ML20066D298
| ML20066D298 | |
| Person / Time | |
|---|---|
| Site: | Grand Gulf  |
| Issue date: | 12/31/1990 |
| From: | Amos C, Breeding R, Tony Brown, Helton J, Higgins S, Jow H, Shiver A ARIZONA STATE UNIV., TEMPE, AZ, SANDIA NATIONAL LABORATORIES, SCIENCE APPLICATIONS INTERNATIONAL CORP. (FORMERLY |
| To: | NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| References | |
| CON-FIN-A-1322 NUREG-CR-4551, NUREG-CR-4551-V6R1P1, NUREG-CR-4551P1, SAND86-1309, NUDOCS 9101140307 | |
| Download: ML20066D298 (305) | |
Text
{{#Wiki_filter:_ ---- ___ _ _ _ NU REG /CR-4551 SANDS 6-1309 Vol. 6, Rev.1, Part 1 Evalua~: ion 0:? Severe Accicen: Risis: ' Granc Gulf, Uni': 1 Main Report Prepared by T. D. Ilrown, it. J. lirceding, !!.-N. Jow, J. C. IIciton, S. J. liiggins, C. N. Amos, A. W. Shiver ,_ Sandia National Laboratories Operated by Sandia Corporation Prepared for U.S. Nuclear llegulatory Commission [0II$0S0k000 j> P _ _ - _ _ _ - _ _ _ _ _ _ _ _ _ - _ .I
I AVAll ADILITY NOTICE Avalbbihty of Reference Materials Cited h NRC Put4 cations Most documents cited in NRC pub 5 cations wtB be evallable from one of the foDowing sources:
- 1. The NRC Public Document Room,2120 L Street, NW, Lower Level, Washington, DC 20555
- 2. The Superintendent of Documents, U.S. Government Printing Office, P.O. Box 37082 Washington, DC 20013-7082
- 3. The National Technical information Service, Springfield, VA 22161 Although the isthg that fogows represents the majority of documente cited in NRC publications, it is not -
Intended to be exhaustive. Referenced documents available for hspection and copying for a fee from tra NRC Publo Document Room include NRC correspondence and internal NRC memorandal NRC Office of Inspection and Enforcement bunctins, circulars, information notices, hspection and hvestigation noticeal Licensee Event Reports; ven. dor reports and correspondence; Commission papers; and app 8 cant and beensee documents and corre. spondence. The fo8owing documente h the NUREG series are avaRable for purchase from the GPO Sales Program: formal NRC staff and contractor reports, NRC sponsored conference proceedings, and NRC booklets and brochures. Also avalable are Regulatory Guides, NRC regulations in the Code of Federal Regularlons, and Nuclear Regulatory Commission issuances. Documents avaRable from the National Technical information Service include NUREG earles reporte and technical reports prepared by other federal agencies and reports prepared by the Atomic Energy Commis-sion, forerunner agency to the Nuclear Regulatory Commiselon. Documents available from publo and special technical librarles include.aR open Iterature items, such as books, )ournal and periodical articles, and traneactions. Federal Hegister notloes, federal and state legista-tion, and congressional reports can usuaRy be obtained from these Exarles, Documents such as theses, dissertations, foreign reporte and translations, and non-NRC conference pro-ceedings are avaRable for purchase from the organlzation sponsoring the pub 5 cation cited. Single copies of NRC dratt reports are avalable free, to the ex+ent of supply, upon written request to the Office of Information Resources Management Distribution Section, U.S. Nuclear Regulatory Commission, - Washington, DC 20555. -- Coples of bdustry codes and standards used in a substantive manner in the NRC regulatory process are maintahed at the NRC Library,7920 Norfolk Avenue, Bethesda, Maryland, and are avaRable there for refer-ence use by the pubGc Codes and standards are usualy copyrighted and may be purchased from the originathg organization or, if they are American National Standards, from the American National Standards Institute,1430 Droadway, New York, NY 10018 DISCLAIMER NOTICE This report was prepared as an account of work sponsored by an a00ncy of the United States Government. Nolther tho United Statos Govemment nor any agency thereof, or any of their employees, makes any warranty, exprosed or implied, or assumos any legal liability of responsibility for any third party's uso, or the reusts of such uso, of any information, apparatus, product or process disclosed in this report, or represents that its use
, - by such third party would not infringo privatoly owned rights.
I NUREG/ Cit-4551 SAND 86-1309 Vol. 6, Rev.1, Part 1 Eva nation of Severe A;cident Ris<s: Granc Gu f, Unit 1 Main Report Manuscript Completed: December 1990 Date Published: December 1990 Prepared by T. D. Ilrown, R. J. lirceding, II.-N. Jow, J. C. IIcitoni, S. J. Iliggins, C. N. Amos2, A. W. Shiver Sandia National Laboratories Albuquerque, NM 87185 Pre;)ared for Division of Systems Research Omce of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission Washington, DC 20555 NRC FIN A1228
' Arizona State University, Tempe, AZ 2 Science Applications International Corporation, Albuquerque, NM
'I ABSTPACT l
In support of the Nuclear Regulatory Commission's (NRC's) assessment of the risk from severe accidents at commercial nuclear power plants in the U.S. reported in NUREG-1150, the Severe Accident Risk Reduction Program (SARRP) has completed a revised calculation of the risk to the general public from severe accidents at the Grand Gulf Nuclear Station, Unit 1. This power plant, located in Port Gibson, Mississippi, is operated by the System Energy Resources, Inc. (SERI). The emphasis in this risk analysis was not on determining a "so called" point estimate of risk. Rather, it was to determine the distribution of l risk, and to discover the uncertainties that account for the breadth of this distribution. The offsite risk from internal initiating events was found to be quite low, both with respect to the safety Soals and to the other plants analyzed in NUREC-1150. The offsite risk is dominated by station blackout type accidents (loss of all ac power) in which core damage occurs shortly af ter the initiation of the accident. The low values for risk can be attributed to the low core damage frequency, the good emergency response, and plant features that reduce the potential source term. Given that ecre damage occurs, it appears quite likely that the containment will fail during the accident. Hydrogen combustion events are the dominant causes of containment failure. Considerable uncertainty is associated with the risk estimates produced in this analysis. l iii/iv
k CONTENTS
SUMMARY
.................. ........................................... S.1
- 1. INTRODUCTION................................................. .. 1.1-1.1 Background and Obj ec tives of NUREG 1150. . . . . . . . . . . . . . . . . . . . 1.1 1.2 Overvie w of Grand Gulf Nuclear S tation, Unit 1. . . . . . . . . . . . . 1. 3 1.3 Changes Since the Draft Report.............................. 1.5 1.4 Structure of the Analysis................................... 1.8 1.5 Organiza tion o f this Repo rt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ,1,16 1.6 Re f e r e nc e s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . l .18 -
- 2. ANALYSIS OF THE ACCIDENT PROGRESSION . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 2.1 Plant Features Important to the Accident Progression at Grand Gulf................................. 2.1 2.2 Interface with the Core Damage Frequency Anal'ysis.......... 2.3 2.2.1 Definition of PDSs.................................... 2.3 2.2.2 PDS Frequencies....................................?... 2.7:
- 2. 2. 3 High-Level Grouping of PDS s . . . . . . . . . . . . . . . . . . . . . . . . . . 2.13 2.2,4 Variables Sampled in the Accident Frequency Analysis.......................... 2.15 2.3 Description of the APET.................................... 2.15 2 . 3 .1 Ove rvi ew o f the APET . . . . . . . . . . . . . . . . . . . . . . . - . . . . . . . . . 2 , 2 0 2.3.2 Overview of the APET Quantification.................. 2.22 2.3.3 Variables Sampled for the ~~
Accident Progression Analysis....................... 2.33 2.4 Description-of the Accident Progression Bins............... 2.55 2.4.1 Description of the Bin Characteristics............... 2.55 2.4.2 Robinning............................................ 2,66 2.4.3 Summary Bins for Presentation........................ 2.66 2.5 Results of the Accident Progression Analysis............... 2.70 2.5.1 Results for Internal Ini tia tors. . . . . . . . . . . . . . . . . . . . . 2. 70 2.5.1.1 Results for PDS 1............................... 2.70 2.5.1.2 Results for PDS 2.............................. 2.72 2.5.1.3 Results for PDS 3.............................. 2.72 2.5.1.4 Results for-PDS 4............................. .2.74 v
i i 2.5.1.5 Results for PDS 5....... ...... .......... .... 2.77 2.5.1.6 Results for PDS 6.............................. 2.77 2.5.1.7 Results for PDS 7.............................. 2.80 2.5.1.8 Results for PDS 8.............................. 2.80 2,5.1.9 Results for PDS 9.............................. 2.83 2.5.1.10 Results for PDS 10............................. 2.83 2.5.1.11 Results for PDS 11..................... ....... 2.85 , 2.5.1.12 Results for PDS 12............................. 2.88 2.5.1.13 Core Damage Arrest and Avoidance o f Vc s sel Breach . . . . . . . . . . . . . . . . . . . . . 2. 88 2.5.1,14 Early Containment Failure ...................... 2.91 2.5.1.15 Early Drywell Failure -.......................... 2.95 2 . 5 .1.16 S umma ry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.97 2.6 Insights from the Accident Progression Ana l y s i s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 .10 0 2.7 References.................................................,2.102
- 3. RADIOLOGICAL SOURCE TERM ANALYSIS......... ...................... 3.1 ,
}
3.1 Grand Gulf Features Important to the Source Te rm Analysi s . . . . . . . .. ........................ 3.1 3.2 De sc rip tion o f the GCSOR Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3
- 3. 2.1 Ove rview of the Parametric Mode 1. . . . . . . . . . . . . . . . . . . . . . 3.3-3 . 2 . 2 De s c rip t ion o f GG S0R . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4 3.2.3 Variables Sampled in the Source Term Analysis.................................. 3.9 .
3.3 Results of Source Term Analysis............................. 3.14
- 3. 3.1 Re sults for Internal Initia tors . . . . . . . . . . . . . . . . . . . . . . . 3.14 3.3.1.1 Re s ul ts fo r PDS 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 ,14 3.3.1.2 Results for PDS 2 ............................. 3.17 3.3.1.3 Re s ul t s fo r PD S 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ' ~ 3 .17 3.3.1.4 Re sul ts f o r PDS 4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.18 3.3.1.5 Re s ul t s fo r PD S 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 .18 3.3.1.6 Re sul ts fo r PDS 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . 3 .19 3.3.1.7 Results for PDS 7 ............................. 3.19 3.3.1.8 Results for PDS 8 ............................ 3.20 3.3.1.9 Results for PDS 9 ............................. 3.21 3.3.1.10 Results for PDS 10............................. 3.21 3.3.1.11 Results for PDS 11............................. 3.22 3 . 3 .1.12 Re sul ts fo r PD S 12 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.22 3.3.1.13 Results for Generalized APBs .................. 3.23 3.3.1.14 Summary........................................ 3.24 3.4 Partitioning of toe Source Terms for Consequence Analysis...................................... 3.59 3.4.1 Results for Internal Initiators....................... 3.59 vt
3.5 Insights from the Source Term Analysis...... .............. 3.71 3.6 References......................................-...........3.72
- 4. CON S EQU EN C E AN A LYS I S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 4.1 Description of the Consequence Analysis...................... 4,1 4,2 MACCS Input for Grand Gulf................................... 4,3 4.3 Results of the MACCS Consequence Calculations................ 4.5 4.3.1 Results foc Internal Initiators...................... 4.5 4 . 4 R e f e r e nc e s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 .15
- 5. RI S K RESU LT S FOR GRAND GULF , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1 5.1 Results for Internal Initiators............................. 5.1 5.1.1 Risk Results.............,,... ....................., 5,1 5.1.2 Contributors to Risk................,,................ 5.9 5.1,3 Contributors to Uncertainty......................... 5.17 5.2 References......................................,,.......... 5.21 6, INSIGilTS AND CONCLUSIONS......................................... 6 .1 -
FIGURES F.1 Back End Documentation for NUREG 1150....................... xv S.1 Overview of Integrated Plant Analysis in NUREG-1150..................................... ........ S.5 S.2 Mean Probability of APBs for the Summary PDSs. . . . . . . . . . . . . . S 10 S.3 Probability of Core Damage-Arrest........................... S.11-S.4 Probability of Early' Containment Failure.................... S.12 S.5 Exceedance Frequencies for Release Fraction.................'S.18' S.6 Consequences Conditional on Source Terms....................'S.21 S.7 Exceedance Frequencies for Risk............................. S.23 S.B Distributions of Annual Risk. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . S . 2 7 S.9 Fractional PDS Contributions to Annual Risk................. S.29 S.10 Fractional APB Contributions to-Annual Risk............ .... S.30 1.1 Section of Grand Gulf Containment...................... .... 1.4 vii
l.2 Overview of Integrated Plant Analysis in NUREG 1150..................................... ........ 1.9 1.3 E x am p l e Ri s k CC D P . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.15 2.2 1 Core Damage Frequency Distributions................... 2.9 2.2-2 Core Damage Frequency Distributions for the Summary PDS Groups.........................., 2.14 2.5 1 Probability of Core Damage Arrest..................... 2.90 2.5-2 Probability of Early Containment Failure............., 2.92 2.5-3 Mean Probability of CF Before Vessel Breach. . . . . . . . . . 2.93 2.5-4 Hean Probability of CF at Vessel Breach.......... ... 2.94 2.5-5 Mean Probability of Drywell Failure Before Vessel Breach................................ 2.96 2.5-6 Mean Probability of Drywell Failure at Vessel Breach ...................................... 2.97
- 2. 5- 7 Mean Probability of APBs for the Summary PDSs. . . . . . . . . 2.99 2,5 8 Distribution of Frequencies for AFB Groups........... 2.101 3.2-1 Blood Flow Diagram for CGS 0R........................., 3.8 3.3 1 Exceedance Frequencies for Release Fractions for Grand Gulf Internal Initiators PDS 1: Fast SB0...................................... 3.25 3.3-2 Exceedance Frequencies for Release Fractions for Grand Gulf Internal Initiators PDS 2: Fast SB0.....................,............... 3.28 3.3 3 Exceedance Frequencies for Release Fractions for Crand Gulf Internal Initiators PDS 3: Fast SB0..................................... 3,30 3.3-4 Exceedance Frequencies for Release Fractions for Grand Gulf Internal Initiators PDS 4: Slow SB0...................................... 3.32 3.3-5 Exceedance Frequencies for Release Fractions for Grand Gulf Internal Initiators PDS 5: Slow SB0.....................................'.~3.34 3.3 6 Exceodance Frequencies for Release Fractions for Grand Gulf Internal Initiators PDS 6: Slow SB0..................................... 3.36 3.3-7 Exceedance Frequencies for Release Fractions for Grand Gulf Internal Initiators PDS 7: Fast SB0..................................... 3.38 3.3-8 Exceedance Frequencies for Release Fractions for Grand Gulf Internal Initiators PDS 8: Slow-SB0.................. .................. 3.40 3.3-9 Exceedance Frequencies for Release Fractions for Grand Gulf Internal Initiators PDS 9: Fast ATWS.......................... ......... 3.42 3.3-10 Exceedance Frequencies for Release Fractions for Grand Gulf Internal Initiators PDS 10: Slow ATWS... ............. .. .............. 3.44 vill
3.3-11 Exceedance Frequencies for Release Fractions for Grand Gulf Int 9rnal Initiators PDS 11: Fast T2....... ........................ ..... 3.46 3.3 1? Exceedance Frequencies for Release Fractions for Grand Gulf Internal Initiators PDS 12: Slow T2........................ .......... . 3.48 3.3-13 Exceedance Frequencies for Release Fractions for Grand Gulf Internal Initiators S u mm a ry A P B 1 . . . . . . . . . . . . . . . . . . . . . . . . . ............. 3.49 3 304 E meedance Frequentiot for F.olteca Fcuctions for Grand Gulf Internal Initiators Summary APB 2................................ ....... 3.50 c 3.3-15 Exceedance Frequencies for Release Fractions for Grand Gulf Internal Initiators Summary AFB 3......... .............................. 3.51 3.3 16 Exceedance Frequencies for Release Fractions for Grand Gulf Internal Initiators Summary APB 4......................... ............ 3.52 3.3-17 Exceedance Frequencies for Release Fractions for Grand Gulf Internal Initiators S ununa ry A P B 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ......... 3 53 3.3-18 Exceedance Frequencies for Release Fractions for Grand Gulf Internal Initiators Summary AFB 6................. ..................... 3.54 3.3-19 Exceedance Frequencies for Release Fractions for Grand Gulf Internal Initiators Summary AFB 7............................... ....... 3.55 3.3-20 Exceedance Frequencies for Release Fractions for Grand Gulf Internal Initiators Summary APB B.......................... ............. 3.56 3.3 21 Execedance Frequencies for Release Fractions for Grand Gulf Internal Initiators All Internal Initiators........ ............... ..... 3.57 3.4 1 Distributions of Nonzero Early and Chronic Health Effects Weights for ~~ Internal Initiators................................... 3.61 4.2-1 Consequences Conditional on Source Terms,...... ..... 4.9 5.1-1 Results of the Integrated Risk Analysis for Internal Initiators at Grand Gulf: Statistical Measures of the 250 Exceedance Frequency Curves for Six Consequence Measures.......................... 5.2 5.1-2 Annual Risks. Grand Gulf: All Internal Initiators... . 5.7 5.1-3 Fractional PDS Contributions to Annual Risk; Grand Gulf.............................
.... . 5.13 5.1-4 Fractional AFB Gontributions to Annual Risk; Grand Gulf.... . .... ............... ....... 5.15 ix
TABLES F.1 NUREG 1150 Analysis Documentation........................... xvi S.1 Design Features Relevant to Severe Accidents................ S.3 S.2 Grand Gulf Core Damage Frequencies.......................... S.7 S.3 Two Methods of Calculating Contcibution toHeanRisk...............................................S.28 , 2.2 1 Grand Gulf PDS Gharacteristics....................... 2.5 2.2-2 PDS Core Damage Frequencies for Grand Gulf........... 2,8 2.2 3 Plant Damage brate Comparison or Grand Gulf..... .................................... 2.10 2.2-4 Relationship Between. PDSs and Summary Groups...... .. 2.14 2.2 5 Variables Sampled in the Accident Frequency Analysis for Internal Initiators.......... ......... 2.16 2.3-1 Questions in the Grand Gulf APET. . . . . . . . . . . . . . . . . . . . 2. 23 2.3-2 Grand Gul f APET Quantification Summary. . . . . . . . . . . . . . 2. 33 2.3-3 Variables Sampled in the Accident Progression Analysis............................................ 2.35 2.4-1 Description of Accident Progression Bin Characteristics................................. 2,59 2.4 2 Description of Summary Accident Progression Bins Charactertistics............................... 2.68 2.5-1 Results of the Accident Progression Analysis for Grand Gulf Internal Initiators-PDS 1: Short Term SB0............................... 2.71 2.5 2 Results of the Accident Progression Analysis ' for Grand Gulf Internal Initiators: PDS 2: Sh o r t - Te rm S B0. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.73 2.5-3 Results of the Accident Progression Analysis for Grand Gulf Internal Initiators: PDS 3: S h o r t Te rm S B0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ~. 2 . 7 5 2.5 4 Results of the Accident Progression Analysis for Grand Gulf Internal Initiators: PDS 4: Lo n g Te rm S B0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , 2 . 7 6 2.5 5 Results of the Accident Progression Analysis for Grand Gulf Internal Initiators: PDS 5: Long Term SB0................................ 2.78 2.5-6 Results of the Accident Progression Analysis for Grand Gulf Internal Initiators: PDS 6: Lo n g - Te rm S B0. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.79 2.5 7 Results of the Accident Progression Analysis for Grand Gulf Internal Initiators: PDS 7: Short-Term SB0......................... .. . 2.81 2.5 8 Results of the Accident Progression Analysis for Grand Gulf Internal Initiators: PDS 8: Long-Term SB0..................... ........ , 2.82 X _- -- ___ _. l
i 2,5-9 Results of the Accident Progression Analysis for Grand Gulf Internal Initiators: PDS.9: S h o r t T e rm ATWS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.84 i 2.5 10 Results of the Accident Progression Analysis for -Grand Gulf Internal Initiators: PDS 10: Long Te rm ATW S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.86 2.5 11 Results of the Accident Progression Analysis ', for Grand Gulf Internal Initiators: PDS 11: S ho r t Te rm T2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.87 2.5 12 Resultr of the Accident Progression Analysis for Grand Gulf Internal Initictors: PDS 12: Lo ng T e rm T 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.89 3.2-1 Isotopes in Each Radionuclide Releaso Glass ......... 3,4 3.2 2 Variables Sampled in the Source Term Analysis..................-................... 3.10 3.3 1 Mean Source Terms for Grand Gulf Internal Initiators: PDS 1.................................. 3.25 3.3 2 Mean Source Terms for Grand Gulf Internal Initiators: PDS 2............ ..................... 3.27 j 3.3 3 Mean Source Terms for Grand Gulf Internal Initiators: PDS 3.................................. 3.29 3,3 4 Hean Source Terms for Grand Gulf Internal Initiators: PDS 4....... .......................... 3.31 3.3 5 Hean Source Terms for Grand Gulf Internal Initiators: PDS 5.................................. 3.33 3.3-6 Mean Source Terms for Grand Gulf Internal I n i t i a to r s : - PDS 6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.35 3.3-7 Mean Source Terms for Grand Gulf Internal Initiators: PDS 7.................................. 3.37 3.3 8 Men.n Source Terms for Grand Gulf Internal Initiators: PDS 8.................................. 3.39 3.3-9 Mean Source Terms for Grand Gulf Internal Initiators: PDS 9.................................. 3.41 3.3 10 Mean Source Terms for Grand Gulf Internal '~ Initiators: PDS 10................................. 3.43 3,3-11 Mean Source Terms for Grand Gulf Internal Initiators: PDS 11................................. 3.45 3.3-12 Mean Source Terms for Grand Gulf Internal Initiators: PDS 12..................................-3.47 3.4-1 Summary of Early Fatality & Chronic Fatality Effect Weights fo r Inte rnal Ini tia tors . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. 60 3.4-2 Distributions of Source Terms with Nonzero Early Fatality and Chronic Fatality................................... 3.64 3.4-3 Dlutributions of Source Teius with Zero Early Fatality and Nonzero Chronic Fatality................................... 3.66 3.4-4 Mean Source Terms Resulting from Partitioning......................... ...... .. 3.67 xi I, 1
4.1-1 Definition o f Consequence Analysis Results . . . . . . . . . . . 4.2 ' 4,2 1 Site Specific Input Data for Grand - Gul f MACCS Calcul a tions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 4.2-2 Shielding Factors used for Grand Gul f MACCS Calcul a t i ons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 4.3 1 Mean Consequence Results for Internal Initiators..........................................- 4.7 5.1 1 Distributions for Annual Risk at Gra'ad Gulf ......... 5.8 5.1-2 Fractional PDS Contributions (in percent) to Annual Risk...................................... 5.11 5.1-3 Fractional APB Contributions (in percent) to Annual Risk..................................... 5.14 5.1-4 Summary of Regression Analyces for Annual Risk........................................ 5.19 ( xii /
I l FOREWORD This is one of numerous documents that support the preparation of the final ' NUREG 1150 document by the NRC Office of Nuclear Regulatory Research. Figure 1 illustrates the documentation of the accident progression, source term, consequence, and risk analyses. The direct supporting documents for the first draft and for the revised draft of NUREG 1150 are given in Table
- 1. They were produced by the three interfacing programs at Sandia National Laboratories (SNL) that performed the work: the Accident Sequence Evaluation Program (ASEP), the Severe Accident Risk Reduction Program (SARRP), and the PRA Phenomenology and Risk Uncertainty Evaluation Program (PRUEP). The Zion volumes were written by Brookhaven National Laboratory and Idaho National Engineering Laboratory.
The Accident Frequency Analysis, and its constituent analyses, such as the Systems Analysis and the Initiating Event Analysis, are reported in NUREG/CR 4550. Originally, NUREG/CR 4550 was published without the designation " Draft for Comment." Thus, the current revision of NUREG/CR-4550 is designated Revision 1. The label Revision 1 is used consistently on all volumes, including Volume 2 which was not part of the original documentation. NUREG/CR 4551 was originally published as a " Draft for Comment". While the current version could have been issued without a revision indication, all volumer of NUREC/CR-4551 have been designated Revision 1 for consistency with NUREC/CR-4550. The material contained in NUREG/CR-4700 in the original documentation is J now contained in NUREG/CR-4551; NUREG/CR-4700 is not being revised. The contents of the volumes in both NUREG/CR 4550 and NUREG/CR 4551 have been altered. In both documents now, Volume 1 describes the methods utilized in the analyses, Volume 2 presents the - elicitation of expert judgment, Volume 3 concerns the analyses for Surry, Volume 4 concerns the analyses for Peach Bottom, and so on. The Grand Gulf analysis is contained in Volume 6 of a NUREC/CR-4551. Note that the Grand Gulf plant was also treated in Volume 4 of the original Draf t for Comment version of NUREG/CR 4700. In addition to NUREG/CR-4550 and NUREG/CR-4551, there are .s eve ral ~other ' reports published in association with NUREG-1150 that explain the methods used, document the computer codes that implement these methods, or present the results of calculations performed to obtain information specifically for this project. These reports include: NUREG/CR-5032, SAND 87-2428, "Modeling Time to Recovery and Initiating Event Frequency for Loss of Off-site Power Incidents at Nuclear Power Plants," R. L. Iman'and-S. C. Hora, Sandia National Laboratories, Albuquerque, NM, January 1988. NUREG/CP.-4840, SAND 88 3102, " Procedures for the Exturaal Emmt- Core Damage Frequency Analysis for NUREG 1150," M. P. Bohn and J. A. Lambright, Sandia National Laboratories, Albuquerque, NM, December 1990 NUREG/CR-5174, SAND 88-1607, J. M. Criesmeyer and L. N. Smith, "A Reference Manual for the Event Progression and Analysis Code (EVNTRE)," Sandia National Laboratories, Albuquerque, NM, 1989. xiii f 1 i _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - _ - - _ i
l l NUREG/CR 5380, SAND 88 2988, S. J. Higgins, "A User's Manual for the Post Processing Program PSTEVNT," Sandia National . Laboratories , Albuquerque, NM, 1989. l NUREG/CR-4624, - BMI-2139, R. S. Denning et al., "Radionuclide Release Calculations for Selected Severe Accident Scenarios," Volumes I.V, Batte11e's Columbus Division, Columbus, OH, 1986. NUREC/CR 5062, BMI 2160, M. T. Leonard e t. al,, " Supplemental Radionuclide Release Calculations for Selected Severe Accident Scenarios," Battelle Columbus Division, Columbus 0H, 1988. NUREG/CR-5331, SAND 89 0072, S. E. Dingman et al., "MELCOR Analyses for Accident Progression Issues," Sandia National Laboratories. *
~
Albuquerque, NM, 1990. j NUREG/CR-5253, SAND 88-2940, R. L. Iman, J. C. Helton, and J '. D. Johnson, " PARTITION: A Program for Definin6 the Source Term / Consequence Analysis Interfaces in the NUREG-1150 Probabilistic Risk Assessments User's Guide," Sandia National Laboratories,. Albuquerque, NM, May 1990. 1
~
l l l xiv
I- / .. . .. . . .. 1 SUPPORT DOCUMENTS TO NUREG-1150
/
tJUREG-1150
/(tJRC< Staff)
~-
EVALUATION OF SEVERE ACCIDENT RISKS NUREG/CR-4551 METHOOS MAJOR INPUT PARAMETERS SURRY PEACH BOTTOM SEQUOYAH GR AND GULF ZION NUREG/CR-4550 Vol 1 Vol 2 Vol3 Vol.4 Vol. 5 Vol.6 Vol. 7
, M M BU R U M of at 61 MM 8 m E =
> 8 $ 'E' '
2 G E w e m 8 y a E 0 '$ ? ? ? 8 C h a s d d d s e e s e S
- o -
8 m s 8 d z E J ! 3 o w $ 8
" m
" e
" e
" o
" e s 8 5 3 s a
- 5 a
- a 3 0 2 0 2 0 0 0 $
a 8 i s g e a 5 i ! O a a a 8 B 9 E 9 2 e 2 2 5 i 9 e 2 i i ra e E w o M < z a = a z a z a sl o 3* a: .;,
- J:cJ D D 2
*.) J =
3" 33 3* D~ 3=
~
Figure 1. Back-End Documentation for NUREG-1150.
s . Table'l. NUREG-ll50 Anmlysis Documentation Oricins1 Documentation NUREG/CR-4700 NUREG/CR-4550 NUREG/CR-4551 Containment Event Analysis Analysis of Core Damage Frequency Evaluation of Severe Accident Risks and the Potential for Risk Reduction for Potential Severe Accident From Internal Events Vol. 1 Surry Unit 1 Vol. 1 Surry Unit 1 Vol. 1 Methodology 2 Sequoyah Unit 1 2 Sequoyah' Unit 1 2 Summary (Not Published) 3 Peach Bottom Unit 2 3 Surry Unit 1 3 Peach Bottom Unit 2 4 Grand Gulf Unit 1 4 Grand Gulf Unit 1 4 Peach Bottom Unit 2 5 Sequoyah Unit 1 6 Grand Gulf Unit 1 7 Zion Unit'l Revised Documentation NUREG/CR-4551, Rev. 1. Eval. of Severe Accident Risks NUREG/CR-4550, Rev. 1, Analysis of Core Damage Frequency
- Methodology Vol. 1 Part 1, Methodology; Part.2, Appendices
-Vol. 1 Part 1 Expert Judgment Elicit. Expert Panel 2 Part 1 In-Vessel Issues 2
Part 2 Expert Judgment Elicit. Project Staff Part 2 Containment Loads'and MCCI Issues Part 3 Structural-Issues Part 4 Source' Term Issues Part 5 Supporting Calculations Part 6 Other Issues-Part 7 MACCS Input Part 1 Surry Unit 1 Internal Events 3 Part 1 Surry' Analysis and Results 3 Part 2 Surry Unit 1 Internal Events App. Part 2 Surry Appendices Part 3 Surry External Events , 4 Part 1 Peach Bottom Analysis'and Results 4 Part 1 Peach Bottom Unit 2 Internal Events Part 2 Peach Bottom Appendices Part 2 Peach Bottom Unit 2 Int. Events App. Part 3 Peach Bottom Unit 2 External Events 5 Part 1 Sequoyah Analysis and Results 5 Part 1 Ses,uoyah Unit 1 Internal Events Part 2 Sequoyah Unit 1 Internal Events App. Part.2 Sequoyah Appendices 6 Part 1 Grand Gulf Analysis and Results 6 Part 1 Grand Gulf Unit 1 Tnternal Events Part 2 Grand Gulf ' Appendices Part 2 Grand Gulf Unit 1 Internal Events App. Zion Unit 1 Internal Events 7 Part 1 Zion Analysis and Results 7 Part 2 Appendices c
ACKNOWLEDGMENTS We wish to thank the many people who worked in various capacities to support this analysis: E. Corham Bergeron (SNL), who was the program manager and provided many helpiul suggestions in methods and techniques; i F. T. Harper (SNL), who provided the day to day leadership of the project and worked wherever help was needed; J. L. Sprung (SNL), J. D. Johnson (Applied Physics, formerly SAIC), and D. I. Chanin (Technadyne), all of the consequence analysis team, who performed the MACCS - analysis ; R. . L. Iman (SNL) for his work in designing the overall computational strategy and~the codes to be used in implementing that strategy and J . D. Johnson for constructing some of those codes; S. E. Dingman (SNL) for the many computer calculations that she performed in support of this analysis and for her help in suggesting ways to model various aspects of the accident progression in the APET; and R. A. Garber for her technical editing of the report. We also wish to thAuk the other plant analysts, A. C. Payne (SNL), and G. J. Gregory (SNL), for their many helpful suggestions. Several members of the Quality Control Team, K. D. Bergeron (SNL), G. J. Boyd (SAROS), D. R. Bradicy (SNL), R. S. Denning (BMI), S. E. Dingman, J. E. Kelly (SNL), D. M. Kunsman (SNL), J. Lehner (BNL), S. R. Lewis (SAROS), and D. W. Pyatt (NRC), reviewed various parts of the analysis and we thank them for their constructive suggestions for improving the overall quality of the analysis. We are particularly thankful to them for their review of the Grand Gulf APET and its user functions. The authors also acknowledge the. efforts of the Level I Grand Gulf analysts, M. T. Drouin (SAIC) and T. A. Wheeler (SNL) for their efforts in making the interface between the Level I and Level II internal events analyses work efficiently, Finally, we wish to thank M. A. Cunningham, J. A. Murphy, and P. K. Niyogi-of the NRC for their funding of this project and program and management support. xvil j
t ACRONYMS AND IMITIALISMS ADS automatic depressurization system APB accident progression bin APET accident progression event tree ASEP accident sequence evaluation program ATVS anticipated transient without scram BAF bottom of active fuel l BNL Brookhaven Naticaal Laboratory BWR boilin6 water reactor l- CCF common cause failure. i CCI core-concrete interaction CCDF complementary cumulative distribution function CDF cumulative distribution function CF containment failure CFW chronic fatality weight CS containment spray system CST condensate storage tank DCH direct containment heating DG diesel generator l ECCS emergency core cooling system (s) , EF- carly fatalities l EFW early fatality weight E0P emergency operating procedures EPRI Electric Power Research Institute-EVSE ex-vessel steam explosion FSAR final safety analysis report IVS firewater system
-HEP human error probability --
HIS hydrogen ignition system HPCS high pressure core spray HPME high pressure melt ejection HRA human reliability analysis INEL Idaho National Engineering Laboratory
'LCF latent cancer fatalities LHS Latin Hypercube Sampling LOCA loss-of-coolant accident LOSP -loss of offsite power ,
LPCI low pressure coolant injection LPCS low pressure core spray LTSB long term station blackout LWR light water reactor HACCS MELCOR Accident Consequence Code System MCDF mean core damage frequency xviii
1 I MDP motor-driven pump MOV motor operated valve MSIV main steam isolation valve NRC Nuclear Regulatory Commission PCS power conversion system PDS plant damage state PRA probabilistic risk analysis PRUEP Phenomenology and Risk Uncertainty Evaluation Program PWR pressurized water reactor RCIC reactor core isolation cooling RCS reactor coolant system RHR residual heat removal RPS reactor protection system RSS Reactor Safety Study RPV reactor pressure vessel SAIC Science Applications, International Corporation SAROS Safety and Reliability Optimization Services, Inc. SARRP Severe Accident Risk Reduction Program SB0 station blackout SERC steam explosion review Sroup SERI System Energy Resources, Inc. SLC standby liquid control SNL Sandia National Laboratories SORV stuck-open relief valve-SPC suppression pool cooling SPMU suppression pool makeup SRV safety relief valve SSW standby service water STSB short term station blackout TDP turbine driven pump TEMAC Top Event Matrix Analysis Code i VB vessel breach l l XiX
SUMMARY
S.1 lifG.cd ae tion The Uni te c' States Nuclear Regulatory Commission (NRC) has recently completed a major study to provide a current characterization of severe accident riska from light vater reactors (1NRs). This characterization is derived from integrated risk analyses of five plants. The summary of this study, NUREC 11502, has been issued as a second draft for comment. The risk assessments on which NUREG 1150 is based can generally be characterized as consisting of four analysis steps, an integration step, and an uncertainty analysis scep: , 1. Accident frequency analysis: the determination of the likelikood and nature of accidents that result in the onset of core damage.
- 2. Accident progression analysis: an investigation of the core damage process, both within the reactor vessel before it fails and in the containment afterwards, and the resultant impact on the containment.
- 3. Source term analysis: an estimation of the radionuclide transport within the reactor coolant system (RCS) and the containment, and the magnitude of the subsequent releases to the environment.
- 4. Consequence analysis: the calculation of the offsite consequences, primarily in terms of health effects in the general population.
- 5. Risk intestation: the assembly of the outputs of the previous tasks into an overall expression of risk.
- 6. Uncertainty analysis: the propagation of the uncertainties in- the initiating events, failure events, accident progression branching ration and parameters, and source term parameters through the first three analyses above, and the determination of which of these uncertainties contributes the most to the uncertainty in risk.
This volume presents the details of the last five of the six steps listed above for the Grand Gulf Nuclear Station, Unit 1. -The first step is described in NUREG/CR 4550,2 S.? uverview of Crand Culf Nuclear Station. Unit 1 l The Grand Gulf Nuc1 car Station. Unit 1 is operated by System Enorgy Resources Inc. (SERI) and is located on the east bank of the Mississippi river in southwestern Mississippi, about six miles northwest of Port l Gibson, Mississippi. The nearest large city is Jackson, Mississippi, l approximately 55 miles to the northeast of the plant. S.1
1 1 The nuclear reactor of Grand Gulf Unit 1 is a 3833 MWt BWR 6 boiling water reactor (bWR) designed and supplied by General Electric Company. Unit 1 constructed by Bechtel Corporation, began commercial operation in July 1985, i 4 Table S.1 summarizes the design features of the plant ti.. are relevant to severe accidents. As is evident from this table, there is considerable redundancy and diversity of coolant injection and heat removal features at Grand Gulf. Grand Gulf has a Mark III BWR containment. The contaitvnent is a steel lined reinforced concrete structure. In the Mark III design the reactor pressure vessel is housed in the drywell, which is in turn completely enclosed in the containment structure. The drywell and the containment communicate through passive vents in the suppression pool. Although the free volume of the containment is comparable with a large PWR containment, the design pressure of the Grand Gulf containment is fairly low (15 psig). S.3 Lescription o Lthe Integrated Risk Ann 1vnig Risk is determined by combining the results of four constituent analyses: the accident frequency, accident progression, source term, and consequence analyses. Uncertainty K risk is determined by assigning distributions to important variables, g-nerating a sample from these variables, and propagating each observation of the sample through the entire analysis.
'We sample for Grand Gulf consisted of 250 observations involving variables
, the fit uc three constituent analyses. The risk analysis synthesizes t Jesults of the four constituent analyses to produce incasures of offsite ik and the uncertainty in that risk. This process is depicted in Figure S.1. This figure shows, in the boxes, the computer codes utilized. The interfaces between constituent analyses are shown between the boxes. A mathematical summary of the process, using a matrix representation, is given in Section 1.4 of this volume.
The accident frequency analysis uses event tree and fault tree techniques to investigate the m, nner in which various initiating events can lead to core damage and the frequency of various types of accidents. Experimental dat', past observa',ional data, and modeling results are combined to produce frequency estianes for the minimal cut sets that lead to core damage. A minimal cut vc is a unique combination of initiating event and individual hardware or .sperator failures. The minimal cut sets are grouped into plant , damage ctates (PDSs), where all minimal cut sets in a PDS provide a similar set of initial conditions for the subsequent accident progression analysis. Thus, the PDSs form the interface between the accident frequency analysis and the accident progression analysis. The outcome of the accident frequency analysis is a frequency for each PDS or group of PDSs for each observation in the sample. The accident progression analysis uses large, complex event trees to determine the possible ways in which an accident might evolve from each PDS. The definition of each PDS provides enough information to define the initial conditions for the accident progression event tree (APET) analysis. l Past observations, experimental data, mechanistic code calculations, and expert judgment were used in the development of . the model for accident i i "2 . l
\
L , . _ _ - _ - _ __
l progression that is embodied in the APET and in the selection of the branch probabilities and parameter values used in the APET. Due to the large number of questions in the Crand Culf APET and the fact that many of these questions have more than two outcomes, there are far too many paths through the APET to permit their individual consideration in subsequent source term and consequence anslysis. Table S.1 Design Featura: Relevant to Severe Accidents Crand Culf Unit 1 Coolant Injection High Pressure Core Spray System (HPCS) Systems One train, one MDP' Dedicated diesel Benerator 5 Reactor Core Isolation Cooling System (RCIC) j One train, one TDP* Low Pressure Core Spray System (LPCS) One train, one MDP* j Low Pressure Coolant Injection System (LPCI) Three trains, three MDP* Backup Coolant Injection Systems Standby service water system Firewater system Condensate system Automatic Depressurization System (ADS) Eight relief valves , Requires de power Heat Removal Residual Heat Removal System -- Systems Suppression pool cooling mode: Removes decay heat from suppression pool - two trains, two MDP* i' Shutdown Cooling System Removes decay heat during accidents in which reactor vessel integrity maintained and reactor pressure vessel (RPV) is at low pressure - two trains, two MDP' Containment Spray System: Suppression pressure in containment - two trains, two MDP* Reactivity Control Control Rods Standby Liquid Control System S.3 _m +
.c m7,..m ,
l Table S.1 (continued) Emergency Electrical Electrical Power (ac) Fower Two diesel generators (DGs) HPCS diesel generator has crosstics Electrical Power (de) 12 hour station batteries Containment Structurc BWR Mark III Reinforced concrete structure with steel liner Decign pressure of 15 psig Volume is 1,67 million ft 3, Free volume of 1.4 million ft3 Dryvell Structure Completely enclosed within containment structure Communicates with wetvell through horizontal vents Internal design pressure of 30 psid Free volume of 270,000 fta Reactor Pedestal Cylindrical cavity located directly below RPV Cavity Water on drywell floor will drain into the cavity Volume of the cavity is large enough to contain any core debris released from the vessel Containment Systems Hydrogen Igniter System (HIS) Prevents the buildup of large quantities of hydrogen in the containment requires ac power -- Containment Venting Used when suppression pool cooling and containment sprays have failed to reduce primary containment pressure requires ac power
*MDP - motor driven pump TDP - turbine-driven pump The paths through the trees are grouped into accident progression bins (APDs), where each bin is a group of paths through the event tree that defino a similar set of conditions for source term analysis. The properties of each accident progression bin define the initial conditions for the estimation of a source term. The result of the accident S.4
ACCIDENT ACClCENT SOURCE FREQUENCY PROGRESSION TERM CONSEQUENCE NJALYSIS NJALYSIS ANALYSIS NJALYSIS
# OOLCGCat CNMr P Ap aME Ta'O mFOsws& Ae'JJT M aLN ES* EC ?S t ava Pt ANT sYs4M (gg,f=ygg y mg W E @ DtoGFCaLOa'a oatm a rom r mtua! P"7 MSS'ON F:S$2ON PROI?JCr TRMOFST uoOEts(isttomst M Rf v0V AL O(W M C AT A
#.9 t u ip%at ( yf g F$} Q A TA E WE % ? WT
, y a0CCENT , 9 Pt eN T D AuAr# pnocntssrfw i MM ' N ** l i 1
f./3 u 1 sers cur se1
- TEv e s 1,-
e vumt j6 v: :
- msca
/5 -
came=s a i r unces costex=ct s l "*e
= ,,
s i N rnE0' E=C5ES Nh m N N ratoutwoES t reas ss+ oms
., ., . ...c L_
-o varma raov :ce= taw .vtec~c s et=<cava ee <ssues; 5 " 054""'SS'"
cor=w=Ts o - rac~r two rssues :- ,, ,,, a wi.r e=ecarsse rssues i et Tauo ccess umvet= cues sarstr f Figure S.I. Overview of Integrated Flant Analysis in NUREG-1150.
i progression analysis is a probability for each APB, conditional on the i occurrence of a PDS, for each observation in the sample. A source term is calculated for each AFB with a non zero conditional I probability for each observation in the sample by GCSOR, a fast running j parametric computer code. CGSOR is not a detailed mechanistic model; it is , not designed to inodel the fission product transport, physics, and chemistry from first principles. Instead, CGSOR integrates thu results of many detailed codes and the conclusions of many experts. Most of the parameters used to calculate fission' product release fractions in GGSOR are sampled i from distributions provided by an expert panel. Because of the large number of APBs, use of a fast executin6 code like GGSOR is necessary. The number of APBs for which source terins are calculated is so large that it is not computationally practical - to perform a' consequance calculation for every source term. As a result, the source terms have to be combined Each source terin group is a collection of source into source terin groups. terms that result in sitnilar consequences. The process of _ determining
> vhich APBs are included in which source torin group is called partitioning. l This process considers the potential of each source term group to cause early fatalities and latent cancer fatalities. The result of the source ,
term calculation and subsequent partitioning is that each AFB for each observation is assigned to a source term group. A consequence analysis is performed for each source term group, generating both inean consequences and distributions of consequences. Since cach APB ; is assigned to a source t e rrn group, the consequences are known for every APB of each observation in the sample. The frequency of each PDS for each observation is known from the accident frequency analysis, and the conditional probability of each APB is determined for every PDS group and observation in the accident progression analysis. Thus, for each APB of each observation in tho' sample , both frequency and consequences are 1 determined. The risk analysis. u.embles and analyzes all these separate estimates of offsite risk. S.4 Results of the Accident Frecuency Analysis . The accident frequency analysis for Grand Gulf is documented elsewhere.2 This section only summarizes the results of the accident frequency analyses since they form the starting point for the analyses that are covered in. , this volume. Table S.2 lists four summary measures of the core damage i l frequency distributions for Grand Gulf for the twelve internally initiated PDSs. The four summary measures are the mean and the 5th, 50th (median) _ l and 95th percentiles. PDSs 1, 2, 3, and 7 involve station blackout scenarios in which coolant' injection is lost early such that core damage occurs in the short term with the RPV at high pressure. For PDSs 1, 2, and 3, - .offsite power is recoverable and the operators can depressurize the RPV. For PDSs 2 and 3 heat removal via the containment sprays is failed and not recoverable. For PDSs 1, 2, and 3 the core damage process may be arrested before the vessel fails if offsite power is recovered and coolant injection is restored to ; 1 S.6
Table S.2 Grand Gulf Core Damage Frequencies Internal Initiators Core Damare Frecuency (1/R vr) % Mean TCD PDS 5t - Median Menn 954 Frecuency PDS 1 2.6E 08 5.1E 07 3.2E 06 1.1E 05 79 PDS 2 6.4E 11 2.1E 09 4.6E-08 1.9E 07 1 PDS 3 1.3E-09 3.4E 08 1.5E 07 6.7E 07 4 ; PDS 4 5.3E 11 2.3E-09 3.7E 08 1.6E 07 1 PDS 5 7.4E 13 3.2E 11 2.3E 09 3.0E 09 <<1 PDS 6 1.4E 12 1.3E 10 1.4E-09 7.2E 09 <<1 PDS-7 2.8E 08 2.4E-07 4.2E 07 1.6E 06 11 PDS 8 2.6E 10 8.4E 09 6.3E 08 2.7E 07 2 PDS-9 3.2E-10 7.9E 09 5.0E 08 1.9E 07 1 , PDS 10 3.9E 10 8. 9) 09 6.2E-08 2.3E 07 2 PDS 11 3.1E 11 1.2E 09 1.8E-08 5.3E 08 <1 PDS-12 4.9E 12 6.8E 11 2.9E 10 1.2E-09 <<1 a Total 1.8E 07 1.1E 06 4.1E 06 1.4E-05 - -- D the core. PDS 7 is different from tha first three PDS in that both ac and de power are lost and cannot be recovered. Except for.the unlikely event l that a safety relief valve (SRV) sticks open and depressurizes the RPV l which then allows the fire water system to be used as a backup source of I coolant injection, accidents that progress from this PDS always proceed to vessel failure. The PDS group that includes these four PDSs is referred to as the short term station blackout (STSB)-or STSB group. l PDSs 4, 5, 6, and 8 involve station blackout scenarios in which coolant injection is lost late such that core damage occurs in the long term. For PDSs 4, 5, and 6 core -damage occurs with the RPV at low pressure and offsite power is recoverable. For PDSs 5 and 6 heat removal via the containment sprays is failed and not recoverable. For PDSs 4, 5, and 6 the ; core damage- - proces s may be arrested before the . vessel fails if offsite power is recovered and coolant injection is restored to the core. PDS 8 is ; l S.7
+
--w.. ,e v.,, . . _ . ~ . , , . --- , _ . . . , _ , , . , . . , - _ _ , . . - . . _ , , - - . . . . .- -~_-,, -,-..
different from the other 3 PDS in that both ac and de power ere lost and cannot be recovered. Thus, for accidents that progress frotn this PDS, the vessel always fails at high pressure. The PDS group that includes these four PDSs is referred to as the long-term station bicchout (LTSB) or LTSB group. PDSs 9 and 10 involve anticipated transient without scram (ATWS) scenarios. For PDS 9 coolant injection is lost early such that core damage occurs in the short term whereas for PDS 10 injection is lost late such that core damage occurs in the long term. 'or i both PDSs, core damage occurs because the operators fail to depressurize the vessel to allow low pressure injection systems to cool the core. If the operators correct this error sufficiently early in the accident, the core damage process can be arrested before the vessel fails. The PDS group that includes these two PDSs is referred to as the ATWS group. PDSs 11 and 12 involve transient scenarios where the power conversion system (PCS) is lost (i.e., T2). For PDS 11 coolant injection is lost early such that core damage occurs in the short term whereas for PDS 12 injection is lost late such that core damage occurs in the long term. For both PDSs core damage occurs because the operators fail to depressurite the vessel to allow low pressure injection systems to cool the core. If the operators correct this error sufficiently early in the accident, the core damage process can be arrested before the vessel fails. In both PDSs heat removal via the containment sprays is possible. The PDS group that includes these two PDSs is referred to as the transient or T2 group. S,5 Accident Progressien Annivsis S.5.1 Descriotion of the Accident Pforressjon Annivsis g The accident progression analysis is performed by means of a large and detailed event tree such as the APET. This event tree forms a high level model of the accident progression, including the response of the containment to the loads placed upon it. The APET is not meant to-be a substitute for detailed, mechanistic computer simulation codes. Rather, it is a framework for integrating the results of these codes together with experimental results and expert judgment. The detailed, mechanistic codes require too much computer time to be run for all the possible' accident progression paths. Furthermore, no single available code treats all the important phenomena in a complete and thorough manner that is acceptable to all those knowledgeable in the field. There fore , the results from these codes, as interpreted by experts, are summarized in an event tree. The resulting APET can be evaluated quickly by computer, so that the full diversity of possible accident progressions can be considered and the uncertainty in the many phenomena involved can be included. The APET treats the progression of the accident from the onset of core damage through the core concrete interaction (CCI). It accounts for the various events that may lead to the release of fission products due to the accident. The Grand Gulf APET consists of 125 questions, most of which have more than two brancher.. Four time periods are considered in the tree. The recovery of offsite power is considered both before vessel failure as S8
, . _ . . _ _ _ _ _ . _ _ _ _ . _ _.. _ _ _ _ _.. _ _ _ _ _ _ .._ _._. _ _ ~ . . _ _
i l 1 1
! well as af ter vessel failure. The possibility of arresting the core degradation process before failure of the vessel is explicitly considered.
Coro damage arrest may occur following the recovery of offsite power or when depressurization of the RPV allows injection by a low pressure injection system that previously could not function with the RPV at high pressure. Contairment failure is considered before vessel breach, around the time of vossol breach, ar,d late in the accident. The dominant events l that can cause containment failure are hydrogen combusti'on events (both de flagr ations and detonations) and the accumulation of steam and/or
; noncondensibics in the contaitueent.
The APET is so large and complex that it cannot be presented graphically and must be evaluated by computer. A computer code. EWTRE , has been written for this purpose. In addition to evaluating the APET, EWTRE sorts the myriad possible paths through the tree into a manageable number of outcomes, denoted accident progression bins (APBs). 4 S.5.2 Results of the Aceldent Prorression Analysis Results of the accident progression analysis for internal initiators at Crand Culf are summarized in Figures S.2, S.3, and S.4 Figure S.2 shows the mean distribution among the summary accident progression bins for the summary PDS groups. Technically, this figure displays the mean probability of a summary AFB conditional on the occurrence of a PDS group. Since only mean values are shown, Figure S.2 gives no indication of the range of i values encountered. The distributions of the expnted conditional probability for core damage arrest for a given PDS group cre shown in-Figure 5.3. Similarly, the distributions of the expected conditional probability for early containment failure (CF) for a given PDS group are displayed in Figure 5.4. Early CF any time before vessel breach, at vessel breach, or shortly following vessel breach, l Figure S.2 indicates the mean probability of the possibic outcomes of the accident progression analysis. The width of each box in the figure indicates how likely each accident progression outcome is for each type of accident. Because roughly 90% of the total mean core damage frequency is attributed to the short term station blackout (SBO) summary PDS group, the results presented in the frequency weighted average column are heavily influenced by the short term SB0 results. If the accident procaeds to core damage, containment failure during the accident is a likely outcome. The mean conditional probability of early containment failure is approximately 0.50 and half of this maan value is associated with accidents that also involve some bypass of the suppression pool (i.e., drywell failure). If the accident proceeds to vessel breach and the containmer.t ooes not fail early, there is still a fairly high probability that the containment will I fall late in the accident. Events that can fail the containment late in , the accident are hydrogen burns and the accumulation of noncondensibles and steam in the containment. In the SB0 PDSs ac power may not be available late in the accident and. thus, the containment sprays will not be availabic to condense the steam. Furthermore, even if the sprays are
- . available, the accumulation of noncondensibles generated at vessel breach and during CCI may _ still fail the contaitunent.
S9
-. - - , , . - . - . . - . - - . - . . - - . - - .. . - . - - ~ - ..._ - - - .
..- . _ . - . . ~ . - . .- _ - .. . _ - . .
l STSil 1.TSD ATWS Transients 'All (3.85E-06) (1.04E-07) (1,12E-07) (1.87E-08) (4.09E-06) VB.enrly CF. 0.100 0.292 0.000 0.011 0.150 carly SPD, no CS ,,,, ,,,__ _ VD. enrly CF. 0.031 0.017 0.237 0.202 0.049 - ! carly SPD CS __ _,, , I VB carly CF. 0.006 0.005 0.003 0.003 0.007 late SPD
~
T l VD, early CF. 0.182 0.531 0.500 0.3 31 0.218
- no SPD _, ,,,_
i VD, late CF 0.300 0,129 0.074 0.232 0.284 __.J . _ _ VD, venting 0.032 0.003 0.109 0.075 0.038 I VD, No CF 0.053 0.003 0.036 0.092 0.050 l l No Yb 0.201 0.015 0.025 0.050 0.100 l CF = Containment Failure Grand Gulf CS = Containntent Sprays CV = Containment \enting SPD = Suppression Pool Dypass '~ VD = Vessel Dreach Figure S.2. Mean Probability of APBs for the Summary PDSs. Contair. ment venting is not -a -likely outcome in this analysis. There are several reasons for this result. First, the dominant PDSs are the short - term SB0s. In these PDSs the suppression pool remains subcooled during core damage and, therefore, the containment is - not pressurized by the - accumulation of steam. During core damage and af ter vessel breach a significant quantity of radionuclides will be released to the containment. Af ter vessel breach it is unlikely that the operator will :ver.t these releases to the outside environment. The results of this analysis indicate that-there is a high. likelihood'that the reactor cavity will contain_ water at vessel breach. With respect to-containment integrity and radionuclide release, this situation has both disadvantaSes and advantages. The presence of water allows for the-l- S.10
. - - . - _ - . . - - . - . . ~ - . . .- . - _ . . . . _ - . . - - - . . . ~ - . . - - - - . _ - .
t 1.E0 . 95th
~
. 3 I 1
. onth.1 1
, M = mean !
m = median j th = percenule l , 1 i-DT gg . w. _ Il
. .O . 9 5 t h. ,
) 2o <L . u ot> o ]. l A
- 1.E-1_ m..
m E -
. 95th a m .l j
e
,0 a -
06th, i i eu . l l ]O - u-.
- f. O Y> -
, u..
- m-- l
- M.,
I l 1.E-2_ m 7 I" l' l I PDS Group STSB LTSB ATWS Transients Total l Core Damage Preq. 3.9E-06 1,0E-07 1.1 E-07 1.9E 4.1E-06 F16 ure S.3. Probability of Core Damage Arrest. i possibility of ex vessel steam explosions which can indirectly threaten the integrity of the drywell through the failure of the reactor- pedest.al. An ex vessel steam explosion also contributes to radionuclide release at. vessel breach.- On the other hand, this water also contri)utes to the high - probability that core debris released from the vessel will be cooled. If CCI does initiate, the release will be scrubbed by the overlaying pool of water. s Core Damace Arrest. For the short term SB0 group the probability of core damage arrest is driven by the likelihood that ac power is recovered early in the accident. Injection to the RPV generally follows ac power recovery.
- Although the mean probability of recovering . ac power .is high . (0.60) for most of the short term SB0 PDSs, there 'aro several factors - that tend to reduce the probability of core - damage arrest. First, . res toration - of coolant injection to the RPV does not guarantee that the vessel will not 0
S.11
,,n.. . . , - -.- ~4 e- r - - -+ r E w e + ww w
- a v- -u-+-w-*s*-e'= - * - * *
- -ww.y v e -----em-** 1*yW- er e w**=m=-1rt-+r"ww- --'rt-
1 i l i .\ . 1 E0 95th. - 90gg %n- bth, 36th. - w -
... p l w
. I u
l stL J h
-l m . l
~
~
i T i E 1 a -l
\
y& 3 ,,,, 6t 6
- i h I g 51.E-1. r a .5 -
)
1 75 : f c C l S.0 g l Eh - l 86 1 5 i i etu I Grand Gulf 1.E-2. 5t h ] " * * "" 6th ] *th a percentile PDS Group STSB LTSB ATWS Transients All Core Damage Freq. 3.9E-06 1.0E-07 1.1E-07 1.9E-08 4.1E-00 Figure S.4. Probability of Early Containment Failure. fail. In some cases the core debris is not in a coolable configuration when injection is recovered and, therefore, the accident continues to vessel breach. In addition, an in vessel . steam explosion may fail the vessel before the core is cooled. There are other cases in which only-low pressure injection systems are recovered; however, the - operators have failed to depressurize the RPV. With the vessel at system pressure these-low pressure systems are unable to provide coolant to the core . and, ; l therefore, the accident proceeds to vessel breach. . Finally, in PDS 7, which is a significant contributor to the mean frequency of this summary group, ac power cannot be recovered. Therefore, except-for the infrequent case which involves a stuck open SRV that depressurizes.the RPV and allows , firewater to be injected into the vessel, . accidents in this group progress to vessel failure. l q S.12 e
. - - . . _ _ _ _ . m m,- - - ,...,_~~..,,_,s,,..,....,_,.-,,,.~s.,_,., _.._,,._..-_,...--m . - - - . ..,. ...__m.4,
4 As with the short terrn SB0 group, the probability of core dainage arrest for the long terra SB0 group is also driven by the likelihood that ac power is recovered. The probability of core damage arrest for the long term SB0 group, however, is significantly lower than the corresponding value for the shor t. te rin SB0 group. Two fe; tors are responsible for most of this difference. First, the probability of recovering ac power during a long terrn SB0 is considerably lower than the probability of recovering ac power during a sMt term SBO. Second, in PDS 8, which accounts for approxitnately half of this Group's incan frequency, ac power cannot be recovered and the accident always proceeds to vessel breach, For both the ATVS PDS group and the T2 PDS group, the probability that core damage is not arrested is driven by operator errors. In these PDSs low pressure injection systeres are available; however, the operator fails to depressurize the RPV. It must be remembered that core damage arrest does not necessarily mean that there will be no radionuclide releases during the accident. Both hydrogen and radionuclides are released to the containment during the core damage process. If a large atnount of hydrogen is generated during core damage and is subsequently, ignited, it is possible that the resulting load will fail the containment. If the containment fails, a pathway is established for the radionuclides to enter the outside environment. This radionuclide release is generally small, however, because in the majority of the cases in which vessel breach is averted these releases are scrubbed as they pass through the suppression pool. Furthermore, if the vessel does , not fail, there are no ex vessel releases (e.g., CCI releases). Enriv Containment Failure. The early fatality risk depends strongly on the probability of early CF. Early CF includes both failures that occur before ' vessel breach and during the tiine period around vessel breach. The Grand Gulf containment is a fairly weak structure when considering the loads that can potentially occur during the course of the accident. The design , pressure is only 15 psig and the assessed incan failure pressure is 55 psig. . Because of its low failure pressure, the Grand Gulf containment is not mnly : s.usceptible to loads from hydrogen deflagrations and detonations, but can , also be threatened by slow pressurization events that are associated with the accumulation of steam and noncondensibles. The production of hydrogen during the core damage process and later during vessel breach, should it occur, is a key factor that affects the probability of containment failure. In a BVR core there is a large inventory of zirconium. Large amounts of hydrogen are produced from the , oxidation of this metal during the core damage process. If the llIS is not operating, which is the case in a SBO, the hydrogen will accumulate in the containment. For accidents in which the euppression pool is subcooled, the steam released frorn the RPV is condensed in the pool. The lack of steam in
-the containment atmosphere in combination with the large amount of hydrogen released during the core degradation process allows mixtures to form that have a high hydrogen concentration. Subsequent ignition ot' this hydrogen by either random sources or by the recovery of ac power can result in loads that not only can threaten the containment but also can pose a significant challenge to the drywell structure.
