ML20128C063
| ML20128C063 | |
| Person / Time | |
|---|---|
| Site: | 05200001 |
| Issue date: | 01/29/1993 |
| From: | GENERAL ELECTRIC CO. |
| To: | |
| Shared Package | |
| ML20128C059 | List: |
| References | |
| NUDOCS 9302030298 | |
| Download: ML20128C063 (331) | |
Text
{{#Wiki_filter:ABWR mm3 Standard Plant p ,, 4 SECTION 19E.2 CONTENTS Section Title Page 19 E.2.1 Methods and Assumotions 19E.21 19 E.2.1.1 Code Description 19E.2.1 19E.2.1.1.1 MAAP3.08 19E.21 19E.2.1.1.2 ABWR Modifications 19E.2 2 19 E.2.1.2 ABWR Configuration Basis 19E.2 3 19E.2.1.2.1 ABWR Configuration Assumptions 19E.2 3 19E.2.1.2.2 Station Blackout Performance 19E.2 4 DE.2.1.2.2.1 Summary 19E.2-4 19E.2.1.2.2.2 Core Cooling 19E.2 4 19E.2.1.2.23 f rimary Containment Vessel (PVC)latepity 19E.2 6 19E.2.1.2.2.4 Operator Actions 19E.2 6 5 19E.2.1.2.2.5 Recovery Following Restoration of Power 19E.2 6 19E.2.1.2.2.6 Conclusions 19E.2 7 19E.2.13 Phenomenological Assumptions 19E.2 7 19E.2.13.1 Steam Explosions 19E.2 7 19E.2.13.2 Depee of Metal Water Reaction 19E.2 7 19E.2.133 Suppression Pool Bypass Due to Additional Failures 19E.2 7 19E.2.13.4 Effeet of RHR Heat Exchanger Failure in a Seismic 19E.2 7 Event 19E.2.13.5 Radiation Heating of the Equipment Tunnel 19E.2 7.1 19E.2.13.6 Basemat Penetration 19E.2 8 19E.2.13.7 Hydrogen Burning and Explosions 19E.2-8 19E.2.13.8 Mode of Vessel Failure 19E.2 8 l 19E.2 ii [v) Amendment 9302030290 930129 PDR ADOCK 05200001 A PDR -
l 8 ABWR maims Standard Plant Rev A l SECTION 19E.2 (Continued) l CONTENTS Ses11An Iltle East 19E.2.1.4 Definition of Base Case Anumptions 19E.2? 19E.2.1.4.1 Core Melt Progreulon and Ilydrogen Generation 19E.2 8.1 19E.2.1.4.2 In.Venel Reconry , 19E.2 8.1 _ 19E.2.1.4.3 System Recovery After Veuel Failure 19E.2 8.1 And Normal Containment 1.cakage ] 10E.2.1.4.4 Early Drywell Head Failure 19E.2 9 4 19E.2.1.4.5 Consequences of Supptession Pool Dralo 19E.2-9 : 19E.2.1.4.6 Deleted 19E.2 9 19E.2.1.4.7 Deleted 19E.2 9 19E.2.1.4.8 DeleIed 19E.2 9 19E.2.1.5 Resolution of Phenomenological Uncertainties 19E.2 9 O 19E.2.1.5.1 Identification and Screening of Phenomenologicalissues 19E.2 9 19E.2.1.5.2 Sensitidly Studics 19E.2 9 19E.2.1.5.2.1 Core Melt Progrenion and Hydrogen Generation 19E.2 9 19E.2.1.5.2.2 Fission Product Release from the Core 19E.2 9.1 i 19E.2.1.5.2.3 Cs! Revaporization 19E.2 9.1 19E.2.1.5.2.4 Time of Vessel Failure 19E.2 9.1 - 19E.2.1.5.2.5 Recriticality During In Veuel Recovery 19E.2 9.2 19E.2.1J.2.6 Debris Entrainment and Direct Containment Heatlag 19E.2 9.2 19E.2.1.5.2.7 Fue! Coolant Interaetion 19E.2 9.2 19E.2.lil Amendment
- l. ..
1 ABWR mum , StaIndard Plarti a. u ; SECTION 19E.2 (Continued) ;. O coureurs . so.E - e. 19E.2.1.$.2.8 Cote Concrete Interaction and Debris Coolability 19E.2 9.2 19E.2.1.5.2.9 Fiulon Product Relene location 19E.2 9.2 j 19E.2.1.$.2.10 Fiulon Product Release Flow Area 19E.2 9.2 19E.2.1.5.2.11 Supptchslon Pool Bypass 19E.2 9.2 19E.2.1.5.2.12 High Temperature Failure of the Drywell 19E.2 93 19E.2.1.$.2.13 Suppression Pool Decontamination Factor 19E.2 93 19E.2.1.6 Uncertalnty Analyses 19E.2 93 19E.2.1.6.1 Direct Containment Heating 19E.2 93 19E.2.1.6.2 Core Concrete interaction 19E.2 93 19E.2.1.63 Pool Bypau 19E.2 9.4 19E.2.2 Accident Sequences 19E.210 19E.2.2.1 Loss of All Core Cooling With Vessel Failure At 19E.212 14w Preuure (LCLP) 19E.2.2.2 Lou of All Core Cooling with Vessel Failure At 19E.214 High Preuure (LCitP) 19E.2.23 Station Blackout with RCIC 19E.216 19E.2.2.4 loss of Containment Heat Removal 19E.217 19E.2.2.5 Large LOCA with Failure of All Core Cooling 19E.218 19E.2.2.5 Concurrent Lou of All Core Cooling and ANS with 19E.219 Veuel Failure at Low Preuure 19E.2.2.7 Concurrent leu of All Core Cooling and ATWS with 19E.2 20 Venel Failure at High Preuure
+
19E.2.2.8 Concurrent Station Blackout with ANS 19E.2 21 19E.2.2.9 Summary 19E.2 21 19E.2 iii.1 Amendment l s
- l. - , _ _ . _.,____.-.......-_.-_-_.._,_...,....m. . . . . - . - . . . . . . , . . . . . _ . _ - _ , . ~ - - . - _ - - -- . .----. - .
1 i- AB%R atms Standard Plant Rw A
' SECTION 19E.2 (Continued) i 4
CONTENTS l l Section g g I 19EJJ Justincation of Phenomenotanical Assumntions gpE.2 22 2 19E.2.3.1 Steam Explosions 19E.2 22 i f 1 b 4 19E.2111.2 4 Amendment l:: t v e --. . . . .,., ,-r .v.,-c,.-r4.rm_,. . . . - , ..-..r, ,m, , , _ ,. . . . . . , , . . . ._,-,m,.E,.,_.-w_ ,' ._
..,J-,_,,.v,m,,-,,---._ . , - . .. . . , _ . ,'
ABWR msi=^s Standard Plant Rn A SECTION 19E.2(Continued) CONTENTS Section Title Page 19E.23.1.1 The Steam Explosion Process 19E.2 22 19E.2.3.1.2 Predous Studics 193.2 22 19E.23.13 Thentetical Considerations , 19E.2 23 l 19E.23.2 100% Metal Water Reaction 19E.2 27 19E.233 Suppreuion Pool Bypus Paths 19E.2 28 19E.2.3.3.1 Introduction 19E.2 28 19E.23.3.2 Identification and Description of 19E.2 30 Suppreulon Pool B) pass Pathways 19E.2.333 Evaluation of B) pass Probability 19E.2 31 19E.233.4 Suppreulon Pool Dypus Resulting from 19E.244 External Event Analysis 19E.23.4 Effect of RHR Heat Excharger Failure in a 9E.2 35 O 19E.23.4.1 Seismic Event RHR Equipment Room Flooding 19E.2 35 19E.23.4.2 Dynamic Loads Induced by Chugging 19E.2 35 19E.23.43 RHR Equipment Room StructuralIntegrity 19E.2.M 19E.23.5 Potential for Fluhing During Venting 19E.2.%.1 19E.23.5.1 Critical Time Constants for Blowdown Response 19E.2.%.1 19E.23.5.2 Pool Swe11 19E.2.M.1 19E.23.5.2.1 Pool Swell Due to Suppreuion Pool Mashing 19E.2.M.1 19E.23.5.2.2 Pool Swell Due to Row From Dr>well 19E.2.M.2 19E.23.5.23 Steam Source 19E.2.M.2 19E.23.5.2.4 Application to ABWR 19E.2.%3 Carryover Due to Entrainment 19E.2 %3 ( 19E.23.53 t 19E.2.iv Amendment 1
- - - , -,--r-,- , < . , . . - - , , , , . , ~ , , - - , . - , , - - ..n,. ,-,,,.-~.,.4e , . , -,.e,~.. ~ ,. - - , . - , - . + , , , , ,,w,.,. ,,.,n-.--- ,-m., , . , , , ,
ABWR m.m. Standard Plant kW A , SECTION 19E.2(Continued) O coureurs Secilon Title Page 19E.2.4 Analnin for Reco,erv Seanences 19E.2 37 19 E.2.4.1 Time of Drywell Spray Initiation 19E.2 37 19E.2.4.2 in Veuel Recovery 19E.2 38 19E.2.43 System Recovery after Venel Fallute and Normal 19E.2 38 Containment 1/ .Jage 19E.2.4.4 Early Dr>well Heud Failute 19E.2 39 19E.2.4.$ Suppreulon Pool Drain 19E.2-39 19E.23 Idtat[fication and Screeninn of Phenomeno. logical Ianues 19E.2 42.2 19E.2.5.1 Review of NUREO/CR.4551 Grand Gulf and Peach Bottom Analysis 19E.2 42.2 19 E.2.5.1.1 Orand Gutf 19E.2 42.2 19E.2.5.1.2 Pcach Bottom 19E.2-423 19E.2.5.13 Application of NUREO/CR 4551 Results to ABWR 19E.2 42.3 19E.2.5.2 - Review of NUREO.1335 19E.2 423 19E.2.53 Review'of Recommended SensitMry Analyses for an Individual Plant Examination Using MAAP 3.0B (EPR!) 19E.2 42.4 19E.2.5.4 Review of ALWR Requiremr.nts Document 19E.2 42.4 19E.2.5.5 Summary and Conclusions 19E.2 42.4 19E.2.6 sensitMtv A==Inis and Set plag Studies for Phenouemological Innuis 19E.2 42.5 19E.2.6.1 Core Melt Progreulon and llydtogen Generation 19E.2 42.5 19E.2.6.2 Fiulon Product Release From Core 19E.2 42.6 19E.2.63 Csi Re evaporation . 19E.2-42.7 19E.2.iv.1 O Amendment m ...y._e ._
, . . - . , , , , , , . _ . . . _ . . -m~ , . . . - . . . .n , . . _ , , _ - . . .__._,...,,y.-. .~% ,_ .
ABWR m ai m s Sla_ndard Plant bA SECTION 19E.2(Continued) O courEurs Section Title Page 19El.6.4 Time of Vessel Failure 19E.2 42.7 19E.2.6.5 Recriticality During in Vessel Recovery 19E.2-42.8 , 19E.2.6.5.1 Potentialior Reeriticality , 19E.2-42.9 19E.2.612 implications of Recriticality 19E.2 42.10 19E.2.6.53 . Conclusions 19E.2 42.11 19 E.2.6.6 Debris Entrainment and Direct Contain-ment lieating 19E.242.11 19E.2.6.7 Fuel Coolant lateractions 19E.2 42.11 19E.2.6.8 Core Concrete lateraction and Debris Coolability 19E.242.12 19E.2.6.9 Fission Product Release location 19E.2 42.12 19E.2.6.10 Fission Product release flow Area 19E.2 42.12 19E.2.6.11 Supptession Pool 1)ypass 19E.2 42.13 19E.2.6.12 High Temperature Failure of Drywell 19E.2 42.13 19E.2.6.13 Suppression Pool Decontamination Factor 19E.2 42.14 19EJ.7 Detailed Phenomenological Uncertainty Studies 19EJ42.15 19E.2.7.1 Direct Containment Heating 19E.2 42.15 19EJ.7.2 Debris Coolability 19E.2 42.15 19E.2.7.3 Suppression Pool Bypass 19E.2-42.16 19E.2.8 Severe Accident Deslan Feature Considerations 19E.2 42.17 19E.2.8.1 Containment Overpreuure Protection System 19E.2 42.17 19E.2.8.1.1 Pressure Setpoint Determination 19E3-42.17 19E.2.8.1.2 Variability in Rupture Disk Setpoint 19E.2 42.18 l 19E.2 iv.2 O Amendment l i
<---o- , , , - _ . . _ , , _ . , . _ . . . . _ - _ - _ - . . ~ , . - , , _ , . - , -. . - . . . . - _ - ,r,. ,. e.-~.. .-4
ABM usu 4, ! t Standard Plant n,, 3 i 1 1 l SECTION 19E.2(Continued) 1 i . CONTENTS 1 i Section Title Page l.; 19E.2.8.1.3 Sizing of Rupture Disk . 19E.2 42.18
, 19E.2.8.1.4 Comparison of ABWR Performance With <
and Without COPS 19E.2 42.19 19E.2.8.1.5 Suppression Pool Bypass 19E.2 42.19 19E.2.8.1.6 Summary 19E.2 42.20 i 19E.2.8.2 tower Drywell Mooder 19E.2 42.20 - i i l 19E.2.8.2.1 Introduction 19E.2-42.20 19E.2.8.2.2 Minimum Acceptance Mow Rate 19E.2 42.21 } l 19E.2.8.2.3 Expected Mooder Mow Rate 19E.2 42.21 19E.2.8.2.4 Time to Fill Lower Drywell 19E.2 42.22 19E.2.8.2.5 Consequences of One Mooder Line Opening First 19E.2 42.22 4 19E.2.8.2.6 Valve Opcning Time 19E.2 42.23 19E.2.8.2.7 Estimate of Net Risk 19E.2 42.23 19E.2.8.2.8 Summary 19E.2 42.24 19E.2.8.3 - - Corium Shield 19E.2 42.24 19E.2.9 References 19E.2-43
- 19E.2 v -
Arnendment
,_.~....,,,,.w.. , _ , , , _ , _ , _ _ , , , _ ,_ _ _ ,
-- . - - . . - . - - - - - , ,n-, , __ __
ABWR zmmri Standard Plant PM A SECTION 19E.2(Continued) l j TABLES Table T.itle East - 19E.21 Potential Suppression Pool Bypass Lines 19E.2 44 i I 19E.2 2 ABWR Plant Ability to Cope with Station 19E.2 47 Blackout for up to 8 Hours l l 19E.2 3 Definition of Accident Sequences Codes 19E.2 48 h i 19E.2-4 Grouping of Accident Classes into Base Sequences 19E.2 50 19E/44 Sequence of Events for LCLP PF D H 19E.2 51 l I 19E.2 6 Sequence of Events for LCLP FS R N 19E.2 $2 ! i 19E.2 7 Sguence of Events for LCHP PS R N 19E.2 53 ' 19E.2 54 19E.2 8 Sequence of Events for LCHP PF R M a 19E.2 9 Sequence of Events for SBRC FA R O 19E.2 55 19E.210 Sequence of Events for SBRC PF R N 19E.2 56 19E.211 Sequence of Events for LHRC OO R O 19E.2 57 19E.212 Sequence of Events for LBLC PF R N 19E.2 58 19E.213 Sequence of Events for NSCL PF R N 19E.2 59 19E.214 Sequence of Events for NSCH PF P M 19E.240 19E.215 Sequence of Events for NSRC PF D M 19E.241 19E.216 Summary of Critical Parameters for Severe 19E.242 Accident Sequence i__ 19E.217 Important Parameters for Steam Explosion 19E.243
. Analysis ,
19E.218 Potential Bypass Pathway Matrix 19E.244 , 19E.219 Flow Split Fractions 19E.245 19E.2 20 Failure Probabilities 19E 2-66 19E.2 21 Summary of Bypass Probabilities 19E.247 19E.2 vi - . Amendment
.,_._.._,.......__~,_.,,,-.,.,y.,,,,,_.-.c_,._..m,--.m.,_ ,,,-.m,.._m,-,--m ,v,.-m,,--,y~
ABWR m ai m s Standard Plant n,t A SECTION 19E.2(Continued) O r^11'es cco ii##ea) Table Iltle P.agt 19E.2 22 NUREO/CR-4551 Grand Gulf APET Events By Category 19E.2 68.1 19E.2 23 NRC Identified Parameters for Sensithity Study From NUREG 1335 , 19E.2-68.6 19E.2 24 Issues To Be lovestigated in ABWR l Sensithity Analysis 19E.2 68.7 l 19E.2 25 Comparison of Volatile Fission Product Relene 19E.2-68.8 19E.2 26 Comparison of Low Pressure Core Melt Performance With and Without Containment Overpressure Protection System 19E.2-68.9 19E.2 27 Probability of Release Mode With and Without COPS 19E.2 68.10 19E.2 28 Sensithity Studies for Paulve Mooder Reliability Frequencies of important CET Results 19E.2 68.11 l !O i i 19E.2 vi.1 'O Amendment r ,,,...--r ..-.,~.,---e--n,.-- . - , , . , -. . , , , - - , . . , . , , . . - , . . , . , , , , . . ,,,,,.,,,,,-.,,..,--,,,,,,,.cna, ,_m,,.-.n.~,n,.we v ..,,n-.m,n.,, e
< ABWR mmu a,a Standard Plant I 4 SECTION 19E.2 (Continued i O ILLUSTRATIONS i i i EI M t U1.11 f.ASC l 19E.21 Simplified Sketch of NySupplies to Safety 19E.2 68.12 i Orkde ADS Valves i 19E.2 2A LCLP PF R N (Vessel Preuure) 19E.2-69 t 19E.2 2B LCLP PF R N(UO Temperaturc) 19E.249 l 2 19E.2 2C LCLP PF.R N (das Temperature) ISE.2 73 19E.2 2D LCLP.PF R N(UO Temperature) 19E.2 70 ; 2 19E.2 2E ! CLP PF R N (Venel Preuvre) - 19E.2 71 19E.2 2F LCLP PF R N (Mau of Non Condensables) 19E.2 71 19E.2 20 LCLP PF R N (Noble Oases) - 19E.2 72 19E.2 2H LCLP PF.R N (Volatile Fluion Products) 19E.2 72 19E.2 3A LCLP FS R N (Drywell Preuure) 19E.2 73 0 19E.2 3B LCLP FS R N(GasTemperature) 19E.2 73 19E.2 3C LCLP FS R N (Water Mus) 19E.2 74 . 19E.2 3D LCLP FS-R N (Noble Gas) 19E.2 74 19E.2-3E LCLP FS R N (Volatile Fission Product) 19E.2 75 19E.2 4A LCHP PS R N (Veuel Preuure) 19E.2 76 l I i 19E.2 48 LCHP PS R N (Drywellire! .urc) 19E.2 76 , 19E.2 4C LCHP PS R N(UO Temperature) 19E.2 77 2 19E.2 4D LCHP PS R N (Gas Temperature) 19E.2 7/ i 19E.2 4E LCHP PS R N (Voy Mau) 19E.2 78 19E.2 4F ' LCHP PS R N (Water Mau) 19E.2 78 l 19E.2 40 LCHP PS R N (GlobalMau) 19E.2 79 I
-- 19E 2 vil l O
l Amtadmeet i 1 4 x 1.. _ . .. .- - . , ,....--..-y--,,e. ..- ..
.w__,..-_,-.,,-,..~.r,. , . , , , - . _ , , . . . - , , , , . . , _ , . , . . .,w,,m..-,r . ~v. ,.m.,#--w-nw, ..w,-..w,m
ABWR msms Standard Plant MA l SECTION 19E.2 (Continued O ILLUSTRATIONS
- Elmire Illie East 19E 2 4H LCHP PS R.N (Noble Oases) 19E 2 79 .
! 19E.2 41 1.CHP.PS R.N (Volatile Finion Product) 19E.2 60 4
l 19E.2.$A 1,CHP.PF.P.M (Drywell Preuure) 19E.2 81 , < -\ I } 19E.2 $B LCHP.PF.P M (Gas Temperature) 19E.2 81 J 19E.2.$C LCHP.PF.P.M (UO Temperatase) 19E.2 82 l 2 19E.2 82 ! 19E.2 5D LCllP.PF.P.M (Water Mass) 19E.2 5E LCHP PF.P.M (Fission Products Release) 19E.2 83 1 19E.2 6A SBRC FA.R.O (Drywell Pressure) 19E.2-84 19E.2 6B SBRC FA.R O (Water Temperaturn) 19E.2 84 , l i 19E.24C SBRC FA.R.O(UO Temperature) 19E.2 fl5 2 19E.2 6D SBRC FA.R O (Vessel Water Height) 19E.2 85 3, - 19E.2-6E SBRC FA.R O (Water Man) 19E.2 86 ; ! 19E.2 7A SBRC.PF.R.N (Vessel Pressure) 19E.2 87 ! 19E.2 7B SBRC.PF.R.N (Upper Drywell Pressure) 19E.2-87 4 . ~ 19E.2 7C SBRC PF.R N (Gas Temperature) 19E.2-88 f 19E.2 7D SBRC PF.R.N(UO Temperature) 19E.2-88 2 19E.2 7E SBRC PF.R.N (Water Mast) 19E.2 89 19E.2 7F SBRC PF R N (Volatile Fission Product Release) 19E.2 89 19E.2-8A LHRC 00 R 0 (Drywell Pressure) 19E.2 90 19E.2-8B LHRC 00 R 0 (Water Temperature) 19E.2 90
- 19E.2 8C LHRC-00-R 0 (Water Mass) 19E.2 91 19E.2. vill O Amendment r, ->r-r- , v. re,-e -.-#s + w.m<.-c.--ee- e--e-- ,,,,,,.-.y.=-w.., - - --,,-w .m_e... ...w,w.y.w...,ew.--e.w-r-. - , . . . . - - + r-- e--m..=..w...w=--.re<=--e--,w- '
ABWR ms Standard Plant %4 SECTION 19E 2 (Continued) ILLUSTRATIONS (Continued) Figure Title East 19E.2 9A LBLC PF R N (Drywell Pressure) 19E.2 92 19E.2 9B LBLC PF R N (Gas Tempercure) 19E.2 92 t 19E.2 9C LBLC-PF.R N (Water Mass) 19E.2-93 19E.2 9D LBLC PF.R N (Volatile Fission Product Release) 29E.2 93 19E.210A NSCL PF R N (Drywell Pressure) 19E.2 94 19E.210B NSCL PF R N(UO Temperaturc) 19E.2 94 2 19E.210C NSCL PF R N (Water Mass) 19E.2 95 19E.210D NSCL PF-R N (Volatile Fission Product Release) 19E.2 95 19E.211A NSCH PF.P-M (Drywell Pressurc) 19E.2 96 19E.211B NSCH PF P M (Gas Temperature) 1LE.2 % 19E.211C NSCH PF P M (U 2Dioxide Mass) 19E.2 97 19E.211D NSCH PF P M (Fission Product) 19E.2 97 19E.212A NSRC PF R N (Vessel Pressure) 19E.2 98 19E.212B NSRC PF R N (Drywell Pressure) 19E.2 98 19E.2-12C NSRC PF R N (Power) 19E.2 99 19E.212D NSRC PF-R N (UO Temperature) 19E.2 99 2 WE.212E NSRC PF R N (Water Mass) 19E.2100 19E J W NSRC PF R N (Volatile Fission Product Release) 19E.2100 19E.2 D Steam Explosion Process 19E.2101 19E 214A interfacialInstability 19E.2102 19E.214B Coriumn Stream in Liquid 19E.2-102 19E.215 important Response Times 19E.2103 19E.2-ix O Amendment
ABWR uAnooxs-Standard Plant Rev A SECTION 19E 2 (Continued) ILLUSTRATIONS (Continued) Figure Iltle Eagg , 19E.216 Self Triggering Process 19E.21M
- 19E.2 17 Conditions for Steam Explosion 19E.2105 19E.218 Application to ABWR 19E.2106 19E.219A Suppression Pool r ms Paths and Configurations 19E.2107 19E.219B Suppres!.lon Pool Byg * . Paths ana Configurations 19E.2107 19E.219C Suppression Pool Bypass Paths and Co.a. figurations 19T .2108 19E.219D Suppression Poo' Bypass Paths and Confqmrations 19E.2108 19E.219E Suppression Poal Bypass Paths and Configurations 19E.2-100 19E.219F Suppression Pool Bypass Paths and Configurations 19E.2109 19E.219G Suppression Pool Bypass Paths and Confqrurations 19E.2110 19E.219H Suppression Pool Bypass Paths and Configurations 19E.I $ J 19E.2-191 Suppression Pool Bypass Paths and Configurations 19E.2111 19E.219J Suppreuion Pool Bypass Paths and Configurat4ons 19E.2111 19E.219K Suppreuion Pool Bypass Paths and Configurations 1st.2112 19E.2 20A Small LOCAs Outside Containment 19E.2113 19E.2 20B Medium LOCAs Outside Containment 19E.2-114 19E.2 20C 1.arge LOCAs Outside Containment 19E.2115 19E.2 21 Whole Body Dose at 1/2 Mile as a Probability of Exceedence 19E.2116 j
19E.2-22 Impact of COPS on Risk 19E.2117 19E.2 23 Lower Drywell Flooder System 19E.2118 l 19E.2 24- Flooder Valve Assembly 19E.24119 19E.2-x 0 Amendment
, w, , - - - - - , . - ., s.-- , ._ - -.-.e. , ..
ABWR: mass Standard Plant ne A j 19E.2 - Deterministic. Analyses of their transport to the containment and beyond. , Plant Performence molten core slump into the lower plenum of the
- RPV, vessel failure, corium concrete interactions i 19E.2.1. Methods and Assumptions and further release and transport of fission products.
l h!AAP models all of the engineered safety systems This subse: tion summarizes the methods and as- such as emergency core cooling, automatic
- sumptions that were used in evaluating the Reactor depressurization, safety relief valves, and decay heat i Pressure Vessel (RPV) and containment responses removal, h1AAP also allows the user to model i and determining the resulting source term. The operator behavior and deviations in system
{ operation. 4 l (Reference hiodular Accident
- 1) was theAnalysis Program primary tool used to (h1AAP) l determine the fission product source terms, hiAAP has a modular structure in which separate included in this subsection is a brief description of subroutines are dedicated to modeling specific j the code, the basic assumptions about the ABWR regions and physical phenomena. The main program configuration, a discussion of those phenomena not directs the program execution through r,everal high explicitly modeled in the htAAP analysis, and the level subroutines. The program calls a sepence of definition of the base case, system and region subroutines at each time step.
l These subroutines,in turn, call phenomenology 19E.2.1.1. Code Description subroutines as required. The simulation of an entire accident sequence does not require any user inter. hiAAP was used to determine the vessel and vention during the running of the program. A set of containment responses and the source terms for the built in property-library subroutines provide physical properties. l l ABWR under severe accident conditions. h1AAP-3.0B was modified to model the configuration of the ABWR. An overview of htAAP3.0B is provided (1) High LevelSubroutines below, followed by a discussion of the changes made in the code to model the ADWR, This new version The high level subroutines include the main pro-of the code will be referred to as htAAP3.0B- gram, the input and output subroutines, the data
- ABWR. storage and retrieval subroutines, and the numeri-calintegration subroutines. Also included in the 19E.2.l.l.l. h1AAP3.0B high level subroutines is a controlling routine, BWROP, which allows user interventions that de-htAAP is a computer code developed as a part of scribe the actions occurring during an accident the industry Degraded Core Ru!: making (IDCOR) sequence. The high level subroutines pass global program to investigate the physical phenomena that variables by common blocks (not argument lists) might occur in the ' event of a severe light water and do not contain physical madels for the reactor reactor accident leading to core damage, possible plant. The time integration subroutines,INTGRT reactor pressure vessel (RPV) failure, and possible and DIFFUN control the time steps and call failure of containment integrity and release of fission system and region subroutines at each time step products to the environment. MAAP development during an accident transient, was sponsored by the Atomic Industrial Forum, h1AAP includes models for the important phenom- (2) System and Region Subroutines ena that might occur in a severe light water reactor accident. The system and tegion subroutines include the EVENTS subroutine which sets the event flags
- hlAAP is an integrated code which tracks the - -(Boolean variables) giving the status of the system progression of hypothetical accident sequences from and the status of operator interventions. The a set of initiating events to either a safe, stable and event flags control code execution. Region coolable state or containment structural failure and subroutines, one for each physical region of the fission product release to the environment. hiAAP reactor system, define the differential equations models a wide spectrum of phenomena including for the conservation of internal energy and mass.
steam flashing, water inventory loss, core heatup, Other systems subroutines examine the inter-cladding oxidation and' hydrogen evolution, fission region gas flow rates and calculate the core
] product release from the degraded fuel rods and Amendment 19E.21
L ABWR- nums i Standard Plant w I { temperatures and fuel cladding. coolant interac- routine INITAL and phenomenological i tions. The systems and region subroutines pass' routine M3 POOL. I global variables by common blocks and operate j on them by calling the phenomenology subrou-
~
(b) Lower Drywell: Several alterations were I tines. required in order to model the ABWR lower j drywell, Flow paths were added to model the > i (3) Phenomenology Subroutines sacuum breakers from the wetwell, the vents , to the suppression pool and overflow from the :
- The phenomenology subroutines describe the suppression pool through the wetwell dr>well rates of the physical processes occurring in each connecting vents. Core concrete attack in the l
1 region of the reactor plant model. The phenom- lower drywell region can result in penetration ! enology subroutines pass variables by argument of the pedestal to the wetwell drywell 5 lists, and generally do not use or alter global connecting vents. When penetration occurs,. l variables. The phenomenology subrct' tines are flow between the twer drywell region and the generic in nature and can be called by any of the suppression pool will occur Models for this
- region subroutines or by other phenomenology flow were incorporated which employ a user
! subroutines. supplied concre:e penetration limit.- The ! PEDSTL region subroutine was affected as (l) Property Library Subroutines was the PDFP region fission product i - subroutine. } The property library subroutines give the physical properties (e.g., specific heat and (c) Upper Drywell: This region required the saturation presst're) of the important materials. removal of the flow path which represented
- These subroutines use argument lists to pass the vacuum breaker in the Mark Il model,'and variables and do not have side effects on global the addition of steam and gas venting to the
, variables. Property subroutines are called by the suppression pool via the lower drywell. phenomenology subroutines. Affected subroutines are the DRYWEL region and DWFP fission product region g subroutines. 19E 2.1.1.2 ABWR Modifications Several modifications to the MAAP3 0B code (d) Wetwell: The wetwell fission product trans, were required to adequately model the ABWR The port subroutine WWFP was modified to- . starting point for the modifications was the correctly model the ABWR, MAAP3.0B Mark 11 models. The modified version ' of the code is referred to below as MAAP3.0B- (e) Horizontal Vents: The M3VENTA phenome-ABWR. Specific ABWR features which required nological subroutine model for the horizontal ' code changes are listed below. vents in a Mark Ill were applied to model the horirontal vents connecting the wetwellf-(1) Containment Configuration drywell vents and the wetwellin the ABWR, 3 i The ABWR configuration is different than previ- (2) RHR Heat Exchangers ous BWR configurations. MAAP3.0B ABWR models the flow paths between the containment ABWR has heat exchangers in all three RHR compartments correctly. The high level loops. Previously, heat exchangers were modeled subroutine DIFFP was modified. The affected in only two loops of the RHR system. Addition of regions are: the third beat exchanger required a change in the ECCS system subroutine; (a) Suppression Pool Configuration: The ABWR suppression pool configuration required (3) LOCA Location changes in the models to accurately reflect the relationship between water level and MAAP3.0B ABWR directs the flow from all volume. The ABWR suppression poolis LOCA breaks into the upper drywell. However, modeled by applying the Mark Ill pool since there is a small possibility of LOCAs which a model. The affected subroutines are system Amendment 19E22
_ _ _ ~_ _ _ . - _ _ _ _ __ _ _ . i i ABWR ms Standard Plant p, 4 blowdown into the lower drywell, the (1) Condensate Storage Tank. The configuration l MAAP3.0B ABWR allows the user to input the for the condensate storage tank is assumed to RPV Failure Event Code to simulate this event. be consistent with the description in 19.9.9. This change was accomplished by modifying the This is sufficient to satisfy the station blackout performance requirements discussed in high level subroutine BWROP and the region subroutine EVENTS. Subsection 19E.2.1.2.2. l. I . (4) Recirculation Pump Trip (2) Deleted 4 i in the ABWR, font of the Recirculation Pumps (RIPS) trip on either High Vessel Pressure or on Level 3, with the remaining six RIPS tripping on Level 2. MAAP3.0B ABWR allows the user to (3) Type of Concrete Used for Containment. 4 input these different setpoints. Region Limestone Sand concrete was assumed to be l j subroutine BWRVSL was modified to allow this used for all portions of the containment capability, building except the lower drywell floor. This l assumption will affect the conduction of heat 1 (5) Evaporation from a Pool Surface into the containment walls. However, since concrete has very low thermal diffusivity there {
- The evaporation modelin MAAP3.0B was found will be negligible impact on containment to be_ non conservative for the ABWR, The performance. Limestone Sand concrete is
{ problem arises when pedestal penetration occurs representative of the concrete which might be or the passive flooder operates and water from used in much of the United States, j f the wetwell floods the lower drywell. The vapor ! pressure in the lower drywellis much below the (4) Deleted i saturation point since there was no water in this region while the corium was attacking the i concrete and pressurizing. Therefore, steam will I begin to evaporate off the surface of the poolin the lower drywell. J j In MAAP3.0B the water in the suppression pool (5) Battery loading profiles will be developed to
- had to heat to the boiling point before evapora- define appropriate load shedding during i tion was permitted off the surface of the pool. In Station Blaekout (see Subseetion a
MAAP3.0B ABWR, the vapor pressure is con. 19E.2.1.2.2.2(3) ). This item has been serva;ively assumed to rise to saturation in two ider.tified as a COL Action item in Section l
- time steps. This model was applied to the 19.9.9.
4 wetwell, upper and lower drywells. The PEDSTL, DRYWEL and BWM2WW region (6) RCIC room temperature will not exceed subroutines were affected. equipment design temperature viithout room i cooling for at least 8 hours (See Subsection 19E.2.1.2 ABWR Coanguratlee Basis 19 E.2.1.2.2.2(5) ). This item has been identified as a COL Action item in Section 19E.2.1.2.1 ABWR Configuratloa Assumptions 19.9.9. This subsection provides a description of the (7) Control room temperature will not exceed assumptions which were made about the configura. equipment design temperature for at least 8 tion and systems of the ABWR. These assumptions hours without room cooling (See Subsection l d- were made where the design detail was not yet 19E.2.1.2.2.2(6) ). This item has been available or outside the scope of this submittah for identified as a COL Action item in Section example, the type of concrete to be used in the plant 19.9.9. is not included in this submittal. j 19E2 3 Amendment i i i
ABWR mes Standard Plant Rn ^ (8) Operator action during station blackout is 3) DC battery capacity h. V consistent with the EPGs as specified in Subsection 19E.2.1.2.2.4. 4) Water source inventory (condensate storage tank or suppression pool) (9) Deleted
- 5) RCIC room temperature
- 6) Contr01 room (s) temperature (10) Deleted Each of these functions is addressed below.
- 1) Reactor Monitoring Function.
(11) Deleted The teraor monitoring of vessel water level and pressure is performed using local detectors with control room indication. Instrument power supply (12) Deleted is from the station batteries as either DC or constant voltage constant frequency (CVCF) sources. 19E.2.1.2.2 Station Blackout Performance
- 2) Steam Supply to the RCIC Turbine.
19E.2.1.2.2.1 Summary The reactor vessel is the source of energy for the A station blackout is defined as the loss of offsite RCIC turbine which operates the RCIC pump, electrical power and the unavailability of onsite AC maintaining vessel water level. The RCIC turbine electrical power (i.e., failure of diesel generators, in willisolate (i.e., trip) at low pressure (50 psig). However, since the operator will be maintaining p) t most cases). During this period the important phnt performance characteristics to be considered are maintenance of core cooling and containment vessel pressure near 945 psig in accordance with the emergency procedure guidelines (EPGs), integrity, there will be more than adequate RCIC turbine pressure for operation. The RPV pressure will bc l l The analyses summarized in this subsection show controlled manually at this level (by opening 1 or that the ABWR can withstand a station blackout more SRVs) below the first SRV setpoint to avoid without core damage or loss of containment ittegrity SRV cycling, SRV operability during station for a period of at least 8 hours. If AC power is still blackout is dependent on a DC supply source and unavailable beyond this period, the core cooling a nitrogen supply and these are evaluated in the function is assumed to be lost. This accident following discussions. It should be noted that the sequence is discussed in Subsection 19E.2.2.3. SRVs will cycle on the spring setpoint if the l operator fails to manually control pressure. i The key requirements of core cooling and primary containment vessel (PCV) integrity are a) Availability of DC Power for SRV Solenoids. tn:ated separately below. Based on the following evaluation, it is 19E.2.1.2.2.2 Core Cooling concluded that there is no practical limit on the availability of DC power for operating The reactor core isolation cooling (RCIC) system SRV solenoids. provides water to the reactor vessel dumg a station blackout. The following areas are considered to The control power for six of the 18 SRVs is assure RCIC functionality during station blackout: taken from the Division 1 battery. The valves have been considered as part of the load on
- 1) Reactor monitoring function the Division 1 battery for purposes of calculat-ing the time the RCIC would be operable
- 2) Steam supply to the RCIC turbine during station Mackout. This evaluation leads Amendrnent 19fi.24
ABWR uxu=4s Standard Plant n,4 to the conclusion that the 4000 ampere hour b) SRV Operability and High Preuure Contain-
! capacity of the Division 1 battery is sufficient ment Conditions During Station Blackout.
3
\ for 8 hours of coping during station blackout.
- The SRV actuators can open the SRVs with a of the remaining 12 SRVs,6 have thsir pressure differential of approximately 70 psi control power supply on the Divisions II (nitrogen supply pressure above containment battery and 6 are on the Division 111 battery. pressure) without assistance from internal Each of these batteries have a capacity of steam pressure. The SRV accumulators used j 3000 ampere hours. Since Divisions 11 and 111 for the ADS function (see Figure 19E.212),
; would normally be shut don during a station which are gharged to a pressure of 170 psig blackout situation, these batteries and their (12 kg/cm g), have insufficient pressure and associated power distribution equipment capacity to fully open the SRVs at 100 psig 1 would be available to supply power to the pressure in the containment and RPV, so SRVs if necessary. additional gas at 170 psig is needed from outside the containment to ensure the pres- ; The ambient temperature for Divisions !! and sure control and depreuurization function.
- lit batteries should remain acceptable as i there would be very little load on these The normal supply of N gas to the SRVs batteries during station blackout. For this from the atmospheric con rol system outside j reason, ambient temperature rise due to the the containment is shut off due to low pnessure lack of HVAC should not be a problem for caused by loss of AC power to the heaters or i the batteries and their anociated equipment. heating boiler which is used to gasify the liquid N, supply. However, there is a backup supply of'N., gas from stored bottles at 2130 to 850 3 psigImaximum to minimum) pressure which can be used to open the SRVs in the ADS l
system. Use of thq stored nitrogen bottles requires Based on the above, Divisions 11 and III DC operator action to manually open a closed supplies should be available on an supply valve at the valve location. Gas is then intermittent basis for use in operating SRVs, fed to the SRV actuators through the DC as desired. The 6000 ampere hour total powered ADS solenoid valves inside the j capacity of the two batteries would be containment automatically. The ADS supply adequate for many days of operation beyond lines from the N, bottles must also be isolated the 8 hour capability of Division I. from the normal'N SUPP yl to other systems by
- 2 local manual closure of the motor operated Further, eight of the 18 SRVs are used for crosstic valves which are otherwise inoperable the ADS function and thus have alternate . on AC power loss.
power sources. Five of the eight can be
, supplied by either of tw dmasons (Didsions The high pressure gcs from the N bottles is I or !!). The other three can be supplied by automatically reduced to 170 psig y a self.
, any of three divisions. Control power for actuated pressure regulating valve. If the
. each of the ten SRVs which are not used for SRVs do not opeo with the pressure supplied -
1 the ADS function is supplied by one division by the self. actuated pressure regulating valve (four from Division I, three from Division II, (for example,if containment pressure were and three from Division 111). Thus the ability above 100 psig or if somewhat less than 170 to control reactor pressure is very reliable. psig were supplied), the operator could adjust the set point of the pressure regulating valve above 170 psig at the local station (the relief valves are set at 210 psig). The capacity of a group of ten 45 liter high ! Amendment :2 19EL2 5 4
i ABWR u^6 mas Standard Plant Pev A pressure N, gas bottles at 850 psig minimum for at least 8 hours. g pressure is about 16 times that needed to open j h the 8 ADS SRVs, each of which has an actuator 6) Control Room Temperatures. V piston volume of 16.4 liters (1000 cubic in). Additionally, there are 10 other N, bottles that The safety related equipment required to function can be valved into service by lo* cal manual during station blackout and located in the main, operation. After the 8 ADS valves are opened lower and computer control rooms will be there is sufficient N, gas to account for at least 7 designed for a maximum operating temperature of days leakage frorii the valve actuators, after 122 F (50 C). The ABWR plant will be designed vfaich the N bottles must be replaced to hold to prevent the room temperature from reaching 2 the ADS valve open. Based on the foregoing, it is this equipment design temperature for at least 8 concluded that the ADS valves can be operated ho to depressurize the reactor on loss of normal AC 79,urs, starting at the normal room temperature of F (26 c) power supplies with the containment at 100 psig. The operator has to manually close and open 19E.2.1.2.2.3 Primary Containment Vessel (PCV) valves at the valve locations to supply nitrogen Integrity from outside the containment to open the 8 SRVs used for the ADS function and to hold Containment pressure and temperature analysis them open when the pressure in the RPV drops were performed to determine the containment to near containment pressure. atmospheric conditions after 8 hours of station blackout conditions assuming event initiation at
- 3) DC Battery Capaci:y. 100% thermal power. An analysis was performed which assumed the RCIC suction was taken from the The DC batteries will be sized to be capable of condensate storage tank for the duration of the operating the RCIC system for a minimum of 8 event. The dryv ell and wetwell pressure and hours assuming the expected loading profiles for temperature were calculated to be less than their station blackout. These loading profiles, design basis of 45 psig and 340 F (drywell)/219 F including load shedding, will be defined in detail (wetwell) after 8 hours. Therefore PCV integrity is i as the ABWR design progresses. maintained.
(G
- 4) Water Source Inventory. 19E.2.1.2.2.4. Operator Actions The primary water source for the RCIC System The loss of normal AC power will lead to indirect during station blackout is the condensate storage turbine trip and reactor scram due to high condenser tank (CST) which has been sized to provide pressure on loss of circulating water. The subse-sufficient inventory for a minimum of 8 hours quent loss of feedwater will cause rh RPV to isolate during this scenario. In the event the CST on low water level. Failure of the snergency diesel became depleted, the backup source is the generators to initiate will leave the RCIC system as l suppression pool. The RCIC system must be the only source of makeup water to the core. The l manually overridden to assure that the conden- RCIC system will automatically restore the RPV wa.
sate storage tank will be maintained as the ter level. Operator action are specified in the EPGs primary water source for core cooling and to control the RCIC system (to avoid repeated makeup. restarting of the RCIC turbine) and maintain the RPV level between Level 3 and Level 8.
- 5) RCIC Room Temperature.
In addition, the operator will be instructed to
- Failure of the AC cooling power supplies will maintain RPV pressure below the high pressure allow the RCIC room temperature to rise. The scram setpoint (below first SRV setpoint) to avoid
, ABWR plant will be designed to prevent the SRV cycling by controlling 1 or more SRVs coom temperature from reaching the equipment manually. The PCV pressure and temperature will l design temperature of 151 F (66 C)d starting at not approach design values for at least 8 hours. l the normal room temperature Of 104 F (20 C.), Failure of the RCIC (core uncovery) will require the j operator to blowdown through the SRVs at the ll i fm steam cooling pressure and thereby avoid a high ( ) Amendment 19IL M
4 I ABWR ms L Standard Plant %4 i pressure core melt. fashion using htAAP. These phenomena fallinto ,g two categories: those which are ruled out as being - j 19E.2.1.2.2.5 Recovery Following Restoration of incredible for the ABWR and others which are. i AC Power neglected because they produce an insignificant. change to the overall performance of the'ABWR 1 All equipment necessary for restoration of power under t.evere accident conditions. A more detailed + is located external to the reactor building secondary explanation of some of these phenomena is given in - { containment. With the exception of the control - Subsection 19E.23. j building and the RCIC room, all heat generating ! sources external to secondary containment are 19E.2.IJ.1 Steam Explosions f shutdown during station blackout so that the rooms i should be at temperatures which allow restart of the Large scale steam explosions are deemed incredi. ! support systems under their automatic or manual ble. The geometry of the ABWR will prevent a suffi. l modes following restoration of power. Temper- ciently large contiguous mass of corium from falling l atures in the control building should be such that into water in either the vessel or lower drywell re-l restart can be accomplished by the operators from gions. A more detailed description of this phenome-l the control room. Also, restart could be initiated non as well as the justification for its neglect is ! from the remote shutdown panel or even by local provided in Subsection 19E.2.3.1. Small steam explo. ] control at the motor control centers and switchgear, sions which do not in themselves threaten the integ. - Following restoration of power and initiation of rity of the vessel or containtnent are calculated by l4 operation of the reactor building closed cooling blAAP. Additionally, a scoping calculation is water system, the ECCS areas of secondary con- performed in Subsection 19E.2.6.7 to determine the tainment will be cooled by their safety grade room mass of core material which could participate in a ! coolers so normal operation of the safe shutdown steam explosion without damaging the containment. ! systems could be restored, The turbine building i electrical systems and the non safety related 19E.2.1J.2 Degree of Metal Water Reaction j secondary cooling system provide a backup means of
- restoring cooling to the ECCS equipment areas - The metal water reaction rate used in the
- within secondary containment, integrated analysis is that calculated by the blAAP models. One limit on the generation of hydrogen
. 19E.2.1.2.2.6 Conclusions occurs when all of the zirconium in the cladding is { assumed to react with steam to form zirconium oxide
- The ABWR plant is being designed to be capable and hydrogen gas, The separate effects calculation j of maintaining core cooling and containment integ- in Subsection 19E.23.2 shows that the containment is i rity for at least 8 hours following the loss of offsite capable of withstanding the static pressure that l and onsite AC electrical power. This capability would be generated were this maximum hydrogen l assessment follows the general criteria of: production to occur, as required by 10 CFR 5034(f).
- 1) Assuming no additional single failures 19E.2.1.33 Suppression Pool Bypass due to Additional Failures
- 2) Realistic analytical methods and procedures This assumption covers one of the potential types j A summary of the key plant parameters, design basis of suppression pool bypass. Subsection 19E.233
! values and capability assessment is shown in Table shows that the totalincreased risk due to suppression 19E.2 2. Note that the response of the ABWR con. pool bypass caused by additional failures is less than tainment to Station Blackout would be successful 10% with the exception of the wetwell/drywell j' even if the design basis values were exceeded, as long vacuum breakers. This is judged to be within the
- as the ultimate capability were not exceeded. uncertainty of the PRA. Therefore, only the failure
.' of the wetwell/drywell vacuum breaker needs to be 19E.2.lJ Phenomenological Assumptions considered explicitly. A sensitivity study was
- performed in Subsection 19E.2.6.11 to examine the This subsection contains a summary of those impact of vacuum breaker leakage and failure on
- phenornena which are not considered in an integral fission product release. Subsection 19 5.2.7.3 I
Amendment 19EJ-7 9
-. . -, _ m - - _ +.
a 23A6100AS - Standard Plant Rev A - lf I present's an uncertalnty analysis which determines the impact of bypass on risk. 19E.2.1.3..l Effect of RHR Heat Euchanger Failure j- in a Seismic Event , 1 ! During a seismic event it is possible for the RHR j' heat exchangers to fail by shear of their anchor bolts. ! ! This could ootentially lead to drainage of the j suppression poolif the RHR suction lines are not- i i isolated, Calculations were performed which show j that the operator has about half an hour to isolate
- the heat exchanger.
l 1- -if the heat exchanger is not isolated then the l RHR pump rooms will be subjected to additional j' loading caused by the static head of the water, and potentially by chugging loads as steam discharges from the broken pipe. It is seen that the RHR pump i room integrity will not be breached by these loads, i l Additional details about the pool drainage and l structural loading may be found in subsection . 19E.23.4, The impact of suppression pool drainage on fission product release, should this event occur is found in subsection 19E.2.4.5. O- 19E.2.1.3.5 Radiation Heating of the Equipment lU Tunnet A potential concern for the ABWR during severe
- accidents is radiation heating of the equipment j tunnel. . After vessel failure, the corium in the lower i
t l '. I n l , l i Amendment 19E.2 7.1 J t-
MIN u AstooAs Standard Plant g,, 3 drywell could radiate energy directly to the walls of is not necessary as discussed in Section 19D.5. the equipment tunnel. This could potentially reduce b the structural material strength, eventually resulting in the tunnel buckling under its own weight. 19E.2.1.3.8 Mode of Vessel In the unlikely event of a core melt sequence with The adoption of the passive flooder (Subsection substantial relocation of debris which leads to vessel 9.5.12) precludes this occurrence since the flooder failure, the vessel failure location is expected to be in opens when the temperature reaches $33 K (500 F), the bottom head. A failure of the RIPS has been Upon opening, water from the suppression pool proposed, however, as discussed below, this is not a would flood the lower drywell, covering the corium, credibic mechanism for the ABWR. Figure SA 2 This stops any radiation heat transfer from the gives a pictorial description of the location of the corium to the tunnel walls. Therefore, no significant RIPS in the RPV. Figure SA 1 shows more RIP material strength reduction of the equipment tunnel detail. caused by increased temperature is possible. Since the core melt progression is expected to 19EJ.l.3.6 Basemat Penetration contain the corium inside the core shroud, debris would not approach the RIP impellers or RPV RIP Basemat penetration by the core debris will not nozzles which are located outside the shroud, lead to contalnment failure. In each of the However, if the shroud is perforated by the corium, sequences considered the debris will be quenched the corium might than enter the top of RIP impellers and cooled before basemat penetration can occur, and possibly enter the stretch tube / shaft annulus. The passive flooder opens when the lower drywell This is extremely unlikely since this annulus temperature reaches $33 K (500 F). Even were this thickness decreases in the downward direction to to fail, when the sideways penetration of the pedestal 1.5mm (the variance between the 215mm diameter walls reaches 8 inches, water from the suppression RIP shaft and the 218mm inside diameter of the pool would flood the lower drywell. stationary stretch tube). Any molten material relocating through the RIP would quickly freeze or The pedestal cavity design meets the 0.2 sq. flow through the pump rather than flowing along the O meters /MWt specification of the EPRI Debris Cootability Requirements for Advanced Light Water pump shaft. y reactors (Reference 2) A conservative analysis was in the event the corium did flow down the stretch ( g
~
performed following the methods of the ARSAP tube / shaft annulus, the mo or housing to RPV I Debris Coo! ability Requirement (Reference 3) and nozzle weld might fail allowing the RIP / motor to S utilizing the concrete ablation rate from CORCON drop. Figure 1.2 3b shows the two RIP vertical D j (Reference 4). Assuming a 10 hour delay in adding restraints which connect the bottom of each RIP water to the drywell, this resulted in an ablation motor housing to the RPV bottom head. These depth of 0.9 m (3 ft) before the corium is completely retraints prevent the RIP / motor from dropping out quenched and cooled by the water from the suppres- of the RPV in case the motor housing weld fails for l sion pool any reason. Therefore,in the exceedingly unlikely event of RIP failure, the pump will not fall from the , Additionally, uncertainty analysis was performed vessel, and the penetration through the vessel would . in Subsection 19E.2.7.2 to assess the patential for be small. continued core-concrete attack. This study concluded that debris cooling is highly probable for the ABWR Nevertheless, the corium is expected to freeze design and that there is little impact of contained and, consequently, not flow down the annulus into core conerete interaction on coatainment the motor housing. Therefore, the RPV RIP nozzle performance, motor housing reactor coolant pressure boundary would not be breached. Failure of the vesselin the 19E.2.1.3.7 Hydrogen Burning and Explosions lower head region is the expected mechanism for the release of core debris from the vessel. The ABWR containment is inerted. Hydrogen burning and explosions are not possib!c in an inerted 19E.2.1A Definition of Base Case Assumptions containment. An explicit consideration of the short periods of time when the containment is not inerted In the context of this study the phrase " base case" [ s t9E.2 8 Amendment
ABWR u^6im^s Standard Plant Rev A is used to describe those studies which determine the 19E.2.1.4.3 System Recoscry After Vessel Failure O b nominal response of the ABWR to severe accident conditions using best estimate phenomenological and Normal Containment bakage models and no credit for system recovery. Several All of the base analyses assume that any failed accident sequences were considered using the base case assumptions. The effects of the base case assumptions on the results of the analysis are determined by means of sensitivity studies and uncertainty analyses as necessary. 19E.2.1.4.1 Core hielt Progression and Ilydrogen Generation Critical to the melt progression of the fuel is the question of blockage in the core. In the base cases it was assumed that blockage occurs as predicted by MAAP using the default core melt progression input parameters. This decreases the generation of hydrogen in the core, since there will not be steam flow past the hot zirconium during the later stages of the melt process. The effect of this assumption on the overall response of the plant is determined by turning off the core blockage modelin MAAP. This is done with the sensitisity study in Subsection 19E.2.4.1. For this
,, case steam continues flowing past the fuel rods as
! j they melt. The production of hydrogen continues \/ until there is no more water available for reaction. This leads to a somewhat higher partial pressure of hydrogen, and higher containment pressure. 19E.2.1.4.2 In. Vessel Recovery For sequences in which there is no core cooling available at the onset of the accident it may be possible to recover core cooling at some later time. It is important to know the time which allows for in vessel recovery in order to determine the proba-bility of system recovery in the containment event trees, Subsection 19D 5.12. Recovery is of particular interest for the study of Loss of Offsite Power and Station Blackout sequences. The base sequences do not modelin vessel re-covery. This possibility is considered using a sensitivity study. The MAAP code calculates in-vessel recovery only if a core cooling injection source is recovered before channel blockage occurs. However, the effects of in vessel recovery can be simulated by the use of a wetwell failure as discussed in Subsection 19E.2.4.2. Amendment 19 E.2-81
n r ABWR mim ! Standard Plant %.4 j y !- I system will remain inoperable throughout the from the containment will be investigated in detail.
- _ [ duration of the accident. However, in order to Each parameter will be considered indiviJually, !
j determine toe appropriate accident management although interactions between some key phenomena ' -are considered.' strategy,it is necessary to understand the behavior of . the system if a system were to recover. The recovery of any ECC system would be like the use of the - The uncertainty analysis is a four step process. 3 [ firewater system. Only the recovery of the RHR . The first step is a literature survey which identifies system will prevent containment structural failure, if . all severe accident issues. Second, these issues are 4 j structural integrity is maintained, the only fission screened for their applicability to the ABWR. These j product release mechanism is normal containment two steps are combined in this study. Next sensitivity } leakage. This mechanism is discussed in Subsection studies have been performed over a credible range of 5 ) key parameter values to determine the potential for a. A 19E.2.43. j significant impact on fission product release and 19E.2.1.4.4 Early Drywell Head Failure timing. If such impact is demonstrated, then the issue Q' is carried forward into the final step, a detailed One type of loss of containment structural uncertainty analysis.The propagation of uncertainty e integrity in'the containment event trees is early distributions will not be carried out as done in 3 - drywell head failure following a high pressure melt NUREG 1150. ] sequence. The consequences associated with this event are discussed in Subsection 19E.2.4.4. 19E.2.1.5.1 Identification and Screening of Phenomenological Issues 19E.2.1.4.5 Consequences of Suppression Pool Drain The first step in performing an uncertainty analysis is to identify the key phenomena and their in the seismic event trees, a mode of RHR heat associated uncertainties. To do this, GE has surveyed exchanger failure was identified which could the available literattee as discussed in subsection i potentially result in the draining of the suppression 19E.2.5. Some of the severe accident issues are 3 poolinto the RHR pump rooms. An analysis was screened out, as they are not applicable to the , performed to examine the impact of this on pump ABWR design. For example, hydrogen combustion j . room integrity (Subsection 19E.23.4) which showed phenomena are not important in the ABWR since q i that the room would remain intact, the containment is inerted. Issues identified which i
- could have impact on the severe accident j Therefore, the suppression pool may be viewed performance are included in the sensitivity studies H;
as having moved into the pump rooms. The pump which follow. room will have no ability to withstand the increase in a pressure due to decay heat.~ Rather the room will 19E.2.1.5.2 Sensitivity Studies . 4 seak and the pressure will remain near atmospheric .J pressure. Thus, there will be no holdup of noble Sensitivity studies are performed for the ABWR E gasses. However, since all of the fission products will response to severe accident phenomena in order to pass through the poolin the pump room, significant determine those issues which may have significant h, fission product scrubbing of the volatile fission impact on the offsite risk associated with the ABWR ; J products will occur. Subsection 19E.2.4.5 examines design. Given this goal, the ultimate measurement of j the resulting dose from this type of sequence. sensitivity is the offsite dose. At a given site the . primary factors which influence the dose are the- j x 19E.2.1.5 Resolution of Phenomenological magnitude and time of release.Therefore, changes in
< Uncertalaties these parameters will be used to determine the need J for detailed uncertainty analyses, n
.j The ABWR is designed to limit the sensitivity to "
~
various phenomenological unccrtainties. 19E.2.1.5.2.1 Core Melt Progression and Hydrogen 9 Nevertheless, an uncertainty study was performed. Generation h i Severe accident phenomenological uncertainties are d addressed in an engineering sense. This means that Critical to the melt progression of the fuel is the only those parameters that have a major impact on question of blockage in the core. In the base cases it
'] the timing and magnitude of fission product release was assumed that blockage occurs as predicted by Amendment 19E.19
_ ,__m. - .- - -
E l ABWR- 2-s - Standard Plant %4 MAAP using the default core melt progression input containment. Therefore, no further consideration of - j parameters. This decreases the generation of Csl revaporization is needed. 4 hydrogen in the core, since there will not be steam l flow past the hot zirconium during the later stages of 19E.2.1.5.2.4 Time of Vessel Failure - the melt process.
- The detailed melt progression of a severe accident j The effect of this assumption on the overall is subject to considerable uncertainty. The melt i response of the plant is determined by turning off the progression assumed in MAAP retains the molten j core blockage model in MAAP. This is done with the core material above the core plate until a local 1 sensitivity study in Subsection 19E.2.6.1 For this case failure of the core plate occurs which results in a 2
steam continues flowing past the fuel rods as they large pour of core debris into the lower plenum of i melt. The production of hydrogen continues until the vessel. As a result of this model, the lower head j there is no more water available for reaction. This of the vessel fails almost immediately, even though 4 leads to a somewhat higher partial pressure of there is water in the lower plenum at the time, in
- hydrogen, and higher containment pressure. There is ' other melt progression models, the molten fuel drips virtually no impact on source term, and the time of down the fuel rods in a process called candling.
3 j fission product release is not substantially altered. Under this assumption,it is possible for molten j Therefore, it is judged that the ABWR severe corium to be relocated in the lower plenum slowly, i accident performance is not sensitive to in. vessel where it is quenched. This results in a delayed vessel j hydrogen production. failure after the water in the lower plenum has boiled
- off.
I 19E.2.1.5.2.2 Fission l'roduct Release from the Core A sensitivity study was performed to determine The base sequences use the Cubicciotti model for the impact of the time and mode of vessel failure on fission product release from the fuel. If the release containment performance. it was observed that there from the fuel occurs at a different rate, any potential is little impact on the base scenarios. However, it was i release frem the containment could be affected noted that the mode of vessel failure could impact i through the containment residence time and other phenomena such as direct containment heating {g suppression pool scrubbing. The effect of the release and core concrete interaction. Discussion of these-rate on source term is examined in Subsection relationships ay be found in subsections 19E.2.7.1 3
! 19E.2.6.2. The study indicates that there are modest and 19E.2.7.2 respectively.
l differccccs in the loca'tica of the fission products { within the containment. However, because of the 19E.2.1.5.2.5 Recriticality During in Vcssel j depth and subcooling of the suppression pool and the Recovery. i presence of the COPS, there is no appreciable j' variation in the release from the containment. A potential challenge to the containment has been l Therefore, no further investigation of the impact of identified for accidents in which the core melt is
- fission product release from the fuelis required. arrested in the vessel. Experiments have indicated 4 the potential for the boron carbide in the control j 19E.2.1J.2.3 Csl Revaporizatlos blades to form a cutectic with steel at 1500 K and relocate before the fuel relocates. Thus,if core
! An important aspect of fission product behavior cooling is recovered after the control material has . is the propensity of the aerosols to adhere to the relocated, there is a potential for the core to return !- relatively cooler surfaces of the vessel and- to a critical condition.- A sensitivity study was ! containment. Whilc the deposition process is fairly performed in Subsection 19E.2.6.5 to examine the i well understood, there is considerable uncertainty in potential for recriticality and the implications of its
- the revaporization of the fission products, occurrence for the ABWR design. The study i particularly that of Csl. A sensitivity study was concluded that there was a very short time window
] conducted, as reported in Subsection 19E.2.6.3, to during which a return to criticality was possible. examine the impact of delayed revaporization. A Further, even if it should occur, recriticality is not
- I variation of fission product behavior inside the likely to lead to containment failure. Thus, j- containment was observable. However, there is not a recriticality does not pose a significant threat to the l substantial difference in the release fraction from the ABWR design and need nat be considered in an uncertainty analysis.
Amendment 19E.2 9 I 4 i
ABWR ms Standard Plant %4 + 19E.2.1.5.2.6 Debris Entralnment and Direct has a large drywell floor area and redundant systems Containment Heating which can flood the lower drywell Howeser,
} experiments performed to date have been unable to -
If a core melt accident occurs in which the provide conclusive evidence that these features cool reactor pressure vesselis at high pressure at the time the debris sufficiently to prevent core concrete of vessel failure, the debris may be entrained out of interaction. Therefore, uncertainty analysis was the lower drywell. If the debris is finely fragmented performed as discussed in Subsection 19E.2.6.2. I as it is dispersed, tne pressure in the containment can rise rapidly. This process is known as direct 19E.2.1.5.2.9 Fission Product Release Location - containment heating. If the magnitude of the pressure rise is high enough, the containment may be The adoption of the containment overpressure challenged. This would lead to an early failure of the protection system (COPS) in the ABWR containment structure and large releases of fission containment design serves to significantly reduce the products. Therefore, uncertainty analysis was uncertainties in the timing, location and area of any performed. The conclusions of this study are given in fission product release. The setpoint of the rupture Subsection 19E.2.1.6.1. disk was selected such that there is a small probabili'y of containment failure before the rupture 19E.2.1.5.2.7 Fuel Coolant interactions disk opens. The probabilities for containment failure depend on the accident progression.They were Containment challenges from fuel coolant calculated as described in subsection 19E.2.8.1.1. interactions may occur when molten debris reacts These values were used, along with the appropriate rapidly, perhaps explosively, with water. Fuel coolant source terms, in the containment event trees. interactions are most likely to challenge the containment when molten debris falls into water. 19E.2.1.5.2.10 Fission Product Release Flow Area Examination of the containment event trees indicates that only 0.3% of all sequences have, water in the The presence of the COPS serves to substantially lower drywell before vessel failure. Despite this low reduce the uncertainties associated with the flow are probability, scoping studies were conducted for the release of fission products from the Q considering both the impulse and static loads. As discussed in Subsection 19E.2.1.6.7, the shock wave containment. The limiting flow area was chosen such that any slight variation would not affect the ability of transmitted to 'he structure provides the limiting the system to relieve the containment pressure. loads. Using conservative estimates for the impulse However,if the drywell head fails before the COPS load capability of the pedestal, the structure can opens, there is a great deal of uncertainty in the size withstand the loads associated with a steam of the opening. A sensitivity study was performed, as explosion involving 9.5% of the core mass. This is reported in Subsection 9E.2.6.10, which concluded three times the mass of a credible fuel coolant that there is a smallimpact on the fission product interaction in the ABWR, Therefore, the ABWR is release. In addition, only a small fraction of all very resistant to fuel coolant interactions. This failure releases occur as a results of drywell head failure, mechanism need not be considered further in the Therefore, no further consideration of containment containment event trees or the uncertainty analysis, failure area is necessary. 19E.2.1.5.2.8 Core Concrete Interaction and Debris 19E.2.1.5.2.11 Suppression Pool Haass Coolability The suppression pool bypass study of Subsection The issue of debris coolability has long been an 19E.2.3.3 was not able to show conclusively that a area of considerable uncertainty in the progression stuck open vacuum breaker would not lead to an of a core melt accident. If core concrete attack increase in risk. Section 19E.2.6.11 considers the continues, the timing and magnitude of potential potentialimpact on fission product release of a fully fission product release can be miected: the pedestal or partially stuck open vacuum breaker. The study could be eroded which could threaten containment concludes that there may be a substantialincrease in structure, non condensable gasses could pressurize offsite dose if a vacuum breaker sticks open. the containment leading to early rupture disk Therefore, this issue is examined using a detailed opening, and additional fission products could be uncertainty analysis. The results of this examination i released from the molten core. The ABWR design are summarized in Subsection 19E.2.1.6.3. l Amendment 19E.2-9.2
A 4.4 .-- e b eu-e.r,5hu--M----ha.em.m--+s- Ar 42 ha -4 sa a A- i- -+whm----+M44lh-R :14Ja 4h4M#4 # aid-a.iLL a Mu-A-5--Atr4__.,.+s-. ,-26.--Am4 -.- 4, g* . .+.,J4-s tWR- ms Standard Plant n, 4 19E.2.lJ.2.12 High Temperature Failure of the impact on the ABWR severe accident performance. Drpell As a result of this screening, three issues were identified for more detailed examination as being One of the failure modes identified for the potentially risk significant. The following provides a containment was the degradation of the seats for the discussion of how Direct Containment Heating moveable penetrations in the drywell due to high (DCH), pool bypass, and Core Concrete Interaction temperature, in the base analyses discussed in s'CCI) each impact the containment failure Section 19E.2.2, the only sequences which exceeded probability and risk profile, the threshold temperature of 533 K (500 F) were those in which debris was entrained into the upper 19EJ.l.5J.1 Direct Containment Heating dr)well and sprays were not available. A scruitivity studies were performed to determine the potentital A large number of calculations were performed to for other sequences to exceed the threshold determine the impact of DCH on the probability of temperature which could lead to early fission product containment failure and offsite risk. The analysis release. The largest increase in drywell temperature investigated uncertainties in a variety of phenomena: was only 5 K, which left ample margin to a high - hiode of vessel failure temperature failure, Therefore, no further study of hiats of molten core debris at the time of vessel this area is necessary. failure
- Potential for high pressure melt ejection 19E.2.lJ.2.13 Suppression Pool Decontamination Fragmentation of debris in the containment Factor Additional sensitivity studies were performed to The pressure suppression poolis a very effective examine other phenomena which could affect DCH.
means of removing fission products from the gas The study concluded that a deterministic best space in a severe accident. The efficiency of the estimate for the peak pressure from DCH would not t scrubbing process is typically characterized in terms lead to containment failure. Consideration of the I of a decontamination factor (DF) defined by the uncertainties in the phenomena lead to an estimated m mass of debris which enters the pool divided by the mass of debris which leaves the pool. htAAP.A3WR CCFP of 0.1% for all core damage events. Since the probability of containment failure due to DCH is ( , uses correlations based on the SUPRA code to very low, there is no measurable impact on offsite ? calculate the DF. In order to investigate the dose. O sensitivity of the offsite consequences of a severe accident to the suppression pool decontamination 19E.2.1.53.2 Core Concrete Interactions u factor, a simple sensitidty study was performed. The y blAAP ABWR code was modified to allow a A large number of calculations were performed as constant DF of 100, a very conservative value for the part of the investigation into core conctetc ABWR configuration, to be used for all species interactions in the ABWR These calculations I (except noble gasses, for which the DF is 1.0). This addressed uncertainties in the following parameters: resulted in an increase in fission product release of Amount of core debris about four orders of magnitude. Nonetheless, there - Debris-to-water heat transfer was no notable increase in offsite dose above a - Amount of additional stcclin the debris conditional probability of 0.04. Thus, there is not a Delayed Gooding of the lower drywell significant impact on dose, even for a DF of 100. Fire water injection instead of passive Gooder Thus, no further :onsideration of suppression pool decontamination factor is required in an uncertainty The conclusion from all of these uncertainty analysis, calculations were: 9' l
. 19E.2.1JJ Uncertainty Analyses 1. For the dominant core melt sequences that b release core material into the containment,90%
3 A systematic examination of severe accident result in no significant CCI. An insignificant y challenges was performed as part of the ABWR number of sequences are expected to experience PRA development. After screening the challenges dry CCI.
- -$ for their applicability to the ABWR, a sensitivity study was performed to examine their potential Amendment 19E2 93 l
l
-, --, . -,.- +,- -. - - ,e .e--q:--v<ey-- 4 yw, s- - ----ov==rw-->-+vW-w +i.v.wr- --+-- e--w't e ,ne*- w 'r-* v
ABWR mums Standard Plant . ReyJ
- 2. Even for those low Ircquency cases with The sum of the frequency of pool bypass (3) v significant CCI, radial crosion remains below the structural limit of the pedestal. After sequences with no drywell spray available is 7.4E.11; 0.05% of all core damage events. Since this value is consideration of uncertainties only 1.5c"c of the extremely low there is no impact on offsite dose.
sequences with significant CCI will suffer pedestal failure. Combining this conclusion with the first, only 0.15% of all core melt sequences with vessel failure willlead to additional drywell failures as a result of CCI.
- 3. The time of fission product release is not significantly affected by continued CCI.
4 The fission product release is dominated by the noble gasses when the containment overpressure protection system operates. This conclusion is unaffected by assumptions on debris coolability. Therefore, the offsite dose for sequences with rupture disk operation is not impacted by core concrete attack. These conclusions would indicate that the uncertainties associated with CCI have an insignificant influence on the contair. ment failure probability and risk.
$ 19E.2.1.5.3.3 Pool Bypass Analyses performed in subsection 19E.2.3.3.3(4) indicate that the only significant mode of suppression O pool bypass occurs via the vacuum breakers.
8 Uncer*2inty : alyses and sensitivity studies were performed to assess the effect of pool bypass on risk. Some of the key conclusions of these studies are summarized below.
- 1. The probability of a large leakage path between the wetwell and drywell ia approximately 0.4%.
- 2. There is a 2% probability that there is a small leakage path between the drywcIl ar.d wetwell.
Based on the Morowitz plugging model,90% of these sequences are expected to plug before the rupture disk setpoint is reached. In sequences with plugging, there is no significant increase in the time of fission product release or in offsite dose.
- 3. Use of the firewater spray system can prevent early opening of the rupture disk for a bypass path of any size.
l ['
' mm um l
l
ABWR miss Standard Plant %s 19E.2.2. Accident Sequences depressurire the vessel quickly enough to allow the low pressure systems to operate without ADS. Furthermore, the low frequency of Class lilA ( The accident sequences are chosen such that both the core damage accident classes and the events allows their consideration here. containment esent tree classes are well represer.ted. SBRC: Station blackout with EIC operating for Clagses of accidents with frequencies greater than 10' were considered in selecting the accident 8 houts is class tB.2. sequences to be studied. LHRC Loss of heat temovalin the sontainment A complete accident sequence is designated by sequences are characterited by a cooled care but an eight digit character. The first four characters the containment structure fails due to loss of indicate the general conditions of the accident. The containment heat removal. This sequence next two digits are used to identify any mitigating embodies class II. systems used. The seventh digit indicates the mode of release, and the eighth character indicates the LBLC: Large break LOCA with loss of all core magnitude of the release. A summary of the gooling represents elass IIID. accident sequence codes is given in Table 19E.2 3. NSCL: Transient with no acram or core gooling; The fitst consideration in selecting accident vessel fails at low pressure models class IC. sequences for analysis was to represent the core damage event trees. To accomplish this each NSCH Transient with no scram or core gooling; accident class was examined to determine the most sessel fails at high pressure represents class IE, severe sequence. The frequency of the event was then considered. If the freJuency of the most severe NSRC: The station blackout with no acram or sequence was less than 10 and if it was significantly boron injection sequence assumes that the EIC smaller than the overall frequency of the class then system is available for core cooling. The reduced the next most severe case was examined. Noje that flow to the core reduces the reactor power. Also the sequences with frequencies of less than 10 were mode:4d by this sequence are other loss of offsite not completely dismissed. They were retained in the power sequences where the operator manually O sum of the event class frequencies. Eight accident sequences were selected for reduces flow to the reactor in order to reduce power. This sequence portrays class 1%1. analysis with htAAP. Table 19E.2-4 shows how each For eacta base sequence, there are a variety of accident class relates to the accident sequences mitigating systems which could be used to prevent or l analyzed.- Each of the eight accident sequences is reduce the release of fission products to the described below. environment. The fifth and sixth digits of the accident sequence indicator describe the mitigating LCLP: Loss of all core gooling with vessel failure features which were assumed to operate. occurring at low gressure represents accident class ID and some IB 1 and IB-3 sequences. 00: This symbol is used when none of the t mitigative features are operated, due to failure of l LCHP: Loss of all core gooling with vessel the system or the operator, or the absence of the failure occurring at lLigh gressure models initiating condition for the system. accident classea IA and Illa as well as some IB 1 and 18 3 sequences. The results are somewhat IV: There are several means by which the non conservative for some of the Class IIIA operator may arrest the core melt ja the yessel. If sequences because the rate of water loss from the any ECC system is recovered or if the firewater vessel may be somewhat faster for medium break system started before vessel failure occurs it may l-LOCAs. Small break LOCAs will be accurately be possible to prevent vessel failure, assuring that modeled by this case. Even for the case of the any fission products generated are scrubbed medium break LOCA the results should be through the suppression pool via the SRV lines, reasonably accurate, because the definition of a In vessel recovery is treated as a sensitivity study medium break LOCA is that which does not in subsection 19E.2.4.2. i 19Ello Amendment
-\
ABWR :-s Standard Plant %3 1 PF: The passive [looder system is described in emergency procedure guidelines in Appendix ISA. l m Subsection 9.5.12. This system automatically opens a connection between the suppression pool Information about the hardware connections are (V) and the lower drywell region when the supplied in the description of the RHR system in te mperature of the lower drywell airspace SSAR Subsection 5.4.7.1.1.10. In particular, reaches 533K (500 F). This serves to keep the Figure 5.410 shows the connections from either corium temperature low, preventing the diesel driven pumps or the fire truck to the core. concrete interaction, and prevents radiation RHR system. The connection to the diesel driven heat transfer from the corium to the containment pump are in the RHR valve room. Opening structures and atmosphere, valves F101 and F102 allows water to now from the Gre protution system into the RHR piping. C The passive flooder system is designed to cause Periodic stroke testing of these valves is required 6. the lower drywell to be flooded when there is no by Table 3.9 8 of the SSAR to ensure vai e - water overlying core debris in the lower deywell. operability. The fire truck connection is located T If there is no overlying water pool the fusible outside the reactor building at grade level. Both h materialin the valve will heat up, and melt the connections to the RHR system rw protected by g fusible plug. If there is water overlying the debris check valves (F100 and F1N) tc .ec that RCS pool, the lower drywell will not heat up suffic- pressurization does not r sult i ' reach of the
- iently to cause the passive flooder to open. injection path. The required . Ste for the p Examination of the Containment Event Trees firewater addition system in speuned in Section j (Subsection 19D.5.11) shows that the firewater 2.15.6 of the ITAAC.
, addition system is expected to operate in most of the accident sequences. Therefore, the passive HR: Containment heat temovalis provided by
$e flooder in not needed in the majority of the RHR system. For the base analyses the RHR d accidents. Rather, the lower drywell tiooder is system is conservatively assumed to be viewed as a passive backup system which floods unavailable.
the lower drywell, in order to keep the temper. ature in the drywell tow, and in order to allow PS: f.assive flooder and drywell spray both (q ' quenching of the core debris. operate. The drywell sprays are one function of the RHR system. During severe accidents, FA: The (irewater addition system, described in especially those which cause vessel failure to occur Subsection 5.4.7, allows the operator to manually at high pressure, the drywell sprays keep the tie the fire protection system into the residual upper drywell cool. This prevents degradation of heat removal (RHR) injection line if this action penetration seals which could result in leakage + is performed within about 15 minutes this will through the movable penetrations and the release prevent core damage, as described in Subsection of fission products below the pressure capacity of 19.3.1.3.1. The firewater system also acts as a the containment. The upper drywell drains into mitigating feature after core damage. Under the suppression pool. Therefore, the use of the these circumstances, the water from the firewater drywell sprays will not prevent the temperature in system pours through the vessel and onto the the lower drywell from increasing. Therefore, the corium on the floor of the lower drywell. This passive flooder will open when the lower drpell stops core-concrete attack and radiation heating becomes sufficiently hot. in the same manner as the passive flooder, in addition, the firewater system adds water to the FS: A {irewater addition spray function was I containment increasing the thermal mass. This added to the firewater system as a backup to the reduces the rate of containment pressurization RHR drywell spray, Used in spray mode the and delays or prevents significant fission product firewater system adds external water to the release. The operator is instructed to turn off the containment increasing the thermal mass of the I firewater system when the water levelin the system, and it provides cooling of the upper suppression pool is at the vessel bottom drywell region. The operator is instructed to elevation, unless firewater is the only means operate the spray system if the vessel failure has 3 available for core cooling and the vesselis still occurred, as determined by the temperature of the
- d. intact. Operator actions governing use of the drywell and the inability to maintain water lesel in the vessel. The firewater spray causes the p 5 firewater addition system is specified in the V 6 19E 2.l t l d Amendment l
ABWR l Standard Plant **[^5. l i pressure and the temperature of the upper drywell to decrease rapidly. When the water levelin the f
! suppression pool reaches the level of the bottom of '
i the vessel the operator is instructed to turn the j firewater system off, turning it on again only as 3 necessary to prevent the upper drywell temperature j l from exceeding $33K (500 F). If drywell head ~ failure occurs the firewater spray system is to be I restarted. This causes any fission product acrosols to j agglomerate on the_ spray droplets, reducin5 the lission product release to the endronment.
! There are several mechanisms whereby fission 6
) pioducts may be released from the containment to
- the environment. The mode of release is designated 4 by the seventh character in the accident sequence indicator.
N: tiormal containment leakage does not allow l significant release to the environment as l j discussed in subsection 19E.2.4.3. l P: Leakage through movable genetrations in the - drywellis assumed to occur when the gas temperature exceeds $33K (500*F) and the
; pressure exceeds 0.515 MPa (60 psig). Further l discussion of this type of leakage is given in j Appendix 19F. If containment heat removalis not recovered drywell head failure or rupture disk opening could follow the onset of leakage.
R: An overpressure protection relief rupture disk is described in Subection 6.2.5 and in j 19 E.2.8.1.1. D: The drywell head is assumed to fait before < the rupture disk opens. The mean failure ! pressure of the drywell head is 1.025 MPs (134
- psig)if the temperature in the upper drywellis below 533K (500 F). However, as discussed in Attachment A to Appendix 19F,there is a small probability the drywell fails at lower pressure. At higher temperatures, the drywell head is assumed a
O Amendment 19E.2 Il i
I j . ABWR , m.s - i 4 Sta.pdard Plant %, [ to f ail at lower pressures as descri$ ed in lowed b; .sactor scram, The feedwater is conserva. Appendix 19F. tively assumsd to trip, with a coastdown of 5 seconds,
- Four of the Reactor Internal Pumps (RIPS) trip on
( high vessel pressure. The SRVs cycle open and closed t relieve the steam pressure. As the w;;ter { leici falls, the remainder of the RIPS trip on low le"cl. The ECC injection systems are assumed tu i E: Early structural failure of the containment fai. . has been proposed for casec which result in the failure of the vessel at high pressu;c. The effect The sequence of events which includes passive of an early containment stru-tural failure is flooder and rupture disk opening for this accident is I examined in Subsection 19E.2.4A shown in Table 19E.2 5 Figures 19E.2 2A through 19E.2 2H show the system behavior throughout the l accident sequence. , S: Suppression pool drainage into the RHR pump rooms may be possible following an earthquake. For these cases the release will be About one half hout after accident initiation. scrubbed but the release of fission products will sufficient decay heat has been generated to lower the begin with the onset of fuel damage. These cases water level to two thirds core height, and the-are considered in a sensitivity study in Subsection operator opens one SRV to provide steam cooling. 19E.2A5. The vessel blows down (Figure 19E.2 2A), while the
" fuel heats up (Figure 19E.2 2D) and begins to melt.
