the nominal case. It should be noted that the nominal case only credits conservatisms that can be readily quantified. A similar sensitivity for the net effect of parameters important in the MSLB analysis is also provided. Section 11 describes the sensitivity of EGOTHIC to changes in the calculated timestep size. The timestep selection logic was modified to reduce the calculated timestep by one half and by one-quarter in separate .:ases. The results of this sensitivity show that the solution is stable, in that the pressure transients did not change appreciably as the timestep size was reduced. Thb result supports the conclusion that the timestep logic used in ECOTHIC is acceptable. Section 12 examines the sensitivity of the predicted AP600 containment pressure transient to changes in clime noding. Results support the noding used to represent volumes, elevations, and azimuthal segments in the extemal annulus, as well as the numerical mesh pattern through conductors. The methodology specified in the above sections is used, together with design inputs specified in Section 1.5, to perform the licensing basis calculations in Section 6.2 of the AP600 SSAR. 1.4 USE OF LST AND VALIDATION RESULTS In the mid 1980's, Westinghouse developed the LST as an integral test to provide long term heat and mass transfer data for a geometrically similar model of the AP600 containment vessel. The focus was on long-term transient behavior, because that is where the AP600, with no credited active heat removal system, differed significantly from current operating plant databases. Because of limitations of scale (power to-volume and power to-area ratios, and steam supply), the LST matrix was selected to vary boundary conditions parametrically to obtain data over the AP600 range of parameters. Specific AP600 transients were not simulated with the LST. The LST was designed to provide long tenn heat and mass transfer data in an integral setting, that is, with external evaporation and intemal condensation acting simultaneously, for a p geometrically similar model of the AP600 containment vessel (Reference 1.5, Section 3.2.4.2). The U use of the LST has been supported through the application of scaling methodologies that have introduction October 1997 c:\3006w.1.non;nw101497 Revu6on 1
1 10 l evolved during the 1990's. The LST is being used to examine AP600 containment performance from two perspectives: 1) addressing specific phenomena a3 an SET; and 2) system level pressure transient simulation for the purpose of examining lumped parameter code biases. As discussed in subsection 1.2.2, local data from the LST has been combined with other SETS and integlal effects tests (IETs) at different scales to provide supporting data for the following phenomena:
- Dry extemal riser annulus heat transfer
. External liquid film stability j . Intemal condensation mass transfer . Intemal stratification l Scaling has been used to identify portions of the AP600 transient that are represented by LST transients and distortions in the system pressure transient (Ref.1.2, Section 11). Results indicate that there are distortions from a system perspective over the transition from initial blowdown to long term, quasi-steady behtvior. The distortions arise from compromises associated with testing at the reduced scale. System-level distortion occurs due to the need to simultaneously preserve power-to-volume and power to-area, which are mutually exclusive requirements, when volume compliance and condensation rates are both significant. Varying the boundary conditions with step changes in extemal water flow, extemal fan, and steam flow addresses the distortion sufficiently to use the LST data for code validation. The following summarizes the LST test matrix. 1.4.1 LST Matrix Tests Recognizing at the outset that the LST would contain distortions in terms of representing a single AP600 transient from beginning to end, the test matrix was developed to contain parametric variations that examined various extremes and combinations of AP600 boundary condition effects. In this way, the LST was ranged similarly to an SET. Scaling also shows that the combination of LST matrix tests can be used for understanding the application of lumped parameter biases to AP600. The variations in extemal boundary conditions give a range of relative values of condensation rates (sinks) and steam flows (sources), such that the relative values during the distorted portion of the AP600 transient are effectively covered. in addition to the more obvious matrix test parameters, such as steam flow, experience with the intemational containment test database pointed to the need to examine the effects of boundary condition parameters on distributions of noncondensables inside containment. The following provides a brief overview of the parametric variations included in the LST matrix (Ref.1.6, Tables 1.31,1.3-2, and 1.3-3). Introduction Cktober 1997 o:\3006w 1.norv1b 101497 Revision 1
1*]1 b,c O O O_ -
' Introduction m % gg7 . c:\3006w 1.non:1b 101497 Revision 1
1 12 O jb,c 1.4.2 Use of LST Separate Effects Data Scaling has been used to assess the use of the LST to supplement the smaller scale separate effects data (Ref.1.2, Sections 10.1 and 11.3). Separate effects test data from the LST is used to support validation of the condensation correlation applied to the inner steel shcIl surface (Ref.1.3, Section 3.9) and to examine potential stratification effects in an enclosed volume in an integral setting with external cooling (Section 9). Water coverage and film stability data were used to develop a bounding model to address the effects of film stability (Section 7). External dry heat transfer data have been used to supplement convective heat transfer data (Ref.1.3, Section 3.5). 1.4.3 LST Confirmation of Phenomena The LST data have been used to validate the system scaling equation used to support the identification and ranking of phenomena (Ref.1.2, Section 10.2). 1.4.4 Code Comparison to LST as an Inl ?gral Test O Analyses of the LST have been completed using the EGOTHIC lumped parameter momentum formulation. In the LST calculations, nominal properties and nominal test boundary and initial conditions were used to isolate the biases inherent in the computer code, independent of conservatism included in the Evaluation Model. This allowed the examination of the known lumped parameter biases, and quantification of the effects of compensating errors in lumped parameter results. The method to address the lumped parameter biases, as well as the method used to address phenomena for the Evaluation Model are documented (Ref.1.1, Section 4.4). The AP600 containment DDA analysis approach is based on the lumped parameter formulation. Examination of LST EGOTHIC lumped parameter results identified compensating effects (velocity and steam concentration) that have been bounded in the application to AP600 by using free convection on interior surfaces. By using free convection, the effect of computed velocity is eliminated, and effects of steam concentration distribution can then be separately bounded. O Introduction Cktober 1997 o \3006w.1.non:1b-101497 Revision 1 1' . _ _ _ _ _ _ ._
1 13 1.5 INTERFACE WITil SSAR CALCULATIONS (3 Q The licensing basis containment DDA pressure analysis reported in Section 6.2 of the AP600 SSAR is performed with the WCOTHIC Evaluation Model, defined by methodology described herein. The following design inputs are required as input to the Evaluation Model methodology:
- PCS delivered flow as a function of time assuming failure of one PCCWST drain valve to open.
- Conservatively calculated mass and energy releases as a function of time, using approved methodology (SSAR 6.2.1.3.2 for LOCA and SSAR 6.2.1.4 for MSLU).
- Appropriate Technical Specification and Site Interface Parameters for initial and boundary conditions (SSAR 16.1, Section 3.6 and SSAR 2.3).
The results of the Evaluation Model are used for design pressure evaluation and equipment qualification condition specifications, as reported in SSAR Appendix 3D. Evaluation Model methodology considers DBA phenomena so that the predicted containment pressure has sufficient margin to bound uncertainty in important parameters. The temperature of the break room node is the maximum temperature in containment and is used for input to equipment qualification envelopes to bound the effects of temperature distributions. V
1.6 CONCLUSION
S This report defines a methodology which yields conservative pressure calculation and temperature envelopes. Evaluation Model methodology is cross referenced to PIRT phenomena in Reference 1.1, Section 4.4. The licensing basis DBA calculation is presented in Section 6.2 of the SSAR along with specific design parameters that are input to the methodology.
1.7 REFERENCES
1.1 WCAP-14812, " Accident Specification and Phenomena Evaluation for AP600 Passive Containment Cooling System" Rev.1, June 1997. 1.2 WCAP-14845, Revision 2, " Scaling Analysis for AP600 Containment Pressure During Design Basis Accidents," June 1997, 1.3 WCAP-14326, " Experimental Basis for the AP600 Containment Vessel Heat and Mass Transfer Correlations" Rev.1, May 1997. (~N 1.4 WCAP-14382, "WGOTHIC Code Description and Validation," May 1995. D) Introduction cktober 1997 o;\3006w 1.nortib 101497 Reviske 1
1 14 C 1.5 WCAP 14141,"AP600 Test and Analysis Plan for Design Certification," Rev.1, April 1995. 1.6 WCAP 14135, " Final Data Report for PCS Large-Scale Tests, Phase 2 and Phase 3," O Revision 1, April 1997. O l l l l l l l O' Introduction Cktober 1997 o:\30(4w-1.nortit>101497 Revukm 1
WESTINGilOUSE NON PROPRIETARY CLASS 3 O Section 2 Test and Analysis Process Overview and High and Medium Ranked Containment Phenomena O O c:\3(Kbw 2.non:Ib101497
111 TABLE OF CONTENTS r
\
LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii 1.lST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ill
2.1 INTRODUCTION
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 2.2 ELEMENT 1 - AP600 PCS REQUIREMENTS AND CODE CAPABILIT1ES . . . . . 21 2.3 ELEMENT 2 - ASSESS CODE VERSUS TESTS AND IMPORTANT PROCESSES . 23 2.4 ELEMENT 3 ASSESS UNCERTAINTIES AND DEVELOP BOUNDING M ODE LS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 2.5 ELEMENT 4 PERFORM DBA CALCULATIONS AND COMPARE TO SUCCESS CRITERIA . . . . . . . . . . . . . . . . . . . 4 ................................28 2.6 CONC LUSION S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-8
2.7 REFERENCES
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-8 List OF TABLES Table 21 Phenomena identification and Ranking Table - Summary of High <
and Medium Ranked Phenomena . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 LIST OF FIGURES Figure 2-1 PCS Test and Analysis Process Overview . . . . . . . . . . . . . . . . . . . . . . . . 22 O G Test and Analysis Process Overview ociober 1997
' o:\3006w.2.non:1b 101497 Revision 1
1 w 21 i ) 2.1 -INTRODUCTION 1 O The Evaluation Model for the passive containment cooling system (PCS) design basis accident l (DBA) has been developed using elements of scaling (top down and bottom up modeling of the integrated components), testing, and analysis (bottom up phenomenological models and j evaluations), similar to the methodology for Code Scaling Applicability and Uncertainty (Ref.
- 2.1). Results have been used to identify bounding models and input values for use in the DBA
- Evaluation Model. The results of the DBA analyses provide conservative predictions of design j basis trarslent pressare and temperature response for the containment.
i' !_ The development of the PCS DBA methodology _ has followed an approach which can be i organized into the four elements shown in Figure 21. The elements include tasks, that together - I provide a structured, traceable, and practical method for j-
- Specifying the scenario .
(
- Identifying phenomena important to the transient j
!
- Evaluating data and scale effects q
- Documenting and validating the computer code
{ ;
- Assessing margins and uncertainties ;
- Developing and applying the Evaluation Model
[
'Ihe process is represented by a once-through flow diagram for simplicity. The actual procvss included many iterations between the various tasks. For example, to better represent the ,
{ observations of the large-scale containment test (LST) dome temperature distribution, due to the j subcooling of the film applied to the LST, the initial EGOTHIC code version used in 1992 was i augmented by the addition of a model for convective heat transport for the liquid film. In j - addition, extensive review by represtntatives of regulttory agencie6 industry, and academia I l were incorpo:ated into the process (Ref. 2.2). The end result is documentation which describes i the PCS DBA Evaluaten Model and its bases in an auditable, traceable manner. Following is [ a brief description of the four elements of the process used to develop the methodology, l 2.2 ELEMENT 1 - AP600 PCS REQUIREMENTS AND CODE CAPABILITIES l The PCS DBA methodology development process began with a review of the AP600 design and ' DBA- scenarios and an identification of phenomena important for AP600 containment
- pressurization. From this review, an initial test program was defined and a computer code was j
- r. elected.
L i A PIRT was developed to identify the key thermal-hydraulic phenomena which govern the [ transients of interest. The PIRT (Ref. 2.2, Section 4) ranks phenomena according to their relative
, importance to the particular transient phase of interest. The PIRT process included input and L review by representatives of academia and regulatory authorities, and cross functional Test and Analysis Process Overview october $
- oA3006w-2.non:1b-101497 Revise : C 3-1
,w 5,--* i-=- ,,----~- ,,,,w . v ..~.. ..---.--Eww..--..e,w.wm,-4 -% , +-.,,m, , --, ..~r,vc-- ~2w-y--vc--n-,-v,e
3-2 , PCS Test and Analysis Process Overview Element Objective OUTPUT
+ scenartoldentmeauon Select and freeze computer code based on phenomenological + Pm I modeling reaulroments for AP600 + code documentauon containment pressure prodletions + EvaluationModel requirernents y ->- scaling assessment Assess code capability to model 2 important phenomene by compartoon + code documenistion (update) to test data and select bounding analysis approach cevalen report and noding guiasnee U
Assess uncertainties and range of + Appropriate methM for bounding each 3 parameters to develop boundina key mput omup models
+ Frozen AP600 noding U
Perform AP600 DBA calculations + Conflrmationthat 4 and compare to success criteria
, ," *mt 4
Figure 21 PCS Test and Analysis Process Overview Test and Analysis Process Overview october 1997 o:\3006w-2.non.1twl01497 Revision 1
23 Westinghouse technical reviews. The bases for high, medium, and low rankings are documented in the PIRT. A key result of the PIRT is that the dominant phenomenon for transferring heat from the containment is mass transfer - condensation on the inside and evaporation on the outside. The mass and energy release boundary condition imposed on the problem is the , primary driver of the containment pressure response, and is ranked high. For the loss-of-ec,lant ., accident (LOCA) scenario, pressurization is mitigated primarily by intemal volume compliance during blowdown, and by internal heat sinks below deck, from blowdown through the transition period when the PCS cooling begins to dominate and turns the pressure around. PCS heat removal dominates the long tenn LOCA response. The main steamline break (MSLB) transient is mitigated primarily by volume compliance and internal heat sinks. A summary of the high and medium ranked phenomena is shown in Table 21. In parallel with bottom-up phenomena evaluations, the EGOTHIC computer code was selected, upgraded, and frozen to allow explicit modeling of many of the phenomena identified in the initial review. As the scaling analysis and testing programs progressed, code upgrades were completed to better model experimental results according to guidelines consistent with computer code lifecycle management. Hand calculations and spreadsheets were used to verify correct programming of the upgrades as documemed within the Westinghouse QA program. Documentation of the code used in the Evaluation Model consists of base GOTHIC 4.0 documentation (Refs. 2.3,2.4,2.5) and upgrades to create EGOTHIC 4.1 (Section 3), 2.3 ELEMENT 2 - ASSESS CODE VERSUS TESTS AND IMPORTANT PROCESSES Analyses and computer code validations were used to identify the most appropriate models and biases to use in the PCS DBA Evaluation Model. The PCS test results were documented, including separate effects (Ref. 2.6) and integral effects (Ref. 2.7). The PCS test data and other data from the literature were used to provide input to code validation (Ref. 2.8). Validation was used to study how the oversimplification inherent in the lumped parameter EGOTHIC model applies to the AP600. The lumped parameter limitations lead to the potential for compensating errors, so that a methodology to bound the effects of compensating errors was identified (Ref. 2.8, page 8-9). The effect of lumped parameter momentum formulation and noding on EGOTHIC results was an important output of validation. Insight from validation was used to develop a bounding Evaluation Model in Element 3. O NJ Test and Analysis Process Overview october 1997 o:\30lEw.2 non.lb-101497 Revuion 1
2-4 Table 21 Phenomena identification and Ranking Table . Summary of liigh and Medium Ranked Phenomena Phenomenon ' Effect on Containment Pi Groups Where Addressed Break Source Mass and The mass and energy source for n p,gt,rk.enth Scaling Analysis Energy (I A) containment pressurization n pgt,,g,,,,g Ap, work,d E ri worQ Gas Compliance (2C) Stores mass and energy in np,, Scaling Analysis atmosphere, increasing pressure initial Conditions Inside Temperature, humidity, pressure parameter Initial Conditions (4 A, 40, 4C) affect noncondensables and energy Section 5 storage Containment Solid }{ cat Store energy (and re nove mass np,gj Scaling Analysis Sinks (3), Pool (5), Drops from atmosphere) reducing up ,,o,y (1), and Shell (7) pressure Internal licat Sink Limits conduction heat transfer parameter Scaling Analysis Conduction (3D, into heat sinks, shell, or pool, and SE,7F) and lleat through shell. Stratification in the Capacity (3E, SA, break pool can affect the effective 7G) heat capacity of the pool. Heat Transfer Water on and noncondensible parameter Scaling Analysis Through Horizontal gases near upward facitig Liquid Films (3C) horizontal surfaces limit heat and mass transfer to horizontal heat sinks Condensation Mass The first-order transport process n p ,,,,y Scaling Analysis Transfer (3F, SD, 7C) that removes mass and energy from the containment gas Break Source Direction, elevation, density, and parameter Circulation and Direction and momentum can dominate Stratification, Elevation (10), circulation and affect condensation Section 9 Momentum (1C), rate. and Density (1D) Circulation and Intercompartment Flow Stratification (2A) (Circulation) and stratification en
............. ....... affect the distribution of steam parameters Intercompartment (and noncondensables) near heat sinks for condensation heat Flow (2D) removal.
Source Fog (2D) Affects circulation and parameter stratification via buoyancy e Test and Analysis Process Overview octotier 1997 o:\3006w.2.non:lt>.101497 Revinon 1
25 Pbeaomena identification and Ranidng Table . Summary of liigh and Medium
^
Table 21 1 (cont.) Ranked Phenomena Phenomenon
- Effect on Containment Pi Groups Where Addressed k Evaporation Mass First-order transport process that n,, ,, m Scaling Analysis Tranrfer (7N) removes mass and energy from neb ,5 rLwork the evaporating extemal shell n r. work,d A p work,p PCS Natural Convective air flow provides parameter Scaling Analysis Circulation (9A, convect!ve heat and mass transfer 13A) from containtnent shell.
1.iquid Film Flow Affects the upper limit for water parameter Film Stability, Rate (8A), Water coverage on the external shell and Section 7 Temperature (8B), amount of water available for Film Stability (SC) evaporation. IJguld Film Energy inside: n,,,,,, Scaling Analysis Transport (7E,7M) Carries condersation energy to the IRWST and break pool. See note 1 Outside: Absorbs energy rejected by the external shell surface. Convection licat Transfer A second order transport process np,q) Scaling Analysis (3G, 7H,10A,10B) that removes energy from the n,,qm+ n,,q, q containment gas, and from the gote2 tg external shell. Radiation Heat Transfer A second order transport process np,q) Scaling Analysi-(3H. 71) that removes energy frem the n,,qng n,,q, containment gas and from the di extemal shell, dote 2 Baffle Conduction (10D) Conduction through the bafile into n,,q 3,, PIRT Sections and Baffle 1.cakage Paths downcomer volume and leakage n e.g.bt, 4.4.10D and (10G) paths can influence the extemal None for 4.4.10G natural circulation flow rates leakage Indicators in parentheses refer to phenomena in the " Phenomena identification and Ranking , According to Effect on Containment Pressure" (Reference 2.2, Table 41) Note 1. 9.e fraction of the intemal condensation carried away by the liquid film is defined by the ratio:
/
n,j i (n,,,,i+dn,,,g,i), suticoole for each heat heat capacity of thesinkexternal
- 1. The fraction liquid of the is external definedshell by heat the rejection ratio: that goes into the We ==/(R e, Se tic.n S. quee A ,q m+8 e
.fg.en+"e q du). The pi group values for AP600 are presented in Refetcace 2.9, Note 2. Inside containment agi represents the pressure effect of senale heat transfer. The sensible heat transfer is approdmately 1/2 radiation heat transfer and 1/2 convection heat transfer. Outside m+n .e du represents the sum of the dry and evaporating shell sensible hest transfer, that containment is approximate n,fy 1/2 radiation heat tra.nsfer and 1/2 convection heat transfer.
O V
^
Test and Analysis Process Overview ociater 1997 oAX4w-2.nortit> 101497 Revinon 1
> 24 A scaling evaluation of AP600 was performed (Ref. 2.9) which provided additional confirmation of the PIRT phenomena and ranking. Scaling identified the appropriate nondimensional
~ parameter:,'.he effecta of facility scales, and the ranges of parameters expected in AP600. Scaling was also ucd to identify distortions in the LST facility and to evaluate the effect of distortions on the use of tha LET for studying lumped parameter code biases. T'ie results of scaling, test!ng, and code validation were used to establish a boundim; analysis apped tos each ci the PlRT phenomena, documented in Reference 2.2, Section 4.4. 2.4 ELEMENT 3 - ASSESS UNCERTAINTMS AND DEVELOP BOUNDING MODELS Uncertainties were assessed, and together with the results of code validation, were used to develop a method of applying the WGOTHIC lumped parameter formulation to create a bounding DBA Evaluation Model. Key results are summarized as follows. i it is worthwhile acting how representative high ranked phenomena are addressed for the AP600 PCS in the context of understanding this overview. In this regard, some background on lumped parameter containment codes follows, and then a summary is given of how uncertainties are i handled for two representative phenomena, the heat and mass 'ransfer rate correlations, and { d circulation and stratification. , Th. application of EGOTHIC lumped parameter formulation for the PCS Evaluation Model has been justified by conservatively addressing lumped parameter biases (Appendix 9L, Section 9.C.3.4). Lumped parameter contahunent codes have been used for nuclear power plant licensing calculations for over 30 yeacs. Limitations of the lumped parameter approach for containment modeling are documented in the literature. Generally, lumped parameter codes can reasonably predict global parameters, such as pressure, but the lumped parameter formulation oversimplifies physics when local details are important. Far containment analysis, details within a mlun e are impcrtant when the physics of stratihcadon within a volume or entrainment into jets or plumes is important. Coupling of the MGOTHIC lumped parameter nodes, with one or rnore distributed parameter volumes to gain some resolution of the details within a volume, can increase the accuracy of the solution. However, while distributed par 4 meter calculations were used to help undeistand test results, the use of such more detailed models was not practical for PCS DBi calculatiens due to computing requirements. Complex thermal hydraulic models may pmduce results that match or bound test data but may also include compensating errors. Sufficient data were obtained on the important variables in the LST to isolt.te compensating errors in the lumped parameter model. Studies of IST calculations have shown that the compensating errors in lumped parameter calculations arise from offsetting effects of steam concentration and velocity. Because the jet source is numerically expanded to uniformly fill the volume flow area in a lumped perameter node, numerical h october 1997 Test and Anahsis Process Overview c:\300bw-2mIt>101497 Revision 1
27 entrainment leads to high predicted velocities in the above-deck region and a resultant homogenization of the containment. Mixing of noncondensables from the below-deck region in the LST penalizes PCS heat transfer because the noncondensables from below-deck penalize condensation rates. Overpredicting velocities benefits PCS heat transfer because of forced convection enhancement. In the Evaluation Model, the competing elfects are addressed by using only free convection inside containment, thereby climinating the influence of vehicity overprediction. Tids results in a bounding prediction relative to the potential for compensating errors. After developing an understanding of lumped parameter model performance, bounding approaches to address important phenomena, summarized in Table 21, were developed. Uncertainties are addressed by quantifying a bias and distribution for a phenomenon or by studying the range of expected AP600 conditions and establishing an upper bound approach. Example of the two approaches follow, using mass transfer correlation and circulation and stratification. Separate effects tests (SETS) and LST data have been used to select appropriate heat and mass transfer correlations from the literature and develop biases to bound the data (Ref. 2.6, Section 4.5). A lower bound for heat transfer through the containment shell to the ultimate heat sink is therefore used. One of the more complex issues is the coupling of circulation and stratification, break direction and momentum, and intercompartment flow, and the impact of those parameters on internal heat sink utilization. Circulation and stratification are complex physical processes that are not easily solved by numerical methods. Since the AP600 relies on passive cooling by natural circulation, there are no active systems to force the atmosphere to homogenize. Based on a study of plausil.le break scenarios (mass and energy, momentum, direction, and elevation), bounding, or extreme ca as are identified for further study. The extreme cases are studied using first principles calculations and sensitivities to specific flow patterns of interest. The lumped parameter plant model, with above-deck noding based on noding frozen for the LST evaluations, is used to calculate the containment response for the specified flow patterns. Based on the sensitivities, a limiting scenario is chosen for use as the PCS DBA to bound the impact on mass transfer of the strongly coupled phenomena. Blases are introduced with lumped parameter compartment nodes to bound the effects of stratification, and an asses > ment of stratification effects on PCS heat a tmoval through the shell shows that no net penalty on heat removal from the above-deck region need be applied. This is discussed in more detail in Section 9 (See Table 91). Similar evaluations have led to the definition of a bounding Evaluation Model for important phenomena identified in the PIRT and documented in Reference 2.2 Section 4.4. O b Test and Analysis Process Overview cetober 1997 oA3006w.2.non:ll*101497 Revision 1 1
28 l 2.5 ELEMENT 4 - PERFORM DBA CALCULATIONS AND COMPARE TO SUCCESS CRITERIA gl The Evaluation Model was developed as previously described to produce conservative, bounding pressure transients for each accident phase. The acceptance criteria are that the peak pressure must remain below the design pressure of 45 psig, and pressure should be rapidly reduced, consistent with assumptions in radiological release calculations, which is typleally interpretated as the pressure at 24 hours should be less than one half of the design pressure. Documentation is provided in Reference 2.2 that shows for each phenomenon:
- Relevant model i. e code
. Test basis . Report references . Sum. mary report conclusions . Validation basis summary . How validation results are used . How uncertainty is addressed
2.6 CONCLUSION
S A structured, traceable approach has been followed to develop the AP600 PCS DBA Evaluation Model. The PIRT has been used to develop a bounding Evaluation Model and the PIRT has been used as the basis for a road map to relevant supporting information for each phenomenon.
2.7 REFERENCES
2.1 Boyack, D.E., et al., "An Overview of the Code Scaling, Applicability, and Uncertainty Evaluation Methodology," Nuclear Engineering and Design, Volume 119 (1990) No.1, May 1990, pp 1-15 2.2 WCAP-14812, " Accident Specification and Phenomena Evaluation for AP600 Passive Containment Cooling System," Rev.1, June 1997 2.3 NTD-NRC-95-4563, " GOTHIC Version 4.0 Documentation", September 21,1995 2.4 NTD-NRC-95-4577, " Updated GOTHIC Documentation", October 12,1995 2.5 NTD-NRC-95-4595, "AP600 }VGOTHIC Comparison to GOTHIC," November 13,1995 2.6 WCAP-14326, " Experimental Basis for the AP600 Containment Vessel Heat and Mass Transfer Correladons," Rev.1, May 1997 Test and Analysis Process Overvir..v october 1997 c:\3006w.2..wn:1b 101497 Revision 1
29 2.7 WCAP 14135, " Final Data Report for PCS Large Scale Tests, Phase 2 and Phase 3," Revision 1, April 1997 2.8 WCAP 14382, *}YGOTlilC Code Description and Validation," May 1995 2.9 WCAP 14845, " Scaling Analysis for AP600 Containment Pressure During Design Basis Accidents," Rev. 2, June 1997 h O Test and Analysis Process Overview odober tw7 a\3006w-2.non:1b-101497 Revision 1
WESTINGHOUSE NON-PROPRIETARY CLASS 3 O Section 3 Overview of WGOTHIC O O n\3006w 3.non:1b-101497
111 TABLE OF CONTENTS - LIST OF TA BLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
- LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
3.1 INTRODUCTION
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 3.2 OVERVIEW OF THE CODE DEVELOPMENT AND VALIDATION . . . . . . . . . . 3-1 3.3 THE WGOTHIC CLIME MODEL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-10 3.4 GENERAL CLIME EQUATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-13 3.5 INTEGRATION OF THE WESTINGHOUSE CLIME MODEL INTO GOTHIC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 18
3.6 REFERENCES
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-20 LIST OF TABLES Table 3-1 GOTHIC Validation Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-5 Table 3-2 GOTHIC Phenomenolagical Models Validated by Test . . . . . . . . . . . . . . . 3-6 LIST OF FIGURES Figure 3-1 Summary of GOTHIC Historical Development , . . . . . . . . . . . . . . . . . . . 3-2 Figure 3-2 GOTHIC Modeling Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-3 Figure 3-3 Summary of WGOTHIC Historical Development . . . . . . . . . . . . . . . . . . . 3 7 Figure 3-4 Westinghouse-GOTHIC Clime Wall Source Term Models . . . . . . . . . . 3-12 Figure 3-5 Clime Finite Difference Model Defialtions . . . . . . . . . . . . . . . . . . . . . . . 3-14 Figure 3-6 Clime Routines Flow Control Outline . . . . . . . . . . . . . . . . . . . . . . . . . . 3-19 Overview of WGOTHIC october 1997 o;\3006w.3.ron:1b 101497 Revision 1
31 3.1 IN1RODUCT10N O The panive containment cooling system (PCS) phenomena were identified and ranked by order of importance in determining the vessel pressure in a phenomena identification and ranking table (PIRT). The important phenomena are summt.rized in Section 2. Existing containment analysis codes were reviewed to determine which most closely met the requirements identified in the PIRT. Although none of the codes met all of the requirements, the GOTHIC code package (Reference 3.1) was selected for further development based on its validation history and modeling capability. This section provides an overview of the GOTHIC code and describes the changes made to the GOTHIC solver program to incorporate the special heat and mass transfer correlations, liquid film tracking, and the wall to-wall radiation model for performing design basis analyses for PCS .ype containments. 3.2 OVERVIEW OF 'INE CODE DEVELOPMENT AND VALIDA110N The GOTHIC code is a state-of the art program for modeling multi phase flow. & GOTHIC code has been developed through a long history from other qualified thermal-hydraulle-- computer codes (as shown in Figure 31). GOTHIC consists of three separate programs, the preprocessor, solver, and postprocessor. The preprocessor allows the user to rapidly create and modify an input model. The solver performs O the numerical solution for the problem. & postprocessor, in conjunction with the preprocessor, allows the user to rap!dly create graphic and tabular outputs for most parameters in the model. The GOTHIC rolver program calculates the solution for the integral form of the conservation equations for mass, momentum, and energy for multi-component, two-phase flow. .The conservation equations are solved for three fields: continuous liquid, liquid drops, and the steam / gas phase. W three fields may be in thermal nonequilibrium within the same computational cell. This would allow the modeling of subcooled drops (for example, containment spray) falling through an atmosphere of saturated steam. The gas component of the steam / gas field can be comprised of up to eight different noncondensable gases with mass balances performed for each comynent. Relative velocities are calculated for each field, as well as the effects of two-phase slip on pressure drop. Heat transfer between the phases, surfaces, and the fluid are also allowed. The GOTHIC solver program is capable of performing calculations in three modes. A model can be created in the lumped parameter nodal-network mode, the two-dimensional distributed parameter mode, or the three-dimensional distributed parameter mode. Each of these modes
. may be used within the same model (as shown in Figure 3-2). The lumped parameter nodal-network mode is used for the AP600 containment Evaluation Model.
O Overview of WGOTHIC . odober 1997 oA3006w.3cortib.101497 Revwon1
^
i 32 ! O COBRA IV 1973 4 U y U 77 75 979 U U 85 982 U u 3 1986 u u 983 985 u 1986 u 19 9 Figure 31 Summary of GOTHIC lijstorical Development Overview of lyGOTHIC octcrier 1997 n\3006w.3.non:1b 101497 Revakm I
1
- 33
.i IQ i h ! l l i ro b \ 8 i ( )i 1 i I. ' lG n f,.>.,:....'g..:._-
, ~>,.... .~ .. , . ,, , , .u,. ; E i ..: N,, ;
l v ;. . .: .,. . . .;, . l ...q...?.. 4
,;..s ,,. . ,, , e , ;
q.......;.... i
- ,l;,,l. .....
, v,. ; i . , e
+ ,. , ,,, ; - ....,... ;...q.. 7...;... q....
4
, . ,, ,l.< ... l. <.l i l l
- l
, .l,.. .<....... .
- . , . , , i ....,. , e
..c, i . ,
i
...>...:...;..s...a,...'...
l MA !!! ! ! ! ; 5r , I 1 J i l , 1 a , Figure 3 2 GOTHIC Modeling Features J Overview of WGOTHIC Octoter 1997
' Revision 1 oA3006w 3.non:1b-101497
I 3-4 The GOTHIC code also contains the options to model a large number of structures and components. These include, but are not limited to, heated and unheated conductors, pumps, fans, a variety of heat exchangers, and ice condensers. These components can be coupled to represent the various systems fo md in any typical containment. They are not used in the AP600 analysis desc11 bed in this report. The GOTHIC code ha.= an extensive validation history which was an important consideration in the selection of the code for further development for mode %g of the PCS. The GOTHIC code validation program includes both e comparison of code-caletdated results with analytical solutions to specified standard problems and a comparison of code-calculated results with experimental data. The results of the EPRI sponsored GOTHIC code validation program are presented in Reference 3.1, Enclosure 1. Table 3-1 lists some of the tests used in the GOTHIC code validation program. The phenomenological models validated by each test are cross-referenced and presented in Table 3-2. In addition, industry experience using GOTHIC in the lumped parameter mode, as well as attempts to improve results using muld-dimensional analyses, are described in Appendix 9.C.3. Westinghouse purciased Version 3.4c of the GOTHIC code in 1991 and began modifying it to include mechanistO convection heat and mass transfer correlations, a liquid film tracking model, a one-dimensional wall conduction model, and wall-to-wall radiant heat transfer to model heat removal by the PCS The code with modifications, is called Westinghouse-GOTHIC and is abbreviated as WGOTHIC. The WGOTHIC development h' story is shown in Figure 3-3. The PCS heat and mass transfer models developed by Westinghouse were incorporated into the GOTHIC version 3.4c pre-processor and solver programs to create the WGOTHIC version 1.0 pre-processor and solver programs in 1993. Between 1991 and 1993, while Westinghcase was developing the PCS heat and mass transfer models, GOTHIC version 3.4d underwent an EPRI-sponsored peer review. The purpose of the review was to establish a reference point for placing the GOTHIC code package under a 10 CFR 50, Appendix B Quality Assurance Program. The peer design review group reviewed the documentation, coding, convergence, pre-/ post-processor, code qualification package, and the code's adequacy for containment analysis. The conclusions from the review are presented in Section 2.2 of tne GOTHIC Design Review Final Report (Reference 3.2). Overall, the GOTHIC containment analysis package was found to be adequate for containment analyses, and that the code package offered the ability to provide more accurate and mechanistic results than with other currently available containment codes. This conclusion was qualified with the statements that the nodal and junction treatment, as well as the range of the e Overview of LVGOTHIC cktober 1997 c:\3006w-3.non:1b-101497 Revision 1
3-5 o U Table 31 GOTillC Validation Tests Battelle-Frankfurt Tests D-1, D-15, D-16 Modeling: 7 lumped parameter volumes, junctions Phenomena: Blowdown transients, subes,mpartment pressurization, wall differential pressures Battelle-Frankfurt Test 6 Modeling 1 distribut-d parameter volume (55 cells), conductors, junctions Phenomena: Hydrogen transport by convection and diffusion Battelle-Frankfurt Tests 12,20 Modeling: Combination of 5 lumped and I distributed parameter volumes (2 cells), conductors, jurictions Phenomena: Hyarogen transport by convection and diffusion Battelle-Frankfurt Tests C-13, C-15 Modeling: 10 lumped parameter volumes, conductors, junctions Phenomena: Main steamline break, pressure / temperature response Hanford Engineering Development Modeling 1 distributed parameter volume (300 cells), Laboratory Tests HM 5, HM-6 conductors, junctiotu p) ( Phenomena: Hydrogen mixing in a large, simulated containment Light Water Reactor Aerosol Modeling: Combination of 1 lumped and 1 Containment Experiments Tests LA-5, distributed parameter (2 cells) volumes, conductors, LA-6 junctions Phenomena: Severe accident response to sudden containment failure Marviken Full-Scale Containment Modeling: 21 lumped parameter volumes, conductors, Tests 17,24 junctions Phenomena: Pressurized high temperature steam blowdown Carolina's Virginia Tube Reactor Tests 3, Modeling: 2 lumped volume and a 2 distributed 4, 5 parameter volume (20 cells) models, conductors, junctions Phenomena: Steam blowdowns (T31.5 includes hydrogen / helium) Heissdampfreaktor Tests V21.1, T31.1, Modeling 37 lumped parameter volumes, conductors, T31.5, V44 junctions Phenomena: Steam blowdowns (T31.5 includes hydrogen / helium) O V Overvie.v of }VGOTHtC october 1997 c:\3006w-3.non:1b-101497 Revision 1
3-6 Table 3 2 GOTHIC Phenomenological Models Validated by Test item BFMC HEDL LACE MARV CVIR HDR Fluid momentum X X X Energy transport X X X Noncondensable gases X X X X X X Equations of state X X X Pressure response X X X X X X Temperature response X X X X X X Humidity response X X X X X X Hydrogen transport X Energy sources X X X X X Subcompartment analysis X X High energy line breaks X PWR standard X containment BWR pressure X suppression Fluid / structure interaction X Conductors X Subdivided volumes X Turbulence X 3-D calculations X X X l Overview of }VGOTHIC October 1997 0:\3006w-inon:1b-101497 Revision 1
3-7 i, t !O - GOTHIC 3.4c p 1991 j u j i f IP-GOTHIC 3.4d 1 Developes Basic i 1991 Clime Methodology l u i t I 1 Adds Mixed Convection i Heat Transfer y Correlations to l Clime Methodology l EPRI Sponsored g Peer Review - for Nuclear Q/A WGOTHIC ' 1 i u ! O l X Developes
- Subcooled Film Clime Methodology -
U l 199 ! u i ! l Validates
! WGOTH1C for AP600 Containment Analysis ~
1995 i U _.7 Integrates { l 'E"> "deied;; v 1996
=C ini'D {
l X Corrects Clime ' _ Dryout Error !- U .
- - i i
s G '"'97 '9 a Figure 3-3 Summary of HGOTHIC Historical Development t i I Overview of HGOTHIC october 1997 o:\3006w-3.non:1b-101497 Revision 1 4 i
= - .--
l 3-8 qualification database, need to be justified for each intended application; as was done via the large-scale (LST) and separate effects tests (SETS) (Reference 3.3) and various scale integral tests (Appendix 9.C.3) used to qualify WGOTHIC for use in the licensing of the AP600. The review group had three recommendations. The first was the addition of dynamic memory allocation, so that the code would not need to be recompiled for different sizes of models. The second was the inclusion of an iterated Newton method to aid in convergence. The third was to incorporate a fog model to simulate condensation of vapor when regions go from superheated to saturated. As described in Reference 3.4, the conclusions and recommendations from the GOTHIC design review also apply to EGOTHIC. None of the recommended chan<;es were incorporated in EGOTHIC. The first recommendation, dynamic memory allocat 1, wasn't incorporated in EGOTHIC, since it is a user convenience option and does not affect the solution technique. The second recommendation, to include an iterated Newton solution option to aid in convergence, was not incorporated in EGOTHIC, since satisfactory convergence was supported by the comparisons presented in the GOTHIC code qualification test report (Reference 3.1) and the EGOTHIC validation report (Reference 3.3). The third recommendation, to include a fog model, was not incorporated in EGOTHIC because it was concluded that, based on the GOTHIC CVTR qualification test case results (Reference 3.1) and an assessment of fog modeling as it relates to the AP600 (Reference 3.5, Sections 4.4.2D and 4.4.9C), it is conservative with respect to the prediction of containment temperature and pressure to not include the fog model. Westinghouse updated the PCS models to account for subcooled films and incorporated the GOTHIC software error corrections that were provided by NAI to create EGOTHIC pre-processor version 2.0 and solver version 1.2 in 1994. Westinghouse validated this version of WGOTHIC for performing AP600 analyses in 1995 (Reference 3.3). The WGOTHIC validation program consisted of four parts:
- 1. The subset of GOTHIC validation tests that was identified as sensitive in the original acceptance tests was rerun with WGOTHIC. These tests were run with the same input options selected in the original GOTHIC validation calculation (that is, the PCS models were not exercised) to determine if any of the code changes made to incorporate the PCS models would affect the transient results. This comparison is presented in Appendix D of Reference 3.3. It shows that the code changes Westinghouse made to incorporate the PCS models do not affect the GOTHIC calculation results.
- 2. The PCS model one-dimensional conduction equation solution technique was validated by comparison with an analytical solution for a test problem. This comparison is presented in Section 4.1 of Reference 3.3. The code calculated results match the analytical solution.
O Overview of }yGOTHIC october 1997 o:\300ow-3.non:1b-101497 Revnion 1
4 1 3. The PCS model heat and mass transfer correlations were validated by comparison with separate effects test data from the Westinghouse Flat Plate Tests, the Westinghouse [ Large-Scale Tests, the Wisconsin Condensation Tests, and publicly available published reports. These comparisons are documented in Reference 3.6. The correlations are acceptable for modeling heat and mass transfer from the AP600 PCS.
- 4. The EGOTHIC code, including the PCS'models and nodalization, was validated by
. comparison with transient test data from the Westinghouse Large-Scale tests. l- Comparison with test data from the LSTs validated the ability of WGOTHIC to represent j internal flow fields and noncondensable gas distributions and to calculate the net heat
- removal from the vessel in an integral system. The validation provided insight for the -
l applicability of documented lumped parameter biases (Appendix 9.C.3) to AP600 and l l identified a bounding approach to address compensating errors. This comparison is
- presented in Section 8 of Reference 3.3. Section 9, Table 9-1 summarizes how lumped j parameter biases have been addressed.
. y j in 1996, the source code for the PCS heat and mass transfer models for EGOTHIC solver i version 1.2 and pre-processor version 2.0 was incorporated into the GOTHIC solver and pre-processor version 4.0 source code to create the EGOTHIC version 4.0 pre-processor and solver I programs. This was done to incorporate all of the GOTHIC design review code changes into
~WGOTHIC.
O A series of verification tests, including the most sensitive GOTHIC code qualifications test cases, were run to validate EGOTHIC version 4.0. The results of the GOTHIC code qualification test l'
. cases that were run using the EGOTHIC version 4.0 all compared very well with the results obtained using GOTHIC version 4.0, indicating that the incorporation of the Westinghouse PCS model did not significantly affect the GOTHIC calculations.
Version 4.1 of the WGOTHIC pre-processor, solver, and post-processor programs was created in 1997 to correct an error that was discovered in the PCS heat and mass transfer model and several other non-calculational code problems. ' The error caused the PCS heat removal to be overpredicted' at the point of dryout. Verification _ test cases. performed using WGOTHIC
-version 4.1 demonstrated that the dryout error was corrected. Version 4.1 of WGOTHIC has c been used for all of the analyses presented in this report except as specifically noted for 1 sensitivity studies in Section 11.
i 4 i q u l Overview of WGOTHIC october 1997 k o:\3006w-3.non:15101497 Revnion 1 l s
3-10 3.3 THE WGOTHIC CLIME MODEL A solution technique that includes wall to-w;11 radiation at the conditions expected for the 9 AP600 plant design necessitates a close coupling of the participating walls. This coupling is accomplished by assigning boundaries that define the portions of the various walls that radiate to one another. Consistent with the basic formulation implemented for the GOTHIC code that considers conductors or heat sinks to be energy sink or source terms, code modifications that include wall-to-wall radiant heat transfer can be thought of as the addition of a special type of conductor group. This special conductor type or group consists of a set of walls that radiate to each other and interface with GOTHIC fluid cells through mass and energy source terms. The term clime, meaning region, is used to differentiate and distinguish this special conductor type from those already existing in GOTHIC terminology. For the AP600 containment model, a clime is a horizontal slice of the containment structure consisting of the following: The heat and mass transfer source terms from the containment volume to the shell Liquid film mass and energy conservation and thermal resistance on shell, baffle, or shield building surfaces. Conduction through the shell Heat and mass transfer source terms from the exterior shell to the riser air flow channel
. Radiation from the exterior shell to the interior baffle Heat and mass transfer source terms to the interior baffle from the riser air flow channel
- Conduction through the baffle Heat and mass transfer source terms from the exterior baffle to the downcomer air flow channel Radiant heat transfer from the exterior baffle to the interior surface of the shield building Heat transfer source terms to the interior surface of the shield building from the downcomer air flow channel Conduction through the wall of the shield building O
Overview of WGOTHIC october 1997 ! o \3006w-3.non:1b-101497 Revision 1
1 - i 3-11 ; 9 i
* - Both radiant and convective heat transfer from the exterior surface of the shield building
- to the environment
' A representative AP600 three-conductor clime is shown schematically in Figure 3-4. The iniernal ;
- containment vessel volume, riser air flow channel volume, downcomer alt flow channel volume,
!- ' and environment volume are separate computational cells or fluid volumes in the model. The [ shell, baffle, and shield building walls are one-dimensional conductors representing solid wall structures between the computational cells. These conductors are further subdivided into regions i of different materials with different mesh sizes. Each conductor surface may have a liquid film present (not shown) depending on thermodynamic conditions. l l The climes are stacked vertically through the PCS to model the effects of changing properties both inside and outside the containment shell. Usually there are at least two stacks of climes ! a wet stack and a dry stack. The only difference between a wet and dry stack is that a time- [ dependent, water flow rate boundary condition is specified for each conductor surface of the top l- clime in a wet stack. Because condensation can occur on either wet or dry conductor surfaces, [ an initially dry stack of climes could contain some wet conductor surfaces and/or a partially wet ! - conductor surface due to condensation. Likewise, an initially wet stack of climes could contain
- . some dry conductor surfaces and/or a partially dry conductor surface due to evaporation.
e i j- The user must specify values for the area and circumferential perimeter for.each conductor of [ each clime in both the wet and dry stacks. For the AP600, the input values for the area and j circumferential perimeter for the clime conductors in the wet stacks are based on measurements of the water coverage from the full-scale water distribution tests. - The water coverage input l modsl, which conservatively bounds results from several test facilities, for the AP600 l containment Evaluation Model is described in Section 7. j . The WGOTHIC clime model calculates the temperature, flow rate, and thermal resistance of the -
- ' water films on the various conductor surfaces of a clime. Liquid mass is conserved whenever I. the film reaches the bottom clime in a stack or a conductor surface dries out. The clime model takes the film flow rate from each conductor surface of the previous clime in the stack as input, l
j then adds the local condensation rate, or subtracts the local evaporation rate to determine the [ output water flow rate 'on each of its corresponding conductor surfaces. Any liquid film
- remaining on the conductor surfaces of the last clime in a stack is added to the liquid field of j the GOTHIC cell in contact with the conductor surface, or an altemate drain cell specified by
- user mput.
Dryout occurs when either the film flow rate is low enough or the heat flux is high enough to
- _ result in complete evaporation of the film before it can exit the conductor. The clime model calculates the evaporation heat and mass transfer and the location of the dryout elevation; the remainder of the conductor surface below the dryout elevation is treated as a dry surface.
4
' Overview of EGOTHIC october 1997 o:\3006wanon:1b-101497 Revision 1
3-12 O 1 1 1 1 I i 1 1 Inside i Air i I Air i Environinent Containtuent l l'pflow ! ! Downflow l 1 Clnannel j 1 Cliannel i 1 1 8 8 I I I I i 1 1 1 1 1 1 1 Shield Vessel i I I Baffic 1 Building Wall i I I Wall I Wall l l l 1 l l 4 l
! i ! Clime i I I I Boundary Qvensect }eonvect Oronaense Q mass (raatnet) raat net i h
4 Clitne Boundarv i l i i 1 i 1 l i GOTillC Clitues CLIME Ileat ana Mass Transfer Hautines Interfere G0711:0 Volume Properties are Used in will4 West 6nghnuwe-GOTHIC Volume snurre Terms the CLIME Heat and Ma.m Transfer ruitelations. Figure 3-4 Westinghouse GOTHIC Clime Wall Source Term Models Overview of }yGOTHIC oct<ter 1997 o:\3006w-3.non:1b101497 Revnion 1 L
4 3-13 3.4 GENERAL CLIME EQUA'I1ONS O The energy equation for the film must balance the heat from the wall into the film, the heat 4 conduction through the film, and the heat and mass transfer from the film surface to the ambient, with the change in energy of the flowing film. Assuming constant fluid properties over L the node surface, one-dimensional film flow along the wall, one-dimensional conduction across - the film, and that the viscous dissipation term can be neglected, the general energy transport equation for the film can be written in terms of temperature as: i i. 2 ! BT 0T T.k OT 2
+ y*BZ_ (3-1) l pep 0x -
i l For computational purposes, the water film is divided into two layers as shown in Figure 3-5, j- The inner boundary of the film touches the wall and its temperature equals the wall temperature.
- The outer boundary touches the atmosphere and its temperature is coupled to the temperature i of the atmosphere through the heat and mass transfer boundary layer correlations. The temperature at the center line between the layers represents the average heat stored in the film.
- Referring to Figure 3-5, the film energy transport equation can be expressed in a finite difference l form as follows
i Tava Tavs, old , 4kgi , Tsurf,1 - 2T avs
+Twall,1 Tin - Too, at pmm Cp, film -
2 AZ 6x 1m (3-2)
- where:
kmm = film thermal conductivity (Bru/ft-*c *F) 6x mm = film thickness (ft) j c,mm p
= film heat capacity (Btu /lbm *F)
- pmm_ = film density (Ibm /ft3)
Tin = inlet temperature of film at the top of the clime (*F) ! Toet . '= exit temperature of film at the bottom of the clime ('F)
, Tavs = temperature of the center of the film (*F) H T,,ii,3 = temperature of first wall' node (*F)
F Tsurf,1 = film surface temperature ('F) p AZ = height of the clime (ft) vz = film velocity (ft/sec) lt. Overview of WGOTHIC October 1997 i - o:\3006w4.non:1b-101497 Revision 1 l
- ::w; y - --r -
em m g +
3-14 _ t O Clime Boundary IIN 1 I
/ / /
I
)'X /
[
/
l l l l / 1 l I
/ / z 0
l/
/
l lOX WAU. !
! l / / ~ ~
I i i i / bXmmll T
/ " I I 3yo /
l l 6,/ j 3 e iI I
' / /
IWAIL T,8311oTf,itz Tsvar.1 sunr.a
/
T.AIR I 1
/ ! !I / !I / /
u I
/ / [
74 _ _ _ _ _ . Clime Boundary Figure 3-5 Clime Finite Difference Model Definitions Overview of WGOTHIC october 1997 o:\3006w-3.non:1b-101497 Revision 1
- _ _ _ .- - - - .. . - _-- - - _. - - . ._- _ - - . - . . - -~-
3 15 The film inlet temperature is given, either from a boundary condition or from the outlet
._[ temperature of the preceding clime in the stack. As a stability criterion, the film outlet temperature is defined to be the same as the average temperature.
Tout - T,yg (3-3)
.The inner film surface boundary condition forces the heat flux from the outer surface of the conductor wall to equal the heat flux into the inner layer of the film. The inner film surface
, boundary condition is: l Twgi,1 - Twall,2 T*"8 " T**II'I
= 2k g (M)
AXwa!! 6 xfit, where: 1 Axwau = wall node layer thickness (ft) Tw.ti,2 = temperature of second wall node ("F) kw.ii = wall thermal conductivity (Btu /sec-ft 'F) The outer film surface boundary condition equates the energy leaving the outer film layer
- surface to the energy entering the atmosphere. The energy leaving the film surface m.'.y enter the atmosphere through a combination of convection, evaporation, and radiation. The outer film i surface boundary condition is
2k alm Isurf.1 - T*"8 hc(Tsurf,1 - T,1,) + hy h is(P'[-Pg 4 4
)+cc(Tsurf,1 - Tsurf,2) @5)
X film where: he = convection heat transfer coefficient from the film to the air (PR1/sec-ft 2.op) T ,, = air temperature ('F) hu = mass transfer coefficient (Ibm /sec-ft2-psi) hfs = latent heat of vaporization of the film (Btu /lbm) 7 P,*,$ = Partial pressure of steam in the air (psi) t )
\J
- Overview of WGOTHIC october 1997 o
- \3006w-3.non:1b 101497 Revision 1 1
3-16 film P = saturation pressure of steam at the film surface temperature, Tsud,1 (PSI) c = emissivity of film surface o = Stefan-Bolzman constant Tsud,2 = temperature of second radiative surface (*R) The four film equations are: T,yg - Tavg,old 4k film Tsurf,1 - 2Tavg + Twii,3 Tin - Tot,,
= +V Z (3-6)
At pfilmcp, film 2 E 6 xf. T w,gg,3 - Twall,2 T*VK T**ll 1
= 24 (3-7) ball 6xfilm 1 "f 2k am a =h e(Tsud.1 - T,i,) + hy his(P"l'n-P*) + co (Tfud,1-Tfud,2) (3-8) 6xfilm Tout = T,yg (3-9)
The wall conduction equation is tightly coupled to these film equations. For points within the wall, the conduction equation is simply a one-dimensional partial differential equation: 2 b=k BT (3-10) dt pcp 3x 2 r e Overview of LVGOTHIC ektober 1997 m\3006w-3.norutt> 101497 Revision 1
3 17 By replacing the derivatives with finite differences, this partial differential equation is replaced with a system of algebraic equations. The superscript "n" identifies the point (node) at which
- the derivatives are to be calculated.
Iwall.n -Iwall,n,old , kall Iwall,n4 ~ 2Iwall,n + Iwall,n-1 3, At pw gic p,w g3 This equation, along with Equations (3-6 through 3-9), can be considered to be the system of equations for a clime. ( O (3-12) (3-13) Ja,e Equations 3-6,3-9, and 3-11 through 3-13 represent the complete system of equations for a clime
. as used in WGOTHIC. See Section 4 for the description of how climes are implemented for the AP600.
Overview of EGOnilC october 1997 c \3006w-3.nortib-101497 - Revision 1
3-18 3.5 INTEGRATION OF TIIE WESTINGHOUSE CLIME MODEL iNTO GOTHIC The Westinghouse clime model is composed of a set of subroutines. These subroutines were added to the GOTHIC solver program to create the 31 GOTHIC solver program. The GOTHIC solver program logic was modified to incorporate the clime model as follows:
. A call to the subroutine that reads the clime input was added . A call to the subroutine "gshell", the main calling routine for the clime model, was added
- A call to the subroutine that generates the clime output was added The clime model flow control outline is shown in Figure 3-6. Subroutine "gshell" is the main calling routine for the other subroutines of the clime model. Separate subroutines in the clime model compute the heat and mass transfer coefficients between the conductor surfaces and the corresponding volumes, the surface-to-surface radiation heat transfer, the conductor wall temperature distribution, and the changes to the source terms for the GOTHIC mass and energy conservation equations.
The interface between the clime model and GOTHIC takes place through the source terms for the GOTHIC mass and energy conservation equations. The GOTHIC vapor mass and energy source terms are updated to include the mass and energy transfer due to convection, radiation, evaporation, and condensation within the climes. The GOTHIC liquid mass and energy source terms are updated to include the liquid mass and energy transfer due to runoff or stripping of the liquid film from the climes. O ! Overview of EGOTHIC october 1997 o:\3006w-3.non:1b-101497 Revision 1
. . . . _ _ . _ . . . . . . . _ _ _ . - _ _ . _ . . _ . . . - . _ . . _ . . _ ~ ~ . . _ . . _ . . . . - . - _ . . _ . . . . _ - . . . _ . _ . . . - _ .
4 4 l t 3-19 , a a,b.c - 4 J, - d. ,2 . i- t Y1-y v i
)
4 : T r-f i t i A
- =
i-i
. r 1
i l i i. ( A t 4 l .. i r N t I l -: 1 r l' (
- t. .,
I. l' i i j .- l
- 3. -
4 Figure 3-6 Clime Routines Flow Control Outline J Overview of LVGOTHIC october 1997 o:\3006w.3.non:1b-101497 Revision 1
. . . -- - - - . ..-. - .: - . . . - . , - .. . . - , ~ , .. . . . . . . - - . - . . .- . - . _ . - - . . . . . ~ . . . ~ .
3-20
3.6 REFERENCES
3.1. Westinghouse Letter NTD-NRC-95-4563, B.A. McIntyre to Quay (NRC), " GOTHIC Version 4.0 Documentation," September 21,1995 3.2. Westinghouse Letter NTD-NRC-95-4462, N.J. Liparuto to T.R. Quay (NRC), EPRI Report RA-93-10, " GOTHIC Design Review, Final Report," May 15,1995 3.3. WCAP-14382, "EGOTHIC Code Description and Validation," M. Kennedy, et al., May 1995 3.4. Westinghouse Letter NTD-NRC-95-4595, B.A. McIntyre to Quay (NRC), "AP600 WGOTHIC Comparison to GOTHIC," November 13,1995 3.5. WCAP-14812, " Accident Specification and Phenomena Evaluation for AP600 Passive Containment Cooling System," Revision 1, June 1997. 3.6. WCAP 14326, Rev.1, " Experimental Basis for the AP600 Containment Vessel Heat and Mass Transfer Correlations," F. Delose, et al., May 1997 9 O Overview of LVGOTHIC october 1997 . c:\3006w-3.non:ll> 101497 Revtsion 1
_ _ . . _ _ - _ _ _ _ . . . _ . . . _ . _ _ . . _ . _ . . _ . . . _ _ . . . _ _ . . _ _ . _ _ - _ . . -- _ .._.. _ ___ __ =__ _ _ a 8 i-7 i i WESTINGHOUSE NON PROPRIETARY CLASS 3 1-4 i i. e i e I Section 4 1 .. i F Description of WGOTHIC Evaluation Model i I ! 1 t-t. P t P (, 0:\3006w-4.nostib101497
,9 O -
This entire section (4) is proprietary to Westinghouse Electric Corporation.(a,c) s O Description of )YGOTHIC Evaluation Model g g,3,7 c:\3006w-4.nortib-101797
WESTINGHOUSE NON-PROPR1ETARY CLASS 3 _ _ _ _ ( A 1 > i L Section 5 1 i ! Initial and Boundary Conditions t i t 1 i i i e i P 1 . l i f i I l-e iO i 0:\3X)6w-5 non:1b-101497
al-1 - TABLE OF CONTEN15
\
LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
5.1 INTRODUCTION
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1 i
5.2 INITIAL CONDITION SENSITIVITY CASES . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1 5.3 INITIAL CONTAINMENT HUMIDITY . . . . . . . . . , . ................... 5-3 5A INITIAL CONTAINMENT PRESSURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 5.5 INITIAL CONTAINMENT TEMPERATURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-9 5.6 AMBIENT HUMIDITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-12 5.7 AMBIENT TEMPERATURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-15 5.8 SENSITIVITY TO DROP MODELING ASSUMPTIONS . . . . . . . . . . . . . . . . . . . . 5-20 0
5.9 CONCLUSION
S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-20 O initial and Boundary Conditions October 1997 o:\3006w-5.rwn:1b 101497 Revision 1
iv LIST OF TABLES Table 5-1 Initial Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1 Table 5-2 Initial Conditions Sensitivity Analysis Cases . . . . . . . . . . . . . . . . . . . . . . 5-2 Table 5-3 Summary of Pressure Results for LOCA Initial Condition Sensitivity Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-3 Table 5-4 Senunary of Pressure Results for MSLB Initial Condition Sensitivity Stu dies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-3 O 1 O Initial and Boundary Conditicas october 1997 o:\3006w-5.non:1b-101497 Revision 1
v _ LIST OF FIGURES
~
Figure 5-1 Case 1 - Initial Containment Humidity Sensitivity - LOCA . . . . . . . . . . . 5-5 Figure 5 2 Case 8 - Initial Containment _ Humidity Sensitivity - MSLB . . . . . . . . . . . . 5-6 Figure 5-3 Case 2 - Initial Containment Pressure Sensitivity - LOCA . . . . . . . . . . . . 5-7 Figure 5-4 Case 9 - Initial Containment Pressure Sensitivity - MSLB . . . . . . . . . . . . . 5-8 Figure 5-5 Case 3 - Initial Containment Temperature Sensitivity - LOCA . . . . . . . . . 5-10 Figure 5-6_ Case 10 - Initial Containment Temperature Sensitivity - MSLB . . . . . . . . 5-11' Figure 5-7 Car.e 4 - Ambient Humidity Sensitivity at 115'F Ambient Temperature - LOCA ..'.....................................5 , Figure 5-8 Case 11 - Ambient Humidity Sensitivity at 115'F Ambient Temperature - MSLB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-14 Figure 5-9 Case 5 - Ambient Temperature Sensitivity - LOCA . . . . . . . . . . . . . . . . . 5-16 Figure 5-10 Case 12 - Ambient Temperature Sensitivity - MSLB . . . . . . . . . . . . . . . . . 5-17 O, Figure 5-11 Case 7 - Film Temperature Sensitivity - LOCA . . . . . . . . . . . . . . . . . . . . 5-18 V Figure 5-12 Case 14 - Film Temperature Sensitivity - MSLB . . . . . . . . . . . . . . . . . . . . 5-19
- Figure 5-13 Blowdown Drop Fraction Sensitivity - LOCA . . . . . . . . . . . . . . . . . . . . . 5-21 Initial and Boundary Conditions October 1997 c:\3006w-5.norulb-101497 Revision 1 1
5-1
5.1 INTRODUCTION
U The purpose of this section is to describe a series of sensitivity analyses performed for the design basis LOCA and MSLB cases in order to examine the effect of initial conditions on containment pressure response. Sensitivity evaluations are performed on initial containment humidity, pressure, and temperature, as well as ambient (outside containment) humidity and temperature. In addition, a sensitivity to drop modeling assumptions - a boundary condition for LOCA -is presented in this section. Initial conditions assumed in the WGOTHIC Evaluation Model are conservatively set to - maximize containment pressure response and are consistent with Technical Specifications and site interface parameter limits. Initial conditions assumed in the sensitivity evaluations are set at the opposing end of the Technical Specifications and site interface parameter limits for all sensitivity cases in this section except for the external temperature sensitivity, which was examined over a more limited range to be consistent with the film temperature range. 5.2 INITIAL CONDITION SENSITIVITY CASES The initial conditions considered in the sensitivity studies are summarized in Table 5-1. The reference values for the initial condition parameters were selected in the Evaluation Model to o maximize peak containment pressure. The reference value in Table 5-1 corresponds to the d Evaluation Model. Table 5-1 Initial Conditions Reference Sensitivity Initial Condition Sensitivity Case Value Value Containment Relative Humidity, % 0 100 Containment Pressure, psia 15.7 14.5 Containment Temperature, 'F 120 50 Ambient (Outside) Relative Humidity, % (Based on 22 0 80 F wet bulb temperature at 115"F) 100% Relative Humidity is the maximum value when 0/100 the ambient temperature is 40*F Ambient (Outside) Temperature, 'F 115 40 Water Film Temperature on Outside Shell Surface, 'F 120 40 l O LJ Initial and Boundary Conditions october 1997 o:\3006w-5.nortitr101497 Revishn 1
5-2 The sensitivity cases considered for the LOCA and MSLB transients are summarized in Table 5-2 with the initial condition parameters assumed in each case. Only values noted in Table 5-2 were varied in each of the cases. The reference cases for the LOCA and MSLB are described in Sections 4.5.2.1 and 4.5.2.2, respectively. The mass and energy releases are the same for all LOCA and MSLB cases. A summary of the pressure results are summarized in Table 5-3 for the LOCA and Table 5-4 for the MSLB. A discussion of each sensitivity case is provided in the following sections. Table 5-2 Initial Conditions Sensitivity Analysis Cases inside Contaisunent Outside Containment T air P RH T ht sink T-air T-fdm RH Case Transient ('F) (psla) (%) ('F) ('F) ('F) (%) > Reference 120 15.7 0 120 115 120 22 1 LOCA 120 15.7 100 120 115 120 22 2 LOCA 120 14.5 0 120 115 120 22 3 LOCA 50 15.7 0 50 115 120 22 4 LOCA 120 15.7 0 120 115 120 0 5 LOCA 120 15.7 0 120 40 40 100 6 LOCA 120 15.7 0 120 40 40 0 7 LOCA 120 15.7 0 120 115 40 22 8 MSLB 120 15.7 100 120 115 120 22 9 MSLB 120 14.5 0 120 115 120 22 10 MSLB 50 15.7 0 50 115 120 22 11 MSLB 120 15.7 0 120 115 120 0 12 MSLB 120 15.7 0 120 40 40 100 13 MSLB 120 15.7 0 120 40 4e 0 14 MSLB 120 15.7 0 120 115 40 22 O Initial and Boundary Conditions october 1997 o:\3006w.5.non:1b-101497 Revision 1
5-3 Table 5-3 Summary of Pressure Results for LOCA initial Condition Sensitivity Studies Peak Pressure Peak Post Blowdown During Blowdown Pressure Pressure at 24 Hours (psig) (psig) (psig) Eval. Model M.4 43.9 18.9 Case 1 34.0 42.5 16.6 Case 2 33.0 42.1 17.2 Case 3 35.2 42.5 19.4 Case 4 34.4 43.9 18.9 Case 5 34.4 43.7 1A6 Case 6 34.4 43.7 16.6 Case 7 M.4 43.7 16.6 d
- Table 5-4 Summary of Pressure Results for MSLB Initial Condition
/]
v Sensitivity Studies Peak Pressure (psig) Evaluation Model 44.8 a Case 8 43.6 Case 9 42.9 Case 10 44.6 Case 11 44.7 Case 12 44.7 Case 13 44.7 Case 14 44.7 5.3 INITIAL CONTAINMENT HUMIDITY The purpose of this sensitivity analysis is to illustrate the effect of initial containment humidity on containment pressure response, laitial humidity affects the initial mass of air in the containment and the concentration of air inside containment during the accident. In general, the Initial and Boundary Conditions Cktober 1997 o:\3006w-5.non:Ib-101497 Revision 1
5-4 presence of noncondensible gases reduces the effectiveness of internal heat sink structures to absorb energy, since the condensing vapor must diffuse through the gas before it can condense on the surface. The upper and lower bounds on relative humidity are 100 percent and 0 percent, respectively. The minimum, initial containment relative humidity (0 percent)is used for the Evaluation Model, since this value produces a higher peak containment pressure. The maximum relative humidity (100 percent) is assumed for the sensitivity case in order to quantify the effect of initial containment relative humidity on containment pressure response. rhe sensitivity of containment pressure to initial containment humidity is illustrated in Figure 5-1 for the LOCA (Case 1) and Figure 5-2 for the MSLB (Case 8). The sensitivity and reference cases are compared, corresponding to 100 percent and 0 percent relative humidity, respectively. A higher containment pressure response is predicted for zero percent relative humidity than for the sensitivity case at 100 percent relative humidity. The effect of relative humidity on contalruent pressure is explained by the influence of air on the rate of condensation on intemal heat sink structures, and the additional mass of air in containment. Lower relative humidity corresponds to lower vapor partial pressure and hence, to lower water vapor concentration. Since the total initial pressure is fixed, the partial pressure and, therefore, the concentration of air is greater at 0 percent than at 100 percent relative humidity. The higher mass of air also contributes to the pressurization as it heats up in thermal equilibrium with the steam. A greater quantity of air in the condensing vi.por also results in greater resistance to heat transfer, since the vapor must diffuse through the gas before it can condense on the surface. This factor reduces the overall heat removal capability of intemal heat sink structures, and results in greater containment pressures for the initial 0 percent relative humidity case. 5.4 INITIAL CONTAINMENT PRESSURE Initial containment pressure directly affects the containment pressure response. The range of initial containment pressures is bounded by the Technical Specifications limits. The initial internal containment pressure is set to the maximum Technical Specifications limit of 15.7 psia (1.0 psig) in the Evaluation Model. The lower bound (sensitivity case) initial containment pressure is set at the minimum Technical Specification limit of 14.5 psia (-0.2 psig). The sensitivity of containment pressure to the initial containment pressure is shown in Figure 5-3 for the LOCA (Case 2) and Figure 5-4 for the MSLB (Case 9). As expected, greater initial containment pressure results in greater containment pressure response throughout the transient. A higher initial pressure results in a greater mass (and hence concentration) cf air and results in higher contair ment pressures. O Initial and Boundary Conditions october 1997 j oA3006w-5.nowlb-101497 Kevision 1
55: A V initial Containmeni H um i d i t y -- S e n s i t i v_ i t y (LoCA) Evoluotion Model
---- Se n s i t i v i t y Cose 50 .? : ,,;-:%
g - g 's s D
~
e 30
.b 6
- s
[ s s
~
- 20 ^
e -
- s\./ - ?
u . ss
- o- : ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' '
10 ' ' ' ' ' ' ' ' 1 - 2 3 4 5 10 10 10 10 10 Time (seconds) ql Figure 5-1 Case i Initial Containment Humidity Sensitivity - LOCA
- O.
Initial and Boundary Conditions october 1997 o:\3006w-5.norrib-101497 Revision 1
5-6 O initial Containment Humidity Sensitivity (WSLB) Evoluollon Wodel
----Sensitivlty Case 50 m - ~
v a : ,' h
- s e 30 --
sA h 3 , M . 's ' m 20 '
$ : 's
- o. . ,
~ ' ' ' ' ' ' ' ' ' ' ' ' 7 1 '
10 ;- O 1000 2000 3000 4000 5000 Time (seconds) Figure 5 2 Case 8 - Initial Containment Humidity Sensitivity - htStil Initial and Doundary Conditions ortt+er 1997 c:\3006w $wc1;.101497 RevWon 1
57 Initial Containment Pressure Sensitivity (LOCA) Evoluotion Wodel
---- Se na l t i v i t y Case 50 c:n .
gQ 1" : g' 's
- f___ .
y O e 30 u 3 (s , s s
\ ~
20 -
~\/-d s '
y . \/
- a. : ,
in , , , , , , , , , , , , , , , , , , ,,,,, , , , ,,,,, 1 2 3 4 5 10 10 10- 10 10 Time (seconds) Figure 5 3 Case 2. Initial Containment Pressure Sensitivity . LOCA O initial and Boundary Conditions october 1997 c:\3006w-$.rmit>101497 - Reybion 1
5-8 . l O1 l l 1 Initial Containment Pressure Sensitivity (MSLB) ! Evoluotion Hodel l
---- Sensitivity Case 50 m - ~
m 40 . /
- o. ./ 's\s's v ./
e 30 u
'Q '
h
- _ 's '
m 20 ' u :
- a. .
10 - LL ' ' ' ' ' ' ' ' ' 4' 0 1000 2000 3000 4000 5000 Time (seconds) Figure 5-4 C 9 Initial Containment Pressure Sensitivity MSLB Initial and Boundary Conditions oct(eer 1997 o:\3006w 5.nortib-101497 Rntion 1
5-9 5.5 INmAL CONTAINMENT TEMPERA *IURE O - The purpose of this sensitivity analysis is to quantify the effect of initial containment temperature on containment pressure response. Containment air, internal heat sink, and containment shell initial temperature are simultaneously varied. A change in the initial air temperature affects the - concentration of air inside containment. A change in _ tim initial containment heat sink . temperature directly affects the heat absorption capacity of these structures. The initial containment temperature is set to the maximum Technical Specification limit of 120*F in the Evaluation Model. The lower bound (sensitivity case) initial containment temperature is set to a value of 50*F. The sensitivity of containment pressure to initial containment temperature is shown in Figure 5 5 - for the LOCA (Case 3) and Figure 5-6 for the MSLB (Case 10). As indicated, a higher peak containment pressure is predicted for the Evaluation Model case at 120'F, than for the sensitivity case at 50'F initial temperature. As illustrated M Figure 5 5, the pressure is higher for the 50*F initial temperature case during the blowdown phase of the transient, lower at the time of
- maximum pressure, and higher beyond approximately 5000 seconds. For the MSLB case shown in Figure 5-6, the peak pressure is slightly lower for the 50'F initial temperature case, but is higher during the initial pressure rise and beyond approximately 2000 seconds.
This pressure response behavior is predominately due to two competing influences: (1) the elfeet of initial temperature on the amount of air in the containment, and (2) the effect of initial temperature on the heat absorption capacity of internal heat sink structures. A lower initial temperature results in a higher air mass which contributes to the pressurization as- the containment heats up. The increased concentration inalbits condensation of vapor on internal heat sinks and results in higher containment pressures. In contrast, a lower initial temperature results in increased heat absorption capacity of internal heat sinks that tend to lower containment pressures. Initially the noncondensible gas concentration factor dominates, and the sensitivity case exhibits a slightly higher containment pressure. When the heat absorption capacity of internal heat sinks becomes the more dominant factor, a higher containment pressure results for the Evaluation Model case. As the internal heat sinks ssturata, the air concentration factor again becomes the governing influence, and the pressure for the sensitivity case exceeds that for the
- Evaluation Model. The Evaluation Model uses the maximum temperature assumption in order to maximize the more limiting post-blowdown peak containment pressure.
l Initial and Boundary Conditions - october 1997
. oA3006w-5.non:1b 101497 - Reviskm I s
i_... . _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ . _ _ '
5 10 O Initial Containment Temperature Sensitivity (LOCA) l Evaluotion Wodel !
---- Se n s i t I v l t y Case j 50 en -
u> 4 0 s ) a : ' v -
- r_ _
3 =
7
- s
~ ' '^
20 c) u V N o- : ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' 10 1 2 3 4 5 10 10 10 10 10 Time (seconds) Figure 5 5 Case 3 -Initial Containment Temperature Sensitivity LOCA Initial and Boundary Conditions Octoter 1997 o:\3006w-5.non:1b-101497 Revision 1
i 5 11 L 4 i l 4 1 l Initial Contolnment Temperature Senaltivity (WSLB) Evoluotion Wodel
---- Se n s i t i v i t y Cose ,
50 - m -
" 40
- %s v
! a> 3 0 I
* 20 '
O
~
N 10
- 0 1000 2000 3000 4000 5000 Time (seconds) i i
Figure 5-6 Case 10 Initial Containment Temperature Sensitivity - MSLB
-b initial and Boundary Conditions octotwr 1997 o:\3006w 5,rmlb 101497 IL% ion 1
5 12 5.6 AMBIENT llUMIDITY licat is removed from the containment atmosphei by condensation and convection heat transfer to the shell, where it is conducted through the snell and rejected to the atmosphere on the outside of containment, licat rejection to the atmosphere is achieved by convection to the buoyant cooling air, radiation to the baffle, and evaporation of the external PCS film to the cooling air. Evaporation of PCS water is the most significant of these heat removal mechanisms. Evaporation mass transfer is driven by the concentratim gradient, or equivalently, the vapor partial pressure difference between the film and riser air. Changes in ambient or outside atmospheric conditions (e.g., relative humidity) can influence, to some degree, the vapor partial pressure difference. The purpose of this sensitivity analysis is to evaluate the effect of ambient humidity on containment pressure response. The upper limit of ambient humidity is defined by the site interface parameters to be a maximum wet bulb temperature of 80*F. This corresponds to a relative humidity 0122 percent when the ambient temperature is 115'F. These boundary conditions are assumed in the Evaluation hiodel. Two sets of sensitivities to relative humidity are presented. The first provides a comparison of the Evaluation Model to the case with 0 percent relative humidity at an ambient temperature of 115'F. The second sensitivity compares relative humidity of 0 percent and 100 percent at an arnblent temperature of 40'F. The sensitivity of containment pressure to ambient humidity is depicted in Figure 5-7 for the LOCA (Case 4) and Figure 5-8 for the MSLB (Case 11). The sensitivity and reference cases are compared corresponding to 0 and 22 percent relative humidity, respectively. These figures illustrate that containment pressure is not sensitive to initial inlet humidity. This result is consistent with the small effect of inlet humidity on the main factors goveming the process of evaporation between the wetted shell and the riser air flow. The ree of evaporation is principally driven by the concentration gradient or, equivalently, the difference in vapor partial pressure between the film interface and the bulk air mixture. The partial pressure of vapor at the film interface is equal to the saturation pressure at the film temperature. Neause the concentration of water vapor in the bulk air mixture is small in comparison, the partial pressure gradient is essentially given by the saturation pressure at the film interface. Consequently, initial inlet humidity las no significant effect on the rate of film evaporation or on containment pressure. The sensitivity performed at 40*F, comparing 0 and 100 percent relative humidity, exhibited the same behavior, indicating almost no sensitivity to ambient humidity. A comparison of Case 5 to Case 6 for LOCA, and Case 12 to Case 13 for MSLB indicates a nearly identical pressure response. A comparison plot for these cases is not provided. The differences in these cases compared to the Evaluation Model are due to ambient temperature differences which are discussed in the Section 5.7. h Initial and Boundary Conditions octotier 1997 oA300tiw-5.non:lt> 101497 Rmsion 1
5-13 /~'T (v ) Amblent Humidity Sensitivity at 115 F Amblent Temperoture (LOCA) Evoluotion Wodel
---- Se n s i t i v i t y Case 50 ^ .
C7' - m 40
- o. :
~ /-~'s e 30 O " / => ~
- 20 ~
x m u N/ O. ~ 10 1 2 3 4 5 10 10 10 10 10 Time (seconds) Figure 5-7 Case 4 - Ambient flumidity Sensitivity at 115'F Ambient Temperature . LOCA (] L/ Initial and Boundary Conditions october 1997 o:\3006w-5.non:lb-101497 Revision 1
5 14 O Ambient Humidity Sensitivity at 115 F Ambient Temperature (WSLB) Evoluotion Model Sen Itivity Case 50 m -
~ =
M 40 A s O.
\
v _ c) 30
" h N
3
2 0 N 0- ~ ~N 10 O 1000 2000 3000 4000 5000 Time (seconds)
Figure 5-8 Case 11 - Ambient Humidity Sensitivity at 115'F Ambient Temperature - MSLB Initial and Boundary Conditions october 1997 o:\3006w-5.non:1t*101497 Revision 1
5-15 5.7 AMBIENT TEMPERATURE b The purpose of this sensitivity analysis is to illustrate the effect of ambient temperature on containment pressure response. Cooling air and PCS water temperature are simultaneously and independently varied in order to investigate the effects. A change in the amb!cnt air temperature primarily affects heat rejection by convection to the riser air flow. A change in the PCS water temperature affects the amount of energy absorbed by sensible heating.
'Ihe site interface parameter limits on ambient air temperature are 115'F and -40'F. The minimum PCS water temperature is limited by the Technical Specifications to a value of 40'F.
Since a higher ambient temperature and PCS water temperature produces a slightly greater containment pressure, the maximum ambient temperature (115'F) and PCS water temperature (120'F) are assumed for the Evaluation hiodel. The temperature for both inlet air and PCS water (sensitivity case) is set equal to 40'F. The sensith*ity of containment pressure to ambient temperature is shown in Figure 5-9 for the LOCA (Case 5) and Figure 5-10 for the hiSLB (Case 12). As indicated, lower containment pressures are predicted for the densitivity case at lower ambient temperatures late in the transient for the LOCA case. There is little impact on the peak pressure or pressure early in time. The containment pressure for an MSLB is less sensitive to external conditions and therefore, there c is a smaller impact on pressure for the entire transient. \ The reduction in the long term pressure is primarily attributed to liquid subcooling with a small contribution due to forced convection heat transfer effects. The external liquid film absorbs sensible heat from the point of PCS flow application to the point where significant film evaporation occurs. The subcooled heat capacity is dependent on water source temperature and external water flow rate. A lower source temperature results in greater subcooled heat capacity of the external film and, hence, more energy removed from containment. Forced convection heat transfer exists in the riser post-wetting as a result of the high buoyancy-driven air flow rate. The rate of energy transfer by forced convection is dependent on the heat transfer coefficient and the temperature difference between the liquid film and bulk air. Of these parameters, the temperature difference is influenced to a greater extent by bulk air temperature. A lower bulk air temperature results in greater forced convection heat transfer and, therefore, more energy removal from containment. The combined energy absorbed by liquid subcooling and forced convection represents a small fraction of the total energy removed from containment. Consequently, lowering the ambient air and source water temperatures to 40'F results in more total energy removed from containment, and, therefore, results in a decrease in containment pressure relative to the Evaluation Model. Case 7 (LOCA) and Case 14 (MSLB) considered only the change in PCS water temperature, shown in Figures 5-11 and 5-12, respectively. These cases confirm that the air temperature v impact is less important than the >CS r water temperature. Initial and Boundary Conditions october 1997 o:\3006w-5.non:1b 101497 Reviskm 1
5-16 O Amblent Temperature Sensitivity (LOCA) Evoluotion Wodel
----Sensillvity Case 50 ~
m [ 5 : _/
* : 7_
U
/ O N ~ $ 20 ^
a> w
~
g-Q
~
C1. ~ 10 1 2 3 4 5 10 10 10 10 10 Time (seconds) i l l Figure 5-9 Case 5 - Ambient Tempetiture Sensitivity - LOCA { initial and Boundary Conditions October 1997 o:\3006w-5.non:1tr101497 Rn'ision 1 l
5 17 O Ambient Temperature Sensitivity (MSLB) Evoluollon Wodel
---- Sens i t i v i t y Case 50 m .
5 4n b v a : N a) 30 6. 3 20
\ '"
CL. - so 0 1000 2000 3000 4000 5000 Time (seconds) Figure 510 Case 12 Ambient Temperature Sensitivity MSLB Initial and Boundary Conditions . October 1997
- o:\3006w 5.nortib-101497 -. %m1
_ _= -. _ t.