S.13
i Figure S.4 shows the probability distribution for early CF at Grand Gulf. The probability distributions displayed in this figure are for accidents that proceed to vessel breach and are conditional on core damage. The weakness of the containment, relative to the loads that are imposed on it, is reflected in the relatively high containment failure probabilities. l tlydrogen combustion events are the dominant events that cause early CF in the short term SB0 and T2 PDS groups. The mean probability of early containment failure for these two groups is roughly 0.5. The majority of i these failures are caused by hydrogen deflagrations rather than by detonations. In both of these summary PDS groups ~ the suppression pool is subcooled before vessel breach and, therefore, there is no significant ' accumuistion of steam in the containment. This virtually eliminates the possibility of early CF from clow pressurization events (e.g., accumulation of steam) . Because the itis is not available during a short term SBO, severe hydrogen combustion events before vessel bre di are p;sibic. In the shore term SB0 PDS group, about half of the mean probability is associated with CFs that occur before vessel breach and the other half with failures that occur shortly af ter vessel breach. In the T2 PDS group, on _ the other hand, almost all of the early- CFs occur at the time of vessel-breach. For accidents in the T2 group, it is likely that the operator l terned on the llIS before core damage and, therefore, the hydrogen generated t betere vessel breach is usually burned such that the resulting load is b e n i g., . The rapid combustion of hydrogen generated at vessel . breach, 3 however, can still lead to early CF. For the long term SB0 PDS group, the mean conditional probability of early CF is 0.85. Less than half of these early CFs are caused by hydrogen ! combustion events. In this summary PDS group the suppression pool is saturated and the containment is pressurized by the accumulation of steam i that is generated by the hot pool. In most of these accidents hydrogen burns before vessel breach are not possible - because the containment is steam inert. Approximately two thirds of this mean probability results from early CFs that occur before vessel breach and the preponderance of these CFs are caused by pressurization - events associated with1the accumulation of steam in the containment. There are a few cases, however, in which the containment sprays are recovered before vessel breach and a combustible mixture is formed by the condensation of the steam. Subsequent igniticn of this mixture can result -in containment failure. The remaining third of the mean probability results from early CFs that occur at vessel breach and the vast majority of these failures are caused by hydrogen combustion events. For the ATWS PDS group, the mean conditional probability of early CF is 0.76. Similar to the long term SB0 group, less than half of the early CF probability associated with the ATWS group is caused by hydrogen combustion events. This PDS group consists of both a long term PDS and a short term PDS. In the long term PDS the suppression pool is saturated and'either the operators vent the containment or the - - containn ent fails . before vessel breach from the accumulation- of steam in the containment. This PDS- is responsibic for a little more than half of this group's mean frequency. In the short term PDS,- on the the other hand, almost all of the early CF S.14-
l 1 probability is associated with failures that occur at the time of vessel breach. The pool is subcooled in the short term PDS, Although combustible mixtures can form in the containment before vessel breach in this PDS, the 111 5 is typically on during core damage and, therefore, the hydrogcn generated before vessel breach is usually burned such that the resulting load is benign. k:1v Drfwell Failure. Early drywell failure is an important attribute of the ac e.ident progression because failure of the drywell establishes a pathway fer radionuclides in the drywell to bypass the suppression pool. Bncause accidents that result in early drywell failure coincident with carly containment failure are generally the dominant risk contributors, it is appropriate to discuss the events that can lead to early drywell failure. Before vessel breach the only significant event that causes drywell failure is hydrogen combustion. Slow pressurization events associated with the accumulation of steam in the containment are not a threat to the drywell structure, for the short term SB0 PDS group, most of the failurcs are caused by deflagrations. A relatively small fraction of these failures is caused by detonations. The mean probability of drywell failure before vessel breach is considerably less for the other PDS groups. There are several reasons for the lower failure probability in these groups. In the long term SB0 PDS group the containment is frequently steam inert during this stage of the accident. In the ATWS PDS stoup, the containment is steam inert in some of the cases and in many of the other cases the 1115 is operating during core damage. In the T2 PDS group, the llIS is also generally operating during the core damage process. For dryvell failures that occur at vessel breach, loads accompanying vessel 4 breach are respor.sible for the majority of these f r.ilures. These quasi-static loads, which were provided by the Containment Loads Expert Panel, i include contributions from: DCll , ex vessel steam explosions, hydrogen burns, and RPV blow down. At vessel breach these events pressurize the drywell volume before the suppression pool vents clear. Nearly half of the drywell failures that occur at vessel breach are caused by these loads. In addition to directly pressurizing the drywell volume, these loads can also pressurize the reactor cavity and fail the pedestal. The loss of reactor support can induce dryvell failure. Roughly a quarter of the drywell failures that occur at vessel breach can be attributed to failure of the reactor pedestal. S,6 Source Term Analysis S.6.1 Descrintion of the Source Term Analysis The source term for a given bin consists of the release fractions for the I nine radionuclide classes for the early release and for the late release, I and additional information about the timing of the releases, the energy l associated with the releases, and the height of the releases. This source term comprises the information required for the calculation of consequences in the succeeding analysis. A source term is calculated for each APB for S.15-
l cach observation in the sartple. The nine radionuclide classes are: inert gases, iodine, cesium, tellurlwn, strontium, ruthenium, lanthanum, cerium, and barium. The source terin analysis is performed by a relatively small computer code: CCSOR. The purpose of this code is Im1 to calculat" behavior of the fission products from their chemical and physical prs. 2es and the flow and temperature conditions in the reactor and the cc .. ainment. Instead, GGSOR provides a means of incorporating into the analysis the results of the- more detailed codes that do consider these quantities. This approach is needed because the detailed codes require too inany computer resources to be able to compute source terms for the numerous accident progression bins and the 250 observations that result from the sampling approach used in NUREG 1150. CCSOR is a fast running, parametric computer code used to calculate the source terms for each ApB. for each observation for Crand Gulf. As there are typically about three hundred bins for each observation, and 2$0 observations in the samplo, the need for a source term calculation method that requires few computer resources for one evaluation is obvious. GGSOR provides a framework for synthesizing the results of experiments and mechanistic codes, as interpreted by expetts in the field. The reason for
" filtering
- the detailed code results through the experts is that no code available treats all the phenomena in a manner generally acceptable to those knowledgeable in the field. Thus, the experts are used to extend the code results in areas where the codes are deficient and to judge the applicability of the model predictions. They also factor in the latest experimental results and modify the code results in areas where the codes are known or suspected of oversimplifying. Since the majority of the parameters used to compute the source term are derived from distributions ,
determined by an expert panal, the dependence of GGSOR on various detailed codes reflects the preferences of the experts on the panel. It is not possible to perform a separate consequence calculation for each of the approximately 75,000 source terms computed for the Grand tulf integrated risk analysis. Therefore, the interface between the source term analysis and the consequence analysis is formed by grouping the source terms into a much smaller number of source term groups. These groups are defined so that the source. terms within them have similar properties, and a single consequence calculation is performed for the mean sourcs term for each group. This grouping of the source terms is performed with the PARTITION program, and the process is referred to as " partitioning. The partitioning process involves the following steps: definition of an early health effect weight (Ell) for ecch source term, definition of a chronic health effect weight (Cil) for cach source term, subdivision
-(partitioning) of the source terms on the basis of Ell and Cil, a further subdivision on the basis of the time the evacuation starts relative to the start of the release, and calculation of frequency weighted mean source terms.
S.16 a l
The result of the partitioning process is that the source term for each accident progression bin is assigned to a source term group. In the risk computations, each accident progression bin is represented by the mean source term for the group to which it is assigned, and the consequences calculated for that mean source term. S.6.2 Results of the Source Term Analysis When all the internally initiated accidents at Grand Gulf are considered together, the plots shown in Figure S.5 are obtained. These plots show four statistical measures of the 250 curves (one for each observation in the sample) that give the frequencies with which release fractions are exceeded. Figure S.5 summarises the complementary cumulative distribution functions (CCDFs) for all of the radionuclide groups except for the noble gases. The mean frequency of exceeding a release fraction of 0.10 for iodine and cesium is on the order of 10*S/ year and for tellurium and strontium it is on the order of 104/ year. The mean frequency of exceeding a release fraction of 0,01 for the La radionuclide class is on the order of 10'8/ year . S.7 Consecuence Annivsis S.7.1 Description of the Consecuence Analysis offsite consequences are calculated with the MELCOR Accident Consequence Code System (MACCS) for each of the source term groups defined in the partitioning process. MACCS tracks the dispersion of the radioactive material in the atmosphere from the pls.nt and computes its its deposition on the ground. MACCS then calculates the effects of this radioactivity on the population and the environment. Doses and the ensuing health effects from 60 radionuclides are computed for the following pathways: immersion or cloudshine, inhalation from the plume, groundshine, deposition on the skin, inhalation of resuspended ground contamination, ingestion of contaminated water and ingestion of contaminated food. __ MACCS treats atmospheric dispersion by the use of multiple, straight line caussian plumes. Each plume can have a different direction, duration, and initial radionuclide concentration. Cross vind dispersion is treated by a multi step function. Dry and wet deposition are treated as independent processes. The weather variability is treated by means of a stratified sampling proc.ss. For early exposure, the following- pathways are considered; immersion or cloudshine, inhalation from the plume, groundshine, depositicn on the skin, and inhalation of resuspended ground contamination. For the long term exposure, MACCS considers following four pathways: groundshine, inhalation of resuspended ground contamination, ingestion of contaminated water and ingestion of contaminated food. The direct exposure pathways, groundshine and inhalation of resuspended ground contamination, produce - doses in the population living in the area surround!.ng the plant. The indirect exposure pathways, ingestion of contaminated water and food,-produce doses in those who ingest food or water emanating from the area around the accident site.
~
S.17
l ' ,.i(' ![it i rE , t s r o
- t 0 0 E
a E i 0 0 t t . t i e . 1 n
. I
. . . . 'E0..
- t l i .
t.:..- - E z t n h.nC et oi0 h. l a O . L s , cer 5. .d 5S 9.ie f.
. ..'. a.
1 r S n r C . e
.P . ...' 2
' - 2
- r - - r t E o E o n 0 f 0 f I
, - 1 n
. 1 1 no '
N o l
,. ..- - 3l- c -
i
. 3t i
- c E or l
A
.- - E or ,
- 0 f - 0 t T
, . 1 , -
- a. e :
- a- e ,' .
s f
- s l 4 o 4 o
- e ,
Ef
-e G u
.- - Ef o e
- O R e s -
tR ,
. . . .a.' .
t d n _ - - 5 - 5 a
- r
_ - - E ,
. E G
. - 0 ,- , _.' _
0 1 1 r o 4 5 6 7 8 9 4 5 6 7 8 f E E E E- 0 E E E E E 0- E n 0 o 0 0 0 0 0 0 o 0 0 1 0 t o 1 t t t 1 1 1 1 t 1 i t mg,xIssoo.s. .se4w*t eocoy..UO i
?o*>'bp8.uueayw*heocov.oO t c
a r F 0 0 E e E s 0 0 a i] 1 , , - ,- - , - ,;~ , 1 e l 1 1 e
- R a-
- E-0 1 , ,.. . E 0
1 e r o
.- 2 i
- m. -
2 T f
- r .=
- r s
..
- 1 E0 of .'
- . E O f
o i e
- d- 1 n .
L n c o o n 3t i 3f l e E- e
- - '. u
- c
- E or ,
OF or q e
- 0 t T -
' L r
- a- e , - -
- a. e F
- s -
.- s
_ - 4 a - . 4 a El
-e ,
. E- e l e
c
- O e ,- 0 e n L R , . 1 R a a-
. d e
.- - 5 _ - 5
- e
- - - E- ,
. - . E c c: r 0 - _ .y -
O x
- - - - ~
1 t E 4 5 6 7 8 9 4 5 6 8 ,
- - - - - 7- .
E- E- E- E E E E E E- E E- c_ S o 0 0 0 0 0 0 0 0 0 0 O . t 1 1 1 t 1 1 1 t 1 L t S
^g,>Isooo.s6e4w*cevcov.UO .
to*Iboges.seayv2.eucOy..oa t e r t g i F m.5
lL , ~l ' - , , , 0 0 E E 0 0 I t ' t J _ _ _
.e _ .
t
. l
.f i
r i h 1 E
- E n et h oI h __ 0 a-O L
2 t o
. P c5sD r9u55___
e
- ~
t
. . ., .~ '.' -
1 2 B o
- r - -
- r
- E o - E o 0 f 0 f
,,*., s- 1 n o -
a-1 n o
- 3t i
- 3t i
' - c E- ocr
- E or 0 0 T t T
- e s
'4 1 e
s 4 a _
. . -~ -- e a
- El
-e .
. '*, El 0 e - O R e
,- 1 R -
'5 t
. ,. ,,. 's- ,- -
5 1 E , - E 0 0 t 1 4 5 G 7 8 9 4 5 6 7 8 9
- - - E E E E E 0 E E E 0 E E 0 0 o 0 0 0 0 0 0 0 0 D )
1 1 t 1 1 t t t 1 1 1 L d e
^'o* Lsge seawty vcove.ona r r u T,>es.pune ueaMt*& eucove.oO n n
i t n D 0 o E 0 0 ( c D 1 _ , - ,- L ~ ,i- ,_ - _ --
- 1
- 1
- S.
- E . - E- S 0 0 1 1 e e u " r 2 C
- R '
u
. .1- 2 ,....-
- r g
- r E- o
. .*..* - E o O f 0 f i
F L n 3t io ,
' 3t io t n
- c E- o,t Mo r - r (T - 0T e
s
' ' 14 a
e s 4 a El
-e ,
=
= E- e l
,- 0 e ,- 0 e t R t R
...',, -L _.
- - 5 - 5
.- ,- - E
- - E-
.w , 0 c : .'- , - , .- ,: 0
- ~
1 1 4 5 6 8 9 4 5 6 7 8 9
- 7- - -
E E- E E-E- E E- E E E- E 0 O 0 0 O 0 0 0 0 0 D 0 O L 1 L L 1 1 1 t 1 L 1 L T,>. es oToog s.oaMgEu .ocooeeuO -
^'oE[o ges. eavr y evcoveeonw t m.~@
,j; i,. .]J ',
1,;v-
. _. ~ _ _ ____ _ . _ - __. _. _ . _ _ _ ___ __
I i The contamination of water bodies is estimated for the washof f of land-deposited material as well as direct deposition. The food pathway model includes direct deposition onto the crop species and uptake from the soil. Both short term and long term mitigative measures are modeled in MACCS. Short term actions include evacuation, sheltering, and emergency relocation out of the emergency planning zone. Long term actions include relocation and restrictions on land use and crops. Relocation and land decontamination, interdiction, and condemnation are based on projected long term doses from groundshine and the inhalation of resuspended radioactivity. The disposal of agricultural products and the removal of farmland from crop production are based on ground contamination criteria. The health effects models link the dose received by an organ to morbidity or mortality. The models used in MACCS calculate both short term and long-term effects to a number of organs. Although the variables thou6 h t to be the largest contributors to the uncertainty in risk are sampled from distributions in the accident frequency, accident progression, and source term analyses, there is no analogous treatment of uncertainties in the consequence analysis. Variability in the weather is fully accounted for, but the uncertainty in other parameters such as the dry deposition velocity or the evacuation rate is not considered. The MACCS consequence model calculates a large number of different consequence measuraa. Eceults for the following cir, censequence measures are given in this report: early fatalities, total latent cancer fatalities,
- population dose within 50 miles, population dose for the entire region, l carly fatality risk within one mile, and latent cancer fatality risk within i 10 miles. For NUREG 1150, 99.5% of the populction evacuates and 0.5% or the population continues normal activity. For internal initiators at Grand Gulf, the evacuation delay time between warning and the beginning of evacuation is 1.25 h. -
S.7.2 Pesults of the Consecuence Annivsis l The results presented in this section are conditional on the occurrence of a source term group. That is, given that a release takes place, with release fractions and other characteristics as defined by one of the source term groups, then the tables and figures in this section give the consequences expected. This section contains no indication about the frequency with which these consequences may be expected. Implicit in the tcsulta given in this section are that 0.5% of the population does not i evacuate and that there is a 1.25 h delay between the warning to evacuate and the actual start of the evacuation. CCDFs display the results of the consequence calculation in a compact and complete form. The CCDFs in Figure S.6 for early fatalities and latent cancer fatalities display the relationship between consequence size and consequence frequency due . to variability in- the weather for each source term group which has a non zero frequency. Conditional on the occurrence S,20
i jQ Mmesmosaurs git kva -. 7? ., * *" *
*at-"+'*r***
I -k , N 'N'h \ l 10" N
'h
\ -
x
\
s ,
'N x gg r .I C
. \ \ T U 10
T \\ g k t
\
10 _ m \ 10 10 10" 10 10' # 10" 10 10 10* # 10* ' X, EARLY FATALITIES
,, GRAND GULF -
INTERNAL EVENTS
*'t' N 10 ' '
' \ -
x
$ \\
t 5 \ 10'# ' -
>l C ,
,\;
h10 - e m
\-
\
10 - l 10 10 10 10' ' 10 *
- 10 * ' 10 10 10 10
- x. LATENT CANCER FATALjTIES Figure S,6. Consequences Conditional on Source Terms.
Grand Gulf: Internal Initiators. S.21
. - -. _ . _ _ _ ._ . . _ . _ _ _ _ . _ _ _ . _ _ _ _ _ . _ _ . . _ . . _ _ _ _ _ ~ . . _ _ . _ _ _ _ _ _ _ _ .
I 1 l
\
of a release, each of these CCDFs gives the probability that individual ' consequence values will be exceeded due to the uncertainty in the weather conditions existing at the time of an accident. Figure S.6 shows that there is considerable variability in the consequences that is solely due t;o the weather. There is, of course, considerabic variability between source term groups that is due to the size and tirning of the reicase as well. S.8 Interrated Risk Analysis S.8.1 Determinntion of Risk Risk is determined by bringing together the results of the four conatituent analyses: the accident frequency analysis, the accident progression analysis, the source term analysis, and the consequence analysis. This process is described in general terms in Section S.2 of this summary, and in inathematical terms in Section 1.4 of this volume. Specifically, the accident frequency analysis produces a frequency for each PDS for each observation, and the accic'ent progression analysis resultn in a probability-for each APB, conditional on the occurrence of the PDS group. The absolute frequency for each bin for each observation is obtained by summing the product of the PDS frequency for that observation and the conditional probability for the APB for that observation over all the PDSs in the APB. A source term in calculated for each APB for each observation; this. source terrn is then assigned to a source terin group in the partitioning process. The consequences are then computed for each source term group. The overall result of the source term calculation, the partitioning, s.nd the consequence calculation is that a set of consequence values is identified with each APB for each observation. As the absolute frequency of each APB ir. known from the accident frequency and accident progression results, both frequency and consequences are known for . each APB. Tho ~ risk analysis assetables and analyzes all these separate estiinates of offsite risk. S.8.2 Resuirs of the Risk Analys{g Measures of Rink. Figure S.7 shows the basic results of the integrated risk analysis for internal initiators at Grand Gulf. This figure shows ! four statistical ineasures of the families of complementary cumalative distribution functions (CCDFs) for early fatalities, latent cancer fatalities, individual risk of early fatality within one mile of the site i boundary, and individual risk of latent cancer fatality within 10 tuiles of l the plant. The CCDFs display the relationship between the frequency of the consequence and the inagnitude of the consequence. As there are 250 l observations in the sampic for Grand Gulf, the actual risk results at the I rnost basic level are 250 CCDFs for each consequence measure. Figure S.7 displays the 5th percentile, toedian , inc an , and 95th percentile for these 250 curves, and shows the relationship between the tnagnitude of the consequence and the frequency at which the consequence is exceeded, as well as the variation in that relationship. The 5th and 95th percentile curves provide an indication of the spread between observations,- which is often large. This spread is due to l I S,22
. . . . . . - - . . _ _ . - - . . . _ . ..- - _. . . ~ . . . . . - - - - - -. - .. - . ~..
LOE-S , ,
. e m onia.
7 b tot-6 '......,,, . . {51b. . .- y,-h 3 -
..y9.... .
10E-7 'a., g
~
y$ ' 10E-8 '.,
)
e . l Lot-9 '., . 3
}. ....,5.., N.,, ,
O tot-10 '=> ' - "' > tot-4 tot-3 tot-2 LOC-t toto t001 1002 EOrly feiolltles tot-4 m T 7
- v. -
[ tot-s ,................................... ,, . 1 y .., . E tot-6 \. 1 s '. s . 1 y/c ,,,, r , f* tet-e Percenille i* 1 e
. 35.ttt.. i ',
r7FF ',' . , 1 q( tot-e . . N b. . . . ', i :
" 'UC" '" '
toc-51.0c-2 tot-i " oto 10ti t tot 2 tocs tot 4 tot 5 -
- LOleni COncer foiQlitit$
GRAND GULF BaseCase Figure S.7. Exceedance Frequencies for Risk, Grand Gulf: All Internal Initiators. i
?
i S.23
. _:. - _ .._~_. - -.. _ _ , . . _ - - - _ . - ..-_.._;
.- . . _ . . . . .-- - - . ~ . . - - _ _ - . .- .- . .. - . . ._. --
ih m6 tof-S g . . .a , g Percenille f ... 5.'h . . u ..................., n . i a tot-6 .... 33 , 1 l .. ,," ' ... .
~
.. >.S....
~
(toe-7 _
'.. . 1 10E-6 ; 3*
- i. -
{t0C-9 . ,,'*... , \, y
.,,s. ;- -
( ,% v 1 i i a ii tf
* ' I i*
t0E-s 100-7 tot-6 t0[- 5 t0E-4 t0[-3 t0E-2 Coriy Foiolity Risk Within 1 Wile t0E-4 , , , .., , .., , 1 =. 4 t. [ tot-s n...............................,,, 1
] ".. .
I t0[-6 ,. 1-g .', l v tog,7 , ............................... ,, . i i E
- 5 '. **, ;
ILOC-8 '
" I . 1 Mn.or.tita ~.
h..
-~
10E-9 een i. '..
,.~30m f
i
\
+
1 -
..P.%.... .\ '
b t0E-10 ' + ' ' ' "' ' 10E-8 tot-7 tot-6 t0E-5 t0E-4 100 t0E-2 Latent Concer Fotolity Risk Within 10 Miles GRAND GULF BaseCase Figure S 7. (continued) 1 S.24
. , - . _ _ _ _ _ . - _ _ _ . _ _ . _ _ _ _ - - _ . ~ - . . .- . -. - -
. - . - - . . _ . . . - . . . - - . - - . . - . . - .-. - . - . . . ~ . . - . . - - . - - . _ . - - - .- ._.
l l l
, tot-4
! b I 10C-5 I 3, 5 i 10[-6 ,
)
i '.-
- 1,,,,
1
., 1 tot-6 . '.
l y Percentile '. i, y i \ : 6 tot-9 . .-{5)h een . . '. i
=50tn
}y (
. . Pf b. . .
w tot-so 1 - 4 "
\,
\, ,l LOLD 10C1 LOC t0C3 t0[4 LOCS t0[6 tnt?
Population Dose (person-rem) Within 50 Miles 10t-4 T , E 1 t0E-5 b .." " " " " " " " " * " * " * " " * " " ' . . . . . , . 1-
%g
. t0E-6 ,
v [ - ( tot-7 ,....................................,'. r
\
3-t a., tot-B \ .
\' )
I Percentile i i . : -- tot-9 . . {8 0 ".ih . . 't '
', 3
, = botn i i:
..))b.... g t.
b10E-10 , , L tot, "I.ou' tots LOL' 1006 '" '" " Population Dose (person-rem) Within Region GRAND GULF BaseCose Figure S.7. (continued) i S.25
, . . . _ - - . _ . . _ . . -- .- - , - - . . , . . . _ , - - . - - ~ . _ , _ . . . _ . . - _ - . , _ .
1 i uncertainty in the sarnpled variables, and not to differences in the weather at the titne of the accident. As the tnagnitude of the consequence measure increases, the mean curve typically approaches or exceeds the 95th percentile curve. This results when the mean is dorninated by a few observations, which of ten happens for larte values of the consequences. Only a few observations have nonzero exceedance frequencies for these large 4 consequonees. Taken as a whoir, the results in Figure S.7 indicate that large consequences are relatively unlikely to occur. 4 Although the CCDFs convey the most information about the offsite risk, summary incasures are also useful. Such a summary value, denoted annual i risk, may be detertnined for each observation in the sataple by summing the a product of the frequencies and consequences for all the points used to j construct the CCDP. This has the effect of averaging over the different { weather states as well as over the different types of accidents that can
- occur. Since the complete analysis consisted of a sample of 250 observa-j tions, there are 250 values of annual risk for each consequence measure.
l These 250 values may be ranked and plotted as histograms, which is done in < a Figure S.8. The same four statistical measures used above are shown on
! these plots as well. Note that considerable intorination has been lost in going from the CCDFs in Figure S.7 to the histograms of annual values in j Figure S.8; the relationship between the size of the consequence and its frequency has been sacrificed to obtain a single value for risk for each
- observation.
l j The plots in Figure S.8 show the variation in the annual risk for internal
- initiators for four consequence ineasures. Where the inean is close to the l 95th percentile, a relatively small number of observations dominate the
, mean value. This is inore likely to occur for the early fatality consequence measures than for the latent cancer fatality or population dose consequence measures due to the threshold effect for early fatalities. l The safety goals are written in terms of mean individual fatality risks. The picts in Figurc S 8 for individual early fatality risk and individual latent cancer fatality risk show that essentially the entire risk
- distribution for Grand Gulf falls below the safety goals and the means are j also well below the safety goals.
l
; A single measure of risk for the entire sample inay be obtained by taking the mean value of the distribution for annual risk. This ineasure of risk is commonly called mean risk, although it is actually the average of the annual risk. Mean risk values for internal initiators fer four consequence measures are given in Figure S.8, S.8.3 Inmortant Contributors to Risk There are two ways to calculate the contribution to mean risk. The fractional contribution to mean risk (FCMR) is found by dividing the average risk for the subset of interest for the sample by the average total risk for the sample. The snean fractional contribution- to risk (MFCR) is i found by determining the ratio of the risk for the subset of interest to the total risk for each observation, and then averaging over the sample.
S.26
- , , . - - , , - - , , ,--, ,-,---,-,-re -
--,--,.n_ ,.,~-+.--v,, n.-, -, re- -
,n , ,
i i
^
x 1.E-6 F IE 1 E-6 , g 1 Ofety Goal LE-7 3 1 E2
'II~
N ' = L C E1ED 9'Wh . 3 d II-8 3
$,1E-8' ,
u__
~
1 O 1E-9' 1 :: i S .E-9l "-*
- gz-1 h I 3 2 1 ". ""- i f 1E-i% , y lE-Id I 1.E-2, Rh_. $ '
I E-tt, 8 ~ S IE-It c m- ,
'~
1E-t( . 1 tE-12' .
<1 E-12 ' '
g ,g PrebeMBRy PrebeMitty NT i ^ M - mean ; to m = median r l 6 th = p m
" II-5 j Iz_1 ' . 1 E3 ,
i a - l, u
,, Sofety GoeS \
l
. LE-2, 1.E2 ' -1 I l $.
. , ruh , n IE-7 3 l 2
-~
9%h - o 1 En ' 3 To W.- g u_. 3 iz-s, g 11-s, Y * ' s 2 1E-4_ g In ' a y
,1-9, = -
l , u i c g 1 E-1, wh_. 3 m_ , o j IE-fg , 3 iE-s- "' 8 - g IE-2, y tE-st 5th_. 3n. t ! i' v 1E-s 2 1E-3 S 1E-12 e prob W hty Pre.abihty Prot >.Nuty I , r l Figure S.8. Distributions of Annual Rfsk. Grand Gulf: All Internal Initiators. ) 4 i e ="w_ e _ _ . _ _ _ _ . _ _ _ _ _ _ _ _ _ _ . _ _ _ . _
Results of computing the contributions to the incan risk for internal initiators by the two methods are presented in Table S.3. Percentages are shown for early fatalities and latent cancer fatalities for the four svaunary PDS groups. l l Table S.3 Two Methods of Calculating Cor.tribution to Mean Risk l i Contributors (%) to Mean Early Fatality Risk for Internal Initiators fDS Croeo E.011B tiEra Fast SB0 93,2 84,1 Slow SB0 4.7 6.5 ATWS 2.0 7.9 T2 Trans. 0.2 1.5 Contributors (t) to Mean Latent Cancer Fatality Risk for Internal Initiators PDS Group EfEB MFCR i Fast SB0 91.3 85.3 Slow SB0 4.8 5.0 i ATWS 3.5 8.2 i T2 Trans. 0.4 1.5 l Pie charts for the contributions of the summary PDS- groups to inean risk- for internal initiators for these two risk measures for both methods are shown in Figure S.9. Figure S.10 displays similar pie charts for the contributions of the summary APBs to mean risk. Not surprisingly, the two methods of calculating contribution to risk yield different values. Because both methods of computing the contributions to rink are conceptually valid, the conclusion is clear: contributors to mean risk can only be interpreted-in a very broad sense. That is, it is valid to say that the short term SB0 groups is the maj or contributor to mean early fatality risk at Grand Gulf. It is not valid to state that the short term SB0 group contributes 93.2% of the early fatality risk at Grand Gulf. Although the exact values are different for each method, the basic conclusions that can be drawn frotn these results are the same. For all of the consequence ineasures, the mean risk is dominated by the short term SB0-PDS group. This group is the dominant contributor to the core damage irequency and because ac power is not initially available in these PDSs, there is a significant probability that these accidents will involve early l S.28 l l
Earl 8.2E-9/yReactor FotoLLtyyear nrCR FCMR
~..
v 4 I .3n 4 1: ._ x 3 2
.- 2 Latent Concer Fatalities
' % 9.5E-4/ Reactor year MFOR FCMR Summary PDS Group , 1: Short-Term 580 fy , 2: Long-Term 500 - 3: flTH5 * " 4: T2 4 I - 211_ 4 I -- , 3 3 p'S ,,,:. h . . .%. 2 i Figure.S.9. Fractional PDS Contri'outions to Annual Risk. Grand Gulf: All Internal Initiators. (MFCR - Mean Fractional Contribution to Risk; FCMR - Fractional Contribution to Mean Risk).
Early PotoLLty i FCMR 8.2E-9/ Reactor year MFco E i 1 4 < -
;- e 4
- 2. (([ NMf
/ :gg 4 _
, 3 <.
N __c
-~,.~
~
?}}
- '~[:;P4
% ~~:
~ ~
- 2 '4
." Latent Cancer Fatalities 8 9.5E-4/ Reactor year FCMR MFCR Summary ficcident Progression 3 ,, 7.
1: VB,Eorty CF, Early ,.:d s a SP Dypass, No CS 2: VB,CorL3 CF, Corty ig : -: \ SP Byposs, CS f:::__:::~:1, fg@_ :: :.: -s 3: VB,Eorly CF, l ;.- Lot
- SP BSPoSS 1-: 'w~
3--:-.
~
4 in ' 4: VD,Corly CF, : :~ k [ No SP D3poss :M 8 Si VB, Lote CF 4k -][- 6: VB, Venting ( , 6 7: VO, No CF ' t; 8: No VB 5 . 3 i Figure S.10. Fractional ABP Contributions to Annual Risk. Grand Gulf: All Internal Initiators. (MFCR - Mean Fractional Contribution to Risk; FCMR - Fractional Contribution to Mean Risk) .
l' l l I containment failure and vessel breach. Thus, these accidents are not only the most frequent but they also involve accidents that can potentially result in a large early release. Tne long term SB0 group and the ATUS group contribute considerably less to these risk measures and the T2 group is a very minor contributor. For early fatalities, which depend on a large early release, the risk is dominated by accidents that progress to vessel breach and that involve early containment failures. Accidents in which the containment fails late are much less significant. In Figure S.10 the first bin (vessel breach, Early CF, Early SP Bypass, No CS) is the dominant contributor to these risk measures because the containment fails early and the releases at vessel breach and after vessel breach are not scrubbed by either the pool or the containment sprays, Although the fourth bin in Figure S.10 (vessel breach, Early CF, No SP Bypass) does not involve drywell failure, its contribution to early fatality risk is higher than the second bin (vessel breach, Early CF, Early SP Bypass, CS Avail . ) in which the drywell fails early in the accident, The reason f or this is that the mean probability of the fourth bin is roughly four times the mean probability of the second bin. Thus, i although the fourth bin does not involve drywell failure, the probability of this bin coupled with the fact that the containment fails early is sufficient to make this bin a significant contributor to early fatality risk. Latent cancer fatalities depend primarily on the total amount of radioactivity released. Thus, unlike early fatality risk, the timing of containment failure is not particularly important for this risk measure. If the suppression pool is bypassed there is a greater likelihood that the-release will be large, Thus, accidents in which some of the releases are not scrubbed by either the pool or the sprays tend to contribute more to latent cancer fatality risk than accidents in which the drywell remains intact. It is for this reason that the first bin in Figure S.10 (vessel breach, Early CF, Early SP Bypass, No CS) is the dominant contributor to the latent cancer fatality risk. The bin that involves accidents in which the vessel does not fail makes a minor contribution to the early fatality risk; however, it- makes a noticeable contribucion to the latent cancer fatality risk. It must be remembered that although the vessel does not fail in these accidents, the containment can selli fail early in these accidents from the combustion of hydrogen in the wetwell. Early failure of the containment will allow a portion of the in vessel releases to escape into the environment. The combination of the threshold effect associated with early fatalities with the fact that the releases associated with this bin are fairly small results in few early fatalities. For latent cancers, on the other hand, there is no threshold effect. Thus any releases that are not trapped by the suppression pool or removed by the containment sprays can contribute to the latent cancer risk. S,31
I l S 8.4 Important Contributors to the Uncertainty in Risk The important contributors to the uncertainty in internally initiated risk are determined by performing regression based sensitivity analyses for the o mean values for risk. The regression analyses = for. carly fatalities and individual risk of early -fatality within 1 mile only account for about 45% ' of the observed variability. The independent variables - that account for this variability are those that determine the frequency and the magnitude of an early release. The regression analyses for the other four consequence measures are somewhat more successful as they are able_ to account for about 60% of the variability, _The. independent variables that account for this variability are predominantly those variables that determine the frequencies of ;he accident. 1 S9 Innights and Conclusions il Core Damare Arrest. For the dominant summary PDS group, short term SBO, there is a significant probability that the core damage _ process _ will be arrested. and vessel failure will be averted. - For the accidents in which the vessel does not fail, there are no ex vessel fission product releases (e.6., DCil or CCI). Furthermore, loads accompanying vessel breach, which ' pose a significant challenge- to both- the drywell and the containment, are avoided. . The conditional probability of core damage arrest in the short-term SB0 PDS group is driven by the ac-power-recovery probability. In the other summary PDS groups (i.e. , long-term SBO, ATWS, and T2) it is unlikely that core damage process will be arrested. - The core damage arrest probability for ' the long term SB0 group -is low because the -probability of recovering ac power early in the accident is fairly low for this PDS group. . In the ATWS and T2 PDS groups the low values for. core damage arrest are attributed to fairly high likelihood that _. the operators fail to depressurize the RPV to allow coolant injection to be restored to the core. Containment Failung. Given that core damage occurs, itiis likely that the containment will fail during the course of the accident. Furthermore, for the dominant PDS summary group, short-term SBO, there is a substantial probability that the containment will fail'early in the accident. liydrogen combustion events are the dominant events that cause early CF in the short-term S_BO . and T2 PDS groups. The combinatior. of a relatively weak containment, the copious production of hydrogen during core damage, and the unavailability of the llIS during _ a SB0 leade to a high ' conditional - probability of containment failure. For these two groups , the mean probability of early containment failure is approximately 0.5. In the short term SB0 group _ about half of the early_ CFs _ occur before vessel breach and the other half _ occur shortly after vessel breach. In the T2 PDS group the vast majority of the early containment failures _ occur around the time of vessel breacl,. For both the long-term SB0 PDS group and the ATVS PDS group, hydrogen combustion events and pressurization of the containment from the accumulation of steam contribute to their high - conditional probabilities of early containment failure. Drywell Failurg. Early drywell f ailure is an important attribute of the accident progression because failure of the drywell establishes a pathway-S.32
- _ _ _ _ - _ _ - _ _ _ _ _ - _ - __ _ 1
for radionuclides in the drywell to bypass the suppression pool. The suppression pool offers an important mechanism for redt'r ir.g dc = L'.s co nn . Accidents that result in early drywell failure coincident with early containment failure are generally the dominant contributors to risk. Of the accidents that result in early containment failure, roughly half of them also involve early drywell failure. Early drywell failures include failures that occur before vessel breach and failures that occur at vessel breach. Only the short-term SB0 PDS group has significant probability . of drywell failure before vesael breach. The vast majority of these drywell failures are caused by hydrogen combustion events. All of the PDS groups have a significant probability of drywell failure at the time of vessel breach. The majority of these failures are caused by loads accompanying ' vessel breach. These quasi-static loads include contributions from DCH, ex vessel steam explosions, hydrogen burns and RPV blow down. Flssion Product Releases. 'here ** vonsiderable uncertainty in the release fractions for all types o. ,nts. There are several features of the Grand Gulf plant that tend tv mitigate the release. First, the in vessel releases are generally directed to the suppression pool where they are subjected to the pool decontamination factor. Provided the drywell has not failed, the radionuclides released into the drywell will also pass through the pool. Although generally not as effective as the suppression pool, the containment sprays and the reactor cavity pool also offer a mechanisms for reducing the release of radionuclides from the containment " hen the suppression pool has been bypassed. The largest releases tend to occur when the suppression pool is bypassed and the containment sprays are not operating. Bigli. The offsite risk from internal initiating events wos found to be quite low, both with respect to the safety goals and to the other plants analyzed in NUREC-1150. The offsite risk is dominated by short-term SB0 PDSs. The long-term SB0 group and the ATWS group contribute considerably less to these risk measures and the T2 group is a very minor contributor. The low vnhes for risk can be attributed to the low core damage frequency, the good emergency response, and plant features that reduce the potential source term. Uncertainty in Risk. Considerable uncertainty is associated with the risk estimates produced in this analysis. The largest contributors to this uncertainty are the uncertainties in the parameters that determine the frequency of core damage and the uncertainty in some of the parameters that determine the magnitude of the fission product release to the environment. Propagation of the uncertainties in the accident frequency, accident progression, and source term analyses through to risk - allows the uncertainty to be quantitatively calculated and displayed. Conmarison with the Safety Coals. For both the individual risk of early fatality within one mile of the site boundary and the individual risk of latent cancer fatality within 10 miles, the 95th percentile value for annual risk falls nearly three ordera of magnitude below the safety goals. Furthermore, for both of these risk measures, the maximum of the 250 values that make up the annuni risk distributions also falls well below the safety goal. S.33 ) l __-__-_________- . _ l
Re fe rences
- 1. U.S. Nuclear Regulatory Commission, " Severe Accident Riska: An Assessment for Five U.S. Nuclear Powec Plants", Second Draft for_raer-Review, NUREG 1150, June 1989.
2 M. T. Drouin, J. L. LaChance, B. J. Shapiro, S. Miller, and T. A. Wheeler, " Analysis of Core Damage Frequency: Grand Gulf, Unit 1 Internal Events," NUREG/CR 4550, SAND 86-2084, Vol. 6 Rev. 1, Sandia-National Laboratories, September 1989. S.34
- 1. INTRODUCTION l
; The United States Nuclear Regulatory Commission (NRC) has recently completed a major study to provide a current characterization of severe accident risks from light water reactors (LVRs). The characterization was derived from the analysis of five plants. The report of that work, NUREG-11501 has recently been issued as a second draf t for comment. NUREC-1150 is based on extensive investigations by NRC contractors. Several series of reports document these analyses as discussed in the Foreword.
These risk assessments can generally be characterized as consisting of four analysis steps, an integration step, and an uncertainty step.
- 1. Accident frequency analysis: the determination of the likelihood and nature of accidents that result in the onset of core damage.
- 2. Accident progression analysis: an investigation of the core damage procass, both within the reactor vessel before it fails and in the containment afterwards, and the resultaat impact on the containment.
- 3. Source term analysis: an estimation of the radionuclide transport within the reactor coolant system (RCS) and the containment, and the magnitude of the subsequent releases to the environment.
- 4. Consequence analysis: the calculation of the offsite consequences in terms of health effects and financial impact.
- 5. Risk integration: the combination of the outputs of the previous tasks into an overall expression of risk.
- 6. Uncertainty analysis: the determination of which uncertainties in the preceding analyses contribute the most to the uncertainty in risk.
, This volume is one of seven that comprise NUREC/CR-4551. NUREC/CR-4551 l presents the details of the last five of the six analyses listed above. The analyses reported here start with the onset of core damage and conclude with an integrated estimate of overall risk and uncertainty in risk. This volume, Volume 6, describes these analyses, the inputs utilized in them, and the results obtained, for Grand Gulf Nuclear Station, Unit 1. The methods utilized in these analyses are described in detail in Volume 1 of this report and are only briefly discussed here, l 1.1 Backcround and Obiectivas of NUREG-1150 Assessment of risk from the operation of nuclear power plants, involves determination of the likelihood of various accident sequences and their potential offsite consequences. In 1975, the NRC completed the first comprehensive study of the probabilities and consequences-of core-meltdown accidents--the " Reactor Safety Studyd (RSS).2 This report showed that the probabilities of such accidents were higher than previausly believed, but that the consequences were significantly lower. The product of probability 1.1
and consequence--a measure of the- risk of core molt . accidents was estimated to be quite low when compared:with natural events such as floods and earthquakes and with other societal risks such ~ as automobile and-airplane accidents. Since that" time, . many risk _ assessments of -specific, plants have been performed. In general, each of these 'has progressively- i reflected at least some of the advances thati have been made in reactor safety and in the ability, to predict the- frequency Tof several accidents, the amount of radioactive material released as a result of such accidents, and the offsite consequences of such atrelease. In order to investigate the significance - of more recent _ developments -in_ a : comprehensive fashion, it was concluded that ' the current efforts: of research programs being sponsored by- the- NRC should be . coalesced to produce an updated representation of risk for operatin6 -nuclear power plants.
" Severe Accident Risks: An Assessment for Five U.S. Nuc1 car- Power Plants"1 is the result of this program. The five nuclear power plants are Surry, Peach Bottom, Sequoyah,' Grand Gulf, and Zion. : The analyses _ _of - the . first four plants were performed by Sendt e ' Nadend Laboratories . ' (SNL) . The -
analysis of Zion was _ performed by Idaho National Engineering' Laboratory (INEL) and Brookhaven National Laboratory (BNL). The following are overall objectives of the NUREG 1150 program.
- 1. Provide a current assessment of the severe accident risks to the public from five nuclear power plants, which will:
-l
- a. Provide a " snapshot" - of : the risks reflecting _ plant = design = and
~
operational characteristics, related . failure data,' ~and severe accident phenomenological information extant-in 1988;
- b. Update. the estimates of the NRC's : 1975 risk assessment the-
" Reactor Safety Study".;2
- c. Include quantitative' estimates of. risk uncertainty, in response to the principal criticism of_the " Reactor Safety _ Study"; and -
- d. Identify plant-specific-risk vulnerabilities, in.the context of the NRC's individual plant examination process.--
- 2. Summarize the perspectives gained in performing - these risk:
analyses, with respect _to:
- a. Issues- s ignificant - to severe accident frequencies,-_
consequences, and risk;
- b. Uncerteinties - for which the risk'is significant.an'd which may merit-further research; and'
- c. Potential for risk reduction.
- 3. Provide a set of methods for the prioritization of potential safety issues and related research.
1.2
l t These objectives required special considerations in the selection und development of the analysis methods. This report describes those special considerations and the solutions impl ew nted in the analyses supporting l NUREG 1150. 1.2 overview of Grand Gulf Nuclear Station. Unit 1 The subject of the analyses reported in this volume is the Grand Gulf Nuclear Station, Unit 1. It is operated by System Energy Resources Inc. (SERI) and ic located on the east bank of the Mississippi river in southwestern Mississippi, about 6 miles northwest of Port Gibson, Mississippi. Inc nearest large city is Jackson, Mississippi, approximately 55 miles to the northeast cf the plant. The nuclear reactor of Grand Gulf Unit 1 is a 3833 MWt BWR-6 boiling water reactor (BWR) designed and supplied by General Electric Gompany, Unit 1, constructed by Bechtel Gorporation, began commercial operation in July 1985. Grand Gulf has three diesel generators (DGs) that are used to supply emergency ac power in the event that offsite power from the grid is lost. One of these DGs is dedicated to the high pressure core spray inj ection l system (HPCS); the other two DGs supply ac power to two trains of emergency l systems. In the event of an accident there are several systems that can ! supply coolant injection to the core. Two systems are available to provide ' ! high pressure coolant injection: the high pressure core spray system (HPCS) and the reactor core isolation cooling system (RCIC). HPCS has a motor-l driven pump and can supply injection when the vessel pressure is either high or low. RCIC, on the other hand, uses a turbine driven pump and can only be used when the vessel pressure is high. Both tbc low pressure core spray system (LPCS) and the low pressure coolant injection system (LPCI) can provide coolant injection to the reactor vessel during accidents in which the system pressure is low. Both systems use motor driven pumps. LPCS has one train whereas LPCI consists of three trains. Additional i systems that can be used as backup sources of coolant injection arc ~ the standby service water crosstie system, firewater system, control rod drive system, and the condensate system. To allow low pressure injection systems i to supply coolant to the vessel, the automatic depressurization system l (ADS) is used to depressurize the reactor vessel. This system uses eight relief valves to direct the vessel steam to the suppression pool. The Grand Gulf containment is a Mark III BWR containment. The containment I is a steel-lined reinforced concrete structure. In the Mark III design the reactor pressure vessel (RPV) is housed in the drywell which is in turn completely enclosed in the containment structure. The drywell and the containment communicate through passive vents in the suppression pool. Figure 1.1 shows a section through the Grand Gulf containment. During an accident, steam from the vessel -is - directed through the safety / relief valves and is discharged through a sparger into the suppression pool. The steam is condensed in the pool and any noncondensible gases pass through the pool inte the containment atmosphere. Similarly, any steam and noncondensible gases released into the drywell are vented into the suppression pool. The design pressure of the Grand Gulf containment is 1.3
Containment Structure O O O O e # Containment -O O Sprays O O e e e Hydrogen O O ' igniter (90) r
. e e e M .
..' Polar Crane .- ,
g - 9. 4 l;o e e 8 ;;
;.g.
o
.o 9 , .i.
* :w t .
,. l , _ - - -
._' % .~,. .
, o i .. .,
s e g p..r.... c ,s . .I i i
' . 3. . : . . . . . ,
. E. '
- i . .
- n.
s
- c. ',[". . . . .-
?.. .- --
. 3.. ... .. . .. .: ,..
.,. ,. ,, Reactor Shield '.'c
?:-
', / Wall e i. .
? e >;; . ..', e 2:
- ..- J..
Vent ' 4 ,;l.~.1.;;; g Ig - < Welt -' l AnnulusN n .. Wall ;
' . ,- -.N .
m x.
? .-
" .i (.
o 1.
.-Y/
(.e.-. , ,. t'
/
Horizontal Vents
.. ?.. . /
Suppression ; }; Reactor
*- 15--
5 -
. .k.i -
- :1 Pool % l4.. 4
- w - i Cavity ': ?.;
- . t k=
. 9. u.
- .~d.:' '. . <
4:.
- .. ; o . .r .
i n' t 2 m
>t onta nment p9,,, y,,
; .' .s
. a. . ; . ;;. .:.; p
- .'.': *. t'. _'
'. e...
,.*S.
.. 1, r. y..:'.. ; ;?
.. .*:/
' . . * .. ., ;*. .,. :* . ,. .y:;
. , .*72 :/.;; '.':
Figure 1.1. Section of Grand Gulf Containment. i l 4 1.4 l
1 l l 15 psig (103 kPa). Although the design pressure is fairly low, the volume of the containment is comparable with a large PWR containment (1.67 million cubic feet). To suppress the pressure in the containment during an accident, two trains of containment sprays are located in the Grand Gulf containment. The containment spray system is one mode of the residual heat removal system (RHR). In the event that the RHR system fails to suppress the pressure in the containment, the containment can be vented. To reduce the potential of a severe hydrogen combustion event during an accident, the containment has a hydrogen ignition system (HIS). This system is designed to prevent the buildup of large quantities of hydrogen inside the containment. Igniters are located throughout the containment-and drywell volumes. Section 2.1 of this volume contains more detail - on the plant's features important to the progression of the accident and to the containment's performance. 1.3 Chances Since the Draft Renort The Grand Gulf analyses for the February 1987 draf t of NUREG-1150 were presented in Volume 4 of the original " Draft for Comment" . versions of 9Ah/CL Mil EM "WR/f!1000, pdlish02 in April lE. % as.aly... performed for NUREG ll50, Second Draft for Peer Review, June 1989, and reported in this volume, are completely new. While they build on the previous analyses and the basic approach is the same, very little from the first analyses is used directly in these analyses. This section presents the major differences between the two- analyses. Essentially, the accident g progression analysis and the source term analysis were completely redone to [' incorporate new information and to take advantage of expanded methods and analysis capabilities. Quantification. A major change since the previous analyses is the expert e11 citation process used to quantify variables and parameters thought to be large contributors to the uncertainty in risk. This process was used both for the accident progression analysis and the source term analysis. The sizes of the panels were expanded, with each panel containing experts from industry and academia in addition to experts from NRC contractors. The number of issues addressed was also increased to about thirty. Separate panels of experts were convened for In Vessel Processes, Containment Loads, Containment Structural Response, Molten Core-Containment Interactions, and Source Term Issues. To ensure that expert opinion was obtained in a manner consistent with the state of the art in this area, specialists in the process of obtaining expert judgments in an unbiased fashion were involved in designing the clicitation process, explaining it to the experts, and trainin6 them in the methods used. Tbc experts.were given several months between the meeting at which the problem was defined and the meeting at which their opinions were elicited so that they could review the literature, discuss the problem with colleagues, and perform independent analyses. The results of the 1.5
i elicitation of each expert were carefully recorded, and the reasoning of each expert and the process by which their individual conclusions were ag regated into the final distribution are thoroughly documented. Accident Procression Annivois. Not only was a substantial fraction of the Accident Progression Event Tree (APET) for Crand Gulf rewritten for this analysis, but the capabilities of EVNTRE, the code that evaluates the APET, were considerably expanded. The maj or improvements to EVNTRE were the ability to utilize user functions and the ability to treat continuous distributions. A user function is a FORTRAN subprogram which is linked with the EVNTRE code. When referenced in the APET, the user function is evaluated to perform calculations too complex to be handled directly in the APET. In the current Grand Gulf APET, the user function is called to: determine the containment baseline pressure during the various time periods; compute the amount of hydrogen released to the containment at the time of vessel breach and during CCI; compute the concentration and the flammability of the atmosphere in the containment and drywell during the various time periods; calculate the pressure rise due to hydrogen burns; determine whether the containment fails and the mode of failure; determine whether the drywell fails and the mode of failure. These problems were handled in a much simpler fashion in *:he previous analysis. The event tree used for the analysis for the 1987 draft of NUREG-1150 could only treat discrete distributions. In the analysis reported- here auG%m distributions are used. Use of continuous distributions removes a significant constraint from the expert elicitations and elitainates any errors introduced by discrete levels in the previous analysis. The event tree that forms the basis of this analysis was modified to address new issues and to incorporate new information. Thus, not only was the structure of the tree changed but new information was used to quantify the tree. A major modification was the way hydrogen combustion events were modeled and quantified. The amount of hydrogen in the - containment is tracked throughout the accident. The ignition probability, detonation probability and the loads from a combustion event are all a function of the hydrogen concentration. In the current APET, loads are assigned to both deflagrations and detonations. These loads are then compared to the structural capacity of the containment to determine whether it fails or not and the mode of failure. In this analysis, drywell failure from deflagrations is also considered. In addition to combustion events, another major change in the APET is the section that addresses vessel breach. In-vessel steam explosions are now addressed in the tree. Furthermore, the tree was modified to incorporate new information supplied-- by the Containment Loads Expert Panel on loads accompanying vessel breach. Pressurization of the drywell and pressurization the reactor cavity from events at vessel breach are considered. Failure of the reactor pedestal at vessel breach was not included in the previous analysis. Because of changes in the accident progression analysis and the source term analysis, the definitions of bins used to group the results from the aceident progression analysis have also changed. 1.6
Source Term Analysis. While the basic parametric approach uced in the original version of CGSOR, the code used to compute source terms, has been retained in the present version of GCSOR, the code has been completely rewritten with a different orientation. The current version of GGSOR is quite different. First, it is not tied to the source term code package (STCP) in any way. It was recognized before the new version was developed that most of the parameters would come from continuous distributions defined by an expert panel. Thus, the current version does not rely on results from the STCP or any other specific code. The experts utilized the results of one or more codes in deriving their distributions, but CGSOR itself merely combines the parameters defined by the expert panel. Finally, a new method to group the source terms computed by GGSOR has been devised. A source term is calculated for each accident progression bin for each observation in the sample. As a result, there are too many source terms to perform a consequence calculation for each and the source terms have to be grouped before the consequence calculations are performed. The " clustering" method utilized in the previous analysis was somewhat subjective and not as reproducible as desired. The new " partitioning" scheme developed for grouping the source terms in this analysis eliminates these problems. Consecuence Analysis. The consequence analysis for the current NUREC-1150 version 1 does not differ so markedly from that for the previous version of NUREG 1150 as does the accident progression analysis and the source term analysis. Version 1.4 of MACCS was used for the original analysis, while version 1,5 is used for thi analysis. The major difference between the two versions is in the data used in the lung model. Version.1,4 used the lung data contained in the original version of " Health Effects Models for Nuclear Power Plant Accident Consequence Analysis",3 whereas version 1.5 of MACCS uses the lung data from Revision 1 (1989) of this report.' Other changes were made to the structure of the cck in the transition from 1.4 to 1.5, but the effects of these changes on the consequence verlues calculated are small. Another difference in the consequence calculation is that.the NRC specified evacuation of 99.5s of the population in the evacuation area for this analysis, as compared with the previous analysis in which 95% of the population was evacuated. Risk Analysis. The risk analysis combines the results of the - accident - frequency analysis, the accident progression analysis, the source term analysis, and the consequence analysis to obtain estimates of risk to the This
~
offsite population and the uncertainty in those estimates. combination of the results of the constituent analyses was performed essentially the same way for both the previous and the current analyser. The only differences are in the number of variables sampled and the number of observations in the sample, 1.7
/
1.4 Structure of the Analysis The analysis of the Grand Gulf plant for NUREG-1150 is a Level 3 probabilistic risk assessment composed of four constituent analyses:
- 1. Accident frequency analysis, which estimates the frequency of core damage for all significant initiating events;
- 2. Accident progression analysis, which determines the possible ways in which an accident could evolve given core damage;
- 3. Source term analysis, which estimates the source terms (i.e.,-
environmental releases) for specific accident conditions; and 4, Consequence analysis, which estimates the health and economic impacts of the individual source terms. Each of these analyses is a substantial undertaking in itself. By taking care to carefully define the interfaces between these individual analyses, the transfer of information is facilitated. At the completion of each constituent analysis, intermediate results are generated for presentation and interpretation. An overview of the assembly of these components into an integrated analysis is shown in Figure 1.2. The NUREG-1150 plant studies are fully integrated probabilistic risk assessments in the sense that calculations leading to both risk and uncertainty in risk are carried through all. four components of the individual plant studies. The frequency of the initiating event, the conditional probability of - the paths leading to the consequence, and the value of the consequence itself can then be combined to obtain a risk measure. Maasures of uncertainty in risk are obtained by repeating the calculation just indicated many times with different values for important parameters. This provides a distribution of risk estimates that is' a measure of the uncertainty in risk. It is important to recognize that a probabilistic risk assessment is a procedure for assembling and organizing information from many sources; the models actually used in the computational framework of a probabilistic risk assessment serve to organize this information, and as a result, are rarely as detailed as most of the models that are actually used in the original generation of this information. In order to capture the uncertainties, the first three of the four constituent analyses attempt to utilize all available sources of information for each analysis component, including past observational data, experimental data, mechanistic modeling and, as appropriate or necessary, expert judgment. This requires the use of relatively quick running models to assemble and manipulate the data developed for each analysis. To facilitate both the conceptual description and the computational implementation of the NUREG-1150 analyses, a matrix representationb6 is used to show how the overall integrated analysis fits together and how the progression of an accident can be traced from initiatin6 event to offsite consequences. 1.8
= _ . __ _. -
ij i i j E - S _ R [ S m I U m A - R P m m S E C N - t
, u O
A A E S E T T A A AT N C LD DA LO O C 0 - N S ATA C1 5 OCr CC I E S OEGM S C 1 U Q L Y
- L F O A 0 C 1
- FEOR L A G
E A D A THR O M S N R AT L Eu OO E N A E R O E ED U U _ HM A N C R T T N S
- O EI n
C Tt i RT UR OA s SP i s y T 3h l UT - - a S OR ; n E RO A I A P CM S NS R R YL
- ON
- A T L t
U E A OT AR w T A v n a S N R T C O R O s l A 0JM 5 t E i S P D X NOR CRPN D d R' T N A N y S e E O O E . , t MIS T S F C 'S a A R S t N E S N E O r A F D E N r E U S g C GR B S P e I _ Q # S C O E E E t A R C U E R P F N S A S L P n I U I MN A N _ 7O$h _ R O S f TO
. : O
- E I
S F R S S S o N S ! E E U C E S S N P R G R w e D E Y 0E E & E
,]E
- T O L N9 GR P i C R A I
R Y E S T A mP H
,t v
C G N DE T OR N T N C T A L N r A RO A C G F A ROv ,bE v OE L u T A e v E L P T L O P : N E E E S G S A E uS u A E T C' N E mENV T 2 CA U I T T M N S C C 1 _ E A E l
- L R D RQ P F S e CltA E U
r T N F A C A S S u g T CR y s Y uO A V D i T C S e T t A D E T N f F N N S I uR OE T E D I E U LY
- ED O T
- S N
O N C R C Q A I E S F S S C E N ) T E - A R A F L S T U C R P X AN E uN Ev E A T F S TL VO Y N A S T O R A S NS S O N p F E E TN SL A i S N ST O A E x { DN C t D t T f D E P O A u. Et UDN MR A t A R E T OP F X E N E C H~*
l l Accident Frecuency Analysis. The accident frequency analysis uses event tree and fault tree techniques to investigate the manner in which various initiating events can lead to core damage. In initial detailed analyses, the SETS program 7 is used to combine experimental data, past observational data and modeling results into estimates of core damage frequency. The ultimate outcome of the initial accident frequency analysis for each plant is a group of minimal cut sets that lead to core damage. Detailed descriptions of the systems analyses for the individual plants are available elsewhere.08.1b u.12 For the final integrateo NUREG 1150 analysis for each plant, the group of risk significant minimal cut sets is used as the systems model. In the integrated analysis, the TEMAC programu u is used to evaluate the minimal cut sets. The minimal cut sets themselves are grouped into PDSa, where all minimal cut sets in a PDS provide a similar set of conditions for the subsequent accident progression analysis. Thus, the PDSs form the interface between the accident frequency analysis and the accident progression analysis. With use of the transition matrix notation, the accident progression analysis may be represented by fPDS - fIE P(IE*PDS), (Eq. 1.1) where fPDS is the vector of frequencies for the PDSs, fIE is the vector of frequencies for the initiating events, and P(IE*PDS) is the matrix of transttion probabilities from initiating events to the PDSs. Specifically: fIE - ( flE g, ..., f1Enig ) , fIE 3 - frequency (yr'1)_ for initiating event i, nIE - number of initiating events, fPDS - [fPDS3 , .. , fPDS yp3), n fPDS) - frequency (yr-1) for plant damage state j , nPDS - number of PDSs, pFDSn ... pFDS ,npos 2 ._ pFDSnig,t ... pPDSnIE.nPDS and pPDSy - probability that initiating event i will lead to plant damage state j. The elements pFDS y of P(IE*PDS) are conditional probabilities: given that initiating event i has occurred, pFDSg- is the probability that plant damage state j will also occur. The elements of P(IE*PDS) are determined by the analysis of the minimal cut sets with the TEMAC program. In turn, both the cut sets and the cata used in their analysis come from earlier studies that draw on many sources of information. Thus, although the elements pFDS U f P(IE*PDS) are represented as though they are single numbers, in practice these elements are functions of the many_ sources of information that went into the accident frequency analysis. 1.10
Accident Progression Annivsis. The accident pro 6ression analysis uses event tree techniques to determine the possible ways in which an accident might evolve from each PDS. Specifically, a single event tree is de'reloped for each plant and evaluated with the EVNTRE computer program.15 The definition of each PDS provides enough information to define the initial conditions for the accident progression event tree (APET) analysis, Due to the large number of questions in the Grand Gulf APET and the fact that many of these questions have more than two outcomes, there are far too many paths through each tree' to permit their individual consideration in sub s equent source term and consequence analysis. Therefore, the paths through the trees are grouped into accident progression bins, where cach bin is a group of paths through the event tree that define a similar set of conditions for source term analysis. The properties of each accident progression bin define the initial conditions for the estimation of the source term. Past observations, experimental data, mechanistic code calculations, and expert judgment were used in the development and parameterization of the model for accident progression that is embodied in the APET, The transition matrix representation for the accident progression analysis ic fAPB - fPDS P(PDS*APB), (Eq. 1.2) where fPDS is the vector of frequencies for the PDSs defined in Eq. 1.1, fAPB is the vector of frequencies for the accident progression bins, and P(PDS4APB) is the matrix of transition probabilities from PDSs to accident progression bins. Specifically: fAPB - ( fAPB 3, ..., fAPBnArali l fAPBx - frequency (yr*1) for accident progression ! bin k, nAPB - number of accident progression bins, pAPBit . . . pAPB ,nxp3 1 P(PDS4A2-pAPBnpp3,3 ... pAPB npos,ngp3 and pAPB3- probability that plant damage state j will lead to accident progression bin k. The properties of f PDS are given in conjunction with Eq.1.1. The elements pAI'B3 of P(PDS-+APB) are determined in the accident progression analysis by evaluating the APET with EVNTRE for each PDS group. Source Term Analysis. The source terms are calculated for each APB with a non zero conditional probability by a fast-runaing parametric computer code 1.11
entitled CGSOR. GCSOR is not a detailed mechanistic model and is not designed to simulate the fission product transport, physics, and chemistry from first principles. Instead, CGSOR integrates the results of many detailed codes and the conclusions of many experts. The experts, in turn, i based many of their conclusions on the results of calculations with codes such as the Source Term Code Package,18.17 MELCOR,18 and MAAP.18 Most of the parameters utilized calculating the fission product release fractions in-GGSOR are sampled from distributions provided by an expert panel. Because of the large number of APBs , use of fast executing code like CGSOR is absolutely necessary. The number of APBs for which source terms are calculated is so large that it was not practical to perform a consequence calculation for every source term. That.is, the consequence code, MAG G S ,20,21,22 required so much computer time to calculate the consequences of a source term that the source terms had to be combined into source term groups. Each source term group is_a collection of source terms that result in similar consequences. The frequency of the source term group is the sum of the frequencies of all the APBs which make up the group. The process of determining which APBs go to which source term group is denoted partitioning. It involves considering the potential of each source term group to cause early fatalities and latent cancer fatalities. Partitioning is a complex process; it is discussed in detail in Volume 1 of this report and in the User's Guide for the PARTITION Program.23 The transition matrix representation of tha source term calculation and-the Brouping process is fSTG - fAPB P(APB*STG) (Eq. 1.3) l where fAPB is the vector of frequencies for the accident progression bins l defined in Eq. 1.2, fSTG is the vector of frequencies for the source term groups, and P( AP B-+STG) is the matrix of - transition probabilities from accident progression bins to source term groups. Specifically, fSTO - (fSTG 1, ..., fSTGnstol e
- fSTGA - frequency (yr-1) for source term group 1, l
nSTG - number of source term groups, pSTG11 ... pSTG ,n37o 1 P(APB4STC) = , . pSTGnAra.1 ... pSTGnArs.nsto
~
l and 1.12
i zj i pSTG,j - probability that accident progression bin .k will be assigned to source term group 1.
}
U
'l= if accident' progression' bin k_is-; -
assigned to source term group-A- ]
,0 o therwise .-
The properties of fAPB are- given in conjunction with Eq. 1.2. Note that the source terms themselves do not~ appear in Eq. 1.4 The source terms are- j' used only to assign -an APB to a source term group. The consequences for each APB are computed from the average source term for the. group to' which i the APB has been_ assigned. V Consecuence Analysis. The consequence analysis -is performed for - each ) source term group by the MACCS program._ The results for each source term {
~
group include - estimates' f'or_ both- mean' consequences i and distributions of ~ consequences. When these . consequence _ results are combined with the .l frequencies for the source term groups, overall mea.sures of risk are i obtained. 'The consequence analysis . differs from the preceding _-three ; constituent analyses in that uncertainties - are not explicitlyz treated in-the consequence analysis.- That is , important values and parameters are l determined from distributions by --a; sampling process : in the accident. frequency analysis, the accident progression' analysis; and the source term analysis. This is not - the case for the _ consequences in the analyses; j performed for NUP.EC 1150. i In the transition matrix notation. -:the risk may be expressed by 1 (Sq. 1.4)
~
r0 - fSTG cSTG
=
where fSTG is the vector of frequencies f.or-the-source--term groups defined- ; in Eq. 1.3, r0 is the -vector. of risk . measures, and -cSTG Dis . the" matrix of
~
mean consequence - measures conditional . on the occurrence -' of indivitlual [ source term groups. Specifically, r0 - ( rC3 , . . . , -rCncl i rc, - risk -(consequence /yr) for consequence measure m, 4 nC - number of consequence measures, cSTG 31 ... cSTC 3,nc cSTGnsra,g ... cSTG nsto,nc i and - - 1.13'
l cSTGj,- mean value (over weather) of consequence measure m conditional on the occurrence of source term group A. The properties of fSTO are given in conjunction with Eq. 1.3. The elements cGTGy, of cSTO are ducermined from consequence calculatiotts with HACCS for individual source term groups. Computation of Risk. Equations 1.1 through 1.4 can be combined to obtain the following expression for risk: rc - fIE P(IE-*PDS) F(PDS-*APB) P(APB-*STG) cSTG. Eq.(1.5) This equation shows how each of the constituent analyses enters into the calculation of risk, starting from the frequencies of the initiating events and ending with the calculation of consequences. Evaluation of the expression in Eq.1.5 is performed with the PRAMIS2' and RISQUE codes. The description of the complete risk calculation so far has focused on the computation of mean risk (consequences / year) because doing so makes the overall structure of the NUREG 1150 PRAs more easy to comprehend. The mean risk results are derivad from the frequency of the initiating events, the conditional probabilities of the many ways that each accident may evolve and the probability of occurrence for each type of weather sequence at the time of an accident. The mean risk, then, is a summary risk measure. More information is conveyed when distributions for consequence values are displayed. The form typically used for this is the complementary cumulative distribution function (CCDF). CCDFs are detir.ca by pairs of f values (c f), where c is a consequene.e value and the f is the frequency I with which e is exceeded. Figure 1.3 is an example of a CCDP. The construction of CCDFs is described in Volume 1 of this report. Each mean risk result is the outcome from reducing a curve of the form shown in Figure 1.3 to a single value. While the mean risk results are often useful for summaries or high level comparisons, the CCDF is the more basic mea 3ure of risk because it displays the relationship between the size of the consequence and frequency exceedance. The nature of this relationship, i.e., that high consequence events are much less likely than low consequence events is lost when mean risk results alone are reported. This report utilizes both mean risk and CCDFs to report the risk results. Propagation of Uncertainty through the Analysis. The integrated NUREG-1150 t;. analyses use Monte Carlo procedures as a basis for both uncertainty and the sensitivity analysis. This approach utilizes a sequence: Xp X, 2
..., X ny (Eq. 1.6) of potentially important variables, where nV is the number of variables selected for consideration. Most of these variables were considered by a .
panel of experts representing the NRC and its contractors, the academic world, and the nuclear industry. For each variable treated in this manner, two to six experts considered all the information at their disposal and provided a distribution for the variable. Formal decision analysis 1.14
1 i 1.0E-3 , O o1.0E-4 . o 3 I u o . s
*o1.0E-5 .
o 1 e : u - 8,1.0E-6 . - v x i o : C ~
@1.0E-7 .