The final character in the accident sequence There is little generation of hydrogen gas due to the designator is assigned after the sequence has been metal water reaction during the in vessel portion of simulated with MAAP. This eighth character. the accident (Figure 19E.2 2F) because thc vessel indicates the magnitude of the release predicted by blowdown limits the available steam when the MAAP. Negligible, low, medium, and high ' cladding is hot. About 2 hours after initiation of the ategories were established as follows according to transient, the lower vessel head fails, the amount of noble gasses and volatile fhslon . products released: The corium falls into the lower drywell along with the , water that had been retained in the lower plenum O Noble Gas Volatiles of the vessel. Rapid corium to water heat transfer quenches the corium (Figure 19E.2 2D) and results N < 100% <0.1% in non equilibrium steam generation causing a L < 100% <1% pressure increase in the drywell (Figure 19E.2 28). M < 100% < 10% Then the pressure decreases slightly as the contain-H < 100% > 10% ment temperature and pressure equilibrate with the pool conditions. Just under one hour is then Additionally, the character 0 indicates that no core required to boil away the water in the lower drpell damage occurred, therefore there is no release of (Figure 19E.2 2E) before the corium begins to heat radioactidty. up (Figure 19E.2 2D). After the water in the lower drywell boils off the drywell pressure decreases In the following subsections cach of the accident because steam is condensed on the containment heat classes is considered in turn. For each general sinks but there is no steam generated. accident condition several possible mitigarms actions are considered as suggested by the accident (a) Passive Flooder Operation (PF) progression. After the corium in the lower drywell is uncovered, the corium and the gas above it begin to 19E 2.2.1.14ss of All Core Coollag with Vessel heat up. When the lower drywell atmosphere. reaches $33K (500 F) at about 5 hours (Figure - l Failure at Low Pressure (LCLP) 19E.2-2C), the passive flooder opens. Water then - The initiating event selected for this sequence is a pours from the wetwell into the drywell (Figure Main Steam Isolstion Valve (MSIV) Closure, fol- 19E.2 2E) to the level of the upper horizontal sent. 19E212 Amendment G
ABWR -m Standard Plant Ris_3 This covers the corium, quenches it, and generates a pools, with overlying gas spaces at potentially different pressures. In the large scale of the plant. (] small pressure spike (Figure 19E.2 2B). Following this there is again a slight decrease in pressure as the the cool water enters the lower drywell pool
'Q drpell returns to equilibrium with the pool. underneath the surface boundary layer of the pool.
Since the density is slightly higher than that of the Since the peak corium temperature during this bulk pool,it will tend to sink. This will tend to damp process is 1600K (2400*F) no significant core the oscillation. concrete attack occurs during the heatup of the corium, therefore no additional non condensable The size of the oscillation is dependent,in part on gasses are generated. When the corium is quenched the time step because the decrease in the bulk pool the generation of additional non condensable gasses temperature is a function of the amount of cool is prevented (Figure 19E.2 2F). water added to the lower dryweu. To determine the sensitivity of the containment response to the time After the passive flooder opens the corium is step used by htAAP, a representatise sequence was cosered by an overlying water pool, thus, the run using very small time steps. While the results temperature of the lower drywell gas decreases showed a slight decrease in the magnitude and (Figure 19E.2 2C). The small, periodic oscillations period of the oscillations, no significant effect on the seen in the lower dryweb water level after the passive overall tramient response was obsersed. flooder opens (Figure 19E.2.2E) are due to a physicalinstability caused by the small pressure and The upper drywell temperature continues to density differences between the lower drywell and increase since the remaining fuelin the vesselloses the wetwell. its decay heat energy to the vessel walls and drywell via radiative and convective heat transfer. The pres. The oscillations begin when there is a small surization of the containment continues (Figure pressure differential between the wetwell and the 19E.2 2B) because the corium is now transferring lower drywell. The pressure differential causes heat directly to the water wHe renits in stearring. relatively cool water from the suppression pool to flow into the lower drywell. This reduces the bulk The containment continues to pressurire until the temperature of the lower drywell pool. Since MAAP wetwell pressure reaches 0.72 MPa (90 psig) at p assumes the pool is well mined, the surf ace 20.2 hours (Figure 19E.2 2B), when the rupture disk i i]
\ temperature also decreases, resulting in a decrease opens. No penetration leakage (Appendix 19F) is in the partial pressure of steam in the lower drywell predicted since the temperature in the upper drywc!!
gas space. This pressure decrease (19E.2.2B) draws remains oclow 533K (500 F), until well after th: rupture disk opens (Figure 19E.2 2C). I 4 additional water into the lower drywtli pool from the y4 suppression pool. Figures 19E.2 20 and 19E.2 2H give the release l When the elevation of water in the lower drywell fractions of the noble gases, cesium iodide, and
} is sufficient to eliminate the pressure differential, the cesium hydroxide as functions of time. The release flow from the wetwell stops. The cooled water in the of noble gases is nearly complete one hour after the o rupture disk opens. The release of the volatile 4 lower drywell then begins to beat back up to saturation due to heat loss from the debris bed, species, Csl and CsOH, occurs over a much longer g
a Once saturated pool conditions are reached, period of time and is nearly complete at 100 hours, y p steaming begins and the lower drywell pressure increases. This could cause reverse flow through the The release fraction of Csl at 72 hours is less than IE 7. flooder line. The subsequent loss of mass in the T lower drywell would cause the region to heat up There is approximately a 2% probability that the 2 more quickly, exacerbating the amplitude and period drywell head will fait prior to the rupture disk of the oscillations. Therefore, the MAAP flooder opening for this case. Assuming the drywell head line modelincludes a check valve which prevents fails as the wetwc!! pressure reaches 0.72MPa (90 flow from the lower drywell into the wetwell. psig) at 20.2 hours, drywell head failure will preclude rupture disk opening. Fission product release begins While this instability is based on physical directly from the dr)well. Noble gas release is nearly phenomena, MAAP over predicts its severity. complete at 32 hours, and the volatile fission product MAAP models this system as two perft.ctly mixed release continues until 120 hours. The duration of 19E.LD p Amendment L 1 V
- ABWR uxsixxs
- Standard Plant ,,, 4 the release is significantly longer for the drpell head The firewster addition system continues to add failure sequence since the heat source in the drywell water, first filling the wetwell to the level of the suppression pool return path. At 7 hours, the allows only a slow depressurization of the wetwell j which contains the noble gases. The Cs! release suppression pool overflows. Then water begins to fraction at 72 hours is 7.5E.2, v.hich is much greatet spill into the lower drpell and the mass of water in
' than the release for the corresponding rupture disk both the wetwell and the lower drywell incresse in
. case. proportion to their surface areas (Figure 19E.2 3C).
i During this time the increase in pressure (Figure (b) Firewater Spray (FS) 19E.2 3A) is due to the slow compression of the i non.condensable gases above the water. If the operator fails to initiate the firewater Side calculations have shown that the pressure i.e addition system in the first 20 minutes of the accident to prevent core damage, there is still potential for the centainment is minimized when the water Iml is significant benefit from its use after vessel failure is near the bottom of the vessel, assuming that the
- assumed to occur. The results of a sequence using drywell and wetwell are at the same pressure. For i the firewater addition system are given in Table this reason, the operator is dire ted to turn off the 19E.2 6 and Figures 19E.2 3A through 19E.2 3E. firewater systern when the water levelin the 1
j suppression pool reaches the elevation of the bottom i The firewater system adds water to the of the vessel, which occurs at 23.6 hours (Figure . containment through the RHR injection lines. When 19E.2 3C), { trying to prevent vessel failure the operator is j instructed to inject water to the vessel via the LPFL After the firecater spray is turned off the
- line, if this is not accomplished in time to prevent d
vessel failure, the valses are realigned to the drywell spray. The water then pours from the upper drywell into the wetwell via the wetwell drywell connecting vents, and eventually overflows into the lower
' drywell. This cools the corium, preventing core. concrete attack and additional metal. water l reaction. Since external water is used, the effective 4
heat capacity of the containment is increased. Furthermore, since the decay heat in the corium is delivered by convection to the water, no significant radiation heat transfer takes place, and the lower and upper drywell atmospheres remain cool. Therefore, no degradation of the movable penetration seals is expected, and no leakage through these penetrations will occur. In this case it was assumed that the operator - starts the firewater system four hours after the initiation of the event. The first four hours of the transient are identical to the LCLP.PF.R.N sequence discussed above. When the firewater system starts a pressere spike (Figure 19E.2 3A) is ' observed in the drywell which is caused by the evaporation of droplets in a superheated atmosphere. After the containment atmosphere is cooled (Figure 19E.2 3B), the pressure drops fairly rapidly to match the droplet temperature, this causes some water to spill over fro.n the wetwell to the l lower drywell(Figure 19E.2 3C). I90'I'I3 I Amendment I
\
-..w- ,y,m n y yp +,, , . . - -,m,
4 i ABM nutus l Standard Plant %4 pressures in the drywell and wetwell increase (Figure drywell spray operating is shown in Table 19E.2 7. l 19E.2 3A) to values consistent with the temperature Figures 19E.2 4A to 19E.2 41 show the system of the suppression pool and non condensable gas response to the presumed accident. is pressure. The pressure in the drywell regions continues to increa e as steam is generated by the The early stages of this transient are identical to corium in the lower drywell. This forces water to be those of a LCLP accident. The MSIVs close, the a displaced from the lower drywell to the suppression reactor scrams and the feed *ater coasts down. The pool sia the wetwell/drywell connecting sents. When core becomes uncovered at 17 minutes, and metal l l water can no longer flow directly from the drpell water reaction begins generating hydrogen (Figure into the werwell the drpell region begins steaming. 19E.2 4G) as the core heats up. The sessel This steam Dows to the suppression pool where it is continues to cycle on the SRV setpoints (Figure
- quenched. During this period the pressure in the 19E.2 4A) as the water in the core boils away, and j wetwell stays nearly constant while that in the the core melts. Since the suppresion pool g
- drywelt region increases (Figure 19E.2 3A). temperature is below the suppression pool heat ;
- capacity temperature limit at the time of sessel rg
+ At 26 hours the wetwell becomes nearly failure,5PV loads are not a concern. e 5 saturated an:t the pressure in the wetwell begins to y increase along with that in the drpell. At 31.1 hours d the pressure in the wetwell has reached 0.72 MPa (90 psig) and the rupture disk opens, After the rupture At 2.0 hours the vessel fails. The initial discharge
- disk opens the pressure decreases rapidly (Figure of corium and water from the lower plenum is j 19E.2 3A) and fission product release begins. At entrained by the steam from the vessel into the i about $7 hours the water in the lower drpell boils upper drywell and wetwell because the sessel fails at J away leaving the corium uncovered. The gas high pressure (Figure 19E.2 4A). As the steam is
- j temgerature in the lower drywell increases to 533K . driven from the lower drywell the corium is carried 1 (500 F)(Figure 19E.2 3B) and the passive flooder into the upper drywell and wetwell (Figure opens at 61 hours, allowing water to flow from the 19E.2 4E) That portion of the corium which is
- suppression pool to the drywell(Figure 19E.2 3C). blown into the witwellis immediately quenched. It The noble gas release is nearly complete at 35 hours, heats up only very slowly, as the suppression pool and the volatile fission product release h nearly heats (Figure 19E.2 4C). The corium which is complete at 76 hours. The release fraction of Csl at transferred into the upper drywellis initially cooled j
72 hours is about IE 7 (Figure 19E.2-4C) by the atmosphere and by contact with the Door of the upper drywell. } There is approximately a 5% probability that the drywell head will fail before the rupture disk opens (e) Passive Flooder and Drywell Spray Operation for this case. Assuming the drywell head fails as the (PS) l
- wetwell pressure reaches 0.72MPa (90 psig) at 31.1 i hours, drywell head failure will preclude rupture disk The passive flooder opens 30 seconds after the opening. Fission product release begins directly vessel fails as the temperature in the lower drpell l
from the drywelt Noble gas release is nearly reaches $33 K (500 F) (Figure 19E.2 4D). This complete at 69 hours, and the volatile fission product allows water from the suppression pool to Good the l release continues until 90 hours. The Cs1 release lower drywell, cooling the corium in the lower 3 fraction at 72 hours is 5.3E 2, which is much greater drywell. This does not, however, ensure that the j than the release fraction for the corresponding upper drywell remains cool, since there is corium in 1 rupture disk case. this region. In order to prevent !cakage through the movable penetrations in the upper drywell ti.e sprays 4 19E.2.2.2. Loss of All Core Coollag with Vessel must be initiated within the first 4 hours of the l Failure at High Pressure (LCHP) transient. j
- The initiator used for this analysis is a station When the drywell spray is turned on the blackout with loss of all core cooling. For this temperatures of both the corium and the gas in the sequence the operator is assumed to fail'to upper drywell drop sharply (Figures 19E.2 4C and l depressurize the vessel. The complete sequence of 19E.2-4D). The containment pressure also drops as events for this accident with the passive flooder and steam is condensed by the spray droplets (Figure 19E114
/ wn&nen:
. kw A
, -r--- - ..n.- ----
m w -- ,- E e.e,---.+
- AB\M ursimas Pe* A Slandard Plant 19E.2 4B). Tbc rapid depressurization of the lower drywell also causes water to flow from the (j~'T suppression pool to the lower drywell through the
( ' open passise flooder (Figure 19E.2 4E). After the drywell sprays are turned on the
- containment slowly repressurites (Figure 19E.2 4A).
The pressure difference between the wetwell and the C) V l i g9g),g4 ; AmeMment
=j \ <
w/ l
= _ - .
AB W wims Standard Plant n,, 4 drywell is very small because the recirculation of Water is present in the lower drywell for the g water from the suppression pool to the drpell keeps remainder of the sequence since the passise flooder g the steam near the saturation pressure of the is open. suppression pool water. If at any time during this sequence the RHR heat exchangers begin to operate When the suppression pool water level teaches l the containment would depressurize. Containment the bottom of the vessel, at about 22 hours, the failure and fission product release would be averted. operator is assumed to turn off the firewater system. The corium in the upper drywell then causes the , temperature in the upper drpell to increase. When - If the heat exchangers are not recovered the rupture disk is assumed to open when the wetwell the temperature in the upper drywell again reaches pressure reached 0.72 MPa (90 psig) at 25.0 hours. 500 K (440*F) the operator restarts the drywell I Upon rupture disk opening, fission products leave spray. This causes the pressure and the upper the containment. The release of noble gases drywell temperature to decrease. After 15 minutes continues for about 8 hours after the rupture disk the operator turns the system off in order to opens. The volatile fission product release continues minimize excess water addition to the containment. The cycle is repeated many times for about 25 hours. The release fraction of Csl at 72 hours is less than IE 7. Due to MAAP code limitations the firewater There is approximately a 2% probability that the spray was switched to drywell spray from the RHR drywell head will fait prior to rupture disk opening system in the !.PCI mode after the water level in the for this case. Assuming the drpell head fails as the suppression pool reached the bottom of the sessel. wetwell pressure reaches 0.72MPa (90 psig) at 25.0 The effect of this change is an increased rate of hours, drywell head failure will preclude rupture disk suppression pool heating and containment opening. Fission product release begins directly pressurization, leading to an earlier containment from the drywell, Noble gas release is nearly failure than would be predicted if the spray complete at 35 hours, and the volatile fission product continued to be suppl 4d by Grewater addition. The release continues beyond 5 days. The Csl release werwell pressure reaches 0.72MPa (90 psig) at 50 fraction at 72 hours is 2E 4 hus and the rupture disk opens. The volatile fission pioduct release continues fu the next 75 f (b) Firewater Spray Operation (FS) hodrs and the Cs! release fraction at 72 hours is less than IE 7. It is possible for the operator to delay the time cf ~ containment structural failure and reduce the fission There is less than a 5% probability that the product release by adding water to the containment drywell head will fail before the rupture disk opens after a loss of core cooling with vessel failure at high for this case. Assuming the drywell head fails as the wetwell pressure reaches 0.72MPa (90 psig) at 50 , l LCHP pressure.PS R N, aConsider loss of all coreacooling caseoccurs which and begins identically hours, drywell head failure to will preclude rupture the reactor scrams. The operator is assumed to fait disk opening. Fission product release begins directly to blowdown the reactor, vessel failure occurs at high from t!!e drywell. The noble gas release is nearly pressure and corium is entrained into the upper complete at 55 hours and the volatile fission product-drywell and wetwell. release continues for more than 5 days. The release fraction of Cs! at 72 hours is 1.5E 4. It is assumed that the operator turns on the I firewster addition spray system 1.9 hours after the (c) Passive Flooder Operation start of the accident,just before the passive flooder would operate. The pressure and the upper drywell If the operator takes no actions after a high temperature decrease rapidly. The additionalwater pressure coro melt, high temperatures will ensue in from the spray is initially directed to the suppression the upper drywell and leakage will occur through the pool. Since the flow from the sprays does not large movable penetrations as discussed in Appendix initially enter the lower drywell, the passive flooder 19F. The sequence of events for this case is opens at 2.0 hours. This begins to flood the lower summarized in Table 19E.2 8 and is depicted in l drywell. The containment then remains in a stable Figure 19E.2 5A through 19E.2 5E. condition for eeveral hours with the containment pressure and suppression pool mass increasing. 19E.213 Amendment
- - -- ~ - . ._.- ,
, .- .- - - . . - . ~ . - _. - .- . - -.-. -.... - . ~ . . . . . . . - .
4 .I
.~ . . - + - . -
-- .~ _.
- d. 'I ABWR mm ;
[ Standard Plant - n,g j } Tha passive floodet opens when the temperature i in the lower drywell reaches 533 K (500*F) at 2.0 l hours (Figure 19E.2 58)e Water then Dows from the 7
-wetweilinto the lower drywell(Figure 19E.2 5D).-
.qeunching the corium in the to ser drywell(Figure i 19E.2 5C)./ la contratt, the corium in the upper - :
- j. drywell heats up, after an initial beat loss to the upper I drywell atmosphere and structures (Figure 19E.2 5B).
5 This heats the upper dr>well atmosphere. The seal !- degradation temperature of $33 K'(500'F) , i determined in Appendix 19F is reached about in-2.1 hours (Figure 19E.2 5B), but leakage does not l start at this time because the pressure is still relatively l
~
- low (Figure 19E.2 5A).'
e b The containment continues to pressurire, and j l 18.1 leakage hours. Thisthrough initiatesthe movable the release penetrations begins at of fission l products (Figure 19E.2 5E). However, since the 2 leakage is not sufficient to pass all of the decay heat '- ' energy, the containment continues to pressurize , (Figure 19E.2 5A). ! l At about 40 hours, the pressure in the drywell dips ,' by about 0.1 MPa (14.7 psid). This dip is caused by : l the flow of water from the suppression pool into the p 1 i i s a i. 1 3, iO
~~
n.
. . - ,, m. ., ,_,-- .-.r.,. - .- _ _ ..-. . _ , _ _ . - ~ . . . . . .m.__,.-~,m~.... - - - , - - .
ABWR Standard Plapt ux6tx4s efv % lower drywell which reduces the average temp. events for the case in which core cooling is erature of the water in the lower drywell. The maintained is summarized in Table 19E.2 9 and is O depicted in Figures 19E.2 6A through 19E.2 6E. temperature decrease results le a decrease in pressure because the drywellis filled with saturated The more serious sequence of events is that in which steam. The initial flow of water from the the operator fails to inject with the firewater system. suppression pool causes the pressure of the lower This case is summarized in Table 19E.2.10 and is drywell to drop, which in turn causes more water to shown in Figures 19E.2 7A through 19E.2 7F. flow frer' the suppression pool. '!he flow stops when enough water has been added to the lower A reactor scram occurs immediately upon loss of d pell such that the static head above the flooder power. The MSIVs close and the RIPS coast down. balances tiie pressure decrease. While this may be a Feedwater pumps also coast down and the water mathematical artifact of the calculation,it has no level bgins to fall. When the water bvel reaches sericus impact on the analysis. Level 2, the RCIC system initiates. The steam boiled off in the core is routed to the suppression pool At about 69 hours into the accident the drywell through the SRVs. gas ternperature has reached a steady value of 830 K (1035 F) iFigure 19E.2 5B) and the drpell pressure initially, the RCIC suction is taken from the has reached a steady value of 0.66MPa (81 psig) condensate storage tank (CST). After 1.3 hours the (Figure 19E 2 5A). The containment does not reach suppression poollevel high high alarm is reached, the wetwell rupture disk setpoint pressure of and RCIC suction switches to the suppression pool. 0.72MPa (90 psig), nor does it reach the pressure Later, at 4.4 hours, the high suppression poof l necessaiy to fail the drpell head. The dryweil head temperature alarm occurs, and the operator failure pressure at 830K is reduced to o.75MPa (94 manually switches RCIC suction back to the CST. psig) because high temperatures in the drywell The reactor is maintained in this quasiotudy weaken the drywcil head seal as liscussed in condition, with the supprescon pool heating up, a d Appendix 19F. the containment pressurizine for 8 hours. After the RCIC system is presumed :o Sil, the water in the The fission product release begins at 18.1 hours vessel continues to boil off a the suppresion pool. The pool begins to overflow ta the lower drywell at p) (v (Figure 19E.2 5E). The noble gas release continues well beyond 5 days, while the volatile fission product release is nearly complete at 70 hours. The release about 9 hours. fraction of Csl at 72 hours is 8.8E 2. The data in Figures 19E.2 6, in which core cooling is maintained by the addition of firewater, begins at 6 19E.2.2.3. Station filackout with RCIC (SBRC) hours. The sequence in which core cooling is main, tained is identical to the sequence in which it is not This accident initiator, SBRC, represents a until the addition of firewater at abou: 10 hours. station blackout serguence. These are characterized Thus, Figures 19E.2 7 can be substituted for Figures by the unavailability of all AC Power. Therefore, the 19E.2 6 for the first 6 hours. RCIC system aad firewster are the only systems available for core cooling. This sequence assumes (a) Firewater Addition Prevents Core Damage RCIC operates for 8 hours, providing core cooling (pet Subsection 19E.1.2.2). After the RCIC fails, the It is possible for the operator to prevent core operator depressurizes the vessel and begins damage during an SBRC sequence by using the injection with the firewater addition system which firewater system to inject water into the vessel after can maintain core cooling ir. definitely. However, no the RCIC is assumed to fail. To do this, operator containment cooling system is available since all the must manually depressurize the reactor and align the diesel generators were assumed to fail, valves to begin injecting with the firewater addition system. Two types of station blackout sequences are considered. In the first, the operator successfully The depressuri74 tion causes the water to flash to initiates the firewater addition system. This steam, lowering .he water level in the vessel (Figure sequence is then similar to class 11 events. There is 19E.2-6D). M/ AP predicts that the core heats up to no core damage unless the containment structural about 1150 Y. (1610'F) during this time (Figure l failure leads to core damage. The sequence of 19E.2 6C). "'herefore, severe core damage will not 19E416 A m nd a ni i 1
'%)
ABWR nuius StandariElant , ,, 4 occur. When the pressure teaches the shutoff head of r- the firemater addition sptem,1.96 MPa (270 psig), wster injection begins and the core cools rapidly The ( water levelin the sessel then rises n ntilit reaches lesel 5 (Figure 19E 2 6C). The o,$ctator then maintains water lesel between lesel 8 a id level . During this time, the containment pressure increases slowly while the RCIC operates (figure O 19E3161 p Amendment f s 1 i
i I ABWR %s l Standard Plant n, 4 19El 6A). After RCIC failure the water leselin the gas temperature reaches 533 K (500 F) at 23.5 hours. m sessel drops quickly. At 9 8 hours the wster lesci the paune flooder opens. [V} reaches 2/3 core height and the operator depres. surues the scuel. As the seuel pressure falls, the Whea the pauhe flooder opens water pours from containment pressute increases qukkly. Durmg the the welwellinto the lower drywell(Figure 19El 7Eh blowdown the water lesclin the suppression pml has This quenches the corium and causes the wetwell beco'ne sufficient to cause the water to begin to pressure to increase rapidly to the rupture disk merflow from the wetwellinto the lower drywell rupture prenure,0.72 MPa (90 psig)in about region (Figure 19E.2 6E). After the blowdown, minutes. when the firewster system is injecting, the pressure fines more slowly since only decay heat is being The fission product release for this sequence added to the suppression pool (Figure 19E.2 6A). (Figure 19E.2 7F) begins at 23.5 hours, the time of o rupture disk opening. The noble gas release lasts h The decay heat addition causes a slight volumetric
- d. espansion of water in the suppreulon pool. Since about 3 hours The volatile fiuion products are the water leselin the suppression poolis already at releasert slowly over the next 75 hours. The C$l 7
the overflow point, the espansion results in flow to release fraction at 72 hours is leu than IE.7. 4 g the lower drywell and causes a slight decrease in There is approximately a 24 probability that the j suppresion pool man. drywell head will fail prior to rupture disk opening When the wetwell pressure reaches 0.72 Mpa for this case. Auuming the drywell head fails as the (90 psig) af ter 32.3 hours, the rupture disk opens, wetwell pressure reaches 0 72MPs (90 psig) at 23.5 flowever, because no core damage has occurred hours, drywell head failure will preclude rupture disk there is no release of fiuion products, opening. Fiuion product release begins directly from the drywell. Noble gas release is nearly (b) Passive Flooder Operation complete at 38 hours, and the volatile finion product release continues until105 hwrs. The Csl release if the operstor fails to use the firewater addition ftaction at 72 hours is 3.4E 1, which is much greater system after the RCIC fails, then core damage will than the release for the corresponding rupture disk occur. The sequence of events for this case is shown case. T s j in Table 19E.210. The system response to this 19E 2,2.4. Less of Contalnment liest Remoial l accident is shown in Figures 19E.2 7A to 19E.2 7F. (LilRC) Eight hours after the loss of offsite power, RCIC is auumed to fail. The water level begins to fall, This case, LilRC, was simulated using an MSIV although the rate of the water lesel decrease is closure event with lou of the drywell coolers, since slower than that for the LCLP sequence because the this event isolates the reactor immediately, and will decay heat is lower. The operator depreuurizes the therefore direct the most heat to the suppression seuel when the water level reaches two thirds core pool of any Clau lt event. The sequence of events is height by opening one SRV (Figure 19E.2 7A) at shown in Table 19E.2411. Figures 19E.2-8A through l 9.7 hours (SRV operability is discussed in Subsection 19E.2 8C show the system response to this sequence. 19E.2.1.2), if the operator fails to begin injection using the firewater system then the fuel melts slowly, For most of these sequences ECCS suction is l and the vessel fails at 123 hours. Initially drawn from the CST. When the high suppression poollevelis reached the suction a The corium and the lower plenum water then fall switched to the supprenion pool. Howeser, for to the lower drywell floor The containment cor, tin- simplicity, no credit was taken for the CST insentory ues to pressurize as this water boils (Figure The effect of this assumption is to underestimate the l 19E.2 7B). At 21.1 hours the lower drywell dries out mass of water in the suppression pool, thus overpre. (Figure 19E.2 7E) and the corium begins to heat up dicting the increase in suppression pool temperature (Figure 19E.2 7D). The corium radiates energy to and containment preuure. Additionally,in the later the lower drywell gas (Figure 19E.2 7C). When the stages of this transient the operator could switch the suction for the ECCS back to the condensate storage pool or use the firewster addition system, either of 90 m
/N Am namm \- )
ABWR "*7;I^j Standard PlanL_. i which provides a source of makeup water to the f suppression pool. , I ( hislV closure causes scram and feedwater trip. l As the water lesel falls core cooling (RCIC)irutiates. l
/
Since the reactor is isolated ett of the decay heat is directed to the pool, causing ine pool temperature to increase (Figure 19E.240). When the suppression pool temperature reaches f/C (1$5*F) the operator q
=
blows down the reactor in accordance with the EPGs. As the veuel pressure f alls. RCIC trips due V 19t' 2 l?I Amendment O V
ABWR nA61%A1 Standard Plant Rev A to insufficient turbine pressure. The wster lesel falls. The system response to this esent is given in Figures and the HPCF system initiates. 19EO.9A through 19E.2 9D. The containment preuurites sery slowly. At The feedwater system h conscrsathely auumed to 21.7 hours. the preuure reaches 0.72 MPa (90 psig), trip at the initiation of the event for this analph, (Figure 19EO.8A) and the rupture dhk opens. After The reactor serams on a high drymell preuure signal, r rupture disk opens the supprenion pool begins to and the MSIW close as the vessel preuute drops. boil off(Figure 19E.2 8C). The system will remain The core uncovers in 2.8 minutes and the fuel begins in this quasiateady state for a verylong time. to beat up. Venel failure occurs at 1 A hours. If at any time during this transient a source of At the time of vessel failure, the corium and water makeup wster to the containment can be used, the from the lower plenum fallinto the lower drywell reactor can be maintained indefinitely in this state. The corium is queached by this lower drywell water. As mentioned abose, either the firewater addition The water in the lower drywell then begins to bml ll system or the water in the CST could preside a away (Figure 19E.2 9C), pressurizing the source of makeup water to the containment. containment. (Figure 19E.2 9A). If makeup water is not supplied the water levelin (a) Passise Flooder Operation the supprenion pool will eventually become so low that the cote cooling pumps are unable to draw After the water in the lower drpeU is boiled away sufficient suction, and core cooling could be lost, by the decay heat energy in the corium, the corium begins to heat up, raising the lower drywell l The transient was simulated for 72 hours in thistemperature (Figure 19E.2 9B). When the gas analysis and that condition was not reached. When there is insufficient suppression pool suction the temperature in the lower drywell reaches $33 K operator could still maintain core cooling by (500 F) at 5,7 hours the passive flooder opens switching the ECCS suction back to the CST. The automatically. Water flows into the lower drywell CST has at least 8 hour capacity for core cooling (Figure 19E.2 9C) and the temperature drops as based on the station blackout performance steam is generated (Figure 19E.2 9B). auessment (Subsection 19E.I.2.2). O if core cooling is lost, the water in the vessel will begin to boil off slowly, and eventually, core melt will After the passive flooder opens the containment pressurizes slowly (Figure 19E.2 9A) as steam h senetated in the lower drywell, The eniire occur, no earlier than three hours after the loss of containment remains cool (Figure 19E.2 98) since core cooling.,The analysis of this transient was not the corium is conret carrit i sny further because there is a very long time for the operstar to tah & Lecessary action to When the wetwell pressure reaches 0.72MPa (90 terminate the event. psig), at 19.1 hours (Figure 19E.2 9A), the rupture dhk opens. ne fhslon product release occurs oser the next 105 hours (Figure 19E.2 9D). The C$l 19E.2JJ. Large LOC A with Failure of All Coet relcue fraction at 72 hours is less than IE 7. l Cool %g tLBLC) There is approximately a 2% probability that the A main steam line break is assumed to represent drywell head will fait prior to rupture disk opening - the LBLC case, since it has the largest flow area and for this case. Auum*mg the drywell head fails as the will cause the most rapid lou of coolant from the wetwell pressure reaches 0.72MPa (90 psig) at 19.1 vessel. The sequence of events for this case is similar hours, drywell head failure will preclude rupture disk to that for LCLP (loss of core cooling with vessel opening. Fission product release begins directly failure at low preuvres, hwever, the core melt will from the drywell.- Noble gas release is nearly occur earlier for the LBLC cue. T'ne sequence of complete at 31 hours, and the volatile fluion product esents for the LBLC case with the passive flooder release continues until 100 hours. The Csl release fraction at 72 hours is 2.2E.2, which is much greater and dr>well head failure is shows in Table 19E.212. than the release for the corresponding rupture disk case. IDIS Amendment v
ABWR DA*^$ Standard Pj' ant Pev A (b) Firewater Spray if the operator initiates the firewater addition system to add water to the containment through the ( RHR line then the time to containment structural f ailure will be delayed. For this analysis it is assumed that the operator begins injection 4 hours after the start of the accident. The sequence of esents after sessel f ailure for this sequence is similar j to that for the LCLP fS R N sequence shown in Figures 19E.0 3A to 19E.2 3E.
/
,,n. t. t O\
l V t
ABWR m si m s Standard Plant ,o 4 When the firewster sptem is initiated there is Upon loss of power, the MSIVs close and the , some spluhing of water into the lower drywell. This feedwater and recirculation pumps trip. All O presents the code from predicting operation of the pusive flooder, automatic and manual attempts to insert control rods are assumed to fail. The SRVs open to relieve the vessel preuvre. Furthermore, allinjection pumps, Esentually, at about 11 hours, the suppreulon including the RCIC and SLC pumps fail to inject pool osernows into the lower drywell. Water it wster into the vessel. Because of the increased added to the containment sia the firewater splem power lesel the water levelin the seuel falls rapidly l until the *ater leselin the suppression pool reaches and the core is uncosered in 3.7 minutes. l isthe lesel inofthethe no boiling sessel lower drywell.bottom. The containment During thisThetime thereof the uncovered core now . temperature preuurires slowly due to the compteulon of the begins to rise (Figure 19E.2108), and core damage non.condensable gaues. At 23.4 hours the firewster begins. At 30 minutes the operator is suumed to initiate ADS and the vessel blows down. When the 1 intem is shut off. As in the LCLP.FS.R.N case ' (Subsection 19E.2.2.l(b), the containment preuure veuct f alls at 1.3 hours, the preuvre is sufficiently l first increases very slowly as the water in the lower low to prevent entrainment. The corium, together drywell heats to saturation. Then after boiling with any water in the lower plenum, falls into the begins, the preuure rises more rapidly, lower drywell(Figure 19E.210C). The wetwell preuurs resc,hes 0.72MPa (90 psig) The corium is quenched in the lower drywell by at 29.5 hours, the rupture disk opens, and fission the water from the lower plenum. The water then product release begins. At about 62 hours the lower boils, causing the drywell preuvre to rise (Figure drywell has dried out leaving the corium uncovered. 19E.210A). All of the water is boiled off at L9 l This causes the gas W rerature in the lower drywell hours (Figure 19E.210C), to increase to S.J /t (%?) causing the passive flooder to open. N wm of volatile fission (a) Passive Flooder products is nearly comAu pJ hours. The release fraction of Cal at 72 hours is less than 1E.7. If no actions are taken by the operator to initiate the firewster system, the passive flooder will open O There is approximately a $% probability that the drywell head will fail before the rupture disk opens for this case. Assuming the drywctl head fails u the when the temperature of the lower drywell reaches
$33 K ($00*F) er 4.4 hours. At that time water from l the wetwell will pour into the lower drywell, coscring wetwell preuure reaches 0.72MPs (90 psig) at 29.5 the corium. This prevents core concrete attack and hours, drywell head failure precludes rupture disk metal water reaction from occurring because the opening. Fission produc. release begins directly corium is not st.fficiently hot for either reaction to from the drywell. Noble gas release is nearly occur (Figure 19E.2108).
complete at 60 hours, and the volatile fission product release continues until 95 hours. The Csl release The containment pressure then begins to rise fraction at 72 hours is 2.4E.2, which is much greater slowly as steam is generated (Figure 19E.210A). than the release fraction for the corresponding The rupture disk opens at 18.7 hours, and the fission rupture disk case. products are reteesed (Figure 19E.210D). The noble gas release lasts about 2 hours. The volatile 19E.2J.6. Concocevet IAan of All Coet Cooling and fission product releue lasts about 85 hours. The Csl A1WS with Vessel Falises at law Prveaure (NSCL) release fraction at 72 hours is leu than IE.7. The sequence chosen to represent the NSCL case There is approximately a 2% probability that the is a station blackout case with failure to scram. This drywell head will fait prior to rupture disk opening
.cquence is analogous to the LCLP case, with the for this case, Assuming the drywell head fa,ture will additional failure of reactivity control. The sequence preclude rupture disk opening. Fission product of events for this case,if the operator does not release beg;ns directly from the drywell. Noble gas initiate the firewster addition system is given in release in nearly complete at 33 hours, and the Table 19E.213. Some of the important parameters volatile fission product release continues until 100 are depicted in Figures 19E,210A through hours. The Cal release fraction at 72 hours is 8.5E.2.
19E.210D. 19ESt9 Annamni
ABWR m. ,,, Standard Plant m3 which is much greater than the release for the corresponding rupture disk case. O (b) Firewstet Spray if the operator begins injection using the Grewatet additions system after vessel failure has occurred, then the time of dr>well head f ailure can be delayed. l The sequence LCl.P.fS.R.N of esents case shown for 19E.2 in Figures this case 3A is similar to the through IVE 2 3E. For this sequence, where neither scram or core cooling was successful, the operator is assumed to initiate the Grewater system within 4 hours. When the Grewater system is initiated, there is some splashing of water into the lower drywell. This prevents the passive Gooder from opening. The Grewster addition serves to keep the drywell cool, and increases the thermal mass of the suppression pool, slowing the containment pressurintion rate. The water levelin th esuppression pool reaches the spillover height at about 15 hours. When the water level of the suppression pool reaches the bottom of the veuel, at 23,7 hours, the operator is assumed to turn off the system. The containment pressurlaation rate then increases, and the rupture disk opens at 30.7 hours. O At about 57 hours the water over the corium boils away ! caving the corium uncovered. The gas temp"erature in the lower drywellincreases to $33 K (500 F) and the passive Gooder opens at 61 hours. The volatile fission product release continues for the next 8 hours. The release fraction of Cat at 72 hours is leu than IE.7. There is approximately a 5% probability that the drywell head will fail before the rupture disk opens for this case, Assuming the drywell head fails as the wetwell pressure reaches 0.72MPs (90 paig) at 30.7 hours, dryweu head failure wiu preclude rupture disk opening. Finaion product release begins directly from - the drywell. Noble gas release la acarly complete at 62 hours, and the volatile fission product release continues until 85 hours. The Csl release fraction at 72 hours is 6.4E 2, which la much greater than the release fraction for the corresponding rupture disk Case. 19E119 I Amendmeat
3 ABWR u si m i ny 4 Standard Plant .,,_ sharnly by about .1 MPs (14.7 psid). This dip is 19E.2J.7. Concureent L4ss of All Core Cooting and caute when the pressure difference between the ATWS with Venel Failurt at liigh Fressure wet *cil .cd lower drywell sides of the passise flooder allows water to flow from the suppreuion l (NSCH) pool into the lower drywell, which is now fdled with The NSCH sequence is analogous to the LCHP wster. The initial flow of water from the suppreuion sequence described in 19E.2.2.2 with the additional pool causes the temperature of the lower drywell f ailure of reactivity control. The main effect of pool to decrease which in turn results in failure to scram or inject boron is to decrease the depreuurization of the lower drywell. This induces time of venet failure, since the teactor stan at power more water to flow from the suppreulon pool. The for the first few minutes of the transient. However, now stops when enough water has been added to the the power lesel soon drops due to additional voiding lower drywell so the static head above the flooder in the core. The sequence of esents for the NSCH balances the preuure decrease. While this may be sequence where the passive flooder is the only only a mathematical artifact of the calculation, it has mitigating sptem is given in Table 19E.214. Figures no serious impact on the analpis. 19E.211A through 19E.211D illustrate the key parameters. At about 67 hours into the accident the drywell gas temperature has reached a steady value of 850K Following an isolation event the water in the (1070'F). At the same time the drywell preuure has c veuel boils rapidly, and the core becomes uncovered resched a steady value of 0.67MPa (82.5 psig) g in 3.6 minutes. If the operator fails to blow down to (Figure 19E.211A). The containment does not low pressure, a high pressure venel melt occurs in reach the rupture disk setpoint pressure of 0.72MPa 4p 1.3 hours. Since the suppreulon pool temperature is (90 psig), not does it reach the pressure necessary to
;! below the suppreulon pool heat capacity temper. f all the drywell head. The drywell head failure sture limit at the time of venel fallute, SRV loads pressure at 850K is reduced to o.71MPs (88 psis) 3 G are not a concern. As with a LCHP event, corium is because high temperatures in the drywell weaken the entrained into the wetwell and upper drywell(Figure drywell head seal u discuued in Appendix 19F.
19E.211C). Fission product release begins when drywell O (a) Passive Flooder (PF) If the operstor does not initiate the firewater penetration leakage starts, at 17.8 hours. The initial release rate is very small (Figure 19E.211D) because of the small penetration leakage. The noble l sptem then the pauive flooder will open at 1.4 hours gas release continues well beyond $ days, while the when the tersperature of the gas in the lower drywell volatile fiulon product release is nearly complete at l reaches $33 K (500*F) (Figure 19E.2118). This 65 hours. The release itaction of Csl at 72 hours is immediately cools the corium in the lower drywell 7.3E 2. and the gas temperature in this region drops to near the saturation temperature. (b) Firewster Spray Addition (FS) The only heat sinks available to remove the decay The scenario in which the operator begins the heat generated by the corium in the upper drywell firewster spray after veuel failure has occurred is the region are the concrete walls and the atmosphere, analog to the LCHP.FS.R.N case considered in l Since the heat transfer to the concrete is not very subsection 19E.2.2.2(b). The only major differences after the spray is initiated are the temperature of the l, effective the Shortly increases steadily. gas temperature after the passive flooder la the upper pool, and drywell consequeatly the pressure in the opens the temperature in the upper drywell exceeds containment, the penetration leakage temperature threshhold (Figure 19E.211B). However, since the pressure is Comparison: of the LCHP.PF.P.M and only 0.25 MPa (22 psig), leakage does not occur at NSCH.PF.P.M pressure histories (Figures 19E.2 5A this time but is delayed until 17.8 hours when the and 19E.211A, respectively) shows that the drywell pressure reaches 0.46 MPs ($2 paig) as additional power generated in the ATWS sequence
. shown in Figure 19E.211A. causes the pressure for this case to be about 0.02 MPs (3 psi) higher than the non ATWS At 47.7 hours, the preuure in the drywell dips Ara.w .
iseno ____ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ .1
ABWR m.imi Rev A Standard Plant sequence. This increase in pressure represents the additional power generated in the first hours of the ATWS transient. After this time the power fevel will 4 l i
- O 19E2 201 Amendment O
9
y ABWR :mius Standard Plant %i drop to decay heat levels because of a strong water lesel in the vessel will drop and the core will q negative void coefficient in the core. begin to melt, as seen by the increasing fuel " Q Therefore, ince the difference in the pressures temperature in Figure 19E.212D. At the ume time the power will drop to the decay heat level because of the two cases is small, the transient considered of increasing soids (Figure 19E.L12C). i:ere, a simultaneous loss of all core cooling and failAre of reacthiry control with vesset failure at high (a) Passive flooder l pressure,in which the operator start the firewster The operstor depressurires the reactor l 1.CHP.FS.R.N spray sptem after case seuelinfailure, considered willNobehne 19E.2.2.2(b). hke the 10 minutes after the RCIC is tripped. Vessel failure further analpis of this sequence was performed. ensues at 5.6 hours. Corium and water fallinto the l
' lower drywell(Figure 19E.212E). A short time 19E.2.2J Concurrent Station Blackout Wth ATWS later, at 8.6 hours, the wetwell pressure reaches 0.72MPa (90 psig) and the rupture disk opens. The l INSRC) containment begins to depressurite (Figure The final sequence considered here, NSRC, is 19E.212D) and fission product release begins. The the case where a station blackout occurs and all lower drywell dries out at about 30 hours and the reactisity control fails, in this sequence the RCIC is passive flooder opens soon after. The noble gas the only system available to provide core cooling. release occurs within the first 5 hours after the The sequence of events for this case is given in Table rupture disk opens, while the volatile lhsion product 19E.215 and some of the key parameters are shown release continues for 100 hours. The release fraction l in Figure 19E.212A through 19E212F. of Cs! at 72 hours is less than IE 7 Upon the loss of power the reactor isolates There h approximately a 2% probability that the immediately. The vessel panssure increases and drywell head will fait prior to rupture disk opening
- SRVs cycle to controi pressure (Figure 19E.2.12A). for this case. Assuming the drywell head fails when l The water level falls rapidly and at 1.1 the minutes, the reaches 0.72MPs (90 psig) at wetwell pressure RCIC system begins injec'hg. The water level 8.6 hours, drywell head failure will preclude rupture continues to fall and at 2.2 minutes the top of the disk opening. Fission product release begins directly
(^) core becomes uncovered. from the drywell Noble gas release is nearly complete at 19 hours, and the volatile fission product Although the top few noder heat up to about release continues until 50 hours. The Cal release 850 K (1070 F), the core dens not melt at this time fraction is 4.8E 1 at 72 hours, which is much greater due to steam coollug (Figure 19E.212D). The than the release for the corresponding rupture disk power level during this time is about 4% (Figure case. 19E.212C). This amount is that required to boil the water injected by RCIC. During this time the (b) Firewater Sprays Operated containment pressurires fairly rapidly due to the relatively high rate of steam generation (Figure The operator can delay the release of fission 19E.2128). products by initiating the firewater spray before the rupture disk opens. If the firewater spray begins at All of the water added by the RCIC system is 6.1 hours,30 minutes after vessel failure, the fission converted to steam in the core. The steam flows product release does not begin until 26.4 hours, through the SRVs to the suppression pool where it is Upon firewster spray initiation the containment quenched, adding to the mass of the pool. At 1.9 pressure and temperature decrease. At 22 hours the hours the suppression pool begins to overflow into level has reached the bottom of the vessel and the the lower drywell. operator is instructed to turn off the spray. The containment begins to pressurite until, at 26.4 hours, if the operator is unable to shutdown the reactor the wetwell preuure reaches 0.72MPa (90 psig) and by means of either the rods or boron injection then the rupture disk opens. The containment rapidly the containment pressure will reach the RCIC depressurizes and fission product release begins. At l turbine exhaust pressure limit in 3.6 hours (Figure 49 hours the lower drywell dries out leaving the 19E.212B). This causes the RCIC to tsip. As there corium uncovered and the passive flooder opens at is no other source of vesselinjection available, the 52 hours. The noble gas release is nearly complete 19E 2 21 A Amendment
n - l ABWR m.m.
,, ,, 3 Standard Plant at 33 hours, while the volatile fission product release continues until about 120 hours. The Cat islease O fraction at 72 hours is less than IE 7.
There is approximately a $% probability that the drpell head will fait before the rupture disk opens for this case. Assuming the drpell head fails as the wetwell pressure reaches 0.72MPa (90 psig) at 26.4 hours, drywell head f ailure will preclude rupture disk opening. Fission product release begins directly from the drpell. Noble gas release is nearly complete at 38 hours, and the volatile fission product release continues until 85 hours The Csl release fraction at 72 hours is 2.0E.1, which is much greater than the release fraction for the corresponding rupture disk case. 19EJ.2.9. Summary Table 19E.216 gives a summary of the critical pa. tameters for the accident sequences discussed above. For each sequence considered in the analysis which results in fission product release, the time of vessel failure, the start of itssion product release and the time of drywell head failure are givea. Also shown are the duration of the release and the release l fraction of Cal after 72 hours. O
\
19tL221i Amendment ()
ABWR m ., Standatd Plant nu ; 19E.2.3 Justincation of Phenomenological triggering can occur if the energy transfer from i Assumptions droplets forms additional droplets from the debris O Several separate effects studies were performed to stream and rapidly mixes them with surrounding water. Both external triggering and self. triggering can cause steam explosions. External triggering has supplement the MAAP analyses of sesere accident sequences. These studies were performed to address been employed in experiments by the use of the technicalissues which could potentially have submerged explosive devices, which include impact on the ABWR response to postulated severe exploding wires, primicord, and blasting caps. The accidents. They were selected for consideration based corresponding energy release of external triggers on the results of past PRA experience within the promotes the rapid breakup of molten debris into i industry. small particles. I 19E.23.1. Steam Explosions Self triggering sometimes is caused by debris stream impingement and shattering on submerged A steam explosion is caused by thermal energy structures, if triggering suddenly creates increased release to water, which causes rapid steam formation, surface area for heat transfer, the rapid formation of expansion, and substaatial pressure or impact loads the steam causes water acceleration, which can on structures. It is possible that the high thermal create substantial pressure and impact forces. energy content of molten core debris can cause a steam explosion if it enters water under conditions 19E.2J.lJ. Prwlous Studies favorable to rapid heat transfer. Analytical and experimental studies of stearn The potential for an ex-vessel steam explosion for explosion IDCOR study phenomena (References are 5,6).summarized Analytical modelsin a 1983 l_ a postulated severe accident in the ABWR plant is evaluated in this subsection, and is found to be and experimental studies reported in the literature extremely low, are discussed from the standpoint of necessary
- conditions required to produce large scale steam 19E.23.1.1.'the Steam Explosion Process explosions. It was determined that the following O' Figure 19E.213 helps explain the process of specific conditions had to be satisfied for steam explosions to occur:
steam explosions. It is postulated that a loss of cooling mechanism causes the reactor core to melt, 1. Many tonnes of molten core debris must enter the followed by vessel breach and discharge of molten water, core debris with high thermal energy into the lower drywell, which is assumed to contain a stagnant pool 2. The debris must be coarsely fragmented into of water. The energy transfer rate to water depends about one centimeter diameter or smaller on the volume of submerged debris and available particles and thoroughly mixed with water. surface are for heat transfer, if many small particles of molten debris enter or form in the water, heat 3. A trigger must initiate a localized explosion which transfer will be rapid. Larger particles have less subsequently fragments adjacent particles into surface area per unit volume, and correspondingly submillimeter size, and rapidly mixes them with slower heat transfer. Moreover, internal heat the surrounding water in less than a millisecond. diffusion in large particles can limit the heat transfer promoting rapid vaporization. rate to the water.
- 4. A continuous liquid slug must cover _ the High velocity discharge of liquid debris into air vaporization zone so that it can be propelled can form spray size droplets before they enter a water upward like a missile by the explosive interaction.
pool. However, for most cases debris discharge in the ABWR is expected to occur by gravity draining from It was concluded in the IDCOR study that for a depressurized vessel, which could form larger both in vessel and ex. vessel steam explosions, the droplets in air, about 24 mm. Smaller droplets would formation of tonnes of coarsely fragmented mohen be formed by a stream of molten debris falling core debris dispersed in water, with the associate.1 through water. An eyent c alle d large steam generation rates,is fundamentally Amendment tE2 22 s i
ABWR 2mimAs Standard Plant Itev A inconsistent with a continuous overlying liquid slug melt, ara coolant, were compared on a normalised required for efficient energy transfer. That is, steam time scale, and show that for properly scaled molten O. esplosions do not provide a set of credible physical procerses leading to failure of either the primary debris pours, comparable behavior can be expected in scales as low as 1/8 of full sire. Most of the j system or the reactor containment building. The IDCOR conclusion and the conclusion of this experiments (Reference 5) werediscussed in the at smaller scales, whichIDCOR leaves tcportl analysis differ from the earlier WASH 1400 report the scaling question short of full resolution. l (Reference 7),in which energetic steam explosions However,it is expected that the basic theoretical were believed possible, leading to early containment formulations, which are consistent with experimental i failure, phenomena at small scales, can be extrapolated to l evaluate the potential for steam explosion in full sire hf olten debris discharge from a reactor vessel at applications. That is largely the IDCOR approach high pressure is more likely to be atomized and enter which leads to the conclusion of low steam explosion the pool as small droplets, which can rapidly transfer potential during a severe accident in a full scale thermal energy and increase the potential for a reactor. steam explosion. However, the major conclusion from data and analytical models discussed in the 19EJ.3 lJ. Theoretical Considerations l IDCOR study (Reference 5) imply that low vessel pressure and gravity discharge of molten core debris The theoretical considerations of this study are in ABWR has an extremely low potential for based on simplified, bounding analyses which tend to generating a steam explosion. be conservative in the promotion of steam explosions, These considerations are used to The IDCOR study (Reference 5) reports that in evaluate the geometric conditions expected for a l various experiments it was possible to cause a steam molten debris pour into water, the heat transfer and explosion with an external trigger, which broke the steam formation rates, pressure rise and water molten metal into small drops and mixed them with hydrodynamic response time. It is concluded from surrounding water. Several experiments were these considerations that the potential of an ex vessel reported in which iron thermite was observed to steam explosion in ABWR is extremely low. O 1 undergo self triggering prior to a steam explosion. However, the thermite at a temperature of 3000 K (a) Estimated Debris Droplet Formation Size apparently remained liquid during the triggering process, offering only surface tension resistance to Hydrodynamic instability causes droplet formation molten droplet formation. Molten core debris is when two parallel, adjacent liquid streams with expected to be discharged at the liquidus temperature different densities travel at different velocities. of 2600 K. The outer surface of small droplets freezes Figure 19E.214A shows the heavier liquid of density rapidly after entering water, perhaps even while p n the bottom, flowing horitontally with velocity falling a long distance through air, so that further v h, underneath the lighter liquid of density p , h droplet division requires more energy to fracture the 11owing with velocity v . The condition for unstable outer crust formed than it does to overcome liquid surface tension. This helps explain why scif triggering interface waves cali be obtained (Reference 9)in the form can be obserwd with some highly superheated metals, but is much less likely with molten core debris. : 8W3 0 2Kg(p'#) d# #ht# .IYht #) h t Experimental work reported in the IDCOR study 3
+ -- <0 (1) was performed in small scale test facilities, which A A A,(p *# )
h L leads to questions about how accurately the experiments represent full size severe accident steam where o is the surface tension of the heavier liquid l explosion response. Theofanous (Reference 8) and A is the wave length. An unstable wave can grow addressed the scaling concern by formulating the to the amplitude at which it detaches from the basic phenomena of steam explosions, and comparing heavict liquid and forms a droplet of approximate computer solutions for different scales. Calculated diameter A,or radius r A/2. pressure and volume frac' ions of steam, Figure 19E.214B shows a corium stream of density p falling vertically at velocity V through O' stationary fluid of density p,. Here, the gravity term 19E.2 23 Amendment __J
ABWR nisioor, Standard Plant , ,4
- dor.s not play a role in wave growth, and equation (1) surface is obtained by lategrating equation 4 with
' O gives the approximate minimum stable radius of V, = V, which gives droplets formed as A Cd#eY K0(p + p.) - = erp (5)
> (2) A, pL r ,., ,
pp,V The stream area broadens about 11 percent for an if the debris stream discharge is determined by approximate C of 1.0. A debris strearn which gravity draining ,its donward velocity at distance z broadens in the I would re. absorb small interface from the debris surface in the reactor is V = /(2gr). droplets formed by instability in the water. This l Droplet sizes formed in air and water would be action would tend to reduce the formation of many different because density p, plays a strong role in small interface droplets for high heat transfer into equation (2), the water. It follows that substantial droplet formation in the water pool would have to occur by When the debris stream first enters the water self. triggering. The dynamics of steam formation pool, it undergoes deceleration, a, due to the drag and the triggering process are discussed after a force. Under these conditions, the forces on the consideration of steem formation from a single debris stream resemble those of Firure 19E.214A,
~
debris droplet, except that the term g in equation (1) must be replaced by -(a+ g). Stability of the frontal surface of (c) Steam Formation, Single Debris Droplet liquid debris in contact with water can be evaluated
=v = 0 in equation (1). It follows The amount of steam formed if all debris droplet by that setting the expe v'dsed shable droplet size formed thermalis energy is transferred to an associated water mass M, at P,is given by r r (3) E'
" M=
(, * (a + g)(p p,) (6) 8 hg(P ,)
- If a mass of debris M caters the pool at velocity Ve, the drag force causes a deceleration where E' is the energy remaining for steam formation after heating M from a subcooled state CdA#,Va to saturation. That is If E is the droplet total a= (4) thermal energy, 2M E' = E E'# (7) where A is the projected area of M and Cd is the drag coefficient. where E,, is the energy required to saturate the water mass, (b) Debris Stream Broadeningin Water E,, = M,e y(Tsat sc) (8)
Equation (4) gives an approximate deceleration of a debris mau entering the pool. If an average The maximum volume of steam formed at ambient debris mass preuure is M=AoLp E' vg(P,) l' = (9) 8 decelerates at hgg(P ,) a = dV/dt = (dV/dy)(dy/dt) = VdV/dy if the steam volume is spherical,its radius is 3 in the pool,its velocity at depth y below the water R, = 3V, /4x (10) e 19E.2 24 Amendtsent 22
ABWR ur6tmas Sandard Plant 4A (d) Thermal Response Time of Corium Droplet expansion, which can be estimated from the Rayleigh equation (Reference 9) for a spherical bubble, l An idealized spherical debris droplet of radius t et temperature T undergoes convection cooling to the RR* + (3/2)(R') = (P b 'I )/# 1 (I4) ambient fluid at a heat transfer rate. where primes indicate der,ivatnes with respect to . time. q = 4xhrt (T
- T,).
Since the pressure of the gas inside the bubble is Assuming uniform droplet internal temperature, the not known it is necessary to introduce additional droplet internal thermal energy relative to its equations for the growth rate of the bubble. The surroundings, rate at which mass enters the bubble may be approximated by E = (4/3)pMc y r (T.T,) 3 (11) h m' = hA p(Tg T,) cxp ( t/Th) (15) 8 is diminished at a rate q, for which where T is given by equation (12). h T T, = (T; T,) exp ( t/Th ) An energy balance for the bubble growth may also where the ti.ie constant th gives the convective time be written: response as , f PVhm# b 8 8
+U=0 (16) r (12)
Th* ##v / 3h Assuming an ideal gas The internal conduction of heat occurs with an l approximate time constant (Reference 10), U = Pbb Y /@l) (17) 7* 2rt /a (13) If the bubble is further assume to be spherical, one O Either 7 or T may control the heat transfer h rate to surrotinding water. Figure 19E.215 gives the may combine equations (16) and (17) to yield
, 3 12nkP R R,+ 41P R = 3m, h (k 1) (18) {
b b I8 conduction and convection response times in terms of droplet radius for convection bounds defined by an Combining this equation with the mass rate enhanced film boiling coefficient of 3.0 times the equation (15) and the Rayleigh bubble equation (14) l Berenson horizontal flat plate value (Reference 11), forms a system of three differential equations in the ar.d with a nucleate boiling coefficient. Debris two dependent variables P and R. These equations droplets at the liquidus temperature of 2600 K with were solved numerically assuming values of the surface waviness are expected to undergo enhanced constants which are typical of a corium steam film boiling heat transfer. Enhancement factors system. The initial conditions and other assumed between 3.0 and 6.0 have been observed at liquid parameter values are shown in Table 19E.217. A l surfaces disturbed by gas bubbling (Reference 12). hydrodynamic time constant The convective response time is seen to be proportional to the droplet radius in Figure 19E.215. 7 = R/R' (19) l It is seen that laternal conduction could become limiting for droplet sizes above 0.2 mm radius if nucleate boiling occurred, and above 10 mm radius if was obtained which is plotted in Figure 19E.215. enhanced film boiling dominated the surface heat transfer. Figure 19E.215 shows that 1 is less than the convective heat transfer response' limes for either (c) Hydrodynamic Response Time nucleate or enhanced film boiling. Therefore,in cases where the heat transfer from the debris A steam bubble formed by a single debris droplet droplets is controlled by convection, the surrounding grows to an equilibrium radius R, at ambient water with a shorter dynamic time response gently pressure. The growt 5. time depends on its rate of expands with the steam bubble without permitting a O 3 high pressure difference to form. It follows that 19t M Amendment
ABWR u^ m ^s Standard PlanL Rev A steam volume formation for the range of debris (12), that is, R /T y it follows that the debris q droplets shown in Figure 19E.215 is primarily droplet radius wIich could promote self triggering i Q determined by the droplet surface heat transfer rate. can be estimated from s (f) Conditions for Self Triggc&g 20 gj3 (hg ) js se trig Self triggering could occur if the mechanical 9p'h [v78(T; T*)) l energy DW released from a molten debris droplet (:4) 4 was sufficient to form additional droplets and mit them with surrounding water. The process of (g) Conditions for a Steam Explosion self. triggering is shown in Figure 19E.216. A debris droplet of radius r rapidly transfers its thermal energy It is assumed that many droplets of molten debris I to an associated water region from which steam is have formed and are in the process of forming a formed. The expanding steam performs a net amount submerged solume of steam, as shown in Figure of work on its surroundings. If part of the expansion 19E.217. The steam fortnation time corresponds to work is sufficient to form one or more debris droplets the convection response time T i equation (12), b and mix them with surrounding water, a propagating since all droplets transfer heat umultaneously. The event could occur, creating the potential for a steam provides an equivalent explosion. The work required to form a debris totalinvolved inertia during steam waterexparmass M,ision. The equation of droplet of radius r is approximately motion for M gcan be written as AWo = 4xar (20) (P P,) A = M y* (25) where o is the surface tension of the liquid. If the A on an work required for mixing a new droplet with where average press M = p'0rk,L. g is y =The ((P solution P for y, basep/2, for surrounding water is conservatively neglected, the which the approximate hydrodpamle)/p L)trekpo condition for triggering is for the overlying poolis O t () AW = AW, (21) 2 2I#LL 7 = (26) P An estimate of the expansion work done by a steam (P P,) I4 bubble which expands to volume V,is given by The average pressure is estimated from AW = (P P,)V, (22) P as P,(2Vg/A L) (27) here P is an average pressure during expansion. The term (P P,) can be approximated from the Rayleigh The total steam volume V g if formed at ambient bubble equation, written as pressure, can be obtained from,pEquation (9) with E'
,a replaced by l P P, = p (RR* g + 3(R ) / 2) (23)
(28) E' total = NE' The bubble wall acceleration, R*, is negative during the expansion. This can be shown from a where N is the total number of debris droplets large amplitude solution to the Rayleigh equation for participating in the steam formation process. It is the sudden appearance of a high pressure bubble possible for a steam explosion occur if the condition l which expands adiabatically (Reference 13), if the term RR is neglected, Equation (23) yields a higher 7 Th I9I l { P P,, resulting in a conservatively high estimate of expansion work. The bubble wall velocity is is satisfied. That is,if water motion is sluggish estimated from the maximum size given by Equation relative to the submerged steam formation, then it is (10) and the convection response time of Equation to high velocity, possible accompaniedto byaccelerate high pressure aM,hd impact.
/3 1911M6 Amendment
i ABWR m. no x Standard Plant 19E.2J.l.4. Applicatloa to ARWR lt was assumed that a debris stream which extended throughout the pool depth was the O Table 19E.217 gives approximate salues of the important parameters, partially explained in Figure participating mass, corresponding to j 19E.218, which were used in evaluating the potential Md = ppRe L = 213 kg ' for an ex vessel steam explosion in the ABWR, j with a total thermal energy of l
+
First, the expected corium droplet sires were found. The debris stream velocity and radius entering E = M d'v(Tdi . T,) = 256 MJ the water pool were obtained as V = 11 m/s, and Ro = 3.7 cm. This then allowed the computation of The mechanical work of steam expansion is the stable droplet tires formed by the debris stream v : falling through air and water: W = P,E g(P,)= 193 MJ - r,;, = 24 mm hg(P,) r = 0.03 mm This indicates that about 7% of the total thermal
**'" energy is converted into mechanical energy. This is flowever, debris stream broadening in the water will far higher that the 1% to 3% range reported in prevent small droplets from forming at the interface. experiments (Reference 5), and is therefore highly l The stream deceleration when entering the water was conservative for assessing the potential for a steam about 178 m/s 8, based on a cylindrical debris mass of explosion.
approximately 3.7 cm radius and an equallength. This yielded droplet sizes of If the participating liquid mass is equivalent to the totalin the lower drywellif the passive flooder were r 2.5 mm somehow to fail open before vessel failure, decd The expected average debris droplet size in the Mg= p Ag Lg= 485,000 kg water corresponds to the instability of deceleration. With this information, the important response times Then the corresponding hydrodynamic response time for bubble grewth were determined: is Th= 9.2 s convection 7p= 038 s 7, = 1.8 s internal conduction The convection heat transfer response time, found previously is Yg= 0.006 s bubble growth, single droplet l Th= 9.2 s That is, steam bubble growth from debris particle energy is limited by the convective heat transfer rate, it iollows that the condhion for a steam explosion given in Equation (29) is not satisfied, even for this l Equation (24) shows that self triggering could bounding case, which employs the highly - occur if a debris droplet radius is greater than conservative 7% thermal energy conversion. 83 mm, and therefore is unlikely in the ABWR for Therefore, the steam explosion potentialin ADWR the expected droplet size of 2.5 mm- is extremely low. A conservative, bounding analysis was considered in which it was assumed that a debris mass in the pool 19E.2JJ 100% Metal Water Reaction l was broken up into small droplets of the expected 2.5 mm radius. The resulting heat transfer and An analysis of the capability of the ABWR to hydrodynamic response times were evaluated withstand 100% fuel clad metal water reaction was I according to the conditions for a steam explosion performed in accordance with 10 CFR 50.34(f). l given in equation (29). Since the system is inerted it the containment l O. atmosphere will not support hydrogen combustion. Amendment 1E3 27 l
ABWR mioux. Standard Plant u Therefore, it la neceuary only to consider static loads total plant risk and therefore do not need to be on the containment. specifically evaluated further in the PRA. A simple analysis was performed to detertree the (1) Dennition of Supptenion Pool Bypass effect of the added hydrogen mass and heat energy anociated with 100% fuel. clad metal water reaction. Suppreulon Pool Bypass is defined as the Since the design basis accident for peak contaiament transport of fission products through pathways which p" wure is a large break LOCA, this accident was do not include the suppression pool. In such cases, ch sen as the basis for the analysis. the scrubbing action for fission product retention is lost and the potentialconsequences of the release in order to simplify the analysis several arc higher. conservative suumptions were made. Since it is not possible to release the hydrogen before the first The potential for suppression pool bypass has 1 pressure peak, caly the second peak is considered, been a subject of analysis since the early days of. The hydrogen is distributed in the same manner as WAsil 1400 (Reference 7). The *V" sequence which l the nitrogen. All of the metal water reaction heat represented a break of the low pressure line outside - energy is auumed to be absorbed by the supprenion of the primary containment was one of the more pod "/atet. Finally, no credit was taken for the dominant release sequences in WASit 1400, The dri n 1 'd wetwell heat sinks. IDCOR analysis and BMI.2104 also reviewed sequences in which the suppression pool scrubbing onsideration of 100% fuel clad metal water action was not obtained in the release pathway. reaction results in a peak pressure of about 75 psig. The governing service level C (for steel portions not in order to review the importance of suppreulon backed by concrete)/ factored load category (for pool bypass pathways, the potential mechanisms, concrete portions including steellinct) pressure probabilities and source locations were reviewed to capability of the contairiment structure is 77 psig identify where fission products might be r: leased which is the internal preuvre required to cause the outside of the containment._ The analysis he conser. 4 O maximum stress intensity in the steel drywell head to reach general membrane yielding according to serdce level C limits of ASME.III, Division 1 Subarcticle vatively focused on the station blackout event because it leads to a higher likelihood of suppression pool bypass and because it is considered one of the NE.3220. Therefore, the ABWR is able to withstand more probable initiating events for core damage ' 100% fuel clad metal water reaction as required by sequences. 10 CFR 5034(f). The principle conclusion of the review is that, with the exception of certain lines addressed in 19E.2J.3 Suppnasion Pool Bypass Paths containment event trees of the PRA, suppression pool bypass pathways do not contribute significantly 19E.2.3.3.1 Introductica to risk. Consequently, the probabilistic risk assessment does not require a separate evaluation of This section reviews the potential risk of certain bypass sequences, unless the sequences develop suppression pool bypau paths and demonstrates that, during the course of an event, for example, as a with the exception of the wetwell drywell vacuum result oflow suppression pool water level. Such breakers, and certain other lines, bypasa paths present cases are considered in Section 19D.5.7. no significant risk following severe accidents. Because of this insignificance, only the vacuum Nevertheless, certain bypass lines which result breakers and the other lines require further from piping failures outside of the primary consideration in the ABWR PRA. The approach containment are lo2ded in this redew in order to used in this evaluation is similar to that submitted tc. assess their significance. l the NRC in support of the GESSAR (Reference 14) review. (2) Mechanisms for Suppression Poul Bypass - The results of the evaluation is that bypass lines - All tines which originate in the reactor vessel or the primary containment are required by sections of A evaluated contribute no more than almut 10% of the 10CFR50 to meet certain requirements for contain, V ment isolation. Lices which originate in the reactor 19tl.24a
? Amendment
i
- ABM Standard Plant 88^*"5^s km A I
f I vessel or the containment are required by General . Design Criteria 55 and 56 to have dual barrier , , proteetion wbleb is senerally obtsined by redundant
- isolation valves Lines which are considered .
i non essentialin mitlgating an accident are also ] required to automatically isolate in response to j diverse isolation signals. Other lines which may be i usefulin mitigating an accident are considered i I
- l l
4 1 4 i 4 0 O __m ,, _ , 4
-,_,,,,,,x_.-,.y-,,+ , . , , , - - ,.y._,m...y,,,,.nmm,-_,.y,-.4,--_,.1,__. , - . _ _ . - - , _ - , - , , . . - . , , - . . ~ . . , - - . . y . . - - . , - - - - - . . . - - - - <
ABWR nu ms Signdard Plant Rd exceptions to the General Design Criteria (NUitEG pathways will be insignificant, OMO, Section 6.2.4) and are permitted to have remote (c manualisolation valves, provided that a rneans is available to detect leakage or breaks in these lines The justification for this approach is as follows: outside of the primary containment. Itisk = Total [ Event Frequency x Consequence) (30) A potential mechanism for suppression pool =F xC nbp + Fbp xC (31) nbp bp bypass is the *Ex. containment LOCA* which results
- I from the combined f ailure of a line outside of the where
- F*P = The total core damage primary containment along with the failure of its frequency of non bypass redundant isolation valves to close. If this events combination of events occurs, the operator is made aware of the situation throug,h leakage detection C = The consequence of a non-
"bP alarms and is instructed by plant procedures to bypass event manually isolate the lines,if possible, when the sump
, water levelin areas outside containment exceeds a F = The total core damage bP frequency by bypass events predetermined point. which are equivalent to a Because of these provisions the probability of complete bypass of the suppression pool bypass occurring from the suppression pool
*Ex co.ntainment LOCA* is extremely small since it requires the simultaneous failures of a piping system, C = The consequence of a pP redundant and electrically separate isolation valves complete bypass event and the failure of the operator to tak. action.
Subsection 19E.2.3.3.4 summarires an evaluation of If the total bypass risk is to be insignificant, the the core damage frequency from ex container. tat last term in equation (31) must be much less than the LOCAs. first, or: O V The plant design criteria ensure a highly reliable system for containment isolation. Nevertheless, even F bP C nbP (32) 4 though there is diversity in the types of valves, all F C types have experienced failures at operating nuclear "t'E bP plants and certain events, such as station blackout The total bypass and non bypass event frequencies event, may make the early isolation of some lines (F) noted above are the total core damage impossible. This section evaluates the significance of frequencies for the,e events assuming that all events bypass paths in order to justify that no sdditional have the same consequence. Since this is seldom the treatment in the PRA is neccessary, case, the bypass frequency must be defined such that the proper consequence is applied. This is accom-plished through evaluation of flow split fractions (f) as discussed below. The total bypass frequency can be expressed as: (3) Methodology for Evaluation of Suppression Pool Bypass F bp
= Fg x IP g (33) l The evaluation of suppression pool bypass pathways is based on a methodology which evaluates where: F = The total core damage fre.
ed the potential relative increase in offsite consequence quency from bypass events over those events with suppression pool scrubbing. Then, knowing this P *b . - The total conditional proba, amount of increase, if it can be shown that the E' bility of full suppression pool probabiSty of bypass is sufficiently low as to offset the bypass path i, given a core increased consequence, the added risk from these damage event, en\ Amendment 19ti.2 29
ABWR nuimas Standard Plant Ret A The conditional probability of full bypass can be if equation (36)is satisGed. then the total bypass risk further refined by the expression: is insignificant. P gP;
- P ; x f; (34) (4) Criteria for Exclusion of Dypass Sequences in the PRA where: f. - The fraction of fission products gener-
' As noted previously,if it can be shown that the ated duiing a core damage event which pass through line i (subsection probability of bypass is suf0cicut 8% as to offset 19E23.3.3 (1) discusses this term in the increased consequence, the risk resulting from more detail) release throue.h bypass pathways will be insignificant.
The Sw split fraction (f) is defined as To establish a threshold for this frequency, the the ratio of the flow rate which paues consequence ratio (right side of equation %) was out of the bypass pathway to the total evaluated using the MAAP 38 AllWR and CRAC flow rate of aerosols generated during codes to establish the approximate order of magni-the core melt process. The line flow tude for evaluation purposes. To establish a split reduces the consequence associ- threshold for this frequency, the consequence ratio ated with smaller lines due to inherent [right side of equation (M)] was evaluated using the flow restrictions in those lines as com- MAAP 3D ABWR and CRAC codes to establish the pared with the consequence of larger approximate order of magnitude for evaluation lines. The flow split fraction accounts purposes, for this consequence reduction by re-ducing the equivalent bypass probabil- For non bypass case, the offsite dose from normal ity. containment leakage following core damage was used as a basis. 'NClf, described in Appendix 19P, and P .= The conditional ptobability of bypass in is the consequence from normal containment g'P'
, line i (Section 19E2.3.3.3 (2) discusses leakage;' Case 7' may be used as an approximation this term in more detail). of the full suppression pool bypass consequence.
The conditional probability of bypass is The corresponding ratio based on values in Table established through a detailed evalua- 19P.211s 8.4E 4 which can be used in the evaluation tion of each potential bypass pathway, of pool bypass significance. Further evaluation of establishing the failure which must 'Ex containment LOCA* suppression pool bypass occur for a bypass path to develop and paths in the PRA is not necessary if it can be shown assigning a probability to that failure. that the total bypass probability is significantly less than this consequence ra'!o. Core damage events result in essentially two types of release: releases which bypass the suppression 19E.2.3.3.2 Identification and Description of pool and those that do not. With this simplification, Suppression Pool Bypass Pathways the total non bypass frequency can also be defm' ed as: Identification of the potential suppression pool F (35) bypass pathways was based on information in the gP =Pcd PbP ABWR Standard Safety Analysis Report and Inserting equations (33), (34) and (35) into equation supporting piping and instrument diagrams. The (32))ields: potential pathways are shown in matrix form in Table 19E.218. P C nbp/Cbp (36) bpi x f.i < < Table 19E.21 summarites the results of reviewing Assuming Fg is much less than F which would the ABWR design for lines which are potential cd be consistent Tith the basis for containment pathways. For each line the table provides the line isdolation. sizes, pathways and type of isolation up to the second I isolation valve. The bypass lines identified in Table y 19E.21 were derived from a systematic review of the ) 1 ABWR P&lDs and other drawings. ! V Amendment 19ENO
)
ABWR imius Standard Plant hA Several lines in Table 19E.21 were excluded from Wa k the vent flow rate in a single line (SRY further consideration on the basis of a variety of or drywell vent) which passes to the l O judgements discussed in the table notes. In general, the exclusion was based on deterministic rather than suppression pool probabilistic arguments, For instance, the RWCU n = the number of flow paths to the suppres. i return line to feedwater and LPFL Loop A were sion pool included in Table 19E.21 and excluded from further analysis because the bypass path is protected by the This can be simplified into the form: j feedwater check valves. I = f/1+f (38) l l l The remaining lines are considered potential l sources for significant fission product release where f'= i following severe accidents. Although the probability W)/nWk [ I that these lines could release a significant amount of From the formula for turbulent compressible Duld now l fission products is extremly small, they are reviewed (Reference 15) I further in Subsection 19E.2.3.3.3 to assess the 2 importance of these releases. W = 1891 Yd [(dP)/KV]t/2 {39) where W = j or k(ib/ht) l Y = Expansion factor d = Internal diameter (in) l l 19E.2J,3J Esaluation of Hypass Probability (dP) = Differential pressure (psid)- K = Resistance coefficient = f"L/D + K' Equation (36) of Section 19E2.3.3.1 establishes f" = friction factor the need for evaluation of the now splits and failure L/D = pipe length to diameter ratio, including probability for each line not cxcluded in corrections for valves, bends , Table 19E.21. This section prosides the basis for the K' = additional factors for entrance and exit evaluation of each of these factors. effects V = Specific volume of fluid (cf/lb) (1) Evaluation of Bypass Flow Split Fraction (f,) To assess the fraction of aerosol release which. . bypass the suppression pool a flow split fraction is f'= 1891 Y.d.2[dp/KV]I/2 /1891 nyk d k [dp/KVlII needed, the now split fraction (f) is defined as the 1J 2 ratio of the flav
- rate which passes out of a bypass = Y)d /nY dg k [dp/K]1/2 j [dp/K]1/2 (40) pathway to the total flow rate of aerosols generated during the core melt process. Two generalized bypass Equation (40) may be rearranged to show; paths have been evaluated: 1) a path from the RPV which passes to the reactor building with the f' = (1/n)[Y)/Yk][d;/d k] x[dP)/dPk I remainder passing to the suppression pool through the SRVs and 2) a path from the drywell to the [Kk/Kj I ,
3 I4II reactor building with the remainder passing to the suppression pool through the drywell vents. The expressions in equation (41) were evaluated riu. merically for the actualline configurations to arrise at The flow split fraction may be defined as: the flow split fractions used. The following assumptions were made in this analysis: (37) 1. Containment pressure following the core melt is f = W)/W) + nWk assumed to be at an average of 45 psig during the post core melt period. Although the containment pressure could evectually increase to a higher icsci, where W) = the flow rate which passes through the bypass pathway the average is used to assess the total amount of release since a release would be occurring
- throughout this period. This pressure is typical of Amendment 19fi.241 i
l
! ABWR nuou j Standard Plant MA ! those calculated in severe accident analyses (see the drywell sources, the path to the suppression pool j Figures 19E.2 2 through 19E.212). is estimated Io be 5 ft. (1.5 M) ,
- 2. Prior to RPV melt.through, the reactor pressure Other values used in the calculation are listed '
l j vessel (RPV)is maintained at a relativelylow below.
! pressure (100 psig) by the automatic depreuurim tion system or equivalent manual operator action. Parameter Assumed Value Rasis 1
! Four ten inch safety relief valves (ADS valves) are i conservatively assumed to be open to release RPV Resistance Coefficient (K= P'L/D) {
{ cf0uent to the suppression pool. This is consistent ;
) with the minimum instructions in the EPGs. Ten Friction Factor .011 to .018 Ref 14 (pg A 25) I 24 inch drywell vent paths are consistent with the Site dependent) l ! ABWR design configuration. For conservatism i j the vents ate assumed Io be one quarter Line Diameter (D) Various Line s!re (see i uncovered. Table 19E.21) i 1
j 3. The pressure drop in the bypau path between the Other Resistances (K) Ref 14 (pg A 30) fluion product source and the release point is a j function of whether the line produces sonic or Gate valve 13 sub sonic velocities. For RPV sources, an average Check valve 135 l 1 100 psig internal RPV pressure is assumed during Globe valve 340 { the core melt process. This is based on an Entrance effects .5 1 average 45 psig drywell pressure and an anumed Exit effects 1.0 l SRV design which closes the SRV when a differ. ! ential pressure of about 50 psid exists between the Expansion Factor (Y) .6 to .9 Ref 14 (pg A022) main steamline an the SRV discharge line. (dP, K dep.) {, j Depressurization of the RPV or contalament Table 19E.219 shows sample results (f
- from equa.
j . through the bypass path is not considered. The tion 41) for a line with two motor operated valves. In auumption is made that pressure is continuously the evaluation of individual bypass lines the actual generated during the severe accident in sufficient configuration is used. The evaluation of flow split
- quantity to uncover the SRV discharge or drywell fractions is considered to be conservative for several i vents. reasons
- 4. The pressure from in the non bypass path be. (a) Bypass release paths would normally be expected
- tween the fission product source and the suppres. to be more restricted than evaluated due to sion pool release point depends on the suppres. smaller lines, more valves and pipe bends, valves l
j slon poollevel. The suppression poollevelis being partially closed or pipe breaks being assumed to be higher than normal because of the smaller than the piping diameter. l ' ! depressurization of the RPV to the Suppression l pool through the SRVs. For RPV sources, the (b) No credit is taken for additional retention of j SRVs experience about a 20 foot (6.0M) elevation fission products in the reactor building, in piping
. head over the SRVs during the core melt process. or through radioactive decay.
For drywell sources a 15 foot (4.5Y) elevation
, head is experienced over the upper horizontal (c) For drywell sources, a higher than analyzed , vent. For the station blackout sequence, the effect differential pressure should exist between Ihe !. of ECCS system operation on suppression pool drywell and wetwell. This will icad to lower flows level has been ignored. through the bypass path.