5 18 O Film Temperature Sensitivity (LOCA) Evoluotion Wodel
---- Se n a l t I v l t y Case 50 = ^ :
en
._ : yT 5" :
- f_
/
- = 3o
./ '
9
- s m
m 20 w _
,_Q 1 .
10 1 2 3 4 6 10 10 10 10 10 Time (seconds) Figure 5-11 Case 7 Film Temperature Sensitivity LOCA initial and Boundary Conditions Cktober 1997 . o \Xmw.5.nortib101497 Revision 1
)
5 19 O Film Temperature Sensitivity (MSLB) Evoluotion Wodel
---- Se ns i t i v i t y Case 50 m - ~ " 40
- 'N v .
e 30 N
~
- 20 "
o-N N' 10 O 1000 2000 3000 4000 5000 Time (seconds) Figure 512 Case 14 Film Temperature Sensit!vity . MSLB O Initial and Boundary Conditions octoter 1997 c:\3006w-5.rostib 101497 Rn h 1
5 20 5.8 SENSmVITY TO DROP htODELING ASSUhtPTIONS During a LOCA blowdown, the liquid and entrained droplets enter the atmosphere saturated at O j the containment total pressure where they are exposed to the containment gas mixture of air and steam at the steam partial pressure. Since the liquid and drops are initially superheated, they evaporate quickly to reach thermal equilibrium with the gas mixture. A sensitivity study was l performed for the LOCA to determine the impact of the modeling assumption in WGOTIIIC of ' the fraction of liquid converted to drops on the containment pressure. The mass released during the MSLB does not contain droplets. The fraction of liquid assumed to be tumed into droplets during the LOCA blowdown was varied from 0 to 100 percent. These sensitivities showed that the impact of assuming no droplets released, had a significant impact on the calculated pressure response cornpared to the cases where droplets were modeled. With no droplets assumed, the blowdown pressure was higher, but the peak pressure was lower. liowever, the sensitivity to the asst.med fraction of droplets was very weak above a level of approximately 5 percent. The drops are strongly coupled to the containment atmosphere temperature due to the large surface area of the drops. The presence of drops in the atmosphere at approximately the 5 percent level maintains the atmosphere in a saturated condition and the presence of additional drops has little impact on containment pressure.
'this sensitivity indicates that it is important to model the presence of drops in the containment atmosphere but the specific fraction assumed has a minor impact on the resulting pressi re. The containment pressure response for assumed droplet fractions of 0 and 100 percent along with the Evaluation hiodel assumptions for drops (discussed in Section 4.5.2.1) is illustrated in Figure 513.
5.9 CONCLUSION
S A series of sensitivity analyses has been carried out for the reference design basis LOCA and MSLB transients to determine the effect of initial conditions on containment pressure response. Sensitivity evaluations were performed on initial containment humidity, pressure, and temperature, as well as ambient humidity and temperature and PCS water (film) temperature. These sensitivities demonstrate that the initial conditions assumptions in the Evaluation hiodel result in a conservative prediction of containment pressure. The containment pressure is more sensitive to internal conditions than to ambient conditions. The sensitivity to intemal conditions is due primarily to the effect of these conditions on the amount of air in the containmerit. A sensitivity was performed for the LOCA to determine the impact of the drop modeling assumption in ECOTHIC on the calculated containment pressuie. The results show that it is I important that the droplet formation be modeled, but at fractions above approximately 5 percent, the fraction assumed to be released as drops has a small impact on the calculated pressure. l l 6 l l Initial and Boundary Conditions october 1997 oA3006w 5.nonttwl01497 Revisbn 1 l
5 21 O Blowdown Drop Fraction Sensitivi.y (LOCA) Evoluotion Wodel
----NoDrops ----Ai1 Drop 50 m -
CL
- sm v
- /s _ %~, 4-
['/ a) 30 3
/
m 20 y s Q- - jo i ... . . . . . . . . . . . . . . .. 1 2 3 4 6 10 10 10 10 10 Time (seconds) y e Figure 513 Blowdown Drop Fraction Sensliivity . LOCA laitial and Boundary Conditions October 1997 a.\3006w-5.non:ll>101497 Revision 1
WESTINGHOUSE NON PROPRIETARY CLASS 3 O Section 6 Meteorological Effects on PCS Performance O O c:\3006w-6.nortib 101497
ill TABLE OF CONENTS r_
\ ]
LIST OF FIG URES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ill ) l
6.1 INTRODUCTION
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 l
6.2 WIND-INDUCED TUF BULENCE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 6.2.1 Summary of Wind Tunnel Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-1 6.2.2 Tracking of a Wind Driven Particle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-3 6.2.3 Containment Time Constants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 6.2,4 Wind Induced Or.cillation Effect on Heat Transfer Coefficient . . . . . . . . . 6-5 6.2.5 WGOTHIC Evaluation Model Basis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 5 6.3 RECIRCULATION OF CHIMNEY EFFLUENT . . . . . . . . . . . . . . . . . . . . . . . . . . 67 6.3.1 Summary of Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-7 6.3.2 Evaluation of Effect of Rec'rculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-8 6.3.3 WGOTHIC Evaluation Model Basis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-9
. 6.4 CONC LUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 10 6.5 RE FE RENC ES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 10
- LIST OF FIGURES 4
Figure 6-1 Particle Path Through AP600 PCS With and Without Wind . . . . . . . . . . . 6-4 Figure 6-2 1D Containment Shell Model Inside Temperature Results . . . . . . . . . . . . 6-6 i. O
.Q Meteorological Effects on PCS Performance Detcher1997 a:\3006w4norulb 101497 Revukm 1
61 E1 INTRODUC110N O Meteorological conditions which could be postulated to degrade the performance of the AP600 PCS design have been investigated. The design includes, within the chimney, a shield plate which protects the containment surface from direct impingement of rain. Screens on the PCS inlets and around the entrance to the chimney protect the PCS from birds or larger debris which may be blown by wind. Meteorological effects that are evaluated, are wind induced turbulence and the potential for recirculation due to wind or temperature inversions. This chapter shows that the assumption of a quiescent atmosphere in the evaluation model conservatively neglects enhancements to heat and mass transfer due to wind, it is also shown that the potential effects , of recirculation produce a negligible effect on containment pressure. 6.2 WIND INDUCED 'I1JRBULENCE 6.2.1 Summary of Wind Tunnel Tests A goal of the containment building design is that wind not advert.ely impact heat removal from the building. The PCS is designed for wind to either have a nominal effect on PCS flow (wind neutral) or enhance PCS flow (wind positive). To verify the wind positive performance, a series of wind tunnel tests were performed. The wind tunnel tests, perfarmed at the Boundary Layer O Wind Tunnel Laboratory at the University of Western Ontario (UWO), were designed to test the aerodynamic response of air flow past the AP600 containment under a variety of conditions. The tests occurred in four phases. Phase 1 testing (~1:100 3cale) examined the effects of verious design c; dons on the wind induced pressures. In Phase 1 testing, although the flow through the building annulus was not modeled, the pressure difference between inlets and chimney, Ap, was measured. The inlet-minus-chimney Ap is the pressurt; driving flow through the PCS, and a pressure coefficient, ep , is defined based on free stream wind velocity and Ap: AP = 1/2 ep pamb Ybt where Pamb = ambient air density V,oor = free stream wind velocity In Phase 2 tests, the air flow path was modeled for two different building designs: the most wind neutral design found in Phase 1 testing and the current design of the building. The
. purpose of the Phase 2 testing was to provide information for the design of the baffle wall.
Meteorological Effects on PCS Performance october 1997 oM006w4.non:1b-101497 ReiWon 1
6-2 Iluoyancy was not considered in the wind tunnel tests, since the driving pressure due to buoyancy amounts to only about 1 to 5 percent of the wind induced driving pressure for the design wind cases. h At the end of Phase 2 of the wind tunnel program, several ques 6cas remained. In Phase 3, analysis was used to address the potential effects of wind and thermalinversion on recirculation of the chimney efauent back into the inlet, using evallable literature from mechanical and natural draft cooling towers. l Three additional questions were addressed with testing in Phase 4. The first question regards l the effect of Reynold's number on the resu'ts. Reynold's number effects could only be addressed l definitively by testing a larger model (1:30 scale) in a higher wind speed tunnel, such that the Reynold's numbers were in the same range as expected full-scale values. The becond question I was the effect of a tornado wind profile (near uniform) on the results. Tornado profile effects could be obtained using the same test model as in previous phases, but with a uniform flow model. The third question addressed the blockage effects of a hyperbolic cooling tower relative to the UWO wind tunnel size. Cooling tower blockage could be addressed by testing the model in a larger wind tunnel where blockage would be small. The final question, the effect of severe terrain, was the subject of Phase 4 testing, in which a smaller scale (1:800 scale) was chosen to allow modeling of larger areas around the site. Test results indicated that the AP600 design was wind positive for average PCS flow The O testing included a variety of terrain and conditions, including open country terrain, tornado loading, modeling of the cooling tower (s), and simulation of several types of severe terrain. Open country terrdn yielded the most beneficial results for PCS heat removal, indicating a significant contribution to PCS air flow due to wind induced driving pressures. The effect of the cooling tower, however, was to reduce static pressure at both the chimney and the inlets, resulting in lower mean wind Ap. Thus, the likelihood of flow in the PCS changing directior (flow reversal) was greater when the plant was in the wake of the cooling tower, giving the least positive mean PCS driving force due to wind. The three Phase 4 severe terrain scenarios included an escarpment with mountain backdrop, e river valley site, and a river valley site with two cooling towers. Each terrain scenario caused durations and magnitudes of negative wind Ap, which could lead to flow reversals within the PCS flow path. The winci positive response of the AP600 has been shown (Ref. 6.1) to be beneficial for containment heat removal for the limiting terrain configuration. IncreasM wind speed drives more flow through the AP600 annulus and increases heat and mass transfer coefficients. Three questions have been addressed regarding the results of the wind tunnel tests: Meteorological Effects on PCS Performance october 1997 oA3006w4.non:ltul01497 Revision 1
63 c The model scale aerodynamic response versus full-scale response
- The effects of wind induced flow oscillations on PCS heat removal and containment pressure response
- The effect of near zero average wind Ap for certain wind angles in some of the severe terrain tests Due to the shape of the AP600 (sharp edges on the shield building which initiate flow separation), the model to full-scale aerodynamic response is relatively insensitive to model size in the ratige tested. A review of the literature has indicated that pressure oscillations in heat transfer generally improve heat transfer rates. In addition, time constants associated with the containment shell and internal volume minimize any benefit or penalty on containment pressure due to oscillations. The effect of wind-induced pressure oscillations has been evaluated with sirnple calculations.
6.2.2 Tracking of a Wind Driven Particle Using the measured pressure coefhcients, density of air, and design wind speed of 214 mph, wind Ap was calculated and conv< rted into annulus veLeities using the momentum equation, n which balances the driving fore : with the unrecoverable losses. Fi5ure 6-1 presents the V calculated path of the first element to travel from the inlet to the outlet of the PCS. Figure 61 also presents the path of the element neglecting the wind, and using an assumed buoyancy-driven annulus velocity of 15 ft/sec. Note that the wind-driven element shows a ne* positive flow response to pressun oscillations (net flow is from the inlet to the chimney). 6.2.3 Containment Time Constants A review of the literature ha3.>dicated that oscillating flows gem. rally increase heat transfer. The effect t ' the wind Ap oscillations on the AP600 post LOCA pressure response is limited by time constants associated with the con.ainment shell and the containment volume. The shell time constant gives the response of the contairunent shell to changts in its environment. Using a lenped mass approach, the time constant compares the thermal capacitance of the shell to the heat removal rate from its surface and has a value of about 255 seconds. The shell time constant is significantly higher than the frequency of pressure fluctuations, which are on the order of several seconds for high wind speed cases. The time constants show that the thermal response of the containment shell is sufficiently slow so that high speed oscillatione will not Mgnificantly affect PCS '4 eat removal. At lower wind speeds, oscillatioru are much slower. However, at lower wind speeds, the wind Ap is much lower. As wind speed reduces, the wind Ap decreases
- i. , idly, as a function of the square of the wind velocity. Thus, oscillations will not have a (7 sigilficant impact on PCS heat removal. Since PCS heat removalis relativcly unaffected, G'
Meteorological Effects on PCS Performance ociceer 1997 c:\30t%w4non:1b 101497 RevWon 1
i 6-4 O! 4 350
- Ch""*'F 300 l ... ****'
230 _
/ ....... ~
n - o C ' o Rner "o200 -
/
y .-
,, 150 - ,/
O *,l 100
,a', ,' Dowwoner 50 - o**,
0 0 0 10 20 30 40 50 Time (sec) 214 mph Wind Speed No Wind , , l t Figure 6-1 Particle Path Through AP600 PCS With and Without Wind O Meteorological Effects on PCS Performance octoter 1997 l o:\30tew4.non:ll> 101497 Revision 1
65 containment pressure response to a postulated LOCA will not be significantly affected by ( pressure osdllations. Thus, heat transfer fluctuations occur relatively faster than the ability of the wall material to transmit oscillations through the shell. 6.2.4 Wind Induced Oscillation Effect on Heat Transfer Coefficient Pressure fluctuations affect the heat transfer coefficient on the containment surface. In particular, oscillations result in short periods where the heat transfer coefficient may be lower than the value assumed in the no-wind case, followed by periods of higher heat transfer coefficients. The heat transfer response to wind oscillations has been investigated using a 1 D plane wall conduction model. The conduction model was used to estimate the effeet of pressure oscillations on heat transfer through the containment shell. The model sirr,. dates the containment shell and a liquid water film on the outside of the shell. The 1 D conduction model was subjected to the heat and mass transfer coefficient on the outside of the plane wall calculated from the time-varying annulus velocity. Only forced convection correlations were used, so that heat and mass transfer rates on the outside of the plane wall approached zero as annulus velocities approached zero. The use of a forced convection correlation is conservative since, even as velocities in the annulus pass through zero, heat transfer would still occur. To further impose a conservative bias in the calculation, heat and mass transfer rates on the outside of the wall were assumed to be zero whenever the annulus velocity was negative. The response of the containment shell to the imposed velocity was calculated. Figure 6-2 presents the surface temperature of the inside of the plane wall versus time. The figure compares the response of the wall to the annulus velocity oscillations versus the response assuming a steady buoyancy-driven annulus velocity, Note that, despite neglecting heat removal from the wall during periods of ne6ative annulus velocity, the temperature of the inside of the plane wall is still about the same as a typical steady velocity case, showing that the response of the containment shell is limited by the time constants discussed in previous sections. 6.2.5 HGOTHIC Evaluation Model Basis The wind tunnel testing of the AP600 indicates that the average wind ap tends to be positive under a variety of conditions. Wind flowing towards and over the containment building will ten.1 to increase average flow rates through the AP600 PCS. The wind induced flow rate inci,eare will improve heat transfer rates in the AP600 PCS. In addition to the open country terrain, several highly turbetent severe terrain scenarios were tested to obtain data on the AP600 subjected to limiting site conditions. For the severe terrain, positive wind Ap_ that averages near zero may be seen. In addition, the wind Ap tends to oscillate, giving periods of negative. wind op. Negative pressures indicate the possibility of flow p reversals within the PCS annulus. Assessment of the current literature has indicated that flow V Meteorological Effects on I'CS l'erformance october 1997 c:VX106w4rm1b 101497 Reviskm 1
1 I i 1 6-6
)
l 1 O 300
~ ~ ~ ~ ~ ~ ~ ~ ~ - ~ " " " " " " " " " " " " " " " " " " " - ' " - - ~~~~~ ~~"--- -"""
280 b -- e 260 rs,o a 0 240 - " - ~"~~~~ ~ ~ ~-~~ ~ ~~ - - - - " ~~ ~~ ~ -
~~~ --"" ",~~~ ~ """"~" " " ~
1 _
' ' '- - ~ - ~ ~ ~ ' ' '
E u t- 220 - - ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ " " - ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ " ~ " " ~ ~ ~ - h200 -- - o "O m I80 - - -~~~ ~~~~~~~~~~ ~~~-- - ~~ - -- ~~~~-- ~~~~ -" ~~~~~~~~
.5 160 - -
1 i 1 0 1,000 2.000 3.000 4.000 Time (sec) No Wind 214 mph Wind Speed Figure 6 2 1D Containment Shell Model inside Temperature Results Meteorological Effects on PCS Performance octotier 1997 tr\3006w-6.non:1b-101497 Revision 1
l 67 4 oscillations will tend to increase heat transfer primarily by enhancing mixing across the riser annulus flow channel. While periods of negative pressure may result in short periods of flow reversal within the annulus, the literature indicates that turbulent conditions may continue to 1 exist. Turbulent conditions would continue to provide significant heat transfer rates despite the oscillating flow. Time constants were calculated for the containment shell which indicated ' hat the shell time const. ats were of significantly higher magnitude than the period of the T cessure
] ' l oscillations. T hus the pressure oscillations in the annulus would be damped in their c.iects on l the conta' ament heat removal rates at the inside of the containment shell. A 1 D conduction
! model of the containment shell, subjected to oscillating heat transfer rates, was solved using the l wind op from a particularly turbulent angle of the limiting test site. The 1 D conduction i calculation used the forced convection correlation to conservatively determine heat transfer rates i in the annulus. Heat transfer was also assumed to be zero when the flow reversed. The results of the calculation indicate a slight hnefit in PCS heat removal and containment pressure due . to wind for the limiting case. The effect of the containment shell was to dampen the oscillations occurring on one side of the shell. Thus, a conservative calculation of the AP600 containment ! response to a LOCA could assume a quiescent atmosphere. ! 6.3 RECIRCULATION OF CHIMNEY EFFLUENT After the PCS cooling air flow passes over the cordainment shell surface, the air and evaporated ! water exhaust through an opening in the roof of the shield building and through the chimney. js The potential for recirculation of the chimney effluent back to the PCS inlets, due to temperature inversions or strong winds has been evaluated (Ref. 6.2) through a review o. literature and
- shows the negligible effect of a conservatively high assumcd recirculation.
i 6.3.1 Summary of Literature Review
- Many references were found in the literature to address potential recirculation due to strong winds or thermal inversions. References are available for natural draft hyperbolic cooling towers, typicaPy hundreds of feet tall, and for mechanical draft cooling towers, typically 10 to 20 feet tall.
2 Strong winds can cause the formation of a recirculation cavity on the leeward side of a building or cooling tower. It was found that there are some intermediate wind speeds which can be sufficient to bend the plume horizontally, yet not strong enough to carry all the effluent away. 4 Analytical and experimental research in the literature was conducted to determine the extent of the recirculation cavity behind a natural draft cooling tower and its effect on the plume. Curves are provided in the literature based on a normalized temperature difference that indicates the
- increase in the mixed mean ambient inlet temperature due to mixing with the plume. Such
, curves suggest a maximum normalized temperature increase of 10 percent for recirculation.
! p
- (J Meteorological Effects on PCS Perforrnance october 1997 oA3006w4non.lb 101497 Revision 1 ]
l _ . _ _ , - _ - _ . - . - - - , ~ - . --
68 Similar studies for mechanical draft towers suggest recirculation of 3 to 7 percent reaching a maximum of 15 percent. Thermst inversions, and combinations of wind and temperature inversions were cited. Results showed that an inversion, by itsuf, does not induce the downflow necessary to recirculate chimney effluent. Adverse inversion conditions are associated with calm or light winds. Using simplified plume rise equations, the approximate effluent conditions resulted in plume rise above the shield building chimney for atable atmospheric conditions (inversions). The plume rise was sufficient to raise the plume, in light wind, abeve the recirculation zone of structures the size of l those associated with the AP600 design. Consequently, the maximum expected recirculation I would be determined from the strong wind case. Based on the literature review and evaluations of the AP600, the upper limit for recirculation of l the AP600 cidmney effluent is [ ]*# To account for the uncertainty in choosing a value i for recirculation, the more conservative value of [ )*# has been assessed, which would result in the mixed mean ambient inlet temperature increasing from the safety analysis basis of 115'F to [ ]*#. 6.3.2 Evaluation of Effect of Recirculation The effect of a recirculation ratio of [ ]*# has be:n assessed with WGOTHIC sensitivity calculations. The base case calculation used an inlet temperature of 115'F and inlet humidity of 20 percent. Two sensitivities were run: one with only the inlet temperature increased, based on the recirculation ratio, and one with both the mlet tcmperature and inlet humidity increased. Results show that the pressure transient is insensitive to temperature and humidity in this range due to the self regulating performance of the PCS. The base case used for recirculation sensitivity differs from the evaluation model in the details of intemal noding, azimuthal segregation of the annulus into quadrants, modified mass and energy releases, the use of 22 percent relative humidity, an initial PCS flow profile starting at 220 gpm, and in the use of nominal heat and mass transfer correlations. Since these sensitivity results are used to examine relative effects of changes in the annulus inlet conditions, the sensitivity results are judged to provide a reasonable estimate of tne potential effect of recirculation. The base case chimney outlet temperature reaches a maximum of [ )*#at about 2100 seconds and decreases almost linearly to [ )*# at about 8700 seconds, after which it gradually reduces to [ )** at 24 hours. For simplicity, a conservative assessment of the potential effect can be based on an assumed chimney outlet temperature of [ )"#, which indudes [ )*# to account for the increase in outlet temperature when the inlet temperature is O Meteorological Effects on PCS Perfonnance ociate 1997 c:\XKbwAnon1t 101497 Revision 1
6-9 increased in the sensitivity run. Using the definition of the recirculation ratio, the mixed mean inlet temperature, accounting for the effect of effluent recirculation, is Tin = T. + h (T w , T.) Tin = 115 + [ ]*# So the inlet temperature to be assumed in the sensitivity cases is [ ]*# which is applied for all annulus quadrants, consistent with the definition of R from the literature. The first sensitivity case used a constant [ ] temperature and essentially unchanged inlet humidity [ ]*#. Results from the sensitivity show that the pressure transient changed by a negligible (<0.1 percent) amount due to the [ ]*# increase in inlet temperature, and confirmed the initial guess for the corresponding increase in outlet temperature. The second sensitivity included the increase in inlet temperature combined with the inlet humidity set to 98 percent. Again, there is a negligible effect on the containment pressure. The lack of sensitivity of the pressure response is due to the self regulating performance of the PCS. ITy comparing the annulus conditions in going from 20 to 98 percent inlet humidity, it is seen that the annulus mass flow rate increased by about [ ]*#.The higher mass flow increases the capacity to move vapor out of the annulus and is due to the increase in vapor pressure at the annulus outlet from [ ]*# of the approximately 14.*/ psia total pressure. Since steam density is more sensitive to temperature O increases than air density is, and steam density is less than air density at annulus conditions, the increased steam content provides a greater density driving head for flow through the annulus.
'Ihe increased mass flow results in a greater velocity through the annulus, which increases the PCS mass transfer coefficient. Thus, the mass flow increase offsets increases in intet humidity.
Similar, self regulating performance results from an increase in inlet temperature alone. It may be expected that an increase in hilet humidity would suppress the evaporation rate from the film. Such an effect is actually small since the driving force for evaporation is the difference between the vapor pressure of the film and the bulk saturation pressure in the annulus. Since the vapor pressure of the film is on the order of ten times that of the annulus, a relatively large percentage change to annulus humidity corresponds to a relatively small percent of the driving force. Sensitivities to the effects of increasing both inlet temperature and humidity to account for potential recirculation show that there is a negligible effect on the containment pressure transient. 6.3.3 HGOTHIC Evaluation Model Basis Since the effect of effluent recirculation is negligible, the EGOTHIC evaluation model does not O consider any additional penalty due to recirculation. Meteorological Effects on PCS Performance october 1997 oA3006w-6.norrib 101497 Revision 1
6-10
6.4 CONCLUSION
S Wind induced pressure oscillations have been shown to provide a benefit to PCS heat removal because of the wind positive design; that is, wind induces more heat removal than a quiescent atmosphere. The effects of recirculation due to thermal inversions or strong winds has been shown to have a negligible impact on PCS heat removal. The lyGOTHIC evaluation e.odel bounds the postulated effects with no input modifications.
6.5 REFERENCES
i 6.1 NTD-NRC f 3-4467, J. Narula, " Analysis of AP600 Wind Tunnel Testing for PCS Heat Removal," June 2,1995. ! l 6.2 NTD-NRC-94-4166, R. Haessler, "AP600 Passive Containment Cooling System Letter Reports," June 10,1994. O O Meteorological Effects on PCS Performance october 1997 c:\3006w 6.non:lb-101497 Reybion 1
WESTINGHOUSE NON-PROPRIETARY .MS 3 r Section 7 Basis and Method for Calculating the PCS Water Evaporation Rate for the AP600 Containment DBA Evaluation Model O 2 1 o:\MXMw-7.non:lb101497
iii . TABLE OF CON *EN'IS LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v LIST OF FIGURES . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi EXECUTIVE
SUMMARY
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vil
7.1 INTRODUCTION
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-1 7.2 WATER APPLICATION AND DISTRIBUTION . . . . . . . . . . . . . . . . . . . . 7-3 7.2.1 Containment Shell Surface Coating - 7-3 7.2.2 PCS Water Distribution Weir Description and Operation . . . . . . . 7-5 7.2.3 -- PCS Water Distribution Testing Results . . . . . . . . . . . . . . . . . . . . 7-8 7.2.4 Delivered Water Flow Rate versus Time . . . . . . . . . . . . . . . . . . . . 7-10 7.3 AP600 WATER COVERT.GE BASIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-13 j 7.3.1 Water Distribution Film Flow Rate, rdist 7-14 7.3.2 Minimum Film Flow Rate, Emin _ . . . . . . ; . . . . . . . . . . . . . . . . . . . 7-16 l
g 7.4 EFFECT OF TWO DIMENSIONAL (2-D) HEAT CONDUCTION a THROUGH THE CONTAINMENT SHELL , . . , . . . . . . . . . . . . . . . . . . 7-18 7.4.1 Geometry of the Wet and Dry Vertical Stripes on the Containment Outside Steel Surface . . . . . . . . . . . . . . . . . . . . . . . . 7-18
-7.4.2 Inside and Outside Heat Transfer Boundary Conditions for the Conduction Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-19 7.4.3 2-D Conduction (ANSYS) Model Description . . . . . . . . . . . . . . . . 7-20 .
7.4.4 Enhanced Evaporation due to 2-D Conduction . . . . . . . . . . . . . . . 7-21 7.4.5 --Insights from the PCS Large-Scale Testing - . . . . . . . . . . . . . . . . . . 7-21
- 7.5 - THE AP600 CONTAINMENT EVALUATION MODEL TREATMENT OF WATER COVERAGE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-31 7.5.1 PCS Film Coverage Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-31 7.5.2 - WGOTHIC Model . . . . . . . . . . . . . . . . . . . . . . . -. . . . . . . . . . . . . . 7-36 u - 7.6
SUMMARY
OF SUPPORTING TESTS AND SELECTED ANALYSIS . . . . 7-40 7.6.1 Westinghouse Wet Flat Plate Test . . . . . . . . . . . . . . . . . . . . . . . . . 7-40 7.6.2 Small-Scale Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-40 7.6.3 Large-Scale Tests (LSTs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-41 7.6.4 E;timated AP600 Range of Film Coverage Parameters . . . . . . . . . 7-50 7.6.5 AP600 Containment Shell Heatup Analysis . . . . . . . . . . . . . . . . . 7-52
. Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o:\3006w 7.non:1b 101497
iv TABLE OF CONTENTS (Cont.) 7.7 AP600 CONTAINMENT DBA EVALUATION MODEL O FILM COVERAGE SENSITIVITIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-55 7.7.1 Sensitivity of the Evaluation Model to the input PCS Film Flow Ra te . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-55 7.7.2 Sensitivity to the Water Coverage Area . . . . . . . . . . . . . . . . . . . 7-55 7.7.3 Consen atism in the Assumed Time 1 . lay for Application of the PCS Film . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-61 78 CONCLUSIONS AND
SUMMARY
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-64 7.9 NOMENCLATURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-67 7,10 REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-68 APPENDIX 7A PHYSICS OF LIQUID FILMS ON THE AP600 CONTAINMENT SHELL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7A-1 9
s O Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o;\3006w 7.non:1b101497
l y LIST OF TABLES , Table 71 - Ranges of the Film Coverage Parameters in the PCS Tests . . . . . . . . . . . . . 7-4 Table 7-2 - Summary of the Phase 3 Water Distribution Test . . . . . . . . . . . . . . . . . . . . 7-9 Table 7-3 PCS Time Sequence of Events (Based on 440 gpm flow rate) . . . . . . . . . . 7-13 Table 7-4 Elevation D Heat Flux Comparison From PCS Large-Scale Tests RC048C and RC050C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-29 Table 7-5 Elevation E Heat Flux Comparison From PCS Large-Scale Tests RC048C and RC050C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Table 7-6 Clime Wetted Perimeter and Basis for EGOTHIC Model . . . . . . . . . . . . 7-37 Table 7-7 Summary of STC Heated Flat Plate Tests . . . . . . . . . . . . . . . . . . . . . . . . . 7-42 Table 7-8 Summary of Small-Scale ksts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-44 Table 7-9 Summary of Large-Scale Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-48 Table 7-10 AP600 Shell Temperature and Outside Heat Flux . . . . . . . . . . . . . . . . . . 7-51 Table 7-11 Comparison of the Range of Film Coverage Parameters . . . . . . . . . . . . . . 7-52 . Table 7-12 Transient Dry Shell Temperature Increase . . . . . . . . . . . . . . . . . . . . . . . . 7-54
\
Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o:\3006w 7.non:1b 101497
vi LIST OF FIGURES Figure 7-1 Illustration of PCS Water Distribution Weir Assembly . . . . . . . . . . . . . . . 7-7 O Figure 7-2 Conservative Estimate of the Gravity-Driven PCS Flow Rate . . . . . . . . . . 7-11 Figure 7-3 Weir Outflow ............................................. 7-12 Figure 7-4 Comparison of Water Distribution Model to Phase 3 Water Distribution Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-15 Determination of Gamma-Min from LST, SST, and Flat Plate Data . . . . . 7-17 Figure 7-5 Figure 7-6 Normalized Water Evaporation Rate (2-D/1D Conduction) versus Overall Containment Wetted Fraction . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-22 Figure 7-7 Containment Steel Shell Temperature Gradients (*F) with 2-D Heat Conduction; 20 psig, 25% Wetted . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-23 2 Figure 7-8 Containment Steel Shell Thermal Flux Gradients (Bru/hr-ft ) in Y-Direction; 20 psig, 25% Wetted . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-24 Figure 7-9 Containment Steel Shell Total Thermal Flux (Btu /hr-ft 2); 20 psig, 25% Wetted . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-25 Figure 7-10 Thermal Flux in Y-Direction on Outside Surface of Containment Wall 2 [ B tu / hr-f t ] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-2 6 Figure 7-11 Thermal Flux in Y-direction on Inside Surface of Containment 2 Wall [ Btu / hr-ft ] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-27 Figure 7-12 Large-Scale Test Water Coverage Pattern . . . . . . . . . . . . . . . . . . . . . . . . . 7-46 Figure 7-13 Sensitivity to the Input PCS Film Flow Rate ...................... 7-56 Figure 7-14 Comparison of Peak Containment Pressure as Function of PCS Coverage Area ............................................ 7-58 Figure 7-15 PCS Runofi Flow Rates as a Function of Coverage Area . . . . . , . . . . . . . 7-59 Figure 7-16 Comparison of Evaporation Model Peak Pressure with 100% and 50% Constant Coverage Models ............................... 7-60 Figure 7-17 Comparison of Evaporation Model PCS Runoff Flow Rate with 100% and 50% Constant Coverage Models ..............,. ...... 7-62 Figure 7-18 Difference in the Integrated Energy Transferred to the Top of the Dome (with PCS Film Applied at 35 and 337 Seconds) . . . . . . . . . . . . . . 7-63 O Basis and Method for Calculating the PCS Water Evaporation Rate oct.ter 1997 for the AP600 Containment DBA Evaluation Model Revision 1 l o:\3006w.7.non:ltr101497
vii EXECUTIVE
SUMMARY
O The basis and calculational method used to determine the amount of water that evaporates from the AP600 containment steel shell during the operation of the passive containment cooling system (PCS) are conservative; both with respect to the individual elements of the EGOTHIC code and the PCS film coverage model, as well as the method of combining these elements in the Evaluation Model.
= The amount of water that can be evaporated from the AP600 containment shell is input to the EGOTHIC portion of the containment Evaluation Model. The amount of water evaporated-determines the caletdated effectiveness of the PCS in limiting peak containment pressure, as well as_ the capability of the PCS to reduce and maintain low containment pressure following . postulated limiting design basis events.
The basis for determining the amount of water that is evaporated has been developed based on PCS test data. Since the water evaporated at a given containment pressure (temperature) is dependent on the containment surface area that is wetted, the area used in the Evaluation Model is conservatively determined using the following: The portion of the containment shell perimeter that is wetted versus the amount of water. being delivered from the PCS water storage tank to the containment dome has been
- based on testing of the Phase 3 Water Distribution Test (Reference 7.2). This test was !
performed with prototypic water distribution devices on a full sized segment of the dome and the top of the sidewall, using cold water. PCS tests performed with heated surfaces with evaporating water have demonstrated - that cold water on a cold surface conservatively underpredicts the coverage that occurs with heated water on a heated
. surface. # The minimum water film flow rate per foot of wetted perimeter used to determine when water streams begin to narrow in width, conservatively bounds the minimum film flow rates observed in' the PCS tests over the range of anticipated heat fluxes.
The calculational methods for determining the evaporated water flow rate have been developed and are consistent with or conservatively bound PCS test data and observations, and include the following: The evaporation of water due to the conduction of heat in the circumferential direction through the containment steel shell (i.e.,2-D conduction) has been calculated for the alternating, vertical, wet and dry stripes observed in the PCS testing at reduced delivered water flow rates.
. Basis and Method for Calculating the PCS Water Evaporatiori Rate - caober tw7 for the AP600 Containment DBA Evaluation Model Revision 1 0:\3006w-7.non:1b-10H97
vill
- The decrease in the dry surface convective and radiative heat transfer that is calculated to occur with alternating, vertical, wet and dry stripes on the containment shell has been conservatively considered in the AP600 containment Evaluation Model.
Bounding assumptions and conservatisms for the operational characteristics of the PCS have been incorporated in the Evaluation Model. The most significant of these is that the portion of the containment shell surface wetted by the initial 440 gpm PCS-delivered water flow rate is assumed to be no greater than the 90 percent coverage observed in tests with 220 gpm PCS flow. A sensitivity study has shown that the containment design pressure will not be exceeded when only 70 percent of the containment surface is wetted. O t O Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 0:\300ew-7.non:1b-101497
71 7.1 IN'IRODUCTION O The energy released to the containment atmosphere following a postulated design basis high - energy line break for the AP600 is removed from the exterior containment shell surface by a combination of convection and radiation from dry surface areas and by convection, radiation, and water evaporation from wetted surface areas, to a naturally circulating air stream. - The
= energy removal due to water evaporation dominates the PCS total heat removal and is a function -
of the PCS flow rate, the wetted area, and the external shell temperature. Since these parameters '- vary with time, the energy removal rate due to evaporation also varies with time. The conteinment shell outer surface is wetted with water that is stored in a tank located above the containment. Piping and two parallel valves provide a flow path from the tank to the top
. of the containment shell. The valves open upon receipt of a high pressure signal, allowing water from the tank to drain by gravity through the piping to a central distribution bucket located above the center of the containment shell. This water flow fills the distribution bucket, overflows out onto the dome, and spreads outward on' the nearly horizontal surface at the top of the containment shell. As the applied water spreads outward from the center of the dome, it runs down the increasingly sloped dome surface where it is collected and redistributed by weirs located at the ~24-foot and ~51-foot radius of the dome. These water distribution weirs reapply the collected water at a regular uniform spacing'around the containment shell perimeter.
The PCS water flow rate into the distribution bucket and onto the containment surface is contrclied by the inlet elevations of three standpipes within the PCS water storage tank. As the tank drains and each standpipe is uncovered, the PCS flow to the containment surface is reduced in a step-wise fashion. The standpipes are located so that the PCS flow results in sufficient heat removal to match the decreasing rate of heat release to the containment, and to achieve the desired decrease in containment pressure.
- Because the ability of the PCS to remove heat at a given containment pressure (temperature) is
- largely dependent on the amount of water applied and the surface area that is wetted, the method of water application and the behavior / stability of the liquid film are important.
Therefore, this section describes the testing and analyses utilized to define a conservative water flow rate input to the WGOTHIC Evaluation Model, including:
- 1. Water distribution testing used to demonstrate the weir design and how the resulting wetted surface area is affected by the applied water flow rate and surface irregularities in the containment shell stru.ture.
- 2. PCS testing performed with heated wetted surfaces to determine how the water film is
- affected by post-accident AP600 containment operating conditions, including the steel q 'shell surface temperature, the water film temperature, the water film mass flux (mass D
Basis and Method for Calculating the PCS Water Evaporation Rate October 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o:\3006w-7.nort1b-101497
72 _ flow rate per foot of wetted perimeter, hereafter referred to simply as film flow rate), and cooling air flow velocity.
- 3. The method used to predict the containment shell wetted area and water film behavior conservatively compares with test data in order to conservatively calculate the amount of water that can be evaporated from the containment shell.
- 4. The method used to calculate the effect of heat conduction, in the circumferential direction though the steel containment shell (2-D conduction), on th. water evaporation rate from the surface with vertical wet stripes.
The liquid film application, flow rate, area wetted, and film behavior are evaluated in the "PCS film coverage model," separate from the WGOTHIC Evaluation Model. The film coverage model permits a conservative determmation of the amount of supplied water that evaporates from the shell, considering the aspects of water application, and film behavior and stability. The resulting amount of water is input to the EGOTHIC Evaluation Model. The methodology bounds data from tests of an unheated, full-scale portion of the containment dome and 4 feet of sidewall, and from various scale heated tests. Evaporated water flow rate is calculated using a simple model that is consistent with test observations and uses as inputs the parameters Edist and rmin which are selected to conservatively bound test data. rdist represents the film flow rate (mass flow rate per unit wetted perimeter) of water applied by the weir distribution system at the second weir. Fmin represents the minimum stable film flow rate, below which water coverage is assumed to decrease, and is selected to bound heated film stability test data. The database from which conservative values for rdist and Fmin are determined is discussed, as well as how these parameters are implemented into the simple evaporated water model. The model for evaporated water is input to the EGOTHIC Evaluation Model, and a heat flux calculation is performed (Section 7.5). The use of an applied water flow, representing a lower bound on evaporated water, limits the amount of evaporation cooling credited in the Evaluation Model. The Evaluation Model conservatively neglects heat removal during the initial period from the first spillage from the bucket to the time when steady-state coverage has developed on the containment shell (Section 7.5.2.2). The time to develop steady-state coverage is conservatively overestimated. The effects of surface temperature during the initial application are also addressed (Section 7.6.5). The supporting tests for water coverage are shown to span the range of AP600 nondimensional parameters, so that the database is sufficient. O Basis and Method for Calculating the PCS Water Evaporation Rate October 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o:\3006w-7.non:1b-101497
73 7.2 WATER APPLICATION AND DISTRIBUTION O The wetting characteristics of the containment coating and the application and distribution of water onto the containment steel shell outer surface are important design features of the AP600. The AP600 containment is covered with an inorganic zinc coating, and an assembly of devices on the containment dome are used to collect and redistribute water to maximize the containment surface wetted area at a given delivered water flow rate. The Phase 3 PCS water distribution test (Reference 7.2) was performed to demonstrate the operation of the prototype of the AP600 water distribution devices on a full-scale sector of the containment ' dome. Other PCS tests were performed to quantify the heat removal capability of the PCS The test results provided information to understand and characterize the behavior of 3 water films on the outside of the containment surface, in addition to the containment coating and the water distribution devices, other parameters that characterize the water film behavior are the delivered water flow rate, the water film flow rate (per foot of wetted perimeter), the water film temperature, and the evaporative heat flux. The film Reynold's number provides a dimensionless measure of the film flow rate, and the Marangoni number is a dimensionless ' measure of heat flux. The range of dimensioned and dimensionless parameters for PCS testing used to understand and characterize containment surface wetting are summarized in Table 7-1. 7.2.1 Containment Shell Surface Coating The AP600 containment shell surface is covered with an inorganic zine coating for corrosion protection. Prototypical coated surfaces were obtained for testing by following the manufacturers' specifications for preparation of the metal surface and for application of the coating for each test article in the tests described in Sections 7.2.3 and 7.6. The surface was prepared for coating application according to the coating numufacturer's . requirements by sandblasting to a white metal surface finish. The coating was then sprayed onto the surface to a thickness range within the required specification of 4 to 10 mils. Coating thickness measurements were taken to verify that the coating thickness was within specification. Local or spot recoating of the surfaces was performed if the surface of the test article was affected by changes to the facility, such a= the installation of additional instrument
. penetrations.
O
- Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o:\3006w-7.nortib-101497 s.
7-4
~
~ a,c Table 71 Ranges of the Film Coverage Parameters in the PCS Tests 9 Note: (1) Dimensionless groups are defined in Nomenclature, Section 7.9. O Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 l for the AP600 Containment DBA Evaluation Model Revision 1 o:\3006w 7.nortitr101497
75 ,, Although no specific aging simulation of the surfaces was performed prior to testing, matrix tests
) were performed over a period of time using the original coated surface, where aging of the surfaces ocatrred due to operation and exposure to the environment. For example, the small-scale test vessel was erected in 1986 and tests were performed until late 1992 using the same test vessel with the original coating. The LST matrix tests were conducted from late 1991 un'.il the end of 1993 and further operation took place through 1996, with the original coating. An estimate of the equivalent service time cannot be evaluated since a large number of tests were performed during this period in each test facility, no noticeable degradation of the surface was noted during the testing.
In consideration of the above, the surfaces tested are considered prototypic of the AP600 containment shell exterior surface. Measurement and/or observations of film coverage on the prototypical surface were made in each of the PCS tests. 7.2.2 PCS Water Distribution Weir Description and Operation An assembly of devices for distributing the water applied to the containment shell is provided to maximize the outside surface area of the containment shell that is wetted during PCS operation. The PCS water distribution devices include a distribution bucket located above the center of the containment dome, eight divider plates that extend radially from the center of the dome to the first set of water distribution weirs, the first set of water distribution weirs located E v at the ~24 foot radius of the dome, and the second set of water distribution weirs located at the
-51 foot radius of the dome.
The PCS water is delivered to the water distribution bucket at the center of the containment dome. The bucket has 16 vertical slots, such that two slots meter water flow to each of the eight pie-shaped segments on the dome created by the eight divider plates that originate at the distribution bucket and extend radially along the surface of the dome to the first distribution weir ring. These divider plates are required because the center of the dome is relatively flat, and maldistribution of flow due to localized imperfections in plate welds or alignment, or variations in the slope at the center of the dome could otherwise occur. Thus, the dividers distribute the water applied to each one-eighth dome segment and to the corresponding one of eight weir assemblies that comprise the first ring of weirs. The first weir ring consists of eight weir assemblies located at the ~24 foot radius. This radial position is just below a circumferential weld around the containment dome at the 22-foot radius from the dome center. Thus this discontinuity will have no lasting effect on water distribution, since this first set of weirs, just below the weld line, will collect the applied water and redistribuie it. Each of the eight first weir assemblies consist of two water collection dams that direct the applied water, in its one-eighth segment from the dome center, into a collection box. p Each of the eight collection boxes meters flow to two distribution troughs, one on either side of U Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 0:\3006w-7.nortib-101497 l l
7-6 the collection box. Each distribution trough meters the water from the collection box back onto the dome surface via nine V-notches spaced at 1-foot intervals. The eight weir assemblies are installed with the distribution boxes end-to-end, so that each forms one-eighth of the weir ring which completely circles the containment dome at the ~24 foot radius, and which apply water at the 24-foot radius in 144 streams with an ~1 foot stream spacing. Because the containment dome has sufficient downward slope at the 24-foot radius, radial dividers are not required below the first weir ring and the water applied at the first set of weirs will follow the natural fall line to the second weir ring. The second weir ring is located circumferentially on the containment shell at 50.7 feet from the center of the dome, just below the second circumferential weld on the containment shell. This assembly again corrects any uneven distribution of flow that may have occurred below the first weir ring due to weld discontinuities or deviations in the dome shape from the ideal shape. Also, since the containment dome is steeply sloped at this radial position, the water applied by this second weir ring is not significantly affected by local surface imperfections or deviations from ideal shape, since gravity rather than allowable surface variations becomes controlling. Thus, the second weir ring creates an even distribution over the rest of the dome and the vertical portion of the containment shell. The second weir ring consists of sixteen weir assemblies; each with two collection dams, a collection box, and two distribution troughs. The 16 weir assemblies are again arranged end-to-end to form a distribution system that completely circles the containment. Water that runs down the dome from each of the 16 distribution troughs in the first weir ring is collected by the dams, flows into the collection box, and is metered to two distribution troughs. In this second weir ring, the distribution troughs each have 18 V-notches spaced at 6.5-inch intervals. Figure 7-1 is an illustration of a weir assembly. The dams collect all the water flowing from above them and direct this water into their corresponding collection box. As the water rises in the collection boxes, it overflows via three V-notches on either side of the top of the box, effectively dividing the collected water into six equal portions. Each portion of the water overflowing through the six collection box V-notches, flows into one of the three parallel flow channels in each of the two distribution tmughs. As the parallel flow channels fill with water, each flow channel overflows via another set of V-notches arranged equidistantly along the back wall (facing the containment axial center-line) of the distribution trough, onto the containment shell. The eight weir assemblies comprising the first weir ring, have 16 distribution boxes, each with nine V-notches equally spaced at 1-foot intervals; resulting in 144 individual streams of water applied to the dome at the 24-foot radius. The 16 weir assemblies comprising the second weir ring, has 32 distribution boxes, with each having 18 V-notches equally spaced at 6.5-inch intervals; resulting in 576 individual streams of water applied to the containment dome at the 50.7-foot radius. Note that in each weir assembly the spacing of the two streams, one on either side of the collection box, is greater than the uniform V-notch spacing along the distribution boxes. Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o:\300ew.7.non:1b-101497
7-7 (3 f w/ E %
$5, M 5
{k b
,C 20 58 Bb $"3 s N
h 3 I P3 ~k j k
% C e-) .gs N %
E ME @
)eE o YC D N 5 l
k \ 6 e 5
\\
O 6 5
'w) 4 /
J o j de 5x
#IEe i N j k III #sg[
EE1 8 5c,
\ ~s Figure 7-1 111ustration of PCS Water Distribution Weir AsseM )
Basis and Method for Calculating the PC.S Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o:\3006w-7.non:1b 101497
U All components of the water distribution system are seismic category I, and designed to withstand thermal and pressure expansion / contraction of the containment without failure. The system is capable of functioning adequately during PCS accident operation under extreme low or high ambient temperatures. The weir distribution systems are constructed of stainless steel to limit concerns over blockage due to corrosion products or pamt/ coating degradation. The water distribution weir system is designated an AP600 safety class C component based on its containment cooling function. 7.2.3 PCS Water Distribution Testing Results The Water Distribution Test (Reference 7.2) was used to determine the effectiveness of water distribution devices, to determine the water coverage as a function of the flow rate on the prototypical surface, and to determine the time to establish steady-state coverage on the AP600. A full-scale test section, representing a 1/8 sector of the containment dome to the ~50 foot radius and a 1/16 s.ector of the full containment dome and a 4-foot long portion of the vertical sidewall, was built. The test section included both meridional and circumferential joints, with the maximum allowable plate misalignment, and was coated with the prototypic inorganic zine coating. Testing included simulation of the maximum allowable deviation in dome shape from ideal shape, by tilting the distribution troughs. There was no source of heat to simulate mass and energy removal by evaporation for these tests. Two water distribution weir designs were tested. The final weir design was tested in Phase 3 of the Water Distribution Test (Reference 7.2) and is the weir described in Section 7.2.2. These tests demonstrated that the water coverage just below the weirs consisted of discrete streams after the water was ce!!ected, redistributed, and re-applied at a fixed spacing around the contailunent dome perimeter by the water distributian weirs. These individual streams were sufficiently wide at the higher applied flows (35 and 27.5 gpm) to join just below the weirs and provide high water film coverage over the portion of the test section below the weir. However, at reduced applied water flow rates, the streams were sufficiently narrow in width that the water coverage consisted of vertical alternating wet and dry stripes. Below the sacond set of weirs at the ~51 foot radius, where the downward slope of the contairunent dome is 35', the stripes remained discrete from the weir to the springline and down the vertical sidewall. At the lowest flow rate tested,6.9 gpm (equivalent to 55 gpm of water applied to the AP600), the 32 weir V-notches in the lower weir ring distribution troughs produced 29-30 discrete vertical wet stripes with an average width of ~2.5 inches. It is noted that several streams joined together only because of specific worst case surface defects that were simulated on the test section. The water coverage was measured just above the second weir (at the 49-ft radius) and at the springline (65-ft radius at the top of the vertical sidewall). Measurements of stripe widths accounted for only the traverse where flowing water was observed, not the wider wetted Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 on3006w.7.non:1b-101497
7-9 traverse. The Phase 3 test data are summarized in Table 7-2, where the wetied perimeter of the i flowing water was observed and is ILsted as a percent of total area or water coverage.
"" ~
a,C Table 7 2 Summary of the Phase 3 Water Distribution Test
, mummmmmm-Q Q The coverage listed at 49 ft and 65-ft are the measured coverage just above the second weir and at the springline and for the water flow rates delivered in Phase 3 of the Water Distribution Test.
The coverage decreases as the delivered flow rate decreases. The flow rate was not adjusted to c account for the water lost at sampling points upstream of the springline. This correction would increase the water coverage percentages slightly. The surface area that was wetted at a flow rate of 27.5 gpm (equivalent to 220 gpm on the AP600) was estimated to be [ }'# from the top of the dome down to the first weir, based on a review of the video tapes for the Phase 3 Water Distribution Tests. About [ ]"# of the vessel was wet between the first and second weirs, and the entire vessel was wet at the bottom of the test section. This test also demonstrated the time required to fill the prototypic water distribution devices and establish steady-state water coverage on the AP600 containment at a flow rate equivalent to 220 gpm delivered water flow in the AP600. Based on a review of the video tapes of the test, water began to spill from the first set of weirs at about 2.5 minutes, and spilled from the second weir ring at about five minutes after flow into the bucket was initiated. The total time to completely fill the weir devices and establish steady-state coverage on the dome and sidewall was conservatively estimated to be 11 minutes. Since the AP600 initial PCS-delivered flow rate Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Modei Revision 1 o:\3006w-7.r.on:1b-101497
7 10 has been increased to 443 gpm, the above time required to achieve steady-state water coverage will be decreased as discussed in Section 7.2.4 below, h 7.2.4 Delivered Water Flow Rate versus Tm* te The AP60W 443 gpm initial delivered flow rate fills the water distribution devices and wets a sufficient portion of the containment surface to limit the calculated peak containment pressure to less than the containment design pressure following the postulated design basis events. This high initial delivered flow also results in the containment pressure being quickly reduced following the pressure peak. At 3 hours after the PCS water flow is initiated, the delivered flow rate is decreased based on the decreased core decay heat while maintaining sufficient flow to maintain low containment pressure. The PCS delivered flow rate is again reduced at 30 hours after PCS flow initiation. A conservative estimate for the gravity-drive, PCS water flow rate is calculated assuming the single failure of one of two valves (located . parallel) to open. This flow rate is shown in Figure 7-2. A combination of three standpipes . 3 along with orifices to adjust the delivered flow rate with time. Note, the largest flow resistance is in the orifices, so the single failure assumption reduces the gravity-driven flow rate by less than 2 percent. The flow rate decreases with time as the water level in the PCS storage tank decreases. A value of 440 gpm is used for the maximum PCS flow rate in calculating the time required to achieve steady-state water coverage on the containment surface following initiation of PCS water flow. A simple analytical model was developed and the weir outflow rates were calculated for the 440 gpm (0.98 lbm/sec) PCS water flow rate. The results are shown in Figure 7-3. The figure shows the time delay that occurs as each device fills before overflowing, and shows that one to two minutes is required for each device to reach an outflow rate near the maximum. The results in Figure 7-3 help illustrate the time sequence of the flow rates from the bucket, first set of weirs, and second set of weirs that are important considerations for water coverage on the dome and sidewalls. The times for water coverage that are used in the Evaluation Model are based on the actual prototype test results discussed in Section 7.2.3 above, rather than these simple calculations. Increasing the PCS flow rate reduces the time to fill the weirs and establish steady-state coverage. At the 440 gpm PCS initial delivered flow rate, the weir volume will be filled twice as fast as measured in the 220 gpm Water Distribution Test. Therefore, steady-state coverage will be established in approximately 5 minutes after the distribution bucket is filled, as shown in Table 7-3. O Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 0:\3006w.7.non:1b 101497
7 11 r i i \
.~,
80 70 . . . . . . . . . . . 60- - - - - - - - - - - - - - - -- w - - - - - -- - m I t:. 3m 40-C O,g 30 .. . . . LL. l l 20 . . . _ .- ~ . . . . . . . . . . . . 7
,0 . . .
_............q..... 0
. . . . . . . . . . 2 ..... . . . . . . . . . . . . . . . . . . . . . . . . . . . .....u 1 10 100 1000 10000 100000 1000000 Time (secondo)
Figure 7-2 Conservative Estimate of the Gravity-Driven PCS Flow Rate t Basis and Method for Calculating the PCS Water Evaporation Rate October 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o \3006w.7.rmrtib.101497
l
".12 0 ~
a,c e Figure 7-3 Weir Outflow O Basis and Method for Celculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o \3006w-7.non:ltr101497
7 13 Table 7 3 PCS Time Sequence of Events (Based on 440 gpm flow rate) [c )/ Event Time (sec) Signal Actuation 0 Valve Strokes Open 20 Bucket Fills & Spills 36 The weirs are filled and steady-state coverage is 337 established 7,3 AP600 WATER COVERAGE BASIS The PCS film coverage model was developed to calculate the amount of water that evaporates from the AP600 shell, consistent with conservative models for film stability. Inputs to the coverage model are the total PCS flow rate, the sidewall height and diameter, and an estimated evaporation mass flux. The output from the modelis the amount of water that evaporates from the AP600 sidewall. The amount of water that evaporates is input to the EGOTHIC model, rather than the total delivered PCS flow rate. The EGOTHIC code calculation is used to determine an average evaporation mass flux that is input to the PCS film coverage model. Thus, an iteration is required between the two models. The iteration can be made to converge on the (O conservative side (underestimated evaporation flux input to the PCS film coverage model which results in an overestimated amount of water runoff, and therefore underestimates the amount of water that can be evaporated for the input to EGOTHIC). 1 The PCS film coverage model described in Section 7.5 assemes there is no evaporation frun the center of the dome to the second weir. The model further sirnplifies the real geon2try by mo.leling the side wall as a right circular cylinder 108.5 feet high and 130 feet in diameter. This model has the same surface area as the AP600 containment below the second weir, The " film flow rate", represented by the parameter P,is the water film mass flow rate divioed by the circumferential wetted perimeter, that is, T = ih/W. Water is applied to the AP600 shell by a series of streams around the circumference. When the PCS-delivered flow is red: ced from the high initial flow rate at 3 hours, the application of streams typically results in stripes that flow down from the application elevation. Such stripes start with a film flow rate, Edist, that is greater than the stability limit. Thus, the stripes flow down the wall at a constant width until the film stability limit, Emin is reached due to evaporation of the film. Once Emin is reached, film stability cat.ses the width of the stripe to reduce as additional evaporation takes place. In the analytical model, the film flow rate is applied to the shell at a value referred to as Edist and the film stripe only narrows when F reduces due to evaporation so that F = Emin. The bases for Edist and Fmin are presented in this section. (q 4
%.)
Basis and Method for Calculating the PCS Water Evr.poration Rate october 1997 for the AP600 Containment DBA Evaluation Model Revisian 1 c:\300nw-7.non:1b-101497
7-14 7.3.1 Wate Distr %ution Film Flow Rate, Fdist The PCS film cove age model uses a value of P at the top of the side wall to represent the film O flow rate i codt. d by the second weir. Values of water coverage at the springline were measured in Phase 3 of the Water Distribution Test (Reference 7.2). The data, presented in Figure 7-4, show the coverage increased with the total water flow rate to 88 percent coverage at 220 gpm, then increased to 92 percent at 280 gpm. Thus, a model that limits the coverage at the top of t*" side wall to 90 percent bounds the test data at flow rates greater than 280 gpm. Modeling the coverage at lower flow rates with a value of Ed u = 293 lbm/hr-ft bounds the test data for IN gpm and lower flow rates. Since the delivered flow rate in AF600 drops from 426 gpm to pproximately 123 gpm at 3 hours,90 percent coverage for flow rates above 280 gpm and rdist = 293 lbm/hr-ft for flow rates below 123 gpm are reasonable bounds that cover all wual PCS flow rates. The aom temperature isothermal Water Distribution Test data are applicable to AP600 operation with the shell at 180* to 190 F. The basis for using the data is that the water applied to the shell bey.and the second weir during actual PCS operation is heated to the shell temperature while flowing down to the second weir. The scaling analysis estimated less than 1600 ft2 of heat transfer surface area is required to heat up the subcooled water, whereas there is 4400 ft 2 og c wetted surface area above the second weir, and approximately 10,000 ft2 of total surface area (Reference 7.8, Section 7.6.6). Consequantly, heated water is applied to a heated shell at approximately the same temperature at the second weir, so the water and shell are nearly mhanal, as in the Water Distribution Test. The decreased stability exhibited by the application of cold water to a hot surface (References 7.10 and 7.12) is not an issue in this case, , so it is assumed that tha coverage measured in the cold, isothermal tests is conservative for the nearly isothermal application below the second weir in AF600. The room temperature (65' to 68'F) Water Distribution Test coverages are a conservative basis for Fdist in AP600 due to the effect of increased temperature on the film properties. The film spreads where the water spills from the weir V-natch and impinges on the shell surface. The spreading is a momentum-dominated process that is opposed by friction and surface tension. At higher temperatures 180 to 190 F, the film viscosity is decreased by a factor of 2 to 3, and the surface tension is decreased by 15-20 percent, while the impingement momentum is essentially unchanged. The reduction of the friction and surface tension both allow the film to spread more at high temperature than at low temperature, z Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 c:\3006w-7.non:1b-101497
d MS 4 ( J 9 1 1 1 , Modeh Coverage = 90% o - 0.9--- --g o
=
C 0.0- - - ~ ~ ~ ~ - ~ ~ ~ - - ~ ~ - - - ~ ~ - - ~ ~ ~ - - - - - ~ ~ ~ ~ - - - - G
.C g Q,7 ....................................................................................................................
Q. U)
;;; 0.6- ~ ~ ~ ~ - - ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ - ~ ~ ~ ~ ~ - ~ ~ ~ ~ -
C
.O Q $. . . ~ . . . ~ ~ . . . . - ~ ~ ~ ~ ~ . - ~ . ~ ~ . . ~ . . . . - ~ . . ~ _ _ ~ ~ - . . . - - . . ~ . . ~ . ~ . - ~ ~ . . .
o O
@ Model: Gamma-Diot =
- u. O'4- ~ ~ ~ ~ ~ - - - - - ~ ~ ~ ~ ~ ~ ~ - - ~ ~ ~ -293 - - - - - - - - ~ ~ ~ ~ ~ - - ~ ~ ~ - ~ ~ - - ~ ~ ~ - -
tbm/hr-ft a o O V (O Q,g. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B i O. 2 --- -- 1 OM---- - - - - - - - - - - - - - - - - - - - - - - J
- O . . . . . .
. O SO 100 150 200 250 300 Equivalent AP600 Flow Rate, gpm .i + 1 , Figure 7-4 Comparison of Water Distribution Model to Phase 3 Water Distribution Test Results
]
U + Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1
'o:\3006w-7.non:1b-101497
7-16 7.3.2 Mi am Flow Rate, P,in Observatic evaporating film flow on heated surfaces shows the film flows in constant O width stripes until evaporation causes the film flow rate to reach a minimum value, E min after which, the film width narrows with additional evaporation. Most of the LST, SST, and Westinghouse Flat Plate tests produced constant width stripes, or constant coverage. The lowest values of film flow rate, r, either were above r in, m or at most were close to Emin. Consequently, the film measurement data for each of the tests is a record of values of F m2 r in APlot of all the test values of r will show by its lower bound where r m in lies for each test. The LST wetted circumference, heat flux, and runoff flow at the bottom were measured, so these data represent r2Emin. Several of the lowest measured values of F ar presented in Figure 7-5. To be conservative, it is necessary to develop an upper limit on the lower bound of the measured values of P. The work of Bohn and Davis (Reference 7.10) shows the critical film Feynolds number is an increasing function of the heat flux. Since the range of AP600 heat fluxes is relatively low, and since a single value of the critical Reynolds number is easici to use as a limit, it was elected to use a single value of the critical Reynolds number that bounds the highest sidewall heat flux in AP600 (approximately 5000 B/hr-ft2 ). Furthermore, since Re = 40/p and viscosity does not vary much over the range of heated tests and AP600 temperatures, the lower range of film flow rates is bounded with the single value mr in = 120 lbm/hr-ft. The data in Figure 7-5 shows this value to be 5 to 10 times higher than the lower values of r for each test, and is therefore conservative for AP609 use. O Ilasis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Moc'el Revision 1 o:\3006w-7.non:lb-101497
i 7 17 9
- . ._ C
( r b i i l 1 i s i 4 i 5-4 i l Determination of Gamma-Min from LST, SST, and Flat Plate Data Figure 7 ) l Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o-\3006w-7.non:1b-101497
7-18 7.4 EFFECT OF TWO-DIMENSION AL (2-D) HEAT CON DUCTION THROUGH THE CONTAINMENT SHELL h The AP600 PCS transfers heat from the containment atmosphere to the outside environment. Cooling water is applied to the outside surface of the shell to facilitate the heat removal process by evaporation of the applied water. Early in the postulated event, the water applied to the shell exterior provides at least 90 percent coverage of the external surface. As the transient progresses, the applied flow rate is reduced and the water coverage of the external surface area of the shell is reduced as discussed in Section 7.2 and 7.3. As evidenced by test data, the flow distribution weirs develop attemating wet and dry vertical
" stripes" on the containment surface. These stripes become clearly segregated as the applied water flow rate is reduced. Heat removal from the wetted areas is greater than from the dry areas and results in the wetted surface area being cooler than the dry surface (evaporative cooling in the wetted area is much greater than convection and radiation from the dry surface).
This difference in temperature results in heat conduction in the circumferential direction through the thickness of the containment shell. Thermal energy is conducted from the hotter dry stripe areas into the adjacent portions of the containment shell cooled by a wet stripe. The transfer of additional thermal energy to the wet stripe increases the temperature of the wetted steel which increases the water film temperature, which increases the water evaporation rate, the containment heat removal rate, and the use of the delivered water. Since the water evaporation rate calculated by EGOTHIC only widers heat conduction in the O radial direction through the containment steel shell, evaporation rates calculated by WGOTHIC are enhanced by considering the effects of circumferential two-dimensional heat conduction. A description follows of the method used to calculate the effect of circumferential two-dimensional heat conduction on the water evaporation. Section 7.5 describes how the enhanced evaporation rate is used in the AP600 containment Evaluation Model. 7.4.1 Geometry of the Wet and Dry Vertical Stripes on the Containment Outside Steel Surface The AP600 water distribution testing, as discussed in Section 7.2.3, showed that the outside surface of the cont-inment shell will be partially wet when the PCS-delivered water flow rate is reduced below the high initial flow rate. At cold, unheated conditions, the observed side wall wetting was 47 percent with 100 gpm and 24 percent with 55 gpm equivalent applied flows for the AP600. The limited percentages of wetted area were a consequence of the water being applied to the surface at discretely spaced locations, and the fact that the water spread to a stream width that resulted in a bounding I'dist of 293 lb/hr-ft. Therefore, the observed stream width and wetted surface areas were directly proportional to the water flow rate. At the above Basis and Method for Calculating the PCS Water Evaporation Rate October 1997 f for the AP600 Containrnent DBA Evaluation Model Revision 1 l oA3006w-7.non:1b-101497
7 19
/i g flow rates, the stream widths were observed to be less than'the distance between weir slots and .therefore attemating, vertical, dry and wetted stripes formed down the containment below the F
second distribution weir. . 4
- The occurrence of altemating wet and dry vertical stripes on the containment outside surface j was also documented on a hot surface with evaporation in progress in the PCS large-scale test' l- (Reference 7.7).-
In the water distribution test, the streams initiated by the second (lower) set of weirs had a i j eenter-txenter spacing on the vertical sidewall that corresponded to the spacing of applied
- water streams at the weir, multiplied by the ratio of the containment radius at the sidewall to l _ the radius at the weir. For example, the 6-inch weir slot spacing at the ~50-foot radius of the 4- doma produced stripes at a spacing of inches at the sidewall radius of 65-feet, in the LST, with l heat transfer occurring, wet stripes were observed to flow ~ vertically at constant width tc, the F bottom of the sidewall unless almost all of the applied water was evaporated.
i This~ evaluation of the effects of two-dimensional conduction on the wet steel surface j temperature, and resulting water evaporation rate was based on the same alternating wet and l dry stripe pattem and spacing produced by the weir (s) in the water distribution test. However, the location of the second weir ring and the weir ring slot spacing used were updated to correspond to the AP600 plant. Specifically, the weir slots on the backwall of the distribution troughs in the second weir ring are at the SG.7 foot radius, and the spacing between weir slots j is 6.5 inches. This results in an 8.35-inch center-line to center-line stripe spacing at the vertical j sidewall. In a idition, a wider dry stripe directly under the 16-weir collection boxes was taken L into account. p
~7.4.2 Inside and Outside Heat Transfer Boundary Conditions for the Conduction Model ;
L - - g The boundary conditions used in the two-dimensional heat conduction model were established i .by_ a series of one-dimensional, steady-state calculations of the PCS heat transfer process . performed at steady-state containment pressures ranging from 10 psig- to 65 psig (24.7 to._
- 79.7 psia). These calculations were performed using the same heat and mass transfer correlations -
F as used in WGOTHIC. The heat transfer and the temperature differences from' the steam / air mixture -inside. containment through. the steel shell,L and from the wet and - dry outside ! " containment surfacet to the air are provided. The heat transfer and temperature differences were
' used to establish heat transfer coefficients for each containment pressure condition for both the
! inside hent transfer to the inside water film, and for the outside heat transfer from the outside
- water film. These- heat transfer coefficients were reduced based on the conservative multiplication ~ factors (Reference 7.11, WCAP-14326, Rev.1) applied in WGOTHIC, and were F then further decreased to account for the water film and paint layer conductivities and ,
j thicknesses. The outside heat transfer coefficient versus the outside steel shell temperat tre 454 Basis and Method for Calculating the PCS Water Evaporation Rate October 1997
- for the AP600 Containment DBA Evaluation Model Revision 1 I
o:\3006w-7.non:1b-101497 ~ l
7 20 obtained for each pressure condition for the wetted surface, was fitted using a second degree polynomial for use in the conduction model. A constant dry scrface heat transfer coefficient (with a fixed outside cooling air temperature) that accurately modeled the pressure conditions analyzed, established the outside heat transfer boundary conditions. These boundary conditions were reviewed to assure that the heat transfer rates at all containment pressure / temperature conditions were higher than the corresponding heat transfer calculated by lyGOTHIC in the containment analysis. Tids assures that any increase in heat transfer, as compared to the heat transfer with only radial conduction through the containment steel shell, is underpredicted. 7.4.3 2-D Conduction (ANSYS) Model Description The effect of circumferential conduction through the AP600 steel containment shell on the shell surface temperatures and the resulting effects on the condensing heat transfer on the inside surface, the evaporative heat transfer on outside wetted surfaces, and the convective heat transfer from the dry outside surface; were quantified using the ANSYS computer code. The ANSYS computer code is a multi purpose, finite element program that has been used commercially since 1970. For this calculation ANSYS revision 5.3 was used. The ANSYS calculation was a two-dimensional, thermal, steady-state analysis of a periodic half-cell (cross-section) that consisted'of a two-dimensional block [0.1354-feet (1.625 inches)}" C thick and 0.3479-feet (4.174 inches) wide; corresponding to the AP600 containment steel shell thicknese and the spacing of water streams at the containment sidewall perimeter imposed by the PCS water distribution weirs. A thermal conductivity of 24 Btu /hr-ft 'F was used for the steel , material. Adiabatic boundary conditions were used for the right and left side of the half-cell model to represent symmetry and periodicity of the cell. For each steady-state containment pressure analyzed, a half-cell model was established for each water coverage fraction ranging from 0.05 to 0.95. The noding density was increased on each side of the wet / dry interface on the outside surface to increase the accuracy of the heat transfer calculation near the wet / dry interface. In addition to these partially wetted half-cell models, the heat transfer with a completely wetted
..d completely dry half-cell model was analyzed for each containment pressure using the same inside and outside boundary conditions. Smce the half-cell has a 1-D solution when fully wet or dry, these cases provide the heat flux with only radial conduction through the containment shell. The heat flux r, i results of these fully wetted and fully dry cases were used to normalize 'he heat flux rate obtained from the partially wetted cases, where 2-D heat conducticn occurs.
O Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 0:\3006w-7.non:ltul01497
, 7 21 g 7.4.4 Enhanced Evaporation due to 2-D Conduction O ^
- ~ . The heat flux from the wetted portion of the half-cell model was compared with the wetted heat
- flux that occurs when only radial heat conduction (one-dimension)is assumed. Figure 7-6 shows
- the water evaporation rate with two dimensional conduction versus the fraction of wetted area,
- normalized' to the- evaporation _ rate, calculated with only radial heat conduction (one-dimensional) outward through the steel shell, for containment pressures of 10,15, 20, and j 25 psig. These calculational results are bounded by the following polynomial expression which
- is used in the PCS film coverage model to determine the evaporation-limited applied water flow l rate that is input to WGOTHIC (See Section 7.5)
i [ ]* C (7-1) e i _where I
- M= the wetted area heat transfer rate enhancement or multiplication factor x= fraction of containment surface wetted, = W/Wo 1 Several additional plots to illustrate the effect of two-dimensional conduction on the PCS heat
- transfer process are provided for the 20 psig containment pressure,25 percent wetted case. A
=-
j temperature distribution contour plot is shown for the ANSYS half-cell model in Figure 7-7, with -
- the surface inside containment at the top of the page. Figure 7-8 shows the thermal flux from the inside to outside surface (-y direction), perpendicular to the containment shell, and Figure 7-9 l shows the total heat flux (x, -y directions) that occur in the steel shell. Figures 7-10 and 7-11 3
show the thermal flux distribution on the outside and inside surface of the wall, respectively. 7.4.5 Insights from the PCS Large-Scale Testing The large-scale PCS heat transfer tests were largely conducted wit':Egh water coverage fractions . such that circumferential conduction would have little or no effect en the water evaporation rate. 7 An exception is test run RC050C of matrix test 213.1. A clear indication of 2-D conduction i
- effects is seen by comparing the results of RC050C with test run RC048C of matrix test 212.1.
In these tests, the containment pressure and other boundary conditions were essentially the j same, with the exception that the amount of water applied to the external surface of the test vessel was [ ]a,b gpm in test RC048C and only [ ]a,b gpm in test RC050C. i The reduced water flow rate in test RC050C resulted in a reduction in the wetted area observed l at the bottom of the test vessel sidewall, [ ]a,b percent for test RC050C versus [ ]a,b percent for test RC048C. In spite of the reduced wetted area in test RC050C, the total heat removed from the test vessel and the amount of water evaporated in this test was equal to test RC048C. 2 Basis and Method for Calculating the PCS Water Evaporation Rate October 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o:\3006w-7.non:1b 101497
.---m.- ,-m -a v -, -e., ~ ,w , . _ , - - - - - , -
v-
7 22
- ~
a,e O I O Figure 7 6 Nonaallred Water Evaporation Rate (2 D/1D Conduction) versus Overall Contcinment Wetted Fraction Basis and Method for Calculating the PCS Water Evaporation Itate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 0:\3006w 7.norrib-101t>7
WESTINGHOUSE PROPRIETARY CLASS 2 7 23 ANSYS 5.3 MAY 0 1997 11:26:54 PLOT No. 2 WCDAL SOLUT1oh1 STEtel SVD el INSIDE CONTAINMENT I'0*
- d ttrea 381897
$MN e193.528 SMX *199.098
- 183.$20 5 185.251 it WET l DRY OUTSIDE CONTAINMENT
- e x containment wall 2 D t he rma l conduction. 2 0paio, 2 5% overall coverage i
4 Figure 7 7 Containment Stcel Shell Temperature Gradients ('F) with 2 D IIcat Conduction; 20 psig, 25% Wetted
, V ' Basis and Method for Calculating the PCS Water Evaporation Rate July 1997 for the AP600 Containment DBA Evaluation Model o:\3006w.7.wpf.it#.597
7 24 hT.STINGHOUSE PROPRIETARY CLASS 2 0 3 ANSYS.6,.3,, m , 11826414 PLOT No. ) WCDAL COLUTION , t?EP*1 i SUb *1 INSIDE CONTAINMENT U**** mvu R$YS*C SMN **2963 f EMNDea3131 , SMX **227.237 EMXB**118.195 WET I DRY OUTSIDE CONTAINMENT , i b containment wall 2D thermal eenduction. 20pm15 _ 2S% overall coverage Figure 7-8 Containment Steel Shell Thermal Flux Gradients (Btu /hraft2 ) in Y Direction; 20 psig,25% Wetted Basis and Method for Calculating the PCS Water Evaporation Rate July 1997 for the AP600 Containment DBA Evaluation Model o:\3006w 7.wpf:1t>472597
) i i o WESTINGHOUSE PROPRIETARY CLASS 2 7 25 ) i 1 1 4 l ANCYS $.3 MAY 4 1995 11a29i32 P L,C'? NO. 4 NCDAL COLUTION j ETEP=1 3 S'.'*. *1
?!MU+1 INSIDE CONTAINMENT tr$va <^ m EMW *233.81b CMKt*227.166 CMX *6072
, CMX3*4909
' g 233.41%
l ~ g 882
- 3,.632
]
~
C:3 .. .
, e g 2429 3 34*7 - 4124 . 6
- a s.. . . . -
- l. WET ! DRY I
I l b' OUTSIDE CONTAINMENT l i f contal w ent wal. 2D t h e rtna l conawetion. Oopsic. *n overal, eeversee ? l l ? ! Figure 7 9 Containment Steel Shell Total Thermal Flux (Blu/hr.ft 2h 20 psig,25% Wetted 4 Basis and Method for Calculating the PCS Water Evaporation Rete July 1997 ! for the AP600 Containment DBA Evaluation Model { o:\3006w.7.wpf;1W.597 2 i
7 26 WESTINGliOUSE PROPRIETARY CLASS 2 I ANSYS $.) MAY 8 9997 17:01:30 Pt47 NO. 1 POST 1 STEPe1
-- SUD e1 *328.998' g T1Mte1 PATH PL77 NODial **** . 4 30 ' NOD 2sG4 *Te9.849' J tv e1 Digte.1$
XF e.%
. . . . . . I; ::: $* DUFFER . i me.ned '
1979.933' e l 449.640 '
+8tM .979' *a740.1FF' .sese.sys ' - - . ~ . i ..t .. s.e4s t... i t .wp
- s. . ;
3.a us s .t., i 3.631 c- -" DIST containment well 3D therenal conduction. 25% overall coverage Figure 710 Thennal Flux in Y. Direction on Outside Surface of Containment 'tiall [ Btu /hr-ft2 j Basis and Method for Calculating the PCS Water Evaporation Rate July 1997 for the AP600 Containment DBA Evaluation Model o:\3006w-7.wpf.1b472597 l
i 7 27 An'sYS 5,3 uf 8 1997 17:03 90 PLOT loo. 2 PCST) 578Pa1
. n...e.' '?',:t FATH PL*J, . ..eee' 31:!!! -e t . .- iY.,:',. -e '
I; ::: 8*Df>FFBR
.ees.ter' *ess. nee ' *ene.eeg' .oet. ree '
e nese. ore' asets.tvi'
. nee.svi , ,
t two o e I .**e I 4.M6 I h.es? I h.tesI h.79
.MF t.ees t.rse s.4Je s.est DIST Conteirunent well 20 thermal condusttien, 25% overall coverage Figure 711 Thermal Flux in Y direction on inside Surface of Containment Wall
[ Btu /hr ft2 j O Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 oA3006w 7.non:1b 101497 _ =
7 28 Furthermore, the lower portion of the vertical sidewall in test run RC050C was wetted with [ la,b This stripe geometry is similar to that obsen*ed in the water distribution test discussed in Section 7.2.3 and assumed for the 2 D conduction g model. A comparison of the LST vessel shell wall temperatures for these two test runs using the thermocouple pairs (one thermocouple at the inside wall surface matched to a thermocouple at the outside wall surface), used to derive local heat flux rates, provides insight to the effect of circumferential conduction. Tables 7-4 and 7-5 provide comparisons of the inside and outside LST shell temperatures, and the local heat flux derived from the temperature difference across the 7/8-inch thick steel shell; for the two lowest elevations on the LST sidewall. Table 7 4 shows that at the Level D elevation in test run RC04SC all the inside and outside wall temperatures are relatively uniform. This indicates that the outer wall is wetted at all the thermocouple pair locations. The average outside wall surface temperature is [ la,b F, and the average local heat flux is [ la,b Btu /hr ft2 based on the thermocouple pair ATs. In comparison, only four of seven outside wall thermocouples appear to be wetted in test run RC050C (dry outside wall temperatures are very high, [ la.b.F, and the wall AT is rmall). In this test run the average wetted outside wall temperature is [ la,b.F, and the average wetted 2 local heat flux is [ la,b Btu /hr ft . Similarly, Table 7-5 shows that at the Level E elevation (just above the runoff collection gutter) the test run RC048C uniformly wetted outside wall average temperature is [ ]a,b*F and the average heat flux is [ la,b Btu /hr-ft 2. In comparison, test run RC050C which is [ la,b percent wetted at this elevation; indicates that only two of the outside wall thermocouples are clearly wetted, and the average wetted outside wall temperature is [ la.b.F and the average heat flux is [ la,b Btu /hr-ft 2. Note that the test run RC050C thermocouples at the 240' circumferential location show an outside wall temperature and heat flux that is intermediate to the clearly wetted or dry locations. This thermocouple pair may be adjacent to a wet stripe, where circumferential heat conduction would cause these observed intermediate temperatures and AT. There tables show that with the striped water coverage on the outside surface, the wall temperatures and heat flux of the wet portions of the shell are higher than when the outside surface is completely wet. octoter 1997 O Basis and Method for Calculating the PCS Water Evaporation Rate for the AP600 Containment DBA Evaluation Model Revision 1 c:\300hw-7.non:ll*101497
7-29
~ ~ \ a,c
( Table 7 4 Elevation D lleat Flux Comparison from PCS t,arge Scale Tests RC048C and RC050C C V Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 i o;\3006w-7.nortItwl01497
i 7 30 l Table 7 5 Elevation E Heat Flux Comparison From PCS Large-Scale Tests RC048C and RC050C O l 1 i Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o:\3006w 7.nortib 101497 l
7 31 7.5 THE AP600 CONTAINMENT EVALUATION MODEL TREATMENT OF WATER COVERAGE The AP600 PCS heat removal at a given containment pressure is largely determined by how much of the delivered water evaporates. Heat removal is maximized if all the water delivered to the shell evaporates. Consequently, the determination of how much water evaporates from the containment shellis necessary to determine the containment heat removal. The flow rate of water that runs off the shell is the difference between the delivered water flow rate and the evaporation rate. The AP600 containment Evaluation Model includes the PCS film coverage model and the ECOTHIC code. The PCS film coverage model calculates how much of the delivered flow evaporates and how much runs, unevaporated, off the shell. The evaporation and runoff calculation is performed consistent with the simple film coverage model described in Section 7.5.1. Since the simple film coverage model requires heat flux as an input, iterations on heat flux are performed between the film coverage model and the WGOTHIC code. The calculation is performed at enough times to define a transient evaporation flow rate history for input to the EGOTHIC code. The EGOTHIC code calculates the resulting containment pressure history as discussed in Section 7.5.2. The separate PCS film coverage model is used so the EGOTHIC code is not required to determine wetted area as a function of time. The use of only the evaporated flow to calculate containment pressure introduces conservatism because it O discounts sensible heating of any runoff flow, and because a heat flux lower than the heat flux calculated by EGOTHIC is used to calculate the amount of water evaporated. in the AP600 plant, as the water coverage of the containment shell decreases due to the decrease in the delivered PCS flow rate with time, alternate wet and dry stripes are formed on the containment shell exterior surface and two-dimensional (radial and circumferential) heat conduction is established in the containment shell. Accounting for two-dimensional conduction increases the temperature of the wetted steel surace, and therefore also increases the temperature of the liquid film, over what is calculated for one-dimensional (radial) conduction only. The increase in the temperature of the liquid film, in turn, results in the evaporation of more water, reducing the calculated runoff from the shell. Section 7.5.1.3 describes how the increase in water evaporation effectiveness of the PCS is accounted for in the AP600 containment Evaluation Model, when both radial and circumferential heat conduction are important in the steel containment shell. 7.5.1 PCS Film Coverage Model The simple analytical model used to calculate the rate at which water evaporates from the sidewall of the containment shell is described in this section. The model assumes water is delivered to the sidewall consistent with the initial distribution spreading data described in Basis and Method for Calculating the PCS Water Evaporation Rate octoter 1997 for the AP600 Containment DBA Evaluation Model Revision 1 oA3006w-7.nortib-101497
7-32 Section 7.3.1. That is, [ )** percent of the shell circumference is wet for delivered flow rates greater than 280 gpm. When the delivered flow rate decreases to less than 125 gpm, the wetted circumference is calculated using a water film flow rate oi[ }^h Ibm /br ft (rdist). The water g flows in constant width stripes down the sidewall as the water evaporates, until r is reduced to Emin, defined in Section 7.3.2. Thereaf ter, evaporation reduces the .'ilm width while r remains constant at rmin-The evaporation rate model starts with a simple definition that relates the total film flow rate, th; the wetted circumference, or width of the wetted surface, W; and the film flow, or mass flow rate per unit width, P. Each of these is assumed to be a function of the parameter Z, the distance below the top of the sidewall. The equation is: th = r W (72) which, rearranged, also defines F. The derivative of the mass flow rate with respect to vertical distance is also used. Using the chain rule for derivatives: N =WS +r dW (7-3) dZ dZ dZ The wetted coverage and runoff flow rate are calculated based on the following assumptions and boundary conditions:
. The delivered water flow rate boundary condition at the top of the sidewall, sh on is presented in Figure 7 2. T*te initial film flow rate at the top of the sidewall is specified to te r = rhon/0.9W ,o where W is o the containment circumference, for 611vered flow rates greater than [ }*h gpm. For delivered flow rates less than [ ]** gpm, the film flow rate at the top of the sidewall is specified to be rdist = [ ]* lbm/hr ft. The other boundary condition, the width of coverage at the top, W,op is determined from W,op =
0.9Wo, or W,op = rh /rdist, depending on the delivered flow rate.