3 o- : 8 L. 81.0E-8 . C 3
-o ,
D o - 01.0E-9 x : ua . 1.0E-10 - ~' '- - 1.0E-3 1.0E '2 1.0E-1 1.0E0 1.0E1 1.0E2 1.0E3 - 1.0E4 1.0E5 Latent Cancer Fatalities Figure 1.3. F.xample Risk CCDP. 1.15
l l techniques 25 (also in Vol. 2 of this report) were used to obtain and record each expert's conclusions and to aggregate the assessments of the individual panel members into summary distribution for the variablo. Thus, a sequence of distributions D 3 , Dz, .... D,y, (Eq. 1.7) is obtained, where Di is the distribution assigned to variable X g. From these distributions, a stratified Monte Carlo technique, Latin Hypercube Sampling,26.27 is used to obtain the variable values that will actually be propagated through the integrated analysis. The result of generating a sample from the variables in Eq.1.6 with the distributions in Eq. 1.7 is a sequence S3- (Xii, X iae e X ,ov), 1 - 1, 2, i
. . . , nulS , (Eq. 1.8) of sample elements, where Xg is the value for variable X 3 in samplo element i and nulS is the number of elements in the sample. The expression in Eq. 1.5 is then determined for each element of the sample. This cred 7s a sequence of results of the form rC3= IIE Pg (IE*PDS) g P (PDS*APB) P (APB*STG) cSTG, i g (Eq. 1,9) ;
where the subscript i is used to denote the evaluation of the expression in l Eq. 1.5 with the ith sample element in Eq, 1.8. The uncertainty and sensitivity analyses in NUREG 1150 are based on the calculations summarized l in Eq. 1.9. Since P(IE*PDS), P ( PDS-*APB) and P(APB*STG) are based on results obtained with TEMAC, EVNTRE and GGSOR, determination of the expression in Eq. 1.9 requires a separate evaluation of-the cut sets,-the APET, and the source term model for each element or observation in the sample. The matrix cSTG in Eq. 1.9 is not subscripted because the NUREG-1150 analyses do-not include consequence modeling uncertainty other than the stochastic variability due to weather conditions. 1.5 Orvanization of this Renort l This report is published in seven volumes as described briefly in the Foreword. The first volunie of NUREG/CR-4551 describes the methods used in the accident progression analysis, the source term analysis, and the consequence analysis, in addition to presenting the methods used to assemble the results of these constituent analyses to determine risk and the uncertainty in risk. The second volume _ describes the results of convening expert panels to determine distributions for the variables thought to be the most important contributors to uncertainty in risk. Panels were formed to consider in-vessel processes, loads to the containment, containment structural response, molten core-containment interactions, and source term .ssues. In addition to documenting the results of these panels for about 30 important parameters, Volume 2 includes supporting material used by these panels and preser s the results of distributions that were determined by other means. 1.16
Volumes 3 through 6 present the results of the accident progression analysis, the source term analysis, and the consequence analysis, and tne combined risk results for Surry, Peach Bottom, Sequoyah, and Grand Gulf, respectively. These analyses were performed by SNL, Volume 7 presents analogous results for Zion. The Zion analyses were performed by BNL, 1 This volume of NUREG/CR 4551, Volume . 6, presents risk and constituent analysis results for Unit 1 of the Grand Gulf Nuclear Station, operated by , the Sys tem Energy Resources Inc., Part 1 of this volume presents the - analysis and the results in some detail; Part 2 consists of appendices, which contain further detail. Following a summary and an introduction, - Chapter 2 of this volume presents the results of the accident progression analysis for internal initiating events. Chapter 3 presents the result of the source term analysis, and Chapter 4 gives the result of the consequence analysis. Chapter 5 summarizes the risk results, including the contributors to uncertainty in risk for Grand Gulf, and Chapter 6 contains che insights and conclusions of the complete analysis.
~
t l l 1.17 l
. _ _ - - _ . .= . - . - . - - , . . - . . ~ - . . - -- - . - . - . . - -.- ,
.1 l
1.6 References
- 1. U.S. Nuclear Regulatory Commission, " Severe Accident Risks: An-Assessment for Five U.S. Nuclear Power . Plants ," NUREG 1150, June 1989.
- 2. U.S. Nuc1 car Regulatory Commission, " Reactor Safety Study -_ An Assessment of Accident Risks in U.S. Commercial -Nuclear Power Plants," WASH 1400 (NUREG 75/014), 1975.
- 3. J. S. Evans et'al., " Health Effects Models for Nuclear Power Plant-Accident Consequence Analysis," NUREG/CR-4214, SAND 85-7185, Sandia National Laboratories, August 1986.
- 4. J. S. Evans ee al,, " Health- Effects Models for Nuclear - Power ' Plant Accident Consequence Analysis," NUREG/CR 4214, Revision 1.--SAND 85--
- 7185, Sandia - National - Laboratories , and Harvard University, Cattbridge , MA, (Part I published January 1990; Part II published:May-1989),
- 5. S. Kaplan, " Matrix Theo y Formalism for Event Tree Analysis:
Application to - Nuclear-Risk Analysis," Risk Analysis . 2, pp. 9 18, 1982, [
- -6. b. C. Bley, S, Kaplan, and B. J . - Garrick, "Assemblin6 and Decomposing .
c PRA Results: A Matrix Formalism," in Proceedines of the Internationni Meetine on - Thermal ~ Nuclear Reactor- Safety, NUREG/CP-l 0027, Vol. 1, pp. 173-182, U. S. - Nuclear Regulatory - Commission, l Washington, D.C., 1982. l-7, R. B. Worrell, " SETS Reference Manual," NUREG/CR 4213, SAND 83 2675, Sandia National Laboratories, May-1985. 8, R. C. Bertucio and A. J. Julius, Analysis of Core Damage Frequency: Surry, Unit 1 Internal Events," NUREG/CR 4550, Vol. 3, SAND 86 2084, T Sandia National Laboratories,-April 1990.. *~ ~
- 9. Bertucio and S. -R. Brown, " Analysis of Core Damage Frequency:-
R. C. Sequoyah, Unit 1 Internal Events," NUREG/CR 4550, - Vol . 5, SAND 86-: 2084, Sandia National Laboratories, 1989.
- 10. A. M. Kolaczkowski et al . , " Analysis of C' ore - Damage Frequency': Peach-Bottom, Unit-2. Internal Events," NUREG/CR 4550, Vol;E4, SAND 86-2084, Sandia National Laboratories, April 1989,
- 11. M. T. Drouin et al. , " Analysis of Core Damage Frequency: Grand Gulf, Unit 1 Internal Events," .NUREG/CR-4550, Vol, 6,, SAND 86-2084, Sandia National Laboratories,. September 1989.
- 12. M. B. Sattison- and K. W. - Hall ,- " Analysis of Cots Damage - Frequency':
Zion, Unit 1 Internal Events," NUREG/CR-4550, Vol. - 7, EGG-2554, EG&G Idaho Inc., (Idaho Nationa1' Engineering Laboratory), May 1990. o 1,18 l t 4m,. .mm,,..,... _. ,, ,c. . _ . . . - . , .,_~,%I~,-m..,-.. ...,u . ,.!....,m.,. , . . .
t' < q
- 13. R. L. Iman, "A Matrix Based Approach to Uncertainty :and Sensitivity Analysis for Fault Trees ," Risk Analysi s . 7,'pp. 21 33, 1987.
14 R. L, Iman and M. J. Shortencarier, "A User's Guide for the Top Event Matrix Analysis Code (TEMAC) ," lNUREG/CR 4598, . SAND 86 0960, Sandia .; National-Laboratories,-April 1986. .j
- 15. J. M. Griesmeyer and- L. N, Smith,="A Reference. Manual for the Event l Progression Analysis Code (EVNTRE)," NUREG/CR 5174, SAND 88 1607, Sandia National Laboratories, September 1989 _.
- 16. R. S. Denning, J. .A. Gieseke,_P.-Cybulskis, K. W. Lee, _H. Jordan, L. A. Curtis, R, F.-Kelly, V. Kogan, and P. M. Schumacher, "Radionuclide Calculations = for $616cted Sever 6 Acuideut. - Scenerios ,*
NUREG/CR-4624, BMI-2139, Vols.1-5, Battelle Columbus -Division,1986. I
- 17. M. T. Leonard et al.. " Supplemental Radionuclide Release Calculations-for Selected Severe Accident S_cenarios , " NUREC/CR 5062, _ BMI 2160,_
Battelle Columbus Division, 1988. 9
- 18. S. E. Dingman et al . . "MELCOR Analyses for. Accident Progression .
Issues," NUREG/CR-5331, SAND 89-0072, Sandia NationalL Laboratories, December 1990.
- 19. Industry Degraded Core Rulemaking Program, "Medular Accident Analysis Program - (MAAP) User's ' Manual," IDCOR Technical Report on subtasks 16.2 and 16.3, Fausts 6 : Associates' Ine , for- the Atomic Industrial -
Forum. Bethesda, MD,' 1987.
- 20. D. I. Chanin, J. L . - S p rung , L. T. Ritchie , ' and H. -N Jow, "MELCOR-Accident Consequence Code System (MACCS): _ User's Guide," NUREG/CR-4691, SAND 86-1562 Vol. 1, Sandia- National Laboratories, February _
1990.-
- 21. H . - N . - J ow , J. L.-Sprung, J. .A. Ro11stin, L. T.LRitchie, and-D. I. Chanin, "MELCOR Accident-Consaquence Code-System-(MACCS)i hodel:
- Description," NUREG/CR-4691, SAND 86-1562, Vol . - 12, Sandia i National Laboratories, February 1990. y
- 22. J. A. Ro11stin, D. I . Chanin,- and H. -N. Jow "MELCOR'_ Accidentf Consequence Code System (MACCS): Programmer's Reference ' Manual,"
NUREG/CR-4691, SAND 86-1562, Vol. 3, Sandia' National Laboratories, February 1990.
- 23. R. L. Iman, J. C. Helton,_and J. D. Johnson, "PALTITION: A Program Defining the Source Term / Consequence Analysis Interface in the NUREG .
1150 Probabilistic Risk Assessments User's Guide ,-" -NUREG/CR 5253, SAND 88 2940, Sandia National Laboratories, May 1939. >
- 24. R. L. Iman, J . D. Johnson, and .J . ' O. Holton, "PRAMIS : Probabilistic Risk Assessment - Model Integration System User's Guide," NUREG/CR-5262, SAND 88-3093, Sandia National Laboratories, May 1990.
1.19
- 25. S. C. Hora and R. L. Iman, " Expert Opinion in Risk Analysis -
The NUREC 1150 Methodology," Nuclear Science and Fanineering. 10'2: pp. 323 331 (1989).
- 26. M. J . McKay, W. J . Conover, and R. J . Beckman, "A Comparison of Three l Methods for Selecting Values of Input Variables in the Analysis of Output from a Computer Code," Ieshnometries. 21, 239-245, 1979.
- 27. R. L. Iman and M. J. Shortencarier, "A FORTRAN 77' Program and User's l Guide for the Generation of Latin Hypercube and Random Sampics for l Use with Computer Models," NUREC/CR-3624, SAND 83-2365, Sandia l National Laboratories, March 1984.
l l l l l l l t 1.20
- 2. ANALYSIS OF THE ACCIDENT PROGRESSION This chapter describes the analysis of the progression of the accident, starting from significant core uncovery (i.e., 2 ft above the bottom of the active fuel (BAF] with imminent re-flooding of the core not expected) and continuing for about 24 h or until the bulk of the radioactive material that is going to be released has been released. As the last barrier to the release of the fission products to the environment, the response of the containment to the stresses placed upon it by the degradation of the core and failure of the reactor vessel is an important part of this analysis.
The main tool for performing the accident progression analysis is a large and complex event tree. The methods used in the accident progression analysis are presented in Volume 1 of this report. The accident progression analysis starts with information received from the accident frequency analysis: frequencies and definitions of the plant damage states (PDSs). The results of the accident progression analysis are passed to the source term analysis and the risk analysis, Section 2.1 reviews the plant features that are important to the accident s progression analyris and the containment response. Section 2.2 summarizes the results of the accident frequency analysis, defines the PDSs, and presents their frequencies. Section 2.3 contains a brief description of the accident progression event tree (APET). A detailed description of the APET is contained in Appendix A. Section 2.4 describes the way- in which the results of the evaluation of the APET are grouped together into bins. This grouping is necessary to reduce the information resulting - from the APET evaluation to a manageable amount while still preserving the information required by the source term analysis. Section 2.5 presents the results of the accident progression analysis for internal-initiators. 2.1 Plant Features Imoortant to the Accident Procression at Grand Gulf The entire Grand Gulf plant was briefly described in Section 1,2 of this volume. This section provides more detail on the features that are important to the progression of a core degradation accident and- the - response of the containment to the stresses placed upon it. These features are:
. The Containment Structure;
- The Drywell Structure and Suppression Pool; The Reactor Pedestal Cavity; The Hydrogen Ignition System (HIS);
- The Containment Heat Removal System; and The Automatic Depressurization System (ADS).
The Grand Gulf Containment Structure. Grand Gulf has a Mark III
. containment. The Grand Gulf containment is a reinforced concrete structure with a steel liner. An important feature of the the Mark III containment is its large free volwne (1,400,000 ft 3) which allows it to have a _ loc design pressure (15 psig). The assessed mean failure pressure of the containment is 55 psig. Because of its large volume, the Grand Gulf containment is not inerted. Thus, during accidents in which the HIS is not available, combustible hydrogen mixtures can be present in the containment.
2.1
! t
?
The Drywell Structure and Sunnression Pool. The Grand Gulf drywell houses the reactor pressure vessel (RPV) and is completely surrounded by the containment structure. The drywell structure is a reinforced concrete structure and has a design pressure of 30 psid. The free volume of the drywell is 270,000 ft 3. The assessed mean failure pressure of the drywell structure is 85 psid. The drywell volume communicates to the containment volume through the vapor suppression pool. Passive vents allow the passage of steam and air into the vetwell after first passing through the pool which provides the condensing action. The RPV safety / relief line (including those associated with the ADS) discharge through spargers into the suppression pool, which again provides condensation of any s team teleases. Thus, in vessel releases are first passed through the pool before being released to the wetwell air space. The steam is condensed in the pool and the noncondensibles (i.e., hydrogen) are passed to the wetwell air space. Similarly, releases accompanying vessel breach are directed to the suppression pool (assuming the drywell structure is intact) before being A released into the containment. This process reduces the pressure in the p containment; however, it also allows combustible mixtures of hydrogen and 8 air to accumulate in the . containment. The HIS is designed to burn this hydrogen at low concentrations so that the accompanying containment pressurization is negligibic. The Reactor Pedestal Cavity. The reactor pedestal cavity is located directly below the RPV. The upper section of the cavity is formed by the 5.75 ft thick pedestal wall and the lower section of the cavity is recessed into the drywell floor. The pedestal cavity is essentially a right cylinder with a diameter of 21.17 f t and a depth of approximately 28 f t. The upper section of the cavity contains the control rod drive (CRD) housings. The major pedestal penetrations are the CRD piping penstrations at the top of the pedestal and the CRD removal opening which is a 3 f t by 7 ft doorway and is located 9.5 f t above the cavity floor. The cavity can contain all of the core debris released at the time of vessel breach. Thus, direct attack of the drywell vall by core debris is not an issue at Grand Gulf as it is for the Mark I coatainments. When the drywell is completely flooded a water depth- of 22.8 ft can be established in the cavity. There are two pathways by which water in the dryvell can enter the reactor cavity. The first pathway is through the drywell floor drains. There are four 4 inch drains in the drywell floor that connect to the equipment drain sump in the pedestal. The second pathway is through a door (3 f t by 7 f t) in the pedestal located 3'-4" above the drywell floor. The potential for lar6e amounts of water to be in the cavity has two major implications. First, when core debris is released from the vessel at the time of vessel breach the potential ex:. s t for large fuel-coolant interactions (FCIs) to occur if the cavity is full of water. These FCIs can fail the drywell directly from quasi-static pressure loads or can fail the RPV pedestal, which can then lead to drywell failure (e.g., penetration failure). On the other hand, a large amount of water in the cavity can cool the core debris that is released from the reactor vessel and thus mitigate the releases associated with core-concrete interactions (CCIs). 2 2.2 I l
l i l liydroren Irnition S y s tgtt._GLS.). . The Crand Gulf containment has an llIS. igniters are located throughout the containment and drywell volumes. The funct' n of the llIS is to prevent the buildup of large quantities of hydrav n inside the conteinment during accident conditions. This is accomplished by igniting, via a spark, small amounta of hydrogen before it has had a chance to accumulaee. The llIS consists of 90 Generr.1 Motors AC Division glow plugs (Model 70). 45 powered by each ac power division. The llIS is manually actuated. The glow plugs would not perform their function without ac power. Thus, the hts will not be available either during a station blackout or if the operatcts fail to actuate the system if ac power is available. Anntainment llent Removal Systemg. Suppression pool cooling (SPC) and the
- containment spray system (CS) are two modes of the residual heat removal (RlIR) system. The RilR system is a two train system with motor operated valves and pumps. Both trains have two heat exchangers in series downstream of the pump. In either the SPC or the CS modes of operation, the PJtR system can remove heat from the suppression pool by passing water from the pool through heat exchangers (with service water on the shell 9ide). In the CS mode, water is sprayed into the containment. The SPC system is manually initiated and controlled. The CS system, on the other haiM. is automatically initiated and controlled. Both the SPC and the CS modes of Ri!R require ac power and are, therefore, unavailable <during a station blackout.
The Automatic Deoressurization System. The Automatic Depressurization System (ADS) is designed to depressurize the reactor vessel to a pressure ' St which the low pressure injection systems can inj ect coolant to the reactor vessel. The ADS consists of eight relief valves capable of being manually opened. For the system to be automa tically - initiated a low pressure injection pump must be running. Thus, the ADS will not b. automatically initiated during a station blackout. The operator can also manually initiate the ADS, or he may depressurize the reactor vessel using the 12 safety relief valves (SRVs) that are not connected to the ADS logic. Each valve discharges into the suppression pool. The ADS valves are located in the drywell and pressures of approximately 100 psi vill prevent opening the ADS valves. The' assessed containment failure pressure at the 99th percentile h only 97 psig and, thus, failure of the ADS because of high pressure is not considered in this study. The ADS does, however, require de power. Therefore, the RPV can not be depressurized in sequences that involve failure of de power, 2.2 Interface with the_ Core namace Frecuency Analysi,g 2.2.1 Definitic,.p of PDSs Information evout the maay different accidents that lead to core damage is passed from the core damage frequency analysis to the accident progression analysis by means of PDSs. Because ' most of the accident sequences identified in the core damage frequency analysis will have - accident progressions similar to other sequences, these sequences have been grouped together into PDSs. All the sequences in one PDS sh"11d behave similarly
- in the period af ter core damage has begun. - For Grand Gulf, the PDS is 2.3
1 denoted by a 12 Ictter indicator that defines six charact ,Aca that lar6cly determine the initial and boundary conditions of ' m eldent prot,res sion. More information about the accident sequences may be found in NURtc/CP. 4550, Volume 6.1 The methods used in the accident frequency analysis are ptesented in NUREG/CR 4550. Volume 1.2 Table 2.21 lists the six characteristics used to define the PDSs. Under each characteristic are given the possible values for that characteristic. For example, the first characteristic denotes the initiating event and the status of ac and de power at the time core damage begins (assumed to be when the water level is 2 f t above the BAF) . Table 2.2 1 shows that there are four possibilities for this characteristic: B1 for a station blackout with of fsite power not recoverable because there is no emergency de power; B2 for a station blackout with offsite power recoverable; T2 for locs of power conversion system - (PCS) transient; and TC for an anticipated transient without scram (ATWS). The first characteristic denotes the initiating event and the status of ac and de power. The station blackouts are separated based on the ,. availability of de power. The loss of PCS transient and the ATWS event have both onsite and offsite power. The second characteristic denotes the reactor vessel pressure at the time of core damage The reactor pressure can be either high or low. llish pressure is detined as system pressure (approximately 1040 psig) . Low-pressure is defined as being less than 200 psia. The third characteristic denotes the type of coolant inj ection that is , avaliable or recoverable. This characteristic indicates if the coolant injection system is a high pressure system or a low pressure system. The availability of the firewater system and the - condensate system are also indicated because these systems require operator actions to align these systems. The fourth characteristic denotes the availability of the containment spray (CS) mode of RHR. In this analysis, the RilR heat exchangers are always available when the containment sprays are available. Therefore, there are no scenarios that involve spraying hot water because the heat exchangers are not available.. The fifth characteristic denotes the availability of the containment venting system, the standby gas treatment system, containment isolation system and the hydrogen ignition system. All of these systems require ac power and, therefore, their availability is directly related to the availability of ac power. The sixth characteristic denotes the time of core damage. Two times are considered in this analysis: core damage occurs in the short term (atal h), l and core damage occurs in the long term (at h 12 h). When the core damage l oce"n in the short term the accident is referred to as a short term or l fu accident (e.g. , short term station blackout or fast station blackout). Similarly, when core damage occurs in the long term the accident is referred to as a long term or slow accident (e.g. , slow TC), ! 2.4
v Tabic 2.2 1 Grand Gulf PDS Characteristics
- 1. What is the initiating event and what is the status of ac and de ,
power? ! B1 - Station blackout transient has occurred. Offsite power is not recoverable because there is no emergency de power. B2 - Station blackout transient has occurred. Offsite power is recoverable. l T2 - Loss of PCS transient has occurred. Offsite or onsite power is available. TC - ATWS has occurred. Offsite or onsite power is available.
- 2. Is the reactor vessel at hi 6h or low pressure?
P1 - The reactor vessel is at high pressure at the onset of core damage and depressurization is now pessibic. P2 - The reactor vessel is at high pressure at the onset of core damage because the nyerator failed to deprorsvrine; depressurize: tion is possible. P3 - The reactor vessel could be at high pressure at the onset of core damage. The operator depressurizing the vessel (which is possible) was not included in tire model. P4 - The reactor vessel is at low pressu.e.
- 3. What type of coolant injection is available or recoverable?
II - Injection to the reactor vessel is not available af ter the c onset of core damage. 12 - Injection with the firewater system is available before and after the onset of core damage. 13 - Injection with the condensate system is recoverable with the restoration of offsite power. 14 - Injection with the low pressure systems (LPCS and. coolant i inj ection) is recoverable with the restoration of offsite power. 15 - Inj ection with both the high and low pressure systems is recoverabic with the restoration of offsite power. 2.5 M A er'--
l I l l Tabic 2.2 1 (Continued) 16 - Injection with the high pressure systems (reactor core isolation cooling [RCIC) and CRD) and the low pressure systems (LPCs ard coolant injection) is recoverable with the restoration of offsite power.
- 4. Is Containment Spray (CS) mode of idlR available or recoverable?
111 - CS is not available at the onset of core damage, neither is it recoverable. 1i2 - At least one train of CS is recoverable with the restoration of offsite pcwer. 113 - At least on train of CS is available at the onset of core darage. i 5. Are the following systems available: venting, SBGT, CI, and }{211 i M1 - Miscellaneous systems (venting, SBGT, CI, and 111) 2 are not available at the onset of core damage. I X2 - h' .s c ell aneous systems (venting, SBGT, CI, and 11 1 2) are recoverable with the restoration of offsite power. l M3 - Miscellaneous systems (venting, SBGT, C1, and 11 1 2 ) are available at the onset of core damage.
- 6. When does core damage occur?
ST - Core damage occurs in the short term (at a 1 h). LT - Core damage occurs in the long term ( at a 12 h). - I I ) i 4 2.6
,aw i.,,---y w,-4-y,,y _+-s-,-y.aw- , - -.: - ,gy s,m ,-w, _ -w--r
.. -_ .-. - - . - . - - _ - _ =
- . - - - ~ ._ . -
l l l l l 2.2.2 PDS Freauene.ics ! In 'his subsection the 12 PDSs are described and their core darnage frequencies are presented. The accident frequency analysis for internal i initiators was performed with inore observations per satople than were the j accident progression analysis and the subsequent analyses. Since the sainplex were different in the random seed as well as the number of observations, the core damage frequencies differ slightly as is to be expected. The PDSs used in the Grand Gulf accident progression, source te rrn , and risk integration analyses are presented in Table 2.2 2. The mean core damage frequencies presented in this table are based on a sarnple size of 250. The core damage frequency distributions for the 12 PDSs, based on a sample size of 250, are presented in Figure 2.2 1. The accident frequency analysis - reports the PDS frequencies based on a sample siz> of 1000 (see Section 5 of NUREG/CR-4550, Vol. 6 Part 1) .1 When cot.sidered as a separate entity, a great inany variables could be sainpled in the accident frequency analysis, and a sarnple size of 1000 was used. A sample this large was not feasible for the integrated risk analysis. Based on the results from the 1000 observation sainple, those variables which were not important to the uncertainty in the core damage frequency were eliminated from the sampling, and the cut sets were re-cvaluated using 250 observations for the integrated risk analysis. As some variation from sample to sample is observed, even when the satuple size and the variables sampled remain the same, there are variations between the 1000 observation sample utilized for the stand-alone accident frequency analys!- and the 250 observation sample used for the integrated risk l analysis. These differences are suminarized in Ohle 2.2 3. For each PDS group, the first . line of Table 2.2-3 contains the 5th per-centile, rnedian, mean, and 95th percentile core darnage frequencies for the 1000 observation sample used in the stand alone accident frequency analysis. These values are taken from Table 5.3 1 of NUREG/CR 4550, Volume 6, Part 1. Samples containing 250 observations are used for the integrated risk analysis at Grand Gulf. The 5th percentile, median, mean, and -95 th percentile core damage fre'quencies for this sample are shown on the second line of Table 2.2 3 for each PDS. The differences between distributions for_ core damage frequency for the two sampics are within the statistical variation to be expected. Note that the fractional coatributions of each PDS to the THCD in Table 2.2-2 are I slightly different from those in Table 2.2 3. This is due to the fact that the PDS fractional contributions in Table 2.2-2 are based on the sample of-250 observations, and tha contributions in Table 2.2 3 are based on the - sarnple of 1000 observations. The remaining portion of this subsection describes the essential characteristics of each of the 12 PDSs. The descriptions of the PDSs were extracted from Chapter 5 of NUREG/CR 4550, Volume 6, Part 1.1 2.7
l l l t Table 2.2 2 PDS Core Darnage Frequencies for Grand Gulf t Mean PDS CD Frequency PDS t PDS Name (1/vri IMCD Frea. PDS Der;criotor Number 1 Fast Blackout 3.2E 06 79.2 B2 P3-IS-il2 M2-ST 2 Fast Blackout 4.6E 08 1.1 B2 P315 lll M2 ST 3 Fast Blackout 1.5E 07 3.7 B2 P3 I3 ill M2 ST 4 Slow Blackout 3.7E 08 0.9 B2 P4 IS Il2 M2 LT
$ Slow Blackout 2.3E 09 1 B2 P4 15 ill M2 LT
- 6 Slow Blackout 1.4E 09 1 B2 P412 Ill M2 LT 7 Fast Blackout 4.2E 07 10.3 B1 P1 11 ill M1 ST 8 Slow Blackout 6.3E 08 1.5 B1 P1 11 til M1 LT 9 Fast ATWS 5.0E-08 1.2 TC P2 16 il3 M3 ST 10 Slow ATWS 6.2E 08 1.5 TC P2 14 il3 M3.LT 11 Fast T2 1.8E 08 0.4 T2 P2 I5 il3 M3 LT 12 Slow T2 2.9E 10 ul T2 P2 I5 il3 M3 LT i
l l l i i i t i 2.8
. _ . . . .. . . - - - - . . _ ._. . . . _ . _ ~ . . . . . . . . - - -
i I 1 E-4 Ctand Gulf W mean m . m.di.n th . percentue u 1.E-t, esa.
*~ .
o.a. m f- 1 E-7. y , 1 96th.) t T. sa. J
$1E-8, ~
] ,
,,u,1 L
{ ***" _ esp. .
"~'
C 1E-D, "'
]
. ] 3 IE-to
..] ..,.
'"+
M
~ i It-11 , '.I I PDS1 PD5 2 PDS 3 PDS 4 PDS6 PDS6 l
l Figure 2.2-1. Core Darnage Frequency Distributions, t 1E-4. Grand Gulf W . mean m
- median esth.
, th a percentile M
oses. . . . I IE-6, %
. sa. esu o.a. _. -
l S N 5% I,iE l ,$% v u.. L 7z 1.E-8 sih . E-. s , %
?
c1E-9, ~ ' N - ' ' ' ' ' sth s% - r -. 1E-10 - 5% 1E-It i ! PD3 7 PDSB PDSD PDS 10 PDS 11 PDS 12 Total Figure 2.2-1 (continued). 2.9
Tabic 2.2 3 Plant Damage State Comparison for Grand Gulf Ills Sample Core Damare Frecuency (1/vri % TMCD PDS gj d 54 jiedian Mean 951 Fre a . (2) PDS1 1000 2.5E 08 5.0E 07 3.2E 06 9.6E 06 79 Fast SB0 250 2.6E 08 5.1E 07 3.2E 06 1.1E 05 PDS2 1000 6.9E 11 2.3E 09 4.8E 08 9.4E 08 1 Fast SB0 250 6.4E 11 2.1E 09 4,6E 08 1.9E 07 PDS3 1000 1.5E 09 3.1E 08 1.8E 07 5.8E 07 4 Past SB0 250 1.3E 09 3.4E 08 1.5E 07 6.7E 07 PDS4 1000 6.4E 11 2.7E 09 3.9,t 08 1.0E 07 1 Slow SB0 250 5.3E 11 2.3E 09 3.7L 08' 1.6E 07 PDS5 1000 5.5E 13 3.6E 11 1.3E 09 2.7E 09 <<1 Slow SB0 250 7.4E 13 3.2E 11 2.3E 09 3.0E 09 PDS6 1000 2.4E 12 1.4E 10 2.0E 09 5.8E 09 <<1 Slow SBo 250 1.4E 12 1.3E-10 1.4E 09 7.2E 09 PDS7 1000 2.9E 08 2.3E 07 4.3E 07 1.4E 06 11 Fast SB0 250 2.8E 08 2.4E 07 4.2E 07 1.6E 06 PDSB 1000 3.0E 10 9.2E 09 6.6E 08 2.0E 07 2-Slow SB0 250 2.6E 10 8.4E 09 6.3E 08 0.7E 07 PDS9 1000 3.9E 10 8.9E 09 5.0E 08 2.3E 07 1 Fast ATWS 250 3.2E l'0 7.9E-09 5.0E 08 1.9E 07 - PDS10 1000 4.9E 10 1.0E 08 6.3E 08 2.8E 07 2 Slow ATWS 250 3.9E 10 8.9E 09 6.2E 08 2.3E-07 PDS11 1000 2.5E 11 1.3E 09 1.2E 08 4.4E 08 <1 Fast T2 250 3.1E 11 1.2E 09 1.8E 08 5.3E 08 PDS12 1000 1.6E 13 .9.9E 12 2.7E 10 8.3E 10 <<1 Slow T2 250 4.9E 12 6.8E 11 2.9E 10 1.2E 09 To**1 1000 1.7E 07 1.2E 06. 4.0E 06 1.2E-05 - 250 1.8E-07 1.1E 06 4.1E 06 1.4E 05 Notes: tui'ho Accident Frequency Analysis used a uts sample size of 1000 The Accident Progression Analysis used a 111S sample size of 250 (2) Percentages based on the UlS sample size of 1000 2.10
[DS 1 (B2-P3 15 H2 M2 ST). This PDS involves station blackout scenarios where loss of offsite power (LOSP) is recoverabic (B2). Coolant injection is lost early such that core damage occurs in the short term (ST) and with the vessel at high pressure (P3) because depressurization did not have an effect in the prevention of core damage. If offsite power is restored then the following functions are available; either high pressure injection or low pressure injection or both (15), heat removal via the sprays (H2), and the toisc ellane ou s systems venting, standby gas treatment (SBCT), containment isolation (CI), hydrogen ignition (H21) (MP). This PDS also includes cut sets with either one or two stuck open relief valves (SORVs). With the restoration of offsite power, the following coolant injection syst. ems are recoverable: HPCS, condensato, Low Pressure 1 Coolant Injection (LPCI) and Low Pressure Core Sprcy (LPCS). In some casen, HPCS and LPCS are recoverabic, but only foi around 12 h; they are then lost on room heatup. The firewater system is available in overy cut set. For those cut sets with two SORVs, the RCIC system is available but is not su#ficient to prevent core da:nage. PDS-2 (B2 P3 15 H1 M2 ST). This PDS involves station blackout scenarios where LDSP is recoverable (B2) . Coolant injection is lost early so that core damage occurs in the short term (ST) and with the vessel at high pressure (P3) because depressurization did not - have an effect in the prevention of core damage. If offsite power is restored then the following functions are available: either high pressure injection or low pressure injection or both (15), and the miscellaneous systems -venting, SBCT, CI,- H21 (M2). Heat removal via the sprays is not available with the recovery of offsite power (H1). This PDS also includes cut sets with either one or two SORVs. With the restoration of offsite power, the following coolant injection systems are recoverable: HPCS and condensate. In some cases, LPCS is recoverable, but only for approximately 12 h, at which time they fail as a result of room heatup. The Firecater system is available in every cut set. For those cut sets with two SORVs, the RCIC system is available but is not sufficient to prevent core damage. PDS-3 (B2 P3 13-M1 M2 ST). This PDS involves station blackout scenarios where_1DSP is recoverable (32). Coolant injection is lost' early so that core damage occurs in the short term (ST) and with the vessel at high pressure (P3) because depressurization did not have an effect in the prevention of core damags ' If offsite power is restored then the following functions are available: low pressure injection only with condensate (13) and the miscellaneous systems venting, SBGT, CI, H21 (M2). Heat removal via the sprays is not available with the restoration of offsite power (H1). This PDS also includes cut sets with either one or two SORVs. With the restoration of offsite power, the following coolant injection system is recoverable: condensate. HPCS and LP( .1 are available with the recovery of offsite power, but only for approximately 12 h, at which time they fail as a result of room heatup. The Firewater system is available in every cut set. For those cut sets with two SORVs, the RCIC system is available but is not sufficient to prevent core damage. I 2.11 l
I l t l i l PDS 4 (B2.P4-15-il2 M2-LT). This PDS involves station blackout scenarios where LCSP is recoverable (B2). Coolant injection is lost late so that core darnage occurs in the long term (LT) and with the vessel at low pressure (P4). If offsite power is restored then the following functions are available: either high pressure injection or low pressure injection or both (I$), heat removal via the sprays (112), and the miscellaneous systems.
-venting, SBGT, CI,112I (M2),
k'i th the restoration of offsite power, the following coolant inj ection systerns are recoverable: HPCS, condensate, LPCI and LPCS. In some cases, itPCS and LPCS are recoverable, but only for approxitnately 12 h, at which tirne they fail as a result of room heatup. The Firewater syt m is available in every cut set. PDS 5 (B2 P4-IS l?1 M2-LT). This PDS involves otation blachout scenarios in which MSP is recoverable (B2). Coolant injection is lost late so that core damage occurs in the long term (LT) and with the vessel at low pressure (P4). If offsite power is restored, then the functions of high pressure injection or low pressure injection or both (15) are available, as well as the miscellaneous systems of venting, SBGT, CI, and 1121 (M2). Feat removal via the sprays is not available with the restoration of offsite , power (111) . I There are some cut sets in which heat removal sprays are available with offsite power restoration, but these have negligible contribution and wete not removed. PDS 6 (B2-P4-12 ill M2-LT). This PDS involves station blackout scenarios where MSP is recoverable (B2). Coolant inj ection is lost late so that core damage occurs in the long term (LT) and . with the vessel at low pressure (P4). Firewater is recoverable (12). If offsite power is restored, then the following functions are available: the misec11aneous , systems venting, SBGT, CI, !!21 (M2). Heat removal via the sprays is not I available with the restoration of offsite power (H1). HPCS is available with the restoration of offsite power, but only for around 12 h; it is then lost on room heatup. PDS 7 (B1 P1 11 H1 M1 ST). This PDS involves station blackout (without any de power) scenarios where LOSP is not recoverable (B1). Coolant injection is lost early so that core damage occurs in the short term (ST) and with I the vessel at high pressure- (P1) and depressurization is not possible. Since offsite power is not recoverable, the functions of injection (11), heat removal (H1), and those of the sniscellaneous systems (M1), are not available. PDS 8 (B1-P1-11-H1-M1 LT1 This PDS involves station blackout (without any de power) scenarios where LOSP is not recoverable (B1). Coolant injection is lost late so that cotu damage occurs in the long term (LT) and with the vessel at- high pressure (P1), and depressurization is not possible. .Since offsite power is not recoverable, functions (i.e., inj ection (11), heat removal [111] and the miscellaneous Systems (Ml}) are not available. 2.12
E PDS 4 (TC P2 16 H3 M3 ST). This PDS involves ATWS transient scenarios (TC). Coolant injection is lost early so that core damage occurs in the short term (ST) and with the vessel at high pressure because the operator failed to depressurize (P2), liigh pressure injection with RCIC is available (16). llent removal via the sprays is available ( 11 3 ) and the misec11ancous systems (i .e. , venting, SBGT, CI and 1121) are available (M3) . PDS-10 (TC P2-14 il3-M3 LT). This PDS invalves ATWS transient scenarios ("'C ) . Coolant injection is lost late such that core damage occurs in the
;ang term (LT) and with the vessel at high pressure because the operator failed to depressurize (P2). Low pressure injection is recoverable with reactor depressurization (14), lleat removal via the sprays is available
( 11 3 ) and the miscellaneous systems (i.e., venting, SBGT, C1 and 112I) are available (M3). PDS-11 (T2 P2-15 il3 M3 ST). This PDS involves transient scenarios where the PCS is lost (T2). Coolant injection is lost early so that core damage occurs in the short term (ST) and with the vessel at high pressure because the operator failed to depressurize (P2). Both high pressure and low pressure are recoverable (15) since the failures involved operator failures. lle a t removal via the sprays is aval'able ( 11 3 ) and the misec11aneous systems (i.e. , venting, SBGT, CI and 1121 are available (M3). TDS -12 ,(T2 P2 -I S-H3 -M3 LT) . This PDS involves transient scenarios where the PCS is lost (T2). Coolant injection is lost late so that core damage occurt. in the long term (ST) and with the vessel at high pressure because the operator failed to depressurize (P2). Both high and low pressure are recoverable (15) since the failures involved operator failures, llent removal via the sprays is availabic ( 11 3 ) and the miscellaneous systems (i.e. , venting, EMT, C1 and 1121) are available (M3). 2.2.3 ilich Level Groupinc of PDSs To provide simpler, more easily understood summarios for NUREG 1150, the 12 PDSs described above were further condensed into the following four groups:
- 1. Short Term Station Blackout
- 2. Long Term Station Blackout
- 3. ATWS
- 4. T2 Transients These four groups are denoted summary PDS groups or collapsed PDS groups.
The mapping from the 12 groups described in the previous section into the four summary groups used in the presentation of many of the results is given in Table 2.2-4. In combining two groups to form one summary group, frequency weighting by observation is employed. The percentages of the total mean core damage frequency given above provide only approximate weightings. The core damage frequency distributions for the four summary PDS groups are presented in Figure 2.2 2. 2.13
._ .. _ - . . _ _ _ _ ~ ._. _ _ _ _ .. . _ _ _ _ _ -
Table 2.2-4 Relationship between PDSs and Summary Groups Sumtria ry Group -- % TMCDP PDS Groups t TMCDF 1
- 1. Fast SB0 95 1. Fast Blackout 79
- 2. Fast Blackout 1
- 3. Fast Blackout 4
- 7. Fast Blackout 11
- 2. Slow SB0 3 4. Slow Blackout 1
- 5. Slow Blackout <<1
- 6. Slow Blackout <<1
- 8. Slow Blackout 2
- 3. ATWS 's 9. Fast ATWS 1
- 10. Slow ATWS 2
- 4. Transients <1 11. Fast Transient <1
- 12. Slow Transient <<1 1.E-4 Grand Gulf u u..
1.E-6, n. I,$ia. N l . ..i,._ u -.
-, 6th J K1.0-7 3 ,, u J ,,,,,
^
*~ '
k ,
- n. ~. --
. , 7_ ,
E 5 - 0 .E-9 1 3 sta i% 1.E-10 g W = mean gg, m = median th . percentile 1.E- 1 L PDS Group Fust Slow ATWS T2 Tol61 FB0 SD0 Figure 2.2-2. Gore Damage Frequency Distributions for the Summary PDS Groups. Grand Gulf: Internal Initiators. l 1 2.14
- - , . - - - .-_,,-------.,.,,,,,,,.w..,,.. . , , , . . , - - , , ,s , ,.,nx, .- + , .-
-. -. - -- . = . - . - _
2.2.4 varinbles Sanmled in the Accident Frecuenev Analysis In the stand alone accidat forcrey relyris Mr inurrf1 meter, n larga number of varables were sampled. (A list of these variables may be found in NUREG/CR 4550, Vol. 6, Part 1.)1 Only those variables found to be important to the uncertainty in the accident frequencies were selected for sampling in the integrated risk analysis. These variables are listed and defined in Table 2.2 5. The first column in Table 2.2 5 contains the variable name which is an eight character identifier. Where these differ from the identifiers used in the fault trees, these identifiers are listed in the description in brackets. Generally, the eight character identifiers have been selected to be as informative as posaible to those not familiar with the conventions used in systems analysis. The second column in Tc'.,le 2.2 5 gives the range of the distribution for the variable and the third column indicates the type of distribution used and its mean value. The fourth and fifth columns in Table 2.2-5 show whether the variable is correlated with any other variable and the last column describes the variable. More complete descriptions and discussion of these variables may be found in the Grand Gulf accident frequency analysis report (NUREG/CR 4550, Vol. 6 ) ,2 This report also gives the source or the derivation of the distributions for all these variables. 2.3 Description of the APET This section describes the APET that is used to perform the accident progression analysis for Grand Gulf. The APET :itself forms a high 1cyc1 model of the accident progression. The APET is too large to be drawn out in a figure as smaller event trees usually are. Instead, the APET exists only as a computer input file. The APET is evaluated by the code EVNTRE, which is described elsewhere.8 The APET is not meant to be a substitute for detailed, mechanistic codes such as the STCP, CONTAIN, MELCOR, and MAAP. Rather, it is an integrating framework for synthesizing the results of these codes together with expert judgment on the strengths and weaknesses of _ the codes. The detailed, mechanistic codes require too much computer time to be run for all the possible accident progression paths. Therefore, the results from these codes are represented in the Grand Gulf APET, which can be evaluated l relatively quickly, In this way, the full diversity of possible accident progressions can be considered and the uncertainty in the many phenomena involved can be included. The following section contains a brief overview of the Grand Gulf APET. l Datails, including a complete listing of the APET and a discussion of each ! question, may be found in. Appendix A of this volume. Section 2.3.2 is a summary of how the APET was quantified, that is, how the many numerical values for branching ratios and parameters were derived. Section 2.3.3 presents the variables that were sampled in the accident progression analysis for Grand Gulf. I 2.15 l
1 I l l, l' , Taole 2 2-5 i Variables Sampled in the Accident Frequency Analysis for Internal Initiators f Correlation Correlated With Description Variable Rance, Distribution 1.5E-05 Iognormal None Probability of failure to open MOV-FOP 0.085 Mean-0.003 (per demand) for motor operated valves (generic). , i 4.0E-6 Ingnormal None Probability that the motor operated 4 MOV-MAIN Mean-7.8E-4 valve is out for maintenance (per 0.023 demand) (Seneric)..[MOV-MA] 1.5E-5 legnormal None Probability of failure to start MDP-FSTR 0.085 Mean-0.003 (per demand) for motor-driven pumps (generic). [MDP-FS} i F Probability of failure to run (per MDP-FRUN 3.6E-6 Lognormal None j % demand) for.motoc-driven pumps ! 0.020 Hean-7.2E-4 1 (generic). {MDP-FR] j M~)P-MAIN 9.9E-6 Legnormal None Probability that the motor-driven l 0.057 Mean-l.9E + pump is out for maintenance (per demand) (generic). [MDP-MA] ,
~
1.5E-4 lognormal None Probability of failure to start ; TOP-FSTR (per demand) for turbine-driven
~
0.85 Mean-0.029 pump (RCIC) (generic). {TDP-FS} 0.012 Max Entropy None Probability for failure to run (per T0P-FRUN Mean -0.12 demand) for turbine-driven pump , 1.0 ' (RCIC) (generic). {TDP-FR] t Lognormal None Probability for failure to run (per DDP-FRUN- 9.4E-5 i ' Menn-0.019 demand) .for diesel-driven pumo 0.54 (FUS) (generic). [DDP-FR} , e
Table 2.2-5 (continued) Correlation Correlated V! h Description 1 yariable Ranre Distribution , DGN-FSTR' O.003 Iognormal None Probability that a diesel f.enerator 0.19 Mean-0.030' fails to start (per demand) j (generic). [DGN-FS] DGN-FRUN 7.9E-5 legnormal None Probability that a diesel generator l fails to run-(per demand) 0.45 Mean-0.016 (generic). [DGN-FR]
'DGN-MAIN 3.0E-5 Lognormni None Probability that a' diesel generator 0.17 Mean-0.006 is out for maintenance (per demand)
(generic). [DGN-MA] F BAT-FDP 1.0E-4 Im normal None Probability that a battery fails to C 0.006 Mean-0.001 deliver power (per demand) (generic). [ BAT-LP) SSW-MAIN 2.4E-6 legnormal None Probability to fail to restore the ' 3 0.079 Mean-0.0017 SSW train after maintenance (HRA). ' [SSW-XHE-RE-TAB 2,4]
'- k',IC- DEP 0.0041 Max Entropy None Probability to fail to depressurize 0.41 Mean .041 the RPV via the RCIC steam line after 12 h (HRA). [RA-RCICDEP-12HR]
R -DPRES 3.3E-7 Iognormal None Probability of common cause 0.0019' Mean-6.9E-5 miscalibration of drywell pressure sensors (HRA). [CCF-MC] . BETA-2DG O.0039 Lognormal. , None. Beta factor for common cause ; 0.24 'Mean-3.8E-2 , failure of two diesel generators i (generic). BETA-BAT 4.1E-4 Lognormal None Beta factor for common cause 0 025 Mean-0.004 failure of three batteries (generic). [ BETA-3 BAT] 1 4
?
Table 2.2-5 (continued) Correlation Correlated With Description r
' Variable Rance Distribution' I
BETA-SSW 0.0014 Lognormal None Beta factor for common cause ; 0.088 Mean-0.014 failure of three service water .; system motor driven pumps (generic). [ BETA-3SSU] , . i Initiating event: frequency (1/yr) i IE-T2 0.16 Lognormal None
'Mean-1.6 of a transient with loss of PCS.
10.0 , 1
.Lognormal None Probability to fail to manually RA-INJ-1 1.7E-6 Mean-3.4E-4 actuate injection within one hour ;
0.0099 after an auto-actuation failure ! (HRA). i
.Legnormal None -Probability to fail to manually ,
1
% FWSACT12 1.5E-4 0.85 Mean-0.029 align and actuate the FWS after 12-i-
hours (HRA). .[RA-FWSACT-12HR} l a None Probability to fail to recover PCS RA-PCS-1 .0.01 Max Entropy
,,within one hour (generic).
1.0 Mean-0.1 i Iognormal None Initiating event: frequency (1/yr) IE-TC 0.72 of a transient (combination of TI, 45 Mean-7.2-T2, and T3s). f Probability of mechanical failure F-RPS S'.0E Iognormal. . None ' Mean-9.9E-6 of the reactor protection system. l 2.8E-3 i - [CM} 4 i F-ADS 0.0125 Max Entropy None Probability that the operator fails ;
- f. to depressurire the RPV during an j 1.0 Mean-0.125 ",
ATVS. [ ADS-XHE] ; t i 1 i
+
4 1 i 1
- i
, Table 2.2-5 (continued) . Rance. ' Distribution Correlation Correlated With Description ! ! Variable IE-LOSP 6.3E-5 LOSP Data None Initiating event: frequency (1/yr) [ 0.58 Mean- 0.1 of LOSP. [IE-Tl] ! AC-ST-M 0.086 wSP Data Rank 1 AC-LT-n Probability of failure to restore 0.35 Mean-0.19 AC-ST-n AC power within 1 h. [RA-IDSP-lHR 1 ; i I AC-LT-FE 3.5E-4 LOSP Data Rank 1 AC-LT-n Probability of failure to restorr 0.10 Mean-0.015 AC-ST-n AC power within 12 h. [RA-IDSP- . 12HR] l 4 h 1 i j M a t* e-I 1 .. ! .l 1 a i 2 l ! t i, 4 . !p- , I
=f i
i i E i i
-.-- . - ~ . _ - - - - . - . .- -. _ - - - - - - . - - - . -
( 2.3.1 Overview of the APET The APET for Grand Gulf considers the progression of the accident from the time core damage is imminent (i.e., 2 ft above the BAF) through the CCI. Although the CCI may progress at ever slower rates for days, the end of this analysis has been arbitrarily set at 24 h. Except in very unusual accidents, almost all of the fission products that are going to be released from the containment will have been released by 24 h after the initiator. This event tree is based on the Grand Gulf containment arrangeacnt, systems, and procedures. In addition, emphasis was placed on modeling the accident progression for the dominant PDSs presented in the accident frequency analysis described in NUREG/CR 4550, Vol. 6, Part 1.1 Table 2.3.1 lists the 125 questions in the Grand Culf ' APET. In this APET four tino periods are considered. To facilitate understanding of the APET and referencing between questions, each branch or every question is assigned a mnemonic abbreviation. The mnemonic branch abbreviations for teost branches start with a character or characters which indicate the time period of the question. The time periods and their abbreviations are: El Initial Questions 1 through 22 determine the conditions at the beginning of the accident (i.e., before core damage). E2 Early Questions 23 through 57 address the progression of the ~ accident from the beginning of core damage to just before vessel breach. Questions in this time period consider the status of various systems (coolant inj ection , ac power, i IIIS , etc.), the molecular composition of the containment and drywell atmosphere, hydrogen burn phenomena (e.g., ignition and loads), and the containment and drywell structural response to containment loads. Questions 53 through 57 establish the conditions in the containment and drywell just before vessel breach. These - questions determine the amount of water in the reactor cavity, the containment pressure, and whether the drywell L atmosphere,is combustible. I Intermediate Questions 58 through 98 determine the progression of the accident from immediately before vessel breach to the time
- of significant CCI, The potential for core damage arrest j -(i.e., no vessel breach) is addressed in this time period.
- The inaj o rity of these questions address the loads accompanying vessel breach and the containment and drywell structural response to these loads, llydrogen combustion is considered both at the time of vessel breach and during the time period before significant CCI begins. liydrogen phenomena associated with the hydrogen produced during CCI is addressed in the next time period.
2.20
- , - _ . -.. = , . . . .
. . - - . . .-. - . .. . _ ~ . _ - . . - . _ - . _ - .- . - . . - -
J Quections -96 through 98 establish the conditions in the containment and drywell for the next time period. These . questions determine the containment pressure and the amount of water in the reactor pedestal cavity. L Late Questions 99 through 125 determine the progression of the accident during the CCI. Containment failures from hydrogen combustion and late overpressure (i.e., from steam and noncondensibles) are addressed in this time period. Similarly, drywell failures from hydrogen combustion and reactor pedestal failure (caused by concrete erosion) are also considered during this time period. The clock time for each period will vary depending upon the type of accident being modeled. This APET does not contain .-any questions to resolve core.vulnerabic sequences. These are PDS6 which have failure of containment haat removal only. The continual deposition of decay heat in the containment by operation of the emergency core cooling system (ECCS) in the recirculation mode is predicted to lead to eventual containment failure in many hours or a few days. Containment failure, in turn, may lead to ECCS failure. T*is is not the case in this study. In the accident frequency analysis it was determined that deformation of injection lines does not occur, and since the systems that take suction from the suppression pool can pump saturated water and, thus, continue to operate, loss of injection does not occur as a result of containment failure. Thus, there are no coro-vulnerable , sequences. In several places in the evaluation of the APET, a User Function is called. l This is a FORTRAN function subprogram which is executed at that point in l the evaluation of the APET. The user function allows computations to be i carried out that are too complex to be treated directly in the event tree. The usu function itself is listed in Appendix A.2, and the general types l of calculations performed by the user function are described below. - The I user function is called to:
- 1. Determine the containment baseline pressure during the various time periods;
- 2. Compute the amount of hydrogen released to the contaitunent at the time of vessel breach and during CCI;
- 3. Compute the concentration and the flammability of the atmosphere in the containment and drywell during the various time periods; 4 Calculate the pressure rise due to hydrogen burns;
- 5. Determine whether the containment fails and the mode of failure;
- 6. Determine whether the drywell fails and the mode of failure.
2.21 I
j i l l 2.3.2 Overview of the APPT Ounntification This section summarizes the ways in which the questions in the Grand Gulf APET were quantified and discusses these methods briefly. A detailed l discussion of each question may be found in Appendix A.1.1. i In addition to the number and name of the question, Table 2.31 indicates if the question was sampled, and how the question was evaluated or
! quantified. In the sampling column, on entry of P - indicates that a
! parameter is sampled from a distribution. The entry Zo in the sampling i column indicates that the question was sampled zero one, and the entry SP means the question was sampled with_ split fractions. The difference may bc . illustrated by a simple example. Consider a question that has two , branches, and a uniform distribution from 0.0 to 1.0 for the probability for the first branch. If the sampling is zero one, in half the observations, the probability for the first branch will be 1.0, and in the other half of the observations it will be 0.0. If the sampling is split f raction,. the probability for the first branch for each observation is a random fractional value between 0.0 and 1.0. The average over all the fractions in the sample is 0.50. The implications of Zo or SF sampling are discussed in the methodology volume (Volume 1) of this report. If the sampling column is blank, the branching ration for that question, and the parameter values defined in that question, if any, are fixed. The branching ratios of the PDS questions change to indicate which PDS is being considered. Some of the branching ratios depend on the relative frequency of the PDSs which make up the PDS group being considered. These branching ratios change for every sample observation, but may do so for some PDS groups and not for others. If the branching ratis change from observation to observation for any one of the seven PD&- groups, SF is-placed in the sampling column for the PDS questions.
- The number of questions associated with each type of quantification is summarited in Table 2.3 2.
l in some cases, a question may have been quantified by more tha6 one source, If this is the case, the entry under Quantification in Table 2.3-1 represents the major contributor _ to the quantification. For example, Question 70, which addresses the loads accompanying vessel breach, was quantified by the Containment Loads Expert Panel and by the project staff. The majority of cases were quantified by the expert panel. There were several cases, however, which the Expert Panel felt were not important. These cases were quantified internally by the project staff. However, because the majority of the cases were quantified by the Expert Panel, the entry in Table 2.3-1 for Question 70 indicates that this question was quant: fled by the Containment Loads Expert Panel. l I 2.22
l Table 2.3 1 Questions in the Crand Culf APET ] Question Number Ouestion Sampling Ounntifiention 1 ) 1. What is the initiating event? PDS
- 2. Is there a Station Blackout? PDS
- 3. Is de Power not available? PDS
- 4. Do one or more S/RVs fail to reclose? SF PDS
- 5. Does llPCS fail to inject? PDS
- 6. Does RCIC fail to inject initially? PDS -
j 7. Does the CRD hydraulic system fail to inject? PDS l 8. Does the condensate system fail? PDS
- 9. Do the LPCS and LPCI systems fail? PDS
- 10. Does RHR fail (heat exchangers not available)? PDS I 11. Does the service water system or
- cross tie to LPCI fail? PDS i
i 12. Does the fire protection system cross-tie to LPCI fail? PDS -
- 13. f.re the containment (wetwell) sprays failed? PDS 14 What is the status of vessel depressurization? PDS
- 15. When does core damage occur? PDS
- 16. What is the level of pre-existing leakage or isolation failure? AcFrqAn
- 17. What is the level of pre-existing AcFrqAn suppression pool bypass?
^
- 18. What is the structural capacity of the containment? P Struct 2.23
t I j Table 2.3 1 (continued) l Question Number Ouestion Samnlinc Ouantification
- 19. Uhat is the structural capacity of the drywell? P Struct i
- 20. What type of sequence is this (summary of plant damage)? Summary
- 21. Do the operators turn on the llIS before core damage (CD)? AcFrqAn
- 22. Is the containment not vented beforn CD? AcFrqAn
- 23. Does (do) any SRV tailpipo vacuum SF Internal breaker (s) stick wide open?
24, Does ac power remain lost during core degradation? SF ROSP
- 25. Is de power available during core degradation? AcFrqAn
- 26. What is the RPV pressure during core dc6radation? AcFrqAn
- 27. What is the status of the 111S before vessel breach (VB)? AcFrqAn
- 28. Is RPV injection restored during core degradation? AcFrqAn -
- 29. Is the core in a critical i
configuration following-injection recovery? Internal
- 30. What is the status of containment sprays? Internal
- 31. What amount of oxygen is in the votwell during CD?- Internal
- 32. What amount of oxygen is in the drywell during CD? Internal l
- 33. What amount of steam is present in the containment at core damage? Internal 2.24
Table 2.3 1 (continued) Question Number Ouestion Sampling Ounntification
- 34. What amount of stese is prraent in the drywell at Jnre dunage? Internal
- 35. Total amount of H2 released in-vessel during CD? P In-Vessel
- 36. What is the level of in vessel Summary zirconium oxidation?
- 37. What is the containment pressure ducing CD? UFUN Int
- 38. What is the level of containment leakage due to slow pressurization before VB? Z0 UFUN Int
- 39. What is the maximum hydrogen concentration in the wetwell before VB? UPUN Int
- 40. To what level is the wetwell inert during CD? UFUN Int
- 41. Do diffusion flames consume the hydrogen released before VB7 SF Internal
- 42. What is the maximum hydrogen concentration in the drywell -
before VB7 UFUN Int
- 43. Do deflagrations occur in the wetwell prior to VB7 SF Loads
- 44. Is there a detonation in the wetwell prior to VB7 SF Loads
- 45. What is the level of containment impulse load before vessel breach? Summary
- 46. With what efficiency is H burned 2 prior to VB? P Loada
- 47. What is the peak pressure in containment irom a hydrogen burn? UFUN Int 1
2.25
--_ ~. - - . . - . - -- .- . - . . . - - . _ _ . - _ . . . . _ - . -. --. .. .-.
Table 2.3-1 (continued) Question _Uumber_ Ouestion Samuling Ouantification i
- 48. What is the level of drywell leakabc induced by an early detonation in containment? 20 UFUN-Str
- 49. What is the icvel of containment leakage induced by an early detonation? ZO UFUN Str
- 50. What is the level of containment I leakage before vessel breach? Z0 UFUN Int !
4 l
- 51. What is the level of drywell leakage induced by containment pressurization? 20 UPUN Int
- 52. What is the level of suppression pool bypass following early combustion events? Z0 Internal
- 53. Has the upper pool dumped? Summary
- $4. Is there water in the reactor i
cavity? Z0 Internal
- 55. What is the containment pressure before VB? UFUN Int
- 56. To what level is the DW steam inert at VB7 UFUN-Inc-
- 57. Is there sufficient H2 for UFUN Int.
combustion / detonation in the DW before VB7
- 58. Does an Alpha Mode Event fail both the vessel and the containment? SF Note 1
- 59. What fraction of the core participates in core-slump? Internal
- 60. Is there a large in vessel steam explosion? Internal
- 61. What fraction of the core debrio v:.*vid be mobile at vessel breach? 20 Internal 2.26
} Table 2.3 1 (continued) Question Number Ouestion Samnling Ouantifiention
- 62. Does a large in vessel steam explosion fail the vessel? Zo Internal
- 63. What is the mode of vessel breach? 20 Internal
- 64. Does high pressure melt ejection occur? 20 Internal
- 65. Does a detonation occur in the DW at VB? Summary 6b. Does a deflagration occur in the DW at VB7 Summary
- 67. Does a large ex; vessel steam explosion occur? Internal
- 68. What amount of H a is released at vessel breach? P In Vessel
- 69. How much hydrogen is released at vessel breach? UFUN Int
- 70. What is the peak drywell/
l vetwell pressure difference resulting from VB7 P -Loads
- 71. What is the peak pedestal pressure at VB? P Loads -
- 72. Is the impulse loading to the drywell at VB sufficient to cause failure? -Z0 UPUN Str l 73. Is drywell pressurization at VB sufficient to cause failure? Z0 UFUN Int 74 Does the RPV pedestal fail.due to pressurization at vessel breach? P Internal
- 75. Does the RPV pedestal fall from an SF Internal ex-vessel steam explosion (impulse loading)?
- 76. Does the RPV pedestal failure induce drywell failure? Z0 Struct 2.27
Tabic 2.3-1 (continued) Question llumh.tL. QutAt.l.QD,. _ . . -- EMEl.1Lg Ouantifieation
- 77. What is the pressure in the containment at VB prior to a hydrogen burn? P Internal
- 78. What is the concentration of hydrogen in containment immediately after VB7 UFUN Int
- 79. Is ac power not recovered following vessel breach? SF ROSP
- 80. Is de power available following vessel breach? AcFrqAn
- 81. What is the status of containment sprays following vessel breach? 20 Internal
- 82. To what level is the wetwell inert after VB7 UFUN. Int
- 83. Is there sufficient oxygen in the UFUN Int containment to' support combustion 64 Does ignition occur in the containment at VB7 SF Loads
- 85. Does ignition occur in the containment following vessel breach? SF Internal-
- 86. Is there a detonation in the wetwell I following VB? SF Internal
- 87. What is the level of containmen*,
impulse load following vessel breach? Summary
- 88. With what efficiency is H2 burned following VB7 P Internal
- 89. What would be the peak pressure in containment from a hydrogen burn at VB7 UFUN Int
- 90. What is the level of containment UFUN Int pressurization at vessel breach?