- 5. The length of lines discharging to the suppression pool and through the bypass paths affects the resistance coefficient Equation (39). Based on the 1 ABWR arrangement drawings this length is estimated to be approximately 85 ft. (25 M). For
! Amendment 19E.2 32 l f ,
.--.wv-~,- -.mw-.-.<..www-*-- 4 u---- - - + - - + . . . , , , . .-- - - - .,-s.*_,,+-n.--- - . , - - - ~ -- .-v2.-o.o . ----...----+.---w--- - . - ~
ABWR 2mims Standard Plant ,,, 3 (2) Evaluation of Failure Probabilities (P ) U The failure probabilities used for the detailed calculation of the bypass probabilities are summariicd in Table 19E.2 20. The bases for these probabilities are provided below-(a) Current operating plant htSIV failure to close probability is about 4E 3/ demand with a common mode failure probability (P;) of about 1E-4/de-mand For this evaluation the common mode l failure probability of IE 4 is assumed for failure of both valves in a single line to close. l (b) Current operating plants evaluate htSIV leakage against a leakage requirement of 11.5 SCFil per valve. About 50% of the valves typically fail this localleak rate test at this level and about 10% are believed to typically exceed the 640 SCFil level allowed by ADWR proposed technical specifica. tions. The leakage probability (P2) used in this analysis was based on three leakage groups: Probability Group Leakage Per Valve Estlins G1 < 11.5 scfh .5 .5 G2 11.5 to 640 scfh .4 ,2 ( N C3 > 640 scfh .1 .01
,O 19E.242.1 Amendment
ABWR m. u Sinndard Plant The MSIV leakage probability (P2) is assigned a among check valves was considered for lines b V value of .71 to correspond to the total line leak. age probability. Flow split fractions were deter. containing redundant series check valves. Only Fced water and the SLC paths contain more than mined for each of the groups and a weighted one check valve. For these lines a lleta factor of aserage flow split fraction (weighted by the line .18 was used for the failure of the second vahe, leakage probabilities) was determined for use in the evaluation. (h) When power is available, some normally closed valves open during an event in response to an (c) The probability of flow paning to the main con. injection signal, even though the actualinjection denser is judged to be governed by the failure of fails (a requirement for a core damage to the bypass valve to close. This probability (P3) occur). )- is taken at 4E 3 from Reference 16. Once flow l The probability that ECCS valves are not closed passes to the main condeuner, the condenser is assumed to fall (P4) via the relatively low by an operator (P10) is considered remote positive pressure rupture disks. during a severe accident. A value of 0.5 is judged reasonable especialy considering the (d) The main steamline 1,reak probability (PS) was potential for room environment degradation. line break probability (P15). For station blackout events, since the valves do l not open, these lines do not contribute to (c) Normally open pneumatic (P6) and DC motor potential bypast risk, operated valves (P7) have failed to close. Causes include improper setting of Iorque (i) Some normally closed valves may be open at the switches leading to valve stem f ailure, beginning of the event. The failure probability undetected valve operator failures and improper (P11) for these valves assumes they are open 4 packing materials or lubricants. GE has issued hours during a 7000 hour operating cycle and several service information letters on valve that the operator fails to recognize the open problems and recommended actions to prevent path and close the vale. A 0.5 probability is recurrence of the failures. The industry failure judged reasonable for the operators failure to t v rates for motor operated valves is about act during the core damage event. 3.6E 3/ demand and 4.1E 3 for air operated valves. These failure rates are not significantly (j) Some valves may be opcaed by the operator affected by the vahe environments. A common during the course of the event. Such action may cause failure among air operated valves was be in compliance with written procedures or it considered for lines containimg redundant series may occur due to confusion in following a procc-valves. For these lines a !! eta factor of.18 was dure. The probability that valves are inadvert-used for the failure of the second valve. ently opened (P12) is considered a violation of planned procedures. A value of 1E 3 is jadged (f) AC solenoid and motor operated valves are reasonable during a core damage event. subject to a common mode failure (P8) if motive power is unavailable such as dusing a Station (k) Pipe rupture is extremely rare in stainless steel Blackout event. For station blackout everts piping. Ilowever, carbon steel piping has been these valves will have a conditional failure observed to fail under certain conditions. The probability of 1.0. For this analysis a failure frequency of these failures has been widely probability of 1.0 was conservatively assumed. studied and shown to be in the range of IE 7 events / year. The probabilities ofline rupture as (g) Check valves have been observed to failin such a function of line size (P13, P14, P15) are taken a way as to permit full reverse flow, a condition from Reference 14. Four line segments outside I necessary to permit suppression pool bypass for of the containment are assumed for each bypass some lines. Maintenance errors associated with line. The intermediate line size (3 to 6 inches) testable check valves have also been observed. probability is assumed to be twice that of the The industry failure rates for ebeck valves large line size (greater than 6 inches), allowing complete reverse flow (P9), based on r3 7000 hours of operation per operating cycle,is For pipe failures in an ladividual bypass line, it V l about 8.4E 3 per cycle. A common cause failure was presumed that an undetected break in an 19C M Amendment
ABWR ums Standard Plant m unpressurized line could occur at any tinie, both slows the break flow and terminates any long Therefore, the conditional probability of a term release from the break. Therefore if the EPO O bypass path was then taken to be the same as the failure rate during a one year period (whlch actions are taken, no additional consequence of the event occur, was estimated to be 7000 hours). This approach of estimating pipe failure probability h judged to The system arrangement routes the RWCU lines be conservative, above the core to avoid a potential siphon of the core inventory. In the event of an unisolated RWCU line The failure probabilities used ln the evaluation break, lowering the RPV level to below the should be considered conditional probabilities, given shutdown cooling suction and depressurizing the a core melt. In general the above probabilities are RPV would be sufficient to terminate the break flow not affected by the core melt process itself and can without causing core damage. This action should be therefore be considered independent of the event possible prior to any impact on other ECCS process. equipment.These actions are included in Section 19D.7. Whether the bypass path is the initiator or occurs simultaneously with the event is inconseq. (3) Evaluation of Bypass Probability ventialin the evaluation based on the following discussion. The approach taken in the bypass study Table 19E.2 21 summarizes the results of these is to consider the presence of a bypass path as an evaluations. For each potential bypass pathway,it independent event from the events which caused the shows the flow split fraction based on the line site core damage in a speci1c sequence. This approach is and valve configuration, the equation to calculate the acceptable because for large breaks the associated systems are not in general relied upon to prevent bypass probability, calcule.tlons using the datathe fromresultsTable 19E.2 of the probability 20 and l core damage and no consequence of these failures . the bypass fraction for the line. The table also have been identified which would affect the systems includes reference to the sketch (Figure 19E.219) preventing core damage. Therefore whether the which illustrates the potential pathways. The evalua. break is an initiator or consequetitial does not affect tion is based on the conservative assumption of a O the final evaluation. Similarly, none of the systems associated with the smaller bypass lines are station blackout event since it is believed to be the dominant core damage sequence and gives the high. associated with preventing core damage. Therefore est bypass fractions. they too are not associated with the cause of the core melt. The ACRS has expressed concern regarding the failure of the RWCU suction in combination with failure of the isolation valves to close. The concern is that there may be a flooding situation that could have a high consequence if it leads to an eventual loss of suppression pool and CST inventory or flooding of other ECCS rooms. Such an ewnt would not be consistent with this presumed independence of 6e assumed conditional probabilities, if a break in the RWCU suction line were the postulated 1,0CA, the containment isolation valves would be expected to close, terminating the event. NRC concerns over Motor Operated Valve (MOV) closure capability are being addressed as an industry actisity.- In this evaluation it was assumed that the valves fail to close due to a Station Blackout event. Furthermore, should the isolation valves fail to close, the system arrangement assures that the core is not uncovered and EPGs require depressurization which Amendment 19E.2-331
- _ - - . -_- . _ - - _ _ . -_ - 1
f ABWR mas StandardEant Rey _A (4) Evaluation of Results pipe break frequency is prosided in Appendis 19E.2333 (2)(L). Section 19E2.33.1 (4) provides a conscrsative justification th.t bypass paths with a total bypass X 3Line isolation The conditional probability of I fraction less than 8.4E-4 do not substantially increase automatic isolation valves failing to close given the offsite risk. As is shown in Table 19E.2 21, the the ex containment LOCA. Values used and bypass probability is about 3.9E 5 for all potential the manner in which probabilities were . paths not addressed in the Containment event trees. comoined are shown on Table 19E.2 21. l This totalis well within the goal. PgOper. Action The conditional probability that Potential bypass through the Wetwell Drywell operator fails to act to manually isolate the Vacuum Breakers and the inerting lines are included ex containment LOCA. Such a failure to act in the containment event trees. (Section 19D.5). could be due to a lack of instrumentation availability or mechanical failure. For most ' I Based on the above discussion,it can be bypass paths considered, the very conservative concluded that suppression pool bypass paths and assumption was made that no operator action is i Ex Containment LOCAs not addressed by the taken. For ECCS discharge lines and warmup Containment event trees do not contribute a lines the operator is assumed to act to clost; an significant offsite risk and do not need further open valve. if needed. The basis for the value evaluation in the PRA. chosen (P10 in Section 19E.233 (2)) is based on general operator awareness of the potential for 19E.2JJ.4 Evaluation of Ex containment LOCA these paths to be unisolated. Although the leak Core Damage Fregoency detection system is adequate to alert the operator of a break in the system, (1) Introduction lastrumentation failure is not considered to provide a strong contribution to the failure To provide a separate assessment of the probability. O importance of bypass paths, a more comprehensive V analysis of the frequency of core damage from LOCAs outside containment was conducted using O gSecond Dhision not Affected For most lines it is conservatively assumed that the LOCA affects event tree and fault tree techniques. the division in which the break occurs. This factor represents the conditional probability that Conservative and simplified event trees of the LOCA also affects the required makeup for LOCA outside containment events were developed core woling fron' a second electrical dhision. It and included as Figures 19E.2 20s through is assumed that such failure results from 19E.2 20c. These trees show that the total core environmental effects from flooding or damage frequency due to LOCAs outside of pressurization effects. containment is about 13E-8 per year. The end point for these trees is core damage with or without bypass A systematic evaluation of potential cold of the containment. flooding due to ex containment breaks was summarized in Appendix 19R, Probabilistic (2) Assumptions Flooding Analysis. Flooding in the reactor building is notrd to disable the system affected The following definitions and considerations were and potentially flood the Reactor Building applied in development of the trees. corridor, but not disable other makeup equipment due to the water tight doors V1 Line Break Outside The frequencyof piping contained in the design. The analysis of an breaks in small, medium or large breaks outside unisolated RWCU break in subsection 1911.4.5 of containment and which communicate directly shows that no cooling systems will be damaged. with the reactor vessel. The lines are grouped by type of isolation. The basis for each event Compartment pressurization nd emironmental initiation frequency is the line size and the total effects of high pressure LOCAs in secondary nember of tines considered. The basis for the containment were considered in the development of Figures 19E.2 20a through c.
%.)
Amendment 24 191L2 M
ABWR mas Sluttdjlrd Plant Rm A Equipment in the ABWR design is arranged For LOCAs which occur in the reactor building, with consideratio of divisional separation. A the event is auumed to fail the division in which high energy line brerA in a dhision would cause the break occurs. For other LOCAs, such as the blowout panels from the division to relieve LOCAs in the turbine building, no divisional the initial pressure spike to the t, team tunnel. impact is assumed. Subsequent pressurization of the room could egentually cause a release cf the energy into the Consideration of inventory depletion due to the next adjacent division in a clockwhe progrenion LOCA outside containment is addressed by EpGs through the reactor building. which specify that coolant makeup sources using inventory sources outside of containment be used As doors from the corridor and penetrations are as the preferred source, in the ABWR design forced open, the environment of the adjacent small breaks can be accommodated by any of the divisions could be affected by the presence of high pressure coolant makeup systems (RCIC, steam. However, the qualification of the HPCF 11 and ilPCF C) which are in separate equipment to 212 degrees F and 100% humidity divisions and which diaw water from the makes the probability of further system condensate storage. Since condensate is unavailabihty unlikely. Where a LOCA could effectively an unlimited supply and makeup occur in an area adjacent to a separate division, capability exists, no additional conc;rn is a value of IE 3 was assumed for 0,, based on necessary for the small break LOCAs outside of conservative engineering judgdment, to containment, represent the remote pouibility for failurc of these adjacent systems. Medium and large breaks outside of containment can be accommodated by any of the three i For line breaks in the turbine building the effect divisions in the short term following a break of the break would not impact the divisional without concern for inventory loss in the RPV. power distribution and, for these sequences, the All penetrations, except the RPY/RWCU bottom O gvalue was judged to be negligible, head drain (a unique situation addressed separately in Section 19.9.1 by an event specific Q(~~N Altbough line routing ate not specified, the analysis assumes that breaks inside reactor procedure), are above the top of active fuel so that core uncovery due to inventory depletion is not a building equipment rooms affect the dhision in concern. In the longer term, tt e break will which the breaks occur; LOCAs outside of the depressurize the RPV which effectively reduces secondary containment are not assumed to fail a the loss of inventory from the break to a level well division of equipment. within the makeup capacity of other availhole systems which makeup from sources outside of Q Coolant Makeup This factor represents the containment, such as firewater. Due to the conditional probability of core cooling failure by reduction in loss rate through the break, all sources of cooling with consideration to those significant time is available for operators to affected by the ex containment LOCA. The compensate for the usage of wat. sd flooding in values used are derived from an evaluation of the affected area. Furthermore, ., :stors are the PRA fault trees and are summarized below: assumed to follow plant procedures in isolating the break or lowering RPV level to a level below COOLANT MAKEUP FAILURE (Og) the affected penettation,if neceuary. Adequate BREAK SIZE instrumentation and long term makeup from Small Medium Largt firewater and condensate sources would normally Div. not Affected 2.2E 7 6.2E 7 6.1E 7 be available. 1 Div. affected 1.1E 6 8.6E-6 8.5E 6 2 Div. affected 3.6E 4 3.7E 3 3.7E 3 (3) Conclusion The conditional probribility when one or more For each 3f the event trees shown in Fi6ures electrical divisions are affected were derived by 19E.210a through c the total non bypass and bypass disabling the most limiting dhision in the LOCA core damage frequencies are shown and are event trees and then calculating the resulting summarized below; b) V conditional probability. Artndment N 19E2W t
4 i ABN Standard Plant 2mi As m ,1 l Core Damage Frequency (events /p) Non-Bvnnas Bvoans Total - j ) 1 V Small LOCAs 1.268 1.1D9 1.3E-8 !' Intermediate LOCAs 2.3610 126t0 35D10 j Large LOCAs 20610 4BE-13 20E-10 j TOTAL 1.268 1.259 1.3D8 3 j Ex containment LOCA events without bypass j represent a small fraction of the total core
- damage frequency (1.6E 7) are therefore j justMied as not being further evaluated in the
, PRA. Althoagh the consequence from bypass events is )- greater than for non bypass events, the total l
- frequency of bypass events concurrent with core j damage is extremely small. The core damage i frequency of ex containment 1 OCAs with j bypass is less than 1% of the total evaluated j core damage frequency. Large LOCAs can be j excluded from iurther consideration o the basis 1 of low probability. Exclusion of Medium and
{ Small bypass sequences is based on the i additional consideration of the reductions in 2 consequences of the ex containment LOCAs I due to the flow splits provided .5 v restrictions i dJe to line sizing. This is discussed in Section
- 3 19E.2333.
4
- in addition, since significant margin exists i between the current PRA results and the safety j goals. it can be concluded that the bypass events i do not significantly contribute to the offsite exposure risk.
l 1 ! 19E 2335 Suppression Pool Bypass Resulting from i External Events t-i The effect of external events on the Suppression Pool Bypass evaluation is dtscussed in Appendix 191 to determine if a significant potential for bypass'mg j the suppression pool results from component failures induced by a seismic event. Only seismic events were 3 considered to provide a significant challeone to the i creation of bypass wths beyond that already j considered in the PRA. i i 5 4 4 l l 19E.2-N 2 I Amendment i-
ABWR 2-mas Standard Plant m., A wall resulting from a postulated break of the RHR Q pump discharge pipe were estimated using applicable test data. This wall runs parallel with the discharge V pipe at a distance of approximately 4 ft. The length of this wall is 43 ft, the height is 20 ft. The RHR pump discharge piping is assumed to run 2 ft above the equipment room, with the rupture located exactly opposite to the middle of the wall (worst case). 19E.2.3.4 ENect of RilR lleat Exchanger Failure in a Seismic Esent The dynamic loads result from the discharge of the containment atmosphere through the broken I A failure of the RHR heat exchanger mounting, pipe into the wa"r gool in the RilR equipment can conservatively be postulated to shear the pipe room. It was c r4 atively assumed that the entire between the RHR pump diseharge and the RHR heat volume of the es tt room was flooded with the exchanger. About 30 minutes is available for the suppression pool aer. operator to close the RHR suction valve to the i l suppression pool. If no power is available, or if the The gas discharged from the broken pipe will be 4 operator failed to close the suction valve (s), the initially almost pure nitrogen, later a mixture of
- suppression pool will drain to the RHR equipment nitrogen and steam with decreasing nitrogen content, rooms. and finally, after all the nitrogen is purged out of the containment, pure steam. The mean flow rates through the broken pipe will be a function of This subsection describes the analysis of these pressure in the containment, which in turn will sequences which concludes that structural integrity of initially depend on the accident scenario. In the long the RHR equipment room will be re .dnad s
and that, term, however, the mass flow rate will be driven by in effect, the suppression pool scrubbig a transferred the steam generated from the decay heat. Itis from the suppression pool to the Riit? equipment assumed that there will be no pressurization of any ' O rooms, airspace remaining in the RHR equipment room. i V 19E.2.3.4.1 RilR Equipment Room Flooding This situation is similar to the discharge of the drywell atmosphere through the drywell vents into The RHR equipment room drains to a sump the suppression pool during a LOCA. The test room below. This sump room also receives drains results from LOCA tests conducted by GE for a wide from the HPCF equipment room (in two cases) and range of break sizes demonstrate that the highest from the RCIC room (in one case). The design of wetwell pressure loads due to this discharge are these drains is assumed to have a device to prevent experienced late in the event during the
- chugging
- one sump from filling and backflowing up to the regime charagterized by low mass fluxes HPCF or RCIC rooms. The device is assumed to (< 10 lbm/s.'* ) and high steam / air ratios function without AC power. This prevents the loss of (<1% air). At hgher mass fluxes the condensation HPCF or RCIC resulting from RHR equipment room oscillation regime) and higher air contents, the loads flooding. were substantially lower.
The analysis of the resulting loads in the RHR To estimate the chugging loads on the RHR room equipment and the basis for concluding that the room wall, the Mark Ill PSTF test data were used. The will remain intact is described in the following Mark 111 data were chosen because of the horizontal paragrapls. orientation of the ven's and because no pressurization of the airspace above suppression pool 19E.2.3.4.2 Dynamic Loads Induced by Chugging which approximates the situation in the ABWR RHR room. The highest chugging loads on the wall The dynamic loads on the RHR equipment room seen during the Mark Ill experiments were 100 psi. These pressures were observed on the drywell wall adjacent to the vent exit into the pool. Because of g the close proximity of the pressure sensor to the
] Amendmcet 24 1911205
r ABM 23xsiens, m Standard Plant source of the pressure disturbance (the collapsing which leads to the beat exchanger mounting failure causing the postulated room flooding. No structural O steam bubble) this pressure can be considered to be h the actualbubble pressure, damage is predicted, although some concrete cracking is inevitable, After the earthquake, the wall The period between the pressure spikes was would be structurally sound to withstand the loads typically 1 to 5 seconds or more. Following the peak imposed by flooding as described below. pressure spike, a ser!:s of lower amplitude pressure oscillations were observed, with frequencies that The seismic. induced flood imposes loadings to the were in the range of the natural frequencies of the room in the form of hydrostatic and hydrodynamic vents and water pool. The maximum arnplitude of pressures. It is assumed that no damaging these oscillations was typically less than 10% of the aftershocks would occur during flood. From the maximum pressure spike. above discussion the most significant hydrodynamic load is caused by chugging. The pressure transient Given the RHR equipment room geometry, and on the wallis idealized by a sharp pressure spike using a conservative pressure attenuation model with a maximum amplitude of about 4 psig preceded (supported by the Mark !!! experimental data), it by a half cycle sinusoidal and followed by a decay was calculated that the peak, spatially averaged, sinusoidal with much smaller amplitudea. dynamic wall pressure will be below 4 psi, if the maximum bubble pressure of 100 psiis assumed. To find the dynamic effect on the wall response, With higher flowrates and higher non condensable the wall was modeled as an equivalent single degree contents in the discharge, the loads are expected to of freedom system subjected to the pressure be lower. Therefore, this conclusion should also transient described above. The results show that the cover a range of severe accidents during which maximum dynamic amplification factor is 0.26. The non condensable gases (e.g., H , CO2 ) are equivalent static chugging pressure is thus about 1 2 generated from metal water reaction and/or psig. corium-concrete interaction. Under the combined hydrostatic pressures of a 19E.2.3.4.3 RHR Equipment Room Structural fully flooded condition and equivalent static chugging Integrity pressure uniformly distributed over the entire wall, the stress analysis *was performed by treating the wall The structuralintegrity of the RHR equipment as a flat plate with fixed supports along the edges, room structure was evaluated for the loads resulting The resulting maximum moment is found to be from the seismic induced flood. The RHR room is about 56% of the ultimate moment capacity in located at the reactor building basemat level in ea:h accordance with the ultimate strength design method of the three divisions. The most critical wallin a for reinforced concrete. The maximum shear stress typical room was chosen for this investigation. This is within the ACI.349 code allowable. The leaktight wall runs along the 90 270" direction of the reactor RHR room access door was also evaluated and is building and connects to the exterior wall and the found to be structurally sound against flood loadings, containment at both sides. The wallis approximately 13 m (43.64 ft) wide,6.5 m (21.32 ft) tall, and 0.5 m la summary, the structuralintegrity of the RHR (1.64 ft) thick. room can be maintained for the seismic induced flood. This ensures that fission products can be The wall was examined for its ability to withstand scrubbed by the entrapped water. a 2g earthquake which is more severe than that O V Amendment 22 19E.2 X
'~
, 4 ABWR m einois Standard Plant REV ^ r6 19 E.2.3.5 Potential for Flashing During the wetwell will conservatively be underestimated. If s [ Venting v one further neglects the effects of any temperature
$ change which results from the blowdown (a second O The adoption of the Containment Overpressure order cffect), the rate of depressurizadon is:
Protection System (COPS) tn the ABWR design limits the potential release from the containment in the dP 0.665 ART]Pp, unhkely event that co' ainment failure is immanent. In -= (42) the absence of significant suppression pool bypass, the di V , M,,w fission products will be scrubbed as they pass through the suppression pool. The predominant conditions in where: P = Pressure, the suppression pool yield very high decontaminadon factors for all fission products except the noble gasses- A = Rupture Disk Flow Area, Given the extremely low releases from the gas space which result from suppression pool scrubbing before R = Universal Gas Constant, the rupture disk opens, the potential release resulting from the rapid depressuritation at the time the rupture p8 = Density of Gas, disk opens must be considered. Comparison of the time constant for blowdown V. = Volume of wetwell airspace, with the time constant for the pressure wave propagation around the wetwell demonstrates that the M.,w = Molecular weight of gas suppression pool acts as a one-dimensional body for the species in wetwell, purpose of this analysis 'This allows the calculadon of the pool swell height. Comparison of this level to the Conservatively assuming the wetwell vapor space location of the containment penetration indicatr that has only steam, for a tilowdown from 0.65 MPa to there is no potential for water to enter the COPS atmospheric conditions, the assumptions above yield a piping. This climinates the need for consideration of time constant on the order of 9 minutes. A typical time (q
,bj both water loads on the COPS piping and of fissmn product transport with water, it is also necessary to constant for a pressure wave going around the torus which comprises the wetwell is about 0.5 seconds.
consider the potential for water carried into the piping Comparison of these two numbers indicates clearly that to become entrained out of the COPS at the stack- the entire suppression pool will participate in the Calculation of entrainment at the surface of the blowdown. Thus, two dimensional effects may be suppression poolis also considered using the work of neglected. Rozen, et. al. (Reference 17) and is found to have an insignificant impact on fission product release. 19 E.2.3.5.2 Pool Swell 19 E.2.3.5.1 Critical Time Constants for in order to maximize the potential level in the lilowdown Response suppression pool, the analysis assumes that the firewater system has added enough water to fill the pool The time constant for the depressurization of the to the level of the bottom of the vessel (Elevation 0.0 wetwell airspace is calculated from critical flow meters). The two sources of steam which may lead to considerations. Comparing this value to the time level swell are in,:luded in this discussion. The first constant for propagation of a pressure wave around the steam source is the flow from the drywell through the wetwell annulus allows one to determine if non- connecting vents into the suppression pool. The second uniform effects in the suppression need to be considered source of steam which could lead to level swell is the in calculating the suppression pool response, flashing of the pool itself as the system depressurizes. The depressurization time constant for the wetwell 19 E.2.3.5.2.1 Pool Swell due to airspace is estimated based on the critical flow through Suppression Pool Flashicg the rupture disk opening and the ideal gas law. There are two sources of steam to the wetwell airspace: the Pool swell due to the flashing of the suppression blowdown through the vent system of steam and non. pool may be estimated by nw of a drift flux model. A condensible gas from the drywell, and the boiling or uruform void generation raw as assumed at each point steaming of the suppression pool which results from in the liquid. The average void fraction is then given the pressure decrease, if both of these sources are by-
/ T neglected, the time constant for the depressurizadon of Amendment ?? 19E.2 36.I
i 1 1 . ABWR 2mim, Standard Plani REV, A
! J, / U After the void fraction has been determined the ! k 5p = 2 + C 3 / U" p I Iml in pdndple, can M calatated usbg the
] 08 . . relationship in Equation (46). However, the difficulty 1 in applying these equations to the case with flow from 4 (Reference 1) where the mass flow rate, W,, at the top the drywell is the determination of the appropriate area of the pool determines the superficial gas velocity: which participates in the pool swell. Therefore, in order to determine if pool swell is a concem, the problem is j, = W, / Ap8 (44) considered in reverse. That is, the increase in pool height needed to raise water to the elevation of the ! vents is assumed to be present. This allows the
- and the drift velocity, U,,, is given by
- calculation of a void fraction and effective area for flow, l If one then assumes that there is a semi. circular region 1 - , V!M of influence around each of the vents, the critical radius U ,=1.53 og 78 (45) - may be detemuned.
1
. ( Pt s.
2A where: o = surface tension ofliquid, t= (48) i ! g = acceleration due to gravity' If this value is less than the distance between the inner and outer walls of the suppression pool, then
- P1
= density ofliquid. pool swell is not expected to lead to carryover of water
- into the COPS.
1 Then, by assuming the mass of the pool is approximately equal to the initial pool mass, the 19E.2.3,5.2.3 Steam Source average void fraction is used to calculate the average j pool height: The gas flow through the rupture disk comes from } three possible sources: the wetwell vapor space, the l V h= drywell vapor space and flashing of the suppression (46) ! Ep, + (1- E,)pi pool, la this calculation of pool swell, the wetwell vapor source is neglected. This results in a somewhat conservative estimate of the pool swell. In order to i where: h o = imt
. . ial pool height, determine the fraction of flow from each of the sources, the respnnse of the suppression pool and the drywell to j 19E.2.3.5.2.2 Pool Swell due to Flow From a change in wetwell pressure is calculated. Comparison l Drywell of these values allows the ratio of the flow rates from the suppression pool flashing and drywell throughflow A drift flux model is also used to determine the to be determined.
void fraction in the region of the pool above the horizontal vents due to flow from the drywell. The The pool flashing rate is determined by horizontal vents are located at the inner wall of the consideration of the conservation of energy equation in suppression pool torus, if quenching of steam in the the suppression pool:
- suppression pool (which is subcooled at the onset of
= the blowdown) is neglected, the void fraction in the d region above the vents is a constant: g(mphr) = Wph, (49) t
- a= J'/U" wLw
- mp = mass of water in the (47) 1 + Coj, / U., suppression pool, (Reference 1) where the terms are analogous to those hf = specific enthalpy of saturated
- defined for Equations (43) through (45), but now refer liquid,
- to drywell conditions. Comparison of Equation (47) to j Equation (43) indicates that the pool swell elevation is hg = specific enthalpy of saturated much more sensitive to through flow from the drywell vapor.
than it is to flashing of the suppression pool. } Amendment n 19E.2-36.2 4
5 , ABWR Standard Plant *'*7v^2 Taking the derivative on the left hand side of the alone will not lead to flooding of the COPS equation and introducing the derivative of enthalpy penetration. along the saturation curve, one concludes that-If the pool level were to rise an additional 3.28 mp g meters near the outer wall of the suppression pool due dP P to flow from the drywell, the COPS penetration could Wp= (50) be flooded. A void fraction of 20% due to through flow
'8 from the drywell is required for this additional pool swell. Applying Equations (47). (48) and the upper The ideal gas law is used in the drywell to derive bound value from Equation ($2), one arrives at an the relationship: radius of 0.78 meters for the region affected by flow from the drywell. This area would be located near to tk Wo = PVpM.,o/RTo (51) horizental connecting vents at the inner wall of the suppression pool. Since the distance between the inner and outer walls of the suppression pool is 7.5 meters, where all terms were defined previously and the one may safely conclude that pool swell will not subscript D refers to the conditions in the drywell, threaten the COI,S under these conditions.
De ratio of the flow rates from the drywell to pool flashing as found by combmtng Equations (50) and 19E.2.3.5.3 Carryover due to Entralament (51) The entrainment of water droplets by the steam fl w through the suppression pool is potentially a W_D. . VDM ' Dh concern since the water could carry fission products (52) W RTom, dh.L P through the COPS to the environment. A very simpic dP estimate analysis based on the work by Kutateladze (Reference 18) indicates the potential entrainment for a Pool swell is of chief concern for cases in which pool of water sparged from below. nc threshold for the N the firewater addition system has been used to add water entrainment of a droplet is based on the velocity of the to the containment. The suppression pool mass for this steam from the surface of the suppression pool: case is about 7.0E6 kg. An upper bound estimate of the mass flow ratio assumes that the drywell contains - e v4u nitrogen at relatively low temperature (100 C) and that U*"""'d = 2' 7 (53) the suppression pool is hot (160 C), Under thesc
- ' p/ 6 conditions the flow rate ratio is 0.065. These conditions will not occur in the ABWR, since the drywell cannot be cool when the containment pressure Assuming the properties of steam at the rupture is high. However, this value is useful to gain an disk setpomt, the threshold velocity is about 6 m/s.
understanding of the range of Equation (52). ne The superficial velocity from the surface of the bounding calculation shows that less than 10% of the suppression pool is 0.02 m/s, assuming all of the now flow through the COPS is being drawn through the through the COPS was passed through the suppression horizontal connecting vents. Therefore, the primar5 pool Rus, there is more than two orders of magnitude contributor to pool swell is flashing of the suppression between the superficial velocity which would bc P00;- observed under the conditions of interest and the threshold for entrainment. This indicates there ull be no significant entrainment from the surface of the pool. 19 E.2.3.5.2.4 Application to ABWR A m re s phisticated analysis is possible using the Pool swell is maximized at high pressure and work oMozen, et. al. @eference 17) to estimate pen temperature conditions. It is presumed that the rupture disk has just opened. Since the pool swell elevation is very I w amounts f entrainment. This method uses the superficial velocity of steam rising from tlw pool more sensitive to flow from the drywell, the upper and the pressure of the system to determme th: typical bound value for the mass flow ratio found above is
.df0Pl et size and the ratio of hquid mass to vapor mass used. For this condition, the average void fraction duc which ,s i entrained from the surface of the pool. Usmg to pool flashing is about 4%. This results in a pool this correlation, the ratio of h, quid mass to vapor mass swell of about 0.59 meters. Since the bottom of the ts about 4E4 K one considers an energy balance on the COPS penetration is at 3.87 meters, this mechanism /G suppression pool before and after the rupture disk V opens, it can be determined that just over one tenth of Amendment 71 19E.2 36.3
( . ,. , , . ,:: x;:- - - -- -
-- 3t i
b BWR 23A6100AS Standard Plant ! REY.A i j [ the suppression pool flashes to steam during the i blowdown. Thus, the fraction of suppression pool j liquid which might be transported from the suppression
- pool as a liquid is 4E.7. .
j The fission products in the suppression pool will i exist as a dissolved salt and as sediment on the bottom i of the pool. Therefore, the fraction of the fission products which can be carried out the COPS by i entrainment will be some fraction less than the ratio of 2 the liquid entrained from the pool surface. However, a j release fraction of 4E 7 will not lead to significant
; offsite dose.
I i i i 2 f f 1 iO e 4 -l 1 5 l J t i 4 i I i- [ i( l i Anahunt M 19E.2 36.4 l I t
.. . - _ . _ _ . . . - . . . . , . . . _I 1
-ABWR .
2mos Standard Plant Rev A 19E.2A Analysis for Recovery Sequences product release would be negligible.' Although the later time of release might argue for delaying the In order to determine the sensitidty of the PRA initiation of the firewater s). tem, the effect on risk is to various assumptions made in both the deterministic judged to be outweighed by the simplicity of telling and probabilistic portions of the analysis a series of the operator to initiate the firewater system as soon sensitivity studies were performed using htAAP. as possible in all circumstances. Additionally, some sequences with unusual characteristics, such as those having no containment The operator is instructed to initiate the firewater structural failure are considered in this section. addition system as soon as it is determined that the i water levelin the vessel cannot be maintained using % other systems. However,if the firewater system is - 19E.2.4.1. Time of Drywell Spray initiation not initialized quickly, the passive flooder will open I allowing the lower drywell to be flooded from the 9 The drywell spray initiation times used in the suppression pool. Thus, the assumed time for base analyses are simply assumptions used for the initiation of the firewater addition system does not 3' purpose of the study. This subsection examines the have a significant impact on the accident progression possible variation in accident progression which or on any eventual fission product release, would result if the time of spray initiation is varied from that assumed in the base studies. For example,in some cases the firewater system is not initiated for 2 hours. As a consequence of the accident progression, as modeled in the CETs, it is known that the operator failed to initiate the firewater injection system. Thus, it is logical to assume that the operator does not initiate the system immediately after vessel failure, if the system were operated immediately, the containment water level would reach g-
~
the level of the bottom of the vessel somewhat sooner a maximum of two hours earlier in this example). At h t(his time the operator would be directed to terminate injection. As seen in Figure 19E.2 3A, the T er containment pressure rises at this time eventually leading to opening of the rupture disk. The change in h@about two hours earlierthan that in the base analysis. time of rupture disk opening in this case would be On the other hand,if the operator did not initiate the firewater addition system in the assumed two hour period, more of the water initially in the lower drywell would boil off. Eventually, the debris in the lower drywell could begin to heat up. This would lead to actuation of the passive flooder in the lower drywell. This would quench the debris and keep the drywell cool. If at some later time the firewater system in initiated, the thermal mass of the suppression pool would be increased as in other sequences with firewater addition. Since the ' containment water level would reach the bottom of the vessel later than in the nominal case, the firewater injection would be terminated later, leading to later opening of the 6 rupture disk. The effe t on the magnitude of fission ( ( \ Amendment 19E.2-37 l i L l ,. -- -- .- - _ _ _ _ _ . - - - . -_. - . -
2 MN DA6100A5 ! 5 Standard Plant w4 l 2 1
- damage liinjection is begun within 20 minutes after the loss ofinjection.
3 O U in MAAP,it is not possible to halt core damage once the first channel region has blocked. Since this occurs very shortly after the onset of core damage, it ! is very difficult to determine the effects ofin. vessel < recovery on fission product release directly. 1 However, the salient feature of core melt arrest in the vesselis suppression pool scrubbing. If the core j , melt is arrested in the vessel then all of the fission products which leave the vessel must do so via the
- SRVs. These discharge through quenchers at the j bottom of the suppression pool, ensuring fission j product scrubbing.
i
- Fission product scrubbing is also provided if the release is from the wetwell airspace, as would occur ,
if the wetwell airspace were prodded with a rupture l disk. The release fractions associated with this type of release is examined in Subsection 19E.2.4.8. The l results of that study are applied to this case in the
- _ cffects analysis of Section 19E3. l 1
19E.2AJ Sptem Recovery after Vessel Failure and ) Normal Contalament teskage a This subsection describes the determination of
- containment leakage when pressures are below the ultimate pressure capability of the containment.
{ The majority of accidents for the ABWR do not 19E.2A.2. In Vessel Recovery lead to containment structural failure. In these accidents the RHR system is recovered to cool the This subsection examines the in vessel recovery containment following core damage, These l i sequence to determine how fission product scrubbing sequences are indicated by the characters HR in the should be modeled for these sequences. fifth and sixth digits of the accident sequence designator in the containment event trees. Although The potential for recovery of vesselinjection there is no structural failure of the containment in systems before vessel failure occurs is believed to be these cases, there will still be a small release of an important feature in the mitigation of severe fission products due to normal containment leakage. l These sequences are binned as NCL in the
- accidents. The sequences with fifth and sixth characters IV in the accident sequence designator in containment event trees.
j the containment event trees have core melt arrest in i the vessel. For the ABWR any of the ECC systems or To estimate the fission product release associated the firewater addition system is capable of adding with normal containment leakage following core sufficient water to the vessel to prevent core damage, damage a sensitivity study was performed using l MAAP. A loss of all core cooling with vessel failure and in theory, to halt the core melt progression once j it has begun. It is expected that the ECC systems can at low pressure case was chosen for the analysis. prevent core damage if injection is delayed up to half The transient was run for two days. The drywell and hour. The firewater system prevents core head failure pressure was raised to prevent containment rupture. The containment leakage area was chosen such that the leak rate was equal to the Amendment
i ABM muoaxs Standard Plant %3 technical specification limit of 0.4% per day at rated much higher than that for the base case pressure. (LCHP-PF D H). Fission product release begins at
- 6 the time of vessel failure (1.4 hours). The noble gas The average pressure during the transient is release is very slow since most of the noble gasses 0.55 MPa (65 psig). The first appreciable Ussion are trapped in the wetwell. After % hours only 85%
i product release occurs at 2.3 hours. After 50 hours of the noble gasses have been released to the the release fraction of noble gasses is 0.54%. The environment. The volatile fission product release is release fractions of the volatile fission products are predominantly governed by the revaporization of the
- much lower,5.2E-6 and 3.SE 6 for CsOH and Csi, fission products from the vesselinternals. After j respectively. The results of this ana!ysis is used in the 72 hours this revaporization is nearly complete. The a consequence analysis of section 19E.3 with the Csl and CsOH release fractions are both about 47%.
] sequence name NCL,
- Since the fission product release is significantly l higher than that for the base case, this information
- 19E.2.4.4. Early Drywell Head Failure will be included in the consequence analysis of i Seetion 19E.3.
} This subsection describes the modeling of fission
- product release for cases with early drywell head failure resulting from a high pressure core melt. 19E.1.4.5. Sappression Pool Drain in section 19D.5 the probability of the vessel This subsection describes the modeling of i failing at high pressure leading directly to loss of sequences in which the suppression pool water drains
! containment integrity was estimated. These to the RHR pump rooms, i sequences are indicated by the character E in the
! seventh digit of the accident sequence code. This The draining of the suppression pool following a j sensitivity study examines the potential fission product seismic event has been proposed as a potential release associated with such an event. Only two types . of sequences can lead to this occurrence: a loss of all mechanism following a seismic for the loss event. of sequences These containment are integrity l core cooling with vessel failure at high pressure designated with the seventh digit S in the accident " ' (LCHP), or a concurrent ATWS and loss of all core sequence code. The water from the suppression pool
- cooling with vessel failure at high pressure (NSCH). would flood the pump rooms as discussed in The LCHP event was chosen to represent this case as Subsection 19E.2.3,4. This analpis indicates that the i a has a higher probability of occurrence, pump room integrity will not be lost.
f' The history of this event until the time of However, there is a pipe chase that leads up from containment structural failure is identical to the the top of the pump room which has no capacity to LCHP events described in Subsection 19E.2.2.2 until withstand high pressure. There is no effective fissiou
- the time of vessel failure. At this time it is assumed product holdup if heat exchanger failure and
- that the drywell head fails. There is no significant suppression pool drain occur. This sensitisity study effect of the drywell failure on the entrainment of evaluates the fission product release associated with corium into the upper drywell, or on the opening of this structural failure mode, the passive flooder.
Since this failure is caused by a scismic event it is The pressure in the containment remains low, assumed that if one heat exchanger fails, causing the usually less than 0.2 MPa (15 psig). Just before the 2 suppression pool to drain into the RHR pump rooms i hour mark MAAP predicts that the drywell tear then all three heat exchangers fall. A comparison of becomes plugged by aerosols using the Morowitz the total floor area of the pump rooms to that of the plugging model. The pressure rises to a peak value of suppression pool shows that the water level in the 0.3 MPa (29 psig) before the acrosob are blown out pump rooms was nearly equal to the initiallevel of and the containment pressure falls to about 0.2 MPa the water in the suppression pool. Therefore, the (15 psig). suppression pool may be envisioned as being displaced to the pump rooms rather than being lost. The fission product release for this sequence is j Amendment 19E.249 a
-m _-._m_._ _.
AB R usum s Standard Plant ,,, 3 Any release of fission products to the atmosphere must pass through the RHR suction line,into the O pump room, and are then scrubbed in the pool now located in the pump room. For simplicity, the fission product release following heat exchanger failure and suppression pool drain was taodeled by assuming a large opening
'in the wetwell above the normal water level. No significant pressure head was allowed to develop in the wetwell. A loss of all core cooling event with vessel failure at low pressure and passive flooder operation was chosen (LCLP.PF) to model the transient. Dryout of the lower drywell, which could occur if no water was added to that region. was not modeled since the suppression pool elevation in this analysis was sufficient to prevent this occurrence.
The fission product release r curs as fission products are released from the fuel. The fission g products exit the vessel through the SRVs.' Scrubbing 4 occurs as the fission products are blown through the 1j pool. The only delay associated with any release of it fission products which are not trapped in the pool is the dilution effect brought about by a large wetwell i gas volume. }e The release of fission products begins as the fuel y_ ( begins to melt at about 0.5 hours. The noble gas release was essentially complete at 8 hours. The g release of volatile fission products was very small due 4 to scrubbing. The final release fraction after 84 hours , was less than 1.E.5. j d3 6 i 19E.2-40 Amendment l-
ABWR nA61%AS Standard Plant Rev A O O Amendment 19E.0-41
l ABWR MAOTAS Standard Plant 2e O 4 4 j 4 O j i 4 O' Amendment 19E.2-12 4
ABWR u^6 mas Standard Plant Rev 3 0 v 4 4 Os.J Amcodtnea 19E.2 82.1
l ABWR 2m Standard Plant j 19E.2.5 Identification and Screening (3) Systems Behavior /Opertior Actions
? of PhenomenologicalIssues
~3 -
Twelve events defined operator actions and he first step in performing an uncertainty analysis systems availability during the course of the f
/> is to identify the key phenomena and their associated accident progression including whether hydrogen uncertainties. To do this, GE has surveyed these igmtors were available, the status of containment vanous sources (References 19 through 27). sprays and whether the containment was vented.
These event questions were generally asked pnot to core damage, during core damage, at vessel he following provides a summary of the key literature reviews. Some of the severe accident issues failure and late after vessel failure. Other events are screened out as not being applicable to the ABWR c nsWred were RV pressure dunng core damage, design. At the end, a list of sensitivity issues will be upper pool dump, SRVs sticking open, and presented for investigation in the ABWR PRA. restoration of in-vessel injection during core damage. 19 E.2.5.1 Regnew of NUREG/CR 4551 (4) AC/DC Power Availability Grand Gulf and Peach Bottom Analysis Six events were related to AC and DC power The ABWR containment shares some similarities in design to the Mark Ill BWR containment. The availability / recovery during core damage, f 11 wing vessel failure and late in the accident NUREG ll50 study of Grand Gulf was used to identify phenomena and issues which may need to be addressed pmgnsson. _ in the ABWR uncertainty analysis, in addition, the Peach Bonom (Mark 1) analysis was also reviewed for (5) Criticality insights. The results of the NUREG 1150 Grand Gulf and Peach Bottom containment analysis are presented One event assessed whether the debris would be in below, a critical configuration after core injection recovery. O 19E.2 5.1.1 Grand Gulf (6) Hydrogen Related Phenomena / Issues The Grand Gulf accident progression event tree (APET) consists of 125 event headings. The events Forty-eight events in the GG APET were related treated in the Grand Gulf APET can be grouped into ten to assessing the impact of hydrogen production categories based on similar accident progression and combusuon on containment and drywell phenomena or characteristics. This grouping is integrity. These hydrogen event questions were summarized on Table 19E.2-22 along with the Grand asked at numerous time periods throughout the Gulf APET events which fall into each group. A accident progression: during core damage, at vessel summary of the phenomena and issues addressed by failure, following vessel failure and late in the each event group are dimwi below: accident sequence. (1) Plant Damage State Grouping Events The hydrogen production event questions considered hydrogen production in-vessel during com damage and that released at vessel failure and De first fificen events in the GG APET and Event 20 were somng type events which summarized the during core cete interactions (CCI). Several plant damage state for a sequence based on the events were included to assess the transient availability of various core injection and concentrations of hydrogen, oxygen and steam m containment systems, the timing of core damage, the drywell and containment throughout the the availability of AC and DC power and the accident progression and to determme if regions vessel pressure, were inert (or non inert) to deflagrations or detonations during various time penods. (2) Structural Capacity / Initial Containment Status For distinct time periods throughout the accident Four events (Events 16 19) summarized the early Progression the probability of ignition of status of containment integrity and pool bypass hydrogen diffusion flames, uncontrolled deflagrations, and detonations were considered and defined the ' structural capacities of the al ng with the efficiencies of the burns and the containment and drywell to quasi static and v Peak burn pressures (and detonation impulse impulse loading.
^=n*nat M 19E.2 42.2
4. ABWR u-s Standard Plant REY.A i 1 O loads). Additional events compared these loads 19 E.2.5.1.2 Peach Bottom C with the containment and drywell structural capacities and determined if failure or leakage The major phenomena considered in Jie Peach would result. Bottom APET which were not addressed in the GG APET were liner mell-through and over temperature (7) Containment /Drywell Pressunzation and Failure failure of the containment (drywell) pc.etrations. Twenty two events assessed containment and 19 E .2.5.1.3 Application of NUREGl drywell pressure and level of leakage resulting CR 4551 Results to ABWR from a combii ation of loads (gradual overpressurizatica from steam . and non- Since the ABWR containment is inerted, the GG condensible gases) not directly associated with APET events associated with details of hydrogen hydrogen combustion. This set of evertts also production and combustion are not relevant. assessed the response of the reactor pedestd and drywell to the pressure loads resulting from The remaining GG APET areas are generally energetic events which may occur at vessel failure considered applicable to the ABWR, insights from the including steam explosions and rapid steam GG APET have been factored into the ABWR generation in the reactor cavity, blowdown of the containment event tree analysis considering differences reactor vessel from high pressure and high between the two designs. pressure melt ejection. De design of the ABWR lower drywell is very (8) Core Concrete Interactions / Pedestal Failure different than the Peach Bottom pedestal cavity. The manway used to gain access to the lower drywell is Seven events were directed at assessing the about 5 meters above the floor. ne liner, which behavior of debris in the reactor cavity following represents the containment boundary, in the lower vessel failure. Dese events determined whether drywell is protected by one meter of sacrificial concrete. there was a water supply to the debris, whether nerefore, the debris will not come in contact with the A the debris was coolable,(and if not) the nature of liner in a manner which could lead to liner melt-
! ! the resulting CCI and whether the CCI would through. Therefore, liner melt @. rough is not addressed 'd result in pedestal failure. in the ABWR analysis.
(9) Steam Explosion Related However, in the unlikely event of vessel breach with the vessel at high pressure, it is considered Five events assessed the likelihood and possible that debris transported into the upper drywell consequences of steam explosions occurnng in. may threaten containment integrity as a result of a vessel or ex vessel in the reactor cavity. in vessel general heatup of the upper drywell atmosphere if the steam explosions which failed the upper reactor drywell sprays are not available. As discussed in vessel head, drywell and containment (alpha mode Attachment 19EA, the debris will not be transported as failure) or which failed the lower head of the a contiguous mass. Therefore, the formation of a debris vessel were considered. He pinbability of large pool in the upper drywell is not a credible event. ex-vessel steam explosions occurring and failing However, there may be some debris in the upper the pedestal (by impulse loading) were also drywell which could lead to long term high temperature evaluated. failure of the containment. The effects of high upper drywell temperature are considered in the CET in (10) Core Damage Progression and Vessel Breach assessing the probability of drywell failure. Four events were related to assessing the general 19 E .2.5. 2 Review of NUREG 1335 in. vessel accident progression and vessel failure characteristics. These events evaluated the amount Table A.5 from NUREG 1335 is included here as of core debris in the initial core slump, the Table 19E.2-23. This table includes a list of the amount of debris mobile in the lower head at parameters identified by the NRC to be addressed in an vessel failure, the mode of vessel failure and IPE. All of these will be addressed in the final list of whether an HPME occurred. sensitivity analyses to be carried out for the ABWR except for those discussed below:
/^
( (1) Combustion in Containment v Amendmem ?? 19E.242]
ABWR nw REV A s Standard Plant i ) As noted above, the ABWR containment is 19 E. 2.5.4 Review of ALWR V inerted and, therefore, combustion mil not result Requirements Document in a challenge to containment. The EPRI ALWR Requirements Document (2) Induccd Failure of the Reactor Coolant System includes a top level section referred to as the Key Assumptions and Guidelines (KAG) which defines the This is mainly an issue for PWRs. The thin walls manner in which a probabilistic risk acessment is to of the reactor coolant system outside of the vessel be performed for advanced plants. Paragraphs 6.2 and may fail to due extended exposure to elevated 6.3 address those parameters which could be important temperature and pressure. For typical conditions for the containmert n:sponse: in a BWR dunng an accident, induced failures are judged to not occur. (1) Parameters related to hydrogen burns, (3) Direct Contact of Debns on Containment (2) Core Debns Coolability, Due to the configuration of the ABWR cavity, (3) Pressure capacity and failure location of the under a low pressure vessel failure scenario, core containment, debris will be retained in the cavity and will not come in direct contact with the containment (4) High Pressure Melt Ejection, boundary. For a high pressure melt scenario, debns that is entrained into the upper drywell will (5) Ex vesselcombustible gas generacon, be dispersed and will not result in the coherent flow of debris to the containment shell needed to (6) Operator Actions, cause containment failure. 19 E.2.5.3 Review of Recommended Sensitivity Analyses for an Individual Plant (8) lodine composition and revaponzation. Examination using SfAAP 3.0B (EPRI), As stated previously, hydrogen burning is This document was reviewed to ensure that there precluded in the ABWR design by use of an inerted were no new tssues that had not previously been containment. Operator actions are being considered in a identified in the above documents. In this document, separate study. The remainder of these issues are the following key issues are highlighted for BWR neluded in the ABWR sensiuvity and uncertainty sensiuvity analyses: analyses. (1) Hydrogen Generation In. vessel, 19 E.2.5.5 Summary and Conclusions (2) Mass of Molten Core released at vessel failure. Table 19E.2-24 is the list of issues to be investigated in an ABWR sensitivity analysis and has (3) Csl re-vaporization, been denved from the documents desenbed above. (4) Debris Cootability, (5) Containment Failure Mode. All of these issues are being addressed in the ABWR sensitivity and uncertainty analysis. Some issues are being considered indirectly in the framework of the phenomenological issues they affect. F(,r example, the mass of molten core released at vessel failure is considered in terms of the impact on high pressure melt ejection, direct containment heating and core debris coolability. Amendment ?? 19E.2 42.4
ABWR 23A6100AS Standard Plant REV.A r ( 19E 2,6 Sensillvily Analysis and and two at high pressure (LCHP.PF P M and LCHP-Scoping Studies for Phenomenological PS R-N). These cases were identical to their respective gS
- 3 base cases, described in Section 19E.2.2, except that the model parameter for blockage and hydrogen generation (FCRBLK, see Reference 1) was set to Sensitivity studies are performed for the ABWR prevent blockage and to cause the metal water reaction response to severe accident phenomena in order to to continue past the eutectic temperature of the corium, determine those issues which may hase significant impact on the offsite nsk associated with the ABWR For the cases at low pressure, the amount of design. Given this goal, the ultiraate measurement of z rconium oxidation increased from 6.3% of the active c
sensiuvity is the offsite dose. At a given site the clad to 15.8%. ne time of vessel breach decreased from e pnmary factors which influence the dose are the 1.8 hours to 1.1 hours. For the dominant case with 3 magnitude and time of release. Therefore, changes in the firewater system operating, the rupture disk opens g these parameters will be used to determine the need for at 30.6 hours as compared to 31.1 hours for the base 4 detailed uncertainty analyses. The issues to, be. case. The Csl release to the environment increases mvest gated in the ABWR sensitivity analysis is given slightly to about 1.E 6; however, the release is still in Table 19E.2 negligible and will not affect the offsite dose. For the case with passive flooder operation, the time of rupture 19 E.2.6.1 Core Melt Progression and disk opening decreased from 20.2 hours to 16.7 hours. Hydrogen Generation The change in the magnitude of fission product release This subsection examines the effect of the MAAP core melt progression modeling on the hydrogen ne blockage model had a more pronounced effect generauon due to metal water reaction- on the amount of zirconium oxidized for the high pressure cases. De fraction of zirconium oxidized for ne progression of a severe accident during the the no blockage case was 35.9%, increased from 5.1% period when the core is melting is important in for the case which included the blockage model. For the A predicting the amount of hydrogen produced during the core melt. The standard melt progression using MAAP LCHP-PS-R N case, the time to rupture disk opening () _ g is characterized by molten corium fonning blockages in is decreased frem 25.0 to 20.0 hours. The impact on the magnitude cifission product release is negligible.
; the channels which prevent steam from Howing in the channels. Bis model has two major effects on the rnett Hew <er, for the LCHP PF P H case the effect of 4 y progression. First, once a region has been blocked, it is an increase in pressure is more significant because
~ impossible for that region to be cooled since no water leakage through the movable penetrations is assumed to E
can flow into the channel to arrest the core melt, occur at 0.46 MPa (52 psig). The time fission product g Therefore, a core melt can not be arrested in the vessel release begins for this case is reduced from 18.1 hours j after the onset of core damage. Secondly, the blockage n the base case to 7.1 hours with increased hydrogen i of the channel prevents steam from flowing past the production, Additionally, the magnitude of the Csl hot, uncovered portion of the fuel. This serves to limit release fracuon at 72 hours is increased from 8.7% to the metal water rescuan which can occur in the vessel. 12.5 %. Metal-water reaction in a BWR is dominated by the ne difference in the effects of blockage on oxidation of zirconium. This reaction has two hydrogen production can be best explained by important consequences in a severe accident. First, the considering the steam flow past the hot fuel cladding. reaction is exothermic, that is it adds energy to the For cases with vessel failure at low pressure, the containment. Second, as oxygen from the steam is operator blows the reactor dowe before significant consumed in the oxidation reaction, hydrogen gas is heatup of the cladding has occurred. Although the generated which adds to the parual pressure of the non- blockage model does not predict complete blockage condensable gasses in containment. Both of these until shortly before vessel failure, the loss of water in effects tend to increase the pressurization rate of the the core region which occurs during the blowdown containment and shorten the time to fission product effectively terminates the metal-water reaction after release, only 6.3% of the active cladding has been oxidized. The conditions found in the high-pressure vessel failure A sensitivity study was performed to determine the cases are more conducive to hydrogen generation for effects of the blockage model on hydrogen generation- three reasons-Four cases were examined, two at low pressure (V) (corresponding to LCLP FS R N and LCLP PF-R N) p 19 E.242.5
ABWR 23A6100A$ Standard Plant asy. A d (O d (1) Higher steam temperature in the vessel prior to vessel failure, the suppression pool earlier. His will result in more cificient scrubbing of the fission products. (2) A greater mass of water in the core region, and The effect of the release rate can be modeled in MAAP ABWR by use of the variable SCALFP (3) A longer time before vessel failure. (Reference 1) which decreases the release rate. Since early releases will result in lower releases from Despite these conditions, the blockage model containment, this possibility will not be examined. In causes slightly less of the zircomum to be oxidized by order to investigate the sensitivity of the dose to the MAAP ABWR for base cases with vessel failure at release rate from the fuel, the LCLP-PF R N sequence high pressure than for cases with vessel failure at low was run with SCALFP changed from its nominal value pressure. The blockage model used in the base cases of 1.0 to 10.0. This reduces the rate of release by an presumes that molten material forms blockages in the order of magnitude, core which prevent steam now past the fuel cladding. This terminates zirconium oxidation and limits The behavior of the noble gases is not noticeably hydrogen production. The core is fully blocked in the :.ltered by the slower release. Some variation of the high-pressure melt sequence at 1.2 hours, while in the volatile release is observed. De most risk significant low-pressure sequence full blockage is delayed until of the volatile fission products, Cst, is used as the 1.8 hours, measure of the behavior of the fission products. In the nominal case approxirnaiety 65% of the fission When the blockage model is disabled, the effect of products are carried into the suppression pool shortly the blowdown becomes more apparent. De lower water after vessel failure. A small percentage of the Csl is level m the low pressure core melt sequence results in found in the drywell at this time, but the majority of less steam generation from decay heat and less hydrogen the remaining fission products remain in the vessel generation. Therefore, much more hydrogen is generated where they are slowly revaporized. Finally, after the in the high pressure case which has more steam rupture disk opens, the flow through the vessel is available for metal water reaction, sufficient to cause vaporization of the remaining 25% O Csl in the vessel. The final release fraction of Csl V In summary, the blockage and eutectic cutoff models used in MAAP reduces the hydrogen generation through the rupture disk to the environment is less than 1.E-7. by a factor of 2 to 7 compared to the cases w here these models are not used. For the more dominant LCLP-FS. The same basic trends may be observed in the R-N, LCLP-PF-R-N and LCHP PS R-N sequences behavior of the sequence with SCALFP equal to 10. there is very liule change in release and time to rupture However, the amount of material in each location disk operation. The only case -hich resulted in a varies substantially during the progression of the significant impact on the timing and magnitude of accident. At the time of vessel failure only 25% of the fission product release is the LCHP-PF-P-H sequence. Cs! has been swept to the suppression pool. About However, examination of the containment event trees 20% of the Cs! is still present in the corium which in Section 19D.5 indicates the probability of this event relocates to the lower drywell. The remaiaing 55% of is very small. Herefore, it is judged that the ABWR the material remains in the vessel, either in the fuel severe accident perfccmance is not sensitive to in vessel itself or on the various cool surfaces of the vessel. hydrogen poducuen. Slow release of Csl from the vessel then occurs until a the time of the rupture disk opening when the fraction ci 19E.2.6.2 Fission Product Release from of Csl in the vessel and that in the suppression pool are a Core both about 40%. The amount of fission products in the
, drywell remains relatively unchanged during this penod.
As in the nominal case, the remain Csl leaves the 7 The base sequences shown in Section 19E.2.2 use vessel soon after the rupture disk opens. The final o the Cubiccioni model for fission product release from release fraction of Cs! to the environment is also 1.E-7 f the fuel. If the release from the fuel occurs later than for this case,
;g the time predicted by the MAAP model then there could be more airbome fission products available for release k from the containment Also, as the accident progresses, Despite the large variations in the location of the the decontamination factor associated with the fission products within the containment during the suppression pool will decrease as the pool heats up. accident, there is no appreciable variation in release from the containment due to the presence of the O Conversely, if the release is more rapid, the fission containment overpressure in the design. Therefore, no products will pass through the SRVs or the drywell to
() Amendment 77 19E.2 42 6
ABWR MA6to0AS Standard plant REY A O further investigation of the impact of fiuion pnduct plenum water pool is very small. Consequently, the release from the fuelis required. lower plenum is nearly full of water at the time of core plate failure.
- 19E.2,6.3 Csl Re esaporation
" Due to the large amount of core debris poured into An important aspect of fission product behavior is the vessel head at the time of core plate failure, local
$ the propensity of the aerosols to adhere to the relatively failure of the instrument tubes is predicted very soon f cooler surfaces of the vessel and containment. While after debris enters the lower plenum. Derefore, there is M the deposition process is fairly well understood, there is insufficient heat transfer to the corium to quench it in
$ considerable uncertainty in the revaporitation of the the vessel; and, molten corium and water are relocated d fission products. MAAP assumes that the fission to the lower drywell. Figure 19E.2 2E shows that
{ products are revaporized such that the local vapor approximately 200,000 kg of water falls into the lower i pressure is consistent with the temperature of the drywell at the time of vessel failure. surface. However, it has been proposed that chemical reactions may occur on the deposition surfaces which in other melt progression models the molten fuel bmd the fission products. This could result in delayed drips down the fuel rods in a process called candling, revaponzation as the heat sink temperature slowly rises Under this assumption, it is possible for molten due to the decay heat of the fission products. corium to be relocated in the lower plenum slowly, where it is quenched. Vessel failure could then be in the vessel of a BWR, most of the fission delayed until all water in the lower plenum is boiled off product deposition occurs on the steam dryers. After the and the corium is reheated. Dis delay allows more time fission products are deposited, they slowly begin to for operator action which could prevent vessel failure heat the dryers due to the decay heat they carry. As the from occurring. temperature of the dryer increases, the fission prtducts are revaporiicd. Thus, the impact of chemical bmdmg During the time when the water in the lower of fission products to the dryers may be simulated by plenum is boiling, steam would continue to flow past assuming a larger dryer mass. This causes the dryer the fuel rods which could result in increased hydrogen 7._ temperature to rise more slowly, which in turn s% production. The impact of hydrogen production on the V) ( the re-evolution process. For this study, the dryu mass was doubled and the base sequence LCLP-PF-R-N was containment response is discussed in Subsections 19E.2.3.2 and 19E.2.6.1 which conclude that increased recalculcaed. mel water reaction will not have a significant impact on ibe offsite risk. As in the discussion of fission product release .n Subsection 19E.2.6.2, the Csl will be used as the More important than the hydrogen generation is representative fission product compound. There is no the behavior of the fission products assuming this type real difference in the timing of the key events. of core melt progression. As modeled in the MAAP However, comparison of the results of this calculation program, a significant fraction of the volatile fission to the base sequence described in 19E.2.2.1 shows that products are not swept into the suppression pool as there is 2% to 5% more Csl in the vessel at any time they are Tleased from the melting fuel. Rather, they are during the transient. Nonetheless, there is not a retained on the relatively cool surfaces in the vessel substantial difference in the release fraction from the such as the steam dryers. Later, as these structures heat containment. In both cases the release fracuan of Csl at up, the fission products are revaporized. If the vessel is 72 hours is about 1.E.7. Based on this small release still intact, the fission products will be swept directly fraction, no further consideration of Cs! re- into the suppression pool via the safety relief valves evaporization is necessary, w here most of the volatile species will be retained. 7 19 E.2.6.4 Time of Vessel Failure For typical sequences using MAAP ABWR, up to 80% of the volatile fission products are deposited on d vessel surfaces just prior to vessel failure. These fission
< The detailed progression of a core melt during a severe accident is subject to considerable uncertainty, products would be released to the drywell atmosphere 4 The core melt progression assumed in MAAP retains very slowly and would only be swept into the the corium above the core pla'e until local core plate suppression pool gradually as steam is generated in the
[g failure occurs, resulting in a large pour of core debris drywell and the containment pressurizes. A sequence y into the lower plenum of the vessel. Before this time, was rerun with a modified version of MAAP-ABWR in Y water in the lower plenuni has sery little impact on the w hich vessel failure was delayed until the water in the (N 4 accident progression because heat transfer to the lower lower plenum had been boiled dry. In this sequence, Amendment ?? 19 E.2-42.7
ABWR m ms Standard Plant REV.A (m) only about 30% of fission products remained in the vessel at the time of vessel failure. This indicates that There are several mechanisms which tend to reduce the potential that the core becomes critical. First, when MAAP ABWR base analysis may overpredict the the cold water is injected into the hot core, it is likely airborne fission products in the drywell. This could that the any fuel which had been at very high result in a significant conservatism for sequences in temperature will shatter and form a rubble bed. w hich the drywell head is presumed to fail. Therefore, Analyses perfonned by PNL (Reference 29) indicate the base analysis is conservatise as regards the in vessel that the rubble bed geometry is suberitical since it is effects of debns coolability in the lower head and time undermoderated. Similarly,if there has been substanual of vessel failure, relocation of fuel from the upper part of tle core me lower portion of the core will be undermooerated and The assumed core melt progression model will will probably be subcritical. Finally, if recriticality have mmor impact on the long term ex. vessel portion occurs, boron can be injected via the SLC system to of a severe accident, in the base analyses shown in return the core to a subentical state. Section 19E.2.2, there is an initial quenching of the corium in the lower plenum followed by a period of Presuming that the core recriticality occurs as a time in which the wa:er in the lower plenum boils off. result of in vessel recovery, the power level would rise ne corium then reheats and the passive Gooders open, to a lesel determined largely by the rate of injection. The influx of water through the flooders quench the Thus, in effect, a partial ATWS condition develops. As corium. If the corium is retained in the vessel until the with any ATWS condition, voiding in the core tends to water from the lower plenum was boiled off, then the limit the power generation. Thus, the more injection initial quenching of debris in the lower drywell will not available to the core, the higher the power level. occur and the passive flooder will open earlier relative Depending on the precise configuration of the core and to the time of vessel failure. However, this will not control material, it is possible that some of the fuel is have a significant effect on the overall plant response to damaged locally. However, since coolant is necessary a severe accident, for power generation above the decay heat level, widespread melting of the fuel is inconsistent with the ne potential for the debris to be cooled in the increased power levcl associated with recnticality. f] lower plenum may have an important effect on some of Q the phenomena which are important immediately after vessel failure, if the debns is not coolable, as was The steam generated in the core would flow through the SRVs to the suppression pool which assumed in the base analyses, there may be a large would begin to heat up, pressurizing the containment. amount of molten debns at the time of vessel failure, he emergency operating procedures direct the operator If, on the other hand. the debris is cooled in the lower to inject boron via the SLC system and to reduce the plenum, the penetrations may be expected to fail when water level. Boron injection terminates the recriticality only a small fraction of the matenal is molten. Both of event. Lowering the water level reduces the power these possibilities are considered in the direct generation to a level which can be removed from the containment heating and debns coolability uncertainty containment via the containment heat removal system. studies contained in 19E.2.7. If no steps are taken to reduce the power level or to terminate the event, the containment will continue to 19 E.2.6.5 Recriticality During In Vessel pressurize leading to opening of the rupture disk and Recovery possibly to containment failure, in either case, any fission products released from the fuelin the penod in D A potential challenge to the containment has been which the core was melting and not retained in the 7 identified for accidents in which the core melt is arrested suppression pool could be released from containment. in the vessel, Experiments have indicated the potential 4 for the boron carbide in the control blades to form a in order to examine the potential for recriticality to
$ eutectic with steel at 1500 K and relocate (Reference the ABWR containment a low-pressure core melt
% 28). This is considerably less than the temperature at sequence was examined in detail to estimate the length G which the fuel relocates (2500 K). Thus, as the core of time in which recriticality is possible. Qualitative heats up and begins to melt, there may be regions of judgements are made about the potential for fuel the core which are uncontrolled. If the vessel were shattering and the effects of fuel relocation.
re0ooded after the onset of control blade relocation there Additionally, a transient was run using a modified is a potential for regions of the core to become critical version of MAAP ABWR to provide a conservative l raising the power level. His could increase the rate of estimate of the minimum time available for the containment pressurization and could potentially lead to injection of boron should recriticality occur. operation of the rupture disk or to containment failure. (Vc) Amendment 7 19E 2 C 8 i
ABWR 23 A6100AS Standard Plant REY.A
/D 19 E.2.6,5.1 Potential for Recriticalit.v Du, = (FA)u,o(TL - TL,) (2) in examining the potential for recriticality it is important to recognize that the heating and relocation where: FA = Effective area for radiation heat of the core does not occur uniformly. Variations in the transfer, time of uncovery, heat transfer to other structures and the decay power cause the core heatup to progress from o = Stefan.Boltzman coefficient, the top central portion of the core to the outer and lower regions. In general, once a pertion of the core T = Temperature.
begins to heat up, it heats quickly until it reaches its melt point and begins to relocate. The view factor from the control blade to the fuel is taken to be one w hich neglects the ef fects near the A MAAP-ABWR calculation of a low. pressure center of the cross. These assumptions tend to core melt sequence was examined in detail to minimize the temperature difference. Assuming a decay investigate the heatup and melting behavior of the core. heat level of 29e rated power and incipient melting of The ABWR core has been modeled using a mesh of ten the control blades, the lower bound on the temperature axial and five radial nodes such that each cell has equal difference between me fuel and the control blades is vclume. Each node is assumed to have a single about 15 K. Even if different assumptions were made, temperature. The relocation of the baron carbide is not maximizing the temperature difference, the fuel and modeled in MAAP. However, judicious examination of control material temperatures would be very close to the MAAP analysis can give useful insights, each other at these high temperatures. Therefore, the use of a single temperature for the fuel and the control Before taoking at the MAAP.ABWR analysis, blade is a reasonable assumption. consider the possibility for temperature differences between the fuel and the control blades. The source of A MAAP.ABWR calculation of a low. pressure energy for the heating of the core is the decay heat in core melt sequence was examined to determine the core the fuel. This leads to the observation that the heatup and relocation characteristics. The core was O temperature of the control blades should be less than nodalized using 5 radial rings and 10 axial levels. [Q that of the bulk of the core. Any temperature difference between the control material and the f uct would tend to About 48 minutes after the start of the accident, the temperature in the inner rings of levels 8 and 9 exceeded decrease the time window for recriticality. In order to 1500 K, the temperature at which relocation of the estimate the temperature difference, a simple radiation control material might begin. Within a minute levels 6 calculation is performed which neglects heat transfer to and 7 also exceed 1500. At 52 minutes, the fuel exceeds the steam and assumes that the heat transfer between the temperature for zirconium melting (2100 K): and, the fuel and the control blade w ill cause both to heat up by 55 minutes, there is widespread melting of the core at the same rate. Thus, in this region. (mcp )u, After the fuel exceeds the melt point for zirconium, 6m* = (1) any remaining cladding will be highly oxidic. It is Qomy (mep)u, + (mc p)u + (mc p)an judged to be highly likely that the rapid addition of cold water to the vessel would result in local shattering and w here: Q = Rate of heat addition, relocation of the fuel. Thus, one would not expect a region w hich has exceeded the zirconium melt point to mc p = Thermal mass. c m n'cnsal. As time progresses, the region which might be J Using approximate values for the thermal masses, devoid of control material moves downwards. At the only about 10% of the decay energy will go the control same time, fuel from the upper regions of the core also blade. relocates filling these regions with fuel. This reduces the mass ratio of the moderator to the fuel reducing the For an indication of the temperature difference potential for recriticality. Therefore, it is judged that the between the blade and the fuel when the blade begins to critical interval for recriticality is a period of about 7 melt, a simplified radiation heat transfer calculation is minutes. performed. The channel box walls are neglected and
,. black body radiation is assumed- The probability of recovering core cooling in this
[ } interval is fairly small. In order for recriticality to
%/
Amendment ?? 19R.2 42.9
ABWR St:nd:rd Plant REV.A (q occur, there must be a system (or operator) failure that V) deprives the core of all cooling for about 50 minutes, then injection must be recovered in a time window of the operator has ample indication that the reactor is critical since the containment pressure is rising very rapidly. about seven minutes. Based on the standard models for recovery of systems and operator error, it is concluded that the probability of this occurrence is small. This estimate overpredicts the power level and, Therefore, the probability of a recriticality event thus, underpredicts the time until the rupture disk occurring is small. might open for several reasons: 19 E.2.6.5.2 Implications of Recriticality (1) As discussed above, it is expected that only a small region in the core will become critical. Despite the judgement of a low potential for Most of the core will be shut down. Thus, the recriticality, an assessment of the effects of a bulk of the core will generate power at decay heat level. The Chexal Layman correlation represents reenticality event are examined if vessel injection is recovered and some portion of the core becomes critical, the condition where the entire core is the power level would rise above the decay heat level, uncontrolled. Bus, the power level associated with recnticality will be a fraction of the ATWS As long as core injection continues, the fuel would be power predicted by Chexal Layman. cooled, thus, no significant fuel damage would occur. However, the addnional power generation could increase (2) The Chexal-Layman correlation is based on the rate of containment pressurization. The operator could terminate the recriticahty event by initiation of conditions at rated reactor pressure. At low l the SLC system or mitigate the event by controlling pressure, the void fraction will be considerably the vessel injection flow rate. higher. This causes the power level to be substantially reduced at low pressure. Many of the To bound the impact on the containment, a recovery scenanos will occur with the vessel at low pressure. For these cases, the use of Chexal-calculation was performed to determine the earliest time at which the rupture disk could open given a Layman is conservative. If the vessel is at high recriticality event. This time indicates the time pressure, the LPFL systems will not have O available to terminate recriticality via the stand-by sufficient head to inject and the power level will (j
' liquid control system or, as a minimum, to reduce the be lower than that calculated here.
power level via flow control and slow the rate of pressunzation. (3) It is highly unlikely that the all of the ECC systems will be recovered at the same time. As shown in Section 19D.5, the dominant core MAAP was not designed to analyze recovery scenarios. The model does not contain entic@y damage event in the ABWR is initiated by a models, nor can it assess power associated with a transient with failure of all core cooling (Classes IA and ID). These sequences represent about 70% degraded core configuration. However, with one minct modification, it is possible to force an ATWS to war of all core damage events. The simultaneous recovery of all ECC systems is not credible for late in an accident which, in effect, is a recriticality these scenarios. At a given pressure the power event with the entire core uncontrolled. MAAP- ABWR level is directly proportional to the flow rate, includes the Chexal Layman correlation for power Thus, the power should be about one fifth that during an ATWS. This result will bound the power generation in a recriticality event. given here since it is highly likely that only one ECC system is recovered. The low-pressure core melt scenario discussed above was used to estimate the time to rupture disk (4) Even if all injection systems were to inject, the operator is instructed to reduce the core flow if the opening during a recriticality event. It was assumed that power rises above decay heat level. Studies of recovery of injection occurred at approximately 50 ATWS at high pressure have shown that the use minutes, in order to determine the minimum time to l rupture disk operation, all of the LPFL and HPCF of flow control will reduce the power to about 4%. Analyses performed for the success critena in systems were presumed to be available. A full ATWS condition was forced at the time of injection recovery, Subsection 19.3.1.3.1(2) show that the Based on the Chexa! Layman correlation MAAP. containment can be maintained below Service ABWR predicts a power level of 15E The containment Level C by use of flow control and the containment heat removal system (Reference 30). pressunzes to the rupture disk setpoint about 55 g minutes after recovery of injection. During this time k Ammdment ?? 19E.2 42.10
ABWR 23A6l00AS Standard Plant asy. A p Thus, one hour is a very conservadse estimate of coolant interactions are most likely to challenge the the time until the opening of the rupture disk, it is containment when molten debris falls into water, expected that the actual time until the containment Examination of the containment event trees indicates pressure reached the rupture disk setpoint would be that only 0.3% of all sequences have water in the lower several hours. If the operator initiates SLC injection as drywell before vessel failure. Both the impulse and directed in the Emergency Procedures, the recriticality stade loads are considered. Fuct coolant interactions event would be ter.ninated. Therefore, the risk (FCI) may occur either at the time of vessel failure associated with a recriticality event is not judged to be when corium and water fall from the lower plenum of significant. the vessel, or when the lower drywell flooder opens after vessel failure has occurred. 19 E.2.6.5.3 Conclusions Fuel coolant interactions were addressed in the The potential for recriticality, as well as the early assessment for the ABWR response to a severe implications of its occurrence, was examined. It was accident. Subsection 19E.2.3.1 examined the concluded that the time window in which recriticality hydrodynamic limitations for steam explosions and , could occur is very small and that on'y a small portion concluded that there was no potential for a large scale j of the core could become critical at any time of steam explosion. The pressurization of the containment d recovery cf injecdon. A very conservative calculation from non-explosive steam generation was calculated in J' was performed which assumed that the entire core was the analyses for the accident scenarios. He following uncontrolled and all ECC systems were used. This sections examine the available experimental data base , bounding calculation indicates the containment does not for its relevance to the ABWR configuration, and j exceed the rupture disk setpoint for at least one hour provide a simple, scoping calculation to estimate the after recovery. It is expected that the actual time until rupture disk operation would be several hours. This ability of the ABWR contairt nent to withstand a large, energetic fuel coolant interaction.
}
w allows ample time for the operator to terminate the event by use of the stand.by liquid control system or to Four potential failure modes are considered. The - (p V mitigate the event by reducing the rate of injection to the vessel and initiating containment heat removal. Thus, it is concluded that recriticality does not pose a transmission of a shock wave through water to the structure may damage the pedestal. Similarly, a shock wave through the airspace can cause an impulse load. d sigmficant threat to the ABWR design. However, since the gas is compressible, the shock i wave transmitted through the gas will be much smaller ;- 19E.2.6.6 Debris Entrainment and Direct than that which can be transmitted through the water. J' Containment Heating nerefore this mechanism is not considered here. Third, loading is caused by = lugs of water propelled into if a core melt accident occurs in which the reactor containment structures as a result of explosive steam S pressure vessel is at high pressure at the time of vessel generation. Finally, the rapid steam generation may 9 failure, the debris may be entrained out of the lower lead to overpressurization of the drywell. drywell, if the debris is finely fragmented as it is The details of the analysis are presented in 3 dispersed, the pressure in the containment can rise Attachment 19EB. De studies show that the limiting m rapidly. This process is called direct containment
$ heating (DCH). The magnitude of the pressure rise is loading mechanism is the shock wave transmitted to the structure. Using a conservative bound for the y dependent on the amount of debris involved in the impulse load capability of the pedestal, the structure y event. If a large fraction of debtis participates in the DCH event the containment may be challenged. Since can withstand the loads associated with a steam
; this would lead to an early failure of the containment explosion involving 9.5% of the core mass. This is structure in the drywell. The fission product releases three times the mass of a credible fuel coolant l from this type of scenario are judged to be high. interaction in the ABWR. Therefore, the ABWR j nerefore,it was decided to bypass the performance of a pedestal is very resistant to fuel coolant interactions, his failure mechanism need not be considered further j sensitivity study for this case and perform a detailed in tns cont *.mment event trees or the uncertainty
; uncertainty analysis. The results of this uncertainty
- analysis can be found in Subsection 19E.2.7.1. analysis.
Y 19 E.2.6.7 Fuel Coolant Interactions Challenges of the containment during a severe [G accident may result from fuel coolant interactions. Fuel Amendment 77 19E.2 4211
s ABWR oms Standard Plant REY.A C" 19 E.2.6.8 Core Concrete Interaction and containment have an estimated failure pressure of 9 Debris Coolability 180 psig. Thus, it is expected that most fission
- product releases will be via the n pture disk.
3 The issue of debris coolability has long been an W area of considerable uncertainty in the progression of a A fragility cu vc for the drywell head, Figure 2, core melt accident. In the ABWR design the lower 19FA 1, shows the uncertainty in the failure pressure
- 2. drywell floor area is large in order to facilitate the for the drywell head, The uncertainty of the rupture
$ spreading of the cote debris. The firewater addition pressure for the COPS is very small as 'iiscussed in
" system, as well as the passive flooder design, ensure Subsection 19E.2.8.1.1 and 19E.2.8.1.2. Integrating that debris will always be covered by water in the event over these two distributions, one can determine the of a severe accident. probability that the drywell head fails before the COPS actuates. Because of the pressure difference between the However, experiments performed to date have been wetwell and the drywell, two cases must be considered.
unable to provide conclusive evidence that these For sequences in w hich the firewater system is used and features cool the debris sufficiently to prevent core water is added to the containment, as described in concrete interaction from occurnng. If core concrete Section 19E.2.2, there is approximately a 5% chance interaction were to contintc unabated, there are two that the drywell head will fait. For sequences without possible challenges for the ABWR containment design, water addition to the containment, the drywell head First, the generation of non condensible gas would failure probability is about 2%. These probabilities are contribute to the slow pressurization of the used in the quantification of the containment event trees containment, even if containment heat removal is in Section 19D.5. available. Second, if the concrete were croded to a sufficient depth, the pedestal walls could be weakened 19 E.2,6.10 Fission Product Release Flow g' to the point that the vessel was no longer sufficiently Area supported. If the vessel then tipped or fell, the piping 4 attached to the vessel could cause the containment The presence of the COPS servc3 to substantially penetrations to tear, most likely in the drywell region reduce the uncenainties associated with the flow area for - of the containment- release of fission products from the containment. In the i s/ unlikely event that fission products are released from 2 The time of fission product release from the the containment, the release will almost always be via Sl contamment for either of these mechanisms would be the COPS. Since this is an engineered feature of the % fairly late but is dependent on the heat transfer from the plant, the uncertainties associated with the available { corium to the overlying pool of water. Additionally, flow area are very small. The COPS is designed to continued core concrete interaction can lead to an allow steam now equivalent to 3% rated power. Since increase in the amount of fission product release. Since the decay heat level will be less than 1% at the time core concrete interaction can lead to a mode of drywell COPS operation is required, it is judged that the failure and because of the high visibility of this issue, containment response is not sensitive to any small it was decided to bypass the sensitivity study and to variation in the COPS cffective flow area. perform detailed uncertainty analysis for the dual issues of debris coolability and core concrete interxtion. However, for the few cases discussed in Subsection 19E.2.6.9, the pressurization of the containment leads 19 E.2.6.9 Fission Product Release Location to failure of the drywell head. For these cases there is substantial uncertainty in the failure area. Therefore, cr- The adoption of the rupture disk in ti.e ABWR two sensitivity cases were analyzed. In the first case the cf containment design serves to significantly reduce the nominal failure area of 20 square inches (.0129 m2 ) a uncertainties in the timing, location and area of any was increased by a factor of two. In the second case the fission product release. As discussed in Subsection failure area was divided by two. This broad range should 19E.2.8.1, the Containment Overpressure Protection bound any possible variations in the failure flow area. g'. System (COPS) is highly reliable. The setpoint of the
$ rupture disk,90 psig (0.72 MPa), was selected such The results for the three cases are identical until the l
% that there is a very small probability that the time of drywell head failure. After drywell head failure a containment structure fails. As shown in Appendix the basic trends of the data are unchanged. The
; 19F.3.1, the weakest portion of the ABWR containment pressure is larger for cases with the t containment is the drywell head. The median failure smaller failure area than for those with larger areas, pressure of the drywell he'ad is estimated to be 134 psig There is also a small variation in source term for the fi (1.03 MPa abs). The other portions of the three cases. In the nominal case the release fraction of V
Amendment ?? I9E.2 42.I2 f
.--_-.-------.-._-._------.--n--
- ~ ~ - ~
_M 4 1 4- ABWR- 23A6100AS i Standard P! ant - REV.A-Csl is 9.7E For the larger How area, the release (5) - Bypass leakage is present from the beginning of j- fraction increases to 12.6%; w hile, for the smaller flow the accident and the operator initiates the firewater. e-
- area, the release drops 4.2% Considering the upper - spray system. l l bound, doubling the flow area increases the release by .