. The water flows in constant width stripes below each weir slot while the flow rate decreases due to evaporation for ra ist > r > Tmin. The distance, Z, the water flows down the containment sidewah is bounded 0 < Z s Zmin 5 Zmas, where Zmin is the distance down the containment wall at which r = Tmin, and Zmax is the distance to the bottom of the shell. . Atr = Tmin, evaporation causes the film stripe to narrow while r remains c< >stant at Imin-O Basis and Method for Calculating the PCS Water Evaporation Rate ociober 1997 for the AP600 Containment DBA Evalua* ion Model Revision 1 0:\X106w-7.non;1twl01497
7 33 7.5.1.1 Constant Width Coverage O After the water distribution is established at the top of the sidewall by the weir, the film evaporates at mass flux, $m, as it flows down the shcIl in stripes of constant width. The basis for the constant width stripe is the observations of the stripes on the LST, and the physleal explanation in Appendix 7.A 3. For a constant width stripe dW/dz = 0, and dr/dz = - $m. The change equations for th, r, and W for the constant width portion of the stripe are:
= -4mW (74) l = -$m (7-5) l 1
dW =0 dZ (74) With the boundary conditions listed above, and Equations 7-4,7 5, and 7-6, the water mass flow rate, rh, and the film flow rate, r, can be calculated for the constant width evaporation portion Q of the coverage. For the case with $m = constant, the simple analytical expression for the mass flow rate is: rh = rho , - $mW ,Z (7-7) Equation (7-4) can be written in terms of difference equations for a numerical solution where Arh = rh2 - rh ,3 AZ = Z -2 2, and 1 $m is a variable: Arh = - W,op$mAZ (7-8) or rh2 =rhi-Wiop$m(Z 2~Z i)- (79) Knowing th, the film flow rate is determined from Equation (7-2) where r = rdist = rh /W. Basis and Method fo* Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Ce uaintnent DBA Evaluation Model Revision 1 o:\3006w 7.non:1b-101497
7 34 The value of Z when r reduces to rmin is Zmin. The value of Zmin can be determined l from Equation (7 7) when $m is constant: iho,/Wdai -Imin (7-10) im 7.5.1.2 Constant I' min Coverage When T = r m in, the stripe widtn W begins to narrow, while Emin is maintained at a constant value. The resulting change equatiore for ih, r, and W for this portion of the stripe are:
= -$,W (7-11) bdZ=0 (7 12) dW , , &mW dZ f When $m = constant and r = Tmin = constant, the solution to Equation 7-13 is the simple exponential function:
W =W dat e* #" (7-14) When omis not constant with height, the analytical expression for W depends on the functional form of $m and a general expression is written for numerical integration where $m is calculated for each AZ: AW = - W$mAZ (7-15) Imin O Basis and Method for Calculating the PCS Water Evaporation Rate cetoter 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o;\3006w 7.non:1t*101497
i l 7 35 or W2 = W:
, (7 16) min Knowing W from Equations (714) or (716), the mass flow rate at any Z is simply calculated from Equation (7 2). The runoff flow rate is rhog = Wr in,m where W is the wetted circumference at the bottom of the e.ontainment shell, Z = Z,,x.
By inspection of Equation (714), it is noted that W, the film flow per unit width, is always 4 greater than zero. Thus, for constant values of $m and rmin, Equation (714) always predicts some water runs off the wall without evaporating. However, from experimental observations, l all the water delivered to the containment shell is evaporated for some transient conditions. Thus, the preceding calculation method is conservative in its execution. l. 7.5.L3 Spreadsheet Calculation The equations developed in Sections 7.5.1.1 and 7.5.1.2 are solved in a spreadsheet for both one-dimensional and two-dimensional shell heat transfer. During the initial high flow ITS period I when water coverage is high (~90 percent circumferential coverage is used ) and two-Smensional effects are small, only one-dimensional heat transfer is calculated. However, at three hours the
- PCS-delivered flow is reduced from 440 gpm to 123 gpm, and the wetted circumference at the top of the containment shell is predicted to decrease to approximately 51 percent. Figure 7-6 shows the two-dimens.'onal conduction enhances evaporation at this and further reduced coverage 3. Thus, for time greater than three hours, as the delivered flow rate wntinues to decrease, two-dimensional conduction is included in the calculation of the evaporation rate.
The two-dimensional conduction model discussed in Section 7.4 calculated the evaporation rate for a range of wet stripe widths for both one and two-dimensional conduction. The calculation used the same overall temperature difference (and steam partial pressure difference) between the bulk containment and the bulk riser as boundary conditions for both cases. The effective heat transfer coeffici .sts, for mass transfer, were determined for each case. The comparison shows for a given stripe width, the enhancement of the evaporation rate when the real physical case of two-dimensional conductic n is considered. It was found that the enhancement varied with stripe width, but had little effect on the overall temperature difference between the bulk containment and riser. Consequently, the family of curves representing the bulk temperature difference were lower bounded, thereby eliminating the dependence on the bul,k temperature difference. The only dependent variable is the wet stripe width. The enhancement of evaporation is characterized by the multiplier, M, that is a polynomial function of the wet stripe width. M is defined by Equation 71. O Basis and Method for Calculating the PCS Water Evaporation Rate octote 1997 for the AP600 Containtnent DBA Evaluation Model Revision 1 o:\3006w-7.non:1t>101497
7 36 l i
= The multiplier, M is used in the spreadsheet as a multiplier on $m for times greater than 3 hours, to produce a better estimate of the actual evaporation flux from the film stripes.
The evaporation rate calculated with the PCS film coverage model spreadsheet is input to the MGOT HIC cede. Note that the evaporated flow rate is the total delivered water flow rate to the ! l shell minus the runoff flow rate. The MGOTHIC pressure calculation is thereby limited to the amount of flow that is independently r.hown to evaporate. 7.5.2 lyGOT111C Model The WGOTHIC model is described in detail in Section 4. Features specific to water coverage are discussed in this section. The WOOTHIC code uses a special type of heat conductor called a " clime" to model the convection, radiation, conduction, evaporation, and condensation heat and mass transfer processes from the inside of contairunent to the outside of containment. Each clime consists of a horizontal slice of the shell, riser, baffle, downcomer, and shield building of the PCS. [ )*# climes are used to represent the PCS in the containment DDA Evaluation Model. The WGOTillC model uses the following input to compute the evaporation heat removal rate from the shell: the applied PCS water flow rate (the evaporation rate), the temperature of the applied PCS water temperature, and the area and wetted perimeter for each clime. The vertical variation in the wetted perimeter and the resulting wetted area were conservatively evaluated h in the PCS film coverage model, so the WGOTillC code does not calculate the change in these values as a function of time or position. Rather, the PCS water evaporation rate input value is determined in the PCS film coverage model, and is input to WGOTHIC to account for changes in the evapocation rate due to anticipated changes in the coverage with time and location on the shell. The WGOTHIC code uses wetted perimeter inputs for each clime as described in Section 7.5.2.1. The PCS evaporation flow rate that is input to the WGOTillC code is calculated in the PCS film coverage model described in Section 7.5.1. The application of the PCS flow is assumed to be delayed until 337 v c, based on the estimated time required to reach steady state coverage at a PCS flow raw of 440 gpm as described in Section 7.3. 7.5.2.1 Wetted Perimeter Inputs The wetted perimetet of the water on each dime is input to the WGOTHIC model. The model I allows the water to flow at constant width until it reaches the next lower clime, or it evaporates entirely. When it evaporates entirely before reaching the bottom of the clime, the code tracks the distance traveled and breaks the clime vertically into wet and dry portions with temperatures tiasis and Method for Calculating the PCS Water Evaporation Rate octotc 1997 for the AP600 Containment DBA Evaluation Model Revision 1 oA3006w 7.non:1t>101497
7 37 calculated using the appropriate wet or dry heat and mass transfer models. The wetted perimeter input values for the EGOTHIC Model are based on the measured water coverage values from the Phase 3 Water Distribution Tests on the dome, and 90 percent of the shell circumference wetted on the sidewall. Use of the evaporated PCS water flow rate from the film coverage model as the applied water flow rate input to the EGOTlilC model, as described in Section 7.5.1, eliminates the need to vary the wetted perimeter input values with time. The sensitivity analyses, presented in Section 7.6, demonstrate that this approach is conservative as compared to using the actual PCS film flow rate with variable coverage area. The fixed wetted perimeters input for the WGOTHIC climes provide a conservative coverage fraction for the initial three hours of PCS operation when the delivered flow rate is high. However, the amount of water that is evaporated is calculated in the PCS film coverage model, so the area used by EGOTHIC to evaporate the applied water does not affect the evaporation rate. The wetted perimeter value for the top of the dome down to the first weir is estimated from the video tapes of the Phase 3 Water Distribution Test. The coverage area and wetted perimeter change over the diverging area between the first and second weirs. The wetted perimeter specified for this region is based on the average of the value just below the first weir and the minimum measured value just above the second weir. The wetted perimeter does not change much over the steeply sloped region between the second weir and the top of the vertical y sidewall. The wetted perimeter input values are the same for each clime representing the vertical sidewall. The percent of the perimeter wetted is summarized in Table 74. The values listed represent the measurements at the 27.5 gpm flow rate (which is equivalent to a 220 gpm flow rate on the AP600). The use of these wetted perimeters percentages for the 440 gpm PCS flow case is a conservatism in the AP600 containment Evaluation Model. Table 7 6 Clime Wetted Perimeter and Basis for WGOTHIC Model Cilme Percentage Location Method of Determination a,b,c 1 Top of dome to second weir Visual inspection and calculation 2 Second weir to top of vertical Measured sidewall 3-7 __ Sidewall Measured O V Basis and Method for Calculating the PCS Water Evaporation Rate cktober 1997 for the AP600 Containment DBA Evaluation Model Revision 1 oA3006w-7.non.11w101497
7-38 7.5.2.2 EGOTitlC lteration with Spreadsheet The flow rate applied to the shell surface in the EGOTHIC model is the evaporated flow rate O from the spreadsheet calculation. An iteration between thi. epreadsheet and the ECOTHIC model is necessar" to converge on the same evaporation rate in both. The iteration between the spreadsheet and the WGOTHIC cai:ulations proceed as follows:
- 1. An average evaporation heat flux,4,h at selected times is determined from the wet WGOTHIC climes below the second weir. ,
- 2. The evaporation mass flux, $n, = $h /h fg , is input to the spreadsheet. The evaporation rate,rh,,,p is calculated in the spreadsheet for each time using Equations (7-7) and (7-14) for problems with constant evaporation mass flux, and Equations (7 8) and (715) for !
problems with variable evaporation flux. Spatially constant flux calculated from one-dimensional conduction are assumed for the first three hours of the transient, and thereafter variable mass flux and two-dimensional conduction are assumed.
- 3. EGOTHIC is run with th,,,p from the spreadsheet and the calculated results are used to define $hfor input to Step 2 to recalculate the water evaporation rate.
When the EGOTillC calculated values of $h are sufficiently close to, but higher than, the values assumed for input to the spreadsheet under Step 2, the solution is converged. That is because g a higher heat flux input to the spreadsheet will predict more water evaporated. The use of the lower evaporation rate input results in ECOTHIC pressure predictions that are slightly high. 7.5.2.3 Dry Convection and Radiation *icat Transfer Predictions The ANSYS two-dimensional heat conduction results show that the temperature of the dry surface area is decreased compared to the dry surface temperature when only one-dimensional radial heat conduction is used. This results in less radiation and convection from the dry regions. Although WCOTHIC utilizes one-dimensional radial heat conduction, the dry area convection and radiation is not overpredicted because EGOTHIC must use a wet surface area that corresponds to the evaporated water flow rate calculated by the PCS film coverage model. The mporated water flow rite calculated by the PCS film coverage model includes the enhanced evaporatian characterized by the multiplier, M (Section 7.5.1.3). Thus, EGOTHIC must use more cooler wet surface area to evaporate the water flow rate from the PCS film coverage model. This results in less hotter dry surface area, and therefore, EGOTHIC underpredicts the nct radiation and convection from the dry surface. O Basis and Method for Calculating the PCS Water Evaporation Rate octotwr 1997 for the AP600 Containment DBA Evaluation Model Revision 1 oA3006w 7.non ll>101497
7 39 The EGOTHIC conservatism can be estimated using values from the ANSYS two-dimensional calculation, which is the best representation of the heat conduction through the shell. It is assumed the EGOTHIC temperatures and heat fluxes are the same as ANSYS one-dimensional cases.
- At containment pressures of 15 and 25 psig, and with 50 percent wet coverage, the two-dimensional ANSYS model predicts dry heat transfer (radiation and convection) is 73 and 82 percent respectively, of the one dimensional value. EGOTHIC will predict 67 percent of the one-dimensional dry heat transfer, since it reduces the dry surface area available by 33 percent.
- For the same containtnent pressures, at 25 percent wet coverage, the two-dimensional model predicts the actual dry heat transfer (radiation and convection) is 83 to 91 percent of the onedimensional value. EGOTHIC will predict 57 percent of the one-dimensional dry heat transfer, since it reduces the dry surface area by 43 percent. Thus, EGOTHIC again predicts less dry energy removal than two-dimensional model predicts.
It is concluded that the MGOTHIC model pred!sts less dry heat transfer than the two-dimensional ANSYS model. O Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evat: .ation Model Revision 1 c:\3006w-7.nortit>101497
7 40 7.6
SUMMARY
OF SUPPORTING TESTS AND SELECTED ANALYSIS This section provides a summary of the PCS tests and data that are relevant to water film coverage and film behavior, and which support the Evaluation Model. In addition Section 7.6.4 provides an estimate of the range of film coverage parameters that can occur in the AP600, and compares this parameter range to a composite of the ranges tested. Section 7.6.5 summarizes an estimate of the heatup of the AP600 containment shell versus time. This heatup versus time is utilized in the ,ensitivity study of PCS flow initiation time presented fn Section 7.7.3. l 7.6.1 Westinghouse Wet Flat Plate Test The primary purpose of the Westinghouse wet flat plate test was to generate heat and mass transfer data for evaporative cooling with parameters that bound the expected conditions on the AP600 containment shell. A secondary purpose was to observe the film hydrodynamics l including possible formation of dry patches due to surface tension instabilities. The test article is described in Reference 7.3. Tests were performed in two orientations, vertical (to represent the sidewall) and 15 degrees from horizontal (to represent the upper portion of the dome) with various cornbinations of air velocity, film flow rate, and heat flux. A stable, wavy laminar water film was formed easily on the hot, coated, steel surface, even in the vertical orientation. A description of the test section and results from the various tests are given in Reference 7.3. The test data are summarized in Table 7-7. Two of the heated flat plate tests were run with very low film flow rates at relatively high heat flux (6000 - 8000 DTU/hr ft 2) to force the film to completely evaporate before reaching the end of the test section. The observations given in Reference 7.3 state the following: "The upper part was 80 percent wetted and fingers of water film extended down 4 feet to within 2 feet of the end of the heated plate. The bottom of the fingers slowly moved up and down. The dry patch between fingers was between 1/4-inch and 1-1/2 inches wide. As the width varied in time, the lateral, slow flow of liquid could be seen feeding the thinnest parts of evaporating film. These two tests showed tnat the end point of water films on the containment would still be stable film evaporation, even with very thin films and high heat fluxes." 7.6.2 Small-Scale Tests The small-scale tests were designed to provide heat and mass transfer data for both the inside I and outside of the test vessel. The test apparatus consisted of a 3-foot diameter,24-foot high steel pressure vessel filled with air at atmospheric pressure into which steam was supplied to I Basis and Method for Calculating the PCS Water Evapontion Rate ocioter 1997 l for the ANO0 Containment DBA Evaluation Model Revision 1 o:\3006w-7a non:lt>101497 { 1
l N1 maintain various pressures. Water was applied to the extemal surface to simulate evaporation in AP600. 'lhe pressure vessel was surrounded by a clear, plexiglass shield that formed a 15-inch wide annulus for either forced or natural circulation-driven ai flow and allowed observation of the applied extemal film flow. The tests were conducted with varying steam supply flow rates, water film ilow rates, water film temperatures, cooling air flow rates, and cooling air inlet temperature and humidity, instrumentation was provided to measure intemal steam condensation rates, extemal water evaporation rates, inner and outer wall temperatures, film temperatures, air velocity, temperatures, and humidity. A summary of the test data from Reference 7.4 (lor tests with measured water coverage) is provided in Table 7-8. The following observations and conclusions (with respect to the water film) were drawn from these tests:
- A stable, uniform, wavy laminar film was formed on the inorganic zinc-coated steel surface using simple welrs.
* 'Ihc (2m remained stable and uniform on the vertical sidewall of the vessel at average evaporating heat fluxes in the range of those expected on the AP600, 7.6.3 Large Scale Tests (LSTs)
The Westinghouse large-scale PCS test facility was built to provide heat and mass transfer test data for a geometrically similar model of the AP600 containment vessel. The wsts provided experimental data used for evaluating the physics in containment, determining the relative importance of various parameters that affect heat and mass transfer, and validating computer codes and models. The following provides a discussion focused on the use of LST data to develop a bounding film coverage model. Three series of tests were run at the Westinghouse large-secle PCS test facility. The steady state pressure, annulus air flow rate, extemal water flow rate, injected steam flow rate, injection velocity, location and orientation, and noncondensible gas concentration were varied between the tests. Test conditions were t. elected to provide heat and mass transfer validation over a range of post accident containment operating conditions for the AP600. O Basis and Method for Calculating the PCS Water Evaporation Rate cetoter 1997 for the AP600 Containment DBA Evaluation Model Revision 1 oA3006w.7a rmitib-101497 l
= _ _ _ . _ . - _ -
7-43 l l
~
a,C O l 1 l
- 2 h
2 l d E 1 d A
'E b
E E 0> R a Basis and Method for Calculating the PCS Water Evaporation Kate octc6er 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o;\Nuw 74.rmn:ll*101497
l 7-43 a,c . b v I O 2 g d
~
d E I y o 6 E 5 E
.e. 3 F-l ~ -
Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o \3006w 7a.rcrt1L*101497
7 44 A,C O 9 2 h 4 J! 4 Ji t b E E O 2 s t # I _ g Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 c:\3006w 7a.non:ll*101497
7 45 The large-scale PCS test facility is a 20-foot tall,15 foot diameter pressure vessel that simulates the AP600 containment vessel. The geometry is approximately a 1/8-scale of the AP600 containment vessel. A plexiglas cylinder is installed around the vessel to form the air cooling annulus. Air flows upward through the annulus via natural convection to cool the vessel, resulting in condensation of the steam inside the vessel. A fan is located at the top of the annulus shell to provide the capability to induce higher air velocities than can be achieved with purely natural convection. Water is applied to the elliptical dome surface by two rings of J- l tubes. This method of application resulted in a series of spaced, wavy laminar flow stripes. At low tnst pressures the stripes spread within a few inches of their application point to form a continuous wavy laminar film. At high pressure the continuous film separated to form discrete stripes. The following important observations with respect to film behavior were made during the tests:
- The J tubes resulted in a non uniform distribution of water on the surface of the LST, similar to that observed in the Water Distribution Test. i l
1 Some J tubes dripped and others had noticeably lower flow rates. This resulted in some j regions of the dome and sidewall that were just we; or had a very low film flow rate. l g . As the pressure and temperature increased inside the pressure vessel, dry spots first V began to form in the wet, but low flow regions on the dome and sidewall. With increased pressure and heat flux, the dry spots grew vertically (both upward from the gutter and downward from the dome, between dripping or low flow J-tubes), separating the original continuous film into wavy laminar flow stripes. At higher heat fluxes, dry spots also formed just below, and in line with the J tube location. A typical coverage pattern for high heat flux and high flow rate is shown in Figure 712. The central, wavy laminar flow region of the individual film stripes was surrounded by a region of laminar flow (with no visible waves). The thickness of the laminar flow region appeared to continually decrease out to the very edge (or bottom) of the film stripe.
. The widths of both the wavy laminar and laminar flow regions of the stripe were observed to decrease with increasing heat flux. At high flow rates, the width of the stripe was observed to remain relatively constant with elevation as the film flowed down the vertical sidewall. At lower flow rates, the stripe width was observed to taper uniformly with elevation as the film flowed down the vertical sidewall.
O 4 V Basis and Method for Calculating the PCS Water Evaporation Rate octater 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o:\3006w-7a.non:1tr101497 l t
7 46 9 b i v . i d, > Q f* , lf' ~M$}
):5; $s? }ll{k'"h; $ $l. $f 1- ; if*}
pin V;'f I
?
ll${ ;{ QQf ;% .
';[
10
}f ': l%il k
i. g })l i y! f 90' {p N j:) - d g U h d [ M klIM [p fh g[f! pte :ppf ( 9
; idA # gLu . p igi. e, ii S- :
q
- g. iti j<LF p . A f. 4 's i .
i : 4 it , 8: En bd i NW )l1 Q 9 l bs. jif; eQg Ni ]ls & N0 ! Wi e{ sl] "
;l!
pg j i:n 9 : T gg ! l Egi l ; lq %11bl b 4 [ji @
^
I : , t kl a y N @g p g Kb l; $ i [!
! 8: Sir ,j. , @f' .1 } 9% ((s! 'I C
it?!? [Rif gg - Figure 712 Large-Scale Test Water Coverage Pattern Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluat!on Model Revision 1 oA3006w 7a.nortit>101497
7-47
+ The film stripes remained stable (i.e., they did not split or bunih up to form thick, narrow rivulets) as they evaporated on the vertical sidewall.
The applied PCS flow rate was observed to vary or oscillate at a slow but regular period during some tests. This phenomenon was the result of sharing a common water source with a boiler feedwater valve that opened every two minutes, thereby reducing the PCS flowrate to the Ftube header. From observations made during testing the flow oscillations had an effect on the water coverage fraction;it was most noticeable at the bottom of the sidewall. Tiv length of the narrow film stripes and the width of the wider film stripes both decreased when ti:e flow was observed to decrease. The dryout point of the narrow film stripes was observed to rise up and fall down the sidewall as the flow oscillated. At no time were the stripes observed to become unstable due to the oscillations; the process remained well behaved and repeated itself with the periodicity of the applied flow. After the Baseline LSTs were completed, instrumentation was added so the transient inlet and outlet cooling water flow rates could be measured and recorded by the data acquisition system. All of the Phase 2 and Phase 3 tests, with the exception of the blind test, (220.1, RC062) were included in the evaluation. The steam injection location, velocity, and initial pressure were much different in the Phase 3 tests, and subsequently, the level of stratification within the vessel was different than the PW 2 tests. The differences in stratification resulted in changes in the coverage from the top to the bottom of the vessel; the top was less well covered than the bottom p during some of the Phase 3 tests. The test numbers that were evaluated are listed in Table 7-9. b Measurements of the dry stripes on the vessel were taken just above the gutter during the defined steady-state periods of each test. The time the measurement was recorded on the data sheets for each test. This could have been either the time the measurement was started or finished. In test 221.1D, the time of measure:nent does not match with the stated steady-state time period. The test engineer postulated the recorded time to be one hour off, i.e.,12:45 was recorded as 1:45 by mistake. The following assesses the effects of variations in flow during the time taken to record coverage data at the gutter. From recorded test data the maximum and minimum exit mass flow rates were determined over the approximate time the wetted perimeter measurement was made. The time taken to perform these measurements was related to the namber of dry stripes; more stripes took longer to measure. A 15-minute band on either side of the stated time of measurement was used in this evaluation to bound the time it took to make the measurement. The maximum and minimum exit film flow rates were calculated by dividing the maximum and minimum mass flow rate by the measured wetted perimeter value. Because the film flow rate is calculated by dividing the mass flow rate by the wetted perimeter, if the wetted perimeter were slightly less than measured (due to a reduction in the mass flow rate), the film flow rate
,3 would be higher than calculated with this method. Similarly,if the wetted perimetce were I <
V Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 a\3006w-7a.non:1b-101497
7-18
~ ~
a,C Table 7 9 Summary of Large-Scale Tests 1 1 1 1 l 9 g Basis and Method for Calculating the PCS Water Evaporation Rate october tw7
- for the AP600 Containment DBA Evaluation Model Revision 1 o.\3006w 74.non
- ltwl01497 1
7-49 a,e Os Table 7 9 Summary of Large Scale Tests (cont.) l ( O V Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 c:\3006w 7a.non.1t*101497
\
7-50 I slightly higher than measured (due to an increase in the mass flow rate), the film flow rate would be lower. In either case, the difference between the maximum and minimum film flow rates would be smaller than calculated. The maximum and minimum film flow rates for the tests are tabulated in Table 7-9. An evaluation of the LST data (Reference 7.5) yielded some additional important conclusions with respect to film coverage and heat removal:
. Evaporation is the primary usode of heat removal from the outside of the vessel. Sensible heating of the subcooled liquid film, convection, and radiation are second order. . Striped film coverage prcwided better heat removal than forced quadrant coverage for the same wetted coverage. . The highest heat flux occurred near the top of the dome at the elevation where the extemal film was applied for all of the wetted LSTs (except the horizontal, high-velocity, steam jet injection case). Although the dome represents about 30 percent of the heat transfer surface area, approximately 40 percent of the total heat removal occurred on the dome and 60 percent on the cylindrical sidewalls.
Injection csf high-velocity steam (similar to a steamline break) resulted in a well-mixed vessel (both above and below the operating deck), and thus, a relatively uniform wall temperature and heat flux over the evaporating surface. The test data related t 3 water coverage from References 7.6 and 7.7 are summarized in Table 7-9. Tests 207.1,207.3,208.1,216.1A, and 216.1B were conducted with water coverage by quadrants and are not representative of AP600 conditions and are therefore excluded from the table. The data of Table 7-9 are used to develop a bounding film stability model as described in Section 7.3.2. 7.6.4 Estimated AP600 Range of Fihn Coverage Parameters The estimates for the maumum and minimum values for the range of AP600 film coverage parameters during a DBA are calculated using the simple approach des:ribed below. The AP600 range of film coverage parameters is compared with the range of the PCS tests. To determine a maximum sidewall film flow for the AP600 rate, none of the initial 440 gpm PCS water flow is assumed to evaporate on the dome. Mear.urements from the unheated, Phase 3 Water Distribution Tests indicate that at least 90 percent of the perimeter at the top of the sidewall will be wetted with 220 gpm, assuming this same wetted parameter at the higher actual PCS delivered flow rate results in an estimated maximum sidewall film flow rate of 1 Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 0:WO6w-7a.non:lb-101497 6
7 51 ! 590 lbm/hr-ft. The maximum sidewall Re mm would be 32N O ' .;a estimated maximum 200'F film temperature. The liquid film Reynolds numbers range up to 20,000 in the test data (Referenec 7.11). The shell heat flux provides the boundary conditions for the evaporating film. The steady-state, shell average heat flux and film temperature were calculated for tne subcooled, evaporating, and dry portions of the shell, assuming an initial ambient air and film temperature of 120*F. These calculations were performed at the containment design pressure,45 psig, to bound conditions at the expected DBA peak pressure and at 22.5 psig, for conditions representative of 24 hours after blowdown. The results are presented in Table 7-10. Table 710 AP600 Shell Temperature and Outside Heat Flux Avg Subcooled Avg. Evaporating Avg. Dry Film Shell Containment Heat Flux Film Heat Flux Temp. Heat Flux Temp. Pressure (psig) (BTU / hr-ft2) Ten p. (*F) (BTU / hr ft 23 (.F) (BTUIhr ft 23 (.p3 45 7740 155 5500 190 455 260 22.5 3700 147 2500 '75 300 220 V To account for stratification, the maximum wet uhell heat flux is estimated to be 50 percent higher than the average subcooled value (at 45 psig); this would be about 11500 BTU /hr-ft2 . The minimum wet shell heat flux would be 0 BTU /hr-ft2 , l* The initial PCS film temperature di be between 40 F and 120*F. The 120 F value is used in the DBA Evaluation Model to minimize the benefit of heat removed by heating the subcooled film. The film temperature will increase as the film flows down the dome. The maximum evaporating film temperature was estimated to be less than 200 F. Since the resistance to heat transfer through the thin film is very small (compared with the mass transfer and internal resistance, as shown in Reference 7.8) and the heat flux is relatively low, the maximum wet shell surface temperature is estimated to be less than 20 F higher than the maximum film temperature even at the conservatively high 11500 Bru/hr-ft2 beat flux. The resultit g estimated range of the AP600 film parameters during a DBA is summarized in Table 7-11 and compared with the composite test data range. The test data parameter ranges are sufficient for evaluating the film stability model. It is important for the test data to cover the higher range of heat flux and the lower range of the sidewall film Reynolds number for evaluating the film stability model. Films with high
, k ynolds number values on low heat flux surfaces are more stable than films with low Reynolds v
Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 0:\3006w-7tnon:lb-101497
7 52 Table 711 Comparison of the Range of Film Coverage Parameters l 4 I number values on high heat flux surfaces. The maximum tested heat flux is almost twice as high as the estimated maximum AP600 value. Tests were run at low film flow rates and to dryout, so the lower range of film Reynolds numbers are also covered. 7.6.5 AP600 Containment Shell Heatup Analysis This section summarizes an analysis of the heatup of the AP600 containment shell versus time. This analysis will be utilized in the sensitivity strdy on the importance of the time at which PCS flow is put on the containment dome following a DBA (see Section 7.7.3). The shell surface temperature begins to increase following a DBA. The time for the dry outer chell to reach a given temperature is a function of the internal containment gas temperature, the internal energy transfer coefficient, and the shell thickness. The time can be calculated using the properties of the [ ]' c steel shell and Figure 4-8 from Kreith (Reference 7.9).
/
The initial shell temperature is assumed to be 120*F. The time for the dry external shell surface temperature to reach the boiling point (212*F) can be calculated with the following input: T = 212 F (external shell surface temperature) Tg = 120 F (initial shell temperature) T., = 250*F (internal containment gas temperature) ( = (T - T,.)/(T i - T ) = 0.292 O Hasis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 0:\3006w.74.non;1b-101497
i 7 53 l l The properties of the steel shell are given below: k,- =. 25 BTU /hr-ft-F (thermal conductivity) L =
- [0.1354)*# ft (thickness) 2 a =- - 0.49 ft /hr (thermal diffusivity)
If the internal energy transfer coefficient (primarily condensation in air) is assumed to be much larger than the external energy transfer coefficient (primarily forced coiwe: tion) such that the dry outer shell surface can be considered adiabatic (insulated) over the tine period of interest, then the time to reach 212*F will be minimized. The results of the calculation are tabulated as a function of the internal energy transfer coefficient below (Note: Bi = Blot Number):
- U (BTU /hr-ft2-F) 1/Bi at/L2 t (seconds) ~
5 36.8 43 < 10 18.4 22 50 3.68 - 5 100 1.84 -2.6 Based on the scaling analysis calculations (Reference 7.8), the average, internal energy transfer coefficient for AP600 initially following a large LOCA event is estimated to be about 2 100 BTU /hr-ft -F. Therefore, the time for the dry outer containment shell temperature to reach 212*F is estimated to be between 300 and 400 seconds. The EGOTHIC Evaluation Model calculates shell surface temperatures at the top of the dome, before application of the PCS film can be compared to the above hand calculations. During the initial 5.5 minutes of the transient, the containment gas temperature (and therefore the maximum possible internal shell surface temperature) is maintained at about 250'F by condensation on the
. heat sinks inside containment. The dome surface temperature is predic+ed to be 174*F at 337 sec, and without external water is projected to reach 212'F at 500 sec, which is reasonable agreement with the estimates above. The heatup rate from the EGOTHIC calculation is about 0.2'F/sec and falls between the-50 and 100 BTU /hr-ft 2-F internal energy transfer coefficient values -
assumed in the hand calculation.
.The calculated temperature increase in the dry external shell surface is compared to water coverage events as a function of time in Table 7-12.
At: the maximum time delay for initial water application to the shell (36 seconds, from - Table 7-12), the outer shell temperature is calculated to increase less can 4*F. The temperature increase of the dry portion of the outer shell is less than 70 F at the time the weirs are filled and O
- Basis and Method for Calculating the PCS Water Evaporation Rate odober 1997 for the AP600 Containment DBA Evaluation Model Revision 1 oA3006w-7a.nortit>101497
7 54 - steady-state coverage is established (337 seconds, from Table 7-12). Therefore, the wall temperature is less than 190 F at the time steady-state coverage is established. Water coverage is not adversely affected by application of the film to a hot, dry, shell surface. Both the STC wet flat plate tests and the LSTs verified the ability of the water film to wet and rewet a hot, dry surface (temperature exceeding 240*F) with the inorganic zine coating. Video tape records of the Westinghouse wet flat plate tests show the initial wetting, dryout, and re-wetting of a hot, dry plate in both a vertical and inclined position. The dry plate temperature wa3 estimated to be about 240'F (based on the maximum heating fluid temperature). An applied wavy laminar film quickly covered the hot, dry plate. As the flow rate was reduced, the waves in the film became smaller and eventually disappeared. The plate remained visibly wet until after the film flow was turned off, then dry patches appeared and grew in circumference as the plate dried out. Video tapes also show the initial wetting of the LST vessel. The measured shell ' surface temperature was about 260*F at the time the water was applied. The film front was observed to '3izzle" as it quickly advanced downward and covered the surface of the elliptical dome. Table 7-12 Transient Dry Shell Temperature Increase increase in Dry, External Event Time (sec) Shell Temp. ('F) Signal Actuation 0 0 Valve Strokes Open 20 0 Piping Fills 34 2 Bucket Fills & Spills 36 4 Weirs are filled and steady-state 337 68 coverage is established O Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o:\3006w-7a.non:1b 101497
7-55
- 7.7 AP600 CONTAINMENT DBA EVALUATION MODEL FILM COVERAGE SENSITIVITIES Sensitivity analyses performed with the AP600 containment DBA Evaluation Model are provided in this section. The model's sensitivity to the PC, film flow rate and water coverage are studied.
An estimate of the conservatism in the assumed time delay for PCS film application is also studied. 7.7.1 Sensitivity of the Evaluation Model to the Input PCS Film Flow Rate Calculations were performed using the WGOTHIC code with the AP600 containment Evaluation Model described in Section 4. The delivered PCS flow rate presented in Figure 7-2 was applied to the EGOTHIC model. Sensitivity calculations were performed by decreasing the input PCS flow rates to 75,60,50, and 25 percent of the nominal value. Recalling that the tir,ie it takes to fill the headers and writs is inversely proportional to the film flow rate, the time of film application was adjusted in each case to account for the decreased film flow rate. The water wetted perimeter input value was kept the same (90 percent) for each case, assuring the difference in calculational results was due only to applied PCS flow. Figure 7-13 presents the change in peak containment pressure as a function of percent change in applied PCS flow rate. As expected, the peak containment pressure increases as PCS flow rate decreases from its nominal value. Decreasing the PCS flow rate results in the following, The time of film application is increased. The heat removed from the containment to heat the cool applied PCS water is reduced. The amount of evaporation from the containment shell can decrease. These reductions in the energy removed from containment during the early portion of the transient, result in an increase in the calculated peak pressure for the transient. Note that the containment pressure decrease is very modest un;.1 the applied flowrate is significantly decreased. This is because the initial decreases in applied flow only decrease the runoff flow rate and not the wetted area. Therefore, the amount of water evaporated remains constant. 7.7.2 Sensitivity to the Water Coverage Area A sensitivity study was performed to determine the effect that PCS water coverage area has on the AP600 peak containment pressure for a DBA LOCA as calculated by EGOTHIC. The AP600 containment Evaluation Model described in Sections 4 and 7.5 was used to perform the calculations with only one-dimensional heat conduction through the shell. The sensitivity study considered a range of sidewall water coverage fractions from 20 to 100 percent. These input coverage fractions were kept constant over the entire transient. The delivered PCS water flow rate shown in Figure 7-2 was used in each case. D Basis and Method for Calculating the PCS Water Evaporation Rate October 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o:\3006.v-7b.non:1b-101497
7-56 O SensitMty to the input PCS Film Flow Rate 10 I S ps E i 5 ED N O s 6 (s) , , 0.2s oJ e.7s 1 Normalized PCS Flow Rate Figure 7-13 Sensitivity to ;he input PCS Film Flow Rate O ! Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o:\3006w-7b.non:1b-101497
7-57 Although the ultimate heat removal capability of the PCS is not affected by the coverage fraction l () the heat removal at a given pressure is affected. The transient pressure comparison is shown in Figure 7-14. As the water coverage fraction decreases, the peak containment pressure increases. For the 100 percent coverage case, the peak pressure is about 43 psig. The containment design pressure limit,45 psig, is exceeded at 70 percent and lower coverage. Decreasing the coverage fraction results in a decrease in the amount of evaporation at a given ecstainment pressure (temperature). As the coverage fraction decreases, the reduced evaporative heat removal causes the containment pressure to increase until the evaporation rate per unit area increases sufficiently to remove enough heat to match the energy input into contairunent. The transient PCS runoff flow rate is shown in Figure 7-15. The runoff flow rate is the difference between the PCS delivered flow rate and the evaporation rate. As the input coverage area decreases, the amount of evaporated water decreases and the runoff flow increases. Figure 7-16 presents a comparison of the pressure transients for the 50 rnd 100 percent coverage cases to the Evaluation Model. The level in the PCS water storage tank drops below the first standpipe at about 10,800 seconds causing a substantial reduction in the PCS flow rate (from 423 gpm to 123 gpm). For the 100 and 50 percent coverage cases, this results in a large decrea;.e in the runoff flow rate, but no OV change in the evaporation rate, which is dictated by the containment pressure (temperature). Note that all the delivered water is not being used. Pressure continues to decrease, although at a slower rate in both the constant coverage cases since in both cases evaporation is removing more heat than is being released to containment. But in the Evaluation Model, the containment pressure increases when the delivered flow decreases. This occurs because the water coverage partion of the Evaluation Model decreases the wetted perime'er to ~51 percent, (i.e., the wetted surface area is decreased in accordance with the decrease in the applied water flow rate). The increase in pressure reflects the increase in the evaporation rate required to achieve a balance between the heat removed from and the heat input to the containment. Therefore, the Evaluation Model containment pressure approaches the same pressure as the 50 percent fixed coverage case. Pressure then begins to decrease again when the evaporative heat removed at the area dictated by the delivered flow rate exceeds the heat input. At about 40,000 seconds, the IRWST is assumed to empty. After the IRWST empties, the flow for core cooling is provided by the sump, which is assumed to be at saturation. Since most of the internal heat sinks (except concrete) are saturated, the PCS is the primary heat sink at this time and must now absorb the energy that had previously been absorbed by sensible heat addition to the cool IRWST water. The containment pressure increases until the heat removal rate (primarily evaporation from the PCS) exceeds the heat generation rate. The pressures for all three cases remained below the 24-hour goal of 1/2 design pressure. n V) ( Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 oA3006w-7b.non:lb-101497
)
7-ofB 5
- \o a 3 r s 8 0t i 0hs 6
w ea 7An F Eoe OO$ 4E$# b P d g i n6M o0 r u t 0e t r e t 1 h om ~ bCo 1 od 7-0n 1 , 1 t f 4 t
\
4 7irao oo 7n , mC e a C
}.1 i! I. a q c. ic 'i$
J1 nc l o We , i
.i' t
u m
/ /,.-.,. ** /./ (
Da l , t p , Bi a Wo Ang r i s
./ .
Et o ( n
/- '.c.j . >
vh c . B 9e . , . ae l o i , i uP f s , aC P d (o t , iS o e ) 9 . nW a o
- f. '.- .>
k J Ma t C o n , oe s Ne dr n
\
s . - eE l v a t a i n e J
/,. , . .3 I
I 1 p d No s . _ o m
. :. 'i,a ; _
r e , n _ a . '!. _ t t ve ) f ;. N _ i
=
o P , n r e R a s s go t u - e r , e a . s e , F , u n , c t i
- o : :
7 ~ - -
~_r ~ ~ -
r . - i . : ~ : ; . o n o
. 0 eE ,
- I iE sg
- g f
P F e8oj~ C . S C g X g I g _ o v f , f I e Roc r .:.. i ;!!! i8 3 i et a vo g ib s e e ir o 1 A r n9 9 e 1 7 a e 9
)1l; l !
O ofB
\ro a 3f s Xt M hs i
- a. o tn oc m u.
.$ =g 8' $. "
w ea
- n 7Ad
- b. P F
i g n6 o0 M u r n0e l t e , h bCo l , l
- 7- , l i 1
0n 1 od 1 oo , i f l l 5 Ll 4 t f 9 iao r , 7n , mC el a P a's . .. '
!l.. _
nc C S, , t u W. . S ( ,
;!l; Dla R o , .
Bi Ang t u e ,
* ,: ;. i . ' i ; n.-
n o g s , Et . . ' vh ae f f u , l uP F l o qS i
)
aC , g t iS o w , nW R A.E.....- a u ,
- s Ma t
o e t e s m no ,, ,. , dr e E a o ,
- s l
v s l 3
'
- s- -
a a , p o F g , r a u n o , c u n no
)
t i o t i , n o n g , I l I g Ra o t f e , C , vo o i "is;, s.:. t :
!a i v ,
i: e ,
':1;.
r , a , g , e , A e r e r ~ -
- _ - - - . 3 a
, .g v. $ c ,g - l _
va
- . "i _ -l.- .
I f ( { g 1 E Ro ec ;i.i1; '.i g :itl. l I , t vo i b s ir e /' o 1 5 n9 9 9 1 7 1 li,
l
%J omm '> E g:
g l 9 5' Comparison of Evaluation Model Peak Pressure with l g*[ g'g j' 100% and 50% Constant Coverage Models p e , a y a E 9 8, T' su "
- 55
, -gr _ 4 5' ' sn e = nn 50 - - - - --- P y' oo Containment Pressure Limit = 45 poig gB 45 -_ ou -a A...- , . . '..~ - s!
~
E i-N' '- h$ [!. 3 35 l . _ ,, k if E 30 .
--- f r .- e Fe '
- c. a.
- '25 ....'. .
- 8. 2 '
[ l 2-T I 5' E : N' j y 220 m
; N .; ? ._ - - ~
45 ; g g e w 10 :-
'? :
3 5 E -
, . 0 . . .....i . . ....i . ....i 10 100 1,000 10,000 100,000 $ 1 8 Time (sei.onds) ae =
IN % h% g c c pp p . . . . . . . . . . . . . . . . .
<E G'g E ,,,
a ,4 w4 O I O O
. . - . .~ . - . -. . . . . .
7-61 The transient runoff flow rate for these three cases is shown in Figure 7-17. The runoff flow rate for the 50 percent coverage case is higher than the 100 percent coverage case. The lower evaporative heat removal in this case results in a sustained higher containment energy content and subsequently higher pressure. Note that there is vtually no runoff flow in the Evaluation 3 Model case since the water coverage portion of the mode! 'imits the applied water to the amount that can evaporate. 7.7.3 Conservatism in the Assumed Time Delay for Application of the PCS Film A delay in application of the PCS film is assumed in the DBA Evaluation Model to cover the time it takes to fill the weirs and establish steady state coverage, as described in Section 7.2. The coverage delay time 4s conservative in that it neglects energy removal from the shell while ' steady-state film overage is being developed. The following assessment shows the amount of conservatism in the predicted energy removal is small. To quantify the amount of energy removal neglected during the development of steady-state film coverage, the }YGOTHIC calculation used to access the heatup of the containment shell, described in Section 7.6 was extended to 1,800 seconds. The heat removal results from this case with the water film applied at 337 seconds were compared to the results from a second case in which the assumed water coverage delay time was reduced to 35 seconds. The same input water coverage fractions were used in both cases. Note that the }yGOTHIC Evaluation Model assumes that steady-state water coverage develops instantaneously after a specified time required to fill the weirs and develop steady-state coverage. The 35-second delay case is a more realistic estimate of the film application delay time for the top of the dome, but will overestimate heat removal from the rest of the dome and sidewall. Therefore, only the heat removal from the top of the dome will be compared for the two cases to estimate the effect on heat removal. Figure 7-18 compares the integrated energy removal rate from the top of the dome as a functior of time. There is very little difference in the energy removal rates for either case. This is the result of the fact that the time required to significantly heat the containment extemal shell is comparable to the 33-second delay time for water application. Recall that, from Table 7-12, the external shell surface is calculated to heat up about 68*F after ~5 minutes when steady flow conditions develop. The energy release difference at the lower portions of the dome and sidewalls will be even less. Therefore, the assumed water coverage delay time, although conservative, has a minor effect on containment pressure for the 440 gpm PCS delivered water flow rate. Basis and Method for Calculating the PCS Water Evaporation Rate October 1997 for the AP600 Containment DBA Evaluation Model Revision 1 0:\3006w.7b.non:1b-101497
i I[1l 7-x fB 6 3
\or as 2 0t Chsi 3ea oo E E wn g B W> ,, E E e , ng a!. f w
7Adn
- b. P g F
i "8$5Dg# oEgu$ge& I n6M o0 u r n0e 1
- t e h
1 6Co od 7- } E g= 0n t f 1 " 1 4 7 : s 'a ;, 9 irao E 7n , mC el a aC nc no , t u Da l dm 5p E Bi t 0a ( . . Ang %rs i m Et Co s , vh on g ae l n s o u
%P.
uP t f , .' ' aC t a nE q$ l , ' iS o t v ) , a . nW Cp g a , Ma t oo vr e , oe e a J , M8
/
dr r ti a e E g o o - e n l v a l d
\
p o r MM o o # a d d o t i o n el e l s P u n o ,, l,_ . R a C S y , t , e R , u , n o , o f f , F , _:__.s l o ,
~:
w - o ___:r -
- _r 5r ~
R a t e
- e8 #8o vo
- 8, es r k w E* g i
t h 1 0 0 g g 5 fIE 4 Ro % ;
;6: 8 ;.
et c vo i b s e ir o 1 n9 9 1 7 O h I
7-63 g c5 8 ! 3 - I_ i _l< Sj i a , 2s 13 En _li ll IM e FW * ' v3 h - f a O 3 }3 , 11 IE h . _g i - at
.E - ,
S l - W ! e -
- E :
o , l ,
,,,i ,, ,,, ,, ,,, ,, i, ,
a 4 S*$ suolllIn (n18) eWoc o} peJJe;sueJ1 ASJous pensJBa}ul Figure 7-18 Difference in the Integrated Energy Transferred to the Top of the
, Dome (with PCS Film Applied at 35 and 337 Seconds)
Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 0:\3006w-7b.non:1b-101497
1 7-64
7.8 CONCLUSION
S AND
SUMMARY
The basis and calculational method used to determine the amount of water that is evaporated from the AP600 containment steel shell during the operation of the passive containment water cooling system is conservative; both with respect to the individual elements of the EGOTHIC code and the PCS film coverage model, as well as the method of combining these elements in the Evaluation Model. The amount of wate'. that can be evaporated is the important input parameter to the EGOTHIC portion of the Evaluation Model. The amount of water evaporated determines the effectiveness of the PCS in limitmg peak containment pressure, as well as the capability of the PCS to reduce and maintain low rontainment pressure following postulated limiting design basis events. The basis for determining the evaporation limited flow rate has been developed based on PCS test data and observations, and includes the following: The portion of the containment shell perimeter that is wetted versus the amount of water being delivered from the PCS water storage tank to the containment dome has been based on testing of the Phase 3 Water Distribution Test. This test was performed with prototypic water distribution devices on a full sized segment of the dome and top of sidewall. The relationship of wetted perimeter to delivered flow is . -aservatively bounded by the linear equation, rdist = Delivered Flow / Wetted Perimeter where: Edist is a constant = 293 lb m /hr-ft This wetted perimeter divided by the actual perimeter (but which is 50.9) provides the initial containment water coverage fraction which is directly related to the amount of j water evaporated at a given containment pressure. The several PCS tests performed with hot surfaces with evaporating water have demonstrated that the above Edis, obtained with cold water on a cold surface conservatively bounds the actual ra ist that will occur with heated water on a heated surface during operation of the PCS.
- In the heat flux range of PCS operation, water streams on the containment surface are observed to become narrower in width only when most of the water in the stream has been et sporated. The Evaluation Model uses a rmin of 120 lbm/hr-ft, as the film flow rate at which water streams will become narrower. This minimum film flow rate Basis and Method for Calculating the PCS Water Evaporation Rate O
Octobu 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o:\3006w-7b.non:1b-101497
7-65 conservatively bounds the observed minimum film flow rates observed in the PCS tests over the entire range of anticipated heat fluxes.
' Water or streams of water on the containment below the second water distribution weir ring (2 35' downward slope) and on the vertical containment sidewall are always observed to flow downward, following the natural fall line of the dome surface. In addition, these water streams are only observed to split or become unstable when the water film ibw rate is ul' min-The calculational methods 'or determining the amount of water evaporated have been developed and are consistent with or conservatively bound FS test data and observations, and include the following:
The evaporation of water due to the conduct *on of heat in the circumferential direction througn the centainment steel shell has been calculated for observed alternating, vertical, wetted and dry stripes observed in the PCS testing at reduced delivered water flow rates. The reduction in dry surface convective and radiative heat transfer that is calculated to ; occur with alternating, vertical, wet and dry stripes on the containment shell has been determined to be conservatively considered in the WGOTHIC portion of the Evaluation Model. O Bounding assumptions and conservatisms for the operational characteristice of the PCS delivering and applying water to the containment surface have been incorporated in the Evaluation Model including: The portion of the containment shell surface wetted by the initial high PCS-delivered water flow rate is conservatively assumed to be 90 percent although at the high initial flow rate,-100 percent coverage is expected. A sensitivity study has shown that the containment design pressure will not be- exceeded when only 70 percent of the containment surface wetted. Coverage at lower flow rates is based on cold water data - which are believed to underestimate the coverage area. The minimum delivered PCS flow rate used in the Evaluation Model assumes the single failure of one of two parallel valves in the PCS water storage tank discharge flow path to open. A 337-second delay time is used to account for filling the water distribution devices and for establishing steady-state water coverage over the containment shell. No credit is teken for any containment heat removal due to heating the delivered water or due to evaporation, prior to when the steady-state water coverage is established. Basis ar.'d Method for Calculating the PCS Water Evaporation Rate October 1997
- for the AP600 Containment DBA Evaluation Model Revision 1 c:\3006w-7b.non:1tr101497
7-66 I
- The temperature of the delivered PCS water is assumed to be 120 F,5'F higher than the design basis maximum ambient temperature, to minimize the amount of containment heat removed in heating the water to the temperature at which it is being evaporated.
- The evaporated water flow rate calculated by the PCS film coverage model neglects the subcooled heat capacity of the runoff flow.
O O Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o-\3006w-7b.non:1b101497
7-67 7.9 NOMENCLATURE Dimensionless Groups l Bi
- Convection hL l Biot Number: gm *7 Ma = Surface Tension Force 8o BT 6 2 Marangoni Number: Viscous Force TTTCW 4r Reynold Number: Re = Momentum Force =_
p
]
Parameters g = ' gravitational constant h = convection heat transfer coefficient k = conductivity L = characteristic length, rh = mass flow rate > M = multiplier representing the ratio of 2-D to 1-D heat transfer V (1
# = surface heat flux T = film temperature; W = width of water film stripe Z = vertical distance from top of sidewall Greek Characters a = thermal diffusivity, = surface angle of inclination relative to horizontal P = film flow rate = mass flow rate per unit width of film, 6 = film thickness, p = liquid density a' = liquid surface tension & = heat or mass flux 0 = contact angle between the surface and film p = liquid viscosity /^\
Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model ' Revision 1 o:\3006w-7b.non:1b-101497 l
7-68 7.10 REFERENCES 7.1 A. T. Pieczynski, W. A. Stewart, WCAP-13884, " Water Film Formation on AP600 Reactor Containment Surface," February 1988 7.2 J. E. Gilmore, WCAP-13960, "PCS Water Distribution Phase 3 Test Data Report," December 1993 7.3 W. A. Stewart, A. T. Pieczynski, L. E. Conway, WCAP 12665 Rev 1, " Tests of Heat Transfer and Water Film Evaporation on a 14cated Plate Simulating Cooling of the AP600 Reactor Containment," April 1992 7.4 R. E. Batiste, WCAP-14134, "AP600 Passive Containment Cooling System Integral Small-Scale Tests Final Report," August 1994 7.5 R. P. Ofstun and D. R. Spencer, PCS-T2R050, "Large-Scale Test Data Evaluation," May 1995, Westinghouse Electric Corporation 7.6 F. E. Peters, WCAP-13566, "AP6001/8th Large-Scale Passive Containment Cooling System Heat Transfer Test Baseline Data Report," October 1992 7.7 F. E. Peters, WCAP-14135, Rev.1, " Final Data Report for PCS Large Scale Tests, Phase 2 and Phase 3," April 1997 l 7.8 D. R. Spencer, WCAP-14845, Rev. 2, " Scaling Analysis for AP600 Containment Pressure During Design Basis Accidents," June 1997 7.9 Frank Kreith, " Principles of Heat Transfer," 3rd Edition,1973 7.10 M. S. Bohn and S. H. Davis, "Thermocapillary breakdown of Falling Liquid Films at High Reynolds Numbers," International Journal of Heat and Mass Transfer, Vol. 36, pp 1875-1881 (1993) 7.11 F. Delose, R. P. Ofstun, D. R. Spencer, WCAP-14326, Rev.1, " Experimental Basis for the AP600 Containment Vessel Heat and Mass Transfer Correlations," May 1997, Westinghouse Electric Corporation 7.12 T. Fujita and T. Veda, " Heat Transfer to Falling Liquid Films and Film Breakdown Parts I and II," International Journal of Heat and Mass Transfer, Vol. 21, pp. 97-108 and 109-118 (1978) O Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o:\3006w.7b.rntib-101497
APPENDIX 7A O PHYSICS OF LIQUID FILMS ON THE AP600 CONTAINMENT SHELL O O Basis and Meti od for Calculating the PCS Water Evaporatior, Rate october 1997 for the AP600 ( ontainment DBA Evaluation Model Revision 1 o:\3006w 7b.non:1b-101497
7.A 1 TABLE OF CON 111NTS
^
7 A.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.A 2 7.A.2
SUMMARY
OF GENERAL LIQUID FILM BEHAVIOR . . . . . . . . . . . . . . . . . . . 7.A 7.A.3 CONTACT ANGLE AND SURFACE WETTABILITY . . . . . . . . . . . . . . . . . . . 7.A-8 7.A 4 DESCRIPTION OF LST OBSERVATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.A 14 7.A.S CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.A 18 7.A.6 REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.A-20 LISI' OF TABLES Table 7.A-1 Sunur = * . s ,' u ten. . >~ Contact Angle . . . . . . . . 7.A-12 tyq t' e _, .. s Figure 7.A i t eitio.; in Sw ha iNm Dect the S:afac~ ' r. Heated Li.itud Pi!m . . . . . . . . . . . . . . . . . . . . . . . . , 7. A-6
- Figure 7.A- >*'-
T..-i W1 Onbem Covt - + . .,x'ancing and Recading f.entae w .2 .... .....................................7.A-10 Figure 7.A Sketch of Qualitative Wavy Lamin.:e Film Flow Characteristics on Heated LST Shell Water Stripe . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . 7. A-15 Figure 7.A-4 Large-Scale Test Water Coverage Pattern at High PCS Flows . . . . . . . 7.A-17 , Figure 7.A-!i Sketch of LST Observation of Vessel Exterior at Water Flows Similar to LST 213.1 Showing Complete Dryout of Some Stripes . . . . . 7.A-19 . A U Basis and Method for Calculating the PCS Water Evaporation Rate . october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 0:\3006w.7b.non:1b-101497 1 i
7.A-2 7.A.1 INTRODUCTION The total evaporation from the external shell is the parameter of interest for mass transfer, the O dominant means of removing heat from the containment. Total evaporation is equal to the integral of the mass flux over the covered, or wetted, area. The mass flux for a given set of parameters (surface and film temperature, film flow rate, annulus conditions) is given by correlations presented in Reference 7.A.1. The subject of this appendix is the wetted area of the external shell surface, and how the wetted area is limited by film stability effects. Note that the initial application of water to the external surface at safety analysis basis surface temperatures is discussed in Section 7.6.5, so that quasi-steady water coverage is assumed to be established in the discussions of this appendix. The introduction and Section 7.2 provide a brief overview of the AP600 design, as it relates to film stability considerations. The test program for AP600 water coverage is discussed in Section 7.6, where it is shown that the range of nondimensional parameters for AP600 is adequately covered in the test program. Subsequent Apper. dix 'A sections give a summary of literature findings on film stability, a discussion of the contact wetting angle that addresses the wettability of the coated surface in the context of surfaces studied in the literature, and a description of LST observed liquid film behavior for high and low flow tests. The physics summarized in this appendix were considered in the development of the PCS film coverage model. The coverage model for the AP600 Evaluation Model is biased to conservatively bound test data that inchsde cold full-scale tests and smaller-scale heated surface tests. The AP600 double dam-weir system is designed to evenly distribute the PCS water onto the surface of the dome. The elliptical shape of the dome and corresponding area divergence helps spread the stripes of water flowing from the individual V-notches in the weirs. Water coverage on the top of the AP600 dome is the most difficult to quantify, but water coverage on this portion of the dome is also the least important to the successful operation of the PCS; the area between the top of the dome and the second weir is only about 10 percent of the total shell external surface area and is neglected in the PCS film coverage model calculation of evaporated flow rate. The distribution system applies water to the shell in discrete, evenly spaced streams. Water from the PCCWST drain header falls inta a bucket suspended just above the center of the AP600 deme. Slots on the side of the bucket allow water to spill at discrete locations around the circumference onto the containment dome. From there, the water flows outward and downward, spreading due to the area divergence, until it is collected and redistributed by a series of two weir rings. Weir outflow rates as a function of tirn including t'ie initial filling of the bucket and dams, is shown in Figure 7-3. The method of water application on AP600, by weir slots, O Basis and Method for Calculating the PCS Water Evaporation Rate odober 1997 for the AP600 Containment DBA Eve.luation Model Revision 1 0:\3006 v.7b.non:1b.101497
7.A - Induces discrete water streams that can remain discrete at low PCS water flow rates and merge to form continuous circumferential water coverage at higher PCS water flow rates.
- The initial application of water flowing from a weir slot hits the surface and spreads until..
surface tension and skin friction dissipate the momentum. If the film is significantly subcooled ' relative to the surface at that point, thermocapillary effects (see Section 7.A.2.2) may also affect how wide the stripe is as it flows down from the point of application. The AP600 has two weir rings on the dome. By the time the water exits the second weir ring, the water has been heated to a temperature relatively dose to that of the shell, so that thermocapillary effects are less important. Therefore, the focus for film stability is on evaporating film stability. 4 Evaporation of the PCS water results in a reduction of the mass flow rate as the fnm advances down the containment structure from the second weir. As the mass flow decreases, the wetted perimeter of the stable film flow also changes. From observation of tests, the wetted perimeter typically decreases as the mass flow decreases. The physical processes that limit the amount of stable film coverage on the AP600 containment shell are discussed in this appendix. 7.A.2
SUMMARY
OF GENERAL LIQUID FILM BEHAVIOR This section provides a summary of available literature on models and data for liquid films and 1 provides a discussion of the various aspects of liquid film behavior, , O 7.A.1.1 Literature Summary The study of movement in a fluid interface has been studied over 150 years. In studying the -
. spreading of a drop of alcohol on the surface of water, British engineer and physicist James Thompson correctly explained the phenomena as a surface-tension-driven flow. The name of =
Italian physicist Carlo Marangoni has been associated with two distinct but related surface c effectsc The first is motion in a fluid interface caused by local variations in interfacial tension which were, in tum, caused by differences in composition or temperature. The second phenomenon is a conjugate of the first;it is the departure from equilibrium surface tension that is produced by the extension or contraction of an interface. Both of these phenomena are important to the understanding of the behavior of liquid films. The stability of liquid films has been studied by many analysts and experimenters within the last 50 years. These studies may be grouped in two general categories; 11 Determining the minimum flow rate required to rewet a stable dry patch.-
' 2. Examining the thermocapillary breakdown of a thin film.
Il u Basis and Method for Calculating the PCS Water Evaporation Rate October 1997 for the AP600 Containment DBA Evaluation Model Revision 1 oA3006w-7b.nostit>101497
7.A-4 Films are generally categorized t 4 saturated films or subcooled films, due to differences in stability, or wetting performance. Films that are applied at or near the temperature of the surface are typically referred to as " saturated films." Such films, when applied to heated surfaces as is done on AP600, have a significant evaporation component and are thus called " evaporating films." Norman and McIntyre (Ref 7.A.2) reported data showing that a large increase in the minimum film wetting rate was required as the temperature difference between the surface and film was increased (that is, subcooling of the liquid film relative to the surface was increased). Hallet (Ref. 7.A-3) also observed this phenomenon and developed a film breakdown correlation that was related to the film surface tension difference, the wave number, and the heat transfer coefficient. Fujita and Ueda (Ref. 7.A-4) measured the breakdown of both subcooled and saturated liquid films on heated, vertical, polished, stainless steel tubes. A comparison of the results from their tests also showed that the highly subcooled films are unstable at flow rates several times higher than that observed for saturated films. More recently, Bohn and Davis (Ref. 7.A-5) measured the breakdown of subcooled water films on heated, vertical, polished, stainless steel tubes and developed a film breakdown correlation that was Apendent on thermocapillary effects. Thus, there is clearly a basis for separately considering film stability for subcooled and evaporating films. The conclusion that thermocapihary effects influence the early breakdown of subcooled films is based on the following. Subcooled films having liquid temperatures much lower than the solid surface temperature absorb heat, causing the film temperature to increase. Evaporating films that are more nearly in thermal equilibrium with the solid surface, transfer mass and energy from the film surface to the gas atmosphere. Thus, one explanation for the apparent reduced h stability of subcooled films is the existence of significantly higher temperature gradients through the film that give rise to increased thermocapillary forces (see Section 7.A.2.2). The manner in which data has been presented in the literature is also of interest. In general, the surface heat flux is recognized as the dominant independent parameter, and properties have a strong influence on film behavior. The literature presents data most often as film flow rate (mass flow rate per unit wetted perimeter) versus heat flux. To account for the effect of viscosity on wettability, the AP600 data reduction uses film Reynold's number as the dependent parameter, with surface heat flux as the independent parameter. The performance of the coated surface to be used for AP600 can be compared to the performance of the typical surfaces studied in the literature, polished steel and polished copper. The use of polished materials in laboratory tests allows careful characterization of the important parameter, the wetting angle. The coated surface does not lend itself to characterization of a single local wetting angle (Section 7.A.3.2). Therefore, the data for film flow rate versus heat flux give an appropriate means of comparison of film stability data. Stable film flow rates on the order of Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Contai.unent DBA Evaluation Model Revision 1 i l oT3006w-7b.nort1b-101497
i 7.A 5 ! l l 20 to 50 lbm/_hr ft are noted on the LST and other AP600 test surfaces, even with heat fluxes up ) to 10,000 Bru/hr-ft2. Comparison to Fujita Ueda data shows that the coated surface is significantly better at wetting, and is less sensitive to heat flux than the polished surfaces. The list of papers reviewed and considered for application to the AP600 containment Evaluation ) 3 Model is extensive and will not be given here. However, in a summary article (Ref. 7.A-6), Bankoff provided an extensive list of relevant papers, many of which have been considered for 4 application to AP600. The current state of the art is focused on the " moving contact line," which
- was also considered for application to the AP600 containment Evaluation Model, but is generally I not very practical for engineering application. i
) 7.A.2.2 Thermocapillary Effect 1 ! Based on discussions with Bankoff (Reference 7.A-8), the thermocapillary effect is a result of the variation of surface tension with temperature in moving from the contact line to the free film surface (see Figure 7.A-1). For a stable stripe shape, the forces in the horizontal direction must sum to zero. The surface tension decreases as temperature increases, so the minimum stable film flow rate has to be greater to prevent the hotter liquid at the surface from causing the film stripe width to contract. The thermocapillary effect on the force balance is sometimes estimated (as in equation 7.A-2) by replacing the actual c(T) fur.ction with a much simpler function using the temperature drop through the film which can be related to the heat flux as l
%rf - Ofilm " %rf -(%rf Tfiim) = k 1
l This simplification becomes increasingly inaccurate as the film subcooling increases, since the ! j' sensible temperature increase of the film invalidates the approximation q"/k, used to estimate l the film surface temperature. l Overall, investigators have identified momentum, surface tension, body (hydrostatic) force, j thermocapillary, and vapor thrust as the dominant forces affecting film stability. These forces are typically expressed as functions of flow rate, heat flux, fluid properties, and wetting angle. Vapor thrust can be neglected in AP600 because the heat flux is low, less than 10,000 BTU /hr-ft2,. I Consequently, for AP600 liquid films, film stability may be considered to be controlled by a
- balance between momentum, surface tension, hydrostatic, and thermocapillary forces.
O Basis and Method for Calculating the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o:\3006w 7b.non:1b-101497
)l
7.A-6 O O-(Tfilm) 0-(Tsurf) O
~
t"ttt"tttttt Applied Heat Flux, q" Tsurf > Tfilm O surf I 9ilm Figure 7.A-1 Variation in Surface Tension Over the Surface of a Heated Liquid Film Basis and Method for Calculating the PCS Water Evaporation Rate october '997 O for the AP600 Containment DBA Evaluation Model Revisio 1 o:\3006w.7b.norcib 101497
7.A-7 7.A.2.3 /.vailable lheoretical Analytical Models O O The available analytical theoretical models have not been found to be practical for application to AP600.; Rather, the AP600 Evaluation Model includes a film coverage model that is consistent with the physics of liquid films, and is developed to provide a conservatively bounded total water coverage. However, models proposed in the literature can be used to gain insight into film behavior, h Zuber Staub model (Ref. 7.A-7) considers the stability of a dry patch located .within a uniform, flowing film, i.e., the inability of the liquid film to rewet the dry patch. -h , mathematical formulation of the model includes three of the dominant terms identified above: momentum, surface tension, and thermocapillary. The model uses a vertical force balance at the tip of a postulated dry patch to determine the minimum uniform film thickness required to rewet the dry patch. This minimum film thickness is a funcuon of the surface heat flux, the film properties (including the contact angle between the film and sum.
- One of the Zuber-Staub formulations treats the film thickness as the dependent parameter from which film stability criteria can be derived. Although film thickness is not easily measured, film thickness is related to the film flow rate through continuity. Therefore, the discussions that follow will treat the film flow rate as the controlling parameter from which film stability criteria ,
for AP600 may be derived. O According to the Zuber-Staub model, if the film flow rate !s greater than the minimum stability value, any dry patch created in the film would be washed over and would readily disappear after formation due to the momentum of the flowing film. Conversely,if the film flow' rate was equal to or less than the minimum stability value, a dry patch, if formed, would be predicted to be stable (i.e., the film would not be able to recover the dry patch). The Zuber-Staub model does not consider the effects of waves in recovering the dry patch. The concept of a force balance can be used to develop insight into controlling parameters for film stability. A force balance more specific to the AP600 design that includes momentum, surface tension, thermocapillary, and body forces (and thus, surface inclination angle, ) to account for - spreading on the inclined surface of the elliptical dome, but neglects the vapor thrust term, may be' written in terms of the film flow sate, F. Since the relationship is for a stable film width, equilibrium between the various forces is assumed. If the film flow rate is greater than the value ' of r in the equation, the film will wash over any dry patch which happens to form. The equation, which can be used to examine the minimum stable film flow, rmin, is: Basis and Method for Calculating the PCS Water Evaporation Rate octobn 1997 for the AP600 Containment DBA Evaluation Model Reviaton 1 0:\3006w 7b.non:1b-101497 l
7.A-8 1 9g sin Emin 3pg cos3 prmin o(1-cose) doq"cose ___ + = - (7.A-2) 15 pp 8 sin 1 dT k 3F.E mm 3 2
,g sinp p Note that the formulation given above assuraes a laminar film with uniform film thickness and does not consider the effect of waves in wavy laminar flow. Waves in wavy laminar fiow typically have a peak to valley distance of about 3 times the average film thickness, but occupy only a small fraction of the flowing volume. Waves carry momentum as they pass, but do not significantly affect the calculated average film thickness. Waves will wash through the region of flowing film, effectively wiping out any history effect of the method of application or other upstream effects. Therefore, film stability can be considered to be a local phenomenon, governed by local force balances at the point of interest on the contact line.
Equation 7.A-2 predicts higher values for the minimum stable film flow rate on surfaces that wet poorly, that is, those that have large contact angles, than for surfaces that wet readily. For surfaces that are heated, heat flux is destabilizing. The equation also shows that as the film heats up, it becomes more stable due to property changes. Since the theoretical models available in the literature are not practical for application to AP600, the insight gained from examining those approaches is used to support development of an empirical bulk coverage model. That is, the film stability can be characterized using a criterion for a minimum film flow rate, Emin, that will maintain a stable stripe. Data from tests at different scales, wherein the range of AP600 dimensionless parameters is sufficiently covered, can be used to empirically derive a bounding value for Fmin. As discussed in 7.A.2.1, data <.an be represented using the film flow rate, and plotted against the dominant independent parameter, heat flux. 7.A.3 CONTACT ANGLE AND SURFACE WETTABILITY A discussion of contact angles in general and observations from test coupons are provided to gain insight into the performance cf the coated surface relative to surfaces in the literature. Finally, factors which can affect surface wettability are discussed. 7.A.3.1 Advancing and Receding Contact Angles The place where the wet and dry regions intersect is called the contact line. For example, in a liquid film flowing down a wall in a constant width stripe, the contact lines are the two vertical O Basis and Method for Calculating the PCS Water Evaporation Rate October 1997 for the AP600 Containment DBA Evaluation Model Revtsion 1 l 0:\3006w-7b.nortib-101497
7.A 9 lines defining the width of the stripe. The contact angle is defined as the angle between the , solid and the liquid surface at the contact line. The contact angle between a water film and the surface to which it is applied is an indication of the surface wettability. Typically, better wetting occurs on surfaces with small contact angles, in practice, contact angles are measured for both advancing and receding films. Usually the two values are quite different, with the advancing contact angle being much larger than the receding contact angle. The relation between contact engle and velocity !s qualitatively depicted in Figure 7.A 2. Of interest is the hysteresis between advancing and receding contact angles. 'Ihere is actually a range of stable contact angles for a static contact line. Thus, if a droplet starts out by spreading, such as when it is dropped onto the surface, it will spread to a diameter govemed by the advancing contact angle. Then as the droplet evaporates, it may be expected to remain at a constant diameter, with the contact line anci'ored, until the mass lost contracts the droplet such that the receding contact angle is reached. Further evaporation would then cause the droplet diameter to decrease. It is general practice to measure contact angles of a liquid on a smooth or polished surface, such as glass or polished steel having surface profiles measured in microns. High magnification is used to meamre the contact angle as it meets the surface. The surface on the external containmmt shell is an inorganic zine coating applied on a carbon steel structure. The surface o! the inorganic rine coating is not smooth, having a surface profile of several mils. With a surface profile of several mils, the magnified image shows significant peaks and valleys, making it impossible to measure a single contact angle that is applicable over the entire surface. Thus, the significance of a representative contact angle for the organic zin: coating used for the exterior of the AP600 containment shell is diminished. The interest for AP600 application is on bulk coverage performance over a large surface area, so larger scale integral tests are used. It is desired, howev:r, to understand and relate the bulk wetting performance of the coated surface to thtt of surfaces in the literature. Therefore, measurements were taken to characterize a bulk static contact angle on the prototypic surface by observing a drop on sample coupons under various conditions as described below. 7.A.3.2 Static Contact Angle Measurements of Coated Surface The bulk contact angle for a drop of water was meacured as a function of temperature and age of the surface coating selected for the AP600. Two samples were prepared for these measurements. The first test coupon was supplied to Westinghouse by the coating vendor. This sample was prepared by the vendnr and was not subjected to weathering. The second sample was a 12 in2 section of a steel platc that was painted by Westinghouse and weathered for two years. O G' Basis and Method l'or Calculating the PCS Water Evaporation Rate october 1997 for the AP500 Containment DBA Evaluation Model Revision 1 a\XX4w ?b.rottitw101497
7.A 10 o Wetting Angle
/
4 / 8'\ ' Br Hysteresis j j
)O2 Receding contact ilne Awancing contact line
(-) Velocity (+) Contact Angle Behavior with Velocity 61 02 7 9 r1 r2 Static Co'nlact Angle Velocity m Receding Advancing Br Oa Dynamic Contact Angle Figure 7.A 2 Typical Qualitative Contact Angles for Advancing and Receding Contact Lines O Basis and Method for Calculating the IU Water Evaporation Kate ociater 1997 for the AP600 Containment DBA Evaluation Model Revision 1 c:\3006w-7b.non.lb 101497
l '/.A 11 l { The following procedure was used to determine the static contact angle for both samples at ! ambient conditions:
- The test coupons were cleaned per coating vendor specifications and dried.
- The test coupon was placed in a horizontal position. E
- A drop of water was placed on the test coupon.
Using an optical comparator, the average angle between the sample surface and the drop at the interface. Measurements were repeated using several drops to ensure repeatability and consistency in the measurements. Additional measurements were taken with the test coupons held at different temperatures. This was dane to evaluate the effect of the surface temperature on the cantact angle. The test coupons were heated with either hot water or a heat gun. The static contact angle measurements taken are summarized in Table 7.A 1. They show that 7 the contact angle for inorganic zine coated surface decreasr3 both with an increase in age and ( an increase in temperature. At high temperatures, the contact angle was observed to be initially larger than that observed for lower temperatures. It was observed, however, that the drops quickly spread and flattened out to a quasi-steady shape, thereby reducing the measured contact angle. From the measurements listed in Table 7.A 1, it is concluded that a representative bulk or average contact anele for the inorganic zine coated containment shell surface is between [
]*# for a new sur.dce, and between [ ]*# after just two years of weathering.
Basis and Method for Calculating the PCS Water Evaporation Rate ociater 1997 for the AP600 Containment DBA Evah ation Model Revision 1 o:\3006w 7b.rmrtit*101497
7.A i? -
~
l
- a,c 1 I
Table 7.A 1 Summary of Test Results to Determine Static Contact Angle l
=$.88 A small drop of water spread around on the inorganic zine-coated surface was not observed to contract, or snap back into a drop. This observation indicates that the receding contact angle for this surface is nearly zero. These observations also suggest that the film breakdown to form a dry spot occurs at a lower film Reynolds number than the critical Reynolds number for rewetting.
Static wetting angle measurements indicate that the coated surface is clearly more wettable than surfaces reported in the literature, and based on the force balance it is expected to be less sensitive to heat flux. 7.A.3.3 Relative Magnitude of Surface Tension Effects A solid surface will be wet with liquid if the free surface energy of the solid is greater than the free surface energy of the liquid. Surface tension, o,is defined as the work required to expand the surface of a liquid by a unit of area, it is a measure of the strength of the intermolecular forces in the fluid, similar to the latent heat of vaporization. liydrogen bonding is the strongest type of irstermolecular force. Liquid water has relatively strong intermolecular forces due to the strong hydrogen bonds; 80 percent of the intermolecular attraction in water is attributed to hydrogen bonding. In a water molecule, the electrons spend more time in the vicinity of the oxygen atom than the hydrogen atoms because oxygen is more electro-negative than hydrogen (3.5 versus 2.1 for hydrogen on a scale of 4.0). This results in an electric dipole within the molecule. For this reason water is said to le a polar molecule. Basis and Method for Calculating the PCS Water Evaporation Rate octoba 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o \3nw 7b non Ib 101497
7.A 13 As its temperature increases, the mean spacing between molecules in a liquid increases, causing the density to decrease and a reduction in the intermolecular forces. Therefore, both surface tension, o, and the latent heat of vaporization, his, decrease with increasing liquid temperatute. For example, the surface tension of water is about 4.97x10'3 lbf/ft and the latent heat of vaporization is about 1054 BTU /lbm at room temperature. The value of these two parameters decreases to 4.0x10~3 lbf/ft and 970 UTU/lbm, respectively, at 212'F. 7.A.3.4 Factors Affecting Surface Wettability The wetting of a solid surface by water is improved by reducing the surface tension of the water (by use of a wetting agent such as a detergent), by making the surface more porous (to improve the spreading by capillary action), or by using a polar surface (increasing the intermolecular forces between the surface and the polar liquid water). The use of a surfactant was ev.amined during the Water Distribution Tests. it was found that surfactants offered no effective improvement in coverage. This has been postulated to be due to the turbulence of the flowing film which would not allow the surfactant to influence the surface of the film significantly. The porosity of the inorganic zine coating is believed to be the primary factor affecting wetting early in the coating's life, adding a significant capillary effect at the contact line, it was postulated that the buildup of polar molecules (e.g., oxides of zine) on the solid surface improved its wettability with age. Photographs were taken of both new and weathered surface coating samples using a scanning electron microscope with an energy dispersive X-ray spectrometer to identify the chemical species present on the surface. More oxides of zine were fottnd on the weathered surface than the new surface, supporting the hypothesis that the increase in wetting is due to the surface becoming more polar as it ages. A buildup of some surface contaminants can result in a reduction in wettability. The worst surface contaminant for the inorganic zine coating is silicone; it has both low surface energy and low polarity. Sources of silicone in air pollution are rare. Other surface contaminants that could result in reduced wetting include hydrocarbons such as oils, members of the PTFE family (Te.'lon), polypropylene, and polyethylene residues. To combat surface contaminants, the coatings vendor has developed and made available a standard cleaning procedure and a specially developed detergent that emulsifies these types of surface contaminants so they can be washed away. Although the number of potential contaminants that would adversely affect wetting of the inorganic zine coating surface is probably limited to a dozen or so, it would be very difficult to analytically predict the wetting degradation over time. The degradation of surface wettability would have to be estimated as a function of the concentration of each potential contaminant, the deposition rate of each as a function of the local or worst case atmospheric conditions, and the assumption that the degradation is additive, etc. Therefore, periodic in service inspections will i v Basis and Method for Calculating the PCS Water Es aporation Rate ch:tober 1997 for the AP600 Containment DBA Evaluation Model Revision 1 oA3006w 7b nortitwl01497
7.A 14 be performed to look for corrosion and surface contaminant buildups to assure surface wettability. The frequency and procedures for testing and the minimum acceptance criteria prior j to cleaning the surface are defined in the AP600 Reliability Assurance Program. 7.A.3.5 Summary of Wetting Angle Assessment The contact angle between a water film and the surface to which it is applied is an indication of the surface wettability. Although the surface provided by the inorganic zine coating applied to the external surface of the AP600 containment is not smooth relative to other materials used to measure contact angles such as glass or polished steel, measurements were taken to characterize a bulk static contact angle of a spreading film on the prototypic surface to relate to literature cl ata. The static angle was measured by observing the spreading of a drop on two coupons, one weathered and one not weathered, under ambient and heated conditions. Results showed that a surface weathered for two years is significantly more wettable [ l'h than surfaces for which data exists in the literature (in the range of 60 degrees). 7.A.4 DESCRIfTION OF LST OBSERVATIONS LST observations to characterize wetting behavior were made during shakedown tests, video tapes were recorded, and sketches were made for the test records. During these shakedown tests, quasi-steady heat flux and water flow rate conditions were achieved, and then water flow was slowly valved down in stages with constant steam flow. At each stage, when quasi-steady conditions again were reached, observations and notes were taken. Subsequent similar cycles h were done at several steam flows (heat fluxes). The objective was to observe the behavior of the liquid filtn as it varied from a moderately high flow down to n stly complete evaporation. Since the majority of the LST matrix tests were rim with a high flow rate, the qualitative discussion starts with a description of water coverage on a high flow test. Finally, the water coverage on a low flow test is described. Observations are consistent with the physics of liquid films discussed above. 7.A.4.1 liigh Flow LST As discussed in Section 7.6.3, the water is applied to the external shell in stripes around the circumference at each of the weir elevations. Stripe widths for a given steady state test were relatively constant, varying by fractions of an inch as the delivered flow rate varied (see Section 7.6.3). Based on Reynolds number, the flow regime is wavy laminar, which has been confirmed by test observations. The wavy laminar regime is discussed in the literature. A simple sketch is provided in Figure 7.A-3, showing qualitative characteristics observed for a representative film stripe on a heated LST surface with a high flow rate. High flow rate LST typically exhibited' constant width stripes, as discussed further below. Stripe widths varied from O Basis and Method for Calculating the l'CS Water Evaporation Rate Octoter 1997 l for the AP600 Containment DBA Evaluation Model Revision 1 oA%0nw ?b non It*101497
7.A 15 l 4
- Wavy Laminar Laminar Wet Stripe Edge Film
=
Width - -
~1/8 - 3/4 inch
- p. . - ,
y ,,(
Y _) g!./ ;,j, l
, y ' 'y .[ Dry Surf J,99. '::p , . .. ; _ L,"Y :,;l % /.,2 .