2.28
Table 2.3 1 (continued) Question _Eupber._ Ouestion Samoling Ouantification
- 91. What is the level of dryvell leakage induced by a detonation in containment at VB? 20 UFUN-Str
- 92. What is the level of containment leakage induced by a detonation at VB7 Z0 UFUN Str
- 93. What is the level of containment leakage following vessel breach? Zo UFUN Int
- 94. What is the level of drywell leakage induced by containment pressurization? Z0 UPUN Int
- 95. What is the level of suppression pool bypass following VB7 Z0 Internal
- 96. What is the contain:. cat prescuro after VB7 UFUN Int
- 97. Is water not supplied to the debris late? Z0 Internal
- 98. Is there water in the reactor cavity after VB7 Internal i
- 99. What is the nature of the CCI? Internal 100. What fraction of core not
(- participating in HPME l participates in CCI? P . Internal 101. How much 2H (& equivalent CO) and CO2 are produced during CCI? UFUN-Int 102. What is the level of zirconium oxidation in the pedestal before CCI? Summary 103. Is the containment not vented following VB? Internal 104. Is ac power not recovered late in the accident? SF ROSP
, -105. Is de power available late in the accident? AcFrqAt>
2.29
- - - - = ... ._ . . . . . , , - -
, , , , , . . ,m,---... , . - , -
l 1 l Table 2.3 1 (continued) Question Number _ _ Ouention Sampling Qgnntification. 106. What is the late status of containment sprays? Z0 Internal 107. What is the late concentration of combustible gases in the containment? UFUN Int 108. To what level is the vetwell inert after VB7 -UFUN Int 109. Is there sufficient oxygen in the containment to support late combustion? UFUN Int 110. Does ignition occur late in the containment? SF Internal 111. Is there a detonation in the vecwell following VB7 Internal 112. What is the late level of Internal containment impulse load? 113. What is the late gas combustion efficiency? Internal 114. What is be the peak pressure in containment from a late hydrogen burn? UFUN Int-115. What is the level of drywell Z0 UFUN Str leakage induced by a late detonation in containment? 116. What is the level of containment leakage induced by a late detonation? Z0 UPUN Str 117. What is the level of containment leakage induced by late combustion events? 20 UFUN Int 116. What in the level or crywell leakage induced by late combustion? Z0 UFUN-Int i 119. Is the containment not vented late? Internal 2.30 l l
, _, ._._...,1,_..__ . . . , , _
, , _ _ . _ . . , , , . . . . . ...u.....__,..,..._...,_.. .._.....,._,....,.;......_..._...,;._._
l 1 i i Table 2.3-1 (Continued) Question Number Ouestion Samplinn Ouantification 120. How much concrete must be eroded to cause pedestal failure? P Struct 121. At what time does pedestal failure occur? P MCCI 122. What is the level of late supprsssion pool bypass? Z0 Struct 123. What is the late containment pressure due to non-condensibles or steam? P Intern
- 124. Does containment failure occur late due to non condensibles or steam? 20 UFUN Int 125, What is the long-term level of containment leakage? Summary Notes to Table 2.3-1 Note 1. The Alpha mode of vessel and containment failure was previously considered by the St.eam Explosion Review Croup. The distribution used in this analysis is based on information contained in the report generated by.
this group. See the discussion of Question 58 in Appendix A.1.1. i*g to Abbreviations and Initialisms in Table 2.3 1 AcFrqAn The quantification was - performed by the Accident Frequency Analysis project staff. Internal The quantification was performed at Sandia National 1.aboratories by the project team with the assistance of other members of the laboratory staff. g' In Vassel This question was quantified by sampling an aggregate distribution provided by the Expert Panel on In Vessel Issues.* Loads This question was quantified by sampling an ag6regate di.=,tribution provided by the Expert Panel on Containment Loads Issues. MCCI This question was quantified by sarpling an aggregate distribucion prnvided by the Expert Panel on Molten-Core /Contain.nent Interaction. Issues. 2.31 1 1
l Y,ey to Abbreviations and Initialisms in Tabic 2.3-1 (continued) P i, value, sampled from a distribution, is assigned to a patsmeter, PDS The quantificatica follows directly the definition of the PDS. ROSP This question was quantified by sampling a distribution derived from the offsite power recovery data for the plant. SF Split fraction sampling: the branch probabilities are real numbers between zero and one. Struct This question was quantified by sampling from a aggregate distribution provided by the Expert Panel on Structural Issues. Summary The quantification for this question follows directly from the branches taken at preceding questions, or the values of parameters defined in proceding questions. UFUN-Str This question is quantiff ad by the execution f . module - in the Osor Function subroutino, u ing distribut. ins from the Structucal Expert Panel UFUN Int This question is ~' 4ntified by the execution of a module in the User Functior ,abroutine using models and data generated by the project st - , 20 Zero One sampling: tne branch probabilities are either 0,0 or 1.0, l 2.37
I Table 2.3 2-Grand Gulf APET Quantification Summary Type of Numt_r of-Ounntificetion Ouestions~ Comments PDS 15 Determined by the.PDS. AcFrqAn 10 Determined by the Accident Frequency Analysis. . InternM 37 Ooantif 4.a<1 j et arnpilv ' this analysis. Summary 9 The branch taken at - th x s question follows directly from the branches taken at previous questions. ROSP 3 This question wr o quantified by sampling a distribution scrived from the offsite power recovery data for the plant. UFUN-Str 7 Calculated-in the User Function using distributions frcm the Structural Expert Panel. UFUN Int 29 Calculated in the User Function using models and data generated by ' the pioject staff; In-Vessel 2 Dis. -ibutions from the In-V. ssel Expert Panel. Loads , 6 Disttibutions from the 4 4ntainment Loads Expert Panel MCCI 1 Distributions from the Molten Core-Containment - Interaction Panel. Struct 5 Distributions from the Structural Expert Panel. Other Expert 1 See Note 1. Table 2.3-1, 2.3.3 Variables Samnled for the Accident Progression Analysig About 186 variables vere sampled for the accident progression analysis. That is, every time the APET was evaluated by EVNTRE, the original values of about 186 variables were replaced with values selected for- the particular observation under consideration. These values were selected by the Latin flypercube Sampling (LHS; pro 6 rata from distributions thar were defined before the APET we.s evaluat(d. Many of these distributions were determined by exper c panels, Tabic 2.3-3 lists the variables in the APET which were sampled for the accident progression analysis. Some of them are 9 33
- 1
. . . _ _ _ _ . _ _ _ . _ ___ _ _ _ . . _ ._ -J
branch fractions; the others are parameter values for use in celculations, performed while the APET is being evaluated. In Tabic 2.3-3, the first column gives the variable abbreviation or identifier, and the question (and case if appropriate) in which the variable is used. Where several variables are correlated, they are treated as one variable in the sensitivity analysis (see section 5.1,4), but are different variabica as .far as the accident progt;ssion analysis and sehpling process are concerned. The second column gives the range of the distribution for the variable. The minimum and maximum values o' the distribution are listeo in this col umn An entry of "Zero/One" ta this column indicates that the variablo was sempled Zero One, i.e., it took on only the values 0.0 and 1.0. . ~i n each observation, one of these two values would be assigned. The third column in Tabic 2.3 3 indicates the type of distributin- sed. For uniform distributions from 0.0 to 1.0, the mean is obvious and so is not listed. Otherwise. the mean is given, if appropriate. The entry
" Experts" for the distribution indicates that the distribution came from an expert panel and the entry " Internal" distribution indicates that the distribution was determined by some method other than the formal expert clicitation process. (None of the distributions obtained by aggregating the conclusions of experts can be described succinctly in words, Plots of the aggregate expert distributions are contained in Volume 2 of this report. A listing of the input to the LilS program that contains many of these distributions in tabular form is given in Appendix E.) For Zero One variables, an indication of the probability of each state is given in this column.
6 The fourth and fif th columns in Table 2.3-3 show whether the variabic is correlated with any other variable, " Rank 1" indicates a rank correlation , of 1.0. An "n" is used to indicate any integer. In- the entry for ll2INVES1, ll2INVESn in the " Correl, with" column indicates that ll2INVES1 is correlated with H21NVES2, H2INVES3,...., and H2INVES6. For further information on each of the variables listed in the table, see the detailed discussion of the indicated APET question in Appendix A. A 2.34 1
i Tab le 2. 3- 3 Variables Sampled in the Accident Progression Analysis Variable Correlated Question Description
& Case Rance Distribution Correlation Uith Rank.1 IPDUF Cor.tainment failure pressure (kPa).
PCFall 195 Internal 755 Mean 383 EPDUF Q18 Cl Q173 C3 IPDWF 260 Internal Rank 1 PCFail Dry 4 ell failure pressure (kPa) when the 963 Mean - 588 EPDUF pressure loading is inside the drywell. Q19 C1 EPDUF 260 Internal Rank 1 PCFail Drydell failure pressure.(kPa) when the , 963 Mean - 588 IPDUF pressure loading is outside the drywell ! Q19 Cl (i.e., in the wetwell). l 9 Io CFRan 0.0 Uniform Rank 1 DUFRan Random number used to determine-the 1.0 mode of containment failure (Quasi-Q18 Cl static loads). i DUFRan 0.0 Uniform Rank 1 CFRan Random number used to determine the 1.0 mode of drywell failure (Quasi-static ; Q19 Cl loads). , r 0.0 Experts Rank 1 IMPDUF The failure impulse (kPa-s) of the IMPCF. 102.5 Mean - 19.5 c ntainment. . Q18 Cl IMPDu? 2.5 Experts Rank.1 IMPCF The failure impulse (kPa-s) of the 125 Mean - 33. dryvell. Q19 C1 > IMRanC 0.0 Uniform Rank 1 IMRanD Randon number used to determine the' mode of containment failure (Impulse loads) . .f Q18 C1- 1.0 , i i ! IMRanD 0.0 Uniform Rank 1 IMRanC. Randon number used to determine the mode' of dryvell failure (Impulse' loads). j Q19 Cl l'. 0 'l j
Table 2.3-3 (continued) , Variable Question Correlated Correlation With Description
& Case Raneq Distribution SRVBkrl 0.01 Uniform None The failure probability of a SRV Q23 C2. 0.50 tailpipe vacuum breaker (RPV at high ~
pressure). SRV3kr2 0.01 Uniform N'ne The failure probability of a SRV 0.10 tailpipe vacuum breaker (either ATUS or Q23 C4 RPV at low pressure). H2INVES1 0.0 Experts Rank 1 'H2INVESn The amount of hydrogen .(kg-moles) Q35 Cl 955 Mean - 222 produced in-vessel during an ATUS in which coolant injection is restored to g the RPV. L es H2IrNES2 0.0 Experts ' Rank 1 H2INVESn The atount of hydrogen (kg-moles) Q35 C2 1267 Mean - 461 produced in-vessel during an ATUS in which coolant injection 'is not restored to the RPV. H2INVES3 0.0 Experts Rank 1 H2INVESn The amount of H2 (kg-moles) produced in-1042 Mean - 333 vessel. The RPV is at high pressure and Q35 C3 coolant.is restored to the RPV. The PDS is not an ATUS. F2INVES4 0.0 Experts Rank 1 H2INVESn The amount of H 2(kg-moles)' produced in-903 Mean - 283 vessel. The RPV is at low pressure and Q35 C4 coolant is restored to the.kPV. The PDS is not an ATWS. I H2INVESS 36.4 Experts Rank 1 .H2INVE$n' The amount of H 2 (kg-moles) produced in-1251 Mean - 450 vessel. -The RPV is at high pressure and Q35 C5 coolant is not restored to the RPV. The PDS is not an ATUS. ___ _ _ _ _ _ _ -_. __ _ __ _._ ._ a
i i I' s :l Table 2.3-3 (continued) Var. Ques. Correlated Correlation Uith Description ,i
& Case Ranee Distribution lI H21NVES6 0.0 Experts Rank 1 H2INVESn The amount of H2 (kg-moles) produced in- ;
vessel. .The RPV is at low pressure and- '! Q35 C6 1285 Mean - 466 coolant is not restored to the RPV. The ,
- - PDS is not an f G j Dif-nsB 0.5 . Uniform None The probabilty that the H2 in the '[
.Q41 C3 1.0 burns as a diffusion flame when the PDS is not a station blackout and the HIS.is
- off. ,
0.0 Internal . Rank 1 Dif-SBn The probability that'the H2 in the Dif-SBl 0.17 Mean .12. containment burns.as diffusion flame Q41 C4 during a SB'in which ac power is
,N recovered and the HIS is on.
y
.Dif-SBn
+
Rank'1 The proba'.ility that the H2 in the Dif-SB2 0 .' O Internal
.'I 3.085 Mes. .06 containment burns es diffusion flame Q41 C5' during a SB in.which ac power'is recovered and the'HIS is off. s DflgBVB1 0.0 ~ Experts. Rank'1 -
DflgBVBn Probability of_ hydrogen ignition before. ~ l 0.72 Mean ;.18 VB. - The.RPV is;at-high pressure, there - Q43 C4 'is no ac. power and Hz < 4%-
~
+
DflgBVB2' O.0 Experts - : Rank 1' ;Dfl6BVBn Probability .of hydrogen. ignition before ' 0.74 Mean - .23 VB. The RPV is at high pressure sad. Q43 C5 there is no fac power and 4t '< H2 < 'St. t
' DflgBVB3 -0.0- - Experts . . -
Rank 1- . DflgBVBn- ' Probability .of hydroget ignition. ' The RPV is 'at low pressure and there is
. Q43 C6' O.72 Mean .21 *
- no a'c power and.4% < H2 :< 8%.
QB5'C6 , q DflgBVB4 0.0 Experts- Rank 1- ~DflgBVBn' Probability of. hydrogen -ignition. before 10.75 .Mean.- .28 VB. The RPV is at high pressure s nd
' Q43 C7 there-is no ac. power and 8% < H2 < 12%. .l
.g.,
-7= - or -
w-- a + 3.e - . - r ----,w--4---s eh,_-.----- = y- .g._,, .t-.y- 4 ..- , . _ _ , -___g _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _
Table 2.3-3 (continued) Variable Ouestion Correlated Correlation Uith Description
& Case Rance Distribution DflgBVB5 0.0 Exports Rank 1 DflgBVBn Probability of hydrogen ignition,
- 0.75 Mean .28 The RPV is at low pressure and ttere is Q43 C:I no ac power and 8% < H2< 12%-
Q85 C:s DflgBVB6 0.0 Experts Rank 1 DflgBVBn Probability of hydrogen ignition before 0.75 Mean .39 VB. The RPV is at high pressure, there Q43 C9 ' is no ac power and 12% < H2< 165 DflgBVB7 0.0 Experts Rank 1 DflgBVBn Probability of hydrogen ignition. 0.75 Hean .38 The RPV is at low pressure and there is Q43 C10 no ac power and 12% < H2 < 16%. y QS5 C4 DflgBVBS- 0.0 Experts ' Rank'l DflgBVBn Probability of hydrogen ignition before 0.75 Mean .50 VB. The RPV ic at high pressure and Q43 C11 there is no ac power and H2> 16%. DflgBVB9 0.0 Experts Rank 1 DflgBVBn Probability of hydrogen ignition. 0.75 Mean - 49 The RPV is at low pressure and there is Q43 Cl2 no ac power and Hz > 16%. Q85 Cl-DronB7B1 0.0 Experts Rank 1 DtonBVBn Probability that the hydrogen detonates Q44 C2 .66 Mean .22 given that the H2 has ignited, the steam concentration is high, and 12% < H2< Q86 C4 i 16%. s DronBVB2 0.0 Experts Rank 1 DtonBVBn Probability that the hydrogen detonates Q44 Cu .75 Mean .25 given that the H2 has ignited, the steam Q86 CG,8 ,- concentration is high, and H2> 10%. DronBVB3 00 Expert Rank 1 DtonBVBn Probability that the hydrogen detonates Q4A C5 .65 Mean .t6 given that the H2 has ignited, the steam concentration is. low, and 16% < H2 < Q86 C7 20%. i
Table 2.3-3 (continued) Variable Correlated Question With Description
& Case Rance Distributier .'orrelation Rank 1 DtonBVBn Probability that the hydrogen detonates DtonBVB4 J.0 Experts
.70 Mean .45 given that the H2 has ignited, the steam j Q44 C7 concentration is low, and H2> 20%-
Q86 C9 Rank 1 ImpLoada "the impul. isad (kPa-s) on the d gwell ImpLoadl 2. 8 Experts
- from a is . ion when 12% < H2< 16%.
i Q44 C2 12.3 Mean - 5.8 Q86 C4,6,8 Rank 1 ImpLoadn The impulse load (kPa-s) on the dgwell lepLoad2 0.0 Experts 63.0 Mean - 12.4 from a detonation when H2> 16%. m Q44 C5 o Q86 C2.3,7,9 e EffBrnPn The-effect' , burn efficiancy in the EffBrnP1 0.0 Experts Rank 1
.079 wetwell giw n that the steam Q46 C2 0.24 Mean concentratio.. is high and Ii2 < 4%_
EffBrnPn The effectiv turn efficiency in the EffBrnP2 0.0 Experts Rank 1
.28 wetwell givro that the steam Q46 C4 0.64 Mean concentratio.' is high and 4% < H2< 8%.
QS8 C1 EffBrnPn The effective burn efficiency in the EffBrnP3 0.0- Experts Rank 1
.28 wetwell given that the steam Q46 C5 0.65 Mean .
concentration is low and 4% < H2 < 8% . Q88.C2 Rank 1 EffBrnPn The effective burn efficiency in the EffBrnP4 0.0 Experts wetwell given that the steam
~
Q46 C6 0.93 Mean - .46 , concentration is high and 8% < H 2'< 12%- i Q88 C3 Rank 1 EffBrnPn The effective burn efficiency in the EffBrnPS 0.0 Experts
.57 wetwell given that the steam Q46 C7.. 0.83 Mean concentration is low and 8% < H2 < 12%.
Q88 C4 w
Table 2.3-3 (continued) Varieble Correlated Question Description _& _ Case Rance Distribution Correlation Uith Rank 1 EffBrnPn The effective burn efficiency in the EffBrnP6 0.0 Experts 0.84 Mean .48 wetwell given that the steam Q46 C8 concentration is high and 12% < H.. < Q88 C5 16%. EffBrnPn The' effective burn efficiency in the EffBrnP7 0.0 Experts l'enk 1 0.94 Mean .73 wetwe11 given that the steam Q46 C9 concentration is low and 12% < H 2'< 16%. Q88 C6 Renk 1 EffBrnPn The effective burn efficiency in the 9 EffErnPS 0.0 Experts b 0.78 Mean .49 wetwell given that the steam Q46 C10 concentration is high and H2> 16%. Q88 C7 EffBrnP9 0.0 Experts Rsnk 1 EffBrnPn The effective burn efficiency in the 1.0 Mean .75 wetvell given that the steam Q46 C11 Q88 C8 concentration is low and H2 > 16%. Brncmpil- 0.0 Experts Rank 1 BrnCapin The actual burn completeness in the 0.69 Mean .27 vetvell given that H2 < 8% . Q46 C2 Q88 C1,2 Rank 1 ErnCmpin The actual burn completeness in the-BrnCap12 0.37' Experts
- 0. E') Mean .74 wetwell given that 8% < H2 < 12%.
Q46 C6 Q88 C3,4 0.53 Experts Rank 1 BrnCmpin The actual burn completeness in the BrnCmpl3 1.0 Mean .88 i vetvell given that 12% < H2 < 16%. Q46 C8 Q88 CS,6-BrnCmp14 0.59 Experts Rank 1 BrnCmpin The actual burn completeness in the 1.0 Mean .93 wetwell given that H2> 16%. Q46 C10 Q88 C7,8
't , r T
I Table 2.3-3-(continued) Variable f Question Correlated ; l- With Description i~ & Case Rance- Distribution Correlation l DWVacBkr. Zero Fail 0.05- :bne The, probability that theidrywell xacuum l Q52.C2 One. breaker will fail to.reclose after a , hydrogen burn in the the wetwell-(ac~
- i. Q95 C2,3 Q122 C3,5 power must be available). .
DUF1dDif- ~ Zero F1d 0.45 Wne The probability : that a hydrogen Lt rn .; '
-One Wet 0.45 (diffusion flamer pushes suppression r--- :Q54 C4 Dry 0.'10 pool water in'theldrywell
- DWFldH21 Zero Fld' O.50 - Rank 1- DWFldHa n Tbc probability that the accumulation of _j H ..in the wetwell' pushes pool water:into - .i
-w Q54'C5 .One Het 0.50- .. 2 4* the-drywell given that the upper' pool- l
' has dumped. j DWFldH22 Zero. Wet 0150_. - Rank 1 DWFldH ' 2n The probability.that the' accumulation'of ;
'i
. 2H in"the wetwell pushes pool water into Q54 C8 One' Dry 0.50 the drywelligiven that the upper. pool '
has not dumped. - t ALPHAl- 0.0 Experts Rank ~1 jALPHAn. Probability 7that an Alpha mode event:
"Q58 C2 ' 1.0 :Mean -L.01'- occurs; given that:the.RPV is at low 1 l - pressure.. -t
'i 0.0. Experts. Rank 1 ALPHAn- Probability.-.that"an Alpha mode event f ALPHA 2_-
- occurs, given .that: the RPV is at high i Q58 Cl- 0.1~ " Mean - 001 l.
-pressure.- 'l
. .i Probability:that thereiis~a large. amount
~
g
- LiqVB1- Zero HILiq.0.025- 1$one-One
- LoLiq 0.9752 . of molten core debris-(HiLig) at UB l -Q61.C1 ~
~ given that coolant injection -is being, .. r y supplied to the RPV.
}
?
m ' $ ~ ., -
,v ,
. . - - _ ; w ,.v .-
g- %,.. -ny,-~._+.,~_49.n_..v..a, u.~.- a.n- .-,v .y: _ _ _ _ _ _ _ _ _ . . _ _ _ __:_, ._ _ x,.,
Table 2.3-3 (cont:inued) i Variable Correlated Question Description
& Case _ Rance Distribution Correlation With _
LiqVB2 Zero HiLiq 0.10 None Probability that there is a large amount One LoLiq 0.90 of molten core debris (HiLig) at VB Q61 C2 given that coolant is not being supplied to the RPV. Zero BtHd 0.2 None The probability that an in-vessel steam F-F2V-5E One LgBrch 0.2 explosion will fail the RPV Q62 C2 SmBrch 0.3 nFail 0.3 2ero BtHd 0.124 Rank 1 F-RPVn The probability that the RPV will fail F-RPV1 N Q63 C5 One LgBrch 0.005 given that a large amount of the core SmBrch 0.371 is molten and coolant is being. injected nFail 0.500 into the RPV. Zero BtHd 0.249 Rank 1 F-RPVn The probability that the RPV will fall F-RPV2 Q63 C6,C7 One LgBrch 0.005 .given that there is no coolant C9,C10 SmBrch 0.746 inj ection. nFail 0.000 Zero BtHd 0.062 Rank 1 F-RPVn The probability that the RPV will fail F-RPV3 . Q63 C8 One LgBrch 0.005 given that a small amount of the core is SmBrch 0.188 molten and coolant is being injected i nFail 0.745 into the RPV. . HPME Zero HPME 0.8 None Tb e probability of an HPME event given One ', that the RPV fcils at high pressure.
-Q64 C2
) F 2AVB1 0.0 Experts Rank 1 H2AVBn The amount of H2 (kg-moles) produced at 781 Mean - 61 H21NVESn VB during an ATUS in which coolant Q68 C2 injection is restored to the RPV. a
, < 7
4 Table 2.3-3 (continued) Variable Question Correlated With Description
& Case Rance Distribution Correlation 0.0 Experts Rank 1 H2AVBn The amount of hydrogen (kg-moles)
H2AVB2 642 Mean - 89 H21NVESn produced at vessel breach during an ATUS < Q68 C3 in which coolant injection is not restored to the RPV. 0.0 Experts Rank 1 H2AVBn The amount of H2 (kg-moles) produced at H2AVB3 VB. The PDS is not an ATWS, the RP'/ is 260 Mean - 53 H2INVESn Q68 C4 pressurized, and coolant is restored to the RPV during CD. 0.0 Experts Rank 1 H2AVBn The amount of H 2 (kg-moled produced at H2AVB4 VB. The RPV is at low press.tre and
" 156 Mean - 27 H21NVESn Q68 C5 coolant is restored to the EPV. The PDS is not an ATUS.
0.0 Fynerts Rank 1 H2AVBn The amount of H2 (kg-moles) produced at H2AVB5 VB. The RPV is at high pressure and 625 Mean - 234 H21NVESn Q68 C6 coolant is not restored to the RPV. The PDS is not'an ATUS. j 0.0 Experts Rank.1 P2AVBn The amount of H2 (kg-moles) produced at H2AVB6 H2INVESn VB. The PDS is a not an ATUS, the RPV is Q68 C7 417 Mean - 62 pressurized, and coolant was restored to the RPV during CD. 0.0 Experts Rank 1 DWPVB2,5,6 The peak drywell/wetwell pressure DUPVB1 differentia.1 (kPa) at VB. RPV fai.ls at 2000 Mean - 434 CP-VB1 Q70 C2 high pressure into a wet cavity (Expert i t
..w --a - . - .-.A
i Table 2.3-3 (continued) l l Variable Correlated Question With Description
& Case Rance Distribution Correlation Experts Rank 1- DUPVB1,5,6 The peak drywell/wetwell pressure DVPVB2 0.0 2000 Mean - 332 CP-VB1 differential (kPa) at VB. RPV fails at Q70 C3 high pressure into a wet cavity (Expert Case 1-hC).
Rank 1 DUPVB4,7,8 The peak dryvell/wetwell pressure DWPVB3 33. Experts differential (kPa) at VB. RPV fails at CP-VB2 Q70 C4 950 Mean - 392 high pressure into a dry cavity (Expert Case 2-HC). eo Rank 1 DUPVB3,7,8 The peak drywell/wetwel?_ pressure
'e DWPVB4 20. Experts differential (kPa) at VB. RPV f.ils at CP-VB2 Q70 C5 531 Mean - 242 high pressure into a dry cavity (Expert Case 2-hC).
0.0 Experts' Rank 1 DWPVB1,2,6 The peak drywell/wetvell pressure DWPVB5 2000 Hean - 425 CP-VB1 differential (kPa) at VB. RPV fails a*; Q70 C6 high pressure into a wet cavity (Expert Case 1-He). Experts Rank 1' .DWPVB1,2,5 The peak dryvell/wetwell pressure DWPVB6 0.0 2000 Mean - 311 CP-VB1 differential (kPa) at VB. RPV fails at Q70 C7 high pressure into a wet cavity (Expert Case 1-he). Rank 1 DUPVB4,5,8. The peak drywell/wetwell pressure DUPVB7 33. Experts 850 Hean - 336 : CP-VB2 differential (kPa) at VB. RPV fails at Q70 C8 high pressure into a dry cavity.(Expert Case 2-He).
,:.-.s---,:
- ' , .,s,
, 3 Table 2.3-3 (continued)
Variable Correlated Question With Description
& Case Range Distribution Correlation l
- 20. -Experts Rank 1 DUP 7B4,5,7 The peak drywell/wetwell pressure DWPVB3 531 Mean - 222 CP-VB2 differential (kPa) at VB. RPV fails at Q70 C9 high pressure into a dry cavity (Expert Case 2-hc).
Rank 1 DUPVB9-12 The peak drywell/wetwell pressure DWFVB9 0.0 Experts 2000 Mean - 295 CP-VB3 differential (kPa) at VB. RPV fails at Q70 C10 low pressure into a ver cavity (Expert case 3-HC). Rank 1 DUPVB9-12 The peak drywell/wetwell pressure DUPVB10 ~0.0 Experts
. to 2000 Mean.- 242 CP-VB3 differential (kPa) at VB. RPV fails at b
Q70 C11 low pressure into a wet cavity (Expert Case 3-hC). Rank 1 DUPVB9-12 The pess 1 ywell/wetwell pressure 0.0 Experts DWP'2311 2000 Mean - 290 CP-VB3 differeu-i.tl (kPa) at VB. R"V feils at-Q70 C12 low pressure into a wet cavity (Expert Case 3-He). i Experts Rank 1 DUPVB9-12 The peak drywell/wetwell pressure DUPVB12 0.0 2000 Mean - 239 CP-VB3 differential (kPa) at VB. RPV fails at Q/O C13 low pressure into a wet cavity (Expert Case 3-he). 550 Experts Rank 1 PedVB2,5,6 The peak pedestal cavity pressure'(kPr) PedVB1 at VB. RPV fails at high pressure into a Q71 C2 8370 Mean - 3580 , we t , wity (Expert Case 1-HC). 468 Experts Rank 1 PedVB1,5,6 The peak pedestal cavity pressure (kPa) PedV82 at VB. RPV fails at high pressure into a Q71 C3 8370 Mean - 2780 wet cavity (Expert Case 1-hC).
' '"' " me -v1.-
- ___ -..~.. . - - - - - - -.. .. .-- --- - - - --- - - --
.{
i i Table'2.3-3 (continued)
-I Variable .
Correlated
-Question Description !
& Case Rance ' Distribution Correlation Vith PedVB3 385 Experts Rank 1 PedVB4,7,8 The peak pedestal cavity pressure (kPa) l Q71 C4 6000 Hean - 3080 . at VB. RPV fails at high pressure into a !
dry cavity (Expert Case'2-HC). ] 4
- PedVB4 0 Experts' Rank . PedVB5,7,8 The peatt pedestal cavity-pressure (kPa)~
4980 Mean - 1720 at VB. RPV fails at high pressure into a Q71 C5 dry cavity (Expert Case 2-hC). [ 440 . Experts . Rank 1 PedVB1,2,6 The peak pedestal cavity. pressure (kPa) -j PedVB5 ' 6700-Mean - 3250 at VB. RPV fails at high pressure into a
.Q71 C6 '
wet cavity (Expert Case 1-Hc).- + n _
'n :PedVB1,2,5 The peak pedestal cavity pressure (kPa)..
~
PedVB6 374 Experts ! Rank 1 '];
,5690 Mean 2170 at VB.'RPV fails at high pressure'into a'-
Q71 C7 wet ' cavity (Expert Case 1-he) . .. PedVB7 "308 L Experts .
~ Rank'1 PedVB5,6,8 ! The peak pedestal cavity pressure (kPa)- ,
6000 Mean'- 2850 at VB. ii.PV fails at high pressure into.a Q71 C8' dry cavity.(Expert Case.2-He). l l-q 262- Experts -Rank 1 PedVBS,6,7. ' The' peak pedestalicavity pressure-(kPa) ' Pe4VB8- . 3990 Mean - 1430 at VB.'EPV fails;at high pressurerinto.a
- . .Q71 C9-4
~
dry. cavity (Expert Case 2-he). l i ..
- The peak pedestal cavity pressure (kPa)-
s i PedVB9 '.200 , Experts' Ranks 1 fPedVBn' j
..Q71 C10.. 4200 Mean - 1120'- n-11-17 at VB. RPV fails at: low pressure into a l- . wet cavity (Expert Cases 3-OHC and
!. Q71 C12' ,
, 3-OHO).-
- i. ';
j .PedVB10' -138) Experts. Rank-l' ' 'PedVBn The peak' pedestal cavi.ty pressure (kPa)
-2400 Mean - 734;. cn-11-17' . at VB. RPV fails at low-pressure into a-
- Q71 C11 '
- wet cavity.(Expert Case 3-OhC). .
~
t j j i I
. t
, . -u, 4 . r, - - , y . ..w.- -
...w.. -e r. . . . , . .--r , u,.. ,
i i i 4 1able 2.3-3 (continued)
't Varis ble .t Ques ; ion -
Correlated Description ~
&- Case - Rance Distribution -Correlation Uith j PedVB11 69 Expertsc Rank 1 PedVBn- The peak pedestal cavity pressure (kPa) ;
n-11-17 at VB. RPV fails'at lew pressure into a l Q71 C13 2400 Mean - 557 -
- wet cavity (Expert Case 3-ohC). !
i PedVB12 100-~ Experts . Rank 1 PedVBn The peak pedestal cavity pressure (kPa). Q71 C14' 4200 Mean - 1000 - n-11-17 at.VB. RPV fails,at low pressure into a ,
- vet cavity (Expert- Case : 3-OHe) . l PedVB13' 100' Experts. - Rank 1 PedVBn The peak pedestal' cavity pressure (kPa). i
..Q71 C15 2100.Mean - 606 n-11-17 at VB. RPV fails- at low. pressure into a w
wet cavity'(Expert Cases 3-Ohe and Q71 C16 3-oHc); b PedVB14 69 Experts'. Rank 1 ; PedVBn- The peak _pe'destal. cavity pressure (kPa)'
. n-11 at VB. RPV fails'.at low pressure into a Q71 C17 1600 Mean - 436 wet: cavity (Expert Case.3-ohe)..
l' None: Pedestal failure pressure:(kPa) i PedFail' 900 Uniform , Q74 C1 . 1700 y PedExSE' OO Uniform' None: The probability that the ' reactor l
.Q75 C11. 1.S' pedestal fails from an ExSE!given that
'i an ExSE occurs at VB. ,
CP-VB1 3.35 Internai . Rank 1 : DVPVBd Wetwell pressure rise- (kPa) at VB prior . 1
' to a burn..RPV fails at high pressure' .J; 227 Mean - 50. n-1,2,5.6 Q77: C2 into a wet cavity;,the suppression pool .1
- : is bypassed at VB. I CP-VB2. ; - 4.36 Internal' ' Rank 1 'DWPVBn' Wetwell pressure rise.(kPa) at VB prior 1 41- n-3,4,7,8 ~to a burn. RPV fails at high pressure- , i
;Q77 C3 :
-92.5'Mean into a' dry cavity; the suppression' pool is bypassed a: VB..
s
.s. -- -
_y.-,.- r-e ._.~._i.<=- ~ , ,.oc
.. s+ e . y
- g. --s- - r .<.
r P+ " ;""2 -
'---m--n- t"'"i- .W:"
Table 2.3-3 (continued) Variable Question Correlated Rance Distribution Correlation Uith Description
& Case i
CP-VB3 2.36 Internal Rank 1 DUPVBn Uetwell pressure rise (kPa) at VB prior 227 Mean - 35 n-10-13 to a burn. RPV fails at low pressure
- Q77 C4 into a wet cavity; the suppression pool is bypassed at VB.
CP-VB4 0.0 Uniform None Wetwell pressure rise (kPa) at VB prior 113.5 to Lurn Either HPME or ExSE occurs Q77 C6 at Q77 C7 VB; the suppression pool is not bypassed at VB. CSFaill Zero Fail 0.50 Rank 1 CSFail2 Probability that the energetic events One Reevy 0.50 that ruptured the containment at VB also s QS1 C2 io fail the CS (ac power is not available). CSFail2 Zero Fail 0.50 Rank 1 CSFaill Probability that the energetic events One Oper. 0.50 that ruptured the containment at VB also Q81 C4 fail the CS (CS were operating before VB). CSFai13' Zero Fail 0.50 None Probability that the energetic events One Avail 0.45 that ruptured the containment at VB alzo Q81 C6 Q106 C6 Oper. 0.05 fail the CS (ac power is recovered following VB). IgnAVB1 0.1 Expert Rank 1 IgnAVBn Tne H2 ignition probability at VB given 0.92 Mean .63 that the RFV fails at high pressure or Q84 C3 there is an ExSE and H2 > 16 % . IgnAVB2 0.04 Expert Rank.1- IgnAVBa The H2 ignition probability at VB given 0.87 Mean .56 that the RPV fails at high pressure or-l Q84 C4 there is an ExSE and 12% < H2 < 16%. l L_ -_ _ - - _ _ _ _ - . - -
a ___ Table 2.3-3 (continued) Variable Correlated , Question With Description I & Case Range Distribution Correlation 0.02 Expert Rank 1 IgnAVBn The H 2 ignition probability at VB given IgnAVB3 that the RPV fails at high pressure or Q84 C5 0,67 Mean .43 there is an ExSE and 8% < H2 < 12%- , Rank 1 IgnAVBn The H 2 ignition probability at VB given IgnAVB4 0.0 0.6 Expert . Mean .29 that the RPV fails at high pressure or ] 4 Q84 C6 - there is an ExSE and 4% < H2 < 8%- 0.0 Expert Rank 1 IgnAVBn The H2 ignition probability at VB given IgnAUBS that the RPV fails at high pressure or Q84 C7 0.035 Mean -0.005 there is an ExSE and H2 < 4%. Zero Fall 0.175 None The probability that pedesta) failure DU-Ped-F induces drywell failure given that the Q76 C2 One - pedestal fails. Q122 C2,5 Zero noWat 0.50 None The probability that a coolant injection LDBWatl system supplies water to the debris Q97 C2 One Lgust 0.25 SmWat 0.25' after VB given that ac power is not
-allable. ,
Zero noWat 0.33 None The probability that. a coolant injection LDBWat2 One LgWat 0.33 system supplies water to the' debris Q97 C4 after VB'given that an. injection system SmWat 0.33 was working before VB. 2ero novat 0.50 None The probability that a coolant injection LDBWat3 One Lgvat 0.25 , system supplies, water to the debris Q97 C5 after VB given that there.was no Smvat 0.25 injection before VB.
..___M
Table 2.3-3 (continued) Variable Question Correlated Correlation With Description
& Case Rance Distribution CD-CCIl 0.6 Uniform None The fraction of core debris that 1.0 participates in CCI; given that a large Q100 C2 amount of core debris participates in an Ex5E.
CD-CC12 0.9 Uniform None The fraction of core debris that 1.0 participates in CCI; given that a small Q100 C3. amount of core debris participates in an ExSE. L-CIgni 0.0 Expert . Rank 1 L-CIgnn The H2 ignition probability late in the y
, Q110 C4 0.7f. Mean - 0.51 accident given that there is no ac power c) - and H2> 16%.
L-CIgn2 0.0 Expert Rank 1 L-CIgnn The H2 ignition probability late in the 0.73 Mean - 0.42 accident; there is no ac power and 12% Q110 C5
<H2 < 16%.
L-CIgn3 0.0 Expert Rank 1 L-CIgnn The H2 ignition probability late in the 0.75 Mean - 0.33 accident; there is no ac power and Q110 C6 8% < H2 < 12%. L-CIgn4 0.0 Expert Rank 1 L-CIgnn The H2 ignition probability late in the 0.73 Mean - 0.29 accident; there is no ac power and Q110 C7 4% <H2 < 8% - ConErPed 0.3 Expert None The depth (H) of concrete erosion that 2.1 Mean - 1.1 will fail the reactor pedestal. Q120 C1 PedFlG1 0.0 Expert Rank 1 PedFnCn The depth of concrete eroded (M) in 1 h Q121 C3 0.53 Mean - 0.19 during CCl--Expert Group 1.
+
Table 2.3-3 (continued) Variable Question Correlated
& . Case Rance Distribution Correlation Uith Descrintion PedF1G2 0.0 Expert Rank 1 PedFnCn The depth of concrete eroded (M) in 1 h Q121 C5 0.39 Mean - 0.14 during CCI--Expert Group 2.
PedF1G3 0.0 Expert Rank 1 PedFnCn The depth of concrete eroded (M) in 1 h Q121 C4 0.53 Mean - 0.16 during CCI--Expert Group 3. PedF1G4 0.023 Expert Rarik 1 PedFnCn 'Ihe depth of concrete eroded (M) in 1 h Q121 C9 0.43 Mean - 0.20 during CCI--Expert Group 4 PedF1G5 0.023 Expert Rank-1 PedFnCn The depth of concrete eroded (M) in 1 h 0.61 Mean - 0.26 during CCI--Expert Group 5. Q121 C8 The depth of concrete eroded (M) in 1 h
$ PedF1G6 0.023 Expert . Rank 1 PedFnCn Q121 C6 0.60 Mean - 0.20 during CCI--Expert Group 6. >
PedF1G7 0.023 Expert Rank 1 PedFnCn The depth of concrete eroded (M) in 1 h Q121 C7 0.61 Mean - 0.26 during CCI--Expert Group 7. PedF3G1 0.0 Expert Rank 1 PedFnCn The depth of concrete eroded (M) in 3 h Q121 C3 0.75 Mean .0.32 during CCI--Expert Group 1. PedF3G2 0.0 Expert Rank 1 PedFnCn The depth of concrete eroded (M) in 3 h Q121 C5 0.68 Mean - 0.26 during CCI--Expert Group 2. PedF3G3 0.0 Expert Rank 1 PedFnCn . The depth of concrete eroded (M) in 3 h . Q121 C4 0.75 Mean - 0.29 during CCI--Expert Group'3. PedF3G4 0.075 Expert Rank 1 PedFnCn The. depth of concrete eroded (M) in 3 h Q121 C9 0.85 Mean - 0.40 ' during CCI--Expert Group 4 4 PedF3G5 0.075 Expert Rank 1 PedFnCn The depth of concrete eroded (M) in 3 h Q121 C8 0.85 Mean - 0.47 during CCI--Expert Group 5. 4
~
_ - - . _ . _ . _ _ . - _ _ _ _ _ _ . _ - L %
~ . __
Table 2.3-3 (continued) l Variable Correlated Question Description
& Case Rance Distribution ,_C_ptrela tion Uith 0.075 Expert Rar.k 1 PedFnCn The depth of concrete eroded (M) in 3 h PedF3G6 0.85 Mean - 0.41 during CCI--Expert Group 6.
Q121 C6 l' 0.075 Expert Rank 1 PedFnCn The depth of concrete eroded (M) in 3 h PedF3G7 during CCI--Expert Group 7. Q121 C7 0.85 Dean - 0.47 0.15 Expert Rank 1 PedFnCn The depth of concrete eroded (M) in 6 h PedF6G1 during CCI--Expert Group 1. Q121 C3 1.26 Mean - 0.55 Rank 1 PedFnCn The depth of concrete eroded (M) in 6 PedF6G2 0.15 Expert 1.26 Mean - 0.49 during CCI--Expert Group 2. Q121 C5 U 0.15 Expert Rank 1 PedFnCn The depth of concrete eroded (M) in 6 h PedF6G3 1.26 Mean - 0.52 during CCI--Expert Group 3. Q121 C4 0.23 Expert Rank 1 PedFnCn The depth of concrete eroded (M) in 6 h PedF6G4 during CCI--Expert Group 4. Q121 C9 1.26 Mean - 0.62 0.28 Expert Rank 1 PedFnCn The depth of concrete eroded (M) in 6 h PedF6G5 1.26 Mean - 0.71 during CCI--Expert Group 5. Q121 C8 0.23 Expert Rank 1 PedFnCn The depth of concrete eroded (M) in 6 h PedF6G6 1.26 Mean - 0.66 during CCI--Expert Group 6. Q121 C6 0.28 -Expert Rank 1 PedFnCn The depth of concrete eroded (M) in 6 h PedF6G7 1.26 Mean - 0.73 during CCI--Expert Group 7. Q121-C7 0.36 Expert Rqnk 1 PedFnCn The depth of concrete eroded .(M) in 10 h PedF10G1 1.41 Mean .0.83 during CCI--Expert Group 1. Q121 C3 0.25 Expert Rank 1 PedFnCn 'The depth of concrete eroded (M) in 10 h PedF10G2 1.41 Mean -.0.74 during CCI--Expert Group 2. Q121 C5
l 4 Table 2.3-7 (e- unued) Variable Question Correlated Correlation With Description
& Case Rance Distribution _
PedF10G3 0.25 Expert Rank 1 PedFnCn The depth of concrete eroded (M) in 10 h Q121 C4 1.41 Mean - 0.79 during CCI--Expert Group 3. PedF10G4 0.30 Expert Rank 1 PedFnCn The depth of concrete eroded (M) in 10 h Q121 C9 1.57 Mean - 0.82 during CCI--Expert Group 4. PedF10G5 0.37 Expert Rank 1 PedFnCn The depth of concrete eroded (M) in 10 h 1.57 Mean - 0.92 during CCI--Expert Group 5. Q171 c8 PedF10G6 0.29 Expert Rank 1 PedFnCn The depth of concrete eroded-(M) in 10 h w Q121 C6 1 37 Mean - 0.83 during CCI--Expert Group 6. PedF10G7 0.37 Expert Rank 1 PedFnCn The depth of concrete eroded (M) in 10 h " Q121 C7 1.57 Mean - 0.92 during CCI--Expert Group 7. Lt-Pres 250 Uniform None The pressure (kPa) in the containment 550 late in the accident due to Q123 C2 noncondensibles. AC-LT-CD 0.00 Internal Rank l AC-ST-n The probability that ac power is Q24 C2 -0.31 Mean .19 AC-LT-n recovered before VB during a long-term SB given that it was not available at CD. AC-ST-CD 0.39 Internal Rank 1 AC-ST-n The probability that ac power is AC-LT-n recovered before VB during a short-term Q24 C3 0.82 Mean .o2 SB given that it was not available et CD.
+
AC-LT-VB 0.00 Internal Rank 1 AC-ST-n The probability that ac power is Q79 C3_ 0.23 Mean .10- AC-LT-n recovered after VB during a long-term SB given that it was not available before VB.
^-' _.__-__mm'
._ ._____-___.m__m.
l Table 2.3-3 (continued) Variable Question Correlated Correla: ion Uith Description l
& Case Rance Distribution AC-ST-VB 0.14 Internal Rank 1 AC-ST-n The probability that ac power is 0.58 Mean .38 AC-LT-n recovered after VB during a short-term Q79 C4 SB given that it was not available before VB.
AC-LT-LT 0.00 Internal Rank 1 AC-ST-n The probability that ac power is 0.19 Mean .09 - AC-LT-n recovered late in the accident during a Q104 C3 long-term SB given that it was not available after VB. AC-ST-LT 0.38 Internal Rank 1 AC-ST-n The probability that ac power is
.77 'LC-LT-n recovered late in the accident during a 9 . Q104 C4 0.87 Mean short-term SB given that it was not y
available after VB. 1 4 4 h
2.4 Description of the Accident Progression Bins As each path through the APET is evaluated, the result of that evaluation is stored by assigning it to an _- Accident Progression Bin. This bin describes the evaluation in enough detail that a. source term (release of radionuclides) can be calculated for it. The accident progression bins are the means by which information is passed from the accident progression analysis to the source term analysis. A bin is. defined by specifying the attribute or vslue for each of thirteen characteristics or quantitics which define a certain feature ' of the evaluation' of the APET. Section 2.4.1 describes the 13 characteristics, and the values that each characteristic can assume. A more detailed description of the binner, discussing each case in turn, is contained in Appendix A.l.3. The binner itself, which is expressed as a computer-input file, is -listed .in Appendix A.l.4 Section 2.4.2 contains a discussion of rebinning, a process that takes placc between evaluating the, APET (in which binning takes place) and the source term e.nalys is . Section 2.4,3 den ribes a reduced set of binning characteristics, which is used in pres?.ncing the results of evaluating the APET. 2 4.1 Description of the Big _Q,baracteristics The binning. scheme for Grand Gulf utilizes 13 characteristics. That is, there are 13 types of information required to define a path through the APET. A bin is defined by specifying a letter for each of tl e 13 characteristics, where each letter for each characteristic has a meaning defined below. For a characteristic, the possible states are termed attributes. The Grand Gulf binning characteristics are: Characteristic Abbreviation Descriotion 1 Aseq Type of Accident Sequence 2 Zroxid- Fraction of zirconium oxidized in-vessel 3 VB Vessel Condition at Vessel Breach 4 DCH SE Fraction of core participating in direct containment heating or steam explosions 5 SPB-L The mode.and timing of suppression pool l bypass 6 CLeak-L The mode and timing of containment failure 7 Sprays Period in which containment sprays operate 8 MCCI Type of CCI 2.55
Characteristic 6hhr.eviation' Description 9 SRVBkr Occurrence of a stuck open SRV tailpipe vacuum breaker t 10 CF BVB Events - causing containment failure before vessel breach , 11 CF-VB Events causing containment failure at vessel breach 12 DF BVB Events causing-- drywell failure before vessel breach 13 DF VB, Events causing =drywell failure at vessel breach Most of this information, organized in - this manner, is needed by GCSOR to calculate the fission product source terms. Characteristics 10 through 13 , are not used by CGSOR, but have been retained because they provide useful information on the types of events that cause containment and drywell -> failures.
- A description of each attribute for each characteristic is presented in table 2,4-1. The remainder'of this section consists,of a brief description
- <
l of each characteristic and an explanation of an example bin. Characteristic 1 addresses the type of accident sequence that has occurred. Six attributes are defined. The attributes are based on the initiating p event and the time at which core damage . occurs . The initiating events ! include station. blackout, loss of PCS transient, and ATWS. For each > initiating event there are two times at which core damage occurs: short-term and long-term. - i-Characteristic 2 addresses the fraction of - in vessel zirconium that oxidized before vessel breach. .There are .- two . peasible values for this characteristic: low and high. The demarcation point between'the two ranges is 21%. Characteristic 3 addresses the RPV pressure before vessel breach and the availability of coolant . inj ection- at vessel breach; there are five possibilities, . including no~ vessel . breach. The RPV can either be at high or low pressure before vessel breach. High pressure is - system pressure (i.e., approximately '1000 psia) and low pressure is less than 200 psia. There are two possibilities for coolant inj ec tion: coolant is- being
. injected into the RPV at or immediately after vessel breach, or coolant is not being injected into the RPV at or immediacely after vessel breach, i
2.56
- -- ~ .. .. - - - - . - . - - -- . . - - - . . . - I l 1 l Characteristic 4 addresses the fraction of core participating in DCil or an l ex vessel steam explosion. There are five attributes associated with this characteristic. There - are two levels for DCil: low (10% of the core) and high (40% of tho' core) . Similarly, there are two levels for steam explosions: low (St of the core) and high (20t ' of the core) . The fif th attribute is for the case when there are no DCil events or ex vessel steam explosions, if a DCli event and a steam cxplosion both occur during an accident tae attribute associated with the DCl! event is assign to this characteristic. The reason for this is that. .aore radionuclides are -- released during a DCil event l than are release from a steam explosion. Characteristic 5 addresses the amount of suppression pool bypass and time that pool bypass occurs. There are eight attributes. The bypass can be either nominal (no change), small (leak), or large (rupture). Three time periods are addressed: early -intermediate, and late. Characteristic 6 addresses the size of hole that results from containment failure and the time period in which the containment failed. The are nine a ttribute s .- The hole size can be either small (leak) or large -(rupture). Three time periods are addressed: early, intermediate , - and latc. Containment venting before vessel breach and af ter vessel breach are also addressed by this characteristic. The last attribute is no containment failure. Characteristic 7_ addresses the period in which containment sprays operate; there are four attributes. Two time periods are addressed: early and intermediate. The intermediate tire period' does not include sprays operation in the late time period. If the sprays come on during the late time period it is because ac powet was previously unavailable and -it was restored during the late ; time period. By' this time the majority of the fission products have already been released and the sprays are no longer-effective in scrubbing the radionuclides.- Sprays that operate only.in the . late period are grouped with rh : m -in which the sprays never operate. The four possibilities are: _no containment :: prays. only early containment sprays, only -intermediate contalament --sprays ,- and ear'y _ and intermediate containment sprays. Characteristic -8 addresses CCI. There are five- attributes, including no CCI releases. The first four ath ibutes are concerned 'with - the amount of water in - the- reactor- pedestal cavity. The cavity can be dry, wet, or flocded. If the core debris is initially coolable but there is not a re;Renishable_ water supply, then delayed CCI occurs. That is, the core debris ~ io ; initially cooled until all the water is -boiled from the cavity.
.After the water is boiled-away,' CCI begins. It is estimated that CCI'will be delayed for 3 h'for this case.
Characteristic 9 addresses the occurrence of:a stuck open SRV tailpipe vacuum breaker. There are two possibilities: the vacuum breaker is stuck open or_the vacuum breaker is closed.. Characteristic 10 - addresses - the events that can cause containment failure before vessel breach. There are eight attributes. The containment failure 2.57 L
+ + -
,-,-,, n.e, - .r,,- w - inn , , , - - - , , ,--m, ,,r- r e-,---e n - r,- 4
_ _ ._7
.[
i can be caused by a s1(v pressurization event, a hydrogen deflagration,lor a^ ! hydrogen detonation. The failure size can be-either a leak or a rupture. In addition, venting can cause a breach in the containment boundary. -The last attribute is no containment failure before vessel breach. Characteristic 11 addresses the events -- that can cause - containment failure at vessel breach. There are eight attributes. Containment failure during, this time period _ can b'e caused by an alpha mode event, a hydrogen deflagration or a hydrogen detonation. The failure size can be either a' leak or a rupture (except for_ the ' alpha - mode event which is _always ! considered as a rupture).
-Characteristic 12 addresses the events that . can cause drywell fsOwe before vessel breach. There are five attributes. Drywell~ failure can be, ,
caused by hydrogen deflagratiens or detonations. The failure size can be either a leak or a rupture. The last attribute _is no drywell . failure-before vessel breach. Characteristic 13 addresses ' the events that can _cause drywell ' failure -.at : the time of vessel breach. There are twelve attributes. Drywell failure can be caused by an alpha modo event, a hydrogen defla5 ration, a hydrogen detonation, or by quasi-static loads accompanying -vessel breach ; The failure size can be either a leak or a rupture.(except for the alpha modo event which is always considered as a rupture). In addition, reactor pedestal failure can in some instances lead _ to _drywell failure (e.g., movement of RPV causes a penetration failure) . Pedestal failure during i this time period are caused by either loads _ accompanying vessel breach or , by dynamic loads associated with ex-vessel steam explosions. Drywell ! failures that are induced by pedestal failures are always assumed to be , ruptures. A typical bin might be ABBDAACCBGEEL which, using the information presented in Table 2,4-1, is: { A Fst SB Accident sequence is a short-term station blackout-B LoZrox A small fraction of the zirconium was oxidized in-vessel. B lop nLPI The RPV was ' at low- pressure before vessel breach . and-there was no injection' to the RPV af ter vessel breach ' D- LoEXSE A small fraction of the core. participated in an ex - vessel steam explosion; there was no-DCH event I. A- SPBEOLO There was no-suppression pool bypass-1- -A CE Lk The1 containment failed early from the development of-a-leak C -LCS Spray operation.was; recovered after vessel breach-C FLDCCI CCI proceeded in a flooded reactor cavity-B cSRVBkr A.SRV tailpipe vacuum breaker did not stick open . . O CL DEF A deflagration caused a. leak in; the - containment before vessel breach E E-Leak- The; containment failed from the ~.developmer.t o f - - leak before vessel breach E nDFa11 The drywell did not fail before vessel breach
- j. L nIDWF The drywell did not fail at the time:of vessel breach 2.58
Table 2.4-1 Description of Accident Progression Bin Charatteristics Attribute Mnemonic Description _ j Characteristic 1: Type of Accident Sequence A Pst-SB Short term station blackout B Slw SB Long term station blackout C Fat T2 Short term loss of PCS transient D Slw T2 Long-term loss of PCS transient E Fat-TC Short term ATWS F Slw TC Long term ATWS Characteristic 2: Fraction of Zirconium Oxidized In-Vessel A liiZrox liigh: Creater than 21 % of the In-Vessel zirconium has been oxidized before vessel breach B LoZrox Low: Less than 21% of the In-Vessel circonium has been oxidized before vessel breach Characteristic 3: Vessel Condition at Vessel Breach - A liiP nLPI RPV is at high prersure and there is j no coolant inj ec tion af ter vessel i breach - - l B lop-nLPI RPV is at low pressure and there is no coolant inj ec tion after vessel breach C 111P-LPI RPV is at high pressure and coolant is being inj ec ted after vessel breach D lop-LPI RPV is at low pressure and coolant is being inj e c te d after vessel breach E nVB There is no vessel breach (i.e., core damage arrest) i i l i l 2,59 1
Table 2.4 1 (continued) 1 Attribute Mnemonic Description l Characteristic 4: Fraction of Core Participating in DCH or Steam Explosions A HiDCil 40% of the core participates in DCil B LoDCll '10% of the core participates in DCll C HiEXSE 40% of the core participates in ex-vessel steam explosions D LoEX3E 10% of the core participates in ex-vessel ~ steam explosions E nDC11- SE There are no DCil or steam explosions events Characteristic 5: Mode and Timing of Suppression Pool Bypass A SPBE0LO Nominal leakage B SPBE0I3 Early nomincl, intermediate rupture C SPBE0L2 Early nominal, late leakage D SPBE0L3 Early nominal, late rupture E SPBE2L2 Early leakage F SPBE213 Early leakage, intermediate rupture G SPBE2L3 Early leakage, late' rupture 11 SPBEJL3 Early rupture Characteristic 6: Mode and Timing of Containment Failure A CE-Lk Leak before vessel breach (VB) B CE Rrt Rupture before vessel breach C CE-VENT Containment vented before vessel breach D CVE LK Leak at vessel breach E CVB Rpt Rupture at vessel breach F CL-Lk Late Leak 2.60
Table 2.4 1 (continued) Attriburg __ Mnemonic Desqr 4on Characteristic 6 (continued) C CL Rpt Late Rupture ! 11 CL VENT Containment vented late I CnFall No containment failure Characteristic 7: Period in wh!.ch Containment Sprays Operate A noCS The containment sprays do no operate during the accident B ECSnoL The sprays only operate before vessel breach (VB) C LCS The sprays only operate after vessel breach D ECS The sprays both before vessel breach and cfter vessel breach Characteristic 8: Type of Core-Concrete Interactions (CCI) A DryCCI CCI occurs in a dry reactor pedestal - cavity B WetCCI CCI occure in wet cavity C FLDCCI CCI cccut: in a flooded cavity-D DlyCCI CCI releases are delayed E noCCI There are no CGI releases a Characteristic 9: Occurrence of a Stuck Open SRV Tailpipe Vacuum Breaker A oSRVBkr An SRV tailpipe vacuum breaker sticks open during core damage B cSRVBkr There are no stuck open tailpipe vacuum breakers ; l I l l 1 2,61
---~~n..
Table 2.4 1 (continued) Attribute Mnemonic Description I Characteristic 10: Events Causing Containment Failure Before Vessel Breach A E-VENT The containment was vented before core degradation (considered as a large hole). B CR-SP The containment failed from either an isolation failure or f;om a slow pressurization event (i.e., steam buildup) which led i.o the develop-ment of a large '.iol e or rupture; cominal hole size is 7 ft 2, C CR DET The containment failed from a hydrogen detonation which led to the development of a large- hole or rupture; nominal hole size is 7 ft 2, D CR-DEF The containment f ailed - from a hydrogen deflagration which led to the development of a large hole or rupture; nominal hole size is 7 ft 2, E CL-SP The containment failed from either an isolation failure or from a slow pressurization event- (i.e., steam buildup) which led to the l development of a small hole or leak; nominal hole size is 0.1 ft 2, F CL-DET The containment failed from a hydrogen detonation which led to the development of a small hole or leak; nominal hole size is 0.1 ftz, C CL-DEF The containment failed from - a l hydrogen deflagration which led to the development of a small hole or leak; nominal hole size is 0.1 ft 2, H nCFail The containment did not fail before vessel breach. 2.62
I Table 2.4-1 (continued) Attribute Mnemonic Descrit tion Characteristic 11: Events Causing Containment Failure at Vessel Breach
. A Erupt The . containment. - f ailed by - _the development'of-a large hole before_-
- vessel breach.-
B ALFilA The containment failed from an alpha j mode = event - which led to- the
-development.of a - large ' hole or rupture; nominal; hole size is 7.ftz,,
C IR Det The containment- failed from a hydrogen detonation which led to the-
- development ;of a: large / hole. or rupture; nominal hole size-is 7 fe ,z D IR Def The containment . failed from -a hydrogen deflagration - which led - to the development of'a large hole or-rupture;_nominaluhole size is 7 fta, E E Leak The containment _ f aile'd _by the-development of -a small hole before vessel breach..
F - IL Det The containment-- f ailed -from a hydrogen detonation which led to the~ ' development of a'small hole or leak;
-nominal = hole size-is 0,1 ft ,- 2 C IL-Def The :' containment - f ailedL from - a hydrogen deflagration which led . to -
the ; development -of -_a small hole o r .'
-leak; nominal _ hole size is 0.1~ft 2, 11 nICFail The containment did not fail before or at the time of vessel breach.
Characteristic 12: Events Causing' Drywell' Failure Before Vessel' Breach ., A DR Det .The drywell- failed from a hydrogen l detonation which: led to. _ the I: ' development of a;large hole: .or
- rupture; nominal - hole size: is '.1. 0 -
1-f ta , i 2,63 j -- _.n ._...._,.._.._.u._., _-_ - .. u . . . . . _ . , _ - _ _ _ - _ , _ .
l Table-2,4-1 (continued) Attribute Mnemonic Description B DR Def The drywell failed from a hydrogen l deflagration which led to the development of a large hole or rupture; nominal hole size is 1,0 ft. 2 C DL-Det The drywell failed from a hydrogen detonation which led to the development of a small hole or leak; nominal hole size is 0.1 ft 2, D DL-Def The drywell isiled from a hydrogen deflagration which icd to the development of s small hole or leak; nominal hole sit e is 0,1 fe z, E n nFail The drywell did not fail before vessel breach, Characteristic 13: Events Causing Dryvell Failure at Vessel Breach A EDWRpt The dryvell failed by the development of a large hole before vessel breach, - B ALPHA The drywell failed from an alpha modo event which led to the development of a large hole or rupture; nominal hole size 'iis 1 ~, 0 l f ta , l C R-DWOP The drywell failed f rom - loads accompanying vessel breach which led to the development of a large hole l or rupture; nominal hole size is 1.0 f ta , D R-PedP The drywell failure was induced'by ) the reactor pedestal failure which ; led to the development of a large i hole or- rupture;- nominal hole size ) is 1,0 f t* , The pedestal failed l from loads accompanying vessel l breach. l 2.64 l
Table 2.4 1-(continued) Attribute Mnemonic Description Characteristic 13'(Continued) E R PedSE The drywell failure was induced by the - reactor _ pedestal- failure - which-led. to _ the - development of a large holo f or rupture; nominal hole size is 1.0 ftz. The pedestal failed from dynamic: loads . associated - with
- an .ex vessel steam explosion. in the reactor cavity.
F DR-Det The drywell failed from - a= hydrogen de tonation which led: to .the development ;of a large hole or-rupture; - nominal hole size -is 1.0
' f tz ,
G DR-Def The drywell failed from . a hydrogen deflagration which : led to - the development of a -large . hole' or
- rupture; nominal hole sizob is 1.0 fe z, 11 EDWLk The drywell failed by the ;
development ' of-- a small hole - before vessel breach, I LDWOP The drywell f ailed :from loads accompanying vessel breach which led to 1the development of .a smal'1 hole - -
- or leak; nomina 11 hole size is , 0 1 --
' fe z, l
J DL-Det Tho' drywellL failed from _ a hydrogen de tonation' which led -to ' the - development.of a smallihole or leak; nominal hole size is'O 1 fe z, K DL Def The drywell" failed from_- a- hydrogen deflagration 4 ;which : led. to the
- development of 'a . small hole 'or leak;.
nominal hole size-is 0.1 ftz, L nIDWF The drywel1[did -not fail -before or- y
.at the time of vessel breach.
l 2.65
'l
, _ _ ,, ,_ , , _ . . _ _ _ _ _ _l
2.4.2 Rebinninn 1 ! The binning scheme used for evaluating the APET does not exactly match the l input information required by CGSOR. The additional information in the l initial binning is kept because it provides a better record of the outcomes l of the APET evaluation. Therefore, there is a step between the evaluation of the APET and the evaluation of GGSOR known as "rebinning". In the rebinning, a few attributes in some characteristics are combined because there are no significant differences between them for calculating the fission product releases. In the rebinning for Grand Gulf, there are no changes for characteristics 1 through 9. That is, for these nine characteristics, the information produced by the APET is exactly that used by GGSOR. The last ic.ur characteristics, 10, 11, 12, and 13, provide additional information on the-types of events that caused containment and drywell failure. This additional information is not used by GGSOR and, therefore, has been deleted in the rebinning process. Thus, the rebinning process converts the example bin, ABBDAACCBGEEL, to: A Pst SB Accident sequence is a short-term station blackout B LoZr0x A small fraction of the zirconium was oxidized in vessel-B lop nLPI The RPV was at low pressure before vessel breach and there was no injection to the RPV after vessel breach D LoEXSE A small fraction of the core participated in an ex vessel steam explosion; there was no DCH event A SPBE0LO There was no suppression pool bypass A CE Lk The containment - failed early from the development of a leak - C LCS Spray operation was recovered after vessel breach C FLDCCI CCI proceeded in a flooded reactor cavity B cSRVBkr A SRV tailpipe vacuum breaker did not stick open i 2.4.3 Summary Bins for Presentation For presentation purposes in NUREG 1150,4 a set of " summary" bins has been adopted. Instead of the 13 characteristics and thousands of possible bins that describe the evaluation of the APET in detail, the summary bins place the outcomes of the evaluation of the APET into a few, very general groups. . The eight summary bins for Grand Gulf are: 2.66
l vessel breach,- Early CF, Early SP' Bypass, CS Not Available:- ] VB, Early CF, Early SP Bypass, CS Available VB, Early CF, Late-SP Bypass VB, Early CF, No SP Bypass VB, Late CF VB, Vent VB, No CF No VB f In the summary binning scheme there are- essentially four characteristics: vessel breach, containment failure, . suppression pool bypass, and containment spray operation. Each of these characteristics and their-
- associated attributes are defined in Table 2.4 2.