1 only 30% Since less than 5% of all releases are a Suppression pool bypass can lead to a significant l result of drywell head failure, the change .in offsite increases in fission product release. Releases can be on . 9 -i consequences will be small. Therefore, no further the order of 10% for a fully stuck open vacuum breaker, 9 i consideration of containment failure area is necessary. For sequences in which the firewater addition system is G ., j used in. spray mode, the time to release is not t i 19E.2.6.l l Suppression Pool Bypass significantly affected. However, for sequences without l sprays, the time from the beginning of the accident i' The BWR containment is designed such that all until the onset of the release can be significantly k[ gas generated in the vessel and the drywell passes reduced. The use of the Morowitz blockage model , { I through the suppression pool. This serves to quench results in a significant improvement in the calculated - 3-l the steam in the gas stream, which substantially risk associated with suppression pool bypass, ' j decreases the pressurization rate of the containment. In Nonetheless, there is a substantial increase in
- addition, any fission products carried in the gas stream consequences associated with large bypass areas,
} are scrubbed and retained in the suppression pool Since Therefore, suppression pool bypass is examined with a
- the ABWR is designed such that any fission product detailed uncertainty analysis in subsection 19E.2.7,
- release is from the wetwell airspace, this substantially I reduces the risk in the unlikely event of a severe IS E.2.6.12 liigh Temperature Failure of i accident. Subsection 19E.2.3.3.3(4) examined Dr yw ell q i mechanisms which could result in suppression pool '.
bypass, and determined that the only pathway which One of the failure modes identified for the .M # cuald significantly increase risk is vacuum breaker containment was the degradation of the seals for the l failure or leakage. The results of a sensitivity study moveable penetrations in the drywell due to high - w 1 performed to examine the impact of vacuum breaker temperature (Subsection 19F.3.2.2). In the base
- j. ,) Nrformance is summarized in this subsection. Details
~
of & analysis can be found in Attachment 19EE.3. analyses sequences which exceeded discussed in Section the threshold temperature of ' 1
$33 K (500 F) were those in which debris was The dominant severe accident sequence [ Loss of all entrained into the upper drywell and sprays were not ',
-core Cooling with vessel failure occurring at Low available. In these cases the debris can radiate directly to Pressure (LCLP)] was chosen to evaluate plant the upper drywell structures. For the other sequences, performance. MAAP ABWR runs were made with the debris is covered by water so elevated temperature in f effective vacuum breaker area. ANK, varying from 0 to the upper drywell is dependent on heat transfer from i
d 2030 cm2 (315 in2). The upper bound corresponds to remaining fuel in the vessel to the upper drywell. R one fully open vacuum breaker. Five variations were analyzed. In cact case the overpressure relief rupture To ascertain the sensitivity of the drywell q disk opened when the wetwell pressure reached 0.72 temperature to parameters which could affect it, several MPa (90 psig). The five scenarios were: sensitivity studies were performed. All of the studies Q were performed using a low-pressure core _ melt (1) Bypass leakage begins after passive flooder sequence. The LCLP PF-R N sequence, with passive activation; aerosol plugging is neglected. flooder operation, was selected since cases with firewater spray available are not expected to result in (2) Bypass leakage is present from the beginning of high drywell temperatures, the accident; aerosol plugging is neglected. In the first calculation performed, the mass of (3) Bypass leakage begins after passive Dooder equi pment in the drywell was decreased to reduce the activation; aerosol plugging of the vacuum thermal mass in the upper drywell, The mass was breakeropeningis considered, arbitrarily decreased to half of the nominal value used in the base adlyses.Tbc temperature in the upper drywell , (4) Bypass leakage is present from the beginning of at the time the rupture disk opened decreased from its - the accident; aerosol plugging of the vacuum r.ominal value of 500 K (441 F) to 487 K (418 F). breaker opening is considered. While this result is somewhat counterintuitive, it can be easily explained. In the early stages of the accident, the temperature in the drywell is higher in the-
. Amendment r? 19E.2 42.13
-- - . , _ _ _ n. _ _ _ _ , _ . _ . _ _ _ , _ _, a .. _ ,
ABWR o iouxs Statidard Plant REV.A (A O t sensitivity case. This results in a small increase in the amount of fuel which melts and relocates into the lower pool for all species of interest except the r,0ble gasses which have a DF of 1.0. drywell. Consequently, there is less heat generation in the vessel and less radiative heat transfer to the upper in order to investigate the sensitivity of the offsite drywell. The overall containment performance is not consequences of a severe accident to the suppression affected by the slight decrease in temperature. pool decontamination factor, a simple sensitivity study was pe, formed. The MAAP ABWR code was modified A second analysis was then performed in w hich the to allow a canstant DF to be input for all species mass of equipment ia the upper drywell was increased except the noble gasses. Two calculations were then by a factor of two. In this case the upper drywell repeated assuming a conservative DF of 100. None of temperature at the time of rupture disk opening is the other Ossion product removal mechanisms were virtually unchanged from the nominal case, in the very affected by the change. long term, well after the rupture disk opens, there is a slight increase in temperature compared to the nominal The two cases selected for study were both low-case as one would expect based on the previous result. pressure core melt sequences, in the first sequence, However, there is no significant impact on containment LCLP FS-R-N, the firewater system is assumed to be performance. available, while in the second case, LCLP-PF-R-N, the passive Gooder operated to cool the debris. Both cases ne final sensitivity case performed considered the indicated a significant increase in the fission product impact of increasing the convectise heat losses from release. For the case with the firewater system the vessel to the drywell 50% above its nominal value. available, the fraction of Csl release increased from A slight incr:ase in the upper drywell gas temperature 1.5E-7 to 1.2E-3. For the case with the passive flooder was observed in this case. At the time the rupture disk the results were similar, the Csl release increased from opened, the upper drywell temperature was 505 K 1.2E-7 to 1.6E-3. (450 F) as compared to 500 K (441 F) in the nominal case. The overall containment performance is not CRAC cases were run in order to determine the affected by this slight change. effect of these changes on the consequences of release. The results of this calculation are shown in Figure j In summary, the three sensitivity studies performed to assess the sensitivity of the drywell temperature to 19E.2-21. Case 1 is the nominal case and Case 4 uses the release fractions from this sensitivity study. The the detailed modeling assumptions indicate that the conditional probability of exceeding the offsite dose ABWR is not sensitive to those parameters which indicated on the x axis is shown. The probability of the affect drywell temperature. Therefore, no further study dose is dependent on the weather. The curve shows that of this area is necessary. there is virtually no impact until a conditional probability of 0.04. Thus, there will not be a 19E.2.6.13 Suppression Pool significant impact on offsite dose, even for this very Decontamination Factor conservative DF of 100. Rus,it has been shown that g the consequences of a severe accident are not very N From the standpoint of severe accidents, one of the sensitive to variation in the suppression pool most important features of a pressure suppression decontamination factor. No further consideration of this containment is the suppression pool. The suppression phenomena is required in uncertainty analysis. g g pool not only quenches any steam which enters it, d reducing the rate of containment pressurization, it also
% traps the fission products carried with the gas flow.
d This process, known as scrubbing, significantly reduces l the amount of fission product acrosols available for i release from the containment. The efficiency of the scrubbing process is typically charrterized in terms of a decontamination factor (DF) defined by ths mass of debris which enters the pool divided by the mass of debris which leaves the pool. MAAP-ABWR uses correlations based on the SUPRA code to calculate the DF. These correlations typically result in very high retention of fission products in the b Amendment 77 19E.2 42.14
ABWR
- 23 A6100AS Standard Plant gay. 3 0
19E.2,7 Detailed Pher.omenological failure due to DCH is very low, there is no measurable Uncertainty Studies impact on offsite dose. 19E,2.7.2 Debris Coolability 19 E.2.7.1 Direct Containment Haating De issue of debris coolability has long been an Direct Containment Heating (DCH) is the sudden area of considerable uncertainty in the progression of a heatup and pressurtzation of the containment resulting core melt accident. In the ABWR design, the lower from the fragmemation and dispersal of core material in the containment atmosphere. DCH is a concern for drywell floor area is large in order to facilitate the spreading of the core debris. The firewater addition sequences in which the vessel fails at hign pressure system, as well as the passive flooder design, ensure since the steam Dow from the vessel provides the that debns will always be covered by water m the event motive force for entrainment. In the event of a of a severe acciden sufficiently large DCH event, the containment could fail at the time of vessel failure. This would lead to However, experiments performed to date have been very high releases to the environment. In the past DCH has been addressed for Pressurized Water Reactors. unable to provide conclusive evidence tnat these features cool the debris sufficiently to prevent core BWRs have very reliable vessel depressurization concrete interaction from occurnng. If core concrete systems. Thus, the frequency of accidents with the interaction were to contmue unabated, there are two vessel remaining at high pressure is extremely low, However, with the many sources of low. pressure possible challenges for the ABWR containment design, injection available to the ABWR to prevent core First, the generation of non.condensable gas would damage, the frequency of all core damage sequences is contribute to the slow pressurization, even if containment heat removal is available. Second, if the very low. Derefore, high pressure core melts a epear as contnbutors to the total core damage frequency, albeit concrete were eroded to a sufficient depth, the pedestal walls could be weakened to the poire that tne vessel with a very low probability. was no longer sufficiently supported. If the vessel then A detailed uncertainty analysis utilizing tipped or fell, the piping attached to the vessel could cause the containment penetrations to tear, most likely decomposition event trees (DETs) was performed to m the drywell region of the conta'nment. Additionally, assess the peak drywell pressure resulting from a DCH continued core concrete mteraction can lead to an l event. This analysis is given in Appendix 19EA. A increase in the amount of fission prodtet release. large number of calculations were performed to determine the impact of DCH on the probability of containment failure and offsite risk. The analysis A detailed uncertainty analysis utilizing investigated uncertainties in a variety of phenomena: decomposition event trees (DETs) was performed to determine the potential for continued core concrete interaction and its impact on the containment response. (1) M<xie of vessel failure, This analysis is given in Appendix 19EC. A large l number of calculations were performed. These (2) Mass of molten core debris at the time of vessel calculations addressed uncertainu,es m the followmg failure, parameters: (3) Potential for high pressure melt ejection, (1) Amount cf core decris, (4) Fragmentation of debris in the containment. (2) Debris-to-water heat transfer, Additional sensitivity studies were performed to (3) Amountof addin.onal steel in the debris, examine other phenomena which could affect DCH. The study concluded that a deterministic best estimate (4) Delayed floodmg of the lower drywell, for the peak pressure from DCH would not lead to containment failure. Consideration of the uncertainties in the phenomena lead to an estimated CCFP of 0.1% . (5) Fire water injection instead of passive flooder. for all core damage events. Additional sensitivity analyses were considered which indicate that an upper The conclusion from all of these uncertainty bound on the impact of DCH is 1.5%. Even in this calculations were: limiting case, the probability of DCH failing containment is well below the goal of 10%. (1) For the dominant core melt sequences that release Furthermore, since the probability of containment core material into the containment,90% result in Amendment ?? 19E.242 I5
, AInVR .ms S_landard Plant RI:V. A no significant CCl. An insignifrant number of low probabihties of occurrence. No leakage and, 1 sequences are eyected to experieme dry CCI. correspondingly, no irnpact on plant nsk is eslweted to occur for almost all(approsimately 98 percent) of the (2) Even for those low frequency cases with accident demands. Small amounts of leakage have a trobability of 1.8 percent per event, art. can result in significart CCl, radial crosion remai% below the struct2allimit of the pedestal. After consideration medium volatile fission product releases (one to ten l of uncertainties only !.5% of the sequences with percent of initial inventory). Volatile fission product significant CCI will suffer pedestal failure. f. ' ases on the order of 10 to ' percent of initial Combining this conclusion with the first, only inventory can result when large a..sounts of suppression 0.15% of all core melt sequences with veuel pool bypass are prnM. Ilowever, the impact on plant failure will t;ad to additional drywell tailures as a risk is still negligible because the probability of large result of CCI. leakage is only 0.39 percent.
(3) The time of fission product release is not significantly affected by continued CCI. , I (4) The fission pnxtuct release is dominated by the i l noble gasses when the cemsnment overpressure protection system operates. This conclusion is unaffected by assumptions on debris coolability. Therefore, the offsite dose for sequences with rupture disk operation is not impacted by core concrbe attack. These conclusions would indicate that the uncertainties associated with CCI have an insignificant influence on the containment failure probability and ( risk. 19 E.2,7.3 Suppression Poo! Ilypass Suppression pool bypass (the passage of gas and sinr from the drywell directly into the wetwell space) can lead to increased fission product releases.
, shown in Subsecuon 19E.2.3.3.3(4), the only mode suppression kol bypass that has the possibility of significandy increasing risk is vacuum breaker leakage.
Attachment 19EE determined the probabilities and consequences for vacuum breaker leakage areas from rero to that correspondir.e to one vacuum breaker stuck fully open. Fission product : lease fractions were determined with MAAP ABWR using the dominate accident sequence (Loss of all core Coohng with vessel failure occurring a Low Pressure (LCLP)) modified to include a path between the drywell and the welwell airspace. Plugging of leakage paths by fission products was considered for small pathways. Leakage probabilities were determined by reviewing recent operating esperieace of wetwell to drywell vacuum breakers in - BWRs with Mark I,11 and 111 containments. Suppression pool bypass does not significantly add [,) to the risk associated with the ABWR because the b bypass areas resulting in increased releases are offset by Amendmem 't? 19E.2 4216
1 ABWR MA6HoA5 . Standard Plant . RIN 4 ( G Q [ 19E.2.8
Feature Considerations Severe Accident Dnign penetrauons in the drywell rather than dr)well head f ailure.). To compare the consequences of severe accidents resulting in fission product releases via drywell head failure to those with releases through the S Although the frequency of core damage is very low COPS. MAAP was used to simulate a series of severe j in the ABWR design, features wcre added to tir design accident sequences fedoth release inechamsms these to ensure a robust response of the containment to a severe accident sequences are descnbed in Section severe accident. This sectmn discusses the imponant 19E.2.2. Failure pressure nf the drywell head was considvations for the severe accident design icatures.
assurned to be equal to it2 median ultimate strength, 1.025 MPa (134 psig). The sesults of these runs show 19 0.2.8.1 Containment Oserpressure releases of volatile fission products, after 72 hours, for Protection Splem tin COPS cases to be several orders of magnitude less Aar' fcr the corresponding drywell head failure cases. ABWR has a very low core damage frequency
- The Csl release fractions are compared in Table Furthennore, in the unlikely event of an accident 19E.2 25. Most accident sequences show this large resulting in core damage, the fission products are difference in releases between drywell head fatlure and typically trapped in the containment and there is no COPS cases, release to the environment. Nonetheless, in order to mitigate the consequences of a severe accident which 19E:2.8.ld Pressure Setpoint results m the release of fission products and to limit the Determination effects of uncertamues in severe accident phenomena, ABWR is equipped with a Containment Overpressure Sod b tors were considered in determining the Protection System (COPS). This system is mtended to optimum pressure se*Wint for the rupture disk. The provide protection against the rare sequences in which results of the previous analysis show that it is desirable structural integrity of the containmen' is challenged by to avoid drywell head failure. This can be assured by
, overpressurir.ation. It has been determined that these providing a rupture disk pressure setpoint below the i rate sequences comprise only 16 gercent of the
, , g gg g hypothestred severe accident sequmces.
integnty of the containment. However, as the pressure setp int is reduced, the time to containment failure and De COPS is pa.1 of the atmospheric control fiss n pr duct release is also reduced. Thus, the system and consists of two 8. inch diameter setpomt of the rupture disk must optarmre these overpressure relief rupture disks mounted in senes on a c mpeting factors: miniminng the probability of
- 14. inch line which connects the wetwell airspace to the drywell head failure while maximiting time before stack. The COPS provide. a fission product relcase Assmn pr duct Men to k emironrnent.
point at a time pnor to cent nment structural failure. Thus, the containmer,! structure will not fail. By De service level C capability of Se containment engineenng the release point in the wetwell airspace' serves as onc indication of the structuralintegrity of the the escaping fission products are forced through the containment. As shown in Appendix 19F, the service suppression pool. In a core damage event initiated by a level C for the ABWR is 97 psig, limited by the transient in which the vessel does not fail, fission drywell head. Thus, it is desirable to set the rupture products are directed to the suppression pool via the disk setpoint below this salue. SRVs, scrubbing any potential release in a severe accident with core damage and vessel failure or in a he distribution of drywell head failure pressure LOCA which leads to core damage, the fission prvducts and the distnbution of rupture disk burst pressure were will be directed from the vessel and drywell through the also considered in determining the burst pressure. As drywell connecting vents and into t .e suppression pool stated in Attachment A to Appendix 19F, the drywell again insunng any release is scrubbed. Eventually, if head failure pressure is assumed to have a lognormal the containment pressure cannot be controlled, the distribuuon with a median failure pressure equal to its rupture disk opens. Any fission product release to the ultimate strenge of 1.025 MPa (134 psig). The environment is greatly reduced by the scrubbing variability of rupture disk opening pressures is best provided by the suppression pool. modeled with a normal or Gaussian distribution. Typical high quality rupture disks exhibit a tolerance of In the absence of the COPS, unmitigsted 1% of the mean opening pressure. Tests have shown overpressurization of the containment will result in eat Ms M toluance spans 12 to 123, standard
. failure of the drywell head for most severe accident deviations of the rupture disk population.Ris analysis
[b scenarios (Some high. pressure core melt sequences result in fission product leakage through the moveable f the Containment Overpressure Protection System l
^"*"* 77 19tw ?
A Il W R -s Standard Plant uv. A conservatively assumes that only 12 standard des ianons lhe clapsed time to rupture disk opening was are included within the 5% tolerance, within 0 8 hours of the baw case value of 20.2 hours for both caws tested. liigher rupture disk temperatures A critical peameter in determining the risk of (i.e. lower pressure setpoints) teduce the tituc to rupture dr>well head failure before rupture disk opening is the disk opening and lower rupture disk temperatures (i.e. pressure difference between the drywell and wetwell, higher pressure setpoints) increase the time to rupture Late in an accident the drywell is at higher pressure disk opening. There were no significant changes in than the wetwell. For a given rupture disk setpoint the fission product release. For both cases the Csl release probability of drywell head failure increases as the fraction at ?2 hours remained less than lE 7. ; l pressure difference increases. The masimum dryw eli to wetwell pressure dif ference is 0.1 h1Pa (14 psi). This Another parameter affected by the variation in the ( pressure difference occurs for cases in which firewater rupture disk temperature is the probability of drywell spray was activated af ter vessel failure but terminated head failure prior to rupture disk opening in a severe before containment failure, Cases without fireaater accident. Using the rupture disk and drywell head failure ; distributions,it was determined that the probability of spray have pressure dif ferences of no more than 0.05 i h1Pa (7 psi). drywell head failure prior to rupture disk opening increased from about 2% for the base case to about 3% A rupture disk setpoint of 0.72 h1Pa (90 psig) at for the case with the rupture disk temperature of 311 K 366 K (200 F) was chosen. The residual risk o r drywell (100 F). With a rupture disk temperature of 422 K head fadute may be calculated by combining the two (20 F), the probability decreawd to about 1.5% The distributions with an offset corresponding to the rupture disk temperature variation has a similar effect pressure difference between the wetwell and the dryw ell. on the severe accident sequences in which the firewater A 90 psig setpoint results in a 5% probability of spray system is activated. The probability of drywell drywell head failure prior to rupture disk opening for a head failure prior to rupture disk opening increases from 0.1 h1Pa (14 psi) drywell to wetwell pressure about 5% for the base case to about 6.5% for the case difference. For a drywell to wetwell pressure dif ference with the rupture disk temperature of 31i K (100 F) and of 0.05 h1Pa (7 psi), the drywell head failure decreases to atuut 4% for the case with the rupture disk probability prior to rupture disk opening is 2E This temperature of 422 K (300 F). G is judged to be an acceptable level of risk, The results of this sensitivity study show that 19 E.2.8.1.2 Variability in Rupture Disk variations in rupture disk temperature, which cause Scipoint small variations in rupture disk opening pressure, have a minor effect on the performance of the ADWR Nickel was chosen as the material for the rupture Containment Overpressure Protection System. disk for evaluation purposes due to its relative insensitivity to changes in temperature. At 191L2.8.1,3 Siring of Rupture Disk temperatures above room temperature the opening pressure of a typical nickel rupture disk will decrease by The sire of the rupture disk has also been about 2% for a 56 K (100 F) increase in temperature, optimized, if the rupture disk is too small, it could be Thus, in order to estimate the uncertainly due to incapable of venting enough steam to prevent further variations in the temperature of the ADWR rupture containment pressuriention. On the other hand, if the disk, a sensitivity study was performed in which the rupture disk is too htrge, level swell in the suppression pressure setpoint of the rupture disk was varied. pool could introduce water into the COPS piping, if this were to occur, the piping could be damaged or there The nominal pressure setpoint of the rupture disk could be carryover of u aterborne fission products from is 0.72 MPa (90 psig) at 366 K (200 F). Two cases the containment. were examined using h1AAP in this sensitivity study. For both cases the LCLP PF R sequence was used as An eight inch rupture disk was selected. This is the base case. First, the rupture disk pressure setpoint sufficient to allow 35 kg/see of steam flow at the was reduced to 0.708 MPa (88 psig) which corresponds opening pressure of 90 psig (0.72 51Pa.a) and to a rupture disk temperature of 422 K (300 F); and, corresponds to a energy flow of about 2.4% rated second, the pressure setpoint was increased to 0.735 power, For virtually all severe accident sequences, the h1Pa (92 psig) which corresponds to a temperature of rupture disk would not be called upon until about 20 311 K (100 F). This temperature range, from 311 to hours after scram. The decay heat level at this time is 422 K (100 to 300 F), bounds all anticipated rupture less than 0.5E Thus, there is ample margin in the (m)
%J s'isk temperatures. sir.ing of the rupture disk for severe accidents.
Amendment ?? 191L2 4218
ABWR 23A6tnoAs j Statidard Plant uv. A L 4 1 An additional accident was considered in the drywell head failure. In the 3.9 hours between rupture selection of the rupture disk size. In the event of an disk opening and hypothetical drywell head failure for l j ATWS with the addidonal failure of the standby liquid the LCLP FS sequence, the probability of recovering j control system, the operator is directed to lower water RilR capability is only 4% (see Subsection 19.3.2.7). I level to control power. Analysis has shown that the This represents the probability that the COPS was } RilR system is espable of removing the energy opened unnecessarily since RilR would have been i generated by the ATWS from the containment recovered in this time period. d (Subsection 19.3.1.3.1). If the additional failure of l containme beat removal is assumed, a simple For cases with passive flooder operation, the 4 calculat: n im tlat an the rupture disk area is just Guion product release occurs about 6 to 8 hours sooner sufficir Wit the containment pressure below than it would have if the drywell head was allowed to service km L. pressurite to 1.025 MPa (134 psig). For the range of 1 severe accident sequences described in Secdon 19E.2.2, ! Calculations were also performed to invesdgate the the probability of RilR recovery in a similarly defined j potential effects of pool swell and fission product time window is about 11E
- carryover at the time of COPS operation. These analyses (Subsection 19E.2.3.5) indicate that pool For both cases, there is a small probability that
) swell does not threaten the integrity of the COPS RilR will be recovered before the time at which i j piping and that no significant entrainment of fission containment would fail if the rupture disk setpoint has j j products will occur due to carryover, been surpassed in light of this fact and given the i
! difference in magnitude of the fission product release,it i 19 E.2.8.1.4 Comparison of AllWR is clearly preferable to direct the fission products
! Performance With and Without COPS through the rupture disk, i i The results of the MAAP calculations for the 19 E.2.8.1.5 Suppression Pool flypass l various accident scenarios were investigated in Section i 19Ea.2 and the releases are summarited in A comparison of performance for cases with Tabk 19E.2 25. Comparisons of Csl release fraction at suppression pool bypass flow through an open vacuum jl 72 hours show large differences between the COPS and breaker valve was also considered. Cases were run with drywell head failure cases. Csl release fraction at 72 bypass effective area varying from 5 to 2030 cm2 l- hours for drywell head failures is on the ordet of 0.1% -(.0054 to 2.19 ft 2). A fully open vacuum breaker has ! to 15% For all cases with release via the COPS, an effective area of 2030 cm2 . The dominant the Loss
- MAAP predicts release fractions of less than IE 7. of All Core Coolant with Vessel Failure at Low j Table 19E.2 26 summarires several critical parameters Pressure sequence was considered with Passive Flooder a for the dominant low pnsssure core melt scenario. Operation since previous analysis has shown that the l firewater system is capable of mitigating bypass.
- There is, of course, some reduction in the clapsed i time to fission product release for the COPS cases No credit was taken for aerosol plugging of the j when compared to the drywell head failure cases. For bypass leakage in this analysis; and, therefore, the i the dominant accident sequences in which the operator results are conservative. Also, it was assumed that the i initiates the firewater spray system prior to ' bypass leakage was present from the beginning of the i overpressurir.ation, the time difference between rupture - accident sequence. As the bypass area increases, the
! disk opening and drywell head failure is only 3 to 4 fraction of fission product aerosols which pass through hours. A typical example is the Loss of All Core the suppression pool decreases. Thus, the benefit of a' Coolant with Vessel Failure at Low Pressure with wetwell release of fission products is significantly Firewater Spray addition sequence (LCLP.FS), as reduced as the bypass area increases. described in Subsection 19E.2.2.1. For this sequence the wetwell pressure will reach 0.72 MPa (90 psig) and For bypass effective areas less than 50 cm2 (,034
- the rupture disk will open at 31.1 hours. Without the [g2 ). Csl releases at 72 hours from the COPS cases rupture disk, the drywell will reach 1.025 MPa (134 were smaller than for the corresponding drywell head i psig) at 35.0 hours, failure cases. However, the differences in Csl releases at -
- 72 hours were only factors of 2 to 4 rather than several l The potential for increased risk due to the rupture orders of magnitude. The time difference between '
disk opening early has been considered. It is assumed drywell head failure and rupture disk opening was 4 to S that recovery of RIIR. capability is sufficient to hours for these small bypass areas. For bypass effective 3 terminate containment pressurir.ation and prevent areas greater than 50 cm 2 (.054 ft2 ) Csl release i Amendment ?? 191L2 42,19 I i
- e. . - . , ,...- .__ . _ _ . . . , . . .,._~.._,-.~.,_..,__,w a_.,____,_.-,_..________..,_,...,_
ABWR nos Standard Plant RLVA ( fractions at 72 hours are on the order of 10ci for luth reduced as a result of the COPS implementauon mio the drywell head failure cases and the COPS cases. One the design. the other hand, the ume difference betw een rupture disk opening and dryw;11 head failer is only 2 to 4 hours 19 E.2.N.2 Lower Drywell Flooder for these larger bypass areas. These relauvely small time differences will not significantly affect the 19 0.2.8.2.1 Introduction magnitude of the offsite dose. Attachment 19EE has a complete discussion of suppression pool bypass flow This section provides the bases for siring the lower dirough vacu.im breaker valves- drywell flooder system, ne system is desenbed in detail in Section 9.5.12 of the ABWR SS AR. 19 E.2.8.1.6 Summary The lower drywell flooder provides an alternate A wetwell pressure setpoint of 0.72 MPa (90 psig) source of water to the lower drywell once it contains for the overpressure relief rupture disk meets the design core debris. The pnmary water source is the firewater a goal. The probabihty of containment structural failure addition system. Water present in the lower drywell - is minimited while maximizing the time to fission cools the core debns and establishes a water pool above ? product release in a severe accident. The 5.1% the debris. Water absorbs heat by first heating up to I maximum probability of containment structural failure saturation conditions and then boiling away. Debris '. if the pressure reaches the rupture disk setpoint in a cooling requires that the water absorb the heat generated 1 severe accident, combined with the already low core in the debris bed and the latent and sensible heat released A damage frequency and reliable containment heat by the debris as its temperature decreases. Quenching removal, produces an extremely low probability of prevents or mitigates core concrete interaction (CCl). significant fission product release, in addition, the An overlying water pool also scrubs fission products elapsed time to rupture disk opening is greater than 24 which may be released from the debris bed. hours for most severe accident sequences. The flooder system is comprised of ten piping The net risk reduction associated with the lines. Each line originates in one of the ten vertical
- h)
( implementation of the COPS system in the design of the ABWR is summanred in Table 19E.2 27 and Figure 19E.2 22. All sequences which would result in pipes which are part of the drywell to wetwell connecting vent system. The vents are arranged symmetrically around the perimeter of the lower COPS operation were assumed to lead to failure of the drywell. The flow through each flooder line will be drywell head. This may slightly overpredict the initiated by triggering a fusible plug at the line exit - P probability of drywelll~ad failure since there will be (lower drywell side). Since four inch diameter fusible A somewhat more time available for the recovery of disks may be commercially available, the flooder line containment heat removal if the COPS system were diameter was chwen as four inches, f- not present. Table 19E.2 26 indicates a low probability 3 of RHR recovery in the interval between the time of The tenon disk resides between the stainless steel "j COPS initiation and the time of drywell head failure if COPS were not present. For the case with firewater disk and the fusible plug in the flooder valve. its purpose is to insulate the fusible plug from the addition to the containment, the probability of RHR relatively cold suppression pool water, if insulation recovery during the period of interest is 4%. Therefcre, was not provided, melting of the plug might not be j . no significant error is intmduced into the calculation, uniform and operation of the flooder valve might be : impaired. The disk will not melt or stick in the valve ! '. Table 19E.2 27 indicates that the probability of because tenon has a softening temperature of 4 drywell head failure increases by a factor 50 for approximately 400aC and a maximum continuous sequences with core damage (Classes I and !!!) if the operating temperature of 288'C both of which are COPS system is not present. For Class 11 sequences, above the plug melting temperature of 260 C. I the lou of containment heat removal may lead to core Furthermore, teflon has high chemical resistance and i damage for those sequences which have drywell head w 11 not adhere to the stainless steel plug nor the failure. Since the probability of drywell head failure fusible plug. increases by a factor of 100 without the COPS system, the core damage probability astwiated with Class !! The minimum acceptable Dow rate for the flooder ' events also increases by a factor of 100. Figure system corresponds to the flow rate which can just 19E.2 22 shows the probability of exceedence versus absorb the heat generated in the debris bed. Minimum whole body dose at 1/2 mile for the ABWR and for the acceptable now is calculated in Section 19E.2.8.2.2 ABWR without the COPS system. The offsite dose is The expected flow rate in the Gooder system can be ,
%wn 19E1 4 N
s q ABWR 2 mms Standard Plant REV.A The expected flow rate in the flooder system can be active, this system flow corresponds to a minimum obtained by applying Bernoulli's equation to the 'looder individual line flow of 2 liters /sec. geometry. Th!s calculation is presented in Section 9 19E.2.8.2.3. 19 E.2.8.2.3 Espected flooder Flow Hate - q 4 19E.2.8.2.2 Minimum Acceptable Flow De now rate through the flooder system will be i i,
~
Rate governed by the flow area, the hydrostatic driving head l 4 and head losses in the lines. ? I L lleat is generated in the debris bed by fission V t product decay and tirconium osidation. Any Gooder T1e flow area depends on the diameter of the floo(lcr flow in excess of the amount required to remure lines and the number of lines that are participMing. '3 generated heat ull participate in quenching the debris Assuming that one flooder fails to operate, the flow age and establishing a water pool above the debris bed. As is shown in Attachment 19EC, the time seguired to quench the debris is not a critical parameter in determining n containment performance. Therefore, the minimurn d (1) acceptable flow rate for the lower drywell Gooder system Ar=7lnt is the rate which will completely absorb all the heat generated in the debris bed. = 0.073m 2 The decay heat generation rate at the time when where dr = diameter of lines (0.1016 m.n in), and debris is expected to first enter the lower drywell during credible accident scenarios is approximately one percent nr = number of lines (9, assuming one fails). of rated power (39 MW). Birty.nine megawatts can be used as a first approximation of the decay heat The elevation of the flooder line exit t< low the generation rate of the debris bed in the lower drywell. water level in the drywell to wetwell connecting vents e This assumption is highly conservative because the determines the hydrostatic head, see Figure 19E.2 23. entire core mass will never completely relocate into the (gV) lower drywell. Furthermore, noble gasses and volatiles Due to steaming in the drywell, the drywell pressure is greater than the wetwell pressure and the u ater level in will escape from the molten debris, carrying away the the drywell.lo.wetwell connecting vents is assumed to decay heat associated with these two constituents be depressed to the bottom of the first row of horizontal l (approumately 20 percent of the total). vents. This leaves a hydrostatic head, Az, of 0.375 meters to the inlet of the flooder lines, lleat can also be generated in the bed by exothermic reactions of the debris constituents. He most energetic Form and frictional head losses decrease the flow reactions involve oxidation of zirconium by water vapor through the Hooder lines. Form losses are due to and carbon dioxide. The only source of significant entrance and exit effects as well as the 906 cibow and amounts of oxidizing agents is the concrete beneath the valve, A loss coefficient, k, of 3 conservatively debris bed. The water above the bed will not contribute accounts for all the head losses in the floocky system, significantly to oxidation because the surface of the bed will form a crust which will quickly be depleted of Applying Bernoulli's equation to steady, rirconium. NUREG-5565 indicates that a typical irrotational flow and assuming that the level of the ablation rate for concrete is two inches per hour, The suppression pool does not change (since the surface area generation rate, assuming that the 110 2 and C72 of suppression pool is much greater than the flooder released during ablation completely react with flow area) yields a flooder now rate of zirconium, is 3.6 MW, Combining these two sources l of heat yields a debris bed heat generation rate of 43 MW. 2W sn= Ar 1+k (2) The heat absorption capability of the suppression 3 pool water is 2,350 hU/m3. Therefore, the minimum = 0.099m /sec acceptable flow rate for the lower drywell flooder system is 0.018 m 3/sec (18 liters /sec). Assuming a four inch where t nis the total volumetric flow raic through nine throat as discussed in Section 19E.2.8.2.3, this flow lines and g is the acceleration of gravity. For a liquid can be provided by two lines of the lower drywell density of 980 kg/m 3, this corresponds to a system (Q 1 . flooding system. Alternatively,if nme flooder imes are now rate of 97 kg/sec and an individual line now rate l Amendmem .n 19E.2 42 21
l ABWR 23A6to0As Standard Plant su.v. A o l of 10.8 kg/sec. This is the expected now rate through to the flooder waler, the rate and time to fdl the lower ! the flooder system assuming complete expulsion of the drywc!! are i fusible plug and minimum hydrostatic drivmg head. tra, = 0.080m' / sec
,4 19 E.2.8.2.4 Time to Fill Lower Dr)uell a
tra, = 21 minutes
? Water that enters the lower drywell provides
" cooling to the debris bed. It also establishes an overtymg liquid layer. Neglect ng the subcooling of the The maximum heat flux from the surface of a v ikuler water, heat transfer from the debris bed to the debris bed that has been experimentally observed (see 15 water will result in vaporitation. De amount of fbler Section 19 Ell.2.2) is 2 MW/m2 . He lower drywell has i flow w hich is vaporized is a surface area of 88.25 m2, Thus, the maximum cooling g rate of the debris bed, Omn, is 177 MW. For this heat o transfer rate, the rate and time to fill the lower drywell Q OI D k eep " hgg p We v ram, = 0.022m' / sec where ,,, a volume rate at w hich Ikuler water is vaporiral, Ira.., = 1.3 hours D = heat transfer from the debns bed to in practice, this high heat flux is not expected to be the flooder water, snaintained as the debris is quenched. Nonetheless, the time to fill the lower drywell to the elevation of the hrs
= latent heat of vaporitation of water, Ihler exit will be bounded by these two values,21 minutes and 1.3 hours. This difference in timing will pq = density of water, not have a significant impact on the fission product (o
\ ne amount of '%xler flow w hich can contribute to release from the containment since the steam produced during debris quenching will carry any fission products a cased during this time into the suppression pool. filling the lower dr, ellis m 19 E.2.8.2.5 Consequences of One Flooder d f ra = in - 0..p (4) Line Opening First j The time to fill the lower drywell to the exit of the Core debris that enters the lower drywell will bc x Dooder is distilbuted fairly uniforml" The lower drywell floor lf was designed so that debris spreading would not be g
,fS , h g hindered. The temperature of the lower drywell air space g and suuctures should be even fr. ore uniform because of vfg convective and radiative heat transfer from debris m terial, Conler regions will tend to absorb more heat where Vrin is the volume of the lower drywell below an wannemnes m ng n tewum equahzahn, the Dooder exit. The flooder exit will be 1.15 meters above the lower drywell floor.,The surface area to the liowever, if highly non umform debris dispersal lower drywell floor is 88.25 m*. Rus' occurs, it has been postulated that one ikxxler line could open and its operation could delay or even prevent the 3
Vrm = 101.5m other lines from activating. In the worst physical case, the initiation of one flooder line causes crust formation Flooder actuation is expected to occur without completely quenching the debris. The crust approximately five hours after reactor scram during limits heat transfer from the surface of the debris bed, most severe accident scenanos. T me decay heat level at Core-concrete interaction (CCI) will occur if surface this time is approximately one pes :ent (lot) of the rated heat transfer is reduced enough. power. Assuming the entire core relocates to the lower drywell, the debris bed will have a decay heat generation CCI results in large quantities of gases being O rate,Q3 of 39.26 MW. If all of this heat is transferred formed under the surface of the crust. The gases will h increase in pressure due to continued generation until AmenJmeni ?? 3gE.2 42.22
ABWR 23 , ,, Standard Plant nry A O the crutt ruptures or they escape from the edres of the bed. In either case, the gases will pass from the debns bed into the lower drywell airspace. The panagt (th r espels the remainder of the plug. the stainless steel dak and the teflon disk. will be unobstructed with gasses exiting the uebn? The valve opening ume is the time requ red to melt above the water elevation or through an overlying layet the fusible metal in the annular groove. To eshmate the of water. Since only one Dooder line is presumed acuve, opening time, a calculation has been made for a pure the water layer, if it exists, will be thin and no bismuth plug. Bismuth was used because it has the significant amount of heat will be transferred from the closest melttng point 10 533 K. gas to the liquid. Heat transfer from the surroundmg stainless stect Concrete has an ablation temperature of pipe to the plug is by conduction. Heat transfer from approximately 1500 K. The released gases from core steam in the lower drywell to the stainless steel pipe is by convection. The pipe also receives radiative heat l concrete interaction Higher ternperatures maywill be be at least reached by theatgases this as temperature. from the debris on the lower drywell fkor. Heat transfer they interact with debris material in their etit. Thus, to the bouom of the valve was neglec'ed. The debris bed m gases enter the lower drywell air space at very high surface temperature and lower drywell gas temperature 3 temperature. The CCI gases will increase the were estimated using a representative MAAP.ABWR temperature of the lower drywell air space. More Gooder sequence. Using these assumptions, the valve opening , lines will become active as the lower drywell time was calculated to tu less than approximately 10 y minutes depending on the steam absorblivity, This is a l ternperature increases. a single flooder line For thisatreason, is transient condition worst and therepresentative activation time of from when the lower drywell gas t is not expected to adversely affect the operation of the space reaches 533 K until the flooder line becornes
,} other lines. active.
w 19 E.2.8.2.6 Valve Opening Time 19E.2.8.2,7 Estimation of Net Risk 9 N &
% The fusible plug valve is de !gned to open when in order to assess the net risk of the passive flooder t-p) g U
g the lower drywell temperature reaches 533 K. The fusible material is made up of an alloy mixture of two system, a sensitivity study was performed using three failure probabilities for the passive flooder node, P. in 5 ch hor more o { f anumony,f the the containment event trees. In these cases, the failure f tellurium, zinc following metals: and copper, Alloy contents tin, siint, ofbismuth, probability the passive flooder was increased froin its $ are chosen so that the plug melts when its temperature base case value of 0.001 t0 0.01,0.I, and 1.0. jf reaches $33 K. } As indicated in Table 19E.2 28, the overall results d The melting points of the individual metals are as are not sensitive to this parameter. Failure of the follows: passive flooder leads to an increase in the probability of Melting Dry CCI. Thus, the probability of Dry CCI increases Mdal Point 00 by one, two and three orders of magnitude, respectively for the three sensitivity cases. However, the base case Antimony (Sb) 903 results for Dry CCI are so small that a three order of Bismuth (Bi) 544 magnitude increase does not impact other results significantly. Copper (Cu) 1356 The principal conclusions of the sensitivity studies Silver (Ag) 1233 are Tellurium (Te) 722 (1) Pedestal failure does not increase since it is Tin (Sn) 505 dominated by the Wet CCI sequences. Zinc (Zn) 692 (2) The only probabilistic output which shows any The basic configuration of the fusible plug valve is significant variation is drywell head seal shown in Figure 19E.2 24. The plastic cap has a overtemperature leakage (Pen OT) which exhibits melting point much lower than that of the fusible plug. a two fold increase for a two orders of magnitude Flow initiation occurs when the small annular grove, increase in the pssive flooder failure probability, 2.0 mm in depth, melts. Hydrostatic pressure then and a ter fold increase for a three order of A magnitude increase. The change in seal leakage is V; Amencknem 77 1E2423
ABWR 2mims Standard Plant REV.A much less than the change in passive flooder failure probability since high RPV pressure De ABWR has two drain sumps in the periphery sequences with entrainment of debris to the upper of the lower drywell floor which could collect core drywell and failure of the u per drywell sprays debris during a severe accident if ingression is not dominate the seal leakage sequences in the base prevented. If ingression occurs, a debris ted will form analysis. in the sump which has the potential to te deeper than the bed on the lower drywell floor. Debris cootability (3) Even for the caw where the passive flooder is tecomes more uncertain as the depth of a debris bed assumed to be unavailable, the probability increases. Therefore, debris should te kept out of the associated with the Dry CCI is only 3.5E.10. sumps. Since only the Dry CCI cases have failure of the O fassive flooder, this frequency represents an upper The two drain sumps have different design C bound for the impact of passive flooder failure on objectives. One, the loor drain (HCW) sump, collects offsite dose. water which falls on the lower drywell floor. The other, 5 the equipment drain (LCW) sump, collects water 4 Thus,it is seen that the lower drywell flooder does leaking from valves and piping. Both sumps have not affect net risk for probabilities above 3E 10. pumps and instrumentation which allow the plant Therefore, no chart of the impact on risk was created. operators to determine water leakage rates from various The value of the COPS system is not in a direct impact sources. Plant shutdown is required when leakage rate on risk. Rather, it should be viewed as a passive limits are exceeded for a certain amount of time. A system which serves to limit the impact of uncertainty more complete discussion on the water collection in operator actions and allows the ABWR design to system can be found in Section 5.2.5. mitigate a severe accident in a purely passive manner. Debris will be prevented from entering into the 19 E.2.8.2.8 Summary lower drywell sumps by shield walls (corium shields) built around their periphery. De shields will be The passive flooder meets its design goal of constructed from material which will prevent or O preventing or, at least, mitigating core concrete interaction in the lower drywell. The flow rate required to remove the heat generated in the debris bed is minimize interactions with the core debris. nc shield for the floor drain sump will have channels at floor level that allow nearly unrestricted water flow at rates 0.018 m3/sec which can be provided by two of the ten on the order of and somewhat greater than the leakage flooder lines. De espected flow rate is 0.099 m3/sec limits. The channels will be sired so that they plug ] (nine of the ten lines active),if the expected flow rate is with core debris during a severe accident; thus 2 achieved, a one meter layer of water will be established previnting debris ingression into the sump. The above the bed in a time between 21 minutes and 1.3 equipment drain sump will be solid. A complete g hours after flow initiation. One flooder line opening description of corium shields can < found in w first is not espected to prevent the other lines from Attachment 19ED.
'2 opening during a severe accident in which significant amounts of core debris is present in the drywell. The flooder lines will become active within ten minutes of the lower drywell gas space reaching 533 K. De passive flooder has negligible impact on the net risk of the plant since it provides a redundant function to the fliewater addition system.
19E.2.8.3 Corium Shield During hypothetical severe accident in the ABWR, mohen core debris may be present on the
~
lower drywell floor. De EPRI ALWR Requirements G Documeni specifies a floor area of at ienst E 0.02 m2/MW, to promote debris cootability, his has { been interpreted in the ABWR design as a requirement 4 for an unrestricted lower drywell floor area of 79 m2, .S Amendmeni ?? 19E.2-42.24
7 _ _ _ - _ ._._ _ _ . _ 4 ABWR
- Send:rd Plant 234 m as uv. 4 19E,2,9 References
- 15. Crane, Plow of fluids Through Valves, Fitung, and Pipe. Technical Paper 410,1969.
- l. MAAP.3.0B Computer Code Manual. EPRI NP.
7071 CCML, November 1990.
- 16. NEDE 22056 Revision 2, Reliability Analysis
- 2. Advanced Licht Water Reactor Utility Data Manual General Electric Company, January 1986.
' Reouirements Document, EPRI Report NP4780 L.
- 17. A.M. Rozen. S.I. Golub and T.1. Vitintseva.
Calculating the Trarsported Entrainment During
- 3. Advanced Reactor Severe Accident Program, Sparging, Translated by Polyglot Language Technical Supportfor the EPRI Debris Coolability SI
Requirementfor Advanced Light Water Reactors, Task 8.3.J.6, Fauske and Associates, Inc., Burr 18. S. Xutateladze, Elements of the Hydrodynamics of Ridge,IL, August 1988. Gas Liquid Systems, Fluid Mechanics . Soviet Research, 1, 4, 1972. 4 NUREG/CR.3920, SAND 841246, Rev. 3, CGRCON MOD 2 User's Manual August 1984, 19. Nuclear Regulatory Commission. Severe Accident Risks: An Assessment for Five U.S. Nuclear
- 5. R.E, Henry, Key Phenomenolgical Models for Power Plants. NUREG 1150, June 1989.
I O cchn cal Re 4IA 1983. 20. Nuclear Regulatory Commission. Evaluation of Severe Accident Risks: Peach Bottom, Unit 2,
- 6. R.E. Henry, Key Phenomenological Models for NUREG 4551 Volume 4, December 1990.
Assessing Explosive Steam Generation Rates, IDCOR Technical Report 14.1B,1983, 21. Nuclear Regulatory Commission. Evaluation of Severe .1ccident Risks: Grand Gulf, Unit 1,
- 7. NUREG 75/014, WASH 1400, Reactor Safety NUREG-4551 Volume 6. December 1990.
O Study: An Assessment of Ace! dent Risks in U.S. CommercialNuclear Power Plants,0ctober 1975. 22. Nuclear Regulatory Commission. Individual Plant Examination: Submittal Guidance, NUREG.I 335,
- 8. T.G. Theofanous, Scaling Condderations in Steam AUGU I989' Explosions, Procedures 1987 National Heat Transfer Conferenc'e, Pittsburgh, August 912, 23, Nuclear Regulatory Commission, Generic Letter 1987, pp. 58-67.
No. 88 20, Individual Plant Examination for Severe Accident Vulnerabilities, November 23,
- 9. Sir Horace Lamb, Hvdrodynamics, Dover. I988*
- 10. F. Kreith, Princinlet of Hem Trnncfer. 3rd Editinn. 24. Probabilistic Risk Ateettment for Kuosheng IEP A Dun Donnelley Publisher, N.Y.1976, u er n Repun oh Atomic Energy Council, July 1985.
I1. PJ. Berenson, film Boiling Heat Transferfrom a
" " ^ ' 25. Electric Power Resean;h Institute, Recommended Sen s o ?8 , 96l' Sensitivity Analyses for an Individual Plant Examination Using MAAP 3.08, Draft 1992.
I2. G.A. Green. Experiments on Melt Spreading and Bubbling Heat Tranger, Severe Accident Rescan;h 26. Nuclear Regulatory Comm. ission, MAAP 3.0B Partners Meeting, Bethesda, MD, October 17 21, Code Evaluation Final Report Draft, December 1981. I99I'
-13. FJ. Moody, Pressure Suppression Containment 27. Advanced Linht Water Reactor Utility Thermal Hydraulics State of the Art, NUREG/CP. Reauirements Document. Volume !!, Chapter 1, 0014, Volume 1,1980.
Appendix A: PRA Key Assumptions and Guidelines, EPRI Report NP-6780-L. 14, 22A7007 Revision 21, GESSAR II,238 Nuclear Island. General Electric Company. Amendment ?? 19E.2 43
, __ . ._ - . . . _ . . . _ . . . . . . _ _ _ - _ . , - . . - . ~ . . - . . , , . . _ . . . , _ . _ - _ , _ _ . . . . . --.-_.m.... .. .,
s ABWR Standard Plant
**$$,j
- 28. R. O. Gauntt, R. D. Gasser, L. J. Ott, The DF-4 l.
()) Fuel Damage Experiment in ACRR sah a BnR Control Blade and Channel Box, NUREGICR- \ 4671, SAND 86-1443, November 1989
- 29. R. Tokrar and R. Libby, Recriticality in a Boiling
%'ater Reactor Following a Core Damage Accident,
- Proceedings of the 17th Water Reactor Safety 1 Meeting, October 1989.
- 30. K.C. Wagner, Analysis of a High Pressure A1WS with Very Low Alakeup Flow, DOEllD 10211, f
October 1988. I
/
( ) s._/ I \
'%)
* *"** U 19 E.2 -4 3. I e
4 4 ABWR 2mimas j Standard Plant Rn A a i ! TAHLE 19E.21
; POTENTIAL SUPPRESSION POOL HYPASS LINES i 1 i
l' PATHWAY BASIS FOR , 1 NUMBER SIZE (mm) ISO 1ATION EXCLUSION ! DESCRIPTION OF LINES E110M IQ (1 in. - 25.4 mm) VALVES (SEE NOTES) i 1 Main Steam 4 RPV ST 700 (AO, AO) -
- Main Steam Line Drain 1 RPV ST 200 MO, MO 3 ;
l Feedwater 2 RPV ST 550 CK,CK - ! Reactor Inst. Lines 30 RPV RB 6 CK - - CRD Insert / Withdraw 103 RPV RB <1 CK, MA 1 ilPCF Discharge 2 RPV RB 200 CK, MO - HPCF Warmup 2 RPY RB 25 MO, MO - HPCF Suction 2 SP RB 400 MO -2 l Supp PoolInstrumentation 6 SP RB 6 CK 2 SLC Injectica 1 RPV RB 40 CK, CK - RCIC Steam Supply 1 RPV_ RB 150 (MO, MO) - RCIC Discharge 1 RPV RB 150 CK, MO 5 RCIC Min. Flow 1 SP RB 150 MO 2 RCIC Suction 1 SP RB 200 MO 2 RCIC Turbine Exhaust i SP RB 350 MO,CK 2 RCIC Turb. Exh Vac Bkr 1 SP RB 40 CK, CK 2 MO, CK O RB RCIC Vac Pump Discharge 1 SP 50 2 , RHR LPFL Discharge 2 RPV RB 250 CK, MO - RHR Warmup Lines 2 RPV RB 25 MO, MO - RHR Wetwell Spray 2 WW RB 100 MO 2,4 RHR Drywell Spray 2 DW RB 200 MO, MO 4 RHR SDC Suction 3 RPV RB 350 MO, MO 3 Amendment 24 19fL2-14
ABWR mi s Standard Plant n,, 4 q TABLE 19E.21 (Continued) LJ POTENTIAL SUPPRESSION POOL BYPASS LINES PATIIWAY BASIS FOR NUMBER SIZE (mm) ISOLATION EXCLUSION OF LINES FROM IQ (1 in. = 214 mm) VALVES (SEE NOTES) l DESCRIPTION RWCU Suction 1 RPV RB 200 (MO, MO) . RWCU Return 1 RPV RB 200 MO, MO 5 l RWCU licad Spray Line 1 RPV RB 150 CK, MO, MO 3 RWCU Instrument Lines 4 RPV RB 6 CK - Post Accident Sampling 4 RPV RB 25 (MO, MO) - RIP Motor Purge 10 RPV RB <1 CK, CK 1 l RIP Cooling Water 4 RPV RB 50 MO, MO 1 LDS Instruments 9 RPV RB 6 CK - SPCU Suction 1 SP RB 200 MO,CK 2 SPCU Return 1 SP RB. 250 MO, MO 2 l Cont. Atmosphere Monitor 6 DW RB 20 MA 8 LDS Samples 2 DW RB 30 (SO,50) . Drywell Sump Drains 2 DW RB 100 MO, MO . HVCW/RBCW Supply 4 DW RB 100 CK,MO 1 HVCW/DWCW Return 4 DW RB 100 MO, MO 1 DW Exhaust /SGTS 2 DW RB 250 AO,AO 7 Wetwell Vent to SGTS WW RB 250 AO,AO 2 O 1 DW Purge 1 DW RB -300 AO - WW Inerting/ Purge 2 WW RB 550 AO 2 Instrument Air 2 DW RB 50 CK, MO 1 l SRV Pneumatic Supply 3 DW RB 50 CK, CK 1 Flamability Control 1 DW RB 100 (MO,MO) 3 ADS /SRV Discharge 8 RPV WW 300 RV . ACS Crosstie 2 DW WW 550 AO,AO . l WW/DW Vacuum Breaker 8 DW WW 500 CK . Miscellaneous Leakage 1 DW RB ... NCNE 6 Access Tunnels 2 DW RB ... NONE 6
' 19E2-l$
Amendment 24 1
i
- ABWR maim ^s nu x Standard Plant TABLE 19E.21 (Continued) i POTENTIAL SUPPRESSION POOL HYPASS LINES LEGEND AND ACRONYMS i PATilWAY significantly reduced due to decay and other
! Source (From) Termination (To) removal mechanisms j RPV Reactor Pressure Vessel WW W e t w e il 3 DW Drywell RB Reactor Dldg 4. Some lines which originate in the primary j SP Suppression Pool WW Wetwell containment are designed for operating pressures ST Steam Tunnel higher than would be expected in the containment isolation Valve Types during a severe accident. These lines (with design j pressures greater than about 100 psig) were j AO Air Operated excluded since the probability of a break under
- MO Motor Operated less than normal operating pressures and i RV Relief Valve coincident with the severe accident is extremely CK Check Valve small, i MA Manually Actuated i SO Solenoid Operated 5. Some lines return to the feedwater line. These j () Common Mode Fallure Potential (See pathways (such as LPCF loop A and RWCU) are Section 19E2333 (2)) excluded since they are bounded by the evaluation
; of feedwater.
Bates for Exclusion
- 6. Acceptable long term leakage from the drywell to
- 1. Closed systems such as closed cooling water the reactor building following a design basis i systems which do not directly connect to the accident is specified at .4% of drywell volume per RPV or containment atmosphere require two 24 hours. During severe accident conditions this
,O failures to become a bypass pathway; a leak or break within the cooled component and a line break outside of containment. Very low leakage could be somewhat greater due to higher than design basis containment pressure. Ifowever, the contribution of this leakage to overall risk is l
flow is expected out of the break or leak at ignored because this leakage is through numerous l the cooled component is likely due to the tortuous passages of small diameter which provide high degree of restriction. These pathways ample opportunity for plateout and plugging are not considered further on the basis of this effects (see subsection 19E2.13.4) A discussion of ! very low flow rate. Similarly, extreme the drywell access tunnels is included in section j restrictions in CRD seals provides the basis 19F. for excluding those lines. !' 7. Drywell purge lines are normally closed and fail
- 2. Pathways which originate in the primary closed. The potential for inadvertent opening is
- containment wetwell airspace or the considered remote and is addressed by suppression pool are excluded because fission Emergency Procedure guidelines.
product aerosols would first be trapped in the suppression pool and would thus not be
- available for release through the bypass path.
- 3. Some lines are closed during normal plant operation and would not be expected to be opened in the short term following a plant accident. These lines are excluded on the basis of low frequency of use. Furthermore, should a bypass pathway develop later when the line is used, the fission product sourc:
~ term would be ex' pected to have been already Amendment 24 19E.246
.~ ,
i i ABWR nuims i Standard Plant MA 1
] TABLE 19E.2 2 ABWR PLANT ABILITY TO COPE WITH STATION BLACKOUT i FOR UP TO 8 HOURS
- Plant Parameter Deslan Ratin Value Station Blackogtjagg
- ) RPV Level Core covered Core covered RPV Preuure 50 psig RCIC trip > 150 psig 150 psig RCIC rated flow b) D.C. Battery Capacity 11,400 amp brs Sufficient with load
{ 1 Div.1,2,3 & 4 shedding c) RCIC Water Source 1) CST 20 x 10 ft CST sufficient with
- 2) Suppresgorgpool- RPV preuurirzJ l 126 x 10 ft j d) RCIC Roorn Temperature 151 F <1$1 F c) DrywcuTemperature 340 F < 340 P f) Dryweu Preuure 4$ psig < 4$ psig g) WetweUTemperature 219F < 219 F h) Wetweu Preuure 45 psig <45 psig i) Contr<d Rooms i yfain 122 F c 122 F 4
lower 122 F < 122 F i Computer 122F < 122 F i a i O Amendment 22 19E.247
- - . . _ _ . _ _ - . . . . _ _ _ .-_ . . _ . _ . _ . _ . . , _ . - _ . _ . _ __t
23A4100A5 bA
- . Standard Plant i
TABLE 19E.2 3 j DEFINITION OF ACCIDENT SEQUENCE CODES ! charneters 1 to 4 General condition Indientor i LCLP Loss of All Core Cooling with Veuel Failure occurring at low Preuure LCHP Loss of All Core Cooling with Veuel Failure occurring at High Preuure j SBRC Station Blackout with RCIC operating for 8 hours i i LHRC Loss of Heat Removalin the Containment
- LBLC Large Break LOCA with IAu of All Core Cooling I NSCL Transient without Scram and with Failure of All Core Cooling, Vessei Failure occurs at Low j Pressure d
NSCH Transient without Scram and with Failure of All Core Cooling. Vessel Failure occurs at High ! Pressure 5 NSRC Station Blackout without Scram, RCIC operates i j Characters 5 and & Mltinatina Features i 00 No mitigating features operated IV in Veuel Recovery i j PF Passive Flooder FA Firewater Addition System injects into the Veuel l HR Containment Heat Removal ! PS Pauive Flooder and Drywell Spray Firewster Addition System switched to Drywell Spray Mode f FS 1 19 M 4 Amendment 22 i
i 1 ABWR Standard Plant nx.iao4, x,y 3 i, TABLE 19E.2 3 (Continued) i DEFINITION OF ACCIDENT SEQUENCE CODES '
)
i i ,i i --Character 1: Mode of Release N Normal Containment Leakage i First release via leakage through Moveable Penetrations l lP R Overpreuure Protection Relief Rupture Disk l D Drywell Head Failure { 4 t E Early Containment Structural Failure S Supprculon Pool Failure-1 i Character 8: Mannitude of Release O No core damage, no fiuion product release l l N Negligible: Leu than 0.1% volatile fission products !- L Law: 0.1% to 1% volatile fluion products l M Medium: 1% to 10% volatile fission products 1 H High: More that.10% sotatile fluion products l , l ! 1 1 i 4 i i 1 i 19E.2 49 Amendment 1- n i 4 t d de - a.r . ,,w -. ,r-+-,, - . -
- eewwe-,w-.-,--r, wr-- et.,+<-.+, - * - - ,r,~c-w-+,y,m-c.-.-s. -,rw--,e-y w,-we--->-w.-mm- , ,.-,-v--gwey - % y-
ABWR HAHMAS Sinndard Plant neu l TA11LE 19E.2 4 Grouping of Acciden'. Classes into llase Sequences Accident Class Initiator Code Base Scauence Subsection Number IA LCilP 19E.2.2.2 18 1 - LCLP 19E.2.2.1 LCllP 19E.2.2.2 111 2 SilRC 19E.2.2.3 iB.3 LCLP 19E.2.2.1 LCilP 19E.2.2.2 IC NSCL 19E.2.2.6 ID LCLP 19E.2.2.1 IE NSC11 19E.2.2.7 11 LilRC 19E.2.2.4 lilA LCHP 19E.2.2.2 lilD LDLC 19E.2.2.5 IV 1 NSRC 19E.2.2.8 ( Y Amendment 22 19E 2 30
ABWR nuiw4s
,,v s Standard Plant TAllLE 19E.2 5 Sequence of Esents for LCLP PF R N bas of All Core Cooling with Vessel Failure at Low Pressure l Passhe Flmder Operates and Rupture Did Opens list fdint 0.0 MSIV Closure 4.2 see Reactor Scrammed 0.4hr Indicated Water Level at :/3 Core lleight One SRV Opened by Operator 1.8 hr Vessel Failed 2.7 hr Water in Lower Drywell Boiled Off Corium lieatup Begins 5 4 hr Passive Flooder Opens 20.2 hr Rupture Disk Opens O
5 19E151 Amendment c)
ABWR n^um^s Standard I'lant Me, A TAllLE 19E.2 6 O seu#e eeerc'e isree'c'e es a s Lou of All Core Cooling with Vessel railure at Low Pressure Firewater Addition Sptem injects and Rupture Disk Opens l limt foun! 0.0 MSIV Closure 4,2 sec Reactor Scrammed 0.4 hr Indicated Water Level at 2/3 Core lleight One SRV Opened by Operator 1.8 hr Veuel Failed 2.7 hr Water in 14wer Drywell Boiled Off Corium }{eatup Began 4.0 hr Firewater Spray Started 7.0 hr Suppreulon Pool Overflows to the Lower Drywell 23.6 hr Firewster Spray Stopped 31.1 hr Rupture Disk Opened p& 56.6hr Water in Lower Drywell Boiled Off 61.1 hr Passive flooder Opcned 19EIS2 Amendment g v)
ABWR nuiwu peu Standard Plant Table 19E.2 7 l Sequence of Events for LCllP PS R N Loss of All Core Cooling with Vessel railure at liigh Pressure l Passise nooder and Dry 4cli Spray Operates, Rupture Disk Opens list Eid 0.0 M SIV Closure 4.2 see Reactor Scrammed 0.3 hr Core Uncovered 2.0 hr Vesnel Fails Corium and Water Entrained into Upper Drywell 2.0 hr Passive Mooder Opens 4 0 hr DrycIl Spray Initiated 25.0 hr Rupture Disk Opens O V 19EM3 Amendment f
' w );
ABWR n46im4s Standard Plant N tv] Table 19E.2 8 O O I Sequence of Events for LCilP PF P.M Loss of All Co e Cooling with Vessel Pailure at High Pressure Passisc flooder Operates, Penetration leakage Occurs l lima IIJCD1 0.0 MSIV Closure 4.2 see Reactor Scrammed . 1 03 hr Core Uncovered 2.0 hr Vessel rails Corium and Water Entrained into Upper Drywell l 2.0hr Passive Flooder Opens 2.1 hr Seal Degradation Temperature Reached 18.1hr Leakage Begins through Moveable Penetrations Ftssion Product Release Begins O 19E1.4 Amendment
/ \
t 1 V
. i ABWR DA6lWA5 pn A j Standard Plant l O Table 19E.2 9 V i i Sequence of Events for SilRC FA R 0 Station Blackout with RCIC Operational for 8 llours
- l Firewater Addition to Vessel Used to Prevent Core Damage, Rupture Disk Opens Ilmt L.tal 0.0 MSIV Closure 4.2 See Reactor Scrammed
$2 see RCIC Injection, Suction Itom CSP l
1.3hr RCIC Suction Switched to Suppression Pool 4.4 hr RCIC Suction Switched to CSP l 8.0 hr RCIC railure 9.0hr Suppression Pool began to overnow to Lower Drywell 9.8 hr Manual ADS 9.9 hr Collapsed Water Level Fed to 2/3 core height Firewater A*ddition System Injection Degan O
- L) 32.3 hr Rupture Disk Opened 19E 2 SS Amendment
\
v
_ . . _ _ _ _ __. _ _ _ _ _ - - - _ _ _ - _ _ _ m.. _ . - - . _ . 4 J l ABWR n4uuoi, . Standard Plant a,a i !. Table 19E.210 l l Sequtnce of Events for SBRC.PF.R.N ] Station Blackout with RCIC Operational for 8 Hours , l l Passive Flooder Operates and Rupture Disk Opens l l ! j liDit EitId 1 j 0.0 MSIV Closure j 4.2 See Reactor Scrammed i
- 32see RCIC Injection, Suction from CSP l
l 13 hr RCIC Suction Switched to Suppression Pool I 9 l 4,4hr RCIC Suction SMtched to CSP l 1 T
- 8.0hr RCIC Failure i 93 hr Core Uncovered l 9.7 hr One SRV opened by operator 123 hr Vessel Fails l 21.1 hr Lower Drpell Water Boils Away
- =
j' 23.5 hr Passive Flooder Opens Rupture Disk Opens 1 4 i N Amendment 19E.2 M d
T'w '-tw=tr- q v --W y r e-+e-- v e, v m y- 4,, g.m-m.,, .,v.m . , , , , , , _ , , _
tGWR nuims i ,o l Standard.Elant I Table 19E.2.ll i o I Sequence of Events for LilRC 00.R 0 Isolation with Loss of Containment flest Removal l Rupture Disk Opent { I lima Eun! 0.0 MSIV Closure 4.2 see Reactor Scrammed 1.1 min RCIC Injection 2.9 hr Manual Open 1 SRV 3.0 hr llPCF Injection 3.1 hr RCIC Trip on Low Turbine Pressure 21.7 hr Rupture Disk Opens
> 72 hr Potential Loss of Core Cooling O
C/ I9b.) $7 Amtndment 6
'N_f
ABWR niaiwas Re u Standard Plant Tal.le 19E.212 l Sequence of Esents for LDLC PF R N Large Break LOCA With Loss of Core Cooling Pauive Mmder Operates and Rupture Disk Opens l litut E1181 0.0 Main Steam Line Dreak b.2 see fligh Drywell Pressure Signal l 4.4 see Reactor Scrammed 14.9 see MSIV Closed 2.8 min Core Uncovered ) i Vessel Failed ! 1.4hr l 5.7 hr Passive nooder Opened 19,1 hr Rupture Dith Opens O 19E: 58 Amendment [ \
ABWR 23mimu Standar31.Elat11 Rev A TAllLE 19E.213 r3 1 Sequence of Esents far NSCL PF II N (y I Concurrent ten of All Core Cooling and ATWS with Vessel failure at Low Pressure l Pauive Monder Operates and Rupture Disk Opens limg thint 0.0 MSIV Closure 17 min Core Uncovered 0.5 hr One SRV Opened by Operator 13 hr Vessel Failed 1,9 hr Water in lower Drywell Boiled Off Corium lleatup Begins 4.4hr Pauive nooder Opens 18.7 hr Rupture Disk Opens i v Amendment 19FJ 59 f%
ABWR nA6imas Standard Plant Rev A Table 19E.214 o ld l Sequence of Events for NSCil PF P M Concurrent Loss of All Core Cooling and ATWS with Vessel railure at liigh Pressure Passive rimler Operates, Penetration Leakage Occurs l , um ca j 0.0 MSIV Closure 3 6 min Core Uncovered 1.3 br Vessel Fails Corium and Water Entraired into Upper Drywell 1.4hr Passive Flooder Opens 1.4 hr Seal Degradation Temperature Reached 17.8 hr Leakage Begins Through Moveable Penetrations Fission Product Release Begins G 4 Amendment 19E W (M.
\ )
ABWR nyim, Standard Plant , Table 19E.215 Sequence of Events for NSRC PF R N Concurrent Station Blackout and ATWS Passive Flooder Operates, Rupture Disk Opens l limt E11nt 0.0 MSIV Closure 4.1 min Core Uncovered 1.9 hr Suppression Pool Began to Overflow to Lower Drywell . 3.6 hr RCIC Tripped 3.8 hr SRV Opened 5.6 hr Vessel Failed 8.6hr Rupture Disk Opened Fission Product Release Began [ v 19E.241 Amendment L
23A6100A$
,,, x
- Standard Plant-Table 19E.216 i
, . Summary of Critical Parameters for Severe Accident Sequences i Time of Fission Time of Vess:1. Product - Rupture Disk End of Csl Release Fraction of Release Time Opening Csl #t 72 hours i Acetdent Emilutg Release 1 i LCLP PF R N 1.8 hr 20.2 hr 20.2 hr 100 hr < 1E 7 - LCLP FS R N 1.8 hr ,31.1 hr 31.1 hr 76 hr < 1E 7 LCHP-PS R N 2.0 hr 25.0 hr 25.0 hr 50 hr <1E 7 LCHP FS R N 2.0 hr So hr* $0 hr* 125 hr' <1E 7* LCHP PF P M 2.0 hr 18.1 hr N/A 70 hr 8.8E 2 i SBRC PF R N 123 hr 23.5 hr 23.5 hr 100 hr <1E 7 LBLC PF-R N 1.4 hr 19.1 hr 19.1 hr 125 hr <1E 7 LBLC FS R N 1.4 hr 29.5 hr 29.5 hr 67 hr < 1E 7 NSCL PF R N 13 hr 18.7 hr 18.7 hr 105 hr <1E 7 NSCL FS R N 13 hr 30.7 hr 30,7 hr 69 hr <1E 7 13 hr 17.8 hr N/A 65 hr 7,3E-2 NSCH PF P M NSCH FS R N 13 hr 50 hr* $0 hr* 125 hr* < 1E 7* l l NSRC-PF-R N 5.6 hr 8.6 hr 8.6 hr 110 hr < 1E 7
\ NSRC FS R N 5.6hr 26.4 hr 26.4 hr 120 hr <1E 7 4
9 i 1 Release param:ters are approximate See sequence discussion for more detail. l I I j i 4
. 19E.242 l Amendmem O
i ABWR i m imas Sfandard Plant _ a., 4 l
- Table 19E.217
! Important Parameters for Steam Explosion Analysis t Syrnbol Value Description m' 500 kg/s Mass Flow Rate of Corium from Vessel 0 0.056 m3 /s Volumetric Flow of Corium from Vessel ! e 7.E 6 m8 /s Thermal Diffusivity of Corium c 480 J/kg K Specific Heat of Corium l y
- ! q 9000 kg/m 3 Density of Corium I
r 4.0 N/m Surface Tension of Molten Corium-h 390 W/me K Heat Transfer Coefficient for Corium Droplet Tg 2600 K Initial Temperat'ure of Corium Droplet q,;, 1.1 kg/m 3 Densityof Air i a 1000 kg/m3 Density of Water vg(P ,) 1.7 m3 /kg Specific Volume of Evaporation for Water hg (P,) 2257 kJ/kg Specific Enthalpy of Evaporation for Water j L 5.5 m Height of Water in Lower Drywell A 88.2 mi Area of Lower Drywell l H 6m Distaace from Bottom of Vessel to Surface of Water in Lower Drywell Amendment 22 19E.243
i I ABWR usumxs a, 4 l Standard Plant i Table 19E.2.18 i Potential Bypass Pathway Matrix 4 3 1 e FROM i ' Wetwell Suppression Dnwell Airsnace Pool {= .IQ REY ? - Drywell No NA NA NA l i ! Wetwell . . l Airspace Yes Yes NA NA l Reactor Building Yes Yes Yes Yes Turbine i Building Yes Yes .Yes Yes I i l
- j. The above matrix shows the paths that potentially 1 bypass the suppression pool if i r i
4 s s
- Pathways which originate in the drywell and potentially release into the wetwell are potential bypass paths if the containment is vented or the wetwell fails during the severe accident.
r-19E.2M Amendment 22 1 1 1
1 ABWR zu6iOOrs i Standard Plant nea j j o Table 19E.2-19 i') Flow Split Fractions Line size now Split Fraction am m RPV Snurre Dnweil snurre 6 0.25 1.5 E 05 54E45 12 0.5 9.4 E-05 3.4 E44 25 1 5.7E44 2.0E43 50 2 3.3E 03 1.2E-02 100 4 1.8FA2 6.2E42 150 6 4 8E 02 15E41 200 8 8 9E-02 2.5 E-01 250 10 1.4E 01 3.6E-01 300 12 2.0E-01 4.6E 01 350 14 2.6E 01 5AE41 (V
\
400 16 3.2E 01 6.2E 01 450 18 3.8E-01 6.7E-01 500 20 4.3E41 7.2E-01 700 28 6.lE-01 8.4 E41 1000 40 7.7E41 9.2E-01 r'~x 4 (' ') Amendment 24 19E.2-65
ABWR me.s Standard Plant no 4- q i 2 .
. _ Table 19E.2-20 4
A
- V a
Failure Probabilities e j_ Symbol Description Prob / Event Egh-P1 Steam pilot operated valve (MSIV) 1.0E 3 a j P2 MSIV Icakage probability z - 7.1E 1 b P3 Turbine Bypass Isolation 4.0E 3 e i~ P4 Main condenser failure - 1.0 'e PS MSL br'eak outside containment 8.0E 6 d. 1 ! P6 Air operated valve (NO) 4.1E 3 e i i P7 DC Motor operated valve (NO) 3.6E 3 e j P8 AC Motor Operated valve (NO SBO) 1.0 f: l P9 Check Valve 8.4E 3 g i j P10 Motor operated valves (NC) ~ 5.0E 1 h (O i - k) P11 Motot operated valves (NC)- 2.8E 4 i P12 _ inadvertent opening 1.0E-3 '- j' 4 P13 Smallline break 2.4E 4 k l P14 Medium line break 1.6E-5 . k-P15 Large line break' 8.0E-6 k i d-i S $i a l
- (h .
Amendment 24 19E.24 3 4-V- 'siw.- -- -%,+w. w --_ w.
y ABWR nuimu Standard Plant n, 3 1 l Table 19E.2 21 Summary of Bypass Probabilities i j Lines from the RPV 4 Bypass
- Flow Split Probability Bypass Bypass Figure Eghs . Fraction Eauation Probability Fraction 10E 2-19 Main Steam 6.7E 1 4'Pl*(P3'P4 + PS) 1.6E5 1.1E 5 A
- Main Steam Leakage 2.2E 5 4*P2'(P3'P4 + PS) 1.1E 2 2.5E 7 A
! Feedwater 5.2E1 2*P9'P15 2.4E 8 13E 8 B Reactor Inst. Lines 3.1E 5 30'P13'P9 6.0E 5 1,9E 9 D { 4
; HPCF Discharge 1.1E 1 2'P9'P10'P14 1.3E 7 - 1.5E 8 ' C HPCF Warmup 1.0E 3 - 2*P10*P11*P13 6.7E 8 6.7E-11 C 4
SLC Injection 3.0E 3 1*P9'P13 3.6E 7 1.1E-9 B l RCIC Steam Supply 6.9E-2 l'P8'P14 1.6E-5 1.!E-6 E LPFL Discharge 1.7E 1 2'P9'P10*P15 6.7E-8 1.1E-8 C LPFL Warmup Line . 1.0E 3 2'P10'P11'P13 6.7E.8 6.7E 11 C l RWCU Suction 1.2E-1 l'P8'P14 1.6E-5 2.0E 6 E . RWCU last Lines 3.1E 5 4*P13'P9 8.1E-6 2.5E 10 D l l 4 Post Ace Sampling 1.0E 3 4*P8'P13 9.6E 4 9.9E 7 J-d LDS Instruments 3.1E-5 9'P13'P9 1.8E5 5.7E 10 D 1 SRV Discharge 6.9E 2 8'P14 13E 4 8.8E 6 K i Total 2.4E 5 I~ 4 , These lines may be excluded for station blackout events Amendment 24 19E..M? J J
1 ! ABWR uuimis i i Standard Plant Rev.A
- , Table 19E.2 21 Summary of Bypass Probabilities (Continued) i e
! Lines from the Dnwell i { Bypass
- Flow Split Probability Bypass Bypass Figure
! Pathway Fraction Eauntion Probability Fraction 19E.2 19 i Cont Atmos Monitor 8.9E-4 6*P9'P13 1.2E5 1.0E 9 D s 3 . LDS Samples 1,7E 3 2*P8'P13 4.8E-4 63E 8 E ! Drywell Sump Drain - 3.0E-2 -2*P8'P13 4.8E 4 1.2E-6 J l
- 'DW Inerting/ Purge 1.4E1 2*P6' 1.5E 3 2.1E 4 1 1
- ACS Crosstic 1.1E 1 2'P12 1.5E-6 1.6E 7 H l WW DW Vac Bkr 2.6E 1 8'P9 6.7E 2 1.7E 2 G SRV Discharge 6.9E 2 8'P14 13E-4 9.0E 6 K
\
Total excluding vacuum breaker 1.5E 5 and ????? lines i l ! Grand Total excluding vacuum breaker 3.9E-5 i and ????? lines I Goal 8.4E-4 c 4 1-4 i
- Addressed on Containment Event Trees.
1 i h 1
- l Amendment 24 19E.2428 l i
~
. , .m..-.
ABWR u uimis REV.A Standard Plant Table 19E.2 22
- NUREG/CR 4551 GRAND GULF APET EVENTS BY CATEGORY q
. et h
- Event d Number Description J
Plant Damace State Grouoine Events
! Initiating Event Type 2 Station Blxkout 3 DC Power Availability 4 S/RV Fails to Reclose -
5 HPCS Failure 6 RCIC Failure Initially 7= CRD Injection Failure 8 Condensate System Failure 9 LPCS/LPCI Systems Failure 10 RHR Failure i1 Service Water /LPCI Crosstic Failure 12 Fire Protection Crosstic Failure 13 Containment Spray Failure 14 Vesxl Depressurization 15 Time Core Damage 20 Plant Damage State Summary Structural CanacirvAnitial Containment Statm O 16 17 Containment Isolation (Pre-existing Leakage) Extent of Pool Bypass Initially 18 Containment Capacity (Quasi static / Dynamic Loadmg) 19 Drywell Capacity (Quasi static / Dynamic Loading) Svstems Behavior'Ocerator Actions 21 Ignitors Tumed On Before Core Damage 22 Containment Vented Before Core Damage 23 SRV Vacuum Breakers Stick Open 26 RV Pressure During Core Damage 27 Status of Hydrogen Ignitors Before Vessel Breach 28 RV Injection Restored During Core Damage 30 Containment Spray Status 53 Upper Pool Dump 81 Containment Spray Status Following Vessel Breach 103 Containment Vented Following Vessel Breach 106 Containment Spray Status Late 119 Containment Vented Late Amendment ?? 19E.2 68.1
AB'WR uwas REV.A
. Standard Plant Table 19E.2 22 NUREG/CR 4551 GRAND GULF APET EVENTS BY CATEGORY (CONTINUED) '
1 j Event 4 Number Description j AC/DC Power Availability AC Power Recovered During Core Damage ) 24 DC Power Available During Core Damage 25 79 AC Power Recovered Following Vessel Breach 80 DC Power Available Following Vessel Breach . A 104 AC Power Recovered Late
- 105 DC Power Available Late i
Criticality 4 29 Core in Critical Configuradon Following injection Recovery Hydrocen Related Phenomena /ISSIES i 31- Amount Oxygen in Wetwell During Core Damage 32 Amount Oxygen in Drywell During Core Damage 33 Amount Steam in Containment During Core Damage j 34 Amount Steam in Drywell During Core Damage g t 35 Amount Hydrogen Released in-vessel During Core . Damage ) 36 Level In-vessel Zirconium Oxidation 39 Max. Hydrogen Concentration in Wetwell Before Vessel Failure 40 Extent Wetwell Inert During Core Damage
- 41 Diffusion Flames Consume Hydrogen Before Vessel Brtoch I
42 Max. Hydrogen Concentration in Drywell Before VesselFailure ! 43 Deflagrations in Wetwell Before Vessel Breach 44 Detonation in Wetwell Before Vessel Breach ! 45 Containment Impulse Load Before Vessel Breach 46 Hydrogen Burn Efficiency Before Vessel Breach ' 47 Peak Hydrogen Burn Containment Pressure . 48 Extent of Drywell Leakage Due to Early Detonation in Containment 49 Extent of Containment Leakage Due to Early Detonation in Containment 56 Extent DrywellInert at Vessel Bresh 57 Sufficient Hydrogen in Drywell for l Cocibustion/ Detonation Before Vessel Breach 65 Detonation in Drywell at Vessel Breach 66 Deflagration in Drywell at Vessel Breach 68 Amount Hydrogen Released at Vessel Breach 19Eif>8 2
! Amendment 77 4
ABWR mems REV A Standard Plant i Table 19E.2 22 NUREG/CR 4551 GRAND GULF APET EVENTS BY CATEGORY (CONTINUED) Event l Number Description Hydrocen Related Phenomena / Issues (continueQ 69 dow Much Hydrogen Released at Vessel Breach 78 Hydrogen Concentration in Containment Immediately After Vessel Breach 82 Extent Wetwell inen After Vessel Breach 83 Sufficient Oxygen in Containment for Combustion 84 Hydrogen Ignition in Containment at Vessel Breach 85 Hydrogen Ignition in Containment Following Vessel Breach I 86 Hydrogen Detonation in Wetwell Following Vessel Brtsch 87 Impulse Loading to Containment Following Vessel
- Breach
; 88 Hydrogen Burn Efficiency Following Vessel Breach i 89 Peak Containment Pressure From Hydrogen Burn at i
Vessel Breach 91 Extent of Drywell Leakage Due to Detonation in i Containment at Vessel Breach 92 Extent of Containment Leakage Due to Detonation at
!% Vessel Breach 101 Hydrogen (and CO) Produced During CCI 102 Level Zirconium Oxidation in Pedestal Before CCI 107 Late Concentration Combustible Gases in
- Containment 108 Level Wetwell Inert After Vessel Breach 109 Sufficient Oxygen in Containment for Late Combustion
, 110 Hydrogen Ignition in Containment Late 111 Detonation in Wetwell Following Vessel Breach 112 Containmentimpulse Load Late 113 Hydrogen Bum Efficiency Late i14 Peak Containment Pressure From Late Hydrogen Bum i 115 Extent of Drywell Leakage Due to Detonation in Containment Lae 116 Extent of Containment Leakage Due to - Late Detonation 117 Level of Containment Leakage Due to Late Combustion ' i18 Level of Drywell Leakage Due in Late Combustion i 1 O Amendment ?? 19E.2 68 3
. __. _ 1-
U l 23A6100A5 Standard Plant REY.A l Table 19E.2 22_ .