Yuj{
- .. 4 igg 13,,
f.
.. ymp
- g; ..
Ay g , dL N '. . . >- y Alternating waves with a lateral velocity component Figure 7.A 3 Sketch of Qualitative Wavy Laminar Film Flow Characteristics on Heated LST Shell Water Stripe Basis and Method for Calculating the PCS Water Evaporation Rate octote 1997 for the AP600 Containment DBA Evaluation Model Revision 1 o:\300tiw 7b.non:11w101497 J
--- - . . - , , - , ,.. ..n..,_- -~
7.A 16 an inch or less to complete circumferential coverage, depending on the test delivered flow and heat flux. Within a stripe, the majority of the width is flowing water with wavy laminar conditions. In that portion, waves are generated by upstream disturbances and advance down g the stripe with a velocity faster than the average film speed, consistent with continuity flow theory. The waves generally alternate with slight left and right horizontal velocity components. Water stripe edges exhibited a narrow [ l** region of laminar flow. Visual observation indicated that the edges were wetted but not obviously flowing. When an obstruction was placed within a wet edge, a " bow wave" built up above the obstruction, , confirming that indeed liquid was flowing downward in that region. The film flow, and thus thickness, in that region is small enough that viscous forces damp out any disturbances. For example, the waves are damped by viscous forces in the stripe edge. Note that the laminar edge was a'.so observed to occur on stripes which narrowed as their film flow rates decreased dae to evaporation. This indicates that there is a very thin layer near a stripe edge, or in fact the equivalent wetting angle at the contact line is very small. This is consistent with the consideration in 7.A.3 that the receding wetting angle likely governs film stability of an evaporating stripe on AP600. Since the water is applied as stripes at the dome with J tubes (see Figure 7.A-4), and there is significant liquid film subcooling over much of the LST dome for high flow tests, the width of stripes that reach the vertical sidewall is less than can be supported by a stable film at the given film flow rate. Therefore, it can be postulated that the initial width at the top of the vertical sidewall is sufficiently greater than the evaporating film stability limit and that evaporation from h the stripes does not cause the rcceding contact angle to be reached. Rather the film stripes in high flow LST tests are believed to remain within the region of hysteresis over the entire height, consistent with the observed constant width stripes. 7.A.4.2 Low Flow LST Figure 7.A-5 shows a composite of typical film characteristics on a portion of the LST shell at relatively low flows typical of the water flow applied to LST 213.1. The tests described here have film flows that are low enough that evaporation causes the receding contact angle to be reached, and further evaporation leads to narrowing of the stripes As for the high flow LST, the water is delivered to the vessel shell surface via J tubes, as a subcooled film. The application method and subcooled film stability set the initial stripe width, similar to the high flow tests. However, the film heats up to become an evaporating film before it reaches the sidewall. Observations were made of shakedown tests at conditions (steam flow, external water flow) similar to those for LST 213.1. During the initial setup prior to heating the vessel, the film flow was established and gradually valved down. As very low flows were Basis and Method for Calculating ne PCS Water Evaporation Rate cktober 1997 9 for the AP600 Containment DBA Evaluation Model Revision 1 oM00twanorthwl01497
7.A 17 lO l ) Film is ov of
~
l Film width at top estabi shed ,, $e ! "'*1%; T'( i g}ii f, g . b. Nbh S p0S Vr lO 5
~ , L *i**# i. o h
l Hlfe ( t l .h,5 j Figure 7.A 4 Large Scale Test Water Coverage Pattern at liigh PCS Flows !O for e AP C nta Es lu on i 1 5 ion 1 j o:\3006w-7b.rmlb-101497
7.A 18 . reached, some J tubes were seen to stop delivering water before others, indicating that there was some asymmetry in delivered flow per stream. This is consistent with observations of heated tests that indicated stripe widths and vertical extents varied around the circumference of the g vessel. In Figure 7.A 5 the width of the two outer stripes shown remain approximately constant down to a certain elevation, varying only as the delivered flow rate varied. At some elevation on the sidewall, which may be different for different stripes, the film width began to narrow until the gutter was reached. For some stripes, shown as the two innermost stripes in Figure 7.A-5, the delivered flow was low enough that the stripes completely ew.porated before reaching the gutter elevation. The slope of the changing width as a function of height, dW/dZ, was carefully observed. Qualitative observation indicated that the dW/dZ of each stripe around the circumference was nearly constant at a specific quasi-steady-state test condition. Of most interest in these Mt relative to water coverage, is the fact that stripes that evaporated complet:.ly did so withou, changing their characteristics near the point of complete dryout. Thus, for the surface tested, the liquid films did not snap, or draw up, into a thick film. The edges of the film, including the bottom edge remained as wavy laminar film up to within a fraction of an inch from the edge,incitiding the lower edge. As the water flow rate was valved down, the bottom edge moved gradually up, and when the flow was increased to its original value, the vertical extent of the stripe returned to a consistent elevation. Therefore, the film was well behaved as it completely evaporated. 7.A 5 CONCLUSIONS A comparison of coated surface data to literature data for polished surfaces shows the coated surface is more wettable than surfaces reported in the literature. Literature models are not sufficiently developed to be considered reliable. The literature provides an indication of the appropriate parameters to study film breakdown data: Remm or l' and heat flux. A practical approach taken to bound the data from the various test 3 is to establish a minimum stable film flow rate, l', that can be used to define a minimum coverage. History effects are washed out by waves, so breakdown can be considered to be a local phenomenon. Therefore, LST (Section 7.6.3), SST (Section 7.6.2), and heated flat plate tests (Section 7.6.1) can be said to represent the bottom portions of liquid film stripes on AP600 that dry out due to evaporation. Basis and Method for Calculating the PCS Water Evaporation Rate O october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 oA3006w.7b nortitul01497
7.A 19 Stripe widths govemed by momentum spreadng at application point and umhed ed gy s g, -
- /
_g water appiled as stripes s - ( ,. ' 5 ',Nb^ f,,,-;. 5:-'
, ?'.
- -Relatively constant width k
k Y. . s /' I - Evaporating film [- more stable than F/ subcooled film
,0. -
I - '.> '. L m, .. W(Z)
,s- , .*
Q f\ /- : ,..- a , -:S n ':: /:-q , i f.
-:f ,
l 1 k :. - b..? 9' I
,' J ,J J ! \
some stripes narrow as they Films reach reach evaporating ,p, y film stabildylima (
\ Slope (dW/dZ) without changing approdmately charaderistiorn equal for narrowing stripes at fixed test conditions Figure 7.A 5 Sketch of LST Observation of Vessel Exterior at Water Flows Similar to LST 213,1 Showing Complete Dryout of Some Stripes 's Basis a'n d Method for Calculating the PCS Water Evaporation Rate octoter 1997 for the AP600 Containment DBA Evaluation Model Revision 1 oA3006w-7b.norultr101497
_. ~_ 7.A-20 Observations of tests are explained based on physics of liquid films on heated surfaces. At high enough applied flows, the applied stripe maintains constant width until the film stability limit is reached, governed by the receding contact angle, then the stripe begins to narrow consistent g with the minimum film flow rate required to maintain a stable film. Observations of tests show that cornplete dryout occurs while maintaining a stable stripe geometry, gradually decreasing in width until it disappears. 7.A.6 REFERENCES 7.A.1 WCAP-14326, Revision 1, " Experimental Basis for the AF600 Containment Vessel Heat and Mass Transfer Correlations," May 1997. 7.A.2 W. S. Norman and V. McIntyre, " Heat Transfer to a Liquid Film on a Vertical Surface," Trans. Inst. Chern. Engrs. Vol. 38, pp 301-307 (1960). 7.A 3 V. A. Hallett," Surface Phenomena Causing Breakdown of Falling Liquid Films During Heat Transfer," Internattonal Journal of Heat and Afass Transfer, Vol. 9, pp 283 294 (1966). 7.A-4 T. Fujita and T. Ueda," Heat Transfer to Falling Liquid Films and Film Breakdown Parts I and 11," International Journal of Heat and Mass Transfer, Vol. 21, pp 97-108 and 109118 (1978). h 7.A 5 M. S. Bohn and S. H. Davis,'Thermocapillary breakdown of Falling Liquid Films at High Reynolds Numbers," International lournal of Heat and Mass Transfer, Vol. 36, pp 1875-1881 (1993).
/.A-6 S. G. Bankoff, " Dynamics and Stability of Thin Heated Uc;.id Films," Transactions of the ASME - Journal of Heat Transfer, Vol.112., pp 538 546, (1990).
7 A-7 N. Zuber and F. W. Staub, " Stability of Dry Patches Forming in Liquid Films Flowing over Heated Surfaces," International Journal of Heat and h1 ass Transfc., Vol. 9, pp 897-905 (1966). 7.a 8 Dr. S. G. Bankoff, discussions held at Westinghouse Science and Technology Center during observations of LST,1996. O Basis and Method for Calculatir g the PCS Water Evaporation Rate october 1997 for the AP600 Containment DBA Evaluation Model Revision 1 0:\30th.7b.non.1t*101497
WESTINGilOUSE NON PROPRIETARY CLASS 3 O Section 8 l l AP600 Containment Pressure Sensitivity During Blowdown O O
~
o;\300bw-8.noitlbl01497
ill TABLE OF CONTENTS LIST OF FIG URES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ill
8.1 INTRODUCTION
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 8.2 M ETH OD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 8.3 AN A LYSIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 8,4 CON C LUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 LIST OF FIGURES Figure 8-1 Comparison of Single Node Model with EM Pressure Curve . . . . . . . . . . 83 Figure 8-2 Comparison of Response with No Heat Sinks to EM Response . . . . . . . . 8-4 O
r (
. s AP600 Containment Pressure Sensitivity During Blowdown October tw7 . o:\3006w4.ncatlb 101497 Reviskm 1
81 8.1 INTRODUC110N O '!he purpose of performing the single-node EGOTHIC analysis is to show that the containment pressure during the blowdown phase of the AP600 (predicted using the EGOTHIC code) is essentially the same as if Standard Review Plan (SRP) methodologies were utilized for the analysis. This comparison supports the use of EGOTHIC during the analysis of the blowdown phase of the transient, since it is expected that the presence of external heat removal from the containment shell during the first 50 seconds of the transient has little impact on the pressure-transient. The containment shell time constant is long, as compared to the transient time, and passive cooling system (PCS) film flow is assumed to be delayed until well after the end of blowdown.- ' The purpose of performing the sensitivity to heat sinks during blowdown is to confirm that volume compliance is the dominant means of mitigating pressure increase during blowdown. 8.2 METHOD The evaluation model (EM) described in Section 4 was used for comparison in'this study. The EM was converted to a single node containment model, consistent with SRP 6.2.1 methodology and comparable to the licensed Westinghouse methodology by the following input modifications:
*- All of the climes were removed.
- All of the flow paths, except for those associated with the mass and energy release forcing functions, were deleted. The mass and energy forcing functions were not changed.
- All control volumes which represent the outside containment regions were deleted.
- A single-node containment control volume, containing all of the thermal conductors from 1 the base case and the two mass and energy release forcing functions, was created.
- A conductor representing .the containment shell was added to the single node containment control volume.
*- The Uchida heat transfer correlation with revaporization was used on the shell and enductors.
O AP600 Containment Pressure Sensitivity During Blowdown ek1ober 1997 c:\3006w 8.non;1b.101497 RevWon 1
8-2 i The Ehi was modified to eliminate heat removal from the containment gas volume by internal l heat sinks and the steel shell. The only modification to the Ehi was to delete all thermal conductors within containment and to effectively eliminate the clime conducters for the shell g itself by assuming an adiabatic inner surface. 8.3 ANALYSIS The blowdown phase pressure results for the single-node analysis are compared to the Ehi containment pressure in Figure 81. The blowdown phase pressure response without heat sinks is ecmpared to the Ehi results in Figure 8 2.
8.4 CONCLUSION
S The conclusion of tl,o blowdown noding sensitivity is that the single-node model (utilizing SRP 6.2.1 methodologies) essentially provides the same results during the blowdown phase as the EM. The conclusion of the sensitivity to eliminating heat sinks during blowdown shows a relative pressure increase at the end of blowdown of only 3.6 psi relative to the Eht. This compares to the Ehi pressure increase of about 33 psi during the blowdown phase, which confirms the dominant pressure mitigation during blowdown is energy storage due to pressure increase of h the volume, or volume compliance. O AP600 Containment Pressure Sensitivity During Blowdown october 1997 c \ x n w-8.rmitit*101497 Revtskin 1 j
O oA \ 3 P 0 6 0 0 . b 0 w 8 C F ai 1I!B aE" g . o mn nt gi u S
- a 1 i i ,
t un e 0me l 1 8-1 4 n Wt -
'r P
e s s C o S u r 0 e m p , S a e r n s i s i o ( n t i v 6 y i t o f l sR d - S ) D i n u r g e J l g i n e n N s , Bl o s o w d e eR J d d , o M w o , d n e l i w i i t h S E . M P r e s s u r e e - _ : ~
~ 4 .
C v u r o m =
- E $ R e
5*- c I1 !1 y Ro e t vb ie s r k1 m99 8 1 7 84 _. n o
~
l ! e ! l l - 1
. - Nm } \
i: I. - lii 8 '.. v \ i -
'g@
i.
}. . - \ t i! .E w i. _mS !
v g ]'.
] G ~ '.'. i=
j S
# i 3 .\. o E
e - - m . s _m 4 M9 o o o o o o o its 4 M N e (61sd) oJnsseJd Figure 8-2 Comparison of Response with No lleat Sinks to EM Response i Al'600 Containment Pressure Sensitivity During Illowdown october 1997 o;\3006w-8.non.1t-101497 Rnision 1
WESTINGHOUSE NON-PROPRIETARY CLASS 3 O Section 9 Circulation and Stratification Within Containment O O c:\30%non\3006w-9.mwb '01497
i lil TABLE OF CONT 11NTS LIST O F TA B LES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv LIST O F FI G U R ES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V 9
SUMMARY
OF CIRCULATION AND STRATIFICATION EVALUATION . . . . . . . 91
9.1 INTRODUCTION
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-4 9.1.1 Definitions ...........................................914 9.1.2 Lumped Parameter Diases and Capabilities . . . . . . . . . . . . . . . . . . . 914 9.2 LARGE SCALE TEST RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 17 9.2.1 LOCA Configura tion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-18 9.2.2 MSLB Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 19 9.2.3 Method to Address Distortions in LST Stratification Data . . . . . . . . 9 20 9.2.4 Application of Modeling Methods Developed for NUPEC M-4-3 Lumped Parameter Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 21 9.3 CIRCULATION AND STRATIFICATION AS5J.SSMENT FOR THE LOSS-OF-COOLANT ACCIDENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-26 9.3.1 LOCA Dreak Scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-27 9.3.2 AP600 LOCA Evaluation Model . . . . . . . . . . . . . . . . . , , . . . . . . . . 9-34 9.4 MAIN STEAMLINE DREAK (MSLB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-44 9.4.1 B reak Loca tions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-44 9.4.2 AP600 MSLB Evaluation Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-45 9.4.3 MSLD Sensitivity Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-46 9.5 C ONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-4 8 9.6 REF E RENC ES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-50 APPENDIX 9A THERMAL AND CIRCULATION EFFECTS OF DROPS DURING A LOCA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 A-1 APPENDIX 1B EFFECTS OF STRATIFICATION ON HEAT SINK UTILIZATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9B-1 APPENDIX 9C ADDITIONAL INFORMATION ON AP600 CONTAINMENT CIRCULATION AND STRATIFICATION . . . . . . . . . . . . . . . . . . . 9C-1 >
Circulation and Stratification Within Containment october 1997 c.\30$non\3006w 9.non:1t*101497 Revision 1
iv 1 LIST OF TABI ES I Table 91 Circulation and Stratification Evaluation Summary . . . . . . . . . . . . . . . . . 95 O Table 9 2 Flow Areas Connecting to North and South CMT Compartments (excluding Dead Ended Compartment Connections) . . . . . . . . . . . . . . . . 9 24 l l O l O Circulation and Stratification Within Containment ociot,cr 1997 o: \30bnon \ 3006w-9.norv l ti-101497 Revnion 1
v i LIST OF FIGURES O Figure 91 Measured Steam Concentrations for LST . . . . . . . . . . . . . . . . . . . . . . . . . 9 52 Figure 9 2 LST with Diffuser Under Steam Generator - Steam Flow 0.110.17 lb/sec Intemal Fluid Temperature - Group 1 . . . . . . . . . . . . . 9 53 Figure 9-3 LST with Diffuser Under Steam Generator - Steam Flow 0.11 0.17 lb/sec - Saturation Temperature - Group 1 . . . . . . . . . . . . . . . . 9-54 Figure 9-4 LST with Diffuser Under Steam Generator - Steam Flow 0.11-0.17 lb/sec - Intemal Steam Pressure Ratio - Group 1. . . . . . . . . . . . 9-55 ) Figure 9 5 LST with Diffuser Under Steam Generator - Steam Flow 0.11 0.17 lb/sec - Intemal Fluid Temperature - Group 2 . . . . . . . . . . . . . 9-56 Figure 9-6 LST with Diffuser Under Steam Generator - Steam Flow 0.110.17 lb/sec - Saturation Temperature Group 2 . . . . . . . . . . . . . . . 9-57 Figure 9 7 LST with Diffuser Under Steam Generator - Steam Flow 0,11-0.17 lb/r.ec - Intemal Steam Pressure Ratio - Group 2 . . . . . . . . . . . . 9-58 Figure 9 8 LST with Diffuser Under Steam Generator - Steam Flow 0.27 0.36 lb/sec - Intemal Fluid Temperature . . . . . . . . . . . . . . . . . . . . . . 9-59 Figure 9-9 LST with Diffuser Under Steam Generator - Steam Flow 0.27 0.36 lb/sec - Saturation Temperature . . . . . . . . . . . . . . . . . . . . . . . . 9-60 Fir 9ee 910 LST with Diffuser Under Steam Generator - Steam Flow 0.27-0.36 lb/sec - Internal Steam Pressure Ratio . . . . . . . . . . . . . . . . . . . . 9-61 Figure 911 LST with Diffuser Under Steam Generator - Steam Flow 0.49-0.62 lb/sec - Intemal Fluid Temperature - Group 1 . . . . . . . . . . . . . 9-62 t'igure 912 LST with Diffuser Under Steam Generator - Steam Flow 0.49-0.26 lb/sec - Saturation Temperature - Group 1 . . . . . . . . . . . . . . . . 9-63 Figure 913 LST with Diffuser Under Steam Generator - Steam Flow 0.49 0.62 lb/sec - Intemal Steam Pressure Ratio - Group 1. . . . . . . . . . . . 9-64 Figure 914 LST with Diffuser Under Steam Generator - Steam Flow 0.49-0.62 lb/sec - Intemal Fluid Temperature - Group 2 . . . . . . . . . . . . . 9-65 Figure 9-15 LST with Diffuser Under Steam Generator - Steam Flow 0.49-0.62 lb/sec - Saturation Temperature - Group 2 . . . . . . . . . . . . . . . . 9-66 Figure 916 LST with Diffuser Under Steam Generater - Steam Flow 0.49-0.62 lb/sec - Intemal Steam Pressure Ratio - Group 2 . . . . . . . . . . . . 9-67 Circulation and Stratification Within Containment cetober 1997 cA3006 non\3006w-9.nortit*101497 Reviskm 1
vi LIST OF FIGURES (Cont.) Figure 9-17 LST with Diffuser Under Steam Generator - Steam Flow O 0.76-0.84 lb/sec - Intemal Fluid Temperature . . . . . . . . . . . . . . . . . . . . . 9-68 I Figure 9-18 LST with Diffuser Under Steam Generator - Steam Flow ! 0.76-0.84 lb/see - Saturation Temperature . . . . . . . . . . . . . . . . . . . . . . . . 9-69 l Figure 919 LST with Diffuser Under Steam Generator - Steam Flow 0.76-0.84 lb/sec - Internal Steam Pressure Ratio . . . . . . . . . . . . . . . . . . . . 9 70 l Figure 9-20 LST with Diffuser Under Steam Generator - Steam Flow 1.10-1.20 lb/sec - Internal Fluid Temperature . . . . . . . . . . . . . . . . . . . . . . 9 71 Figure 9-21 LST with Diffuser Under Steam Generator Steam Flow 1.10-1.20 lb/sec - Saturation Temperature . . . . . . . . . . . . . . . . . . . . . . . . 9-72 Figure 9 22 LST with Diffuser Under Steam Generator - Steam Flow 1.10-1.20 lb/sec - Intemal Steam Pressure Ratio . . . . . . . . . . . . . . . . . . . . 9-73 Figure 9-23 LST with Diffuser Under Steam Generator - Steam Flow 1.54 1.68 lb/sec - Internal Fluid Temperature . . . . . . . . . . . . . . . . . . . . . . 9-74 Figure 9-24 LST with Diffuser Under Steam Generator - Steam Flow 1.54-1.68 lb/sec - Saturation Temperature . . . . . . . . . . . . . . . . . . . . . . . . 9-75 Figure 9 25 LST with Diffuser Under Steam Generator - Steam Flow 1.54-1.68 lb/sec Internal Steam Pressure Ratio . . . . . . . . . . . . . . . . . . . . 9-76 Figure 9 26 LST with Diffuser Up 6 Feet - Steam Flow 0.76 & 1.68 lb/sec - Internal Fluid Temperature . . . . . . . . . . . . . . . . . . . . . . . . . 9-77 Figure 9-27 LST with Diffuser Up 6 Feet - Steam Flow 0.76 & 1,68 lb/sec - Saturation Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-78 Figure 9-28 LST with Diffuser Up 6 Feet Steam Flow 0.76 & 1.68 lb/sec - Internal Steam Pressure Ratio . . . . . . . . . . . . , . . . . . . . . . . 9-79 Figure 9 29 LST with Steam Injection: 3 Inch Pipe - Steam Flow 0.76 - 0.95 lb/sec - Internal Fluid Temperature .......... . . . . . . . . 9 80 Figure 9-30 LST with Steam injection: 3 Inch Pipe - Steam Flow 0.76 - 0.95 lb/sec - Saturation Temperature . . . . . . . . . . . . . . . . . . . . . . . 9-81 Figure 9-31 LST with Steam Injection: 3 Inch Pipe - Steam Flow 0.76 - 0.95 lb/sec - Internal Steam Pressure Ratio . . . . . . . . . . . . . . . . . . . 9 82 O Circulation and Stratification Within Contairunent october 1997 o;\3006-non\3006w-9.non:lb-101497 Revision 1
vil L LIST OF FIGURES (Cont.) 'O Figure 9-32 LST with Steam injection: 3 Inch Pipe Steam Flow 1.25 1.31 lb/sec - Internal Fluid Temperature . . . . . . . . . . . . . . . . . . . . 9-83 Figure 9 33 LST with Steam injection: 3 Inch Pipe - Steam Flow 1.25 1.31 lb/sec - Saturation Temperature . . . . . . . . . . . . . . . . . . . . . . . 9 84 Figure 9 34 LST with Steam injection: 3 Inch Pipe - Steam Flow 1.25 - 1.31 lb/sec - Internal Steam Pressure Ratio . . . . . . . . . . . . . . . . . . . 9-85 Figure 9 35 CMT Compartment Layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-86 Figure 9-36 Simplified AP600 Containment Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . 9 87
- Figure 9-37 = - HGOTHIC Calculated LOCA Blowdown Steam Pressure Ratio for Jet Momentum Dissipated in SG East Compartment . . . . . . . . . . . . . . . . . . . ._9-88 Figure 9 38 WGOTHIC Calculated AP600 Containment Pressure - Sensitivity to Loss Coefficients for LOCA Jet Momentum Dissipated in SG East Compa rtment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-89 Figure 9-39 EGOTHIC Calculated Flow Pattem - Sensitivity to Loss Coefficients for LOCA Jet Momentum Dissipated in SG East Com t tt O 20 Seconds - . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .par . . .men . . . .a. . . . . . . 9 90 Figure 940 ECOTHIC Calcu6al Flow Pattem - Sensitivity to Loss Coefficients for LOCA Jet Momentum Dissipated in SG East Comp. at 1000 Seconds . . . . 9 91 Figure 9 41 EGOTHIC Calculated Flow Pattem - Sensitivity to loss Coefficients for LOCA Jet Momentum Dissipated in SG East Comp, at 1550 Seconds . . . . 9 92 Figure 942 EGOTHIC Calculated Flow Pattern - Sensitivity to Loss Coefficients for LOCA Jet Momentum Dissipated in SG East Comp. at 80050 Seconds . . . 9 93 Figure 9-43 WGOTHIC Calculated AP600 Containment Heat Removal Rates - LOCA Jet Momentum Dissipated in SG East Compartment . . . . . . . . . . . . . . . 9-94 Figure 944 WGOTHIC Calculated AP600 Containment Steam Pressure Ratio for LOCA Jet Momentum Dissipated in SG East Compartment . . . . . . . . . . . 9-95 Figure 9-45 EGOTHIC Calculated AP600 Cont.= Fressure - Sensitivity to Heat Transfer Coefficient for Study of Undissipated Jet Effects During a LOCA . . . . . . 9-%
Figure 946_ EGOTHIC Calculated AP600 Containment Pressure-LOCA Jet Momentum Dissipated in SG East Compartment . . . . . . . . . . . . . . . . . . . 9-97 Circulation and Stratification Within Containment october 1997 c:\300& non\3006w-9.ncat l b-101497 Revision 1
vili LIST OF FIGURES (Cont.) O Figure 9-46A EGOTHIC Calculated AP600 Containment Below-Deck Compartment Pressure for LOCA Jet Momentum Dissipated in SG East Compartment . 9 98 Figure 9 47 EGOTHIC Calculated Flow Pattem - LOCA Jet Moinentum Dissipated la SG East Compartment at 20 Seconds . . . . . . . . . . . . . . . . . . . . . . . . . . 9-99 Figure 9-48 EGOTHIC Calculated Flow Pattem - LOCA Jet Momentum Dissipated in SG East Compartment at 1000 Seconds . . . . . . . . . . . . . . . . . . . . . . . 9100 Figure 9-49 EGOTHIC Calculated Flow Pattem - LOCA Jet Momentum Dissipated in SG East Compartment at 1500 Seconds . . . . . . . . . . . . . . . . . . . . . . . 9-101 Figure 9 50 EGOTHIC Calculated Flow Pattem - LOCA Jet Momentum Dissipated in SG East Compartment at 8000 Seconds . . . . . . . . . . . . . . . . . . . . . . 9 -102 Figure 9 51 Details of EGOTHIC Flow Paths to Above-Deck Region from CMT, Refueling Canal, and IRWST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 103 Figure 9 52 WGOTHIC Calculated AP600 Containment Pressure - LOCA Plume Rising into CMT Room . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 104 Figure 9-53 EGOTHIC Calculated Flow Pattem - LOCA Plume Rising into CMT Room at 1000 Seconds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 105 Figure 9 54 EGOTHIC Calculated Flow Pattem - LOCA Plume Rising into CMT Room at 1400 Seconds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-106 Figure 9-55 EGOTHIC Calculated AP600 Containment Heat Removal Rates - LOCA Plume Rising into CMT Room . . . . . . . . . . . . . . . . . . . . . . . . . . 9107 Figure 9-56 EGOTHIC Calculated AP600 Containment Pressure - LOCA Plume Rising into CMT Room and SG Compartments . . . . . . . . . . . . . . . . . . . 9-108 Figure 9-57 EGOTHIC Calculated Flow Pattem LOCA Plume Rising into CMT Room and SG Compartments at 1000 Seconds . . . . . . . . . . . . . . . 9109 Figure 9-58 WGOTHIC Calculated Flow Pattem - LOCA Plume Rising into CMT Room and SG Compartments at 1500 Seconds . . . . . . . . . . . . . . . 9-110 Figure 9-59 EGOTHIC Calculated AP600 Containment Heat Removal Rates - LOCA Plume Rising into CMT Room and SG Compartments . . . . . . . . 9-111 Figure 9-60 EGOTHIC Calculated AP600 Containment Steam Pressure Ratio for MSLB Above-Deck . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-112 O Circulation and Stratification Within Containment October 1997 o:\3006*n\3006w-9.non It>101697 Rmsson 1
I - ix LIST OF FIGURES (Cont.) Figure 941 EGOTHIC Calculated AP600 Containment Pressure - MSLB Above Operating Deck . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 113 Figure 9 62 WGOTHIC Calculated AP600 Containment Pressure - MSLB in CMT Room . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1 14 O Circulation and Stratification Within Containment October 1997 a\3006-non\3006w 9.nortib101697 Revision 1
9-1 ! 9
SUMMARY
OF CIRCULATION AND STRATIFICATION EVALUATION c Design basis accident (DBA) evaluations of AP600 containment pressurization transients follow l an approach that bounds uncertainties in parameters important for containment response. In this regard, the assessment of circulation and stratification inside the AP600 examined a range
- of possible break elevations, orientations, and momenttun to determine the worst case set of assumptions. A summary of the evaluation results and a cross reference to supporting subsections is given in Table 91.
The effect of break parameters on mass transfer to heat sinks, the dominant means of pressure mitigation, is e,aluated. The evaluation results in both the selection of a limiting scenario for large-scale circulation, and also a conservative handling of potential effects of stratification. The objective is to perform a bounding, or worst case analysis. The effects of circulation and stratification do not lend themselves readily to quantification of a bias and distribution for uncertainty, such as would be done for a best-estimate analysis. For example, it would be very difficult to quantify the probability of a break being directed in any particular direction. Rather the simplest DBA approach is to examine the range of possible AP600 break conditions, to select a limiting scenario, and use modeling techniques to bound the potential for reduced heat sink mass transfer, g For equipment qualification (SSAR Appendix 3D.5.5.1.5), the simple bounding approach is taken which uses the temperature in the break compartment as input to the qualification envelope. This temperature is the maximum value in containment. For containment pressure, the evaluation in this section has been performed, summarized as follows. The containment pressure transient is potentially affected by parameters which influence the dominant heat removal mechanism, mass transfer. Mass transfer has as its primary parameters steam concentration, and, in the case of forced convection conditions, velocity. Large-scale circulatior and entrainment into jets or plumes can drive circulation and can affect local values of steam concentration and velocity near heat transfer surfaces. Jet and plume entrainment within compartments or the above-deck region can also result in stratification, or the existence of a vertical steam concentration gradient. Therefore, an assessment of the effects of circulation and stratification should focus on how the steam concentration and velocity are affected. Since the Evaluation Model assumes only free convection inside the containment, the potential benefit of forced convection, when it exists, is neglected. Therefore, the assessment can be further focused on the potential effects on steam concentration distributions. For the main steamline break (MSLB), the containment vessel shell never becomes the dominant heat removal mechanism before break releases are over; therefore, known biases inherent in the lumped pcrameter Evaluation Model are used to minimize the internal heat sink p effectiveness. Lumped parameter model biases, supported with LST comparisons, are used to V Circulation and Stratification Within Containment cktoter 1997 oA3006-non\3006w-9.rmit>101497 Revision 1
9-2 impose a conservative break release boundary condition location in the Evaluation Model for MSLB pressure responses. For the loss of coolant accident double-ended cold leg guillotine break (LOC A DECLG), temporal partitioning has been used to further refine the evaluation fo- blowdown (0-30 sec.), refill (30 to 90 sec.), peak pressure (90 to 1200 sec.), and long-term (1200 sec. and beyond). During blowdown, volume pressurization is the dominant energy absorber, so the details of mixing and stratification effects are not dominant. During the long-tern the passive containment cooling system (PCS) is the dominant heat removal mechanism, so mat increasing the concentration of noncondensables in the above-deck region would reduce the PCS heat removal capability and result in higher calculated containment pressures. The peak pressure period, where both the below-deck heat sinks and the PCS surface are significant contributors, has been assessed by examining extreme release scenarios and examining the range of AP600 conditions to select a limiting scenario for peak pressure. The evaluation includes a logical sorting and organization of extreme break scenarios that are quantified by various analytical models and selected experimental results. The analytical models include hand calculations and the use of the EGOTHIC Evaluation Model for sensitivities to the range of the extreme break scenarios considered. Entrainment into a jet or plume and large-scale, dersity-driven circulation between compartments can force some degree of homogenization between and within compartments. Entrainment into a jet or plume can reduce the vertical density gradient occurring due to stratification because of the induced circulation. The assessment of large-scale circulation and compartment density gradients is summarized below. - Large-scale circulation is evaluated by examining a range of extreme release scenarios, including break location, elevation, orientation, and momentum. A limiting, large-scale circulation scenario for the peak pressure period can be shown to result from the assumption of dissipation of the break momentum within the steam generator compartment, at the elevation of the primary system pipe. The scenario is limiting because other scenarios were shown to have improved heat sink utilization, and thus lower peak pressures. For example, the extreme postulated scenario of an undissipated forced jet exiting the upper steam generator ccenpartment opening would drive significantly more convection on the steel shell (PCS) surface, and data indicates that the kinetic energy exiting the steam generator compartment would drive circulation below deck. Mass transfer would be greater than that for a buoyant plume. For a buoyant sc,urce and a break low in containment, it is reasonable to use a lumped parameter formulation to model the large-scale, or intercompartment circulation. A review of possible release locations and the expected circulation patterns led to the selection of four potentially limiting cases for further evaluation. The lumped parameter EGOTHIC Evalut tion Model was then used to examine those potentially limiting buoyant source release locations. Results from the sensitivity cases were consistent with < the expected circulation patterns in each case, which supports the use of the EGOTHIC lumped parameter model for those sensitivities. Results also showed that the postulated scenarios Circulation and Stratification Within Containment cktober 1997 o \3006 non\3006w.9.non:1 tut 01497 Reymon 1
9-3 examined a wide sange of posalble transient evolutions of steam concentrations throughout the dominant circulating compartments. An assumption of a buoyant release within the broken steam generatcr compartment reduced the steam access to a large fraction of heat sinks compared to the other locations for a buoyant release, which reduced below-deck heat sink effectiveness and led to the maximum calculated containment pressure. The use of lumped parameter models can introduce a bias in heat and mass transfer calculations when details within a compartment or region may be important. The simplified momentum formulation can lead to overmixing when multiple lumped parameter nodes are used to represent a single region, such as is done for the above-deck region in the Evaluation Model. Thus, density gradients larger than those predicted by the model in the above-deck region are expected and are assessed independently from the Evaluation Model. The calculation uses a single calculational node to represent each below deck compartment. The single node s representing each compartment allows only an average value of steam concentration for that compartment. For both above- and below-deck regions, density gradients larger than those predicted by the Evaluation Model are evaluated to gain insight into the effects of extreme gradients on heat sink utilization. Showing how sensitive the heat sink utilization is to extreme gradients provides greater confidence that the simplifications inherent in the Evaluation Model have been conservatively bounded. Since stratification within compartments is not considered explicitly in the WGOTHIC lumped
-g parameter model, it has been evaluated for its effect on total compartment heat sink utilization.
The potential for degraded heat sink effectiveness has been examined using a simple calculation for the vertical heat sink disn ibution and an extreme vertical density gradient. Results show that the total heat sink effectiveness within a compartment or region is affected by the assumed vertical gradients. Evaluations al o showed that mass transfer to upward facing surfaces in circulating compartments may be degraded very early in the transient, and heat sink effectiveness within dead-ended compartments may be overestimated by the lumped parameter model after blowdown Biases have been introduced into the Evaluation Model to bound these effects. The conclusion of the circulation and stratification assessmer. provides specific guidelines for the Evaluation Model ta bound the effects. The guidelines are summarized in Table 9-1, noted in the conclusions in Section 9.5, and are implemented as noted in Section 4 in tl e Special Modeling Assumptions subsection for each compartment or region. O Circulation and Stratification Within Containment october 1997 c:\3006-non\3006w.9.non:1b-101497 Revision 1
9-4
9.1 INTRODUCTION
The rupture of the primary system or main steamline piping has the potential to release a significant amount of mass and energy into the containment atmosphere. The AP600 is designed to withstand a loss-of-coolant accident (LOCA) or a main steamline break (MSLB) through a combination of a high containment design pressure and passive heat removal mechanisms. The passive heat removal systems include energy absorption by internal heat sinks as well as heat removal by the AP600 passive containment cooling system (PCS). A containment analysis is performed to verify the adequacy of the containment heat removal mechanisms to maintain post-accident containment pressure below the design limit. In this regard, the WGOTHIC code (Reference 9.1) has been developed as the containment code for performinb he design basis containment analysis. An appropriate Evaluation Model (Section 4) has been created which considers important input parameters such as mass and energy releases, containment volume, internal heat sinks, and PCS heat removal to calculate post-LOCA and post-MSLB containment pressure and temperature response. To obtain a conservative containment analysis, the effects of circulation and stratifiution must be bounded by the Evaluation Model. Circulation and stratification are natural processes that occur in the AP600 during postulated containment pressurization transients and have been identified as important phenomena to be addressed in support of the Evaluation Model for containment pressure calculations (Reference 9.2). The circulation and stratification that occur in the AP600 during a high energy p?e break transient, have the potential to reduce heat and mass transfer rates by transporting and concentrating noncondensables. The degradation of heat and mass transfer may reduce the effectiveness of the AP600 heat sinks and the PCS at mitigating the peak containment pressure. The eifects of circulation and stratification rnust be addressed to justify the approach used in the contairurent Evaluation Model. This section presents an overview of the effects of circulation and stratification for the AP600 containment Evaluation Model for LOCAs and MSLBs. The evaluation results are summarized in Table 9-1. As the table shows, the LOCAs and MSLUs are evaluated separately. The LOCA event is divided into four temporal phases based on heat sink utilization: the blowdown phase, the refill phase, the peak pressure phase, and the long-term phase. During each of these phases, important phencmena, such as mass and energy release rates, break source direction, and heat removal mechanisras are considered for impact on circulation and stratification. Unlike the LOCA events, the MSLB cases are not divided into temporal phases. The MSLB is characterized by a single, high-intensity blowdown phase. However, different piping rupture locations are considered in the MSLB evaluation. O Circulation and Stratification Within Containment october 1997 o:\3006-non\3006w-9.non.lb-101497 Revision 1
9-5 ; B Table 91 Circulation and Stratification Evaluation Summary { Element Summary of Eysluation Relevant WCAP. PIRT 14407 Parameter
- fection Reference l
l General Approach i l Circulation and stratification evaluated because of Circulation / the potential to degrade heat sink effectiveness via stratification 9.0 the condensation parameters: (2A),
. Steam concent: An condensation
- Velocity (3F, 7C)
High kinetic energy sources, such as during Circulation 9.0 LOCA blowdown and MSLB result in forced (2A), convection component of mass transfer condensation (3F) Effects of v : sty eliminated in calculation by Circulation / 9.0 assuming osay free convection internally. Focus, stratification therefore, is on impact of circulation and (2A) N stratification on steam and noncondensible distributions Equipment qualification temperature is Circulation / 9.0 conservatively taken from the break compartment stratification (containment pressure is therefore, the focus of (2A) the evaluation in Section 9) For the DBA LOCA, volume compliance is the Gas 9.0, primary pressure mitigator during blowdown, compliance 9.3, intemal heat sinks and the containment steel shell (2C), 9.3.2.1, are the primary mitigators during the peak condensation 9.3.2.4 pressure phase, and the steel shell surface is the (3F, 7C) dominant mitigator during the long-tenn phase For the DBA MSLB, the internal heat sinks are the Condensation 9.0, 9.4.3 dominant pressure mitigators. (3F) Circulation and Stratification Within Containment october 1997 o:\3006-non\3006w-9.non:lb-101497 Revissa 1
9-6 Table 91 (cont.) Circulation and Stratification Evaluation Summary Test Data A. Method to Power to volume ratio: using only quasi steady. Stratification 9.2.3 Address state data for circulation and stratification, (2A), Distortions in the therefore, no impact of this distortion on these int. heat sink LST for results, conduction Circulation and (3D), shell Stratification conduction g- Assessment Power to area ratio:
'7F)
Circulation / 9.2.3, I A.1
+ Steam flow was ranged and external boundary Stratification I_ conditions were ranged (2A)
- Considering the matrix of the LST, a range of
] power-to-area (or condensation rate) ratios were considered, which nummizes the degree of the distortion ] L_J
= Distortion addressed by considering stratification and condensation data from LST matrix tests and supplementing LST with assessment of international test data for stratification Circulation path impact on circidation : cannot use Circulation 9.2.3 the LST data for assessment of circulation. (2A),
Addressed by supplementing LST with assessment interemprt of international test data for circulation flow (2B) Circulation path impact on stratification : Stratification 9.2.3
- Lack of LST SG compartment circulation results (2A),
in LST stratification more extreme than if a interemprt circulation path existed flow (2B)
- Addressed by supplementing LST with assessment of using international test data for stratification O
Circulation and Stratification Within Containment octoter 1997 o:\3006-non\3006w-9.non:ltr101497 Revision 1
9-7 l Table 91 (cont.) Circulation and Stratification Evaluation Summary Q v B. Usage of LST IST above-deck separate effects style data for Mass and 9.2.1,9.2.2 Data for condensation and stratification is considered energy (I A), Circulation and direction and Stratification elevation (IB), Assessment momentum (IC), Circulation / stratification (2A) LOCA - applicable tests had diffuser under the SG, Stratification 9.2.1 reference case had elevated diffuser. (2A) Key LST result - above-deck stratification data used to support development of a bounding stratification gradient for evaluation of heat sita utilization for peak pressure /long-term phases MSLB - applicable tests had elevated 3" pipe Circulation / 9.2.2 pointing vertically / horizontally. stratification (2A) P Key LST results - kinetic energy drives circulation below-deck, forced convection significantly enhances mass transfer (factor of 1 to 10 over shell surface relative to free convection mass transfer) Lumped Parameter Biases implemented in lyGOTHIC Evaluation Model A. International / Lumped-parameter modeling uses a simplified Circulation / 9.1.2 Industry momentum formulation, which biases calculated stratification IVrience pressure with respect to circulation and (2A) stratification. These biases are evaluated and bounded by the Evaluation Model. NUPEC modeling experience is applied to Circulation 9.2.4, Evaluation Model compartment flow connections, (2A) Inter- Appendix C resulting in reasonably predicted circulation compartment fection patterns. Flow (2B) 9.C.3 f% Circulation and Stratification Within Containment october 1997 o \300ernon\3006w-9.non:1b-101497 Revmon 1 l
9-8 Table 91 (cont.) Circulation and Stratification Evaluation Summary B.1.OCA Biases The effects of stratification on heat sink utilization Stratification 9.3.1.1 4 are negligible for compartments experiencing (2A) j downflow of heavier ambient atmosphere mixture I
- Dead-ended compartments with no assumed Stratification 9.3.2.1 ;
thermal gradients stratify (2A)
- Condensation and convective heat transfer l tumed off in dead-ended compartments after j 30 seconds
- Effect of stratification on steel shell Stratification 9.3.1.1 ,
condensation assessed with extreme gradient (2A) [
- Stratification effect bounded by removing upward facing surface of operating deck as a heat sink
- Effect of stratification on heat sinks in a below. Stratification 9.3.1.3 deck compartment assessed with extreme gradient (2A)
- CMT room (most heat sinks) evaluated for case in which LOCA plume is rising in room
. Stratification effect bounded by removing floor as heat sink (bias applied in all compartments regardless of assumed break location)
C. MSLB Biases LST data indicates: Circulation / 9.4.2
- Kinetic energy drives some circulation below- stratification deck (2A)
- Forced convection is driven by high kinetic energy jet above-deck
- No significant stratification above-deck, therefore no bias required O
Circulation and Stratification Within Containment october 1997 o:\3006-non\3006w-9.non:1b-101497 Revision 1
9-9 Table 91 (cont.) Cheulation and Stratification Evaluation Summary Lumped parameter model code biases: Circulation / 9.4.2
. Evaluation Model places break node at stratification operating deck level mimmizing circulation and (2A) steam access to below-deck heat sinks
- Momentum dissipated in each node (Evaluation Model uses only free convection)
Density-driven circulation as plume rises resulting in relatively homogeneous region above modeled break node results in steam rich region above modeled break node and steam-deficient region below modeled break node, which bounds effects of stratification conservatively, the LOCA stratification biases are included for the MSLB Evaluation Model LOCA Evaluation Results A. Considerations LOCA blowdown (0 to 30 seconds): Intercompart 9.3.2, ( by Time Phase for
- Blowdown pressurizes compartments and ment Flow 9.3.2.1 s Evaluation Model drives significant circulation above and below-deck (28)
- Lumped parameter modeling adequate for pressure-driven flow Gas
- Containment pressure insensitive to noding compliance (multi-node vs. one-node model) (2C)
- Low sensitivity to heat sinks because volume storage is dominant pressure mitigator Break source
- Fr indicates significant forced convection on momentum 9.2.2 steel shell, Evaluation Model conservatively (1C) assumes only free convection
- Steam driven into dead-ended compartments.
Assuming thermally uniform heat sinks results in no circulation, therefore, condensation and convection heat transfer ia dead-ended compartments neglected after 30 seconds. LOCA refill (30 to 90 seconds): Break source 9.3.2.2
- Break releases are negtgible mass and
- Containment depressurizes during this phase energy (1A)
- Conservatively ignore containment pressure reduction by neglecting this phase to maximize g initial pressure for the peak pressure phase s
Circulation and Stratification Within Containment October 1997 o:\30thnon\3006w 9.non:lb-101497 Revision 1
9-10 Table 91 (cont.) Circulation and Stratification Evaluation Summary LOCA peak pressure (90 to 1200 seconds): Break source 9.3.2.3
- Steam source location changes to ADS Stage 4 (IB, IC) valves in both SG compartments at approximately
[ 1000 seconds
- Condensation on steel shell becomes dominant heat removal mechanism towards end of peak pressure phase
- Compartment filling reduces heat transfer for affected compartments during peak pressure phase and long-term phase (compartment filling is modeled by code)
LOCA long-term (1200 seconds to 24 hours): Condensation 9.3.2.4
- Condensation on steel shell remains dominant (7C) heat removal mechanism
* }VCOTHIC predicted steam gradient becomes Break pool essentially homogeneous in less than 24 hour , filling (SF) excluding the SG compartments (due to ADS Stage 4 valves releasing steam) Stratification
- Evaluation using extreme stratification gradient (2A) shows nearly negligible increase in heat removal by the iteel shell relative to homogeneous steam concen'. ration case - bounded by removing the non-grating operating deck floors
- Evaluation using extreme stratification gradient shows a decrease in heat removal by the below-deck compartment heat sinks relative to the homogeneous steam ccacentration case - bounded 9.3.1.3 by removing the compartment floor.
B. Range of Break Jet dissipated in SG East compartment Break source 9.3.1.1, Scenarios and
- Limiting scenario (IB, IC) 9.3.2.S Effects
- Post-blowdown flow into CMT room is downward with steam / air mixture Undissipated jet in SG East compartment Break source 9.3.1.2
- Forced convection above-deck improves (IB, IC) condensation on containment shell
- Significant kinetic energy-driven circulation below-deck
- Minimal stratification in above-deck region O
Circulation and Stratification Within Containment october 1997 o:\3006-non\3006w-9.non lb-101497 Revision 1
9-11 1
/N Table 91 (cont.) Circulation and Stratification Evaluation Summary b memummmmmmusummmum ummmmmmmmmmm l Jet to RCDT cavity - dissipated plume rises in Break source 9.3.1.3 CMT North room (1B,1C)
- Good steam access to below-deck room with most internal heat sinks Jet to RCDT cavity - dissipated plume rises in SG Break source 9.3.1.3 West compartment (IB, IC)
- Same scenario as dissipated jet rising in SG East compartment Jet dissipates in RCDT cavity Break source 9.3.1.3
- Flow split based on fMw area and loss (IB, IC) coefficients
- Better steam access to CMT room and SG West compartment compared to break in SG East compartment C. Sensitivity Break locations (all located low in containment): Intercompart 9.3.2.5 Cases Run with
- Jet undissipa'ed in SG East compartment - ment Flow the Evaluation forced convection benefit on steel shell assessed to (2B)
Model estimate effect of undissipated jet
- Jet dissipated in SG East compartment - limiting case for maximum containment pressure
- Jet into RCDT cavity - plume rises in CMT North room
- Jet dissipated in RCDT cavity - plume rise determined by flow path resistances Loss coefficients: Intercompart 9.3.2.1
- Loss coefficients for several flow paths changed ment Flow to modify blowdown-predicted flow direction (2B)
- Modeled dissipated jet in SG East compartment End of blowdown conditions changed with negligible change to maximum containment pressure
>O \j Circulation and Stratification Within Containment October 1997 c:\3006-non\3006w-9.non:1b-101497 Revision 1 n
9-12 Table 91 (cont.) Circulation and Stratification Evaluation Summary Thermal and circul' tion effects of drops: Break source 9.2.3.6
- Drops only created during LOCA blowdown droplet / liquid
- Thermal effects flashing (IE) 5.8
- 5 percent drop formation enough to saturate containment atmosphere Strc.ufication - O percent drops less limiting for maximum (2A) containment pressure - Negligible change in containment pressure Intercompart between 100 percent drops and Evaluation ment Flow Model(approximately 50 percent drops) (2B)
- Circulation effects examined for 0 and 100 9.2.3.6 percent drop formation Containment
- - Presence of drops increases density of vohune fog atmosphere increasing relative buoyancy of (2D) plume
- - Containment atmosphere entrainment into plume is significant for both 0 and 100 percent cases D. Conclusions
- Evaluation Model with dissipa < break in SG 9.2.3.5,9.5 East compartment is the limiting scenario
- Calculated containment pressure is not very sensitive to break location due to heat sink utilization prior to maximum pressure
- Biases included in Evaluation Model to bound effects of stratification MSLB Ev-laation Results A. Break
- Selected based on routing of steamline pipe 9.4,9.4.1, Location
- MSLB above-deck 9.4.1.1, 9.4.2 Scenarios and - High kinetic energy release with relatively Stratification Effects short duration, which drives circulation below (2A) the source
- High Fr number (comparison provided to LST Intercompart Fr number) ment Flow 2B) - LST data indicates forced mnvection Break source enhancement to mass transfer (only free (1B, IC)
- convection modeled)
- Break in MSLB Evaluation Model located in node just above-deck, which limits steam access to below-deck heat sinks O
Circulation and Stratification Within Containment Otteber 1997 o \3006-non\3006w-9.non:lb-101497 Revtsion I
9-13 ('~ Table 91 (cont.) Circulation and Stratification Evaluation Summary
- MSLB in CMT North room 9.41.2
- Break in CMT room would significantly dissipate due to equipment in room and rise as a plume - CMT room contains most of the internal heat sinks - Good steam access to CMT room heat sinks, therefore case is expected to be less limiting II. Sensitivity
- MSLB located just above deck 9.4.3 Cases
- MSLB in CMT North room C. Conclusions
- MSLB in CMT North room calculated 9.4.3,9.S containment pressure significantly less limiting
- MSLB located just above-deck used for the MSLB Evaluation Model (1) PIRT parameters are identified in Reference 9.2, Table 4-1 O
C\ V Circulation and Stratification Within Cc,ntainment october 1997 o:\3006-non\3006w-9.non:lb-101197 Revu'on 1
9-14 9.1.1 Definitions Several terms used to discuss circulation and stratification are defined, as they relate to containment analysis. Stratification is a state characterized by strata, or horizontal layers, of different density. Stratification is stable when the lower layers are increasingly dense due to composition and/or temperature. The term stratification does not indicate the magnitude of the density gradient. Mixing is a collective term for convective transport processes that reduce temperature and/or concentration diherences within a volume or between volumes. Convective transport processes in containment include jets, plumes, wall layers, turbulent diffusion, and entrained flow. Molecular diffusion also contributes to mixing but is considerably less effective than convection, except in boundary layers. Diffucion also centributes to mixing in stratified conditions. Circulation is a term used to describe gross, overall convective flow patterns that occur on a compartment scale and on a large scale (or containment scale). The compartment-scale circulation is due to wall layers, jets, plumes, and entrained flow. The large-scale circulation is due to interactions between compartments induced by pressure, density, elevation, and momentum differences such as intercompartment flow. The break source jet or plume can induce both compartment-scale and large-scale circulation. Segregation is a sta.te characterized by a different air / steam concentration in one compartment O than in another. For example, the heavier air may reach different concentrations in separate compartments, especially the dead-ended compartments if the intercompartment circulation is low. 9.1.2 Lumped Parameter Bia'.es and Capabilities Lumped paramebr biases and capabilities have been identified based on industry experience, as documented in the literature (Appendix 9.C, Section 9.C.3.4). The documented experience base includes facilities at different geometric scales, from that of the LST to nearly full-scale AP600 height (Appendix 9.C, Figure 9.C.2-1). The lumped parameter biases and capabilities, summarized below, have been reported consistently across the range of facilities, indicating that the biases and capabilities are applicable to the AP600 Evaluation Model. The consistency across scales also indicates that the LST facility is a reasonable basis on which to study the biases and capabilities as they apply to AP600, reported in WCAP-14382 (Reference 9.1). The following provides a summary of the method used in the development of the AP600 Evaluation Model to address each documented bias and capability.
- 1. Single node models were not capable of modeling stratification, or the passing of a stratification front through horizontal vents.
Circulation and Stratification Within Containment october 1997 I o:\3006-non\3006w-9.non lb-10M97 Revision 1 l
9-15 (A,C) -
- 2. Sump liquid level and sump temperature were not well predicted (a,c)
O-
- 3. Some codes produced results which were not correct due to missing or oversimplifying buoyancy terms (a,c)
Circulation and Stratification Within Containment October 1997 o:\30(Wnon\3006w-9.non:1b-101497 Revision 1 j
9 16
- 4. To account for recirculation flows, the applied lumped parameter model used double junctions in the horizontal direction. (This did not help in the case of an elevated release and resulting stratified containment.)
l See discussion for item 3 above regarding the impact of lumped volume static pressure profile on the use of double junctions in the Evaluation Model. The lumped parameter model is used for AP600 LOCA cases of low releases only. For the main steamline breaks, releases contain high kinetic energy. Therefore, the break node used in the lumped parameter model is a node that minimizes kinetic energy driven circulation to below-deck heat sinks, thus overestimating calculated containment pressure.
- 5. For releases low in containment, typical for the AP600 LOCA DECLG, the lumped parameter model well predicted pressure, temperature, and helium concentrations inside the compartments, which were affected by the global circulation loop, while predictions needed improvements to account for postulated circulation effects inside dead-ended compartments (a,c)
O
- 6. Scenarios with h<amogeneous containment atmosphere (like HDR E11.4 and E11.5) can be simulated successfully with lumped parameter models. (Such conditions typically result from breaks located within the bottom 20 percent of the containment height.)
See discussion for item 5 regarding the use of lumped parameter models for bounding design basis analyses.
- 7. Circulation effects due to sump boiling (releases generated at the bottom of containment) were well-simulated.
Sump boiling is not a consideratL.n for containment DBA, since long-term primary system energy rejection is through the ADS Stage 4 valves and the sump is therefore a relatively insignificant heat source. O Circulation and Stratification Within Containment October 1997 o:\3006-non\3006w-9.non:1b-101497 Revision 1
9 17
- 8. The order of magnitude of computed velocities matches data and it can be concluded that trends in the direction of the flow are predicted well; however, predicted velocities differ by as much as a factor of two.
Calculated velocities using lumped parameter codes are strongly dependent on the noding used. Experience with validating the W. GOTHIC lumped parameter model of the LST (Reference 9.1, Section 8.2) shows that the noding used can result in calculated velocities that differ from measured by an order of magnitude, showing that the particular test facility and noding used can have a strong influence on calculated velocities. Therefore, a bounding approach is used in the WGOTHIC Evaluation Model, as follows. The effects of predicted velocities in the AP600 pressure transient are eliminated by considering only free convection heat and mus transfer in the containment. This conservatively biases the Evaluation Model when forced convection would occur during the LOCA blowdown and the MSLB transients.
- 9. The lumped parameter method does not have the capability to predict the hydrogen distribution in a stratified containment atmosphere, as in HDR E11.2 with high-positioned release, in a break scenario with buoyant plume (released at about 50 percent of containment height), the steam and gas transport to the lower parts of the containment were over-predicted. (Artificial limitation of convective flows by decreasing flow areas improved predicted concentrations in the lower regions, but overestimated the containment pressure in upper compartments.)
O Hydrogen distribution predictions are not a consideration for contahunent DBA (Reference 9.2, Section 4.4.2E).- 9.2 ' LARGE SCALE TEST RESUL'IS In the AP600, interest is focused on how much the jet kinetic energy affects gradients inside containment. If the jet kinetic energy is sufficient to disrupt stable stratification, it may also be sufficiently energetic to virtually eliminate vertical gradients in the upper containment volume and to induce circulation between the above-deck and below-oeck regions. The Westinghouse AP600 Large-Scale Test (LST) data was tsed to understand the effect of jet kinetic' energy'on stratification gradients above the operating deck. The Westinghouse large-scale PCS test facility was built to provide integral test data for a geometrically similar model of the AP600 containment vessel and PCS. The tests provide experimental data that can be used for evaluating the physics in containment, determining the relative importance of various parameters that affect heat and mass transfer, and validating computer cod.s. Three series of tests (References 9.5 and 9.6) were run at the Westinghouse large-scale PCS test facility. The steady-state pressure, annulus air flow rate, water coverage, steam flow rate, injection velocity, location and orientation, and noncondensible gas O concentration were varied between the tests. Circulation and Stratification Within Containment october 1997 o:\3006-non\3006w-9.non:1b-101497 Revision 1 ,
9 18 it is desirable to use a Froude number fonnulation that relates momentum phenomena in both the AP600 and the LST to penrdi scaled inferences between the tests and the AP600. A volumetric Froude number can be defined as the square of the jet Reynolds number, divided by the containment Grashof number: P.U?d! Fr' = g(p. - P.)W where p, = density of ambient containment U, = velocity of jet at source d, = hydraulic diameter of jet at source g = gravitational cceeleration p, = density of jet source H = height of volume above steam source The following sections first describe test configurations as they represent LOCA and MSLB configurations and then provide data that can be used to examine gradients in the above-deck region. 9.2.1 LOCA Configuration Twenty-five LSTs were conducted in the LOCA configuration with the diffuser located under the steam generator model. A diffuser was used to provide a uniform s elocity profile. The tests do not apply to the LOCA blowdown phase, but they do apply to the peak pressure and long-term phases. The volumetric Froude numbers ranged from approximately 5x104 to 5x10~3 Steam concentrations just above the deck and below the deck near the bottom of the vessel are presented in Figure 9-1, which can be used to see test-to-test variation in above-deck gradients. The plotted values are the ratios of the measured local steam partial pressure to the partial pressure of steam assuming perfect mixing. A value of 1.0 indicates perfect mixing. The values show the above-deck ratios generally range from 0.6 to 1.0 and below-deck values range from 0.1 to 0.4. The below-deck values are an indication of the distortion in the LST due to lack of a simulated steam generator compartment flow path. The distortion leads to an air-rich mixture in the LST below-deck. Stratification data for LSTs with the diffuser under the simulated steam generator compartment are shown in Figures 9-2 through 9-25. Tests have been grouped by steam flow and plotted so that the temperature axis spans the same range for all the tests to simplify test-to-test comparison. For each group of tests, three plots are shown. First is the azimuthally-averaged l Circulation and Stratification Within Containment october 1997 tr\3006 non\3006w4.non:1b-101497 Revision 1
9-19 temperature data from thermocouples located one inch inside the vessel shell, called the " fluid thermocouples." Data is available fron, cine elevations above the operating deck; fluid thermocouple data was not taken below the deck. Second is a plot of the saturation temperature = obtained based on the third plot of measured steam mole fractions, or pressure ratio (p,./P% ). Also, a reference test to examine the physics of stratification (test 222.2), with an elevated diffuser, is included as Figures 9-26 through 9-28. These test data are reviewed in Section 9.2.3 - to develop insight into an appropriate bounding stratification gradient. 9.2.2 MSLB Configuration Phase 3 of the LST program included a series of tests designed to simulate a main steamline pipe rupture. LST data from baseline and Phase 2 tests suggested that noncondensible concentrations increase dramatically below the elevation of steam injection with considerable steam mixing above the operating deck. One could postulate that the effect of the higher steamline elevation , could be to create a larger volume of rich air mixture which extends above the operating deck, and reduces the active heat transfer area. Test series 222 addressed the impact of the elevation and direction of the steamline break on the response of the test vessel and included a high flow transient to a steady-state condition. The kinetic energy available in an MSLB is seen to be an important parameter. The four configurations in this test series were: 222.1 Low velocity steam flow from under the operating deck 222.2 Low velocity steam flow above the operating deck (a reference condition to examine the physics, not a realistic AP600 configuration) 222.3 High velocity steam flow with horizontal discharge above the operating deck 222.4 High velocity steam flow above the operating deck directed upward Stratification data for LSTs with high kinetic energy above the operating deck are shown in Figures 9 29 through 9-34, also grouped by steam flow, and showing measu. J internal fluid temperature, saturation temperature, and measured steam pressure ratio, as described in Section 9.2.1 for the LOCA configuration. These data are referenced in the development of a bounding MSLB Evaluation Model (Section 9.4.2). To understand the effects of kinetic energy on circulation and stratification, it is useful to note the stratification pattern observed for a test with a buoyant source (low Froude number) versus a test with a high Froude number. For example, test 222.4 can be used to assess the effects of steam releases with Froude numbers representative of an MSLB occurring above the steam - generator. Test 222.4 is compared to test 222.2, which had a similar setup, but a diffuser was used to provide a low velocity elevated steam source. The elevated buoyant source in test 222.2 produced a significantly stratified vessel, with very little steam penetration below the elevation Circulation and Stratification Within Containtnent October 1997 - o:\3006-non\3006w-9.non:ll> LCM 97 Revision 1
9-20 of the break. In contrast, the high kinetic energy-elevated source of test 222.4 induced a substantial amount of circulation in the test vessel, including substantial steam ingress into the below-deck regions. The decrease in the steam concentration stratification for test 222.4 compared to test 222.2 is due to the high kinetic energy of the injected fluid because that is the only significant difference between the two tests. Mass transfer data from LSTs with high velocity jets (forced convection) has been compared to that from the low velocity diffuser under the simulated steam generator (dominated by free convection) in Reference 9.9, Figure 3.9-5. Results indicate that in the LST forced convection effects enhanced the mass transfer rate by a factor of 1 to a factor of 10 in the direction the jet is directed. 9.2.3 Method to Address Distortions in IST Stratification Data Intemal momentum effects were distorted in the LST due to the lack of a simulated flow path for entrainment near the bottom of the steam generator compartment. Thus in the LOCA DECLG configuration, the LST effectively stratified into two regions - separated at the elevation of the steam generator compartment exit (Section 9.2.1). Therefore, the LST cannot be used to examine intercompartment circulation. There is also a system level distortion in the LST with respect to power-to-volume and power-to-area (Reference 9.7, Section 11). Since only quasi-steady state data for circulation and stratification were used, there is no impact of power-to-vc,lume distortion on this evaluation. The LST quasi-steady data was taken with a range of break flow rates, and the extemal wall boundary condition was ranged using controllable variables (tuming extemal water and fan on and off). The intemal release configuration also allowed varying the release elevation, momentum, and direction. Initial noncondensible content ranged from near vacuum to two atmospheres. Thus the LST provides a valuable database to examine the physics of potential stratification mechanisms that may be postulated to occur in AP600. Because of the momentum-related distortions in the LST, available intemational test data has been reviewed (Appendix 9.C, Section 9.C.2) to supplement the database for examining stratification effects. The supplementing of LST data with additional tests at various scales, combined with the use of LST matrix tests, sufficiently addresses the system level power-to-area distortion. The following summarizes conclusions that may be drawn from LST c.d n the intemational databases, leading to the selection of an extreme stratification gradient to be considered in thermal calculations of Appendix 9.B. It is desired to gain insight into vertical steam concentration gradients that may occur within the region above the operating deck and within compartments below-deck during a LOCA. (The
' bounding approach for an MSLB is given in Section 9.4.2.) The region above the operating deck in the LST can be considered to be an enclosure with a plume and wall boundary layers Circulation and Stratification Within Containment october 1997 l o:\3006-non\3006w.9.non:ll> 1014o7 Revtsion 1
9-21 (Appendix 9.C, Section 9.C.1.4.1). The relevant vertical profile data is presented in Figures 9-2 through 9-25. Comparisons of internal fluid thermocouple data (1-inch inside the vessel wall) and steam concentration measurements show that the gas is within a few degrees of saturation, so that the vertical temperature profiles provide a good measure of the vertical steam concentration gradient during the LSTs. Clearly, for the diffuser under the steam generator model, there is only about a 3 to 12 F temperature gradient from the steam generator exit elevation to the dome. The plotted data is at the fluid thermocouple location. A review of the internal rake temperature data shows that the bulk fluid vertical temperature difference is equal to or several degrees less than that given by the fluid thermocouples. Comparison of the vertical temperature profile from the elevated diffuser case in the LST (Figure 9-26) shows that the stratification in the above-deck region is more pronounced than that in any of the tests with the LOCA configuration. Such stratification from an elevated diffuser is similar to that observed in the CVTR tests (Appendix 9 C, Section 9.C.2.3) which had a similarly elevated, low momentum source. Tests in the much larger HDR and NUPEC facilities indicate that stratification gradients from diffuse releases low in containment in fact produce temperature gradients above the operating deck similar in magnitude to those quoted above in the LST with a low diffuser. However, because of the distortions in LST mentioned above and uncertainties in transferring stratification data from HDR and NUPEC to AP600, an extreme stratification gradient, well beyond that which would occur in a containment with natural convection and a low elevation release, has been considered for thermal calculations. O V The steam concentrations used for thermal calculations presented in Appendix 9 B assume a three region distribution - nearly pure steam at the top (steam fraction 0.98), the average value at the middle (steam fraction 0.63), and the balance of the air content at the bottom (steam fraction 0.28). The elevated diffuser case in the LST shows a steam pressure ratio (equal to steam mole fraction) of 0.10 near the operating deck and 0.90 under the dome. The distribution chosen is consistent with that indicated by the LST elevated diffuser, considering that the Appendix 9.B calculation represents an average steam concentration calculated for AP600 transient conditions, it should be noted that the LST elevated diffuser test produces an extreme, or bounding, test configuration for the real situation of a buoyant plume released low in containment, such as for the LOCA DECLG post-blowdown. Thermal calculations in Appendix 9.B are used to develop appropriate biases to bound the effects of stratification within the AP600 lumped parameter compartment nodes and the above-deck region. 9.2.4 Application of Modeling Methods Developed for NUPEC M-4-3 Lumped Parameter Model The following is a brief summary of the experience gained in developing the WGOTHIC lumped parameter model of the NUPEC natural circulation test, M-4-3, and application of the experience n to development of the WGOTHIC lumped parameter Evaluation Model. Justification is provided b Circulation and Stratification Within Containment october 1997 oA3006-non \ 3006w-9.noru i t-101497 Revision 1
9-22 for using the lumped parameter Evaluation Model for performing sensitivity studies. The sensitivities are used to examine the effects of circulation in containment from a LOCA DECLG. NUPEC Lumped Parameter Modeling Experience Actual circulation was interpreted based on data provided by NUPEC for the detailed time history for gas temperature and hydrogen concentration as wc!! as a video of processed data to aid visualization. As shown in Figures 9.C.2 32 (flow pattern) and 9.C.2-38 (data for one circulation loop) of Appendix 9.C, the break flow rose from the affected steam generator loop, spread through the upper portion of the large vertical opening into the adjacent steam generator loop, and rose from those two compartments into the dome. The large-scale natural convection loop continued with continuity driving circulation down through the opposite steam generator compartments and other openings through the operating deck, and then down to the level of the break release. From the break release level, the convection loop was closed by entrainment into the rising plume. This result is consistent with results of international tests at several scales and is rather simple and straightforward. However, careful development of the lumped parameter noding structure is necessary to allow the code to predict the observed qualitative behavior, as follows, it should first be noted that for the M-4-3 calculations, best estimate condensation correlations were used to better isolate the biases of lumped parameter noding on predicted parameter distributions and the effect of those biases on containment pressure. For general application of WGOTHIC lumped parameter,it is necessary that the vertical noding be defined by a set of horizontal planes that cut through the entire modeled region, as described in Reference 9.4, Section 16.12.1. This is done to prevent artificial flows driven solely by the method used to estimate a static pressure profile using the single value of density availab'e within a lumped parameter cell. The successful elimination of such artificial circulation is confirmed when a new model is developed by runnmg a null problem (uniform temperatures in heat sinks and volumes, and no heat or mass source) and verifying that there is no predicted circulation. (a,c) Circulation and Stratification Within Containment october 1997 c:\3006-non\3006w-9.non:lt>101497 Rmsion 1
~... _ _ _ _ __._. - . _ _ _ . _ - _ _ . . - . _ _ _ .
9 1 l (a,c) -- Application of NUPEC Test Experience to Containment Etmluation Model. l S The Evaluation Model has >een verihed to have no significant artificial flows in a null problem. In the ft.rther development of the WGOTHIC lumped parameter noding used in the containment , pressure Evaluation Model, experience with the NUPEC tests was used qualitatively in representing the CMT compartment.
~
(a,C) h - _
'i O , - Circulation and Stratification Within Containment October 1997 - i o:\3006 non\3006w-9.non:ltr101497 Revision 1 -!
9-24 i Table 9 2 Flow Areas Connecting to North and South CMT Compartments (excluding Dead Ended Compartment Connections) ummmmmmmmmmmmmurummmmm mmmmmmmmmmmmmusamm Flow Path End B Flow Path Number Flow Path End A Node Node Flow Area (ft') (a,c) O O Circulation and Stratification Within Containment october 1997 o- \3006-non\ 3006w-9.non:1b.101497 Revision 1
.,~ . . . . . . . . - - ~ ~ . _ . . . - - . . . - _ _ - . . - . . . . - . . . . _ . . . . . . - . . . . . . . . - . . . . . . ~ . . . . - . . . . . - . ~ . . . - .
a - i,. .
- 9 25 '
f 4,p 1 1
-i
? I
,' 6 4 I i 1 1, -
i a. l' e i l i i
,t ,
4 _} i n 4 5 k 4 f Circulation and Stratification Within Containment october 1997 -
- o:\3006-non\3006w-9.non:1tr101497 Revision 1 -
9-26 9.3 CIRCULATION AND STRATIFICATION ASSESSMENT FOR THE LOSS-OF-COOLANT ACCIDENT h' The rupture of primary system piping can lead to a significant release of mass and energy into the AP600 containment. A containment analysis is performed to verify the ability of the passive containment systenu, to mitigate the consequences of a hypothetical LOCA. The }.VCOTHIC code, in conjunction with an AP600 Evaluation Model,is used for the containment analysis. The effects of circulation and stratification must be bounded by the containment analysis calculations to ensure a conservative containment analysis. For purposes of evaluating the effects of circulation and stratification on the LOCA containment analysis, the LOCA event is divided into four temporal phases: the blowdown phase, the refill phase, the peak pressure phase, and the long-term phase, based on Section 3.4.2.2 of Reference 9.2. The blowdown phase is the period immediately following the rupture of the primary system piping: For the design basis event, a double-ended, cold leg guillotine (DECLG) break is assumed, which results in the complete severance of the pipe. This phase is characterized by a rapid depressurization of the reactor coolant system (RCS), as the RCS inventory is expelled into the containment volume. The containment gas volume rapidly pressurizes due to the tremendous release of mass and energy. This phase is short in duration (about 30 seconds) and ends when the RCS pressure has equilibrated with containment. 4 The refill phase immediately follows blowdown. After blowdown, the accumulators refill the lower plenum of the reactor with a high flow rate of cold water so that releases from the break cease for about 60 seconds. As the reactor water level rises through the core, water is turned to steam. The resulting steam and water flow rates from the break are very low and increase with time. The mass and energy release rates are two orders of magnitude less than the blowdown rates, and can be approximated as 0 from approximately 30 to 90 seconds into the event. With a negligible steam source rate and a high condensation rate, the containment pressure drops by a few psi from its peak at the end of blowdown to the end of the refill phase at approximately 90 seconds. (It should be noted that the Evaluation Model used for sensitivity studies conservatively neglects the period of no releases.) The phase following refillis the peak pressure phase. During the beginning of the peak pressure phase, a continuing pressurization of the containment building accompanies the release of mass and energy. Containment pressurization is mitigated by the containment gas volume and the presence of the substantial number of heat sinks in the AP600. Hot steam condenses on the cold steel and concrete surfaces, which transfers energy into the heat sinks. As this phase continues, the temperature of the intemal heat sinks increases and their effectiveness is reduced. By this time, however, water flow onto the AP600 containment shell has initiated. The PCS provides the path to the ultimate heat sink, and represents the only assumed path through wluch energy can be removed from inside the containment building. A key feature of the peak pressure phase Circulation and Stratification Within Containment Octotwr 1997 l o43006-non\3005w-9.non:lb-101497 Revision 1
9-27 is the second, more limiting pressure peak. The combination of internal heat sinks and the PCS act to limit the containment pressurization, and containment pressure begins to drop. Later in this phase, the PCS b:comes clearly dominant. The peak pressure phase extends from 90 seconds to about 1200 seconds. During this phase, at about 1000 seconds, ADS Stage 4 actuates and becomes the source of mass and energy release. The long-term phase is the period from the second peak pressure (about 1200 seconds) to twenty four hours and beyond after the accident initiation. During the long-tem phase, core decay heat continues to create steam, which exits the fourth stage automatic depressurization system (ADS) as a buoyant plume. The containment continues to depressurize as a result of energy removed by the PCS. As containment pressure drops, internal heat sinks may begin to reject some of their heat back into the containment atmosphere. Thus the long-term phase depressurization is govemed by PCS heat removal. To facilitate an understanding of the relative positions of the various compartments, a simplified compartment diagram is provided in Figure 9-?5. Figure 9-36 shows the relative location of various important compartments, such as the steam generator compartment, the core makeup tank (CMT) compartment, and the above-deck volume. Noding used to represent these compartments within the Evaluation Model is described in Section 4. The compartment features are discussed in Section 4 and summarized in Table 3-1 of Reference 9.2. Figure 9-35 presents a diagram of the CMT compartment. The CMT room contains most of the below-deck containment heat sinks (approximately 52 percent of below-deck heat sinks by area). Although (V) 48 percent of the heat sinks are not in the CMT room, no other single below-deck compartment contains as many heat sinks. Also, the CMT room is the largest (volume) of the below deck compartments and contains many flow paths. These flow paths mean that the CMT room is of significant importance with respect to both above- and below-deck circulation patterns. Therefore, the effect of circulation and stratification on heat sink utilization in the CMT room plays an important part in the transient pressure mitigation in AP600. 9.3.1 LOCA Break Scenarios The DECLG rupture is the design basis LOCA event for the AP600. The circulation and stratification patterns associated with this break will depend on the direction of the break jet momentum. Although leak-before-break has been implemented for AP600, the conservative design basis analysis evaluation assumes the broken pipe can be pointed in any direction from its nominal position. Three scenarios may be postulated: the jet momentum is locally dissipated in the steam generator compartment, the jet exits undissipated up through the steam generator compartment, or the jet momentum is dissipated in the reactor coolant drain tank (RDCT) cavity (stainvell). rh
' )) \
Circulation and Stratification Within Cordainment october 1997 oM006-non\3006w-9.non:1b-101497 kevision 1
9-28 9.3.1.1 Jet Momentum Locally Dissipated in Steam Generator Compartment During the blowdown phase, a tremendous amount of mass is released as shown in Figures 4-96 O and 4-98 of Section 4.5.2. For the case where the jet momentum is locally dissipated, the source flow rate is so high that it increases the local pressure by several psi. This results in a high-pressure source in the break compartment, with the fluid flow distribution governed by the relative resistances through flowpaths. This forces the source mixture through the RCDT cavity, CMT room, the steam generator compartments, and into the above-deck volume. Pressurization will also drive steam into dead-ended compartinents during blowdown (See subsection 9.3.2.1). As the event progresses into the peak pressure phase, the source flow rate drops by two orders of magnitude. The jet momentum locally dissipates. This brings the source flow velocity to near zero, including a local pressure increase that is the same order of magnitude as the buoyant forces. The pressure source may be opposed or aided by buoyancy in other flow paths. The resulting flow pattern is the solution to the flow in a network with buoyancy and heat / mass transfer in the network branches. Superimposed on the large-scale flow, the mixture within a given compartment is most likely stratified (Reference 9-8). Widtin compartments, the gas may stratify with air concentrating in lower regions and steam concentrating in upper regions, resulting in a vertical steam concentration gradient. If the circulation is sufficient to entrain significant bulk mixture, the gradient may be expected to be small. Entrainment-driven circulation rates in the CMT room are shown, for example, in Section 9.3.1.3. Significant circulation occurs over the height of the CMT room. h Stratification is expected in the AP600 based on LST data. Low Froude numbers during the long term indicate a low kinetic energy buoyant plume source. This type of plume is not , sufficiently energetic to disrupt stratification in the AP600. The physics of buoyant plumes and wall layers leads to the existence of recirculating stratification (Appendix 9.C, Section 9.C.1.4.1) in the above-deck region. Plumes rise from the release point and entrain significant volume of mixture as they rise. The heavier bulk air / steam mixture is drawn through the top of the CMT and other deck openings and through compartments to be entrained into the rising plume. Stratification is assumed to have a negligible impact on heat removal in compartments which experience the already air-rich downflow. A very conservative assessment of the effects of stratification on heat removal through the steel shell by the PCS has been performed (Appendix 9.B). An extreme stratification gradient is assumed, to bound the potential for distortions in test data relative to AP600 (9.2.3). The homogeneous case total heat sink utilization results are nearly equal to those for the stratified case, with the homogeneous case giving less than 0.5 percent les:, instantaneous heat removal rates. A simple bias of removing operating deck floors is included in the Evaluation Model to bound this effect. The containment pressure was calculated for this case using the WGOTHIC Evaluation Model, (Section 4). It was assumed that the jet was dissipated in the East steam generator compartment, Circulation and Stratification Within Containment october 1997 c: \3006-non \3006w-9.non:1 b-101497 Revision 1 l
9-29 so no specific break orientation was modeled. The break was located in Volume [ )(") at l U-f_) elevation [ l'") The results are discussed in Section 9.3.2.5. 9.3.1.2 Jet Directed Up With No Dissipation A jet directed upward, that passes through the steam generator compartment undissipated, is considered unlikely. Releases are initially from the break and, later in the transient, releases exit from the fourth stage ADS and the break pipe is covered with liquid. The AP600 design calls for a steel plate to cover half the flow area in the steam generator compartment above the celd leg pipe and ADS Stage 4 valves. This plate and other structures in the steam generator compartment such as gratings, supports, and the steam generator itself make it ' doubtful that the break jet could pass through the steam generator compartment unobstructed. Despite the improbability of this scenario, it will be considered as an extreme case to support the selecdon of a limiting scenario for circulation and stratification. For the case in which a jet is postulated to pass undissipated up through the steam generator compartment, there is no entrainment into the Steam Generator compartment due to chimney or momentum effects because these effects would act to dissipate the jet. An undissipated jet would enter the above-deck region at the top of the Steam Generator compartment with approximately the same diameter as the broken cold leg pipe. This scenario is similar to two of the IST MSLB configuration tests 222.3 and 222.4. To assess the effects relative % the mass ( transfer in the above-deck region, volumetric Froude numbers (Fr,) for the undissipated jet are determined and compared to the LST. An examination of the magnitude of AP600 pressure improvements is provided with sensitivities, relative to condensation results discussed in Reference 9.9, Section 3.9. For a LOCA DECLG, a postulated undissipated jet will have the same mass flow rate as the AP600 design basis LOCA DECLG exiting the top of the steam generator compartment. The two cases differ in the flow area and exit velocity. For the design basis case, the flow area is the area at the top of the Steam Generator compartment. For the undissipated jet, the flow area is the area of the cold leg pipe. For a constant mass flow rate, the product of the flow area times the exit velocity will be equal for the two cases (Uotetx Aosct = Umoex Amos, vchere U is the velocity, A is the area, subscript occo designates the design basis case, and subscript mos designates the undissipated jet case). Fr, defined in Section 9.2 is proportional to U2d2, and is 2 therefore proportional to U2A . For the two cases, the other terms in the Fr, equation will be the same and Fr,.mos can be expressed in terms of Fr v .oect, using Uotetx Aoect = Umos x Amos. The relationship is Fr,.mos = Fry .otet (Aotet / Amos)2 The area of the top of the Steam Generator compartment is approximately [ ](") and the area of the cold leg pipe is approximately [ ]("' This results in Fr,.mos ~= Fry.otet X[ }" g Reference 9.7, Section 6.5.2 presents Fr y as a function of time for the design basis LOCA in Figure Q 6-2, At 24 hours Fr, is approximately 3E-06 (the minimum value during the transient excluding Circulation and Stratification Within Containment october 1997 o:\3006-non\3006w-9.non:lt>101497 Er <ision 1
9-30 the refill phase). For an undissipated jet, Fr, is estimated to be 3E-06 x [ ]"*) which equals [ )(**). This value is at the lower end of the LST Fr range in the MSLB configuration as shown in Reference 9.7, Figure 6-3. For such high values of Fr, data from the LST in the MSLB configuration (Section 9.2.2) shows that there is minimal deviation from a homogeneous steam concentration in the above-deck region. For the MSLB, Reference 9.9, Figure 3.9-5 shows that use of the Evaluation Model free convection correlation underpredicts condensation on shell surfaces by a factor of [ ]"') for the LST. A multiplier of [ ]"") is a reasoaable factor to assess based on the data. To address postulated uncertainty in scaling the LST condensation results to AP600, a range of potential forced convection benefits in AP600 shell heat transfer are considered by examining the sensitivity of predicted containment pressure to condensatien multipliers in the Evaluation Model. A sensitivity study examined the effects on containment pressure of using condensation multipliers of [ ]"*) These sensitivity cases show that taking credit for improved condensation provides a significant benefit in the calculated containment pressure. The results are discussed in Section 9.3.2.5. 9.3.1.3 Jet into RCDT Cavity (Stairwell) During the blowdown phase, a jet into the RCDT cavity will create a pressure source in the RCDT cavity compartment. As with the jet dissipation in the East steam generator compartment, the high pressure source will force fluid through all available openings. The source mixture will flow into the above-deck volume through both the CMT room and steam generator compartments. Following the blowdown phase, the source will rise from the RCDT cavity as a buoyant plume and split, based upon flow areas and resistances, with part of the flow rising through the West steam generator compartment and the remaining fluid flowing through the CMT compartment. The post-blowdown flow split between the West steam generator and the CMT compartment will depend on flow areas and loss coefficients associated with both flow paths. A range of flow splits can be postulated varying from all the fluid rising through the steam generator compartments to all of the fluid rising through the CMT roem and everything in between. The first scenario is an extreme case which postulates that all the fluid rises through the West steam generator compartment. This scenario is identical to the scenario that assumes the jet momentum is locally dissipated in the East steam generator compartment. The case of the jet momentum dissipated in the East steam generator compartment is discussed in Section 9.3.1.1. The buoyant plume rising from the RCDT cavity into the West steam generator compartment is essentially the same scenario. The second scenario is a split of the flow entering the RCDT cavity, with part of the break flow rising through the West steam generator compartment and part rising through the CMT compartment. The flow split is dependent on the relative flow path resistances. In this scenario, both the steam generator compartments and the CMT compartment would be subjected to a Circulation and Stratification Within Containment october 1997 o:\3006-non\3006w-9.non 1b-101497 Revision 1
9-31 i steam rich break plume. The CMT and steam generator compartments contain the majority of lO the below-deck heat sinks. The flow split will result in good heat sink utilization subjecting both the steam generator compartments and the CMT compartment to the steam source. Thus, the case with the jet momentum dissipated in the ItCDT cavity and a plume flow split between the CMT and steam generator compartments, will not be limiting. This is confirmed in the j sensitivity calculations of Section 9.3.2.5. The third scenario is an extreme case which postulates that the plume from the RCDT cavity rises into the CMT rcem. For this scenario, the buoyant plume rises from the floor to the ceiling of the CMT room, entraining gas from the bulk concentration present in the CMT room. An examination of entrainment into a CMT plume can be used to gain insight into the potential for stratification. Calculation of CMT Room Plume entrainment Rates For the case of the LOCA jet being dissipated in the CMT room, the rate of entrainment of mixture in the CMT into the incoming break flow plume, Q,, can be estimated based on the work of Peterson (Reference 9.15). In particular, Peterson gives the following relation for the volumetric entrainment rate into a buoyant plume, Q, = k, B "5 z 5 (91) b where k, is a constant equal to approximately 0.15, : is the height of the plume, and B is the buoyancy flux, given by:
~
B=g ' Q, (92) P.