The summary bins are listed roughly.in decreasing order of the severity of: . 1 the resulting source term. The eight summary bins may now be defined as follows: vessel breach, Early CF, Early SP Bypass,.'CS Not Available -i Vessel breach occurs' and both the containment and the drywell- have failed either before or at the time of vessel breach. .The containment sprays do not operate before or at the time of vessel breach. vessel breach,~Early CF, Early SP Bypass, CS'Available Vessel breach occurs and both the containment and the drywell failL either before or at the time of vessel breach. In -this bin, however, the containment sprays do operate before or at the time of vessel breach. vessel breach, Early CF,-Late SP Bypass Vessel breach occurs and the containment fails cither ~before or at the - time of vessel breach. The drywell does not' fail until the late time ~ period and, thus, both the in-vessel releases and the releases- i associated with vessel breach are scrubbed by the s~uppression pool. Therefore, the availability of containment sprays during the time period that .the suppression pool is not bypasseL is t.nt very important and, thus, the CS characteristic has been dropped. vessel breach, Early CF, No SP Bypass = 4 Vessel breach occurs and .the containment fails- either before or ne cne time of vessel breach,- The 'drywell does not fail and, therefore, all' of the radionuclide releases pass through the = suppression pool. Because the pool has not been bypassed, the - availability of the sprays . is not very important and, thus, the CS - characteristic has been dropped. 2.67
-ra-,,.e-n e er, -n. -,- e-,- , , , , , , , -a-,,e,i--,Nm-,., , ew'-,an, ,.--ne w.m..,n. ,--,e. oNa , - - . ~ ,
1 l vessel breach, Late CF Vessel breach occurs, however, the containment does not fail until the late time period. If the containment did not fail early it is unlikely that the drywell will fail early. Thus, the suppression pool bypass characteristic and tihe containment spray characteristic have been ' l dropped. vessel breach, Vent This summary bin represents the case in which vessel breach occurs and - the containment was vented during any of the time periods in the accident. VB, No CF Vessel breach occurs but there is no containment failure and any releases associated with normal containment leakage are minor. Thus, the suppression pool bypass characteristic and the containment spray
- characteristic have been dropped. The risk associated with- this bin will be negli ftble.
l No vessel breach l l Vessel breach is averted. Thus, there are no releases associated with l vessel breach and there are no CCI releases, It must be remembered, however, that the containment can fail even if vessel breach is ( averted. Thus, the potential exists for some of the in-vessel releases to be released to the environment. It follows that there will be some risk associated with this bin. i Table 2,4 Description of Summary Accident Progression Bin Characteristics 9
- Attribute Description Characteristic 1: Vessel Breach sVB)
VB Vessel breach occurs No VB Vessel breach does not occur. Characteristic 2: Containment Failure Time (CF) Early CF The containment fails either before or at the time of vessel breach from the development of a leak or a rupture. Late CF The containment fails during the late time period l from the development of either a leak or a rupture. 2.68 l l
Table 2.4 2 (Continued) Attribute Description Characteristic 2 (Continued) Vent The containment is vented during any of the time periods. No CF The containment does not fail. Characteristic 3: Suppression Pool (SP) Bypass Early SP Bypass The drywell fails either before or at the time of vessel breach from the development of a leak or.a rupture. Late SP Bypass The.drywell fails during the late time period from the development of either a leak or a rupture. No SP Bypass The drywell does not fail. Characteristic 4: Containment Spray (CS) Operation CS Not Avail. The containment sprays do not operate during the early or intermediate time periods. CS Available The containment sprays operate during either the early time period, the intermediate time period, or during both time periods. l i i l l l l 2.69 l
l 2.5 Ittsylts of the Accident Progression Analysis r This section presents the results of evaluating the APET. As evaluating the APET produces a number of accident progression bins (APBs), the discussion is primarily in terms of APBs. Some intermediate results are also presented. Section 2.5.1 presents the results for the internal initiators. External events (fire and seismic) were not considered in the Grand Gulf analysis. The tables in this section present only a very small portion of the output obtained by evaluating the APETs. Complete listings giving average bin conditional probabilities for each PDS group, and listings giving the bin ! probabilities for each PDS group for each observation are available on computer media by request. 2.5.1 Results for Internal Initiators 2.5.1.1 Results for PDS 1: Short-Term SBO. This PDS involves station blackout scenarios where loss of offsite power (LOSP) is recoverable. Coolant injection is lost early so that core damage occurs in the short term and with - the vessel at high pressure. If offsite power is restored, then the following functions are available: either high pressure injection or low pressure injection or both, heat removal via the sprays, and the miscellaneous systems-venting, standby gas treatment (SBGT), containment isolation (CI), hydrogen ignition (H 1)2 . In addition, the firewater system is available. This PDS also includes cut sets with either one or two stuck open SRVs. Table 2.5-1 lists the five most probable APBs for this PDS, the five most probable APBs that have vessel breach, and the five - most probable APBs that i have containment failure (CF). The " Order" column gives the order of the bin when ranked by conditional _ probability. The " Prob." column lists - mean- APB probabilities conditional on the occurrence of the PDS group. That is, this table shows the results averaged over the 250 observations that form the sample. If Bin A ocentred with a probability of 0.004 for each observation, its probability would be 0.004 in Table 2.5-1. If - Bin B occurred with a probability of 1.00 for one observation and did not occur in the other 249 observations, its probability would also - be ' O.004 The remaining eight columns explain 8 of the 9 characteristics in the APB indicator for the rebinned results. The first characteristic, the accident sequence indicator (ASeg), has been omitted since this is defined by the PDS. The abbreviations for each APB characteristic are explained in Section-2,4 above, The first part of Table 2,5-1 shows the first five bins when they are-ranked in order by probability. Evaluation of the APET produced 3837 source term bins for this PDS. To capture 95% of the probability, 1812 bins are required. The five most probable bins capture only 13% of the probability. 2.70
t Table 2.5-1 Results of the Accident Progression Analysis for Grand Gulf Internal Initiators: PDS 1; Short-Term SBO Order Bin Prob . ** ZrOxid VB _ DCH-SE SPB CF Sprays ECE SRVBkr Five Most Probable Bins
- 1 ABBDDGCCB 0.032 LoZr0x LoPnLPI LoExSE SPBEOL3 CL-Rpt LCS FLDCCI cSRVBkr 2 ABEEAICEB 0.029 IoZrOx nVB nDCH-SE SPBEOLO CnFail LCS noCCI cSRVBkr '
3 ABEEAGCEB 0.027 .LoZrOx nVB nDCH-SE SPBEOLO CL-Rpt LCS noCCI cSRVBkr 4 ABEEAFCEB 0.026 LoZrox nVB nDCH-SE' SPBEOM CL-Lk LCS noCCI cSRVBkr ABEEAHCEB 0.019 loZrOx nVB nDCH-SE SPBE01D CL-Vent LCS noCCI cSRVBkr 5 Five'Most Probable Bins that have VB* 1 ABBDDCCCB 0.032 LoZrOx LoPnLPI LoExSE SPBEOL3 CL-Rpt LCS FLDCCI cSRVBkr 9 ABDDDCCCB 0.012 loZrOx- lop-LPI LoExSE SPBEOL3 CL-Rpt LCS FLCCI cSRVBkr , 12 ABBDDCACB 0.010 LoZrOx LoPnLPI LoExSE SPBEOL3 CL-Rpt noCS FLCCI cSRVBkr 13 ABBDDGCCA 0.010 LoZrox LoPnLPI LoExSE SPBEOL3 CL-Rpt LCS FLCCI oSRVBkr ABBDAICEB 0.008 IoZrOx LoPnLPI LoExSE SPBEOLO CnFall IIS noCCI cSRVBkr 14 Five Most Probable Bins that have Early CF* AAEEABAEB 0.013 HiZrOx' nVB nDCH-SE SPBE0LO CE-Rpt noCS noCCI cSRVBkr 7 0.011 HiZrOx nVB nDCH-SE SPBE2L2 CE-Rpt noCS noCCI cSRVBkr 10 AAEEEBAEB 0.008 HiZr0x nVB' nDCH-SE SPBEOLO CE-Lk LCS noCCI cSRVBkr 15 AAEEAACEB 18 AAEEHBAEB 0.007 HiZrOx nVB nDCH-SE SPBE3L3 CE-Rpt noCS noCCI cSRVBkr 0.004 HiZrOx LoPnLPI LoEXSE SPBEOLO CE-Rpt noCS FLDCCI cSRVBkr 31 AABDABACB
~
A listing of all bins, and a lis$ ting by observation are available on computer media. Hean probability conditional on the occurrence of the PDS.
L 1 1 Four of the five most probable bins have no vessel breach and in all of j these five bins the containment either fails late or does not fail. The ; five most probable bins with vessel breach all occur with the RPV at low ( pressure and again, the containment either fails late or doec not fail, j The last part of Tabic 2.51 shows the five most probable APBs with early CF. (Early CF means CF before, at, or immediately af ter vessel breach.) In four of these five bins vessel breach does not occur. In the one bin that vessel breach occurs the drywell does not fail and, therefore, all of the releases pass through the suppression pool. For this PDS the probability of recoverin6 offsite electrical power before vessel breach (i.e., in the early time period) is 0.62. The probability of recovering coolant injection before vessel breach _ is 0.87 which includes the recovery of injection systems when ac power is recovered and the use 'of the firewater system for those accidents in which ac power is not recovered. If coolant injection is restored to the RPV, it is possible to arrest the core damage process and avoid vessel breach. For this PDS the probability that vessel breach is averted is 0.32. The probability that the containment fails early, with early defined as before or around the time of vessel breach, is 0.36. 2.5.1.2 Results for PDS 2: Short-Term SBS. PDS 2 is the same as PDS 1 except that heat removal via the sprays is not available with the recovery of offsite power. Table 2.5-2 lists the five most probable APBs for this PDS, the five most probable APBs that have vessel breach, and the five most probable APBs that have early contaitunent failure (CF). Evaluation of the APET produced 2571 source term bins for this PDS. To capture 95% of the probability,1066 bins are required. The five most probable bins capture only 16% of the probability. In four of the five most probable bins, vessel breach is averted. In the bin that has vessel breach, the containment fails in the late tiro period. In all of the five most probable bins that have vessel breach the containment either fails in the late time period or does not fail. Similarly, in all of the five most probable bins that have early ' containment failure vessel breacl. is averted. Only two of-the fim most probable bins that have early containment failure have coincident ';ivell failure. Furthermore, in these two bins vessel breach is averted ar , there are no stuck open SRV tailpipe vacuum breakers. Thus, there are oaly in-vessel releases and these pass through the suppression pool. The probability that offsite electrical power is - recovered before -vesse_1 breach is 0.62. For this PDS the probability that coolant injection is recovered and vessel breach is averted is 0.32. The probability that the containment will fail early is 0.36, 2.5.1.3 Results for PDS 3: Short-Term SBO. PDS 3 is the same as PDS 1 cxcept that heat removal via the sprays is not available with the recovery of offsite power and the only injection system that is available with the reco"ery of offsite power is the condensate system. 2.72
l
.. L i
Table 2.5-2 ; Results of the Accident Progression Analysis for Grand Gulf ! Internal Initiators: PDS 2; Short-Term SB0 i i t i Order bin Prob . ** ZrOxid VB __ DCH-SE SPB CF Sorays MCCI SRVBkr l < l Five Most Probable Bins * ; i 1 ABBDDGACB 0.043 LoZrOx loPnLPI LoEXSE SPBEOL3 CL-Rpt noCS FIDCCI cSRVBkr I 2 ABEEAIAEB 0.035 loZrOx nVB nDCH-SE SPBEOLO CnFail noCS noCCI cSRVBkr i 3 ABEEAGAEB 0.034 ToZrOx nVB nDCH-SE SPBEOLO CL-Rpt noCS noCCI cSRVBkr j i ; 4 ABEEAFAEB 0.032 seZrOx nVB nDCH-SE SPBEOLO CL-Lk' noCS noCCI cSRVBkr 5 ABEEAHAEB 0.021 LoZrOx nVB nDCH-SE SPBEOLO CL-VENT noCS noCCI cSRVBkr '{ l Five Most Probable Bins that have VB* i 4 . . cSRVBkr
~
w -I ~ABBDDGACB 0.043 LoZrOx LoPnLPI IoEXSE- SPBEOL3 CL-Rpt noCS FLDCCI ! 0.015 IoZrOx IoP-LPI LoEISE SPBEOL3 CL-Rpt. noCS FLDCCI cSRVBkr 'l y 7 lABDDDGACB 0.012 IoZrOx IoPnLPI- LoEKSE SPBEOL3 CL-Rpt noCS FLDCCI .oSRVBkr. .. 10 'ABBDDGACA 13 ABBDAIAEB 0.010 IoZrOx LoPnLPI. -IoEXSE SPBEOLO CnFail noCS. noCCI cSRVBkr-ABBDAFAEB 0.010 LoZrOx IoPnLPI- LoEXSE SPBE01D CL-Lk noCS noCCI cSRVBkr i 14 , Five Most Probable Bins.that'have Early CF*
.AAEEABAEB 0.014 .HiZrOx' nVB nDCH-SE SPBEOLO CE-Rpt noCS noCCI -cSRVBkr- l
, 8.l
'O.012 HiZror nVB nDCH-SE SPBE2L2 .CE-Rpt noCS noCCI cSRVBkr Ll
- 12. 'AAEEEBAEB
'O.009: HiZrOx- nVB nDCH-SE SPBE3L3 CE-Rpt: noCS noCCI cSRVBkr 17- AAEEHBAEB-
.18- AAEEAAAEB 0.009- HiZrOx- nVB' nDCH-SE SPBEOLO CE-Lk noCS noCCI cSRVBkr !
.0.005 LoZrOx nVB nDCH-SE SPBEOLO CE-Lk noCS noCCI cSEVBkr -i 30 ..ABEEAAAEB E
l ,
- i l --
LA listing of all bins, and.a listing by observation are available on computer media. l' Mean probability conditional on 'the occurrence of the PDS. ; i , 4
.. - --- -. _ . - - . -~ . ~ ~ - _ .
Table 2.5 3 lists the five most probable APBs for this PDS, the five most probable APBs that have vessel breach, and the five most probable APBs that have early containment failure (CF). Evaluation of the APET produced 2669 source term bins for this PDS. To capture 95% of the probability, 1216 bins are required. The five most probable bins capture only 12% of tha probability. In three of the five most probable bins vessel breach is averted and in the ; two bins that have vessel breach the contaitunent fails in the late time i period. In four of the five most probable bins that have vessel breach the containment either fails in the late tine period or does not fail. Only two of the five most probable bins that have early containment failure have coincident drywell failure. Furthermore, in these two bins vessel breach is averted and there are no stuck open SRV tailpipe vacuum breakers. Thus, there are only in vessel releases and these pass through the suppression pool. The probability that offsite electrical power is recovered before vessel breach is 0.62. For this PDS the probability that coolant inj ection is recovered and vessel breach is averted is 0.21. The probability that the containment fails early is 0.44. The early containment failure probability is lower for PDS 1 then it is for this PDS because PDS 1 has a higher probability that vessel breach will be averted. 2.5.1.4 Results for PDS 4 Lont-Term SBO. This PDS involves station blackout scenarios where LDSP is recoverable. Coolant injection is lost late such that core damage occurs in the long term and with the vessel at low pressure. If offsite power is restored, then the following functions are available: either high pressure injection or low pressure injection or both, heat removal via the sprays, and the miscellaneous systems- venting, SBCT, CI, H 21. In addition, the firewater system is recoverable. Table 2.5-4 lists the 10 most probable APBs for this PDS and the five most probable APBs that have early containment failure and early suppression pool bypass. Evaluation of the APET produced 2353 source term bins- for - this PDS. To capture 95% of the probability, 842 bins are required. The ten most probable bins capture 23% of the probability. In all of the ten most probable bins vessel breach occurs, the RPV is at low pressure, and an ex vessel steam explosion, which involves a small amount of the core, occurs at vessel breach. Containment' sprays are not available in any of the ten most probable bins. In all of the five most probabic bins that have early containment failure and early suppression pool bypass vessel breach occurs with the RPV at low pressure followed by an ex-vessel steam explosion. There are 'no stuck open tailpipe vacuum l breakers in these five bins so all of the in-vessel releases pass through ( the suppression pool. However, because there is early drywell failure,.the , ex-vessel releases will bypass the suppression pool. Although sprays are ! not available in these five bins, CCI either. proceeds in a flooded cavity (3 bins) and, therefore, the CCI releases are scrubbed, or the core debris is cooled and there are no CCI releases (2 bins). 2.74 I l l
Table 2.5-3 Results of the Accident Progression Analysis for Crand Gulf Internal Initiators: PDS 3: Short-Term SBO VB SPB CF Sorays MCCI SRVBkr Order Bin hob . ** ZrOxid _ DCH-SE Five Most Probable Bins
- IAPnLPI LoEXSE SPBE0L3 CL-Rpt noCS FLDCCI cSRVBkr 1 ABBDDGACB 0.041 LoZrOx nDCH-SE SPBEOLO CL-Rpt noCS noCCI cSRVBkr 2 ABEEAGAEB 0.024 LoZrOx nVB nDCH-SE SPBEOLO CnFail noCS noCCI cSRVBkr 3 ABEEAIAEB 0.022 LoZrox nVB nVB nDCH-SE SPBEOLO CL-Lk noCC noCCI cSRVBkr 4 ABEEAFAEB 0.020 LoZrOx lop-LPI LoEKSE SPBEOL3 CL-Rpt noCS FLDCCI cSRVBkr 5 'ABDDDGACB 0.014 LoZrOx Five Most Probable Bins that have VB*
LoPnLPI leEXSE SPBEOL3 CL-Rpt noCS FLDCCI cSRVBkr ABBDDGACB 0.041 LoZrOx
-[ l 0.014 LoZrOx leP-LPI LoEXSE SPBEOL3 CL-Rpt noCS FLDCCI cSRVBkr v' 5 ABDDDGACB cSRVBkr 0.013 LoZrOx HiPnLPI LoDCH SPBEOLO CVB-Rpt noCS noCCI.
6 ABABAEAEB FLDCCI oSRVBkr 0.013 LoZrox LoPnLP1 .IeEXSE SPBEOL3 CL-Rpt noCS 8 ABBDDGACA cSRVBkr IAPnLPI LoEXSE SPBEOLO CnFail noCS noCCI 10 ABBDAIAEB 0.010 IeZrOx Five Most Probable Bins that have Early CF* HiPnLPI LoDCH SPBEOLO CVB-Rpt noCS noCCI cSEVBkr 6 ABABAEAEB 0.013' LoZrOx nDCH-SE SPBEOLO CE-Rpt noCS noCCI cSPVBkr 16 AAEEABAEB 0.007 HiZrOx ,nVB nDCH-SE SPBE2L2 CE-Rpt noCS noCCI cSRVBkr 18 AAEEEBAEB 0.007 HiZrOx nVB HiPnLPI LoDCH SPBEOL3 CVB-Rpt 'noCS noCCI cSRVBkr
.20 ABABBEAEB 0.007 LoZrOx
'nVB. nDCH-SE SPBE3L3 CE-Rpt noCS noCCI cSRVBkr 22 AAEEHBAEB 0.006 HiZrOx A listing of all bins, and a lit)ing by observation are available on computer media.
Mean probability conditional on the occurrence of the PDS.
Table 2.5-4 Results of the Accident Progression Analysis for Crand Culf ; Internal' Initiators: PDS 4: Long-Term SBO j l
' Order- -Bin ' Prob . ** ZrOxid VB ._ DCH-SE SPB CF Soravs MCCI SRVBkr j Ten Most Probable Bins
- h r
LoPnITI CL-Rpt FLDCCI cSRVBkr '. 1 BABDAGACB 0.032 HiZrOx LoEXSE SPBEOLO noCS i 2 0.031 HiZrox LoPnLPI LoEXSE. SPBEOLO CVB-Rpt noCS FLDCCI cSRVBkr I BABDAEACB cSRVBkr' I' 3 BABDHBACB .0.026 HiZrOx 'LoPnLPI LoEXSE SPBE3L3 CE-Rpt noCS FLDCCI .f 4 BBBDAGACB 0.026 LoZrOx loPnLPI LIoEXSE SPBEOLO CL-Rpt noCS FLDCCI cSRVBkr 5 BABDAEAEB 0.014 -HiZrOx LoPnLPI LoEXSE SPEEOLO CVB-Rpt 'noCS noCCI- cSRVBkr 6 BABDBEACB 0.020 HiZrOx LoPnLPI leEXSE SPBE013- CVB-Rpt- noCS FLDCCI cSRVBkr IoPnLPI loEXSE SPBEOLO CL-Rpt- noCS noCCI cSRVBkr 1
-7 BABDAGAEB 0.020 H1ZrOx 8 BABDHBAEB 0.018 H1Zrox LcPnLPI IoEXSE SPBE3L3 CE-Rpt noCS noCCI cSRVBkr :
.4 '9 ,BBBDAEACB. 0.016 LoZrOx LoPnLPI LoEXSE SPBEOLO CVB-Rpt noCS FLDCCI cSRVBkr' l
%- 10 BBBDAGAEB 'O.015 LoZrox' LoPnLPI IoEXSE SPBEOLO CL-Rpt noCS noCCI cSRVBkr- l
'l Five-Most Probable Bins that have Early CF and P.arly Suppression Pool Bypass * ;
i 3 BABDHBACB 0.026' ~ H1ZrOx IoPnLPI IoEXSE SPBE3L3 CE-Rpt noCS FLDCCI cSRVBkr 6 BABDBEACB :0.020- HiZrox IoPnLPI IoEXSE SPBE013 CVB-Rpt noCS FLDCCI 'cSRVBkr .
- 3- BABDHBAEB' 0.018~ HiZrOx 'LoPnLPI. LoEXSE SPBE3L3 CE-Rpt noCS noCCI cSRVBkr *
{ 11- BABDBEAEB 0.013. HiZrox -LoPnLPI LoEKSE SPBE013 CVB-Rpt noCS noCCI cSRVBkr. l 12 'BBBDBEACB 'O.012 IoZrOx. LoPnLPI . IoEXSE - SPBE0I3 .CVB-Rpt noCS FIDCCI ,cSRVBkr i l . .. q , g.
)
4 A' listing of .all bins, and a listing by observat-ion are available on' computer media. l Mean probabilityiconditional on the occurrence of the PDS. .; l r ' i . e ; . Li
Because this is a slow SB0 (i.e.,-core damage occurs a 12 h) this PDS has a much lower probability of recovering offsite power than did the fast SB0 in which core damage occurs in approximately 1 h. The probability that offsite electrical power is recovered before vessel breach is 0.19. For ~ this PDS the probability that coolant injection is recovered and vecsel breach is averted is only 0.05. The probability that the containment fails early is 0.65.
'2.5,1.5 Results for PDS St Long-Term SBO. PDS 5 is the same as N9 4-except that heat removal via the sprays is not available with the recovery of offsite power, lloweve r , because there is a low probability of recovering offsite power in this PDS this difference is not very important. ,
Table 2.5-5 lists the 10 most probable APBs for this PDS and the five most_ ' probable APBs that have early contairment failure and early supprension . pool bypass. Evaluation of the APET produced 1468 source term bins for this PDS. To capture 95% of the probability, 482 bins are required. The 10 most probable bins capture 26% of the probability. In all of the 10 most probable bins in which vessel breach occurs, the RPV is at low pressure, and _ an ex-vessel steam explosion, _ which involves a small amount of the core, occurs at vessel breach. Containment sprays are not available in any of the 10 most probable bins. In all of the five most-probable bins that have early containment failure and early suppression pool bypass vessel breach occurs with the RPV at low pressure followed.by an ex vessel steam explosion. There are no stuck open tailpipe - vacuum , breakers in these five bins so all of the in avessel releases pass through , the suppression pool, llowever, because there is early drywell failure, the I ex vessel releases will bypass the suppression pool. Although sprays are not available in these_five bins, CCI either proceeds in a flooded cavity (three bins) and, therefore, - the CCI releases are scrubbed, or the core debris is cooled and there are no CCI releases-(two bins). The probability that offsite electrical power is - recovered before vessel breach is 0.19. For this PDS - the probabili.ty that - coolant- injection is - l recovered and vessel breach is averted is only 0.05. The probability that the containment fails early is 0.64. 2.5.1.6 Results for DS 6: Lone-Term SBO. PDS 6 is the same as'PDS 4 except that heat removal via the sprays is not available with the recovery of. offsite power and the only injection system that is recoverable is the , firewater system. llowever, because the operators did not use the firewater l system during the many hours before core damage it is assumed that there is - a negligible probability that they _will use this system duririg core - damage. Thus, there is no coolant injection to the RPV. Table 2.5-6 lists the 10 most probable APBs for this PDS- and the five most l probable APBs that have early containment failure and early suppression - l- pool bypass. Evaluation of the APET produced 1127 source term bins - for l- this PDS. To capture 95% of the probability, 356 bins- are required. The ! 10 most probable bins capture 31% of the probability. ; 2.77
.l
Table 2.5-5 Results of the Accident Progression Analysis for Crand Culf Internal Initiators: PDS S: Long-Term SBO VB SPB CF Soravs MCCI SRVBkr Order Bin Prob . ** Zr0xid __ DCH-SE Ten Most Probable Bins
- LoPnLPI LoEXSE SPBEOLO cL-Rpt noCS FIl)CCI cSRVBkr 1 BABDAGACB 0.036 HiZr0x 0.034 HiZrOx IoPnLPI IoEXSE SPBEOLO CVB-Rpt noCS FLDCCI cSRVBkr 2 BABDAEACB HiZrOx LoPnLPI ,LoEXSE SPBE3L3 CE-Rpt. noCS FLDCCI cSRVBkr 3 BABDHBACB 0.030 LoZrOx LoPnLPI loEXSE SPBE0LO CL-Rpt noCS FLDCCI cSRVBkr 4 BBBDAGACB 0.029
-LoZrOx LoPnLPI LoEXSE SPBEOLO CVB-Rpt noci FLDCCI cSRVBkr 5 BBBDAEACB 0.027 HiZrOx LoPnLPI LoEXSE SPBEOLO CVB-Rpt noCS noCCI cSRVBkr 6 BABDAEAEB 0.024 LoPnLPI leEXSE SPBE01D CL-Rpt noCS noCCI cSRVBkr 7 BABDAGAEB 0.022 HiZrox LoPnLPI LoEXSE SPBE013 CVB-Rpt noCS FLDCCI cSRVBkr 8 BABDBEACB 0.022 HiZrOx y
HiZrOx LoPnLPI LoEXSE SPBE3L3 CE-Rpt noCS noCCI cSRVBkr 9 BABDHBAEB 0.020 y LoZrOx LoPnLPI LoEXSF SPBEOLO CL-Rpt noCS noCCI cSRVBkr m 10 BBBDAGAEB 0.020 Five Most Probable Bins that have Early CF and Early Suppression Pool Bypass
- LoPnLPI LoEXSE SPBE3L3 CE-Rpt noCS FLDCCI cSRVBkr 3 BABDHBACB 'O.030 HiZrox HiZrOx LoPnLPI IeEXSE SPBE0I3 CVB-Rpt noCS FLDCCI cSRVBkr 8 BABDBEACB 0.022 LoPnLPI LoEXSE SPBE3L3 CE-Rpt noCS noCCI cSRVBkr 9 BABDHBAEB 0.020 HiZr0x' LoZrOx LoPnLPI IAEXSE SPBE013~ CVB-Rpt noCS FLDCCI cSRVBkr 12 BBBDBEACB 0.015 HiZrOx LoPnLPI LoEXSE SPBE0I3 CVB-Rpt noCS noCCI cSRVBkr 13 BABDBEACB 0.015 A listing of all bins, and a listing by observation are available on computer media.
Mean probability conditional on the occurrence of the PDS.
9 4 i Table 2.5-6 l Results of the Accident Progression Analysis for Crand Gulf Internal Initiators: PDS 6: long-Term SBO Prob . ** ZrOxid VB __ DCH-SE SPB CF Sprays MCCI SRVBkr, Order Bin Ten Most Probable Bins * !
.0.044 HiZrOx LoPnLPI. IoEXSE SPBE3L3 CE-Rpt. noCS FLDCCI cSRVBkr 1 BABDHBACB
- 0.041 HiZrOx -LoPnLPI LoEKSE SPBEOLO CL-Rpt noCS FLDCCI cSRVBkr i 2 BABDACACB 0.037 -HiZrOx LoPnLPI LoEXSE SPBEOLO- CVB-Rpt noCS FLDCCI cSRVBkr l 3' BABDAEACB '
0.033 IoZrOx IoPnLPI -LoEXSE SPBE01A CL-Rpt noCS FLDCCI cSRVBkr' 4 'BBBDAGACB ' 4 0.029 HiZrOx IoPnLPI LoEXSE SPEE3L3 CE-Rpt noCS noCCI cSRVBkr 5 BABDHBAEB HiZrOx LoPnLPI loEXSE ,SPBEOLO CVB-Rpt noCS -noCCI~ cSRVBkr . 6- BABDAEAEB 'O.026 0.026 .HiZrOx IoPnLPI- LoEXSE- SPBEOLO' CL-Rpt: noCS noCCI cSRVBkr 7, BABDAGAEB
.O.026 LoZrOx :LoPnLPI loEXSE- SPBEOLO CVB-Rpt noCS FLDCCI cSRVBkr 8 BBBDAEACB !
O.024 HiZrOx LoPnLPI IoEESE SPBE0I3 CVB-Rpt noCS .FLDCCI cSRVBkr 9 BABDBEACB' CL-Rpt noCS noCCI cSRVBkr-
- 5 10 BBBDAGAEB 0,020 LoZrOx. -LoPnLPI LoEXSE =SPBEOLO' Five Most' Probable Bins that.have Early CF and'Early Suppression Pool 2 Bypass
- l U ,
0.044 HiZrOx IoPnLPI -LoEXSE SPBE3L3 CE-Rpt noCS. FLDCCl- :cSRVBkr' i- 1' BABDHBACB cSRVBkr
-BABDHBAEB ~0.029 -HiZrOx lePnLPI LoEXSE SPBE3L3 .CE-Rpt' noCS -noCCI 5
cSEVBkr
~
0.024 HiZrOx LoPnLPI- LoEXSE SPBE013 CVB-Rpt noCS FLDCCI ,
'.9 BABDBEACB 0.016 LoZrOx: IoPnLPI .. LoEXSE SPBE013- .CVB-Rpt noCS FIDCCI .cSRVBkr.
12 'BBBDBEACB
- HiZrOx IoPnLPI LoEXSE SPBE013 CVB-Rpt. noCS noCCI cSRVP' . l 13- BABDBEAEB 0.016 A listing g of all bins, jand .a- listing by observation are available ou computer media.
Mean. probability conditional on the occurrence:of the PDS. v ! -l 3-4
;)
4 u i O. .. c :
i In all of the 10 most probable bins vessel breach occurs with the RTV at low pressure followed by an ex vessel steam explosion that involves a small f raction of the core. The containment sprays do not operate during the accident but because there are no stuck open SRV tailpipe vacuum breakers all of the in vessel releases are still scrubbed by the suppression pool. In all of the 10 most probable bins the core debris released from the vessel is ooled and there are no CCI releases. The probability that offsite electrical power is recovered before vessel breach is 0.19. Ilowever , because there is no coolant injection to the vessel the probability of vessel breach is 1.0. The probability that the containment fails early is 0.68. 2.5.1.7 Results for PDS 7! Short-Term S1Q. This PDS involves station blackout (without any de power) scenarios where LOSP is not recoverable. Coolant injection is lost early such that core damage occurs in the short term. The ADS requires de power. Thus, the operator cannot depressurize the vessel before core damage. Also, because offsite power is not recoverable, the functions of injection, heat removal, and those of the misec11aneous systems are not available. This PDS also includes cut sets
, - with either one or two stuck open SRVs. If the RPV is depressurized i through the stuck open SRVs, the firewater system can be used as a source of low pressure injections.
Table 2.5 7 lists the 10 most probable APBa for this PDS and the five most probable APBs that have early containment failure and early suppression pool bypass. Evaluation of the APET produced 1473 source term bins for this PDS. To capture 95% of the probability, 552 bins are required. The 10 most probable bins capture 91% of the probability. In all of the 10 most probable bins, vessel breach occurs with the RPV at high pressure followed by a DCH event that invnives a small fraction of the core. The containment sprays do not operato during the accident but because there are no stuck open SRV tailpipe vacuum breakers all of the in-vessel releases are still scrubbed by the suppression pool. - Because de power is lost, ac power can not be recovered and - the ADS is-unavailable such that the RPV is at high pressure. There is a small probability (4%), however, that a SRV will stick open and depressurize the RPV. Once the RPV has been depressurized, the firewater system can be used to provide coolant- injection to the RPV. The firewater system has its own power supply. Thus, the probability that vessel breach is averted is only 0.01. The probability that the containment fails early is 0.60. 2.5.1.8 Results for PDS 6; Lonr_-Term SBO. This PDS involves station blackout (without any de power) scenarios where LOSP is not recoverable. Coolant injection is lost late such that core damage occurs in the long term. The ADS requires.de power. Thus, the operator canne't depressurize the vessel before core damage. Since offsite power is not recoverable, the injection and heat removal functions and the misce 11ancor.s systems are not available. Table 2.5-8 lists the 10 most probable Ards for this PDS and the five most probable APBs that have early containment failure and-early 2.80
Table 2.5-7 .. l Results of the Accident Progression Analysis for Grand Culf Internal Initiators: PDS 7: Short-Term SB0 , I Prob . ** ZrOxid VB _ DCH-SE SPB CF Soravs MCCI SRVBkr Order Bin Ten Most Probable Bins
- 0.041 LoZrOx HiPnLPI' LoDCH SPBEOLO CVB-Rpt noCS noCCI cSRVBkr .l 1 ABABAEAEB-0.028 HiZrOx HiPnLPI LoDCH SPBE0LO CVB-Rpt noCS noCCI cSRVBkr '!
t 2 AAABAEAEB 3 AAABAIAEB 0.025 HiZrox. HiPnLPI loDCH EPBEOLO CnFall noCS noCCI cSRVBkr 0.024 HiZrOx HiPnLPI LoDCH SPBEOLO CL-Lk noCS noCCI. cSRVBkr I 4 AAABAFAEB LoZr0x- HiPnLPI leDCH SPBE013 CVB-Rpt- noCS noCCI cSRVBkr j S ABABBEAEB 0.018 .o.- f O.015- HiZr0x' HiPnLPI IoDCH SPBE2L2 CE-Rpt noCS noCCI cSRVBkr-6 AAABEBAEB
- 0.015. LoZrOx HiPnLPI LoDCH SPBEOLO CL-Rpt noCS' noCCI cSRVBkr' .
7 ABABAGAEB 8 ABABAFAEB' 'O.014 LoZrOx. HiPnLPI LoDCH SPBEOLO CL-Lk noCS' noCCI' cSRVBkr 1 y AAABABAEB 0.014- HiZrOx ' HiPnLPI. IoDCH SPBEOLO CE-Rpt noCS noCCI' cSRVBkr.-
.6 9 ;
i ~ 0.013 HiZrOx hip-LPI loDCH SPBEOLO' CL-Lk noCS noCCI cSRVBkr 10 AACBAFAEB t Five Host Probable-Bins that have'Early CF and Early Suppression Pool Bypass * , 70.018- LoZrOx HiPnLPI LoDCH ~SPBE0I3 CVB-Rpt noCS noCCI cSRVBkr. ;; 5 'ABABBEAEB O.015 - HiZrOx '. HiPnLPI- LoDCH :SPBE2L2 CE-Rpt- noCS noCCI cSRVBkr. 4 ! 6 AAABEBAEB.' 0.012 HiZrox- HiPnLPI inDCH SPBE013 CVB-Rpt noCS noCCI cSRVBkr 13 AAABBEAEB HiZrOx . HiPnLPI LoDCH SPBE3L3 CE-Rpt noCS noCCI cSRVBkr 16 AAABHBAEB 0.011.
'HiZrOx- hip-LPI leDCH SPBE3L3 - CE-Rpt. - noCS noCCI cSRVBkr-r 18 AACBHBAEB 0.010 t i:
t~ i i; . A listing"of all bins, and a listing by observation are available on computer media. 4 Hean' probability conditional on the occurrence of the PDS. i ' i e
- q m-- - - ~ , u ,.w.,- y._r
i Table 2.5-8 Results of the Accident Progression Analysis for Grand Gulf Internal Initiators: PDS 8: Long-Term SBO Order Bin Prob . ** ZrOxid VB _ DCH-SE SPB CF Sorays MCCI SRVBkr Ten Most Probable Bins
- 1 BAABAAAEB 0.067 HiZrOx HiPnLPI LoDCH SPBE01A CE-Lk noCS noCCI cSRVBkr-2 BBABAAAEB 0.040 LoZrOx HiPnLPI IoDCH SPBEOLO CE-Lk noCS noCCI cSRVBkr 3 BAABAEAEB 0.030 HiZrOx HiPnLPI LoDCH SPBEOLO CVB-Rpt noCS noCCI cSRVBkr 4 BACBAAAEB 0.030 HiZrOx hip-LPI LoDCH SPBE0LO CE-Lk noCS noCCI cSRVBkr 5 BBABAEAEB 0.027 LoZrOx HiPnLPI LoDCH SPBEOLO CVB-Rpt noCS noCCI cSRVBkr 6 BAABABAEB 0.027 HiZrox HiPnLPI LoDCH SPBEOLO CE-Rpt noCS noCCI cSRVBkr 7 BAABAAAEA 0.021 HiZrOx HiPnLPI LoDCH SPBE0IA CE-Lk neCS noCCI oSRVBkr 8 BAABAAACB 0.017 .HiZrOx HiPnLPI leDCH SPBEOLO CE-Lk noCS FLDCCI cSRVBkr u
a, 9 BAABBBADB 0.016 HiZrOx HiPnLPI IoDCH SPBE0I3 CE-Rpt noCS D1yCCI cSRVBhr
" 10 BAABAGAEB 0.016 HiZrOx HiPnLPI LoDCH SPBEOLO CL-Rpt noCS noCCI cSRVBkr Five Most Probable Bins that have Early CF and Early Suppression Pool Bypass
- 9 BAABBBADB 0.016 HiZrOx HiPnLII LoDCH SPBE013 CE-Rpt noCS DlyCCI cSRVBkr 13 BAABBAADB 0.014 HiZrGx HiPnLPI LoDCH SPBE0I3 CE-Lk noCS DlyCCI cSRVBk-18 BAABEAAEB 0.010 HiZrOx HiPnLPI LoDCH SPBE2L2 CE-Lk noCS -noCCI cSRVBkr 21 BBABBAADB 0.009 IoZrOx HiPnLPI LoDCH SPBE0I3 CE-Lk noCS DlyCCI cSRVBkr 23 BBABBBADB 0.009 LoZrOx HiPnLPI LoDCH SPBE013 CE-Rpt noC.i DlyCCI cSRVBkr A listing of all bins, and 'a listing 'by observation are available on computer :nedia.
1 Mean probability conditional on the occurrence of the PDS. L - _ _ _ _ _ _ _ _ _ _ _ _ _ . _
1 1 suppression , bypass. Evaluation of the APET produced 494 source term bins for th:u DS. To capture 95% of the probability, 232 bins are required. The < most probable bins capture 29% of the probability. In all of the 10 most probable bins, vessel breach occurs with the RPV at high pressure followed by a DCll event that involves a small fraction of the core. The containment sprays do not operate during the accident. There is only one bin that has a stuck-open tailpipe vacuum breaker; however for this bin the drywell does not . fail. Thus, all of the in-vessel releases are scrubbed by the suppression pool. Only one of the 10 most. probable bins has drywell failure. Because de power is lost, ac power can not be recovereo and the ADS is unavailable such that the RPV is at high pressure. Because Lns. e is no early coolant injection to the RPV, the probability of vessel bread is 1.0. The probability that the containment fails early is 0.54. 2.5.1.9 Results for PDS 9: Short-Term AWS. This PDS involves AWS scenarios. Coolant injection is lost early such that core damage occurs in the short term with the vessel at high pressure because the operator failed to depressurize it. The low pressure injection is recoverable with reactor depressurization, lleat removal via the sprays is available and the miscellaneous systems (i.e. , venting, SBGT, CI and 1121) are. available. Table 2.5-9 lists the 10 most probable APBs for this PDS and the five most probable APBs that have early containment failure and early suppression pool bypass. Evaluation of the - APET produced 1793 source term bins for this PDS. To capture 95% of the probability, 477 bins are required. The 10 most probable bins capture 33% of the probability. In the 10 most probable bins vessel breach occurs with RPV at high pressure. In nine of the 10 most probable bins a DCll event occurs at . vessel breach, and in the other bin an ex-vessel stee.m explosion follows vessel breach. In all but one of the 10 most probable bins the containment fails at vessel breach. Containment sprays are operating during- the intermediate time period in all of these 10 bins. There are no- CCI releases in all but one of these bins and in the bin that CCI does cccur the releases are scrubbed by a flooded cavity. Electrical power is always available in this PDS. The probability that the RPV will be at high pressure during core damage is 0.84. The probability that coolant injection will be restored to the RPV and vessel breach will be averted is only 0.04. This low probability of core damage arrest is driven by the failure of the operators to depressurize the RPV. The probability that the containment fails early is 0.67. 2.5.1.10 Results for PDS 10: Lonn-Term ATWS. This PDS. involves ATUS scenarios. Coolant injection is lost late such that core damage occurs in the long term with the vessel at high pressure because the operator failed to depressurize it. Low pressure injection is recoverable with reactor , depressurization. Ile at removal via the sprays is available, and the misec11aneous systems (i.e. , venting, SBCT, CI and 112I) are available. 2.83
i i , i i Table 2.5-9 Results of the Accident Progression Analysis for Grand Gulf Internal Initiators: PDS 9: Short-Term ATUC i Bin Prob . ** ZrOxid VB _ DCH-SE _ SPB CF Sprays MCCI SRVBkr t Order I Ten Most Probable Bins * ( 0.087 HiZrOx HiPnLPI LoDCH SPBEOLO CVB-Rpt LCS noCCI cSRVBkr
! 1 EAABAECEB 0.055 IeZrox HiPnLPI LoDCH SPBEOLO CVB-Rpt LCS noCCI cSRVBkr j 2 EBABAECLB 0.035 HiZrOx hip-LPI LoDCH CFBEOLO CVB-Rpt LCS noCCI cSRVBkr 3 EACBAECEB 0.033 HiZrOx- HiPnLPI IeDCH SPBEGI3 CVB-Rpt -IIS noCCI cSRVBkr 4 EAABBECEB 0.028 LoZrox HiPnLPI leDCH SPBE013 CVB-Rpt LCS noCCI cSRVBkr 5 EBABBECEB 0.021 IoZrox hip-LPI .IoDCH SPBEOLO CVB-Rpt LCS noCCI cSRVBkr 6 EBCBAECEB 0.021 HiZrOx HiPnLPI LoDG SPBEOLO CVB-Rpt LCS FLDCCI cSRVBkr
- 7 EAABAECCB 0.018 H12rOx HiPnLPI LoEKSE SPBEOLO CVB-Rpt IIS noCCI cSRVBkr w 8 EAADAECEB O.017 HiZrOx HiPnLPI IeDCH SPBEOLO CL-12 IIS noCCI cSRVBkr "w 9 EAABAFCEB
" O.017 HiZrOx hip-LPI IoDG SPBE0I3 CVB-Rpt LCS noCCI cSRVBkr 10 EACBBECEB i Five Most Probable Bins that have Early Cr and Early Suppression Pool Bypass
- 0.023 HiZrOx HiPnLPI LoDCH SPBE0I3 CVB-Rpt IIS noCCI cSRVBkr 4 EAABB1 . _.S 0.028 IoZrOx HiPnLPI LoDG SPBE0I3 CVB-Rpt IIS noCCI cSRVBkr
- 5. EBABBECEB 0.017 HiZrOx hip-LPI ImDCH SPBE0I3 CVB-Rpt LCS noCCI cSRVBkr 10 EACBBECEB D.017 HiZrOx hip-LPI IoDCH SPBE3L3 CVB-Rpt LCS noCCI cSRVBkr 15 EACBHECE6 0.017 HiZrOx hip-LPI LoDCH SPBE2L2 CVB-Rpt 14S noCCI cSRVBkr j 20- EAABEECEB A listing of all bins, and a listing by observation are available on computer media.
Mean probability conditional on the occurrence of the PDS. i l
Table 2.5 10 lists the 10 most probable APts for this PDS and the t ivo mos.t probable APBs that have early containment failure and early suppr$ssion pool bypass. Evaluation of the APET produced 1661 source term bins for this PDS. To capture 95% of the probability, 496 bins are required. The 10 most probable bins capture 21% of the probability. In all of the 10 most probable bins, vessel breach occurs with the RPV at high pressure followed by a DC11 event that involves a small fraction of the core. In all of these bins the containment fails early; however, there is coincident drywell failure in only one of these bins. The containment sprays operate before vessel breach in all of these bins and continue to operate during the entire accident in all but two of these bins. Electrical power is always available in this PDS. The probability that the RPV will be at high pressure during core damage is 0.97. The probability , that coolant injection will be restored to the RPV and vessel breach will be averted is only 0.01. This low probability of core damage arrest is driven by the failure of the operators to depressurize the RPV. The probability that the containment fails early is 1.0. The containment always fails in this PDS because the energy dumped into the suppression pool from the RPV during an ATWS transient exceeds the capacity of the RHR system which results in a large buildup of steam in the containment. 2.5.1.11 Results for PDS 11: Short -Terpt Il. This PDS involves transient scenarios where the PCS is lost (T2). Coolant injection is lost early such that core damage occurs in the short term with the vessel at high pressure because the operator failed to depressurize it. Both high and low pressure inj ection systems are recoverable since the failuret involved operator failures, lleat removal via the sprays is available and the miscellaneous systems (i.e. , venting, SBGT, CI and 111)2 are availabic. Table 2.5 11 lists the 10 most probable APBs for this PDS and the five most probable APBs that have early containment failure and early suppression pool bypass. Evaluation 'of the APET produced 2136 source term bins for this PDS. To capture 95% of the probability, 705 bins are required. - The 10 most probable bins capture 22% of the probability. In all of the 10 mos[ probable bins, vessel breach occurs with the RPV at high pressure followed by a DCH event that involves a small fraction of the core. The containment fails early in all but two of these bins. Only two of thcae bins have coincident early containment failure and early dryvell failure. The containment sprays operate during the intermediate time period in all of these bins and there are no CCI release in all but one of these bins. Electrical power is always available in this PDS. The probability that the : RPV will be at high pressure during core damage is 0.84. The probability I that coolant injection will be restored to the RPV and vessel breach will l be averted is only 0.05. This low probability of core damage arrant is driven by the failure of the operators to depressurize the RPV. The probability that the containment fails early is 0.56. 4 2.85
4 ; i Table 2.5-10 i Results of the Accident Progression Analysis for Crand Gulf Internal Initiators: PDS 10: Long-Term ATUS i Order Bin Prob . ** ZrOxid VB _ DCH-SE SPB CF Soravs MCCI SRVBkr Ten Most Probable Bins
- 1 FAABAADEB. 0.047 H1ZrOx HiPnLPI loDCH SPBEOLO CE-Lk ECS noCCI cSRVBkr 2 FACBAADEB 0.026 HiZrOx hip-LPI LoDCH SPBEOLO CE-Lk ECS noCCI cSRVBkr 3 FACBABDEB O.025 HiZrOx hip-LPI loDCH SPBE01D CE-Rpt ECS noCCI cSRVBkr 4 FAABABBEB -0.024 HiZrOx HiPnLPI loDCH SPBEOLO CE-Rpt ECSnoL noCCI cSRVBkr' 5 FBABAADEB 0.016 IoZrOx HiPnLPI LoDCU SPBEOLO CE-Lk ECS noCCI cSRVBkr 6 FACBBADEB O.OlG HiZrOx hip-LPI loDOI SPBEOI3 CE-Lk ECS noCCI cSRVBkr 4 7 FAABAADDB 0.015 HiZrOx HiPnLPI LoDCH SPBE01D CE-Lk ECS DlyCCI cSRVBkr 8 FAABABDEB O.015 HiZrOx HiPnLPI loDCH SPBEOLO CE-Rpt ECS noCCI cSRVBkr
- u 9 FBABABDEB 0.014 LoZrox HiPnLPI IoDCH SPBEOLO CE-Rpt ECS noCC1 cSRVBkr 6 cSRVBkr
- 10 FAABABBDB 0.013 HiZrOx HiPnLPI IoDCH SPBE01h CE-Rpt ECSnoL D1yCCI Five Most Probable Bins that have Early CF and Early Suppression Pool Bypas3*
6 FACBBADEB O.016 HiZrOx hip-LPI loDCH SPBE0I3 CE-Lk ECS noCCI cSRVBkr 13 FBABBADEB 0.011 LoZrox. HiPnLPI loDCH SPBE013 CE-Lk ECS noCCI cSRVBkr 16 FAABBBBEB 0.010 HiZrOx HiPnLPI IoDCH SPBE0I3 CE-Rpt EC5noL noCCI cSRVBkr 17 FAABBADEB O.010 HiZrOx HiPnLPI leDCH SPBEDI3 CE-Lk ECS noCCI cSRVBkr 22 FAABBBDEB 0.008 HiZrOx HiPnLPI IoDCH SPBE013 CE-Rpt ECS noCCI cSRVBkr A listing.of all bins, and a listing by observation are available on computer media. Mean probability conditional on the occurrence of the PDS.
=t-1 e e e. 3 m p e y- .+ a w .
t-" w<-- .- % '- w i9' * --e u ------- u---------- --
4 1 I Table 2.5-11 Results of the ticcident Progression Analysis for Grand Gulf Internal Initiators: PDS 11: Short-Tern T2 Order Bin Prob . ** ZrOxid VB _ DCH-SE SPB CF Sprays MCCI SRVBkr Ten Most Probable Bins
- 1 CAABAECEB 0.060 HiZrOx HiPnLPI ToDCH SPBEOLO CVB-Rpt LCS noCCI cSRVBkr 2 CBABAECEB 0.030 IoZrox HirnLPI LoDCH SPBEOLO CVB-Rpt LCS noCCI cSRVBkr ,
3 CACBAECEB 0.025 HiZrox hip-LPI loDCH SPBEOLO CVB-Rpt LCS noCCI cSRVBkr 4 CBABBECEB O.018 IoZrOx HiPnLPI loDCH SPBE0I3 CVB-Rpt LCS noCCI cSRVBkr 5 CAABBECEB 0.018 HiZrOx HiPnIEI feDCH SPBE0I3 CVB-Rpt IES noCCI cSRVBkr 6 CBCBAECEB 0.016 IoZrox hip-LPI LoDCH SPBE01D CVB-Rpt LCS noCCI cSRVBkr 7 CAABAECEA 0.015 HiZrOx HiPnLPI ImDCH SPBEOLO CVE-Rpt LCS noCCI oSRVBkr 8 CAABAECCB 0.014 HiZrox HirnLPI leDCH SPBEGIA CVB Rpt ILS FLDCCI cSRVBkr !
- o 9 CAABAFCEB 0.014 HiZrOx HiPnLPI IoDCH SPBEOLO CL-Ik LCS noCCI cSRVBkr ie 10 CAABAICEB 0.013 HiZrOx HiPnifI IoDCH SPBEOLO CnFall IIS noCCI cSRVBkr l
Five Most Probable Bins that have Early CF and Early Suppression Pool Bypass
- 4 CBABBECEB O.018 IeZrOx liiPnLPI loDCH SPBE0I3 CVB-Rpt ILS noCCI cSRVBkr CAABBECEB 'O.018 H1Zr0x HiPnLPI LoDCH SPBE0I3 CVB-Rpt LOS noCCI cSRVBkr 5
13 CACBBECEB O.011 HiZrox hip-LPI IoDCH SPBEGI3 CVB-Rpt LCS noCCI cSRVBkr 18 CAABBECEA .0.009 HiZrOx HiPnLPI IoDCH SPBE0I3 CVB-Rpt 145 noCCI oSRVBkr 20 CACBHECEB 0.008 HiZrOx hip-LPI leDCH SPBE3L3 CVB-Rpt LCS noCCI cSRVBkr A listing of all bins, and a listing by observation are available on computer media. Mean probability conditional on the occurrence of the PDS. t
2.5.1.12 gesults for PDS 12: Long-Term T2. PDS 12 is the same as PDS 11 except that core damage occurs in the long term. Table 2.512 lists the 10 most probable APBs for this PDS and the five ruost probable APBs that have early containment failure and early suppression pool bypass. Evaluation of the APET produced 2136 source term bins for this PDS. To capture 95% of the probability, 705 bins are required. The 10 tnost probable bins capture 22% of the probability. In all of the 10 most probable bina essel breach occurs with the RPV at high pressure followed by a DCil event diat involves a small fraction of the l core, the containment fails early in all but two of these bins. Only two l of these bins have coincident early containment failure and early drywell i failure. The contaitunent sprays operate during the intermediate time ! period in all of these bins and there are no CCI release in all but one of ! these bins. ! l Electrical power is always available in this PDS. The probability that the RPV will be at high pressure during core damage is 0.84. The probability that coolant injection will be restored to the RPV and vessel breach will be averted is only 0.05. This low probability of core damage arrest is driven by the failure of the operators to depressurite the RPV. The probability that the containment fails early is 0.56, 2.5.1.13 Core Damage Arrest and Avoidaneo of Vessel Breach. Once core damage has begun, the only way vessel failure is prevented is if coolant injection is restored to the RPV. Restoration of coolant injection to the RPV, however, does not necessarily preclude vessel breach. If injection is not recovered until late in the core damage process, it is unlikely that the addition of water will revent vessel breach. In addition, there is the possibility that the ce ra debeis that slumps into the bottom head of the vessel will trigger a steam explosion. Although steam explosions do not guarantee vessel failure, they do pose a significant challenge to the integrity of the RPV and in some cases do result in vessel failure, Figure 2.51 shows the probability that core damage is arrested before the lower head of the vessel fails for the four collapsed PDS groups (super-groupa). For the short term station blackout super group the probability of core damage arrest is driven by the likelihood that ac power is recovered early in the accident. Injection to the RPV generally follows ac power recovery. Although the mean probability of recovering ac power is high (0.62%) for short term station - blackout PDSs, there are several factors that tend to reduce the probability of core damage arrest. First, restoration of coolant injection to the RPV does not guarantee that the vessel will not fail. In some cases the core debris is not in a coolable configuration when inj ection is recovered and, therefore, the accident continues to vessel breach. There are other cases in which only low pressure inj ection systems are recovered; however, the operators have failed to depressurize the RPV. With the vessel at system pressure these low pressure systems are unable to provide coolant to the core and, therefore, the accident proceeds to vessel breach. Finally, in PDS 7, which is a significant contributor to the mean frequency of this saper-group, ac power cannot be recovered. Therefore, except for the infrequent 2.88
i Table 2.5-12 Results of the Accident Progression Analysis for Grand Gulf Internal Initiators: PDS 12: Iong-Term T2 Bin Prob .** ZrOxid VB __ DCH-SE _ SPB CF Sorays 11C.C_1 SRVBkr Order i Ten Most Probable Bins * , i 1 DAABAECEB 0.060 HiZrox HiPnLPI LoDCH SPEEOLO CVB-Rpt LCS noCCI cSRVBkr 0.030 LoZrox HirnLPI LoDCH SPBEOLO CVB-Rpt IES noCCI cSRVBkr 2 DBABAECEB ' 3 DACBAECEB 0.025 H1ZrOx hip-LPI IoDCH SPBEOLO CVB-Rpt LCS noCCI cSRVBkr
'4 DBABBECEB 0.018 IoZrOx HiPnLPI loDCH SPBE013 CVB-Rpt LCS noCCI cSRVBkr 5 DAABBECEB 0.018 HiZrOx HiPnLPI LoDCH SPBE0I3 CVB-Rpt IES noCCI cSRVBkr 6 DBCBAECEB 0.016 IeZrOx HIP-LPI LoDCH SPBEOLO CVB-Rpt 145 noCCI cSRVBkr DAABAECEA 0.015 HiZrOx HIPnLPI LoDCH SPBE01A CVB-Rpt LCS noCCI oSRVBkr 7
8 DAABAECCB 0.014 H1ZrOx HiPnLPI leDCH SPBEOLO CVB-Rpt IES FLDCCI cSRVBkr y 9 DAABAFCEB 0.014 HiZrOx HiPnLPI leDCII SPEEOLO CL-Lk LCS noCCI cSRVBkr
- e. 10 DAABAICEB 0.013 HiZrOx HiPnLPI IeDCH SPBEOLO CnFail LCS noCCI cSRVBkr a
Five Most Probable Bins that have Early CF and Early Suppression Pool Bypass
- 4 DBABBECEB O.018 IoZrox HiPnLPI LoDCH SPBE013 CVB-Rpt IIS noCCI cSRVBkr i 5 DAABBECEB 0.018 HiZrOx HiPnLPI IoDCH SPBE013 CVB-Rpt LCS noCCI cSRVBkr DACBBECEB 0.011 HiZrOx hip-LPI IoDCH SPBE013 CVB-Rpt LCS noCCI cSRVBkr ;
13 0.009 HiZrOx HIPnLPI IeDCH SPBEGI3 CVE-Rpt LCS noCCI oSRVBkr
'18 DAABBECEA 20 DACBHECEB O.008 HiZrOx hip-LPI loDCH SPBE3L3 CVB-Rpt IES noCCI cSRVBkr A listing of all bins, and a listing by observation are available on computer media.
Mean probability conditional'on the occurrence of the PDS.
, s i
4 i~ - i
l 3 l cases which involve a stuck open SRV that depressurizes the RPV and allows firewater to be injected into the vessel, accidents in this group progress to vessel failure. 1.00 Grand Gulf 95tk 3, 6fith
)
w . me.n m - median
- th = percenule b
=1 v u- A 1 3b 96th. ~* ],
E< ] J iO L ' A E1.E-1. e m.. eeth. 3; l ..ts. , l , t u
]6 u-.
6i u..
- m. I k
, 1 1 IE-2 " 7 F* I ~l 7 pDS Group STSB 1.TSB ATWS Translerus Total Core Demare fYeq 3.9E-06 1.0E- 07 1.l E-07 1.9E-08 4.1E-06 Figure 2.5 1. Probability of Core Damage Arrest. As with the short-term station - blackout super group, the probability of core damage . arrest for the long term station blackout super group is also driven by the likelihood that ac power is recovered, The probability of- , core damage arrest for the long term station blackout super group, however, ! is significantly lower than the corresponding value for the short term - station blackout super groQp. Two factors are responsible for most of this difference. First, the mean probability of ac power recovery for the long-term station blackouts (given that power can be recovered) is roughly a third of the corresponding probability for a short-term station blackout, d The conditional probability of recovering ac power is defined as the probability of recovering power during the core degradation process given that power was not available at the initiation of core damage. The greater 2.90
._ _ - __ . _ _ _ . _ , _ - ~ , . . .- _ . _ _ . . -
- - .-. . -- - _ . - - - _ = . - - . - _ . - - ..- - . - .
I I h the amount of time that elapses without power recovery (i.e. , the start of the time interval), the smaller the probability that power v111 be recovered in an ensuing time interval. For a short term SB0 accident, core dawage occurs one hour af ter the initiating event whereas for a long term l SB0 accident, core damage occurs 12 hours after the initiating event. l Second, in PDS 8, which accounts for approximately half of this group's mean frequency, ac power cannot be recovered- and the accident always
- proceeds to vessel breach, For both the ATWS super group and the T2 super-group, the probability of core damage arrest is driven by operator errors. In these PDSs-low pressure injection systems are available; however, the operator fails to depressurize the RPV. The mean probability of core damage arrest for the ATWS super group is slightly lower than the mean value for the T2 super-group. There are two reasons for this difference. First, the operators are more susceptible to errors during the accidents in the ATWS super group than they are in the T2 super group. Second, in the ATWS super group the probability that the core debris is cooled when injection is restored is lower than the corresponding probability in the T2 super group.
l It must be remembered that core damage arrest does not necessarily mean . that there will be no radionuclide releases during the accideint. Both hydrogen and radionuclides are released to the containment during the core l damage process. If a large amount of hydrogen is generated during core damage and is subsequently ignited, it is possible that the resulting load will fail the containment. If the containment fails, a pathway is established for the radionuclides to enter the outside environment. This radionuclide release is generally _ small, however, because in the majority of the cases in which vessel breach is averted these releases are scrubbed as they pass through the suppression pool. Furthermore, if the vessel does not fail, there are no ex vessel releases (e.g., CCI releases). 2.5.1.14 Early Containment Failure. The early fatality risk depends strongly on the probability of early containment failure (CF). Early containment failure includes both failures that occur before vessel breach and during the time period around vessel breach. The Grand Gulf l containment is fairly weak structure when compared to the loads that can. potentially occur during the course of the. accident. The design pressure is only 15 psig and the assessed mean failure pressure is 55 psig. Because of its low failure pressure, the Grand Gulf containment is susceptible not-only to loads from hydrogen deflagrations and ' detonations but can also be threatened by slow pressurization events (i.e. , ' the accumulation of steam and hydrogen generated during the- core -degradation. process) during accidents that do not have adequate containment heat removal capacity (e.g., long term SB0s and long term ATWS). The production of hydroge durir; the core damage process and later during. vessel breach, should it occur, is a key- factor that affects the probability of containment failure. In a BWR core there is a large inventory of zirconium. The Grand Gulf- core, 'for example, which contains l approximately 80,000 kg of zirconiam, has- nearly five times as much zirconium as does the Surry core (which is a PWR), Large amounts of hydrogen are produced .from the oxidation of this metal during the core 2.91
1 l damage process. If the IIIS is not operating, the hydrogen vill 4.ccumulate in the containment. For accidents in which the suppression pool is subcooled, the steam released from the RPV is condensed in the pool. The ! lack of steam in the containment atmosphere in combination with the large
- amount of hydrogen released durin6 the core degradation process allows
] mixtures to form that have a high hydrogen concentration. Subsequent ignition of this hydrogen by either random sources or by the recovery of ac 1 ! power can result in loads that cannot only threaten the containmant but can i also pose a significant challenge to the dryvell structure. Figure 2.5-2 shows the probability distribution for ear'y CF at Grand Gulf. The probability distributions displayed in this figure atc for accidents 4 that proceed to vessel breach and are cenditional on core damage. 1 1.E0 . sotu 95gg a* m abtb 26th - I -
- u ; . ., -
! w ,F l, W.. i y ] l stu J h c _J 1 8 "j i- 2 ~ F ~ ! b ;; - ! ::: L i go
- 6th., a 4
c C , ' 4 o ) e E1.E-1.
- a .c ~
[l 7 .3 .. I e a f a
$0 o
~
i 5hcL - i i Od y i
~
'l etu l j Grand Gul! N**'""
" * *
- dI""
1.E-2. stu ] stu ] 'th = percentile PDS Group STSB LTSB ATWS Transients All Core Damage Freq. 3.9E-06 1.0E-07 1.1E 07 1.9E-08 4.1E-06 I Figure 245-2. Probability of Early_ Containment Failure. 2.92
(.- t i i s i I
- Although the mean conditional probability of early containment failure for l
accidents in which vessel breach is averted is approximately 0.25, the l radionuclide releases are generally small because there are only in vessel releases and these releases are typically scrubbed by.the suppression pool. Thus, the early fatality risk is not strongly influenced by the cases in 4 which vessel breach is averted and, therefore. . these cases - have not been l- included in the early containment failure probabilities, ' a l Figure 2.5 3 shows the mean probability of containment failure before
- vessel breach sorted by events that can load to containment failure, a Figure 2.5 4 presents the same type of information for containment failures that occur at vessel breach. These mean values are conditional on core damage.
! The weakness of the containment, relative to the loads that are imposed on l it, is reficcted in the relatively high containment failure probabilities.