-NUREG/CR 4551 GRAND GULF APET EVENTS BY CATEGORY i (CONTINUED) ,
i ! Event j: ' Number - Description t. Containment /Drywell Pressurintion/ Failure i j 37 Containment Pressure During Core Damage - t 38 Extent of Containment Leakage Due to Slow Pressunzation Before Vessel Breach L 50_ - LevelContainment I eakage Befe .I Breach { 51 Level of Drywell Leakage >
.o Containment Pressurization -
- '52 Level Pool Bypass Following Early_ Combustion j Events
- 55- Containment Pressure Before Vessel Breach l 70' Drywell/Wetwell Pressure Differential Resulting from Vessel Breach -
7I Peak Pedestal Pressure at Vessel Breach jL .72 Drywell impulse Load at Vessel Breach Sufficient to 1 Cause Failure [ j 73 Drywell Prr surization at Vessel Breach Sufficient to - Cause Fail.tre - 74 Pedestal'e'ailure Due to Pressurization at Vessel Breach .( 76 Pedests. Failure Causes Drywell Failure 4 77 Conts.nment Pressure at Vessel Breach Prior to } Hydesen Burn i 90 Leve1 Containment Pressurization At Vessel Breach 4 93 Leve i Containment Leakage Following Vessel Breach l- 94 Level of-Drywell Leakage Due to Containment j Pressurtzation _ j 95 - Level Pool Bypass Following Vessel Breach ) % Containment Pressure After. Vessel Breach ' } 122 levelLate Pool Bypass j 123 Late Containment Pressure Due to Non condensibles
- or Steam 124 Late Containment Failure Due to Non-condensibles or
!. Steam l 125 Long Term Level Contamment Leakage i Care Concrete Interactions /Pedestn1 Failure 54 Waterin Reactor Cavity
; 97 Water Supplied to Debris late
{ 98 Water in Cavity After Vessel Breach - t 99 Nature of Core Concrete Interactions (CCI) - 100 Fraction of Core Not Participating in - HPME Participates in CCI 120 Amount Concrete Erosion to Fail Pedestal t i 121 Time of Pedestal Failure i- ? 1 ^ Amendment 77 19E.2 68.4 i- ~.~,-~ , m -- _- , . - . - - . . _ _ , - . - - - - .. - .- - - - - . . - . - ~ - . . - . - - - - -- ~ - - - .
l ABWR uuimu Standard Plant REY.A i-Table 19E.2 22 , i NUREG/CR 4551 GRAND GULF APET EVENTS BY CATEGORY ) (CONTINUED) l i Event j Number . Description Steam Exelosion Related l 58 Alpha Mode Event Fails Vessel and Containment
- 60 Large In-vessel Steam Explosion i 62 In-vessel Steam Explosion Fails Vessel .
- 67 Large Ex-vessel Steam Explosion 75 Pedestal Failure From Ex vessel Steam Explosion Core Damace Proeression and Vessel Brench j 59 Fraction of Core Participating in Core Slump
) 61 Fraction Core Debris Mobile at Vessel Breach 63 Mode of Vessel Breach 64 High Pressure Melt Ejection l i 4 J 5 a i 1 i s I. 1 a l Amendment ?? 19E.2 68.5
ABWR 234.iooAs REY.A Standard Plant
/
( Table 19E.2 23 n
~
NRC IDENTIFIED PARAMETERS FOR SENSITIVITY STUDY FROM NUREG 1335 h e a Performance of containment heat removal systems during core meltdown accidents e in vessel phenomena (primary system at high pressure)
- H2 production and combustion in containtnent
- Induced failure of the reactor coolant system pressure boundary Core relocation characteristics
- Mode of reactor vessel mell-through
. In vessel phenomena (primary system at low pressure)
- H2 production and combustion in containment Core relocation chameteristics
- Fuel / coolant interactions
- Mode of reactor vessel mell-through
. Ex vessel phenomena (primary system at high pressure)
- Direct containment heating concerns
- Potential for early containment failure due to pressure load Long term disposition of core debris (coolable or not coolabic) a Ex vessel phenomena (primary system at low pressure)
- Potential for early containment failure due to direct contact by core debris
- Water availability in cases with long-term core-concrete interactions
- Coolable or not coolable 3
(3
-Q Amendment ?? 19E.2 68.6 -
ABWR meims REV.A Standard Plant r 4
/Q Table 19E.2 24 v.
IO ISSUES TO BE INVESTIGATED IN ABWR SENSITIVITY ANALYSIS h 1 1 a
.v Invessel Hydrogen generation Core Blockage and Melt Progression Fission Product release from core Cs! re evaporation Time of vessel failure Recriticality following in vessel recovery
! Ex venel Debris entrainmtnt and direct containment heating Mass of molten material at time of vessel failure Mode of vessel breach Potentialfor pedestalfailure Steam explosions Mass of molten material at time of vessel failure Presence of water in lower drywell at vessel failure Potential for pedestal failure Core concrete interaction and debris coolability Debris to water heat transfer Debris to crust heat transfer Mass of molten material at time of vessel failure Presence of water in lower drywell at vessel failure Potential for pedestalfailure Non-condensible gas generation Containment failure location a Contamment failure area Pool bypass High temperature failure of drywell Suppression Pool DFs a 'O Amendment 77 19E.2-68 7
ABWR uw REY.A s Standard Plant -g Table 19E.2 25 COMPARISON OF VOLATILE FISSION PRODUCT RELEASES Cal Release Fraction at 72 hours Cal Release Fraction at 72 hours Accident Sequence w/ . COPS w/o C6PS LCLP PF < 1E 7 4.87o LCLP FS < 1E 7 3.77o j LCHP-PF 8.87a
- 8.87o LBLC-PF <1E-7 0.37o LBLC-FS < 1E 7 0.67o NSCL PF < 1E 7 5.47.
I NSCL-FS < 1E 7 4.27o NSCH PF - 7.37e
- 7.37.
NSRC-PF <IE-7 3.7% NSRC-FS < 1E-7 14.5 7o Leakage through the moveable penetrations maintains containment pressure below the COPS setpoint. Amendment 77 19E.2 68.8
.- . . . _ .= _
ABWR uxeimis Standard Plant REY A Table 19E.2 26 COMPARISON OF LOW PRESSURE CORE MELT PERFORMANCE WITII AND WITiiOUT CONTAINMENT OVERPRESSURE PROTECTION SYSTEM w/ COPS w/o COPS Without Water Addition to Containment AP(Drywell Wetwell) . . 7 psig 7 psig
- Time of fission product release 20.2 hr 27.5 hr Csl release fraction @ 72 hours <l.E 7 4.8%
Probability of RHR recovery in time window N/A 11 %
; Probability of eventual DW head failure w/o CHR 2% 100 %
With Water Addition to Containment AP(Drywell Wetwell) 14 psig 14 psig Time of fission product release 31.1 hr 35.0 hr b <l.E 7 Csl release fraction @ 72 houn 3.7E 7 Probability of RHR recovery in time window N/A 4% Probability of eventual DW head failure w/o CHR 5% 100 % i i 1 i o Amendment 77 19E.2 68 9 i
i ABWR moimu ! Standard Plant REV. A - i
- Table 19E.2 27
$ PROBABILITY OF RELEASE MODE WITil AND WITIIOUT COPS + Class I/Ill Class II RD Opens DW Head RD Opens - DW Head Core l j Failure Failure Damage Base Case (with COPS) . 2.08E-8 5.25E 10 1.09E 7 1.10E-9 1.10E 12 l
- Without COPS 0.0 2.13E 8 0.0 1.10E 7 1.10E-10 t
i i i i i T 1 l i i f a i l I-i i i 6 l-d O i . . Amendment 77 - 19E.2 68.10
ABWR meimAs Standard Plant KEV. A Tal>le 19E.2 28 SENSITIVITY STUDIES FOR PASSIVE FLOODER RELIABILITY FREQUENCIES OF IMPORTANT CET RESULTS Failure rate of passise flooder on demand 0.001 0.01 0,1 . 1.0 Tine of CCI No CCI 6.73E 8 6.73E 8 6.73E.8 6.70E i, Wet CCI 7.11 E.9 7.11E 9 7.10E 9 7.07E 9 Dry CCI 3.45E h3 3.45E 12 3.45E 11 3.45E 10 Pedestal Condition No Ped Failure 7.41E 8 7.41E 8 7.40E 8 7.37E 8 Ped Failure 1.06E 10 1.06E.10 1.06E 10 1.06E-10 FP Release Mode COPS 7.58E 9 7.58 E.9 7.57E 9 7.51E 9 DW llead 3.91E 10 3.91E l') 3.91E-10 3.89E 10 Pen. Overtemperature 3.60E Il 3.91E 11 6.98E Il 3.77E 10 O Amendment ?? 19E.2 68.ll
O O O (f) g TYPICAL OF MO VALVES N a WHICH REQUIRE m _ INSIDE D LOCAL M ANUAL
' CON T AINMENT g d OPERATION 5
C1. b T TYPICAL OF 2 SAFETY GRADE SYSTEMS SAME AS ID bPik *
^c l
' H H H H H H H H H H'- /s n n n// ixK:n n n n n n ,u
- s
,n
,n s-
-; 7 e
s l- s -, - l l X n ( 2- u l t 5 i i i i fw"*lll " H n
" ""5'o Vn;'c*'a* ,,
l IIIIm l c is n l
\bV3^L lN2BOTTL.ES g IBACKUP) s l4 6 k-- A l
Ax I l l l TYPICAL b d y I g Or us t l lli i LWJ i M.'d'Afo s' ll- - 3 ' dT l ( ] T T_ '
;]l MAIN
~oN.SA,e1Y s' N ," 'UN
^"
, T T I R~ l TT, GR ADE N2 SYST EM (ATMOSPHERIC CONTROL) l l
V 88 433-73 $ e g Figure 19E.21 SIMPLIFIED SKETCH OF N2 SUPPLIES TO SAFETY GRADE ADS VALVES >d y'j _ i w a
7 ABWR 23A6100AS Standard Plant a,, 4 l I i l I i - 1400 One SRV Opened
~
8 i 1000 4 I g 6 - 3 2 800 g 4 8 9 600 h 4__ h e o 5 E 400 Rupture Disk Opens ] 2 -- 200
' ' ' 0 4
o I 0 10 20 30 40 50 60 70 - Time (hrs) i Figure 19E.2 2A LCLP.PF.R.N LOSS OF ALL CORE COOL!NG WITil VESSEL FAILURE AT LOW PRESSURE, PASSIVE FLOODER OPERATES AND RUPTURE DISK OPENS: VESSEL PRESSURE .; O 1,0 g l l l l 120 j Rupture Disk Opens 0.8 - l Water 1 Boiled - 80 ! B 0.6 - Away _ S 60 S ! !$ 0.4 - E h o - 40 e
; E k E 20 Passive Flooder Opens '"*"
0 Vessel Fails I I I I I I 0.0
! 0 10 20 30 40 50 60 70 Time (hrs)
- Figure 19E.2 2ft LCLP.PF-R.N: LOSS OF ALL CORE COOLING WITil VESSEL l[ FAILURE AT LOW PRESSURE, PASSIVE FLOODER OPERATES AND RUPTURE DISK OPENS: UO2 TEMPERATURE 1
Amendmem 19E.2-M - _. _ _- , ._ -, , .~, ..
- I ABWR m aaoxs
; . Standard Plant a., 4 1500 g g g ; g g 2000 s
, 1200 - l Vessel - 1600 h - C =,. E c 1200 ea e 900 - s I E E I y 600 Paniv. Flooder Opens Upper Drywell 800 f m e - m 4 8 hw "I^C{-_ - ~ _ _ - _ _ - _ - . _ _ _ . -
@0 y j Lower Drywell
- 300 -
0 i I I I I I I 0 0 10 20 30 40 50 60 70
- Time (hrs)-
! Figure 19E.2 2C LCLP PF R N: LOSS OF ALL CORE COOLING WITH VESSEL l_ FAILURE AT LOW PRESSURE, PASSIVE FLOODER OPERATES AND RUPTURE DISK OPENS: GASTEMPERATURE 1 ! l I I I l l j -
- ^ # #9' ****
2500 _ 4000 } 2 2000 - C 3000 L
? y 1500
-)$k - 2000 h
- Il #
J 8 1000 U
-) g i -
1000 D 500 Ql lt---m Lower Drywell Corium 0 I I I I I I 0 O 10 20 30 40 - 50 60 70
- . Time (hrs)-
Figt.re 19E.2 2D LCLP PF R N: LOSS OF ALL CORE COOLING WITil VESSEL O I b. l FAILURE AT LOW PRESSURE, PASSIVE FLOODER OPERATES AND RUPTURE DISK OPENS: UO2 TEMPERATURE ) Amendment 19E.2 70
. ,i,.,,__ . _ _ . . . , . . .. m , , .-
1 ABWR ux6imis Standard Plant , Rev A 5 g l l l g l 10 Rupture Disk Opens 4 O =- - 8-3 Suppression Pool
- f f 3 - e O
S -6 5 5 s 8
- $m 2 -
! - 4 t a> 2 i h* h I ~
Passive Flooder Opens -- 2 ] Lower Drywell 0 #dH~~ ~-d~~4 +~-~0 1 0 10 20 30 40 50 60 70
- l Time (hrs) _
Figure 19E.2 2E LCLP PF R N: LOSS OF ALL CORE COOLING WITil VESSEL a FAILURE AT LOW PRESSURE, PASSIVE FLOODER OPERATES 4 AND RUPTURE DISK OPENS: VESSEL PRESSURE io 000 g , g g g g } 2000 l 800 - e m l
?
1500 $
- $ 600 -
g m re ! 2 2 l
-1000 -
! O 400 - Vessel Failure O Hydrogen - 500 I I I I I I O O i 0 10 20 30 40 50 60 70 '2 Time (hrs) l Figure 19E.2 2F LCLP PF R N: LOSS OF ALL CORE COOLING WITH VESSEL J t FAILURE AT LOW PRESSURE, PASSIVE FLOODER OPERATES
; \ AND RU14URE DISK OPENS: MASS OF NON-CONDENSIBLES Amendment 19E.2 71 i
. . ~ , - . . - ~ , - ~
.-__._____--_._m.__._.__.. . _ . _ . _ _ _ . _ _ _ . _ .
1 ABWR - 2mi=^s
- Standard Plant ne, A 10 i i i i i
( ir i d i
- 0.8 --
E
$p 06 -
~
Noble Gases
*8
. E o S 0.4 - !' g2 un 4 em o i
- u. E 0.2 -
i l I I ' 0.0 i 0 10 20 30 40 50 60 70
- Time (hrs) 1 I Figure 19E.2 2G LCLP PF R N: I.OSS OF ALL CORE COOLING WITil VESSEL FAILURE AT LOW PRESSURE, PASSIVE FLOODER OPERATES t
AND RUPTURE DISK OPENS: NOBLE GASES
- O 1.2 g ; j l l l l f g , 1.0
! 2S i
Hs Q. E 0.8 - -
- c
!. .9 m 3 4 .50 m i "- E 0.6 - - 5 $O aE '~ 3e 4
~ ~ 0.4 og Em B
s'
'~ CsOH -
g$g
.2 i -
/ csi
~
. /
I I I I 0.0 O- 10 20 30 40 50 60 70 i Time (hrs)
! Figure 19E.2 2il LCLP PF R N: LOSS OF ALL CORE COOLING WITil VESSEL ; FAILURE AT LOW PRESSURE PASSIVE FLOODER OPERATES i .
AND RUPTURE DISK OPENS: VOLATILE FISSION PRODUCTS Amendment 19E.2 72 4
,- r,,-- a,--- ~,w,- s e n r gw
1 ABWR 23 .iooxs Standard Plant Rev A 1.0 g ; g ; g 120 3 g Rupture Disk Opens F
~ -
Lower Drywell 100 Dries Out l Firewater Spray - 80 f Starts i 3 0.6 - - E Lower Drywell - so $ g Reicods g OA - l 40
- E E C Upper Drywell
~
- =
0.2 - -
~R ~
i Firewater Spray Stops Vessel Falls 0 ' I I I I I I O.0 J 0 10 20 30 40 50 60 70 Time (hrs) l Figure 19E.2 3A LCLP FS R N: LOSS OF ALL CORE COOLING WITil VESSEL - FAILURE AT LOW PRESSURE, FIREWATER SPRAY OPERATES AND RUPTURE DISK OPENS: DRYWELL PRESSURE O V 600 - - - 1 I I I I i - m0 Passive Flooder Opens Firewater Spray Starts [ Rupture Disk Opens _ 500 - F
,7 pper Drywell
/ -
400 C
/ \N L
? e s g s 400 -
( '\ _ _ , / V' o r Drywl V 200 g 8 (/ n
- Firewater Spray Stops 0
3ng _ 0 I I I I I I 200 O 10 20 30 40 50 60 70 Time (hrs) Figure 19E.2 311 LCLP FS R N: LOSS OF ALLCORE COOLING WITil VESSEL -
-[
V FAILURE AT LOW PRESSURE, FIREWATER SPRAY OPERATES AND RUPTURE DISK OPENS: GAS TEMPERATURE Amendment Igg.2 73
ABWR 2mius Standard Plant w4 ] 10 ,Q ' l I I I I I
- 20 Firewater Spray Stops 7.5 -
Suppression Pool - 16 n k 8 h S - 12 8 4 5 5 - 5 ! k k- , 2 2
- g 8 g y 2.5 -
Firewater Spray Starts g Passrve Flooder Opens ! Suppression Pool Overflows - 4
; O
~ _ f "I'L_
I Lower Drywell
~
' -I -- "-
0 j 0 10 20 30 40 50 60 70 , Time (hrs) Figure 19E.2 3C LCLP.FS.R N: LOSS OF ALL CORE COOLING WITil VESSEL FAILURE AT LOW PRESSURE, FIREWATER SPRAY OPERATES j AND RUPTURE DISK OPENS: WATER MASS t . i\ 1.0 ; i i i i i 0.8 - - , if i e eC ' y j 0.6 - - Noole Gases l 4 zE o S 0,4 - -
$3 na 1 em 4
4 uCe 0.2 - - ) I I I I I I i 0.0 0- .10 20 30 40 50 60 70 Time (hrs) Figure 19E.2 3D LCLP FS.R.N: LOSS OF ALL CORE COOLING WITil VESSEL 2 iO FAILURE AT LOW PRESSURE, FIREWATER SPRAY OPERATES AND RUPTURE DISK OPENS: NOllLE G AS
. Amendmcra 19E.2 74 3
~. - .
...- .- .-. . . - . - . - - . .--- = . . _ - . - - _ . . - _.
i 4 ABWR ummis Standard Plant ne. 4 5 ;_ ; ; g g g
$C 4 - -
fS5 i C- E c
- .9 3 - -
2 i tr y
- Id 3
i E 2 - p 1 bE 1
/
! O3 c Csl /
.9 & j QS 1 - / -
2e (-
- CsOK /
I o l l l ~ + - ~" " ~l 1 -l-0 10 20 30 40 50 60 70 Time (hrs)- i Figure 19E.2 3E LCLP FS R N: LOSS OF ALL CORE COOLING WITH VESSEL ! FAILURE AT LOW PRESSURE, FIREWATER SPRAY OPERA 1 ES
- - AND RUPTURE DISK OPENS
- VOLATILE FISSION PRODUCTS
- 0 i
%.)
Amendment 19E.2 ~15
3
'ABWR uAsi=As Standard Plant - no A 10 I I I i i I -
1400 I Vessel Fails 1200 n 1000 W 6 - 3 ct E 800 g 4 4
. e-
$ 4 -
600 3 l
$ E g Drywell Spray Starts E 400
! 2 - Rupture Disk Opens 200
; , , Vessel ,
g
, 0 10 20 30 40 - 50 60 70 Time (hrs)
Figure 19E.2 4A LCllP PS.R N: LOSS OF ALL CORE COOLING WITil VESSEL i FAILURE AT IIIGil PRESSURE, PASSIVE FLOODER AND DRY. WEi.L SPR AYS OPERATE, RUI'TURE DISK OPENS: VESSEL PRE.iSURE 1,0
- g g. g l ; g 120 j Rupture Disk Opens 0.8 -
100 f '- 80 g - 0.6 - 3 Drywell Spray Starts 60 i 5 5 i g 0.4 - 40 $ E E 20 0.2 - f Upper Drywell Vessel Fails 0.0 I I I I I' I O 10 20 30 40 50 60 70 ] Time (hrs) Figure 19E.2-4B - LCIIP PS-R N: LOSS OF ALL CORE COOLING WITil VESSEL - 'O=
, FAILURE AT llIGil PRESSURE, PASSIVE FLOODER AND DRY.
WELL SPRAYS OPERATE, RUPTURE DISK OPENS: DRYWELL PRESSURE Amendmem 19E.2-76
ABWR ummis Standard Plant _ _ _ nev 4 O ** i i > ' ' ' 2500 - Average in Vessel - 4000 F
$ 2000 --
_ 3000 E E 1500 -- y,,,,; p,;35 - 2000 g Drywell Sprmy Stops -
- Drywell Spray Starts H f1009 l Upper DrywellCorium
- - - e" ~ -
E 1000 D (~~~~ g y ,. 500 '- Wetwell Corium 0 0 I I- I I f i 0 10 20 30 40 50 60 70 Time (hrs) Figure 19E.2 4C LCllP.PS R.N: 1.OSS OF ALL CORE COOLING WITil YESSEL FAILURE AT lilGil PRESSURE. PASSIVE FLOODER AND DRY. WELL SPRAYS OPERATE RUPTURE DISK OPENS: 003 TEht. PERATURE O 1500 - g ; g i ; ; 2000 1250 - 1600 Vessel N g g 1000 - e - 1200 8 Passive Flooder Opens s 750 Upper Drywell _ - he
-~~
D ll Spray Starts - 800 g g pT rywe -
~
C .w
,,,/ 3 - , q Lower-.-.-
Drywell 400 f 250 - Drywell Si ray Stops _ o Rupture Disk Opens I I I I I I 0 0 10 20 30 40 50 60 70
- Time (hrs)
Figure 19E.2 4D LCilP.PS.R.N: LOSS OF ALL CORE COOLING WITil VESSEL O FAILURE AT lilGil PRESSURE, PASSIVE FLOODER AND DRY. WELL SPRAYS OPERATE, RUPTURE DISK OPENS: G AS TEhl. PERATURE Amendmera . 19P 2 TI m___________.__________.______._________._.-___-____._.____________.____.__m
l
! ABWR uu:ms l Standsird Phmt Rev A 1
1 20 g j l i l l l
- 40 lj - 4 i
i - 15 1 - 30 - 2 3 ! b b l ; % l 10 - i es M j 2 Upper Drywell 2
- g . _ . _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _
g i 3 . D 4 5
-} < " ,,,,. _ _. _ ._ ._ _ _ Vessel -.---._------
__ 10 4 Lower Drywell i / 1, l/ _ , _ _ W e_twell _ ) g I I I I l I 0 , 0 10 20 30 40 50 60 70 Time (hrs) Figure 19E.2 4E LCilP.PS.R.N: LOSS OF ALL CORE COOLING WITil VESSEL, 1 FAILURE AT lilGil PRESSURE. PASSIVE FLOODER AND DRY. WEl.L SPRAYS OPERATE, RUI'TURE DISK OPENS: 002 MASS 6 j I i i 1 I I i - 10 i - Rupture Disk Opent
- 4 -
3 8 h 3
. Suppression Pool 4
3 -
- 2 6 0 i
a
- 5 k
1. 2 -- s
~ #
) 5 Passive Flooder Opens .5 I f - Drywell Spray Starts h 1 _. 2
/~ P~~--
- l. Lower Drywell
; [ ' y -- -- 4. _ -- t - - t - - - ' o
- O 10 20 30 40 50 60 70
- . Time (h
- s)
Figure 19E.2 4F LCHP.PS.R.N: LOSS OF ALL CORE COOLING WITil VESSEL I FAILURE AT II1Gil PRESSURE. PASSIVE FLOODER AND DRY. t WEll, SPRAYS OPERATE, RUPTURE DISK OPENS: WATER MASS
*enendment IE 2M 2
- - - ~ --,,na . ,, , , , - . ~ , s , - ., - - - - , , . - , ,, -n .-- . , . , - , - , , - - - - , . , , , , , -,,,-,,-,v,- -
ABWR Standard Plant nA6tm^s RevA I I I I I I 800 - k - 1500 jr 600 - 2 5 400 - 1000 200 -
$00
, Hydrogen T
. i o l i l l l- 1 I l 0 10 20
.s 0
30 40 50' 60 70 Time (hrs) Figure 19E.2 4G LCHP.PS.R.N: LOSS OF ALL CORE COOLING WITH VFSSEL - j FAILURE AT HIGH PRESSURE, PASSIVE FLOODER AND DR% j WELL SPRAYS OPERATE, RUITURE DISK OPENSt GLOBAL MASS iO ,.0 i l I 0.8 - I j 0.6 - Noble Gases - 1 i E
]S 0 .4 -
n} f bC 0.2 - i
- 0.0 I- 1 I P
I I I O 10 20 30 40 50 60 70 l Time (hrs) Figure 19E.2 4H LCif P.PS.R.N: LOSS OF ALL CORE COOLING WITH VESSEL ! FAILURE AT HIGH PRESSURE, PASSIVE FLOODER AND DR%
- WELL SPRAYS OPERATE, RUITURE DISK OPENS
- NOBLE GASES Amendmera 19E.2 79 0
.-,..._._....--.._m,m_...,__-,._m. . , _ , - ;..., . - . . , , . ,_ ...,____.. _..,_, ,..-:_ .a_ ,.a... _,..._x .em ;,0 --,4 ;, - . . , , ..-.wme-._,
_ _ _ _ . . _ _ _ . _ . . ~ _ . . _ . _ _ _ ~ . _ _ _ _ . _ _ _ _ _ _ _ ! ABWR mmo4s
- Standard Plant __
n,v 4 ; i l 5 g ; ; ; j g 1 N' 4 - - 1 j6 i 5E 1 & 1 e 3 - - l a' 2 ~ ~ CsOH
, .h. , - - - . - - - -
b l
- e .g , -
i 1
$z ?
Csl
> /'
0 I I ' ' i 0 10 20 30 40 50 60 70 1 l Time (hrs) ) Figure 19E.2 4I l.CHP.PS M.N: LOSS OF ALL CORE COOLING WITH VESSEL FAILURE AT HIGH PRESSURE. PASSIVE FLOODER AND DRY. { WELL SPRAYS OPERATE. RUI'IURE DISK OPENS: VOLATILES I i j 1 1 l 1 i i I l. } I f-1 4 d 4 Amendment 19E.2 80 1-
. , - - e, - ~ . ~ - - , , , , . . . ,.,-a,, ,. . - - , . -- , , . . , , ., . , . , - . ~ , - , - , - .,~w,.. . - - , _ , - - - - . - - ...,-+.+-,----.--,.w, w,w.-=.-,
1 ABWR zwim^s Standard Plant Rev A 1.0 g i i i 120 0.8 - 100 Upper Drmell 80 0.6 - { .
- 2. ~
e h 0.4 - 40 Penetraton Leakage Degins
- 20 0.2 --
Vessel Fails
, , 9 O 20 40 60 80 100 Time (hrs)
Figure 19E.2 5A LCHP.PF.P.M: LOSS OF ALL CORE COOLING WITH VESSEL FAILURE AT HIGH PRESSURE. PASSIVE FLOODER OPERATES, PENETRATION LEAKAGE: DRVWELL PRESSURE 1500 j l l l 2000 1250 - 1600 C E 1000 -
?-
g Upper Drywell 1200
<a 750 -
- 800 I'
@ N 5"
_._ _ _ M**dWL _ _ _ z
$ "O 1
250 - - 0 Vessel Fails I I I i 0 0 20 40 _ 60 80 100. Time (hrs) Figure 19E.2 5B LCHP PF.P.M: LOSS OF Al.L CORE COOLING WITH VESSEL t FAILURE AT lilGH PRESSURE, PASSIVE FLOODER OPERATES,
\ PENETRATION LEAKAGE: GAS TEMPERATURE -
Amendmera 19112-81
ABWR umas Standard Plant a, 4 3000 g g l g 2500 -
._ Average in Vessel -
4000 I { 2000
- 3000 1500 2000 H H d' 1000 a Upper Drywell Corium g
3 _ _ ( 3 500 ,,,,,,,, _ ,,,, _ _ _ _ _ _ 10*'f2 D ' 9 3 5 _ _ _ _ _ _ _ .
~
Vessel FalIs , j ; 0 20 40 60 80 100 Time (hrs) Figure 19E.2 5C LCHP.PF.P.M LOSS OF ALL CORE COOLING WITH Vl3SEL FAILURE AT HIGH PRESSURE, PASSIVE FLOODER OPERATES, PENETRATION LEAKAGE: 00 2TEMPERATURE 5 g g g i 10 4 - k -.
~
~
gppression Pool
? 5 g 3 -
g
- 6 5 5" h
2 2 - 2 g 4 g h Passive Flooder Opens 1 - 2 Lower Drywell j--- ,e- - - - - - - - - 0 ' 'I I I 0 0 10 20 - J 40 50 60 70 Time (hrs) Figure 19E.2 5D LCHP.PF.P.Mi LOSS OF ALL CORE COOLING WITH VESSEL
' FAILURE AT HIGH PRESSURE, PASSIVE FLOODER OPERATES,
( PENETRATION 1.EAKAGE: WATER MASS Amendment 19E.2-82
ABWR zmms Standard Plant ._ nn 4 1.0 g g ; l
't 0.0 -
v I 0.6 - E
- ~
Noble Gases 15 } OA
.5
$ Penetraton Leakage Begins CC 0.2 - _
CsOH 0.0-- - t _M- l 0 20 40 60 80 100 Time (hrs) Figure 19E.2< 5E LCitP PF P.M LOSS OF ALL CORE COOLING WITil VESSEL FAILURE AT llIGli PRESSURE, PASSIVE FLOODER OPERATES. PENETRATION LEAK AGE: FISSION PRODUCT RELEASE O O-Amedment 19E.2 83 i
l . ABWR Standard Plant "^T*5, a
I I i i i j q -
120 j
** - Rupture Disk opens ,
1 i 0.8 - _ 1 - 80 ) g 0.c - 3 60 $ l . b 0.4 - N U
- 40 U i
E 20
- Uppet Drywell 0.2 RCIC Fails 0
1 ' I I I l I I O.0
- o 10 20 30 40 50 60 70 Time (hrs)
Figure 19E.24A SBRC FA R 0: STATION BLACKOUT.RCIC RUNS EIGilT ] IlOURS. FIREWATER ADDITION PREVENTS CORE DAMAGE, ! RUITORE DISK OPENS: DRYWELL PRESSURE
- O
- 600 , , ,
1 600 i } E 500 e e Rupture Disk Opens ;
@0
- j )a
. e 1 e g !' e t
$ 400 - Suppression Pool 5 h
200 i I l 1 1 I 3oo o 10 20 30 40 60 60 70 Time (hrs) 3 Figure 19E.24B SBRC FA R 0: STATION BLACKOUT,RCIC RUNS EIGHT l HOURS. FIREWATER ADDITION PREVENTS CORE DAMAGE, RUITURE DISK OPENS: WATER TEMPERATURE J Amedmont 19E2 84
, .-- . . , . . - - . , , - ,, , . , . . . - . , - . . . - - .-,n - r -. -- ,n . . , - , . ,
l
- I
)
ABWR unim4s Standard Plant n,, 4 i j 1 50 g g ; j ; ; 1
- Firewater Additon Starts 1500
, 1000 - 12$0
- g i
E - 1000 r., i
- 750 - e
~
750 ,
~
p 500 - j g _
%. Maximum g
- D -o 250 - L neiC F,ii, -
0 i f n i I I I I I i 0 10 20 30 40 50 60 70 - {- Time (hrs) Figure 19E.24C SBRC FA R 0: STATION BLACKOUT,RCIC RUNS EIGHT j HOURS, FIREWATER ADDITION PREVENTS CORE DAMAGE, RUFTURE DISK OPENS: U02 TEMPERATURE
- O 3
20 ; ; g ; g ; 80 RCIC Fails ! Shroud j E - @ g 10 - Top of Active Fuel 6e
*C
! $ { Bottom of Adive Fuel 20 t - 5 i Firewster Addnion Starts 1 I I I I I I 0 0 0 10 20 30 40 50 60 70 Time (hrs) ( i Figure 19E 24D SBRC FA R 0: STATION BLACKOUT,RCIC RUNS EIGHT HOURS FIREWATER ADDITION PREVENTS CORE DAMAGE, Amh 19E.2 85 l
l ABWR nwwxs
! Standard Plant n.. A i
5 l l l l
- l l 4
- 10 I 4 - .
Suppression Pool j - 8 1 i
? 3 _
nupture oisk opens ) 6 l E 5 i ! $ 0 i i 2 2 - 2 '
~
d 4 g 3 g
- SRV Opens j l
i - - 2 Lower Drywell I I I I I 0 0 0 10 20 30 40 50 60 70 -- Time (hrs) Figure 19E.2 6E SBRC FA R 0: STATION BLACKOUT,RCIC RUNS EIGHT i HOURS, FIREWATER ADDITION PREVENTS CORE DAMAGE, j RUPTURE DISK OPENS: WATER MASS iO d 4 l 1 i i l 1 4 i 4 !O , Amedmers 19E.2 46 i
- - _ . - , , , ...m..re , - . , . _ - _ . , , _.-- -. ..--- , -.w ., .- - - - , - , - - , - - - ,
4 i ABWR Standard Plant **'[j,' I I I I I l - 1400 j One SRV Opened
- 1200 1000
[ 6 - - j g - 8x { l e 7 4 5 l 4 _ 600 l j 5 400 1 Rupture Disk Opens . 2 - 200
' ' Vessel 0 l l i '
0
- O 10 20 30 40 50 60 70 3 Time (hrs)
Figure 19E.2 7A SHRC PF.R.N: STATION HLACKOUT M1TH RCIC OPERA 11NG, PASSIVE FLOODER OPERATES AND RUITURE DISK OPENS: - s VESSEL PRESSURE i O 1.0 ; ; ; j j j i - 120 l Passive Flooder Opens Rupture Disk Opens ! 0.8 - 100 80 g 3.6 - E 60 u
- y 2 a 0.4 -
j e a 20 O.2 - pper Drywell Vessol Fa'is d - 0 0.0 I ! I l l l 0 10 20 30 40 50 60 70 Time (hrs) Figure 19E.2 78
- O.
SBRC PF R N: STATION HLACKOUT WITH RCIC OPERATING, PASSIVE FLOODER OPERATES AND RUITURE DISK OPENS:
'() DRYWELL PRESSURE
? ',' ^"*"*""* 19s.2.s7
~v- ne-e, - . . - - . - -- .-
-.,---n-.
ABWR nuimis Standard Plant nev a 1500 g g g i i l 2000 y,35,l 1600 g C c g , 000 1200 _g E Passive Flooder Opens E
~
800 h 4 600 _3 _ Upper DyweH
/t --~ ~',', ~ _ _ -
0 " 400 c: . 7 f ---- - -- Lowor Drywon
%0 _ -
0 l 0 I I I I ' 0 10 20 30 40 50 60 70 Time (hrs) Figure 19E.2 7C SHRC PF R N: STATION IILACKOUT WITH RCIC OPERATING. PASSIVE FLOODER OPERATES AND RUI'rURE DISK OPENS: G AS TEMPERATURE 3000 g ; ; j i l Average in Vessel 2000 3000 Passive Flooder Opens j
" m 1500 --
D RCIC Fails - 2000 g
& A S 1000 -
l 8 3 j\ - 1000 3 k 500
- - --. r - )
- Lower DrywellCorium 0
l I I I I I
.n O 10 20 30 40 50 60 70 Time (hrs)
Figure 19E.2 7D SBRC PF R.N: STATION BLACKOUT WITH RCIC OPERATING, O PASSIVE FLOODER OPER ATES AND RUI"TURE DISK OPENS: UO2 TEMPERATURE henhers 19E.2 BN
a ! ABWR meima i Standard Plant n,, 4 !o - , , , , , , 10 ) 4 -
- e -
i Suppression Pool 8 _ S b [ 3 - g ) 6 g ! e i l@ j 2 2 - g Vessel Fails - 4 g 5 en y Passive Flooder Opens 1 - 2 } Lower Drywell i 0 PI f' ~ I- ~ ~ l- ~ ~-I ~ ~ ~+ ~ ~ ~ 0 1 0 10 20 30 40 50 60 70 Time (hrs) ] d i Figure 19E.2 7E SBRC PF R N: STATION BLACKOUT WITH RCIC OPERATING,
- PASSIVE FLOODER OPERATES AND RUI'fURE DISK OPENS
1 WATER MASS lO i 1.2 g g g i - - m e 1.0 ! D6 jE
- n. 0.8 - -
a 1 e
.9
$m C E 0.6 -
i 25 CsOH j 3e i 9e gE c 0.4 -
/
/~- -
$b l 4
E y 0.2 -
/. -
Cs, 4 I I I I I I 0.0 0 10 20 30 40 50 60 70 i- -Time thrs) Figure 19E.2 7F' SBRC PF R N: STATION BLACKOUT WITH RCIC OPERATING, 4 O PASSIVE FLOODER OPERATES AND RUlrrURE DISK OPENS: VOLATILE FISSION PRODUCT RELEASE -
- Amendment 19E.2 89 1
- . _ . _ , _ . _ _ . . . _ _ . _ ~ , _ . , , . . . _ . . _ . _ . . - _ , . _ _ . . _ _ _ . . _ . _ . - . - . . _ , - . , . .
' ABWR nui s Standard Plant n,, 4 1.0
; ; ; ; ; j
- SRV Opens ~
Rupture Disk Opens ] 0.0 - _ joo i j {
/s -
80 0.6 -
/
E f
^
3 .
/ -
60
\
l 8
/
1 c l 0.4 -
/ \ _
40 g X 1 / \ e
\ vesset j
/ w--
% po 0.2 -
/ N j
9 - r '~ Upper Drywell - i 0 i 0.0 I I I I I I
, 0 10 20 30 40 50 60 70
} Time (hrs) 1 Figure 19E 2 8A i LilR C 00 R 0: ISOLA 110N M1Til LOSS OF CONTAINMENT IIEAT REMOYAL AND RUITURE DISK OPENS: DRYWELL L I'RESSURE 4O 4 600 l 1 l l I I i _ go ] = *- SRV Opens e 4 T
- 500 -
C g Rupture Disk Opens I 400 g 1 a b y yn N Lower Plenum l. g Tsi 3 7 m,,,,,,, ca
/ -
Suppression Pool 3 ( - 200
/
/
I I I 300 I I ' 0 10 20 30 40 50 60 - 70 Time (hrs) Figure 19E.2 8B LilRC-00 R 0: ISOLATION WITil LOSS OF CONTAINMENT O IIEAT REMOVAL AND RUITURE DISK OPENS; WATER TEM. PERATURE Amendment 19E.2-90 ,4-- O
,-,._..--,....--...,.-,,.-,,..,.,-r... - - , . . . . ~ , . - .,.~_-...,,,-,,..--.-.-_,__m. ...~.rr-,_,..,n-,,,
ABWR muers Standard Plant an 4 5 g g ; g g g 10 SRV Opens Rupture Disk Opens 8 3 { 3 -
~
g E Supptession Pool E h - 2 3 2 ~ B 2* 4 - _ ,. I I I I I I O 0 0 10 20 30 40 50 60 70 Time (hrs) Figure 19E.2 8C LilRC 00 R 0: ISOLATION WITH LOSS OF CONTAINMENT llEAT REMOVAL AND RUITURE DISK OPENS: WATER MASS O 4 i O Amendment 19E2-93
1 ABWR ! Stenderd Plant m oicoss 3 Rev A i 1.0 Rupture Disk Opens 0 0.8 -
- 100 l
g 0.6 - 80 E p e - 60 b
~ e
\
- n. -
40 c. 0.2 - 9,, g, , Upper Drywell 0 0.0 I I I I I l 0 10 20 30 40 50 60 70 Time (hrs) Figure 19E 2 9A LHLC PF.R N: LARGE HREAK LOCA WITH LOSS OF ALL CORE COOLING, PASSIVE FLOODER OPERATES AND RUP. TURE DISK OPENS: DRYWELL PRESSURE 1500 y ; g , , 1250 - 2000 1600 g 1000 - y,33,; j - { 1200 750 - E h Passive Flooder Opens - 800 N / h
$ 500 - - - Upper-'Drywell $
o wk/ , - - - - - ~~ ~~ ~~ ~
---~~ 400 N
-- ---- __ o Lower Drywell 250 -
0 0 ! ! I I I I O 10 20 30 40 50 60 70 Time (hrs) Figure 19E.2 9B t LBLC.PF R N: LARGE HREAK LOCA WITil LOSS OF ALL CORE COOLING, PASSIVE FLOODER OPERATES AND RUP. TURE DISK OPENS: GAS TEMPERATURE Amendment 19E.2 92 l e
, ABWR U W WAS Standard Plant an A I
b 1 I I i 1 l 10
- /- Rupture Disk Opens .
4 - 8 1
- E 2 Suppression Pool 8 g- 3 -
g* a 6 5 5 i h 1 2 - 2 4 a g
- h Passive Flooder Opens
! 3 2 : J j Lower Drywell i 0'-P ' " ~ I~ ~ ~ ~ I ~ ~ ~ ~I ~ ~ ~ N - ~ ~ 4~ ~ ~ ~ 0 0 10 20 30 40 50 60 70
- Time (hrs) i Hgure 19E.2 9C L!!LC.PF.R.N
- LARGE BREAK LOCA WITil LOSS OF ALL l CORE COOLING, PASSIVE FLOODER OPER ATES AND RUP.
1 TURE DISK OPENS: WATER MASS iO 1, U~ l i I I I I d u,c 1s U e6 i 5
- j. E 0.3 - -
4 e i k E nu ... e:. i 38 Eg
; s,s'.
99 CsOH/
/
,' ~ n o, .- - o / l c 9, y' y ').2 - / - I QI i 0.0 I ' I I A' I O 10 20 30 40 50 60 70
- Time (hrs) p Figure 19E.2 9D LBLC.PF.R.N
- LARGE BREAK LOCA WITH LOSS OF ALL' CORE COOLING, PASSIVE FLOODER OPERATES AND RUP.
1 Q TURE DISK OPENS: VOLATILE FISSION PRODUCT RELEASE i Amendmera 19E.2-93
ABWR ux6imis Standard Plant Re A 10 i i l i i i i 120 Rupture Disk Opens 0.8 - - 100 80 g 0.6 --
^
2., - 60 - e 2 0.4 - 40 d[ '
- 20 0.2 --
Upper Drywell Passive Flooder Opens -
/ _
I I I I ' 0.0 -- O 10 20 30 40 50 60 70 Time (hrs) Figure 19E.210A NSCL PF R.N: CONCURRENT LOSS OF ALL CORE COOLING AND ATWS WITH YESSEl. FAllURE AT LOW PRESSURE, PAS. SIVE FLOODER AND RUITURE DISK: DRYWELL PRESSURE 3000 g ; ; ; ; i 2500 - '89* " ****' -- 4000 Passive Flooder Opens _
$ 2000 e
- 3000 b l o E
%" E m
1500 - 2000 llI A 8 1000 8 3 ll- - 1000 3 500 d i Lower Drywell Corium
\---_
0 I ' ' 0 O 10 20 30 40 50 60 70 Time (hrs) Figure 19E.210B NSCL PF R N: CONCURRENT LOSS OF ALL CORE COOLING O AND ATWS WITH VESSEL FAILURE AT LOW PRESSURE, PAS. SIVE FLOODER AND RUITURE DISK: UO2 TEMPERATURE Amendmm 19E.2 94
I' ABWR umms , Standard Plant n,, 4 .
~~
j i l I I I l l 10 i g Rapture Disk Opens -1 4 / S_ -- - 8
~
i -
.F Suppression Pool 3 :
3 - 1 % -
- a 4
5 5 i k k 2 2 - 2 j g 4 g E E a 1 Passive Flooder Opens - 2 Lower Drywell r 0 ' ' '~~'4 ' ~ H ~ ~ '~'I ~ ~ ~h - - "'t-~ ~ ~ +-~~- 0 0 10 20 30 40 50 60 70 Time (hrs) l Figure 19E.210C NSCL PF.R.N: CONCURRENT LOSS OF ALL CORE COOLING AND ATWS %1Til VESSEL FAILURE AT LOW PRESSURE, PAS. I SIVE FLOODER AND RUi40RE DISK: WATER MASS 1.2 ; ; ; ; ; ; e 1 y mg 1.0 - - i t1 6 1a
- c. E 0.8 .- -
! .53 4 g.g 0.6 - -
~~~~~
CsOH, , , 4 fl0.4 - - 3b c yEg
's'
! 0.2 - / Col -
/
/ '
0.0 I d- ' I I O 10 20 30 40 50 60 ?O Time (hrs) n V Figure 19E.210D NSCL-PF.R.N: CONCURRENT LOSS OF ALL CORE COOLING AND ATWS WITil VUSEL FAILURE AT LOW PRESSUkE, PAS. SIVE FLOODER AND RUITURE DISK: VOLATILE I1SSION PRODUCTS Amendmed 19E2 95
1, ABWR mamu Standard Plant new A O '-
. i ' i 120 3
1 0.8 - - jon Upper Drywell 80 g 0.6 -
- E -
m ; 0.4 - 40 . & Penetraton Leakage Begins n. t j - 20
- 0.2 --
1 Vesnel Fails I I I I ] 0.0 0 20 40 60 80 100 ! Time (hrg) Figure 19E.211 A NSCll.PF.P.Mt CONCURRENT LOSS OF ALL CORE COOLING l AND ATWS WITH VFSSEL FAILURE AT HIGil PRESSURE, PAS. SIVE FLOODER, PENETRATION LEAKAGE: DRYWELL PRES. SURE 10M ; j j l l 1200 Upper Drywell i 800 - ., g - 800 p 5 g 600 - 5 g 16 Lower Drywell - 400 _- - -~~~~----- , b b 0 200 - Vessel Fails I I I I 0 r-O 20 40 60 80 100 Time (hrs) Figure 19E.211B NSCII.PF P.M: CONCURRENT LOSS OF ALL CORE COOLING O AND ATWS WITil VFSSEL FAILURE AT lilGli PRESSURE, PAS. SIVE FLOODER, PENETRATION LEAKAGE: G AS TEMPER A. TURE Amendmern 19E.2 %
__ _._._ . .. . _ _ _ . _ ~ . . _ . _ . . _ . _ _ _ . _ _ _ _ . _ _ _ . _ _ _ _ . _ _ . . - - _ . - ABWR 23 .. Standard Plant u,o
' 20 i ; ;
i
- 40 , i 15 30 n g
' b b *
% 9 10 ~ 5 0 - go e
$ Upper Drywell f
i e U 5 '- In Wstel 10 r ~ ~ ~~ ~ ~ ~ ~~ ~~ ~ Loweibrywell lI I ~ ~ ~ ~ ~; ,
~ ~ Wi'wEl ,
, 9 0 20 40 60 80 100 Time (hrs) .
Figure 19E.2. llc NSCif.PF.P4f CONCURRENT LOSS OF ALL CORE COOLING AND ATWS WITH VESSEL FAILURE AT HIGH PRESSURE, PAS. SIVE FLOODER.PENETRA110N LEAKAGE: UO2MASS 1.0 ; ; , , i j 0.8 - - a t! a .m 0.6 - -
.6 88 g - 0.4 .- Noble Gases -
ch
.9 m Penetration Leakage Begins E 0.2 - _
s ._ _. CsOH
~
g I S"#f ; ; Cal 0 20 40 60 00 100 Time (hrs) Figure 19E.2.llD NSCH.PF.P.M: CONCURRENT LOSS OF ALL CORE COOLING O AND ATWS WITH VESSEL FAILURE AT HIGH PRESSURE, PAS. SIVE FLOODER PENETRATION LEAKAGE: FISSION PROD. UCTS-Amaidmers 19E.2.g7 -
. - , , - . , - - - . - . , -, . . . - -, . - ~ - . . - . - _ . - . - . - . - - . - - . - - - . - . . , - , , . - - - , . . .
ABWR UWWAS Standard Plant u,, 4 10 l I I l i I -- 1400 One SRV Opened 1200 'l O - 1000 y 6 - 2 800 g
? e 5
f4 - So f E 400 Rupture Disk Opens 200 ves l ' I ' 0 O O 10 20 30 40 50 60 70 Time (hrs) Figure 19E.212A NSRC.PF.R.N CONCURRENT STATION HLACKOUT WITil ATWS, PASSIVE FLOODER OPERATES AND RUI'I'URE DISK OPENS: VESSEL PRESSURE . O
.0 ; ; ; , , ;
Rupture Disk Opens 120 0.8 - 80 , p 0.6 - 9
=
m A 0.4 - Passive Flooder Opens N 40 $ o- E
- 0 0.2 -
~ Upper Drywell Vessel Fails 0
I I I I I l 0.0 0 to 20 30 40 - 50 60 70 Time (hrs) O Figure 19E.1128 NSRC.PF.R N CONCURRENT STATION BLACKOUT WITil ATWS, PASSIVE FLOODER OPERATES AND RUI'TURE DISK OPENS: DRYWELL PRESSURE Amendmem 19E 2 4
i i ABWR m .i m s Standard Plant n,, 4 10 l l l l ) ; i. 8 i 3 i l g 6 - -
! Z d
j 4 -
\ \
) - j 2 - u TotalCore j 0 I I I I I I 0 10 20 30 40 50 60 70 - Time (hrs) Figure 19E.212C NSRC.PF.R.N: CONCURRENT STAT 10N BLACKOUT WITH ATWS, PASSIVE FLOODER OPERATES AND RUPTURE DISK OPENS: POWER !O I I I I i 1 Average in Vessel 52 2003 - 3000 @ 4 y Passive Flooder Opens
. 1500 -
Dries r \ /ji e ! 8 1000 - SRV Opens i g 3 .1000 3 l -
~
Lower Drywell Corium 500 y -.. - - ~ . J i,..__
~
Vessel Faie I I I i i 0 0 10 20 30 40 50 60 70 Time (hrs) Figure 19E.212D NSRC.PF.R.N: CONCURRENT STATION BLACKOUT W1TH ATWS. PASSIVE FLOODER OPERATES AND RUMURE DISK N OPENS: UO2 TEMPERATURE j Amendmera 19E.2 99
. . . _ _ _ _ .-_ _ _ . _ . _ . . _ . _ _ . _ . . _ _ _ _ _ _ _ _ _ _ _ ~ _ _ _ _ . _ . - _ . _ _
s d ABWR :wim4s
- Standard Plant -
ne, 4 l I I I I I I 10 4 - 8 Suppression Pool ,, 3 A h N 6
$ h l
3 2 - 2 3 g - Suppression Pool Cveritows - 4 g ) j Passive flooder Opens y
- 4 -
_ 2 1 % I j O q/ I N - N ' ' ' ' l '- ~ ~ ~ +Lower - ~yw+' Dr ell
~~~ 0 j 0 to 20 30 40 so 60 70 j Time (hrs)
[ Figure 19E.212E NSRC.PF.R.N CONCURRENT STATION HLACKOUT WITil ATWS, PASSIVE FLOODER OPERATES AND RUI'IURE DISK i OPENS: WATER MASS i j 5 g g g g g g l
- i eo 4 - -
E E
.9 h 3 - -
i
${ Passive Flooder Opens ;
~ ~~~~~
h /,,- i EE 2 - i SQ j' D
.9 M Rupture Disk Opens f :
! U{W 1 f
J y "; ; ; ; Cal ; i 0 .10 . 20. 30 40 50 60 70 Time (hrs) Figure 19E.212F NSRC.PF.R.N: CONCURRENT STATION BLACKOUT M1TH
. ATWS PASSIVE FLOODER OPERATES AND RUI'TURE DISK OPENS: VOLATILE FISSION PRODUCT RELEASE 1
Amadment 19E.21% i-S
, _ ,.... _, _ , ~, -.
- ABWR urumas amA
- Standard Plant i
j
; A. MAIN STEAM 4
NO LEAKAGE (P2) NO STREAMLINE TURBINE BYPASS OR FAILURE TO ISOLATE (P1) BREAK (P5) ISOLATION (P3) OK 1 1 OK l BYPASS i , i ' - BYPASS l i i RPV X X X D. 1 Figure 19E.219A - SUPPRESSION POOL BYPASS PATHS AND CONFIGURATIONS f
- O B. FEEDWATER OR SLC I NO CHECKVALVE NO LINEBREAK l FAILURE (P9) (P13, P15)
OK 1 OK i i I BYPASS I
!, n TUR81HE BUILDING (FW) l RPV N, N, // REACTOR BUILDING (SLC)
, l-i Figure 19E.219B SUPPRESSION POOL BYPASS PATHS AND CONFIGURATIONS 19E.2107 Aitadown: s n e
, _ . - _ - - . . , . , ,,,.c---- -
,.y_, ,,_,-y
l 1 ABWR naumas am ^ i sanslanLElant !O I
- C. ECCS LINES NO CHECKVALVE i FAILURE (P9) OPERATOR NO
! OR CLOSES VALVE LINEBREAK l DYPASS (P11) (P10) (P13.P14,PIS) l OK t l d OK 4 I I i i BYPASS ! ! I
- 1 BYPASS i M i i
I i RPV- N ! X # REACTOR BUILDING i i i Figure 19E.219C SUPPRESSION POOL BYPASS PATilS AND CONFIGURATIONS lO I I 0. INSTRUMENT LINES NO LINEBREAK NO CHECK FAILURE (P13, Pid, P15) (P9) i OK i OK BYPASS 1 I RPV (N; [ ] i Figure 19E.219D SUPPRESSION POOL BYPASS PATilS AND CONFIGURATIONS Amendment 19E.2108
ABWR naumas SIRildAEGUIt Rev_A t V E. STATION BLACKOUT AFFECTED LINES ISOLATlON NO LINEBREAK (P8) (P13 P14, P15) OK l OK 1.0 l BYPASS RPV 4g
-- l x
Figure 19E.219E SUPPRESSION POOL BYPASS PATHS AND CONFIGURATIONS F. CONTAINMENT ATMOSPHERIC MONITOR NO CHECKVALVE OPERATOR NO FAILURE CLOSES VALVE LINEBREAK (P9) (P10) (P13) OK OK OK _d BYPASS DRYWELL N X # 0 Figure 19E.219F SUPPRESSION POOL BYPASS PATHS AND CONFIGURATIONS Amendment 19E41W t
7-.
- ABWR uxuwas Rn A Standard Plant G. DRYWELL WETWELL VAC. BKRS l CHECKVALVE FAILURE (P9)
OK l BYPASS I I DRYWELL g N WETWELL CONTAINMENT VENT l ASSUMED Figure 19E.219G SUPPRESSION POOL BYPASS PATHS AND CONFIGURATIONS O H. ATMOSPHERK) CONTROL SYSTEM CROSSTIE NO INADVERTENT NO AO VALVE OPENNG (P12) FAILURE (P9)
- OK OK BYPASS I
I DRYWELL X , X WETWELL I Figure 19E.219H SUPPRESSION POOL BYPASS PATHS AND CONFIGURATIONS O --
ABWR zwimxs Standard Plant a, A
- 1. ORYWELL PURGE NO AIR VALVE VALVE OPEN FAILURE (P6)- (P11)
.. Og OK BYPASS DRYWELL X, W - STACK Figure 19E.2191 SUPPRESSION POOL BYPASS PATHS AND CONFIGURATIONS - I
\ J. SAMPLE LINES OR SUMPS VALVE FAILURE
, NO STATION NO LINE (SUMPS ONLY)
BLACKOUT (P8) BREAK (P13) (P9)
- 1 OK BYPASS BYPASS-I
) RPV X X [ CSAT l-
~ l . REACTOR DRYWELL X g X [ '/, BUILDING SUMP
/ Figure 19E.2-19J SUPPRESSION POOL BYPASS PATHS AND CONFIGURATIONS Amendment 24 19E.2.li t
- - _ _ - _ - _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ . _ s
, .- i 1' i l !- ABWR naamas i j; Standard Plant nev A l i l 1 l 6 i l 2 ! i i !' i l l K. SRV DISCHARGE NO SRV DL ! ADS /SRV BREAK j OPEN (P14) , OK i t 1 i i OK l 1.0 - l i i BYPASS l ', I i ' I n l C ' " !- l. i RPV
- CONTAINENT VENT ww
, ASSUMED I i i l Figure 19E.219K SUPPRESSION POOL BYPASS PATHS AND CONFIGURATIONS i 1 t 1 4 1 J I ) l i 1 l Amendmen 19E.2 Il2 - l' l i
i ABWR 23^ * ^s Standard Plant Rev. A a LINE BREAK LINE OPER. SECOND DIV. COOLANT } OUTSIDE ISOLATION ACTION NOT AFFECTED MAKEUP (V3) (X3) (P1) (O 1) (Oo) ~ i REACTOR INSTRUMENT LINE (30) j RWCU INSTRUMENT LINE (4) i LCS INSTRUMENTS (9) OK 1.1E 6 1.1E 8 NON BYPASS 1.0E 2
-0 NON BYPASS 8.4E 3 l j
OK ! l1.0 1.1E 6 l 9.2E 11 BYPASS 4 j HPFL WARMUP (2) LPFL WARMUP (2) l OK } ! 9.6E-4 . 1.1 E-6 y 1.1E 09 BYPASS }
,1.4E4 OK I .5 1.1E-4 4 7.4E 14 BYPASS j
POST ACCIDENT SAMPLING (4) l ( I' 0 NON-BYPASS , l 9.6E 4 0 1.0 t i OK i 1.0 i ! BYPASS 1.1E-9 1 SLC INJECTION OK 1.1E-6 2.6E 10 NON BYPASS 2.4E-4
""" 0 I
1.5E4 - OK 1.0 1.1E 6 ~ i 4.0E-13 CYPASS I i TOTALS i ' NON-8YPASS 1.2E 8 BYPASS 1.1E 9 , !. I
- j. Figure 19E.2 20A SMALL LOCAS OtJTSIDE CONTAINMENT
! Amendment 19E.2 !!3 i i 1 i
. . ., - - . - - . . . . . .-.-..1
1-23A6100AS 4 Standard Plant a,, x l l i LINE BP.EAK LINE OPER. SECOND DIV. COOLANT l OUTSIDE ISOLATION ACTON NOT AFFECTED MAKEUP
- (V)) (Xg) (Pg) (Og) (Oo) i HPCF DISCHAAGE (2)
OK 8.6E 6 2.3E 10 NON BYPASS 1 3.2E 5 . 0 OK l 4.2E 3 4 8.6E-6 5.8E 13 BYPASS 1 1 OK 1E 3 ! 3.7E 3 j RCIC STEAM SUPPLY (1) 2.4E 13 BYPASS j RWCU SUCTON (1) 0 OK i i 3.2E-5 OK ) 0 OK 1.0 ' 8.6E 6 2.8E 11 BYPASS 1.0 OK 1E 3 ' 3.7E 3 1.2E 11 BYPASS [ SRV DISCHARGE (8) 0 NON BYPASS l 1.3E 4 0 NON BYPASS i-1.0 4 . OK 1.0 6.2E 7 8.1E 11 BYPASS i t TOTALS i NON BYPASS 2.3E 10 BYPASS. 1.2E 10 l l@ Amendment Figure 19E.2 20B MEDIUM LOCAS OUTSIDE CONTAIN!1ENT 19E.2-114
- . _ _ . . - .. . - - -- . . ~ . . . . - - . . . - _ . - _ . . - - . - . . . . - . - . . -
l
- ABWR 23AMAS Standard Plant Rev.A 4
i LNE BREAK LNE' OPER. SECOND DIV, COOLANT
- OUTSIDE ISOLATION ACTION NOT AFFECTED MAKEUP
- (V1) (Xg) (P1) (O g) (00) 4 i MAIN STEAMLINE (4) i OK
,I 6.1E 7 2.0E 11 NON-BYPASS 3.2E 5
- - 0 f
f og ' 1E 3
- . 6.1E 7 i 2.0E 14 BYPASS-
' 1.0 1 OK Negl i i Negl BYPASS FEEDWATER (2) f j (INCLUDING RWCU RETURN AND LPFL A DISCHARGE) 1 OK 6.1E-7
- 3.9E 11 NON BYPASS 6.4E 5 OK 1.5E-3
! 6.1E 7 5.9E 14 BYPASS - 1.0 i ! OK Wegl i E Negl - BYPASS f SRV DISCHARGE (8) i (LOOPS B AND C) OK ' 8.5E4 1.4E 10 NON-BYPASS 1.6E-5 OK i
'#E4 l i 8.5E4 2.8E 13 BYPASS
,5 OK 1E 3 3.7E 3
- 1.2E 13 BYPASS TOTALS i NON BYPASS 2.0E 10 BYPASS 4.8E 13 i '
[ Figure 19E.2 20C LARGE LOCAS OUTSIDE CONTAINMENT 19E2 It$ Ameneent i 4
, .- , - , ..-,-n.. - .. .n. n.,,a--.., , ,.n.
ABWR 23A6 AS Standard Plant A O E 1 - 2 6- B
> I !
C b
- a. \\
i z 0.1
\\
@ case 1h\
w ix . O l Case 4 g 5 N o 0.01 x 0 t
=!
i
' ~
s x
$ \
C 1 0.001 0 10 N 140 2'O 4'O 6'O 8'O 1$0 160 EARLY WHOLE BODY DOSE IN REM i Figure 19E.2-21' WHOLE BODY DOSE AT 1/2 MILE AS A PROBABILITY OF EXCEEDENCE O i Amendment 77 19E.2116
3, ABWR , Standard Plant 23AsdQs . O() s e _ 1 E-06 cc I T w
- w 0 1 E-07 Z
. w o 0 ~ N u. 4 O 1 E-08 ;N, W/ COPS 'g g t ; 's j ^ s
@ N N , W/O COPS 'I g K N# -
1 E-09 . , 0 5 10 1'S 2'O $5 $0 $5 40 EARLY WHOLE BODY DOSE IN REM Figure 19E.2 22 IMPACT OF COPS ON RISK
-~ 't Amendment 77 19E,2 117
i 4 i t ABWR 23A6100AS Standard Plant REY.A O . O 4 i 3 ys, WEIWELL 1
/ Ne
...................................... V NWL 6)6150'
~~~
1l i ! LOWER DRYWELL SUPPRESSION POOL i 1 l i 4 i i First Row of Vents t EL (-) 9700* -- i e h 9 2m mm L EL (-) 10:00* - ~f- - - i x mm 375 lil Flooder Valve # "' k ""
-~
a v Water Flow i l Second Row of Vents 1 m (min) EL (-) 1IMO* - es es 4 4 4 s iBouom of Drywell A s v. ^ :_+___ EL(-) 11630* j
- Elevations based on RPV bonom at EL 0-2 Figure 19E.2 23 LOWER DRYWELL FLOODER SYSTEM k
- A %, l l
i Amendment 77 19E 2-Il8 l
,l
. 7..
i: ! ABWR 23 ui m s
- Standard Plant nsv. A lO i
I i 1 / s / s i 1 1 I m - ?
- Y///./b V////A o
-V////b W
V////W SUPPRESSION
. POOL j WATER
}
/ ,
Stainless Steel Disk I k Teflon Disk ]Q I R k l k FUSIBLE k METAL \ l N -PLUG ,k i s ! 5- Annular Groove s l . Plastic Cap i i Figure 19E.2 24 g FLOODER VALVE ASSEMBLY 1 . Amendment 77 19E.2119 i N
ABWR meioors Standard Plant REV.A (3 W ' Q f 19EA.1 g
SUMMARY
DESCRIPTION Determination of a new setpoint is not completed (tollowing strengthening of the drywell head.) I
; Direct Containment Heating (DCH) is the sudden Subsection 19E.2.1.2.2.2 (2) (b) indicates the operator actions which could be taken to assure SRV operability
" heatup and pressurization of containment resulting from under these conditions The appropnate operator actions ,
the fragmentation and dispersal of core material in the are specified in the ABWR EPGs. Since the y, containment atmosphere. DCH is a concern for sequences in which the vessel fails at high pressure containment pressurizes very slow ly, over a period of ~- since the steam flow from the vessel provides the about a day, there is ample time for the operators to 9 motive force for entrainment. In the event of a take the appropriate actions. g T sufficiently large DCH event, the containment could d Given the above discussions one may conclude that fail at the time of vessel failure. Since this could lead to very high releases to the environment, a study has the ADS system will not be compromised before vessel been carried out to investigate the uncertainties in the failure in the unlikely event of a severe accident, and challenge to containment due to a DCH event. In the the frequency of severe accident sequences in which the past, this issue has been primarily addressed for vessel fails at high pressure is extremely low. However, with the many sources of low pressure Pressurized Water Reactors (Reference 1) since BWRs have very reliable vessel depressurization systems. injection available to the ABWR to prevent core Thus, the frequency of eccidents with the vessel damage, the frequency of all core damage sequences is remaining at high pressure is extremely low. very low. Therefore, high-pressure core melts appear as contributors to the total core damage frequency, albeit Subsection 19D.6.2.5 provides an evaluation of with a very low probability. the ADS system reliability including the nitrogen, control and instrumentation systems. Additional A detailed study utilizing event trees was performed information about the SRVs and the ADS system may to assess the peak drywell pressure resulting from a be found in Subsections 5.2.2, 7.3.1.1.1.1 and DCH event. The following outlines the analysis and 19 E.2.1.2.2.2. the results.
/9 Subsection 7.3.1.1.1.1 (3) (h) indicates that the V signal cables, solenoid valves, safety / relief valve operators and accumulators are located inside the drywell and are designed to operate in the most severe accident resulting from a DBA LOCA, including the radiation effects. The conditions in the containment during the early stages of a severe accident (before vessel failure) which requires depressurization using the SRVs are less challenging than those specified by a g DBA LOCA. Additional analyses of the ADS system e capability were performed in support of station r- blackout performance analysis. This discussion is 9 included in 19E.2.1.22.2. The conclusions of that
. q analysis are that there is ample DC power for the 7 operation of the SRVs for many days after the 8 hour 3 capability required by the station blackout rule.
- e Section 5.2.5 indicates that the nitrogen accumulator capacity for each valve is designed to be sufficient to open for one actuation at drywell design 4 pressure even if the air supply to the accumulators is lost. The risk significant severe accir'ents in the ABWR PRA remain below the design pressure of the containment in the time period before vessel failure.
Valve operability at high containment pressure conditions are also discussed in Subsection 19E.2.1.2.2.2 (2) (b). Based on the presence of the containment overpressure protection system, the maximum drywell pressure is approximately 100 psig. Amendment ?? 19EA.1 1
j , i ' ABWR' i 23A6100AS i Standard Plant REv. A __ ! 19EA.2 DESCRIPTION OF EVENT 19EA.2.1 Event Headings l TREE ANALYSIS i The important parameters and assumptions which l The early containment failure event tree analysis are considered as headings on the main event tree rmd I 1 consists of a main tree (Figures 19EA.21,19EA.2-2. the DETs and the reason for their use are discussed and 19EA.2-3) and three supplemental decomposition - below. j event trees (DETs) (Figures 19EA.2-4; 19EA.2-5, and l 19EA.2-6). The first two events on the main event tree - 19 EA.2.1.1 Containment Pressure Prior to ' i . sort the sequence classes by reactor pressure vessel- RPV Failure (CONTPRES) l (RPV) pressure and pre-existing containment pressure 4 at the time of vessel failure. These parameters are The pre existing pressure of the containment is i uniquely determined by the accident class attributes. obviously important in the assessment of containment The last event on the main event tree assesses the pressurization following vessel failure. Three pressure probability of drywell head failure following vessel regimes have been selected to represent the range of i failure. The probabilities for this event are evaluated in possible pre-RPV failure containment pressures. I supplemental DETs. Three DETs were constructed to MAAP ealculations (described in Section 19EA.2.2 and l assist in the quantification for accident classes with summarized on Table 19EA.2-1) indicate that ABWR
- high RPV pressure. (Low RPV pressure sequence accident sequences can be grouped into three classes.
! classes are not expected to lead to containment These pressure regimes are similar to those selected to pressures which would challenge the _ integrity.of represent pre RPV failure containment pressures in
- containment.) ne three DETs assess the probability of NUREG/CR-4551 (Reference 2).
! containment failure for different pre-existing containment pressures at the time of RPV failure. Class Pressure Range Examples l g
- The DETs consider the major phenomena which low 15 - 30 psia Non ATWS contribute to early over pressurization of containment (Nominal = 1.5 atm) sequences with l Operable DHR or
- from high RPV pressure sequences including debris entrainment from the lower drywell, Direct with rapid core .
- \
l_ Containment Heating (DCH), the pre-existing pressure damage (i.e. all in. in containment prior to RPV failure, and the pressure vessel mj,ection 1' rise due to blowdown of the RPV. Each pathway failed)
- through a DET represents a possible accident 1 progression pathway given the uncertainties in the Inter 30 - 45 psia Large LOCAS with underlying phenomena. A peak containment pressure is (Nominal = 2.5 atm) early failure of t associated with each pathway. These pressures have DHR. SBO with been estimated from a deterministic DCH model RCIC and failure of (described in Section 19EA.3) with input conditions DHR.
l' !- which reproduce the parameter values and assumptions
- along each sequence pathway on the tree. These High > 45 psia ATWS with RCIC.
pressures were then compared with the containment (Nominal = 4 atm) fragility curve (developed in Attachment 19FA) to l determine the probability of containment failure. This event is quantified based on the sequence i accident class. This is a sorting type event. The - { , The probabilities for each sequence pathway with probability of each branch is either 0 or i depending i similar end states were summed and these results upon the attributes of the accident class. j transferred as the branch probabilities for the last event on the main evenurec. 19 EA.2.1.2 RPY Pressure at RPV Failure
, The spectrum of pressures and associated
- probabilities represented by the quantified DETs The RPV pressure at the time of vessel failure is a
- j. ' represents _ a discrete probability distribution on major parameter impacting a numbar of processes containment pressurization followmg vessel failure, which contribute to containment pressurization at RPV i This distribution is a representation of the uncertainties failure. Blowdown of the reactor vessel following' associated with the estimat,on i of containment failure from' elevated pressure contributes directly to pressurization due to the phenomena occurring at vessel -- containment pressurization. High RPV pressures I
O failure. i Amendment ?? 19EA.2. I 1 g
ABWR 23A6tooAs i Standard Plant REV A 1 (3 l C/ promote entrainment of the debris from the lower drywe3 and debris fragmentation, Small: Initial area equal to the area of one controlmd drive ( i penetration (< 0.1 m2 ), Section 19EA.3.1 describes the mechanism for entrainment and the potential for debris dispersal in the Large: Nominal failure area of ABWR. 2.0 m 2, 4 Two pressure regimes are considered- For quantifying this event, the results from NUREG/CR-4551 were used as guidance. In this High (> 200 psia), analysis, as in NUREG/CR-4551, it was assumed that all breach sizes greater than 2.0 m2 could be treated Low (s 200 psia). identically. Fct all core damage sequences where core damage progressior, is not terminated in-vessel (and For sequences with low RPV pressure at the time vessel failure is predicted) NUREG/CR-4551 indicated of vessel failure, the mechanisms which may lead to that the mean probability of small penetration failures rapid containment pressurization are generally not was 0.75 and the probability of large lower-head failure I operative. As discussed in Section 19E A.3.1, modes was 0.25. entrainment cithe debris is an essentic prerequisite for DCH. The entrainment of debris from the lower drywell Analyses performed subsequent to NUREG/CR-occurs due levitation by the steam expelled from the 4551 indicate that the probability of large creep-rupture vessel after vessel failure. For sequer.ces with the RPV lower head failure modes may have been overestimated at low pressure at the time of vessel failure, there is no (References 3 and 17). Even though the best-estimate driving force for the steam. Consequently, in the event studies indicate a small oenetration failure is expcted, tree for early containment failure due to DCH, the this analysis addresses hole sizes up to 2 m . The probability for early containment failure for low. probability of a larger failure is judged to be quite low pressure sequences is set to zero, based on References 3 and 17. Therefore, the
,,- probability of large lower-head failure modes has been (3'j For high-pressure sequences, on the other hand. decreased to 0.1 in this analysis. nus, mechanisms such as DCH and RPV blowdown may challenge the integrity of the containment. The P(Small) = 0.9, remaining events on the event tree assess those mechanisms which impact containment loading for P(Large) = 0.1.
high-pressure sequences. These prooabilities are considered to be appropriate This event is quantified based on the sequence for both early and late core damage sequences. Vessel accident class. This is a sorting type event. The ablation is primarily consolled by the superheat in the probability of each branch is either 0 or 1 depending core debris, and is less influenced by the time of core upon the attributes of the accident class, damage. Thus, the time of core damage will have little effect on the mode of vessel failure. 19 E A.2.1.3 Mode of RPV Failure (MODRVFAIL) 19EA.2.1.4 Fraction of Core Inventory Molten in Lower RPV Head (RVCORMASS) Following slumping of the molten core debris into the lower RPV head, thermal attack on the lower head His parameter largely defines the potential for and lower head penetrations will eventually result in large scale DCH events. It is generally ccmsidered that bottom head failure (unless the debris is cooled in. only the debris that is molten in the lower head at the vessel). Seveal modes of vessel failure have been time of vessel breach will have the potential for considered to be possible ranging from a limited area dispersal and fragmentation. Thus, only this matenal failure of one er more instrument tubes, drains or can significantly contribute to DCH. control rod drives or creep-rupture failure of the lower head resulting in a large diameter failure. Two regimes are considered: This event is a split fraction, representing Small 0-20% Core Debris uncertainties in the phenomenology. Two size classes Inventory (Nominal 10%), p) ( v were defined for this study: Amendmem 71 19EA 2 2
4 9 i ABWR 23A6100AS Standard Plant asy. A
)
Large 20- 60% Core Debris oxidation of zirconium is much higher than that of i - Inventory (Nominal 40%). other metals, the use of a high zirconium mass bounds i the effects of the other metals. Even at the four minute 4 . These mass regimes are similar to those chosen in mark, the distribution of the masses is conservative due NUREG/CR4551 to represent the Grand Gulf plant, to the relative heats of reaction for zirconium and other i metals. Dus, the table cicarly she ws that the assumed NUREG/CR-4551 provides probabilities for three masses bound the BWRSAR results. l cases which appear to be also applicable to the ABWR. ne mean NUREG/CR 4551 probabilities are presented 19 EA.2.1.5 High pressure Melt Ejection (HPME) ! below: (1) Case 1 For sequences with water injection into For sequences with high RPV pressure at vessel l failure, the core debris is likely to be expelled from the i the reactor sessel prior to vessel breach by 1 low-pressure or high-pressure injection systems: vessel at high velocity. Furthermore, the velocity of i the residual gases blowing down from the reactor vessel
= 0.975, are likely to be sufficiently high to result in significant
- P (Small) entrainment of the debris from the lower drywell and to
= 0.02J. result in dispersal and fragmenta: ion of the debris. His i P(Large) j event is a split fraction indicating the uncertainty in i For high and low-pressure Phenomena. The question evaluates whether a (2) Cases 2 and 3 substantial fraction of the core debris is expelled from i sequences without in vesselinjection:
1 the vessel at high velocity and followed by the blowdown of the vessel Given these precursors,it is l P (Small) = 0.9' believed that material will be lifted from the lower drywell floor. A subsequent t. vent headmg (FRAG) will ! P(Large) = 0.l* assess the extent of debris fragmentation md dispersal nt eUPPC W wcR: fh i %/ Since the majority of ABWR core damage sequences do not involve late water addition to the core, Fodl cases where reactor vessel failure occurred
- it is conservatively assumed that the Case 2/3 results under high-pressure conditions, the probability of an apply to all ABWR core damage sequences. HPME was assessed to be 0.8 for the Grand Gulf plant 3
in the NUREG/CR 4550 study. Based on similarities It can be shown that the core debris discharge rates in the design of the ABWR and BWR-6 vessel bottom l used in the ABWR DCH analysis bound results typical heads, it assumed that these results can also be applied of a BWRSAR calculation (Reference 18). Table to the ABWR. Additional discussion is provided in 19EA.2-2 compares the approximate debris masses Sect on 19EA.3.1. The No HPME value of 0.2
- released from the vessel at selected intervals after the represents the potential that the gas from the RPV will
- vessel has failed. The ABWR DCH analysis column break through the core debris and the vessel will bc shows the values used for a small mass. It should be depressurized prior to the release of the core debris and noted that debris entraiament will occur only until the the potential that the initial vessel breach will be near j vessel has depressurized to about 200 psia. The the melt surface. BWRSAR results (Reference 18) a BWRSAR results mdicate that the RPV depressunzes ndicate that the RPV nepressurizes before any
- in about four minutes. De analysis of this study has a substantial amount of core material is expelled from the 3 much larger vessel failum area due to ablation, thus, the ye33,j.
depressurization is more rapid. The pressunzation of the containment is most rapid before the wetwell =0.8* P (HPME) connecting vents clear. Vent clearing will occur within the first second of the blowdown. Therefore, the very P(No HPME) = 0.2-early stages of the debris pour and entrainment are the most significant. For reasons similar to those discussed above for I W-Pressure sequences, if an HPME event does not The total mass of debris and the zirconium mass , ccur then the loads imposed on the containment used in this analysis are much larger than the masses structure at vessel failure will not result in a seriously calculated by BWRSAR. Indeed, the mass of the threat to containment integrity and the probability of
/ zirconium bounds the eniire zirconium and metal mass early containment faHure is assumed to be uro.
( calculated by BWRS AR for the critical, early stages of the blowdown. Since the heat of reaction for the Amendment 71 - 19EA.2 3 l
ABWR DA6100AS Standard Plant RF.V.A (O ! 19 E A.2.1.6 Fraction of Entrained Debris Fragmented and Transported to the Upper is released at vessel failure will be trapped in the lower drywell. The region of the lower drywell above the Dryw ell (FRAG) downcomers does not have any open now paths. Furthermore, the control rod drive mechanisms are This branch in the DETs is a split fraction event, krated in this region. Therefore, the velocities in this For high. pressure sequences where an HPME has region will be lower than that in the region below % occurred, this event assesses the e< tent of dispasal and downcomers. Material which has been lifted off fragmentation of the entrained Jebris. In 0; der for a Door could become trapped in these more stagnare serious overpressure challenge of the containment by regions of the lower drywell above the downcomer. dircet containment heating (DCH), a significant fracuen Thus, one would not expect that all of the debris of the debris that was molten in he lower RPV head at entrained in the gas flow would exit the lower drywell. vessel failure must be transported from the lower drywell into the upper drywell t nd fragmented. The Once debris is assumed to leave the lower drywell mechanisms which may limit tre transport of the and enter the downcomer, two mechanisms govern the molten debris from the lower cavity to the upper cavity final distribution of core material. These mechanisms are discussed below. These include; are the impaction of core debris on structures and the transport to the suppression pool due to flow toward (1) Trapping of the debris in the lower drywell, the wetwell in the downcomer. The gas transport pathway to the upper drywell is relatively convoluted. (2) Impaction and removal of the debris in the gas For the debris to enter the upper drywell it must be transport pathway connecting the lower and upper entrained off the drywell Door, Dowing vertically along drywell compartments, the drywell wall it then turns 90 degrees to enter the horizontal piping. After flowing a short distance (3) Partitioning of the gas (and entrained debris) How through the horizontal piping the flow will encounter a exiting the lower drywell between the upper Tee type junction with the vertical downcomer. At this drywell and the wetwell, point the entrained debris must again tem 90 degree. There is potential for impxtion on the downcomer wall at each turn. This impacted debris is likely then to now h d (4) Debris dispersal by wave formation rather than by small particles. downwrd along the downcomer wall toward the wetwell vents. This effectively removes the material fr m p rticipating in the DCH event. The above mechanisms can impact the extent of debris dispersal to the upper drywell as small debris in addition, if the horizontal wetwell vents have particles w hich is the critical parameter for determining cleared then the entrained flow will split between that the potential threat from DCH. going upward toward the upper drywell and that going downward toward the werwell vents. If the vents have The basic configuration of the ABWR lower not (yet) cleared, then all the Dow will go upward
! dh *d is shown in Figure 19EA.2-7. Additional ,
toward the upper drywell. The debris that partitions details can be seen in the arrangement elevation drawing
& with the gas now going downward toward the wetwell (Figures 1.2-2 and 1.2-2a), the lower drywell elevation vents will n t p rticipate in DCH.
