ML20203E604
| ML20203E604 | |
| Person / Time | |
|---|---|
| Site: | Seabrook  |
| Issue date: | 12/31/1985 |
| From: | Kreslyon Fleming, Torri A, Woodard K PLG, INC. (FORMERLY PICKARD, LOWE & GARRICK, INC.) |
| To: | |
| Shared Package | |
| ML20203E591 | List: |
| References | |
| PLG-0432, PLG-432, NUDOCS 8607240186 | |
| Download: ML20203E604 (443) | |
Text
{{#Wiki_filter:- - - - - - - 1 O L PLG-0432 [ [ [ SEABROOK STATION , RISK MANAGEMENT AND L EMERGENCY PLANNING STUDY [ [ [ [ [ { I [ Prepared for i NEW HAMPSHIRE YANKEE DIVISION PUBLIC SERVICE COMPANY OF NEW HAMPSHIRE Seabrook, New Hampshire ( December 1985 1 [ [ ' ng72amoas'!<s<$4a a ( , Pic:< arc.,Lowe anc.Garric:<,Inc. ( Engineers e ilpplied Scientists e Afanagement Consultants Newport Beach. CA Washington, DC f --
}
I PLG-0432 I I SEABROOK STATION RISK MANAGEMENT AND I EMERGENCY PLANNING STUDY I I Project Manager Karl N. Fleming I Principal Investigators Alfred Torri Keith Woodard I R. Kenneth Deremer Robert J. Lutz (W) Robert E. Henry (FAI) Key Contributors J. H. Scobel (LV) Jackie Lewis I T. Edward Fenstermacher John G. Stampelos Ali Mosleh , g Daniel W. Stillwell E Kathleen C. Ramp James C. Lin i David J. Richard Subcontractors Westinghouse Electric Corporation @ Fauske and Associates,Inc. (FAI) l l Prepared for l NEW HAMPSHIRE YANKEE DIVISION I PUBLIC SERVICE COMPANY OF NEW HAMPSHIRE Seabrook, New Hampshire December 1985 I ' I Pickard,Lowe andGarrick,Inc. lingineers e Applied Scientists Management Consultants I Newport Beach, CA Washington, DC
CONTENTS Section Page LIST OF TABLES vi LIST OF FIGURES ix ACRONYMS xii EXECUTIVE
SUMMARY
S-1 1 INTRODUCTION 1-1 1.1 Objectives 1-1 1.2 Background 1-2 1.2.1 Overview of SSPSA Results 1-3 1.2.2 Emergency Planning 1-4 l 1.3 Project Overview 1.3.1 Technical Approach 1.3.1.1 Plant Model 1-7 1-8 1-8 1.3.1.2 Containment Model 1-11 1.3.1.3 Site Model 1-13 1.3.2 Project Team 1-15 1.3.3 Technical Review and QA 1-15 I 1.4 Report Guide 1.5 References 1-16 1-16 2 RESULTS AND CONCLUSIONS 2-1 2.1 Rebaselining SSPSA Results for No Immediate Protective Actions 2-2 2.2 Evaluation of Emergency Planning Options 2-6 2.3 Sensitivity Analyses of Key Assumptions 2-9 2.4 Evaluation of Other Risk Management Actions 2-11 2.5 Conclusions 2-13 2.6 References 2-14 3 SSPSA PLANT MODEL UPDATE 3-1 3.1 Interfacing System LOCA Sequences 3-3 3.1.1 Seabrook Configuration 3-5 3.1.2 Initiating Event Analysis 3-10 3.1.2.1 Plant Response 3-10 3.1.2.2 Initiating Event Analysis 3-12 3.1.3 Event Tree Model 3-22 3.1.3.1 Top Event Descriptions 3-24 3.1.3.2 Event Tree Structure and End States 3-26 3.1.4 Event Tree Quantification 3-28 3.1.4.1 RHR Piping and Heat Exchanger Strength 3-28 3.1.4.2 RHR Pump Seal Failure Area 3-30 3.1.4.3 Operator Actions and Emergency Procedures 3-32 I iii 1339P121885
I CONTENTS (continued) Section Page 3.1.4.4 Pump Operation in Adverse Environments 3-36 3.1.5 SSPSA Plant Model Integration 3-37 3.2 Containment Recovery Analysis Following an Extended Loss of All AC Power 3-38 3.2.1 Recovery Model 3-39 3.2.2 345-kV Offsite Power Recovery 3-43 3.2.3 34.5-kV Of fsite Power Recovery 3-44 3.2.4 Recovery of Power from Other Transportable Emergency Power Sources 3-45 3.2.5 Results 3-47 3.3 References 3-49 4 SOURCE TERMS AND CONTAINMENT ANAi.YSIS 4-1 ) 4.1 Source Term State-of-the-Art Assessment 4-1 4.1.1 Introduction 4-1 4.1.2 SSPSA Source Terms 4-1 4.1.3 The Industry Degraded Core Program 4-3 4.1.4 The NRC Source Term Program 4-6 4.1.5 Other Source Term Research Programs 4-7 4.1.6 Evidence and Conclusions 4-8 4.2 Accident Phenomena and Source Term Considerations 4-8 4.2.1 Modelf