}lydrogen combustion events are the dominant events that cause early CF in the short term station blackout and T2 super. groups. The mean probability-of early containment failure for these two PDS is roughly 0.$, In both of 1
- these summary PDS groups the suppression pool is subcooled before vessel breach and, therefore, there is no significant accumu?ation of steam in the containment. Although this virtually eliminates the possibility of CF from
, slow pressurization events (e.g., accumulction of steam), it does allow mixtures to form in the containment during a- short term SB0 that have a fairly high hydrogen concentration. Because. the llIS is. initially unavailable in durfng a short term SBO, it is not uncommon for the hydrogen
SUMMARY
SUMMARY
PDS GROUP i ACCIDENT (wean Core Damage frequency) PROGRESSION 7,'gtffY BIN GROUP sisa Lisa ATws transients Average (3.85E-06) -(1.04E-07) (1.12 E-07) (1.87E-08) (4.09E-06) ; 2 CP: Detonation 0.039 0.002 0.004 0.036 j Cf. Deflegration 0 ,102 0.075 0.000 0.010 0.101-L CT: Slow Press 0.004 0.471 0 450 0.006 0.050_ _ Vent 0 000 0.007 4 No CF Before VB- 0.763 0 452 0.445, 0.973 0.736 CP = Containment Tallure Grand Gulf Figure 2.5 3,- Mean Probability of CF Before Vessel Breach. 2.-93 _ T l-
._.-..._.-a. . _ - . . . _ - . . . . - - . . . ~ _ . __..--_.-.-._.;.;-.._,-. , w -, . . . . _ , --. _ c. --
_ _.- ~ _ _ _ _ _ . _ . - . ~ _ _ _ _ . _ . . _ _ _ _ . _ _ _ _ . . _ _ _ _ _ _ _ _ . _ _ _ _ 1 l
- 1
- l i I i l t
. concentration in containment to be above 16% before vessel breach. In the
, short term SB0 PDS group about half of the early CF probability. results from failures that occur before vessel breach and the other half results
! from failures shortly af ter vessel breach. In the T2 super group, on the the other hand, almost all of the early CFs occur at the time of vessel l breach. For accidents in the T2 super group, it is likely that; the operator turned on the HIS before core damage and, therefore, the hydrogen generated before vessel breach is usually burned such that the resulting load is benign.
SUMMARY
SUMMARY
PDS GROUP Nenn core Damage nemey) ACCIDENT PROGRESSION y,*gt,"d ' # BIN GROUP sisu 1.tsa Atwa 7tannients Average (3 050-06) (1.04E-07) ' (1.12E-07) (1.07E-08) (4.0DE-06)
, CP: Detonation 0.032 0.000 0.010 0.030 CP: Deflagration 0.100 0.300 0.280 0.52 0,208 f _ _
Alpha 0.003 0.004 0.001- 0.003 0.003 CF Defore VD 0.219 0.644 0.552 0 022 0.248 l 4 , No Early CF 0.545 0.14 6 0.151 0.437 0,499 ' (. l CF = Contulnment Failure Grand Gulf . ' Figure 2.5 4. Mean Probability of CF at Vessel Breach. - ! For the long term SB0 super group, the mean conditional probability of , early CF is 0.85. I.ess than half of.this probability comes from CFs caused by_ hydrogen combustion events. In this super group the suppression pool is I saturated and the containment is pressurized by tho' accumulation of: steam-that is generated by the hot pool. In most of these accidents hydrogen-burns are not possible because the containment is steam inert. Thus, the preponderance of the CFs that- occur before vessel breach are caused by pressurization events associated .with the . accumulation of steami in tne ; containment. There are a few cases, however, -in which the containment sprays are . recovered before vessel breach.- In these_ - cases- the' sprays . slowly condense the steam which allows a combustible - mixture to form. Ignition of this mixture can potentially,failithe containment. Roughly - a third of this mean probability results from CFs that occur at vessel breach and the vast majority of these failures are. caused by hydrogen combustion events. l l 2.94 =
For the AIVS super group, the snean conditional probability of early CF is 0.76. Similar to the long term SB0 super groud, less than half of this i probability comes from CFs caused by hydrogen combustion events. This ' super group consists of both a long-term PDS and a short term PDS. In the long terin PDS the suppression pool is saturated and either the operators vent the containment or the containment fails before vessel breach from the aceutnulation of steam in the containment. This PDS is responsible for a little nore than half of this super group's inean frequency. In the short-term PDS, on the the other hand, altnost all of the early CFs occur at the .. tirne of vessel breach. The pool is subcooled in the short terin PDS. Although combustible mixtures can form in thu containment before vessel breach in this PDS, the llIS in typically nn during core damage and, , tnerefore, the hydrogen generated before vessel breach is usually burned such that the resulting load is benign. 2.5.1.15 Early Drywell Failure. Early drywell failure is an important attribute of the accident progression because failure of the drywell establishes a pathway for radionuclides in the drywell to bypass the suppression pool, Radionuclides are released to the drywell atmosphere at vessel breach and during CCI. In-vessel releases can also enter the drywell if a vacuum breaker sticks open on a SRV tailpipe. Although an intact drywell guarantens that all of the releases will be scrubbed by the pool, drywell failure does not necessarily mean that the radionuclides will be released frotn the containment, should it fail, without being scrubbed. The in vessel releases, except from accidents that involve a stuck open SRV tailpipe vacuum breaker and a failed drywell, are released to the pool where they are scrubbed before entering the containment. Furthermore, if the containment sprays are operating, the ex vessel releases will be scrubbed by this system. Similarly, if the reactor cavity contains water, which is a likely event, the pool overlaying the core debris will scrub the ) CCI releases. Because accidents that result in early dryvell failure coincident with early containment failure are generally the dorninant risk contributors, it is appropriate to discuss the events that can lead to _early drywell i failure. Figure 2.5 5 shows the mean probability of drywell failure before ! vessel breach sorted by events that can lead to drywell failure. Figure l 2.5 6 presents the same type of information for drywell failures that occur i at vessel breach. These mean values are conditional on core damage; they are not conditional on either vessel breach or early containment failure. In fact, the tnean probability of early drywell failure is 0.31, however, , the mean probability of coincident early containment and dryvell failure is ? 0.23. Thus, some of the accidents tt,at have early drywell failure do not involve early containment , failure. llowever, these figures provide useful insight into the events that are responsible for early drywell failure. Before vessel breach the only significant event that causes drywell-failure is hydrogen combustion in the wetvell. Although the containment structure is considerably weaker than the drywell wall, rapid deflagrations and detonations in the wetwell can lead to large pressure differentials across the drywell wall which can cause drywell failure. For these rapid combustion events, neither containment failure nor passage of gases through the suppression pool into the drywell occur quickly enough to instigate:the i l l 2.95 l \ l _
l l i l pressure rise from the burn. Slow pressurization events associated with J the accumulation of steam in the containment are not a threat to the drywell structure. For the short term SB0 super-group, most of the failures are caused by deflagrations. A relatively small fraction of these l failures is caused by detonations. The mean probability of drywell failure before vessel breach is considerably less for the other PDS groups. There are several reasons for the lower failure probability in the 9 - groups. In the long term SB0 PDS group the contaitunent is frequently steam inert during this stage of the accident. In the ATWS PDS group, the containment is steam inert in some of the cases and in many of the other cases the HIS is on during core damage. In the T2 PDS group, the llIS is also generally on during the core damage process. t
SUMMARY
SUMMARY
PDS GROUP' (uean core Damese frequeney) ACCIDENT PROGRESSION Ql4ueggy t BIN GROUP sisD t.TSB ATWS Transients Average (3.85E-06) (1.048-07) (1.12E-07) (1.87E-00) (4.09E-06) DWP: Detonation 0.021 0.002 0.019 DWP: Deflagration 0.104 0.048 0.026 0.055 0.097 No DWF Defore VD 0.003 0.951 0.972 0.942 0.874 DWT = Drywell Fhilure Grand Gulf , 1 Figure 2.5 5. Mean Probability of Drywell Failure Before Vessel Breach. d 2.96
SUMMARY
SUMMARY
PDS GROUP (ucen c re Damare Frequency) ACCIDENT PROGRESSION 'Jlip*lacY BIN GROUP RSD LTSB ATWS Transients Averare (3.050-06) ( LO4 E- 07) (1.12 E - 07) (1.07 L- 0 0) (4 ODL-0W DWP: Loads 0 004 0 150 0.150 0.14 6 I 0.0B0 Accornpanying VD , _ _ ,, L DWF. Federtal 1%i1. l 0.051 0.000 0.100 0.090 0.050 (Loads Accornp. VD) J , _ ,, DWP: Fedestal Fall. 0.019 0.010 0 007 0 ODD 0010 (Dynernie Loads) DWF. Detonation 0 017 0.003 0.000 0 010 DWF: Deflagration 0.022 0.011 0.019 Alpha 0 005 0.000 0.002 0.002 0.00$ DWF: Ocfore VB O.100 0.043 0.020 0 OLL 0.090 No Ferly DWF 0000 0 072 0 090 0.003 f 0.000 _ L DWF = Drywell Failure Grand Gulf Figure 2.5 6. tiean Probability of Drywell Failure at Vessel Breach. For dryvell failures that occur at vessel breach, loads accompanyinB Ve8851 breach are responsible for the majority of these failures. These quasi-l static loads, which were provided by the containment Loads Expert Panel, include contributions from: DCH, ex vessel steam explosions, hydrogen burns, and RPV blow down. At vessel breach these events pressurize the drywell volume before the suppression pool vents clear. Drywell failures caused by theso loads are responsible for nearly 50t of the incan probability of drywell failure at vessel breach. In addition to directly pressurizing the drywell volutne , these loads can also pressurize the reactor cavity and f ail the pedestal. In some cases loss of reactor support cau induce drywell failure. This is the second event in Figure 2.5 5 that causes drywell failure and it is responsible for almost 30% of the mean ptobability of drywell failure at vessel breach. As can be seen in this figure, alpha modo events are a negligible contributor to the mean probability of early drywell failure. 2.5.1.16 Summarv. Figure 2.5 7 shows the mean distribution among the summary accident progression bins for the PDS super groups. Only mean values are shown, so Figure 2.5 7 gives no indication of the range of values encountered These m an values are conditional on core damage. The 2.97
_ _ _ _ _ _ _ _ _ __ ._ __.. _ _ ~_ _ . _ _ . - _ _ _ _ _ ----- -__ f distribution for core damage arrest is shown in Figure 2.5 1, and the distribution for early (at or before vessel breach) failure of the containment is shown in Figure 2.5 2. Nonetheless, Figure 2.5 7 gives a good idea of the relative likelihood of the possible results of the accident progression analysis. The summary bins are composed of essentially four characteristics: occurrence of vessel breach, timing of containment failure, timing of suppression pool bypass, and the availability of the containment sprays. The summary bins are listed roughly in decreasing order of the severity of the resulting source term. The last two bins are an exception to this ordering scheme. Because thero are some accidents in the NO vessel breach summary bin that have early containment failure, the releases associated with this bin are higher than releases for the vessel breach, No CF summary bin. A description of these summary bins is presented in section 2.4.3. Because roughly 90% of the total mean core damage frequency is attributed to the short term SB0 super group, the results presented in the frequency weighted average column are heavily influenced by the short term SB0 results. If the accident proceeds to core darnage , containment failure during the accident is a likely outcome. The mean conditional probability of early containment failure is approximately 0.50 and half of this mean value is associated with accidents that also involve some bypass of the suppression pool (i.e., drywell failure). If the accident proceeds to vessel breach and the containment does not fail early, there is still a fairly high probability that the containment will fail late in the accident. Events that can fail the containment late in the accident are hydrogen burns and the accumulation of noncondensibles and steatn in the containment. In the SB0 PDSs ac power inay not be available late in the accident and, thus, the containment sprays will not be available to condense the steam. Furthermore, even if the sprays are available, the accumulation of noncondensibles generated at vessel breach and during CCI may still fail the containment. Containment venting is not a likely outcome in this analysis. There are several reasons for this result. First, the dominant PDSs are the short term station blackouts. In these PDSs, the suppression pool remains subcooled during core damage and, therefore, the containment,is not pressurized by the accumulation of steam. During core damage and af ter vessel breach a significant quantity of radionuclides will be released to the containment. After vessel breach it is unlikely that the operator will vent theo releases to the outside environment. The first two summary bins represent accidsats in which vessel breach occurs and both the containment and the drywall fail early. The only difference between the first two bins is the availability of the containment spray system. For accidents characterized by the first bin, I the majority of the ex vessel releases will not be scrubbed by either the suppression pool or the containtnent sprays whereas releases associated with l the second bin will be scrubbed by the sprays. For the SB0 PDSs, the first i bin is a inore likely outcome than the second bin because in many cases ac l power is not available. The opposite trend is observed for the ATWS and T2 PDSs because ac power is never lost in these PDSs. 2,98
.. . - . . - ~ _ . - . - . . - . . . . . - . - . _ - , - . . - . . - - _ ~ . - . . - - - . . . . . - . .
)
*l:-00 0
- - 07) OI - 07) t I *$lif (4 0 1:-e4 VD, early Or. 0100 0192 0 000 0 011 0.160 early $PD, no CD i
~ VD. early CF, 0031 0.017 0 237 0.202 0 049 i early SI'U. CS __ ,,, VD, early CF, 0.006 0.000 0.003 0.003 0 007 late SPD 1 4 i VB, early CF, 0,162 0 631 0 601, 0.331 0.21e f no SPD ,,,, __ , , , f 0.0'i4 0 232 0 204
- VD, late CF 0000 0.129
.) VD, venting ! 0.032 0.003 0.109 0.076 0 030 1 I , VD, No CF 0.0$3 0.003 0.000 0.092 0,050 )l No VD 0.201 0.016 o.026 0.0LO h 0.100 . CP = Containment Tallure Grand Gulf - CS = Containment Sprays CV = Contamment Venting EPD = Suppression Ibol Dypass VB = Vessel Dreach Figure 2.5 7. Mean Probability of APBs for the Summary PDSs. l l The bin that involves vessel breach, early CF, and no suppression pool
- bypass is a like y outcome for both the long term SB0 super. group and the
^ ATWS super group. The reason for this result is that many of these CFs are caused by the accumulation of steam in the containment. This slow type of . pressurization event can fail the containment but does not pose a threat to the drywell strunture. For accidents characterized by this bin,-both the-in vessel and ex vessel releases will be scrubbed-by the suppression pool.- The short term SB0 super group is the only group that has a significant probability of core damage arrest. The mean probability'that vessel breach is averted in this group is 0.20. Although a quarter. of this tocan value is associa_ed with accidents in which core damage is strrested, the majority of- - the in vessel releases are directed to the suppression pool where they are scrubbed before entering the containment. For the accidents in which the core damage process is not arrested and the accident proceeds to vessel breach, there is a significant probability that core debris is cooled and there are no CCI releases.- If CCI is initiated it will most likely' occur in a flooded cavity. CCI releases that occur in. 2.99 4
~< _
a flooded reactor cavity will be scrubbed by the overlaying pool of water. The mean conditional probabilities for the cases with no CCI and the cases where CCI occurs with an overlaying pool of water are: CCI with Overlavinr-flg No CCI fool of Vater Dry CCI Short-Term SB0 PDS 1 0.64 0,35 0.01 2 0.64 0.35 0.01 3 0.61 0.38 0.01 7 0.77 0.21 0.02 Long Term SB0 PDS 4 0.45 0.$2 0.03
, 5 0.45 0.52 0.03 6 0.41 0.55 0.03 8 0.69 0.20 0.11 I ATWS PDS l )
9 0.76 0.24 <0.01 10 0.62 0.21 0,17 l T2 PDS l f 11 0.76 0.24 <0.01 0.76 12 0.24 <0.01 These mente values are conditional on core damage. Furthermore, the no CCI case includes the accident progressions in which vessel breach is arrested. I Figure 2.$.7 shows the mean frequencies for the summary FDS groups and mean conditional probabilities for the summary APBs, where the mean is taken over all 250 observations in the sample. The mean conditional probability of each summary APB may be computed for each PDS group for each observation. When combined with the PDS group frequency, a frequency for each summary APB for each observation is obtained. The distabuts.st, of these values is displayed in Figure 2.5 8. 2.6 Insir. hts from the Accident Prorression Annivsis Several insights can be drawn from the accident progression analysis. First, for the PDSc analyzed in this study, containment f ailure during the accident is a likely outcome, The predominant causes of these failures are hydrogen deflagrations. In the short term SB0 PDS group, which is responsible for roughly 90% of the total mean core damage frequency at Grand Gulf, ac power is not available early in the accident and, therefore, the llIS is not available at the beginning of core damage for the_ vast majority of the accident analyzed. Without the IIIS the hydrogen that is 2.100
i i 1 11-4, M . mean Grand Gulf ( in . mecan th 34rcehtile 1 IA e nu. 9au. .
.ns.
914 L. 9 H L. DHL. 1.l> 0 , "*
, } v ..
.1 u u-. ,
u V* -. --
,w, u - -
'O1E-7s ,
6% ! - aw - g % N [' I ; l L1E-8, ) ~ l V, ' "* ' j y , l l :
. auJ - -
1 E- D .,.
~J lE-!Q l mJ q
! )
1 g_1 l 10-12 o '" !*n l .o *o l l VD, VD, VD,, VD, VD, VD, VD, No VD APB GrrauJ' ently tr. early CT, early CT, early CF, late Cr venting No tr early EPB. early SPD, late EPB noFPD r.o CS Cf Figure 2.5 8. Distribution of Frequencies for APB Groups. , produced from the oxidation of zirconium during the core damage process can accumulate in the containment. SG 'cquent ignition of this hydrogen by either random sources or by the recovery of ac power can result in loads that cannot only threaten the containment but they can also pose a significant challenge to the drywell structure. For the PDSs-analyzed in this study, the mean conditional probebility of early containment fai-lure is nearly 0.5. Furthermore, half of this mean value comes from accidents that also involve some bypass of the suppression pool. Increased availability of the 1115 will significantly reduce the probability of both l containment and drywell failure before vessel breach. However, because of the, weakness of the containment and the potential for the rapid combustion i of hydrogen at vessel breach, the containment will still be-susceptible to l loads at vessel breach. Furthermore, the integrity of the drywell vill still be challenged by loads accompanying vessel breach. The results of the analysis to determine whether there is water in the i reactory cavity, as described in Appendix' A.1, indicate that there is a l high likelihood that the cavity will contain water at vessel breach. The presence of water in the cavity is important, and has both advantages and disadvantages. The presence of water allows for the possibility of ex-vessel steam explosions. On the other hand, this water also contributes to the high probability that ore dew released from the vessel vill be cooled. If CCI does in_..ste, . release will . be scrubbed by the overlaving pool of water. 2.101 I . . _ _ _ _ . . . - _ _ _ ,_ - _ - - - . .
I J I 2.7 References I
- 1. M. T. Drouin, J. L. LaChance, B. J. Shapiro, S. Miller, and T. A. j Vheeler, " Analysis of Core Damage frequency: Grand Gulf, Unit 1 ;
Internal Events," NUREG/CR 4550, Volume 6, SAND 86 2084, Sandia l National Laboratories, September 1989. 1
- 2. D. M. Ericson, Jr. , Editor, T. A. k er, T. T. Sype, M. T. Drouin, W. R. Cramond, A. L. Camp, K. J. Mali. ,ey, and F. T. Harper, " Analysis of Core Damage Frequency: Internal Events Methodology," NUREG/CR.
4550, Volume 1, SAND 86 2084, Sandia National Laboratories, January 1990.
- 3. J. M. Criesmeyer and L. N. Smith, "A Reference Manual for the Event Progression Analysis Code (EVNTRE)," NUREG/CR.5174, SAND 88 1607, Sandia National Laboratorien, September 1989,
- 4. U.S. Nuclear Regulatory Commission, " Severe Accident Risks: An Assessment for Fivo U.S. Nuclear Power Plants," Second Draft for Peer Review, NUREG 1150, June 1989.
= se -
1 l 2.102 \ a l
- . _ . - ,--~-.-w._ , . , . , . ,,, , .3
) i i ) i i j
- 3. RADIOLOGICAL SOURCE TEi0i ANALYSIS ,
The source term is the information passed to the next analysis so that the , offsite consequences can be calculated for each group of accident I progression bins. The source term for a given bin consists of the release fractions for the nine radionuclide groups for the early release and for , the late release, and additional information about the timing of the releases, the energy associated with the releases, and the height of the - releases. Source term analysis is performed by a relatively small computer code: ' GGSOR. The aim of this code is nel to calculate the behavior of the fission products frorn their chamical and physical- properties and the flow < and temperature conditions in the reactor and the containment. Instead, the purpose is to represent the results of the inore detailed codes that do consider these quantities. A niore cornplete discussion of the source terin analysis, and of CGSOR in , j particular, may be found in NUREG/CR 5360.* The rnethods on which GGSOR is based are presented in INREG/CR.4551 Volume 1, and the source term issues
, considered by the expert panels are described more fully in INREC/CR.4551, Volume 2, Part 4.
i Section 3,1 summarizes the features of the Grand Gulf plant that are : important to the magnitude of the radionuclide release. Section 3.2 presents a brief overview .of the GGSOR code, _ and Section 3.' presents the i results of the source term snalysis. Section 3.4 discusses the partitioning of the thousands of source terms into groups for the consequence analysis. Section 3.5 concludes this section with a summary of ' the insights gained from the source term analysis. 3.1 Grand Gulf Fentures Imqortant to the Source Term Analysis Grand Gulf Unit 1 is a boiling water reactor.6 (B W 6) that is housed-in a i Mark III contaitunent. The contaitunent is a reinforced concrete structure with a steel liner. The RPV is located inside the drywell which is in turn surrounded by the containment structure. The drywell volume communicates to the wetwell volume through the suppression pool. The primary barrier between the radionuclides released from the core and the outside environment is the contain.sent structure. The containment . structure has a design pressure of 15 psig and an assessed mean failure pressure of 55 psig. Because of this relatively low- failure pressure
- (relative to the loads that are imposed on it during the course of the tsccident), it was determined during the accident progression anklysin that +
the contaituent is likely to fail during accidents that progress to core damage. In fact, the containment fails early in roughly half of the : accident progressions analyzed. The drywell structure is considerably i'
- li . N . Jow. W. B. Murfin, and J. D. Johnson, "XSOR Codes Users Manual,"
NUREG/CR 5360, SANDB9.'943,.Sandia National Laboratories, (unpublished). ' 3 ,.1 o
I s I stronger than the containment. The design pressure of the drywell from internal pressurization is 30 psid. Nevertheless, the drywell is e till j susceptible to the loads that occur from hydrogen combustion events and from pressurization events accompanying vessel breach. Of the ace.idtnts that result in early containment failure, half also involve early drywall failure. Although the results of this study indicate that the containment is likely to fail, there are a number of plant characteristics that help to reduce the amount of radionuclides that can potentially be released to the
- environment. Because of the suppression pool's ability to effectively trap
- radionuclides, it provides the potential for substantial mitigation of the source terms in severe accidents. In addition to the suppression pool, other features that can potentially reduce the source term are the containment sprays and the reactor cavity pool.
There are two pathways by which radionuclides enter the suppression pool. 2 The first pathway is through the SRV tail pipes.- Because -the dominant centributors to the core damage frequency were transient initiated events (i.e., LOCAs were not analyzed in the accident progression analysis) the in vessel releases exit the vessel via the steam lines, pass through the SRV tail pipes, and are then discharged into the suppression pool through the T quenchers at the end of the tail pipes. For the in vessel releases to bypass the suppression pool a SRV tail pipe vacuum breaker must stick open during core damage And the drywell must be failed. If the drywell is not failed, the releases will enter the drywell volume and then will . be directed to the suppression pool via the horizontal vents. These horizontal vents are the second pathway for radionuclides to enter the suppression pool. If the drywell is intact, the ex vessel releases will also enter the suppression pool via this pathway. The first pathway is more effective than the second pathway at trapping radionuclides. However, the second pathway still offers a significant mechanism for mitigating the source term. The containment sprays can also be effective at reducing the amount of airborne radionuclides. Because the dominant PDSs are short term SB0s, the sprays are generally not on before core damsge. The unavailability of the sprays early in tne accident is not particularly important because as menticned previously, the majority of- the in vessel releases pass through the suppression pool. In the dominant short term SB0 PDS it is likely that the sprays will be on after vessel breach and, therefore, any release from CCI will be scrubbed. The decontamination factor (DF) associated with the sprays is roughly the same as the DF associated with the suppression pool when the radionuclides enter through the horizontal vents. J The Orand Gulf reactor cavity is roughly a right cylindrical volume that is located directly below the RPV, This volume is large enough to contain the core debris that is released from the RPV should vessel breach occur. (flowever, energetic events- such as DCll and ex vessel steam explosions can-disperse core debris outside the cavity.) Thus, unlike a Mark I containment, the core debris generally remains in the reactor cavity. Because of the geometry of-the Mark 117 containment, it is likely that the-3.2
i l cavity will contain water at the t iine of vessel breach. Water can onter the drywell when pressurization events in the wetwell depress the suppression pool suf ficiently such that water is pushed up over the weir wall. The amount of water in the drywell depends on whether the upper water pool has been dumped and on the transient pressurization of the c ontaitunent . During long term PDSs leaking equipment (e.g., recirculation pumps) can also be an important source of water. Water in the drywell can enter the cavity either through the drain in the drywell floor or through a door in the pedestal wall. The presence of water in the cavity is itoportant for threo reasons. First, if there is a large amount of water in the cavity it is possible that the core debris that is released from the cavity will be cooled and, therefore, CCI will not be initiatsd. Second, if CCI is initiated following vessel breach and the cavity contains water, the pool above the core dobris will scrub the CCI releases. Third, ex-vessel steam explosions at vessel breach are possible if the cavity contains water. An ex vessel eteam explosions will increase the amount of airborne radionuclides in the drywell. The first two effects of cavity water initigate the source term. The last effect increases the radionuclide release. Thus, the presence of water can be both beneficial and detrimental. 3.2 Descriplion of the GG1glLQpA This section describes the manner in which the source term is computed for each accident progression bin (APB). The source term is more than the fission product release fractions for each radionuclide class; it also contains information about the timing of the release, the height of the release, and the energy associated with the release. The next subsection , presents a brief overview of the parametric model used to calculate the e source terms. Section 3.2.2 discusses the model in some detail; a complete discussion of GCSOR may be found in Reference 1. Section 3.2.3 presents the parameters sampled in the nource term portion of this analysis. 3.2.1 Overview of the Paramerrie Model GCSOR is a fast-running, paramotric computer code used to calculate the source terms for each APB for each observation for Grand Gulf. As there are typically a few thousand bins for each observation, and 250 observations in the sample, the need for a source calculation inethod that requires a minituum of computer time for one evaluation is obvious. GGSOR is nol designed to calculate the behavior of the fission products from their basic chemical and physical properties and the flow and temperature conditions in the reactor and the containment. The purpose of GGSOR is to i provide a framework for integrating the results of the more detailed codes l that do consider these quantities. Since many of the parameters CCSOR l utilizes to calculate the release fractions were determined by a panel of experts, the results of the detailed codes enter GGSOR " filtered" through the experts. The 60 radionuclides (also referred to as isotopes, or fission products) considered in the consequence calculation are not dealt with individually in the source tona calculation. Some different elements behave similarly 3.3
i enough both chemically and physically in the release path that they can be considered together. The sixty isotopes are placed in nine radionuclide l classes as shown in Table 3.2-1. It is these nine classes which are I treated individually in the source term analysis. Table 3.2 1 l Isotopes in Each Radionuclide Release Class _Relense Class isototas Included
- 1. Inert Cases Kr 85, Kr 85M, Kr 87, Kr 88, Xe 133, Xe 135 1
- 2. Iodine I 131, 1 132, 1-133, 1 134, 1 135
- 3. Cesium Rb 86, Cs 134 Cs 136, Cs 137
- 4. Tellurium Sb 127, Sb 129 Te-127. Te 127M, Te 129 Te 129M, Te-131M, Te 132
- 5. Strontium Sr 89, Sr 90, Sr 91, Sr 92
- 6. Ruthenium Co 58 Co 60, Mo 99 Tc 99M, Ru 103, Ru 105, Ru.
106, Rh 105
- 7. 1.anthanum .Y 90, Y 91, Y-92, Y 93, Zr-95, Zr 97, Nb 95, l.a .
140, La 141, La 142, Pr 143, Nd 147, Am 241, Cm-242, Cm 244 8, Corium Cc 141, Ce-143, Ce 144, Np 239, Pu 238, Pu 239, Pu-240, Pu 241
- 9. Barium Ba-139, Ba 140 3.2.2 Description of GCSOR Since the consequences will generally depend on the timing of containment failure, CGSOR considers three time regimes in .which the containment can fail: before vessel breach, at or near the time of vessel breach, and late in the accident. Furthermore, CGSOR considers two releases from the containment. The first release occurs roughly at the time of containment failure (assuming the containment fails after core damage). The second release begins after the first release has finished (unless CCI initiation ,
is delayed in which case the second release is also delayed). When . the ' containment fails before vessel breach, the first release-is due to fission - products that escape from the fuel while the ' core is still in the RPV (i.e., in vessel releases). For this case, the second release -includes fission products that are released at the. time of vessel breach and after vessel breach. Releases after vessel breach include fission products from-CCI releases, material revolatiliced from the RPV after vessel breach and iodine released from the suppression pool (and in some cases the RPV cavity 3.4
- _ _ _ _ _ - _ _ - _ _ - - - .)
l vater). These releases will be referred to as the late releases. When the l containment fails around the time of vessel breach the first release i includes in vessel releases as well as fission products that are released l at the time of vessel breach. The second release is due to the late l releases. For situations where the containment fails many hours after I vessel breach, both releases consist of in vessel releases, fission products released at vessel breach, and the late releases. The timing and duration of these releases depend primarily on the PDS and the time and mode of containment failure. For radionuclide class 1, the basic parametric equation for CCSOR has the following form: ST i (Eq. 3.1)
- FCOR3 *FVESi *(RELF1 + RELF2 + RELF3)*FCONVg
+ VBPUF *(RELF4 + RELFS)*FCONCi i
+ (1 FCORg VBPUFi )*FLV*FHPE*FDCilg*(RELF6 + RELF7)*FCONC 3
+ (1 FCORg VBPUFi )*FLV*FHPE*FEVSEi *(RELF6 + RELF7)*FCONCi
+ (1 FCOR VBPUF )*FLV*XCCl*FCCIg*(RELF8 + RELF9)*FCONCi i
4 FCORg*(1 - FVESi )*FREVOL i *(RELF10 + RElfil)*FCONC i 1-2, 3, & 4 ONLY)
+ [FLTI1*POOLI + FLT12*CAWI*(RELF12 + RELF13))*RELF14 where RELF1 - TTIJ*FPLBYE/DFSPRV i RE1J2 - ITLP*(1 FPLBYE)/ MAX (DFCPA ,DFSPRVt )
i RELF3 - (1 ITLP)/ MAX (DFVPA ,DFSPRV ) 3 g RELF4 - FPLBYP/DFSPRCi REISS - (1 FPLBYP)/ MAX (DFCPAi ,DFSPRC3 ) RELF6 - FPLBYD/DFSPRC3 RE1J7 - (1 FPLBYD)/ MAX (DFCPAi ,DFSPRCi ) RE1J8 - FPLBYC/ MAX (DFCAVi,DFSPRC3 ) RELF9 - (1 FPLBYC)/ MAX (DFCAVi ,DFCPAi ,DFSPRCg ) RELF10 - FPLBYC/DFSPRC i '- RE1J11 - (1 - FPLBYC)/ MAX (DFCPA i,DFSPRCi ) RE1512 - FPLBYC/DFCPA i RE1513 - (1 - FPLBYC)/DFCPAi RE1J14 - FCONC i if no containment failure
- 1.0 if containment failure XCCI FHPE if DCH or ex vessel steam explosion occurs j
- 1.0 if neither DCH nor ex vessel steam explosion occurs, l
The first summation term on the right side. of Equation 3.1 represents the in-vessel releases. The second term describes the puff release at vessel breach. The third term represents the DCH release. The fourth term - l represents the ex vessel steam explosion release and is: mutually- I exclusive with the third term (i.e. , the experts said if DCH occurred, l then the ex vessel steam explosion release should not be considered l separately). The fif th term represents the _ CCI release. The sixth term ' is the revolatilization release from the reactor coolant system after vessel breach and is for I, Cs, and To classes only. The last term 3.5
represents the late iodine release from the suppression pool and reactor cavity water after the containment fails. The definitions of the various parameters in Equation 3.1 are as follows: CAVW1 - fraction of initial iodine core inventory scrubbed by the cavity water during CCI release DFSPRC 3 - scrubbing decontamination factor for sprays acting on species i released into containment after vessel breach DFSPRV 3 - scrubbing decontamination factor for sprays- acting on species i released into containment from the vessel before vessel breach DFCAV 3 - scrubbing decontamination factor for aerosol species i released into cavity water during CCI release DFCPA g - scrubbing decontamination factor for aerosol species i flowing from drywell to the suppression pool DFVPA g - scrubbing decontamination factor for aerosol species i flowing from the vessel to the suppression pool i FCC1 g - fraction of material released from the melt during molten CCI FCONC i - fraction of species i released from containment for CCI and other releases after vessel breach, not including the effects of scrubbing by pools and sprays FCONV - fraction of species i released from containment for material released into containment before vessel 1 breach, not including the effects of scrubbing by '~ pools and sprays FCOR i - fraction of initial inventory of species i released from the fuel prior to vessel failure FDCH 3 - fraction of radionuclide in the portion of the core involved in direct containment heating that is released to the drywell at vessel breach, FEVSE i fraction of radionuclides in the portion of the core involved in an ex vessel steam explosion that is released to the drywell at vessel breach FHPE - fraction of core material leaving the vessel that participates in either the direct containment heating or the steam explosion and therefore is not available for molten CCI release later 3,6
1 i FLV - fraction of the core material that leaves the vessel l after the vessel breach FREVOL t - fraction of the core material that is deposited on the surfaces of the-reactor vessel and structural materials that is revaporized and released in the drywell after vessel breach FPLBYC - fraction of CCI releases that bypass the suppression pool ' FPLBYD - fraction of DCH releases or ex vessel steam releases that bypass the suppression pool FPLBYE - fraction of in vessel releases that bypass the suppression-pool FPLBYP - fraction of puff releases at vessel breach that bypass the suppression pool PTLP - fraction of the in vessel releases that are released into the drywell through-stuck open SRV tailpipe vacuum breaches PVES 4 - fraction of material released from the fuel that is released from the vessel FLTIl - fraction of fodine in the suppression pool that is volatilized and released after vessel breach FLTI2 - fraction of iodine in the cavity water that'is volatilized and released after vessel breach POOLI - fraction of initial core inventory for iodine scrubhed by the pool ST 3 - fraction of the initial core inventory of species i that is ultimately released to the environment. '- VBPUF 4 - fraction of initial core inventory of. species i that is released to the drywell as puff at the time of vessel breach XCCI - fraction of core material that leaves the vessel that participates _in CCI. It is expected that accompanying containment failure a substantial portion-of the enclosure building at Grand Gulf will fail. Thus, no credit is given for retention of radionuclides in the enclosure building. A detailed discussion of this equation is presented in NUREG/CR 5360.* The FORTRAN listing of CGSOR is contained in Appendix 3.
*H. N. Jow, W. B. Murfin, and J.
D. Johnson,. "XSOR Codes Users Manual," NUREG/CR 5360, SAND 89 0943, Sandia National Laboratories, (unpublished). 3.7
l i 1 1
%o '
A+., A g s, s s [ t ;
)'
\\q 1
9q, \ k s% 1 g J
!r --
e m l 3,8 '
Figure 3.21 depicts the parametric equations schematically in terms of a flow diagram. Coming in from the left is all the radioactivity in any radionuclido class. The black arrows represent releases to the environment and the white arrows represent material retained in the RCS or in the containment. This figure i t, read as follows: the first division of the radioactive material is indicated by FCOR. The top branch, indicated by FCOR, represents the fraction released from the core before vessel breach, and the lower branch, an ainount 1-FCOR, represents the amount still in the RCS at vessel breach. The FCOR branch is then split into that which leaves the RCS before or at vessel breach, FVES, and that which is retained in the RCS past vessel breach, 1 PVES. Of the masarial retained in the RCS at ! vessel breach, a fraction FLATE is revolatilized later. Of the i revolatilized fraction, a portion is removed by en61 neered removal mechanisms such as sprays, parameter 1/DFL, and another portion is removed by natural mechanisms such as deposition, parameter FCONRL. The part of the revolatilized fraction that is not removed escapes to the environment as indicated by the top black arrow in Figure 3.2 1. FCONRL is the containment release fraction for the late revolatilization release, and is set equal to the FCONC value for tellurium. When evaluated as part of the integrated risk analysis, CCSOR is run in the
" sampling mode". That is, most of the parameters in the release fraction equations are determined by sampling from distributions for that parameter, and the value for each parameter varies from observation to observation.
Many of these distributions were provided by an expert panel. The equation above contains 25 parameters. Nine of them were considered by the Source Term Expert Panel. An additional 12 parameters were quantified either by the expert panel for the previous draft of this report or internally. The values for four of these parameters (i.e., CAWI , FLV, POOLI , XCCI) are determined by various combinations of previously defined parameters. Many of these parameters in the equation above are determined directly by sampling from distributions provided by a panel of experts, see NUREG/CR-4551, Volume 2, Part 4. Other parameters are derived from such values, and still others were determined internally, see the XSOR document.* 3.2.3 Variables Samoled in the Source Term Analysis The twelve parameters sampled for the source terra analysis are listed in 4 Table 3.2 2. When CGSOR was evaluated for all the bins generated by the APET evaluation for a given observation, all the sampled parameters in CCSOR had values chosen spect fically for that observation. These values were selected by the Latin Hypercube Sampling (LHS) program from distributions that were previously defined. Many of these distributions were determined by the expert panel on source terms. Eight issues were considered by the Source Term Export Panel:
- H. N. Jow, V. B. Murfin, and J. D. Johnson, "XSOR Codes Users Manual,"
NUREC/CR 5360, SAND 89 0943, Sandia National Laboratories, (unpublished). I 3.9
i
- 1. FCOR and TVES
- 2. Ice Condenser DF (not applicable to Grand Gulf)
- 3. Late Releases from the RPV
- 4. FCCI
- 5. FCONV and FCONC
- 6. Late Iodine
- 7. Reactor Building DF (not applicable to Grand Gulf)
- 8. DCil Releases
- Table 3.2 2 Variables Sampled in the Source Tert Analysis Variable Descrintion FCOR Fraction of each fission product group released from the core to the vessel before vessel breach. There are two cases: high and low zirconium oxidation. This parameter was assessed by the Source Term Expert Panel.
FVES Fraction of each fission -product group released from the core which is released from the vessel. There are three cases: short-term SB0 with the RPV at system pressure, short term FB0 with the RPV at low pressure, and ATWS with the RPV at system pressure. This parameter was assessed by the Source Term Expert Panel. FREVO Fraction of the deposited amount of each fission product group in the RPV which revolatilized after vessel breach and released to the drywell. There are three' cases: no water injection after vessel breach and s * ;gh drywell temporature, no water injection aRer vessel broach and low drywell temperature, and water injection to the vessel after vessel breach. This paramet6r was assessed by the Source Term Expert Panel. FCONV Fraction of each fission product group released from containment for material released into containment before vessel breach, not including the effects of scrubbing by pools and sprays. There are six cases: early containment leakage and a subcooled > suppression pool, early containment leakage and a saturated-suppression pool, early containment rupture and a subcooled suppression pool, early containment rupture and a saturated suppression pool, late containment leak, and late containment rupture. This parameter was assessed by the Source Term Expert Panel. l l FDCil Fraction of each fission product group in the core material that I participates in a direct containment heating event (DC11) that is released to the drywell. Given the occurrence of DCil, there is j only one case. This parameter was assessed by the Source Term Expert Pa"c3 ! I I I 3.10 l l
1 1 1 Table 3.2 2 (continued) Variable Description FEVSE Fraction of each fission product group in the core material that participates in an ex vessel steam explosion that is released to the drywall. Given the occurrence of an ex vessel steam explosion, there is only one case. This parameter was not assessed by the Sourcc Term Expert Panel. It is assumed that the relea.,o fractions for the ex-vessel steam explosion phenomena are sufficiently similar to the release fractions associated with DCil that the DCil distributions are also used to quantify this parameter. FCCI Fraction of each fission product group in th~e the core material at the start of CCIs that is released to the drywell. - There are four cases: low zirconium oxidation in the core and no overlaying water, low zirconium oxidation in the core with overlaying water, high zirconium oxi<iation in the core and no-overlaying water, and high zirconium oxidation in the core with ' overlaying water. This parameter was assessed by the Source Term Expert Panel. FCONC Fraction of each fission product group released from the containment for CCI and other releases after vessel breach, not including the effects of scrubbing by pools and sprays. There are six cases: early containment leakage and a subcooled suppression pool, early containment leakage and a saturated suppression pool, early containment rupture and a subcooled suppression pool, early containment rupture and a saturated suppression pool, late containment leak, and late containment
.pture. This parameter was assessed by the Source Term Expert P.i iel .
FLTI This variable in the Ills sample is used for both FLTIl and FLTI2 j (i.e., completely correlated). -These parameters were assessed i by the Source Term Expert Pan (1. FLTII: Fraction of iodine in the suppression pool that is volatilized and released af ter vessel breach. There are two cases: the suppression pool is subcooled and the suppression pool is saturated. FLTI2: Fraction of iodine in the cavity water that is ! volatilized and released af ter vessel breach. There are two l cases: the reactor cavity is flooded and the reactor cavity in wet. I. l 3.11
l Table 3.2 2 (continued). Variable Descriotion t DFPOOL This variable in the LHS sample-is_used for both DFVPA and DFCPA (i.e., completely correlated) . . This ~ issue was .not assessed by ! the Source Term Expert Panel. The distributions for these - parameters were_obtained from the draft report of NUREG/CR 4551,: j Volume 7.. T DFVPA:- Decontamination factor for in vessel releases ..that are released into the suppression pool.
.DFCPA: Decontamination factor for-aerosol releases flowing from- -
the drywell to the suppression pool.- ; DFCAV Decontamination factor for aerosols. released into the cavity ! water the CCI release.: Thio DF is applied when the core debris - 1 is not coolable ano CCI proceeds under water; There are' two cases: _the reactor cavity _is flooded and the reactor cavity is- ! only partially filled with -water. . .This' issue - was not assessed by the Source Term Expert Panel.- . The distributions . for this
~
parameter were obtained from the draft report of NUREG/CR-4551, Volumo 7, t DFSPRAY ihis ' variable ;in the LilS sample is used for both DFSPRV and DFSPRC (i.e., completely _ correlated), _
. This --issue was not-assessed by'the Source Term Expert Panel. The distributions'for these parameters were- obtained from - the draf t repo' re o f- ~
NUREG/CR-4551, Volume 8. -i f DFSPRV: Decontamination- factor for ' sprays _ acting on ' fission - product. groups released =into the containment:from the vessel. DFSPRC: Decontamination factor for sprays acting. on fission. product groups released into the r containment aftar vessel breach. , Two of these issues are notrapplicable; to ' Grand Gulf. For each issue - considered by the expert panel, the result is an. aggregate distribution for the nine rndionuclide release classes .. defined in Table 3~.2-l'. :These distributi: are - not necessarily. discreto. .. Whi"a n the e xperts provided, separate dL lbutions for.all nine classes.for FCOR,--for other' parameters, =i for example. they stated that classes -5 through 9'should' be considered - together as an acrosol class, l The - sampling process ' works ~ somewhat differently for the -source term l analysis than it does for the accident progression analysis. In the source ? term analysis. Lils wasrused only to determine a random number between 0.0 - l I 3.12 'e _ _ . _ _ . . . - __ -_ _ _ _ __ _ __ _m.__ . . _ _ . _
and 1.0 for each parameter to be samplod. The actual distributions are contained in a data file (listed in Appendix B) that is read by GGSOR before execution. The variable identifiers given in Table 3.2-2 are used in several ways _ in the source term analysis. Consider the first variable in Table 3.2-2: FCOR. FCOR in the equation for fission product release is the actual fraction of each fission product group released from the core to the vessel i before vessel breach for the observatiu in question. But, FCOR is also used to refer to the experts' aggregate distributions from which the nine values (one for each radionuclide class or fission product group) for FCOR are chosen. Furthermore, in the sampling process, FCOR is used to refer to the random number from the UlS , which is used to select the values from these distributions. That means that, as used in sampling, FCOR defines a quantile in those distributions. The release fractions associated with this quantile are used in CGSOR as the FCOR values. Thus, in Table 3.2 2, the end use of each variable is S ven i although the actual sampled variable is a random number between 0.0 and 1.0 used to select an actual value. The variables selected by UlS are used to define quantiles in' the parameter distributions; the values associated with these quantiles are used as parameter values in GGSOR. In use, the process works like this. Suppose Uls selects a value of 0.05 for FCOR for Observation 1. Referring to the data tables in Appendix B.2, it may be seen that, for low zirconium oxidation in vessel, the 0.05 quantile values for FCOR are 0.084 for inert gases, 0.009 for I, 0.009 for cesium, etc. There is no correlation between any of the source term variables, but complete correlation within a variable. FCOR is not correlated with FVES, FCONV, or any other variable, but the values for the different cases and for the different radionuclido classes are completely correlated. That is, if the 0.05 quantile value is chosen for iodine for low zirconium oxidation, the 0.05 quantile value is also chosen for all the other radionuclide classes and for all values for high zirconium oxidation.
- As all the source term variables are uniformly distributed from 0.0 tb 1.0, i
and are uncorrelated, there are no columns for this information in Table 3.2-2 as there are in Table 2.3-2. There is a separate distribution for each radionuclide class for each variable in this table unless otherwise noted in the variable description. The different cases for each variable are noted in the description. Not all the cases considered by GGSOR are listed in Table 3.2 2; parame ter - values for other cases are determined internally in GGSOR, often from the values for the cases listed. For example, there is no distribution for FVES for long term SB0s. The value of FVES for the long term SBos was derived from the distributions for other cases. For each parameter that was assessed by the Source Term Expert Panel, the distribution for the parameter, the reasoning that led each expert to his conclusions, and the aggregation of the individual distributions are fully described in NUREG/CR 4551, Volume 2, Part 4. The distributions for the r 3.13
j
.i L remaining parameters are presented in Appendix _ B. A discussion ot hese J parameters may be found-in=NUREG/CR 5360.*
3.3 Results of Source Term Analysis This section prosents the: results - of computing the sourco terms- for the . APBs produced by evaluating the' APET. The APET's . evaluation produced a ! large number of APBs, so, as in :Section 2.5, only a. sample of - the; 'more likely and more important APBs are discussed here. --However, source terms were computed for all the APBs for each of the 250 observations - in L the , sample. The source term is composed of release fractions-- for the- nine-radionuclide groups for .a first and a second release as' well as. release t timin5, release height, and release energy. As discussed:above.-the u urce terms are computed by a fast running parametric computer code, CGSOR. , i For . purposes of readability, all tables andl figures related;to this , subsection appear at the end of-it. Section 3.3.1 presents the results for the internal initiators. .The tables in this section present only a very small portion of the output: obtained by computing source terms for each APB.. More detailed results are contained in Appendix B, and complete listings are available on computer media by' request. i 3.3.1 Results for Internal Initiators-In a manner analogous to Section 2.~5.1, _ the results of the source term analysis for internal ~ initiators Lare presented . for ~each PDS group. !The tables in this section only provide' a: sample oE APBs1and their associated mean source terms for the various PDSs. 3.3.1.1 Results for PDS it Short-Term SBO. As : discustied , in Section-2.s.1.1, this- PDS involves SB01 scenarios where LOSP is recoverabla. - Coolant injection is lost early such that core ; damage occurs in thoshort term and with the vessel at'high pressure-because depressurization did not " have an effect in the . prevention .. of core s damage . (the- operators can depressurize the RPV . during ' core damage)~. If 'offsite power is restored then coolant injection' to the -RPV, containment sprays. and the HIS .are all-available. For this PDS the mean probability that. vessel breach is-averted is 0.32. The mean probability lthat the containment fails t early .(carly is defined as before or around the -time of' vessel breach) :is.0.36.
~
Tchle. 2.51 lists the five most probable APBs for PDS 1,L the L five most probable APBs that have vessel breach,-and-the five most probable APBs that have early CF. Table 3.3 l' lists the mean source ' terms for these same APBs. Although the same bins are shown in both tables, and the- structures of'both tables are roughly analogous, there are some important differences
. in the nature of the inceris1- presented. In Table 2.5-1, the-bin itself-was well defined, i.e., the characteristics of'the bin-did net varyf from q
- H. N. Jow, W . -- B . Murfin, and J. D. Johnson; "XSOR Codes Users Manual,"
NUREC/CR-5360, SAND 89-0943, Sandia National Laboratories, (unpublished). 3.14 l
._ . . - ..- ._ ._ _ . - _ , . _ . _ _ _ , _ ~ .
observation to observation. The only item in the table-that varied from observation to observation was the_ probability of the. occurrence of the bin. itself. Tbus, Table 2.51 lists a- conditional probability- averaged over L the 250 observations in the sample. In Table 3.3 1, the bin is still well defined, but, as- many of the parameters that are used in calculating the
- fission product releaso vary from observation to observation, the cerce term for a specific bin varies with the abservation. 'Thus, the entries in-all columns in Table 3.3 1 except the Order and Bin columns: represent averages over the 250 observations in the sample.
For example, consider .the first- APB in Table 3.31: ABBDDGCCB. Of the 250 observations in the sample,- 75 had non-=cro conditional probabilities for this bin'. As_ source terms. are not computed - for zero probability bins,; there are 75 source terms' associated with APB ABBDDGCCB. - These 75 source terms were summed and then divided by 75 to produce the mean source term i given-in the first two lines of Table 3.3 1. . The most probable.APB, ABBDDCCCB,_ involves accidents.that proceed to vessel breach. Once vessel breach occurs the core debris is released into che reactor cavity and CCI takes place under a pool of water. For this APB the containment is ruptured late in the accident. When the containment fails' , in the rupture mode late in the acuident, Gu50R groups 90% of the radionuclides that are available to be released from the. containment (i.e., those radionuclidea Sat have not been trapped by the water pools or plated out in the vessel or containment) in the first release and the remaining 10% in the second release. The next ~ four most probabic : MBe involve accidents that do not proceed to ves sel breach - (i.e. , no ex vessel releases) and the containment either fails late in the accident or does not ' fail. As a result the releases associated wirb these bins are considerably less than those associated with the most ' probable -APB. When the containment develops a leak late in the J accident , (e.g. , fourth most probable bin - ABEEAFCEB), GGSOR releases 50% of the-radionucifdes_from the containment in tho ' first release and- the remaining : 50% h the second release. . , _ . For APBs that involve accidents that do not proceed to vessel fath.re -but , do . result _ in early containment failure, all of the radionuclides,. except iodine, are grouped in the first release. Iodine that'is released;from the vessel _ that is- not trapped: in the _ suppression pool is - contained in the first release _. .A fraction of the . iodine - that was trapped by the
~
i suppression pool .is subsequently revolatilized from the _ pool and released - ; into the containment. The revolatilized-iodine is grouped in the second-release. The mean source terms in Table 3.31 can be used to compare the. releases-associated with specific APBs. . Ilowever , as- these mean source . terms are typically not calculated over the; same sample elements fine distinctions between source terms associated with dif ferent APBs may'be lost 'in the averaging proccos. For.some of the accident progression 1 bins the release energy _ assigned to the bin was wrong. The release energy affects how' high the releases are 3.15 _ - , _ . . _ ,_ _. _. . . . _ , , _ ~ , _
lofted in the atmosphere. For accidents in which the containment does not fail the release energy should have been set to zero but was inadvertently set to value that is used when the containment fails. The release fractions for these accidents, however, are typically very small and, therefore, the effect on risk is expected to be negligible. For accidents in which the containment is ruptured at vessel breach, the release energy was inadvertently set to zeros Because the plume is not lof ted as high as ' it should have been, the early fatalities may be slightly overestimated for these accidents. The latent cancer fatalities are not particularly sensitive to this parameter and, thus, the effect on this censequence measure is expected to be very minor. Table 3.3 1 presents mean source terms but does not contain any frequency information. In contrast,' Figure 3.31 presents information on both source term size and frequency. The frequency of each PDS is presented in Section 2.2. Figure 3.31 summarizes the release fraction CCDFs for the iodine, cesium, strontium, and lanthanum radionuclide classes. It indicates the frequency with which different values of the release fraction are exceeded, and displays the uncertainty in that frequency. The curves in Figure 3.3 1 are derived in the following manner: for each observation, evaluation of the A"FT pnduco:i a c.un61tional probability for each APB. When multiplied by the frequency of the PDS for that observation, a frequency for the APB is obtained. Calculation of the source term for the AFB gives a total release fraction for each APB. When all the APBs are _ considered, a curve of exceedance frequency vs. release fraction can be plotted for each observation. Figure 3.3-1 is a summary presentation of these curves for the 250 observations in the sample. Instead of placing all 250 curves on one figure, only four statistical-measures are shown. These measures are generated by analyzing the curves in the vertical direction. For each release fraction on the . abscissa,- , there are 250 values of the exceedance frequency (one for each sample l clement). From these 250 values it is-possible to calculate mean, median (50th quantile), 95th quantile, and 5th quantile values. When this is done for each value of the release fraction, the curves in Figure 3.3 31 are obtained. Thus, Figure '3.3-1 provides information on the relationship between the size of the release fractions ascociated with PDS 1 and the frequency at which these release fractions are exceeded, as well as the r l variation in that relationship between the observations in the sample, As an illustration of the information in Figure 3.31, the mean frequency (yr-1) at which a release fraction of 10'5 is exceeded due to PDS 1 is roughly 2.9 x 10'8, 2.4 x 10'8, 2 x 10'8 and 1.7 x 10-e for the iodine, cosium, strontium and lanthanum release classes, respectively. For a release fraction of 0.1, the corresponding mean exceedance frequencies are 4.2 x 104, 8.! 10 8, 1,7 x 10-e and 8.4 x 10'11, respectively. The three quantiles (i.e. .he median, 95th and 5th) provide an indication of the spread between observations, which is of ten large. Typically, the mean curves reach a point where they drop very rapidly and move above the 95th quantile curve. This happens when the mean curve is dominate,d by a few large observations; this often occurs for large release fractions because only a few of the sample observations have nonzero exceedance frequencies 3.16
.ey ---
l for these large release fractions. Taken as a whole, the results in Figure j 3.3 1 indicate that the occurrence of large source terms (e.g., release ! fractions ;t 0.1) in conjunction with PDS 1 is very infrequent (less than 10'5 for iodine and cesium, less than 10-7 for cesium, strontium, and lanthanum). 3.3.1.2 Results for PDS 2: Short-Term SBO. PDS 2 is the same as PDS 1 except that heat removal via the sprays is not available with the recovery of offsite power. For this PDS the mean probability that coolant inj ection is recovered and vessel breach is averted is 0.32. The mean probability that the containment will fail early is 0.36. Table 2.5.1 2 lists the five most probable APBs for this PDS, the five most probable APBs that have vessel breach, and the five most probable APBs that have early containment failure (CF). A discussion of the accident characteristics for these APBs is presented in Section 2.5.1.2. Table 3.3 2 lists the mean source terms for these same APBs. Although the containment sprays are not available in these APBs, the in vessel releases are discharged into the suppression pool where they are subjected to the pool DF. Of the APBs listed in Table 3.3-2 only one bin has a stuck open tail pipe vacuum breaker. The stuck open vacuum breaker allows a fraction of the in vessel releases to enter the drywell rather then being discharged directly into the suppression pool. However, in this bin the suppression pool is not bypassed until late in the accident and, therefore, the in-vessel releases that enter the drywell still pass through the pool (i.e. , via the horizontal vents) before entering the containment volt . For the APBs that involve vessel failure, the core debris released into-the cavity is either cooled or CCI takes place under water. Thus, any ex-vessel releases are also scrubbed. Thus, although the containment sprays are not available in this PDS the releases associated with the APBs presented in Table 3.2 2 are still mitigated by the suppression pool and tne cavity water. [ Figure 3.3-2 summarizes the release fraction CCDFs for PDS 2. 3.3.1.3 Results for PDS 3: Short-Term SBO. PDS 3 is the same as PDS 1 except that heat removal via the sprays is not available with the recovery of offsite power and the only injection system available with the recovery of offsite power is the condensate system. For this PDS. the mean probability that coolant injection is recovered and vessel breach is averted is 0.21. The mean probability that the containment fails early is 0.44. ( l Table 2.5.1-3 lists the five most probable AFBs for this PDS, the five most i probable APBs that have vessel breach, and the five most probable APBs that have early containment failure (CF). A discussion of the accident characteristics for these APBs is presented in Section 2.5.1.3. Table 3.3-3 lists the mean source terms for these same APBs. For the APBs listed in Table 3.3-3 the in-vessel releases are scrubbed by the suppression pool and any ex-vessel releases, should they occur, are scrubbed by either the suppression pool or the water in the reactor cavity. Only.one.of listed , APBs involves early failure of both the containment and the drywell. But l l 3.17 i j
for this bin the there are no stuck open tail pipe vreuum breakers (i.e., in vessel releases directed to the suppression pool) and vessel breach is averted (i.e., no ex vessel releases). Figure 3,4-3 summarizes the release fraction CCDFs fer PDS 3. 3.3.1.4 Results for PDS 4: Lone-Term SBO, This FDS involves station blackout scenarios where LOSP is recoverabic. Coolant injection is lost late such that core damage occurs in the long term and with the vessel at low pressure. If offsite power is restored, then coolant injection to the RPV, containment sprays and the llIS are all available. Because this is a slow SB0 (i.e., core damage occurs a 12 h), this PDS -has a much lower probability of recovering offsite power than did the fast SB0 in which core damage occurs in approximately 1 h. For this PES the mean probability that coolant injection is recovered and vessel breach is -avorted is only 0.05. The mean probability that the containment fails early is 0,65. Table 2.5,1-4 lists the 10 most probable APBc for this PDS and the five most probable APBs that have early containment failure and early supprossion pool bypass. A discussion of the accident characteristics for these APBs is presented in Section 2.5,1.4. Table 3,3-4 lists the mean sourco terms for these same APBs, In all of the 10 most probable bins vessd breach occurs, the RPV is at low pressure, and an ex vessel steam e.plosion, which involves a small amount of :ho core, occurs at vessel breach, Containment sprays are not available in any of the 10 mos t probable bins, For these APBs the in-vessel rcleases are directed to the suppression pool and either the core debris in the cavity is cooled or CCI takes place under water, lloweve r , in three of these bins both the containment and the drywell are ruptured early in the accident and, therefore, the releases at vessel breach (i.e., releases associate with DCil) are not scrubbed by either the pool or the sprays. In all of the five most probable bins that have early containment failure and early suppression pool bypass vessel breach occurs with the RPV at low pressure-followed by an ex vessel steam explosion. Thero are no stuck-open tail pipe vacuum breakers in these five bins so all of the in vessel relcases pass through the suppression pool, However, because there is early-drywell failure, a pathway is established which hypasses the suppression pool. Although the releases at vessel breach (i.e. , releases associated with an ex vessel steam explosion) are not scrubbed by either the suppression pool or the sprays , the core debris in the reactor cavity is either cooled or l CCI takes place under water. l Figure 3.3 4 summarizes the release fraction CCDFs for PDS 4. 3.3.1.5 Results for PDS 5: Lonn-Term $BO, PDS 5 is the same as PDS 4 except that heat removal via the sprays is not available with the recovery of offsite power. 11cweve r , because there is a low probability of recovering offsite power in this PDS this difference is not very important. For this PDS the mean probability that coolant injection is recovered and vossal breach is averted is only 0.05. The mean probability that the containment fails early is 0.64. 3,18
Table 2.5.1 5 lists the 10 most probable APBs for this PDS . and the five most probable APBs that have early containment failure and early suppression pool bypass. A discussion of the accident characteristics for these APBs is presented in Section- 2.5.1.5. Table 3.3-5 lists the mean source terms for these same APBs. In all of the 10 most probable bins vessel breach occurs, the RPV is at low pressure, and an ex vessel steam explosion, which involves a small amount of the core, occurs at vessel breach. Containment sprays are not available in any of the 10 most probable bins. For these APBs the in vessel releases are directed to the suppression pool and either the core debris in the cavity is cooled or CCI takes place under water, lloweve r , in three of these bins both the containment and the drywell are ruptured early in the accident and, therefore, the releases at vessel breach (i.e. , releases associated . with DCll) are not scrubbed by either the pool or the sprays, In all of the five most probable bins that have early containment failure and early suppression pool bypass vessel breach occurs with the RPV at low pressure followed by an ex-vessel steam explosion. There are no stuck open' tail pipe vacuum breakers in these five bins so all of the in vessel releases pass through the suppression pool, llowever, because there is early drywell failure, a pathway is established which bypasses the suppression pool. Although the releases at vessel breach (i.e. , releases associated with an ex vessel steam explosion) are not scrubbed by either the suppression pool or the sprays, the core debris in the reactor cavity is either cooled or CCI takes place under water. F18 ure 3.3-5 summarizes the release fraction CCDFs for PDS 5. 3.3,1.6 Results for PDS 6: 'Lont Term SBO. PDS 6 is the same as PDS 4 except that neither coolant injection to the RPV nor the containment sprays are available during the accident. Thus, because there is no coolant injection to the vessel, the mean probability of vessel breach is 1.0. The mean probability that the containment fails'early.is 0.68. Table 2.5.1-6 lists the 10 most probable APBs for this PDS and the five most probable APBs that have early containment failure and 'sarly suppression pool bypass. A discussion of the accident-characteristics for these APBs is presented in Section 2.5.1.6. Table 3.3-6 lists the mean source terms for these same APBs. In all of the 10 most probable bins vessel breach occurs with the RPV at low pressure followed by an ex-vessel steam explosion that involves a small fraction of the core.~ The l containment sprays do not operate during the accident but because there are i no stuck open SRV tail pipe vacuum breakers all of the in vessel releases , are still scrubbed by the suppression pool. In all of the 10 most probable ; bin the core debris released from the vessel is cooled and there are no CCI l releases. Figure 3.3-6 summarizes the release fraction CCDFs for PDS 6. 3.3.1.7 Results for PDS 7: Short-Term SBO. This PDS involves station blackout (without any de power) scenarios' where LOSP is not recoverable. Coolant injection is lost early such that core damage occurs in the short term and with the vessel at high pressure and depressurization is not l l l 3.19 i
-- . . - . . - - - - - - . . . . .- - . - --- - = - -- -_-_ _
possible. Since offsite power- is . not recoverable, neither coolant-injection nor containment sprays are available during the accident. In a: small- fraction of these accidents (4%) a SRV will stick open' and depressurize the RPV. Once the RPV has been depressurized, the firewater i system can be used to provide coolant injection to the RPV. l Thus, the.mean probability that vessel breach is: averted is only 0.01. The mean probability that the containment fails early is 0.60 Table 2.5.17 lists - the 10 most probable ' APBs for - this PDS . and'.the five most probable APBs that have : carly containmenti failure and early - suppression pool bypass. ; A discussion of the accident characteristics for these APBs is presented in Section . 2.5.1,7.- Table 3 3-7 lists the mean = . source terms for these same ~ APBs. In: all of- the 10 most probable bins, vessel breach occurs with the RPV at ht hSpressure followed _by a DCH event that involves a small fraction'of the core. The containment sprays do not operate during the accident but because' there are no stuck open SRV tail pipe vacuum ' breakers all of the in vessel releases are still e scrubbed by: the suppression pool, Furthermore,_the core, debris that accumulates in the reactor cavity is cooled by water and, thus, there are no CCI releases. However, the drywell does fail early - in two of these bins and, - therefore, the releases at vessel breach (i.e., releases associated with DCH) are not scrubbed by either the pool or the sprays. I j Figure 3.3-7 summarizes the release fraction CCDFs for PDS 7. 3.3.1.8 }tesults for PDS 8 Lone-Term - SBO. This PDS involves SB0 (without any de power) scenarios where LOSP is not recoverable. . Coolant injection _ is lost late such that core damage occurs in the long term 'and with the vessel at high pressure and depressurization is not possible. Since offsite power is not recoverable, neither coolant injection nor containment sprays are available during, the accident. .Because there is no - coolant injection to the RPV, the probability of vessel breach is l' 0. The uean probability that the containment fails early is 0.54. Table 2.5.1-8 lists the = 10 :most probable APBs for this PDS and the -five
.most probable APBs- that have early containment failure- andf early -
suppression pool bypass. A discussion of the accident characteristics for these APBs is presented in Section 2.5.1.8. Table ' 3.3 8 lists the mean-source terms for these same APBs. In all.of the 10 most-probable bins, vessel breach occurs with the RPV at high pressure followed by a DCH event that involves a small fraction of the core. The containment sprays'do not operate during. the accident. There is. only one bin that has ,a stuck open - tail pipe vacuum breaker; however for thisEbin the drywell does not! fail. Thus, all of the in-vessel releases are scrubbed by-the suppression pool. Although the in vessel releases for the ninth most probable bin. are scrubbed by the - suppression pool, the ex-vessel releases do lnot benefit-from a pool DF. In this APB both the drywell and the containment are ruptured early in the accident. Thus, the radionuclides released at vessel' breach (c< , g. , from DCH) and the releases- from. CCI bypass the suppression pool, furthermore, the sprays are not available' in this PDS and in this - APB CCI does not take place under a pool of water. _ Thus, the ex vessel - releases are not mitigated by the sprays, the suppression pool; or the 3.20
cavity pool. Therefore, it is not surprising that the mean release fractions associated with this APB tend to be higher than the release fractions for the other nine bins. Figure 3.3-8 summarizes the release fraction CCDFs for PDS 8. 3.3.1.9 Results for PDS 9: Short-Term ATVS. This PDS involves ATWS transient scenarios. Coolant injection is lost early such that core damage occurs in the short term and with the vessel at high pressure because the operator failed to depressurize. The low pressure injection is recoverable with reactor depressurization. The containment sprays are available during the accident. The mean probability that coolant injection will be restored to the RPV and vessel breach will be averted is only 0.04. The mean probability that the containment fails early is 0.67. Table 2.5.1-9 lists the 10 most probable APBs for this PDS and the five most probable APBs that have - early _ containment failure and early suppression pool bypass. A discussion of the accident characteristics for those APBs is presented in Section 2.5.1.9. Table 3.3 9 lists the mean source terms for these - same APBs. In the 10 most probable bins , vessel breach occurs with RPV at high pressure. At vessel breach either a DCH event occurs (nine bins) or there is an ex vessel steam explosion (one bin). In all but one of the 10 most probable bins the containment fails at vessel breach. In all of these 10 bins the in vessel releases are directed to the suppression pool. The drywell fails early in three of these APBs and, therefore, the releases at vessel breach bypass the suppression pool, lloweve r , the containment sprays are operating around the time of vessel breach. There are no CCI releases in all but one of these bins and in the bin that CCI does occur, the releases are scrubbed by a flooded cavity. Figure 3.3 9 summarizes the release fraction CCDFs for PDS 9. 3.3.1.10 Results for PDS 10: Long-Term ATUS. This PDS involves ATWS transient scenarios. Coolant injection is lost late such that core damage occurs in the long term ' with the vessel at high pressure becatrse of operator error. Low pressure injection is recoverable with reactor depressurization. The containment sprays are available during the accident, The mean probability that coolant injection will be restored to the RPV and vessel breach will be averted is only 0.01. The probability that the containment fails early is 1.0. The containment always fails in this PDS because the energy dumped into the suppression pool from the RPV during an ATWS transient exceeds the capacity of the RHR s; stem which results in a large buildup of steam in the containment. Table 2.5.1-10 lists the 10 most probable APBs for this PDS and the five most probable APBs that have early containment failure and early suppression pool bypass. A discussion of the accident characteristics for these APBs is presented in Section 2.5.1.10. Table 3.3-10 lists the mean source terms for these same APBs. In all of the 10 most probable bins, vessel breach occurs with the RPV at high pressure followed by a DCll event that involves a small fraction of the core. In all of these bins the containment fails early; however, there is coincident drywell failure in 3.21
)
i I
'1 l
l only one of these bins. The containment sprays operate before vessel breach in all of these bins and continue to operate -during the entire accident in all but two of these bins. In these APBs both in vessel releases and the ex vossel releases are. scrubbed by either the suppression j pool or the containment sprays. Figure 3.3-10 summarizes the release fraction CCDFs for PDS-10. 3.3.1.11 Results for PDS II: Short-Term T2 This PDS involves transient scenarios where the PCS is lost (T2). Coolant injection is lost early such that core damage occurs -in the short term with the vessel at ' high - pressure because L of_ operator error. The - containment spress are available during the accident. The mean probability;that coolant injection will be restored to the RPV and vessel breach will be averted is only 0.05. The mean probability that the containment fails early is 0.56. Table 2.5,1-11 lists the 10 most probable APBs for this PDS and the five most probable APBs that ' have early containment failure and early suppression pool bypass. A discussion of the accident characteristics for these APBs is presented in Section 2.5.1.11. . Table 3.3-11 lists the mean source -terms for these same APBs. In all of the 10 most probable bins, vessel breach occurs with the RPV at high pressure followed by a DCH event that involves a small fraction of the core. The containment' fails early in all but two of these bins. Only two of the'10. bins have coincident early containment failure and early drywell failure. The bins that have early drywell failure do~not have any - stuck open tail pipe . vacuum ' breakers. Thus, in the 10 most probable bins the in vessel releases are scrubbed by the suppression pool. Furthermore, the containment sprays operate around the time of vessel breach and there are no CCI releases in-all but one of - these b ins . Thus, the ex-vessel releases ' are scrubbed by either the suppression pool, the sprays, or the water in the reactor cavity. Figure 3.3 11 summarizes the release fraction CCDFs for PDS 11. 3.3.1.12 Results for PDS 12: Lonn-Term T2. PDS 12 is--the same ar PDS ! 11 except that core damage occurs in the long-term. The mean probability that coolant injection will be restored to the RPV and vessel breach will be averted is only_ 0.05. The mean probability that- the containment fails early is 0.56. , Table - 2. 5.1 12 .lis ts the 10 most probable APBs for this PDS and the five most probable APBs that have early containment- failure andcearly suppression pool bypass. A discussion of the accident characteristics for these APBs is presented in Section 2.5.1.12. Table 3.3-12 lists the mean
-source terms for these same APBs. . In all of the- 10 most probable bins, vessel breach-occurs with the RPV at high pressure followed by a DCH event that involves a small fraction of the core. The containment fails early:in all but two of these bins. Only two of the 10 bins have coincident' early containment failure and early drywell failure. The bins that.have early drywell failure dc, not have ' any stuck open tail pipe vacuum breakers.
j Thus, in the 10 most probable bins the in vessel releases are scrubbed by the suppression pool. Furthermore, the containment sprays operate _around 3.22 l l 1 ! l I i
. - . . - , m_ . _ . - _ _ , . _ _ , , ~ . , , _, _ , , , , . - - ~
l l l i l l the time of vessel breach and there are no CCI releases in all but one of these bins. Thus, the ex vessel releases are scrubbed by either the suppression pool, the sprays, or the water in the reactor cavity. Figure 3.3 12 summarizes the release fraction CCDFs for PDS 12. 3.3.1.13 Results for Generflized Accident Procression Bins. The preceding twelve subsections presented the source term results by PDS group. It is also possible to group the source terms in other ways. These other groupings are called generalized APBs. These generalized APBs are generated by sorting all of the bins from the 12 PDS on attributes of the accident. The generalized bins are composed of essentially four characteristics: occurrence of vessel breach, timing of containment failure, timin6 of suppression pool bypass, and the availability of the containment sprays. These generalized APBs are listed roughly in decreasing order of the severity of the source term (i.e. , release timing and release fractions). (The last two bins are an exception to this ordering scheme). A description of these reduced bins is presented in section 2.4.3. Figure 3.313 shows the variation of the exceedance frequency with release fraction for the iodine, cesium, strontium, and lanthanum radionuclide classes for all the APBs in which the vessel fails and both the containment and drywell fail early in the accident. In this bin the containment sprays are not available. Although the in vessel releases will generally be directed to the suppression pool, the releases at vessel breach and any ex-vessel releases will not be subjected to the DF associated with either the pool or the sprays. If the reactor cavity contains water, however, any CCI releases will be scrubbed by the overlaying pool. Figure 3,3 14 shows the variation of the exceedance frequency with release fraction for all the APBs in which the vessel fails and both the containment and drywell fail early in the accident. This generalized bin is similar to generalized bin used in Figure 3,3-13 except that in these accidents the sprays ate available. The release fractions associated with this bin tend to be lower than the release fractions presented .in Figure i 3.3-13. l I Figure 3.3-15 shows the variation of the exceedance frequency with release
- fraction for all the APBs in which the vessel fails, the containment fails early, and the drywell falls late in the accident. Failure of the drywell Inte in the accident can be induced by failure of the reactor pedestal caused by concrete crosion from CCI. Thus, for this generalized APB both the in-vessel releases and the release at vessel breach are directed into the suppression pool. Initiation of CCI is relatively likely in this APB.