(Figures 1.2 3b and 1.2 3c) and the arrangement plan ct (Figures 1.2-13e inrough 1.213h). De vessel skirt of Finally, since DCH relies on the rapid heat transfer
% the ABWR is solid, there are no openings in it which O from the corium to the surrounding gas, any debris could connect the upper drywell to the lower drywell, which is transported via a wave type motion will not This precludes water transport from the upper drywell into the lower drywell following a LOCA. Hence, the participate m the DCH event. As discussed above and described in detail in Section 19EA.3, wave formation Gow pmh for gases and debris expelled from the lower ,
drywell will be through the downcomers. The upper is n t expected to be the dominant transport mechanism. Most of the debris transported to the upper drywell to wetwell downcomers are imbedded in the drywell is expected to be in the form of particulate. lower drywell wall. ne downcomers are also connected to the lower drywell gas space via horizontal pipes which penetrate the lower drywell wall at an elevation The fraction of the molten debris in the lower head approximately two-thirds of the height between the that is dispersed into the upper drywell and fragmented lower drywell Goor and the top of the lower drywell. can be represented by the following relationship: Because of the lower drywell configuration it is f = Idowncomer X f unpaci X Ispht upd* X feart Irag (l) [3 Q expected that some fraction of the molten debris which Amendment ?? 19E A.h 4
I I ABWR 23AsiooAs i Standard Plant REY.A ' p) s V where:fr,,, = the fraction of debris trans. fraction of debris that does not get removed due to impaction. ported to the upper drywell and fragmented Prior to clearing of the horizonta' <ents, gas flow f downmner = the fraction of debris which ir.to the vent pipes win ? direcir* up into the upper gets entrained out of the lower drywell. As the upper and lower drywells pressurize, the drywell into the downeomers water level within the vent pipes will be depressed, and (estimated uncertainty range venting into the wetwell will begin. The average vent 0.25 - 0.75), clearing time is 0.5 seconds. After the vents have cleared, the gas would preferentially flow into the
= the fraction of the debris enter, wetwell since the upper drywell would be increasing in f i,n p.n ing the downcomers which pressure. If the wetwell and upper drywell pressures were conservatively assumed to be equal, then a 50/50 does not remain permanently impacted on the downcomer spht would occur based on cqual flow areas m both directions. As described ,m Section 19EA,3, the walls (i.e. eithei is not im, entrainment dispersal time is estimated to be two pacted or is impacted and re- ,
entrained)(estimated uncertain, seconds, if the debns is dispersed imcarly, then the vents would clear after only 25% of the debris was ty range 0.5 - 1)' released from the lower drywell. The debris leaving the
= the fraction of the gas flow lower drywell after vent clearing would then fspla.upa. c nservat vely be split 50/50 between the wetwell and which goes upward toward the the upper drywell. Based on this discussion, a median upper drywell (as opposed to y lue f 0.75 was conservatively selected to represent 4 the wetwell)(estimated uncer- the fr etion of debris that flows to the upper drywell.
tainty range 0.5 - 1), The final phenomenon that could influence the (p.n = the fraction of the debris en- amount of the debris that would participate in DCH is trained into the upper drywell the wave formation, if the debris enters the upper y i which enters m the form of drywellin the form of a coherent wave,it would not be A small particles (estimated un- expected to participate in mixing with the gas, certainty range 0.75 - 1). Experiments performed at ANL for PWR cavity configurations have resulted in this wave type of The uncertainty ranges for these four parameters sweepout. As discussed in Section 19EA,3, wave l were chosen based on the physicallayout of the lower formation is not expected to be the dominant removal drywell along with engineering judgment. A median mechanism for the ABWR configuration. Ilowever, as value of 0.5 has been chosen to represent the amount of debris flows through the wetwell/drywell connecting material expelled from the vessel which exits the lower vents to the upper drywell, it is possible that some of drywell. As described earlier, this represents the the debris forms wave-like sheets. Therefore, potential for material to be trapped in the stagnant engineering judgment has been used to estimate that the region above the horizontal vent pipes. The ANL median value for the fraction of debns that is dispersed experiment described in Section 19EA.3 did not include as particulate debris is 0.875, any below vessel structures. It may be possible to freeze and hold material on these massive structures. Assuming a uniform distribution for each of these Furthermore, the openings from the ANL cavity are parameters between their assessed upper and lower much wider than the openmgs m the ABWR. Since the shown in debris will have to make a turn to enter the ABWR bounds results Figure 19EA.2-8 Thisindistribution the distribution for f has for the frr, frag openings, the smaller area makes the debris less likely to entram. Based on the above discussion, a median follow ng characteristics: value of 0.5 appears reasonable for f downmner. median value 0.2.3, The debris leaving the lower drywell will then 0.34, 80th percentile travel a short distance before entering a tee junction. It is expected that some of the debris will impact on the wall of the pipe at the junction and now down into the 99th percentile 0.60. wetwell. Based on the physical characteristics of the Based on the at>ove results, three regimes were O debris now path, it was judged that a median value of selected to represent this parameter in the event tree: Q 0.75 would provide a conservative estimate for the Amendment 'M 19EA.2 5
ABWR mt.s <- Standard Plant REV.A l
- [
! - Low (f tras s 0.35), intermediate (0.35 < fr s s 0.60), j and High (f tr. > 0.60). In the deterministic DCH l pressure calculations described in Reference (1), the
- nominal values for (fras used to represent these three-i_ regimes were 0.25, 0.5, and 0,75. The estimated j probabilities for each of these discrete regimes for this j parameter are shown below
.' P (Low) = 0.8, 1 j P(Inter) = 0.19, t P(High) = 0.01. j Figure 19EA.2 9 shows the comparison of the !- calculated cumulative distribution, as determined from j the above parameters. and the distribution assumed for j the DCH analysis. As the figure shows quite clearly, the assumed distribution is conservative. The entire j range assumed for high fragmentation lies above the 4 calculated range.E For the low and intermediate fragmentation ranges, only a very small portion of the l 4 assumed distribution lies below the calculated } distribution. Therefore, the discretization of Fr,.a used
- for this analysis is conservative.
1 I 19E A.2.1.7 Peak Containment Pressure 5b !v Following RPV Failure - i This event assesses the peak drywell pressure ] following RPV failure. There is only one branch for ! this event. This event summarizes the deterministically j calculated drywell pressure for the set of conditions and
- assumptions specified in the event sequence pathway
- leading to this event. A description of the calculational methodology and calculated results are presented in Section 19EA,3.
j l 19E A.2.1.8 Drywell Head Fails Following i Vessel Failure i This event assess the probability of drywell head l failure given the pressure determined in the previous event. The probabilities for failure are determined from j
- l the drywell Attachment head fragility curve described below in 19FA.
1 4-i i: l 1 i: ' 19EA,2 6 Amendment 'N '
, .- . .- _ . _ , _ _ _ _ , . _ . --_ . . _ . _ ~ . _ . . . . - _ . . . . ,.
ABWR : Standard Plant m ei o s REV.A Table 19EA.21 ( CONTAINMENT PRESSURE AT RPV FAILURE Pressu re M A AP Calculation MPa (psia) LCLP-PF-D M 0.13 (19) LCLP-FS-D-L 0.14 (20) LCH P-PS-D-N 0.14 (20) LCHP-PF P H 0.14 (20) SBRC-PF-D H 0.24 (35) LBLC-PF-D M 0.29 (42) NSCL PF-D.H 0.13 (19) NSCH PF-P-H 0.13 (19) NSRC PF D H 0.43 (62) v j l'\ V Amendmcot ?? e
e i. 2 ABWR 23A6100AS
- Standard Plant asy. A Table 19EA.2 2 -
l COMPARISON OF ASSUMED DEURIS DISCIIARGE \71Til BWRSAR RESULTS i ' Intenrated Debris Dluharned i ABWR DCil Analysis BWRR AR I j Time Zr Metals Osides Zr Metals Oxides - Vessel Failure 0 0 0- 0 0 0 l i i +2 minutes 6500. (1) 17,000 926 3880 0 i
+4 minutes 6500 (1) 17,000 2316 8172' 0 l (1) Only a small portion of the core plate was assumed to be added to the debris.
i t i f a
- \
k f l 4 4 4 s 1 i-
- O i
i Amendment ?? 19EA.2 8 e
1 ABWR 23 A6100AS i I Standard Plant REY.A o \,
.w, ....
t,. ,
- . . ...:..:.. n s,
. .m . . -. e .3.
.. . . . . . . . . _ , . . . c.
E vE%T -> cohT W 55 A ep,W 55 E aat, v;7 [,j,[, 1 3 : b4 t 's I
~
.4,
%L , 3 3 :.:s o,
. . s i L I
I,
'* J )
i Mr
/
, ') s 2
\2 POS 1
- ~ a - v3 - 3 _]
! i i.
. a 7 o aiEM n 10 MIL ~
3 l l Figure 19EA.2 1 7-- i ) DCII EVENT TREE FOR SEQUENCES WITH LOW CONTAINMENT PRESSURE Amendment ?.' 19EA.2 9
r ABWR i3A6tooAs i !- Standard Plant -REY.A 4
.. . .. .. .. . a :
- .?;"a .'.'"; ame
.,a".!.L
. . . -. . . . , . . . . e. e.
! EvEhY > C CF' TE 33 "'"uE 55 E ast e ry i . - i i. i t e a i a l- g . . . , ., m 1 i t o
- i t-1 f
i si.&g . . . in i i 1, ",
' "~ '
pos t ..e,.*,m i- ; i } w. I I ( i 4 " o 4 A 1 -w.. w nm 7 s i e 2
- r i
1 i n i q .
}- Figure 19EA.2 DCH EVENT TREE FOR SEQUENCES WITH INTERMEDIATE CONTAINMENT l {q. PRESSURE i
i 4 i Amendment M 19EA.2 10 i t 1
'w-r
,- + .- g - -- ,-e;. --f--a., ,.,.32 3h g.- w i r- y .. - -
_ . _ _ _ . . _ . . _ ._.~ ___ __ __ _ __ __ _ . _ . . . 4 !. ABWR moi.s . , i Standard Plant REY A 4
. a
- - . . . .. m s,
. ..m.
s. e e c e m g ,pT ., c cp.rpeut 35 8 d141E 55 D84,*CF 1 e d 3 1 4 , a o i a. , . ,, ., w e a J l }' a o i l i i p 9 t i l % 0 i , pas t : . , . .s m i ' i n 1 I l s o i 4 i t i e*.= , . . . , . q a WyJe d i .r i 6 i j '? L . . ....o., k j
, a . .
7 a 1 1 1 a j e Figure 19EA.2 3 l\ DCH EVENT TREE FOR SEQUENCES WITH HIGH COINTAINMENT PRESSURE-l i-Amendment ?? . 19EA 211
. ~ . . - =.- . -
i i R 23A6100AS 4 Standard Plant REY.A f-xs
; ,e. ,
- e. *m.w n.s.
< e, au
..~ .n.e.e.. . .an.o .s . .
a g w.i .i ... n.ua. e. s a i r's.
%.m
==a nue eas .
s-e.s...
; . s , m.m .z.w.is i . -
es .. I,
! men e towr, . ws e. rem .. n .tm ..n, E.thT .* dODu,vr t h AwCC88. 55 .eeg rm.G ;wpuE 55 EsAvCF f.
I .
- n= s ', t- ett e,t. 81a
, . 41 4
; 've f . L .et
.2%'te M1 ' t. P1 ' . !
is
, ~e ,
' . . L .st 3
.i 4,
< t. L M ,
- -, a2 ' .e +
.,..,,, , ,. .u . ;
, 6 ea < n '
=c r e .eiIf13 =c * . t . M ,
a
...m .o,* . n n ,
- x .....,o
. u..u. i
.a.
j 1
- r.th f
44 AaE g
- ~
, . _i.. .. . , a .u . ,o . . u -
, il 4
- a as ne o . , o .u. ,e . . r.-
- i u, , i, ,,
i
- .ee . t t a en1 *o a.tt sir n
> a 4
n . r . ., ,
,e .u. . . u- ,,
j U ,........ . . . . ..,m ,,
- 1.
a 3 <<.u- ,. i?
.g g5 ,,a g, fi o M3 W.
i 1..s- ,,
'I t
) esc = sped 4 Ill #11. 'ed F.it M gg a c 5
.=gte i, 1
- r a. . ., u . .u .
i g .act 4 GI i . ,a gg s.1L.At e o ,.u- ,, 4) r.. . ,,o o . x .u .
. i.
4 wx ,7
..u x ,,
f LS. L f1 '?G 13 - E' i g. 4,s o it s .st. t 80 k 8. kM ,g w3
=c er . til s1f. Su f & tLM g3 4
e d Figure 19EA.2 4 1p DET FOR PROBABILITY OF EARLY CONTAINMENT FAILURE - HIGH RV
- V PRESS AND HIGH CONT PRESS SEQUENCES Amendment ?? 19E A.2 12
ABWR 23 A6100AS Standard Plant REV,A 5 4
.=m, ...
~. . .
m_. w.
..m.,
. m. m. .
.m e
c .m . . erm ..cca..is - rw ca mis c.uu , 1 e. <: m .c. l 1..u. , g,. m4 % . j 4 31 ,
, . . t. .ae ,
e. 4 n . . r. - . , 4 a -. *F ?,'f t 10 ti ti2 og:. ?
...,.x ,
M
! . . :.a ,
- T .12 3 t' - '
!' i a.p t _ w in ;
j . . t uar , , a. J
; y- _ . t d *i t . ,o .. !L M .,
7
- i.
j N a l Ga
- t tLD11. % ..fL M g
- . u '*1 l' '
i> 4 ,~ r ... tu s, i.
, i u .c. c <.ru = '
t
.de t i3 ria .-- <M t . i d *11. ,0 ..tvM g
,e - . i a .si . .< , . m.
J b
=ts= og o it) est. .. h ugt
(._, *" li s;p ~ o i, i. so . iu .c.
~ i.
O P f_8_a*
,e . . u -
a ,, 5.
' 40 .'
af[ Lg,, 979 g) N 3,g ,g g , r.. tLtM ,g 4 .
.c ~
a
. ,o no s uae. i, i
1 . a p;aw i. 1> m . n. .v. G GI 1
. .I it., ,,
~- --
. q. .s. ., . i a .u.
,o ..rmx ,,
- a. es m e o., .n. . , a ., n * ..nu e a ,,
g
.c,
, , i n .u. ,o . . u-8
' Figure 19EA.2 5 DET FOR PROBABILITY OF EARLY CONTAINMENT FAILURE - HIGH RV l
{ s - PRESS AND INTER CONT PRESS SEQUENCES Amendment Tt 19EA.213 I y.. m , # ~w .--.,w +=-
I i l ABWR 23 A6100AS l I Standard Plant REY. A - l 1
<,e... ~;. . .. .. ,....
. m.
....,a,,
.... e
.m.. . - ..-m . . ,
.. .m e .m ., . ,.a ..c-n - =. c m., n e .= ,c, 2
. n,v syx i
- 1. . m .
' , : ai i i..um i
<..n.
. ~ , u i "6 is.ja 39 ,, i M sq l . 1
! '8.k,M ,
17
- j .
- ..e 4
. .n a.IGn . . .,a..,. e ..w , 1 s~ . , a m. v ..un ,
- j. S i
W nm, au
. , a .u. e ..un ,
1 ! - . ,a v.
. , ..,. v v - .a n
- i.
i Lev '21 11
'to .' 112 *.ta w 8.*gM g
- L
- s. Lf 2 m i
., e- , u .u . a ..um ,, d d
- ,s . . r. -
4 :a aj8a 4 9 %) M2 811. t G J1 t M..# M e .w. e ,, r i i. 3 - . !*'f e m 12 1 2 *11 7
. a i 1
9 ,, ~ ad f e!LW
'I f .et i. !S 8 0 41
- u. .,1 , m .u. I eA I.. rume ,,
u . ~ . , a .u . ..u- ,, s 1 i < 0 ,. r.* ,,
,. v
.n , , , i, .u .
- ou
, . . u- ,, e) . ~ q 3 tw't. $10 , . I t 2 *1 f. .v0 f.uR 20 i .n a ya na v .:. .n .,
. , a .u . c ..u, ,,
e 1 40 me.e . t il #11. 908.hiM gg A Figure 19EA.2 6 i DET FOR PROBABILITY OF EARLY CONTAINMENT FAILURE . HIGH RV j 1 y PRESS AND LOW CONT PRESS SEQUENCES i Amenhent M . 19EA.214
I i
- ABWR u .,
Standard Plant nr.v. A !O i a ! I f l ................................. ., j l SECONDARY g
- CONTAINMENT
! -l SOUNDARY e ' ll i l ,, ORYWELL ! e CONNECTING jl l
,l VENT ORYWELL NEAO i i
. I ll-
,8 >
, i r
- J
' pmtMAmy ,
l 'i . I CONTAINMENT
- O ,
um l
.. i l
! a .a OIAPMt.'AOM zeic.e,c.v. 6.: ! 'i'l umR etooA ""
.. I ORYWELL 1
i
,l.
l I,I o u -
. m.-. g g #
O I CLEAN U W l Q suetessa. cHausEn AIR 8 PAC 8 lj zone 4 u .I
. 1. .
4 a i
~
l DA LL WM i , 1 :I McAl2. N i g i n i j j , m esen si n ei - ineuse SAatMAT ! l
\ AlMAAY P
i coeffAINt00NT ecuNoAAY l t l l s t Figure 19EA.2 7 ABWR CONTAINMENT BOUNDARY NOMENCLATURE i i !O p _ f
4 4 l ABWR' u w ooxs l ! Standard Plant: . REY. A ) j- ) 1-i i i i i-i 4 i j 0.04 - i ,L c ~ o , , = r K, v jy i 5 0.03 - j $ i u, :s f3
< x -
t e. 8 3 1 m' s '; y, ; a i A 0.02 - gj @~ll fj , 1 - .,
- .., x v 1
s '; %- +; ' '
?
!' Gi - 2 j 8 6 s 't :x 1 ::: g 5 3 9 3
- j. ]e 0.01 - s x
f I' d fj ., a 1
=
w
?
w 9 1 v 4 2
@2': a. -
,. ::s x
- 3
.j-6
?;-
y i
, f 4 : & g 0 2 ? e y
0 - i Y E UM--- 1 i i i i : -i i i i i i i i i i i ii a O 0.2 0.4 0.6 0.8 ' l .0 - 1-F j frag i i J l i-y
}
9 4 1 4 Figure 19EA.2-8 CALCULATED PROBABILITY DISTRIBUTION FUNCTION FOR DCH j: PARAMETER Frus I
\
1: Amendmen 7; 19EA,2416 }~ 4
4 ABWR 23A6100AS
- Standard Plant - REV. A a
i i i i I i 4 i
; 1
- Calculated
< 0.9 i. ! 0.8j .--- Assumed i : j 0.7 ' l
-l i 0.64:
i u m - 4 d: 0.5i
- d. : [-
j 0.4j . l .- 0.32 1 f. _ . _ _ _ _ _ _ _I 00; i[d j t 0.1 ' 3 i 0',,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, o- . H q t 9 q h m q - o o o o o o o o o l Cummulative Probability i-i 0 1: Figure 19EA.2 9
- COMPARISON OF CALCULATED AND ASSUMED Ft r.: DISTRIBUTIONS 4
J f i n != Amendment 77 19EA.2. I 7
4 i ABWR 23 A6100AS Standard Plant arv.A e j
, ( 19EA.3 FOR DCH DETERMINISTIC MODEL reasons that will soon become clear, we follow IIenry (Reference 5) and leave the coefficient explicitly in the expression. Typical values of Weber number are 6 to 12 when the drag coefficient is leit out. Strictly -
A computer program has been developed to provide Speaking, these would specify the maximum particle scoping calculations for DCH events in the ABWR. sde. ne ass m median diameter, for example, would be i Several simplifications and aisumptions exist in this about half this value (Reference 6), model. This model, and its application to the ABWR design are desenbed below. If we substitute Equation (3) into (2) and neglect th* g s density compared to the liquid density we 19EA.3.1 . Debris Dispersalin the ,n-ABWR
- The purpose of this section is to briefly summarize u,' = 14)
! the available information on debris dispersal from a ( 3 Cc > Ps i configuration like that of the ABWR lower drywell.
I . Define the Kutateladze number by: I 19 E A.3.1.1 Velocity Required to Transport Debris Particles Ku=_, u' , (5) De velocity required to transport debris panicles Pro s }' out of a compartment by entrainment can be easily ' p' 2 ' estimated (Reference 4). To lift a particle of radius t against the force of gravity requires a velocity given by: so that: ! (Pr -Ps) m's = ca n 2 (2)
'4We#
- s 3cj; where
- pr = fluid density.
A free particle in the high Reynolds number limit ps = gas density, has a drag coefficient of about 0.44. If we assume a j Weber number of 8 (the results obviously depend r = particle radius, weakly on this choice), we obtain: i g = acceleration of gravity, Ku = 2.7 (7) 4
= drag coefficient, When substituted into Equation (5), this gives a l cd velocity close to the experimentally measured value j
- us = gas velocity. {uired to entrain particles off a free surface (Refer If we assume complete hydrodynamic breakup, the Consider a different situation in which gas is
, maximum particle radius is given by equating the force sparging a- pool from below. If we consult j imparted by the gas stream to the surface tension force Figure 19EA.3-1 (Reference 4), we note that the drag
; holding the droplet together. This is usually cast in coefficient for a particle bed is in the range 20 40. If we i tenns of a Weber number: use 30 and leave the We number at 8, we obtain:
We= #E'"' d (3) Ku = 0.2 (8) a His does yield just the experimentally measured where: a = surface tension. velocity required to fluidize a pool (Reference 8). 1 , There is some ambiguity on the form of this Rus, we see that the velocity required to lift liquid equation and the choice of the Weber number, We. droplets is a strong function of the configuration. A gas ( Most authors fold the drag coefficient into We. For stream passing horizontally over a liquid pool requires i Amendment 'M 19EA.3 1
AIMR u m nnAs Standard Plant uv. A i a bed,i.e. a situation in which the gas stream proceeds 19 E A.3.1.2.2 Experiments on Grand Gulf vertically from below. This effect results from the Configuration different drag coefficients which apply to the two utuations. nus,one would espect the required selwity Figure 19EA.3 3 shows a schematic view of the for entrainment in the ABWR to be 109 of the salue Grand Gulf pedestal region. Experiments were also for lhe Zion cavity. conducted at Argonne on a scale model of this configuration. Quite differcnt behavior was observed in 19 E A.3.1.2 Argonne Experiments on Debris these tests. Due to the more or less symmetric Dispersal orientation of the scale model CRD ports around the circumference of the pedestal, the gas jet leaving the Spencer, et. al., at Argonne National Laboratory simulated reactor vessel us observed to stagnate, hase conducted a number of debris dispersal proceed horirontally so as to undercut the entire liquid esperiments in several geometries. Extensive pool, turn and proceed verthally using virtually the information on these is available (References 8,9,10. entire cross sectional area of the pedestal (Only a small ~ and t l). llere we shall brieHy summarite the results of fraction of the area was used by the jet moving two sets of quasi. steady experiments designed to dowr. ward from the simulated vewel.) determine the threshold vetxities required to transport debris from simulated reactor cavity / pedestal regions. As fioted in Reference (M), this configuration is reminiscent of a pool sparged from below, Using this 19 E A,3.1.2.1 Experiment on Zion reasoning Equation (8) would te expected to apply. , Configuration indeed, a Kutateladie number of 0,2 was found to accurately predict the velocity required to initiate Figure 19EA.3 2 shows a schematic view of the removal of debris. Visual observanons of the Zion reactor cavity. Experiments conducted with this experiment support this conclusion, as it was seen that geometry (Reference 9) indicate that the threshold the entire pool became Guidized, the liquid rose up to velocity required to disperse liquid droplets into the gas the level of the ports, and was then swept out by the stream and move them from the cavity is gas stream. ' O approximately given by a Kutateladie number of 2.5. Q This is not unes.ccted, since the geometry of the cavity is such as to cause the gas jet leaving the reactor An experiment was also conducted in this geometry with steel shot. A drag rocificient of about 7 to stagnate at the Door of the cavity, to turn and is required to explain the measured velocity threshold proceed horizontally down the cavity keyway over the for sweepout of ~3.5 m/sec. By consulting liquid pool. Thus, the considerations which lead to Figure 19EA.31, we see that such a drag coef ficient is Equation (7) would seem to apply, appropriate for a bed of porosity about 0.8. This is, in fact, the porosity that would be obtained if the entire it should be noted that sweepout, in which a bed of shot was uniformly Huidized up to the elevation continuous liquid film was observed to be Gooded from of the CRD ports. Thi: saggests that the threshold the reactor cavity, occurred at about the same velocity velocity for swecpout in the Grand Gulf configuration as entrainment of droplets. Thus, as noted in Referercc might be a function of the ratio of the initial pool (12), the amount of material transported itom a Zicn. volume to the total volume w hich exists in the pedestal like cavity as droplets relative to the amount under he elevation of the gas flow paths. In other transported as a film is determined by the relative rates words luiditation could begin at a low velocity (w hen of the two proecsses, the porosity is low and the drag coef ficient is high), but as the pool tries to grow toward the exit nowpaths, the An additional experiment was run in which steel velocity required to continue to levitate droplets would shot of diameter 78') x 10 6 m was used instead of a increase. This conjecture cannot be confirmed by the liquid pool. By subs 6ituting into Equation (2), the drag few tests run with liquid pools, however. coefficient necessary to explain the observed velocity threshold for debris dispersal of ~16 m/sec (Reference 19 E A.3.1.2.3 Application to GE AllWR
- 11) is found to be about 0.3. It is not known w hy this Configuration is less than the expected value of 0.44, but the discrepancy is not considered large. The ABWR configuration, Figme 19EA.2-7, is similar to that for Grand Gulf. Thus, we expect that a Kutateladze number on the order of 0.2 should be applied to calculate the dispersal threshold. With the exception of the ANL Grand Gulf work, the V
Ammdment ?? 19EA.3 2 l
AllWR 2]A6HOAS Siandard I'lani RIX A documented cycnments gerfortned to date base focuwd on PWR I)pe cavities such as Zion. As discussed anne, these are not directly applicable to the AhWR configuraticri. l l l t 4 N i Amendment ?? 19E 6.3- 3 1
ANR uimmis Standard Plant utv. A 19E A.3.2 Pressuritatioti 1)ue to 1)Cll mer the time constant for the event (discussed in Section 19EA.3 A). The fraction of the debris alloveed The pressuritathn of the drywell is af fected by the to otidite is a user input (discussed in Section blowdown of gases from the vessel and by the' heat 19EA.3.5).1hc energy of reaction is taken to be that transfer from fragmented corium into the drywell. An for the zirconium steam reaction. Oxidation of the esplicit method is used to calculate the response of the rirconium participating in the 0C11 event is assumed system. The gas is assumed to be an ideal ga': with the to be instantaneous. rate of change of pressure, P calculated from: TM mpn of k dh is chid W m I the amount of energy remaining with the phase change < Ms RI s Nin RT: energy accounted for and assumed to take place at a p, MW,V MW,V uniform temperature of 2500K. Constant specific heat and latent heat of fusion are assumed. w here: M s = Total mass of gas in containment (steam and non-condensible gas), i l R = Gas constant X N M/kg-mole K), T, = Gas temperature, MW, = Average molecular weight of the gas mixture, V = Drywell volume, O g and a dot os er a variable indicates it rate of change with time. The temperature change of the gas, T,, is calculated by assuming that 11 : gas and the fragmented debris are in equilibrium at each time step. Since the DCil event is very rapid, no credit is taken for heat transfer to containment heat sinks. The specific heat capacity for steam is evaluated using a curve fit to saturated steam properties (Reference 13). Constant specific heat is used for the non condensible gas. The rate of change of mass in the containment. Ni s, considers the gas blowdown from the vessel, any flow to the suppression pool through the connecting vents, and hydrogen generation w hich occurs as a result of the re ction between the steam and the rirconium. The mass flow rates through the downcomers and from the vessel are evaluate 1 using a compressible flow model(Reference 13) 'Ihe pressurit.ation of the wetwell due to any addition of nc,.rcondensible gases is considered. Steam which pass:s through the connecting vents is assumed to be quenched. The debris conditiens are calculated by conservation of energy in the system. The mass of debris participating in the DCil event increased linearly v Amendment ?? 19E A 3-4
AllWR 23 A610nAS Standard I'lant ;&v. 4
.b 19EA.3.3 Calculation of Vent Clearing Tiene The DCil program previously desented includes a ; model to predict the time required to elear the honiontal J vents and begin gas flow to the welwell. The malet. ;
based on analysis by hkuly tReference 14), requires as j input the pressuritation rate for the upper drywell. The l DCll model computes the pressuritation rate for each
; time step. Given this, Moody has derised a simple formula for the water velocity resulting from this ramp pressure. The DCil model then computes the water movement as a function of time; and, based on a table look up of vent area vs. water level, calculates the appropriate dryw cil vent area a'. any po nt in time.
i i 4 J d Q Amendment 77 19EA 19
ABWR m6ms Standard Plant REY.A 19E A,3,4 Calculation of Dispersal Time model is applied to the SNL DCil 1 experiment. O'w/ Constant For this calculauon, however, the entramment , a wen direcdy imm experimental data. In l
, addition, the model was not applied to a full, l For the parametric modeling of DCil in tha reactor scale scenar o, only to DCil 1; analysis, a timescale for dispersal must be input. %s i I
in0uences the rate of containment pressurization by defining the u.trainment rate of the debris. (1) Sienicki and Spencer at ANL have written a rela-O ely sophisticated one dimensional hydrodynam-There is a dearth of good modelt for DCH, The ics model called il ARDCORE (Sienicki and Spencer, undated). Separate mass, momentum and i only models w hich were idenufied are: energy equations are written for the liquid film, I the droplets and the gas. The entraintnent ] (1) %c CONTAIN hkxlel correlation is based on liquid jet tweakup formulas developed by De Jarlais, Ishii, and Linehan. Being This is a lumped parameter model in which the one-dimensional, the model does not, of course, rate of entrainment is input by the user. It taken into account non uniformities in velocity, provides no insights for this study; though there is considuation given to entrainment frorn annular films on the cavity walls. (2) Henry has developed a model for ARSAP (Reference 15) that explicitly compares the time" The model was applied to the ANL CWTI 13 scalc for dispersal due to the acceleration of the esperiment and to DCil l. In both cases, it is liquid filtn as a whole to the time required to stated that the debris entrainment time was entrain the debris ,;5 droplets. predicted fairly accurately by the code (time scales on the order of 0.1 seconds). When the code was his model is very attractive from the standpoint then applied to a full scale Zion TMLB accident, that it produces closed form answers and the predkted timewale for sweep +ut of the debris illustrates that the competition between the two from the cavity was of order 2.5 seconds,i.e. the l modes of debris removal from the cavity may be numerical results are fit rather well by:
- an important consideration for designing and i i V; interpreting experiments. Ilowever, there appear to be several problems with this model. First, it m, = ng(1- eM$} (10) assumes a very schematic debris configuration, i,c. un initially static debris pool lying on the where; t = time in seconds since the Door of the cavity. It seems more reasonable to b
- owdown begins; assume that there is debris splashing throughout the cavity, as point out in Levy's WRSIM papers (5) The recent papers by Levy, mentioned above, (Reference 16) and in Spencer's work at ANL. contain an explicit closed form expression for the Next, it is questionable that the entrainment rate time dependent entrainment of debris from the formula that is used, the one developed by Ricou reactor cavity. This formula has been compared to and Spalding for gas gas entrainment, applies to a wide variety of small scale test data with this situation. There is evidence (cited by Levy) remarkably good results. De formula was applied that non uniform gas velocities in cavities may to calculate the entrainment rate for a full scale play an impor' ant role in enhancing entrainment Z on like cavity in a TMLB-type sequence; if one rate. Finally, only very limited compat! sons to assumes that steam exists in the cavity (the data have bec . offered- results are apparently quite sensitive to the gas density there due to the strong dependence on Taken at fa:.e value, llenry's model tends to Euler number), one obtains the seemingly predict very rapid removal of the debris from the nonsensical result of 100 seconds. Ilowever, it cavity, mainly as a liquid film. Oddly enough, the does not appear ai tnis juncture that a constant in time scale for removal of the film depends only in his expression can be derived from smallacale a very weak way on the hole site in the vessel experiments and applied to full scale cavities as (i.e. 'hrough the gas density in the cavity and w as done in the calculation just mentioned; even this matters only as the 0.25 power);
(6) A code called CORDE is under development in (3) BNL has written a one-dimensiorial model called the UK. We have very little information on its OV DCilVih*, in a summary paper presented at the Pittsburgh lleat Transfer conference in 1987, the models, state of development, or predictions. Amendment M 19EA 3-6
ABWR 23A6100As Standard I'lant nx. 4 C g Thus, based solely on the ANL paper, the
- % auumption uwd in this analysis is a debris remosal
!' c folding time of 0.5 seconds, or the linear debris removal assumed over a 2 second period. This salue appeart, to te conservative, but not remarkably so.
Figure 19EA.3 4 cornpares the fraction of debris discharged for the 2 second l'near rate used at this analysis with the 2.5 second e folding tirne from the ANL study. The sensitivity to this anurnption is investigated in Section 19EA.34. a J 4 t l i 1 t V Amendment ?? g9g;4,3 7
AllWR 23 A610nAS Standard Plant aty. 4 19EA.3,5 Application of DCli Model to The sensitivity to this assumptions is insestigated m Subsection 19EA.3.6.3. AllWi{ (3) The initial containment temperature and pressure ! The model requires a variety of inputs which are assumed equal in the w etw cil and drywell. desenbe the geometry of the vessel and containment, J the initial and toundary conditions for the event and a j few model parameters. The steam mass fraction in the drywell is assumed ~ to be 1.0. The sensitivity to this assumption is investigated in Subsecdon 19EA.3 6,6. The geometric information required by the model is: (4) The initial vessel prc ssure is used to calculate the l I (1) The drywell vessel and wetwell gas free volumes source of steam from the vessel to the I containment volume. w hich are used to calculate pressure. The pressure is assumed to te the nominal vcssel The drywell volume used for this analysis is the total for the AllWR opper and lower drywells, pressuie for nonnal operating conditions. Slight f This effectively auurnes that there is a large flow variations in Llus value (such as might result from , a e nsideration of the SRV setpomts) do not have area between upper and lower drywell regions. The a MgnWeant impact on the results. No attempt is T possible impact on the results frorn this taken in das analysis to take credit for partially assumption is considered in Subsection depressurized vessel conditions. , 19 E A.3.6.4. (5) Vessel gas temperature and vessel steam enthalpy, (2) A table of horiiontal vent area as a function of
- distance from the initial water level and the total vent cleanng depth w hen all vents are available. Iloth values are conservatively taken to be constant. The values used are typical for MAAP These are used to calculate the sent clearing thne, analyses of high. pressure core melt scenarios.
O For this analysis, it is assumed that there is no V mitial pressue dif ference tetween the wetwell and MW parameters are: 4 the drywell. Thus, water level in the connecting vents is high, which conservative'y delays the (1) Fraction of Zr to te Oxidized in The DCil Event, time until the vents begin to uncover and gas can
- flow to the wetwell. Of the debris mass that is tving entrained at any instant,20% is assumed to be Zr. This debris is (3) The vessel failure area which is used to calculate assumed to oxidize immediately as it is entrained.
the blowdown from the sessel. Therefore, if one specifies 0.5 as the oxidation fraction, then half of that 20% mass will oxidire; This value is specified for each branch point on 2 moles of H 0 will be replaced by 2 moles of the DETs. H2 in the drywell volume, and the chemical reaction energy will be added to the debris. The initial and toundary conditions are: sensitivity to this parameter is discussed in Subsection 19EA.3.6.5. (1) Debris Mass involved in DCH Event (2) The time for debris entrainment determines the The value of this variable is specified for each interval during which the specified mass of debris w ill te entramed. case on the DETs. (2) Initi.d Debris Temperature Refer to Section 19EA.3.4 for a discussion of this , parameter. The sensitivity to this parameter is investigated in Subsection 19EA.3.6.2. if this temperature is sixcified above 2500 K, then the latent heat of fusion is used in calculating the initial debris energy. If the (3) Time Constant for DCH. temperature is at 2500 K or below, then the latent if set to zero, the debris will be entrained linearly. heat of fusion is not mcluded in the initial debris If set to non zero value, then the debris will O energy. This value was nomina!!y set at 2501 K. entrain at a rate with an e fold value equal to the
. Q Amet.dment " g 9 gg,3, g fL
f A llWR 23 A6100AS , Standard I'lant nix.4 1 l time constant. This analysis assumes the debris is ( entrained linearly. Any sensitisity to this parameter is imunded by the tirne for debns i entrainment sensitisity discuwed in Subsection
; 19 E A.3.6.2.
(4) The time step for the computer code calculations was selected to be one milkwcond. Since the time constant for the DCil event is on the order of a few seconds, there should be no sensitivity to reasonable variations in this l prameter. i
'l The DET methodology addresses the variation in debris mass, initial contamment preuure, and vessel failure area. Section 19EA3.6 provides a discussion of , the importance of the debris temperature, Zr traction, dispersa. . ate, nalabration and initial dryv.cIl steam frac tion.
4 The code calculates the containment response to
! DCll events. The rnost important output of the calculation is the peak containment pressure. The
! results of the model analysis for each branch of the DETs are summarized in the penultimate column of Figures 19EA.2-4,19EA,2 5, and 19EA.2-6. s 1 ]
\
U Amendment 77 19F.A.3 9
I A BM 23swmss Standard Plant sum 4 19E A.3.6 Sensitivity to Various DCil 19 E A ..t6.3 Debris Temperature Parameters The core debris will interact with a variety of As indicated in Section 19EA.2, the DLT structures as it exits the reactor sessel. Thus, it is expected to experience subuantial cooling by those methodology addresses the variation of several key structures on its way to the upper drywell. The ABWR DCH parameters. This section looks at the importance of the debris temperature, amount v Zr oxidued ex. IX'll analysis conservatively assumed that the core vessel, and the dispersal time constant to the overall debris entering the upper drywell was completely mohen at a temperature of 2501 K. A sensitivity case pressurization. These parameters were assumed to be was run assuming 2601 K for the debris temperature constant in the scoping calculations and were judged entering the upper drywell. The results indicate a peak not to have a significant impact on the results. '!he I results of the sensitivity studies confitm that these containment pressure of 133 psia vs. the base case value of 131 psia, I pararr.eters have a second order cifect on the peak con'ainment pressure. 19 E A.3.6.4 Nodalliation i 19 E A.3.6.1 Itase Case The DCil analysis combines the lower and upper
; For the purpose of comparison. the following case drpell compartments into a single control volume represented by node 1. The second node is set up to i
was analyred using the DCil model: represent the wetwell with the suppression pool in the (1) Fraction of core molten at vessel failure - 40% , path connedng node I to node 2. A wnshty caw was run to investigate local pressuritation in the lower
, drywell compartment. To do this, the volume assumed (2) Fraction of material dispersed into the upper in the DCil model was reduced to that of the lower
- drywell- 50%, drywell with the sent area equal to the vent flow area from the lawer to the upper drywell(11.3 m2 ) S nce (3) Dn, persal time constant - 2 seconds, the junction between n(xie 1 and node 2 does not require p (4) Iru.ua. l containment pressure - 1.5 atm.
vent clearing as in the baw case analysis, the vent path (' was assumed already cleared. All of the gas heating was i d(me within the lower drywell compartment. In order to The result of the analysis indicates a peak drywell calculate the correct down stream pressure, zero initial pressure of 131 psia. Referring to the containment steam was assumed in the drywell compartment. This failure curve, this has a failure probability of about is necessary because the model assumes that all steam 0.11 passing from the first node to the second node is condensed, setting the initial steam fraction to zero will 19 E A .3.6.2 Dispersal Time Constant cortcctly account for the increasing upper drywell pressure. A detailed examination of the results of this
- The above case was re run assuming that the sequence indicates the zirconium oxidation reaction is dispersal time constant was ! and 3 seconds, essentially steam limited. Since the zirconium respectively. The results are
- oxidation process consumes most of the steam exiting the vessel, only a small amount of steam will actually Peak Drywell Pressure (psia) enter the second node.
Base Cae 131 The peak pressure computed for the lower drywell was 111 psia vs 131 psia in the base case. This simply Time Const = 1 163 indicates that the lower to-upper drywell vent area is sufficient to preclude any substantial pressure buildup Time Const = 3 110 in the lower drywell region. Section 19EA.3.4 provides the justification for the 19 E A .3.6.5 Zr Osidation Ex. Vessel 2 sec dispersal time and indicmes that it may be somewhat conservative. Ilowever. a 50% change in the The base case assumed that 20% of the core dispersal time resulted in only a 205 change in eh. material that wat dicharged trem the vessel was Zr peak containment pressure. This does not represent a metal. This is based on a uniform distribution of UO2 very significant change. and Zr within the lower plenum of the reactor vessel. Of this material 50% of the Zr that was involved in [s Amendenent ?? 19EA.3 in
ABWR 23 A6100AS Standard Plant uv. A DCil was allowed to oxidisc and contribute to the same analysis w:.s performed auuming that the initial [' drywell heatup. This is a conicrvathe value since, in steam traction was 0.0 (as compar(d to the base case the time frame of interest, only the Zr on the surface of auumption of 1.0L For this second case the peak the particles would oxidite. pressure increased by 33 psid. Since a 50% steam fraction in the containment more accurately represents As a sensitivity calculation, the amount of ex- the actual conditions, the expected increase in peak vessel Zr oxidation was doubled. This chaage in the prenure resulting from the recombination is 25 psid. amount of ex vessel Zr oxidation is equivalent to This is only a 204 (hange in the peak containment assuming that all of the Zr metal in the debris n pressure, and does no' represent a very significant oxidited. Alternately, this auumption is equivalent to effect. assuming the fracuon of core debris exiting the vessel was 40% Zr instead of 209 assumed in the base case. 19 E A.3.6.N Vent Clearing Thus, this sensitivity addresses both any pouible non-uniform core material distribution within the lower After core debris discharge and before RpV head and the potenual for increased oxidation. blowdown, it is expected that the containment will tegm to pressurize even before debris is dispersed into The peak drywell pressure was computed to be 150 the upper drywell. A sensitivity study was run psia vs. the base case valve of 131 psia. This is a very assuming that the vents had already cleared prior to small effect given a factor of two variation in the debris dispersal. The results show that the peak amount of Zr oxidired during the DCll event. containment pressure is reduced by 32 psid compared to 131 psia for the base case. This represents a decrease in 19 E A.3.6.6 Initial 1)rpell Steam l'rntion the peak pressure of about 20%. While this is not very significant, it does provide a measure of the Since steam pauing through the connecting vents conservatism in the analysis. will condense, the amount of welwell pressuritauon during the DCli event is limited. The base analysis assumed a 100% steam environment in the drywell at (~ the start of the event. To investigate the impact of this ( parameter on the peak pressure, a case was run assuming the drywell environment is initially 100% nitrogen. The wetwell pressure will be expected to increase faster for ihn case resulting in a i,igher drywell pressure. This results of the study mdicate that this is true, although the rise in the peak pressure is small, The peak pressure for this scenario is 142 psia, as compared to a peak of 131 psia for the base case. This variation due to initial d.ywell gas composition does not have a significant impact on the results of this study, 19 E A.3.6,7 Ilydrogen Combustion Technical specifications allow the ABWR to be operated with 4% oxygen in the containment. During a DC}l event, the drywell gas temperature may exceed the auto-igniuon limit of approximately 10(K) F. Burning of the hydrogen in the containment with this residual oxygen could result in an increase in energy of the gas. The appropriate reaction energy was added to the existing corium/ gas mixture in order to predict an increase in the peak pressure du: to hydrogen combustion. No credit was taken for the reduction in rnoles w hich would occur as a result of the burn. 7 The peak containment pressure increased by 15 ( psid relative to the base case pressure of 131 psia. The s Amendment ? 19fA 3.Il
{ ABWR 3 3,, , ,,, Standard Plant arv. 4 i i l . 100,000 . . . .i i iii i iisi i iiig i i .. . ) . - Effective Drag Coefficients . for Dense Dispersions ', 10,000 . . l t 1 5 i h 1 1 2 t000 -
~
o . i E 4 20.37
- s . ~
W j O 100 - C =0.80
- o . / .
o - j -
- 4 l
- a -
o -
- w 10 -
pt =0.s0 -
- p . .
- o * '
! C a0.90
.. m w 1 . .
- . Single Sphere in an Infinite Sea .
0.1 ' ' - 1 ' ' ' ' ' ' ' ' ' ' ' ' ' ' 1 10 100 1,000 10,000 100,000 REYNOLDS NUMBER 1 i l Figure 19EA.31 i EFFECTIVE DRAG COEFFICIENT FOR DENSE DISPERSIONS O AmM 19EA.3 12 9 w.ym.mg,,,,,,,--,<w-m ..v ., ., r,,.- .,4.,.
.w,.,-p,.,.
i i
- ABWR 23 i ,
i Standard Plant REV.A } i 1 l 1 l 4 i i i I e a 4 1
-i.0 :.(. J.:.L 1
man ,
"l Gi - .
i l' !)
,,.. d. 5 p% "I ,1::
..3
, ~ j INGTRUMENT N -
- w. f 1..%pgACTtWt CAVITY to l
l l l 1 i l 1 1 +!
- Figure 19EA.3 2 l ZION REACTOR BUILDING lO y--+-- ,w---e.---..w----r- , , *.r. rw.--,-m-,-- .--,- . . , ., - - - - r-- -w- -+,.,-.-.---.--,--.-__m_, _--+-,-+.-s---we-- wme --+w,--_ + ----
ABWR u ,,i m , Standard Plant REV.A O
- i
, s Reactor
, Preuure f i Veuel I
O '~;il,... rn.1,.i,L:
.l ' ' ,'
s- _ 3 RPV Pedestal e ummum Core ummumu
" / y Weir Wall
/
]
y! e :: If k, Sil l h Wey.EF (.:E l lb$6W6'0?Mes'6$N5W69%'6'6?N#6'6W670) [CRD Acceu Manway CRD Openings (4) Figure 19EA.3 3 b SCIIEMATIC O'i GRAND GULF CONTAINMENT O Amendment 71 19EA314
._ . _ _ 4 i
ABWR .
"^ * ^5 Standard Plant nty. A ;
i' 4 1 i i 1 -l> 4 6 ANL
- , e Assumed i
m - e : i a : j t ~ : > aa - - --
- .c : :
o - - e : o : i e S- - .- i L : ! o : - e : :
- o ~
w- : i w . - a o : : !hV i c 3 : o :
-n -
.I o - O - - 1 L "
- u. - -
4 o:,iiiiiiiiliiiiiiiiiiiiiiiiii,1,ii,,,,,,l,,,,,,,,,l,,,,,,,,,} j
- 0 .5 1 1.5 2 2.5 3
' TIME S i i E I t i Figure 19EA.3-4 COMPARISON OF ASSUMED DEBRIS DISCHARGE TO ANL DATA FIT l 4 . O
# D 19EA 315
I J L ABWR meio, 1 Standard Plant uv. a 1 j 19EA.4
SUMMARY
OF RESULTS 1 19EA.4.1 Quantification of .. { Decomposition Event Trees
- The quantified decomposition event trees are shou l in Figures 19EA.21 through 19EA.2 6. The i relationship between the pressure and the cumulative j probability distnbution are shown in Figure 19EA.41.
, Note that the probability distributian functio is (PDFs) are discrete since the discrete probabilities were assigned f in de~elopir's the trees. The PDFs provide a rneasure of
! certainty that the pressure will not esteed a given i value. They are not, however, unceitainty distributions 1 in a statistical sense. Rather, they are based on j knowledge of DCH and engineering judgment wh ch charactenze the ability to accurately characterize the boundary conditions for the problem. From a de:2rministic viewpoint, the best estimate 4 for the peak containment pressure is given by the - l l median Figures 19FA value 1 andof19EA.41, the PDF. As can this inditates thatbetheseen bv comparing , contaimnent would not be expected to fall for any of the initial containment pressures studied. A measure of ! the uncertainty in this study is found by using the
- weighted sum (mean) of the probability of drywell fr.ilure for each of the branches on the DETs. These a weighted values are transfened to the containment event trees for use as the conditional probability of drywell failure for sequer.ces in which the vessel fails at high pressure. The analysis results are summarized below
Conditional Probability of Drywell Failure High RPV Pressure - tow Cont. Pressure 0.005 High RPV Pressure - Inter. Cont. Pressure 0.014 High RPV Pressure - High Cont. Pressure 0.042 Amenenent ?? 19EA 41 1
.- -= __ -__ -. . _ _ _ _ _ _ - - . . . _ _
ABWR 23A6100AS Standard Plant nov. a l l 19E A.4.2 Impact on Containment DC11 containment railure probability of 0.015 is far Failure Probability less than the 0.10 goal for CCFP. This demonstrates a large margin for the ADWR containment design to withstand contamment challenges. An inspection of the ABWR accident classes shows that the conditional probability of having high , RPV pressure at sessel failure is 0.273. Furthermore, i J . the conditional probabiliues of high, intermediate and 1 low pre existing contammerit pressure at vessel failure ! for these high RPV pressure sequctres are: P(Low) = 0.9998, P (Inter.) = 2 x 10 P (liigh) < 104. Combining the above probabilities results in a calculated probabihty of early containment fatture (from direct containment heating) of I x 10 3 conditional on core dmnage. 19EA.4.2.1 Sensitivity of Containment Failure Probability to Assumptions In order to demonstrate the robustness of the containment failure probability to the peak pressures calculated in the deterministic DCll analysis, three additional sensitivity calculations were performed. First, the DET was requantified assuming that the peak containment pressure for each of the low initial containment prer.sure cases was increased by 30 psid. This could represent the possibility of an initial steam fraction of 50% in combination with a hydrogen burn , and no credit for partial clearing of the wetwell connecting vents before the DCH event occurs. The resulting conditional containment failure protability for DCH was increased from I x 10-3 to 7 x 10-3 The secor.d sensitivity case assumed that the containment would be at intermediate pressure for all cases. This represents potential uncertainties in the hydrogen production during the in-vessel portion of the accident. For this case, the containment failure probability due to DCH increases to 3 x 10 3 Although the conditional probability of failure by DCH is a factor of 3 higher in this case than in the base analysis, DCH does not pose a significant threat to the CCFP goal of 0.10. The third sensitivity calculation assumed that the containment would be at the intermediate pressure for all cases and, in addition, that all peak pressures would be increased by 30 psid. The results show thit for this conservative case, the conditional containment failure
. probability for DCH would be about 1.5 x 10-2. Thus, even with these very conservative assumptions, the Amenanent Tl 19EA 4 2
- n. .
1 ABWR 2)A6100A5 Standard Plant _ REY.A 19E A.4.3 Impact on Offsite Dose i The final measure of the impat of uncertatnties in severe accident phenomena is the effest on offsite dose. The CETs r.re quantified using the weighted sum of the containment fatture probabihty as discussed atove. The results of the CETs are then combined with deterministic accident sequence analysis and j consequence enalysis to determine the dose associated with the spectrum of severe uccidents. Ir. order to indicate the possible variation in dose vue to uncertainty in DCH phenomena, other values must be selected for the probabilny of containment failure due to DCH. Since the probabilities used in developing the DETs are themselves the uncertainties in the , phenornena. one cannot determine the classical 5 50-95 confidence limits. However, one can select pressures corresponding to various cumulative certainty of ncin-exceedance (shown in Figure 19EA.41) and compare these values to the containment fragility curve (developed in Attachment 19FA) to estimate the probability of drywell failure with varying degrees of certainty. Sdocting the 50% and 95% values from Figure 19EA.41,one may draw the dose curves shown q in Figure 19EA.4 2. This flgiare shows, that for the 2 Q accident frequencies and cette.inty levels of interest. DCH has no detectable impact on the offsite dose. 4 4 c) Amendment 77 19EA.44
! 5 l ABWR 23A6100As Standard Plant asy.A
- oi i
3 t . U <- 1 U M i O
} , , - - -
- u .
*g i R .
i ?v . , u 0 .95 - a
$e .
i o . Z
- kledian Drywell
! . g l 2 o . Failure Pressure
,s 0.9 - I
{ E
- aw .
l
- 0 -
,y
! Initial Drywell Pressure f
1 M - 1 i m 0.85 - I High E ,
*EL
, J Intermediate
~
~
- Low i
0.8 l l 100 150 200 250 300 i Peak Dr)well Pressure (psia) i j i
- Distributions truncated below 112 psia i Containment not expected to fail i
i 1 I l Figure 19EA.41 CUMULATIVE DISTRIBUTION FOR PEAK PRESSURE DUE TO DCH O s l Amendment i; 19EA.44
, _., .- , . , m.-,
ABWR 23 A6100AS ggy. 4 Standard Plant 0 l
^
5 1 E-06. i g : f cc w - Q. 1E-07: ' h - m 8 1E 08-i -~ O - w Peak Pressw.
~
menv w w eeskprenews l 1E 09 , , , , , , , , , , , , , , , , , , , , , , , , , , 1 10 100 1000 EARLY WHOLE BODY DOSE IN REM l 3 t 1 F1gure 19EA.4 2 l i UNCERTAINTY IN WHOLF, BODY DOSE AT 1/2 MILE DUE TO DCH ' [d r i,tu.s 1 _n - d
ABWR 23 A6100 A5 Standard Plant gtv. 4 19EA.5 CONCLUSIONS The ABWR has a highly reliable depressurtration system which results in a very low probability of a core damage esent which leads to vessel failure at high pressure. Nonetheless, an evaluation of the potential risk of direct containment heating leading to contamri.uit failure in the ABWR has been performed. This stucy indicates that the design of the ABWR is highly resistant to damage as a result of a DCil event. This is due pnmanly to the general configuration of the ABWR lower drywell and connecting vent configuration and area. No modifications to the containment design are suggested as a result of this
.'talyeis.
O x Amendmcat 77 19E A.51
+. ABWR Standard Plant "^7v^f O O 19EA.6 REFERENCES 12. R.E. lienry, Fission ProducI Release During thgh. Pressure Melt Ejection, Task 3.4.6 report,
- 1. U.S. Nuclear Regulatory Commission. Severe Advanced Reactor Severe Accident Program, Accident Risks: As Assessment for Five U.S. Novembet 1988.
Nuclear Power Plants, NUREO.1150, June 1989.
- 13. MAAP 3.0 B Computer Code Manual. EPRI NP.
- 2. Brow n. T.D et. al.. Evaluation of Severe Acciden, 7071-CCML, Volume 2. November 1990.
Risks: Grand Gulf Uniti, NUREGICR.4551, Vol. 6 Rev.1, Part 2, December 1990. 14.1"rederick Moody, Introduction to Unsteady Thermopuid Mechanics,1990,
- 3. BWR leer Head failure Assessment for CSNI Comparison Exercise, EGO EAST 9609, April 15. R.E. Henry, Modificauons for the Development of 1991, the MAAP DOE Code, DOE /ID 10216 Vol. IV, U.S. Depanment of Energy, November 1988.
4 Fauske and Associates, Key Phenomenological Models for Assessing N'on Explosive Steam 16. S. Levy, Debras Dispersal from Reactor Cavity Generation Rates, IDCOR Technical Repon During Lower Temperature Simulant Tests of 14.1B. June 1983. Direct Containment Heating (DCH), Paper presented at if J; Water Reactor Safety Information
- 5. R.E. Henry, personal communication. Meeting, Gaithersburg MD,1990.
- 6. M. Pitch, et. al., Acceleration Induced 17. S.A. Hodge and LJ. Ott, failure Modes of BWR Fragmentation of Liquid Drops, NUREGICR. Reactor Vessel Bottom Head, ORNLIM 1019, 2247 USNRC, August 1981 quoted in J. Sienicki Let'.cr Report, May 10,1989.
and B. Spencet. A Mulupuid and Multiphase Flow and Heat Transferfor the Prediction of Sweepous 18. S.R7 Greene, S.A. Hodge, C.R.11ymnn, M.L. O from a Reactor Cavity, Proc. Fourth Miami Intet. Tobias, The Response of BWR Mark 11 V Symposium on Multi Phase Transport and Particulate Phenomena, Miami Beach, December Containment to Station Blackout Severe Accident Sequences, NUREO/CR 5565.ORNL/TM ll548, 15 17 1986. May 1991.
- 7. S. Kutateladie. Elements ofIhe Hydrodynamics of Gas Liga 'd Systems, Fluid Mechanics Soviet Research,1,4,1972.
- 8. B Spencer, S. Baronowsky, and D. Kilsdonk, Hydrodynamic Swerpout Thresholds in BWR Mark til Reactor Cavity Interactions, ANULWRISAF.
841, Apnl 1984.
- 9. B. Spencer, et. al., Sweepout Thresholds in Reactor Cavity Interactions. AN1dLWRISAF 82 1, April 1982.
- 10. B. Spencer, D Kilsdonk, and J. Sienicki, Hydrodynamics Aspects of Ex Vessel Debris Dispersal in Zion Type Containment Designs, AN1JLWRISAF 831.
II, B. Spencer, D Kilsdonk, J Sienicki, and G.R. Thomas, Phenomenological Investigation of Cavity Interactions Following Postulated Vessel Meltthrough, Proc. Inter. Meeting on Thermal
/ 'i Nuclear Reactor Safety, NUREO/CP-0027,
() Chicago, August 1982. Amen &nent 77 19EA.61
,ABWR umms W Standard Plant REv. A
.s _
x 4 h 19EH,1 INTRODUCTION 19EH,1.1 Probabillly of Pre. flooded a g Lower Drywell 4
~-
t
.> Fuel coolant interactions were addressed in the -
" early assessment for the ABWR response to a severe The configuration of the ABWR containment, i accident. Subsection 19E.2.3.1 examined the shown in Figure 19EB.61, limits the potential for hydrodynamic limitations for steam explosions and water to be in the lower drywell at the time of vessel &
concluded that kre was no potential for a large scale failure. The vessel skirt is solid and there are no active c , steam explosion. The pressurization of the containment injection systems in the lower dr)well. Therefore, the from non explosive steam generation was calculated in only possible sources of water to the containment are the analyses for the accident scenarios. The following the wetwell/drywell connecting vents, the passive secuons examine the available experimental data base flooder and the vesselitself, for its relevance to the ABWR configuration, and provide a simple, scoping calculation to estimate the ne wetwell/drywell connecting vents connect the ability of the ABWR containment to withsumd a large, upper and lower drywell regions to the suppression energetic fuel coolant interaction. pool. The connecting vent is a vertical channel which has a horizontal branch leading to the lower drywell. 4 Challenges of the containment during a severe nerefore,in order for flow from the upper drywell to accident may result from fuel coolant interactions. Both enter the lower bywell, it would have to fall almost the impulse and static loads are considered here. Fuel 9 m down the connecting vents, then turn to enter the Coolant Interactions (FCI) may occur either at the time lower drywell, his is not viewed to be a credible of vessel failure when corium and water fall fmm the scenario, lower plenum of the vessel, or when the lower drywell Gooder opens after vessel failure has occurred. For the water level in the wetwell to rise sufficiently to overflow into the connecting vents, The critical time constants for a steam explosion approximately 2.2E6 kg (4.8E6 lbm) would have to be are considered in 19E.2.3.1. This analysis concludes added to the containment, if the EPGs are followed, this that the critical rates for heat transfer and energy would occur only if injection was being provided from dispersal preclude a large scale steam explosion which an external source in the event that flow from the could damage the contamment. Nonetheless, this study suppression pool was not available. Dis implies that was performed to examine the potential impact of a the only available injection sources are the firewater and large steam explosion on the ABWR. RCIC systems. The 3~C system may be the only system available in e sents minated by station blackout. Several expertments which have provided insights Examination of th cases in 19E.2.2.3 (SBRC to steam explosions are examined, and features of the sequences) and 197 2.2.3 (NSRC sequences) indicates - ABWR are compared to previous plants to indicate the that enough wata can be added by the RCIC system to relative resistance of the ABWR to steam explosions. lead to overflow from the suppression pool to the lower 7 A scoping calculation is also performed to estimate the drywell. If the station blackout continues and the ,. size of steam explosion the ABWR could withstand. firewater addition system is not used to prevent core : damage, vessel failure into a pre. flooded ekvity can ! , Four potential failure modes are considered. The occur in these sequences. The results of the Level 2 ii transmission of a shock wave through water to the analysis, depicted in Figure 19D.5 3 [previously structure may damage the pedestal. Similarly, a shock submitted as CEB92 39 Figure 2] indicate that SBRC ,, wave through the airspace can can cause an impulse sequences with failure of the vessel (no IV) have a load. However, since LM gas is compressible, the shock frequency of 9.7E 11 or 0.06% of all core damage f wave transmiued through the gas will be much smaller sequences. The Class IV ATWS sequences were treated w than that which can be transmiued through the water, very conservatively in the conuinment event trees. All C Therefore, this mechanism is not considered here. of these sequences were presumed to_ lead to core Bird, loading is caused by slugs of water propelled damage with high releases. De frequency of these into containment structures as a result of explosive sequences is 1.66E 10, or 0.11% of all core damage steam generation. Finally, the rapid steam generation sequences.
- may lead to overpressurization of the drywell.
De passive flooder is designed to open when the temperature in the lower drywell airspace reaches
$33 K (500 F). This ternperature is slightly less than the temperature of the steam in the vessel under normal operating conditions. However, any potential break Amendment n 19EB.It
AllWR 23Arilm As Standard Plant R3V. A 110w would cool by flashing as it reaches the hiwer drywell. Therefore, the passne flooder will not open until after vessel failure. A LOCA in the bottom head of the vessel is aho a source of water which could be present in the lower drywell at the time of sessel failure. All of the penetrations in the lower head are small, and any loss of coolant accident through them is classified as a small break LOCA. A censervative estimate of the core camage frequency for events initiated tiy LOCAs in the bottom head is the frequency of all small break LOCAs which lead to core damage lor the AflWR. Examining Table 19D.41, the fraction of all core damage events initiated by a small LOCA is about 0.lfM . 7 The potential for a fuel coolant interaction which
? could threaten the containment may be bounded by
. A summing the frequencies of the sequences with water in g the lower drywell at the time of vessel failure. Threc
? sequences were identified above. The total frequency of g these sequences is 5.lE.10, equivalent to 0.3% of all p3 core darnage sequences. llecause this salue is very a small, it is judged that fuel coolant interactions will not have a sigmficant impact on risk.
LJ l Amendment ?? 19EB 12
AIMR m mo4s Standard Plant Rn A 19 Ell.2 APPLICAIllLITY OF and Benedick (Relerence 5) performed a large se ies of EXPERIMENTS experiments using iron alumina thermite. ne pressure traces for these experiments indicate an explosive preuure puhe of atout 5 msec. A large number of experiments have been performed to better understand FCI. Most of these experiments have been performed at bench scale with The final, intermediate scale test perfonned at simulant materials. Freon Water and Liquid Sandia (Reference 6) used a corium thermite mass to Nitrogen / Water systems are often used. While these simulate the materials which rnight be typical of a severe accident. As m, the Button and Benedick expenments are necessary to understand the underlying expenment, the duration of the pressure pulse m these physics of FCl, they are not directly applicable to the reactor condition. However, there are also sescral expenments was atout 5 rnsec. Three shakedown tests were performed using iron alumina thermite with water experiments performed with metal and oxides which provide insight to the potential for energetic FCI in a i", a crucible. In all of the tests spontaneous, self-tnggered explosions occurred. In contrast, all four of severe accident. the corium tests were externally triggered which resuhed in one mn with a " weak explosion" and one Other expenments, performed for different reasons, with a ,' mild explosion,, Two hypotheses were also yield some insights to FCI. Some experiments proposed to explain these results: performed for debris cootability and core concrete interaction studies added water to the debns. With one notable exception, these experiments did not result in (1) The non condensible gasses generated by an energetic FCl. Finally, one experiment was oxidation stabiliicd the film boiling blanket, performed to examine the impact of a water solid rnaking it less susceptible to triggering; reactor cavity on direct containment heating, in the following section each of these experiments is (2) The UO2and ZrO superheat 2 was only about examined for the insights into FCI and applicability to 300 K. It is possible that the debris froze before the ABWR, the trigger was initiated. This would prevent fine fragmentation of the debris. 19 Ell.2.1 FUEL COOL ANT Hoth these hypotheses have important implications INTERACTION TESTS for application to the severe accidents, presuming a BWRSAR type melt progression, the early pour of A wide variety of experiments have been perfonned debris from the vessel would be metallic. In this case to investigate steam explosions. This section discusses stabilitation of the gas film around the debris could results from selected experiments. Most of the prevent a large mass of molten material from g experiments are prototypic of the reactor condition participating in a steam explosion. On the other hand, o w herem, debris falls into a preyxisting pool of water. the superheat associated with a large oxidic melt is
. The implications of these expenments on the potential typically less than a few hundred degrees. Therefore, it ts- for large, energetic FCI in the ABWR are als is likely that the surface of the debris droplets would discused. frene. This would slow the heat transfer to the coolant investigations into energetic fuel coolant interxtions and steam explosions date back to 1950.
Early experiments, including those by Long (References 1 and 2) and Higgins (Reference 3), identified the requirements for considerable mixing of the molten debris and wates liiggins and Lemmon (Reference 4) noted that the debris must be superheated and that the violence of the explosion increased with the malt temperature. Unfortunately, the triggers used in manj of these experiments were very large. Thus, informaton about the propagation and energetics of these experiments is not applicable to reactor conditions. One of the important parameters in determining the O potential challenge to the containment from a steam VI explosion is the duration of the pressure pulse. Buxton Amendment ?? 19EH 2-1
J AllWR :murs Standard I'lant nix A O Q 19 Ell.2.2 Stratified Splein Experitnents With a in some of the recent experiments performed to examine core concrete interaction, w ater tus leen added to the debris. As discussed m Subsection 19 Ell.l.1, the probability of a large amount of water in the lower drywell at the time of sessel failure is vety small. Af ter core debris is introduced to the lower drywell, it is flooded either by active systems or the panise lower drywell flooding system. Therefore, this is the most probable configuration for a large FCI event in the AllWR. Far fewer experiments have been perfonned in this stratified geometry than in the configuration of debris poured into water. Work by llang and Corradini (Reference 7) used triggered Freon / Water and Liquid ! I Nitrogen / Water systems. In these studies the interaction zone for the vapor explosion is less than 1 em thick. Assuming this depth is representative of reactor material, this would lead to the conclusion that less M of the AllWR core inventory could participate in an FCI event. Protoytpic materials have been used in a few core-concrete interaction experiments in w hich water is added O to molten debris. The MACE and WETCOR tests added water to a pre-existing pool of debris. These tests involved fairly large masses of molten simulant to w hich water was added. Thus, the initial condition is a stratified pool in which water lies over the core debns. The materials and masses of the experiments are summarized in Table 19 Ell.21 No energetic fuel coolant interactions were observed to occur in the stratified configuration. The experiments 13 pically indicated an early heat transfer phase in w hich the heat flutes were on the order of 1.5 to 2 MW/m 2. Later, presumably after the formation of a crust above the mollen debris pool, the heat fluxes decreased. These heat fluxes are considered in Section 19 Ell.6.2 in bounding the non-explosive steam generation rates. p)
\ \/
Amendment M 19Eh 2 2
ABWR = =s Standard Plant uv. 4 19 Ell.2.3 11 ETA V6,1 drywell connecting vents. In faci, the BETA configuration is also much more restrictive than the l Recently, an energetic FCI occurred in the BETA Bibilus reactor it was intended to represent. This fxility. Experiment V6.1 was intended to represent the resnictWe condition resulted in ingression of water into Babilus reactors. These textors have an annular pol of the melt. Since the ABWR configuration has much water around the pedestal cavity. BETA V6.1 was more vent area, water ingression will not occur. designed to determine the impact of these water pools on corium concrete interaction. The configuration of Additionally, there was no water on top of the V6.1 is shown in Figure 19EB.21. The system debris before penetration into the annulus. Thus, the consisted of a concrete crucible with an annular water rnolten debris in V6.1 was highly superheated. This is pot which was vented back to the inner crucible via a contrasted to the situation in the ABWR, The ability to small path. Molten iron alumina thermite was use active systems, such as the firewater addition introduced into the cavity which was then allowed to system and the presence of the passive lower drywell ablate. Hooder virtually ensure that there will be water above the debris in the ABWR. The area of the ABWR lower The debris croded the concrete in the approximate drywell is also very large which enhances cootability, shape shown in Figure 19EB.21. De superheat of the %c uncertainty analysis of Attachment 19EC Indicates melt was very high since there was no water on the there is a low probability that signincant core concrete debris. Eventually, the sideward erosion caused the attack will occur. Therefore, the initial contact mode debris to reach the annular water pool at one local observed in V6.1 is unlikely. point. Instants later an esplosion occurred. The bottom of the crucible was sheared off. There was severe Twen if CCI wcurs and the pedestal is eroded to the damage to the facility. All of the instrumentauan was wetwell drywell connecting vents. The presence of destroyed and the melt injector was thrown sescral water above the debris will cause a crust to form. De meters up, damaging the ceiling, temperature on the lower surface of the crust will te at the melt point of the debris, Within any molten region, Th> energy required to do the damage has not yet the debris temperature will be nearly equal to the melt O been determined. However, the structure surrounding ternperature due to convection in the debris pool. Thus, C the test facility was fairly weak, unprotected sheet metal. Although the doors were blown open they were the addition of any water to the molten pool will cause the debris to freeze and a steam explosion will not not damaged. Therefore,it is believed that the pressure occur. spike may not have been very large. The conditions which led to the explosion at the The symmetry of the damage to the facility BETA facility are not prototypic of the ABWR, Due to indicates that the explosion was very symmetric. There operation of the flooder there is a small likelihood that was very little irregularity in the shearing of the the debris will ablate the side wall and enter the wetwell bottom of the crucibic. Thus, it is difficult to believe drywell connecting vents. This is demonstrated in that the explosion began on one side of the crucible and Section 19EC. Even if the debris does penetrate the propagated sideward. An alternate hypothesis has been pedestal to the connecting vents, the vent area in the proposed (Reference 8). When the debris penetrated to ABWR is sufficient to relieve the steam generation the annular pool, the steam generation rate increased. caused by the initial contac,of water and debris. Thus, Since the annular compartment vents back to the center water would not be forced into the melt as occurred at of the crucible via a small line, the pressure increased BETA. Finally, the superheat of the melt at the BETA and water was forced back into the debris. The debris facility was very high, whereas the superheat of any was still highly superheated at this time. The debris which contacted water in the ABWR would be confinement of the system allowed for intermixing of low. Thus, debris would be casily solidified, reducing the debris and water and prevented the pressure from the heat transfer to the water and preventing rapid steam being relieved. Thus, the damage caused to the system generation. Thus, the explosion in V6.1 does not was not a result of a shock wave, but rather due to indicate that containment damage will occur in the simple pressurization of a confined region. ABWR as a result of FCI. De steam explosion observed in the BETA facility is not applicable to the ABWR system. Although suppression pool and vent system of the ABWR is p located in an annulus around the lower drywell, there is ( adequate vent area to relieve the pressure in the wetwell Amendment 77 19EH 24
ABWR 2mt.s Standard Plant nr.v A (~) 19 Ell.2.4 liigh pressure Melt Ejection V Experiments Sandia performed a series of experiments to examine the influence of water pools on the tchavior of high pressure melts in a Zion like cavity (Reference 9). I 4 Two configurations wcre examined. In the SPIT 15 test debris was injected into a closed acrylic box. This allowed for visualitation of the phenomena. In the i SPIT 17 and lilPS experiments a Zion like cavity was constructed. The basic configuration of the SPIT 17 and filPS experiments is shown in Figure 19EB.2 2. The SPIT 17 cavity was made of aluminum while the litpS experiments used reinforced concrete cavities. In all of the experiments water was present in the cavity at the tirne of melt ejection. The inertia of the water presented venting of the cavity. Thus, the steam generation in the cavity forced the region to pressurire and the structures were destroyed before gas flow from the end of the structure could relieve the pressure in the cavity. It is interesting to compare these ext. .iments to BETA V6.1. In both instances it appears that large pressure spikes wcre created w hen the debris arid water wcre tightly confined. This early confinement keeps the p) (V water and debris in close contact, and seems to lead to the fragmentation of the hot molten material which is a necessary precondition for steam explosions. The results of this experiment are not applicable to the ABWR configuration. The lower drywell is not l initially full of water and there is ample venting of the region. The extreme damage observed in these experiments appear to be consistent with that in BETA V6.1, both in the mode and magnitude of the damage to the facihties. I
\j Amendment ??
19EH 2 4
AIlWR menos Standard Plant REV. A Table 19 Ell.2 1 COltE CONCRETE INTElt ACTION TESTS WITil WATER ADDITION TO DEllitiS ; 4 l a l Esperiment Simulant Debrin Mau (kg) Water Adtlitkm l I MACE MO UO: ZrO: Zr 130 ihkd after attack started 1 1 l ' Flooded after attack MACE MI UO; ZrO: Zr 4(X) started, upper crust was , not fully molten l MACE MIB UO2 ZrO: Zr 4tK) Flooded after attack started, no crust above debris WETCOR Al:0 Ca0 3 34 Water added at I liter /sec O m Amendment ?? 19ER 2 5 L
ABWR 23xeinois Standard Plant REV.A O s
. Annulus sent q_-
.- Initial cavity configuration
// [ '
/
1;ae cavity geometry
/ i
~
- , , , , , ,3, :,:::::::::::::::
n
.: -7 Annular water pool
/// / '
d s Figure 19EB.2 1 BETA V6,1 CONFIGURATION O Amendment 7? 19EB.2 6
- ABWR 23A6100AS i Standard Plant REY. A -
i 3 \ e 1 i ) ! Pressurized Melt i Ejector i r i 4 } i; i I f#X<X6?Xy.:W&W
' J
! lMm:i,$::. w/'W
'iW::>:,:::i$l}f
- ?. f Ih.<f{f'!-{f5f if"'i.:{< j:'.':::W WN.:t 5 :5 f:I5 5 5 f5' l
v? .l <:/.Nj< , a
$$ ...:99;7 , w\',p:iw:yA.;:\ Veer y&:!![:$['>
A;,-):.:p. 5:::5.':: 0.. : 5lIY: y'f}:<k.:$.?q filled g"q.,q e <. . .A,.,'::, :$ n 0 .:.;;::A$%d . . ;. s- : A . ;;;,; j::!f. dS$c.;N;;N., Casity 5::E:i:y:;y:ie'::)\:?h
$'k:/.:f;f::3s $::.g::n' 1/:'4;,;4;.:::\..;:
'::;j; 3.f,;5jf:/;s$::',$: j.J:. j@[My,::$4;dy:$5:
s l :::'y? Y. r 1 l 4 f
. Figure 19EB.2 2 4
i- IIIPS EXPERIMENTAL CONFIGURATION i I e i Amendment 77 19EB.2-7 4 5-I
!4
l ABWR ms Standard Plant _ REV.A I (~N (v; 19EB.3 GENERATION EXPLOSIVE STEA.\1 freeze rapidly after encountering water. Freezing will hinder further droplet divisiou ' ecause more energy will be required to fracture the outer crust than it does to overe me the liquid surface tension. This, in part, This nction presents a bounding analysis of the explains why self triggering can be observed with some matimum steam generation rate which can occur for a , highly superheated metals, but is much less likely with given mass of corium interacting with water. molten core debris. 19EB.3.1 Phenomenology Corium interactions with water can result in rapid steam generation. The rate of steam generation can be limited by the amount of corium or water present. Maximum generation for a given amount of corium occurs when enough water is present to completely quench the corium. Corium mass, surface area, temperature and heat transier coefficient dictate the maximum rate when ample water is available. Two configurations are possible for quenching in the ABWR. First, corium can exit the vessel when the lower drywell contains significant amounts of water. Corium exit from the vessel can be either by a slow pour (small vessel breach) or by a suddcn drop (catastrophic fai'ure of lower vessel head). Second, corium can enter a dry lower drywell and farm a pool Subsequently, tb- lower drywell is finoded with water 3 and the debris is quenched. This situation, commonly l ) refered to as a stratified geometry steam explosion, is v the expected configuration for any large FCI in the ABWR. Molten core debris is expected to be discharged from the vessel close *.o its liquidus temperature, 2600K. Therefore, the maximum temperature in either the pour or stratified geometries will be 2600K. The actual temperature will be lower due to heat loss by the debris prior to interaction with water, in the pour case, corium will transfer heat to the air surrounding the vessel as it falls. Any residual water in the lower drywell, as well as concrete beneath and air above the debris pool will absorb heat in the stratified geometry. For rapid steam generation to occur in either situation, the ejected corium must break ao into small particles. The analysis presented in 19E.2.3.1.4 demonstnted that corium breakup in the ABWR will tw driven by Taylor instabilities. The smallest part cles formed will be approximately 2.5 tr.m based on the Taylor critical wavelength. Debris breakup in the 3 stratified geometry will also be governed by Taylor instabilities. Crust formation will hinder debris breakup. Since corium is expected to exit the vessel near its liquidus O temperature, any heat loss should contribute to crust Q formation. Furthermore, the outer debris surface will Amendment ?? 19EB3 1 [
ABWR ums Standard Plant REV.A r 19 Ell.3.2 Bounding Analysis transfer coefficient that can be expected is that of ( enhanced film boiling, which is 390 W/m2K. The Moody, et al., (Reference 10) determined the emissivity suggested for use in MAAP (Reference 11) maximum steam generation rate during FCI based on a for corium is 0.85. This value will be used for this simphfied thermal hydraulic methodology. The steam analysis, formation rate from a single corium droplet assuming heat tmnsfer to saturated water is: If a mass of corium, Me, imeracts with water and l breaks up into droplets of average radius, r. the number f droplets, N, will be given by: H Aa(T,, - T ) .ut* m, = e (1) fs 4 M N - nr 3 = (3) where: n,; = steam formation rate, H = heat transfer coefficient. The total steam generation rate of N corium A4 = surface area of a corium droplet, droplets is: T,, = droplet surface temperature, s = Nsg=s y e " W T., = saturadon temperature of water where the maximum generation rate is: at the ambient pressure,
= latent heat of vaporizadon for * "
hrs rn ,,,, = (5) water, EcbI fg
\
tg i = time from beginning of This is the maximum steam Generation rate that , interacdon, can occur for a given amount of corium broken up into small droplets in a large body of saturated water, t3 = thermal response Ome. Heat transfer from the droplet to the surroundmg is domir,ated by convection and radiation. The heat transfer coefficient is: H = H, + H, c (2)
= H, T+, -((Ti-Ti)
T., ) t wtere: H, = convective heat transfer coefficient, H, = radiative heat transfer coefficient, o = Stefan-Boltzmann constant, E = cmissivity of the droplet. Due to the high temperature of corium, convective f'N heat transfer from the surface of the particle wiil be in film boiling regime, The maximum convective heat < Amendment ?' I9ED 12
ABWR meious l Standard Plant REV.A O 19EB.4 IMPULSE LOADS The starting radius for bubble growth can be j estimated by a spherical volume equal to the corium l Rapid steam generation can produce a shock wave volume plus the total volume of water it vaporizes ! which in equation form is: ' whid imparts impulse loads to containment structures, . Energetic FCis, however unlikely, may occur :n the lower drywell of the ABWR. Water in the lower 4 3 M M,c,(T,, - T_ )
+ (7) drywell, which must be present nor rapid steam } nR, a generation, can transmit shock waves from the site of
' '8 '
FCI to the walls of the pedestal. Shock waves which , pass into the gas space above the water will be rapidly where: p, = density of corium, damped due to gas compressibility and will not represer.1 any threat to containment integrity, If the c, = specific heat of corium. impulse load is large enough, the pedestal will fail 4 sausing the vessel to tip. Tipping of the vessel would The maximum pressure predicted by Equation (6) , most likely lead to tearing of the containment s shown in Figure 19EB.41 for participating corium penetrations. The scoping analysis presented m this masses from 0 to 30,000 Kg. The required corium , section estimates the amount of corium which can properties were taken from Table 19E.2-17. The steam participate in a FCI without exceeding the impulse louJ and water properties are saturated conditiens at two capability of the pedestal. atmospheres. Two atmospheres is a likely containment 19EB.4.1 Maximum Impulse Pressure The peak pressure during impulse loading of the Moody, et. al., (Reference 10) determined the ASWR pedestal resulting from fuel coolar.t interactions maximum pressure increase at the site of an FCI based should be bounded by the pressure shown in Figure on the steam generation rate given in Equation ($). llis 19EB.41. The pressure predicted by Equation (6) is analysis applied the Rayleigh bubble equation to a conservative because of the assumptions which went single steam bubble with an equivalent volume of the into its creation, Furthermore, this is the pressure at O V many bubbles formed during interaction with N corium droplets of radius, r. Because the volume varies as r3 , the site of FCI, The pressure experienced by the pedestal wall will be reduced because the shock wave this results in overestimation of the rate of bubble has to pass through some amount of water before it expansion. The bubble expansion rate dictates the impinges on the wall. The pressure will decay as 2r as pressure rise. Therefore, this analysis bounds the it moves away from the source (Reference l2). pressure generated by the maximum steam generation during FCI. The maximum pressure increase of a single submerged steam bubble above the ambient pressure during its formation at the generation rate given in Equation (5) is: a (R,Trn,y)2ln AP nm = 0.178 pi 4 (6) O where: p3 = density of saturated water as the ambient pressure, R, = Universal gas constant for steam, R. = starting radius for steam bubble growth. f')
'N Amendmem 7? 19EB A. I
ABWR 23 A6100AS Standard Plant wtv. A ! l 19EllA.2 Impulse Duration
)
The main difference between energetic fuel coolant interactions (steam explosions) and non energetic interactions is the time in which the energy stored in l 4 the corium is transferred to the coolant. Short transfer times, on the order of milliseconds, indicate explosive reactions. Longer times are indicative of non energetic interactions. Several fuel coolant interaction experiments involving corium simulates were reviewed in Section 19EB.2.2. Pulse widths were observed to be of the order 5 ms or less for FCI. 1 J hp) U T 1 rm
$ \ \,.__,/
Amendment ?! 19Ed.4 2
ABWR m ooxs Standard Plant REY.A O Pedestal Capability E. = Young's Modulus of the 19EB 4.3 i ) pedestal. Detailed calculations of the capability of the ABWR pedests! to withstand impulse loading have not Since the pedestal is a composite structure, the been performed. However, a simple clastic-plastic determination of each of these parameters can be qui.e calculation can provide a capability which can be used complicated. A conservative estimate of the resistance for scoping analysis. This estimate can be compared to to deformation and the natural period can be obtained by the maximum pressure expected during a FCI for a using the following parameters: given amount of participating corium snd the impulse duration. The pedestal in Grand Culf (MARK 111 oy = 175 MPa (value for thc A441 containment) was analyzed in NUREG-1150 (Reference steel plates which define the
- 13) with regards to its ability to withstand pressure boundaries of the pedestal),
spikes generated by steam explosions. Since the ABWR pedestal is expected to be at least as strong as that of a MARK !!!, the impulse capability of the a* = 6 cm (total thickness of the tw A44I steel plates which Grand Gulf pedestal can also be used for comparison. define the boundaries of the pedestal, ignores steel webs and 19 EB.4.3.1 Elastic Plastic Calculation concrete fill), A failure limit estimate based on a simple clastic-R. = 6.15 m (average radius of the plastic calculation has been performed by Corradini (Reference 12). The assumptions made in this analysis pedestal), are: p, = 2.400 Kg/m3 (density of (1) The pedestal wall is thin compared to its diameter, concrete fillbetween steel plates), (2) The pressure loading is uniform both spatially and [m ) temporally, E, = 200 GPa (typical value of steel). 4 (3) Failure is based on a strain criteria of p (failure strain / yield strain) equal to 10 Using these parameters yields: Rm = 1.7 MPa and T = 4.2 ms. (4) The pedestal wall is considered to be free standing. The maximum response of elastic plastic one-The resistance to deformation, Rm, of the pedestal degree sys' ems (undamped) due to rectangular load is: pulses is shown in Figure 19EB.4-2. The ratio of pulse duration,I d. to natural period is the horizontal axis. The o,a, strain criteria, , forms the vertical axis. The (8) R, = g* relatic,nship between these two axis parameters is given by a series of curves defined by the ratio of resistance to deformation, Rm, to the average pressure of an impulse, where: c y = yield stress of the pedestal wall, Fi , The amplitude of the square pulse can be conservatively estimated by the maximum pressure rise a, = thickness of the pedestal wall, expected during a FCI, APmn, which is calculated in Section 19EB.4.1. R. = radius of curvature of the wall. As discussed previously, the impulse duration of a The natural period of the pedestal, T, can be FCI is expected to be approximately 5 ms, see Section calculated from: 19EB.2.1. The ratio of ta/T for this duration is 1.2. Using this ratio and a strain criteria of 10 yields a Rm/F of approximately 1.0. This implies that the p,R-2 pedestal can withstand a APmo of 13 MPa. T=2n 9) E, f% where: p, = wall density, Amendmem 77 19EH A3
ABWR 23 A6100AS l 1 Stancard Plant nty. A : The maximum ratio of Rm/Fi in Figure 10EBA-2 (nL} is 2.0. Using this ratio, thJ maximum pressure rise the pedestal can withstand is estimated to be 0.35 MPa. The uncertainty in pulse duration (assumed to be 5 ms) is irrelevant for the maximum ratio of Rm/F because it is obtained for pulse durations much greater than the natura! period of the pedestal.