' In this equation, g is acceleration due to gravity, p, and p, are the ambient fluid and injected fluid densities respectively, and Q is il e ' >lumetric flow rate from the plume source.
I'3 V Circulation and Stratification Within Containment _ _ octot-r 1997 o:\3006-non\3006w-9.non.ll>101497 Revision 1
9 32 Substitution of equation (9 2) into L; 'ition (91) gives:
~
Q, = k, [ g
- Q,)8 z 8 (93)
P, The ratio of entrained flow to break flow is therefore:
~
I b = k, jg ' __]'/8 z (94) R. P, Q,8 Break flow rates for a LOCA DECLG at transient times of 460 seconds and 1,000 seconds are 1,070 ft'/sec and 266 ftS /sec respectively for steam. The injected fluid density is taken as the density of saturated steam at the CMT room pressure. These densities are 0.128 lb/ft' (based on 54.6 psia at 460 seconds) and 0.135 lb/ft'(based on 58 psia at 1,000 seconds). Ambient fluid density is taken as the total density of gas mixture in the CMT room at the times of interest, inspection of the WGOTHIC output, from the sensitivity case which modeled the break in the CMT coom (see Section 9.3.2.5), indicates densities of 0.158 lb/ft' at 460 seconds and 0.165 lb/ft' at 1,000 seconds in the CMT room. The height of the CMT room is 28.1 feet. Based on this data the applicable entrainment ratios, Q,\Q,, are 0.68 at t=460 seconds, and 1.7 at t = 1000 seconds. An entralnment-driven circulation time constant for the CMT room is calculated by dividing the entraint.sent flow rate into the volume of the CMT. From above Q,\Q, is 0.68 when Q, is 1066 ft'/sec and 1.7 when Q,is 266 ft'/sec. Solving for Q, gives a range of 725 to 452 it'/see for the entrainment rate. The volume of the CMT room is approximately 157200 ft' and the resulting circulation time constant ranges from 217 seconds to 348 seconds (3.6 to 5.8 n inutes). This range is relatively short compared to the time of ADS Stage 4 actuation (approximately 1000 seconds), when the steam source is relocated to the steam generator compartments. Assessment of CMT Room Entrainment Circulation The entrainment rate for this case is relatively large, increasing to over a factor of two relative to break flow later in time. Thus, a significant amount of CMT room mixture is entrained into the break as the plume rises to the ceiling. It may be concluded that vertical concentration gradients in the CMT room would be relatively small due to circulation within the room. It also may be concluded that the break flow circulates within the room, significantly increasing the room average steam concentration. Thus, high steam concentrations are expected in the CMT room compared to other break scenarios. The high steam concentrations for this scenario will result in high heat sink utilization for heat sinks in this important room. O Circulation ond Stratification Within Containment _ october 1997
~
o:\3006 non\3006w-9.rmrt1b.101497 RevWon 1
9 33 l With such low density mixture in the North CMT room, the chimney effect induces flow to the room from connecting flow paths at the floor elevation. Connecting flow paths from the Section 4 Evaluation Model are [ l'") horizontally connecting to the steam generator compartments, and until the liquid level closes the path, [ l'"5 from the RCDT cavity (see Figure 9-47). The density head over almost 30 feet of height outside the CMT room strongly drives circulation through the CMT and upward in this scenario, suggesting that the flow should rise from the North CMT room into the above-deck region. There is little resistance to flow navigating past the CMT room pinch point to access the ceiling openings on the South CMT room opposite the stainvell, suggesting that flow would spread as it rises into the South CMT room, and then rise from all CMT deck openings. It is known from studies of building fires that very little pressure driving force is necessary to drive horizontal flow in a stratified room (References 9.10, 9.11, 9.12). The effect of stratification on heat sink utilization is also evaluated. Room pressu*e, temperature, and steam concentrations were input into a separate calculation to assess the potential effect of stratification in the CMT room. For the calculation, the CMT room was divided vertically into three equal sections. Using free convection heat and mass transfer correlations, room heat sink energy removal was calculated for a room with a homogeneous steam concentration. The applied steam fraction was .63. For the second scenario, the CMT room was subjected to a stratified condition. The top region was assumed to be nearly all steam (steam fraction = 0.98), q the middle region was assumed to have a nominal steam fraction (0.63), and the bottom region Q steam fraction was determined by conserving the total amount of stea: in the total volume (0.28). - Figure 9.B-3 shows the energy absorbed by the heat sinks in the CMT room for; 1) a stratified steam concentration with the CMT floor included, 2) a homogeneous steam concentration with the CMT floor included, and 3) a homogeneous steam concentration without the CMT floor included. As Figure 9.B-3 shows, the homogeneous concentratien with the floor results in the most energy absorbed in the CMT room (top curve). The curve for the stratified concentration with the floor is close to the curve for the homogeneous concentration without the floor. The curve for the homogeneous concentration without the floor is more conservative (less energy absorbed) after 2000 seconds. Given the relative closeness of these two curves, and considering the extreme cases they represent, it is concluded that the lumped parameter Evaluation Model (which uses a homogeneous steam concentration in each volume) without floors provides a reasonably conservative model for heat sink utilization, accounting for the thermal effects of potential stratification. Information on the heat sink utilization calculations is presented in Appendix 9.B. c The break scenaria with a buoyant plume flowing into the CMT compartment will not be a limiting scenario. The evaluation of this scenario has shown only small vertical concentration gradients are expected in the CMT compartment while a bias has nevertheless been implemented by removin;; the floor. Furthermore, high steam concentrations are expected in this 3 compartment due to the large amount of entrainment and subsequent circulation driven by the (V break plume. The high steam concentrations will yield improved heat sink usage in this room. Circulation and Stratification Within Containment orehr 1997 c:\3006-non\3006w &nortit*101497 RevWon 1
9-34 The scenario discussed in Section 9.3.1.1, with the jet momentum dissipated in the steam generator compartment, will have lower steam concentrations in the ChfT room. Thus, the break scenario with a buoyant plume flowing into the ChtT compartment will be bounded by the case hl with the break jet locally dissinated. To further confirm this conclusion, the results of a EGOTHIC analysis using the Evaluation hiodel (Section 4), for a buoyant plume flowing into ) I the CMT compartment, are discussed in Section 9.3.2.5. The analysis confirms that the buoyant plume rising into the CMT compartment is not a limiting scenario. 1 9.3.2 AP600 LOCA Evaluation Model The ECOTHIC Evaluation Model uses lumped parameter noding to model the AP600 containment. Lumped parameter noding simplifies the calculation by assuming homogeneous conditions in each network node. Lumped parameter formulation uses what may be called a scalar form of the momentum equations as follows. Here, momentum flow into each volume is parallel to the junction, and the terms perpendicular to the junction are discarded while junction momentum is dissipated within the volume. Momentum orientation is not tracked, and no tuming losses are represented. During the LOCA blowdown phase, the high break mass flow pressurizes the steam generator compartment and flow exits based on relative loss coefficients. Such pressure-driven flows are reasonably modeled by the lumped parameter node-network formulation. Lumped parameter reasonably.epiesents buoyancy and pressure-driven flows and the resulting large-scale circulations. The effects of stratification within each compartment or region can then be superimposed on the large-scale circulation solution. The Evaluation Model is described in Section 4. Comparison of lumped pt.iameter GOTHIC results to test data, has shown lumped parameter noding to be acceptable for LOCA breaks occurring in low zones of containment. Reference 9.13 discusses the test results and subsequent GOTHIC evaluation of the German Heissdampfreaktor (HDR) hydrogen mixing and distribution experiment E11.5. This experiment simulated a large-break LOCA in the lowest region of the HDR containment. The authors conclude that accident scenarios initiated by large break LOCAs in the low zones of containments can be reliably predicted by the GOTHIC lumped parameter model using only a modest number of nodes (Appendix 9.C, Section 9.C.3.3). The DBA LOCA case models the break in [ ]'"' (lower East steam generator compartment) at the [ )(") elevation (refer to Section 4.5.2.1). The conclusions concerning the use of a lumped parameter for low breaks modeled by GOTHIC, can be readily applied to EGOTHIC because of the similarity between the two codes. EGOTHIC is a descendant of the GOTHIC code. The difference between the two codes relates to the heat and mass transfer correlations applied to WGOTHIC by Westinghouse, to model the PCS phenomena in the AP600. Thus, since the LOCA scenarios of interest are breaks in the lower region of containment, it is reasonable to use EGOTHIC lumped parameter to model these events. Circulation md C. ratification Within Containment october 1997 c:\3006-non\3 K6w.9.non:1t>101497 Revision 1
9-35 g As discussed previously, the LOCA event is divided into four phases: the blowdown phase, the () refill phase, peak pressure phase, and the long term phase. These phases are discussed in Subsections 9.3.2.1 through 9.3.2.4. l 9.3.2.1 Blowdown Phase (0 to 30 seconds) The sumped parameter solution during blowdown is a node. network solution, governed by j pressure differences and flow resistances between nodes. The mass and energy release in the Evaluation Model acts as a high pressure source that forces the steam out through flowpaths connected to the source node. The Evaluation Model also assumes only free convection on inner containment surfaces. Based on high kinetic energy during blowdown (Ref. 9.7, Figure 6-2) significant enhancement to mass transfer due to forced convection occurs (Section 9.2.2). The steam is driven into the below-deck region and the above-deck volume. Figure 9-37 shows the calculated steam concentration of various containment regions during the blowdown phase, using the Evaluation Model described in Section 4 with a dissipated break in the SG East compartment. The paragraphs in this subsection describe several sensitivity cases and an evaluation performed to examine various aspects of the blowdown phase. The first sensitivity case examines the effect of modeling a containment with a homogeneous steam concentration on the calculated n containment pressure. The second sensitivity case examines the effect of removing all internal V heat sinks on the calculated containment pressure. Following this sensitivity case, an evaluation of heat sink utilization in dead-ended compartments is performed. The final sensitivity case examines the effect of varying the flow pattern and steam concentrations on the calculated containment pressure. To show the relative insensitivity to stratification, or heat and mass transfer coefficient during blowdown, a comparison is needed between the containment pressure response predicted by this node-network solution, and the containment response predicted for a homogeneous containment. Section 8, Figure 81 compares the LOCA blowdown pressure results of a one node WGOTHIC AP600 model to the node-network solution. The one-node model assumes the same total containment volume and containment heat sinks as the multi node model. Both models predict essentially identical containment pressure responses during the blowdown phase. Therefore, the details of the flow connections and heat mass transfer rates for the multi node Evaluation Model are not important with respect to the containment pressure results because volume compliance is the dominant pressere mitigator during blowdown. During the blowdown phase, the mass and energy release is mitigated primarily by containment volume via the rapid pressurization of the containment building. Figure 8-2 shows a comparison of the AP600 Evaluation Model results for the blowdown phase versus an identical model with (3 all internal heat sinks removed. At the end of blowdown (30 seconds), the difference between d these two cases is about 3 psi, accounting for only 10 percent of the pressurization. Thus,in the Orculation and Stratification Within Containment October tw7 o:\3006-non\3006w-9.non:lt*101497 Revision 1
9-36 Evaluation Model, the blowdown mass and energy release increases containment pressure by about 35 psi, while the containment heat sinks absorb approximately 3 psi worth of energy. Clearly, the dominant mechanism during blowdown is the pressurization of containment. h , J l The heat sink effectiveness in the presence of a stratification gradient is evaluated in Appendix 9.B. To consen'atively account for the reduced effectiveness of heat sinks in lower room areas, floors are eliminated in the AP600 }yGOTHIC Evaluation Model throughout the transient. The effectiveness of heat sinks in dead-ended compartments is also evaluated. Since only one opening exists for these compartments, interaction with overall containment volume is expected to be minimal unless the compartments have non-uniform temperatures. During blowdown, these compartments pressurize along with the rest of containment. Steam / air mixture from the bulk containment volume flows into the dead-ended compartments during the initial pressurization. Once pressurized, additional steam / air flow into the dead ended compartments only occurs to make up for steam condensing in the compartment. Analysis of the Nuclear Power Engineering Corporation (NUPEC) natural circulation test, M-4 3, showed that asymmetric heating of dead-ended compartment walls can lead to natural circulation flows within the compartment (Reference 9.14). However, a conservative evaluation of dead ended compartments would consider no thermally driven circulation. In such a case, inside the compartments, the condensation of steam leaves behind a heavier air-rich mixture. The air flows to the bottom and blankets the lower heat sinks. The poor circulation within the dead-ended compartments leaves the air rich layer relatively undisturbed. As steam continues to condense, the air rich layer continues to build up and will result in significant stable stratification within the di.ad ended compartments. Although the heat sinks in the dead-ended compartments will contribute somewhat to containment heat removal, to conservatively bound the effects of stratification, condensation and convection on the heat sinks in the dead ended compartments are neglected after 3G seconds in the Evaluation Model. Based on the results of the evaluation,it has been demonstrated that blowdown pressure history is relatively insensitive to the effects of circulation and stratification. The internal heat sinks do heat up during blowdown, however, as discussed above, containment volume pressurization is the dominant mechanism for absorbing the energy released. Since volume pressurization is the governing process, blowdown pressure response is not sensitive to circulation and stratification effects. The Evaluation Model utilizes a conservative lower estimate of containment free volume.
- lhus, the uncertainties in heat and mass transfer or stratification, and flow path effects, do not significantly impact the AP600 LOCA blowdown pressure history and the Evaluation Model adequately models the LOCA blowdown phase.
To assess the effects of varying the steam concentrations and flow rates on the cakulated containment pressure, a sensitivity was performed which varied several loss coefficients in the Evaluation Model. This sensitivity shows how changes in conditions during the blowdown Circulation and Stratification Within Containment october 1997 oA3006w-9mortib-101497 Rension 1
9-37 phase affect the later phases and, in particular, the calculated containment pressure. For this ph sensitivity, the Evaluation Model (Section 4), with a dissipated jet in the SG East compartment, was used and the loss coefficients [
]'"' Figure 9-38 shows the pressure transient for this sensitivity case. The maximum calculated pressure is 43.8 psig, which is 0.1 psi less than the 43.9 psig reported in Section 9.3.2.5.
Circulation plots for this sensitivity case are presented in Figures 9 39 through 9-12. Compared to the circulation plots for the dissipated jet in the SG East compartment (Figures 9-47 through 9-50), the effects of the revised loss coefficients are evident. At 20 seconds, Figure 9 39 shows that most of the break flow goes from the SG East compartment to the SG West compartment, and through the RCDT cavity to the North CMT room. At 1000 seconds (Figure 9-40), flow is rising from both SG compartments and a steam / air mixture is flowing down into the North and South CMT volumes. Figure 9-48, shows flow rising only from the SG East compartment. At 1550 and 80050 seconds (Figures 9-41 and 9-42) the ADS Stage 4 valves are the source of the steam releases and the flow patterns are similar to those in Figures 9-49 and 9 50. This sensitivity altered the flow patterns and steam concentrations early in the transient by changing some of the flow path loss coefficients. The change in calculated maximum pressure was p negligible. d l 9.3.2.2 Refill Phase (30 to 90 seconds) l The refill phase immediately follows blowdown. After blowdown, the accumulators refill the lower plenum of the reactor with a high flow rate of cold water so that releases from the break cease for about 60 seconds. As the reactor water level rises through the core, water is turned to steam. The resulting steam and water flow rates from the break are very low and increase with time. The mass and energy release rates are two orders of magnitude less than the blowdown l rates, and can be approximated as 0 from approximately 30 to 90 seconds into the event. With a negligible steam source rate and a high condensation rate, the containment pressure drops by a few psi from its peak at the end of blowdown to the end of the refill phase at approximately 90 seconds. For the calculation of maximum containment pressure, the Evaluation Model conservatively neglects this pressure reduction by using the peak pressure phase mass and energy releases immediately following the blowdown phase (see Section 4.5.2.1). 9.3.2.3 Peak Pressure (90 to 1200 seconds) During the peak pressure phase, the location of the steam releases changes from the break to the ADS Stage 4 valves in both steam generator compartments. The Evaluation Modelincludes this (~T change in steam release location (see Section 4.5.2.1). In addition, the lower compartments begin to fill with liquid from the break. The reduced heat transfer area due to filling is accounted for Circulation and Stratification Within Containment octotwr 1997 o:' 906w.94.non:1b.101497 Rmsion 1
9-38 in the Evaluation Model (see the Thermal Conduction Description sections in Section 4.2). Figure 9-43, for a jet dissipated in the SG East compartment, shows that the condensation on the steel becomes the dominant mechanism for heat removal towards the end of the peak pressure h phase. The evaluation of break scenarios in Section 9.3.1 led to the conclusion that the case with jet momentum dissipated in the steam generator compartment may lead to stratification within compartments after the blowdown phase. Given this possibility,it is necessary to show that the Evaluation Model bounds the possible effects of this stratification. Lumped parameter models assume no gradients within each volume of the network. Thus,in the Evaluation Model, all heat sinks within a compartment volume see identical environmental conditions. In contrast, actual conditions may lead to a stratified compartment with a region of higher steam concentration on top and lower steam concentration near the bottom. For the effects of stratification on heat sink utilization, the most significant heat sinks are the above-deck region (containment shell) and the CMT room (steel and jacketed concrete). The compartment features are discussed in Section 4 and I summarized in Table 31 of Reference 9.2. In Section 9.3.1.3, the CMT room was assessed for its I sensitivity to stratification. In this calculation, heat sink usage was calculated for a homogeneous room and a severely stratified room. A bias has been defined to bound the potential effects of stratification in compartments as discussed in 9.3.1.3. In Section 9.3.1.1, the containment shell was assessed for its sensitivity to stratification. A bias has been defined to bound the potential effects of stratification above-deck as discussed in Section 9.3.1.1. Appendix 9.B discusses the calculations performed. Based upon the results of the evaluation, a method to bound circulation and stratification effects for the peak pressure phase has been developed. In the Evaluation Model, all floors are neglected throughout the transient and condensation and convection on all heat sinks in dead-ended compartments are neglected after 30 seconds (refer to Section 9.3.2.1). 9.3.2.4 Long Term Phase (1200 seconds to 24 hours) Figure 9-43 shows the condensation on the steel shell remains the dominant mechanism for heat removal during the long term. The results shown are from the Evaluation Model (Section 4) with a dissipated jet in the SG East compartment. During early portions of the transient, internal heat sinks are the primary path of containment heat removal. As the transient progresses, the temperature of the heat sinks increases and their heat removal effectiveness is reduced. PCS heat removal, which dominates in the long term, is dependant on steam concentrations. The effects of stratification on the containment shell heat removal have been evaluated in Section 9.3.1.1 and a bias of removing operating deck floors has been included in the Evaluation Model. In addition, WGOTHIC predicts a slight gradient between the upper and lower compartments (excluding dead-ended compartments). Figure 9-44 shows WGOTHIC predicted steam concentrations for various compartments in the AP600, using the Evaluation Model (Section 4) Circulation and Stratification Within Containment octoberE o.\3006w 9tnon:1b-101497 Revision 1
9 39 with a dissipated jet !n the SG East compartment. As Figure 9-44 shows, at 24 hours WCOTillC ( predicts a homogeneous above-deck region. However, WGOTHIC predicts a slightly lower
- steam concentration below the operating deck, excluding the SG compartments which continue to have steam release through the ADS Stage 4 valves.
i The trend over time for the WGOTHIC calculations leads to a very small steam density gradient between above. and below-deck compartments. The WGOTHIC predicted average steam concentration above the operating deck is approximately 0.47 at 24 hours. Below the operating deck, the average is approximately 0.46 at 24 hours excluding the SG compartments. The calculated steam concentration for a homogeneous condition between the above-deck region and the below-deck open compartments is approximately 0.468. There is a negligible change i between the WGOTHIC calculated above-deck steam concentration and the calculated ! homogeneous concentration (excluding dead ended and SG compartments). Since the predicted
- stratification is slight, and since the volume of the above-deck regions is significantly greater l than the below-deck open compartments, mixing the above-deck volume with the below-deck l open compartments does not significantly change the above-deck steam concentrations. Thus, the WGOTHIC predictions as the transient calculation passes through 24 hours are essentially similar to the assumption of a homogeneous containment. It is conservative to not include the steam generator compartment steam concentration in the homogeneous calculation.
p It is concluded that WGOTHIC predicts a slight segregation between the above- and below-deck V regions, but the deviation from the homogeneous assumptica is insignificant. Based upon the results of the evaluation, it has been shown that the Evaluation Model adequately bounds the effects of circulation and stratification during the long term phase. 9.3.2.5 Evaluation Model Results , Sensitivities have been performed using the lumped parameter Evaluation Model (Section 4) for several postulated, plausible break locations. An evaluation of the sensitivities leading to selection of a limiting scenario for design basis accident calculations follows. It has been determined that to bound circulation and stratification effects, floors are neglected throughout the transient, and condensation and convection on all heat sinks in the dead-ended compartments are neglected after blowdown. The stratification of steam and air within compartments may reduce heat sink effectiveness. These biases are included in the Evaluation Model used to perform sensitivities. Undissipated Jet Rising in SG East Compartment The postulated, undissipated jet directed up the Steam Generator compartment results in
- /7 increased heat and mass transfer, possibly as high as a factor of [ ]* over the steel shell U surface based on the LST, compared to that using the free convection correlation in the Circulation and Stratification Within Containment cktober 1997 o:W106w 9amm:1tul01497 Reviuon 1
,. . - ---- -,- - - ~ -_ - . . . _ - . _ - - -
9-40 Evaluation hiodel, as discussed in Section 9.3.1.2. To estimate the potential benefit for AP600, the heat transfer coefficient multipliers for the inner surfaces of the clime conductors (that is, only the steel shell mass transfer is enhanced) wer: increased to [ }" times the h Evaluation Model values. The Evaluation Model with the break in the steam generator East compartment was used for the sensitivity cases. The postulated, undissipated jet will only occur until the ADS Stage 4 valves are opened at approximately 1000 seconds. Therefore the containment pressure response is plotted for the first 1000 seconds of the LOCA. The containment pressure sensitivity results are shown in Fib ure 9-45, along with the Evaluation Model results. The results show that the pressure response during the blowdown phase is the same for all cases. This is expected because volume compliance is the dominant pressure mitigator during blowdown (Section 9.3.2.1). Compared to the Evaluation Model results at 1000 seconds, the calculated containment pressure for the [
]" These results show that a substantial benefit in containment pressure is gained when the heat transfer coefficient is increased to account for the forced convection from an undissipated jet. Therefore, this case will be less limiting than the other postulated break scenarios in which the jet is dissipated.
Dissipated Jet Rising in SG East Compartment Another postulated break scenario, the design basis case, is a dissipated jet in the SG East compartment (Volume 107, elevation 100 ft.). Figure 9-46 shows the results of the WGOTHIC LOCA Evaluation Model which includes the circulation and stratification biases. Assuming the break momentum is dissipated in the broken loop steam generator compartment, a maximum containment pressure of 43.9 psig is calculated, which is below the design pressure of 45 psig. The pressure transients for compartments directly connected to the SG East compartments are shown in Figure 9-46A. Figures 9-47 through 9-50 show the circulation pattern predicted by ECOTHIC for this case at different times during the transient. The figures show the Evaluation Model flow path connections for the below-deck volumes, the flow rates and directions, volume steam pressure ratio, and liquid level. Figure 9-51 is a depiction of each of the flow connections to the above-deck volumes, in subsequent figures, total flows through the ceiling of each compartment are shown for simplicity. Flow paths that have been grouped have the same flow direction. Figure 9-47 presents data at 20 seconds which is near the end of blowdown. Flow is forced into all of the below-deck volumes and into the above-deck volumes from the East and West steam generator compartments and the North and South CMT rooms. Figure 9-48 presents data at 1000 seconds which is near the time of maximum pressure and prior to ADS Stage 4 valve actuation. Flow to the dead-ended compartments has stopped. The general circulation pattem is fluid from the break flowing up through the SG East compartment while a steam / air mixture is drawn into and through the SG West compartment, the North and South CMT rooms, and the RCDT cavity into the SG East compartment. At 1500 seconds, Figure 9-49 shows the change in circulation pattem due to the actuation of the ADS Stage 4 valves. The steam releases g flow up through both steam generator compartments while a steam / air mixture is drawn into W
~
Circulation and Stratification Within Containment ociate 1997 o:\3006w.9a.non.1t401497 Revision ' l
l 9-41 and through the CMT rooms and the RCDT cavity. This flow pattem develops less than 2 minutes after ADS Stage 4 activation. Figure 9 50 shows the circulation pattern near 24 hours. The flow rate out of the ADS Stage 4 valves is approximately one foutth of the flow at 1500 seconds. The flow pattem remains out of the SG compartments and into the CMT rooms, however, flow through the RCDT cavity has ceased, due to liquid level rising above the top of the flow path. Plume Rising in CMT Room In Section 9.3.1.3, the LOCA with jet dissipation in the RCDT cavity was postulated. it was _ postulated that the entire buoyant plume rises into the North CMT compartment. The evaluation concluded this scenario was not limiting because of the higher steam concentrations expected in the CMT compartment, which would result in better internal heat sink utilization. Furthermore, the evaluation concluded that the relative steam densities would drive the steam to navigate the bend in the CMT compartment. This would lead to a steam-rich environment for the heat sinks in the south end of the CMT room opposite the stairwell. To confirm that this scenario is not bounding, a EGOTHIC evaluation was performed starting from the Evaluation Model (Section 4). The evaluation assumed a LOCA where the jet plume dissipates and rises into the North CMT compartment. This was simulated by applying the break boundary conditions to the North CMT node (Volume 6, elevation 107 ft.), the only change rr ;de to the Evaluation 1.bdel. The circulation and stratification biases of neglecting floors throughout the O transient and condensation and convection in dead-ended compartments following blowdown were included. The containment pressure results of this evaluation are shown in Figure 9-52. The maximum pressure was calculated to be 43.7 psig. As expected, tlus pressure is below the pavious scenario where momentum is dissipated in the East oteam generator compartment. The circulation pattem predicted by WGOTHIC is shown in Figures 9 53 and 9-F4. Figure 9-53 presents data at 1000 seconds which is near the time of maximum pressure and prior to ADS Stage 4 valve actuation. Compar:d to Figure 9-48 (bretk in SG East compartment), Figure 9 53 shows flow out of the North and South CMT rooms into the above-deck region, while a steam / alt mixture flows down into both SG compartments and up through the RCDT cavity into the 1 : orth CMT room. Figure 9 54, at 1400 seconds, shows the change in flow pattem dee to ADS Stage 4 valve actuation. The flow rates and pattem are similar to those in Figure 9-49, as expected. Figure 9 55 shows the heat sink utiP2ation for this sensitivity case. Aa expected, Figure 9-55 shows a greater CMT room (Volumes 6 and 101) heat sink utilization than that shown in Figure 9-43 for a break in the SG East compartment. Both figures show that the PCS shell is the dominant heat sink at the time of maximum containment pressure and beyond. Plume Rising in RCDT Cavity f . In Section 9.3.1.3i a LOCA with }et dissipation in the RCDT cavity was postulated. This scenario assumed the' break flow splits between the Chil' and steam generator compartments. The Circulation and Stratification Within Containment October 1997 c:\3006whnortitq01497 Revnion1
l 942 I 1 l evaluation concluded that good below-deck heat sink utilization is expected because of the high ! steam concentrations in the CMT and steam generator compartments. A EGOTHIC calculation was performed for this scenario using the Evaluation Model (Section 4). The calculation simulated the flow split by placing the break boundary condition directly in the RCDT cavity [ ]" The circulation and stratification biases were included. The pressure prediction from the evaluation is shown in Figure 9 56. The maximum pressure was calculated to be 43.4 psig. This pressure is below both of the previously discussed sensitivities. The WGOTHIC predicted circulation pattern is shown in Figures 9-57 and 9-58. Figure 9-57 presents data at 1000 seconds which is near the time of maximum pressure and prior to ADS Stage 4 valve actuation. With the break in the RCDT cavity, the bulk flow distribution is based on the path areas and loss coefficients. Consequently, at 1000 seconds, the steam flow from the break goes up through the CMT rooms, while a steam / air mixture flows down through both SG compartments and into the North CMT room and RCDT cavity. Figure 9-58, at 1500 ' seconds, shows the change in flow pattern due to ADS Stage 4 valve actuation. The flow rates and pattern are similar to those in Figure 9-49, as expected. Figure 9-59 shows the heat sink utilization for this sensitivity case. Compared to Figure 9-43 for a break in the SG East compartment, Figure 9 59 shows a small delay in the heat absorption from the SG East compartment and the CMT rooms. The heat absorption from the SG West compartment starts a little sooner in Figure 9 59. The effects are due to the break location differences. Consistent with the other cases, Figure 9 59 shows that the PCS shell is the dominant heat sink at the time of maximum containment pressure and beyond. 9.3.2.6 Evaluation of Drops During a LOCA Drops, or fog particles, are created when the blowdown break source steam velocity is large enough to disperse a frac' ion of the break liquid along with the gas. As discussed in Reference 9.2, Section 4.4.2D and Reference 9.7, Section 7.1, drops will be formed during the LOCA blowdow a phase. For the post-blowdown phases of a LOCA and for the main steamline break (MSLB), there will not be any significant drop formation. The thermal and circulation effects of drops on LOCA containment pressure are examined in Appendix 9.A and summarized below. Drop fall times for various size drops were determined in Appendix 9.A, which only account for the gravitational effects on the drops. Fall 6mes range from seconds to hours depending on the drop size and fall height. This provides an indication that the drops will exist long enough that their effect on containment pressure must be considered. In addition, Appendix 9.A estimated plume entrainment rates for 0 percent and 100 percent of the E:vak liquid converted u drops. The entrainment rates and subsequent circulation time constant for both 0 and 100 percent drops show that a large fraction of the containment volume will be entrained in the plume within a few minutes, which is relatively short compared to the time to reach maximum pressure (at approximately 1200 seconds), and very short compared to long term cooling. A g Circulation and Stratification Within Containment octoter m7 cA30th94.nort1t-101497 Revtskm 1
9-43 relatively large entrainment rate within the above-deck region indicates that the steam density gradients above-deck are not large whether drops exist or not. Therefore, the presence of drops will not significantly affect the general circulation and stratification patterns in the containment atmosphere. Section 5.6 s.iows the results of sensitivity cases to assess the Evaluation Model treatment of the thermal effects of drops with respect to containment pressure. The results that show the Evaluation Model assumption of 50 percent of the break liquid being converted into drops pmvides essentially the same containment maximum calculated pressure as assuming 100 percent of the liquid is converted into drops. The 50 and 100 percent drop fractions are both t more limiting with respect to maximum pressure than assuming none of the break liquid is converted into drops. The formation of drops during the LOCA blowdown phase is a physically real phmomenon which may influence the maximum containment pressure calculated by the Evaluation Model. Drop forvation increases the effective density of the containment atmosphere due to the close coupling between smal! .irops and gas by shear forces, making the post blowdown releases relatively more buoyant. A small percentage (25%) of the blowdown break liquid formed into drops is sufficient to saturate the containment atmosphere, at which point, additional drop density has a minor thermal effect. The Evaluation Model treatment of drops, as described in Section 4.5.2.1, provides a sufficiently bounding calculation for maximum and long term O containment pressure. l ) l
- O Circulation and Stratification Within Containment cetoter 1997 o:\3006wea.non:ltwl01497 Revision 1
9-44 I 9.4 MAIN STEAMLINE BREAK (htSLB) The main steamline transports steam from the steam generators within the containment building O to the turbine generators in the auxiliary building. The main steamline path begins at the top of the steam generator, where it bends 180' and follows a downward path to the CMT room. In the CMT room, the steamline bends 90', crosses through die CMT room, and exits the building through a penetration in the containment shell. Rupture of the main steamline inside containment would release high energy steam into the containment. To ecnfirm the design adequacy of the containment, various MSLB scenarios are examined to develop a conservative model accounting for the effects of circulation and stratification in the containment pressure calculations. 9.4.1 Break Locations . An evaluation of circulation and stratification must allow for the consideration of possible break locations. For the MSLB, two distinct break locations may be poshdated: a break in the steamline above the operating deck or a break in the steamline in the CMT compartment. 9.4.1.1 MSLB Above the Operating Deck An MSLB above the operating deck could occur anywhere in the steamline piping from the top of the steam generator to the operating deck penetration into the CMT compartment. The design basis AP600 MSLB mass and energy releases for containment pressure assume a 2 1.388 ft break (due to integral Gow limiters) at 30 percent power. The MSLB event is characterized by a high energy release of short duration. Reference 9.7, Figure 6 3 shows the calculated Froude numbers for the event compared to Froude numbers calculated for the LST. The high Froude numbers indicate a high kinetic energy source which is expected to drive circulation above and below the jet source elevation. High Froude munbers also indicate that a significant forced convection enhancement to mass transfer occurs during an AP600 MSLB. An examination of releases from smaller sized breaks in main steamlines indicates that the reduction in mass flow is more than offset by the reduction in exit flow area. Therefore, the larger size breaks have the lowest Froude numbers. The double-ended rupture MSLB has the limiting combination of mass and energy release and Froude numbers. 9.4.1.2 MSLB in the CMT Compartment A steamline rupture in the CMT compartment would propel a high momentum steam jet into the CMT room. Since the break is within an enclosed compartment, momentum from the jet would be dissipated by the equipment, walls, floors, and ceilings of the CMT room. The effect Circulation and Stratification Within Containment octotier 1997 c:\3006w-9a.non:ltwl01497 Revision 1
9-45 would create a pressure source in the CMT compartment with the fluid following the path of v resistance through the node network into adjacent compartments and the above-deck volume. The steam source in the CMT compartment will create a steam rich environment for this room which contains many heat sinks. The high steam concentration will result in excellent heat sink utilizatien for this scenario. The MSLB in the CMT compartment case is bounded by the scenario of an MSI.B occurring above the operating deck. While the break above the operating deck does produce substantial circulation, the steam concentrations in the CMT compartment will not approach the steam levels for a break directly within the CMT room. Thus, the MSLB in the CMT compartment is not the bounding scenario. To confirm this conclusion, Section 9.4.3 presents the results of a EGOTHIC analysis for a break in the CMT compartment. As expected, the containment peak pressure is lower for the MSLB in the CMT compartment than for an MSLB above the operating deck. 9.4.2 AP600 MSLB Evaluation Model in creating an appropriate and conservative Evaluation Model,it is necessary to understand how the code handles circulation, to bias the model to produce bounding but reasonably representative results. Investigation of the lumped parameter Evaluation Model (Section 4) has n shown that this noding structure tends to mix upwards from the break elevation. U The lumped parameter calculational bias may be attributed to the use of multiple, relatively large lumped parameter nodes to represent the above-deck region in the Evaluation Model. Lumped parameter formulation uses what may be called a scalar form of the momentum equations, as follows. Here, momentum flow into each volume is parallel to the junction, and the terms perpendicular to the junction are discarded while junction momentum is dissipated within the volume. Momentum orientation is not tracked, and no turning losses are represented. This momentum dissipation is the characteristic of the lumped parameter noding which results in the calculated stratification above/below the jet. With momentum diffused throughout the volume node, the vigorous circulation from the high kinetic energy jet does not occur in the model. Circulation above the jet source in the lumped parameter model is driven by the density head tenns in the momentum equation which cannot drive flow below the source. Thus, lumped parameter noding predicts a steam rich atmosphere above the assumed source elevation, and a steam-deficient atmosphere below this source elevation (simulating stratificction). With an understanding of both the AP600 physics, and lumped parameter model biases, a EGOTHIC representation is constructed which conservatively represents the accident scenario. The high kinetic energy of the MSLD will tend to circulate steam through the above-deck portion of the containment vessel and lead to forced convection conditions for the shell. The lumped
^'s parameter Evaluatica Model, however, calculates a steam-rich region above the injection point (d and an air rich region below this point. Figure 9-60 shows the steam concentration results of a Circulation and Stratification Within Containment october 1997 o:\3006w-9a.non:1b 101497 Revision 1
9-46 EGOTlilC MSLB calculation using the Evaluation Model with the source entering [ l'"' which is just above the operating deck (refer to Section 4.5.2.2). The model predicts a small steam density gradient above-deck, consistent with the expectation of h only small gradients in the AP600, based on 151' data (see Section 9.2.2). Evaluation has shown that the effect on shell mass transfer of even extreme stratification, beyond that expected for the AP600 (see Section 9.3.1.3), is very small. Very little steam penetrates into the below-deck region in the model. Steam access into the below-deck compartments in the model is govemed only by the volume pressurization. As the mass and energy releases pressurize the above-deck region, a steam / air mixture from above-deck is pushed into the belove-deck compartments. The use of the EGOTHIC lumped parameter model, with an injection point just above the operating deck, results in a conservative Evaluation Model for the steam line break as a result of reduced steam access to the below-deck heat sinks. The reduced steam access is due to the momentum dissipation in the model which reduces the calculated circulation to the nodes below the operating deck. The Evaluation Model neglects any heat and nas transfer contribution from forced convection, so above-deck velocity predictions become unimportant. Mass transfer is seen to be underettimated by as much as a factor of [ )("> on the steel shall surface relative to forced convection in the LST. To add an additional conservative blas, the stratificadon heat sink biases developed for LOCA scenarios are also included 9.4.3 MSLB Sensitivity Results Based on an evaluation of circulation and stratification, an MSLB Evaluation Model has been constructed to bound circulation and stratification effects. The limiting MSLB scenario assumes a pipe break above the operating deck. In this scenario, test data indicates that the high kinetic energy source jet induces circulation above and below the jet elevation, including substantial steam penetration into below-deck compartments. The lumped parameter Evaluation Model, that bounds circulation and stratification, places the break source directly above the operating deck [ l'"> This results in a well-circulated upper region with little steam access to the heat sinks below the operating deck. To further bound circulation and stratification effects, stratification heat sink biases developed for LOCA scenarios are included (see Table 91). Figure 9-61 shows the results of the EGOTHIC MSLB Evaluation Model described above. A containment peak pressure of 44.8 psig is calculated, which is below the design pressure of 45 psig, in Section 9.4.1.2, the MSLB in the CMT compartment scenario was evaluated, concluding that increased circulation below the operating deck would reduce the calculated containment pressure. This scenario was determined not to be the limiting scenario, because of the high steam concentrations expected in the CMT compartment. The high steam concentration would result in improved heat removal rates by the heat smks in the CMT compartment. To confirm this hypothesis, a EGOTHIC analysis was performed for a break in the CMT compartment [ ]("> As with the break above the operating deck, LOCA stratification biases were included. Figure 9-62 shows the results of the EGOTHIC calculation. Circulation and Stratification Within Containment cwber 1997 oMXX.w-9a.non:1t>10H97 Revision 1 l
9-47
,c. A containment peak pressure of 43.2 psig is calculated, which is 1.6 psi less than the peak
( pressure for the MSLB above the operating deck. As expected, the Evaluation Model predicts the MSLB above-deck to be the limiting location. s 2 Circulation and Stratificatian Within Containment Octotier 1997 o:\3006w-9a.non:1h101497 Rnision 1 _. __ = -
9-48 I
9.5 CONCLUSION
S A WGOTHIC Evaluation hiodel is used which considers circulation and stratification in the calculation of LOCA and hiSLB containment pressures and temperatures. The effects of circulation and stratification on the calculated containment pressure have been examined, and biases have been defined for the Evaluation hiodel. The Evaluation hiodel input deck and specific biases are described in Section 4. In addition, break locations have been examined for LOCA and MSLB to determine the limiting location for each transient with respect to calculated containment pressure. The circulation and stratification evaluations performed re Jt in the following biases which have been incorporated into the Evaluation Model for the LOvA analysis:
- Heat and mass transfer from floors of compartments and the operating deck have been removed to bound the potential reduction in heat transfer due to stratification. Refer to Sections 9.3.1.1,9.3.1.3, and Appendix 9.B.
- Condensation and convective heat transfer in dead-ended compartments are turned off after 30 seconds (i.e., after blowdown) to bound the potential reduction in heat transfer due to stratification. The basis for this bias is provided in Section 9.3.2.1.
- The lumped parameter Evaluation Model considers only free convection for internal heat sinks and shell surfaces and, therefore, conservatively neglects the increase in mass transfer to the containment steel shell due to forced convection during blowdown. Refer to Section 9.3.2.1.
Ranges of LOCA break locations and jet directions were evaluated to determine the limiting case with respect to containment pressure. The limiting scenario is the DECLG break in the East steam generator compartment with the jet momentum locally dissipated. This break location is included in the Evaluation Model desenbed in Section 4. Other break locations, or jet directions, result in increased heat sink utilization which results in lower calculated containment pressures. Based on the results presented in Section 9.3.5.2, the calculated maximum LOCA containment pressure from a dissipated jet is not very sensitive to the break location since internal heat sinks
" reach maximum effectiveness" well before the time of maximum pressure.
For the MSLB, the circulation and stratification evaluations performed result in the following biases which have been incorporated into the MSLB Evaluation Model:
- The break is placed in a node at the operating deck level to minimize circulation and steam access to belowdeck heat sinks, which bounds the potential reduction in heat transfer in below-deck compartments due to stratification. This is discussed in Section 9.4.2.
Circulation and Stratification Within Containment october 1997 1 oA300t w.9a non.b101497 Revision 1
9-49
- The lumped parameter Evaluation Model considers only free convection for internal heat O sinks and shell surfaces and, therefore, conservatively neglects the increase in mass transfer to the containment steel shell due to forced convection during the entire transient. Refer to Section 9.4.2.
The above listed LOCA biases (relative to floors and dead ended compartments) have been included in the MSLB Evaluation Model to further conservatively bound potential reductions in heat transfer due to stratification. Refer to Section 9.4.3. Based on the routing of the steamline pipe, two MSLB locations were evaluated; a break above the operating deck and a break in the CMT room. As discussed in Section 9.4.3, the break above-deck resulted in the higher calculated containment pressure. The break in the Chfr room had triereased heat sink utilization in the CMT room which resulted in the lower calculated containment pressure. The above biases are incorporated into the Evaluation Model as described in Section 4.2, subsections entitled "Special Modeling Aasumptions." Therefore, the effects of circulation and stratification have been conservatively bounded in the WGOTHIC containment pressure calculations. O O Circulation and Stratification Within Containment october 1907 c:\3006w 9a.non:11w101497 Reviskm 1
9 50
9.6 REFERENCES
9.1 WCAP-14382, ")yGOTHIC Code Description and Validation," May 1995. 9.2 WCAP-14812, " Accident Specification and Phenomena Evaluation for AP600 Passive Containment Cooling System," Revision 1, June 1997. 9.3 NTD-NRC-95-4563, " GOTHIC Version 4.0 Documentation, Enclosure 2: Technical Manual," 3eptember 21,1995, 9.4 NTD NRC-95-4563, " GOTHIC Version 4.0 Documentation, Enclosure 3: User Mamel," September 21,1995. 9.5 WCAP-13566, "AP6001/8th Large Scale Passive Containment Cooling System Heat Transfer Test Baseline Data Report," October 1992. 9.6 WCAP-14135, " Final Data Report for PCS Large-Scale Tests, Phase 2 and Phase 3," Revision 1, April 1997. 9.7 WCAP-14845, " Scaling Analysis for AP600 Containment Pressure During Design Basis Accidents," Revision 2, June 1997. 9.8 Wolf, L., Gavrilas, M., Mun, K.,1996, " Overview of Experimental Results for Long-Term, Large-Scale Natural Circulations in LWR-containments after Large LOCAs," DOE - Project, Order Number: DE-AP07-961D10765, University of Maryland at College Park, July 1996. 9.9 WCAP-14326, " Experimental Basis for the AP600 Containment Vessel Heat and Mass Transfer Correlations," Revision 1, May 1997. 9.10 Jaluria, Y. "Suoyancy Driven Wall Flows in Enclosure Fires," Twenty first Symposium (Inte;cational) on Combustion, The Combustion Institute, 151-157 (1986). 9.11 Goldman, D., Jaluria, Y., "Effect of Opposing buoyancy on the Flow in Free and Wall Jets," Journal of Thtid Mechanics, 166, 41 56 (1986). 9.12 Jaluria, Y., Cooper, L.Y., " Negatively Buoyant Wall Flows Generated in Enclosure Fires," Progress in Energy and Combustion Science,15,159-182 (1969). 9.13 Fisher, K., Schall, M., and Wolf, M,, " Simulations of GOTHIC Large-Scale Containment Experiments," published by Battelle Ingenieurtechnik GmbH, October 1995. , Circulation and Stratification Within Containment octoter 1997 c:\3006w 9a.non:ll>101497 Revision 1
i i 9-51 4 9.14 Ofstun, R.P., Woodcock, J., Paulse , D.L, " Westinghouse - GOTHIC Modeling of NUPEC's Hydrogen Mixing and P ibution Test M-4 3," Third International Conference on Containment Design and C .t.on, October 1994, Toronto, Canada. , 9.15 Peterson, P.F.," Scaling and Analys of Mixing in Large Stratified Volumes," International Journal of Heat and Mass Transfer, V<d. 37, Suppl.1, pp 97 '06,1994. O i I i i l l I-
,o e
! Circulation and Stratification Within Containment octoter 1997 c:0006w-9a.non:1b 101497 Revision 1 i
9-52 (a,c) i O Figure 9-1 Measured Steam Concentrations for LST O Circulation and Stratification Within Containment Octolwr 1997 o \3006w.9a.non:1b-101497 Reviskm 1
9 53
~
(a,b) O , Figure 9-2 O- LST with Diffuser Under Steam Generator - Steam Flow 0.11-0.17 lb/ see - Internal Fluid Temperature Group 1 Circulation and Stratification Within Containment Octotwr 1997 n\30%v-9a.rmlb101497 Reviskm 1
~ ' ' '- -
9-54 i (a,b)O O Figure 9-3 LST with Diffuser Under Steam Generator - Steam Flow 0.110.17 lb/ see - Saturation Temperature - Group 1 Circulation and Stratification Within Containment o tober 1997 c:\30thw-9a.non:lb101497 Revision 1
9 55
~ ~
(a,b) Figure 9-4 LST with Diffuser Under Steam Generator - Steam Flow 0.11-0.17 IW see - k Internal Steam Pressure Ratio Group 1 Circulation and Stratification Within Containment octote 1997 a\3006w 9a.norelb 101497 Revision 1
9 56
~(a,b)O O
Figure 9 5 LST with Diffuser Under Steam Generator - Steam Flow 0.11-0.17 lb/ sec . Internal Fluid Temperature - Group 2 Circulation and Stratification Within Containment October 1997 o:\3006w.9a.non;1b101497 Revision 1
I 9 57
- ~
(a,b) O Figure 9-6 LST with
- ser Under Steam Generator . Steam Flow 0.110.17 IW see .
_ Saturation Temperature . Group 2 - Circulation and Stratification Within Containment October 1997
'oA3006w-9a.non:lt>101497 . Reviskm 1
9 58
~ (a,b) O O
Figure 9 7 LST with Diffuser Under Steam Generator Steam Flow 0.110.17 lb/ ecc - Internal Steam Pressure Ratio Group 2 Circulation and Stratification Within Containment october 1979 c;\30lbw 9a.non:1t>101497 Revision 1
9 59
~ '
(a,b) 1
..A U
Figure 9 LST with Diffuser Under Steam Generator - Steam Flow 0.27 0.36 IS'see - s htternal Fluid Temperature Circulation and Stratification Within Containment October 1997
- o:\3006w 9a.norulb 101497 Revision 1
94D
~ (a,b)
G C G Figure 9-9 LST with Diffuser Under Steam Generator - Steam Flow 0.27-0.36 lb/ sec - g Saturation Temperature W Circulation and Stratdication Within Containment October 1997 c:\3006w-9a.nortib-101497 Revision 1
t i 9-61 (a,b) t s Figure 9-10 LSF with Diftuser Under Steam Generator - Steam Flow 0.27-0.36 IW sec - Internal Steam Fressure Ratio Circulation and Stratification Within Containment October 1997 o:\3006w-9a.non:1b-101497 Revision 1
9-62 (a,b)O O Figure 9-11 LST with Diffuser Under Steam Generator - Steam Flow 0.49-0.62 lb/ sec - Internal Fluid Temperature - Group 1 Circulation and Stratification Within Containr;.ent october 1997 o:\3006w-9a.non:1b.101497 Revision 1
63
~
(a,b) i , (- 1 Figure 9-12 : LST with Diffuser Under Steam Generator - Steam Flow 0.49-0.26 lb/ sec - Saturation Temperature - Group 1
- Circulation and Stratification Within Containment October 1997 0:GPw-90.nortib-101497 Revhion 1
9-64
~
l - (a,b)O i 1 9 Figure 9-13 LST with Diffuser Ur. der Steam Generator - Steam Flow 0.49-0.62 lb/ see - Internal Steam Pressure Ratio - Group 1 G l!! Circulation and Stratification Within Containmeent octatwr 1997 c:\3006w-9a.non:1b-101497 Revision 1
945-
~
(a,b) - U
-Q s
4 Figure 944 LST with Diffuser Under Steam Generator - Steam Flow 0.49-0.62 lbf sec - Internal Fluid Temperature - Group 2 Circulation and Stratification Within Containment October 1997 o:\3006weaawm:1b-101497 Revision 1
9-66 i -
~ (a,b)O O
4+ Figure 9-15 LST with Diffuser Under Steam Generator - Steam Flow 0.49-0.62 lb/ sec - Saturation Temperature - Group 2 i Circulation and Stratification Within Containment october 1997 c:\3006w-9a.non:lb-101497 Revision 1
9-67
~
(a,b) ( L s \ Figure 9-16 IST with Diffuser Under Steam Generator - Steam Flow 0.49-0.62 lb/ see -
\
Internal Steam Pressure Ratio - Group 2 Circulation and Stratification Within Containment October 1997 o:\3006w-9a.nortit>101497 - Revision 1 y .. . .. .
9-68 (a,b) 0 O Figure 917 LST with Diffuser Under Steam Generator - Steam Flow 0.76-0.84 lb/ see - Internal Fluid Temperature Circulation and St atification Within Containment october 1997 o:\3006w-9a.non:1t>101497 Revision 1
- 9-69 (a,b) l' i
r t
- Figure 948 LST with Diffuser Under Steam Generator - Steam Flow 0.76 0.84 It/ see -
Saturation Te:nperature Circulation and Stratification Within Containment october 1997 o:\3006-9al.nortit>-101497 Revision 1
9-70 I 1
- ~
(a,b) O' O Figure 9-19 LST with Diffuser Under Steam Generator - Steam Flow 0.76-0.84 lb/ see - Internal Steam Pressure Ratio Circulation and Stratification Within Containment october 1997 o;\3006-9al.rvartIb-101497 Revision 1
L 9-71. . l la,b) o i l Figure 9-20 . LST with Diffuser Under Steam Generator - Steam Flow 1.10-1.20 lb/ see -
'( Internal Fluid Temperature Circuli /_40n and Stratification Within Containment october 1997 o:\N)06-9al.non:1b 101497 Revision 1
- h. ,
a
9-72 (a,b) O O s Figure 9-21 LST with Diffuser Under Steam Generator - Steam Flow 1.10-1.20 lb/ sec - Saturation Temperature Circulation and Stratification Within Containment october 1997 o:\3006-9al.non:ltr101497 Revision 1
9 - (a,b)
)
i O - Figure 9-22 LST with Diffuser Under Steam Generator - Steam Flow 1.10-1.20 IW sec - I Internal Steam Pressure Ratio Circulation and Stratification Within Containment october 1997 c:\30064al.non:1b-101497 Revision 1
9-74
- (a,b) 9 0
Figure 9-23 LST with Diffuser Under Steam Generator - Steam Flow 1.541.68 lb/ sec - Internal Fluid Temperature Circulation and Stratification YGthin Containment October 1997 o:\3006-9al.non:1b-101497 Revuon I
9-75
~ ~
(a,b) l< O Figure 9-24 LST with Dihser Under Steam Generator - Steam Flow 1.54-1.68 lb/ see -
\ Saturation Temperature Circulation and Stratification Within Containment October 1997 o:\30064al.non:1b-101497 Revision 1 l
9-76
~
l (a,b) O O Figure 9-25 LST with Diffuser Under Steam Generatar - Steam Flow 1.541.68 lb/ sec - Internal Steam Fressure Ratio . Circulation and Stratification Within Containment October 1997 o:\3006&1.non:1b-101497 Revision 1
i. 9-77
~
lT (a,b) t O 1 l
/ - Figure 9 LST with Diffuser Up 6 Feet - Steam Flow 0.76 Ac 1.68 lb/ see - Internal Fluid Temperature Circulation and Stratification Within Containment october 1997 o:\3006-9al.non:1b-101497 . . Revkion 1
9 78 (a,b) O O Figure 9-27 LST with Diffuser Up 6 Feet - Steam Flow 0.76 & 1.68 lb/ see - Saturation Temperature Circulation and Stratification Within Containment october 1997 o:\3006-9al.non:lt>101497 . Revision I
9 -
~
(a,b) O
\
e. c- Figure 9-28 - LST with Diffuser Up 6 Feet - Steam Flow 0.76 ac 1,68 IW see - Internal Steam Pressure Ratio Circulation and Stratification Within Containment october 1997 o:\3006-9al.non:1b-101497 Revis6on 1
9-80 (a,b) O O i i l Figure 9-29 LST with Steam injection: 3 inch Pipe - Steam Flow 0.76 - 0.95 lb/ sec - Internal Fluid Temperature x Circulation and Stratification Within Containment October 1997
' Revision 1 o:\3006-9al.non:ltwl01497
9-81 -'
~
(a,b)
- Figure 9-30 LST with Steam injection: 3 Inch Pipe - Steam How 0.76 - 0.95 lb/ see -
5 Saturation Temperature Circulation and Stratificatior Within Containment October 1997 c:\3006-9al.rorulb-101497 - Revision 1 s
9-82 (a,b) O 6 i O Figure 9-31 LST with Steam injection: 3 Ineh Pipe - Steam Flow 0.76 - 0.95 lb/ see - Internal Steam Pressure Ratio Circulation and Stratification Within Containment october 1997
, o \300 Mal.non:1b.101497 Revision 1
.9-83.
l
' ~ ~
(a,b) . f}
\
Figure 9-32 LST with Seam Injection: 3 inch Pipe - Seam Flow 1.25 - 1.31 lbf see - { Internal Fluid T. mperature Circulation and Stratification Within Containment october 1997 o:\3006-941non:1b-101497 - Revision 1
l 9-84
- (a,b)
O O Figure 9-33 LST with Steam injection: 3 Inch Pipe - Sseam Flow 1.25 - 1.31 lb/ sec - Saturation Temperature Circulation and Stratification Within Containment october 1997 l o;\300&9al.non:ll>.101497 Revision 1
9-85
~ ~
(a,b) O O O Figure 9-34 LST with Steam injection: 3 Inch Pipe - Steam Flow 1.25 1.31 lb/ see - Internal Steam Pressure Ratio Circulation and Stratification Within Containment October 1997
' o:\3006 9al.nort1b-101497 Revision 1
9-86 O .k
- 3*7.",'.*IN
s , .. . .
/
o,...,......,o a ~ a-' / I I
/
ocea.., .a si... 7 / -\
/ v w.< a aie oe.a.4,, !
7
/ \ >/ q- " t,y k ,X e o g.N w ;
e
./:. ,. ms I +
C v
;;*s , cus l .,,
l 1 A N I l x I -
-s \ '"l N i
o d i O Figure 9-35 CMT Compartment Layout Circulation and Stratification Within Containment October 1997 o:\3006-9al.nortib101497 Revision 1
+87 -
(a,c) O Figvte 9 36 Simplified AP600 Containment Diagram
- - - Circulation and Stratincation Within Containment ---
october 1997 - o:\30%9a2.nortib 101497 Rntkm 1 ' n ,
9-88 (a,c) O l l l l l Figure 9-37 HGOTIIIC Calculated LOCA Blowdown Steam Pressure Ratio for Jet Momentum Dissipated in SG East Compartment Circulation and Stratification Within Containment october 1997 o;\XKW942.non:1b 101497 Revision 1
, , ': ; iii ' ! : : i . .,4 i J , i ! ! !, . ! ji a : 4 .
( 0 - oC
\i -
3r 0c 0u
&l y 9at do na i
g F i - o u I r t a bc n r e R .
, - i .
1 9-0S 1 t 3 _ y . 4 r 8 9a 7t f i 4 . i , c - a , t CH o S i o eG !i!l .l n f fO i W cT m. i t h i n s iH e ni t fCa c N
/ - .- C o /
o r l ( n c 8 s R y t Lu l
- c. i a Ola ,
n Cte ) d - m Ad e e n M J n eA J !;. t i t t P M6 0 s o0 s . mC e / - e o d J R ( n n / t ut a min Dm * ,
!}il i ie ,
ss n , pi t . P a r ' i1+ e e t ds s , iu ,
, nre S
G-S Ee a n s s t t i i
* . _ ~ _
- - - : ~
Cv i _ 4 ot my , ,3 el "o I pt o . e8* . w o a r t L ) _ Rt o c mo es s e t v n ie s t y
- i r o1 n9 9 9-1 7 8 .
9
,+
w
9 90 l (a,c) l l i e Figure 9 39 lyGOTIIIC Calculated Flow Pattern - Sensitivity to Loss Coefficients for LOCA Jet Momentum Dissipated in SG East Compartment at 20 Seconds Circulation and Stratification Witrin Containment octoter 1997 o:\3006-9a2.nortitul01497 Revision 1
9 91 (a,c) O O Figure 9-40 ECOTHIC Calculated Flow Pattern - Sensitivity to Loss Coefficients for LOCA Jet Momentum Dissipated in SG East Comp. at 1000 Seconds Circulation and Stratification Within Containment _ _ october 1997 c:\30%9a2.norrib-101497 Revision 1
9-92 (a,c) O Figure 9-41 }YGOTillC Calculated Flow Pattern . Sensitivity to Loss Coefficients for LOCA Jet Momentum Dissipated in SG East Cump, at 1550 Seconds Circulation and Stratification Withm Containment oAXxw9a2.mw1b 101497 N*' .,]
_. ...____....___.--..._...___..._._-___._.___m..._.---. _ . _ _ . _ . - _ __ i" 9-93 i e (a,c) t i 4 i i i
}
} i i, i, 1 1 i I , i 4 ? 1 1 i i 4 5 i' Figure 9-42 HGOTHIC Calculated Flow Pattem Sensitivity to Loss Coefficients for LOCA Jet Momentum Dissipated in SG East Comp, at 80050 Seconds Circulation and Stratification Within Containment October 1997
- cx\300&9a2.norutt*101497 Revhkm 1 1' , , . . . . . . . . . , . _ . . . . . _ . ..._...._.m.. - -, - , _ , . . - . - . . - . - . . . _ .,-.,m..,
! . ,I ; j s < i! !t t :!! - 'i;:;: "{ ' ;t I -
ll ; ,!. i. ei
$ToS . ::3EA 0
5 5 4 0 4 5 3 0 3 5 2 0 2 5 1 0 1 5 O 0 0 O 0,
~
{: - j_ ~ n: -
. . 0 0 -- 1 - - a - n _
- As i .
- - Qs ._ c . e . 0 _ w 0 e_
, - i 0, dae " 0 Dp- .
1 .
% :: ' m._
b m.
~
- A v_
- +
s . . e . Ts e --
- Ne X - - e 0 C. .
S4 .
, 0 v .
0, _f 1
~ % E M- ,, )
( c e s We C. T r s. U s.
' , e. ,
e C R v.
,t :
i m
, T _ - y p s**
f i 0 0 1 T Ms... C. d
- g. O
,[,, . Nv . +-
- y. _
^
s e
,e,- (/ - s ma . ^ - - e _
s v , W s. _ g \ < _. , G. . v, N i, , S .
. a vi _. " 0 x e'2 -
r i
' 1 r e
g,g - ma a E4 s
* . ' ce 2' , - , sm o d - v 9 o,, ;, ,'I - ; '..- / ,
- p , - " 3
.es .'e l ~ - ;_ ; - _3 _E - :$ - <:E : ~Iht 1 .. a.
0 0 0 0 0 ) .. s. s. 0 0 0 0 0 0 0 0 C2 0 0 0 0 0 0 0 0 0 0 0 0 4. i 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 5 0 5 0 5 0 5 0 5 0 5 a. s.. 7 5 2 0 7 5 2 0 7 5 2 5 2 2 2 2 1 1 1 1 2 ( e.a e v o 7Iiy t 3.c._nr %*I 3o. cS&?,a$.ez ,gc3cS*C-O- 13 ,n" ls od=nngBE$ $no}g =: !lse! g:
- On 2 IiaEgEiEi=Ep*9 sg-na3E=E.a5.=ng8ER595{B-9%Kgp E 9{cs~ 2 .
9-95 , (a,c) O Figure 9-44 HGOTHIC Calculated AP600 Containment Steam Pressure Ratio for LOCA O. . Jet Momentum Dissipated in SG East Compartment Circulation and Stratification Within Containment october 1997 o:\3006 9ainon:1b-101497._. -- Revision 1 -
9 96 O (a,c) O Figure 9-45 }YGOTillC Calculated AP600 Cont. Pressure . Sensitivity to 11 eat Transfer Coefficient for Study of Undissipated Jet Effects During a LOCA 1 oci der IW7 Circulation and Stratification Within Contairunent o;\3(X69a2 non:1b 101497 Revision 1
, i
( !. , ! i , i ; !! ! ! [ (/] O cC
\i 3r 0c %u 9a l
t 2 i 2 o nn o F g i S I s t a u r
\
,. in e T bd 1 9 0S 1 t 4 r 4 6
*m 9 a f 7t i I f
i c l a t i o D}Y i I $ n s s G pTiO f . W a f i t h eH t d I ne d . i n C 1 C o i nCa Sl ( i i / n t a Gcu S s ne i s i l i n Ea at d
)
m s e l e t d e i n C n ae J t oA i mP 6 s i p 0 s I a 0 r s j t C mo J 2 . d l
>;l e n h
l nt - t a i I n I . m ee i _- e _. n t t I f j r e s co ' s ' _ u r e-L e : i O t C ! A l J e t
- o - ? : ~
4 .- M o I c8 e$ eg eg g t o m Rt e o c e n 3u-ub s r i e t u _ o1 n9 9 m 9- _- 9 17 7 _
h9 b fE 45 As m Higher Pressure due to ga g Liquid Elevaban in
's ;il 7" 40 .gm I / Lowersth.ui g n;- s 35 1 E ?!:E 1 o ao -
h 5S 30 A l a= cm g '/ s n b En W
\-
n
-a. 25 ~
n e- n xf s o O *-- e : V 2~ ng p o - {E yg> g 20 _ L c - E : <! cusoi.r&d Gas Pressure i? n8 15 i g En ' 7~.
/ $E 10 a !E
- 9. '
, " ,s s 5?l - -- -a 5 - = ?n & ,
2? ;" g- , 0 . . . ..., . . . ..., .
. ...;i .
w 0 0.1 1 10 100 1,000 10,000 100,000 E'? ng a Time (sec) n c, o3 gm so itam a-,, sc iten twer sc wemam er sc wen tw am c-iry marrc.*y v s w4 vesimme les vdese s vdumme a veheme nez 7g vd.me ist 2g ........_ n-Q 32a a I2
-~
O O O
9-99 r-( (a,c) em s (^g Figure 9-47 lYGOTHIC Calculated Flow Pattern LOCA Jet Momentum Dissipated in ( ,) SG Gast Compartment at 20 Seconds Circulation and Stratification Within Containment october 1997 n.\30lW9a2.nortib-101497 Revision 1 1
9 100 (a,c) O Figure 948 lVGOTHIC Calculated Flow Pattern - LOCA Jet Momentum Dissipated in SG East Compartment at 1000 Seconds Circulation and Stratification Within Containment & " o;\30tW9a2.rm1tr101497 _=
,v ~
l 9 101 l (a,c) O t f Figure 9-49 EGOTHIC Calculated Flow Pattern - LOCA Jet Momentum Dissipated in
\ SG East Compartment at 1500 Seconds - - Circulation and Stratification Within Containment Octoter 1997 o \30lW962.mth101497 Revis6on 1 . .__. - ==
g - _____.m,___._____
9 102 (a,c) O L. 11gure 9-50 WGOTHIC Calculated Flow Pattern - LOCA Jet Momentum Dissipated in SG East Compartment at 8000 Seconds Circulation and Stratifice tion Within Containment N M 1997 o:\30069a2.non:1b-101497 Rmision 1
9 103 (a,c) O Figure 9-51 O Details of.WGOTIllC Flow Paths to Above-Deck Region from CMT, Refueling Canal, and IRWST Circulation and Stratification Within Containment october 1997 o:\3006-942.ravitwl01497 Revision 1 v n ar er ' n s ,w ww , rw - -r ' - - < - , - wr-a.m-,- -s-<,
I l 9-104
- f. G / l : ,- 8 k' ) ;i 4 , = E
- er m
5 g S
~
I
-8 9
C' . i
~
N
\ [S N
N
,N,,,, ,
S S S E E E E E S (61sd) eJnsseJd Figure 9-52 lyGOTHIC Calculated AP600 Containment Pressure . LOCA Plume Rising . Into CMT Room Circulation and Stratification Within Containment octotwr 1997 o \30049a2.nortit> 101497 Revuion 1
9 105 (a,c) O O Figure 9-54 liGOTHic Calculated Flow Pattern . LOCA Plume Rising into CMT Room at 1000 Seconds Circulation and Stratification Within Contairunent october 1997 c:\3006-9a2.nortit>101497 Rntion 1 _. _ _ _ =
9 106 (a,c) O Figure 9 54 lyGOTHIC Calculated Flow Pattern LOCA Plume Rising into CMT Room at 1400 Seconds Circulation and Stratification Within Containment OctotEI1E o:\3006 942.nort1b-101497 Reykdon 1
l ,li' ,! I llll
?sy O
P _5.g".m_# EL 0 5 0 5 0 5 0 5 0 0 5 4 4 3 3 2 2 1 1 0 0 0,
- : : 24~ : ~ _~
=- - 0 3 - 0 1
s l- nhe -- L iS S Q_F C - 0 -
- = 0 - ' i 0, ke ce - ' e . 0 1
DS- s - e
. e vm - - . ou - - bl s - AoV -
TI M
- Mnem -- - 0 C.
So l u -
, 0 V -
s - 0, -
. 1 - s,' ) y -
c mtu-
- t i -
e
- ~
y - 1 ( s e r Cne - r r I l CoV - R m u -
~ f.. h O - ~ ~ -
0 T T 6e - Mm m -o s
%a'[= , 0 C. oi -
1 Mv - s *- . E 4 N . S
. .s' I - ,4 g( -
ts& e W5se - - 9 3 g'
,~ =. - -
Gm S e h Ve - s \ - et
*S \
s x "*
- -s . 0 1 7 8
1 q m& a s N
-[. /
E 4s Cn S e e I . h r . e y / i v
- I ./"f' f /
g' #
, s_ 3
_E
. : ::~ :_r.: f - ^ _ 1
_~ : 9 - ' ~ 1 &
.3 0 0 0 0 0 0 0 0 0 0 0 0 ) 3 0 0 0 0 0 0 0 0 0 0 0 0 C2 t 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 0 l as 0, dnsme n 5 7
0 5 5 2 0 0 5 7 0 5 5 2 0 0 5 0 5' 5 E e 7 5 2 2 deVa l 2 2 2 2 1 1 1 1 ( e D
?N 3$. .E.c.#liE x8 Cowo.'j4,.gI a:h2c5 ' c-s O :
2m5* ,.E @cE5mhe;;*a2*8 9"e.=g3 Es$EE5 E
- nK[*i mFsjE%2e Foo ny. 5 iE=f 5$ n8 *gd$ .
, n- Ea LyEf =E 5 9EN ss
= .s - 'r ' i
ll I' lli1 ii I8 e 0 0
, 0, . 0 0
1 N . 0 0 0, 0 1 1 _ 0
, 0 . 0, . 1 ) . c e
( s e m
- j i 0 0
i T e -
- 1 0 . 1 - _ ~- : - - - _ - -
1 5 0 5 0 s 0 5 0 5 0 4 4 3 3 2 2 1 1
-2E g.8g .
m$ Z. mo0$e nkir;a8 {e ga 'ja; yF n% . 0 sn s n0gei eon {ag 2 5~= e O!- n E a w., = Eb$r n87lg ! I yu
$- gi7s. x 8 .-
lli' 1li
9-109 O -
<.o O
O Figure 9 57 .WGOTHIC Calculated Flow Pattern LOCA Plu.ne Rising into CMT Room and SG Compartments at 1000 Seconds Circulation and Stratification Within Containment october 1997 a\3004942.non:1b-101497 g% g
9 110 . i (8,C) O Figure 9 58 JYGOTillC Calculated Flow Pattern LOCA Plume Rising into CMT Room and SG Compartments at 1500 Seconds Circulation and Stratification Within Containment (ktot,er1997 o \3006 9a2.nort1b.101497 Revbion )
1ll;i1iIlI' l c$~ DO . Rg 2:eEE s 0 5 0 5 0 5 0 5 0 0 5 4 4 3 3 2 2 1 1 g 0 0 0, _: ~~ -~ - i -
. ~
0
. 0 5~ -
1
- ae nh _
i_S Q5C F .
~_ 0 . - 0 9 .
0, k
. c4 e . 0 . D. se$ _
1 e v m ... _- . u . 1\ bl
- w _- .
Ao v _
- ' 4 .
T10 Mers _A - . _ - C. el . 0 S Va ... 0
* - i 0, .
1
) .
w ty 2 c i8 v . e o1 Cme . s T mu . g ( e Dl Cvo . R I
~ ?- i m ~ t, T T 6 . ~ ' 0 Mn . e . 01 C.eeh .g '
- o%,*=.'
g . MV t s .
,e, I + , u gA.N . t s&
e W5s
~-
i s . Cn e
\ S e - .- h .
rb 9.- i Ve .
\ . - , 0 N
pc' - 1 7 e l
@ TG = -
ts &
- a E 4s t' f Gn e =
e
- S e g' - ' .
h e
. - f j_/fc. .
V
-( . s t
u3 ei ' I r u0 _: - :i '- ' ~- _ : .- : E I
/'
1 s 1 tr a& p3 0 0 0 0 0 0 0 0 0 0 0 ) m3 0 0 o 0 0 0 0 0 0 0 0 0 0 0 C 32 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 de s 0, d e 5 0 5 0 5 0 5 0 5 0 5 mn 5 E iss 7 5 2 0 7 5 2 0 7 5 2 d e 2 2 2 2 1 1 1 1 2 ( av e D 3 . e.a $
- c 5 a $ x8 c S cw E Q g r g o g 5c j c--
2 5 g*,4* g!o!5 E s"0::i E >'$ n*4 9l[ $g2. WB3~ e3e tog > DV( 5 rE E Fi ga E ng?29gw O g gae $ n g. #[ n855B1 - Eh=VNfsE 9g4 s~ o
- h5' -
! llll
9 112 (a,c) 9 Figure 9-60 lVGOTillC Calculated AP600 Containment Steam Pressure Ratio for MSLB Abo'. e-Deck Circulation and Stratification Within Containment october tw7 ot\'MX4%2.nortit 101497 Revision 1
~O O O se .l2 ng =
Q 1 MSLB above Operating Deck sw St
" Ei g-50 -
E. OM : 3o ^45-i
!P5! .m .g "y p ,40--
a i a : E E 35 --
!! 2-E >
m a> 3 0 -- I s_. E C 25 - : - g m
!E c) 20 - - *? 6 :
IC a 15 - : 3 - t f
' 't ' t i1 ' t t if f i t i f f f f f I f f I 10 ' ' ' '
p 0 1000 2000 3000, 4000 5000
. f > Time ( s e e 'I ?
x-f is ao sw
-~
I l!1 i:lI ll)il)ll;l l'
- =
0 O 0 0 t ' 5 n ' e ' 0 m ' 0 0 t 4 a ' p ' 0
\) ! 0 c m '
0 3 e o s
/(
C ' e T 0 0 0 .i m O 2 M ' T C ' n ' 0 0 0 i ' 1 B ' L '
.: 2: : $ - - ~ - : 5 : : - - . ~
S 0 0 5 0 5 0 5 0 5 0 M 5 4 D 4 3 3 2 2 1 1
^ * - uc-2 o s.. .
3 m"oul c ni .h E8 1
- den nIEa. >!8 nae!E 7l,o p!Cs9:i=83 : O nga~5;*EEga8 89s- n E*' b ,,E n yeulI?* EDAc' 5~
5 - l\Illl1!
L. Appendix 9.A e:O Thermal and Circulation Effects of Drops , Dt ring a LOCA ,
\
T E O m O
^
The; mal and Circulation Effects of Drops During a LOCA october 1997 o:\3006e;,pa.non:1b-101M Revision 1
t
~ 9.A-1 TABLE OF CON 111NTS A
9.A THERMAL AND CIRCULATION EFFECIS OF DROPS DURING A LOCA . . . . 9.A-2 LIST OF TABLES 9.A-1 Estimated Drop Fall Times . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ~'2 LO O Thermal and Circulathn Effects of tho[ ur'ing a LOCA october 1997 o:\3006appa.nort1b101497 Revision 1
9.A-2 _ 9.A TIIERMAL AND CIRCULATION EFFECTS OF DROPS DURING A LOCA Drops, or fog particles, are created when the blowdown break source steam velocity is large 9 enough to disperse a fraction of the break liquid along with the gas. As discussed in Section 4.4.2D of Reference 9.A.1 ano %ction 7.1 of Reference 9.A.2, drops will be formed during - the LOCA blowdown phase. For the po. *-blowdown phases of a LOCA and for the MSLB, there will not be any significant drop formatica. The thermal and circulation effects of drops on LOCA containment pressure are examined in this section. The limiting DBA analysis LOCA is a DECLG break. The source flow from the reactor side of the break has more enarby than the source flow from the steam generator side of the break, so more drops are expected from the reactor side. During blowdown, a range of drop sizes will be produced. The percentage of liquid converted to drops will also be wi+.hin some range, the theoretical limits being 0 and 100 percent, although it is anticipated that a significant fraction of the liquid will form drops in AP600. Meny factors affect the length of time that the drops will be present in the atmosphere, such as shear coupling to the moving gas, coalescence, de-entrainment at walls and oil.er surfaces, and the drop size (affecting its fall time). To estimate the fall time for various size drops, a simple calculation was performed which only accounts for the gravitational effects en the drops. Using the terminal velocitv versus drop diameter information in Section 7.4 of Reference 9.A.3, fall times range from seconds to hours dependte.g on thc drop size und fall height. Table 9.A-1 shows estimated fall times for drops with diameters of 0.001,0.01, and.1 inches. This provides an indication that the drops will exist long enough that their effect on containment pressure must be considered. Table 9.A-1 Estimate:1 Drop Fall Times unmuummum===== = . raamm Terminal Fall Time (sec) Drop Size (in) Velocity (ft/ sec) 30 ft 100 ft 0.001 .08 375 1250 0.01 8 3.8 12.5 0.1 20 1.5 5 l s O Thermal and Circulation Effects of Drops During a LOCA october 1997 e\3006appa.non:Ib-101497 Revision 1
- 9.A-3 l hrmal Effects O The drops flash when they enter the containment atmosphere, reaching saturation very quickly.