Furthermore, because of the late failure of the drywell, the CCI release will bypass the suppression pool. This APB has a fairly low frequency of occurrence. Figure 3.3 16 shows the vvriation of the exceedance frequency with release l fraction for all the APBs in which the vessel fails and the containment fails early in the accident. In these APBs the drywell does not fail 3.23
during the accident. In this APB - both the inivessel and: the ex-vessel-releases are directed'to the suppression pool. Figure 3.3 17 shows the variation' of the exceedance- frequency with release fraction for all-- the APBs in which' the vessel fails and the containment fails late in - the accident. This generalized APB has a relatively;;high - _ - frequency of occurrence and includes-a variety of different accidents (i.e.- i those-with.and without drywell failure). Figure 3.3 18 shows the-variation of the exceedance frequency with release fraction for all the APBs in which the' vessel fails and the' containment is-vented during the. accident. Figure 3,3 19-shows 5he variation of the exceedance frequency with release fraction for all the APBs in which the vessel failed but' the containment remained _ intact throughout the accident. -Because :in these = APBs' there is-only nominal leakage from the containmentf the rele' ase fractions tend to be quite low. It should be y- pointed that - some; of the ; APBs in this group involve accidents in which the' containment fails even though vessel breach
~
is averted. Figure 3.3-20 shows the variation of the'exceedance' frequency with release fraction for all'the APBs in which the vessel breach is' averted. Although the vessel does not fail in these APBs, "some of - these: bins- involve early containment failure. Thus,- the release fractions for; these - APBs - are , typically larger than - the release fraction ; presented ' in the . previous - figure. a 3.3.1.14 - Summarv.. When- all _ the. types 7of accidents from internal-initiators at Grand Gulf- are- considered' together,_ the exceedance frequency plots shown in Figure . 3.3-21 ' are obtained. A" plot is not showni for Othe noble-gases since almost all of the noble gases-(xenon and~ krypton)_in the core are- eventually released to :the environment' whether ' the containment: fails or not. The mean frequency _ of exceeding a releaso-- fraction- of -0.10 ~ for iodine and cesium is 'on the ' order of '10 8/ year -and' for'. tellurium and strontium it is on the order of 107/ year. Thes second- sheet of- Figure ; 3.'3-16 shows the release fractions ~ for ruthenium, o lanthanum, . cesium p and.- barium, which -are of ten treated-' together as aerosol ! species . TheLmean frequency of ' exceeding a release fraction of 0.01 for rutheniumFlanthanum, and cesium is = on the order of 104/ year. The releases for the barium class are - slightly? higher than- those for the other Ethree aerosol radionuclide-classes. i I 3,24 i
,-- y- -v- p r , w- v w , ,y , , . , n ~ > , - - - - ~,~+ae-- v v r -n ,a n+w--~~n- m e vre
i ( Table 3.3-1 1 Mean Source Term for' Grand Gulf l Internal : Initiators . PDS 1: Fast SBO. - j
~
-'We'rning Release' Release Release Release Tractiom Time' Elevation Energy Start Duration
' Order ' Bin (s) (m) (W) - (s) (s) NG I ,,,,$ s_ To Sr Ru La' Ce ?e "l
. Fiva Most Probable Eins 't 1 ABBDDGCCB 3.6E+03 3.2E+01 3.0E+07 '5.0E+04 1.8E+02 9.0E-01 1.5E-01 7.3E-03 7.9E-03 3.9E-03. 2.0E-04 2.8E 5.7E-04' 3.2E [
1.4E+05 :5.1E+04 1.4E+04 1.0E-02 1. 7E-02. . 8.1E-04 8.7E-04 4.3E-04 2.3E-05' 3.1E-05 6.3E-05 3.7E-04 2 ABEEAICEB . 3.6E+03 '3.2E+01 8.0E+06 5.0E+04 - 7.2E+03' -1.8E-03 1.3E 1.1E-08 5.2E-09 8.7E-10. 1.7E-10 3.8E-11 1.5E-10 9.7E-10 I 0.0E+00 '5.8E+04 2.2E+04 1.8E-03 1.3E-05 1.1E-08 5.2E-09 8.7E-10' 1.7E-10 3.8E-11 1.5E-10 9.7E-10 , 3 ABIEAGCEB 3.6E+03 .3.2E+01 3.2E+08 5.0E+04 1.8E+02 6.1E-01 6.4E-03 1.8E-03 8.7E-04. 2.1E-04 3.0E-05 8.6E-06 3.4E-05 : 2.2E ! 0.0E+00 5.1E+04 . 1:4E+04' 6.7E-02 7.1E-04 -2.0E-04 9.7E-05 ' 2.4E-05 '_3.3E-06' 9.5E-07' 3.8E-06 2.5E-05 , ? 4 ABEEAFCEB 3.6E+03' 3.2E+01 8.0E+06 5.0E+04 - 7.2E+03 3.4E-01 2.5E-03 2.5E-04 1.7E 8.6E-05 1.2E-05 *6.2E-06 3.0E-05. 8.7E-05 a 0.0E+00 5.8E+04 . 2.2E+04 3.4E-01 2.5E-03 2.5E-04 1.7E 8.6E-05 ' 1.2E-05 . S.2E-06 3.0E-05 8.7E-05 i
- }
5 ABEEAECEB 3.6E+03= 13.2E+01 ^ 3.2E+08 5.0E+04 1.8E+02. 6.2E-01 7.0E-03 1.9E-03 9.8E-04 2.4E-04 3.8E-05 1.2E-05 5.3E-05 2.6E-04 , t 0.0E+00 5.1E+04 1.4E+04 6.8E-02 7.8E-0 2.2E-04_ 1.1E-04 '2.7E-05' 4.3E-06 1.4E-06 5.9E-06 22.9E-05' i W ' live % st Probable Eins That Have VB*.' .
. . (
$o l' APEDDGCCB. . 3.6E+03 '3.2E+01 3.0E+07 L5.0E+04. 1.8E+02 9.0E-01 1.5E-01' 7.3E-03 T 7.9E-03 -3.9E-03 2.0E-04 2.8E-04 5.7E-04 3.3E-03 'l 5"' 2.4E+05 5.1E+04' 1.4E+04 1.0E-01 1.7E-02 , 8.1E-04 8.7E-04 4.3E-04 . 2.3E-05 3.1E-05-6.3E-05 3.7E-04 9 A8DDDGCCBl 3.6E+03 3.2E+01 3.0E+07 .5.0E+04 1.8E+02 "9.0E-01 2.1E-01, 3.8E-03 3.7E-03 1.6E-03 8.8E-05 7.4E-05 1.3E-04 9.7E-04
'1.4E+05 '5.1E+04 - 1.4E+04 1.0E-01 2.4E-02 4.3E-04 4.2E-04 1.8E-04 9.5E-06 8.2E-06 1. 4E-0 5 1;1E-04
' 12 ABBDDGAC3 ' 3.6E+03 -3.2E+01 3.0E+07- 5.0E+04'. ,1.8E+02' 9.0E-01 '1.9E-01 1.9E-02' 1.9E-02 1.0E-02 J4.4E-04 7.0E-04 1.4E-03 8.2E-03
- 7.0E+05 .5.1E+04 1.4Et04 1.0E-01 2.1E 2.1E-03 '2.1E-03 1.1E-03' 4.9E-05 7.8E-05 1.5E-04 9.1E-04 13 ABBDDGCCA 3.6E+03 3.2E+01- 3.0E+07 5.0E+04 1.8E+02 : 9.0E-01' 2.6E-01 1.2E-02 1.1E-02 5.2E-03 3.2E-04 3.5E-04 7.8E-04 4.6E-03 1.4E+05 5.1E+04 1.4E+04 1.0E-01 1.7E-02 1.4E-03 1.3E-03 5.8E-04" 3.6E-05 3.9E-05 8.7E-05 5.1E-04 14 .ABBDAICEB . 3.6E+03 '3.2E+01 7.5E+05 5.0E+04- 7.2E+03 2.0E-03 2.0E-05 1.7E-08 8.1E-09 1.5E-09 6.2E-10 1.5E-10 3.0E-10 1.7E-09 9.1E+04 5.8E+04 2.2E+04. 2.0E-03 2.0E-05 1.7E-08 8.1E-09 .1.5E-09 '6.25-10 1.5E-10 3.0E-10'1.7E-09 Five Most Probable Bins That Have Early CF*' . . .
l 7 AAEEABAEB . 3.6E+03 3.2E+01 1.2E+07 8.3E+03 4.7E+03' 8.3E-01 2.7E-02 2.1E-02. 1.5E-02.'6.3E-03 '1.2E-03 4.4E-04 2.3E-03 . 6.5E-03 0.0E+00 :1.3E+04 '1.4E+04 0.0E+00 13.0E-03 0.0E+00 0.0E+00' O.0E+00 0.0E+00 .0.0E400 0.0E*00 0.0E+00 10 AAEEEBAEB 3.6E+03. 3.2E+01 .'1.2E+07 8.3E+03 4.7E+03. 7.7E-01~1.5E-02 1.4E-02 1.1E-02 6.9E-03 1.1E-03 4.4E-04. 2.0E-03."7.0E-03
' O.0E+00 1.3E+04 1.4E+04- 0.0E+00 2.5E-03 0.0E+00 0.0E+00 0.0E+00 0.0E+00~ 0.0E+00 0.0E+00. 0.0E+00
,15' AAEEAACEB: - 3.6E+03- 3.2E+01 '1.2E+07- -8.3E+03- 4.7E+03 o7.1E-01 3.0E-03 5.9E-03 3.4E-03 .3.7E-04 1.3E-04 '1.6E-05 5.9E-05 4.3E-04 0.0E+00 .1.3E+04 2.2E+04 0.0E+00'3.0E-03 0.0E+00 0.0E+00' O.0E+00 0.0E+00 0.0E+00 0.0E+00 0.0E+00' ,
- 18. AFM*AES 3.6E+03 -3.2E+01 1.2E+07_. '8.3E+03 1 4.7E+03 7.2E-01 2.6E-03' 2.5E-03 5.3E-04 1.7E-05 1.2E-05 5.5E-07 1.2I-06 2.8E j 0.0E+00 1.3E+04 .1.4E+04 0.0E+00 5.2E-03 -0.0E+00 0.0E+00' O.0E+00 0.0E+00 0.0E+00 0.0E+00 0.cE+00 .
31 AABDABACB 3.6E+03 3.2E+01 .1.1E+06 8.3E+03 4.7E+03: 8.6E-01 2.tt-02 = 2.0E-02 x 1.3E 4.1E-03 ~ 8.0E-04" 2.3E-04 1.0E-03 4'2E-03. . . 6.7E406 1.3E+04 ' 3.EE+03 1.4E 8.0E-02 '4.5E-02 .4.6E 4.6E-02 1.1E-03 3.2E-03 6.1E-03n3.EE-02i ~! I 4
- * ~ A listing of source tems for all bins is available on computer media ;
-k
. \
\
.n- r : w _ ., . . . - ca., , -. w; ,
O O w w O. O r i. ga , v.; i i p, i ro q d r en i cm ii ra iii iu ii d
- j .! ~
. . 6O
- T. .cl Cr Cs
-l b O
Df ** s f& <
- s. ecn:s a e 1 g -u .2 1- .i : n 3 -4
,', 4u o& ** ,,,,
- W~
i w y& 1 j i
- / s 5
Wc r _7 & c- 9 1
,.c m .9 -
m .0 o i
^ '
9 l l ,a' ,e' - a g s. T.,p
,.. n..
e
-f - # -
4 Q -- U.Q_ t
$ - t' 4
bN O'O #'
! bN O@
UM UM' i l
$ $ & T '{ d U l
]
J
- J.' . o_ *y m l l .l D -j' ,-
,* D G< f to i,ii -
j O C& D Q. 5 CO - m. O,. m W b 'C0 m. r gg b O b O b O b b b b b b b b b y$ O O O O O .O O O O i
& & J & 4 4 4 4. 4 d. d & yON l (JoeA-aopoe; Jed) bey ocuopaeox3 (JDOAJJ0PD8J aed) bey 4 .gesox3 m u' W W-WO OU O kq C .40 es.
W c! o #..M ea n" i cm . i i r e-,-'- " n iii emoii rou s i i r" 6 5 i i r" a i i d 0 C
~
r " . . .i . WH' e N-
*,' wi H.
O "*, wl O-e
* * -p
~
~
& - g i
,e ,< ; L' g y N- s',' : N T U-( b -
i 6. .O C l
,',,, **. WO O k .,/ wO O k QH o
~t
, ~
& c r o
_3 & c gM. e O } O* La v.4 - ' m~D
.: l 8
< : ,' e :;; D gO-wgl e
. .O
-l - .
. w0 t o Ow
, Og- ;& %
a s-O
'o ., & n C f e m -mq
[
~
vh i, ,/ . g w, h, _ me
*b o
e
,**. oo
! i l' ..I wo -@ -e .g-c l 5 W W. 7 ,/ ,#
.: sE W e -
!j-n-
l m. l b ll l b k'-
., , m. ,. wo ! im, , , u. o , g A, L.,, s m , c, , , uso, s , , ,
3 n- e s. to -m e- n .. s e .- e O b O-b O ~O' b' O O- O b O
- b~
O O bO bO s s .. s s a s s .s a s (JooA-aojooo; Jed)*be;j ocuopecox3 (JD04-J0PD0J Jed)' boy ocuoposex3 3.26
M e OOW MnM4 M MC M W M Mmm*C- MOM OMOCod o DOM M O MOOOO C00 0 000 0 00 0 00 000000 0 e I a e 1- 6 e e e e
- 4 e i e e a e I e e e + a + e + e + e +
M M W W W W N k3 W M M W W ki ka ki ha W b4 N kl ha ha las W W kl W W M m @ W 4 H. M. M. M. . N. N. N. 4. h. N. M. O. W. h. M. M. M. O. O. O. O. m. O. M. O. h. D. ; EmmGM Nhh NN MOnMGM NNHH Woh0 nod OWO M 4 00MennnW M eenMM OOMM MOM OWOMO4 D { C OM M O O O C 00 0000 0 00 I t e e a e e e o O. O O. a Oe eO i O O. eOO e eMM OO t e e e + s + i + i + t + O W N W W M l.J M N kl h3 b3 M W kt kl kl W W kl M kJ W W W W W blk3kl44
- e. n. w. 4. m. M. W. W. O. W. w W
. n. . W W. .W M. O.N. N. M. O. O. O. N. o. p o. w. O.
M MM M NMNNMM M MWh. M M W W MM NMNOMCMOMO
# MMM Wh @ WWW d nd M M MOOMO woe chCMOMO O CMM O OOOOO COOO OOM M OD 0000000000-e e t e a e e e i e e a e e e i a e a e e e + e + a + e + e +
4 W h3 W kl Isl N W W M 43 kl M W W W M N N N N M W W kl W W M M N kl W O. m. W.W.W. M. M. M. N. M. O. m. O. 4. E. m. M. M. W. W. . O. *. C. O. O. W. O. N. O. ' Nh MMhWOOMM hhM4 hWMMNN W OW OMOMOMO 4 MOOMWWWMW W Od OW MOG4 4 MOM OMOW O* C D OM MOOOOO O 90000000 0 0 0 00 0 000000 D e e e t e e e e e e i e a I e e t I a e + e + t + 4 + 4 + m A W kl W bl M N N W W W W kl W W M N W W W M M M ina W M N N W W W
- w. O,W W. W w O. O. N. N. b
. e.h.h. A. M.
G L . - .w. w M. M. N. O. M.O. N.O.M. O. O. b. M W 4 wM. MNN99MW eW NNMOMMHH MOM OMOM OMO V e l. NMOOd enne M NMM *NM@ mnM M CM OMOW OW O-k C OM M OOOOOO COOO O OOOOO 0 00000000 0 W be e e e s i e e e e e e e e e T e e e a e e + e + e + e + e + 8 M M Q MW M W h1 w W N.In3 N b3 kl W M kl ist wat kl h) W lal W b4 W kl N M Int W W N e .M. s M. . M. e. M. M. W. O. M. 4 4. N. M. M. M. w e.
. M. O. m. O. h. o. h. 0. M. O.
8 M MmmHNhhMN HMmOMMNNOS WOWOMOMOWO M
@ NMOSeMW ed e NMNMNMme4 4 NO NOW OMOMC-A 6 00 000 000 0 0000000000 00 0 0000000 e T 4 i e e T e a e a e e e e e e a e e e e + e + e + a + e +
F. W bl lat h3 M W W *W**N M kl W bl W W W W b3 In3 is4 W M M lal bl W W kl N W Da. M. O. w. W. w. 4 . . O. G. M. O. N. M. M. M. M. M. M. M. O. M. O. M. O. w. O. W. O. MNewkmHMON HNNNNNMMW 4 MOMonOMONO NMmeMeet Md NMNMNMemMM NONOMOMOM O O C000000000 e - e e e e e e e e 0000 00000 0 e e e e e e i e e e 0 00 0000000 s + e + e + e + e +
@ m M b' M M N ist W ina w W .k3M bl N 43h3M N M N N W In1 b4 W M kl M ins W -
M U @. m. O. O. n. h.. M. M. G. M. ' m. M. O. N. M. M.N. N. N. N. M. O. W. O. M. O. G. O. O. O. Q MNMM MMNNN N MNNNNNNNH H NOMOMOMcMO M N M NMMM* MMM* MNM NMNAMMM NMNMMMMMM M OW M C000 0 0.0 0 0 0 e e e e e i s e i COcOOOOOOO e s i e e i e e s i C000000000 i e e e i n s i i s ed . En3 WinsA N hlM M M N lINMWW M kl M M M M kl W W ist M M M kl Q c M.M.M. N w.M. M. Q. m. @. M. M. M. m.e.W. W. M. M. h. o. n. M. c. N. O. O. O. m. t N N M. NMMOWNNhh MNNNNNHMW 4 NMMNNOmMb4 A M MMMMMMM H N M M M M M M MMM M N OM OM OMON O C0 000 00000 0 0 0 0 0 0 0000 0000000000 N N e e i s e e e e e e e i + e + i + a + i + o $ n' m
. q
. U. U. U. U. U. k. U. U. N. N.U. U. U. U. U. U. . U. U. N.- U.
eMMM WWMMWW aMmM mMMMMM
$. $. N. U. N. U. U. U. U. U.
WOhohohCWO M M ' g 4 h N4 M4 Nd Mg NW N4 N4 NeM* M 4 M
- M
- M4M4 MW W C yv ghy 8 wM a gym 000700o00 gg+ggg+ggg 0000 0 ggggggg,90g 00g OcOOOOOOOD ggggggg+g+- M e
b 0 M c M. 6,4 NQ
- m. e. . N. m. W. . N. m. e.
M HbNMMhNMM
- m. W. m. g. m. W N. N.
MMMMMNhNbN
. . N. h. w. h. 4 6 4 6 . h.
W HW-HwMWNW N
}h U e N e * ****wwe**
- Ug mo ****e<<<<< M eMeM4Meme oM a ha ewn COOTC
+ ++
kaWWwMMWWWk
+ 0000o
++ + + + 0 0000000 0 0
+ + + + + + +++ +
W h3SIkl W w W w w W 00000000TO
+ + + ++ ++ + +
M cg M ew O M, wh3WWwwwwMw en O. m. o. M. O. w. Q. M. O. M. O. M. O. H. Q. m. O. e. M. n. M. M. n. M. M. M. n. n.- u i g M MOOMMMMOOM OMOOMMMMOM WWWMcM DMcM N v 0 a
%g H e snWOmoWomo S ch MhnnnMM h ohChDh Dh D n 8 0000000000 000 0000000 0000000000 e + ++++++ + + + + + + + + + +++ + + + + + +++ + + +
e S 6m. 3 M 9 4 bl W W W la3 W In3inJ W W N W W W W k1 M M M kl N lai M M Esa N lal M M.4 ini M ~ 1 er Md O OsO. O. N. C. O. O. N. O. Mh WOMOMOMO Q 0 0 0 0 0 0 5 6. h.
. e N. O. N.O.N. O. N. O. N. O. W p
MSMh Mhh *Ne MOMON OM OM O N m e M-8 M M M M N N M M M M et
-b M M M M M c
P e O O O O O O O O O O ko O- O O O e6 + + + + + + + + + + M + + + + + M 4 N u 9 M
. N.
M N. M N. N. p. N. . Y. N. e N. w . . . N M M M M n M M M M M. M M M. G S y k M o M O M O M O M O mm- O M O M O M O M . O MM M M M M -w M ea + + + O O O O O
+ + 4+ + + + -+ 4 + + + + + e AGw bl W
W M kl M NW N W kl M eW W N N M E H e w W W W W W W W W-gp W. W. AW. p -W. AW. H W. . k g 3 4 M. M. M. M M M -M. M. M. M M M. M M M. W n m e G C N # M e M N u-M M h _C E
$ h, k e - k . M . -
A A g A 4 . %, A bl -
- O k W M bm Q b s
"2 s Q Q Q Q "Q e Q Q Q Q
- a 2 a % S. .
M M M & I4 E E L = ' 0 g e O kN N M O M 4 4 N. 6 m O ( T k M 4 kN N M M M ke M M M M M M C W W 4 4 3.27
1 8 8 O O i .a ; d r i c.. , p.- p.. ai '"'
, im $
e e t
, A, - - - -
,=C.C.,C <*** I
~
l p C i i
, - u - u
*Re5 O- O E: ,,
L'en3Dn,
-mo -
ya p 2 l -
- u ,
.- 1Lo 1uo l
Ow
& c * , , , ,
- 1* o . .w-d'c y
/ ~
mO ' ~ p k. 4 l - 6@ Oa ,*, ' , we o 8-
- v. ae 4 & -
o
,e e . .
4 /e s - n o e [ - n o m
-vo w4
- l l - 8s ne
/
.a -
s-oE e Um em . l l , aw .~) , aK . y g.
' ' n
,' ; - e -
.e-e
,i - - wi , -
w a mm o O e t8 g
~
,,is , t t : F. ! W-i 5. > r t i L,. ,,
*8 pe..
e n m e '9 :: e s- e m 9 :: $" i d 6 6 d 1 6 1 1- 1 1 1 u$ o o o o 9-o o o c. 9 9 9-
-- om w.
(JoeA-;otooe; sed) be;j ecuopesox3 - (JoeX-sopoe; Jed)*be;j ecuoposox3 m vi eW
**4 0
-U-p e-O- O C e .e4 W W Du O 9 VM
- c. ,m . , , g.s e. ;.; .
c , im o, , i, , , e, ,, im , ,, , -
-ec MH
*~*.,...-* -
- ,. w
,- 1 e-a a'
, u o ***,,- Q t wl -c_e U C
,a , e. r 4 e L
gp _ec
,' N N t4
,' eC .. l L I -L s l e -
WO , WO eH
- O L *,' O L c'
& C *
& C HM o' *
! l , [ s , *: o ,<.. 'WM e *sll: ,' e'O D l' / - w 0 .O 0 wl 00 l 1 e
. - i .
i e OL Og Nt
- e s v & - o , & s c a e t . tD m e-s #
. .m a m k 4
t s - V O .o 8
- Tl O n' U s e l
w: e
- g. o r
wm o e- g-
?
s { e l i o'E d - l 7 WE W
- D
/.
D g l l l n' b : e' , b k
% , .. ,1 .a . . 9 v-
. ,t. . ., . . . ,. e e b O m O - g N- ' CO 'm 9 "'
'1 l
- w. w l.
w l wl w 1
-w 1
w i w 1 w- -w 1-w1 wi O O O O O O O O O O 'O O s s .s s. s s a s : s -s a (;ooA-Jopoe; sed) be;j ecuopeoox3 (Jo8A-J0Po0J Jed) bo;j ocuopeeox3 i l
'3.28
Y Table 3.3-3 Mean Source Terms for Grand Gulf Internal Initiators. FDS 3: Fast SB0 Warning Release Reisese Release Relees. Freetiens Time Elevation Ener8y Start Duration EM j s) (m) Or) (s) (s)_ NG 1 Cs Te Sr Ru Le Ce Pa Order Five Net. Probable Eins ABE DGACB 3.EE+03 3.2E+01 3.0E+07 5.0E+04 1.8E+02 9.0E-01 1.9E-01 1.9E-02 1.9E-02 1.CE-02 4.4E-04 7.0E-04 1.4E-03 6.2E-03 1 7.0E+05 5.1E+04 1.4E+04 1.0E-01 2.1E-02 2.1E-03 2.1E-03 1.1E-03 4.9E-05 7.EE-05 1.5E-04 9.1E-04 2 ABIEAGAES 3.EE+03 3.2E+01 3.2E+08 5.0E+04 1.8E+02 6.1E-01 6.21-03 1.5E-03 7.EE-04 1.9E-04 2.EE-05 7.4E-06 2.95-05 1.9E-04 0.0E+00 5.1E+04 1.4E+04 6.8E-02 6.9E-04 1.7E-04 8.4E-05 2.1E-05 2.8E-06 8.3E-07 3.3E-06 2.2E-05 3 AEEEAIAEB 3.EE+03 3.2E+01 8.0E+06 5.0E+04 7.2E+03 1.EE-03 1.3E-05 1.CE-08 4.9E-09 8.1E-10 1.EE-10 3.6E-11 1.4E-10 9.1E-10 0.0E+00 5.8E+04 2.2E+04 1.8E-03 1.3E-05 1.0E-08 4.9E-09 8.1E-10 1.EE-10 3.EE-11 1.4E-10 9.1E-10 4 ABEEAFAEB 3.EE+03 3.2E+01 8.0E+06 5.0E+04 7.2E+03 3.5Z-01 2.3E-03 2.1E-04 1.4E-04 7.1E-05 9.7E-06 5.1E-06 2.4E-05 7.2E-05 0.0E+00 5.8E+04 2.2E+04 3.5E-01 2.3E-03 2.1E-04 1.4E-04 7.1E-05 9.7E-06 5.1E-06 2.4E-05 7.2E-05 5 A5Dff.G.CB 3.EE+03 3.2E+01 3.0E+07 5.0E+04 1.8E+02 9.0E-01 2.1E-01 2.0E-02 2.0E-02 8.4E-03 2.0E-04 3.9E-04 6.EE-04 5.0E-03 7.0E+05 5.1E+04 1.4E+04 1.0E-01 2.3E-02 2.2E-03 2.2E-03 9.4E-04 2.2E-05 4.4E-05 7.4E-05 5.EE-04 W Five Most Probable Bins That Have VB* ABSDDGACB 3.6E+03 3.2E+01 3.0E+07 5.0E+04 1.8E+02 9.0E-01 1.9E-01. 1.9E-02 1.9E-02 1.0E-02 4.4E-04 7.0E-04 1.4E-G3 8.25-03 j N 1 c 7.0E+05 5.1E+04 1.4E+04 1.0E-01 2.1E-02 2.1E-03 2.1E-03 1.1E-03 4.9E-05 7.8E-05 1.5E-04 9.1E-04 5 ABDDDGACB 3.6E+03 3.2E+01 3.0E+07 5.0E+04 1.8E+02 9.0E-01 2.1E-01 2.0E-02 2.0E-02 8.4E-03 2.0E-04 3.9E-04 6.EE-04 5.0E-03 7.0E+05 5.1E+04 1.4E+04 1.0E-01 2.3E-02 2.2E-03 2.2E-03 9.4E-04 2.2E-05 4.4E-05 7.4E-05 5.EE-04 6 M A RA 5' AM 3.EEt03 3.2E+01 0.0E+00 1.3E+04 1.BE+02 7.1E-01 1.5E-02 1.4E-02 6.5E-03 2.0E-03 1.3E-03 3.8E-64 5.4E-04 .2.3E-03
-0.0E+00 1.3E+04 1.4E+04 0.0E+00 2.3E-02 9.4E-03 3.5E-03 0.0E+00 0.0E+00 0.0E+00 0.0E+00 0.0E+00 3.6E+03 3.2E+01 3.0E+07 5.0E+04 1.8E+02 9.0E-01 1.9E-01 2.3E-02 2.3E-02 1.2E-02 5.2E-04 7.9E-04 1.6E-03 9.7E-03 8 ABBDDGACA 7.0E+05 5.1E+04 1.4E+04 1.0E-01 2.1E-02 2.5E-03 2.5E-03 1.3E-03 5.7E-05 8.8E-05 1.BE-04 1.1E-03 10 ABEDA1AEB 3.EE+03 <3.2E+01 7.5E+05 5.0E+04 7.2E+03 1.9E-03 1.EE-05 .2.2E-08 1.1E-08 2.1E-09 1.4E-09'3.1E-10 4.5E-10 2.4E-09 4.7E+05 5.8E+04 2.2E+04 1.9E-03 1.EE-05 2.2E-08 1.1E-08 2.1E-09 1.4E-09 3.1E-10 4.5E-10 2.4E-09 Five N st Probable Bins That Have Early CF*
ASABAEAEB 3.6E+03 .3.2E+01 0.0E+00 1.3E+04 1.8E+02 7.1E-01 1.5E-02 1.4E-02 6.5E-03 2.0E-03 1.3E-03 3.BE-04 5.4E-04 2.3E-03 E 0.0E+00 1.3E+04 1.4E+04 0.0E+00 2.3E-02 9.4E-03 3.5E-03 0.0E+00 0.CE+00 0.0E+00 0.0E+00 0.0E+00 16 AAEEAEAES 3.EE+03 3.2E+01 1.2E+07 8.3E+03 4.7E+03 8.3E-01 2.7E-02 2.1E-02 1.5E-02 6.3E-03 1.2E-03 4.4E-04 2.3E-03 6.5E-03 0.0E+00 1.3E+04 1.4E404 0.0E+00 3.0E-03. 0.0E+00 0.0E+00 0.0E+00 0.0E+00 0.0E+00 0.0E+00 0.0E+00 18 AA N AEB 3.EE+03 3.2E+01 1.2E+07. 8 3E+03 4.7E+03 7.7E-01 1.5E-02 1.4E-02 1.1E-02 6.9E-03 '1.1E-03 4.4E-04 2.0E-03 7.0E-03 0.CE+00 1.3E+04 144E+04 0.0E400 2.5E-03 0.0E+00 0.0E+00 0.CE+00 0.0E+00 0.0E+00 0.CE+00 0.0E+00 20 ABA3BEAEB 3.EE+03 3.2E+01 0.0E+00 1.3E+04 1.8E+02 6.7E-01 3.6E-02 3.9E-02 1.4E-02 5.3E-03 4.EE-03 1.3E-03 1.5E-03 6.1E-03 0.0E+0e 1.3E+04 1.4E+04 0.0E+00 4.2E-02 1.8E-02 6.3E-03 0.CE+00 0.0E+00 0.0E+00 0.0E+00 0.0E+00 22- AAEEEEAEB 3.EE+03 3.2E+01 1.2E+07 8.3E+03 4.7E+03 7.2E-01 2.EE-03 1.5E-03 .5.3E-04 1.7E-05 1.2E-05 5.5E-07 1.2E-06 2.8E-05 0.0E400 1.3E+04- 1.4E+04 0.0E+00 5.2E-03 0.0E+00 0.0E+00 0.0E+00 0.0E+00 0.CE+00 0.0E+00 0.0E+00 t.
- A listing of source tems for all tins is available on computer media
8 O
~OB-p.. t , i p, 44 p. - y W n ; i j p. s i a a F: 6 i runii im u i i d
. -! .O- ,
- ,. . . , _ , , e.r g ++ sc C i
).
w- . C .C p - LJ
,,e O- e6 6, . 9 n;,
, 1 , E @cn
@ inim
- D .
s f-sg ou a
,- . : a. . . -
- u
[ .. wo. ak-
. '.... "'- wlO kO= - s
. ( ,
a c , a c 9 e , * *:. O ; O O mO
~
mO ,*', y, l' <,',,- - wl 0 0- - wg1 0
, v O u, / O
&u, 8
-l ,' ,
- a. y *
,' ^: o 5 s : n # -
M *e4
- s n -
4 o f n o uO-l i
.i
- 63 a e ,
,a : 1*
O e j@ aw
, aw - -l-
, .- k u m
.: :. l ..: o ,', ': in ee
-i-
- w. s.. ,7*,
w i- mN e u , ,, , t, , , O a Lh ,, u ,,, s- , r' ,e , , O + a o o = o 9 :: e s m e e s cn '$" 6 1 1 6 O 1- 6 O 6 1 O d-O 1 O 1 O 6 O u8 O a O a O a a O a a O a a a ,a a a om
,a (JoeA-JopoeJ Jed)'beJJ ecuopaeox3 (JoeX-JopoeJ Jed) beJ3 scuopesox3 mJ GW
, H O OM CW o o W W eH M O. O o w. D. r ,i p-a.i e, i s i i g%,iss e- - " yuii i . wo i i i puis i i pm. i i i ym. i i ;" @ C'
,,..**** , , ; MH
/,=- . wm ,,n ,= k g;
*,, ',, e.< -
l l LJ O w O e= oe
* -. ,a ,
^-
a u g Va
, g y
/ ; N- / .: nt M tp
,,.**",~ wQ' l .u ,' g u' GG l -
wo ow
, ,, o e O L o o ,e - a C. -
e' -3 a Oc- yM
< 0 ' . LL3 P-4 l' -
[ MO / I MO U o- g o l i 3 w0 w0 .O , s-s
-,e o&
a - u-Oy-
- a. O m t' e C
' O n'
,l M e
e m m, g
- 4 -
4 9-
-l l i gg
' : 6 .0o .
,,- 1.co .
OO # O @ C
- a x. a w' I
l :: ; , ,; '
, u
! 'l, l - in - . l ,-
- o j'o t .
I :l , 1 l' ! : 1= N j ,!.,, , , , , t. a,,' . . 's. , i.. ,
%. s ,
c3 e s e m-9 ; m. s m cn . 9 ::: wI wI wi .I wi i i wi I e -I w w w1 w w- w w O O O O O .O O O- O. O o -O
.a a a J a a a. a a a 4 a j (JoeX-JOjo00J JOd)'be.U 00UOp080X] (JOod-JOpDOJ Jed)*beJJ OOUOp000X3 1
l l t l l t 3.30 4 - , - _ _ - _ - _ . - , . _ . -
i 1 1 M Q F) N M N F) 4 M O F)Ooo oOooooOoOooo6oOOO NO en M F1 M N O in M N M N M M Pn O M N ooOoCCOOoO 4 6 6 6 i e e I e + e a i 4 a e e e 4 e i a e e t a e + t e E is) 044 Sal M b4 lel M bl is] $63 la) lal Is4 $si M &al Od Ut bad lal la) ' W ls) lal lal M is] lal ist M jnj e4 W et. tr. n. e4. N. N. h. au. o. t s. M. e4 4. ed. e4 . O. h. W. e4. N. O. n. c4 e4. et. o. c4 W.
- 4. in F) e4 e4 N4 4 F) o O v4 F3Flet M N N h e e4 N M e9 H F1in O e N 4 M4 F1 4 M 4 n4 oM M M 40 4 4 4 F1 4 M 4 M F) M4 4 Fl OM M ]
oooOOooooCoCooOOOooo oO oooOOOOo e e 1 4 e 6 e e a e + 4 6 4 4 e 6 e 4 e t 6 e a 4 6 6 e 4 6 e W. W . o. . o. . . . . M. . . . , . . . W <
. . . . . . . h OM N h h e4 h 4. F1 e4 et 10 h W N et M N h et F1. e4 N et r% et o ==e 4. 1 1
4 n4 Fi nM4 n,o4 Fi in e n 4 4 Fi n e nM .e M n 4 .e o Fi M ooOO OoooooOoOoCooOOo OoooOoOoOO ( e e e t 6 4 4 4 i e + t e t t e e # 4 e e e t I t t 4 e + 1 e - 8 M4 985. et .F N Fl .f
. . . . O. .......F.G.
F1 O O P4 th tp 4 h N e4 10 h
- 4. N W
.=4 f h. W O v4 N
. O. . .
4 en M M .9 F1 4 en M C s') M4 O *e M F) M4 O 4 MMM4 F):M o M M OOOooOCoOOOoooOOOoOO OOOOoOOooO p e e i e s e e e t + a i e e e a e e e i e e i e 6 6 + e 4 m gg M In3 $d Ian M M ini lal M til Onl En3 b4 M In314 N r t;, b4 en, h. WM.tr.O.kl O. Me4.M +4 ina Id bl en. 40. klF.isJ in.a M.t. 6 0. o. 4.
- o. o. or. b. o. H. e4. O. b. o. O. o. o. O. N e e N N e.1 e4 N F1 4 4 e4 0 4 4 N N N M e4 v4 N M N Fe .e 4 N M 4 OOn st U
F) 4 F) N M N fr. 4 F) oe rI ( en M F1 F) N4 n F) NMNMM F) OMN G. b . Le O O o O O O O O O O O') OOOOOOOOO e a e e e 6 e I e + 4 6 e e e e e a e e ooCoOOOOOO e i e e 4 e e + e f b. lf) ShJ N ld M la) Isl 14 Bd M lEj laj ld 1sj lal N ld Is) M lal LQ lsJ ? 3 (n4 Ed laj 16 ] M M ld lsj e e4
- e eo. 4. N. o. o. o. n. 4 o. M. 4. w. on. o. en. to. to. N. O. o. C. F) . . O. n. M. o. et. e.
et qg qp p3 N et F3 @ O M C .t N N N f4 N N N h m e4 F3 4 N e4 N 4 O N F) O
@ F) M N N M N N M N M N N M .e F1 N F1 N M =9 MNNNM N N FlN N' 04 O O O O O O o O O O O O O O O O o o O tes OOO oooooOo 9 # # e # # a s # s e i e a 6 e a i i e e e e I i e 4 e 6 e F* M lsi M M 14 lel M 151 Ed M lal tal M laj M M b1 bl Inj M ' M M ls) M in) Enl is] $4 is) M e4 e4. H. N. ess Ep. N. h. @. N. e4. en. eD. M. ta. 34 Ok. e4. O. @. e 4. e4. N. N. **. O. N. F1. en. to.
Ob e4 ee N .y to e4 c4 e4 n eq 4 et H 4 et abnM4 4 to ee ,g my c4 et h eg to M 4 N N M e4 N M NMNNM 4 MNNNM 4 F1 e4 NNMNNNNet OOooOoOOooooOoOOOOOo ooooOOOOoO O e e e e e B 6 e t a e e 4 e e 4 e e 4 e a a i e a 6 4 e a 6 M e In3 isj id Im! Im3 Id M tal Ina ini g U M F1. M. M Ina h. M. in3M. et.Me4.MN.est Id
- e. en. M Id Mr. fal M M ist 161 M 14 . M.lal eetal4
. 4 t.40. M. o. N.4M. O.F)e.. c4. o. en.
F) 4. Q emH 4 hMMMM SNOM 4 54 M4WM hMNGh4 NNMW H o 0 F4 e4 N N en F1 eeee N N N N ee N M M e4 N e4 N M M v4 N e4 es e4 N N N et Om +< oe 0 e ioeoi 0 e e o e0ioi oe e0 i0i0e oi i0 0 0 o o. 0 0 : 0 0000o000m e i e i e e i e i c** n 4 e4
. . O. . . . . . . .
re ce N e< a N m e4 N e N N e e e en N e< < 4
........in. . . . . . . . . . .
e N N N e ** N e M M hM Q es e4 ee e. ce e ci ee u o ee ooooOOOoOOooOooOOOoC P. e4 N e. N es e4 e4 N e4 ce ** e4 ee N es C e4 ee oCOOOOOOOO 4 c e e e e i e e e + e e e e i i e a i 4 i e i e i e i + i e M' o .
$ $. U. U. N. N. E. U. U. U. U. U. U. U. U. N. U. N. U. N.U. U.U. U. U. N. U. U. U. N. N. . m (A @ et M N N N O e4 h O h N N 6 6 e4 h N to to e NNhNNNhOWM M g M o a $.
g p
.et .O N4 N4 M F1 N .t N4 N4 N4 et M N <r N 4 nMMN4M Ft N4 N .*
o o O O o o o o O o o o o C o o o o o e-4 W eoa ee-4 e o.* + + + + + -+ + + + + + + ++ + + ++ + F'o pa N+14 O+ +MnM C +o T +o o
+ + O+ +o .e.
,o 4e H p.e4 u M M., ed M ha M M M M ad ba M sd M b4 ba ad ha M Id, v s h. e. o..e.kd M M M M E.d, b e *e4 e,4 04 Q p o. . Q4 b. o. o. <. o. r. o. d. u. .e. h. e. o. n. O. .
@ et O e4 .e M M e4 Gi e4 G et G e4 4 F3 09 el G e4 4 F# O e4 4.
M Un w e4 o. Ob4.eto. l c 6e s- 4 4 4 .e 4 .e e r*4 4 4 4 4 4 <e 4 .e .e E 4 4 4 ,. ve4e4
- $ a ao
.F oooooooooOOoooponnor .400o00oo0o0
,c M
. ,n h. . h. h. h. h. h. h. h. . h. h. h. h. d. h..h.
hhWO4 WhhWWMOhh4 MOOhh
- h. . h.. h.
h. ( MMO4 MMOOM
- h. . h. h. h. . h.uh.
.c Em e h cocecheoooohneecch e M
secoceeOOOO $N W b ooooooOO OOOoCCoooooo et o C O o o C o O o o 4 ed e -
+ ++ + + + + + ++4 + + ++ + + + + + 64 + 4 ++++ + + + + * *e4 9
. . . . . O. . . . . o. . . 4. O. . . . . M 4 W
. a . . . . . . O. p 04 IsJ N 40 C o 4 h N W o O o O N 43 4 h O O N 40 1* 4 h C O .e h O O O O. e 0
4
~
8
.e e4 e4 e4 eg et og e4 et ce e4 b et M et ce e4 g;
4 e o O O o O O o O o O >O O o o O 4 5 E + + + + + + + + + + e4 + + + + + ,Q e4 . Od . N. . et Ih4 F4 F3. FI F1. F3. F) F1. F1. F3 F3 FB. FI F3 FI F1. e 9
> La N O 80 4 4 0 4 4 4 4 4 4 4 lEJ 9 4 .e 4 4 %4 C C C O o o o O O O o C o o O o
+4 1^ W + + + + + + + + + + +1 + + + + + j
$*. M~ b3 M N N W W (d 114 In1 W inJ W -M In3 l&J H .Ale v F) F1 F) F3 F1 F) g he M. M. M. M. . F). M. E F).
54 F). . M. g
*E 4 4 4 .t. 4 4 -4 4 4 4 4 4 4 .4 4 9 e n C &
C +4 U
*e El 64
"' I j
8 5 h b. h .h b . came mm d '.) = a m M 3 nu2 i a. s
, a
= m m
vo m e 3'm a in s U emw m M
.n4 l
a sa l
+1 m st .a4 l E
Q. i s, e n . - N ",:' c. 6 cM 3 N Ft 4 O in h co ci e4 >n to s> e4 u I 1 O ti sie e 1 l l 3.31 l l
- y
li!ll
!:rlI 0 O E E 0
1 C.
-~ ;1
- ,i~
_ ,F ~ m~ 2
,+
. C .
1 le - ~
.i - gh ~ 1
,+ , , . E
- t nh,3 t E
0 et h- O
. A h t . r9gc 5- O 55'
- t
,. L
- s c s'
e .- 2 C
. P - -
2 L
-.* - r E o 0 F ,.s'.' ,
E o 0F
- r a
1 n 3t i o '.i',,'
- s t n i
o r o
- r
'. , 3t f
- - oc ,
- c
.A E 0F r
. E o s _
Fr 1 O L n
,- '" e ,' a.
e o
- " s \
s i . 4 c . 4 o tO E-e
!e -
Ef -e cB aS
- _ 0 -
O e r -
- _ '4 tR 4 .
LR Fw _' w o
> 5 _
5 el E- _
- sS
_ = O _ r E a _- _ - _ :_ rI L : 7; T2: f _- - .=- - N 0 e 1 l : 6 7- S 9 0 1 g 0 e4 1 1 6 7 9. 1 '11 R E E E E- E E
- - - - - S 0 O 0 E E- E E- E E rD O. O. 0 0 0_ 0 0 o 0 oP 1 1 L 1 1 t 1 t t t 1 1 f
$ XIIo ooOL g4tE r' eoCove oO - ,
Cg>-[o gE _e3&E ,ucop,eh t
- ss er .
io ct 0 na E 0 ei 0 E ut 1 O qi
, h=:- - 7
. ' ,g j, -. a - j: _E ! _ ,= - - _; : I L en rI
' 1
'. 1 . F s'. - l E- ea 0 .s ,
E 0 cn _~ 1
',*'.< a 1 r
. nr ae 2 S
_ - 2 dt
. E-r o .~
1
- r en eI
,' 0 F - E o
_~ +- 1 n 0F c _- - ' a 1 nr xf
- o _. o El
_ - 3t i - 3t i u
- c _
_ E- a r ,'_
- E ar
- c .G O F -
O F' 4d _ - L e - L n
~ _
s ,- _- e 3 a
~ _
4 a , s r
~
- 4 a
-e . 3G E- e
_ 0tR le ._
- El 0 e e
- _ a -
- e t R r
- u 5 - g E-C' 5 i F
E-h-; 7 . - 0
- _ _ _; 5 t ~
f~ _ . : - .s;-
,= : 0 t
6 7 8 9 0 1 0 1 1 6 - 7 8 9 1 1 1 E E- E E E E E E' E 0 0 0 0 0 0 E E E-1 O. 1 t 0 O O 0 0' 1 1 1 t t L L 1 t T,X8$ooeL g4cE eoCovgv$ - . r T >1o 82 Lo4cEL ouCoDeo$' ' uh rJ
]" ' ,lllllIf
_ . . - ~ . . . . - ~ . I l l l I MOM NMNM W M NMO 4 MM NMMW M MNMNM M MNM O l OOO OO OOOOOOO oOOOOOOO O O O O O O-O O O O e e e e e e e a e + e e e e t I e a e e e + T e e e e e e e i e la! M b3 la) laj Gal la{ is) 863 Gn3 M la) b) inj lal Isl is) M is) is) Isj b 16) M j g4 to. e4. isa. el. ce. N. N. kl Is4O.kJ
- 6. m. e) M ls). O. ce. w. O. n. e4 e4. e4. h.N. to.O. en. e4.. ts)ed. e) is). e4 0 enMNe4 NeW N N P3 O F3 F) W) e4 et Mh e et N en et M nGNnO wnd M d M ond M4 On@ M M o*d n W MMMd 4 MMMC e O OOO OO O OOOOO O O.eOOOOOO e I a e e e e a s e e + a e a e o e Oe O s eO O. e e O
O O. eOOO s + () ist M b) inj M lsJ Bnj ls) Enl M bl lal M In3 la) M laj b2 is) M 40 e) l.1 lsJ ls) h) tal is) M M imie. M v1. O. e4 to. vi. et. . O. N. O. O. h. et. N. 90. N. n. 40 ID. v). ce. N. ID. N. Os. @. . O. to h e N N Pa h h.* @ M N O N N et N e4 h e4 ee ** F et NM h e4 4 e4 0 5 d ef) 4 M ef) F) W n e fe W O ef) GD 4 M If) e W) to Wi F) W M tf) 4 M M d O OOO OOO OO O OOOO e p.4 e e e 4 e e O O O O. OO OOO O OOOO OO O e e a e + e e e e e e i e lol Is3 in3 16) inj M b) b8 M laj ls) laj 1431s4 ls) Isl M M M M e e e e t #
- 16) 844 is M ls] ist M M M h) e e t +
- N. F). tp. M. N. F1 N. W. h. O.
- b. e4 . e4 N. M. IG. O. @. W. to. O. sp N. to. F4. N. M. W. N.
M e M e4 4 NM4 N e4 en O ef) . ED e e4
- hek
- N # e4 W h etNeO W tf) M ef) 4 M4 ef) F) NMO d if) M W) eM4 If)
- M F) eft 4 M M eft p) O O OO OOOOOO OOOOOOOOOOO e5 O OOOOO OOO D a e 4 e 4 s e s e e t + e a e e a e e e T e s 4 e e i o e +
8 . . . . . . . . . d. . . . . . . . . . . . . O. . O. . . . . O.
- N N re e4 N F) .e e et M e* O N N 4 4 N M N F# NMW e N rn if) of) eo s)
U M 4 M N F) NM+MNMO W ef) F) NM Md if) MNMNMMMNMO e. b OOOO OOOOO OOOOOOOOOOO OOO OOOO OOO
- b. k i e e a e e e i e e t + 1 i e e i e i 4 e # 4 e e e a e e +
F) Di M is2 N b4 M M lalIs1 M is) Ist la) M bl Is3ls3 laJlet et s istIsl M M M M kl M M laJ G em et. eo. 4. N. O. O. O. v1. W. eD. 4. O.10 4D. n. 4. O. n. N. O. O. O. M. 4. O. V). e4. e). n. O. e @ O F3 N e4 c) n n N N F) CNNW N M N N G3 e4 el e N e4 N h FB
- O er e4 e M MNNMNe4MMNNMM 4 NNMNFB 4 F) NNNMNNNNPi K OOO OOb n OOOOOOOOOOOOO OOOOQOOOOO e e e i e e T T e e i e e e a e e i e e e e e e i e e i a e e h isI lsl is3 Is3 143 b3 thJ lsl Isl Est $al M isI OE3 lsI lsI Isl M SEI Nls) M M M lallal M N M
- m. . o. a. e4. e4.e4. N. eu. n. et. W. on N. N. N. e4 en ee0. N.
. es e4. N. h. et. n. v). su. N. et.
On e4 e4 N e 4D e4 e4 #MetMe4Ne4* W et F3 4 e tp H d d H P4 ID e4 h c) 4 N N c) e4NMNNNMM 4 NNMNM* M ** N N M N N *8 N N O O,eOOOOOO OOOOOO OOO O e e i 4 e i eOOO Oe OOOOOOO O O
# e e e e a 1 4 e e I e e i e e a e e e (Q R Inj k) thl 1sl Ist lal ls) 54] b) is) 16) M b) lal Gal M lsj M M bl M is31sj M N y M. n. h. Fa. n. d. ,e4. M adH. G> e4
. h) M is)so. o. a). N. *. *. n. O. O. n. n.e4. .e. *. M O. v. n. <. n.
e y e e ce w h e4etetce < et @ c) w N Op N e af) n he4Nm5 w M e4 N N
*** Q
! U e4 N et e4 c) e4 et N N es N N N M N et M e4 NM M et N et F) e4 N e4N N O e=4 m C 0 0, 0e 0e 0, s e e 0 e ?0000000000?? e i i G e e e e i i ?000000000 e i e ie i e a i M M lsi N isi M M M M M D is) BaJ is) leJ laj In3 la.I lb) ist M 141 ist bl M M lh3 M 164 N W) C p.* e4. N. ls)et. n. b. W. e4 N. n. c). N. N. N. on. vi. to. N.5.O. @. V) e4. wi.
. 4D. h. O. V1. 10. M.
g e4 e4 N e4 e N e4 e4 N ee N O to tD N N @ Hd W OnNNN@ et M f). NO M QQ et e4 et e4 c4 M e4 e4 e4 e4 e4 O e4 N e4 ** e4 N e4 N O COOC0000000 000 0000 0 e4 et e4 e4 e4 N e4 e4 e4 o O C00000i00 m ki m e e i e e i e e e i e + e e e e e i i e i e e i i e s e + 1 Q bl M N N la4 su M M M M M M ka M M M bl M M M 161 kl M M M lsl M kl M kl Q .
- r. O. O. ID. N. N. e). O. O. e4. @. ep. D. D. GD. W. 4. N. W. Q. 4D. N. GA. 40. .e. N. *. n. h. 4D. D.
- m. W m op M em Nh N e e4 6 N N O N N h N N es to to e NNhNh ettD en b C N g N !
O c: g y e O NW N .e M MN*NwN*N*N*MMN* p M M N
- M e) N*N*
e. F"4 N M eeu
*t e O OO OO OO O O OOOOOO O OOOO
+ + + + + + +++++++ +++++++ noOO + +OOOOOOO+++ + + + ++ .e 4
@ e aw
.O M M lal M M M kl bl Ina M laa laa M lal M M In3 M M M g) M is3 M kl ist M N ist bl M.e q H .e4 p e4 w ep v. . O. <. h. io. O. W. Q. <. O. *. O. w. C. *. h. w. o. w. g; 6. tc O. w. h. m. O. v. O.
e4 . h 'e4 Di 54 Op e4 Os e4 W F) 0) e4 ch et @ et @ et O et e F3 03 e4 0
- M G e4
- F) m et 01 es.
@c U
- o. w e
ki m o .e < * * * .e w .e < v .e e < 4 4 * * **e 84 4 4 4 4 4 w*4 e i a fe ^ h *e** M c e e e. es a w T?7iiEETY?Y?TEiETi.7i .E.M. c) MT@@?@MTEiT e v) < v1 Y 5 y eM M M M M M M N M M M M.e vi N ci e v)M M M .ip M ii M d. M N. ci. *. n. a. c). . . . . e $.M n. M.e.. m. MM n. M M M M M
- e. n.
. M. <. vi. . . . N h W) n 4 nhb g4 5 h n tfl W W) b h tf) n n tf) W eft of) If) e o tf) n tr) If) e M .
Es n e e x hnoceernOOO Ch nOOcchn 000 0 0000000000000000 AmeOOminOOce e4 C O O O i,2 OOOOO
. Te e to ^
e be 3
& + + + + + + + + + ++++++++++
M M M ka kl M M M ls3 b) isJ la) M lal M isi M ist tai lsi - H+ + + ++ + + + + + =m( a kl bl M imi inJ M hl M lal ist 4 et 9 w et e () n. r1. O. O. 4. O. n. n. O. O. O. O. n. n. O. O. d. O. n. n. M 4. @. O. O. 4. @, O. O. O. O. p 131 M N ED C C W h N 80 0 0 C O N 03O O w h N so eNOO4 hOOOO e
=
.m e4 8
et ce e4 e4 et e4 et et e4 et e4 b e4 e4 e4 et et g: - ; Me 'O O O O O O O O O O bc C O' O O we e E + + + + + + + + + + H+ + + + + .n a 9 4 ( e4 M M F) F). Fb. F). F). e). F) F) W .') . ., i F) l M. F). F1 F) gf e 60 4 4 4 4 4 4 4 4 4 4 ggt 4 4 4 .e 4 et s I C O O O Q O O O O O O O Q O O O
- ge M
+ + + + + + +
ist M
+ + + 4+ + + + +
l i F1 u e ki kl Is3 in.1 Is1 lal tal e Is1 M isi laJ (al i H +4 w c) n. .g3 r1 F1 c) ' M be M. n. M. M. M. M. n. ca. g en. n. . g e l 3 4 4. 4 4 4 4 4 4 4 9 4 ne 4 4 e eJ 9 W G C O C;
-4
- w. u si n 93
'a 8 0 6 2 c
0
- e. 5 b
0. N
- 4. .a Gl 0.
b I 0 e 16] b M m .Q a E tu a1 . D:
< .a si fr a> ai o
,A A Q Q O O A O A O O O C3 C3 C) C') C 0 s) E) N 81 NJ w a) kg at es u 5 a. ai m R];
= EM m ne = 5 as n ne 54 El S4 84 84 N M4 4
EI El . El u.l p
.a 4 m M .n.
u e f - e O . N M .: 5: e4 N . e)
- W) to N 80 42 e1 >M ID On e4 e4 U.
6- e e is. 81 l . l l 3.33 1 1
T,10E-8 , ., T,1.0E-8 . . 7 i______________________ E 7 _____________________,_ u o t0E-9 '- _- i_ o LOE-9 , y . o ' r o i o -, _ o _ E
~
E 1: s,_ 1.0E-10 u 1.OE-10 ._. s. o.' _ c. - E t0E-11 _ = T.
- 10E-11 , ,. u_ I it e <
e : }. o ' o c ; c I o 1.0E-12 ' o t0E-12 _ 1 y y s______________,,, l e e o - _ _ _ ' '. < o ' __-
~
."a a +- ,e ,..4 ,,
, "" s " ~ ' - "m . d 1.0E--13 . +-
O t0E-13 a >< "a 10E-5 1.0E-4 10E-3 t0E-2 10E-1 10E0 10E-5 t0E-4 10E-3 1.0E-2 10E-1 1.0E0 Release Frection For i Refeose Fraction For Cs u u T, 1.0E-8 . . . .1. . . a T,10E-8 .. .. .. , . . .. Percentile :
---_________ {
_____ - 95th __'-, o t0E-9 ~ ~ Me6 ~ ~ ~ { 1.0E-9 o y
'_, 50th 8
N _ .8* 5 _;.5_th____ 2 e ',
- u g _ _
u u e 10E-10 _- ', _
=
=_ 10E-10 o
o.