ng of Accident Phencmena in the SSPSA 4-8 4.2.2 Advances in Modeling Accident Phenomena and Source Terms 4-9 4.2.3 Severe Accident Technical Issues 4-9 4.2.4 Issue 8 - Direct Heating of the Containment E Atmosphere by Debris 4-13 5 4.2.4.1 Experiments Conducted to Date 4-14 4.2.4.2 Debris Dispersal Characteristics for the Seabrook Configuration 4-15 4.2.4.3 Material Available for Direct Containment Atmosphere Heating 4-15 4.2.5 Issue 15 - Containment Performance 4-16 4.2.5.1 Pressure Capacity, Leak Area, and Uncertainties for the Failure Pressure of Individual Failure Modes 4-16 4.2.5.2 Containment Failure Categories 4-20 4.2.5.3 Composite Probability Distribution for the Containment Failure Pressure 4-20 4.2.5.4 Conclusions 4-21 4.2.6 Issue 16 - Secondary Containment Performance 4-22 4.2.7 Issue 18 Essential Equipment Performance 4-22 4.2.8 Release Categories 4-23 4.3 Zion /Seabrook Design Comparison 4-25 4.3.1 Purpose 4-25 3 4.3.2 Design Comparison Tables 4-25 E 4.3.3 Conclusions 4-28 1339P121885
I I CONTENTS (continued) Section Page 4.4 V-Sequence Analysis 4-29 4.4.1 Description of Physical Plant and Systems 4-30 4.4.2 Analysis of the Overpressure Event 4-31 I i 4.4.3 Description of Event Analysis and Liodels 4-32 ) 4.4.3.1 Core and Containment Behavior 4-33 4.4.3.2 Fission-Product Behavior 4-35 I 4.4.4 Fission-Product Release 4.4.5 Consideration for Emergency Operating Procedures 4-35 4-36 4.4.6 Summary 4-38 I 4.5 Release Categories 4.5.1 Introduction and Overview 4-39 4-39 4.5.2 Best Estimate Release Categories 4-41 I 4.5.3 Conservative Estimate Release Categories 4.5.4 Release Category Uncertainties and 4-43 4-45 Comparison of Release Fractions I 4.5.5 Enveloping Source Terms 4.6 Accident Sequence Mapping into Release Categories 4.6.1 Guidelines for Accident Sequence Mapping 4-46 4-47 4-47 4.6.2 C-Matrix for Current Source Term Categories 4-47 I 4.7 Treatment of Source Term and Site Model Uncertainties 4-48 4.8 References 4-48 I 5 SITE ANALYSIS 5-1 5.1 Review of SSPSA Site Model and Modifications for I This Study 5.2 CRACIT Postprocessor Function 5.2.1 Assessment of Dose as a Function of Distance 5-1 5-4 5-4 5.2.2 Evaluation of Risk as a Function of Distance 5-5 I 5.2.3 Conditional Cumulative Distribution Functions 5-7 5.3 Sensitivity Analyses and Observations 5-7 5.4 References 5-8 APPENDIX A: DOSE VERSUS DISTANCE CURVES A-1 I APPENDIX B: APPENDIX C: APPENDIX D: HEALTH RISK VERSUS DISTANCE CURVES CONDITIONAL RISK CURVES EXPECTED RISK
SUMMARY
TABLES B-1 C-1 D-1 APPENDIX E: PEER REVIEW GROUP COMMENTS AND CONCLUSIONS E-1 I I I v 1339P121885
LIST OF TABLES Table Page 1-1 Summary of Principal Contributors to Risk in Terms of Accident Sequence Groups and Initiating Events g from the SSPSA 1-19 E l 1-2 Definition of 39 Plant Damage States Used in SSPSA Risk Model 1-20 1-3 Identification of Important Plant Damage States in l the SSPSA Risk Model 1-21 l 1-4 Summary of Accident Sequences with Significant Risk ! and Core Melt Frequency Contributions from SSPSA 1-22 3 1-5 Initiating Events Selected for Quantification of the E Seabrook Station Risk Model 1-24 1-6 Summary of SSPSA Plant Event Tree Modules 1-26 E 1-7 Summary of Technical Review Responsibilities 1-27 E I 2-1 Update of Interfacing Systems LOCA Key Results 2-16 l 2-2 Comparison of Release Categories 2-17 2-3 Evaluation of Emergency Planning Options against NRC Safety Goals 2-18 2-4 Comparison of Core Melt Frequencies and Distributions j of Release Types 2-19 l 2-5 Comparison of Acute Fatality Risks from Different Sources 2-20 l 2-6 Sensitivity Analysis of Early Fatality Risk / Safety Goal Ratio for No Immediate Protective Actions 2-21 l 2-7 Ratio of Mean Health Effects Risk with and without l A V-Sequence Emergency Procedure 2-22 3-1 Update of SSPSA Plant Model Results for Core Melt and Plant Damage State Frequencies 3-51 l 3-2 Update of SSPSA Plant Model Accident Sequences l Ranked by Core Melt Frequency Contribution 3-52 g l 3-3a Update of SSPSA Plant Model - Plant Damage State IF 3 Sequences 3-53 3-3b Update of SSPSA Plant Model - Plant Damage State 1FP