Section 7.1 of Reference 9.A.2 estimates 3.5 percent of a givea drop flashes to steam. Section 7.1-also estimates that the drop diameter only decreases 5 percent due to evaporation in later phases. ;i The drops are strongly coupled to the containment atmosphere temperature due to the large surface area of the total drop population. This strong coupling results in the drop temperature, closely following the containment atmosphere temperature as it changes during the transient. Sensitivities using EGOTHIC show that if. 5 percent or more of the liquid is converted into-drops, then the containment atmosphere will be saturated quickly Given the high velocity of--
' the blowdown releases, much greater than 5 percent is anticipated to be converted into drops.
With the atmosphere saturated, thermal effects such as superheating will not occur and the effect of larger drop fractions does not significantly affect the pressure response. The effects of drops on the Evaluation Model calculation of containment pressure is investigated with a sensitivity - h study described in Section 5.8. Circulation and Strati 6 cation Effects The presence of drops increases the density of the containment atmosphere, which makes the post-blowdown steam release relatively more buoyant. An estimate of the effect of drops on circulation and stratification is made by. calculating the plume entrainment rate and resulting circulation time constant for the conditions at the end of the blowdown phase of the DBA LOCA. As discussed in Section 7.1 of Reference 9.A.2, well-accepted models are not available to predict - the mass of the drops created during blowdown, so the bounds of 0 percent and 100 percent of the liquid will be considered.- To estimate the volume entrained into - the plume _(Q,, in . ft'/sec), Peterson's equations
- (Reference 9.A.4) can be used: , . Q. = 0.15 _' B'"
- Z"' i
- where: Z = elevation (ft.)
3 = g
- Q, * (p 6 - p )/p d g " gravitational acceleration = 32.2 ft/sec2 v,t= volumetric steam flow (ft 2/sec) p 3 = containment ambient density (lbm/ft') -
p, = steam density (Ibm /ft') The entrainment is calculated for a height of 100 feet above the top of the steam generator compartment, so Z = 100 ft. The steam release at the beginning of the peak pressure phase is 2 estimated to be 1870 ft /sec (Q,). For the case assuming 0 percent of the liquid is released as t is approximately 0.275. For the case assuming 100 percent of the drops, the (p. 3 - p,.)/p 6 erm Thermal and Circulation Effec +s of Drops During a LOCA - Octobn 1997 o:\3006appa.. ion:1b-101497 Rmsion 1 m.... -
9.A-4 liquid is released as drops, the density term is approximately 0.60. Using the above equation, the estimated entrainment rate is Q,, = 8239 ft'/sec (0 percent drops) and 10695 ft'/sec (100 percent drops). The estimated entrainment at the end of blowdown is a}: proximately four times the steam flow (Q,,) for the case without drops, and slightly less than six times the steam flow for the case with drops. Knowing the entrairanent rate, a circulation time constant can be calculated for the containment free volume. This time constant will change with time, but it provides an indication of the amount of circulation expected for the releases after the refill phase. The circulation time constant is the volume divided by the entrainment rate, and for 0 percent drops it is 206 seconds and for 100 percent drops it is 159 seconds, it should be noted that the estimated times conservatively neglect volumetric entrainment into the wall layers. These time constants increase as the steam flow decreases, but this estimation shows that a large fraction of the containment volume will be entrained in the plume within a few minutes, which is relatively short compared to the time to reach maximum pressure (at approximately 1200 seconds), and very short compared to long-term cooling. A relatively large entrainment rate within the ot>ove-deck region indicates that the steam density gradients above-deck are not large whether drops exist or not. Therefore, the presence of drops will not significant' affect the general circulation and stratification patterns in the containment atmosphere. Evaluation Model Drop Sensitivity Study The Evaluation Model for the LOCA, with the jet dissipated in the steam generator compartment, 9 was used to determine the effect of drops on the Evtluation Model calculation of containment pressure. The treatment of drops in the Evaluation hudel is described in Section 4.5.2.1. The Evaluation Model converts all of the liquid from the reactor side of the break to drops, and none of the liquid from the steam generator side of the break. Sensidvity cases were analyzed for comparison to the Evaluation Model results. The sensitivity cases are discussed in Section 5.8. One case modeled no drop formation and one case modeled 100 percent of the liquid converted into drops. The containment pressure, as a function of time, was calculated for the sensitivity case. The maximum containment pressure, calculated with the Evaluation Model, is greater than the maximum pressure calculated assuming no drop formation. The p.esence of drops does have a slight influence on the Evaluation Model pressure calculation. Drop formation is expected during the blowdown phase and the sensitivity study indicates that drop formation should be r odeled to provide a bounding calculation for containment pressure. Conclusions The formation of drops during the LOCA blowdown phase is a physically real phenomenon that may influence the maximum containment pressure calculated by the Evaluation Model. Drop Thermal and Circulation Effects of Drops During a LOCA october 1997 o-\3006appa.non.1b-101497 Revnion 1
9.A-5 formation increases the density of the containment atmosphere making the post-blowdown releases relatively more buoyant. A small percentage of the blowdown break liquid formed into drops is sufficient to saturate the containment atmosphere, at which point additional drop density has a minor thermal effect. The Evaluation Model treatment of drops, as described in Section 4.5.2.1, provides a sufficient bounding calculation for maximum and long-term containment pressure. References 9.A.1. WCAP-14812, " Accident Specification and Phenomena 9 valuation for AP600 Passive Containment Cooling System," Revision 1, dated June 1997, 9.A.2. WCAP 14845, " Scaling Analysis for AP600 Containment Pressure During Design Basic Accidents," Revision 2, datcd June 1997. 9.A.3. NTD-NRC-95-4563, " GOTHIC Version 4.0 Documentation, Enclosure 2: Technical Manual," September 21,1995. 9.A 4. Peterson, P., " Scaling and Analysis of Mixing in Large Stratified Volumes," International fournal of Heat and Mass Transfer, Vol. 37, Supplement 1, pp 97-106,1994. O 1 (
'Ihermal and Circulation Effects of Drops During a LOCA october 1997 o:\3006appa.norrib 101497 Revnion 1
l Appendix 9.B Effects of Stratification on Heat Sink Utilization O + O Effects of Stratification on Heat Sink Utilization ca,tober 1997 a.\3006w-9b.non:1b-101497 Revukm 1
l 9.B-1 l m TABLE OF CONTENTS ( 9.B.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.B-2 9.B.2 HEAT SINK ANALYSIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.B-2 9 B.3 CONDENSATION BOUNDARY CONDITIONS . . . . . . . . . . . . . . . . . . . . . . . . , 9.B-2 9.B.4 HEAT CONDUCTION MODELS . . . . . . . . . . . . . . . . .................. 9.B-S 9.B.5 RESULTS .................................. .................. 9.s? 9.B.6 . CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.B-11 LIST OF TABLES 9.B-1 Concrete Model Nodal Thickness . . . . . . . . . . . . . . . . . . . . .............. 9.B-6 9.B-2 AP600 Assumed Room Heat Sink Distribution ........................ 9.B-9 a
-p. LIST OF FIGURES 9.B-1: Condensation Heat Transfer Cwfficietea vs. T,s . . . . . . . . . . . . . . . . . . . . . . , 9.B-13 9.B-2: Containment Shell Heat Sink Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.B-14 9.B-3: CMT Room Heat Sink Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.B-15 C
( ) s Effects cf Stratification on Heat Sink Utilintion october 1997 o:\3006w-9b.ron;1b-101497 Revision 1
9.B 2 9.b.; INTRODUCTION An anMysis was performed to determine the impact of stratification on the relative effectiveness of containment heat sinks during a postulated LOCA. Models weie developed to study transient heat conduction effects for steel and con rete structures anact a variety of containment atmosphere boundary conditions. The models were then used to determine the effects of stratification of steam in the containment atmosphere on heat sink utilization in the CMT room and in the above-deck region. 9.B.2 HEAT SINK ANALYSIS The condensation heat transfer in the containment atmosphere has beer. characterized as a function of the steam fraction, and has been used as boundary conditions to detennine the transient heat absorption rate of the heat sink structures. The results of these analyses are used to estimate the relative effects of stratification on the heat sinks located on the FCS steel shell and in the AP600 CMT room. The purpose of the analysis is to obtain relative effects of stratification for reasonably representative conditions to assess the magnitude of the bias. An extreme stratification gradient is assumed from which the relative effect of stratification on total heat sink energy removal ir a region can be assessed. A bias is developed to bound the non-conservative effects of stratification. 9.B.3 CONDENSATION BOUNDARY CONDITIONS Those sensitivity calculations are performcd to examine the relative effect of a gas mixture that is homogeneous (as in a lumped parameter node) and a gas mixture that is stratified. To keep the calculations sirnple, boundary conditions are assumed constant with time, and the following homogenous atmosphere conditions are assumed: T.,. = 276*F P,,, = 59.7 psia f,, = 0.63 (homogeneous steam mole fraction) These parameters represent approximately time-averaged values over the first hour of the LOCA, since the CMT room steam concentration is relatively constant (Figure 9-44). The heat transfer from the containment atmosphere aad the st;ucture is assumed to be dominated by condensation so that convection and radiation are neglected. The condensation Effects of Stratification on Heat Sink Utilization october 1997 o".3006 web.norrib-101497 Revuion 1 l
. -- .-. - . ~ . . . - .. - - - .. - .
4 9.B-3 heat transfer is determined by first determining the mass transfer for turbulent free convection
- (Reference 9.B.1, Section 4.3): " '
- OPbC rh" ' O.13' (9.B-1)
(u fg) ifs p ,, , p , 2 J J where i th" is the condensation mass flux
- p. is the density of steam at the total pressure and boundary layer temperature -
; AP,. is the difference in the steam partial pressure atmosphere - surface v is the mixture kinematic viscosity l g is gravity Pu is the log mean pressure difference atmosphere - surface Ap is the mixture density difference atmosphere - surface p is the bulk mixture density i Sc is the mixture Schmidt number (typically ~0.51) i e
and D, is the air-steam diffusion coefficient which is given by O (Reference 9.B-1, Section 4.3.2)
, , ui
! 14.2 psi T,,,+T,. (9.B-2) D'= 0.892 P- ,2 x 460*R , , The steam partial pressure in the atmosphe:e is given by: 4 P _ = f,,* P (9.B-3)
- where f,, is the steam mole fraction in the atmosphere and P is the total pressure. . The steam partial pressure at the condensing surface is given by:
P_, = P,, (T,,) (9.B-4)
' wher- P , is the saturation pressure corresponding to T,,,.
Effects of Stratification on Heat Sink Utilizat.on october la97
- o
- \3006w-9b.non:Ib-101497 Revison 1 4
.. 4 - .
9.B-4 The log mean pressure difference between the atmosphere air pressure and the air pressure at the surface is given by: g (P,,,,, - P,,,,,) p* * , (9.B-5) in (Pair $urf / Pair atm) where P,i,,,,, is the air partial pressure at the heat sink surface, P - P,,,,,,, and P.,,,,, is the air partial pressure in the atmosphere, (1 - f,,)
- P.
The densities of air and steam at the atmospheric and surface pressures and temperatures are determined from the ideal gas law. To determine the effect of the steam fraction, three distinct regions based on equal volume are assumed. The top region is assurned to be nearly all steam with f,,, = 0.98. The middle region is assumed to be at the nomina' conditions with f,,.a = 0.63. The bottom region steam fraction is determined by conserving the total amount of steam in the total volume, f w
- 3' f,,_ - f,., - f,..a = 0.28 (9.B-6)
Applying these three steam mole fractions along with the above contahent atmosphere O conditions, a relationship can be determined for the condensation heat transfer coefficient as a function of heat smk surface temperature. An equivalent condensation heat transfer coefficient is calculated from ih" for use as a boundary condition for heat sink condensation, described later. The equivalent condensation heat transfer coefficient is calculated by: h = 5 (9B-7) (I atm
~ ,urf where h,, is the difference between the steam and liquid saturation enthalpy. The relationships for equivalent heat t ansfer coefficient are shown graphically in Figure 9.B-1.
The condensation heat transfer coefficient varies considerably with respect to the steam fraction in the containment atmosphere, f,,, and the surface temperature, T,,,,. For each steam fraction, the heat transfer coefficient increases with increasing T,,,, until the saturation temperature that corresponds to the steam partial pressure at the surface is reached. At this point the O Effects of Stratification on Heat Sink Utilization october 1997 l
. o:\3006w-9b.nortIt>101497 Revision 1
9.B-5
- condensation heat transfer drops to zero, and is zero for all surface temperatures greater than this temperature.
For the case of f, = 0.98, T,,, = 291'F, which is greater than the containment atmosphere temperature. Thus, the condensation heat transfer coefficient increases with surface temperature and no cutoff is reached. For the case of f,, = 0.63, T,., = 264'F, and the heat transfer coefficient drops to zero at this temperature. For the case of f,, = 0.28,217'F, the heat transfer coefficient drops to zero, 9.B.4 HEAT CONDUC110N MODELS Several models were developed to calculate heat transfer to the heat sinks. These include: Steel structures of varying thickness
. Concrete structures .- Steel-jacketed concrete structures
- Steel containment shell A description of each model is given as follows.
Steel Structures
'1he one-dimensional model consists of a 1 ft. by 1 ft. section of steel, modeled by ten nodes of . equal thickness, representing one-half the heat sink thickness. For example, for a one-half inch thick steel plate, the model has ten nodes, each 0.025 in. thick. A convective boundary condition is applied to one surface, while the other surface is assumed to be adiabatic. Connections between the nodes are defined by the area of the interface (1 ft2 ), and the distance from the node center to the interface (0.0125 in.). The properties for steel are taken from Section 4:
p = 490.7 lbm/ft' C = 0.107 Btu /lbm *F- , k = 30 Btu /hr-ft 'F A' tero-volume node is attached to the steel at the surface exposed to the atmosphere. The boundary conditions for the three steam fractions are described in the previous section.
- Concrete Heat Sinks The concrete heat sinks have much lower thermal conductivity and are modeled differently than the steel heat sink. The thermal properties of the concrete are given as:
p= 140 lbm/ft' Efects of Stratification on Heat Sink Utilization october 1997
- oA3th%norrib-101497 Revision 1
l 9.B-6 C = 0.19 Btu /lbm *F i k = 0.83 Btu /hr-ft *F Once again, ten nodes ne used to represent one-half the concrete thickness. For this case, the nodes are not equal volume with the nodes nearest the convecting surface having small thicknesses, and the thickness increasing geometrically as the nodes progress inward to the adiabatic boundary. The thicknesses are summarized for each node in Table 9.B-1. (a,c) Table 9.B 1 Concrete Model Nodal Thickness __ .F As for the steel model, Node #1 is connected to a zero-volume surface node, which is in turn connected to the boundary temperature. The heat transfer coefficient is defined in the previous section as a function of the surface temperature for the three steam fractions conridered. Steel-Tacketed Concrete Heat Sinks The steel-jacketed concrete heat sinks combines the two-foot thick concrete model previously described with a one-half inch steel plate. The condensation boundary condition is attached to the cutside nf the steel plate, that is represented by 10 nodes,0.05 in thick. The inside steel node is attached to the first concrete node with an assumed gap oi 0.036 in. The gap conductance is given by h gap =k./6 gap man (9.B-8) Effects of Stratification on Heat Sink Utilization october 1997 l o:\30fEw-9b.non.1b-101497 Revnion 1 l
p. c ,. 9.B-7 where 8,,, -is the gap thickness - and k. :is the thermal conductivity of the containment atmosphere mixture k, = 0.5' (k, + k ) (9B-9)- For T. = 276*F, and f, = 0.5, k. = 0.03 Btu /hr-ft *F, and h,,, = 10 Btu /hr-f:2 'F.- The concrete is represented by 10 nodes with thicknesses shown in TaVe 9.B-1. Steel Containment Shell The steel containment shell model is somewhat more complex in that the inside boundary-condition is the same as the other models whila the outside boundary condition is not adiabatic, but is xepresentative of the outer shell evaporative heat transfer. The steel shell is assumed to
- be 1.625 in, thick. For this case, a [ ]('d The inner-most node is connected w a zero-volume node upon which the condensation boundary condition is assumed. The outer-most node is also connected to a zero-volume node upon which an n aporation boundary condition is assumed. The outside boundary temperature is assumed to be an average between the inlet air temperature at the bottom of the Passive Containment Cooling System annulus, and the outlet air temperature at the top.
T , = 142'F ; and: : h,,,, = 113 Btu /hr-ft2 p -
. Note that the assumption of a constant value of h over the entire shell surface is very conservative, since in the stratified case, the shell adjacent to the steam-rich top _would heat up and significantly ir. crease the evaporation rate on the outside. No credit is taken in this analysis - ' for the associated increase in external heat transfer coefficient.
For this model, there is a short period of time during'which the shell heats up from the initial-temperature. After this time, a steady-state condition is established as heat is transferred at a E nearly constant rate from the inside to the outside of the'shell. 9.B.5. RESUL'IS For each of the models described above, three transient calculations were performed representing each of the three steam fraction conditions. The results of these calculations were used to examine heat absorption effects for each of the cenditions. Since the models represent one square foot'of heat sink area, the results can be used to estimate the heat sink behavior in a - Effects of Stratification on Heat Sink Utilization october 1997
- c:\3006w-9b.rwrulb 101497 Revision 1
9.B-8 typical room by multiplying the integrated heat removal by the total area for a particular heat sink type. Containment Steel Shell Heat Sink Stratification Sensitivity Figure 9.B-2 shows the heat removal rate for the containment shell. For this case, the areas for the top, middle, and bottom of the shell are not weighted equally (as in Equation 9.B-10). For this case, the volume of the containment above the operating deck is divided into three regions of equal volume, and the associated surface area for each volume is used. For the AP600 containment, Elevation of operating deck = 135.25 ft Elevation of spring line = 218.71 ft Elevation of top of dome = 156.4 ft Containment radius = 65 ft Gas Volume in dome = 336,963 ft' Surface area of dome = 15,552 ft 2 Total volume of gas above deck = 1.45 x 10' ft) The two lower regions both consist of a cylindrical gas volume = 481,582 ft'. This corresponds 2 to a cylindrical section 36.28 feet in length with a surface area = 14,776 ft . The upper reg, ion gas volume is also 481,582 ft', and consists of the dome and a cylindrical section 11.1 feet in length. The total surface area associated with this volume is 19,898 ft'. Thus, the equivalent integrated heat removal rate orough one square foot of the shell is weighted by surface area as 3.%
= (19,898' , + 14,776kg + 14,776()/49,450 (9.B-10)
The results show that the higher weighting of the upper, steam-rich region nearly compensates for the lower heat removal rates in the bottom region, and the heat removal rate is slightly (~0.5% after 200 seconds) higher for the homogeneous case. O Effects of Strattfication on Heat Sink Utdization october 1997 o:\3006w 9b.non:1b-101497 Revision !
9.B-9
,_ Results for the steel shell assessment are presented in terms of instantaneous rate since the
( external boundary condition never allows the steel to saturate. The results also allow interpretation of stratification effects during the quasi steady, long-term, while the steel shell is the dominant heat sink and the balance between instantaneous source and sink heat rates governs the containinent pressure. Since the stratification penalty on the steel shell heat removal rate is nearly negligible, a simple bias is introduced into the Evaluation Model by removing the non-grating operating deck floors to bound the effect. The stratification effect is exaggerated due to the use of an extreme gradient, well beyond what has been observed in the IST (Section 9.2.1 and 9.2.3) and in the international containment database (Appendix 9.C.2). Simulated Room Heat Sink Stratification Sensitivity These models were applied to heat sinks which reasonably represent the AP600 CMT room. The heat sinks for the AP600 CMT room (North and South sections) are summarized in Table 9.B-2. Table 9.B 2 AP600 Assumed Room Heat Sink Distribution Heat Sinks in Simulated Room Thickness Surface Area Region Steel-Jacketed Concrete - Ceiling (single. 0.5 in. / 24 in. 5398.87 ft2 Top sided) f'N eQ Steel-Jacketed Concrete - Floors (single-sided) 0.5 in. / 24 in. 5601.44 it2 Bottom Steel-Jacketed Concrete - Walls (double- 0.5 in. / 24 in. 4596.11 ft2 1/3 in each region sided) Steel-Jacketed Concrete - Wall (double-sided) 0.5 in / 48 in 673.99 ft2 1/? in each region Concrete - Bulk (double-sided) 48 in. 3287.36 ft2 1/3 in each region Steel - CMT (single-sided) 4.874 in. 1848,8 ft2 1/3 in each region Steel - Comminment Shell Wall (single-sided) 1.57 in. 11385.53 ft2 1/3 in each region Steel - Columns (double-sided) 0.39 in. 1656.5 ft2 1/3 in each region Steel - Floor Grating (double-sided) 0.39 in. 3781.69 ft2 1/3 in each region Steel - Elevator (double-sided) 0.2in. 218.96 ft2 1/3 in each region Steel - Platform (double-sided) 0.144 in. 11254.2 ft2 1/3 in each region Steel- Stair & Rails (double-sided) 0.132 in. 181.59 ft2 1/3 n each region As was discussed previously, each heat sink was analyzed using three different steam fractions representing the top, middle, and bottom thirds of the room which is a bounding gradient when the plume rises through the CMT compartment. There is expected to be no f significant stratification penalty in the CMT room with downflow in the Evaluation Model, Effects of Stratification on Heat Sink Utilization october 1997 o:\3006w-9b.non;1b 101497 Revision 1
9.B-10 where the plume rises from the steam generator compartment. For each ladividual heat sink, a homogeneous case and three-region averaged result was obtained for a 1 ft2 section of the heat sink. The energy removal by each heat sink is determined by calculating the heat h ' removal for 1 ft 2, and multiplying by the appropriate surface area. Where appropriate, the heat sinks that are located in a specific volume (i.e., ceilings and floors) are not averaged for the three-region, but are anclyzed solely with the steam fraction of that volume. This becomes important for the ceilings since these heat sinks are located within the high steam fraction volume and higher heat transfer is expected when the room is stratified. The opposite is expected when considering floors. Refer to Table 9.B-2 for the region designatien. Figure 9.B-3 shows the integrated heat removal by all the heat sinks in the CMT room for a one hour transient. As will be discussed below, the stratification bias for this case is a function of the total energy absorbed. This is because the adiabatic boundary condition results in heat sinks reaching a maximum thermal aosorption governed by the saturation temperature for the given steam concentration in a volume. Therefore, results for this scenario are presented in terms of integrated total heat absorption. The results show the CMT room heat sinks including the floors for the homogeneous and stratified cases. In addition, the case where the floors are not included for the homogeneous case is also shown. The stratified, three-region results are lower than the homogeneous case results by 10-15% when all heat sinks are considered. The homogeneous case with floors excluded ir slightly conservative when compared to the stratified case with the floors included. Thus, the combination of assuming homogeneous conditions and neglecting the floors in the total heat sink area results in total heat sink utilization that is neutral at the time of peak pressure, and over the longer term is slightly conservative relative to the expected conditions. The assessment of stratification effects is very conservative because a conservatively low benefit for the uppermost region is used, and the gradient is much more extreme than what has been observed in the LST (9.2.1 and 9.2.3) and in the international containment database (Appendix 9.C.2). The choice of stratified conditions to examine for this sensitivity are conservaue and the results bound other, less extreme postulated stratification gradients. The room temperature is assumed to be 276*F in the stratified case, the same temperature as in the base case homogeneous room. One could, for example, postulate a less extreme, thennodynamically consistent, gradient of 0.77 for the top,0.63 for the middle, and 0.49 for the bottom. The saturation temperature for a region at 59.7 psia and a steam mole fraction of 0.98 (psat of 58.5 psia) is 291*F. The upper regian then would be about 15 F hotter than assumed. Therefore, the upper region conditions are thermodynamically inconsistent in a way that minimizes heat absorption m the upper region of the room, and thus maximizes the stratification bias. Effects of Stratification on Heat Sink Utilization october 1997 e43006w-9b.norrib-101497 RevWon 1
9.B-11 l The bias for the CMT room is govemed by the air content in the lowest region. Results ( ) indicate that steel heat sinks, and the steel on Jacketed concrete, reach a maximum for integrated heat absorption well within the one-hour time frame of the calculation. The concrete continues to absorb heat over a very long term, on the order of days. However, the
- transient skin temperature of concrete increases due to its relatively poor thermal conductivity and a gap between the steel jacket and concrete reduces concrete effectiveness, so that the magnitude of concrete heat absorption is not significant relative to the steel. The integrated heat absorption by heat sinks is then primarily a function of the maximum bulk steel temperature rise, which is related to the saturation temperature of the adjacent region.
While a less severe assumed stratification gradient would result in less rapid heat absorption by sinks in the upper region, the upper heat sinks would still reach their maximum well within the one- hour time frame. The lower region integrated heat absorption is limited by the saturation temperature for the ascumed steam concentration. Therefore, the stratification bias is controlled by the lower region steam concentration and is maximized by the assumption of the extreme strc.tification gradient. Since the exaggerated effect of stratification for the case of a plume rising through the CMT shows a bias on total integrated heat removal, a bias is introduced into the Evaluation Model by removing heat sinks associated with floors in compartments. As an additional conservatism, that bias is retained for the Evaluation Model with a plume rising through the steam generator compartment, as well as all sensitivity cases performed, even though most situations result in downflow through the CMT compartment. For the case of the steel containment shell above the operating deck, the dorae surface area weights the upper, steam-rich volume more heavily than the lower volumes, and compensates for the lower heat removal rates. Thus, the homogeneous case results are nearly equal to those for the stratified case, with the homogeneous case giving less than 0.5% less instantaneous heat removal rates. A simple bias of removing operating deck floors is included in the Evaluation Model to bound this effect. 9.B.6 ' CONCLUSIONS For the case of the steel containment shell above the operating deck, the dome surface area weights the upper, steam-rich volume more heavily than the lower volumes, and compensates for the lower heat removal rates. Thus, the homogeneous case results are nearly equal to the stratified case, with the homogeneous case giving less than 0.5 percent less instantaneous heat removal rates. A simple bias of removing operating deck floors is included in the Evaluation Model to bound this effect. The results of the heat sink utilization analysis for below-deck compartments indicate that in general, the assumption of homogeneous compartment volumes predicts higher overall heat (n) removal by the heat sinks compared to stratified volumes. This is primarily due to the Effects of Stratification on Heat Sink Utilization october 1997 o:\3006w 9b.non:1b-101497 Reymon 1
9.B-12 propensity of the condensation heat transfer to fall off as the heat sink surface temperature approaches the local saturation temperature in the lower steam fraction volumes. Stratification gradients are not expected to be nearly as extreme as assumed in this evaluation. The results of the homogeneous case gives 15 20% higher integrated heat removal than the stratified results. Therefore, a bias is introduced in the Evaluation Model to account for this difference, implemented by removing heat sinks representing floors from the Evaluation Model. References 9.B-1. WCAP-14845, " Scaling Analysis for AP600 Containment Pressure During Design Basis Accidents," Revision 2, June 1997. l f / Effects of Stratification on Heat Sink Utilization octotwr w97 o:\30N>w-9b.non;1b-101497 Revision 1
9.B-13 4 Heat Sink Utilization - Heat Transfer Coefficients 100000 h 10000- - - - - - - - , - - - -. g . -_ h 1000 - mmmm** } -.4----l.- l
-+ Bottorn 5 I --G - Middle I
l 100 i . - - . - . . . - - . . . . _
+ Top c
- 10 - ---- - - - - --. .= - . . . - .
1 100 120 140 160 180 200 220 240 260 280 Surface Temperatu o (F) Figure 9.B 1: Condensation Heat Transfer Coefficients vs. Tu Effects of Stratification on Heat Sink Utilization october 1997 o:\3006w-9b.non:1b-101497 Revision 1
9.B-14 1 l 9 g 1 - I _. .. __ . _
.g i
i S 1 ij e-B c l g i
- Y .
2
- 2 R M 2 2 (m4me)anu m Figure a.B-2: Containment Shell Heat Sink Resu:ts 6C Effects of Stratification on Heat Sink Utilization october 1997
. o:\3006w-Nnon lb-101497 Revuion 1
9.B-15 CMT Room Heat Sink Uti..zation 4.00E+07 , 3.50E+07 - - - - - -- p'<T -- -- -
- 3.00E+07 - - # W g - -
y,Trg..- 3,-
-- 2.50E+07 - - - - -- -- -.
g 2.00E+07 - N w 1.50E+07 - - - - - - - - - - - -- -
.J S 1.00E+07 - - -- -- - .- -- _ . - . . . - - - -Qavg w/o floors 5.00E+06 -- ----- - - - --_ - - - Oavg w/ floors 0#' *# 8 0.00E+00 -
0 500 1000 1500 2000 2500 3000 3500 Time (sec) Figure 9.B-3: CMT Room Heat Sink Mesults Effects of Stratif'ication on Heat Sink Utilization october 1997 o:\3006w-9b.noalb-101497 Revision 1 I _ _ _ - _ _ - _ _ .
Appendix 9.C o Additional Information on AP600 Containment Circulation and Stratification . M. DzoDzo O i O
~
Development of Expected Flow Pattems october 1997 o \3006w 9c.non:1b-101497 Rmsion 1
9.C-1 l 9.C.1 DEVELOPMENT OF EXPECTED FLOW PATTERNS FOR AP600 BASED ON SEPARATE FLOW TESTS IN ENCLOSURES 9.C.1.1 STRATIFICATION PHENOMENA Stratification is the formation of horizontal layers of constant density. Stratified layers are stable if the density of the layers decreases in the upward ver+ical direction (the gradients of density are negative in z direction according to Figure 9.C.1-la) and if forced convection mixing is not suf6ciently strong to disrupt the stable fluid layers. Another more general definition of stratified conditiors is that gradients of density in the horizontal direction are small, except in jets, buoyant plumes, and small regions near the vertical walls inside boundary layers (wall jets). In most of the volume, the density gradients in z direction are negative, while inside the jets, wall jets, and buoyant plumes they could be positive (see Figure 9.C.1-1b). Stratification occurs as a consequence of the temperature or concentration gradients in the vertical direction. Increasing temperattves or decreasing concentrations of heavier minure components with increasing elevatica promote stratification. The existence of flow structures, such as jets, plumes, and vertical wall boundary layers, decreases the " steepness" of the vertical density gradients. Examples of stratified conditions are numerous. Stratified layers are observed as large-scale geophysical phenomena (in lakes, sea, and oceans, in atmosphere - stratus clouds), as well as inside the enclosures. For example, warmer air tends to gather below ceilings in energy storage devices, nuclear reactors, solar collectors, and enclosures under the influence of the spread of fire and smoke. This appendix discusses the stratification phenomena inside a nuclear reactor containment. Possible rearons for the stratification will be specified. Strat fication may occur if:
- 1) The t.pper boundary is at the higher temperature than the lower boundary (see Figure 9.C.1-2a), as well as for other similar combinations of temperature boundary conditions at the outside and inside surfaces (see Figure 9.C.1-2b - d).
- 2) A higher concentration of the heavier or lighter components of the mixtures is maintained (by injecting and removing) near the lower or upper boundaries of the enclosure, respectively.
- 3) A lighter fluid is released (permanently, or from time to time) and captured below the i ceiling of the containment.
O
%J Development of Expected Flow Patterns october 1997 o \3006w oc.nortit>101497 Rmsion 1 l
9.C-2 _
- 4) The release point of the lighter / heavier fluid is closer to the top / bottom.
- 5) The shape of the enclosure promotes stratification (tall elongated enclosure).
- 6) De distribution of the non-complete vertical partitions suppresses fluid flow in the upper portions of the enclosure.
- 7) The distribution and size of the horizoual openings suppresses the fluid flow in vertical direction.
- 8) The internal heat sources (sinks) are positioned in the upper (lower) portions of the enclosure.
Under the conditions above (or a combination of them), the stratification may be stable. The presence of stratified layers inhibits circulation, that otherwise could be induced by a jet, plume, or boundary layers. He conduction and diffusion, heat and mass transfer processes, respectively, are dominant. As a result, the overall heat and mass transfer decreases and the heat transfer through the containment shell is slowed. One way to avoid the stratification is to generate fluid flow patterns inside the enclosure using forced convection. Additional devices such as fans, sprays, or nozz'es are necessary, as well as associated power supplies and controls. Since the AP600 relies on a passiv containment cooling system (PCS), only the effects of fluid O circulation due to the interaction of na'. ural convection with the stratified field are discussed. Modifications to the shape of the enclosure, the distribution and size of the internal partitions, and the openings could be made to avoid stratification. A different distribution of heat sources could also be applied to generate natural convection effects. The fluid flows due to natural convection promote better circulation inside the enclosure. The introduction of jets may also interrupt stably stratified layers through better mixing of the layers of various densities and concentrations. With a jet stratification msy become unstable or, at least, the vertica! padient reduced. With only natural convectbn it .he generated buoyancy forces are strong enough, the entin volume of the enclosure will be affected, resulting in the relatively uniform values of temperature and concentration fields. Natural convection heat transfer is dominant and the more intensive circulation improves the transfer of heat from the containment. Naturai convection flow effects are generated spontaneously due to the gravity (buoyancy forces) when heated sources exist, so that additional control and other devices are not necessary. O Developrnent of Expected Flow Pattems cetober 1997 ] o:\3006w4..non:1b-101497 Revision 1 \ . - _ _ _ _ _ - - - - - ----
9.C-3 . l 1 9.C.1.1.1 Static Stratification
'T (Q Static stratification occurs if the upper horizontal boundary of the domain is maintained at a higher temperature than lower boundary, as in Fig. 9.C.1-la. Stratification also occurs if the concentration of heavy components is low in tl.e mixture in the upper portion of the domain.
i The fluid layers are undisturbed and fluid motion is negligible. The temperature or density distribution in the vertical direction is linear. Heat transfer is predominantly govemed by conduction, while mass transfer is driven by diffusion. The formed fluid layers are stable and communicate only with the neighboring upper and lower iayers. The resulting heat and mass transfer rates are low. Corresponding experimental results are found in Akino et al.,1989 and Hiller et al.,1988. In both papers, stable stratified layers are identified using various colors reflected by liquid crystals (suspended in th" 'luid). Static stratification exists inside a containment vessel if the temperature distribution of tha vertical walls is the same as in the surrounding stratified fluid (adiabatic vertical walls, as in Fig 9 C.1-la). Since, the top of the AP600 containment, as well as the vertical walls, are exposed to the surrounding air and cooled by natural convection, stable stratification is not present. Even small temperature differences between the air inside and outside the containment produce large Grashof (Rayleigh) numbers, due to the height of the containmer.t (H,=109 ft). For example, a p temperature difference of 9'F between the air at the deck level and the air below the dome V ceiling results in Gr, = 2.210". This is in the range of chaotical turbulent flow, characterized by upward and downward plumes (see experimental results by Akino et al.,1989). Static stratified layers are also generated by releasing a lighter gas, e.g., steam or hydrogen, into the upper portion of the containment and capturing the gas benea$ the dome. Hydrogen distribution experiments performed in the HDR facility, tes' group Ell, combine high hydrogen release rates with superheated steam injection into the containment (see Wolf et al., 1994a). A comparisoa of influences of the axial break and gas release positions is obtained with E11.2 (high release position) and E11.4 (low release position) experiments. Although these two specific experiments simulate severe accident scenarios, comparison of results from the two experiments provides insights into the physics of stratification. The tests are characterized by boundary conditions that can promote circulation (especially test E11.4). They also show that relatively small concentrauon gradients can exist in the presence of circulation. Steam release from small breaks generates thermal stratification for break positions located at the higher level, with the hot zone above the break locations. Two mechanisms are used to break up the established thermal stratification. The first mechanism used subsequent steam releases at positions lower than the original release to break up the established thermal stratification. This mechanism did not produce homogeneously mixed conditions. The second [V] method is the application of external sprays on the upper dome. This causes condensation on Development of Expected Flow Patterrs cktober 1997 o:\3006w-9c.non:1b-101497 Revision 1
9.C 'he
. inner surface and a decrease in the temperature in the upper part of the dome. Convective flows form and affect the whole volume of the dome and lower compartments, resulting in a completely homogenized atmosphere, h
As in the HDR E11.2 experiment, condensation on the dome of the AP600 breaks up stratification. The condensation on the vertical walls also contributes to breaking of stratified layers and to entrainment in the vertical boundary layers. The circulation inside the containment affects the lower compartments and promotes circulation due to the natural convection. The shape of an enclosure could also promote stratificatin. One example is natural convection inside romb shaped enclosures (see Figure 9.C.1-3). Stratification is generated if the upper vertical side is at .. high temperature and the inclined top and bottom sides are adiabatic (see, Dzodzo,1993). The overall heat and mass transfer are suppressed by the presence of the stratified fluid in the upper and lower corners of the romb shaped enclosures. When the boundary conditions are reversed, i.e., the lower vertical side is at the higher temperature, the entire volume of the enclosure is effected by circulation. Heat transfer is intensified and stratified layers are not present in the upper and lower comers. A comparison of experimentally and numerically obtaiaed temperature and velocity fields for these two cases is presented in Figure 9.C.1-4. An overview of the numerical results for various angles of the romb (parallelogram-shaped) enclosures, Prandtl numbers, and aspect ratios is presented by (Hyun and Choi,1990). . Although the top of the AP600 containment is somewhat conical in shape, stratification in the L upper portion of the dome would not exist because of the natural convection due to the lower temperatures of the ceiling and vertical walls. Stratification effects are promoted if the containment ceiling is insulated or at a higher temperature. The distribution of the internal heat sources in the upper part and heat sinks in the lower part l of enclosures promotes the formation of the stratified layers (set Figure 9.C.1-2 b, c, d). Examples of the influence of the position and distance between the heat source and .uat sink are provided by A. Kurosawa et al.,1993 and C.J. Ho et al.,1994. An example of the influence of an array of discrete heat sources on natural convection is presented by T. J. Heindel et al., l 1995. l Vertical non-complete partitions inside an enclosure contribute to the stratification. If the non-complete vertical partition is positioned near the ceiling, flow in the upper part of the l enclosure is obstructed and a stagnant stratified region near the ceiling is formed (see Hanjalic l - et al.,1996, and Nowak and Novak,1994 for examples of the two-dimensional numerical l simulation, and T. Fusegi et al.,1992, for the three-dimensional simulation). This is of :;pecial l interest for the analysis of the spread of fire and smoke inside the buildings. Such partitions do l not exist above the operating deck level in the AP600. O Development of Expected Flow Patterns October 1997 o \3006w 9c.non:1b-101497 Revision 1
9.C 5 Narrow horizontal openings between upper and lower compartments also suppress circulation 7 and cause stratification. The results of a two-dimensional numerical simulation (R. Frederick and (" ) A. Valencia,1995) show the influence of the size of the horizontal openings on the natural convection inside the vertically connected enclosures. The potential for stratification in compartments below the operating deck of containments, due to the various sizes of the openings is also studied (see ref. Wolf et al.,1994b). 9.C.1.1.2 Stcatification and Circulation Figure 9.C.1-1b illustrates conditions where a portion of an enclosure is stratified and other portions are affected by strong recirculation zones and currents. Due to the circulation effects, shallow vertical density gradients are present inside the stratified portion of the enclosure volume. Convective heat and mass transfer that results from communication between the stratified and flow-affected zones, contributes to the mixing between the zones with different temperatures, concentrations, and densities. Flow inside the enclosure is promoted by the existence of the entraining wall layers (which are a consequence of the heat transfer), penetrating jets, and buoyant plumea (see reference, Peterson,1994 and Figure 9.C.1-1b). To gain insight into AP600 physics, we will start with small-scale enclosure examples and (m) progress to larger scale. 9.C.1.1.2.1 Interaction of WallJets (Boundary Layers) with Stratified Layers One example of interaction of wall jets with stratified layers is the natural convection inside a square enclosure (see Figure 9.C.1-5). The opposite vertical walls of the enclosure are at the different temperatures and the horizontal walls are adiabatic (see Markatos and Pericleous,1984 arvi Figure 9.C.1-5a). With high Rayleigh numbers (over 10), Pr=0.71, turbulent flow exists inside the enclosure. Velocity and temperature gradients are large in the boundary layers. Velocities have maximum values near the walls, while inside the core of the enclosure they are small. The temperature (density) field in the core of the enclosure is stratified (see Figure 9.C.1-Sc). Communication exists between the boundary layec region and core of the enclosure through the vortices (see Figure 9.C.1-5b), which change in number, position, and intensity for various temperature differences between the opposite walls (various Ra numbers). Temperature gradients are highest in the boundary layers near the vertical and horizontal walls (see Figure 9.C.1-Sc). For the laminar convection (Ra=10** and Ra=10'5), the temperature difference between the highest and Iowest points at the vertical axis of the stratified core is 0.6*(T3 - T,), while for the turbulent regime (Ra= 10'8,10'",10*) it is 0.4*(T3 - T,). The decrease la the vertical tempelature gradients inside the stratified core for the turbulent regime is the result of higher gq (j velocities and strc nger circulation inside the cavity. The temperature field inside the core of the enclosure is stratified, while recirculation due to convection inside the enclosure is Development of Expected Flow Pattems october 1997 o \3006w.9c.non:lt>101497 Revision 1
9.C-6 predominantly near the walls. Despite the presence of the stratified core, for high Ra numbers, a fluid particle travels the entire enclosure (due to the convection) and contributes to better mixing and decreases the vertical gradients inside the core, h The increase of the Rayleigh number corresponds with a decrease in the thickness of the boundary layers, an increase in the temperature gradients inside the boundary layers, and an increase in the heat transfer rate. The dependence of the average Nusselt numbers on the Rayleigh numbers is presented in Figure 9.C.1-6. A similar two-dimensional flow pattern and stratified temperature (density) field is also obtained
- between two opposite vertical line Jets (see Figure 9.C.1-7) as discussed in Baines and Turner, l 1969.
A numerical analysis (Markatos and Pericleous,1984) is performed for a two-dimensional plane, assuming that the influence of the front and back walls of real three-dimensional enclosures is l not significant. For Rayleigh numbers greater than 10', the k-c turbulence model is used. Due l to time-averaging, the numerical results do not show either the instability mechanisms during the transition from laminar to turbulent flow, or the resulting oscillations that would result from solving the time dependeN Navier-Stokes equations. Experimental and numerical results for three-dimensional enclosures are provided by Hiller et al. l 1989, Mallinson and de Vhal Davis,1977, respectively. The results indicate that observed i l vorticew which affect mixing inside the core of the enclosure, communicate behveen the front and back walls through the middle of the enclosure, thus enhancing mixing due to three-h l dimensional circulation effects (see Figure 9.C.1-8). l Reviews of varioes aspects of confined convective flows, including the interactions between boundary layers neu the bounding walls and core and the effects of the cavity aspect ratio, inclination angle, and thermal boundary conditions on flow patterns, are presented by Ostrach 1972,1982, Catton,1978, Hoogendoorn,1986 and Allard,1992. A state of the art review of the analyses of two-dimensional and three-dimensional transient effects on the natural convection flows in sidewall heated enclosures is presented by T. Fusegi and J.M. Hyun,1994. , R. J. Janssen and R.A'W.M Henkes,1995 simulated the instability mechanisms and the transition from laminar to turbulent (oscillatory a-d finally chaotical) flow regimes inside a two-dimensional square enclosure with differenti# heated vertical walls nd adiabatic horizontal walls by solving the time-dependent Naviv . ,kes equations. The results indicate that the transition from laminar to chaotic flow (for h .0) is through periodic and quasi-periodic flow regimes. The periodic, quasi-periodic and chaode flow regimes are established for Prandtl 8 8 number 0.71 and Rayleigh numbers 2'10,3'108 and 7.5'10. Internal waves corresponding to 8 j fluctuations in the temperatures at Ra = 2'10 are preselded in Figure 9.C.1-9. The temperature differences in the entire core of the enclosure are small,0.0M*(Ts-T,). The predicted temperature differences inside the core of the enclosure are much sma?ter than those predicted by k-c l Development of Expected Flow Patterns October 1997 l o:\3006w-9c.non:1b 101497 Revision 1 l
a 9.C 7
' turbulence model (Markatos and Pericleous,1984). This indicates that temperature gradients 1 S inside the boundary layers are greater (isotherms inside the thermal boundary layer are not presented in Figure 9.C.1-9) and heat transfer is more intensive than calculated by k-c model.-
Two instability mechanisms influence the transition to turbulent (chaotical) flow regime. The first instability is a Keh'in-Helmholtz type instability (as in a plane jet with inflection points in the velocity profile) in the fluid layer exiting from the corners (where the vertical boundary _
' layers are turned horizontal). The second source of the instability is related to the instability in the boundary layer near the vertical walls. The instability inside the enclosure vertical boundary layers is mechanically (shear) driven. Both regions of the instability origins (hot and cold intrusions from comers and boundary layer waves) are presented in the Figure 9.C.110 (from -
S. Armfield and R. Janssen,1996). The figure presents temperatures for the initial solution,i.e., immediately after setting the left and right vertical boundaries to AT/2 and -AT/2, respectively. For values of Rayleigh numbers greater than 10', the turbulent oscillatory and chaotical flow affects the stratified layers inside the core of the enclosure. If the radius (H,=65 ft) of the AP600 containment is taken as a characteristic length (as a distance between the hot buoyant jet plume
- in the center and cold vertical wall boundary layers), a 9'F temperature difference results in a
. Grashof number Gr, = 4.7'10". 9.C.1.1.2.2 Interaction of jets or plumes with the stratified layers l.q The penetration of a stratified layer by a jet is another example where a portion of an enclosure t
- Q is stratified and another portion is affected by strong recirculating zones (Figure 9.C.1-1b). ;
j Depending upon the strength of the jet and the depth of the stratified layers, portions of the
- enclosure are affected by interaction between the jet and stratified layers. A portion of the
- stratified fluid is entrained by the jet, decreasing the average jet velocity. The jet penetrates upward (Garrad and Patrick,1983, So and Aksoy,1993, and Porterie et al.,1996), or downward (Markatos and Pericleous,1984, see Figure 9.C.1-5 b and c near the cool wall). . A negatively _
buoyant jet, as presented in Kapoor and Jaluria,1993, is also possible.
- The upward penetrating jet is of interest for LOCA or MSLB accident scenarios.- Scaling and
! analysis of mixing in large stratified volumes for the cases.of upward penetrating jets is presented by Peterson,1994. If the strength of the jet is strong enough, it produces fluid flow [ below the ceiling. After. reaching the vertical side walls, the flow results in downward [ 4 negatively buoyant jeta (see Figure 9.C.1-11 a and b). The downward, negatively buoyant penetrating jet (Kapoor and Jaluria, 1993, see t Figure 9.C.1-11a and 1-11b) is of interest for the analysis of the flow patterns inside the upper-deck region (Figure 9.C.1-11a), as well as for the compartments below the dome floor
' (Figure 9.C.1-11b). If the strength of the negatively buoyant jet is not high, it is not able to reach . compartments - below the deck. The direction of the flow changes as presented in ', Q Figure 9.C.1-11a. The redirection of the flow causes additional entrainment of the surrounding - (/ fluid, thus contributing towards the increase of the circulation (and mixing) inside the upper-Development of Expected Flow Patterns october 1997 V - o:\3006w 9e.non:1b-101497 Revision 1 t , . , ~ ,.y., , _r y-w s-
9.C-8 deck region. The correlations for entrainment rates in the negatively buoyant jets are presented in Kapoor and Jaluria,1993. If the strength of the negatively buoyant jets is high, it is able to penetrate into the below-deck O compartments. There are indications from large-scale tests conducted by Westinghouse (tests 222.3 and 222.4, 3-inch pipe, elevated 6 ft, pointed at the wall and up, respectively, see F.E. Peters, WCAP-14135, July 1994) which simulate the MSLD, that the entire volume of the contairunent has almost the same steam concentration. This occurs despite the fact that formation of the global circulation loop between the lower-deck compartments and upper dome re . o . , is not possible. An explanation is that the kinetic energy of the jets is high enough to pro. 6e downward penetration of the negatively buoyant jets into the below-deck compartments. Depending upon the distribution of compartments below the dome floor and the number, size and distribution of openings between the compartments and the dome region, various flow patterns are possible inside the compartments. A portion of the downward vertical plumes produced by natural convection (wall boundary layers) or the negatively buoyant jets produced by strong vertical upward penetrating jet into the dome region enters horizonal openings in the compartments, thus promoting circulation and flow inside the compartments below the deck. For the AP600, fluid flows upward to the dome through other compartment horizontal openings to preserve overall mass continuity and to close the global circulation loop (see Figure 9.C.1-11b). O O Development of Expected Flow Patterns october 1997 o:\3006w-9c.non:1t>101497 Revision 1
9.C-9 i O Th 87 - o aY Z o I
/ / 3 / /
ADIABATIC l VERTICAL S2 g < WALLS l
-9 \ [ .N , ,dc v
y------/ ----- f
/ $2<$1 g <0 / / =X Tc Figure 9.C1-la ~ 'Ihe formation of the horizontallayers of the constant density due to the stratification L (j.:-.
Development of Expected Flow Patterns october 1997 c:\3006w-9c.non:1b 101497 Resision 1
9.C-10 0 Tn (INSIDE THE fUOY T PLUM EM BUOYANT PLUME) Z - o I aS I / aZ>0 !
] . .
sw (WALL JET N iu
] / fC I l 3 - --- --- -- y/
g_- "
/ C 1' *E / /"/
g>0 / as
/ \ as_as~0 x e Ic W BY ~
at at at < 0 STRATIFIED REGIONS g = gy =0, g7 AND S JET REGIONS f0 gy f 0 az > 0 INSIDE THE ENCLOSURE. Figure 9.C.1-1b Interaction of jets, plumes and wall boundary layers with stratified regions Development of Expected Flow Patterns October 1997 o:\3006w-9c.norrib-101497 Revision 1
9.C O Z n Ih Ih Te g l H- f' #
-X /,
Tc
- 0) b)
O f re Te 1 l / j//7, h i /// ////
/ L, // 'i h Tc Tc C) O)
Figure 9.C.1-2 Combination of the constant temperature boundary conditions at the outside and inside surfaces which will produce stratification inside the
-(f enclosure Development of Expected Flow Patterns october 1997 c:\3006w-9c.non:1b 101497 Revision 1
l 9.C-12 - j l H-o G g .h- 7, F O O Tc / Z X "I-BT aZ , 9 STRATIFIED
',/ CORNERS h "3~ _
Th F Tc / ~ ~~~
~
Z d
-- X E-Figure 9.C.1-3 Stratification inside the upper and lower corners of the romb-shaped enclosure Example of the Stratification Caused by the Shape of the Enclosure and Distribution of the Boundary Conditions Development of Expected Flow Patterns October 1997 o:\3006w-9c.non:1b-101497 Revision I
. . . . - . - . . - . - . . - ~ . - . - . . . _ _ , . - . - - . _ . _ . . _ . . . _ . _ . _ . _ ~ . . . - . . - . . .
F 9.C-13 I L 4 4 pr**~4t gplbl f/B ^T u /*
/4 j k 8
4 4 c y c , , T, y+o
'((
IS THERMS STREAM LINES ISOTHERMS l
/ g Ex' t RuRU) WUMERICAL RESULU) NUMERICAL REULD)
E #
' ~T' *~4 '
T' l
/l4 /~4 ,f
- o ,6 ,v), o .
T, as {4 o T, ,e T3 44 s STREAK LINE3 CL C-B, Y. R STREAM LINES - b) -(ISOTHERMS EXM.RDGNTM, RESULU) WUMERICAL RESED) ISOTHERMS WUMDUCAL RESED) Figure 9.C.1-4 Experimental and numerical results hr the romb-shaped enclosure with the tomb angle 44' and Ra=3.5'10', Pr=5270, a) results for the upper vertical wall at the higher temperature b) results for the lower vertical wall at the higher temperature (Reprinted from: M.B. Dzodzo, " Visualization of laminar natural convection in romb-shaped enclosures by means of liquid cryrtals", in Imaging in transport processes (ed. S. Sideman and O K. Hijikata), Begel House, Inc.,1993, pp.183-193)
\
Development of Expected Flow Patterns October 1997 o:\3006w-9c.non:1b 101497 Revision 1
9.C-14 f=0
///////
g To Te o g Z h xlllllll GT=0 at
/////// ///////
f p l - [ t l -l - Tn
, T. Th Ic L_
TE 3 )
/////// ///////
3
) C)
Figure 9.C.1-5 a) The square enclosure with vertical walls at the different temperatures and horizontal walls adiabatic, b) Streamlines for Ra=10' and Pr=0.71,c) Isotherms for Ra=10' and Pr=0.71
" Reprinted from N.C. Markatos and K.A. Pericleous/ Laminar and Turbulent Natural Convection in an Enclosed Cavity, Int. J. Heat Mass Transfer, Vol. 27, No. 5, pp. 755-772,1984, Copyright 1984, Figure 9.C.5(d) and 6(d), with kind permission from Elsever Science Ltd, The Boulevard, Langford Lane, Kidlington OX51GB, UK" O
Development of Expected Flow Patterns october 1997 o:\3006w-9c.non:lt>101497 Revision 1
9.C-15 l s. l~ l: - LN , i 1 1 l Ro 10Y 10 4 10 8 jo' joa 3go jg2 $g4 1gs r
- Nu 1.108 2.201 4.430 8.754 32.045 156.85 840.13 3624.4 11226 i
9 i-i
- ig. . .
l 10s ,, , , i
,,a c
. log (Nu) .. , , p .. . 1- .. . . , ~~ i l I l- 1 I I I I
- s . s s- . . a w .
to 10 to 10 to to 10 10 10 log (Ro) 4 i g l Figure 9.C.16 Average Nusselt numbers as a function of the Rayleigh numbers for the . square enclosure with opposite vertical . walls at the different temperature and Pr=0.71 (air) 4 h i (according to N.C. Markatos and K.A. Pericleous/ Laminar and Turbulent Natural Convection in an Enclosed Cavity, Int. J. Heat Mass Transfer, Vol. 27, No. 5, pp. 755 772,1984) _ Development of Expected Flow Patterns _ october 1997 o:\3006w-9c.non:1b-101497 Revision 1 l o
9.C-16 4 O SOURCE OF THE UPWARD LINE JET . HORIZONTAL SURFACE
, a r ,,, ~ >>> % f y )
( u
~ + ,r j/,m l'Ims S
- s faQ , /+apo 8
%n a l @
o l'. li < cd ,f STRATIFIED CORE - W 4 ej f lY' )f 5 t /-Mo f , I
, t 4 = - J e e n , , , . , , , . ;
HORIZONTAL SURFACE SOURCE OF EHE DOWNWARD LINE JET Figure 9.C.1-7 Formation of the strati 6ed core in between two opposite vertical line jets (after Baines and Turner,1969)
" Reprinted with the permission of Cambridge University Press from Baines W.D. and Turner, J. SJ Turbulent buoyant convection from a source in a confined region, Journal of Fluid Mechanics, Vol. 37,1969; pp. 51-80, Copyright 1969, Figure 9.C.10" Development of Expected Flow Patterns cktober 1997 o:\3006w-9c.non:1t>101497 Revision 1
, 9.C-17 l i i ( j \ , i D 8e j X s l , . i q s s , .; z o y h( .
)
X t ' R) br , T3 (at X = 6) bz 4 Z=dH=1 i ! \ .
'I
- Jll ?%
g v-
'g ,
D l I'I }' -
"0 y M A.;h.f, ,7 --- = 0 by z,
o,
f .kh J..jj.\
N,
',,y b._Z , n
- .,' M ' OZ i H = y/H = 2 h 'X = x/H = 1 Tc (at X = 1)
Figure 9.C.1-8 Flow in an enclosure with vertical opposite walls at different
- temperatures (Ra = 10', Pr = 0.71).
3
- a) forward flow (towards Y=0 and Y=2)-
I streamlines through the points (X=0.5, Y=0.1, Z=0.49) and CGO.5, Y=1.9, Z=0.49) b) reverse flow (towards Y=1.0) - streamlines through the points
- (X=0.3, Y=0.8, Z=0.65) and (X=0.3, Y=1.2, Z=0.65) after (Mallinson and deVahl Davis,1977) i
" Reprinted with the permission of Cambridge University Press from Mallinson, G.D. and G. de Vahl Davis / Three-dimensional natural convection in a box: a numerical study, Journal of Fluid Mechanics, Vol. 83,1977; pp.1-31, Copyright 1977, Figure 9.C.8" Development of Expected Flow Patterns ocober 1997 ,
o:\300bw-9c.non:lb-101497 Revision 1 4 e
9.C-18 i
' ~
n , "- '" " ,"'"l u 9g*;. ::: p.y -
':::.f~r '
l
) . ""%;{'ll.'."..;':~2 m y
w,.
. u,~\snn:n,?!$h'-
c :>
4"nuon . ....=.. ; .g.'.! " i. - . ..
- 14. . . ~) C ~4 c ..=:::
~ ' % *4w z/H O V h.r ::::==.b k ' b:b-hk.[.. c~~'s ,dl ,. s 5 ,) \
0 l l l , g _= T, . Tc j, T't....; .::Sl'
'24 )
z/H o s .... .. W z=> ( 3
.s. '{;))
(:.::. .3o).\(5. e t r.n .:i c' 58
\.
si..g.:::,,/.%44gg 3
- p. .. . \
" "" r-o m ,' ;:::.A.: y$!,'.as ...0 -
a.,,,,:w mn:::. , f
- e. ..
E
. t.
J:::> _ . ' sum.c.: .._ dH WH Figure 9.C.1-9 Internal waves in the square cavity - fluctuations in the temperature field at Ra = 2'10' and Pr=0.71 (air). Circle with the arrow (in the middle) presents the direction of the consecutive temperature fields. Contour lines corre~por.d to +/ -0.0005'AT, +/ -0.001*AT,
+/-0.0015'AT and +/ -0,002'AT (the dotted contour lines correspond to negative values, where T3= AT/ 2 and T, = -AT/ 2). (After Janssen and Henkes,1995) " Reprinted with the permission of Cambridge University Press from Jansenn, R.J.A. and R.A.W.
Henkes/ Influence of Prandtl number on stability mechanisms and transition in a differentially heated square cavity, Journal of Fluid Mechanics, Vol. 290,1995; pp.319-344, ( - pyright 1995, Figure 9.C.4" Development of Expected Flow Patterns october 1997 o:\3006w.9c.non:1b-101497 Revision 1
9.C-19 O ADIABATIC WALL _w 2 _
>v W - ,
7 WAVES g j HOT INTRUSION j
- 4 k k o l H : A l
O ' O Z COLD BTRUSION O O WAVES
~ 1 ^L - ~K C s ~
i
~
A ' ADIABATIC WALL Figure 9.C.1-10 Temperatures for the initial solution with the hot and cold intrusions and boundary layer waves presented (after Armfield and Janssen,1996)
" Reprinted with permission from Int. J. Heat and Fluid Flow, Vol.17, S. Armfield and R. Janssen/ A direct boundary-layer stability analysis of steady-state cavity convection flow, pp. 539 546,1996. Elsevier Science Inc."
Development of Expected Flow Pattems % ,3997 c:\3006w-9c.non:1b-101497 gg 3
9.C-20 e
# Y2 l.
T NRCAT1YELY BUOYANT JET C
~~
Il FFW^% , d g l p f NEGATn'dLY j
'f lY O '
j /smANT JET tr su' t
.s 11 %.2 '
k ,l $smr.dhy*n l ' samar.d hy.n p gy ,,
'[
AREA dF THE j , 7 b 'f mCREASED l N '[ f P HOT JET l L S. Jd j- )Lg i < f NWdy -l J T':i1 y !, BUOYANrJET I 'l ,
$7 k j k j REntaEe.10N OF AREA OF THE a3) ==r ="" b)
Figure 9.C.1-11 Formation of the downward negatively buoyant jets a) negatively buoyant jet redirected inside the dome region b) negatively buoyant jet penetrating the below deck region Development of Expected Flow Patterns I" o:\3006w-9c.non:1b-101m
9.C 21 z 9.C.1.1.3 -References
. 1. - Akino, N., Kunugi, T., Shiina, Y., Sek!, M. and Okamoto, Y. (1989)- " Natural convection in a horizontal silicone oil layer in a circular cylinder heated from -;
below and cooled from above", (in Japanese), Trans. Jpn.' Soc. of Mech. Eng. 55 509 no.1989-1), no. 88-0901 B: 152-158,1989. . , 2._ - Allard, F., (1992)
" Effects of thermal boundary conditions on natural convection in thermally driven
,. cavities", in - Turbulent Natural Convection in Enclosures, A Computational and Experimental Benchmark Study, (Eds.) R.A.W.M Henkes and C.J. Hoogendoorn, Editions Europeenns Thermique et Industrie, Paris, pp. 234 256, i
- 3. S. Armfield and R. Janssen (1996) "A direct boundary-layer stability analysis of steady-state cavity convection flow" Int, J. Heat and Fluid Flow, Vol.17, No. 6, December 1996.
. 4. W.D. Baines and J.S. Tumer (1%9)- L ' Turbulent buoyant convection from a source in a confmed region" J. Fluid Mech., - Vol. 37, part 1, pp. 51-80.
- 5. - Catton, I., (1978)
" Natural convection in enclosures" Proc. 6th Int. Hea$ Transfer Conf., Vol. 6, pp.13-31.
- 6. Dzodzo M. B., (1993)
" Visualization of laminar natural convection in romb-shaped enclosures by means of liquid crystals" in " Imaging in transport processes", (ed. S. Sideman and K. Hijikata),
Chtpter 15, pp.183-193., Begel House, Inc.,1993, i- 7. R. Frederick and A. Valencia, (1995), " Natural Convection in Central Microcavities of Vertical Pinned Enclosures of very High Aspect Ratios," Int. J. Heat and Fluid Flow,
- Vol.16, No. 2, April 1995, pp.114-124.
i
- 8. T. Fusegi, J. M. Hyun, K. Kuwahara, (1992)
" Numerical simulations of natural convection in a differentially heated cubical enclosure with a partition" Int. J. Heat and Fluid Flow, Vol.13, No. 2,-June 1992, pp 176-183. - 9. T. Fusegi, T. M. Hyun, (1994) " Laminat and Tra uitional Natural Convection in an enclosure with complex and relistic conditions," Int. J. Heat e.nd Fluid Flow, Vol.15, No. 4, August 1994, pp. 258-268.
O-O Development of Expected Flow Pattems october 1997 o:\3006w.9c.norrib-101497 ~ Revision 1
9.C022 _
- 10. A. D. Garrad and M.A. Patrick, (1983)
"The velocity field produced by a submerged jet directed upwards at a free surface" Int.
J. Heat Mass Transfer, Vol. 26, No. 7, pp.1029-1036. g
- 11. K. Hanjalic, S. Kernjeres and F. Durst, (1996)
" Natural convection in partitioned two-dimensional enclosures at higher Rayleigh numbers" Int. J. Heat Mass Transfer, Vol. 39, No. 7, pp.1407-1427,1996
- 12. T.1. Heindel, S. Ramadhyani, and F. P. Incropera (1995)
" Conjugate natural convection from an array of discrete heat sources: part 1 - two and three-dimensional model validation" Int. J. Heat and Fluid Flow, Vol.16, No. 6, December 1995, pp.501510.
- 13. Hiller W.J., Koch St., Kowalewski T.A., (1988)
"Simultane erfassung von temperatur und geschwindigkeitsfelden in einer thermischen konvektionsstromung mit ungekapselten flussigkristalltracern, DGLR - Workshop, 2D-Mestechnik,1988.
- 14. Hiller W.J., Koch St., Kowa!+waki T.A., (1989)
'Three-dimensional structures in laminar natural convection in a cube enclosure" Exp.
Therm. Fluid Sci., Vol. 2, pp. 34-44.
- 15. C. J. Ho, Y. T. Cheng and C, C. Wang, (1994)
" Natural convection between two horizontal cylinders inside a circular enclosure h
subjected to external convection" Int. J. Heat and Fluid Flow, Vol.15, No. 4, August 1994, pp 299-306.
- 16. Hoogendoom, C.J., (1986)
" Natural crnvection in enclosures" Proc. 8th Int. Heat Transfer Conf., Vol.1, pp.111-120.
- 17. J.M. Hyun and B.S. Chol, (1990)
" Transient natural convec3on in a parallelogram-shaped enclosure", Int. J. Heat and Fluid Flow, Vol.11, No. 2, June 1990, pp.129-134,
- 18. R. J. A. Janssen and R.A.W.M. Henkes, (1995)
" Influence of Prandtl number on instability mechanisms and transition in a differentially heated square cavity," J. Fluid Mech., Vol. 290, pp. 319-344.,1995
- 19. K. Kapoor and Y. Jaluria, (1993)
" Penetrative convection of a plane turbulent wall jet in a two-layer thermally stable environment: a problem in enclosure fires" Int. J. Heat Mass Transfer, Vol. 36, No.1,
! pp.155-167,1993. 6 Development of Expected Flow Patterns october 1997 o:\3006w-9c.ncn:1b.101497 Revision 1
__._ . _ _ _ _ _ . . . . _ . _ _ _ . . _ _ . _ _ _ _ _ _ _ ~ _ . _ . . _ . .
- 9.C-23
- 20. _ A. Kurosawa, N. Akino, T. Otsuji, S. Kizu, K.'Kobayashi, K. Iwahori, T. Takeda and
_Y. Ito, (1993) -
" Fundamental. study on krmo-hydraulic phenomena concerning passive safety of r advanced _ marine reactor" Journal of Nuclear Science and Technology, Vol 30 [2],
pp.131 142,- February 1993, i .
- 21. G.D. Mallinson and G. de Vahl Davis, (1977)-
l_ "Three-dimensional natural convection in a box: a numerical study" J. Fluid Mech., Vol.- 83, pp.1-31. l 5
- 22. - N.C. Markatos and K.A. Pericleous, (1984)
- " Laminar and turbulent natural convection in an enclosed cavity" Int. J. Heat Mass l Transfer, Vol. 27, No. 5, pp. 755-772.
L 23. E. S. Nowak and M. H. Novak, (1994) }- " Vertical partitions in slender rectangular cavities" Int. J. Heat and Fluid Flow, Vol.15, j .No. 2, April 1994, pp.104-110. i 24. Ostrach, S., (1972)
" Natural convection in enclosures" Advances in Heat Transfer, Vol. 8, Academic Press, -
New York, pp. 161-227.
- 25. Ostrach, S., (1982)
" Natural convection heat transfer in cavities and cells" Proc 7th Int. Heat Transfer Conf.,
l Vol 1, pp. 365-379.
- 26. F.E. Peters, (1994)
[ WCAP-14135, " Final Data Report for PCS Large-Scale Tests, Phase 2 and Phase 3",
- Revision 1, April 1997.
- i
- 27. P.F. Peterson, (1994)
. " Scaling and analysis of mixing in large stratified volumes" Int. J. Heat Mass Transfer; j Vol.- 37, Suppl.1, pp. 97-106.
- 28. B. Porterie, M. Larini, F. Giroud and J.C. Loraud (1996) -
" Solid-propellant fire in an enclosure fitted with a ceiling safety-vent" Int. J. Heat Mass
- Transfer, Vol. 39, No. 3, pp. 575-601.
'29. - R.M. C So and H. Aksoy, (1993) "On vertical turbulent buoyant jets" Int. J. Heat Mass Transfer, Vol. 36, No.13, pp. 3187-3200.
V Development of Expected Flow Patterns october 1997 o:\3006w-9c.non:1b-101497 Revision 1 2-
9.C-24
- 30. L Wolf, H. Holzbauer, T. Cron, (1994a)
" Detailed Assessment of the HDR-Hydrogen Mixing Experiments Ell" International Conference on New Trends in Nuclear System Thermohydraulics, Pisa, Italy, May 30th -
June 2nd, vol. 2, pp. 91 - 103. h
- 31. L. Wolf, H. Holzbauer, M. Schall, (1994b)
" Comparisons between multi-dimensional and lumped parameter Gothic-contalmr.ent analyses with data" International Conference on New Trends in Nuclear System Thermohydraulics, Pisa, Italy, May 30th - June 2nd, vol. 2, pp. 321 - 330.
O O Development of Expected Flow Pattems october 1997 o:\3006w-9c.non:1b-10149/ Revision 1
[ .
- 9.C-25 4
o - 9.C.1.2? CIRCULATION PHENOMENA -i
- p; j
- Circulation processes inside enclosures are the result of natural or forced convection effects. J Forced convection inside an enclosure is promoted using devices such as fans, nozzles, or sprays
- _ _ of liquid droplets. PCS applications are of primary interest, since no credit is taken for active' l systems in the AP600 (non-safety grade fan coolers) design basis analysis.
2
- A review of possible flow patterns 'due to natural convection effects is presented. Natural
- - convection is generated if
1)-~ The upper boundary is at a lower temperature than the lower boundary or opposite vertical boundaries are at different temperature, as well as for other similar combinations 1 of temperature boundary conditions (or imposed heat flux conditions) at the outside and [ inside surfaces. L i- 2) A higher concentration of the lighter or heavier components of a mixture is maintained - near the lower or upper boundaries of the enclosure, respectively. 3). A lighter fluid is released (permanently, or from time to time) from a source which is .
- closer to the bottom of the enclosure.
T
- 4) The shape of the enclosure promotes natural convection (t>gether with the distribution of other boundary conditions).
- 5) The distribution and size of the horizontal and-vertical internal openings allows or enhances (as with a chimney or staircase effects) the formation of fluid flow patterre due to the natural convection.
- 6) If the internal heat sources (sinks) are positioned in the lower (upper) portions of the enclosure.
Under the conc itions above (or a combination of them), natural convection causes circulation inside the enclosure. The convection increases the intensity of heat and mass transfer, therefore
~ increasing the heat released from the containment. The intensity of heat transfer depends upon the location of the heat sinks and sources, which can exchange positions due to the transient effects. The velocity and temperature profiles inside the formed boundary layers (wall jets) influence the rate of heat transfer due to the convection. - Wall jets entrain the surrounding atmosphere and contribute to better. mixing. In the regions with a higher steam concentration, the increase in the heat transfer rate and the effects of entrainment occur due to the condensation -
inside the boundary layers. O G Development of Expected Flow Patterns october 1997 cA3006w-9c.norulb-101497. Revision 1 i'-
. . , , . . - - . + , . - - . , .
9.C 26 Another contributing factor that promotes circulation inside an enclosure is the interaction of the enclosure atmosphere with the penetrating buoyant plumes or jets and wall layers. In the case of a containment vessel, the plumes or jets could be generated by a LOCA or MSLB. If the break position is inside a narrow corridor or surrounded by additional equipment, the kinetic energy of the jet is dissipated and steam rises in the form of a buoyant plume. The rising plume entrains the surrounding gas and results in circulation inside the volume of the enclosure. If the break position is open and the jet is directed upward, both the kinetic energy of the jet and the buoyancy forces contribute to penetration into the atmosphere. The higher soeeds of the jet affect a greater portion of the volume and both entrainment of the surrounding gas and circulation is stronger. 9.C.1.2.1 Circulation Phenomena Due to the Presence of Boundary Layer = (WallJets) and Buoyant Plumes Formed as a Consequence of Natural Convection Effects Natural convection flow is the most often generated by different temperatures or heat fluxes imposed on the boundaries of an enclosure. Various dist-ibutions on the boundasies produce various flow patterns end temperature fields. Section 9 C.1.1 discusses boundary temperature distributions (upper / lower horizontal plates at the higher / lower temperatures) that produce static stratification. Section 9.C1.2 discusses the case where vertical opposite sides are at constant, but different temperatures. If Rayleigh 4 numbers are greater than 10, this condition produces a recirculated region near the walls and a stratified core of the enclosure. Figure 9.C.1-12 presents a case known as Rayleigh-Benard convection. The upper horizontal boundaries are at the lower temperatures (or cooled). The flow patterns formed depend upon the temperature difference and geometry of the enclosure (in fact the value of the Rayleigh number). For the smaller Ra numbers, vortical cells are formed. An increase in the Ra numbers produces a grez.ter number of vortical cells that start to oscillate, periodically changing the size and intensity. A further increase in the Ra number results in chaotic flow, and produces vertical plumes which reach the opposing horizontal sides of the enclosure. The flow patterns and possible bifurcations produced dudng the transition from the laminar to turbulent (chaotical) flow regimes are described in Koschmieder,1993, Yang,1988, and Ozawa et al.,1992. Some txperimental results (flow patterns and temperature fields) are presented for laminar flow regimes by M. Dzodzo et al.,1994 and M.J. Braun et al.,1993. Flow patterns for turbulent and chaotic flow between two horizontal plates at different temperatures are described in Akino et al.; 1989. O Development of Expected Flow Pattems october 1997 o;\300bw-9cnon:1b-101497 Revision 1
9.C-27 '- Flow in the Hele-Shaw cell is presented as an example of natural convection between two i horizontal plates. A Hele Shaw cell has a square cross-section, but it is narrow in one of the horizontal directions so that three-dimensional convection effects are suppressed (see Figure-9.C,1-13). The upper and lawer horizontal sides are at the lower and higher temperatures,- respectively. I Consecutive flow pattems and temperature fields for a Hele-Shaw cell with various Rayleigh' I i~ numbers are presented in Figure 9.C.1-14 (after Buhlet et al.,1987). If the value of the Rayleigh i number is greater than 4'10', oscillatory f'aw patterns with four vortical cells are present.' The
- large and small vortices expand and contract periodically (see Figure 9.C.1-14). At high Rayleigh ;
l numbers (above 5.9'10'), a reverse transition from the oscillatory to the steady flow patterns
- occurs. This phenomena is probably due to suppressed three-dimensional convection effects.
f For cubic or cylindrical enclosures, with the upper and lower horizontal surfaces at the lower , - and higher temperatures, respectively, three-dimensional convection effects produce turbulent j (chaotical) flow (see Figure 9.C.1-15). In the paper by Akino et al.,1989, the turbulent flow l regime starts at a Rayleigh number of 2'10' (Pr = 200).- For fluids with a Frandt! number close ! to one, the transition to turbulent flow regime occurs at a smaller Rayleigh number (Ra - 10 4). L l The flow pattern consists of vertical buoyant plumes detached from the horizontal sides. The j vertical plumes reach opposite sides of the enclosure and generate opposing plumee (see Figure p 9.C.1-16). Temperature gradients near the horizontal surfaces are high, while temp?ratures in j_ Q j the core of the containment are almost uniform. Figu 2 9.C.1-17a' illustrates an example where the temperature in the middle of the enclosure oscillates between 26 and 29'C with ka = 9.38'10'
- (Ts=35'C and T,=20*C). The highest temperature is registered during the rise of the hot plume
- and the lowest temperature is registered during the downward penetration of the cold plume.
L The amplitude of the temperature oscillations in the middle of the enchsure is three degree Celsius. The temperature interval between 26 and 29'C represents 0.2*(T3 -T,) or 20 percent of j the maximum temperature difference. The temperature in the middle of the enclosure is (27.5*C)
- +/- 1.5'C.
i Rayleigh-Benard convection is relevant to the containment. In the case of a LOCA or MSLB, the { upper portion of the dome and vertical sides are cooled. If the temperature below the ceiling , is 9'F lower than temperature of the incoming steam (at the deck level), the Grashof number I (based on the height of the containment, H,=109 ft) is Gr,=2.2 x 10". This Grashof number is in . the range of the chaotic flow, with the upward and downward plumes (because Gr, > 104 /0.71), i Maintaining the vertical walls of the containment at the lower temperature also promotes j downward vertical plumes near the walls due to separation of the vertical boundary layers (see j Figure 9.C.1-17b). Development of Expected Flow Patterns - october 1997 l o:\3006w-9c.non:1b 101497 Revuton !
9.Ca28 9.C.1.2.2 Circulation Phenomena Due f.o the lateraction With the Ilot Buoyaat Plumes and Jets - The presence of a hot buoyant pume or a jet of the hot steam during a LOCA or MSLB O contributes to the circulation of the containment atmosphere by entraining the surrounding air and other gases. In the case of jet inflow, additional entrainment and circulation are generated by tia jet kinetic energy. Depending upon the strength (initial velocity and mass flow) and direction of the plume or jet, various flow patterns inside the containment are possible. Interaction of the vertical downward plumes generated due to the natural convection (cooling of the shell) produce turbulent flow. This results in good mixing of the dome atmosphere. Examples of vertical plumes and jets are presented by Garrad and Patrick,1983, So an Aksoy,1993, and Porterie et al.1996. The scaling and analysis of circulation in large stratified volumes is presented by Peterson,1994. O I O Development of Expected Flow Pattems october 1997 l o:\3006w.9c.non:lt-101497 Revision 1
9.C 29 O Te L /
" / h 138',,Q@ '
{ s s I
/ ii. . t.i. .i i is ' ,ili -
l I i
/ is i e gi U - .
g-f
/ / .Y/ dd') \N
h/ / = y Th Figure 9.C1-12 Rayleigh-Benard convection example Developn.ent of Expected Flow Patterns y , 3 ,3997 o:\3006-cl.non:1b-101497 Revision 1
, ._ . .. . ~ . .. -_ . ,_ . . . - _ - - . . - . _ _ . - - .
9.C 30 ) 9 i d ! Z i 0 C
\ / p/ \ / /
u
/ /
i / /
- / /
/ /
C" / / /
/
\ u// /
/ 9 / / / / ' < / n /
. //
/ /
7 ~7-X - h i J Figure 9.C.1-13 Hele-Shaw cell Development of Expected Flow Patterns october 1997 o:\3006 's.non:1b 101497 Revision 1
9C31 i STREAMLINES ISOT4fERMS l 5 Ra < l.27*10'
; .% llEAT CONDUCTION , ;. : i b- .. 3 1.27*10' < Ra < 2.9e*10' PATTERN A
- s. O O p 1.95*10' < Ra < 3.98*10' I PATTERN B fii ' -C
3.93*10' < Ra < 3.50*10' PATTERN B OSCILLATORY
; A- , .. ., 5 l l ;- '~^* 3.50*10' < Ra < 5.90*10'
, ; I SUBilARMONIC OSCILATORY y .; ; , . q. PA1 TERN B
~ '
l ll 3.77*10' < Ra M l STATIC PA1 TERN B Jg
. L..$M ,.~. a J - 4, e Figure 9.C.114 Steady end oscillatory convection in fiele Shaw cell (after Buhler et al.,1987) " Reprinted from LBuhler, P. Ehrhard, C. Gunther, U. Muller and Zimmermar J Natural convection in vertical gaps heated at the lower side en ewi atal and numerical study, IITU-Vol. 94, AMD Vol. 89, Bifurcation Phenomena in inwrr Processes and Convection, Winter Annual Meeting of the American Society of Mechanical E- aeers, Boston, Massachusetts, December 13 20, 1987" Developmen* of Expected Flow Patterns octotwr 1997 oAXEW<1.rmlb 101497 Revision 1
9.C 33 O COLD PLUMES T s
< i ) O
(% O,
\
tj 1( a t ( o Cf\ z $ jN 0L
\
J ( ~ TH HOT PLUMES Figure 9.C.115 Turbulent (chaotical) flow with hot and cold plumes interactions (plane cross section of the three-dimensional enclosure is presented) (according to Figure 6 in N. Akino, T. Kunugi, Y. Shiina, M. Sekl, Y. Okamoto/ Natural convection in a horizontal silicone oillayer in a circular cylinder heated from below and cooled from above", Trans. Jpn. Soc. of Mech. Eng. 55 509 no.19891), no. 88-0901 B:, pp.152158,1989 with permission from Norio Akino) Development of Expected Flow Pattems octotwr 1997 st\3006<1.non:ll>101497 Revision 1
9.C 33 O COLD PLUME COLD PLUME j i > ff f i HOT LAYER 4' HOT LAYLR (o) (b) l COLD P1UME COLD PLUME HOT < HOT PLUME O \ l xx HOT LAYER (c) x (d) A 't 4 Figure 9.C.116 Interaction of hot and cold plumes (after Aqino et al.,1989) (according to Figure 12 In N. Akino, T. Kunugi, Y. Shiina, M. Seki, Y. Okamoto/ Natural convection in a horizontal silicone oillayer in a circular cylinde. heated from below and cooled , from above", Trans. Jpn. Soc of hiech. Eng. 55 509 no.19891), no. 88 0901 B:, pp.152158,1989 - with permission from Norio Akino)
- Development of Expected Flow I'atterns october 1997 o-\3006<1.nort1b-101497 Revision 1
9.C 34 1 :, , , ,
'y / }
AVERAGE g s
' TEMPERA'I'URES COLD PLUME. j $
TEMPERATURES , { 5 i 50.5
) ,,
4 e I i* l W 8
\ HOT PLUME I N ! TEMPERATURES \.