's *
- o. N 'M , i _
Y
.. s
- c' o-
- 1.0E-11 _
L E 10E-11 _ t th , e i e : i i : E _ '. _ E ; _ o 10E-12 _ i --
.o
,o t0E-12 _
-y e
e . : e : . o- : ; i
- -, . . M 1
'+ - ' ' - * -' *" 4 , ..
W ,""8 -
- n"' w t0E-13 4 w t0E-13
-1~' <" 4 10E-5 LOE-4 10E-3 10E-2 10E-1 1.0E0 10E-5 10E-4 LOE-3 1.0E-2 t0E-1. 10E0
~
Release Froefion For Sr Release Fracfion For La i Figure 3.3-5. Exceedance Frequencies for Release Fractions for Grand Gulf Internal Initiators,.PDS 5: Slow SBO. , w:~c' **r ~
-:_. _ _ _____ - __ _ .__' - ____._____ _ __ 2 a
MNM 4 M NM 4 MMMO4 MMMMN4 M MNMMMhMNMO 000000000 0 C 000000000 00 000 0 0000 t 6 I t t t t i e 6 e t + t 4 4 4 I e 4 4 4 I e + 4 6 8 e e M laa laa is4 la4 Bal bl Is3 8 4 let kl M b4 W laa 1.I1 laJ h4 la4 ist M M is4 N lal W bl M W bl g4 M. N. W. et. W. M. N. h. H. H. W. C. M. 4. m. Oe O. M. h. W. M. N. et. M. C. M. M. W. H. O. MN4 MMM4 4 MMMOMMNNMMWW MNHMMMWNMO 4 M4 M4 M4 M 4 4 4 CMW4 M M M 4 M W M 4 4 M MMMM O 000 00 0000 00 0 00000000 000 00 000 00
$ I t i 4 i i e e e t a + 6 e i e 6 e e I t e 4 6 6 e 1 e +
U Inl 14 M M N W b) Isl N lsJ lei laa Inl 8 4 W W laJ Ist M 163 Is4 M M la4 laa In1 In3 end InJ Ime W W W. M. W. N. M. N. m. W. M. D.
. M. W. M. c. M. M. . W N. h. o. c. h. m. O. M. N. M. M.
MMWh GNhh M. NNOh h
- M MNMM MMHhM NM W HO MM4 M4 M4 MM4 4 CMW4 M4 MMW MMM4 4 MMM4 C 0 00000000 000 60 e0000000 0000000000 I e +
e i e e i e e i e + t 4 4 6 e i t e e e t e s 6
>3 61 W Ina M In3133.e3 la3 lol M Inl Isa N In1 M M N ist M is4 In1In3halaa M blInt M M 11 N. M. h. M. W. M. W O. N. M.W.O. W. N. m. W. W. Me 4. M.
. N. M. N. M. W. M. N. W. h. o.
W NM4 MM M* 4hMOMWNMG MWh
- N4 hWMM NWO 4 M4 MMM4M4 MM O4 MMMMM4 M W M4 MMMMMMO C00000000000000 00000 000 0000000 5 I e e e e e e a 6 e s + 6 e 1 6 4 4 e e e a e e e a 6 6 4 +
M E NNMMWWWNWMMMMWWMWNWW NNWWWWMNMM r W o C. M.O.N. M. M. M. M. O. O. w.C.4 6 4 . Q h. h. M.
. O. M. o. o. c. h. @. N. O. C.
e NMNNMM4 4 NMMONNM M4 4 NM NMNM* 4 MM4 C U S M NM 4 MNM4 MMMO4 MM NMN4000000 M MNM MMNMNMO 0000000000000 0000000000 s t s e i e i0, h e e I t i e + b h 4 5 6 e + 4 6 s e t i e a t 4 M N ltI N M le) 144 IEI lL I lbl N ILI li4 N Id M M In3 Sal Int N W lalla4Ist M in3 ist bl M e O. O. O. M. M. 4. M. W M. C. e C. O. M. W. 4. N.O. M. O. M.4 0.W. O. W. W. M. 4.N. O. t e MMWWMNOMMNMONNNN4 NMW HMMN4 NhM 4C e M G MNMMNNNMMNNMM4 MNNNM 4 MNM NNNNNNM N OOOOODOOODOOODO O DODO OOOOOO O OOO t t t t t t 4 4 4 I e e e i e 8 5 e t 6 e i e e 6 6 6 I O t t b in) Imi b3 W 154 Is4 In3 Is) In3 Id ini MM N b4 Isa H lal M N 064 laJ M. M 9.M,O. m.M. N M. M.M.W. D. M lal M b) N M NN. W. M.. N e e M.h.W. M. O.N. N. h M. W.N. M. 4 WGMMNMM4 MMMMNWMM 4M4 4G4 HM4 MWNb MMM 4 NNNM MNNMM 4 NNNNM4 MMMNNNNMNN 0000000000000000 0 I 000 0000000000 O 4 4 6 4 I 4 5 e t t 4 0 6 4 0 t I l l I I e 4 4 8 0 8 0 8 A e bl ist kl Isa W im4 bl 14 M lil Is3 M 851 N M Ini Enl 153 Isl N M b4 ist tal Oml laa M Is314 lni m U n. M.M. M. h. M. M. N. M. O. W. M. W. N. W. M. 4 4 0. M. M. M. M. G. 4 4 4. M. 4 M. Q NMWGM 4 mmh 4 HmM4 M 4 NSMM hMh 4 NMMMNN
*O U H MM MNNMMNMMNNNMNM NNNM M M M M N et NN NN Om M 0e i 0 e 0 0, i a e 00000000 e e i e e s e e00000000 e e e e a 0000000 000 e e t 4 i s e e e a y .. W NNNMWMNNNM Cg bi o.blM.M N. M. M. M. N d. h. N.c.M W N. d. . n.d.n.h1 M. W. W W M M b.h.M. o. h. C. M. W. O. M. W. n.
g GNHMNMNM SMNGCWNM NNew GNGMNNMMNG M M MMMMMM M MNNONNM MM MMN HMMNM M MM MC QC 000000000000000 0 0 000 0 000000000 Q 4 e e t t t i e a 6 6 6 + t 6 e 4 e t 6 t t 6 4 6 4 4 e e e + o N N. U. U. U. U. N. U. U. N. U. U. U. U. d. d. d. d. N.N. U.U. U.N. N. U. U. N. N. U. U. m' %
. M S NGMhNOMMMhobbhNbNWW e WNhMhNOMhD O M N
O C e m y e o MMN4 N4 N4 M nN4 N4 N4 N 4 N4 nMMMN4 N4 N4 M ee C 000M0000 0000 00 00000 000 000 0 000 N 9 5 b4 W beh M b4W W W h W 1 W kl N NM i N N bl M bl M M W W N gby M k e p 6. W.O.4 0 4 0 4. N. W. C. 4 04 04 0404 M h.W. h,W Q 4. . 0 4 0 4 e g N M NO W MGN MMGM 4 MGM GMGM GMGM W M4 M@. MGHGM N O Qg h S 4 44 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 h4 4 4 4 4 4 4 4 4 4 Us aa a we C000000000000
+ + + + + + + + + + + +++++++TO00 000
+
M OOOOOO OOOO
=+ + + + + + ++ + +
oM e a o M w M M w M biha M M w sq klha w kab3bi w w a M w w ka ka N w w w M mC9 M o*
- a. M. N. n. 4. .M. e4. M. m. M. 4 M. N. M.4. M. 4. M. N.M. e m. M. Q. M. 4. M. 4. M. 4. M.
u g W e tr} 4 MhhMMhh 4 MMMhhMMMMhh'
- M4 MMMMMMM E. cH e WWN MOONMWWOON MOOOONM M
hWWWWOOOOOO
$a
=b 000000000000000 0 0000 McOOOOOOOOO 5 We + + + + + + + + + + + + + + + + + + + + H+ + + + + + + + + + ,M e
. 3 9
. M. M. . . . . . @. . .
W. hNWOONW4500NWOOOONW
. . . . . . . M. M4h4h000000.
. . . . . . O. O. .
Mm m e e M e E M M M M M M M M M M b M M M M M g M e O O O O O O O O O O ho O O O O e G E + + + + + + + + + + e4 + + + + + 4
> 14 M M 14 N lsi M H lal N M M M O'- N. N. N. N.
Is1 N. N. N. es4 N. N. b4 N. n N. N. N. N. N. N N M M W M M M M M M M M M M M M M M M N O
> h M O 4 4 4 4 4 4 4 4 4 4 4 M4 4 4 4 4 u C C O O O O O O O O O O O M O O H e + + + + + + + + + + 4+ + + + + n C # N bl I.3 M b) in) Id 163 lal W 4 Est W BaJ In3 N fl 4 w L WH M. M. M. M. M. M. ' M. M. M. M. A H M. M. M. M. M. g 3 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 9 e a R C O C M U
* = .H <
N U 1 F m e 4 o 9 k- M M e C M n 4- A
= 4 P1 s' q* m si b b u
= 4 E3 4 k S1 A M g) n m o 40 O O A O O O O 00 O O O O C]
mA =k 2 m b b 3km mk bm km .b &A 9 k dkm .k m M e2 .l M N W i M w b T g et N M 4 M W h W m O M
>M M O N
M M. e b = M CI h b 5
)
3.35 1 1 1
t! 4, [ * , !,c- 's !l ;l!t
-i i: ( * , :
0 0 E E 0 O L
, _: i .
+td -
t , e _ _ - - _= : ~
" 1 .l i _ _ .
1 -
.t _ -
,. ' E . n b_ n h.
_ E 0 0 et a- t . cr95_ t ' t 5._ a
~
.' s .e .
' C
;P1
~
2 L 2 ..A - r
- - r E o E o 0 f O F
,.'-=
a 1 n 3t
- c i
o -
...- i L n 3t i o
c r f o E- o
._ .' E ar s Fr n
- 0 F ,
a O L o -
- - 1 ,
,?. e e i .
s - s tO u_ _ -
- o 4 a e -
4 a
-e El cB aS E- le - 0 e. r
- ' 0. R 1 . -
.a 1 R Fw 2_ - " - .o
_ - 5 el 5 _-
,. i - sS
+
- a
- a._
_ , j
-
- E 0
- 7- - ^ _ :_
- e E 0 e
_- , _~ - = _- - . 1 l : 1 e6 0 1 2 3 8 9 0 1 '2 31 R 8 9 1 1 1 1 1 1 1 S E E E E- E E- E E E- E E E rD 0 0 0 0 0 O O O 0 O 0 O oP t t 1 1 1 L L L 1 -L 1 L f T x1uo h eu'eO 6 y eoComeood ss Co*w[ euueaw7Ew.oComeeod' - er io ct na ei 0 0 ut E E 0 qi
= - - _5 - - #_ 2 :
1: - O. 1 e: : .,;;:_=:~,=E ._ A 1 en rI
',,I "
1 F~ -
~
' 1 s',.- + l E- ' -
E
- ea 0
0 cn nr 1 'd 1 .
- . d .
" - ae 2 S dt
"- 21 -
- r en
, ' . . ' E- or E o eI
- 0 F ,
O F c
- - - L y ri xf
*- . d* 1 n .
' U El
" i o ,
' io u 3t 3t g
- c E ar q E- oc
.G r 6d
- 0 1
F , d 0F 1 n d _ 3 a u - e _ ' e s r
- s .-
- w. -
4 o - 4 a 3G E- e
- E- le e
_..I' _ f
- 0 e - O e r
_ - ~ t R 4 _ d- L R u
. - _ , g u.
a 5 _ _ .- " 5 i~ _ - 4
- - - F
-. - E - E-q
,;- - 9~ : 7~ J.3 .
0 -
- _ - . ._ - , - _; . +0 t .
1 0 1 2 . 3 0 1 '21 3-8 9 1 1 1 1 8 9 1 1 1 - E- E E- E E E E E .E E .E E O 0 O 0 0 O 0 0 0 0 O 0 L t L 1 1 L t 1 1 t L 1 l Te3tuoheu ueO62 eoCooeeo5 - Cos[. eu 'egwo*& eogcoeeo5 - _
+
Y$ -
.- , i; , . 4- ' - i t
i M o M -O @ G W-W McMM M CC d MMW O M cMMM O4 M4 M f o oOOoooOOoo e o i c+ ce o+ ci oe ceoeoeC+ oe O W W bl W e c e oe O. o o. o. o. W W W b3 tal W N M M fr4 W W isa W b4 e e e + e s i + e i e e - 1 W W W W W bl14 W W W W M. o. c. o. . H. M. M. M. c. W. h. M. N. h. h. w. p H. N.
. H. O. W. h. W. o. M. w. W. N.
NOW ONNMnWONh M MNNNMNN WONhmowmeh d onoO O4 eM o4 M 4 MMM4 4 4 4 - M o4 MmO4 MeM coOoMM ooooooooCOoooo Q e e + e + e e s e e + e e e 6 s e e 6 e e os +o o.6 to00 + e e 000 e e 0 W l.1 N bl kl kl M W bl tal W W in) b4 es3 N N W lal b4 b) Sal ist W W N is) W M I.3
- e. O. M. O. W. W. M. M. n. o. M. h. m. M. M. M. o. M. o. n. n. o. M. N.M. o. n. e.h. m.
MOMOd 4 M MM ONM M NMMGWMM MONNNOMMW N W Ot ooOMMM OwmMMMMwW MM mow MMoMMMM O OoOM M cCocoOOooOOOOO C 00000000 0 e + 4 + e e 4 e e + e e a e e e e e e e e + t i e + t e e a
.s J Is4 W
tsJ N N M M M Is4 Ina W ist M tal tsa M M M bl bl W W M W lat Int W h1 W W W
. o. h. o. w. w. M. M.M. O.W. M. M. o. o. c. h.n.M.M. M. o. W. M. O. o. m. O. W. M.
McWO4 whhMOMMMMW eMMMM MOMMNOW et W N MOMOG mW W McW M W MW 4 W MMM MO4 MMOd M4 M p C0000o00000000o0 i + e + t 0000 0000000000 i e e e + t e e e e e i e 6 6 e + 6 6 e + e 6 e e e C E WNNNWMWWMMMWWNWMMWNN MWWWMMWWWW i M c. W. O. M. O. W. 4. W. o. d. N. W. o. N. N. O. N. h. h. W D O. 4. h. m. o. o. w. e. M. l 9 M ONOMMMMW CW WM
- NNwNWW 4. 04WWOMWNW 0
e mnmOMO4 4 M OMM ww4 4 MMW 4 MOMmmO4 M4 M d k 00o00 0000oO0 0o0OooO 4) oOcCoDOOO b k e + e + 6 e t i e + e e a i 6 6 e e e e t + e e e + e 6 e e e m W. .W . W U. M. M. N. N. N. W. W. W. W. N. N. N. W. N. M.U. M. S. $.M.N. - U. N. N. N. N. .1 4 6 NOMOMM W 4 MONWOMNNNNNN MONWWO*hWW -< M 4 ' e MMMmmmMMNMMNM4 4 W MMed NMMNNM
- NMN k Co000 000 00000000000 0 0 0000 00000 e e a a e T s e e e a e e i e a e e o e e e e e e e e e e e e b W W Ia3 W W Iml M isi W M W W M N N N M N In3 Ina M bl N N M M bl N W Iml W
- n. M. M. M. W. W. N. N. W. M.W. E. n @. W. W. Me O. G. @.
e
. M. W.W.o. h. n M.M. n.
e WMh MMMMM MW@ MMMhh4 d OG M@@ HNMENMM l NMNMWWMMNNMNM W MMMNMM NNMNNNMNmN O 00oo0o0o0000 e 6 e e e s e e e a 000 00000 e e e e 6 e e 6 e 0o00000o00 e e 6 e e e e e e t M S W w W W kl {M Inn b4 Im2 M M M M .o.I N Is1 153 M MM M M laa W M M tm3 M w e.lb) H.la4 M. W. w M y
. . m. m. e. W. N. m. M. e3. W. W. W. <. m. m. m. W. N. w. w. M. M. O. W. W.
y MGHWMr 'MMMWW WWM MMMMM MMW4MNHMMM Y *@ NNNNCOMMNNMNNMMMMNMM OW NNmNNNMNMN M 00000o00ooo000000o00 e i e e e i e e s i e e e e e e i e e i o0000oo00o e s i e i s e e e e c.. g s
. . . . .......M. . M. . G. n. . . . M.
M M MNMNWWenneoaMMMnoMew.. . MWeeneNWNw yA M OMcMMM M M oM NMNM MMNM M 00o00o00000 00 0 0 0o000 M cMNMOMNM N N N 4 + e + e i e i e + i e e e e i 0o0o0oo00o e + e s i + i e e a k n' Wm N. U. U. U. N. N. N. N. N. U. N. N. N. N. <. U. N. NOh0HMMMecenceMMMMMM e WOeMh ohNhN N. N.U. N. N.N. U. U. . N. N. N. O y N g y o e E. w mM e N<N<MeM eNemMN<M eMMm< ~N<MMN<MMMM 200 It! pococoOoCoogCococooOMaasta+Aaaawaaaaaa'+ "ROooCocooCo-waaaaaaaaa
- a nbu M. ,
k W
- 4. W. W. . N.
b 0 M WQ MMMMNNNNMM* MMNhN4MhN
. N. W. W. h. W. W. W. N. N. h. W. . . M. W. e. h. W. W. W. N. W. h. W. 1-O c
< OM M*MMM* MW M ( m k M e Ug e a p ******4 ooooooooooOOoooooooO
,4 4M e<weeme<< 64 4 M oooo M44 wn*Me o a ha + + + + + + + + +++ +++++++++
j l M tti g e e M aw en a ks ha M ba bi w W ba w M ba W ks W M w ba W w w M. n. M. M. O. m. o. m. M. M. M. n. o. H. O. W. M. n. o. m. n++++T.oiCOo a ba b mM. wwwWW
+ + + +
u g M ' HMM MMMMMMMMMMMMMWMMM o u M. n.n.M.n. M. n. M.M. n. MM cMMMWMMM N y 0 C D M e M e CoOOMMMMOOWON OMMWOMM .M e h oOoooooooobooooooooO h0CWWOOWWWW A M oooooOOOoo # e bi bl W bl kl b3 a M N b) kl W b bl kl W W W M N
- .w 4 b1 b a blkl W W a W o
Mg G. o. o. O. n. h. n. h. o. o. M. h. o. o. n. h. M. h. n. h. M O. O. M. h. o. o. M. h. M. n..> OOOOh wh4 DOMGMhh4 Meh4- COMWOOM WM W W M 8 b M M M M M M M M -M- M M . ce C o o O O O O O o o hO M 'MO M O M M c 4 0 + + + + + + + + + + O O M e w b4 N M bl bl Isl M b3 M N M +- + + k b4 N tal
+
bl
+
M A N H N. N. H M M
+
M M M N. M N. M N. M N. M N. M N. M m
>g N.
M N. 9 N. M N. M N. M M g e
> k W M M M M N O M M M M M M MM M M M M C O O o O o O O O o o be M oe + + + + + + + + +
o O o O O
+ M+ + + + + n a
.o e
M. - M. M. M. M. M. M. M. M. M. (i n. W M M W M M e a.
# a C G e
- M u
- W k 8
_C W 3 A 0 .0 .k 0 h ,0 D4 N M 4 *. 0
,* Ad ea 0l A 01 and al M
g h e at
- 4 W
- a :. a3 d a 3a s-M e o
- W n M
- b. o e N UM N M 4 M W h W G H M e W <
k e > M W M M M M O H A e 3.37
.icf!tI . ti+ j , tl :;;[;l..*i!it tI' ,l3l.iIIll,; ,;; t' ,:i ,>!l!i}i!:ljl> I!i m+
_= ~ O O _ E E _ C G _
~t ,: ,; _ _::: ,; ,, : L
.'.,'.. * , . 1 E
te -- hnhr 6 h 1 _ E e O , e9 e0 t_ t h_ G
- L r0M 5 4 L s _
e C e -- 5_
.F 2 L
' 2 _
.. r r
- E- o +,,...', E- o OF .*
O F L n> ,
,,' 4 L n r o o 3r x
, *, 3i i f '
.'+
c .
- c E- a
',', r
- E o s O
L F ,- O L Fr n~ e d e o 4 _ s - s i '.
- ' - 4 a -
4 a tO E- e cB E- e
- - - v l .
aS
- O e -
O le r
- ' *- L R ,
- s L R Ft 5
5 ea s
- '.,- - sF
._ - _
- E - E- a
_ , - - . : r_ r - - _L G L
,: - f. :' : _ _ O L
e l : 0 0 e7 5 6 7 8 9 1 5 6 7 8 9 1 R
- - - - S E E E- E- E- E- E E- E- E E- E- rD 3 G O P O O. O O. O O O. O. oF 1 L L 1 L 1 L 1 L L 1 1 f
$e;tuo ooE uenYoy cocoveeoO T >.[o ge6. eo_YcEw eucou*eu5 i ss er w
io 0 vt ei a E C 0 E ut O qi _r : - , - -
.. y,E s3 1 :: ,j- _; - ,3 - - ,2_ _- L en rI
.\'.,*. 1 1
F l
~ E- , ',. ' E ea
. O , O cn
\ ,.
L , ~ e L nr u
- " r ae 2 S 2i -- dt r ,
- r en E- o ,' ,-
E o eI
- OF - O F c
**..*L L n>
x .- L ;I r o xf El 3 r. , 3* i u
.G
_ _. E- w - - c E a r 7d
- O L F 1
O F - n a. e
- L e s -
s - 3 ar 4 c - - 4 a . 3G E- e e
" . i
_ O e LR _ E-O le r e s. d L R u-
..*** - p
- g s
- _ 5 . >
5 i
. E-
- ~*' -
F G
- .', E-f, ,_ : - 7:i _ - ~ : ._j _; : s: - - - ,, : - _ O L L~
o 5 6 7 8 9 j 5 6 8 9 0 1 E- E- E- E E t_ E-
- 7-E E E- E- E-O O O. O O. - o O O- O L L L t O. O. O.
1 1 1 1 1 1 L- - 1 - Te>1so ooE Leoyc] eocovoeo5 - e>io geu s.eo.yo*d euccmeeo5' - _ v g
, j , i1Ii I-4 1l, . ;ij i1g3 ;4 ;, .' ];,i4 . i: i - j- ,
4i,j :
i n ! Y l M A. ...no. no.nnnenen n e n n n n . ~ n ..
? ??????????????????? **????????
A W NTWNUMWWWWWWWWWWWWW WWM%NWWWMW adAAAadAAA A daAidaadasAAAdaadJa
. .n......o......nn n nnenn nnnu !
. ? ???*???*???**?????* ?*??????*? I u M MNWWWWWWUNWWNWWWWNM WWWNWNNMMW '
i dAAidaAdiaddadisiaA daidAidAnd < l n .n..on.. n...n.nn n nnen..nu.n
. ? ?????*??????????*?? ?????????*
- W MMWMUNWWWWWWWNNMWWN MWWWNNWWNN A dAAdadAsiddAnAdddAA adJJaaAAAA ,
.n.no..no.n.n.non 1
n
? ???*???????????????
n nnen.nnnnf
?????????
j at U WWWNWWWWWWNNMWNWMWW d idididdAJAddAAAdidA WMUNNNMWWW didAAAAAAA n . .no .nnnen n n <nnunn -n-nI.?????????? E d u ? ???????? ????*?????
- W WWWMUMMWWWMENWWWNNM WNWWNMWWMN
; a dAJAidadisAAAdidida JJidadJdAA 5 nn.nnenennen. -n-i
- e nnenn
? ??*?????????????*?? n-nn!n
???? ?????
W MMWWNNWWNWMUNNMUNNN- UNMMNWWWWW AAddisiAdd d aAidAddiddAAAAAAAAA n mnb,nnnnuennnnenn-n. n-nnnnn-n-g~ ? ??.?*?????????????? *?????*??? M 0 W N. W. N. N. W. W. W. N. W. W. N. T. M. W. M. W....W M.W W.WN. N.N. N. M. M. W. W. y eN.-e-e-n.-senn-NMn -NnssN--.N 0 0 . n-n-enn-ne Um - ? u n n n n ?n n? E? u?n ;,'
?*????? n n n? n?u?e ?.< ?n ?n ?????????T 5 .. W WNMWWWWWWWWWNWWNMWW NdNNNkNWMN 5m a dadAAdaAdAdiaddJJaa 'id J a A A A A J
$H e n-n-e-n-a-n.<n--~~~n
--<pf????? n---n 2 ? ?????*????????????? ?*?,
m$
"% d R W WMMMWWWWUNNMWMWWMMM a diddiadiadaddddiddi e WMNMMWMWNM iddiiAddid m m W :
~,b8 n
.8
;WT
- .n n.n n
? ???????????????????-h??????????
nne n nnn. Lenn n n nn 3"*
&E N NNNWWMNWWWWWMNNMWWW i dJuiAddiAJAJdadJasA 8diddJdadJA WWWWWWWWWW j
\ 181) H, u g . w . . ................... 8.......... # E
@H lt- ? ??????????????????? W?f????????
m{ 3 .* O M NWWWNWWWNWWWWWFWWNW SWWWWWMWWWW l
&* J daAAddiAAJAdidAdida g
e .
'fdddddAdidi e
g E a H . ewe weeeeeeeece
. b ,-
l a m? .?., +.?., .g, o c e o?o? m
. , .? ,?,?,.?.,m? ? ?e?w? e ? ?e ?e?e et.sm?
** *eS. ?* .?., ? d a * '?., .?.,
* ?., .?.,
- m? .? E, m E, ,", m . . -
- a. *. "l a. a. *. ': *. a. *. S, ,*, m "
- W *. -
26 4 m -.-oo.-oo.s
"*a*-.sne 5 s - - a.-.s t u .
.B. - - - - - - - - - b- - - - - .
n C' W W N N N N N N N EN N N N N d - WWWWWWWWW wW W W 4 4 9 O g . . . . . . . . . . m. . . . . m
? ? ? ? ? ? ? ? ? ? ,7 ?? ? ?
R. n$5 W W M. M. WWWWW M. c*W W. N. M. M. E 2
.e. . . .
. b g a w m .m m 8 c A g, W
. - 2. 3 9=
pd ..0 m a N . N p M Al 83 g
,d
- m. - m - .
ni - m 3. 3. 3 e] a
- 3. 3. 3. 3. -
o e N? O
- s. eW b -
4 gd n n . n e ~e . 2. M N N 5 n C . 3.39 l
0 D E E _ 0 O _
- . 1 , , _ _
I L legti-. .
.e
'.*',..*,..t.
\ 1
,i l t
i ih
. 1 E
0 nhot et p. E-O N
. 1 s
c5f0 r9M 5 t
- L _
e b. o _ C
' 2 P ~~ - .
2 l
- r E o N - r E o OF
- 0 F .
- a 1 n o
',.. .'s L n> r o
3f i \ 3f k f E- c E- e s o . or n _ _ , . 4 0F 1 r e _ _~ 0 1 F e i o
, - s tO
. - s - *I ~
4 c 4 e cB
,', . 7f
- e N
=
-e Et aS r
0 e ' 0 e Fw 1 R 1 P o
- s. s el 5 " 5 sS
- _ ~ a
- - . E- ~ E- e r- O _- [ - : 7 , -
0 l : _ L 1 e8 6 8 9 O 1 6 8 9 o n R
! 1 t S
- - - 7- - rD E E E- E E- E- E E E- E- E- oF 0 0 0 O 0 D 0 0 O o O f 1 1 t L t L 1 t L t L ,
ss 7.>- ;Eu <- &2. egoveeoS T* o gE seS&E eocove od er io ct na ei 0 0 ut E E qi 0 0 en 1 , : y - _- : 1 rI s _= _ ,_i.] - F
' 1 .
1 l ea
.. . E- E cn
, 0 t
O nr L r ae dt 2 S 2i r - r en eI
. E- o E- o c D F .
0T, xf _ s a L n ,
.',' r, 1 n El
'. 3t- co i 3t i o
c .G u . E or E- a 8d N s 0 1 f -
.' 0 1
Fr - n 3 a e e
- _ " s -
s r 4 a -
. 4 a 3G E- le " .
E- le e 0 e 0 e r _
" tR
_ ~ a 1 R u, _
.~
~
r _ i
.'s.
. _ + 5 ~ 3 F
. _ ~
. _ . E- E- ' ..
0 1 _ ~ _" _ 7 .;- 0 1 0 0 1 i 6 8 9 1 1 6 7 8 9 1 t . E E E E E E E E E E r O. 0 1 0 t O L 0_ 't O 0 1 0 1 O L 0 t o 1 1 t T,>- gE,_ u rEwegooeeo5 e - T* o g 2 u,S & E eocoyeeu4 w
. ,f;lll;,Ll' !t
d 4 4 4 i !, O.B.088.E.C.T."S." v v 02.0.E.8.f88 . P"v8.v*8."v800 1 4 e tt e W t' N e W e M e W W W e M M W e NeWeWeNeNe
- c;s.4dida4W456Wdadaaas asidasa4WJ a
- 3. 8. i. f a. E. a, *v0, f
- 5E. 8. 0. a. S e. 0. *v f> E.*B.fa.SE,S."vE.
t 4 u e e "! !! N e N e M e V e N M h tlW W W e NeWeMcNeet l
- a. i .a i .a s i s .a i s a i d a d a .a a i siaiassidi
... .-o.
veveverg.ftf....gF.v.f .v rev. e v.o.e r f .et y. t e. t,
" N e N e w e s e n e M c M N tl e t' e e s w e e e W e N w M t!
i- diaiadidisisdausiias isdiaisias 4
, rev
-o.i.v e f e v .,!-e.fm,.f vvv ver e ff.fe- v v i
; - meeeweeuweNewmetenen etwetenese j y .asa si sa s aia s.4di s a is i asasiaasai
. x .o. -o t ,, .fterf e vf ef e f-- eve f. evre .. vee f.f e e fe f eenereeseeweerNewwee eeeeeeeeNe
- a6s4WJ.;sd6adada44ss4 asidss.44W4
- ;" -.-----.N.---N .. . -.N.-. .N.
1 . 1 vvve??V??v?ve*??vef.t it?v?*?ver NeveNeemen emeeredereeswweemsNe saaaaaaaa; saaaassaaasasadassaa
~,N. .N-~ -NN~.---. -~
v*vivette ?vverv*ve ?? c. v.ve. e. e-h 6 WeMeMeWeNWNNWeveWWWM seNWWMeneM wa aaaaaaaaaaaaaaadaasa asaasassas N NNN--NNNNNN-NNN-M oa - eyevyvvvvveyyvvvvv y fN NVfN-NN-N-y vvvvvvv yw eunesmNueNNeewwemmeN nmedeNxnee
- e.. a.Jadi.Jadiaaa.4daadWsa adaasaasa4 um Q - - - --- --- -c-o- -c m e y e v.gvge vet ere . vevet erv.g m'
"w$ @. R W el!e v en eM eN e M N Ne W We e neNeweeter asasssasasisaassadas ,asasasasss M ed
*Et +!-
is gic;cce:gsygyggspses [s : e.~.. g. ~c. ~a. g~ .e. .g .ee. g m u. e.n.gga aygezegy
- 2. g u eNgegomast 8 , .,
u+
- .e N -------- -------NN--
g......~....
-------.- T s
uc e ..... ... i i g- ,e_ g etterretreeterreerte teeteresee 3 m-3: wwwmmxummwwwwwwweeux :twxxwnnnmxx : l &" aaaaaaaaaaAaaaaAsiaa Luaaaaaaaaa g c f c b. E# o .
- g n o Z s re- o e f g.ee o o f f, oe . e.
ogfg.ee . e mee.efgoe 6.o o g o
. fo t .:
2p aN esteeweverweetestees steresteees sesssssssssssssssies i 7sssssssess : E
- - - - - - - -- - - D- - - - - 3
*i ee e e e e e e e e he e e ee z I N N N N N-W W N N N WM N N N N M i i i i i i i d i i #a a i ai 1 p - - - - - - - - - - l mi - - - - m t
-g- ee e e e e ee e e ,e e ee 6:
i" ewewewewee i- A iii ii ii jeweee E . , iiiii a l E $ $ E a u i t1 A h n a e t .
- ,, 1 v e
bb b i.
- bbhbb .E v 1
.. E E A j
- .. . . a e. <
aeg- ~.o N - N ; c , 4 1 3.41 I i __ _ . _ _ _ - , . . _ . _ . . _ _ . _ _ _ _ . _ _ _ . , _ _ _ _ _ . _ . . _ _ _ -. ~. --
t :!I,!' t. ,-lfi,? ,s i f,,r1i[L(;.; r !!
- jI;t j tt ,-1r [L[;f tti s >llIi!. i;!-
_ ;t:& O 0 _. E C E 0 ._ . L , _ _ t _ te - - 1
'- t i - -
E
- f n h. c g oy -
E
- ~
_ A', 0 e t_ eg th- O
,,s;.;
- 1 s
cr c5_Mns e __ -
$__ s L o ,.
- +
2 C P _ - 2 l r
- r E- o
_. ' E o 0 F U F
;,..*a 3t
~
L n 1 n r =
~ o ... i o o 3t i
. f
- c . c E or 0 T *
,,,,',- E- o 0 Fr s n
t 1 o
,- - - e -
e i .
.' s s t S
_. 4 o . 4 o cU
-e -e El ,. Ef 0 e aT rA
, 0 e 1 R F t
- - '- 1 R . +
' s
_ 5 _- 5 e a s F
- . E - E- a ,
7 0 . ~ ,= _ - - _
- 0 e 1 l 1
0 1 0 1 e: R 9 6 .i 8 9 1 1 6 7 -8 9 1 1 E- ,t E E E E- E- E E E E- E rSD 0 o O 0 0 O 0 0 0 0 0 0 o 1 t t t 1 L 1 1 t 1 t 1 f F
$L8uo voeu seovdy eUcoveeU5 - . s s, Te> uo ooE ieDYvEut eucoveeo5 e r i
c o nta 0 0 e i E E u t 0 0 qi . ,: 1 , ,= - ,- ;
, - _=
1 e n _ , _ _:- ,L .J . r F I
' - 1 ,
1
~
e la i -
,,s E E-c 0 0 n n 1 - .. . * * , , = . 1 r a r 1
2i .* 2 S dee t
- r E o ,.'
E o
- r e n
- 0. F ,.- G F cI L n.,
Exf
, ,- 1 n -
. ,,,4 o '
o l
.' +
3t i -
- 3t i
u
- 4
- c E ar -
E- ac .G
.- r 9 d
' 0 1
F - a 0. 1 F n
- - s e -
e 3 a s s r 4 a 4 a 3 G E- e E- e 0 e l 0 e l e
"+ 1R -
1R r
- a u g
5~ - V.' 5 i F
- - E-
_- . E-0 1 _.- - ,- _ &- 0 1 s 9 0 1 6 7 B 9 0 t 6 7 1 1 1
- - - - - - - - t_
E- ,:. E E E E E- E E E E- t 0 o O O 0 0 0 0 O O. 0 o-1 t t L 1 .1 1 1 L 1 1 t Te>Iuo ooE :.e.eYoEu. eucomeev5
$Lf i o ooeu secv&y eOcoveeUd e
th" , l e
- j ' 1! r1>:1 ,s ,.]jia,.1' , , !e 4 j ,1!iI a .! ,;:)i. i]s;<' 2:i.! j i .: 1:l. Ii 4i
i
..N.em .N se...em.m.
- m.....--.f vveryvere tevvvveeve W M M t! N N U N N N M V N t! N t! b! t! N !!
vvvvvvvvve
>! W t! W ti t! W N W M usaasianaAdanadiadaa daiaansaia J vv?vytevvvvv... vvvvvver ._~, veter-evvy "i T tl k b' d M N W W M M M W t! M b' M M W b! N M lI M W D b! M M W
\ daaAiaadadaAAAiaadaa daaddaiadi
...mee~.. ..- . .....
. tvvvvervee.... eeveveeyey . eveevvvete T W W W W W W W N t! W M U M M W M M W W aadAaiididaaaadAAdia WMMTMNNMWW anddadadaa v v?? tty?gm.~ ~.. .m~.,??vt?Yv?tet? ~e....
* ?????????v
[ M M N W W W M b! W N W W M T M N N N W W WWNNUNWWNM
- essaAwaadidwaaasaidi JeaadswiAa
...... . . .~ --...
s a u ???ty???v m.f~??f~. t???vt? ?????????? WeWNwMeeve j W e e N e W W W N b> e t! e e W N M e r e aAaandaaaAAaaaadadas daAaindida
- --..... . ..eNe.---N m.m..--mem evvvvvveveetevvee?ee eveevvetee M MM N Wt!0 W M W UN N U Mt!M W Mk tl W N M W W M O M N AausiaadadaAdaAdanaa adaaaaaaAA l nnnn...n.N.mnn.N nnnfteryvveteveyyeve vty .nnnmNnnmN vertevet?v a WWWWUTWMUNNNMMMMUWNM NNMWMWMUMU uh
- saaaaaasiaaandaaadas asaaansawa o MNmHMMMN. N. N.M
- - eervetvvvveyey.NMN.vvvettN .NM NMMMNMM vveytetvey yM WNTWNNMWNMWWNNMNMWWW WWNMWMWWWW i.. anuaaadiaAnadaAasian uo uddaandada QM M NM NM NmMM NN NNNNNN NNN O N NNNN NNMHN
"$" ???viiiivvi?iv???iii ??????????
E UNTWWWWMTWWWNWWNNWWN WWMMMMWWUU
- m. w 0 m . iWanaduaiWanadi45Wea ,.aaiWaadAid
. t. :
a! esv 3** m.e.eemme.e. retterevvreev: evere h.-...mme.mm NNWWWWWWNNNNNNNUMWWW
.mmeme, erve xter :
g,3 a8 NWMN#WMNWW ] uo u-JadadidAdadadadidada fJadadidada . ! 85
.................... s.......... 1 m
- t' -
???????????????????? ;?????????? *
~ 2. m: e N. n. n. n. W. n. N. n. N. n. N. n. N. W. cN.N.n.n.n.
g g M MMMMMMMMMMMMM. MMMMM. N.n. n. N. N. W.n. W. n. N. an. N MMMMMMMMMM g
. E m . . . ...~ ..
y W- t m z
~ . .. a r.~.m
.env.,. en.r.- m- .e . ~. t .e.,
. nr, f,.m,. e,
. f .re.,
f., f., l, ~
. e. . , e, m. t,e. t m. . ... mm ~ e.,r vr
.a .
N . N . N - N N N . N . N . N
.n.n.nn.. . - . n. ,r. ,.e. nv. m.
ee N .. N N N . n. N N N . N - . zN 0 :
- - - - - - - - - - b- - - .
3 k Y Y I k k k k k I o I~ N. N. N. NY
*N k Y T k 2 w N. N. N. N. N. N. N. a n. N. N. N.
n M n n n a n n n N. . M M n n n a.
, . . . . . . . . . . m. . . . . e E $" ? ? ? ? ? ? Y T ? ? ,V ? ? ? 7
.N 3
W. W. W.W W W W W W W . . c W. NW W. W W E N .
.N N. N. N. N N. N N N.
.N N
N N. w g a p n
. gil 0 21 0 0 0 !
a n 1- W W em Qde Q J [! 2l 0eme .e i
.o m ai m.
1* mm W 9J m - B*933* m em m m e !Q
- me 3m 3 .3 m I- E" A gM N M . . . ~ . .
ae e C
. - m - N 3.43
.a
- , , I( k i:! { i-l1t I>: r iIf it ! -II;! "f;>!iI )i; ti, l,! !; ,ti,ij! t tI:!
!!I .ii ,
. 0 0 E E 0 0 ._ ,,t ,i 1 , ~ _ _ , _-.- 1 l e - . 1
'* 1 ,i i
,*',,,, - t -
, et n h 6-h _
*,;.*,. E 1
0 ,c5 eg t h. t_ E 0 1
, 4 s ,e roh S 5__ o 2 C 2 l
~
- .- .P ~ _ r r 1 E- o E- o
,' 0 F 0 F r o
1 n
;<i'*'...', ' 3t1 no f
3t io i s
',,.'- - c E ar ,,', - c E ar n
' 0F -
- 0. F i o .
- 1 e t S 1
_ - ,. e 4 o s
- ' 4 o s cU a T e rA E- fe E- fe Fwo e -
0 e . O
- t R -
L R
- _ - ' el s
a S
' 5 - 5 _
E- e _ e E O :-
' _ - . _ _ - - .L 0 l :
- - - : ,s -
L 1 e0 R 1 0 1 0 1 6 7 8 9 1 1 6 7 8 9 1 1 r S E 0 E 0 E-0 E 0 E O E-O E O E-0 E O E E-0 E 0 foD F O. 1 t t L L L 1 L 1 1 1 1 s , e s ,
$exLo 0o u seO6E' e cooee 05 -
$Le o_ooeu ue cE e0cooeeu5 - i c r o
nt e a 0 0 ui qt E E
~ _= _c - - _
_ t a1 <1 0
-[ . - ,- 7 - - - _5 :: -
'1 0
reinI F cela 1 1 E ,.'- E 0 1
.,- 0 na nr
..*e 1 " 1 r d e N '
2i s
" 2 S et ce n
- r .- -
r E o - + E- o x I
, '#= 0. Fn ,', 0 F *,
E f 1 - - 1 n
' i o -
i o l u
' 3t - 3t
- c .
E- ac 0 .G _ E ar 1 r 1 d 0 1 F , a 0 1 F n a
. _ e _ e 3 a
- _ - s _ '
s r 4 a _ 4 a 3 G El
-e
- El-e e
_ 0 e - 0 e r
- _ 4 t R - " 1 R u
' - g
_ - i a 5 E 5 E
- F
_ ~ _ 2- _. - _ : _
._ _1 0 -
' _ - , _ . - % _: : O t L 8 0 1 6 8 9 0 1 6 7 9 1 1 7 1 1 E E E E E- E E E E- E E- E 0 O 0 0 0 0 O. 0 0 0 O 0 t L 1 1 1 1 t 1 1 1 L t Te>[o oOeu uoO&* w e coooeo5 - TLi o ooeu a.o 7y ,ogot e05 v p*
" 8 : 3 8. "4 B*C e+. . T . 8 0. 8 0 8+ 2 0 8 8 8 8 4 8 8 "v8. 8 8 0 8 0i+ 8
.. WWWWNUMWWWWONOWWWWWW MEWWWWWWNU assididiaidsAdadAAda Jiaidadiai
. . 8. :. = e :. 8. :. 8. :. 8. :. 8.s 8. s. s. a. c. a. s. :. 8. :. 8.. : 8 :. 8. :. 8 u WENUMbMWUOMtWENWMhMN MWWWNWNEWa
) dia s addiJiis aia aa ;a a a.;JiasJaii
.enono.o o.o o..eeoo .c o.o o.o
. ??verrverevev?vvvvvv veveveveer weetmeetreesWsuwwwWW eer Westes iiiiiiiiiiiiiidadaad iiisaisias me.
??vfv.er.fngngmonnnne. .omg.omono vet.viveve o e.e.eStevert g - MwNwere=x Wst:MWNxNN ewwsetWez:
addiniidadaiaiauddaa diaidsaiai y c 6. re eee me.te.omfnomfmonNsneeevevvvev
-et eim8.
v e viv onfnove a WezemexturseNeuwwwwe wuxeNwWwww
; didiiddiasiiAidadiaa diadidadas
.: manamnN.N . mnN mN.. mo N.N m N.n.
eveveteve-veveteevet ever??evet WWWNNMMWNTWWWWMMNNWW WENWWNWWWN idadadaiaAddaadaddad usadaAadAa Nmmen NmN N m N a n .e N . m . everv-eteeve N.?9NNmmaaeveeve veteveyeev 0 MWNWWNNWWWWWWMMNNNNE MWWWWWNWUN w0 adiadadaadaaaaaaaaAA anadaddiid HU a M NNNmnNNNNN mNNNNmMeto NNNNMNNNNm 5 g z. iiiiiiiEEEmi$isisiss aaaaaaaaaaaaaaadaaii IEEE555555 aaaaaadaai UM -o-o-o-e-o-o-n----mn revreeve=?terteverev wo-o-o-o-o
??ver????e T$mA R WWWWMUNONWWOdWNMOWNN asisidaiadisaiaaddaa.diadidadai MUNWMNMWWW
- m. " " .
ma g : J05 !?3 E!E$$.$E!!.!!!!$E$EEEEh!$E$$$55E$
- -n. . W. . .W. .t. . . .W N N W -..~..:.:.:.
ge g-----~~--- I a: a- .
--~~-------------NNN w ,
v. u . .................... s.......... j ~e i n ZZZZZZE!!!!!!S553355 ?.!!!!!BZZZE I g N We W
. n n a. n n n n n n n n n. n n n e . o.
----------------OMMn a.-
nnnnnnnnn,
- - - - - - - N -
o W F
# M 3
* "W ** "*
l H N !. s
. to. 3 a
888
++4 h+ 8$8 *8
+ *k k k k h + } 8++ 8 "h + k h b8888*888864 ++++h++++? - -.1.g I 6
*
- L.ht .w.w.kl.bl.gl.wo,o,o,o,o,o,o,o,o,n,N,n
*** w w w w w w b3 w w w w wq w o,o,o,o,o,o,o,o,o,c.
e h3 w bl ka m bl kl 163 w w .4 e ( w4 oooooooooooooooowone oooooooooo C E
- - - - - - - - - -- b- - - - - g
#7 ? ? ? ? ? ? ? ? ? ? h? ? ? ? ? 2 l* N N N N N N N N N N WW N N N N w a x a e s a s a .s x x ~x s a x x W
- - m m m - - - m -
I
- w. - m n - a" S,8 ? ? ?w ?w w? w?w ?w w? w ? ? of ?w w? w ?w ?
p n-
- n. w w ew es s s [.
n s sn 8 en s ns en sn s ns sn e n e n n n n n e 2 . e
. C .
S - E B i o g i .n. m 2 a t s [=.h, so m h 0a< 0
.hhh,bw m- m w
m . w p 33E3EE@$233 -E3E38 O M m a < 5.. 9- N m . m e ~ o e j. n :': 8 o e u. -. 3.45
It!l?l !li!l,.,!- tr {,' {!>I ,' 3l >i.' . j ,.!;r1; tl' l i> 1, l\lll' ,.i;i t _ 0 0 _ E E _ 0 0 _
- 1 _
1 le - i ' 1 1
.. *~ -
t nh6t 6h _
= ,
=* E E
et n_ ' 0 O Se0 L - trQh e ~ 5 u__
,', ~',=+
- t s
- o F ' ,- '
2 C 2 l
- r r E- o
- E c 0 F O F r t n L n o o f
- 3hc ,.s=*'
3t i c s n E- a E- e O r ,'.<'. - O L F r o L F i _ - s . e
,, ' t 2 e
,- s 4 e s , 4 c cT a
-e e Frts Et C e -
E-D fe L R .
- L R a eF s
- S '
5 a - E E
- e:
._ ~ _~ __
D _.- _ ,: _ _ O L l e1 1 L R 8 O 1 2 8 9 0 1 2 S ._ 7 9 i 1 1 7 1 1 1 r D foF E E- E E- E- E- E- E E- E E- E O O. O 0 O O D D O O O D L 1 t 1 L L L L L L L L s s
$ E I D j eu s_eavo*&
- Coveev 0 Te>ts_o ooeu s_eo.Yt*& eoCoveeo5 r
i e ro , ct na ei D D ut qi E E O 0 en r
, - ~ _ ..._:. _
- , - _H .
L , - - _ ,- : _- _= : - 1 F I l
, + 1 1 ea
'*.i E E
cn n r 0 O u e
*~ 1 5 L d t
,- r 2 i s ,*- .,,*' ~ 2 S e
eI c n r s' - E- o r E- o . xf
- 0. F w DF a . E l
' ,~ - ~ ' 1 n '
- s
- L n u
, o .
io .G 3f i ' 3t
-e E er .
E- oc 1 1 d _ .' r n 0 1 F a-O L F a _ . e - e r s s 3. 3 G 4 c '
- 4 c E- e
_ _ l - E- e e _ 0 e " O le r
, _ _ 1 R *- L R u
' L i g
_ _ 5 _- ~- 5 F
- E-0
- '. - E-O c $ _- _ - . _- _ _ _3 :
1 L 6 0 1 2 8 0 1 2 7 9 .1 1 1 7 9 1 1 1 E E E- E E E E- E- L E E E 0 O O. 0 O O D G O O O O t L 1 1 L L L L L. L L L TEI o _ eu teaYt* e $couoeoD r Coexluoooouiec.Yr*&ooCoveu0 _ t u*po
' ,.ll; : 4 j:i,ai,i{i i14 ...,;i. ;J<- ; ;3*;,.a!< - Ii ,f1,i!JL ; :I'
M OoO. OMOM OMOMOMMo O aa MOMocOM OM e O, O + O. + O. + O O. O + 4O +OtO O. O. O. O. O. O. Ot O. O. O. O + O. O + O. O . O. O. O. O. m
. WwNwWwWwwwWwwWwwwWww w W w w w w w ei w w h, D, sm, 4, M, C, q O, ., C, W, C, n, O, m, G, D, O, G, O, n C, M, O, @, D, M O, N, C, nOoOwO4 O N O M O en O M ~ e c. M M . ONO=0 MOM O 4 O4 04 Od O eO eOM OMMMMOO eOd Od O4 O4 O
. ????????????a??????? ?????????
( v WwwwwwwwwwwwgwWw<wWww
- a. a. 9 o. " a. ". a. a. a. ". a. . a. a. . ". ". n n.
gwwwwwgwgw
. a. a. a. " a. . a. . a.
9OMOM C@ OeOWOM CM MMMNN OOeONO4 OMO W DMOnO4 04 OW O 4 O e e to @ OO wOeOeO 4 OW O OOO OOOOOO OO O O OOOOOMN OOOOo
$ 6 + 1 + e + 1 + 1 + 4 + 4 + t e 1 6 0 t i +4 +TO + O. +OOOi +
0 . O. . . . . . O. . d. . . . . . . . . O. . . . . . . . nOeOmOMOMOuceoekhhNN M O M O *< O M C M O. MOeO
- Od OM OM O M O M M M en m m *OMOd OM OM M 4 n ??????????????*?????
W W w bl W W W w W W W M W W kl bl W W w M
*????????E NWwwWMMwWW
- f. A O.N. O. M. O.W D. M.O.e. O. O eO es.e. M. M.40. M N. O. M. O. W. G. d. O. M. O.
v e e
** MCW CMOOONOMONOMN*
- MM OONO* ONOMO e3 U
W M Od O4 O M O M O M O M C M N en en @ @ MCM O 4 OMCMO h he h OOOOOOO. a + i + 4 + OOOOOOOOOOOOO 4 s 4 4 + e + 4 e a 6 6 4 O O n 4 43 OOOeO s + T
- t + 4 + T +
n W W h4 W bl W W k W w N kl M W W W w w N w WWWWWWWwww p 4. O. M. O. ep C.M.C. M. O. M. O. N. O.W.G,G. 9.W. W.
. M. O. M. O. N. O. W. O. O. O. l 8
6 mO9OMOwONOMOMCMMeeMM eONC#ONOMO j.
** I e M *MeMcNeN*McNenNeeee nan <MeNwM* >
" . ????**??*??????????? ?????*????
- l W N.W. N. N. N. N. O.N. W. W. U. N. W.N. N. N. N.W.
. M. S.W. N. W. W. N. N. W. U. N.
#hNhM e M W H M 9 e M M e ete e N N MWHMNMMMMM NMMM M e N M N
- N*e N M N N M I
msg e NMN*M *N4 mW N n W W hl W W W W W W N M W M w W M W W w w MWWWklWwNNW y e4 4 p 40. e. N. M. e. n. H. w. w. w. O. 4
- M. A. en. N. N. O. O.
. 4. w. D. h. n. s2 O. N.
y et M e H
- NNNNONet N M M N M M m en NNNmb so M e a e e"4 N O gggnnggggggnngggnnee ggggnugunn N ??????*?????????*??? ??????????
I t.. c U. M. N. W. U. N. W. N. N. N. N. W. N. N. U. N.W W.
. N. W.N.N. N. N.
W. N. W. W. W. N. gN N N M h en M N M N N N M M M N e M M e e NMN*hM eMNO bd N f,) M OM O et O M OM O ce O et O ce M ee et MM M O M O ce C e4 OMO N M ????????*??????????? ?????????? T o. @ 2 0 0. W. E. N. W.M.O. N. U.W. W. W. W. N. N. N. N.U.N.U. N.N.O. N. E. N. W. W. W. O. . N OOOWoh DhonOh0h NeW NN e- g NCho#Ch 0hc M M M # DW 9 b NWN E NW NW N4 NeNeNoM eme "Nd NeNW N4 No t) M Q MN nM n 9 8I OffODOfC000OOOOO OGfC OfOOOfyCOO 8 e, H w = M. u, b .W. .W W. Lal. ..k.l w i.st.. b . b. h.3 k.1M.... b W b. k.l b b...N. k.a tal N.
. . . N.
N. b.i
. . .N. N.
- b. k.l gy Wo MMMMMMMMMMMMMMMMhNhN DM MMMMM MM MM
}$
Q .e4 6 ( N U e 0eeeew**w4
- . t. .
4 e**************ee4 e e 'l
@~
U$
????????????????????
wwwwwwwwwwwwWwwwww. w 7?????????
.wwawWWWwww r
5 M s2 w N en 4 e. 4 4 44.*4 4 4 4 44 444 4 N. @. H. %. e44 w 4 44 4 4
- 4. e. 4 ,s M M O en e M en en n en M in m en en n n h h h h
%M e Iin en en en an en en. m in en OOOOOOOOOOOOOOOOmmen ACOOgCOOOOO W ^
+ w M e.
as ha nn e. s. t. e. t. t. w. u. t. t. e. s. s. t. a. t. w. N. w. N. OOO O O O O O O O O O O O O O N e4 N N s t. a. e. s. s. a. w. ge. s. ?t. OOOOOOOOOO e
.m e .
.E4 M M M M M M M M M M b M M M M M h 43
- O O O O O O O O O O ho O O O O e.
e E + + + + + + + + + + et + + + + + A I N. N. N. N. N. ~ N. N. N. N. N. M - M M M M M M M M M M M we N. M N. M N. M N. N. N O M M e
., d . . . . . * * . . m, , e . , u E Q O O O O O O O O O O O
** @ ^ - ; O. O. O. + + + + + 4+ + + + +
L' 6 m W M W W W lol W lsJ w W Gw W W M W km w ab M. M. M. M. M. M. M. M. M. M. A g M. M. M. M. n. g 3 4 W W 4 W W 4 4
- 4 4 W W W e e4 e a e G S
** U
.C'4 53 w m3
. 0 f M 3 E.
.c. M
.g3 C
h U 6-4 ae tW 6Q W - A A HJ % g El SE4 m E1 o 13393o o3s 4: o o oB3939 3 3 e
~o o . . o o o o 9 o e M e* a a
- a e e a
- u. O . M . O .c T f. M N M W n tD N W @ H > d O M M N h a h b h e 3.47
_. _ _ . ._._...m. _ m..m.m..m .._m_... ...___-_..m._.___ _ _ _ . . _ _ _ _ _ . _ . . _ . _ . .I. t 4 I J 1 1 i O o o s I n , , , s , ,. r , ,
.s . .- - -
hskl,r cl, '.- l 8 s I . b o s
- - e R. Ths; ,
y, N U
,- N i e i L i L WO -
w0 l o6 oG sc sc 4 M i !' -
,,, 9 ..
,,, e a ,
l' . . L$ o c 8.a*8 -e ,-
,.- , a e m
o ( ,' 6 p> ,f' : *8 y N' - 6 o*3' e
+
{ oT a D:
/ sx u3o N
- l l 7
,e i
*m l l o / o l / 1 l 1 5
_ e < l. i . e. m 3 ,A < . u , m - , s ,$ ;;
. . e a u 0 m . e o 0 5" w i i i i u wi e w wi w w i
t wi w u u u i 6 s o
.J o
s o s o
.J o
s d W o s o s o
.J o
.8 o
.J 32
(;oex-Jolone) Jed)'beJJ eouopeeox3 (JD84-Jolooe; Jed) be;j oouopesox3 00o "o u Ce 8 o 8 o Et v i c , ,,, ,, y, , ,,i, e. , j s , , r , , , r..,, , em ,,,y . , , s *sc k* -
., - T 7
,- w w *se o c
,- o o
- et s Ck
',ni 01
. ,l '
n .- .,.. - n, u k, *r*>c
,/ . . ' u
.' 3uod o' e l ec ,, '
, , s c, o
6y
. . . . -.mi .;:.o ; ,,, D
.o 0
l . l u , i l l l WS es
, .l I
~
88 '
, %, *Q o
,/ m
- m. ug
. . gi .
- a -
4 p 4 6 n l - i e < i e
- l', .
$I aw , ; ,
oO aa bs ;
; l ~
o l .i o- fo '
,' - w i
wi w
, e
, , w ,
-i_.> l t. e ir .' - tm .
i t1: 1 L !t i 1.0, t i i e m 9 = W D e a R = U 0 6o 6 o 6 o 6 o 6. o 1o 1 o 6 o 1 o 4 o 1 o 6o s s s s s a s .J s .J s s , (JoeA-Jolooe; Jed) be;j eauopeeox3 (JoeA-;oloos; Jed) be;j ocuopeoox3 1 , f L 4
? E i I l l 11 8
- 8 o
o
'r s r p- ti - p;g.,t....F n gr I F ' FF 'i 5 P i
-{ l
..- n= ,
c.c c.c -
,' J n
$@!$b lu ,
e' :
- " ,i
( i fe su .- f3 su
}
e ',
,- e .
e o
. : c
.9 :
*m m l
! I T)
I l:
- E.e a c s
- ,i e.
aL e ie wL .o ,y O 3' 0 ,4
- l v s8o -
l , a8 ud e
/
. o. - 8
< i l- i k w. u
,. ; , e , , , e s
- al e
r
- ws e e a ss
. . e e
L. L .iii e L>,i a L. , e r w L, i n
.$ o t
i , / L1 e i i s L. , e L1 iri m Lo i W L. s i
=
3 [e ]u n s "- 1 6 6 6 4 1 1 6 6 6 6 6 1 1 , o o o
,a o o o
o J o o o J o o J o o
& ui o
u{'g e 4 ,J ed (JooA-Jogooe) Jed) beJJ eouoposex3 (JoeA-Jojooe; Jo[) be;j ecuoposex3 EA o
- r.
E*m 8 o 8 o 8g5 Vp a - , , r-,, r, s e g-
- r , ,. ,p r . r- - , n, m , . ,m , , ,
w a,, w e i
~
e aR in ,-
- n. o%m c
, : wi- ,<
- : i E o
su - l 3 w05 o au eb o
; ; 3 2
' & $e- l
; i ;
c-i c . l 0 o M
. 1 = -
l 3 = .
- 1f o
.',' 5
- 1f a
se O. d
. m ,_,
t , av e- r l e -: M *e o i : M e p : O j *' Se- e x c ; e e ee u s kuo
- o i m
- l
. .- : i :
3 1 ! ott L: 1 ,L . i, te v i L. i Lo ir4
~3 . km , L re L, t i r Em , , to ,,']
in e s e a w :: n e s e m W :: 1 6 6 1 1 1 1 6 1 o 4-o 1 o 1 o -o 1 1 o o o o o o o o C a a a ' a a a- a ,a a ~ a a a , i (JoeA-Jogoce; Jef)'beJJ Bouop000X3 (Jo8A-J0;oo8J Jo[) beJJ eouop0DDX3 3.49
,_ -~ - . _ . . - . . _ . . _ _ , _._. . ._ .___ ,,
1 , !. ! i , ' .i !a iT t .vg= t !p rL it! -! ;i j .' ,*--i;y. :[;!,,tLi > : i D E 0 . E 0
- O L ;- : , ,e , 1
- i _ - 1 ~
le'- s~ i t ~n h
,*,'*,:..e nh,f t et
, s
a-Z c5#0 rof 5 a. 2 e - -
- Eo Es P -
,N OC Ol Lr a Lr
,' ~ o Fo
....*;,.1, f <r
~
n , n oP
- f
.o .,' i o S a
" f i
4e f sy
- _ ~ 4eo - -
-o nl o ar EF r ,- '
Er i O - O. eF t Le s -
- a 1 s c E
~
c -
~
e a . e , e rF f e -
!e R
FC R - a ey 6
- - 6' sl
- ~ a r
~ E-E- - ,
0 ea
- - r _
,:_ : - y- - 7_-
0 t _ ri r= - ~ : r: _ c
,r ~
1 t le E . 5 6 8 9 0 1 1 1 5 6 8 9 s 1 R ,e
- 7- t-E- E-7-
E E- E- E- E- E- E E E- E- E-o rhl cb O L O L 0 1 0 1-0 1 0 1 0 t 0 1 0 t 0 t 0 1 r O L O L t fo el aa
.h , >f'o~oOE ue"v oec ,oC S eeoyu n'oeAb o j E uO"vtee eocove.os ri
- sB e a i v lA cn esS 0 0 e u esC E E qVd 0 e O. 1 r _n
'1 _ , ~ - ;:
,..','*,( .*Lw*
=
1 i~7 _ ,E- du - 1: _ - ,a :- a F:2 a
,g' ,
eBs cP s
, .',=,- *
',....=;..,3 2 nAa a p
- 2
~ , ~ -
Er dyy E- e OS e rB
,' - i * ,
- Oi
- Lr ,
r,. Lr c Fo ,', s F o xm n , n
- i' Em o .S o -
- a i .
" 4c i t
- ~ 4c t
4
- ar
- or m N ~ 1- f
. EF
- EF ,
O
- 3. ul
- O -
Le
- a.
Les . a, s. 3 G
- " a
- c.
e e dna
- ~
- e -
. le - ,
te r ur R - s R
~
- g
- 6 - -6 i G
- E E-O F'
O. . ~ rE . :- - -
- r: - - , _
- r: ,5 .~ __ 7= . , .- , -
1 0 hL t C 1 6 8 9 s 5 6 8 9 1 1 5 7 1 7- - - E- E- E E E- E- . E- E- E- E-E- E- E E o 0 1 0 1 0 t O t 0 t 0 1 D L O L - 0 1 0. t 0. 1 r
,- t 0 0 1 t mb,>Iuo ooE ug"v eeg ,oCSeenyu ,'oe>[ojE ue"v eee eocoveeoy N
u vo
,!-il4,i; ,.j l2 ;l ij ,: ' - ,, I
- ,icl ,1la;:;; .
. i ii j' ; ,f '
. - - . . - - . . - - - . ~ . . . - . - _ - . - . . -- . . - -- . - - - - . . - . - - - - ..- . _ -
l l j ? l , i ! f I 1 8 8O ! l O
- ,,.s
.i
< , ,- ,. ,. .J ,,
rg ; , i
!:s$
i
.C ,'
) .- f
- 1 5 N fff e , E N
j hM CL 6 ; b .D; o u o. d r & L .
- l. y a
.q ,- -
e Wm j 1 i C l' C 3 ,
.... * ...: , .9 *o N mm Q
1 5 / l i 9 iIS # *. ,- : s. *k . i c: m
.o,e m-
- O ,
1
- sbs u-aeo , .
, ae a u ,' : 1 .
o oe
., u ,/
y au y - u
- e o e
e w3 l 3 m / , e
- s t e 8 i e
!1 I
. id ED
! , a1 m ,3 s 1 o o m ,-3 .. , m , <, ,
3 es +
=
e W := e s m m W e s m in in h 6 4 4 4 6 6 6 6 6 4O 6 6 O "uo d" 8' b O O O b O O a O a O a O a O a ,a a a a a a a a
. Wj
(;ooA-;o;ooe; and) boJJ ecuopcoox3 (Jo*A- J0Pood J6[) bo;j ocuopcoox3 mo ee see . W 8m 8 8 E-ve 6 s ,, re ,,;s O
-g ;
-- ,m , r . r2.my , - ,, ,.m ,, ,m ,, , . , , r-
- u. . 1
.. .... p.
, .. ... - g .
u.. N DO N i
,'l
- ' ; 4m
,}}