'Ihis simple clastic plastic calculation predicts that the pedestal can withstand a maximum pressure during a fuel coolant interaction of 0.85 MPa. The amount of corium which must participate in a FCI to achieve this pressure can be obtained from the analysis presen'.ed in Section 19ED 4.1 and summarized in Figure 19EB.41.
j The amount is 22,400 Kg. The ABWR contains 235,000 Kg of corium.Therefore, the ABWR pedestal can withstand a FCI involving 9,5% of the corium mventory. 19 EB.4.3.2 Comparison to NUREG llSO Grand Gulf Pedestal The ability of the Grand Gulf pedestal to withstand steam explosions was considered in NUREG ll50 (Reference 13). The smallest impulse load expected to fail the pedestal was reported to be 3.5 psi sec (0.024 MPa sec). This limit can be used for comparison to the .\gO ABWR because the ABWR pedestal is expected to be sturdier than that of a MARK Ill. For a pulse duration of 5 milliseconds, this impulse corresponds to a square j wave pressure of 4.8 MPa. This value is significantly
- higher than the pressure predicted by the clastic plastic
= scoping analysis. Alternatively, the pressure predicted by the clastic-plastic analysis (0.85 MPa) can be applied for 28 milliseconds before an impulse load of 3.5 psi-sec is exceeded. Both of these comparisons imply that the clastic-plastic analysis bounds the 3 impulse load required to fail the pedestal. 5
/"%
q) Ametuimcm ?? 19 Ell.4-4
ABWR 23 Abl00AS Standard Plant nty. 4 e) i 19EIL4.4 Capability of the AllWR to Withstand Pressure Impulse i i The ABWR pedestal has been shown in this scoping analysis to be capable of withstanding a peak pressure of 0.85 MPa during a steam explosion. The amount of coriu.n required to produce this pressure impulse during a fuel coolant iriteraction was shown to be 22.400 kg. This represents 9.5% of the ABWR corium inventory. This is more than three times the maximum amount of debris which could participate in an FCI event based on the observations discussed in Section 19EB.2.2. Therefore, the ABWR pedestal is very resistant to the impulse loading which could oc'ur in a severe accident. This failure mechanism need not be considered further in the containment event trees or the uncertainty analysis. 4
\
( 1
'%J l
)
LJ Amendment ?? 19EU 4 5
i j i ABWR meiwis i Standard Plant REV.A 4 4 4 q 1 i i i 1 1 1 i j-t 1 s,
- 1.0 - -
1
- 0.9 --
1 l 4 70.0-- c., 4 20.7-- l' lo o6 -- i j b e 0.5 -- j iw- o.4 --
.g
- j. g 0.3 -
e o
$ 0.2 - -
1 0.1 --
- o.0 ; ; ; ; ; i 1
0 5 to 15- 20 25 30 I Mass of Corium Participating (Mg) i l t-t i i i~ Figure 19EB.41 PEAK IMPULSE PRESSURE FROM FCI ' J - J k Amendment 77 t 9ER.4-6 i ___2___a__-__ --- - - - , , m .- .,, _ . , - - . . ,
4 l ABWR uni As Standard Plant REV.A I r"' ( J l 0 '0 10 40 100 ,'
'l ; j j
_ 1 / / /
$ _[h eb o/ /
h/ /
/
97 i 5)
/
f / / / 1.02
- / > i > / s - - ---
I / / / // io l l
! , y
,6 -
1.05 a / / i ! ff
~
x ( / / / u - 1.10 . u / / / / y 5 4 / / / // f psy 1.20 4 ' e (/ /& l i,3 o -~
/ /f/k% 1.40 i'o ~' "
s "%',5S , Ela'stic) 2.00 = rem w , ; ;
//L'Aff Fgt) g
., r/yj , y -
~~~~~
rg , , Fan - V '
/ l l
~
l Itn - 1 I t y t-Rectangular Resistance Displacement - ) 0.1 i i i i . .'" , I i !""C[*" i i , , , '""CiiO" , 0.1 1.0 io -4o 1 Figure 19EB.4-2 - MAXIMUM RESPONSE OF ELASTIC-PLASTIC ONE-DEGREE SYSTEMS (UNDAMPED) DUE TO RECTANGULAR LOAD PULSES
- (Reference 14)
?
b i Amendment ?? 19EB.4 7 __ - _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ JL
ABWR meinoss Standard Plant REV.A n WATER MISSILES O' i 19EB.5 g = acceleration of gravity. Submerged steam formation resulting from fuel Maximum missile rise heights are presented in coolant interactions can be rapid enough to propel an Figure 19EB.5-1 for participating corium masses of 0 oveoving liquid mass. Impact loads can be imparted to to 30,000 kg. contamment structures if the liquid mass (water missile) is ejected from the water pool with a great enough velocity. Although a prediction of impact by a water missile does not imply damage, additional analysis would be needed to assess the structural response. The maximum height to which a water missile can rise will be determined in this section for a given amount of participating corium. The rise height will be compared to the distance between the expected wcter surface of a pre flooded lower drywell and the bottom of the reactor vessel to determine if damage to the containnient could occur. No other structures are considered because damage to them will not lead to containment failure. 19EH.5.1 Maximum Rise lleight Moody, et. al., (Reference 10) used the steam generation rate determined in Section 19EB.3.2 to predict the upward propulsion velocity and elevation characteristic of a water missile. The maximum
,_s velocity that a water missile can obtain is the / \ maximum radial expansion rate of the steam bubble 'V formed during FCI. This expansion rate is:
R ., = 3
' 5 R ,T., m ,,,, *
(10) 5,2 4npiR,;, i where: R ., = equihbrium steam bubble radius. It is equal to: It) 4 M,c,(To -T_) R= (11) 3n h r,p, , where: p, = vapor density. Balancing the kinetic and potential energies of a water missile yields: Aym,, =; IV- (12)
'g where: Aym , = maximum rise height a missile n will rise above the water l i surfxe, V
Amendment ?? 19EB.5-I
ABWR 23A6100AS Standard Plant any 3 C' 19EB.5.2 Available Rise lleight The water level in the lower drywell wil! not be greater than suppression pool water levet during . s severe accident, The normal water level of the suppression pool is 6.05 meters below the bottom of the reactor vessel. Consequently, a water missile can rise approximately six meters before encountering any I structure the damage of which could lead to I containment failure. o
\
G q C./ Amendmem 77 19EB 5 2
ABWR 23 A6 A$ Standard Plant A n 19EB.5.3 Capability of ABWR to 4 () Withstand Water Missiles The amount of corium which can participate in a FCI in the ABWR and not generate a pressure impulse which is expected to fail the containment is 22.4 Mg. This amount of corium will produce a wamt missile which will rice 1.75 meters, see Figure 19EB.51. This , rise height is significantly lower than the available rise height of 6 meters. Therefore, the pedestal will fail from impulse loading before the required amount of corium participates to elevate a water missile even to the bottom of the reactor vessel For this reason, water missiles are not expected to play a role in determining if the ABWR containment fails due to fuel coolant interactions. l 1 r , O 4 I i
/
Amendment M 19EB5 1
3 ABWR 23A6100AS l - Standard Plant - nty. A 1 k
~
i I 1 1 1 i. 4 l 2.0 -- 1.8 -- i i 1.6 -- ! $ 1,4 -- A 1.2 -- 1.l2 e 1.o -- M l- o o.e --
- .m
! % o.6 - r i o.4 -- 0.2 -- l ' 0.0 ; ; ; ; ; ; i 0 5 10 15 20 25- 30' l Mass of Corium Participating (Mg) L i i-f 4 l i-Figure 19EB.5-1 3 RISE IIEIGHT OF WATER MISSILE 1 + lO ! l Amendment 77 . 19EB.5 4 1
-- . . _ _ , . . _ . - ~ _ _..-.___-._._....._.__.__.JU
ABWR ms Standard Plant REV.A ) 19 Ell,6 CONTAINMENT time of vessel failure (Section 19EB.I.1). Nonetheless, l
. d OVERPRESSURI7. ATION an evaluation will be performed assuming that corium falls into a pre-existing pool of water and is quenched The final element of this study focuses on the innantaneously. This will provide a limit on the peak 1ressurization of the containment which may occur c niainment pressure which could result from during periods of rapid steam generation which may quenching of debris as it falls ir3to the lower drywell, occur when corium is being quenched. In the nighly For the AB% R, the vast majority of sequences with unlikely event of an ABWR core melt which leads to % fadure m' cur at w pmuum. Therefom, gradty vessel failure, the corium will fall into the lower is the driving force for the flow of corium from the lower head of the vessel to the lower drywell. Both drywell. There are ten connecting vents which join the MELCOR and MAAP predict that the vessel fails at lower drywell, the upper drywell and the wetwell, as the penetrations for low pressure melts. After the initial shown m Figure 19EB.6-l. The pressure suppression hole is fonned, the hole ablates due to the flow of hot contamment prevents large increases in containment corium. In order to determine the sensitivity of the pressure by sparging the steam througn the connecting ABWR containmera to rapid steam generation 40% of vents to the suppression pool w,uch condenses the steam. However,if the pressure rise is extremely rapid, the total UO2 mass is assumed to be molten at the time of vessel failure. This value is consistent with the the vents may not be able to clear before the containment is damaged. At even higher steam upper limit for molten debris used in the uncertainty d analyses for direct containment heating. , generauon rates, the area from the lower drywell to the ; upper drywell could be too small and a pressure t
difference between the crywell regions could occur, Two Potential limits for Eressurization due to failing the lower drywell. This analysis determines the steam generation are considered. First, the pressurization of the lower drywell is determined steam generation rates for different limits on FCI. The maximum rate is then compared to the containment considering the hmit of the vent area from the lower pressure capability to assess the poten:ial for drywell to the upper drywell. This determines any containment damage as a result of overpressure during limits f r the assumption that the upper and lower an FCI event. drywell regions have good communication and will fm respond similarly to the pressurization. Second, the (v ) 19 Ell,6,1 Methodology resp nse f the pressure suppression system is evaluated Drywel: pressurization rates are used to 3 determine the vent clearing response which is in turn This calculation compares the pressurization due to used to determine the peak containment pressure as a rapid quenching of corium to the pressure capability of function of the pressurization rate. the containment. Two non-explosive steam generation l limits are considered. If there is a sufficiently large water mass, then the quenching of corium will provide the steam generation limit. If the mass of water limits i the steam spike then the steam generation will be less than, or equal to, the water flow into the lower drywell. The impulse pressure limited mass, calculated in Section 19EB3.1, is also considered. If there is no water in the lower drywell at the time af vessel failure, then the maximum rate of steam generation at some later point in time is the rate at which water is introduced into the lower drywell. If there is still water in the lower plenum at the time of vessel failure, as predicted by MAAP, then this source of water could react with the corium in the lower drywell. Water addition could also cecut via the passive flooder, the use of the firewater addition system or by means of ECCS recovery. Each of these possibilities will be examined to determine the maximum rate at which water could be added to the lower drywell. For most of the core melt sequences in the ABWR
] PRA there will not be water in the lower drywell at the Amendment ?? 19 Ell.o
l ABWR 23 A6100 As Standard Plant REV.A /3 ! ) 19EB.6.2 Maximum Steam Generation with fusible material at the lower drywell end which Rates opens when it reaches a specified temperature. Tnis is l ! shown schematically in Figure 19E8.6-2. The first step in determining the peak pressures The Dow from the wetwell into the lower drywell that may result from fuel coolant interactions is to . determine the maximum steam generation rates. The is driven by the difference m, the water height, h, betw een the connectir.g vcuts and the flooder De flow steam generation can be limied either by the available rate is givcn by: water or the available corium. Both of these possibilities will be considered separately. m = pay 2gh (13) 19 ED.6.2.1 Water Adried to Debris w here: m = water mass tiow into the lower There are four potential se urces of water addition to drywell(kg/s), the lower drywell. First, in a MAAP. type core melt progressioti, there may be water in the lower plenum at p = denoity of water (kg/m3), the time of vessel failure. Aber the corium f alls into the lower drywell, the water will follo v through the A = total arca of passive llooders ablated hole in the lower plenum. Second, the lower g3)* drywell passive flooder opens when its fusible material melts. Water from the wetwellds then driven by gravity = acceleration of gravity (9.81 into the lower drpell. Bird, the firewater system may be used to add water to either the vessel or the upper "I5 )' drywell. In either case, water w ill eventuaIly Gow into , the lower drywell at the f, m ewater injection rate. Finally,if the ECCS is recovered, these systems could be used w inject water into the sessel w hich again will The maximum now through the passive flooder now into the lower drywell. would occur when the pressure difference between the (O wetwell and the drywell was sufficier.t to open the 19 E D.6.2.1.1 Water Insentory from Lower vacuum breakers, and the suppression pool is cold. Plenum AS'ummg a suppression pool temperature of 30 C, p = 996 kg/m 3The total area of the passive floodeis if there is water in the lower plenum at the 6me of is A = 0.081 m 2. Assuming that the pool is at the vessel failure, then it will fall into the lower drywell hi@ water level, the height of water above the passive afte the coriun . Under these conditions, the flow will flooder is h= 4.753, which yields a maximum flow be driven by grsvity through the ablated vessel failure. rate of m= 780 kg/s. The flow rate will typically be The expected failure mode for a BWR is penetration less than this maximum because the DW press is failure (Reference 15). A parametric study was greater than WW and the first row of vents are clear. performed to determine the final, ablated area resulting from different numben of CRD penetrations. The study 19 Ell.6.2.1.3 ECCS and Firewater Flow was conducted by varying the numtier of vessel penetrations presumed to open at the time of vessel The ECCS and firewater system are both capable failure. Since this affects the initial area of the vessel of adding water to the vessel which would flow into the failure, multiple penetration failures have higher initial lower drywell. The firewater system does not rdy cn debris pour rates. As seen in Figure 19EB.6-3, the final AC power, so it is available even during a station area varied from 0.06 m2 for 10 pencuations failed to blackout event. De ECCS is dependant on AC power; 0.08 m2for one penetration. The final area is smaller and, thus, will not be available during station blackout for cases with multiple penetration openings because but could inject water during recovery late in a severe the duration of the debris pour is shorter. In order tu accident. The ECCS system has a flor rate far greater bound the flow of water into the lower plenum, a value than the firewater system. Therefore, no determination of 0.1 m2 s used which results in a maximum mass of the firewater flow is necessary. The maximum flow rate of 1020 kg/s. ECCS tlow will be bounded by the runout now of the ECCS pumps. The actual flow will be somewhat 19EB.6.2.1.2 Paive Flooder Flow smaller due to the flow losses at higher velocities when all of the pumps are operating simultaneously. (3 The passive flooder is composed of ten pipes () connecting the lower drywell to the suppression pool Amendment ?' 19EB b-2
. ABWR meinas Standard Plant IGV. A A I There are two HPCF systems, each with a runout 19EB.6-4. The maximum rate of debris ejection from (d flow ol 3800 gpm (230 kg/s), and three LPFL the vessel is about 6000 kg/sec, Assuming this systems with flow of 4200 gpm (265 kg/s). The material quenches as it is ejected, the steam generation RCIC system is not considered since the vessel will be rate is about 2800 kg/sec. depressurized.The total water addition rate to the lower drywell is 1250 kg/sec. The experimental heat flux observet when molten 4 core debris simulants are poured into water is on the 119EB.6,2.2 Steam Generation Rate for order of 1.5 ta 2.0 MW/m2 based on the Door area. Pre nooded Lower Drywell Using the upper bound on the experimental observations, the maimum steam generation rate for For the ABWR, it is very unlikely that there is the ABWR is 80 kg/sec. This is far below the value water in the lower drywell at the time of vessel failure, determined above for the instantaneous quenching of Thus, steam generation is usually limited by the debris for a bounding debns pour rate, availability of water, However, there may be sequences for which there is ample water, and the limitation on 19 E D.6.2.3 Explosise Steam Generation the steam generation rate is the energy of the quenching Rates corium. Thus, it is prudent to determine the maximum steam generation from this limit if there were a large Based on the examination of the impulse loading water supply available. A large mass of water is calculation of 19EB.4 3.1, the ABWR can withstand assumed to be present in the lower drywell for this the shock wave w hich corresponds to 22.4E3 kg of core l portion of the analysis. debris. The maximum steam generation rate associated with this amount of debris is 4100 kg/sec (see Section A ioc number of analyses have been performed to 19EB.3.2). determine the mode of vessel failure. While there are still some uncertainties in the details of the analysis, 19 E H.6.2.4 Maximum Steam Generation the work performed to date provides overwhelming indication that a BWR vessel fails at the penetrauons The maximum steam generation rates for each of O (References 16 and 17). Once there is some flow N inechanisms described above are summarized in through a penetration, the molten material will begin Table 19EB.6-1. Based on these results, the limiting to ablate the hole. Since there is little change in the scenario is the maximum steam explosion frcro the driving force for the flow of molten material, the scoping study. Therefore, even though this event is far maximum flow rate will occur when the hole size is larger than the expected steam generation rate, the maximized as the mass is exhausted. containment pressurization will be estimated using this value. In some MELCOR type analyses, the coriem quenches in the lower plenum of the vessel. It subsequently heats up and causes vessel failure. Therefore, there is little corium molten at the time of vessel failure. The flow rate of corium from the vessel is limited by the rate at which the corium melts in the vessel. Conversely, using a MAAP type analysis, the corium does not quench in the lower plenum. Thus, there is a large molten mass at the time of vessel failure. Since this will result in larger flow thes than the MELCOR-type model, the MAAP results will be used to determine the corium now rate for this analysis. MAAP (as well as MELCOR) uses the Pilch model for the ablation of the penetration (Reference 11). The velocity of the corium through the vessel failure is approximately constant; therefore, the ablation rate of the failure is linear. A series of MAAP runs were performed which examined the flow rate of molten debris and vessel failure area as a function of the number of failed penetrations. The results of these (n) N) calculations are shown in Figures 19EB.6-3 and Amendment ?? 19EB 6-3
ABWR u m ooAs Standard Plant Rtv. A O U 19EIL6.3 Containment Pressuritation 19 E R.6.3.3 IIoriiontal vent Flow The containment peak pressures may be calculated After the vents have cleared, steam will begin to l based on the flow rates determined above. The results 00w from the drywell to the suppression pool. The , given below are for the most restrictive pressurization drywell pressure during this time is equal to the rate. Three limits are considered. The first condition is wetwell pressure plus the flow and water heads. Using the flow rate of steam Irom the lower drywell to the conservative assumptions and the maximum steam upper drywell. Second, the time period before the Dow rate, the drywell wetwell pres;ure difference is suppression pool vents open must be considered. found to be 0.16 MPa (23 psid). Finally the quasi steady condition of Gow from the drywell to the wetwell through the suppress!an pool is ] considered. ! 19 EB.6.3.1 Drywell Connecting Vent Flow Consideration of the flow through the drywell/wetwell connecting vents is important to ensure that there is adequate vent area to allow the 4 upper and lower drywells to communicate freely. If the now is restricted a significant pressure difference could exist between the upper and lower drywell regions. This could potentially result in lower drywell region failure, even though this region has a much higher ultimate strength than the drywell head (see 19F.3.1). Using the maximum steam generation rate and an effective area of about 11.25 m2 in the drywell/wetwell connecting vents, the pressure difference between the upper and lower drywell regions is less than 0.15 MPa (21 prid). ] U 19 ER,6,3.2 Vent Clearing if the drywell pressure is higher than the wetwell pressure at the time of the FCI. then steam 00w to the
; wetwell can begin immediately, However, if the vents are not open, the pressure must accelerate the water in the vents to allow steam flow. During this interval the pressure in the drywell will rise quickly. Smce the pressure difference between the upper. and lower-drywell regions is small, the entire drywell volume may be considered when calculating the pressure rise during this periad.
Assuming that the initial drywell and wetwell are at equal pressures maximizes the time for vent clearing. The time to vent clearing is calculated based on analysis by Moody (Reference 18).This model requires the pressurization rate for the drywell. A constant ramp rate is determined by assuming a steam generation rate and using the ideal gas relationship for steam. The pressure rise in the drywell due to steam generation is then calculated using the pressurization rate and the time to vent clear:ng. Using the maximum steam now rate, a pressure rise of 0.26 MPa (38 psid) is Calculated.
%./
Amendmem 7? t9EB 6-4 4
ABWR DMINAS Standard Plant REV. A gi i 19EB.6.4 Summary of Overpressurization Limits Based on the calculations presented above, the maximum pressure rise in the lower drywell due to fuel coolant interactions occurs just before the wetwell/drywell connecting vents c!;.tr. At this time a pressure spike in the lower drywell of 0.41 MPa (59 psi) may occur. FCI events of the magnitude considered here occur when there is a large mass of unquenched debris which comes into sudden contact with water, in the ABWR this only occurs early in the course of a severe accident when the wetwc!! pressure is well below the COPS setpoint, typically at about 30 psia (0.2 MPa). Even if the wetwell pressure were near the COPS setpoint of 90 psig (0.72 MPa), the lower drywell would be below its estimated ultimate capability of 180 psig. Therefore, FCI leading to l overpressurization failure of the lower drywell is not a credtble event. Concerning the upper drywell region, a conservative calculation based on the maximum steam generation rate given in Table 19EB.6-1 indicates that the maximum pressure in the upper drywell is the wetwell pressure plus 38 psi. Again, considering that FCI events of the magnitude considered here occur hh when there is a large mass of unquenched debris which comes into sudden contact with water, the drywell will he well below even the service level pressure (97 psig). l Therefore, one would not expect upper drywell failure as a result of FCI. De only FCI event one could hypothesize to occur late in the accident is the recovery of ECCS just before containment failure. However,in the ABWR design the passive flooder ensures that there is water above the debris. The addition of ECCS water will not cause increased heat transfer from the molten debris. Therefore, FCI leading to containment failure late in a severe accident has been ruled out by design. The rapid steam generation rates which can occur due to bounding fuel coolant interactions do not lead to failure of the containment structure or opening of the rupture disk in the ABWR. Therefore, no further consideration of steam generation rates is required. [ v Amendment ?? ;9EB 6-5
ABWR m ims Standard Plant REY.A Table 19Elt.61 h1AXIh1Uh1 STEAh! GENERATION FOR STEAN1 SPIKES Water I imited Cases Flow from lower plenum at the time of vessel failure 1020 kg/s Passive flocxter 780 kg/s Recovered ECCS 1250 kg/s Debris 1 imited Case Debris falling into casity is quenched instantaneously 2800 kg/s Experimentally limit for debris poured into water 80 kg/s F.soloshe Steam Generation Scoping result for shock wave capability 4100 kg/s O O Amendment M 19ER 6 6
l i i ABWR m ims ] Standard Plant REV,A lO i i i El20750
$ Ih b> ;
f e ii= 10303 DIA-- as
/
; AgL-4 9000 4'{'17
;t +1 Jieds ^
~
! L I I maam T,"
-. n p.a ..
v j PERSONNEL ( R 13850 l ( J ,$ ti g 3 90 M
- u. g u.00 - .rm. _ _ _. ,,
E 11360 -
=.
61 - W iih '""
- E E 8000 --
~
@ E 7350 0 W:bakbEGl % J ' ~2K"_ s='is ?!d'MjL8190 3 11Co .
e.t700 m p El-l1400 fgM (=l48 4 y , I ; I . AETURN 1 Hwtt.I8050
/ murH
b Lwt:-isiso s
y e===lotoo DIA - ' Eg.)3480
=
l-l110 o o / ] - at-iitsoo i Ei g siso
- asooo r2 =
access Tuman g alt-i siao l l l . l F(gure 19EB.6-1 l ABWR CONTAINMENT CONFIGURATION l O Amendment ?? 19EB.6-7
I ABWR moi As Standard Plant REV.A I s 9 11 jj Wetwell Lower $j Drywell jj d q r j.; g h L .,
~
lilillilbumme i l Figure 19EB.6 2 PRESSURE IIEAD FOR LOWER DRYWELL FLOODER FLOW i 1.
. 'N l
Amendment 77 19EB.6-8 ?
ABWR 23A6100AS Standard Plant REY,A O tue _ . i . g . . . . g i i i . g i i i i g i i . i i . i i g i i i i g i i i . g i i i i g i ,. j y use :- Nuriber of penetrations: 1 One ,
/
~
o 2 Two g sw - u' --- 3 Four > d i 4 S*m / _
- t w -on -
5 Eight / l u
- 6 Ten /-- -
> mm f
O
'15 E
h 1- --el- - (j 1
- un-on -- --
g .
~
sma . .
~
,,,,t,,,,I,,,,!,,,,1,,,,I...,t.... ,,,,!,,,,I,,,.'
sum umm 4xe 4.am 4mm 4mm Tune (sec) Figure 19EB.6 3 ABLATED RADIUS OF VESSEL FAILURE O Amendment ?? 19EB.6-9
. __----_-_____=_-_--_-_-______-_-____-______.___________-_--___---_---.____1L
ABWR moi,x, 3 Standard Plant REV.A _~ 2 - 3
=
b s s, t mm ... .i....i....;i.iii. .ii....i.i .i....i....i..... I
~
- i
' uma -- Number of penetrations: -c .- -
- 1 One :
- 2 Two N :
M* j 7 3 Four ' Y* 7 i
- 4 Six I l' -
3 h .am
~
_- 5 Eight / ,
/l#i
_' 3 jj 6 Ten i
- ~
$ == :5
// i , I I '/ ,*< j 'I s
s - / g,i s, '
- r :
=
MW
'E ""
l --
/
Q -
/ '
.'r *'i ,1 .. - e
- / / ,i ,f I
// , , ; y unz e -
/ t -
- '/ , ' ,.
i I i
=
I t i e !t t t t !t i t t !t t t t !I t i t !t t t t ! e ta e *t t t t it t i!matht" " tem 4.m.e 4.me 4. mees (m.e an c Time (sec) J a s 9-1 1 _4 s Figure 19EB.6-4 ' MASS FLOW OF CORE DEBRIS THROUGH VESSEL FAILURE - i j f 2 Amendment 77 19EB.6 10
g - ABWR mm, Standard Plant REV.A O
& ' 19EB.7 REFERENCES 13. Severe Accident Rists: An Assessmentfor rive U.S. Nuclear Power Plants NUREG.-l150 Appendin C.
- 1. G. Long, Explosions of Aluminum and Water, Aluminum Company of America, New Kensington, Pennsylvania, ALCOA Report 2 14. US Army Corps of Engineers, Design of-33, August 1950. Structures to Resist the Effects of Atomic Weacons, Manual 1110-345 415,1957
- 2. G. Long, Explosions of Molten Aluminum and Water, Metal Progress, Volume 71, p.107, May 15. J. L. Rempe, Light Water Reactor Lower // cad 1957. Failure Analysis (Draft), NUREGICR.5642, December 1990,
- 3. H.M. Higgins, The Reac: ion of Molten Uranium and Zirconium Alloys with Water Acrobt Report 16. J. L. Rempe, BWR Lower llead Failure 2914 2, Azusa, Californie, April 1955. Assessmentfor CSNI Comparison Exercise, EGO.
EAST 9609, April 1991. 4 A.W. Lemmm, Explosions of Molten Aluminum and Water, Licht Metnis,1980, (C. Minn, Ed.), P. 17 S. A. Hodge and L. J. Ott, Failure Modes of the 817 Proceedings of Technical Sessions Sponsored 19BWR Reactor Vessel Bottom ficad, ORNL/M. by TMS Light Metals Committee at the 190th 1019, May 1989. AIME Annual Meeting.
- 18. Frederick Moody, Introduction to Unsteadv
- 5. L.D. Buxton and W.B. Benedick, Steam Explosion Hermoquid Mechanics.1990.
Ef]iciency Studies, Sandia National Laboratory, S AND/791399, NUREG/CR 0947, November 1979, p 6. L.D. Buxton, W.B. Benedick and M.L. Corradini, Steam Explosion Eficiency Studies: Part 11 Corium Experiments, NUREGICR.1746, S AND/801324, Sandia National Laboratory, October 1980.
- 7. K.H. Bang and M.L. Corradini, Vapor Explosions.
in a Stratsfied Geometry, Nuclear Science and Engineering, Volume 108, Number 1, May 1991.
- 8. M.L Corradini, personal communication, June 24, .
1992.
- 9. W.W. Tatbell. et. al., Pressurized Melt Ejection into Water Pools, NUREG/CR 3916, SAND 84-1531, Sandia National Labor:, tories, March 1991.
- 10. F J. Moody, R. Muralidharan, S.S. Dua, Assessment of Ex-Vessel Steam Pressure Spikes in BWR MARK 11 Containments,17th Water Reactor Safety Information Meeting, NUREG/CP.
0104, Oct,1989. I1. MAAP 3.0 B Computer Code Manual. EPRI NP-7071-CCML, Volume 2, November 1990.
- 12. M.L. Corradini, et. al., Ex Vessel Steam p Explosions in the MARK II Containment, NUREG 1079 Appendis C December 1985.
Amendment ?? 19EB.7 1 __ ___ __ tJ
ABWR 23A As Standard Plant A n
- 19EC DEBRIS COOLABILITY and water and liquid nitrogen and Frean 11, Crust
'4
" AND CORE CONCRETE formation was observed at low gas velocities but f und t become unstable at high sparging rates.
O r INTERACTION 1t was obsetved that as the pas velocity increased b y Appendix 19E of the ABWR PRA discusses core to a magnitude typical of core. concrete uteraction, the heat transfer rate increased by a concrete interaction. In particular, in Section factor of ten. De heat transfer rates were found to 19E.2.1.3.6, it is stated that the core debris will be approach those associated with entical heat flux. quenched preventmg subsnntial concrete ablation due to opration of the passive flooder. Even if the flooder was Greene 1988(Reference 3) assumed to fail, water from the suppression pool would flood the lower drywell after 8 inches of radial ablation Tesu were run with liquid metals with water and had occurred. His conclusion was based on available Freon Rll. Gases were injected in the melt. It experimental information and the work performed in observed that the water / melt interactions were IDCOR Subtask 15.2 (Reference 1). generally unstable and that the upward heat trans er crease w gas ve ty. W # cal Since the original ABWR PRA was submitted there has been contmued research in the areas of debris upwad heat transfu rates wenJound to W 6 tim ' 8 ' " coolability and core concrete interaction. Recent _ 9rrelauon. expenments performed at Argonne as part of the MACE program have indicated that, due to crust formation, debns coolmg may be limited. His section FRAO(Reference 4) willinvestigate the uncertainties associated with debris coolability in the lower drywell of the ABWR. The nis series of tests perfonned at Sandia National , investigation will begin with a look at applicable Laboratories used 3 mm diameter steel spheres expenmental data. Next, the issue of debris coolability heated and placed in a 20 cm diameter concrete will be decomposed into the controlling parameters and crucible. Tests were performed both with and without water addition. Both limestone and O followed by the development of a decomposition event tree (DET). After creation of the DET, deterministic evaluations will be made to quantify the end points of basaltic concrete types were investigated. The - limestone tests showed that a stable crust made of concrete and steel formed that kept the water from
- the tree. Finally, sensitivities to key assumptions will be investigated. penetrating the rest of the debris bed. The basaltic concrete allowed for some water penetration. he 19EC,1 c nelusi n fmm these tests was that core concrete APPLICABILITY OF attack continued even to the presence of water and EXPERIMENTS TO ABWR that a substantial amount of steel oxidation took Several experiments have been carried out to investigate the influence of an overlying water pool on SWISS (Reference 5) debns coolability. The critical parameter that appears to dominate the behavior in several of the expenments is These tests, also performed at SNL, involved the the formation of a stable crust. This crust is found m interaction of molten steel on limestone concrete.
prevent substantial water ingression and, therefore, The steel was heated at approximately five times debris cooling. The major criticism of these the expected reactor decay heat levels. There experiments is that, due to their small scale, a stable appeared to be no violent melt water interactions crust is preferentially formed. This limitation makes it and the melt did not quench. There was a stable quite difficult to extrapolate the results to a large reactor crust that was found to attach to the MgO cavitv The MACE tests at Argonne have attempted to sidewall. Typical upward heat flux was 800 address this weakness by investigating larger cavity kW/m 2. There was also information from the designs, experiment that the overlying water pool provided The following provides a brief summary of several debris coolability experiments- Mark 1 Shell Failure Experiments (R:fere ice 6) neofanous and Saito - 1980 (Reference 2) Several expeiiments were carried out Fauske and
-O Experiments were performed with liquid nitrogen Associates to investigate the influence of water on debris coolability and specifically to observe AmdmL M 19EC 1 1
I ! ABWR 23A6100AS Standard Plant REY.A l l discharged onto a concrete slab pre flooded with lower drywell of the ABWR. Some insights can, O water The initial heat transfer was found to be however, be extracted. The following shows the quite high (20 times CHF) and leseled off at observed upward heat Oux for three of the tests. 2 about 800 kW/m later. l SWISS - 800 kW/m2 MACE (Reference 7) Mark I Shell Test - 800 kW/m2 l A series of large scale experiments are being performed at Argonne Netional Laboratory MACE Scoping - 600 kW/m2 investigating the coolability of molten-corium by water durtn; its interaction with concrete. The One of the major reasons why these tests are not MACE program has attempted to prototypic is that, due to their small scale, they promote a stable crust formation. The larger scale (1) employ prototypic corium melt materials, MACE tests should generate some useful insights. l (2) employ prototypic concrete types, (3) obtain realistic melt temperatures, l
- (4) obtain realistic MCCI initial conditions, (5) include prototypic chemical and internal heating, and by the increased size, (6) ensure applicability to reactor cavities, in the scoping test, a high initial heat removal b) was observed. De crust that was formed was found to be supported by the electrodes. There '-
were periodic melt eruptions through the crust that lead to substantial melt quenching. However, the melt did not completely cool and continued to erode concrete. One of the major difficulties with the test was that there were larger than prototypic heating rates. The next test, M1, was performed on November . 25,1991. The major difficulty with this test was that not all of the material melted initially and the sintered region on the top kept the water from 3 a penetrating the melt. Low melt water heat transfer 3 rates were observed. Concrete attack continaed l with the debris not cooled. His sintered crust ! 7 configuration is not prototypical of the ABWR. e U The most recent test, M1B, corrected the problems encountered with MI, The melt l temperature was observed to decrease steadily to near the concrete liquidus temperature after the water was introduced. Concrete ablation was found to continue but at a reduced rate (a few mm/hr). The post. test cxamination showed that there were large holes in the top surface. (' The experiments described above are insufficient to enable a full understanding of debris coolabil.ty in the Amendment 'n 19 EC1-2
ABWR s Standard Plant .
"#E^4 19EC.2 DESCRIPTION OF EVENT large amounts of debris (40%).
TREE ANALYSIS For the case of a small debris mass in the lower-RPV, it is likely that either 19EC.2.1 Debris Coolability
. (1) Vessel failure occurred fairly quickly after core A decomposition event tree (DET), shown in slump into the lower plenum (MAAP type failure Figure 19EC.21, was developed to assess the likelihood of debris coalability. This section describes model)' or that the branch points and the quantification of this DET.
(2) The debris in the lower plenum was initially 19EC.2.1.1 Fraction of Debris in Lower quenched by residual water in the lower plenum l a faHure cc er aner the water Drywell Early (COR DW E) - - was boiled away and the debris started to reheat
.. (BWRS AR type failure model).
This event assesses the initial debris mass which relocates to the lower drywell soon after vessel failure. For these situations it is judged likely that the debris The amount of debris which enters the lower drywell temperature will be at, or near, its melting point, early is dependent on the amount of debris molten in Hence, the following probabilities were assigned for the lower RPV head at the time of RPV failure and on s case.. the amount of entrainmetat of the debris from the lower drywell. However, for simplicity, debris entrainment t the upper drywell was conservatively neglected in this Case 1 Small Debris Mass in Lower analysis. For consistency with the DCH analysis, two yweH Early regimes are considered for the fraction of the core inventory which is molten in the RPV at the time of low (Superheat) 0.9 RPV failure (see Subsection 19EA.2.1.4). These regimes are: High(Superheat) 0.1 Low 0 20% (nominal 10%) 0.9 For the case of a large amount of molten debris it could be expected that this resulted from a delayed High 20 40% (nominal 40%) 0.1 failure of the RPV allowing more debris to flow into the lower plenum (MAAP model) or for melting and heating of quenched debris already relocated to the lower 19EC.2.1.2 Amount of Initial Debris Superheat (SUPERHEAT) plenum (BWRSAR model). For both situations the extended time to vessel failure could result in higher m Iten debris temperatures at RPV failure. It is unclear This event is used represent the initial debris what the actual debris temperature would be for this temperature when the debris first contacts the lower drywell floor. It is also used as a surrogate u) represent a m, es M M q MN m ead branch to represent this ILrge uncertainty, the additional metal / water reaction heat production associated with a high metal to oxide ratio in the debris. Superheated debris or debris with a high metal Case 2 Large Debris Mass in Lower content is expected to be more difficult to quench WyweH Early initially and to experience faster initial concrete erosion. In the deterministic CCI analysis discussed in Low (Superneat) 0,5 section 19EC.1 the low superheat cases are represented by (molten) debris at the U-Zr O eutectic melting High(Superheat) 0.5 temperature (approximately 2500 K). High superheat was taken to be temperatures in the range 300-500 K 19 EC.2.1.3 Debris Quenched Early above the melting temperature. This was represented in (QUENCH _E) . the deterministic analysis by increasing the amount of steel added to the melt prior to vessel breach. The probability that long term debris cooling will be established is greatly increased if the initial debris Two cases were considered in the DET analysis, pour is quenched soon after being expelled from the The first case represents sequences with a small amount vessel. Initial quenching of the debris implies either (10% of core inventory) of molten debris in the lower that the debris has been fragmented to sizes which plenum at vessel rupture and the second case represents allow cooling, or if the debris is a continuous " pool" V that it is sufficiently shallow to allow cooling by Amadmat ?? 19EC.21
ABWR _s Standard Plant nix.A conduction through the layer of solid debns. taken for early quenching if a significant amount of debris enters the cavity before lower drywell floodmg The ABWR design makes it extremely unlikely occurs, the order of this question and the late cavity that water will be in the lower drywell pnor to RPV flooding question (CAVWAT_L) question is not failure. hiost of the core damage probability is the important for the BWRSAR cu. a l initiated from a transient. This type of sequence would not result in water in the lower drywell at the time of Four cases were defined in the DET. These case vessel failure. Only a LOCA in the recctor drain line are: would result in water entenng the drywell. All other LOCAs blow down into the upper drywell (which Case 1 Small Debris hiass and Low drains directly to the suppression pool). Hence, water Superheat which enters the lower drywell coincident with the expelled debris must come from residual RPV For this case approximately 24000 kg of molten inventory or from in-vessel injection systems w hich are debris are released from the RPV at vessel failure. Since the debris has s low superheat and the debris depth is l operaung htAAP type melt at (or are initiated progression the lower at) RPVisfailure. plenum For a (< 5 cm) it is highly likely that the very shallow nearly full of water at the time of vessel failure. Thus, debns would be irutially quenched. 70.000 kg of water is available to quench the debns. Quench 0.99 in addition, water may enter the lower drywell at the time of vessel failure via the passive flooder. If No Quench 0.01 water from the vessel does not enter the cavity, the debris will rapidly heat the lower drywell, and the Case 2 Small Debris hiass and liigh flooder will open quickly. For a BWRS AR type melt Superheat progression model, there will not be water in the lower plenum at the time of vessel failure. In this case the As for case I approximately 24000 kg of molten lower drywell will heat up quickly and the passive debris are released from the RPV at vessel failure. In G Gooder will open. A calculation was performed with a modified version of hiAAP-ABWR which simulates the BWRSAR melt progression model, described in this case the debris has a high s'.perheat and although the debris depth is very snallow (< 5 cm) it is somewhat less likely that the debris would be initially subsection 19EC.6 Case LATE indicates that the quenched by residual RPV coolant inventory for this l Gooder will open about 30 minutes after vessel failure case than for case 1. for this case. Thus, it is very likely that water will be available to quench the initial debris expelled from the Quench 0.95 vessel. The major parameters judged to impact the probabtlity of ininal debris quenching are Case 3 Large Debris hiass and Low (1) the mass et debns in the lower drywell following RPV failure, For this case approximately 94000 kg of mo4:n debris are released from the RPV at vessel failure. The (2) the availability of water in the lower drywell, and debris depth in this case would be relatively shallow (< 15 cm). Since the debris pool is relatively shallow (3) the initial temperature of the debris, and the debris superheat is low it is judged that it is likely to be initially quenched by residual RPV coolant The mass of debris retained in the lower drywell is inventory, determined in a preceding event. The initial debris temperature is also determined in a prior event. The Quench 0.75 source of water depends on the the presumed core melt progression model as described above. In a h1AAP type No Quench 0.25 melt progression the initial availability of water is assured. For a BWRS AR model the water comes either Case 4 Large Debris Stan and Low from injection systems which begin to inject at vessel Superbeat G failure or from the operation of the flooder which is considered in the next node. Since no credit will be As for case 3 approximately 94000 kg of molten Amendmmt ?? 19EC : l
t ABWR 23A6100A5 ' l Standard Plant nty. A 1
, debris are released from the RPV at vessel failure and Two cases are considered in the quantification of
'- the event. The timing of residual RPV debris entry into l the (< 15 debris cm). However, depth the debris would superheat beisrelatively high and. shallow the lower drywell is considered to be sensitive to the i it is judged to be indeterminate whether or not the extent of the accident progression in vessel at the time i debris will be quenched by residual RPV coolant of vessel failure. For the case of a small amount of i inventory, rnotten debris in the lower RPV plenum at RPV failure (Event 1 in this DET)it is inferred that RPV failure
- - Quench 0.50 has occurred relatively "carly" in the in vessel accident progression process. Conversely, for a large amount of 0.50 molten debris in the lower RPV plenum at RPV failure No Quench it is more likely that the in-vessel accident progression i 19EC.LIA Water Enters Casily Late is further advanced at the time of RPV failure.
Consequently,it would be expected that for the case of ! (CAVWAT L) - small initial debris pours the timing between vessel failure and later debris pours would be delayed relative I This parameter is used to represent the longer term to the case of large initial debris pours. Based on addition of water to the lower drywell. The lower insights from ABWR specific MAAP analyses and l drywell water addition systems which are considered are the diesel driven firewater system, any vessel injection from a review of BWRS AR calculations for other BWR which is available late in the accident and the passive sequences the following branch probabilities were
" flooder. Initi:. tion of the firewater addition system is the estimated.
y
- most likely means of late water addition to the lower l E dr Case 1 Small Debris Mass in Lower pr.ywell.
ssive If flooder the firewater system will begin system to inject is waternot when Drywell started, the Early h d the fusible valves, located at the ends of the pipes near the drywell floor, melt. The fusible valves on the - After Late lnjech 0.9 passive flooder system are assumed to open when the lower drywell gas temperature reaches 500 F (533 K). BeforelateInjee. 0.1 O (,) Assuming a BWRSAR melt progression model the fusible valves on the passive flooder system would Case 2 Large Debris Mass in Lower open in approximately 30 minutes. For a MAAP type Drywell Early melt progression model the water in the lower dryweU is first boiled off. The debris then begins to heat up, if After Late Injection - 0.5 the debris is quenched during the early boil off phase the debris must reheat resulting in approximately 2 Befam Lateinjection 0.5 hours to flooder actuation. If the debris was not quenched early, the flooder opens about 30 minutes 19 EC.2.1.6 . Heat Transfer Rate to Overlying l after the debris bed dries out. This event is a sorting Water (HT- UPWARD) type event, quantified (either 0 or 1) based on prior , branch decisions in the CET' This event assesses the longer term steady state heat transfer rate which characterizes upwards heat 19EC.2.1.5 Time Remainlag Core Debris transfer from the debris. 'Ihne tegimes are considered Falls Into Cavity (COREDROP) (1) heat transfer limited by hydrodynamics in an Thb cvent assesses the timing of the entry of the overlying water pool (CHF limit), remaining debris into the lower drywell relative to the timing of the addition of water (i.e from the passive (2) heat transfer limited by f'im boiling to an flooder or firewater system). If the majority of the overlying water pool,and i debris is held up in the vessel until aft::r water addition begins, then debris cooling is substantially more likely (3) heat transfer limited by conduction through a than if the bulk of the RPV debris enters the lower debris crust on the upper debris surface. drywell prior to water addition. MAAP calculations indicate that the residual RPV debris will melt and fall Nominal values of the heat transfer rate used in the into the lower drywell very slowly after vessel failure, deterministic CCI model to characterize these three heat This behavior is also typical of BWRSAR type transfer regimes are 900, 300 and 100 kW/m2 , calculations (Reference 8). O y i 19EC.2 3 Amen &nent ?? [ I 4
- - . _- _,. _. , ,,,...,_m,,,,o,,,v_...._y- _,y,
i l ABWR 2mias
- Standard Plant REY.A i
The conduction limit represents conditions where a Case 2 Small Debrls Mass in Lower crust forms on the surface of the debris and water - Drywell Early. Debris Initially Quenched and cannot penetrate into the debris bed. The use of a Residual Core Debris Enters Lower Drywell 100 kW/m: heat flux is believed to be very After Flooding conservative, if the debris is not quenched and core i~ concrete interaction occurs, the upper crust will thin to This case is considered to represent nearly as ! a condition where the upward and downward heat fluxes favorable a set of conditions for establishment of a are nearly equal. This will lead to a heat flux much particulated debris bed as was Case 1. In contrast to - higher than 100 kW/m2. Therefore, this value will lead Case I however, the initial phase of the interaction is i to very aggressive core concrete interaction. characterized by only a small amount of debris which is quenched in the lower drywell. Hence, a larger amount The hydrodynamic limit represents cases where of debris enters the lower drywell after RPV failure than j water can penetrate into the debris bed allowing a much for Case 1, Prior to the entry of the residual RPV ! greater effective debns/ coolant heat transfer area. Under debris the lower drywell is flooded resulting in the ! these conditions the heat transfer rate is limited by the residual debris pouring into a pool of water which is
- ability of the water to penetrate the debris bed. The use likely to lead to fragmentation, quenching and the
- l of 900 kW/m21s much lower than the typical heat establishment of a particle bed. Consequently, as for l fluxes observed in the experiments pemmed to date. Case I a relatively high probability is assigned under these conditions to an upwards heat flux characteristic The film boiling regime is selected to represent an of a particle bed with water ingression.
intermediate heat transfer rate where, for example, the crust is unstable allowing water to penetrate the debris CHF Limit 0.9 bed in a limited fashion. The early phase of the
- l experiments indicate a heat flux well in excess of 300 Film Boiling 0.09 kW/m 2before the fonaation of a crust.
Conduction Limit 0.01 . Four cases were identified for quandf; cation. These l cases are described below. Case 3 No initial Debris Quench and Residual Core Debris Enters Lower Drywell Case 1 Large Debris Mass in Lower After Flooding Drywell Early, Debris initially Quenched and Residual Core Debris Enters Lower Drywell This case is considered less favorable for , Afler Flooding establishment of a particulated debris bed which would be conducive to water ingression and coolability. The This case is considered the most favorable set of initial phase of the interaction is characterized by failure conditions for establishment of a particulated debris bed to quench the debris soon after RPV failure. However, , which would be conducive to water ingression and prior to the entry of the residual RPV debts the lower l cootability. The initial phase of the interaction is drywell is flooded resulting in the r:sidual debris characterized by large amount of debris which is pouring into a pool of water which is likely to lead to initially quenched in the lower drywell. Prior to the fragmentation of this debris. However, since the initial entry of the residual RPV debris the lower drywell is - debris pour was not quenched, long term establishment flooded resulting in the residual debris pouring into a of a cootable particulated debris bed is somewhat' pool of water which is likely to lead to fragmentation, uncertain. Consequently, a lower probability us been quenching and the establishment of a particle bed. assigned for the most favorable debris bed configuration Consequently, a high probability is assigned under compared with Cases 1 and 2. these conditions to an upwards heat tiux characteristic of a particle bed with water ingression. CHF Limit 0.5 CHF Limit 0.95 Film Boiling 0.4 Film Boiling 0.045 Conduction Limit 0.1 Conduction Limit 0.005 Case 4 Residual Core Debris Enters Lower Drywell Prior to Flooding This is considered the least favorable set of Amedmm 7? 19EC24
l
- ABWR 2mims i Standard Plant REY.A conditions for establishment of a pouculated debris bed Case 3 Lower DryweH Flooded,
- which would be conducive to water ingression and Upward Heat Transfer Limited by Film Boiling
- cootability. For this case the bulk of the residual core i debris enters the lower drywell prior to lower drywell For cases where the lower drywell is flooded-
! Gooding. This could lead to formation of a molten pool MAAP analysis and supplemental hand calculations j undergoing concrete attack. Later water addition,instead indicate that if the upward heat transfer is in the range
- of particulating the debris may lead to crust formation of about 300 kW/m 2then the debris bed should be
} limiting the ability of water to penetrate into the coolable. $mce this case represents a range of upward . debris, heat transfer egimes (200-400 kW/m2) and the lower 1 part of this range may not in all cases be coolable the CHF Limit 0.1 following probabilities were assigned. , A Film Boiling 0.6 No CCI 0.75 j Conduction Limit 0.3 Wet CCI 0.25 i i 19EC.2.1.7 Core Debris Concrete Attack Dry CCI 0.0 1 (CCD I Case 4 Lower Drywell Flooded. This event characterizes the nature of the debris Upward Heat Transfer Limited by Conductics i concrete attack. ~Ihree branches are considered. The No ! CCI branch represents cases where the little or no For cases wiiere the lower drywell is flooded ! debris concrete attack woutd be expected. Wet CCI MAAP analysis and supplemental hand calculations j represents cases where CCI occurs in the presence of an indicate that if the upward heat transfer is below about
- overlying water pool and Dry CCI is for cases where 200 kW/m2then the debris bed will not be coolable, j the lower drywell was not flooded.
j No CCI 0.0
, Case 1 Lower Drywell Not Flooded
- Wet CCI 1.0 -
t The Dry CCI case occurs for all sequences where j both active injection and the passive flooder fail to Dry CCI 0.0 j supply water to the lower drywell after vessel failure, j Under these conditions Dry CCI is assured. A ! No CCI 0.0 . Wet CCI 0.0 i l Dry CCI 1.0 t 1 Case 2 Lower Drywell Flooded, i Upward Heat Transfer Limited by CHF j For cases where the lower drywell is flooded MAAP analysis and supplemental hand calculations l ! indicate that if the upward heat transfer is above about 300-400 kW/m2then the debris bed will be coolable, i
- No CCI 1.0 Wet CCI 0.0 l
Dry CCI 0.0 a i Amendment 77 19EC.2 5 i i
. . - , - _ , - -,n , , - - - - - . - ,
ABWR 23i.6toors Standard Plant REV.A the inner surface of the connecting vents. It is V 19EC.2.2 Pedestal Resistance to CCI considered quite likely that this will result in water Ris section describes the decomposition event-tree ingression and flocding of the lower drywell. (DET) analysis used to assess the probability of pedestal failure as a result of radial core concrete (CCI) This event is only significant for dry CCI attack in the lower drywell after reactor vessel failure. sequences where the lower drywell is not initially The DET is shown in Figure 19EC.2 2. Pedestal wall ficoded by either active injection or the passive floder. failure is considered to be sensitive to De probabilities are assigned based on judgement. (1) the nature of the CCI (i.e. w hether wet or dry), SP Ingression 0.95 (2) w hether the debris spreads from lower drywell into No SP Ingression 0.05 the suppression pool following radial penevation through the pedestal wall to the wetwell/drywell 19 EC.2.2.3 Debris Flows From Lower connecting vents, and Drywell to Suppression Pool after Downcomer Penetration (WW DEB) _ (3) the extent of radial crosion compared to downward erosion. His event assesses whether a significant amount of the molten debris will flow from the lower drywell The lower drywell will bi flooded in most cases as a into the s'ippression pool following penetration of the result of either active injection systems such as the wetwell/drywell connecting vents. After 25 cm of radial firewater addition system or via passive injection erosion the ablation front will reach the inner surface of through the lower drywell flooder, the downcomers. The floor of the lower drywell is above the bottom of the connecting vents, which, in tum, are above the floor of the wetwell. Thus, once the
' 9 E C.2.2.1 Core Concrete Attack (CCI) downcomers are breached, a flowpath exists from the lower drywell into the suppression pool. Flow of a p)
( - This event characterizes the nature of core concrete attack. nree bmnches are considered. significant portion of the molten debris into the suppression pool will increase the debris surface area in contact with water and decrcase the debris depth in the No CCI lower drywell. Although there is a great deal of uncertainty in this behavior, it is considered fairly Wet CCI likely that the debris will flow into the suppression E Dry CCI The No CCI branch represents cases where there is little or no concrete attack. Wet CCI represents cases 0.3 No WW Debris where CCI occurs in the presence of an overlying water pool. Dry CCI is for cases where the lower drywell was Ratio of Radial to Axial Erosion 19 EC.2.2.4 not flooded. The rate of CCI is higher for cases with dry CCI. (RAD EROS) Given that CCI is occurring, this event assesses This event is a sorting type event which assigns a the ratio of the radial concrete erosion to the downward probabihty of 0 or 1 depending on the branch taken in erosion. Three branches are considered 1/5,1/3 and 1/l. the previous CET event. CCI experiments have generally demonstrated significantly more downward concrete penetranon than 19EC.2.2.2 Suppression Pool Water Floods
'"d!*I penetration. It is hypothesized that radial erosion Lower Drywell after Downcomer Penetration .
is hmited because the concrete decomposition gasses (SP INGRESS)
- establish a gas film between the debris pool and the concrete walls. This gas film acts to insulate the Th.is event assesses tf suppression pool water will concrete sidewalls, and to convect debris heat upwards.
flood into the lower drywell after the crosion front This limits the heat transfer to, and ablation of the n, aches the wetwell/drywell connecting vents. The concrete sidewalls. Conversely, the gas film at the (' vents are imbedded in the pedestal. If 25 cm of the bonom surface of the pool would be unstable due to the pedestal concrete is croded, the ablation front will reach heavier overlying debris pool. The density difference Amendment n 19 EC.2-6
( ABWR 23A6100AS Standard Plant REV.A b V would cause the lower gas film to collapse, allowing contact of the debris with the concrete. This difference Case 1: Debris Flows into Suppression Pool after Downcomer Penetration in gas film behavior would limit the sideward heat transfer compared to the downward heat transfer. This car stesents sequences where a substantial amount of the cmc debt;s relocates into the supp*ession in the BETA series of debris concrete experiments pool after downcomer penetration. This is represented conducted at the KfK research center in Germany, by deterministic calculations FMX100, FMXCSP and downward crosion rates exceeded sideward erosion rates NFlood. The calculations indicate that the increase in by a factor from 3 to greater than 5. For example, in the pool surface area results in either a coolable debris the high power CCI experiment BETA VI.8, the configuration, or greatly reduced radial erosion rates. downward erosion was measured to be approximately Consequently, the likelihood of sufficient radial 40 cm and the sideward erosion was only about 2 cm penetration to fail the pedestal in this case is considered (1/20 sideward to downward erosion ratio). For the low to be remote. power experiment V6,1, the downward erosion was 35 cm and the sideward erosion w as 10 cm (1/3.5 rauo). No Pedestal Failure 1.0 Based on the CCI experiments, and the generally Pedestal Failure 0.0 accepted model desenbed above, it seems appropriate to assume that downward erosion is strongly favored over Case 2: Wet CCI With No Debris Flow sideward erosion. Consequently, larger probabilities are into The Suppression Pool after Downcomer assigned to the 1/5 and 1/3 branches than for the 1/1 Penetration branch. However, since some residual uncenainty remains as to the appropriate assumption for the extent For sequences where CCI was predicted to occur in of radial erosion for large reactor scale situations, a the presence of an overlying water pool with no debris probability of 0.1 is assigned to the 1/1 erosion branch- relocation to the suppression pool, the maximum amount of downward concrete erosion at 50 hours was Radial to axial erosion ratio 1/5 0.45 1.55 m (Case FMXIP). Using this value for the A amount of axial erosion, the radial erosion depth is Q Radial to axial erosion ratio 1/3 0.45 estimated for the three cases. Comparing this value to the pedestal capability of 1.55 m, the following Radial to axial erosion ration 1/1 0.1 estimates are made for the probability of pedestal failure: 19EC.2.2.5 Pedestal Failure (PED) RAD _ EROS 1/5 1/3 1/1 This branch assesses the probability of pedestal failure as a result of excessive radial concrete erosion of No Pedestal Failure 1.0 1.0 0.5 the lower drywell pede;tal wall. PedestalFailure 0.0 0.0 0.5 Structural analysis of the pedestal indicates that the l loads can be supported without yielding if only the Case 3: Dry CCI With No Debris Flow l outer shell and 15 cm of the steel webbing remains into Suppression Pool and No Late Suppression l intact. Thus, for a total wall thickness of 1.7 m, the Pool Water Ingression into The Lower Drywell ! lower limit for the amount of radial erosion which can be sustained without pedestal structural failure is This case represents case DRY in the deterministic 1.55 m. However, since the total depth of the pedestal analysis, in this case the debris is assumed to remain is 1.7 m, erosion to the full 1.7 m depth will dry for the entire duration of the accident. No flow of obviously result in pedestal failure. Additional either water or debris through the wetwell/drywell l discussion of the pedestal strength under radial concrete connecting vents is presumed to occur when the erosion is presented in Section 19EC 4. ablation front reaches the the vents. For this case the axial ablation depth at 50 hours was calced to be i Analyses were performed to estimate the extent of 2.5 m. Using this value to estimate the radial erosion concrete erosion in the lower drywell under a variety of depth for the three radial to axial erosion ratios, the conditions. The results of these analyses are split fractions are are assigned based on the pedestal summarized in Section 19EC.5. Four cases were capability: considered in the DET for quantification of pedestal l / wall failure. These cases are desenbed below. Amendment 77 19EC.2 7
ABWR 23A6100AS Standard Plant arv. A RAD. EROS 1/5 1/3 1/1. No PedestalFailure 1.0 0.99 0.0 . Pedestal Failun 0.0 0.01 1.0 Ca** 4: Dry CCI With No Debris Flow into The Suppression Pool and Late Suppression Pool Water lagression into The Lower Drywell The case in which the debris is initially dry, but becomes flooded with water after the ablation front reaches the wetwell/drywell connecting vents is considered to be slightly better than Case 3. In this case the debris is assumed to remain in the lower drywell throughout the period of CCI, Therefore, the split fractions assigned are: RAD _ EROS 1/5 - 1/3 - 1/1 - No Pedestal Failure 1.0 0.99 0.5 _ PedestalFailure 0.0 0.01 0.5 Amendment 77 19EC.2-8 _. . ~ . . _ . - - . - ~ . . . _ _ . . _ _ _ _ . ._.. _ ,-
ABWR :mos REY.A Standard Plant L ... . N ,. ... ..
.... .~
. ... . . . ... . . . m..i. .
CJHp-g my >( A T MCs= J '.A vue T -t ;.su w set *@ 485 CCI g ga T -a OHF1imithaMI 5
- ltu
.' t Ml a AImi
= P 12mut a m i y -
l w- - - , y., 12=
.a fil m rP-t!
..s g
awF tm3 Eta mi gg I 3
'ii ap*.
11 Af % 4 Pit d SK8
- T ij Ie -e en taa,f.Pe: ,,
- - Nf tihtfta ut gg 3 '
- i aaras ta, m un -amaSeia fitLLe,e r 'O yY
!qw . .9'* it M t inBP s: ,,
'3 - =>
.. i,
* ' LRa lhaETG K1 l, - * *
- W 211 **
.m gy w m ESL ,
"I 8*
ge 7Y #4
! _a w se t am#mi ,,
"" 1" - sar a wan. -i ,,
,, m L' ' n n,,a =,-*81 u<
*d*" ya , .- ,t
'* - - - N m1M M ut g
.a== = =< p
* - - dartimmiTe as 3, O
enum eru ar m B IM1.# 2' 1 gi su ae , 3
% f% \ Ct gg
' '
- f - w trmram ut y I l
man at . *_ t mL All.e ,r.a t
- l- + s c.,
\ .a m.
- =
n,m l'*
-> = sin,es,at n vi e su ,#*' N i.4 O ft M at g
'" 'nedi ,EaPha at
"," gg 4
' ?,_, ,
= = = > ' - lvv i,, - . < . ,
[ en -ma 88 El . e- __ mm 12mih at! g 5W A mAm - - M bh _, I l m. m'na
" ' " f -M5tiaafha ett y
"' ' f na.- maamm l- ~,,1 , r.,
mur er
,,,,,,, g, e Nf hM um mt g
- C SM N m mAm -~, 9' FW ya h' ' '"--%@lenMatt g "RE M ^$ i sweigna ati y
. . um ni l33
~~ ##
masa n jua -
- n , , r., ,,
. -.= . .
an
-^ as %+ ma e,
_ g, a Mt sem n LAE m
^
~, 94 vi ya l3' , a a., ,,
-- x - - , - . . , ,,
i., M,.,
, 9_ . l,, . (mo r, ,,
. ,a Figure 19EC.21 l k CORE DEBRIS CONCRETE ATTACK DET I9EC.14 Amendment 77 1
I
ABWR-Standard Plant 22^ys'l
- = :. .:::. *?,'.:67 . . , =
a _ . . .. , .. ~,
- i. 7,= %~ e
... .. .. , , _ , , , .. m. .a u m m..
,e .c ....
9
. .e . . .. .
x :.
.;r
.c . . :. . 3 a3
. .e . . u .
i 64 et T!! 19
..'.. ; >n
.c . . . - ..
12 i
,c - 3 . 1.
. e, Xa.it l'1
. - .,,..m.
, J3
.3..m .
JB l '1 8 I
, .q 2 . . t A g
15
. et3 . . it g; es. M, stt i.e ,, u 34
. .n .. u .
I4
,1 .,,,t. ';',, ,
4.e. si
. .s . . t ,.
4M
. . t. m G3 '
.C .. u A _ q 3 GL
. 3 . . e. ,,
J4 i m. m t'1 97" I 3 ..s ., t .. I 1 #9
==
..1<= .,
....n c l3H
=g'*" 7= 1 . . e.. e i >>.
.3..t .4 34
! t#1 t a1 l- f
.c ..tt s o
\ aJ Figure 19EC.2 2 d CONTAINMENT EVENT EVALUATION DET FOR PEDESTAL FAILURE Amendment ?? 19EC.210
ABWR :wms Standard Plant REY.A 0 d 19EC,3 DETERMINISTIC MODEL between the debris pool and the concrete walls. FOR CORE CONCRETE This gas film acts to insulate the concrete l sidewal s, and t c nv ct debris heat upwards. I INTERACTION This limits the heat transfer to, and ablation of
. the concrete sidewalls. Conversely, the gas film at As described above, several key parameters the bottom surfa4 of the pool would be unstable innuence the potential for concrete erosion in the due to the heavier over!ying debris pool. The presence of an overlying water pml. An analytical tool density difference would cause the lower gas film was selected to investigate the impact that these to collapse, allowing contact of the debris with parameters have on CCI, containment pressurization, the concrete. This ddference in gas film behavior opening of the over pressure protection system, and would limit the sideward heat transfer compared to I possible fission product release. MAAP ABWR was . the downward heat transfer.
selected since, with a few minor code modificanons, it was capable of investigating the key parameters In the BETA series of debris concrete experiments identified m the DET. MAAP-ABWR allowed the conducted at the KfK research center in Germany. impact of parameter variations to be carried out through dowr.want erosion rates exceeded sideward erosion containr.ent pressurization and fission product release. rates by a factor from 3 to greater than 5, For example, in the high power CCI experiment A few eimple code modificau.ons were made to BETA VI.8 the domiward erosion was measured allow the user to control the debris coolability and to to be approximately 40 cm and the sideward simplify the specification of the severe accident crosion was only about 2 cm (1/20 sideward to scenano. These changes are summanzed below. downward erosion ratio), For the low power experiment V6.1, the downward erosion was 35 (1) Subroutine PLSTM was modified to allow the cm and the sideward erosion was 10cm (1/3.5 user to specify the upward heat flux. Model ratio). parameter FCHF, was redefined to be the upward heat flux in watts /m2 . All other debris-to-water Based on the CCI experiments, and the generally heat transfer mechanisms were disabled in accepted model described above, it seems PLSTM. appropriate to assume that the ratio of radial to axial attack is 1/5. However, this parameter is (2) The following actions were added to the MAIN included as a parameter in the DET for pedestal routme: crosion since the ratio is still ricertain. (a) If lower drywell gas temperature exceeds $33 - Since MAAP assumed that radial and axial K - open passive flooder, penetration were identical, the axial ablation numbers were multiplied by 1/5 to obtain an (b) If radial erosion exceeds 25 cm - allow debris estimate on the radial anack depth, to spread to wetwell and allow water to flood the lower drywell, (2) The heat transfer from the debris to the water was assumed to be equal to the user specified value (c) If radial erosion exceeds 50 cm fail drywell throughout the transient. with an area of ADWLEK (user input), Other than the changes described above, the (d) If upper drywell wall surface temper,ture standard MAAP.ABWR code was used to quantify the exceeds $33 K - begin to leak out e the CCI decomposition event tree, upper drywell as specified in Subsection 19F.3.2.2. 19EC,3.1 Minimum Heat Flux The major assumptions included in the MAAP The most critical element in determing the r analysis are described below: " potential for core concrete interaction, and the r containment response if it should occur is the c (1) CCI experiments have generally demonstrated minimum heat flux. The heat transfer between the /i significantly more downward concrete penetration water and the debris can be limited by: q than radial penetration. It is hypothestzed that , _O D radial erosion is limited because the concrete decomposition gasses establish a gas film (1) Conduction within the debris, g p 19EC.3-t
ABWR ms Standard Plant REY.A (2) Critical heat flux, or if we v (3) Film boiling. (1) assume nucleate boiling is maintained at the surfxe, The last is of concern if the debris surface ! temperature remains so hot that the water cannot wet (2) conservatively assume that the bottom of the ! the surface,i.e. if an insulating blanket of steam forms. debris in contact with concrete ts adiabatic,
)
Film boiling has been observed in well conaolled laboratory environments using polished surfaces. (3) assume molten debris is at umform temperature, ! However, it has been observed that the smallest of and I surface imperfections or contaminants would quickly result in a transition to nucleate boiling. It seems (4) impose the condition that the debris not ablate I highly unlikely that the irregular surface of the debrir concrete, would be able to maintain itself in film boiling. Therefore, film boiling is not likely to limit upward we have as boundary conditions: heat transfer. Critical heat flux is sufficiently high that it would not impose a practical hmit on detiris coolability. . Therefore, a lower limit on the upward heat Dux may C2 = IME. ' be obtained by consideration of the conduction limit. The biggest unknown is whether the debris remains in T(6%)- 450K an intact sir.b-like confi.quration, an intact configuration with irregulanties w hicn increase the heat transfer area where: 6* = debris thickness. 7 and act as fins, or if the debris develops cracks which u allow water to ingress. The presence of cracks would Substituting into Equation (2), we have for the 7
/ increase the heat flux. Therefore,let us consider the worst situation (intact stab).
limiting debris thickness for coolability: { n.J '
~
,~,
j The temperature distribution in steady state, Sw = 0.08 m C assummg a homogeneous debris mixture, is given by: '] This means that if we are in nucleate boiling at the 2 3T surface, we can just remove decay heat purely by k p + q,, = 0 (1) conduction through the debris slab at a thickness of 8 cm. The surface heat flux is: where: k = thermal conductivity (3.5 W/mK), q" = q"Sa = 100 MW / m 2 q~ = volumetric heat generation. It is sufficient for our purposes to consider the case flux would be substantHly higher than this value. of 1% decay power. For a total debris mass of about Thus, one could view this as the lowest possible 244,000 kg, this implies and average initial volumetric upward heat transfer given the boundary conditions. A heat generation rate. higher temperature at the bottom of the crust or heat transfer into the crust would both ir rease the dans-to-water heat transfer, MW q" =1.3 m3 This rather low heat transfer would be increased if the surface was of non-uniform thickness (fin effects) or In a one-dimensional fie geometry, integrating espec ally if the surface cracked sufficiently to allow Equation (1)twice yields: water to ingress. T= ~9 *2 +Cx+C i 2 (2) t 4 V AmenJmest 71 19EC N
ABWR 23A6100AS Standard Plant arv. A A () 19EC.4 PEDESTAL STRENGTil The configuration of the ABWR pedestalis shown in Figure 1.213e. The width of the pedestal is 1.7 m. The design consists of two concentne steel cylinders connected by steel web stiffeners. Ten wetwell-drywell connecting vents run through the annular region between the cylinders. The remainder of the space is filled with concrete, if significant core concrete attack occurs, the strength of the pedestal could be compromised as the pedestal is eroded. The strength of the pedestal after it has undergone erosion is examined to determine the maximum erosion depth allowable to ensure that the pedestal does not collapse. The pedestal is designed based on the maximum stress obtained in the steel plates. The strength of the concrete is neglected. The allowable stress in the steel plates is 0.6 t;mes the yield strength, neglecting temperature. The calculated stress without seismic loads in the ABWR pedestal is 0.4 times the yield strength. For design analysis the largest single load is the accident temperature, if core concrete interaction were to take place as a result of a severe accident, the inner plate of the pedestal would melt. Without a continuous i I (q
\
inner plate the moment induced by the differential temperature disappears. It is expected that any temperature induced moments acting along the stiffeners will be strain limited. Therefore, they will not reduce the capability of the outer plate. In order to estimate the allowable ablation depth, the seismic and thermal loads are removed and the remaining loads are calculated. No attempt was made to take credit for the relocation of fuel from the vessel onto the floor of the drywell. The strength of the remaining concrete is neglected. The loads are compared to the yield strength of the remaining pedestal steel. Therefore, this calculation corresponds roughly to a service level C type of calculation. The results of the calculation show that the outer shell of the pedestal plus 15 cm of the web stiffeners are required to maintain the pedestal loads below yield. This limit is used as a conservative estimate of the pedestal ultimate capability after erosion. The total pedestal width is 1.7 m. Therefore, pedestal integnty is ensured fcr ablation depths up to 1.55 m. l l f3 Amendmem ?? 19ECA-1
ABWR 23A6100AS Standard Plant REV.A O (.) 19EC.5 APPLICATION OF CCI kW/m2 was selected to represent limited water ingression into the debris bed with the upward heat MODEL TO ABWR transfer being controlled by film bc,iling. The largest value used represents critical heat Oux limits for debris The deterministic code used for investigaung core-c ling. Further discussion of these values is included concrete interaction in the ABWR was desenbed in in subsection 19EC.2.1.6 Section 19EC.3. This section will describe the evaluations that were made to support the quantification of the CCI decomposition event tree. As run in its standard manner MAAP-ABWR calculates that 60% of the total core inventory was ! released from the vessel. The remaining 40% was !
- 19EC.5.1 Sequence Select!on calculated to be held up in the core with the decay heat being radiated to the vessel wall and convected into the The MAAP ABWR code, as modified for this upper drywell. De 40% remaining behind is typically applicauon, allowed for a great amount of Oexibibty in the outer penpheral bundles which have low decay heat.
analyzing the impact of key parameter vanations on To suppon the DET quanufication additional ca;.es were core-concrete attack. The following lists the key run assuming that 100% of the core was discharged parameter variations that were investigated; from the reactor vessel. This has two major induences on the containment behavior. Without the peripheral (1) Upward heat transfer to overlying water pool bundles in the core, the drywell heatup is reduced. Second, the added core mass on the lower dr)well Door (2) Mass of debris discharged from vessel will influence the calculation of core concrete attack, debris coolability and containment pressurization. (3) Mode of fission product release from containment (4) Flooding of lower drywell resulting from radial penetration of vertical connecting vents l D I (5) Debris spreading related to radial penetration of ' gd vertical connecung vents l De base case sequence selected to investigate core ' concrete interaction was the low pressure loss of injection scenario. This event was initiated by a transient with the assumption that all injection was unavailable. The RPV was depressurized manually l when the core level dropped below 2/3 core height. [ Without coolant injection, the core melts and slumps l into the lower vessel head. Local penetration failure I occurs and the debris is discharged into ine lower drywell. Table 19EC.51 provides a chronology of the l i events up until the vessel is failed. Table 19EC.5 2 defines each of the sequences analyzed and provides a summary of the results. The l I first column gives the case designation along with
- reference to specific notes. Columns two through four provide the relevant sequence definition information.
For purposes of demonstration, all cases were executed for the dominant sequence, a low pressure loss of injection sequence with a containment pressure at the time of vessel breach of approximately 1 atm. The upward heat flux was varied between 100. 300, and 900 kW/m 2. A value of 100 kW/m 2was selected to approxirntte the heat transfer associated with a stable
/ T- crust formation where the upward loss is controlled by C conduction of heat through the crust. A value of 300 Amendment 7t 19EC 5 1
ABWR 23A6100AS Standard Plant REY.A A U 19EC.S.2 Summary of Resuhs Table 19EC.5 2 summarires the results of the deterministic analyses for the ABWR. The following i general conclusions are indicated by these results: (1) For all sequences with successful operation of the flooder, radial concrete crosion was less than the structural litnit described in Section 19ECA. Radial attack does not pose a significant challenge to containment. 1 (2) For sequences with operatica of the containment overpressure protection system, due to l suppression pool scrubbing, the fission product release is dominated by noble gas. (3) Release times for cases with the passive flooder are on the order of 20 hours after the initiation of core damage (defined as onset of melting). (4) The extended time period between vessel breach and rupture disk r :tuation (or containment failure) provides f't a substantial reduction in the amount of fission pn' duct released from containment. (~ (5) Using experimentally based values for the upward (y] heat transfer (Se: tion 19EC.1) would result in debris cocling in the ABWR and early termination of the core concrcte attack. Therefore, the lower bound for upward heat transfer is conservatively assumed to be 100 kW/m2 . This is done in order to obtain substantial concrete erosion and demonstrate the robust:4ess of the containment design if the debns is not quenched. (6) For the domitiant scenarios with successful operation of fire water to provide water to the debris, the time from onset of melting to fission product release is 24 hours fnxn the beginning of the accident for all upwwd heat transfer rates. A set of plots for case FMX100 case are included in Figures 19EC.51 through 19EC.F5. This case demonstrates long term core concrete interaction, but is otherwise typical of the conditions analyzed. The 2 depletion of zirconium in this case occurs at about
.hk'N0 seconds, coincident with the onset of CO 7
l' production. The hydrogen gas generation is not n equivalent to the amount which would be generated 5 from a 100% metal water reaction because of a d competing reaction between the niconium and CO2 . O V Amendment 77 19LC.5 2
ABWR 23 A6100 AS Standard Plant arv. 4
'T = Velocity of the carium expelled from 19EC,5.3 Initial Concrete Attack due to where: uo s) Impinging Corium Jet the reactor vessel, At vessel failure, core matenal is discha*ged from 8 = Acceleration of gravity,
. the RPV onto the floor of the lower drywell. At low I RPV pressures, the discharge rate of the debns is and tra is defined by controlled by gravity and the vessel breach area in the i lower head. From analyses performed for FCI j calculaunts, Subsecuon 19EB.6.2.2, it is assumed that uot t a + ; gtfa = z, (6) l ten penetradons failed. This results in a maximum
- corium discharge rate of 6000 kg/s. The total failure area is 0.145 m2 . Assuming a density for conum of when
- : r, = the elevation of the reactor vessel l
above the lower drywell floor. l 8000 kg/m 3, a discharge of 6000 kg/s corresponds to a l conum velocity of 5 m/s. The following calculadon I estimates the miual concrete attack depth resulting A crust of frozen corium forms on the concrete and ! from this impinging conum jet. the ablation prmess is the same as at the reactor vessel penetrauon. Thus, the concrete ablation velocity is l The model from the MAAP subroutine JET given by (Reference 9) was used to compute the concrete attack from an impinging jet of conum. The stagnation point h(T. - T,) heat transfer coefficient tetween tha corium jet and the u, = (7) concrete is appmttmated by the expression, p,. c p(T, - To) + b where: T. = Bulk corium temp:rature, ) Nu = $ = 1.14de (3) k
- p. = Concrete density, p a ,
cp = Concrete specific heat, E " "' (4) h = 1.14 k. = Concrete latent heat,
!.t . D,. L
= Conum thermal conductivity, Tenp = Concrete m:lting teq rure, where: K.
u,, = Conum viscenty, To = Irtitial concrete temperature, ue = Velocity ithe corium stream Substituting the corium velocity and the ABWR impinging on the floor, specific geometrical parameters into the above equauons, results in an ablatior, rate of appraximately 1 cm/sec. Wuh the debris being discharged over 5 l D = Diameterof thejet, seconds, tt.: resulting abladon depth is 5 cm. This l w d only & cur in me ccmral pwuon of the tower h = Heat transfer coefficient, drywell, and would in no way threaten the integnty of a tures.
- p. = Corium density, Nu = Nusselt number, Re = Repolds number.
The corium velocity at the cavity floor is given l t by, [ u, a un + gtt e (5)
&]
i l I9EC.5 3 Amendment ?? l l
-ABWR ,,,,,,,,
Standard Plant REV,A Table 19EC.51
SUMMARY
OF TIMING FOR CORE CONCRETE INTERACTION BASE CASE Time (Sees) Eysnt 0.0 Loss of allinjection 4.2 Reactor scrammed 1097.0 Com uncovend 1138.0 Manual depmssurization 3451.0 Onset of core melt - 5364.0 Slump intolower head 5382 0 Vessel failure (G O Amendment 77 19EC.5-4 E1
I a L";> I E CIlI Table 19EC.5-2 {
SUMMARY
OF CCI DETERMINISTIC ANALYSIS FOR AHWR' y% 3. 2 as Upward Debr*s Mass Radial 112 g Fission Product Release Fraction ; Containment IIcat at Vess. Fail Attack at Generated Time of FP ' Press.at Trans. (Frac.of Tot. 50 hrs. at 50 hrs. Release Mode of from Containment Case # Vesselfailure (kw/m2) leventory) (meters) (Kg) (hours) Release NG Csl Sr ABWR100 1Atm. 100 0.6 0.22 1813 19.1 COPS 1.0 2E-06 3E-09 , ABWR300 1 Atm. 300' O.6 9E-07 122 23.3 COPS 1.0 . 2E.10 2E-12 ABWR900 1 Atm. 900 0.6 7E.06 122 23.2 COPS 1.0 3E-11 2E-12 - FMX100 1 Atm. 100 1.0 0.25 2130 17.6 COPS 1.0 IE-06 IE-08 FMX300 1Atm. 300 1.0 7E-03 154 19 3 COPS 1.0 IE-08 3E-15 FMX900 1Atm. 900 1.0 7E-Ol 111 19.1 COPS 1.0 IE-08 2E-14 FMXCSP(1) 1Atm. 100 1.0 0.25 2126 15.7 COPS 1.0 4E-07 3E 10 SENSITIVITY RUNS DRY. IAtm. N/A 0.6 0.50 4990 19.8 DWT 034 4E-03 IE-05 DWFAIL iAtm. 300 1.0 7E-03 154 19 3 DWF 1.0 8E-04 2E-10 FIRE IAtm. 100 1.0 0.25 2131 24.6 COPS 1.0 SE-06 4E-10 FMX1P (2) 1Atm. 100 1.0 031 2762 17.6 COPS 1.0 IE-06 IE-08 NFlood (3) 1Atm. 100 0.6 0.25 2127 17.4 COPS 1.0 8E-07 5E-10 LATE (4) 1 Atm. N/A 1.0 031 2697 20.0 DWT 0.23 6E-03 9E-09 Notes-COPS - Cantainment Overpressure Protection System DWT -- Drywell I e*Irmge occurs through penetrations. DWF .- Drywell Failure (0.0973 m2 ) (1) - FMX100 Run with five times steel mass (2) - Penetration into connecting vents does not cause debris spread. l (3) - Flooder not operauonal, Radial attack results in penetration to WW and debris spread. (4) - Vessel failure assumed to occur after lower plenum water boiled away and debris reheats.