E! 'm ': 0 20 25 30 35 TEMPERATURE C Figure 9 C.117a Vertical temperature distribution inside the cylindrical enclosure with lower and upper horizontal plate at higher and lower temperatures, respectively (Ra = 9.3890', Pr = 200) (according to Figure 14 in N. Akino, T. Kunugl, Y. Shiina, M. Seki, Y. Okamoto/ Natural convection in a horizontal silicone oillayer in a circular cylinder heated from below and cooled from above", Trans. Jpn. Soc. of Mech. Eng. 55 509 no.19891), no. 88-0901 B:, pp.152158,1989 - with permission from Norio Akino) Development of Expected Flow Patterns cktober 1997 o:WXXRI.non:1b-101497 RevWon 1
. , 9.C 35 i
l s' r i E Cs i - COLJ ?._UME o i
/ COL) P_JME l Tc
- O (
i Z L.t i t i
=
- X HOT LAYER 1
i i i j Figure 9.C.117b Generation of the cold plumes due to the brake (separation) of the vertical i boundary layers near the cold vertical walls. Development of Expected Flow Patterns Nder 199/ 'I , 0:\ht.non:1b-1014 Revisicm 1
's C.w3 References
- 1. N. Akino, T. Kunugi, Y. Shina, M. Seki, Y. Okamoto, (1989)
" Natural Convection in a Horizontal Silicone Oil Layer in a Circular Cylinder Heated h
from iblow and Cooled from Above" Nippon Kikai Gakkal Ronbunchyu,55 Vol. 509 (19891), FP.152 -158,
- 2. M. J. Braun , M. B. Dzodzo, S. B. Lattime, (1993)
" Automatic computer based non-intrusive temperature measurements in laminar natural convection using thermochromic liquid crystals in enclosures with variable aspect ratio" FED-Vol.172, Experimental and numerical flow visualization, The 1993 ASME Winter Annual Meeting, New Orleans, Louisiana, November 28 December 3,1993, pp.111 119,
- 3. L Buhler, P. Erhard, G. Gunther, U. Muller, G. Zimn.ermann, (1987)
" Natural convection in vertical gaps heated at the lower side - an experimental and numerical study. In: Bifurcation Phenomena in Thermal Processes and Convections (eds.H.M. Bau, L.A. Bertram, S.A. Lorpela) ASME, HTD-Vol. 94/AMD-Vol.89,67 74.
- 4. M. Dzodzo, M.J. Braun, S.B. Lattime, (1994)
"A non-intrusive computer automated investigation of natural convection using thermochromic liquid crystals and comparison with numerical simulation" (ed.
G.F. Hewitt) Proceedings of The Tenth International Heat Transfer Conference, Brighton, UK, Volume 2,2-MT-6, pp. 225-230.
- 5. A.D. Carrad and M.A. Patrick, (1983)
"The velocity field produced by submerged jet directed upwards at a free surface", Int.
J. Heat Mass Transfer, Vol. 26, No. 7, pp.1029-1036,
- 6. E. L. Koschmieder, (1993)
"Benard Cells and Taylor Vorticles" Cambridge University Press
- 7. M. Ozawa, U. Muller, I. Kimura and T. Takamori, (1992)
" Flow and temperature measurements of natural convection in a Hele-Shaw cell using a thernwsensitive liquid-crystal tracer" Experiments in Fluids, Vol 12, pp. 213-222,
- 8. P.F. Peterson, (1994)
" Scaling and analysis of mixing in large stratified volumes" Int. J. Heat Mass Transfer, Vol. 37, Suppl.1, pp. 97-106.
- 9. B. Petterie, M. Larini, F. Giroud and J.C. Loraud, (1996)
" Solid propellant fire in an enclosure fitted with a ceiling safety vent" Int. J. Heat Mass Transfer, Vol. 39, No. 3, pp. 575-601.
1 Development of Expected Flow Pattems cktober 1997 oA*4006w 9c.nnn:lb 101497 Revuton1
i 9.C.37
- 10. R.M.C So and 11. Aksoy, (1993)
"On vertical tuibulent buoyant jets" Int. J. lient Mass Transfer, Vol. 36, No.13, pp. 3187-O 3200,
- 11. K.T. Yang, (1988)
' Transitions and Bifurcations in Laminar Buoyant Flows in Confined Enclosures" J. of Heat Transfer, November 1988, Vol.110, pp.1191 1204.
i i O <
+
i I 6 Development of Expected Flow Patterns october 1997 c:\3006w-9c.rm1b-101497 Revision 1 l
9.Co38 9.C.1.3 IMPORTANT DIMENSIONLESS GROUPS 9.t.1.3.1 Important Dimensionless Groups for Stratification and Circulation Phenomena inside Enclosures h For natural convection, the ratio of the buoyancy to viscosity forces is the most important dimensionless (Pi) group, ne Grashof number defines the ratio of the buoyancy to viscosity forces: S Gr = gp (T,-T) H' g(p,-p,) H
=
v2 p,v 2 Natural convection correlations of ten use the Rayleigh number instead of Grashof number, where the Rayleigh number Ra is defined as: Ra = g p (T,4,) W = g(p,9,) W = Gr .y = GrPr av p,a v a Using the Rayleigh number reduces the number of dimensionless groups in the correlations for natural convection. The appearance of Prandtl number inside some correlations could be
- svolded, ne Prandtl number is based on fluid properties, i.e., the ratio of kinematic viscosity to thermal diffusivity.
Pr = 1 a When considering the interaction between the hot buoyant plume and the cold vettical wall boundary layer, the Grashof and Rayleigh numbers are defined with H, as the characteristic length. g (p,-p) H,' Gr = p, This applies to flows generated inside the enclosures with the opposite vertical walls at the different temperatures. It also applies to flow caused by two opposing vertical jets (between the two horizontal plates). The AP600 has a combination of the two cases. O Development of Expected Flow Pattems october 1997 c:\XXEw-9c.non:ltwl01497 Revision 1
9.C 39 Upward flow is caused by the buoyant plume, while downward now is caused by the lower temperatures of the vertical wall if the initial kinetic energy of the plume is small, this Grashof number gives an indication of the formed flow pattern and heat transfer due to the two opposing vertical flow paths. The fortmtion of a recirculating stratified core between the vertical jets is related to this parameter as well. When considering the interaction between the cold ceiling and the hot rising plume (at the bottom of the enclosure), the height of the upper-deck region H, can be used as a characteristic length. The Grashof number is:
'Y' '
l l Gr' = 0 p,y The value of this Grashof number indicates the status of the Rayleigh Benard convection. If the values are above 104, it is possible to form periodic vertical downward plumes which detach from the ceiling. The conditions described above interact. The overall flow pattern is expected to be a superposition of the flow patterns described for enclosures with horizontal and vertical temperature gradients. The prevailing flow pattern is estimated from the ratio of the two Grashof numbers already defined: O Gr, , (p, - p,) 11,3 p, Gr, (p, - p) 11,5 p, Note that both dimensions of the large-scale test (LST) installation (11, and H,) are scaled to AP600 dimensions. 'Iherefore,if the ratio of relative densities (in vertical and horizontal directions) is the same, the flow patterns obtained in LST experiments can be applied to the AP600. Even small temperature differences between the shell and the atmosphere inside a containment produce large Grashof numbers. For example, a temperature difference of 9'F results in Gr,= 2.2'10" and Gr, = 4.7'10" for H, = 109 ft and Hy = 65 ft, respectively, in the case of LST, a temperature difference of 9'F results in Gr, = 3.9'10 and Gr, = 7.2'10' for H, = 13.2 ft and H, = 7.5 ft, respectively. If the Grashof numbers are greater than 10", the Nusselt nuc an be obtained by applying the correlation for turbulent free convection. For jets and buoyant plumes that penetrate the containment, the ratio of inertia forces and buoyant A forces influences the entrainment of surrounding gases. If the initial velocities are high, a constant U Development of Expecttd Flow Patterns october 1997 o-\3006w 9c. rum:lt>101497 Remkm 1 J
9.C-40 spreading angle indicates a jet. As the jet velocities decrease, upward motion results from buoyant forces. Buoyant plume behavior is indicated by different spreading angles at each level. The Froude number represents the ratio of the inertia to gravity forces, or the ratio of kinetic O energy to potential energy: UH2 pU8 v Re8 Fr = _U8= = = gH gapH g op H' Gr pv 2 l For buoyant plumes and jets, the Froude number can be defined as: p'U'3 Fr" = 8(P.P,)d, where the characteristic length is the initial diameter of the jet or plume. The source velocity and density have the subscript (o), while the ambient density has the subscript (a). The elevation of the transition from a forced jet to a buoyant plume is calculated (Peterson,1994 and Spencer, 1997) from the expression: z"" = Frg 'd (p ) The ratio of the square of the jet Reynolds number to the containment Grashof number is a volumetric Froude number: Fr' = p,U,2 d,8 g (P.-P,) H ' If the volumetric Froude numbers are much greater than one, the inertia forces dominate. The inertia forces unstabilize stratified layers, promote circulation inside the containment, and contribute towards the better mixing. However, Peterson,1994, proposes that the jet or plume is not able to disturb the stratified vertical density gradients if: O Development of Expected Flow Patterns Octobu 1997 oA3006w-9c.non:1b 101497 RevWon1
- - ~ - - . - . . . ~ _ .
9.C-41 l 8 - d* Fr,<(14 ) 4gaH O where (a) is Taylor's jet entrainment parameter and where a = 0.05 = constant, j 1 I For volumetric Froude numbers less then one, the inertia forces are not dominant and are not able to unstabilir.e stratified layers inside the containment. Therefore, the buoyancy effects are more important than inertia effects. The reciprocal value of the Froude number or Richardson number is the appropriate dimensionless group. g(p,-p,)H 8 p,U,8 d 2 1 Since inertia effects of the plume are not important (Reynoids number of the plume la small), l only Grashof numbers Gr, and Gr,will influence the flow pattern. q
- Another important factor is the position of the jet (plume) or heat source release location. The ratio of the release point level, H,, to the height of the containment, H,, describes the relative position of the jet _ (plume) or heat source:
o t
' If H,/H,is less than 0.2, the release location is considered low. A global circulation flow pattern -affecting the entire containment is most likely formed, if H,/H,is greater than 0.5, the release elevation is high and stratification effects may occur in a portion of the volume. The result may be that only the upper portion of the enclosure is aflected by global circulation, while the lower may be stratified. Such stratification may be stagnant.- In stagnantly stratified regions, no .
entninment into wall boundary layers or buoyant plumes occurs, and thus little or no vertical' mixmg occurs, while in recirculating stratified regions vertical mixing can be strong and can greatly reduce vertical density gradients. Specified criteria for the H,/H, ratio are based on the international experimental database which
'is presented in the next chapter.
9 C.1.3.2 References P. F. Peterson, (1994),
" Scaling and Analysis of Mixing in Lecge Stratified Volumes," Int. J. of Het and Mass Transfer, Vol. 37, Supplement 1, pp. 97-106,1994 Development of Expected Flow Patterns october 1997 c:\3006w-9c.novelb 101497 Reviskm 1
9.C-42 D. R. Spencer, (1997), " Scaling Analysis for AP600 Containments Pressure During Design Basis Accidents," WCAP 14845, Revision 2, June 1997. O l l O O Development of Expected Flow Patterns Octoter 1997 l, o:\3006w-9c.non:ltwl01497 Reviskm 1 l i
9.C-43 9.C.1.4 EXPECTED FLOW PATITRNS FOR AP600 O V 9.C.1.4.1 Simpilfied Representation of Circulation Regions During Post Blowdown LOCA in AP600 and LST The AP600 containment and the large-scale test (LST) facility include five primary flow regions ! (Peterson,1997 -letter to Woodcock). The regions are presented in Figure 9.C.118 showing a control volume that extends to the condensed fluid film surfaces. This figure is useful for i structuring a discussion of circulation and stratification phenomena and for relating separate effects of enclosures tetts to the various regions. i The volumetric flow rates presented in Figure 9.C.118 at " quasi steady" conditions are: l
. Q,, the steam volumetric flow rate from the break,
- Q,, the flow rate of fluid entrained from inside the below-deck region ink, the steam generator compartment (equivalent to the flow rate delivered to the below-deck region due to the penetration of a portion of the wall boundary layers through the deck gap near the walls),
- Og the flow rate of fluid entrained into the plume in the above-deck region, Q., the flow rate of fluid entrained into the vertical wall boundary layers, Q,, the flow rate of steam condensed on the vertical walls (shown leaving the control volume), and Q,, the flow rate of steam condensed on the dome ceiling (shown leaving the control volume)
For the quasi-steady conditions, the steam flow rate entering in the containment volume Q. is equal to the summation of the steam flow rates condensed cm the dome ceiling Q, and vertical walls Q,. The distances presented in Figure 9.C.118 are. H,, the distance between the }et inflow position into the upper deck region and the dome springline elevation (in vertical direction), O Development of Expected flow Pattems october 1997 c:\3006w-9c.norrib 101497 Rwision 1
9.C-44 l i e lin, the distance between the break location and the jet inflow position into the upper-deck region (in vertical direction), and + 11, the distance between the vertical wall and the jet center (in horizontal direction). O The definitions of the regions relate well to the separate effects of the enclosure tests. Region I is below the operating deck level. In the AP600 configuration, connections exist between the below-deck compartments and the upper-deck region (dome). These connections allow the steam jet (plume) generated-entrainment into the break compartment to produce drculation through Region I. The volumetric flow from the lower to the upper deck regions is Q,. Jet entrainment and the slots around the circumference of the deck floor enable this circulation (see Figure 9.C.1 19). In the LST - LOCA experiments, the release point is also below the operating deck level. 110 wever, the compartment containing the release is not connected with the other below-deck compartments (see Figure 9.C.1-20). The simulated steam generator compartment is cormected only with the upper portion (dome) of the containment. Therefore, the jet injection location for the LST LOCA experiments is effectively at the top of the simulated steam generator compartment, where the flow enters the above-deck region, and entrained volumetric flow Q, is equal to zero (see Figure 9.C.120). The atmosphere in the below-deck compartment is a stably stratified region without recirculation. The heat and mass transfer in the below deck compartments are governed primarily by molecular diffusion. Region !! is defined as the volume between the springline elevation and a horizontal line above the operating deck clevation, and between the wall boundary layers (Region IV) and the plume (Region III). Two entrainment mecha.tisms remove fluid from Region II. Entrainments into the vertical jet (or buoyant plume) and the wa ' boundary layers are compensated for by the inflows from the upper and lower horizontal boundaries. In order to preserve mass continuity and to obtain inflow into Region II, the vertical velocity components (see Figure 9.C.1-21) are negative and positive at the upper and lower horizontal boundaries, respectively. 'Ihe fluid inside the Region 11 is recirculating (see Figure 9.C.122), yet has a quasi-steady dp/dz maintained by balance between the buoyancy and the two entrainment mechanisms. Therefore, Region Il can be called a recirculating stratified region. The horizontal density and concentration gradients are small, but significant recirculation flow exists due to the entratnment into the free and wall jets (see Peterson,1997). Region 11 can be considered as a region where the vertical density, temperature and concentration gradients are dependent on the values of the volumetric Froude numbers (for free jets or plumes) and Grashof (Rayleigh) numbers (for wall boundary layers). This is similar to the case of an enclosure with opposite vertical walls at different temperatures (see section 9.C.1.1.2.1). The recirculation and entrainment from the Region Il contributes to a decrease in the vertical temperature, density, and concentration gradients. O Development of Expected Flow Pattems October 1997 o:\3006w.9c.non:1b 101497 Revision 1
9.C-45 Region 111 contains free jets (plumes) which transport fluid in the vertical direction. The upward motion of a jet (or plume) produces entrainment from Region II. As a result, the jet (or plume) spreads, reduces velocity, and dilutes (decreases the temperature and concentration difference between the core of the jet and the surrounding atmosphere Region 11). Region IV contains wall boundary layers widch also provide transport in the vertical direction. The entrainment into the wall boundary layer transports steam into Region IV. The entralnment from Region 11 into the wall boundary layers enhances recirculation inside the Region 11. This contributes to a decrease in the vertical temperature, density, and concentration gradients inside Region 11. Region V, the dome region, is between the containment ceiling and the elevation of the springlir 'decause the temperature of the containment ceiling is lower than the temperature of the atmosphere below the ceiling, downward flowing " ceiling plumes" are formed (see the Rayleigh Ilenard convection example of r.ection 9.C.1.2.1). The difference in the steam concentrations between the top of the Region V (immediately below the ceiling where condensation occurs) and the top of Region 11 are small due to the circulation (interaction) within Region V, caused by cold plumes falling from the ceiling and the hot plume reaching the ceiling of the dome. The downward plumes increase circulation and reduce gradients inside the dome, Region V. The downward "cciling plumes" interact with the uprising plume (from the Region 111). If the strength of the jet (plume) from the Region 111 is high, interactions occur inside ,. Region V and the influence of the downward plumes does not spread towards the lower regions. liowever, if the plume from Region III is not strong enough to produce good mixing inside Region V, the penetration of the downward " celling plumes" into die lower regions can disturb (from time to time) the recirculating stratified layers inside Region II. This tends to reduce the vertical gradients within Region II. If the plume is very weak or does not exist, the vertical downward " ceiling plumes" affect the entire volume of the upper deck region. The flow patterns formed are the result of superposition of Rayleigh Benard convection (described for the enclosure with cold upper and hot lower surface) and recirculating stratification (described for the enclosure with opposite vertical walls at different temperatures). The cold dome ceiling produces downward vertical plumes as in Rayleigh-Benard convection case, while cold vertical walls produce downward wall boundary layers. Due to continuity, the downward wall boundary layers tend to generate upward flow in the middle of the above-deck region. The wall boundary layer and the upward flow in the middle of the containment form a recirculation zone. Between the wall boundary layers and the upward flow in the middle of the containment, a recirculating stratified core is formed. This is similar to enclosures with opposite vertical walls at different temperatures. Note that although there is evidence from n enclosure tests that a stable non-zero vertical density gradient could exist in Region II, V entrainment flows cause circulation of fluid. Region II is not considered as stagnant. Development of Expected Flow Pattems october 1997 o:\3006w-9c,non:1t>101497 Revision 1 i
9.C 46
'Ihe prevailing flow pattera can be postulated (Rayleigh Benard or recirculating stratified) from the ratio of Grcshof nun. ers Gr,/Gr, [ defined for vertical Ap and distance H, (for Gr,) and horizontal Ap and distance ll,(for Gr,)). Note that turbulent Rayleigh Benard convection starts 4
at Ra, = Gr,Pr>10 (for Pr = 0.71, based on 3D enclosure experiments - see section 9.C.1.2.1), while g turbulent flow (with thin boundary layers and recirculating stratified but almost homogenized core) in enclosures with vertical walls at opposite temperatures starts at Ra,=Gr,Pr>10' (for Pr
= 0.71, based on 2D numerical simulations, see section 9.C.1.1.2.1). This indicates that for small values of Rayleigh numbers (10'< Ra, and Ra,<10'), Rayleigh Benard convection is dominant.
Turbulent and chaotical flow are dominated by falling vertical plumes (see F!gure 9.C.115). l For higher Rayleigh numbers (Ra, and Ra, > 10') combined with a weak source plume, in fact smaller Rayleigh number in horizontal direction, falling vertical plumes (see Figure 9.C.1-15) dominate the flow patterns. For the dominant jet (or plume), or high Rayleigh number in horizontal direction (Gr, high) and moderate Froude number, a recirculating stratified flow pattern prevails in Region 11 (see Figure 9.C.15). Higher and similar magnitude values of both t Rayleigh numbers (in vertical and horizontal direction) result in a flow pattern that is a superposition of the two described pattans (shown in Figures 9.C.1 15 and 9.C.15). Finally, for the case of the momentum-dominated jet (with high Froude number), the circulation flow pattern will be present in the entire volume of the containment (see Figure 9.C.111). 9.C.1.4.2 A Qualitative Model for Recirculating Stratified Region II ' A qualitative model of Region 11 is used to address the issue of recirculating stratification and circulation (Peterson,1997). The model is a coarse, first-principle representation of the effects of various volumetric flows and entrainment rates, it qualitatively examines the influence of various parameters on the difference in steam concentrations from the bottom to the top of the Region !! (AX). Because of the complexity of AP600 physics, two simplifying assumptions are used. Itis assumed that Region 11 is not influenced by falling plumes from Region V and that the recirculation effects inside Region 11 can be neglected. Both assumptions cause overestimated vertical steam gradients AX. Interactions between Region 11 and Region V that result from the penetration of the cold falling plumes (from Region V), improve mixing and decrease vertical steam t;radients AX. Recirculation inside Region 11 (established experimentally and numerically inside the core of enclosures) further decreases the vertical steam gradients. O Development of Expected Flow Patterns cktober 1997 oT30th9c.non:1twl01497 Revision 1
9.C 47 Peterson,1997, provides the following mass conservation equation for the thin horizontal layer i inside Re51on !! with area A(z) (see Figure 9.C.1-23): ! O p(z)A(z)dv(z) = p,(z)u,(z)p,(z)dz - p,(z)u,(z)p,(z)dz l l where v(z) is the vertical velocity, and u,(z) and u,(z) are the entrainment velocities into the j steam plume and wall boundary layer, respectively. The vertical coordinate is z, while p, and , p, are the perimeters of the plume (or }et) and wall boundary layer, respectively, ! i Since molar densities are dependent only on the temperature (assuming constant pressure in the ! entire volume), the differences between the molar densities p(z), p, and p, r.re small. To simplify the analysis, the equation is written without densities. A balance of the volumetric flow rates ! is then used for the remainder of the analysis (instead of a mass balance). To further simplify the analysis (considering only global effects), pu , u,, p,, p., and A are , assumed to be constant, or independent of z (Peterson,1997). This assumption results in a ! linear, vertical velocity distribution. Although the actual entrainment varies with height, the - integrated total should be reasonably close to the average constant values. The calculations of the entrained volumetric flow into the plume Q, and wall boundary layer Q, ; are simplified as: n V
- Q, = g H u,(z)p,(z)dz = u, p, H
- and <
Q, ? , " u,(z)p,(z)dz = u, p, H respectively. The total inflow to the top and bottom of Region II (see Figure 9.C.121) pmvides the boundary conditions for the vertical velocities v(0) and v(H) at the bottom and at the top of the Region II, respectively: A v(0) = Q; - Q, - Q, A v(H). = - (Q, + Q, + Q,)
. where Q, is the volumetric rate of flow into the below-deck region (see Figures 9.C.1-18 and ; 9.C.1-19), Due to mass continuity (conservation) for the below-deck region, this flow rate is Develop:nent of Expected Flow Patterns october 1997 - cr\3006w 9c.nore1b-101497 Revision 1 -n
_ -- n. - . , - .,---.--._+n- - . -- -, , , ~ , , . . , , , , -
, - , - - - - - , , +-
9.C 48 equal to the volumetric flow rate (Q,) entrained into the steam generator compartment by the steam jet (plume). The volumetric flow rate of steam condensed on the vertical wall is Q,. Q, is the flow rate of O steam Londensed on the dome. The total steam volumetric inflow into the containment is Q = Q, + Q, (see Figure 9.C.118 and 9.C.123). The linear, vertical velocity distribution in Region 11 is: Av(z) = (Q, - Q, - Q,) - (Q, + Q,) Downflow exists in the top part of Region II, while in the lower portion, the velocities are positive (upwards flow). This agrees with the previous discussion of Region 11 inflow horizontal boundaries (see Figure 9.C.121). Ikcause the continuity-driven velocities are assumed horizontally uniform upward at the bottom and downward at the top of Region II, there will be an elevation, z, where the two meet and vertical velocity is zero. The z coordinate where the vertical velocity is zero in this model is: z= Q, - Q, - Q, H Qr + Q-The average gas mole fraction in Region 11 is: {= [g" A(z) x,(z) dz = f," x,(z) dz The mole fraction of gas at the bottom of Region 11 is found from a mass balance on the wall boundary layer. (Note that Q, is the volumetric flow of steam that condenses on the vertical wall. It contains no noncondensible gas.) x (0) = f," x,(z) u,(z) p,(z) u, p, dzf," x,(z) dz u, p, H 5 Q, (
= =
Q, - 0, Q, - Q, Q, - 0, Q, - Q, 9 Development of Expected Flow Patterns October 1997 c:\3006wAc.non:1b 10149/ Revision 1
9.C-49 Similarly the gas mole fraction at the top of the Region 11 is: p b QE Q, x Q,+C,_x_ x,(0) Q, + [" x,(z)u,(z)p,(z)dz g, . g Q, + u, p, ,(,n x,(z) dz , Q, + Q, + Q, Q, + Q, + Q, Q, + Q, + Q, The relative difference in the concentrations from the bottom to the top of Region 11 is: Q,0,
+Qr Ax , x,(0) - x,(H) ,
Q, , Q, - Q x, x, Q, - Q, Q, + Q, + Q, The final form of the equation, which is more suitable for qualitative understanding of the influence of various volumetric flow rates, is: Ax x Q, (Q, + Q,) K , ,(0)g- x,(H) (Q - Q,)(0, + Q, + Q,) n The influence of the various volumetric flow rates under various assumed conditions will now lj be examined. 9.C.1.4.2.1 Case 1: Strong Plume, Wall Boundary Layer and Plume Entrainments are Equal If the entrainment volumetric flow rates are approximately equal (Q, - Q ) and are large compared to Q, and Q,, the relative concentration difference is simplified to: Ax x,(0) - x,(II) Q,(2Q,) 2 Q, _. ~ _. . x, x, (Q,)(Q,) Q, These assumptions are valid for the case of the jet-dominated flow. The large plume and wall boundary layer entrainment volumetric flow rates act to reduce the relative, vertical steam concentration gradient. Even if the Cow pattern cannot be defined as jet-dominated (i.e., the equation for the relative difference in the concentration from the bottom to the top of Region !! cannot be simplified), the recirculating stratified Region 11 interacts with the plume and wall jets, Regions 111, and IV (see Figure 9.C.1-23). The relative concentration ditierence will still decrease if the entrainments in both the wall layer Q, and plume Q, are large. Development of Expected Flow Pattems octobe 1997 oA3006w-9c.non:1b 101497 Revision 1
9.C 50 It has bec.i .=hown (Enclosure to Westinghouse Letter NSD NRC-97-4978, February 7,1997) that durir.g the quasi-stesdy portion of a LOCA, jet entrainment rates (Q,)in the AP600 are about i a factor of 10 greater than the condensation rate (Q,+Q,). l 9.C.1.4.2.2 Case 2: Equal Entrainment into the Wall Boundary Layer Q. and the Rate of the S.eam Cor.densed at the Vertical Walls Q, A smad dH4scrite between the crurainment volumetrb flow rate into the wall boundary layer Q, and vclumetric fle a rate of the steam nradensed at the vertical walls Q, produces an increase in the relative differov:ni de cu.wntrations, if al! the steam entrained into the wall boundary layer is condensed at the ve:tice.wAh rdhing is left to be redistributed throuc,h the lower horizontal boundary of Region 11 and conhibute towards a decrease in the vertical < uncentration gradients. 9.C.1.4.2.3 Cne 3: liigh Dome Condensation Rate Q, The volumetric !!ow of the steam condensed on the dome of the containment Q, does not directly affect the relttive concentration difference in Region 11 (it is not present l- the equation). However, Indirect effects are ponible. If the condensation on the dome is high, the ratio of Q,/Q, is high, and the volumett;. flow of steam condensing on the vertical walls Q, decreases. In contrast, a small ratio of Q,/Q, represents cn increased volumetric flow rate condensing on the vertical walls, Q,. A decrease in the rate of steam condensing on the vertical walla Q, (in fact the increase of steam volumetric flow rate condensing on the dome, QJ, decreases the relative concetitration difference. 9.C.1.4.2.4 Case 4: !nfluence of the Below Deck Entrainment Q, Region I also Mteracts with the stratified Region II. 'Ihe effects of utis interaction on the relative concentration difference change are captured by the Q, team. A large below-deck entrainment, Q, reduces the concentration diffen nce. In the AP600, below-deck entrainment contributes to a decrease in the relative concentration difference. 'Ihis effect is not present in the IsT case, where Q, = 0. 9.C.1.4.2.5 Case 5: Domicant Entrainment into the Wall Boundary Layer Q. If Q, = 0.5 Q,, re observed in phase 3 of the LST experiments where Q, is between 0.4Q, and 0.6Q, (see WCAP-14135), and if Q, = 2Q,, i.e., the wall boundary layer entrainment is !vice as stronb as plume entrainment (weak plume scenario), if the entrainment in the below-deck region is negligible, Q,=0, and if i e assume Q,= 10Q, the relative steam concentration is: Ax x,(0) - x,(H) , 0.5Q, (20, + Q,) , { ,~ g (2Q, - 0.5Q )(Q, + 0.5Q, + 0) O Development of Expected Flow Patterns October 1997 oA300bw-9c.non:1b 101497 Rmision 1
9.C 51 L
- A further increase in the entrainment into the wall boundary layers causes an additional decrease j , -in the relative difference between steam concentrations in the bottom and the top of Region 11
- j. . (e.g.,if Q =3Qg the relative concentration is 0.064). The increase in the entrainment into the wall i boundary layers contributes to the homogenization of the containment atmosphere.
1 L i 9 C.1.4.2.6 Conclusion l l The expected circulation within the AP600 containment is segregated into five regions that selate ! to separate effects tests (SETS)in enclosures. Given the presence of the externally cooled shell, l which is assumed in a DBA analysis, there are no regions of stagnant stratification in AP600 i containment. i { 1he proposed conceptual model can be used to structure the containment into regions for j comparison to relevant enclosures SETS. The mathematical representation provides insight into
- the influence of various volumetric flows on the axial steam concentration gradients in AP600.
I l ! } i 'i L i i ! i ] t i i i
+
Development of Expected Flow Pattems October 1997 - oA3006w-9c.non:1b 101497 Revhion 1
9.C 52 Dome O V N A ' l 1 i l Qv _e.____,
+Q r----- , 3 \ g Qv+Qt+Q$e i i /
I l' I /
\t %a q
l l l l I I i Ce !Ill ! I
- Q ,. ; j j Qv I g
g l g H=H, i I ! i-I l l lV IV -I De i e i 11 t i II . t l
\ l l e 1 I I e ~- '! Q.+Q i' QirQcQ e a
_______e.-_- Her Qc , p , Qe Qd QMt VOLUMETRIC FLOW RATES REGIONS Q, steam volumetnc dow rate from the brake I below dock region Q, now rate ennained by the jet inside the below deck repon 11 repon of recirculating stratificanon Q, now rate entrained into the plume in the above dock repa 111 repon of jets or plumes Q, now rate entrained into the vert cal wall boundary layers IV repon of wall boundary Isytts Q steam Cow rate condensed on the verocal walls V repon below dome ceiling ~ Q steam Gew_tsttsadened QiL1llLAQEf3fM' Figure 9.C.118 Primary flow regions and volumetric flow rates for quasi steady containment conditions in AP600 case O Development oOixpected Flow Patterns october 1997 i, o:\3006-cl.non:llwl01497 Rmsion 1
9.C.53 O k s b -- e i I I ' I IV IVl De t l II I I II , 1 i i l l I i
' 1 I i ~
Qw-Qv-Qe 'l Qe+Q i'
. _ __ ______ ___ L__.___ _..
f( jet / plume ( -
/ )._ entraimnent 9' j O* \ s 9a/
l 7 , downflow s p , Qe / %
\ I dow\nflow around around deck floor deck flor,r h' '""
QoPQV+Qt i l Figure 9.C.1-19 The volumetric flow between the upper and lower deck regions Q, o Development of Expected Flow Patterns october 1997 l ct\3006-cl.nortib.101497 xeSh 1
9,C 54 Qi Dome y s
^ I r
ov+os- - - - - . l- - - - - - , s s- , s g m ,i i / g i i i 1 t l l l l t S ! l / l I g Ce i III I i l i o. ! l l Qv H = He l
'z \
i i l 1 I i I l l IV De i IV -( 11 i i II i i t i I i I )
\e l Q.
l Qw*kV
~~~~~~~~~ ._-_-~~~Ee ...............====si Hrs grating I u Q,=0 Fe kcPkV+kt VOLUMETRIC FLOW RATES REGIONS Q, stearn volwnetnc flow rate thun the brake I below dock ret ico Q, " 0 Dow rate entrained by the jet inside the below deck repon !! fusion of rectinlanag strati 6 cation Q,Dow rate entrained into the plume in the above deck repon ggy ,,,g,, g g ,,
Q. Cow rate entratned into the vertical wall bo mdary layers IV <ssion of wa0 boundary layers Q, steam Cow rate condensed on the vertical walls y g g, Q, stearn Gow rate condensed on the dome cetting _ Figure 9.C.120 Primary flow regions and volumetric flow rates for quasi-steady containment conditions in LST case Development of Expected Flow Patterns october 1997 . oA3006-c2.ron:1b 101497 Revision 1 l
9.C 55 I O Dome a y s A /
'. 1 Qv+0 l J +Qe -- _.y I
g--- 6 Be Qv+Qt+Qp+Qe I
\
d .J JJ. " J . t..t J _ U L L L L t 11/ g i I /
\ l i i g downward q, I l f i velocity i I t i Ce i III I e ! Q, I 8 I Qv H=H, g 11 ! I f I ' I upwar'd , f ; gy IV j velocity De i e i 11 f1- f TT T f 11 lITITfi' - J - _\ ' \ n.+ 9 l Qw-QcQe i e
O V Ee t ( Hof Qe I v [ p , Qe /
"" Qa=Q&Qt VOI.UMETRIC Ft.OW RATT.5 RE0lONS Q, steam volumetric Dow rate from the brake I below dock region Q. Dow rate entrained by the jet inside abe below deck region g3 r,gion of recaculating stratincanoo Q, Cow rate entrained into the plume in the above deck region g)) ,,pce of jets or plumes Q, Sow rate entrained into the vertical wall boundary layers IV region of wall boundary layers Q, steam now rate condensed on the vertical walls V repon below dome ceiling Q, steam now rate condensed on the darne ceiling Figure 9.C.121 Inflows and outflows from the Region II O
Development of Expected Flow Patterns october lo97 cx\3006 c2.non:1b.101497 Revision 1
9.C 56 Dome A e V *AN! / I I ________.______, ,l_______
\ W ch Y l \ l I I \ l l I \ l l I I
t t Ce !III I I
'[
( S n t// I I l I l I a u, IV j De ! f lN l \' ig@ #! l** l9"(j/@ ,i
,I ._ , l q,q -
______g.;___ . _________ ; .. g Hu Qe V [ Fe e , 90"kv+Qt VOLUMETRIC TLOW RATES REO!ONS . Q. steam volumetne Sow rue tem abe beske I bdow deck region Q, Sow race entraaned by tbs jet inside abe below deck region II region of rocaulan's senti 5catiwi Q, Dow rate enersaasd isso the plume la abe above deck regian
!!!ission c(lees w planes
- 0. Bow rate emosined into abe vertical was boundary layers !Y sosias of wan boundary layers
- Q, steam Sow rose candensed on the verocal wads V region below dome coling Q, steam now rate ew==d on she dome ceGing Figure 9.C.122 Recirculating flow paths inside the Region II Development of Expected Flow Patterns
- 1"7 o;\XXW<2.non:!b.101497 ,
9.C 57 1 Dome y s A / Q,*0
., J_+Qe__ I i i n y______ 'g * '
Qy+Qt+Qp+Qe j
\ l i / v \ I I I g i i / q=H-z \ % I l I i c. !III ! l Q. !L u
[MT*'i /I l l n u, IV j De ! '/ i (t)A(t)WI) l IV g II i i II , I p u,(t)p)6k \ l l P (d5(Opf(ddE I I ,,, I Z , I - h Qw-QcQe !QpQ}s g Hrs Qt U y Fe /
, Qo&v+Qt VOLUMETRIC FLOW RATES REGIONS Q, steam volumetric Dow rate from the brake I below deck stpon Q, Gow rate entramed by the jet inside the below deck repon 11 region of recirculating strati 0 cation Q,Dow raie entrained into the plume in the above deck repon 111 regimi of jets or plumes Q. Dow rate entrained into the vertical well boundary layers IV region of wall boundary layers Q, steam Cow rate condcased on the vertical walh V repon tielva dome ceiling Q, steam Dow rate condensed on the dome ceihng Figure 9.C.123 Mass conservation for thin horizontal layer inside the Region II Development of Expected Flow Patterns October Iw7 o;\30%c2.rmn:Ib101497 Roton I
9.C-58 9,C.I.4.3 References
- 1. P.F. Peterson,1997 LST Mixing Model Writeup - letter tc, J. Woodcock,02/24/97
- 2. Enclosure to Westinghouse Letter NSD NRC-97-4978
Subject:
Position paper in support of the assumption of complete mixing of aerosols in the AP600 containment atmosphere following a loss of coolant accident Febtuary 7,1997
- 3. WCAP-14135, F. E. Peters, April 1997
" Final Data Report for PCS Large-Scale Tests, Phase 2 and Phase 3", Revision 1 Westinghouse Energy Systems O
O Development of Expected Flow Patterns octoter 1997 c:\3006w-9c.nortib-101497 Revision 1
9.C-59 9.C.2 OVERVIEW OF THE INTERNATIONAL CONTAINMENT EXPERIMENTAL DATA BASE i Tests from the available international containment experimen'.al database that are relevant to the . AP600, are presented in this chapter. Some tests are very close to possible AP600 cases. Others i are presented to emphasize the difference between the AP600 and the test conditions that lead towards stratification. Four experimental facilities are considered to supplement LST data. Table 9.C-1 specifies characteristics of each experimental facility and provides a comparison with the LST and the AP600. A comparison of the sizes of various test facilities is provided in Figure 9.C.2-1. Scaled cross-sections of each facility are shown. l A list of the facilities and the_overviewed experiments !s provided in Table 9.C.2-2, as well as the mah characteristics of each experiment. 5
%./
l l l r. - \j i Development of Expected Flow Panems october 1997 o:\300bw-9c.non:1b-101497 Revnion I s I
] 9.C 60 \ f O s 1 4 l stOn i, ,' g , ,I -,,,s, iii
8 t
i O ,_. D Q i i O -
--T-uq ---s p [h = p'.
[.LI.JI g l LST BMC NUPEC CVTR HDR AP600 l. a 4 4 4 l i J j Figure 9.C.2-1 Comparison of different facilities
; Development of Expected Flow Patterns October 1997 l .
o:\3006w-9c.non:1t>101497 Revision 1 1 s
, ~
9.C41 l l
- i Table 9.C 1 Comparison of Wrious Factitles ;
p muumummmmmmmmmm mw- - - - i V Facility LST BMC NUPEC CVTR HDR AP600 ! Volume m' 8? 1 MO 1300 6428 11300 48710 (used) Height m 6.1 9 17.4 34.7 60 57.9 Diameter m 4.57 11.25 10.8 17.7 20 39.6 Number of 3 9 25 3 62 72 11 compartments Volume of the 79 % 70 % 41% 44 % 81 % dome Contalrunent steel concrete steel concrete steel shell steel shell walls concrete and concrete and steel steel O n V Development of Expected Flow Patterns october 1997 c:\3006w 9c.non:1b-101497 Revision 1
1 9.C-62 I l Table 9.C 2 Overviewed Tests from International Database 1
- ummmmmmuumummuur;m - - I Position of the )
Release Point and Stratification or Facility Experiment Main Feature Other Relevant Data Circulation l BMC F2 set First Phase stepwise steam release point is high circulation through addition Hr/Ht=0.444 the majority of compartments, extemal annulus stratified RMC F2 set Phase 2 inducing the steam release point is circulation through natural circulation low the majority of with steam Hr/Ht = 0.111 compartments, injection external annulus stratified BMC F2 set Phase 3 heater on in R6 to heat source location circulation through reverse circulation is low the majority of Hr/Ht = 0.111 compartments, extemal annulus stratified BMC F2 set Phase 4 steam injection in steam release point is circulation through R6 compartment low the majority of Hr/Ht = 0.111 compartmer ts, extemal anr ulus stratified BMC Test 2 hydrogen low positior of hydrogen uniformly Two injection, hydrogen source distributed, compartments uniform initial Hr/Ht = 0.06 circulttion present (Phase 1) temperature, orifice present, low feed rate BMC Test 4 hydrogen high position of concentration Two injection, hydrogen source stratification occurs compartments uniform initial Hr/Ht = 0.57 (Phase I) temperature, no orifice BMC Test 6 hydrogen low position of stratification present, Two injection, stratified hydrogen source highest hydrogen compartments initial Hr/Ht = 0.06 concentration in the (Phase I) temperature, lower compartments orifice present O Development of Expected Fbw Patterns october 1997 o:\3006w-9c.non:ltr101497 Revision 1
9.C-63 Table 9.C 2 Overviewed Tests from International Database (] V (cont.) mummmmmmmmmmmmmm mmmmmmmmmmmmmmmmmmmm - Positicn of the Release Point and Stratification or Facility Experiment Main Feature Other Relevant Data Circulation BMC Test 12 hydrogen injection high release point hydrogen uniformly Six compartments in R2 room (high), H,/H, = 0.69 distributed, (Phase II) uniform initial circulation present temperatures BMC Test 20 hydrogen injection low position of stratification present, Six compartments in R6 room (low), hydrogen source highest hydrogen . (Phase II) stratified initial H,/H, = 0.06 concentration in the temperature lower compartments BMC RX4 sump heat up low position of the caculation present, and three heat and hydrogen homogenizatic- of hydrogen source temperature w a injections H,/H, = 0.0 concentrations NUPEC M43 simulated break low position of the circulation present inside the low heat and hydrogen during release, steam generator source temperature stratifies O compartment, H,/H, = 0.0 and concentration V steam and homogenizes after hydrogen release, the end of release contamment shcIl , insulated CVTR First test without steam release in high position of the temperature field the intemal water the upper steam release stratifies sprays compartment, H,/H, = 0.525 concrete shell CVTR The second and steam release in high position of the temperature field third test with the upper steam release stratifies internal water compartment, H,/H, = 0.525 but not as strong as sprays concrete shell in the previous case HDR E11.2 high positioned high position of the stratification exists, release point steam release extemal sprays (small break) H,/H, = 0.555 promoted circulation and active extemal spray O Development of Expected Flow Patterns october 1997 c:\3006w-9c.nortib-101497 Revision 1
1 9.C-64 Table 9.C 2 Overviewed Tests from International Database (cont.) ummmmmmmu-muummmmmmmmmmmmmu-- T Position of the Release Point and Stratification or Facility Experiment Main Feature Other Relevant Data Circulation HDR E11.3 low positioned global circulation small break pattern formed closed spital stairway entrance HDR 'c11.4 low positioned low position cf tb global circulation release point steam release formed (small break) H,/H, = 0.18 almost uniform and active temperature external spray distribution except below release point HDR T31.5 simulates DBA high position of the temperatures and gas large LOCA in the steam release concentrations first upper section of H,/H, = 0.526 stratify the containment and latter homogenize HDR V21.1 simulates DBA middle position of Equal heating of both large LOCA in the the steam release staircases first middle section of H,/H, = 0.38 suppressed the containment circulation. Slight (in both staircases) global circulation was generated later. HDR E1L5 simulates DBA low position of the global circulation large LOCA steam release due to the steam in the lowest H,/H, = 0.18 release, gas mixture section of the injection and sump cot,tainment boiling contributed with effects of dry towards heat release and homogenization sump boiling O Development of Expected Flow Patterns october 1997 - oA3006w-9c.nortit>101497 Revision 1
9.C-65 9 C.
2.1 DESCRIPTION
OF THE AVAILABLE BATTELLE MODEL CONTAINMENT (BMC) DATABASE The objective of the Battelle Model Containment (BMC) tests is to obtain data to analyze design basis accidents (DBAs), hydrogen distribution, and aerosol depletion. The total volume of the containment is 640 m' and represents 1/64 of the BIBLIS B containment. Its interior is divided into nine compartments and its walls are made of reinforced co crete. The sizes and locations of openings between the compartments can be adjusted by opening (or closing) the openings with steel plates or mobile concrete structures. Three sets of tests are presented. The first set, the F2 experiments, tests natural convection as a function of release location and type of release (steam, air, dry heat).
'Ihe second set of tests studies the influence of the initial temperature distribution, the location of hydrogen injection, the injection rates, and the size of the vent openings on hydrogen distribution, stratification and global circulation.
The third set of tests examines the effect of sump heatup on global natural circulation. ! (O . U F Q U Development of Expected Flow Pattems october 1997 c:\3006w 9c.non:1b-101497 Revision 1
9.C-66 9.C.2.1.1 Natural Convection Phenomena Inside the Multi Compartment Containment (F2 Experiments) The F2 experiments, performed by Kanzleiter in 1988, study natural convection inside a O multi compartment containment as a function of release location (room) and type (steam, air, dry heat). The DMC configuration used for experiment F2 is shown in Figures 9.C.2-2 and 9.C.2-3. A 48-hour heatup period is the first phase of the experiment - see Figure 9.C.24, (Fischer et al., 1989, and Fischer et al.,1991). Th!s is followed by a three-part, natural circulation phase (phases 2,3, and 4) within the 48- to 75-hour time period see Figure 9.C.2-5, (Fischer et al.,1990 and Fixher et al.,1993). An overview of the results and a comparison with analysis codes is presented by Wolf et al, 1996. Data for pressure, temperature, sump temperature, and liquid level, as well as partial steam pressure is presented for phases 1-4 (up to 75 hours), 9.C.2.1.1.1 F2 - Experiment Heatup Phase - Phase 1 Phase 1 is from 0-48 hours. A steam release inside the R2 compartment provides the heatup (see Figures 9.C.2-6 and 9.C.2-4). The stepwise steam addition results in a stepwise increase of the containment pressure (see Figure 9.C.2-7 for GP 9117 location). During the 48 hours of heatup, the atmosphere in the external annulus (the lower portion of R9 surrounding compartment) stratifies, Figure 9.C.2-8. Since there is no driving force for the circulation of steam into the lower air-rich regions of R9, the two experimental curves in Figure 9.C.2-8 (for temperatures GT9004 and GT9037) show that the heatep was delayed in lower positions in the external annulus behind the missile shield. The lower portion of R9 heats up over a longer period (Figure 9.C.2 8) because of global circulation induced by entrainment in the release. The entrainment is fed by flow from R9, R4 and R1, Over a period of time, the atmosphere of the containment in the external annulus stratifies (after 16 hours it is already stratified). However, after 36 hours the stratifiertion is not as pronounced, i.e., the temperature differences are not greater than 10 C in the external annulus. The initial stratification in the external annulus results from the high position of the steam release, which is inside the R2 compartment, and the closed circulation paths in the lower portion of the external annulus (see Figure 9.C.2 9). The 2xperimental curves for temperature histories of the other subcompartments (except for R4 and R3) are not presented in Wolf et al.,1996. However, consecutive phases of other experiments performed in the BMC indicate that natural circulation effects are present and contribute towards homogenization of the temperature fields among the majority of compartments. The only exception is the external annulus. O Development of Expected Flow Pattems october 1997 o \3006w.9c.rmn:1b-101497 Revision 1
9.C Application to the AP600 Case l f There is evidence that a release high in the steam generator compartment can induce global' circulation flow by entrainment through the CMT room openings. It is difficult to compare time
- scales due to significant differences between the BMC and the AP600 compartment arrangement.
i j 9.C.2.1.1.2 Phases 2 4 of the F2 Experiment (Natural Circulation) After the heatup, the experiment continues through three additional phases (see Figures 9.C.2 4 - 1 and 9.C.2-5 for phases 2, 3 and 4) that use the following methods to induce or amplify , circulation: [ Steam injection to induce natural circulation, Activation of the heater to reverse circulation, ' Injection of steam to amplify reversed circulation. Figure 9.C.2-5 illustrates four add'itional phases (5,6,7, and 8) that are not discussed. The [ circled numbers in Figure 9.C.2-5 represent the type of injection (see also Figure 9.C.2-4). The l flow pattems formed during the particular injection are presented below the circled numbers. Figure 9.C.2-9 shows the two different locations for the steam injection, the location of the heater I and the positions of the anemometers. In addition to the measured velocities in the openings (see Table 9 C-3),-the fluctuations in ] . measured temperatures indicate natural circulation (see Figures 9.C.2-10 and 9.C.2-11). Due to natural circulation, complex flow patterns form and temperatures in the compartments are nearly _ ! homogeneous (i.e., temperature differences are not greater than 4*C, see Figures 9.C.2-10 and j _ 9.C.2-11). The detailed temperatures and velocities during each subphase are given by Kanzleiter, , 1988, t Figure 9.C.2-12 presents the thermodynamic states of the containment dome atmosphere at a high position (H = 7.6 m) during various time. periods, while Figure 9.C.213 presents the conditions at a low position in the external annulus (H = 1.0 m). Except during air injection times, the steam partial pressure in the high position follows the shape of the total pressure curve (0.5 to 1.0 bar lower values than p, ). At the low pos.ition, the steam partial pressure is . almost constant (0.5 bar) after 60 hours. This indicates steam stratification inside the extemal annulus (lower portion of the R9 compartment) behind the missile shield. 1 The temperature distribution inside the external annulus is presented in Figure 9 C.2-14 for the second phase and in Figure 9.C.2-15 for the third and fourth phases. Both figures indicate stratification of the temperature fields. The temperature difference between the upper dome and the lowest position in the external annulus is 30*C at the end of the second phase and 18*C at h the end of the fourth phase. V Development of Expected Flow Patterns octobn 1997 o:\3006w 9c.noruib-101497 Revision 1 l
9.C-68 All other compartments have almost homogeneous temperatures (the greatest temperature differences are 4*C), which indicate the presence of the natural circulation (see Figures 9.C.2-10 and 9.C.211). The values of the measured velocities in the vent between the R3 and R6 compartments during the individual phases are presented in Table 9.C-3. The histories of the velocities in the R7-R9 and R3-R6 vent paths are presented in Figures 9.C.216 and 9.C.2-17, respectively. The directions of the convective flow loops as a function of steam and air injections into the various compartments atid the applications of the dry heater are presented in Figure 9.C.2-4 (arrow in R9 compartment represents positive flow loop direction). The upward (positive) velocities in l Figure 9.C.2-16 produce a positive f.ow direction loop. The upward (positive) velocities in l Figure 9.C.217 produce a negative flow direction loop. l The various injections and the application of the dry heat source generate natural circulation and homogenize temperatures in the majority of the containment compartments. O O Development of Expected Flow Pattems october 199f o:\300ew-9c.non:1b-101497 Revision 1
O O O se YN g is Table 9.C-3 Representative Velocity in [m/ s] in Opening from R3 to R6 in the Dtfferent Expenmental Phases of Expubwat F2,
#g Phases 2-4 7-o Phase No. !@ 2.1 2.2 23 2.4 3.1 3.2 33 4.1 4.2 43 4.4 4
035 -0.4 0.4 -0.4 +/-0.1 +/-0.1 +/.0.1 -035 0./0.6 035 -0.4 , m 5- (reprinted from L Wolf, M. Cavrilas, K. Mun, " Overview of experimental results for long-term, large-scale natural circulations in LWR- [ containments after large LOCAS," University of Maryland at College Park, Final Report for DOE - Project, Order Number: DE-AP07-g 96ID10765-)
;r i
f i x-9 k N E n 8$
-s 6
e i
l 9.C-70 1 1 1 i S H
)
y ? ., at as 6 l l Rl-R3 , Yl
,, y- n, s .iar y,- ,
j ne .- 1 i ,,
. .;_j ~ . - -
j *- "'"" 8
- d ;. . .: ; ,. . . . g ,. g ,,a y .. , ,, v.. . / j[;.:.<. n " . 4.as . >
N > G . E (H e + 3.2 m; H = + 0.1 m) , , , , r aar s A* 4;, ni ,a g.,pq - j l
' ' ne l,5.* .* h 5:'. 5,'.. ,
t 8 . ...
~ - ., :. . ..::.'..... .c.;. .y ,
- ., .: {: :: :.;;.(.;:l .;.,'.Hiil.?..;l
\
- uenw G- '- : g -.
- N
.. . , (H = + t.o m) ama w m me m. . .= sessen esey swet m r2) *'v .o raer Figure 9.C.2-2 Model Containment: vertical cross-section 0/180*, horizontal cross-sections at +0.2 m and +1.7 m for configuration in experiment F2 l (reprinted from L. Wolf, M.Gavrilas, K. Mun, " Overview of experimental results for long-term, large-scale natural circulations in LWR containments after large LOCAs", University of Maryland at College Park, Final Report for DOE - Project, Order Number: DE-AP07 961D10765")
l Development of Expected Flow Patterns october 1997 o;\3006-c3.non:1b-101497 Revision 1
. - . _ - - . .. - - . . _ _ . . . . . . . . _ . .- . . . . _ , . _ . _ ~ - .. _
9.C 71
, f*
e M
\ ,g*, \ ,( ,
N ,. I R1 R3 4 . 1 , g' E M I ' 4
~~ ~ &
O
~:, = \hf$f )
m
** a f Ti .. . .:eitMi_
e g;yy a ' su{r, . .,.* 9 7 g,.*.? '
, * * *. .) \ .. a r . ;,
d m
*, Faer ^M -
gy ". . gg
, 1.en wecean * *# '
i poseon pn F2) , M6
,,, , ., k 'N i +8 ,'. ' me 1, i.I m' ,
[
,,,,. . a.e, b
42 :. j, a-- / '
, < 6 l =* '
as
.i a ,, ',.' .' ' [.' L '. / {t D:.':<.3, .l .1 * * * * .'.-. P.. '.:.' ,.;:: ' : ' .* : *..* : l s.s ne *... - . 'f -,- * .
nsa, g 4' ' _
. 4 er at 3,,, c . ,r..
f Inuo sleem in$ecten pessono arrow tae 2. * %
'"'t e. (H = + 3 2 m; H = + 0.1 m) paaWon @ F2)
J. * . . .
- s. , ,
i Figure 9.C.2-3 Model Containment: vertical cross-section 0/180', horizontal cross-sections at +4.01 m and +5.05 m for conSguration in experiment F2 (reprinted from LWolf, M.Gavrilas, K. Mun," Overview of experimental results for long-term, large-scale natural circulations in LWR-containments after large LOCAS", University of Maryland at College Park, Final Report fur DOE - Project, Order Numben DE-AP07-961D10765") Development of Expected Flow Pattems October 1997 i o:\*!XWc3.non;1b-101497 Revision 1 i
9.C-72 4
- e. : : : . . . .
. . : .a u n O ,{*.i~.i.'i l' - ,. . : .4. *:. ". ,' ..s' . ' .
- h.
..... gg . .' 7 %*'.'.
R2
.'34'3 9 .'- . .'! R7 . h. R5 E .. i.
fs m i.C n, .'.p n.'. m: :. .,.
- h. R8 _ g,.. .. .. R6 ..
.'.\ : ... .':O. . . . ,..- ? R3 .- * ,- . .~ . ,,. : .* ,, . . . ... ' ..' . , . . ..s. ' . ~
Phase Pressure Steam Air Heat Drainage Flow i 1 2.1 2.2 1-3 bar 3 bar 3 bar R2 R3 R7 R6 R9 R3 R3 + Injection 1+M h R3 2 l 2.3 3 bar R6 .
- 3 2.4 3 bar R3 + 1 3.1 3t,ar 3.2 to R6 . 2 3.3 1.8 bar -
2 4.1 1.8 bar R3 + ' 4.2 1.8 bar R3 R6 +- 4.3 1.8 bar R6 - 3 4.4 1.8 bar R3 + 4 Figure 9.C.2-4 Injections ar.d convective flow loop directions (+ sign for flow indicates the same direction of the flow as arrow in R9) (reprinted from L. Wolf, htGavrilas, K. Mun, " Overview of experimental results for long-term, large-scale natural circulations in LWR-containments after large LOCAS", University of Maryland at College Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765") Development of Expected Flow Patterns october 1997 o:\3006-c3.su>n:1b-101497 Revision 1
9.C 73 O F2 @ @ @ TW * %ecke
/( S"N / 'h F" 3 \- / /9 '
F' 3\ 79 ['P '
, heeler ' ~
gl) ,,, _st5,- )f
- h ,,,,,
these 2,1 1.32.32.4 3 4.1.4 .3 .4 I 0*I *I *I *0, I Sol I*I e3, *4 *I
@ @ @ O @ @ @@ @ @ @ @ @ @ @ @ e-Type of irvecnon ,H 4 l-- H e-Air aveceos
- -sie.n ort eu n ers -
q ,, 6" j
- Heir Weetion 4
) ewea % 4 si j f; * \
[ portist eit preeswe teveresel , j o n, 4
} < .... ..... ~
0
- -- [
d k [
,o ,
h ', , ,'o i n h end of Pham 4 s Figure 9.C.2-5 Experiment F2, Phases 2-8, experimental procedures, total and partial pressures (reprinted from LWolf, M.Gavrilas, K. Mun, " Overview of experimental results for long-term, large-scale natural circulations in LWR containments after large LOCAS', University of Maryland at College Park, Final Report for DOE - Project, Order Number: DE-AP07-96ID10765") O Development of Expected Flow Patterns october 1997 o \3006 c3.non:Ib-101497 Revision 1
9.C 74 e
~
steen release location f
+$1smug % / " * } .y - TfT' 1r6gs.i,h~9y':': / - SR 9240 $sg':.s,!,tgi.p.tf T HZ 9258 -
GO 9258
- f:..
, g,eoj. :& .q.g*p GT 9250 . # **~
Go 925e) E ___n j E 92M kg
^- -
CT 9054 i ,, iM m j$ I R7 . R5 GP 7 l4i8l__ _
* !)ii .E.d M 121 GD 7240 - _ gl 4 ,j :I, CjMy I
4:12.3 IL g R1_ _g,gj
= . ----
w - ,'s, g 1 , GT 9037 GD 9241 ,p', '. j p m- .
. .% ,-r. -
Mr* g '4 . GT 3115 W -sh ', ;; h -- .
~ ~ ~ ~
110 GT 3100 N y U u evn - a
%:~d,N WT9060 GD 8240 14 b . . _f [ 2. ' Ri:
b We>Ms pgp..#.. -
* = ===
h[ h!hhhh .
' h; . ! / i GD 3240 DM 122 steam release location Figure 9.C.2-6 Number of compartments, locations of measurement transducers and steam releases (reprinted from L. Wolf, M.Gavrilas, K. Mun, " Overview of experimental results for long-term, large-scale natural circulations in LWR-containments afterlarge LOCAS", University of Maryland at College Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765")
Development of Expected Flow Patterns October 1997 c:\3006-c3.non:lt>101497 Revision )
l 9.C-75 1 l m O
--+1t-J . i
- i- p !!E<--
.+ ...
- =
*M' p. .c c. ~
CP9117
- Test Group F2, Phase 1 (Hectup) n -
,- l J ' /
0u mg. - ) h l
- I y F e
j m. Y" n.
- s. 88 2s.
se. 4 a. se. TIME [ HOURS] Figure 9.C.2-7 Test group F2, Phase 1 (heatup), total pressure in the containment (reprinted from L. Wolf, M.Gavrilas, K. Mun, " Overview of experimental results for long-term, large-scale natural circulations in LWR-containments after large LOCAS", University of Maryland at College Park, Final Report for DOE - Project, O. der Number: DE-AP07-96ID10765") Development of Expected Flow Pattems Ottar 1997 o \3006<3.non:1b-101497 Revisu 1
9.C 76 O M unst6_ g ;= a - A "
,_ [" {= +
l g -. - < - - 2 g= ... . . Q^ - ' ly <-- 1 i~= _. "E q= '! '~
^
ctwee4 1 -- CT9827 2 Test Group F2, Phase 1 (Heatup)
- 2 a , l u 2 f ~
8 l t u ax f = h g 1 ,I E I
<g -
s , g
- e. w. 2s. se. 4a. se.
TIME [ HOURS) Figure 9.C.2-8 Test group F2, Phase 1 (heatup), atmospheric temperature in the compadment R9, H=1.0 m and 2.1 m. (reprinted from LWolf, M.Gavrilas, K. Mun, " Overview of experimental results for long-term, large-scale natural circulations in LWR-containments after large LOCAS', University of Maryland at College Park, Final Report for DOE - Project, Order Number: DE-AP07-96ID10765") Development of Expected Flow Patterns october 1997 o:\3(XWc3.non:1b-101497 Revision 1
,9.C-77 i
l i i , I i ZONE 4 l (RS, R2, R4) 1
*/"/ closed \# "~
i- - -- 7 flow I path I I -
\ ,
zowE 3 M*M ZONE (R6, R5) 1 e = steam injectlon (Rs,R7)lL ~ 2 locations j , SONE 2* M = dry heat supply
- (R1, R3) y g e - ,eM w = relacity measure-
,I '
3 i A 1 I l 3 ment l' - N \ j closed flow path closed flow path s circulation for phases 3 and 4 are shown i 4 f i N i i j Figure 9.C.2-9 Scheme of multi-compartment containment geometry in experiment F2 4-I ] (reprinted from L. Wolf, M.Gavrilas, K. Mun, " Overview of experimental re::ults for long-term, large scale natural circulations in LWR containments after large LOCAS", University of Maryland
;. at College Park, Final Report for DOE - Project, Order Number: DE-AP07-96ID10765")
5 ~, p { Development of Expected Flow Patterns october 1E j o:\3006 c3.non:1b-101497 hevision 1 4
9.C-78 O lV 2 Q)016 K [W 16 4V 2 Ci?G?4 K l.. 24 iV 2 coiis ,rw iis iv i m as r tr 25 F02 8 sv 2 moon w tw n sV t as0n v tv v I l' tufteinspelsund Wrmetauscher. Dasofeln stopp der betrieb speisun9 1, nach R7 e & ' 'Dampfrufuhr l [' .;c kjetu I has t * **l**J" 'M T** nach R6 49 ,f 8 .___i__ J_ l -_J___' +4 +< Aen Q. l I mt'? _ l". .. .. i t ~ s I __M g g i i l o i i i i l I If. . , _ _gf_..J_ U~ L .- _ .. L . . .. .-!- _ .. j _ _ _ . l _ .--- .l- - - l l l .ahase 2 . I i I I l l l 1 I i e iE g. lrfilv111
' 'c " - - . - l- . - - -l - - - - { - - -l - - - -l - - -
I J/-f;,i . . '
-(- . .
g l i l g 1 i I l.2 R3 S -
-c l. _..l.. ..l....--l- 3 na ..
4 R7
. w I 5
r- .. c i ., . i i i i , R
, 6, ** I I I I ==jf b
g x
,. - - - - - l- - -- f - d - - - -l - - - -l - - -g l
S L._ u.s i- I I I I I g. 3 g g l l 3 :!:%. . ; .,, ', i:,... i 8 ._gs . .. . .. s . ..l- J _-.I-__.l___
. . . .T W' l l l l l l . , , , , I ,
s - - l l ! ! ! l N5.00 45.50 45.00 45.50 51.00 55.50 54.0G 55.50 Si.00 5$ 50 65.00 61.50 TINIEIN HOURS Figure 9.C.2-10 Test group F2, phase 2: atmospheric temperatures in zones 1,2 and 3 (R5 + RS, R1 + R4, R7 + RS) (reprinted from L. Wolf, M.Gavrilas, K. Mun, " Overview of experimental results for long-term, large-scale natural circulations in LWR-containments after large LOCAS", University of Maryland at College Park, Final Report for DOE - Project, Order Number: DE-AP07-96ID10765") Development of Expected Flow Patterns october 1997 o;\3006.c3.non:1b.101497 Revision 1
I 9,C 79 L
- q ,
2-4 ~ 9 j ..
- 1 i:
p 5 ) j 1v 2 N3H8 K IK 18 4V 2 M1024 W [F 24 . r~ 2v 2 m3ns y er ni sv 2 asas r (* n F02
'3 V - 2 m8Bf2 E EW - 22 SV 2 W5027 K EN 27
- Phase 3 8 I I I Phase 4
[ l eN M,' g l f t,8 bar) l
' e I O ,) Warsetauscher- j lWarastausg ,1 i
- f M NI" of l'* -
}L .I'g I *I I I id. 8+#. 0ampf tufuhr 30ampfru[uhr Dampfaufuhr -' F g. - l- .3 -. - .l. L . nach &J Jnach R6 $t na R3 42 L ' a-s --'-J
*J*' M l
l
- l l
l l N ' t, 3 4 5w- ' J - - -- l - - - l -- - -- lm
._-! ,sn, n,. ,
4
- l - -a - 'F-ln- -.1, 'd25l .
L
- y. r s'g.[ - -=-
gygr.<.. = - 4 4-- 5 ,t, y m ,~ v_-g3 p . . l- **
-t i l l rAnderuna er ' w ' >
s.: :- . i- l l l , St,.velrtIons Kon n, b ys -14 .e. 4 ;
- i t i J i
4
._ = u .... i i i i . t. c: eur # - a- i i i 1., ,2 i R + r ,-y J, 3 I I l- !=
- a. , . ag . -- \ .-u -- n.
3 ad n -- a--- i
- x::n.:.
g .;.. e,? .c o,
- :: . =. . q . . . : . . .. . .:n,.-
..- s s as, s
j a - ' .
~~-
l l l : : l C,3. . m so n ao u.so es.co an B u a a e a u ei.so 2i.ao w. a i .co 75 3 g TIIA E M H O U R S E I i i i, , f ~s } Figure 9.C.2-11 Test group F2, phase 3 and 4: atmospheric temperatures in zone 1,2 and 3 5 (R5 + R6, R1 + R4, R7 + R8) +
-{
j (reprinted from L. Wolf, M.Gavrilas, K. Mun, " Overview of experimental results for long-term,
- . large-scale natural circulations in LWR-containments after large LOCAS", University of Maryland .
- j. at College Park, Final Repo:t for DOE - Project, Order Number: DE-AP07-%1Dlo765")
l\ Development of Expected Flow Patterns October l')97 9 c:\3006 c3.non:1b.111497 - Revision 1 i
9.C 80 i i e 4 1 bor I F2 -R9 top e ,
..totot pressure pia l
3 '
-soturation steam pressure (100% ret. humidity) p., ~ -~
tiol steam pressure pp. measured values *erior tend) i I E2 a unsaturated ( - gdy w (%= - N , I j _ 3$gypt ' unsaturated ' unsaturated 4 cir injection M [4 periods of altinjection% 0 I ' i i 45 50 60 70 80 90 100 h 110 .; time
- i 7s b end of Phase 4 Figure 9.C.2-12 Thermodynamic state of steam air atmosphere in R9 top (H=7.6 m)
(reprinted from LWolf, htGavrilas, K. Mun, " Overview of experimental results for long-term, large-scale natural circulations in LWR-containments after large LOCAs", University of Maryland at College Park, Final Report for DOE . Project, Order Number: DE-AP07-96ID10765") O Development of Expected Flow Patterns October 1997 o:\3006-c4.ruxtib-101497 Revision 1
~ . _ _ _ _ _ _ _ - - _ _ _ . - - _ _ _ _ _ _ _ _ _ _ _ _ _ _
9.C 81 > (m
,y i
bor I -I F2-R9 bottom 3 f h I L
. , totof Pressure pi.e E \ / % -
4 E saturation steorn pressure 21 (100% rel. humidy 1 peat unsaturated l l i - partiet steam pressure pp. y b t _ _
,(rneasured vetues. error band) fin 7 TN TC -
Q 4
. d air injectlen i
MM3M MMdE fE o!r inject. e%n4[( g kkfj) I i i U i 0 45 50 60 70 80 90 100 h 11 0 i and of phan 4 l 4 f. Figure 9.C.2-13 Thermodynamic state of steam air atr..osphere in R9 bottom (H=1 m) (reprinted from L. Wolf, M.Gavrilas, K. Mun, *^0verview of experimental results for long-term, large-scale natural circulations in LWR-containments after large LOCAS", University of Maryland at College Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765") i-~ Development of Expected Flow Patterns October 1997
- c:\3006 c4.non:1b-101497 Revision 1
9.C.82 , O
~
IV 2 GTe014 w tw 14 4V 2 Cii006 w I 6
- y asoio w tw is sv 2 anon w tw n F02 a sv i anow w tw w av i an0on w tw
- 2 t-r u se si, i i i i i i i
-A, ,2---- g i 13 I I I l ,
l , . Phase 2 . , , , , i 8 ____I___J___ ti_ ' L___L___L__J d_ i I I I I l 1 1 I I l- I i l I I I I S@. _ _~~_~~_"L _ _ J _ _ _ J _ _ .1_ _
'I _ _ J _ _ .L _
oC I I I I I I I I I I I I I I I e 8 i g' .. i. __L___
- g. .- . ',. , , _ L _ _ J, _
, _ _ y _ ,_ .
ss. . ::- J7_.,,
% __y.
i
..L___L___L__J___J l l i , . .g i I I I I ** e, 4 I I I I I l .: . ~ ~~ ~ ~~ ~~~ '.- '8 7. idy 6-.~m. c".wT ~i. .a .. h. ,
l Temperaturschic tung: l g Y.- ,. C. t]i.il; a- It il 7,6 m l l
' . ' - . ,.L__L2 l
1
- g. . Ol;;C *:*. ,, ii = 5.4 m . Jg _ - .L - _ _
. - . . . , . g ;3 11 4.3 m l
- :;. .:::p.i:1..!.. q :.'::: 4 u 3.2 m
- a. .. .: '. :' I i 5 II I s
; g 2.1 m ; g k ==_ i ew , ,6 11 1.0 m , , $5.00 45.50 48.00 45.50 Si.00 55.50 54.00 55.50 St.00 55.50 65.00 61.50 TIMEIN HOURS 2
Figure 9.C.2-14 Test group F2, phase 2: atmospheric temperatures in R9 (Zone 4) (reprinted from L. Wolf, M.Gavrilas, K. Mun," Overview of experi nental results for long-term, large-scale natural circulations in LWR-containments after large LC CAS", University of Maryland at College Park, Final Report for DOE - Project, Order Number: DE-AP07-96ID10765") O Development of Expected Flow Patterns October 1997 o:\3006-c4.non:1b.101497 Revision 1
. .- - - - - ~__
9.C.83 O 1Y -2 Cit 014 F ( F 14 4V 2 C19006 F [F 6 2v 2 cit 0iG F tF io sv 2 can
- tr n F02 8 ,3V, 2 C19030 F [F 30 OV 2 C19m2 K (F 2 Q
~
W, ,n,r. A.4 4/r. 68 C. MM 4 l l l 1egeraturschicitung: ' "i 1 p l l l J ^ ?.i ' g i l l l 1 H e 7,6 m 2 ll = 5.4 m 7
~ ~ ~ "~I l
I l
.1 l 3 11 a 4,3 m '. 8-[ a f
I 4 H e 3,2 m ai J l '" ';E
, , 5 Il = 2.1 m .i 4. % ,8 . _ - _g ' __y 6 ll . 1.0 m -
- . g- ,
- g/.::,.
U , l . l l l : <-.:.,,
' {f. g3l', .,:
i I I m 47_ _w_e. .n ig . e' jy iRg. ___ i oberhitzt..e}___ I l l t i I i E g i 1 I I NWh I a-
- s. __ _ L . $ L _ _ .J _ _ _ J _ _ .' .L _ _ _'
_ _l_ _ _ q _ _ _ q _ _ _ _ . .,.
,O . ._ __ _ % .. __. _. _ _L l_ _J d - 1 l
I l I I I I I I i I 1. I
' I I
I I I I g i i I 1 L__L__ 1
- g. . _L. _ _. l _ .l .l. - ?) . J _ .L _ _.
I ' ' I l l ! . k, i l Phase 3 . . Phase 4 .
. &,$DMeN','"
(1.0 bar) h I . l t
$9.00 65.50 65.00 63.50 6k.00 65.50 65.00 65.50 75.00 75.50 74.00 75.50
- T100tLIN HOURS s
Figure 9.C.2-15 Test group F2, phase 3 and 4: atmospheric temperatures in R9 (Zone 4) (reprinted from LWolf, M.Gavrilas, K. Mun, " Overview of experimental results for long-term, large-scale natural circulations in LWR-containments after large LOCAS", University of Maryland at College Park, Final Report for DOE - Project, Order Number: DE-AP07 96ID10765") 13
- V
! Development of Expected Flow Pattems October 1997 ! o:\3006 c4.non:1b.101497 Revision 1 l l.----_ _ - _ _ _ _ _ _ _ _ - - . _ _ _ . _ _ _ _ _ _
9.C-84 O CS3036 VELOCITY IN OPENING FROM R7 TO R9
.. ... o o u u o o u o u u I )
Jn I.
=
[ ' L { s q *g ' - - U n . N t f [ . S taa
- 46. 60, 66. 80. 86. 70. 76.
TIME [ HOURS)
. gttrysdytth / " ""
il. 9
-- 1 m "" + 0 ==-). .=-q lI .. l g= g -+= =- ;;1 .. =] . !. 7 ,
7-a;-
" ""4 q y r_ "i -l -== .A <
i t * ", ",,
=g e.
_ e e m hat. Figure 9.C.2-16 Velocities in opening from compartment R7 to compartment R9 (reprinted from L. Wolf, M.Gavrilas, K. Mun, " Overview of experimental results for long-term, large-scale natural circulations in LWR-containments after large LOCAs", University of Maryla nd at College Park, Final Report for DOE - Project, Ord.r Number: DE-AP07-96ID10765") O Development of Expected Flow Patterns october 1997 o:\3004<4.non:1b-101497 Revision 1
9.C 85 ) l 1 LO I I ~' cssosa 4 I VELOCITY IN OPENING FROM R3 TO R6 g l Wr [ q, u.J L i . .
,. 1 m .
l 5. o. g*
. MM M F 5 .
4
- 46. 60. 66. 60. 66. 70. 76.
TIME [ HOURS) l' nt, J Lm -- l i
$N( --m . n .,3 . y p =- hf~_=g..:= -~~ ._4 g, .- -
]
=M . . e t._-
Figure 9.C. 2-17 Velocities in opening from compartment R3 to compartment R6 (reprinted from L. Wolf, M.Gwrilas, K. Mun, " Overview of experimental results for long-term, large-scale natural circulatiom in LWR-containments after large LOCAs", University of Maryland at College Park, Final Report for DOE - Project, Order Number: DE-AP07-961D10765") O d Development of Expected Flow Patterns october 1997 c:\3006<4.non:1b-101497 Revision 1
9.C-86 9.C.2.1.2 The Influence of Initial Temperature Distribution, Location of Hydrogen Injection, Duration of Injection, and Size of Vent Openings on the flydrogen Distribution (BMC Tests 2,4,6,12 and 20) Another set of experimental results obtained in BMC is presented in Wolf et al.,1994. The temperature and hydrogen distribution are studied first for two compartments (Phase I) and later for the whole containment (Phase II). These experiments are not directly related to LOCA and MSLB situations, but contribute toward a better understanding of the influence of stratification and circulation phenomenas on the hydrogen distribution inside containments. Although Wolf et al.,1994, compared the experimental results with the GOTHIC containment code, the comparisens are not discussed because of the non-prototypical nature of the experiment relative to the AP600. The results of the experiments with only two compartments (upper and lower) are first presented by Langer et al.,1979. The total volume of the two compartments is 72 m'. The central compartments R1, R3 (form lower compartment) and R2 (upper - see Figure 9.C.2-2, 9.C.2-3 and 9.C.2-18) are used for the test. The opening size between the two compartments can be adjusted. Experiments are performd both with and without orifice (with an effective circular 2 opening of 1 m ) between compartments R1 and R2 (see Fig 9.C.2-18). Uniform injection of hydrogen-nitrogen gas is provided by a flat circular plate with a diameter of 2.5 m. The upper containment is preheated with warm air for several days before the start of some experiments to pmvide stratificatiore . Tests 2,4 and 6 (presented by Wolf et al.,1994) investigate the effects of the vertical hydrogen O distribution. The measurement positions are located near the bottom (levels 1 m and 1.85 m) and at the top (levels 5 and 5.5 m) of the containment. The experiments study the effects of . The hydrogen injection rates Test 2 has a longer time duration than test 6
- The locations of hydrogen injections The hydrogen-nitrogen scme is located above the pool surface in tests 2 and 6 The hydrogen-nitrogen source is at the 3.4 m elevation (above the mid-elevation of room R1)in test 4 The vent flow area (between two compartments)
An orifice plate is present between R1 and R2 in tests 2 and 6 Test 4 is performed without the orifice plate Development of Expected Flow Patterns october 1997 c:\3006w-9c.non:1b-101497 Revision 1
9.C 87 i 4 i
- 'e The initial temperatum distribution in the containment (homogeneous versus stratified)
{ A uniform temperature of 19*C is applied in test 2 W temperature is a uniform 22*C in tot 4 [ A temperature stratification of 19'C in the R3 and R1 (lower rooms) and 35*C in the R2 (upper room) exist in test 6 l t {: 1 l The hydrogen prusure ratios at the top and bottom of the compartments are prnented in Figure
- 9.C.219a, b, e for the second, fourth, and sixth experiments. A comparison of the hydrogen i
- partial prasures shows the effects of the hydrogen release position (tat 4) and the initially
! stratified temperature field (test 6) on the hydrogen concentration stratification.
- j. The experimental findings presented in Wolf et al.,1994 are:
4 +
- 1) The hydrogen is homogeneously distributed through a compartment M the hydrogen
- source is at the floor and the feed rate is low, even if an orifice plate is installed (see the
- results for test 2, Figure 9.C.219a). Note that the feed rate in the second experiment is
[ lower than in the fourth and sixth experiment. Also, hydrogen is released for 225 i minutes in test 2 and for 125 minutes in test 4 and 6. ' !: .i j f 2) Vertical concentration stratification occurs if the source is located above the floor (see Ig results for test 4, Figure 9.C.219b). For low kinetic energy, the diffusl<m procus slowly [ [ equalizes concentrations. ;
- . 3) If the openings between compartments are relatively small, the transport of hydrogen may be obstructed (see results for test 6, Figure 9.C.219c).
- 4) If an initial thermal stratification of air exists and an orifice is installed between the lower -
l and upper compartments, the transport of the lighter H,/N 3gas mixture is prevented.
- The highest hydrogen concentrations exist in the lower, cooler part of the compartments,
- where circulation and mixing occurs (see results for test 6, Figure 9.C.219c). The initially .
[ stratified temperature field is provided by keeping ti.e upper compartment R2 at a higher j temperature ( 35*C ) for several days before the stan af the experiment. l: Figure 9.C.219a (see results for test 2) shows that the buoyancy of the rising plume and the s . circulation resulting from entrainment into the introduced lighter H /N, gas mixture lead to a j relatively homogenized atmosphere. Circulation and mixing are present in both the upper and lower compartments. y Test 4, with an elevated source and reasonably' low ' kinetic energy (Fr is not reported for the , L : tests), showt Aat there is no significant driving force for circulation below the break eleva6on. , Stratification into two ree,lons occurs, one below and one above the break elevation. The lower i
- l. Develo;Iment of Expected Flow Patterns ocueer 1997 3
.. o:\3006w.9c.rmn:1b 101497 Revision 1 n w- m y g v e-e- -- mew q wv y -33+%%ee4 wt -e e- r e a--s em= v t- v ~*r-em r-e"+=m &'e=e C wW,r w w'--*WN'--
v"* e -F -p'f r
9.C-88 region is almost stagnant, while circulation and mixing is present in the upper region (see GOTillC numerical simulation results by L. Wolf,11. liolzbauer, M. Schall,1994). In comparison, test 6, which includes an orifice between R1 and R2 and a stratified temperature O field in the upper R2 compartment, shows an almost stagnant upper region. It also shows an increase in the concentration of the lighter li,/N, gas mixture in the lower region (a result of the circulation and mixing la the lower regions R1 and R3). The lighter gas mixture is not able to penetrate into the upper stratified layers due to the presence of the narrow orifice. The circulation cell formed by the gas mixture injection into the lower compartraents does not communicate with the stratified layers in the upper ccmpartment (see also GOTlilC numerical simulation results by L. Wolf,11. Holzbauer, M. Schall,1994). AP600 Application These tests are not relevant to the AP600. The cold AP600 dome prevents the stratification that results from higher temperatures of either the vertical walls at the high elevations or the ceiling. In the BMC case, the higher temperatuies of the we.ll surfaces are maintained fc,r a long period of time due to the heat accumulated in the concrete walls (which are heated for several days before the start of the experiment). The AP600 containment is made of steel. Natural convcetion at the outer surface of the AP600 containment walls keeps their temperature low, so that a highly stratified initial temperature field is not possible. Even if initial strt.tification exists, the application of water on the outer containment surface decreases wall temperatures and causes circulation inside the containment. O Development of Expected Flow Patterns october 1997 c:\30%w 9c.non;1t>101497 Revision 1
. - _ - - - . _ . _ _ _ - . . . . _ - . _ _ . . . _ - . - . - - . . - - - . - - - - _ _ _ .. - .- ~ ~ _ -
9.C.89 4 1 y: l' .* . : . t *. ' . * . , ., . . '. l
.<.:.. g7.... ,. l s . .' '* '. ' : ' - l . * * *' *. ' . . ' .
s,.*.* ..., 1 .
. ORIFICE s
- R9 9 f _
1 U W. ..., y,\ " ' R1 - Rt ' 0 5 .'. 0
. 53 -
U. l'f.
.gt; -
h.ca n:c..;-
- l. no *.*<
. af "1Tt. '< '* ^ j ul d '< MS S q M6 U T'. 'T #8- ~ ? * ,*. ; g $ .l \ h "'
v It
.4 ,81 3 n v. , )*
i
.. . I, l .. * *.
3" , W , .
=
4 .-l.. ' *
' , . .,'*'l.*.. , . , '. .. * .. *. . . l, ..v* .* a.
j ' *
- ( . .
i una,o u nrrs m m N NITROGEN-HYDROGEN , 4 SOURCE
- (34% N2 ) (66% H2)
E i Figure 9.C.218 Vertical cut through BMC with orifice in between R2 and R1. (reprinted with permission fmm authors from LWolf, H. Holzbauer, M. Schall, " Comparison between multi-dimensional and lumped parameter GOTHIC-containment analyses with data", Proceedings, Volume II- Ihermohydraulics of Containment and Severe Accidents, May 30th -
, June 2nd,1994, pp. 321330.)
Development of Expected Flow Patterns october 1997 o:\3006<4.runtib-101497 Redston 1 1 _ _x_ .. r - - - _ _ . . . _e m ._.. ,,. _ , - , ,,, , . _ . . , - , . ~ . . - .- -
9.C-90 I l 9 Hydrogen Pressure Ratio in R1 and R2
]g . . . .
3
- h. .
i ; E01 s
.I ,
N ***'** ***** ********* d conocaon wnee2A E DPMCWie g
* + ameau. = o. 300. 200. Battelle Test No. 2 Time [ min)
EOI End ofIg}ection Figure 9.C.219a BMC test no. 2: Comparison between experimental data and 2.d GOTlilC computations for hydrogen concentrations (reprinted with permission from authors from LWolf, II. Holzbauer, M. Schall, " Comparison between multi-dimensional and lumped parameter GOTillC-containment analyses with data", Proceedings, Volume II- Thermohydraulles of Containment and Severe Accidents, May 30th - June 2nd,1994, pp. 321330.) O Development of Expected Flow Patterns october 1997 o:\3006<4.non lb 101497 Revision 1
9.t 91 \ Hydrogen Pressure Radoin R1 and R2
}g _ - eme i
o E s -
--- ===.om a mue i. ' </ - .
magami:
+ wue.im.
a mue.u. /ri 'EOI "3"< [ 5 .4....... **
..#. ................... . l. / l /./
5;i f'/ - i 100. f ! ano. D** (# Battelle Test No. 4 g -
] EOl . End of Iq,hedon Figure 9.C.219b BMC test no. 4: Comparison between experimental data and 2-d GOTHIC computations for hydrogen concentrations (reprinted with permission from authors from L. Wolf, H. Holzbauer, ht Schall," Comparison between multi-dimensional and lumped parameter GOTHIC-containment analyses with data",
Proceedings, Volume II- nermohydraulles of Contalnment and Severe Accidents, May 30th - June 2nd,1994, pp. 321330.) m U Development of Expected Flow Pattems octobn 1997 cr.\3006<tnceulb.101497 Revision 1
._- _- _ . - ^ 1 9.C 92 ~ Hydrogen Pressure Ratioin R1 and R2
]g . . . . . . . . . . . . . . . . ....................!.. v . . . . . . l. . . . . . . . ..
e . co m e m (ni) h . .
..*.i oonac.mtu) j- . .- ,,, .
E an4ca n. g . . .
.- m. + an4ca s.
a summamim . 9
~
i -
/,l - -- --],,,. . . *
- 7,Y~/
EO! g , ,
- o. so. 100. no.
Time [ min] A EOI = End of injection Battelle Test No. 6 W Figure 9.C.219e BMC test no. 6: Comparison between experimental data and 2d GOTillC computations for hydrogen concentrations (reprinted with permission from authors from L. Wolf,11. liolzbauer, M. Schall, " Comparison between multi-dimensional and lumped parameter GOTIllC-containment analyses with data", Proceedings, Volume II- Thermohydraulles of Containment and Severe Accidents, May 30th - June 2nd,1994, pp. 321330.) Development of Expected Flow Patterns ociater 1997 cr\3006<4 rmitul01497 Revuton 1
9.C 93 in the second phase (Langer and Baukal,198?), the full model containment is used for experiments. The effects of: (1) the initial temperatures and humidities,(2) the geometry of the
] containment, and (3) the location and rate of hydrogen release are investigated.
l The results of tests 12 and 20 are presented in Wolf et al.,1994 and are compared with the results of three DTHIC modeling strategies. Tests 12 and 20 a. t performed with six compartments (Rl 2, R5-8, see Figures 9 C.2 2 and 9.C.2 3). The hydrogi..vnitrogen mixture is , injected into rooms R2 and R6 in tests no.12 and 20, respectively. ! 1 Test no.12 is performed with a uniform initial temperature. It results in a homogenized hydrogen distribution in the containment (see Figure 9.C.2-20). The stratified iniihl temperature distribution in test 20 results in higher hydrogen distribution in the lower level compartments (R1, R6 and R8 - see Figure 9.C.2-21a, b, c). An explanation for this unexpected result is that the circulation cell formed by the injection of the lighter gas mixture is not able to penetrate upper stratified layers at the beginning of the experiment. This is similar to test 6, which includes an orifice and stratified initial temperature field in the upper compartment. After three hours, there is a tendency toward decreased gradients in the concentration field, especially between R1 and R2 compartments. This indicates that global circulation affects the upper stratified layers. A summary of the experimental results is:
- 1) If the temperature field was uniform (test 12), hydrogen was homogeneously distributed (v ) inside the containments,
- 2) For an initially thermally stratified field (test 20), higher hydrogen concentrations are present in the lower (cooler) compartments at the beginning of the experiment.
Both groups of experiments indicate that good air circulation inside the containment (in fact a uniform temperature field) is crucial for homogeneous hydrogen distribution. Note that in the first group of tests, the stratification is obtained by preheating the upper room with warm air for several days before the start of the experiments. A comparison between this experimental data and the numerical results obtained with GOTillC (with lumped parameter and multi-dimensional analyses) is presented in Wolf et. al,1994. lb Development of Expected Flow Pattems ocic*er 1997 oA3006w.9c.ncalb 101497 ReviAm 1
9.C 94 Battelle Test No.12 8 Hydrogen Pressure Ratio in Room R5 and R6 h j 1 ' AI
~
EOI ' d .......................,.a' i ,. j a g e
, , . .................*..f............ ...- * * .i . l 001 HIC lp(RJ) * '""-" OOT1HC lp (R6) . i d :PMC Mas.
i + smCww S .
, i .- .
- o. 20. ao.
Time [ hrs) g Hydrogen Pressure Ratio in Room R1 and R2
~1 . ...- ^
- ...* EOI%
a f, E .* g a * *
. * . 00THIC 1p(RI) . , . """"" OoTHIC.lp(R2) .
- a smC w.
+ emC us..
8 , ; , .
- o. :o. ao.
Time Jhrs) EOI = End of injection Figure 9.C.2 20 BMC test no.12: Comparison between experimental data and GOTillC-Ip computations for hydrogen concentrations (reprinted with permission from authors from LWolf, II. Holzbauer, M. Schall, " Comparison between multi-dimensional and lumped-parameter GOTHIC-containment analyses with data", Proceedings, Volume II Thermohydraulics of Conta!nment and Severe Accidents, May 30th - June 2nd,1994, pp. 321330.) Development of Expected Flow Patterns % ,3,7 o:\3tXbc4 norritul01497 Revision 1
9.C 95 Battelle Test No. 20 ! Hydrogen Pressure Ratio in Room R1 and R2 i , , , i . I i 00THIC-2D(RI) o
"'"""" 00THIC 2O(R2) < ' ' * <
y*gR* .
~*~
4 E BPMC 83 (RI) ,i' . g + BFMC90 (R2) , e ....'.- , g
" . ,/,,,;:
- . i EOI
- jj _.......................j............g..,Z- . . . . . . . i . . . . . . . . . . .?
f f E o : : - N a : g
- e
g _m......I,,,,.......
+
i i f f
~
C 0. 1 4. Time [ hrs] 4 Figure 9.C.2-21a BMC test no. 20: Comparison between experimental data and 2-d
- GOTillC computations for hydrogen concentrr.tlons (reprinted with permission from authors from L. Wolf,11. IIolzbauer, M. Schall, " Comparison between multi-dimensional and lumped parameter GOTillC-containment analyses with data",
Proceedings, Volume II- Thermohydraulics of Containment and Severe Accidents, May 30th -
; June 2nd,1994, pp. 321330.)
v Development of Expected Flow Pattern:. october 1997 o:\3006<4 non:1b 101497 Rniston 1
9.C 96 O g Hydrogen Pressure Ratio in Room R5 and R6 ci , i r i ; .+++ 1
+ +
g -
,i . +
- i. ..
~
s*..............
.......: ..g ;
g! _...........*............j! .
,v, cr.'. . . . . . . . . . . . . . .l. . . . . EOI ,
3 f...* . g U <
. GOTHIC 2D(RS) 8 ,, + "' " GOTHIC-2D (R6) , +/ .i E BFMC 107 (RS) f BFMC 130 (R6) k I,/ + , l ..;" i e i in
- 0. '
~
1 4. Time [ hrs] Figure 9.C.2 21b BMC test no. 20: Comparison between experimental data and 2d GOTHIC computations for hydrogen concentrations (reprinted with permission from authors from L. Wolf, H. Holzbauer, M. Schall, " Comparison between multi dimensional and lumped parameter GOTillC-containment analyses with data", Proceedings, Volume H - Thermohydraulics of Containment and Severe Accidents, May 30th - June 2nd,1994, pp. 321330.) Development of Expected Flow Pattems october 1997 o:\300&c4.non:ltul01497 Revision 1
9.C.97 l t lO
)
j Hydrogen Pressure Ratio in Room R7 and R8 I I I l l g i . oomic.2o(,,) ....................3...........p . '
........... oomic 2o (u) i. + ....
s- . ame 122 (n,) . !. < g + sme1:2(u) . ,,y 'j
- ~
l . 4
+ +..
- e. _ . . . . . . . . . . . . . . . . . . . . . .*. . . i.. .n : ::' ..
a
,S 35 :
Of. C .-
.."E.n............. ' /... . : e :
E +..- ; . EOI -
+... .
,O H . . . .<:.,.. + i , i d 6, - - 2. 4. Time [ hrs] l EOI = End of Injection i i i i i Figure 9.C.2 21e BMC test no. 20: Comparison between experimental data and 2-d GOTIIIC computations for hydrogen concentrations 4 O i Development of Expected Flow Patterns october 1997 c:\W4.nortib 101497 Revision 1 1
,- ,-- .,-,,vn-.-r:-- , ,,..e-,- , . - - - ,
9.0-98 9.C.2.1.3 Effects of Sump ileatup on Global Natural Circulation (Experiments RXI . RX5) The third set of experiments performed in the DMC (Fischer et al.,1994 and Petersen et al.,1994) examine the effect of sump heatup on global natural circulation inside the ec,ntainment. The h starting and transient behavior of natural circulation for small temperature differences, the influence of natural circulation on mixing of hydrogen relcred during accident conditions, and the effects of stratification on the natural convection formation are also studied. A total of five experiments are performed (RX1 to RX5) at atmospheric pressure. Temperatures are recorded in the sump, in the containment atmosphere, and in the concrete structures. The relative humidity, containment pressure, liquid sump level, velocities (in the vents), and hydrogen concentration are also measured. The objective of long term experiments is to establish at what sump temperature global circulation exists. During these experiments, the containment atmosphere, structure, and sump have nearly identical temperatures. Circulation effects inside the containment are already present with a sump temperature as low as 25'C. Experiments RX2 (without hydrogen injection) and RX4 (with multiple hydrogen injections) are performed as long term tests. The respective initial and boundary conditions for all experiments are given in Table 9.C-4. Results are provided for only the RX4 experiment, since the hydrogen distribution is available for this test. A summary of the results for the RX4 experiment, with the cold containment and multiple hydrogen injections, is presented in Wolf et al.,1996. The perspective view and cross-sections of the BMC containment, illustrating the compartment numbers and the location of the h hydrogen injection, are presented in Figures 9.C.2-22 and 9.C.2 23. The instrumentation plan for the RX4 test is specified in Figure 9.C.2 24. At the beginning of the experiment, the temperatures of the structure range from 20-26*C, The sump temperature is 20*C (see Figure 9.C.2-25). Several consecutive characteristic periods evolve during the experiment. The sump heat up is divided into three periods:
- 1) 0 to 1:48 hr - the sump is heated to 50*C
- 2) 2:43 to 3:39 hr - continuation of sump heating to 60'C
- 3) 3:34 to 4:52 hr - continued sump heating to maintain the temperature at 60 C until the end of experiment (5 hr)
Three hydrogen injections occur:
- 1) 1:11-1:24 hr,236 g of hydrogen is released
- 2) 2:11-2:23 hr,215 g of hydrogen is released
- 3) 4:06-4:33 hr,319 g of hydrogen is released O
Development of Expected Flow Patterns october 1997 o:\3006w 9c.non:1b.101497 RcC ion 1
9.C 99 At the beginning of the sump heatup, the anemometers register velocities between 0.2 0.3 m/s (for sump temperatures 24 27'C), while at the end of the experiment, velocities are 0.6-0.8 m/s ! (see Figure 9.C.2 26) - At the end, the temperature of the dome is 30*C (see Figure 9.C.2 27). i Shaded areas in Figures 9.C.2 26 and 9.C.2 27 represent periods of hydrogen injection. Velocities increase during periods of hydrogen in}ection. Sump and atmosphere temperatures are presented in Figures 9.C.2 25,9.C.2 27, and 9.C.2 28. , l Temperature differences in the area of the center compartment and dome are not greater than ; I 2'C (Figure 9.C.2 27). The temperature difference in the external annulus is smaller than 3'C , (Figure 9.C.2 28), indicating the presence of natural circulation effects. Due to the natural circulation, the hydrogen distribution is almost uniform in the whole containment, see (Figures 9.C.2 29 and 9.C.2 30). After two hours, the relative humidity of the
; whole containment atmosphere is 100 percent (see Figure 9.C.2 31). Even low natural circulation flows provide complete mixing of the hydrogen and steam (evaporated from sump). The heated 4
sump provides sufficient buoyancy force for natural circulation flow. i l } O 1 4 i O Development of Expected Flow Pattems october 1997 o;\3006w-9c.non.1b 101497 Revision 1
4 32 9 gg Table 9.C-4 Test Matrix of Battelle Sump Heatup Experiments {'
? -e gg Containment-Atmosphere 8
l3 Test No. Containment Initial Temperature Total Pressure Structure-Temperature [*Cl Samp-Ten ocrature (*CI H,-Mass [g] Io- Media 1 Cl Ibarl RXI air 235 -285 24 - 26
?}?
41 1.011 1) - 40
- 2) - 50
- 3) - 55 h 4)-60
- E y RX2 air 22 - 285 1.009 245 - 28 1) - 40 g 2) - 50 g 3)-60 RX3 air + steam 48 - 585 1.008 42 - 61 50 - 100 RX4 air + H, 24 - 27 1.014 225 - 26.5 1) - 48 1) 236 2)- 60 2) +215
- 3) +319 _
RXS air + steam + H, 55 - 69 1.001 39 - 64 1) 62 - 19 1) 371
- 2) 63 - 56 2) +390
- 3) +406 1
(reprmted from L Wolf, M. Gavrilas, K. Mun, ' Overview of expuawr.tal results for long-term,large-scale natural circulations in LWR-containments after large LOCAS University of Maryland at College Park, Fmal Report for DOE - Propct, Order Number: DE-AP07-
%ID10765-)
i t 9 3'I h i O O O
9.C.101
' f ss \ .f ~%/
ss u l u j
,I as ,
j ne
" .. , L - ng M .w l c ,(\ <
( h hl / - 4 __ a-- .--- r := ,
~. . .
- . ; . <. . n . i ..c1T ....:e17tn;
. : r.: . .',~ . . .. p. .. 1. \ .s= -
iw., ..-. , 9[
;; t a ]c '" ,, ,,,,l .o n2 -g - - 4*'i.s , ?,r,.
1 n :6
- n. .;
a , .
'1 '
l ps9'-M
,,,,'". 6 "' ....
- n. . . .. . . . o... t..* ..
" '" t .. . . .. ; l. : :*.; .. .. .... : l ) $ f
- w. j- .
o., n..
- l .
s
,, u. .l. . '^
- bur Figure 9.C.2 22 Perspective view, vertical and upper horhontal cross sections through BMC with main i. team feedline (reprinted trom L. Wolf, M.Gavrilas, K. Mun," Overview of experimental results for long term, large-scale natural circulations in LWR-containments after large LOCAS", University of Maryland e at College P. rk, Final Report for DOE Project, Order Numben DE AP07 961D10765")
Development of Expected Flow Patterns miod,.,3,97 vr\3006 c4.nortib-101497 g,yt,g,.,, 3 j l
9.C 102 m sy...
+/ ,
s, =
, ~ .t< _ y , -- un , 4 .. sr 1 .
k / c p,:. . .
.,s . . s.... ,. e., . . :,
- w. . m .w:nw.
.,, ,,' I .. r<, . - , .h . , .. ., v.
r.y
. o .
97 .. .. . :.n. i %x -
,m H. sim! ?
l-. tl2 i ,.; .4 .
.v. . - - .s. - . / .. .::. . i.'L. .: _s:y.wd. .:.C. G::.:
7 7 ., -
., . . ' . ![.* : I' .$ i'f:i . . :! . -4 % -,... .q _ H. s. ., ,,, .-ge '
4
..i .,,, .u 1 4.'""'"
Figure 9.C.2 23 Perspective view, vertical and upper horizontal c oss sections through BMC with auxiliary steam feedline (reprinted from L. Wolf, M.Gavrilas, K. Mun," Overview of experimental results for long term, large-scale natural circulations in LWR-containments after large LOCAS", University of Maryland at College Park, Final Report for DGE - Pro}ect, Order Number: DE AP07 96fD10765") Development of Expected Flow Patterns october 1997 o:\3006-c4.non:1b 101497 Revtsion 1
__. __ _ . . _m _ . _ _ _ . _ _ . _ _ _ . _ _ _ ,_ _ _ _ _ _ _ _ _ _ _ ...__.. _ __..___. ._, _. _ ~ _ 9.C 103 9 n ese W M 9 PL 2M A 12
- 9 it !?! A 12 9 n too u 40 9 TL 080 E to
- N . I l'. ! ! .' '. ' . ' .'l
- 9 TL 2M N to j 2 n m 0 M $$ '8 N **. * ' * * * * *,.* . 9 HK 270 A 37 3 n te0 3 33 . . .
** N '
N.. ' , ' . * * . 9 7K 210 A 37 i N ,' l ' ,, 9 7L 270 A 60 W T TL 900 A 13 W % . 3 TL 000 A 33 W ! 9 TL epel SI W ' . 3 7L 3401 If W
' 4 '
3 7L 270 A 12 W 91L DIO R M 9 TL S80 A M W .% -
- r. w i . ,
9 7L 2M t 33 W C TLIID A M w
@ 3 9 7L IM A M w $h'. . t DTLS80R22 A ; h. ,
f C TL 270 A M W 3 7L 000 M 31
- P #7 -
' -97L2MRM - 9 TL 2M R 22 1 3 7L es0 8 al * ~@: 3 vr am u 2
. 3 ft =0 A N w - .. .-* 4 B TL S80 8 30 W - I T3 080 N il '" I
'N ;;E yk-- g .
1 D TL 000 L O2
~ 3 HK 2M A N Ll ap \ l . .,* ;' a .
4 9 n ge0 R le *
, ;I # q IS TK 270 a M -~ "
9 7L 9801 IO W-- % W HK 2M A ll . 8 These A It w 9 TL WS U 05 ll::' F
! b -- Mse-! hf II .)::';" *'
9 TK 279 A ll CTL 2M A 10 W 9 VT R$ U te su -
. ' , .. **. .. . 9 n 2M A lo w sM'* - 9 7L 270 R 80 >gil::,y3 M -- ' ' . 9,:.14. .:.: / -l.
3 . . . . 9.nmA sm , i it w
. . . . . ....i..... . . . ... . .gg 4 ,inggi ,,. ,,, , ,, , /j 90- m . Ti m u i.
- a n = w .i w ,m U TL ,
- c a ^
- e g ti:.:: n :. 5 9 a 0. u M . a ::.::. ., . 7 m . .;. ..
a 9 a s. u .0 - :.: n:. a. . 9n=== w '
+
3 TL 000 M $3 ,
** N-9 TL 000 R M p ' 1 <
4 3 v7 000 M 42 %
' ,h '
j 3 TL 000 M 31 - -- 1:. l a? '
- 9 7t 100 R 10 i
4 n 000 w 22 -
.a ..
N lfim:
.O L L' ; - - -
g *.* q 9n .
, , . .:' / ., -
i 9 n om o0 uemi0
- s. "
u "R^ v d 9 Yr lai10 U OS - - .
'* \' '-, %'8--'*. E_
v
. I' ,
m ~ .i l '.
.. .* ,/* ** .' . . D F12il 3 TL 000 M 13 - 2:sW, ? *- ei m! 'A a, . * * . .
- M D TL 271
= . 3 TL 000 u IO 8 --
,* ........l.. - . . D PL 271 , , , , , , ,. . .l. , . * * * ** D TL 000 L 02 3 TL 000 M 01 $ - . .
- , .. r r. s pe
, a8 . ....l......;,...,...'.. . .*. .f. ... ..,a a H
W. HWe-Demplainsp.esong i D ri m 0 le0' Figure 9.C.2 24 Instrumentation plan for BMC - sump heatup test RX4 (reprinted from L. Wolf, ALGavrilas, K. Mun, " Overview of experimental results for long term, large-scale natural circulations in I WR-containments after large LOCAs", University of Maryland at College Park, Final Report for DOE - Pro}ect, Order Number: DE AP07 961D10765") Development of Expected Flow Patterns october 1997 oA3006c5 non.1t*101497 Revision 1
9 C 104 0
. l::::N - -
It':t 5 ll [ ,!
..... ., - - . p .. ...
m p e6 y
- u["w;
': ti- Ittit:
s s. a P
**i:=::==a%22s ' f :' , istih - <-
c - s!!!.i 4N g/.. .. .
- . . . . 2 --
i g; ; . ---- - :>7
'gg;s e ' '* ^F_
tatt:--- -
*. ? -
e s'Eli'#_
==::: = *- @" m:::: - = :.
in =::p:==:
.ri . :. : ..:!.f.:. ../: .:. ,.::.:-2;;; , '.i::: #E .n.. . :n:-
r w I" " (si
. .. k. [II y JTLodeMios .$' %p M'y-o nthooMis a maeouso ~
s g r= l g , a m Aeon 3s e FTIA96R1e lH3 Bassetsumal ,a f j I l a t I W . be f l l t
)
3
= -==_ '
1 L JW g }&s
= ~'M; '" ' ~ .q) .. , 4- .
l . . . W
- o. t. 2. s. 4. s.
Thee liners) 1 1 Figure 9.C.2 25 Battelle Experiment RX4: Sump and atmospheric temperatures. ( (reprinted from L. Wolf, M.Gavrilas, K. Mun, " Overview of experimental results for long term, large-scale natural circulations in LWR-containments after large LOCAs", University of Maryland at College Park, Final Report for DOE - Project, Order Number: DE-AP07 96TD10765") Development of Expected Flow Patterns octoter 1997 c:\3006-c5.non:1b.101497 Revision 1 l
9.C 10$ O
!b:r Jg ms j;!!!i '8 me " ^
- s.
!;;;l:
star:::
..... ~ - .
w%.z. - in.,n ii;;;M;;;c
=
c
+(r v ..... 1 !! 7. + .
- sp;!!!
sihme!t m w::a, s1 - --
- a. lie!iz - JT Ag hn i
155:!r=WGF@df &_ 'lii!!!A;@y 0"y adk,4g%,.g. Epi: 5 . l
, ,m,,0,, - . . . .
O 9YT32,US$ A tvT900008 IH2T- 7
=
g * :
,, q_ , , , . - - - - - . . .c
- e. 1. 2. 3. A s.
Figure 9.C.2 26 Batte11e-Experiment RX4: Velocities in side vents (reprinted from L. Wolf, M.Gavrilas, K. Mun, " Overview of experimental results for long term, large-scale natural circulations in LWR containments after large LOCAs", University of Maryland at College Park, Final Report for DOE - Project, Order Numben DE AP0/-961D10765") O Development of Expected Flow Pattems october 1M o:\3006-c5.non:1t>101497 Revis6on 1
9.C 106 l l 0
-innii m m - m n=n o h:H+m. . itinh.. m -' lill, i -N ..!!".'.!%:N .% 'llil@'
gt w g .+, ~ . #, 4 li.1 e " : D--2p1 M__
- Im "
.; I ::---in i c
Gi :pt 1 W g=ksh/.. .. . 'r: i ,:
^
i !!, _ _- w Eq - l$g!In . C L , Q. '!.:gt:4l II h ,,$j! a j;.g::s-::=r a;+t .
. : ,p, m. . . 4.4. : . . ..... . < .. . . .o.. )
p {
. .:[.;;l:
ljg g
;jp . . ,u..g,g ., A ,
q,,g.;g; 3 N A 3TliOOM18 g1 i e 3 TIM 0Mel I e 3TIA00MU i 1 A
I -- 1 A yr t ^ ^ - I .t- '
N W FTLA00KM IH36astesenal 'j l [ f i
.. h.A w ._
- l i
i t W
- 0. 1. 2., 3. 4 8.
*nsme lbeunt Figure 9.C.2 27 Battelle Experiment RX4: Temperatures in center compartment and dome (reprinted from L. Wolf, M.Gavrilas, K. Mun, " Overview of experimental results for long term, large scale natural circulations in LWR containments after large LOCAs", University of Maryland at College Park, Final Report for DOE Project, Order Numben DE AP07-961D10765")
O Development of Expected Flow Pattems octcber 1997 c:\3006 c5.non:1b.101497 Revi.s km 1
9.C 107 O ~ im! . .- =::
- n o
_s ;.g. - ~ g, .li.n:.r.=
.ll 13 m.
5 a litti.:
- ,.,.. v.,s ,.y + ae gg:..
i.
'$ ;pl%:M=,h n ue- . ,' ___i,f. n;= -- I ne4_~ an. n:
- a.
;s um - - +
! t . :. ,. . I.n.
=. ,,g,,,a g . n. n..s. . - ;g:,
i i
- x , ou2,,, , .
- o nm.m. -
i & M L27.R22 )
- n m err.rramu
. um.. - [4 l rupunwoossum M
- E .
4
#s' H -
y Tb(, g ' l r
~
- M "
) g T l l l f1 _ i g * * ,
- 8. . 2. 3.
- 4. 3, TisseSmet)
Figure 9.C.2 28 Battelle Experiment RX4: Temperatures in external annulus (reprinted from L. Wolf, M.Gavrilas, K. Mun, "Ovenlew of experimental results for long term, large-scale natt.ral circulat!ons in LWR-containments after large LOCAs", Universi:y of Maryland at College Patk Final Report for DOE Project, Order Number: DE AP07 961D10765") Development of Expected Flow Patterns october 1997 c:\3006 c5.non.1b-101497 Reviskm 1
9.C 108 9 y 3KP670515 O 3KP000hD1
& 9KP000M60 , B 9KP270R36__
e4 E PKP270R10 - - l _ n a li p r. n l I O. 3i T-i
- 1. 2. 3.
e 4. i 5. Tlanc [baurs) Figure 9.C.2 29 Battelle-Experiment RX4: liydrogen concentration along circulation path (reprinted from LWolf, M.Gavrilas, K. Mun," Overview of experimental results for long term, large-scale natural circulations in LWR-containments after large LOCAs", University of Maryland at College Park, Final Report for DOE - Pro}cet, Order Numben DE AP07-96fD10765") O Development of Lepected Flow Patterns octoter 1997 o.\3006<5.non:1b 101497 Revision 1
9,C.109
- O i
f i 4 y 4KP600M32 ' ' ' ' O 8KP090M17 i A 6KP270M17 - i . E 9KP315UO5 - _ W
~
e4 e 9KP270R10 - jfp !
,,__=_y I
3 j w.
- 1. it. _
j W i !O e. E' Ir a ei
- 0, 1, 2. 3, 4. s, Tisse [bours]
i i i i i: Figure 9.C.2 30 Battelle Experiment RX4: Hydrogen concentrations in intennediate subcompartments (reprinted from L. Wolf, M.Gavrilas, K. Mun, " Overview of experimental results for long term, large-scale natural circulations in LWR containments after large LOCAs", University of Maryland at College Park, Final Report for DOE Project, Order Number: DE-AP07-961D10765") Development of Expected Flow Pattems o.tober 1997 o:\300&c5.nmitr101497 Revision 1
9.C 110
..... ..... 1
- =r: -
~ em ~~- '
- =r: -
" " ~ ~ q:j !!E!-o '!!!!!: k.. J T
list:: v ai+ mrP$ i - bps : !:=E!! a _ie::::'
- ==t:W::.1 OC -
'z _ e- ta = ,:
7
.i ;, - f -- - r,;=::.,
i 1; = ,: ' M ;f k L:e: . .' i.-*. ' ': ,' " .
, , , , , , , , .. . ._s. ' * :::=
122 ".
'::::t k ^ 'V JHK270A26 8 - , - - O ,--E ~i S O 9HK270AS7 A 9HICt?9A11
[H3_3__WJ. ,,,,,,,,,,, i n A - g
]
a s 7
- 1. _
+
L' ^
, i , , <
- 0. 1. 2. 3. 4. s, Tisse [bours]
Figure 9.C.2-31 Battelle-Experiment RX4: Relative humidity in the cont.inment atmosphere (reprinted from L. Wolf, M.Gavrilas, K. Mun, " Overview of experimental results for long teim, large scale natural circulations in LWR-containments after large LOCAs", University of Maryland at College Park, final Report for DOE - Pro}ect, Order Number: DE AP07-96fD10765") Development of Expected Flow Patterns octcrer 1997 c:\3006-c5.non:lkl01497 Res h I
- - . . - . - - - - - - - - - - ~ - . _ . - . . . - -
- 9.C 111 - l
; a 4
[ 9.C.2.1.4 References {
- 1. Fischer, K., Kanzleiter, T., Schall, M., Walf, L.,1989
" CEC Thermal-Hydraulle Benchmark T.xercise on FIPLOC Verificatio, Experiment F2 in ;
- BMC 1.ong Term Heatup Phase - 1pecification ' for Phase I, Battelle Institut e.V.,- .
t Frankurt/ Main, Germany, Sept.1980 t
- 2. Fischer, K., T., Schall, M., Wolf, L.,1991 i j
- CEC Thermal 1*ydraulic Benchmark Exercise on FtPLOC Verification Experiment F2 in I BMC Long Telm Heatup Phase Results for Phase 1, Final Report, EUR 13588 EN,1991
- 3. Fischer, K., Kanzleiter, T., Schall, M., Wolf, L.,1990 ;
I " CEC Thermal Hydraulic Benchmark Exercise on FIPLOC Verification Experiment F2 in I' BMC, Experimental Phases 2,3, and 4 Specification for Phase II, Battelle Institut e.V., Frankurt/ Main, Gc many, July 1990 j '4 . Fischer, K., T., Schall, M., Wolf, L.,1993 l " CEC Thermal-Hydraulic Benchmark Exercise on FIPIDC Verification Experiment F2 in BMC, Experiment F2 in BMC, Experimental Phases Results for Phass 2,3 ana 4, Results ; of Comparisons, Final Report, EUR 14454 EN,1993 l I ! 5. Fischer, K., Hafner, W., Holzbauer, H., Kanzleiter, T., (1994). , L " Experiments for Concerning Natural Convective Flows in Battelle Model Containment, l Documentation of Measured Data" !
- (In German), Battelle Ingenieurtechnik GmbH, EJeorn, Germany, Technical Report
[ V68270.2, Oct.1994 i !. 6. T. Kanzleiter,1988 . , "FIPLOC-Verification Experiments", . l' L (In German), Battelle-Institut e. V., Frankfurt / Main, Germany, Final. Report BleV R-66.614-01, March 1988 i
. i
- 7. - Langer G.,'Jenior R., Wentlandt H.G.,1979 :
L " Experimental investigation of the hydrogen distribution in the containment of a LWR : I l following a LOCA" <
; (in German), Battelle-Institut e. V., Frankfurt / Main, FRG, Report BF R 63.363-3
- 8. Langer G., Baukal W.,1982
" Experimental investigation of the hydrogen distribution in_ a model containment i_ (Preliminary Experiments II)"
(In German), BMFT Research Contract 150.375, Battelle-Institut e. V., Frankfurt / Main, '
- FRG, Report BF-R-64.036-3 Development of Expected Flow Pattems onober 1997 o:\3006w 9c.non:lt>101497.-- Revision 1 4 , . , . - . . , . . - _ _ - . . _ , - . _ . . _ _ _ - . . . .
0.C-112 ; l
- 9. Petert.en, K., Pamme, H., Seyffarth, L., Wolf, L., (1994)
"flydrogen Mixing by Natural Convection in PWR Containments" (In German), atw 39 (1994), 758-769
- 10. L. Wolf, H. Holzbauer, M. Schall,1994,
" Comparisons Between Multi Dimensional and Lumped Parameter GOTHIC Containmenc Analyses with Data" International Confeience on New Trends in Nuclear System Thermohydraulics, Pisa, Italy, May 30th - June 2nd, vol. 2, pp. 321 - 330.
- 11. L Wolf, M. Gavrilas, K. Mun,1996,
" Overview of Experimental Results for Long-Term, Large-Scale Natural Circulations in LWR Containments after Large LOCAs" Final Repa-t for DOE - Project, Order Number: DE-AP07 961D10765, Department of Materials and Nuclear Engineering, University of Maryland at College Park, July 1996.
9.C.
2.2 DESCRIPTION
OF Tile AVAILABLE NUPEC DATA BASE 9.C.2.2.1 M 71 Test NUPEC's Hydrogen Mixing and Distribution Test M-71 is used as OECD/CSNI sponsored Intemational Standard Problem Exercise ISP-35 ( report NEA/CSNI/R'94)29 ) to compare at i k validate the performance of various computer codes. The .WGOTHIC code is included among the codes that have been compared to this test (see ref report NEA/CSN1/R(94)29 ). A detailed presentation of the M 71 test is not provided because the test u not directly applicable to the AP600. Internal sprays, which ire not used in the AP600, are active during the M 71 test. A more relevant NUPEC experiment, the M 4-3 test (reference T. Hirose,1993), is performed without internal sprays. This test investigates the mixing behavior of hydrogen and steam injected into the lower containment compartment. 9.C.2.2.2 M-4 3 Test The test facility represents one quarter of a linearly scaled PWR (four loop) containment model with 25 inner compartinents. The approximate volume is 1600 m' with a height of 17.4 m and an inner diameter 10.8 m. Of the total volume, only 1300 m' is used. The dome compartment is approximately 70 percent of the total volume. Twenty-four smaller compartments occupy from 0.1 percent to 4 percent of the total volume. The containment walls are made of steel plates (concrete structures are nos prescat), so that the response of the heat sinks was very fast. The external containment wall is 12 mm thick steel and is insulated. Three floors are located at the 3.2,5.4, and 7.3 m levels. A large su.np is located below the lowest floor. A gas supply system for steam and helium, a containment internal spray supply system, and measuring systems for temperature, helium concentration, and pressure are installed. O Development of Expected Flow Patterns octoter 1997 . o:V.KXbw-9c.rw n:lb 101497 elevbion 1
9.C 113 l 1
- N M+3 test simulates a break inside the lower D loop of the steam generator compartment. !
his compartment is at the first level of the containment. Figures 9.C.2 32,9.C.2 33, and 9.C.2 34 illustrate the containment compartments and their corresponding numbers on each level. Figure i 9.C.2 32 also shows the location of tha break and typical circulation patterns. ; The H+3 test starts with a pressure of 101.35 kPa and an initial containment temperature of 30*C. A mixture of steam and helium is injected at a constant flow rate during the first 30 minutes. The mass flow rates simulating a small break are 0.33 kg/s and L.a27 kg/s for the ! stearn and helium, respectively. The inflow temperatures are 140*C and 20*C for the steam and ! belium, mpectively. ! Experimental data is recorded for 2 hours. The atmosphere and wall temperatures, the helium f concentrations in the various compartments, and the pressure inside the dome are recorded. The ! pressure inside the dome increases almost linearly during the release and reaches 1.6 bar after 30 minutes (end of the release). The total mass of the steam released is 594 kg. Following the [ release, pressure decreases slowly to 1.5 bar (after 2 hours - see Figure 9.C.2 35). The dome temperature increases from 303 to 341'K during the release and then decreases to 334*K (see Figure 9.C.2-36). The dome helium concentration history is presented in Figure 9.C.2 37, i The temperature and helium concentrations increase first near the release point and the compartments above it, indicating an upward flow direction. This la followed by increases in ; the dome and steam generator compartments at the opposite side of the contaltunent, indicating i a downward flow loop' direction. The global circulation loop la clored in the lower level of the containment. The consecutive positions of the higher temperatures and the helium concentration . fronts indicate the flow path due to the global circulation during the break release (see Figure , 9.C.2 38, from 0 1800 seconds). + f A comparison of the temperatures and helium concentration histories in various compartments, Figure 9.C.2-38, shows the direction of the formed global circulation loop during the break > release. One flow path starts in compartment 8, followed by compartments 15,21,25 (dome), 14, and 7. Following break release, the temperature field stratifies but the temperature difference , between the dome and first floor level is not greater than 14*C at the end of the expe:Iment. After the break release, tha temperatures in all the compartments slowly decrease, despite the fact that the outer surface of the containment is insulated. Figure 9.C.2-38b shows that the i concentration in the containment starts to homogenize after the end of the release (1800 seconds). : This indicates that natural circulation effects are active after the end of release. l Note that the external walls of the containment are insulated and the interior walls are made of steel. This limits their ability to accumulate energy and act as a heat sources (which would promote natural circulation) later in the test. The natural' circulation that exists after the end of release is most likely fonned by the presence of the hot sump at the bottom of the containment. O : A Development of Expected Flow Patterns ociober 1997 oA206w.9c.non:1b 101497 Revision 1 i
,..._w.. ..--,nn,.ne,_-,n,..,-,r, .,,--,,n-,,- --,.n v ,.n . ,n-.-. , -nn--v.~~~~n,,.,,,wn.,v,,-
9.C 114
'Ihroughout the test, global circulation is present inside the containment cnd contributes to the homogenen.lon of the atmosphere over time.
App!! cation to AP600 O This test is similar to the AP600 configuration in both the position of the inje-tion point (break at the low level) and the geometry of the containment (70 percent of the volume is dome). However, overriding the similarities are the absence of concrete structures (able to store heat and later act as heat r,ources) and cooling on the outer surface of the containment shell. This features make the test very dissimilar to the AP600. The external cooling and the hot concrete structures. positioned in the lower portion of the /.P600 containment produce strong global circulation and decrease the effects of thermal stratif cation. O O Development of Expected Flow Pattems october 1997 o:\3006w 9c.non:ltr101497 Revision 1
. . _ = . - - - - - . . - . . - ,. . . . . _ . -
' 9.C 115
/,,a*- \
N, p.m. l l . g.,.ga...r.n.r . E.'"ML'.'.. ~ o - 7 t' , , 4 v / p' /
' ] ',
a.v=. . Vr,. u rg i N<r , (9 D - l i
\ T.*A".*A... l v %<,
l B.REAK RELEASE / c y ' LOCATION q 'y . N ._ w s / i PJ;'"c%".."a'!A 8""' z.g,g.,.;.i.,e.....o.. ! Figure 9.C.2-32 Model containment used in the test, break-release position, and circulation flow paths (see arrows) (Figure prepared according to T. Hirose documentation for M-4-3 test data, NUPEC Nuclear Power Engineering Corporation,ISP35-035,30 Apiil,1993) Development of Expected Flow Patterns october 1997 c:\h5.non:1b-101497 Revuion 1 l d
9.C 116 l l 1 q a-4
&.(..
J .. s3 ,
'..i{..B g \ . d "**' /
i # d D a# ' gs, g ' ~ W s
,~
y
' /
g .. a
- /
g u'* i /
/ .,
A - khQ '*
< h1- e i
Figure 9.C.2-33 Detailed arrangement of each floor (Figure prepared according to T. Hirose documentation for M-4-3 test data, NUPEC Nuclear Power Engineering Corporation,ISP35-035,30 April,1993) l l l i Development of Expected Flow Patterns october 1997 : o:\3006-c6.r m:1b-101497 Revision 1 ___ - - -_ -___- __- l.
9.C-117 C (
~
b . I g 9 . . n r
/
L J g @K@ g j g g 2nd Floor
@g / @ 7@
t a 3 & 3 e = ' O
'~'
0 8
~ti r~ ~-"~
S h g,I'M
- l BR'EAK s
. E g em'4 5 8 l /
, ,_, , 1st Floor @ C: - , ---- member Q
4 4 Figure 9.C.2 34 Distributio.' of the compartments at each floor i . (Figure prepared according to T. Hirose documentation for M 4-3 test data, NUPEC Nuclear Power Engineering Corpo, tion, ISP35-035,30 April,1993) Ot Development of Expected Flow Patterns october 1997 o:\3006-c6mn:1b-101497 Revision 1
9.C 118 O o o N . o . oo n .- co .
- c. -
s g: 2 " -
- s '
g : ,-
- 3. % ~ ......................,,,,,,,,,,,,,,,,,,
- a. -
o - ..
- O
- b. '.
o : - o e . . . . l . . . . . . . . I , . . , 0 1.8 S.6 5.4 7. X1 Time (sec) l l Figure 9.C.2-35 Pressure history ; l (Figure prepared according ta T. Hirose documentation for M-4-3 test data, NUPEC Nuclear l Power Engineering Corporation, ISP35-035,30 April,1993) O I Development of Expected Flow Patterns october 1997 c:\3006<6.non:1b-101497 Revision 1 l
i 9,C-119 i O V S i 4 b n. S o - ag kb -
. g: .. & n -
e , e o cy g M ~ o .. Q . n w e ON < > o - t k 0' , , ' ' i.e ' e .6 5.4 7 l; Emo (8ec) X'1h j 4 i
- Figure 9.C.2-36 Dome temperature history 4
(Figure prepared according to T. Hirose documentation for M-4-3 test data, NUPEC Nuclear Power Engineering Corporatiory ISP35-035,30 April,1993) O Development of Expected Flow Patterns october 1997 o \3006<6.nortib 101497 Revision 1
9.C 120 0 c N i C - s
,Q - $ o . ... ... *... **~...'..".....=+<
E " - e - V : . O to - . O v- - . E : ae
. o -
I - - e : . E . . o .. O 'to . 2 of, , , , i , , , ,i,,,,i.,,.
, 2,
' 0 1.8 3.6 5.4 t. X16 Time (sec) A Figure 9.C.2 37 Dome Helium Concentration history (Figure prepared according to T. Hirose documentation for M-4-3 test data, NUPEC Nuclear Power Engineering Corporation, ISP35-035,30 April,1993)
. O Development of Expected Flow Patterns Octot*' 1997 , o:\3006 c6,norrib-101497 *"I
9.C-121 O g L . e g- ?.. ,......- k , 7 *
, _ _ -._ 15 g .- - .............g. . . . . . . . . . . . . . . . . . ,
y - 7 o -
.E .
- 8) k0 ,,,,i,,,, ii i, 1.8 2,.6 5.4 7,.L -
X16 Time (sec). m n N nt
~
- 8 N g -
? 7 _ ~h_
g l _,-w M. g. ... . '
~ 15 -l c. : 15,-
pong e 3 . I . .
,o-7 o
no ' b) e: , , . .1 . , , , ,,,,i..., 0 1.5 2.5 5.4 7. X1 Time (sec) Figure 9.C.2-38 a) Temperature histories for compartments forming a circulation path (8, ' 15, dome and 7) b) Helium concentration histories for compartments forming a circulation path (8,15, dome and 7) (Figure prepared according to T. Hirose documentation for M-4-3 test data, NUPEC Nuclear
, Power Engineering Corporation, ISP35-035,30 April,1993)
Development of Expected Flow Patterns N o:\3006<6.nort1b-101497 .
9.C-122 9.C.2.2.3 References
- 1. NEA/CSNI/R(94)29: Final comparison report on ISP-35: NUPEC hydrogen mixing and distribution test (Test M-7-1), Committee on the safety of nuclear installations OECD Nuclear Energy Agency, December 1994
- 2. Tadashi Hirose (1993): Letter, floppy disk which contains the M 4-3 test data, an attached document, and a video-cassette, NUPEC Nuclear Power Engineering Corporation, ISP35-035,30 April,1993 O
O Development of Expected Flow Patterns October 1997 o:\3006wec.non:1b-101497 Revision 1 I 1
9.C-123 9.C.2.3 DESCRII'I1ON OF THE AVAILABLE CAROLINAS VIRGINIA TUBE REACTOR CONTAINMENT (CVTR) DATABASE The results of the simulated DBA of CVTR containment system are presented in (Schmitt et al., 1970). The CVTR containment is made of reinforced concrete (see Figure 9.C.2-39). Itis cylindrical in shape with an internal diameter of 57 ft 11 - 1/2 in and an inside height of 114 ft. The cylindrical wall is 2 feet thick. At the top of the cylinder is a hemispherical dome made of 1/2-inch thick steel covered by 20.5 inches of concrete. The structure is divided in three regions (operating, intermeciate, and basement). The free volume is about 227,000 8ft . A water spray i system is installed at the 360 foot elevation (6 inches above the containment bend line). Three simulated DBA tests are performed. The first test does not use the water sprays. The pressure and temperature histories are obtained to establish the effects of the sprays on pressure reduction. The remaining tests include active water sprays with flow rates 290 and 500 gpm, respectively. The water spray flow rate is adjusted by changing the total number of nozzles in the spray header. The spray flow is initiated 30 seconds after the end of steam injection and remains in operation for 12 minutes. The second and third tests are not applicable to AP600 DBA situations due to the spray effects. Superheated steam, at approximately 10 F above the saturation temperature,is injected into the containment. The steam is released above the operating floor through a diffuser made of a p 10 foot pipe installed vertically on the discharge nozzle of the steam line (see Figure 9.C.2-40). tg The end of the pipe is capped. The pipe contains approximately 126 holes (with 1-inch diameter) that direct the steam horizontally (see Figure 9.C.2-40). The injection continues until the containment pressure reaches 17.5 psig above the atmospheric pressure. To obtain the containment response, measurements of the interior pressure are made using seven fast response pressure transducers and a Heise gauge. The containment atmospheric temperature is measured by 34 thermocouples and 15 resistance thermometers. Two additianal thermocouples and pressure transducers are located in two partially isolated regions (the reactor header cavity and the refueling canal) to determine their thermal and pressure response. Twenty three thermocouples are positioned at the containment liner envelope and the interior surfaces to obtain the temperature response of the interior concrete sections. The heat transfer through the liner wall of the operating region of the containment is established using heat flow devices along with thermocouples that measure the temperature distribution in the cross-section of the wall. The weight of condensate collected from the selected areas of the cylindrical liner walls is also measured. The mass flow rates of steam for the three DBA tests range from 360,000 to 380,000 lb/hr (see Figures 9.C.2-41 a, b, c). The injection steam conditions are near to saturated conditions with an enthalpy of 1196 Btu /lb. The steam is injected in the operating region at 335 ft elevation. The _q duration of the release is approximately 180 seconds. The measurements are taken for (,/ approximately two hours, but only the first 50 minutes are presented in the report. Development of Expected Flow Patterns October 1997 o:\3006w-9c.non:1b 101497 Revision 1
9.Co124 Containment pressure histories for the three DBA tests are presented in Figure 9.C.2-24. The influence of the spray application and the intensity of the sprays on the decrease of the containment pressure is obvious, h The containment vertical temperature profiles for various tests are presented in Figures 9.C.2-43, 9.C.2-44 and 9.C.2-45, respectively. For the first test, large vertical temperature gradients exist for the duration of the measurements (about two hours), indicating temperature stratification. The containment air recirculation system is activated one and half hours after the beginning of the first test and it decreases the vertical temperature gradients (but not so efficiently as water sprays). The spray tests (tests 2 and 3 - which are not applicable for AP600 situations) contribute towards the uniformity of the containment atmosphere by decreasing the dome temperatures. After one hour, the dome temperatures were 177,155, and 137 F for tests 1,2, and 3, respectively. One hundred minutes of test data from thermocouples mounted at the same horizontal level inside the operating region are analyzed. Significant temperature gradients are not established in the horizontal direction. This is consistent with the view that stratification gradients are mainly in a vertical direction, with negligible horizontal gradients, except near plumes and wall boundary layers. Liner vertical temperatures are almost the same as the temperatures in the upper regions of the contaimnent atmosphere, while they are higher than temperatures in the atmosphere of the lower regions (see Figures 9.C.2-46 and 9.C.2-47 for lower regions). An explanation for the higher liner temperatures in the lower regions is that the steam, in its downward movement, passed through the annular gap near the containment wall (surrounding the operating floor) releasing heat predominantly to the wall instead of the atmosphere. Data from six thermocouples embedded in concrete surfaces in the containment are presented in Figure 9.C.2-48 for the case without sprays. The operating region structures have a large temperature increase, while the basement region is almost unaffected. The results are in agreement with the atmospheric and liner temperature data and show stratification above the break elevation. The results from two thermocouples installed in two partially isolated regions (the reactor header cavity and the fuel transfer canal) are presented in Figure 9.C.2-49. A small temperature increase is seen during the steam injection. After the injection, temperatures decrease very fast, resulting in the very small final temperature increase. The data also indicates that dead-ended , compartments get some steam during the pressure increase. O Developrnent of Expected Flow Patterns October 1997 oA3006w-9c.nomitr101497 Revision 1
9.C-125 The reasons for the established temperature stratification in the case of the CVTR containment test performed without internal sprays are: The position of the release point is above the 335 foot elevation of containment height (see Figure 9.C.2-39). This could be considered as a release in the upper region of the containment (H,/Hi = 0.525). The external walls and the dome of the containment were made of 23-3/4" thick concrete which accumulated a portion of the heat released in the upper regions (operational region i and intermediate region - see Figure 9.C.2-48 for vertical temperature distribution inside , . the lincr). The upper portions of the walls remain hot as containment cools, producing j boundary conditions for a stable stratification. External spray cooling of the dome and walls (upper half exposed to the air) is not applied. The steam condensate is collected in the buckets so that weight of the condensate can be measured. This eliminated hot sump evaporation and heat source effects. The other experiments (BMC and HDR) indicate that even small temperature differences between the sump (as a heat source at the lowest level) and the containment atmosphere generate
- global natural circulation inside the containment.
O 1 d d (
. O Development of Expected Flow Pattems October 1997 c:\3006w-9c.non:1b-101497 Revision 1
9.C-126 O 29'- 3" 1/2" Steel Uner j Bend Une EL-360' 1 il 6f- O " 0D Concrete -- OPERATING REGION, 58'-0" O D 1/4" Steel Liner Ground Level EL- 324' N "
'*E"3lsI
E f 89'- 6" INTERMEDIATE REGION _________________________ Et. 2,3 S ASEMENT REGION' Bosoment Floor EL- 275' u meammmmmam
\ - Foundation Mot W
Figure 9.C.2-39 Carolinas Virginia Tube Reactor (CVTR) containment structure 1 (reprinted from: R.C. Schmitt, G.E. Bingham, J.A. Norberg," Simulated design basis accident tests ; of the Carolinas Virginia Tube Reactor Containment - Final report", Prepared for the U.S. Atomic Energy Commission, Idaho Operations Office, Under Contract No. AT(10-1) 1230) Development of Expected Flow Patterns october 1997 o \3006 <6.norrib-101497 Revnion 1
9.C-127 i i ) < , ,x
' ~ ;k d*
. * ,.Y ' .A
~ h.
f ..
} 'ha g' ' y .m
- .j. / -/1, g
! i *m s tt j %, f[pi .
v
.t. -
c..
-N
- 3 .,. .jm aammet"tr e
/.7 '
h' ' N l
~ '4 .
[h ;. , { l ik ibJ il j hV' _ C' : k
/ ' $. >p n
l k'Iif'Y /
<; - .. T ... q/ )
l j s !Q
^
3 1 m1
~g: / -7 l . } - 9' ,,
l 6 1
.: i l - . . . ,4 1
h r ,, ;j .t 0 :. 553: l Figure 9.C.2-40 CVTR steam line diffuser ! (reprinted from: R.C. Schmitt, G.E. Bingham,J.A. Norberg, " Simulated design basis accident tests j of the Carolina Virginia Tube Reactor Containment - Final report," Prepared for the U.S. Atomic Energy Commission, Idaho Operations Office, Under Contract No. AT(10-1)-1230)
- O Development of Expected Flow Patterns october 1997 o:\3006 c6.non:1b-101497 Revision 1
9.C-128 O
.I 400,000- " V _._._. 1 . _ . . _ - . _ .
1 g
~~j -see w ;r g = - = -._ = , 500,000-i . . . . . . . . . . . . . , . . ~
e temt- nm.t -
. ,i I . - ,- -
i . s- . 900,000 - . lomespy C ^ 9.em F ese %
- ta.ese men si ,empense.temuney-i.
TWW Eawey . L97 a e 9:s Siu emy.as taa.epy . att gemet
. 4 . . . .
I g - 1
.t0 _t0 .0 10 ,_ ,_, .I. .t0 .I Figure 9.C.2-41a Steam injection histories test 1 (reprinted from: R.C. Schnitt, G.E. Bingham, J.A. Norberg," Simulated design basis accident tests of the Caronlinas Virginia Tube Reactor Containment - Final report," Prepared for the U.S.
Atomic Energy Commission, Idaho Operations Office, Under Contract No. AT(10-1)-1230) O . Development of Expected Flow Patterns october 1997 c:\3006-c6.nortit> 101497 Revision 1
9.C 129 O
-Pfl0 g . -. 1 a_ -
av--aa -- */* ,m=
.oom- ) ,
o . ;- n ,, e . . . . . . . . .- . e l I* l . i t n e , ,i j e ri n
. r g ,.o - a ==,, ' ^ see. e.e.eu t m.,y:
i . . n, . c- *= r= in.1 f.orge . 8.0*s . aD,.to. a s.g. tan.py . Isto 0 /t ms,
, , * * , . , , , , . . +- + +
( < e i t i t i I I
'o ao ao ao so eso ao e.e o .so ans +
Figure 9.C.2-41b Steam injection histories test 2 (reprinted from: R.C. Schmitt, G.E. Bingham, J.A. Norberg," Simulated design basis ac:Ident tests of the Carolinas Virginia Tube Reactor Containment - Final report", Prepared for the U.S. Atomic Energy Commission, Idaho Operations Office, Under Contract No. AT(10-1)-1230) ("\
'b Development of Expected now Patterns october 1997 oc\3006<6.non:11.101497 Revision 1
9.C-130 f i > e I I I I 4 l l 480 g g
, 4_ g g== & & ; 4" ** A a _g '200 400.000- = us0 .*
NN ti t
~~~~~o ~w e , , , , -- ~%,4 ,
- 9 e ao0Aco- a 8 1 a e 1 , , e , ,,
3 e w em nas w 5 w% :
. ~ r- - , .0. 0 . . _ . . _ . en has F W Cas.uunse Seem fee bem langy . t 96 a 1 m t , . . 0,' u. $m.
s e, _ .:. 0
. Figure 9.C.2-41c Steam injection histories test 3 (reprinted from: R.C. Schmitt, G.E. Bingham, J.A. Norberg," Simulated design basis accident tests of the Carolinas Virginia Tube Reactor Containment - Final report", Prepared for the U.S. Atomic Energy Commission, Idaho Operations Office, Under Contract No. AT(10-1)-1230) e Development of Expected Flow Patterns October 1997 o:\3006-c6.non:1b-101497 Revision )
_, _ - . _ . . ~ . . , _ . - . . . - - ~ . . . ._. .. . 9.C.131 i i k0 s
~
i
\
j ,, . sv., o.
.\
i - g g\ \ 1 {. I .
,,_ t \\
l.a i
\, \
s.s s N . .n...-. o.<.r . i 5 i
\ \ s. , ,,0. n . w . - **"
31 - sy.y 0 y s, . g, e .n . . s . r..- *'*"*"""* ""'* = = se,., on I i j i I , I . l . l - i O 60 to 30 40 M P f Figure 9.C.2-42 CVTR containment pressure response (Heise gauge) (reprinted from: R.C. Schmitt, G.E. Bingham, J.A. Norberg," Simulated design basis accident tests of the Carolinas Virginia Tube Reactor Containment - Final report", Prepared for the U.S. Atomic Energy Commission, Idaho Operations Office, Under Contract No. AT(10-1)-1230) Development of Expected Flow Patterns october 1997 o;\3006-c6.non:1b-101497 Revision 1
m, 9.C-132 1 I e;
- - - ~ #90 , , , g i g g 9 TC-9 E4 864 M e Tt t 2449 f 9 7 . . . , .L.....
O T4 88 L*370 h - 200 - - E N o
..0 I" * -+ ,
s - il t I l I i t t 0 000 900 200 8000 3500 9000 3n00 40c0 Thee issel 1 Figure 9.C.2-43 Atmospheric vertical temperature profile, test 1 (reprinted from: R.C. Schmitt, G.E. Bingham, J.A. Norberg," Simulated design basis accident tests of the Carolinas Virginia Tube Reactor Containment - Final report", Prepared for the U.S. Atomic Energy Commission, Idaho Operations Office, Under Contract No. AT(10-1)-1230) Development of Expected Flow Patterns October 1997 o:\300e, c6.non:1b-101497 Revision 1 i I
.~ . . . . - - - . . ~ - . . . - . . . . . . . , , - ~ . . . _ . . . - . . - . . _ , . ~ . . , _ - , . _ . _ . ~ , - . _ ~ ~ . . - , - . . . . , _ . , ,
9.C-133 t lO 1 a } i 1
, , , , , , .i Tt e W , 9g., g g , ,,, ,,
i 9 1..? .t -..? h
- 4 9 it et SL
- 8. M j
) , O ft. 54 850 M
- 888 .pret e* gua E
i : -o-
- A .
'M r N 9
I - 47
- loses en laser off f i t I t i i '
, *e=
! . . >=
= is o
} d i ! - Figure 9.C.2 40 Atmospheric vertical temperature profile, test 2 - (reprinted from: R.C. .Hhmitt, G.E. Bingham, J.A. Norberg, " Simulated design basis accident tests - of the Carolinas Virginia Tube Reactor Containment - Final report", Prepared for the U.S. Atomic Energy Commission, Idaho Operations Office, Under Contract No. AT(10-1)-1230) ?.
- Development of Expected Flow Pattems october 1997 t o
- \A06-c6.non:1b-101497 Revision 1 i
l~
9.C 134 1 O
= * ** '
3 5 I I I i g 9 f t. S g4 964 er 5 147 (L '897 9 4 ftlIl (L *.49 ft O 18 30 (L.370 se
, = . . .
k
- o,, %. x -
r +
.i - ; .\
g i i 1 t g g g
* *= == 0= noo tm m. noo . ,= i Figure 9.C.2-45 Atmospheric vertical temperature profile, test 3 (reprinted from: R.C. Schmitt, G.E. Bingham, J.A. Norberg, " Simulated design basis accident tests of the Carolinas Virginia Tube Reactor Containment - Final report", Prepared for the U.S. Atomic Energy Commission, Idaho Operations Office, Under Contract No. AT(10-1)-1230)
O Development of Expected Flow Patterns october 15 o-\h6.non:1b-101497 Fevision 1
,. . - - . , . . ~ _ . , . . - . . - . _ -. _ . . .. .. . - ~.. ~ .. . . .. , , . - 9.C 135 ,
4
- o d
..}
L 4 3 !
- I I I e i iC
- 1essnesempse
- . TC
- I hasame.sie 8L.397 le i .9C*.0 esseneseos teresse EL.3 s. se
- a. ,
sa - o z. . ! t. t i i I, I .- 3 ,
, %w _
' .. .t . . . . . , , taas tosal 4 1 I i. a v 4 Figure 9.C.2-46 Surface and atmosphere temperatures - intermediate region, test 1 I' (reprinted from: R.C. Schmitt, G.E. Bingham, J.A. Norberg, " Simulated design basis accident tests ~ of the Carolinas Virginia Tube Reactor Containment - Final report", Prepared for the U.S. Atomic
- Energy Commission, Idaho Operations Office, Under Contract No. AT(10-1)-1230)
V f
- Development of Expected Flow Pattems oct<her 1997 o:\Y36-c6.noitit>101497 Rmsion 1 4
- 9. Col 36 0
tio , , TC a t-e 1C
- 6 osseman as pawe gg . 3e4i.
.re.n o i, n....
eoo - E 4%y , -+ - L g i I I i
~
1 l 1 Q o = moo . em ooo sooo . . sw ., Figure 9.C.2-47 Surface and atmosphere temperatures - basement region, test 1 (reprinted from: R.C. Schmitt, G.E. Bingham, J.A. Norberg," Simulated design basis accident tests of the Carolinas Virginia Tube Reactor Containment - Final report", Prepared for the U.S. Atomic Energy Commission, Idaho Operations Office, Under Contract No. AT(10-1)-1230) 9 1 Development of Expected Flow Patterns October 1997 ; c:\30%c6.non:1b-101497 Revision 1 '
- . . . - . . . . . - . . . _ - - , . , , _ . - . . . .x.-- ~.~. .- . . ..
9.C-137 lO 1 ) b , 8SO , i a i s i i tte 1asesamt unst 0 it
- 4 Bessanset Piser EL ,379 h h 5 f t
- et tussert teseus t t
- 3toit
# it . S4 comes een IL . 5et et '
} 800 = 0 II
- 03 ""*"'"'8"8bI IL
- O'I 8' -'
D TC a to Opereeeng Fener at .333 ee { [
' 0 ft a se ennemi essermer sneens t L *380 m I
- n. -
.{ ~ ,, m N~c p
g h% C) so - ' ' i . , ,
* *** '*** **oo me ~
tm ,E ,, 4 1- .e. , i f i l Figure 9.C.2-48 Concrete surface temperatures, test 1 (reprinted from: R.C. Schmitt, G.E. Bingham, J.A. Norberg," Simulated design basis accident tests of the Carolinas Virginia Tube Reactor Containment - Final report", Prepared for the U.S. Atomic Energy Commission, Idaho Operations Office, Under Contract No. AT(10-1) 1230)
.(~
i \,) ' Development of Expected Flow Patterns october 1997 , o:\3004c7.non:1b-101497 Revision 1
= . -.
9.C-138 O IM s e i s i I I g TCa tta. 3 TC
- Il tetusassesse flogese (L = $19 fl e TC *et caeeems fasema t 4 3 Se h
. ,6
- 4 7 8tecess Comes tL* +31. h
.. -j x , , . . . . .. .L. ..
i f
.l .a j
t N . - i% ..
. i i=
{l
\ % ,
I l l I i i f I i
. . .u. noo ==o suo sooo iso, emo tiae ises) i Figure 9.C.2-49 Isolated cont &inment region temperatures (reprinted from: R.C. Schmitt, G.E. Bingham, J.A. Norberg," Simulated design basis accident tests !
of the Carolinas Virginia Tube Reactor Containment - Final report", Prepared for the U.S. Atomic , l Energy Commission, Idaho Operations Office, Under Contract No. AT(10-1)-1230) l l Development of Expected Flow Pattems october 1997 l o:\3006-c7 non:1b-101497 Revision 1 l
- 9.C-139 -
- - 9.C.2.4 ~ DESCRIPTION OF THE AVAILABLE HDR DATABASE The HDR containment (Figure 9.C.2 50) is 20 m in diameter and 60_ m high with a total free
- volurne of 11,300 m' 3 It is made of steel and concrete with a total surface area of 30,000 m' available for heat transfer (condensation). The dome volume, including a crane structure, is 8
about 42 percent of the total volume or 5,000 m. The main and spiral staircase form the dominant vertical flow paths connecting the dome with sump. -The containment-can be .
- subdivided into 62 to 72 subcompartments containing pipe', s pumps, and valves etc. The subcompartments are interconnected by numerous openings. The numbers and locations of
- thermocouples and anemometers for the HDR-E test series are presented in Figure 9.C.2-51.-
- i. 9.C.2.4.1 HDR E11 Test Series l Test group E11 (the hydrogen distribution and mixing experiments) is included as a part of the
- HDR-Safety Program Phase III that was conducted in the summer of 1989 (Cron and Schrammel, 1993)._ This test series simulated severe accident scenarios. The tests provide insights on DBA circulation and stratification. A review of the natural circulation flows formed in experiments E11.0 - E11.5 is presented in Wolf et al.,1996. Circulation is indicated by examination of the
- recorded temperatures and velocities.
- Seven of the eight E11 experinients are initiated with small steam releases. The one exception, the E11.5 experiment, is initiated with_a large-break, two-phase LOCA blowdown. The tests jA provide data for the transient and spatial distribution of the gas concentration (hydrogen / helium mixture) under various initial and boundary conditions.
i The following features are considered: l
- The large scale of the' experimental facility and the multi compartment geometry with a
[ 1 relatively large dome High hydrogen release rates typical of severe accidents c Superheated steam injection into the containment - single and mu!tiple steam and 5
-hydrogen injection phases 4 . . Internal concrete and metal structures and surfaces Different axial positions for hydrogen and steam releases The efficiency of hydrogen mitigating features, including venting (for example blockage of one major vertical flow path)
O! Q Development of Expected Flow Patterns october 1997 o:\3006w-9c.non:1b-101497 ~ Revision 1
9.C-140 = The impact of internal and external sprays Sump heat up effects
- Air heater operation effects 9.C.2.4.1.1 Experiment E11.0 Experiment E11.0 is performed with an externa! steam supply and without a gas mixture release.
However, the external steam supply is not able to produce typical severe accident conditbns. Therefot a blowuown of the primary system is used for all other experiments. The results of E11.0 are not discussed here. 9.C.2.4.1.2 Experiment E11.1 Experiment E11.1 includes a high-positioned, small break LOCA, followed by additional steam and gas releases. The effectiveness of the internel spray for decreasing the pressure and temperature inside the containment is also tested. Since this experiment included internal sprays that are not typical of AP600, the results are not discussed here. 9.C.2.4.1.3 Experin:ent E11.2 Experiment E11.2 also includes a high-positioned release point. However, instead of intemal spray as in experiment E11.1, external spray is applied to the dome shell. A detailed presentation of this experiment is given in Wolf, Mun, Floyd,1995. This experiment, together with E11.4, provides results used as benchmark problems to validate numerical modele (Wolf et al.,1994. The verification of the GOTHIC code by Holzbauer and Wolf,1994, is also based on E11.2 results. Figure 9.C.2 32 provides the sequence of the experimental procedures. This includes the duration of the heatup and the blewdown of steam (p=110 bar and T=318*C)in the compartment R1805 (in the main stairway at elevation 21.05-25.3 m, see Figure 9.C.2-51, right side). The consecutive steam and gas (mixture of 85 percent He and 15 percent H 2) injections are also presented in Figure 9.C.2-52. Compartment 1405, in the spiral stairway at elevation 4.5 m (see Figure 9.C.2-52 right side), receives steam at 770 minutes. The change in the containment pressure with time is sho vn in Figure 9.C.2-53. The temperatures, steam concentrations, and He/H2 gas mixture concentrations at various axial positions of the main staircase are presented in Figures 9.C.2-54, 9.C.2-55, and 9.C.2-56, respectively. Temperature differences between the same axial positions of the two vertical flow 3 paths (main anr spiral staircases) are pronounced during the heatup phase. Later, temperature differences decrease and change signs. The crossing temperature histories in Figures 9.C.2 57, 9.C.2-58,9.C.2-59,9.C.2-60 indicate a change in the flow directions. Development of Expected Flow Patterns October 1997 o:\3006w-9c.non:lt>101497 Revision 1
9.C-141 , ~ A steep thermal gradient indicates the presence of the stratification that exists between the lower compartments and the --uainment dome (see Figure 9.C.2-61). Even an additional steam release in the lower portion of the containment (compartment R1405) at 770 minutes is not able to completely eliminate the vertical thermal gradient, although it is greatly reduced. The injection cf steam into the lower level, together with the effects of cooling the dome with the external e ay, causes redistribution of the He/H gas 2 mixture (see Figure 9.C.2-56). The . increase in the gas concentration in the upper part of the dome is a consequence of the steam condensation on *he inside dome surface Later, the decrease in temperatures in the upper dome promote global circulation (causing unstable stratification) and homogenize the dome atmosphere, Even after the external spray is stopped, global circulation affects the lower parts of the containment (see Figures 9.C.2-54,9.C.2 55,9.C.2-56). Overall, the external spray causes 1 a reduction of 0 44 bar in the containment pressure (see Figure 9.C.2 53). 1 The application of external sprays at the dome surface promotes global circulation, homogenizes hydrogen concentrations, and decreases the containment pressure. The water film applied to the external surface of the AP600 dome produces similar effects. 4 ) l O Development of Expected Flow Pattems october 1997 o:\3006w-9c.non:1b 101497 Revision 1 1 -
9.C-142 i l i i e l so.com 7 y , l 9 l u
=5 I hI ,
n I
*e.com [u ,g,i , i 1eO, \
g j .. ! STAIRCASE h w STAIRCASE l f l} , new \ _
=
d
,p g' M _ .7
! nao ) Ass dih , 3 i {
~ a n.ss.
fwh fA . l j C be.E
~
l "" nos. gg p - g .__ j im n um- d .nW _ u _ inn s , w isti ! 4.50m 7 h 7 %- d n u e/ii; l o.. ~ '
' i, 9 j .
4 . i,. i h 0' av '" 180' I i l ! l I i l i l Figure 9.C.2-50 Cross-sections, elevations and numbers of compartments of the HDR containment, left (180*-O' cross-section), right (270*-90' cross-section) (reprinted from: L Wolf, K Mun, J. Floyd,"HDR hydrogen mixing evaluation for containment safety evaluations"- Phase 1, University of Maryland at College Park, Final Report for DOE, Project Order Number: DE-AP07-951D81401, July 1995) l Development of Expected Flow Patterns October 1997 ! o:\3006 c7.non:ltr101497 Revision 1 l I
9.C 143 I
! l l
l O I am sw,,,, y,., g .!._ y I " I ci0430-lb[. cou-
... f %-clot 3e Cro430 me3. cru,N ~_m 'N-cro438 3,
I F**ead" et0432 -- t.. c10431 Cro432 ' cro43) cio?02 - . I 4
.. -cis302 Crs202 s - j'.E-cre302 ctB401 - ., N "L 7 - cTa501 crs401 , "l - us501 CT',1101 % * ' "
CT7701 CF7601 g f
,yl,] :.p-CT5301 cr % pjh .-Cr7701
, cma ~ ct = 3 ? - u => C11101 %. - .l. Cril01 g 6
.-Cr5301 CT4301-
- J.
]g f -CT4002 Cr4301*- -Cr4602 hc , .- p . , -CT3802 l h{_ i ,,,,-Cr3802 l l l M 270' 90' 270' 90' LOCAT10NS OF THF.RMOCOUPLES LOCATIONS OF ANEMOMETEsts Figure 9.C.2 51 Locations and Enumerations of 'ihermocouples (left) and Anemometers (righ0 for ilDR E Test Series (reprinted from: L. Walf, M. Gavrilas, K. Mun," Overview of experimental results for long term, Large-scale natural circulation 9 in LWR-containments after large LOCAs", University of
, Maryland at College Park, Final Report for DOE. Project, Order Numben DE AP07 961D10765", ( July 1996) Development of Expected Flow Pattems october 1997 o.\30%<7.riortit> 101497 Revision 1
9.C.) 44 a,n v; -s O W If W
- r. =,, _ m]
t "T-
=
Heat - up phase R1805
) - fMos' external steam y _
T 225'c u a I.. p
- 9.2 b i' j j }e}
8" ~~
) .
_A and 7 Y Y 1- add. simultaneous j * *
- 4 bl.*wdown (f 25 mm) s T 318 *C l
p = 110 bar 3 Ii I ' 5 I s i s 1 J 1 0 ) g j ] c
?
us.41. i m
'y l ff ' tM t" F II' r 1*3"to if 'N' F 2P l
l t Ciews)I\ 0 690 740 959 1095 lit 5 leap l$36 750 770 97$ 1135 '1200 1511 N""8 g i 'J"' ST'i i S T2 gp ,sys ,sp2 y y
'dy' SP6 YPIO Nb2 'D a,... n.. i .. o.-+.e.
nur 0490 arv-u+ +s .nsr i j swam m#y w6 m
- 2 06 6ss t . ,........[.......J.,.....
til 64750 Redmond esternal etsom me i i , m = 1.19 ks's in RI 801 'g~~~\ ***" ******** ******** OM 74770 H,41s pas mixarre selease into R t DOS mie 0 009 ka's ST2 770 959 Steam retenas esto RI40$ l *<--......- -...... ' ........ une 3 06 ks's j NCl 954973 Neesrulsooedoms a , f 1 SP6 975109$ Eastmal sprey a
- 6 ks's * '** ** "*
SPS 1095-1835 Eaternal errey a
- 8 he's SPIO .835 1185 Existaal optsy a o 10 ks's SPl2 11851200 Enermal spray a
- 12 kg/s NC2 1241HS Nanarel oooldone VEN 1458511 Cantamment ,entag END 1536 End of data acqueaban Figure 9,C.2 52 IIDR E11.2 Sequence of Experimental Test Procedures (reprinted from: L. Wolf, K. Mun, J. Floyd, "IIDR hydrogen mixing evaluation for containment safety evaluations"- Phase 1, University of Maryland at College Park, Final Report for DOE-Pro}ect Order Number: DE AP07-951D81401, July 1995)
O Development of Expected Flow Patterns October 1997 a:\30at><7.nortib-101497 Revision 1
9.C 145 l l l O i nn.: . .i nn um n i..
~
u n n s l l l 1 I
) 1 1
( l
< 1 1 I
I 1: 1 1 8 E 1 . 1 ' m .__J____________,__ _ _ _ _ _ J _ _ _ _ _ _ _ _ _ _ _'p _ .
, i i 1
l 1
- i. , 1 1
1 i ' i 1 L I: i i . . i ______ ________ ____________s__________-_. . 1 l 1 5 - 1 1 8 O - I i I i i i i i i i s,
, i i
i i I. i lih 11 " iW, 'iis ' l iiis ' 0 W.Di M Ii3 9 45. . . 1535 80,0, ,, m , 959 o775, i. 5 .1200 nbi- sr(24% A2NC1 \J. 3 og hioNi2\ 4 h NC2 f Figure 9,C.2 53 HDR E11.2 Containment pressure history (reprinted from: L. Wolf, K. Mun, J. Floyd, "HDR hydrogen mixing evaluation for containment safety evaluations"- Phase 1, University of Maryland at College Park, Final Report for DOE-Project Order Number: DE AP07 951D61401, July 1995) v Development of Expected flow Patterns o ic6er1997 c:\3006 :7.non:1t>101497 - Revision 1
9.C 146 O
.p Ak -
1 g-) 5* '2
- (IL3 El 4N Mut M N t.m e it8.3 052 Ret can n LW ema - _ l -ema. )
I(1L3 etam MenInn a t.m 3 L3 essen Men taa M La m- g"' ;7 _ cme 3till et4m Mensaa n La e to J trust nos tan n La "[ ~ a r l
.. s ,
i ,a r .:: i
#, k:
I e, ul. .____ __ _____.,_______ '_ ' ___ ____ l 1 H = + 31 m .4 l 8 1 H=+46m
= i 3 H = + 40 m 4 ,
4H=+sm , Q ______,5H=+em__, _ _;{__ 4 s H =- 4 m IL , g i n._ .s A . L ) : Ps" Q R . is n ' "i i n i* n [ "P %)m , "?5, ',88,5)2m , sup I,",' sa utn. sr(w% sh NCI \sk S1bio %3\NC3v2s'mo Figure 9.C.2 54 IIDR E11.2 Atmospheric temperatures along main stairway (reprinted from: L. Wolf, K. Mun, J. Floyd, "IIDR hydrogen mixing evaluation for containment safety evaluations"- Phase 1, University of Maryland at College Park, Final Report for DOE-Project Order Numben DE AP07 951D61401, July 1995) O Development of Expected Flow Patterns october 1997 c:\3006 c7.non:1b 101497 Revision 1
9.C 147 i i e 4 1 i i s u . stem aas san a to esiu sism aan ma e is ' iu .mun na ssa = tm : eu .ms w ass ma e ta l s:ieu .amaamma = tm : .u === ass ea = La E11 :::"\ A. , 1 H = + 31 m !- Tl 2=q
?1 :______g#q.
H . + 4. . - m ;g 9 Mi g ______ qg=+434..g s u.+ e. _r - - 4.,48 n g s H =- 4 m .= 1 - 1 ,
,e ,
gh l1
.. .- _:._______________'___.L_ ___ . __ ! l lR ...) . __.______i . . . _ ..__.7,_ .. ).,d". '
g q ... . . . . . . . . 4 lsi ll' i lH l 11 I "P T,$5. "'J", '" '" . **f.IP' vis, er(w% A hcl \ A 31hio%:\ NC2 vbeh i 1 1 1 f Figure 9.C.2 55 HDR E11.2 Steam concentrations along main stairway (reprinted from: L. Wolf, K. Mun, J. Floyd, "IIDR hydrogen mixing evaluation for containment safety evaluations"- Phase 1, University of Maryland at College Park, Final Report for DOE-Project Order Number: DE AP07-951D81401, July 1995) O b i october 1997 Development of Expected Flow Patterns o:\3006<7.non:1b 101497 Revision 1 4
9.C 148 O
- = t. niu e,ai nse <= = ts e sia m o. ame rm = to saiu .ananno= = to g i.o
- nsi.enmnesem i s mim nas em = to e aiu === nei cm a te E11e- - g r<L **
e p .. s l f i jo y .gf [ .____ _ _ _ _ _ _ _ _ _ . . . s _ . _______. -r i T a %a.
~ =
f f =*
=$
g 1 H = + 31 m
.. _ _ _ _ /_i 4'
is
- Q ~ =8 i
g 2 H = + 40 m ! .
.. A. 3 H = + 40 m t. .
l 4H=+8ml Ik gw a 8H=+em 6H -4m __...=-_-___ ___... i l
. . . 7_ .
4 I 4 . .. 1 e i_ . M ,s s M i t i
; e, ii ' g ? "7" ","m."?5,$"'$",ia[i"ei . , P)* W irbe n(ua% sh hcl \sh 51hio%:h4wo Figure 9 C.2-56 liDR E11.211/ He Gas concentrations along main stairway (reprinted from: L. Wolf, K. Mun, J. Floyd,"IIDR hydrogen mixing evaluation for containment safety evaluations"- Phase 1, University of Maryland at College Park, Final Report for DOE-Pro}ect Order Number: DE-AP07-951D81401, July 1995)
O Development of Expected Flow Patterns cktater 1997 c:\3006<7.nortib-101497 Revision 1
__ - . _ _ . _ . . = . . _ . _ _ _ _ _ _ _ . -_ ___. _.--_ _m . _ _ _ . _ . . - - _ . . _ _ _ _ _ _ . _ _ . . - _ _ - . - _ _ 9.C.) 49 l i su. o or Me tan n ie ' g a i (11.3 tii.: CIgnt MW tan so m nastan n IIS n iis (\} ! l a i i > >l i I l !. l l 3 . . 4. I l a l i M I I I uy _ _i_ _ _ .
- :. t _a - - - 4 - -
l ls l d _ .! l4- - l- %' I I l I .I l j!; g._ _i ;
- ~~
T *"* *"' ~ 7 *'* ~ ~f~ J ~ ~ ~ 7 ~ 1 I
.. F.1/ i .
DNs i I l I 1 l-- yl-i----p---} w _-l1 , 3 l i
.I i i I
l l I
.,. ,/ - i i
5 J .- e--- ---g--- p-- . l .,- i
- I I l j/ l I I l
! E i i i i i I i I NNN l 10b 113 0 ap0 770 939 975 1133 1200
- I t I > II ( u s t. ou s
( Ith S[740' M b kNCI Sk3g 3 P10 12k NC2
. = _ . . -
en7
* == = , - . . * .== ,,
1
.. *?E ? . :;;- ,
3 C '; e= s _~ ,_
==, F g = . . 1.:.::.. . e U2 I I'*"" -llllll n s "!5 . ~1 } + ~2 ' 55 5 25 R ll .
D - _ -
". ) ."I_ =====--- = $$ $ 'f,"l 4
- RE
.. a,,=,,; =
Figure 9.C.2 57 IIDR E11.2 Comparison between temperatures in main and spiral stairways at +31.0 m level (reprinted from: L. Wolf, M. Gavrilas, K. Mun," Overview of experimental results for long term, Large}}