- U R E
? 8
>D
i n 1 j ABWR- 23s,, ,s l Standard Plant .. REV.A .l \ i i 1 1 I LP Melt -
.1 mw/m2, Drop Core I N
) _......,,,...i..i....g...,
- m : :
j E - - tn . o . . ! O - - j s -
- C :
e . . i s - --
- e -
g. !O o : : 4 V E : : o _ _
- in __ __
O . - - x - - C : o 'I''''I''''I'''' 0 .5 1 1.5 2 5 TINE S x10 4 Figure 19EC.51 . SAMPLE CALCULATION FOR CORE CONCRETE INTERACTION UPWARD HEAT FLUX 100 kW/m2 : AXIAL CONCRETE ATTACK l' Amendment n 19EC.$4
i ABWR 23 A6100AS Standard Plant nsv. A O LP Melt -
.1 mw/m2. Drop Core g m_.........i.... ....i... ...i.........
x E E N 5-
-i-m i E o : :
g .
~
v 0 --
-I c : :
5m 5-E
- e. : :
L i i
- n. v _ 2 O :
i i: 3 m *-- J e : : 2 5
- E.
N 3- -E i ( i iiiiiiiiiliiiriii,iliiiiiiiiili,,iViiii l o .5 1 1.5 -2 TINE S x10
- Figure 19EC.5 2 SAMPLE CALCULATION FOR CORE CONCRETE INTERACTION UPWARD HEAT FLUX 100 kW/m2 : WETWELL PRESSURE Amendment ?? 19EC.5 7 l,
i. L ABWR 23A6100AS
. Standard Plant nsv. A 4
i O i e d 4 LP Melt - .1 mw/m2. Drop Core g..........i......i.i........i........_ m - n - . u : N - av
/ 7 e ' b i w -
~
~
2 e 8 ~~ i x .. . Oa t. i e a g D o -- o" : o ''''l''''l''''lr' i' O .5 1 1.5 2 TIME S x10? Figure 19EC.5 3 SAMPLE CALCULATION FOR CORE CONCRETE INTERACTION UPWARD HEAT FLUX 100 kW/m?. UPPER DRYWELL TEMPERATURE l-O i Amendment 7? 19EC.$ 8 l _ _ ._, . - . _ - _ . . - . . .,.
ABWR 23 A6100AS - aty,4 Standard Plant LP Melt -
.1 mw/m2, Drop Core i
n e , ; g _
....i....... .i
. l x :
l 1 a m
- l a : : )
v - l
. ~
O - r n L Oi 2 i 1 o _ _-
.J -
I iiiiiliiiiiiii,l,iiiiii,,li,,iii,ii: o i iii 0 .5- 1 1.5 2 5 TIME S x10 i Figure 19EC.5 4 SAMPLE CALCULATION FOR CORE CONCRETE INTERACTION UPWARD HEAT FLUX 100 kW/m2: LDW WATER MASS-
/
l'
- 19EC.5 9 Amendment 77 e - - ,,_y --.n+ em=+Je
ABWR 23A6100A5 Standard Plant arv. A O LP Melt -
,1 mw/m2 Drop Core
.. ... i. ,,iii
.;iii .iii ;i i; O
m
~
O -- O : : n in : : v" v-4 a.8
- e. O E-en :
H : : E O a sOto : : c- :: : O o C'O ? b
~
o ~ mO : : C : : O o : : to e : iiiiiiii Ii-iiiiiii-iIiiiiiiiiiIiiiiiii,i - o 0 .5 1 1.5 2 5 TIME-S x!O l Figure 19EC.5 5 SAMPLE CALCULATION FOR CORE CONCRETE INTERACTION UPWARD HEAT FLUX 100 kW/m2: AVG CORIUM TEMPERATURE O Amendment 77 -19EC.5 10 _ - . - . . . _ . _ _ _ , . . , - _ . ~ _ . . . - . _ , _ . - _ , _ . -
ABWR 23A61004s Standard Plant REV.A n h 19EC.6 PARAMETERS SENSITIVITY TO VARIOUS Case LATE nis sequence was identical to case DRY except for Also included in Table 19EC.5 2 are other analyses a dela)ed vessel failure. He RPV was assumed to fail after all of the water in the lower plenum had boiled that address possible sensitivities to modelling away and the debris heated up to the cutecuc melting assutoptions. These results are descnbed below. point (2501 K). N essel failure occured at 5.3 hours into the sequence as compared to 1.5 hours for the base case. C,ase DRi, S nce there was no water discharged with the core debris at vessel failuie, the gas temperature quickly increased This case was run assuming that the passive to above the Gooder actuation temperature. The flooder flooder did net open and that, even after radial was assumed not to work for this case. The purpose of penetration of the vertical vent pipes, water was not the nm was to obtain an estimate of the time period introduced into the lower drywell. He drywell began t between vessel failure and flood-r actuation. The leak at about 20 hours and resulted in a slow, low MAAP analysis conservatively assumes that the gas magnitude, release of fission products, must reach 533 K before the Dooder can open. In this case it took about one hour before the gas reached 533 Case DWFAll' K. Factoring in the difference tCween the wall surface and the gas temperature, the floc &r would be expected This case is identical to case FMX300 except that to open within 30 minutes after Wwharge of the core l the drywell was assumed to fail at the COPS set point- debns. All other aspects of this rv were similar to the l Due to the long time between veul breach and DRY case. I I contamment failure, the fission products settle out very effectively and the result is a low magmtude Tlease. The overall conclusions from the sensitivity
)
analyses are that the ABWR containment design is Case 14 f X1P quite insensitive to the uncertainues associated with core concrete interaction. ne concrete erosion rates are p his case was identical to case FMX100 except consistent with other published results (Reference 8) Q that the debris is assumed to not spread into the wetwell after gnetrating the vertical connectmg vents-rnd do not pose a serious thraat to containment integrity. Operation of the COPS provides for a The results indicate no sensitivity to this assumption, scrubbed release of the fission products and greatly The radial attack at 50 hours is 31 cm for a ratio of limbs the risk to the public, radial to axial attack of one to five. 19EC.6.1 Impact of Pedestal Concrete Case NFLGOD Selection This case was idemical to case ABWR100 except that the firewater addition system and passive flooder . The pedestal of the ABWR n, defined as the sidewalls of the lower drywell. This structure supports were not operational. Therefore, the debris was initially the vessel and the wetwell/ upper drywell diaphragm dry. After 25 cm of radial erosion, the debris was floor, ne type of concrete to be used in the pedestal is assumed to spread into the wetwell and water from the mi speciSed. Basaltic cmente is required fm the Dom suppression was introduced into the lower drywell. De f the lower drywell, results indicate more concrete erosion with the COPS actuating at 17,4 hours compared to 19.1 hours. Basaltic concrete was used for the lower drywell in determining the response of the containment to core Case FIRE concrete attack. This type of concrete is often used in the United States. He other type of concrete which is This sequence was identical to FMX100 except frequently used is timestone. common sand. Basalue that the firewater system was used to add water to the concnte is mme ra#y cW Mng cm cucrete debris. Due to the addition of cold water, the imer clim h ,s hmestmeommon i sand cuente. pressunzation of containment due to steam was reduced refwe, me wmid expect that nrnestmecmmm t.nd the COPS was not predicted to open until 24.6 sand comete wm ud in h AN @W 4e. de hours as compared to 17.6 hours for the case with side wal!s), the sideward erosion rate would be slower passive flooder operation. than that presented in Table 19EC.5-2. Therefore, the times estimaad in that analysis for the time at which pedestal integrity could be threatened are expected to be v 1 1 19EC & I Amendment 77
ABWR MA6tmAs Standard Plant REY.A O) conservative if non basaltic concrete is used in die pedestal. The other key impact of the type of concrete is the production of non-condensible gas. Limestone common sand concrete produces more non-condensit'le gas than does basaluc concrete, lloweser, this will not have a significant impact on theis analysis because the surfaco area of the sidewall will be only ten to fifteen pe cent of the floor area if core concrete attack should oecur. Furthermore, the shape of the debris p.vl will be pancake like.1he gas generated at the side wall will not be able to reach into the debris pool and cause more
- rapid metal water teaction in the debns pool. Rather, it v will bypass the debns. Therefore, there will be little ,
l 1 J i impact of the gas generation on the rate of attacir due to any enhanced rnettJ water reaction. 1
, 50 In summary, the type of concrete to be used in the d pedestal side wall is not specified, if non basaltic concrete is used in the pedestal the rate of sideward ablation may be somew hat reduced as compared to the d
analysis presented here. The rate of non condensible gas generation may be slightly higher. Howeser, because of
- the relative areas of the sidewall and the floor the (
) l'ipact will be small. The conclusions of the I I uncertainty analysis will not be affected by a different p r hoice of concrete. V V) I l Amendment ?? 10 EC.6-2
ABWR :3AetmAs Standard Plant nty A D 19EC.7 Impact on Offsite Dose The effect of the maximum core concrete interaction sounc term on a relaease with operation of the rupture disk is shown in Figure 19EC.71. The cases with rupture disk are the only risk significant release categones which would be impacted by core concrete interaction (The other sequences are cases with early containment failure due to DCH.) As the Figure 13 clearly shows,CCI does not have ignificant irnpxt on the offsite dow. O t e.
)
O Amendment ?? 19EC 71 l l
ABWR ' Standard Plant "*'$^4 O E 1 -- g
-s X -
-w
- \\
5 g cas.2 g O W case 1 l 0.1 1 0 ~ d 's T u O \ O e 2 x
\
m 0 C
' 0.01 .
O 5 1'O 1'S 2'O 25 EARLY WHOLE BODY DOSE IN REM. < Figure 19EC,71 WHOLE BODY DOSE AT 1/2 MILE AS A PROBABILITY OF EXCEEDENCE O 19EC7 2 Amendment 77
. . . _ . , . . ~ . . , , ,-- , , . . - . . - . . , _ - _ , , _ -
ABWR UA6100A5 Standard Plant nEv A 19EC.8 Conclusions This section investigated the impact of core. concrete interaction on the ADWR containment response. First. detailed DETs were develcped to address all of the key parameters that innuence CCI. Then, several determinstic anal) sis were carned out to supp rt quantiGcation of the trees. The following summarizes the important conclusions of the CCI investigation: (1) For the domirunt core melt sequerres that release core material into the containment,90"r result in no signincant CCl. Virtually no sequences have dry CCI. (2) Esen for the low frequency cases with significant CC1, radial erosion remains below the structural limit. (3) The fission product release mode is dominated by operation of the containment overpressure protection system. The release, which occurs at - about 24 hours, is not distinguishable from a case . with no CCI. (4) Experimental results indicate that sufficient upward heat transfer to an overlying water pool O, would exist in the ABWR lower drywell to cool the debris. l l [ V I Amendment 77 19 EC.B.I
4 e i i ABWR 23y,,, i Standard Plant arv.4 1 l 19EC.9 REFERENCES ! 1. Final Report on Core Debris Coolability,IDCOR j Task 15.2 4
- 2. An integrated Structure and Scaling Methodology for Severe Accident Technicalissue Resolution, to be published as NUREO/CR. Draft 1991.
i
- 3. 0.A. Greene, C. Finfrock and S.B. Burson, 1 Phenomenological Studies on Molten Core.
l Concrete Interactions, Nucleat Engineering and i Desien, 108,167 177, 1988. 4 M.W. Tarbell. D.R. Bradley, R.E. Blose. J.W. ,
' Ross, and D.W. Gilbert, Sustained Concrete Attack by Low. Temperature Fragmented Core - , Debris, NUREO/CR 3024 SAND 82 2476 R3, !
R4, July 1987. !
; 5. R.E. Blose. J.E. Oronager, AJ. Suo Antilla, aml ,
i 1.E. Brockman, $ustained litated MetailletMelt -
- Concrete Interactions with Overlaying Water l Pools, NUREO/CR 4727, SAND 851546 R3, R4, R7. July 1987.
- 6. R. Henry, Experiments Relating to Drywell Shell
- ~ Core Debris Interaction, BWR Mark I
! Containment Workshop, Baltimore, MD. February
- 24 26, 1988. See also B. Malinovic, R. Henry, I i and B. Sehgal, Experiments Relating to BWR
\ Mark i Drywell Shell Core Debris interactions, i ASME/AIChE National Heat Transfer Conference, Philadelphia. August 1989. i i 7 B.R. Sehgal. ACE Program Phase D: Melt Attack and Cootability Experiments (MACE) Program, l presentation at CSARP meeting, May 1992. f i +
- 8. S.R. Greene, S.A. Hodge, C.R. Hyman, M.L t Tobias, The Response of BWR Mark It i Containnwnt to Station Blackout Severe Accident Sequences, NUREG/CR 5565, ORNL/TM.!1548, May 1991.
l 9.- MAAP 3.0 B Computer Code Manual. EPRI NP-
- . 7071 CCML, Volume 2. November 1990.
O Amendment ?? 19FC 91
-N-t' rmymy-s*"r"-m- ""m" -
ABWR nA6:nors Standard Plant REY.A N P Os , W5.0.1 ISSUE
.f*s
; Dui.nv - .pothetical severe accident in the
' 3Wi<, moltr- ore debris may be present on the
!*er drywth ~ 0) floor. The EPRI ALWR
, v on . c' '
.4 .m.ient specifies a Door area of at
.s ? $ enW g ito promote debris coolability.
as been interpreted in the ABWR design as,a o u ment for an unrestricted LD Ikxx area of 79 m*, } The ABWR has two drain sumps in the periphery I of LD floor which could collect core debris during a severe accident if ingression is not prevented. If ingression occurs, a debris bed will form in the sump which has the potential to be deeper than the led on the LD floor. Debris coolability tecomes rnore uncertain as the depth of a debris bed increases. The two drain sumps have different design objectives. One, the floor drain sump, is designed to collect any water which falls on the LD floor. The other, the equipment drain sump, collects water leaking j from valves and piping. Both sumps have pumps and i instrumentation which allow the plant operators to e determine water leakage rates from various sources. i) Plant shutdown is required w ben leakage rate limits are r~T a escecded for a certain amount of time. A more complete Q$ 3 discussion on the water collection system can be found in Section 5,2.5. f-Amendment 71 19ED l 1 j
ABWR m, , ,,,, , Standard Plant nix.A 19ED.2 PROPOSED 1)ESIGN A protective layer of refractory bricks-a rorium shield-could be built around the sumps to prevent corium ingression. The shield for equipment drain sump would be solid except for the inlet and outlet pipmg which would go through its roof. The shield for the floor drain sump would be similar except that it must have channels at floor level to allow water which falls onto the LD Door to now into the sump. The height of the channels would be chosen so that any molten debris which reaches the inlet will frecie before it exited and spilled into the sump. The width and number of the channels would te chosen so that the required water now rate during normal reactor operation is achievable. A sketch of a concept for Door drain sump shield is shown in Figure 19ED.21. The walls of the equipment drain sump shield (solid shield) only have to be thick enough to withstand ablation, if any is expected to occur, for the chosen wall material. The walls of the Door drain sump shield (channeled shield) must be thicker so that molten debris flowing through the channels has enough residence time to ensure debris solidification. (g Both shields would extend above the LD Goor to an elevation greater than the expected maximum height of core debris. Thus, no significant amounts of debris will collect on the shield roofs. The solid shield can te l placed directly on top of the LD Hoor. The channeled shield will have refractory bricks embedded into the LD floor beneath the shield to prevent core. concrete interaction involving the molten debns in the charmels. The analyses presented in Sections 19 EDA and 19ED.5 provide a basis for sizing the propowd design of the floor drain sump corium shield. / '\ Amendment ?? 19ED L 1
ABWR 2tA6100A$ Standard Plant n:v, A Lower Drywell Corium Shield mrtrcctrostrya 5 Channel g gm - g T e-1 1 Sump Concrete Fill I l mu ! Containment Boundary a) side view Channels
.a-
..M
. = ,. 5. . 5. . =. .'. -
44rM A?*1 M ?*?#
.. m
,I Concrete Fill b) front view Figure 19ED.21 CONCEPTUAL DESIGN OF LOWER DRYWELL FLOOR DRAIN SUMP SillELD V
Amendment M 19ED 2 2
ABWR 23 A61(K)AS Standard I'lant nev.A v 19ED.3 SUCCESS CRITERIA FOR d*is. I l'ROl'OSED DESIGN (7) Seismic Adequxy a For the proposed design to be considered d The seismic adequacy of the corium shields wtil " successful,it must satisfy the following requirements: be detertnined in the detailed design phase. 9 (1) hielting Point of Shield hiaterial Abose Initial Adequacy should be easily meet because the 4 Contact Temperature shields are at the lowest point in the containment, j hiissile generation is not an issue because the
'J shields are not near any vital equipment.
The shield wall material shall be chosen so that ter a tem rat te t tween he de ns d th Section 19ED.6 contains an example of success dtield wall. calculations for requirernents (1) throut (4) for a chosen channel height of I cm. (2) Channel Length The length of the channels in the shield must be long enough to ensure that a plug forms in the channel before debris spills into the sump. De freezing process is expected to take on the order of seconds or less to complete. (3) Shield licight,16, Atme Lower Drywell Floor
%e shield height above the lower drywell floor p shall be chosen to ensure long term debris s
1 solidification. The freezing process will be complete during the time frame when the shield walls are behasing as semi infinite salids, in addition, the shield must be tall enough to prevent debris from accumulating on the roof of the shield. (4) Shield Depth,14, Below Lower Drywell Floor The shield depth of the below the lower drywell floor shall be chosen to ensure long term debris solidification. (5) Water Flow Rate The total flow area of the shield channels shall be great enough to allow water flow rates stated in the Technical Specifications without causing excessive water pool formation in the lower drywell. (6) Chemical Resistance of Shield Walls i e De wall material chosen for the corium shields
? must have good chemical resistance to siliceous d slags and reducing environments. Resistance can p $j be determined to a first degree by companng the i o flibb's free energy of the oxides which make up the shield wall and the oxides present in core Amendment ?? 19ED 3 i
i ABWR mum 4s Standard Plant REY.A i ! I 19ED.4 ANALYSIS OF SillELD until long after a plug has formed. Any heat ! FREEZING ABILITY transfer i the shield material between adjacent-channels enhances the debris frtezing process.
- Heat transfer and phase change analyses are i presented in this section to determine the feasibility of (4) The shield rail acts as a semi. infinite slab with an imtial temperature of 330 K during the initial l a channeled shield to prevent molten debns ingression freezing pmcess.
- into the floor drain sump. Two time frames were considered. First. a freeze frcnt analysis was performed i for early times (seconds or less) to detennine the time The properties of shield cause il to le a poor ,
required to form a plug.The long term ability of a plug conductor of heat. The penetration depth during i i to remain solid was determined using a steady state the short duration of the freezing process is on the I analysis, order of ten millimeters. The small increases in LD temperature prior to the presence of core debris does not significantly alter the shield 19ED.4.1 Assumptions temperature from its value dun,ng normal plant operation. The major assumptions invoked in the analyses and their bases follow: (5) Core debris is not expected to enter the LD until at least two hours after accident initiation at (1) Molten debns enters the channel with negligible w hich time the dec- ( heat level is approximately superheat. one percent of rated power. Molten debris interacts with structural material Core debris will not enter the lower drywell before (steel, concrete, etc.) and the lower drywell about two hours for any credible severe accident environment as it passes from the vessel. contacts (Section 19E.2.2).- the LD floor and spreads to the shield. This !- interaction depletes the , molten debris of,any @ The decay heat generation in the debris is ! superheat and can result in eutectic formations, negligible compared to the rate of latent heat ne melting temperature of core debns which has generadon during the freezing mess. undergone little interaction is approximately i 2500 K. Significant interaction with the concrete g g.;g g l floor reduces the debris melting te.nperature to approximately 1700 K. (7) The thermal conductivity and thermal diffusivity of debris in solid and liquid phases are the same, (2) During the freezing process, the temperature profile of the solidified debris rapidly obtains its steady state value. (8) The contact resistance between the bricks was assumed to be negligible. Contact resistance can be controlled in the final design by varying the j This assumption introduces little inaccuracy beca s ,' thickness of the bricks or by usm, g a high , temperature binder between the bricks. Thicker m bricks tend to minimize overall contact resistance ] a) the heat conduction coefficient in the solidified by reducmg the number of contact points. Some W debris is significantly lar8er than that of the shield contact resistance may be acceptable in the final _ matedal' d[ - design if the composite thermal conductivity is high enough that the shields provide short and-b) the depth of the solidified debris is long term debris solidification. considerably less than the height of the shield.
. '9) The corium shields were assumed to be 2 (3) Heat transfer within the channel and shield is one- - structurally stable. Structural stability is only an 0 dimensional issue during the initial onslaught of debris into y the lower drywell. After debris comes into contact a l- De height of each channel is much less than its length. The heat transfer in the shield material is with the shields. a crust will form and it will tend 4 low enough that any heat transferred from debris to grow in time. Crust formation climinates f "i
buoyancy forces and will hold the individual . contacting the shield wall outside of the channel bricks into place, does not affect the temperature along the channel i Amendment 'n 19ED 41
. , . , _ . - - _ -. . _ ___-_._._._.-a._.u.__.. _ - . - _ . , - m.. , _ . _
ABWR 2mios Standard Plant REV.A O Initial Freeting of Molten V 19E D,4.2 4 .,*"~ ' dT,
"* ('
, Debris in Channel If the floor drain sump shield fulfills its design where: 41 = the latent heat flux. I objective a debris plug will form in the channel before molten conum has a chance to traverse the channel and The latent heat Dux is: reach the sump. Molten debris enters the channel at a significantly elevated temperature (2500 K to 1700 K) dt compared to the shield wall (~ 330 K). The walls qs - dtS- puh m (3) absorb heat from the debris because of the large temperature difference. Since the debris contains whe e: 4 = crust thickness, l negligible superheat, any heat loss by the debris results in freezing. Frecie fronts start at the channel walls and move toward the center of the channel. The leading edge t = time, of the freeze front will stay at the melting temperature of the debns. The freezing piacess is symmetric about pan = density of debris, the centerline of the channel because the same amount of heat is transfened through each wall w hile they are hih = debris latent heat of fusion, behaving as semi infinite slabs. The channel walls behave as semi infinite slabs during the freering Combining these two equations, evaluating the process because the heat conduction rate through the temperature gradient and rearranging yields: wall material is low compared to the release rate of latent heat. A sketch of the freezing process is shown - in Figure 19ED.41. d,x,, , I -k l S-dt p,,h > _ x, (Tu -- Td - 4 *2 , (4) (1) Freezing Time O This is a non linear, non homogeneous, first order Q The temperature profile in the crust (Reference 1), assuming it quickly rexhes its steady state shape, differential equation. Before cffon is expended to solve t, the relative magnitudes of the terms 85: containing the crust thickness will be determined to see if either one dominates, x' +T, -2Tr m x 2 T,(x) 2= g '1' 7' ' {' The initial interface temperature between the wall I (1) of the channel and the debris can be approximated
. T. + Tr m by assuming both the debris and the shield wall l behave as semi infinite solids. The resulting 2
temperature will be somew hat less than the actual where: T,(x) = temperature within the crust, interface temperature because the freezing process will force the crust to stay closer to its initial temperature than it would if it were an semi-x = crust coordinate measured from infinite solid body only experiencing conduction. the crust centerline, The contact temperature between the debris and the channel wall (Reference 2), assuming semi-q = heat density of the crust, nfmite bodies,is: 4 = half thickness of the crust, T. = Tom /(kpc), + Td(kpc),, (5) I
= thermal conductivity of debris, Y(kPC)cm + g(kpc),-
kr T, = interface temperature between the where: c = specific heat, wall and debris, em = debris material properties, Tr,m = melting temperature of debris. O w = w all material properties. . b The energy balance at the freeze front is: Using the debris properties found in the Table Amendment *? 19ED.4 2
ABWR nuimAs Standard Plant RIN.A b V entitled Important Parameters for Steam height 1(,. The frecring time is: Explo. tion Analpis (19E.217) and representative wall properties found in Table 19ED.41, the gj"2p* h
- interface temperature is estimated to be 1390 K. (9) t r , = d, k r (Tr,, - T, )
he debris energy generation density can be found by assuming a decay heat level and a total amount (2) Interface Temperature, T, of corium. De density is: The interface temperature between the debris and the channel wall can be determined by equating gI,Oep" (6) the heat flux from the crust to that which the m* crust can absorb. The heat flux from the crust is: where: Q3 = decay heat level, q" dT' (10;
= -k r m, = total mass of corium,235 Mg. dx .i,t2 Evaluaung this two hours af ter accident initiauon w hich evaluates to:
(decay heat level equals approximately one percent of rated power) yields: q", = 4'
- L + kL (Tr., - T,)
(10 2 x* 6 3 4 = 1.5 x 10 MW / m As shown previously, the temperature-difference De two terms inside the brackets in Equauon (4) term dominates the energy generation term in this can now be evaluated. For a channel height of I equation for small channel heights. Therefore, the cm (x e,m = 0.5 cm) and a debris melting crust heat flux can be simplified to: L.:mperature of 1700 K, these values are: m a = x, (Tr.m - T,) (12) k 6 L(Tr , -T,) = 1.86 x 10 W / m 2 9~ x,
- Indg me expression for x, in Equatbn @) and q.1 = 3.8 x 10' W / m 2 2 rearranging yields:
Therefore, the term containing the temperature
, pMTr m -Tj difference across the crust is much larger than the qg . (13) one containing the heat generation rate. The 2t temperature profile in the channel system ignoring energy generation in the debris is shown The heat flux (Reference 3) absorbed by the in Figure 19ED.41. Equation (4) can be channel wall can be approximated by that which a simplified to: semi infinite solid txxty can absorb. His flux is:
b= (7) q ~* , k p, - M p ha x, (Tr,, - T,) (14) di ynn,t Solving this equation with the initiu condition where: a, = thermal diffusivity of the wall that xc (t=0) = 0, rcveals: material. 2kr(Tr. -T,)t Equating (13) and (14) produces an equation x=i e (8) goveming the interface temperature. It is: 1 Pemh m f gi/2 T, - T, nkrPmh ,3 a " (yN This equation can be rearranged to de ermine the time required to freeze debris in a channel of yTr,, - T, t 2k,2 > (15) Amendmem 77 19ED 4 3
i 1 l AllWR 23 A6100A5 l Standard Plant nov.A p ( Solving this equation for T, using the quadratic formula yields:
-(c, - 2T,)! f(e, - 2T,)2 - 4(T,2 - e,Tr, )
T, = , (16) whe*e:c o = the square of the right hand side of Equation (15L Negative solutions of this equation are physically impossible. Fct a Tr,m of 1700 K and a T4 of 330 K, the interface temperature is 1560 K. Similarly, the interface temperature is 21"O K for Tr,m = 2500 K and T i = 330 K. Since this temperature is higher than the value for two semi infinite solid bodies coming into contact, the dominance of the temperature dif ference term in Equatiom (4) and (l 1) should be reverified. The heat-generation and temperature. difference terms for a interface temperature of 1560 K and channel half height of 0.5 centimeters are: k (T,,, -T,)- 8.4 x 10 5W / m 2 Q(D q = 3.8 x 10' W / m2 Esen though the dominance is not as great as before, the iemperature-difference term is still significantly greater than the heat-generation term and the assumptions made previously are still valid. O V Amendment 7i 19ED.4 4
ABWR m. m Standard Plant urv. A 19 EDA.3 Required Channel Length to where: v, = velocity at the entrance of the Insure l'reezing channel, g = gravitational acceleration The propagation rate of the frecie front was determined ,n Section 19ED.4.2. This allowed constant, determination of the time to completely freeze the debris in a channel of specified height. A simple or = height of detris in the lower approximation of the channel length, required to drywell. provide this residence time,is the product of the initial molten debris velocity and the freeting time. This Expanumg debris height yields: approximation would predict shield dimensions considerably larger than actually required. A more 2gm*t realistic channel length can be obtained by considering ve (t)= (18) the reduction in channel flow area as debris freezes. In EmA W the remainder of this section the following parameters will be determined: debris velocity at channel entrance, where; m. = maximum ejection rate of channel area decreaw resulting from debris freeting, corium from a failed vessel, average channel debris velocity, and the required c hannel length to insure plug formation at the channel entrance Au = fkor area of thelower drywell before corium ingression into the sump. (79m2 ), (1) Debris Velocity at Channel Entrance (2) Channel Area Decrease Resulting From Debris , ne possibihty exists that moller debris will not l even enter the channel after it has come into Since the entrance velocity is assumed to remains , contact with the shield wall. Debris which is constant, the mass flow rate of corium in the ! spreading across the lower drywell floor will have channel decreases in time due to the area reduction i at least a thin crust formed on its leading edge, if resulting from debris iceezing. A conceptual the flow energy of the adsancing debns front is p cture of this area reduction process is shown in roe great enough to break this crust and overcome Figure 19ED.4 2. Conservation of mass requires surface tension on the length scale of the channel that the mass flow rate of corium entering the height, debris vcill not enter the channel. channel per unit length is constant throughout the Unfortunately, the physics of crust formation is channel. %c mass flow rate at the entrance of the not currently understood well enough to support channel and at the location downstream where the I this argument without a great deal of uncenamty. debris front has just amved is: l l The entrance velocity will be governed by the height of corium outside of the channel. my = 6ve)HM 6YMH o (19) l Assuming that the debris spreads uniformly across the lower drywell floor, the height of debris can where: rh, = time varying mass flow rate per be obtained by integrating the volumetric unit width at the entrance of the expulsion rate of corium from the vessel divided channel, by the floor area of the lower drywell. A conservative overprediction of debris depth can be H, = time varying entrance flow obtained by multiplying the maximum expulsion height of the channel, rate by time and dividing by area. The upper bound of the expulsion rate was shown in Section vo = time varying velocity at the 19EB.6.2.2 to be 6000 kg/sec. downstream location in the channel where molten debris has De velocity in the channel without area reduction just arnved, due to debris freezing can be conservatively overpredicted by ignoring frictional effects, his Ho = unobstructed height of the velocuy is: channel. v,(t) = G2 gar (t) (17) His equation requires that: hendmem M 19ED 4-5
ABWR 23A6100AS Standard Plant ncv. A O O This is the average velocity of the molten debris v,(t) = v,(t)II,(t) (20) nio the shield channel. ( equ i ne ngd minsum Frecting The entrance flow height is: The channel length, required to ensure a plug li i(t) = 11. - 21,(t) (21) forms at the channel entrance before debris spills into the sumps, is: Inserting the relationship for x, found in Equation (8) into this expression yields: L r,,,,, = V( t r=. ) t r-. 9
= a,t r,,2,,
aA t,2,,, (26) 8k r(Tr'. - T,) ga li i(t) = 11 - i (22) poh. The product of this equation and the width of the shield channel desenbes the reduction of channel inlet flow area with time. (3) Average Channel Debrts Velocity The velocity of the leading edge of molten debris in the channel can be obtained by combining Equations (20) and (22). It is:
< s f 8k,(Tr , - T,)t i v,(t) = v,(t) 1- I (23,;
L 11, ph3 ( , The average velocitj of debris between the entrance of the channel and the leading edge of molten corium is:
' tv,(t)dt l V(t) = b (24) tdt 0
Evaluating this integral yields: V(t)= a E (25)
"11. t where:
,, , d 2 gm, 5 pA34 2 k ,(T,,, - T,)
l'm\ b, = 5 (j 3) p.h > Amendment ?? 19ED 4 6
- - . - . - . - = _ - . . . . _ _ - ~ . . . . _ _ . - - . - _ - - . . -- . - - - . .
i 1 ABWR 23A6)(OAS j Standard Plant aty. A 1 4 Table 19 E D.4 1 ) MATERIAL PROPERTIES OF A REPRESENTATIVE REFRACTORY BRICK j AND CONCRETE ) Property Representathe lirick Concrete i (Reference 6)
- j Melting Temperature (K) > 2200 1450 l '
i 4 Density (kg/m3) 2700 2300 Thennal Conductivity (W/niK) 4 1.3 l j Specific Ileat (JAgK) 1000 800 Thermal Diffusivity (m2/s) 1.48 x 10-6 7,$ x in 7 4 i O i O Amendment 77 19ED 4-7
- , - . - . ~ . - . __... .. _ ,._ - ._. - _ , - - . . - - , . . . - - _ , . , , . . . _ . . . , - . , . . _ . _ . . , . _ _ . _ . _ - . - , _ - . - - . -
ABWR DA6tmAs Standard Plant RLv. A O G Shield Wall (~ semi-infinite Molten Debris Crust Solid flody)
- anucuuuuuuuuuuu I. ;
- mmtmmmtmmmm i
- m m m m m m m m t::::
J
- M::M:::::N:::utt:M::
- m :: m m m itt m u u t m Muuu"ututunut ::ut R ;
- mmmmumumuutt
( a em:mmtmmmmum: a ammmmuummum: a ;ummmmmmumm: tin -. 3- ((r( 5 ammmmuummum:
)>, ( d ammumumummmt a
um mmummmmmt i 3( ;
- mmmmuummmm:
.:mumtmmmumum:
' 5 ;. v":mmmmm I '; mmt:
'ummtmtmmmt
.f.
1,,,.mummmum:m mmmmm:mum (~,* a m '"mmmm mtmm a am,e ""mmmmm: (b )a,:MuutuMn,21""'""' mmm Ti mmmmmmmmmm: fa c :mmtmtmmm:m:m: mmmmmm a
- mmmmmmm mm
- mmmmm mmmm:
i ututunatuu:untuu :: a auu:nn:ntununut n: ammmmmummm:e r ; ints:nnnnnnnnnn:: i Channel Freeze F,rontm, i Centerline Figure 19ED.41 TEMPERATURE PROFILE IN CilANNEL REGION r\
; 7 d
u Amendment 77 19ED 4 8 _ _ _ _ _ . -u
ABWR 23AsuoAs RLY, A Standard Plant ' l l i Channel Wall
> b Molten % Increasing Time Debris of & 0 H(t) ((,
Velocity y Freeze Front
> t U O
i i i Figure 19ED.4 2 CIIANNEL FLOW llEIGIIT REDUCTION DURING FREEZE PROCESS O
"* 19Ep 4 9 ,
i
1 i ABWR 23A6100AS , Standard Plant nix. A ; ,i 19ED.5 LONG TERM AHILITY OF L, 4**II,,U'~T') (#) DEHRIS TO REMAIN SOLID f Initial debris solidification was considered in where: ql, = steady state heat flux through i Section 19ED.4. The requirements for keeping the the upper shield wall. i debris in the channel froren for an extended period of i time (at least 24 hours) will te determined in thts I section. The height of the upper shield wall (above the II* = height of the upper wall, lower drywell floor) and depth of the lower shield wall Ti = temperature of the upper wall (below the lower drywell floor) will be specified.
- in contact with debris.
, 19ED.$.1 Upper Shield Wall (Above Tn = temperature of the tipper wan Lower Drywell Floor) in contact with the lower De roof of the upper shield wall should te free, or N**" ' "#" at least nearly so, of debris to provide long term Natural convection governs the temperature of the cooling to the debris frozen in the channel. No wall in contact with the lower drywell environment. significant amount of debris will splatter on the roof - The heat flux from the top of the wr.11 can te written dunng ejecu,on from the vessel because the sump is "8,' nest the periphery of the lower drywell. To prevent any debris from flowing ori top of the shield roof, the shield should be taller than the matimum possible debris pool q". = E(T. - Tw) (30) depth in the lower drywcil. This requirement is given i by: where: 5 = riaturalconvection heat transfer coefficient. i O 11,,2
- p. Aw...
(27) Ty = temperature of the lower drywellenvironment where: memu = maximum arnount of corium,
.235 Mg, The natitral convection heat transfer coefficient depends on the Rayleigh number.The Rayleigh number '
s Aia, min = minimum floor area of the lower drywell 79 m2,
" gD(T, - Tu )Ls, Evaluating this expression yields: va i
11,,2 0.33 m (28) where: Rat. = Rayleigh number, in the lcng term (at least minutes after debris = thermalexpansion coefficient i solidification), the lower drywell will be filled with of steam = 1/f at assuming ideal either saturated steam or water. lleat transfer from the gas behavior, shield to the environment is less effcctive when steam is present. Therefore, only steam will be considered in v = kinematic viscosity of 8 team, the remainder of this analysis. A shield wall sited to perform its function when steam is present will also a = taermal diffusivity of steam, perform its function when water fills the lower drywell. L, w characteristiclengthof the The maximum steam temperature in the lower si,ird top, drywell is that of saturated steam at the ultimate containment pressure (180 psig). The steady state heat The characteristic' length of a horizontal heated flux through the upper shield wallis: plate is one-half its width (Reference 4). De floor drain sump is approximately one meter wide; therefore, the O characteristic length of the shield roof is 0.5 meters, Evaluating the Rayleigh number for naturated steam at i Amendment 77 19ED.3 t
, .._ ___ - _ _ _ _ - . . ~, __ _ . _ _ . . . . - - - - - , . _ -._._w
ABWR 23A6tt0A5
!;tandard Plant RL:v. A
[ ( ultimate containment pre,sure (180 psig,190 C) containment pressure is 190 C. The height of the the yic kis: upper shield wall, which will transfer all of the heat generat:d in the channel for these conditions, is: Rat u 5 n10' K-'(T,- Tu) 02) 11** s 3 (39) For 107 s Rat s 1011, the Nusselt number 8.52 W / m K (Refercisce 4) for an upward f acing heated plate undergomg rutural convecuon L: . If the upper shield wall satisfies this inequality,it wtille capable of transferring all of the heat generated us by debris in the channel; and, as a result, guarantec Nut = 0.15 Rat (33) long tenn debris solidification even if the lower drywell has not been fkoded. To te acceptable, the height of The .verage natual cornection heat transfer the shield wall must satisfy the inequalities in cotflicient is: Equations (39) and (28). bk Nu 04) L. wr e: L = thennal conductivity of stearn in the k.wcr drywell. Combitan Eptc6 60) ed 02) through (34) yields: g q", = 8. 79(T. - T;)*#' W / m 2g4/s (33) w hich can le rearranged to: sua T, = T 4i ( 9 ",,* K 3 (36) (8.79 W / m3 , Inuning this into Equation (30) yields: a gs/4 q ", = Ti - Tw K OD g 9 2, This equation can M solved iteratively to determine the heat flux which can be tran=ferred through the upper wall of a given wall height. The wall height requirement for transferring a given heat flux is: r n $3/4 " 11', s b 9" K (38) q ", T' - Tu 8.79 W / m3 j s The decay heat levelin the ABWR 24 hciurs after l accident initiation is approximately 0.6%. The l volumetric heat generat on rate of debris at this time can be determined using Equation (6). It is p) (V 0.9 MW/m .3 The debris / wall interface temperature which will guarantee that the debris remains frozen is 1700 K. The temperature of saturated steam at ultimate l l Amendment 7? 19ED.5 2 _ __ _ i
AInVR 23A6)(OAS Standard Plant nix.A (O V 19ED.5.2 Lower Shield Wall (llelow lia = depth of the lower shield wall below the lower drywett floor. Lwer Drywell Floor) One side of the lowcr shield wallis in contact with The maximum temperature at each interface is debris and the other is in direct contact with the xhieved as t a . The maxiraum termatures at the bawmat. The basemat is constructed of concrete. A wall / debris interface, T.u and the wall /basemat conservative estimate of the lower shield wall depth can interface, Ti .S are: be made by assuming that concrete acts like a perfect insulator. Thus, no heat is allowed to pass from the q "* t q~ shield wall to the basemat. The boundary condition T. u = T,', + +
- H '" (41) between the debris and the wa'l is conservatively E *
- r . * "i* k, assumed to be constant heat flux. The initial burst of energy into the shield wall, caused by debris frecting, mi has ample time to distribute itself throughout the wall.
With these boundary conditions, the temperature q"t q"*H* distnbution in the lower wall can be determined T * = T"* + (42) analytically. P.C,, , H 3 6k, The analytical solution will provide a means for The heat flux through the lower wallis tounded by deterrninmg the time required for each of the interfaces one half of the heat Dux generated in the channel when to reach their allowable tempcnnure limits for a given the sites of the upper and lower wall are comparable. heat flux. The wall /basernat interface temperature lhe actual heat flux will be less because the upper wall sl.ould not exceeded the melting point of concrete is free to convect to the lower drywell environment and (1450 K). Continued debris solidification is guaranteed will accept rnore heat flux than the lower wall. The if the wall /basemat inte face temperature does not maximum heat generation in the charmel corresponds to exceed 1700 K. The wall will be sired so that the a decay heat level of one tercent. Since decay heat limits are not exceeded during the first 24 hours after decreases with time, using the maxirnum value bounds V(7 anitial debris solidificauon. The upper shield wall will be sired so that it can transfer the full decay he:r, load the temperature response of the lower shield wall. Using Equation (6), one half of the heat flux generated after 24 hours has elapsed, as discussed in Section in the channel la: 19ED.5.1, The temperature distribution in a slab q" ,,= = (q",)2"*"'s (43) (Reference 5), subjected to constant heat flux at one ' "b surface (x = Hw) and insulated at the other (x = 0) is: where; m. = total corium mass. T,,(ta)-T ,,. 9 I.' + i. Hi , ' The initial temperature of the shield wall should be 3, g P *
- P.
- l*
- adjusted to account for the energy it absorbs during the
' debns frecting process. If both shield walls have the 2
31' - Hi , ,2 (-1)" k ,n2*n't Ili2t . 1,, same thickness, the adjusted temperature is: 2,{,,2 2 3 , g 6 11 iw (40) Tu, = T, + "" (44) p ,cy ,H w where: T, i
= temperature distribuuon in the lower shield wall, Equations (41) through (44) can be used to determine if a chosen lower shield wall depth will Ti i. = adjusted initial temperature of satisfy the requirement of keeping the debris in the the shield wall, channel frczen for at least 24 hours. After 24 hours has elapsed, the upper shield wall will be able to remove the entire amount of heat generated in the channel 4 :,. = heat Oux through the lower (Section 19ED.5.1),
shield wall, ( v) c, p = specific heat of the shield wall, The process for determining an acceptable wall depth proceeds as follows. First a wall depth is chosen Amendment vt 19ED 5 3
ABWR ,m, ,, Standard I'lant nix.A
" ) which is comparable to the upper shield wall height.
Then, the adjmted initial temperature and heat loads are calculated using Equations (44) and (43), respxiively. The interface temperatures at 24 hours are determined by Equations (41) and (42). If Tod < 1700 K and T.g < 1450 K, the chosen depth is acceptable, if not, a new depth is chosen and the process repeated untti an acceptable depth is determined. An example of this procedure is gisen in part (4) of Section 19ED.6. i L1 O V Amendment 77 19ED 5- 4
ABWR 23Abl00AS Standard Plant arv. A 19ED,6 EXAMPLE CALCULATION temperatures of 1560 K and 2180 K for debris melting temperatures of 1700 K and 2500 K. The sizmg requirements for the floor drain corium respectively. The melting temperature of the shield were set forth in Sections 19ED.4 and 19ED.5 represernadve shield material is over 2200 K; based on a chosen channel height. An example siring therefore, it passes this test. exercise is presented in this section. The selected channel height,11,0 is one centimeter. Representative (2) Channel Length shield wall material properties are shown in Table 19ED.4 1. The equations, needed to determine the channel length required to ensure that a plug is fonned at (1) Melting Point of Shield Material Above Initial the entrance of the channel before debris spills Contact Temperature into the sump, are (9) and (26). These equations combine to gnc: De initial conta:t temperature between the debrts and the channel wall ghen in Equation (16) is: t r . , ,.is/2too . , e gtoo. 2 (45) 2
-(c, - 2Tj i (c, - 2 T - 4fT -c,T r whm 1,
(16) H 2p hm 8k r (Tr,, - T.)
, s a,=- E**"
nkrN hmu, 5 p,, A w ga ,, 2
,k*
^
O 5 2 k r(Tr,, - T,) De parameters required to evaluate this equation M p hm are: and Ho is the assumed channel height (0.0lm). Ti = initial temperature of the shield wall, 330 K. The maximum length results when Tr,m = 1700 K. The contact temperature was Tr,m = debris freezing temperature shown previously to be 1560 K for this freezing ranges trom 1700 K to 2500 K, temperature. The other parameters required to evaluate these equations are: kr = debris thermal conductivity, 30 W/m2 K, rh,. = ejection rate of debris from a failed vessel,6000 kg/sec
- p. = density of corium,9000 kg/m3 , (conservative maximum),
blh = debris latent heat of fusion, A ld.rmn = minimum floor area of the 2,7 x 105 J/kg, lower drywell.specified as 0.02 m2/ MWth in the EPRI
- a. = thermal diffusivity of the shield ALWR Requirements wall material, a representative Document;it is equal to 79 m 2, value of 1.48 x 10-6 m2/ rec will be used.
Using these parameters, the plug formation time k, = thermalconducthity of shield is 7.2 seconds and the reydred channel length is wall material, a representative 1.06 meters. This length was determined using a value of 4 W/mK will be used. highly conservative corium discharge rate. The q analysis assumed a constant discharge rate equal to Q Evaluating Equation (16) yields interface maximum discharge rate predicted using a highly conservative model. De actual discharge rate will Amenanent 77 19ED 6-1 J
ABWR 23A6!00A5 Standard Plant REY.A be lower, if the length requirement is highi De wall /dettis, T./d. and the wall /basernat T,r,, restrictive, the discharge rate could te refined witfi interface temperatures are given by; additional effort. (3) Shield lleight, ilu,, Above Lower Drywell 11oor T.f 4 = Tu. + p g
+ (4l}
k De height requirements for the upper shield wall are given in Equations (28) and (39). These ad equauons are: 4 H,,2 0.33 m (28) T,3 = uT . + p c ,.*fl6k, p i
,-'"'" (42) and- where: I = time asumed to te 24 hours. l k, Evaluating these expressions yicids g"* g 2 (39) T.a = 8190 K and T,fd = 820 K. Since these 8.52 W / m K temperatures meet the requirernents for long term For a wall conductivity of 4 W/mK, these debris solidification (T,ld < 1700 K and inequalities require: T,fd < 1450 F), the ch;an wall depth is acceptable.
0.33 m s 11,, s 0.47 m (46) (5) Summary of Snield Requirements A height of 0.4 meters is chosen. A proposed floor drain sump corium shield with a (4) Shield Depth, Hi., Below Lower Drywell Floor specified channel height of one centimeter and wall material properties shown in Table 19ED.4 1 p wiH prevent corium ingression into the sump if it h De lower shield wall should be sized according to Equations (41) through (44). An initial height of rnects t f H wing regements. 0.4 meters is chosen to begin the determination of hiinimum melting point of shield material: acceptability. The adjusted initial temperature of the lower shield wall accounting for energy 2180 K' absorption during debris freezing is: Channel Length: 1.% m. b' Height above lower drywell floor: 0.4 m, Tu. = T i + [p".c
. p
,,"Hi
, (44)
= 341 K Depth below lower drywell floor: 0.4 m.
where: p, = density of wall material, cp ., = specific heat of wall material. The limiting heat flux through the lower wall is:
, Q a4P=H.
'"uuna 2m. (47) 2
= 7520 W / m mhere: Q3a = 1% of decay heat, 39.26 Mw; l m. = mass of corium,235 Mg.
l Amendment ?? 19ED 6 2 f u
ABWR 23 A6)(0AS Standard Plant REV. A C
\ 19ED.7 DETAILED DESIGN ISSUES i During detailed design of the ABWR. the exact j g shield material and shield dimensions will be chosen.
;, Cursory examination of material properties indicate T alumina may be an acceptable wall material. The 3, requirements for the shield are stated in Section 19ED.3. Example calculations of the requirements are shewn in Section 19ED.6. Interference with under-vessel servicing equipment will be considered in detennining if the proposed dimensions are acceptable; if not, the sizing process will be redonc for a new channel height and/or a new shield wall material. The number and width of channels in the shield will be chosen to meet the design requirements for water flow into the sump dunng nonnal ABWR plant operation.
l [ ( Amendment 77 19E D.7. I
P ABWR 23A6100A5 Standard Plant uv. A 19ED.8 REFERENCES
- 1. Frank P. Incropera and David P. DeWitt, fundamentals of fleat and Afass Transfer,2d Ed.,
John Wiley and Sons,1985, pp. 85 86.
- 2. Glen E. Myers, Analytical Afethods in Conduction llear Transfer, Genium Publishing Corp.,
Schenectady, NY,1987, p. 202,
- 3. Frank P. Inciopera and David P. DeWitt, Fundamentals of fleat and Atass Transfer,2nd Ed.,
John Wiley and Sons,1985, p. 203. 4 Frank P. Incropera and David P. DeWitt, Fundamentals offleat and Afass Tranger,2nd Ed., John Wiley and Sons,1985, pp. 433-435.
$. 11.S. Carslaw and 1.C. ]eaget, Conduction ofIleat in Solids,2nd Ed., Oxford University Press,1959, pp,112113.
- 6. Afark's Standard llandbook for Afechanical Engineers,6th Ed., Theodore Baumeister, Editor in Chief, McGraw liill Book Company,1978, pp. 6-171 to 6177.
O
/3 m)
Amendment 77 19ED 81
ABWR 23A61004S Standard Plant ntv. A U 19EE.1 SUPPRESSION POOL llYPASS g As shown in Subsection 19E.2.3.3.3(4), the only mode of suppression pool bypass that presents any significant risk during a sesere accident is vacuum breaker leakage. Vacuum breaker leakage is the passage of gas from the drywell into the wetwell air space. Vapor suppression and fission product scrubbing by the suppression pool are not available to the gas and vapor w hich passes through the vacuum breakers. m The ABWR contains eight vacuum breakats.
- p. ABWR vacuum breakers are swing check valves
. designed to open passively when wetwell pressure exceeds drywell pressure by 0.0035 MPe (0.5 psid).
? When the pressure differential is less than this, or h drywcli pressure exceeds wetwell pressure, the vacuum
{ breakers should be completely scated and no flow v r.hould be passirig through them. A large pressure U differential will produce a large force tending to close T het vacuum breaker valves. A pressure differential of
+0.048 MPa (+7 psid) is typical in a severe accident after core damage cccurs and the passive flooder opens.
This pressure differential produces a closing force of 9810 N (2200 lbf) on the valves. For severe accident (] o scenanos in which the firewater system is actuated, the V x pressure differential is about +0.096 MPa (+14 psid) which produces a closing force of 19600 N (4400 lbf) 3 on the valves. These ,large closing forces, as well as 1 rouu.ne inspection, maintenance, and testing, ensure the y probability of vacuum breaker leakage after the C actuation of the passive flooder or the drywell spray system is extremely low, m Large amounts of leakage can occur as a result of m catastrophic failure of valve components or a valve
< sticking open. Lesser amounts of leakage can result I . from normal wear and tear including degradation of the
$ valve seating surfaces or retaining magnets. For h sufticiently large amounts of leakage during a severe 3 accident, the time to rupture disk opening or U
l ) containment failure can be reduced and the amount of fission products released can be increased. l A study utiliaing decomposition event trees and deterministic modeling was performed to access the impact of vacuum breakes leakage on the performance j of the ABWR during a severe accident. The event tree ! analysis is contained in Section 19EE.2, Section 19EE,3 contains the deterministic evaluation. l l t [ ( q l Arnendmen 'n 19EE 1 1
i ABWR 23A61n0A5 Standard Plant nEv.4
, 19EE.2 DESCRIPTION OF Information about the valves connecting the l DECOMPOSITION EVENT TREE containment and reactor building were not included Aggyggg because some of these valves are not swing, check i
- valves. The database query provided a short narrative of
'# '" " " * * ' " *
- I I*' # *P "##
i The suppression pool bypass decomposition event g$#, $ tree analysis consists of one decompoution event tree (DET), Figure 19EE.2 1. The DET considers the major ' De database query included BWR Mark 1,11 and !!! d phenomena which influence accident consequences. The 6 ' containments. The vacuum breakers in these first two events on the DET sort out vacuum breaker containments are similar in design to the ABWR
- leakage area. Plugging of vacuum breaker leakage vacuum breakers (passive, flapper type valves attached Y-d pathways by aerosols is considered in the third event. If to hotirontal piping). The ABWR vacuum breakers 9 m leakage exists but the pathway is not very large' will be slightly different in site than some of those 3 x acrosol plugging can significantly diminish the currently in operation, but this does not undermine the "
consequences of suppression pool bypass through the appliduity & h vactum breakers. We last event assesses the amount of o J suppression pool bypass. the failures were culled to exclude failures other than those that could lead to a vacuum treaker sticking M The probabilities for each sequence pathway with pen r catastrophically failing. Failures to open were n. b similar end States were summed arid these results
" #*C """ *# #" " ## *I transferred as the branch probabilities of the rnain cauw. Most faHurcs to open 00 out of 12) were 7 containment event tree, 4 attributed to either the setpomt drift or worn retaining e magnets. Neither of these conditions would prevent the $
19EE.2.1 %,acuum Breaker Stuck Open vacuum breaker from closing once it had open, albeit at u l g (VB) a differenthl pressure outside the normal range. The v remaining failures were due to: 1) a loose set screw on When a vacuum breaker sticks open or the flapper pivot pin and 2) excessive clearance between 2[fcatastrophicaPy fails, a large pathway is established the valve shaft and di:.k. Both of these conditions led to 3 between the drywell and wetwell. The deterministic opening forces greater than technical specification
- analysis desenbed in Section 19EE.3 demonstretes that limits and greater than the forces required to open the
] pr.thway areas greater than 41 cm2 (opening widths other vacuum breakers tested in the same sequence. In I greater than 0.9 cm) can significantly affect accident the ABWR design, the depressurization transients consequences. which lead to opening of the vacuum breakers are very mild. Therefore, if either of the these two failure The suppression pool bypass scoping analysis conditions existed during an accident, the affected valves presented in Section 19E.2.3.3 assumed a failure would probably not open because the other vacuum probah!!ity for vacuum breaker full reverser iow of breakers would open and relieve high differential 6.7E.2/ demand based on pre.1970 U.S. BWR operating wetwell pressure before the force required to open the birtory of general check valves. This failure rate is affected valves was achieved.
highly conservative because: Failures to pass leak rate tests during refueling and (1) The ABWR vacuum breaker design is based on maintenance outages when the vacuum breaker current knowledge which is subs 4ntially proximity switch indicated " closed" were also excluded improved over earher check valve designs. because they represent small leakage paths. These 7 fauures were included in the probability for VB_ LEAK % (2) The ABWR vacuum breaker environment is as described in 19EE.2.2 A " closed" indication will be Y significantly less severe than general check valves given only when the vacuum breaker disk is seated or 4
. the working fluid is gas rather than liquid and very nearly so. Failures to close were included, as were the ABWR vacuum breakers will not experience cases m which excessive force was required to cycle a ;.;
chugging loads. vacuum breaker during stroke capability testing.
]
k The failure probability used in this analysis was 3 based on BWR operating experience from April 1981 to The database query provided the following results: ( 4 March 1991 as contained in a database of Licensing Abnormal operation which ; ig Event Reports. The database was queried for abnormal could lead to failure to close: 18 (Nam). i bj wetwell to drywell vacuum breaker operation. 1 1 AmeMment ?? 19EE.21
- - - , + , - , , - - - - _ _ , - . - - . - . - - - ..- _ _ , _ _ - .,-.- _ _ _ _ __J :
ABWR 23A6100AS Standard Plant Rev. A n h Cumulative vacuc n breaker operating time: 2.66E7 hours (Teio..). The ability of vacuum breakers to open and close in current plants is demonstrated monthly during stroke capability tests (T,ude = 720 hours). Therefore, the probability that one of the eight ABWR vacuum , breakers will fail to close on demand and a large leakage e path will be established between the wetwell and c., drywell can be approximated by: Y
$ P(VB) = '"it-g,p T% (1)
U = 3.f E - 3 / demand This failure probability conservatively over-estimates the probability that one of the ABWR vacuum breakers will fail to close during accident conditions because the closure forces during an xcident will be at least an order of magnitude greater than those present during testing and normal operation. Additional closure force will enhance sealing and overcome some, if not all closing resistance.
, The vacuum breakers in the ABWR will not be (m) suoke tested every month as are those in current %) operation. This is expected to improve vacuum breaker reliability because the monthly stroking increases wear, increases galling potential, imparts impact loads to the valve components, loads the valves in a non uniform manner, and decreases the sealing ability of the soft seats. Reliability will also be increased by improvements made possible by the operational experience of vacuum breakers currently in BWRs with Mark I,11 and III containments. These improvements will include material selection, valve assembly techniques and maintenance procedures. Corrosion on ABWR vacuum breaker components will be negligible because of material selection and cperating environment (nearly pure nitmgen). Since reliability is improved and corrosion will be negligible, the failure probability determined during monthly testing of current vacuum breakers provides a conservative over estimation of ABWR vacuum breaker reliability.
f
\ l %)
Amendment M 19EE.2 2 I
ABWR 23A6100AS Standard Plant nsv. A to 9 19EE.2.2 Vacuum Breaker Leaks probability, The database query provided the fonowing . results:
) (VB_ LEAK)-
The consequences of small leakage paths between Number of Mark I wetwell-to drywell vacuum breaker h the drywell and wetwell are less severe than those for a -abnormal operations which could lead to small d vacuum breaker sticking open. The small leakage arca _ leakage: 42 (N ic.k) cutoff was determined to be 41 cm2 in the sensitivity study contained in Section 19EE.3. The BWR- Cumulative Mark I vacuum breaker operating time: operating history described in the previous section 2.37E7 (19EE.2.1) was also used to determine the probability - boitrs (Ti ,g). of smallleakage, ne actual amount of leakage was not reported in - BWRs with Mark I containments have a single the database and is generally not available. However, passive, flapper type valve attached to the end of each the vacuum breaker leakage area can be roughly - vacuum breaker line. Mark 11 containments have two characterized. Currently, wetwell to-drywell vacuum passive, flapper type valves in series in each vacuum breakers are verified closed by indication lights in the breaker line. Mark Ill containments have a single, control room every reven days. Position is determined flapper type valve in series with n motor operated valve by proximity switches which are generally accurate to (MOV) in each line. All of the valves are attached to within the 0.9 cm disk opening which corresponds to horizontal piping in the wetwell air space. Since the the 41 cm2 cutoff area. De proximity switches used in ABWR has a single, flapper-type valve on the end of conjunction with the ABWR vacuum breakers will each line in the wetwell air space, the operating have even closer tolerances because of the increased experience of BWR's with Mark I containments importance placed on bypass leakage. None of the provides the best indication of ABWR vacuum breaker leakage failures included failure of " closed" indication. leakage. Actual ABWR vacuum breakers will perform Berefore, leakage was occurring when the valve was better than those in Mark I containments because: 1) open less than the cutoff amount.- the ABWR vacuum breaker materials-especially those of the seating surfaces-will be improved because they During the operating period selected in the database , will be based on the many years accumulated vacuum query, refueling and maintenance outages were breaker experience of current BWRs,2) the ABWR conducted every twelve to eighteen months. Thus, vacuum breakers will not experience chugging loads, taking the test time to be eighteen months (T ten = and 3) the ABWR vacuum breakers will not be cycled 13,140 hours) is conservative. The probability that one every month. of the eight .ABWR vacuum breakers develops a small leakage path can be approximated by: The ability oi vacuum breakers to remain leak tight is demonstrated during wetwell to-drywell leakage r 4 tests performed as part of each refueling and P(VB_ LEAK) = 1 - 1- '" Ta
~
maintenance outage. During these tests, the drywell is ( i , (2) pressurized with respect to the wetwell and the pressure
= 0.17 / demand decay rate measured. If the pressum diffemntial decreases too rapidly indicating excessive leakage, the root cause is found and corrected, ne instances when a vacuum The value used in the quantification of the breaker was found to be the leakage pathway are containment. event trees is 0.18/ demand. The difference reported in Licensing Event Reports and included in the between this value and the value calculated above, operating experience database, ne pressurization rate using a more detailed analysis, results in a slight used in the leakage tests are generally slower than those c nservausm m the analysis, experienced during accident conditions. Increased pressurization rates improve the scaling capability of his Probability is a conservative over-estimation soft seats and reduce leakage, since wetwell-to-drywellleakage test are conducted at differential pressures much lower than those expected All failures reported in the selected operating during accidant conditions. The additional differential history of wetwell-to-drywell vacuum breakers in- Pressures win greatly enhance seahng.
Mark I containments except failures to open and those used to determine vacuum breaker stuck open were included in the determination of small leakage Amendment ?? 19EE.2 3
ABWRi 23^6100AS - n Standard Plant REY.A k 719EE.2.3 Aerosols Plug Leakage Path y (LEAK , PLUG) o E Re consequences of leakage pathways between the
,3 drywell and wetwell can be greatly diminished if aerosols plug the path. %e Vaughan aerosol plugging model(Reference 1) was used with MAAP ABWR to determine if and at what time plugging occurred. A full description of this methodology can be found.in Subsection 19EE3.1.
The sensitivity study contained in 19EE.3.2 predicts that if plugging is allowed to occur in small Icakage paths (opening widths s 0.9 cm), accident consequences are not effected by the presence ofleakage paths. Even; though ' plugging may _ reduce the-consequences of it.rger opeisng widths, no credit was taken in the DET. The sensitivity study predicted plugging for opening widths up to 1.63 cm. Herefore, high probability, 0.9, was given to plugging of ~ opening widths up to 0.9 cm.
-J
- p 1
J j i Amendment ?? ' 19EE.2-4 b-
. _ . _ _. _ _ _ . . . _ . . _ . _ _ . . . _ . . - . _ . . __ . _ . _ ..~ _ ._. . . ..
~
-ABWR 23 A6100A5 --
T Standard Plant' REV.A 19EE.2.4 Suppression Pool Bypass
. (POOL _BP)
? -
, 3 This heading on the DET summartzes the amount
, D of suppression pool: bypass, "No Pool Bypass" d . Indicates that either no leakage, an insignificant amount of leakage, or a plugged leakage pathway exists The
, consequences of a particular accident scenario will be unaffected by pool bypass for this condition. "Small Leak" indicates that a small amount of pool bypass is . - present. Small amounts of bypass will have marginal impact on accident consequences, Large amounts of.
pool bypass are indicated by "Large leakage". Accident consequences will increase in severity when large . amounts of pool bypass etist. i j -. 4 i I-I' 1:
. - _ , 19EE.2 5 s
, , , . . . , , - .y , - . . -.,.,w w- ., v i.. e r - - - - -
----.m---r- y up ,- w w . --E
J 4 i 3 ABWR 23A6100AS i Standard Plant REY,A a a i . u.. a s .. w. . - t j .. c.c 1 .. .. .uco unu i .oa i l i ' 4 1 t i , e u .. .m , 4 M I j 4 1 i n 4 1 ...m, t - i
- .* <en , ,
i j 'Hi"= i l ,, u u.. .- .. m. o i t T i ve l'uot opf4 L.m Lt as , GS 39 1-1-
- , O Figure 19EE.21
- CONTAINMENT EVENT EVALUATION DET FOR SUPPRESSION POOL BYPASS Amendtnent ?? 19EE.2 6
- -, . . _ . _ . _ . . - . ..;,.--__,.-.~m._4
i . 7 1-l ABWR 2mims
- ' Standard Plant REY. A ,
i t4 4 s DETERMINISTIC d 19EE.3 scenarios 3 and 4 used a conservative proportionality L 4 ANALYSIS wastant of 50.000 kg/m3 (3115 lbm/ft ). I Although the Vaughan aerosol plugging model-A senritivity study was performed with MAAP. , d es not suggest an upper bound on the size of leak. ! 2 p ths which can be plugged, there is some question l d' ABWR bypass during tosevere access the accident impact of conditions, suppression pool about the applicability of the model for leak paths i greater than -I cm (0.39 in) in diameter. In 19EE.3.1 Method NRC/IDCOR Technical Issue 13A (Reference 2), the W NRC asserted that the data cited by Morewitz 4
- The dominant severe accident sequence [ Loss of all (Reference 3) in support of the Vaughan plugging g core Cooling with vessel failure occurnng at Low mcdel for pathways greater than I cm diameter does not
!- g Pressure (LCLP)] was chosen to evaluate plant adequately simulate severe accident conditions. The ! experiments cited with pathways greater than I cm r M, effective performance. MAAbAaWR vacuum breaker runs area, A/VK, were varying frommade 0 to with(0.39 in) in diameter involved straight ducts with ' 2 4 2030 cm (315 in ).2The upper bound corresponds to lengths greater than 10 meters (32.8 ft). Therefore, due 5 one fully open vacuum breaker. Five variations were to the Ixk of appropriate experimental data, the NRC w analyzed. In each case the overpressure relief rupture has accepted the Vaughan aerosol plugging model only V disk opened when the welwell pressure reached 0.72 lor leak pathways smaller than I cm (0.39 in). The l MPa (90 psig). 'lhe five scenarios were: NRC's position on this issue is stated in the resolution of NRC/IDCOR Technical Issue 13 A (Reference 2). (1) Bypass leakage begins after passive floc ler acuvation, aerosol ptugging is neglected; . In order to accurately simulate aerosol flow 4 through open vacuum breaker valves in the ABWR, (2) Bypass leakage is present from the beginning of experiments should be conducied with ducts of less than the accident, aero;ol plugging is neglected; 2 cm (0,79 in) in length. However, the trends of the j experimental data do not suggest that the Vaughan l (3) Bypass leakage begins after passive flooder plugging raodel is invalid for openings only slightly a:tivation, aerosol plugging of the vacuum target -than I cm. Unfortunately, r o definitive tuaker opemng is considered; cenclusions car be reached regarding 15e applicability limit without aduidonal experimental data. For this (4) Bypass leakage is present from the beginning of reason, studies were performed with and without the accident, aerosol plugging of the vacuum plugging for vacuum breaker bypass widths up to 1.6 breaker opening is considered; cm (0.63 in) corresponding to an effective area of 75 cm2 (11,6 in2). This information is used to (5) Bypass leakage is present from the beginning of indicated the conservatisms which may exist in the i the accident and the operator initiates the firewater analysis. spray system. , The opening of a stuck-open vacuum breaker is l MAAP-ABWR uses the MAAP3.0B aerosol neither circular nor rectangular. Rather it is a crescent plugging model developed by E.U. Vaughan - shape formed by two circular disks sc;arating while (Reference 1). The model predicts the mass of aerosol remaining hinged at one point. The !*ak path width required to flow through the leak path in order to form a used for the Vaughan plugging modelis conservatively plug as a function of the size of the opening. MAAP assumed to be the maximum crack width. The length of I conserynively assumes that the flow rate through the opening is approximated as the effective area divided by vacuum breaker opening is not affected by the growing the width. For vacuum breaker opening widths of up to aerosol plug until the aerosol mass required to plug the 1 cm (0.39 in), corresponding to bypass effective areas leak completely has passed through the opening. For a of up to 46 cm2 (6.45 in2), use of the plugging model circular opening, the mass is proportional to the cube provides the best estimate of containment response. As of the diameter; and, for a rectangular opening, the discussed above, additional calculations were run for mass is proportional to the product of the length and widths up to 1.6 cm (0.63 in). l the square of its width. The proportionality constant i has 10,000 been experimentall{ to 50,000 kg/m 623 to (determined 3!!5 lbm/ft 3), and to range from
, varies with aerosol size, aerosol mass flow rate, and leak path geometry. The MAAP-ABWR runs for Amendment 79 19EE.3 1
'ABWR 23A6100AS Standard Plant -
sun. A
\ Results time continued to decrease to a value of 2.2 hours for a m 19EE.3.2 -
- ; fully open vacuum breaker valve.
5 ' A series of bypass flow areas was analyzed using The 24- and 72-hout Csl release fractions 4 MAAP ABWR for each of the assumed scenarios. A summary of the time and magnitude of fission product asymptotically approached a maximum value for large d releases for each scenario is presented in Table effective areas. De Csl release fractions for the scenario 4 19EE.31. It was not necessary to run all of the 2 cases are very similar to those for the cases of scenario 1, The variations in release are caused by
$ variations in bypass area for each of the five scenaries changes in revaporization behavior due the slight
'q for this analysis.Thus, Table 19EE.31 contains some differences in thermal hydraulic performance, blanks. The characteristics of each scenario is discussed below.
19EE.3.2.3 Late Suppression Pool Bypass 19EE.3.2.1 Late Suppression Pool Bypass with Plugging with no Plugging For the scenario 3 cases, bypass leakage was For the scenario I accident sequence, the passive assumed to begin after the actuation of the passive flooder opent [ based on the gas temperature in the flooder. Plugging of the vacuum treaker opening before lower drywell reaching 533 K (500 F)] at 5.5 hours. the wetwell pressure reached the rupture disk setpoint The pressure in the drywell decreases as cold water was predicted for all cases analyzed. After the leak floods into the suppression pool from the lower plugs, all flow from the drywell is directed through the drywell. Fifteen minutes later, the drywell starts to drywell connecting vents into the suppression pool, repressurize and the suppression pool bypass is There is then a period in which little steam is generated presumed to begin. If there is no bypass leakage, the m the welwell vapor space. The wetwell gas elapsed time before rupture disk opening and fission temperature decreases during this time due to product release is about 20 hours. MAAP predicts that condensation on the walls. This in turn causes the the time to rupture disk opening is not affected for containment pressure to decrease for a short time. effective vacuum breaker bypass areas of up to 5 cm2 Steam generation in the drywell eventually causes the (j~N g 2 (0.78 in ). As the effective area increases from 5 to 50 suppression pool to heat up and the containment cm2 (0.775 to 7.75 in2 ), the time to rupture disk pressure increases again. For cases with vacuum breaker opening steadily decreases to about 10 hours. Above 50 opening widths up to I cm (0.39 in), the elapsed time cm: (7.75 in 2), the time asymptotically approaches 9 to rupture disk actuation is about 20 hours, the same as hours and remains at 9 hours even for a fully open for the case with no bypass leakage. MAAP-ABWR vacuum breaker valve, predicts Csl releases of less than IE 7 at 72 hours for all of the opening widths less than I cm. As expected, fission product releases are much higher for cases with bypass leakage than for the case ne maximum vacuum breaker opening width for without bypass leakage. For non-bypass cases, the which MAAP predicts that the leak path will plug release fraction of Csl at 72 hours is less than IE 7. before the rupture disk opens was determined to be 1.25 - The release fractions of Csl at 24 and 72 hours em (0.49 in). Even if the rupture disk opens before an approach asymptotes as the effective bypass area aerosol plug forms, reductions in source term can be increases. For cases with effective areas greater than observed. After the rupture disk opens, acrosols will 400 cm2,the 24 hour Cs! release fractions are about continue to flow through the vacuum breaker opening 6% and the 72. hour release fractions are about 17%. and can eventually form a plug. This essentially Most of the releases occur late in the sequences as terminates fission product release. The Cs! release fission products revaporize from the vessel surfaces. fractions at 72 hours for cases with late bypass and credit for acrosol plugging are significantly less than 19 EE.3.2.2 Pre existing Suppression Pool for the cases in which no plugging is assumed. Bypass with no Plugging 19 EE.3.2.4 Pre existing Suppression Pool Bypass leakage was assumed to be present from the Bypass with Plugging beginning of the accident sequence for the cases in scenario 2. As with the scenario I cases, the elapsed . The scenario 4 cases, in which suppression pool ume before rupture disk opening is not affected by bypass flow was present from the beginning of the effective bypass areas smaller than 5 cm 2 (0.775 in2). accident, show similar results to those of the scenario 3 Unlike the scenario 1 cases, however, the elapsed time cases. For craes with vacuum breaker opening widths did not reach a 9-hour asymptote. Instead, the clapsed up to 0.9 cm (0.35 in), the bypass leak plugged before Amendment ?? 19EE.3-2
ABWR 23^6:oors Standard Plant REV.A (3 the rupture disk opened and the clapsed time la fission
'Q product release was the same as the case with no bypass (about 20 hours). Also, the fission product release for these cases at 72 hours was less than IE 7, as in the case with no bypass.
The case with an effective bypass area of 46 cm 2 (7.13 in2), opening width of I cm (0.39 in), exhibited a different response. The mass of aerosol passing through the opening was not sufficient to plug the leak before the wetwell pressure reached 0.72 MPa (90 psig) and the rupture disk opened. However, the leak did plug about 30 minutes after the rupture disk opened which reduced the amount of fission products that was released to the environment. MAAP predicts a Csl release fraction of OD4% at 72 hours for this case, which is about two orders of magnitude less than the corresponding case in which no plugging is assumed. The same behavior was observed for the slightly larger 50 cm2case. 19EE.3.2.5 Suppression Pool Bypass with Drywell Spray The last scenario examined the effects of the drywell spray on cases with bypass leakage present from the beginning of the accident. The firewater O addition system was used for these cases since its flowrate is smaller than the drywell spray function of I the RHR system. Assuming the operator initiates the firewater spray within 2 hours of the start of the accident, the elapsed time to rupture disk opening can be delayed to nearly 30 hours. This time is comparable i to the base case, LCLP-FS-R-N, with no bypass leakage (Subsection 19E.2.2.1). The fission product releases for all bypass areas analyzed are on the same order of magnitude as the releases for the cases of scenarios 1 and 2 (with no plugging or firewater addition), but the clapsed time to l release is much longer. The long times to release allow I for a great deal of fission product decay which leads to a substantial reduction in risk as compared to cases in l which the drywell spray is not actuated. i l
)
LJ Amendmen? ?? 19EE3 3
-= - .... , .
ABWR. -
,3yi m REV,A
-7 Standard Plant O N V A 19EE.3.3 - Conclusions of Deterministic d Analysis M Suppression pool bypass can lead to a significant
@ increase in fission product release. Releases can be on .
o the order of 10% for a fully stuck-open vacuum breaker. , For sequences in which the firewater addition system is ' used in spray mode, the time to release is not significantly affected. However, for sequences without sprays, the time from the beginning of the accident : until the onset of the release can be significantly reduced. The use of the Morowitz blockage model results in a significant improvement in the calculated . risk associated with- suppression pool bypass, Nonetheless, there is a_ substantial increase in consequences associated with large bypass areas. Therefore, suppression pool bypass is examined with a decomposition event tree analysis in Section 19EE.2. l 4 l . s
. Amendment ?? 19EE.3-4
. -- .u . _ . . . _ . - . . , ,. c__ , _, tu
J ABWR 23x iooxs Standard Plant REY.A i O V Table 19EE.3-1
SUMMARY
OF VOLATILE FISSION PRODUCT RELEASES FOR SEVERE
- ACCIDENTS WITil SUPPRESSION POOL BYPASS LEAKAGE TIIROUGli VACUUM BREAKER VALVES Eff. Area (cm^2) 0 5 20 41 46 50 58 75 100 400 2030 0.44 0.90 1.00 1.09 1.25 1.63 2.17 8.70 *"
leak Width (cm) 0 0.11 Scenario Time to Fission Product Release (hrs) 15,4 "* *" 9.9 *" 9.1 9.1 9.0 9.0 1 19.9 19.8 1 20.0 *** "* 5.5 "" 4.0 3.5 2,7 2.2 2 19.9 13.1 20.2 *" "* 20.3 20.4 9.2 * *
- 3 19.9 20.2 4 20.2 20.2 20.4 5.9 5.6 *" *" * *
- 19.9
'" *" "* "* 29,7 *" *" *" "* 28.9 5 31.1 Scenario Csl Release Fraction at 72 hours I < IE 7 0.38 % 1.6% *" *** 3.6% "* 6.3% 8.5% 18 % 17 %
h v 2 < IE 7 0.55 % 1.7% *" 4.2% "* 6.5% 8.5% 16 % 18 % j 3 < IE 7 < IE-7 < IE-7 < IE 7 < IE-7 0.06 % * *
- 4 < IE 7 < IE 7 < IE-7 < IE-7 0.Gl% 0.06% '" "* * *
- 5 <lE-7 *" *" *" "* 4.8% *" "* *" *** 14 %
- Plugging presumed to b ineffective
*" Not calculated b
V l Amendment 71 19EE.3-5 i I
. ._ _ . _ . . . . . _ _ . . _ _ _. .. . - . .._ _ _ ..__ _ _ m_.. .. . .
i , ABWR- 23A6100AS . Standard Plant . REY. A - Ik i 519EE.4
?
SUMMARY
OF RESULTS l
*g l9EE.4.1 Quantification of DET
. 4 . l The quantified event. tree is shown' in-l ^ Figure 19EE.21. The probabilities for different ! A leakage areas are transferred to containment event trees.-
- g The probabilities are listed below
I j d No Leakage 0.9782,. i i SmallLeakage 0.0179, f i- Large Leakage -0.0039. . i 1 i i l f e i-i i I i i i l l-i l 1 i i l l I 1 i . 1
.i
- j. . Amendmem 77 19EE.4-l ' l 1
e :
- . .t
. y
- ABWR As 2'^'[y 3
- Standard Plant l- .
et 19EE.4.2 c Impact of Release Fractions
-l d MAAP ABWR predicts the release fraction of Csl
"? for the LCLP case without bypass leakage is less than l
'
- IE 7. 'Ihe effect of leakage on the Csl release fraction x (f)i.s :hown below.
i es g AmountofI akape Rele2e Fraction of Csl a None f < IE 7 Small 1% < f < 10% L1rge f > 10% l Amendment 77 - 39EE.4 2 l
- .a-.- - . ,. ._ , - , . - . - . . .,,,-,..:
. 1 i i ABWR' 23A6100AS Standard Plant , REv.A l 9 1 2 19EE 4.3 Impact on Time to Rupture 4 Disk Opening x ); x i The sensitivity study contained in Section 19EE,3 focused on the Loss of all core Cooling with vessel
@ failure occurring at Low Pressure (LCLP) accident
(' sequence. This is the dominant sequence and its respr>nse to suppression pool bypass should be typical -- ,- of the other accident sequences. Without suppression pool bypass, rupture disk opening is predicted to occur at -20 hours into the , accident for cases with passive flooder operation.-The effect of leakage on time to rupture disk opening, t, is summarized below. Amount of I anknoe - Time to Ruoture Disk Onenina None ~20 hours Small 6 < t < 16 hours Large ' t < 6 hours
.t .
b
%/
' Amerdenent?? 19EE.4 3 4_: - -- _ _ . _ . , _ _ , _ . . _ _
__v.
c ABWR 23A6100AS Standard Plant nv. A im (V\ 19EE.5 CONCLUSIONS Suppression pool bypass (the passage of gas and vapor from the drywell directly into the wetwell air spre) can lead to increased fission product releases. As shown in Subsection 19E.2.3.3.3(4), the only mode of suppression pool bypass that has the possibility of sigmficantly increasing risk is vacuum breaker leakage. This attachment,19EE, determined the probabilities and consequences for vacutun breaker leakage areas from zero to that corTesponding to one vacuum breaker stuck fully open.
), Fission product release fractions were determined x with MAAP-ABWR using the dominant accident sequence { Loss of all core Cooling with vessel failure
? occurring a Low Pressure (LCLP)) modified to include 2 a path between the drywell and the wetwc!) att space.
.? Plugging of leakage paths by fission products was v considered for small pathways. Leakage probabilities a were determined by reviewing recent operating c
experience of wetwell to drywell vacuum breakers in a BWRs with Mark I, II and ill containments.
} Suppression pool bypass does not significantly add y to the risk associated with the ABWR because the r bypass areas resulting in increased releases are offset by i ]y low probabilities of occurrence. No leakage and, corTespondingly, no impact on plant risk is expected to occur for almost all (approximately 98 percent) of the accident demands. Small amounts of leakage have a probabihty of 1.8 percent per event, and can result in medium volatile fission product releases (one to ten percent of initial inventory). Volatile fission product releases on the order of 10 to 20 percent of initial inventory can result when large amounts of suppression pool bypass are present. However, the impact on plant risk is still negligible because the probability of large leakage is only 0.39 percent.
O k Amendment ?? 19 EE.5. I
w ABWR 23A61NAS Standard Plant REV.A
/'%
19EE.6 REFERENCES
- 1. Vaughan, E.U., Simple Afodel for Plugging of Ducts by Aerosol Deposits.Trans. Am. Nuclear.
Em,28,507,1978.
- 1. NRC/IDCOR Technical Issue 13A,1986.
- 3. Morewitz, H.A, Leakage of Aeus;ols from Containment Buildings, Health Physics. Vol. 42, No. 2,1962, pp.195-207.
i I i l k O Amendment 77 19EE.6-1}}