! Sequences 3-54 3-3c Update of SSPSA Plant Model - Plant Damage State 1FV Sequences 3-55 3-3d Update of SSPSA Plant Model - Plant Damage State 1FPV Sequences 3-56 3-3e Update of SSPSA Plant Model - Plant Damage State 2A Sequences 3-57 g 3-3f Update of SSPSA Plant Model - Plant Damage State 3D g Sequences 3-58 3-3g Update of SSPSA Plant Model - Plant Damage State 3F Sequences 3-59 3-3h Update of SSPSA Plant Model - Plant Damage State 3FP Sequences 3-60 3-3i Update of SSPSA Plant Model - Plant Damage State 4A E Sequences 3-61 5 3-3j Update of SSPSA Plant Model - Plant Damage State 7D Sequences 3-62 vi 1339P120685
I LIST OF TABLES (continued) Table Page 3-3k Update of SSPSA Plant Model - Plant Damage State 7F Sequences 3-63 I 3-31 3-3m Update of SSPSA Plant Model - Plant Damage State 7FP Sequences Update of SSPSA Plant Model - Plant Damage State 7FPV 3-64 3-65 I 3-3n Sequences Update of SSPSA Plant Model - Plant Damage State 8A Sequences 3-66 3-30 Update of SSPSA Plant Model - Plant Damage State 8D Sequences 3-67 3-4 Definition of Initiating Events, Top Events, and Boundary Conditions Defined in Technical Specification I 3-5 Update
- and the SSPSA Definition of Initiating Events, Top Events, and Boundary Conditions Defined in This Study 3-68 3-89 3-6 Pump Alignment 3-92 I 3-7 3-8 Summary of V-Sequence Analyzed with MAAP Check Valve Leakage Event Data Base 3-93 3-94 3-9 Statistical Data on Check Valve Leakage Events in I 3-10 PWR, ECCS, and RCS Systems Operator Action Sequences Used in the RHR or V-Sequence LOCA Analysis 3-96 3-97 I 3-11 3-12 Point Estimates for Environmental Failures of the RHR Pumps Point Estimates for Environmental Failures of the 3-98 CBS Pumps 3-99 I 3-13 Point Estimates for Environmental Failures of the Safety Injection Pumps 3-100 3-14 V-Sequence Results - Initially Assigned Plant Damage
- States , 3-101 l 3-15 Cumulative Probability of Containment Failure within l t Hours after a Loss of All AC Power (No Containment l Spray and Recirculation) [cc(t)] 3-102 l 3-16 Cumulative Power Recovery from Offsite Power for l the 345-kV Source within t Hours After the Loss of g Al1 AC Power [&345(t)] 3-103
.g 3-17 3-18 Cumulative Recovery Frequency for the 34.5 kV Source Cumulative Probability of Recovery of Containment Spray 3-104 and Recirculation of Additional Independent Sources within t Hours after a Loss of All AC Power [$0ther(t)] 3-105 l 3-19 Containment Recovery Analysis Results 3-106 I 4-1 4-2 Accident Source Terms and Consequences Calculated by the IDCOR Program for the Zion Station NRC/IDCOR Technical Issues for Severe Accidents 4-51 4-52 4-2a Summary--Technical Support for Issue Resolution 4-53 4-3 Containment Failure Modes and Type 4-58 4-4 Source Term Categories 4-59 I 1339P121685 vii
I LIST OF TABLES (continued) l Table Page 4-5 Containment Design Comparison Table for Seabrook Station and the Zion Station 4-60 g 4-6 Description of Seabrook RHR System 4-74 E 4-7 V-Sequence Chronology 4-77 4-8 Definition of Fission Product Groups 4-78 4-9 Releases to Equipment Vault 4-79 4-10 Releases to Environment - Suppression Pool Scrubbing 4-80 4-11 Releases to Environment - No Suppression Pool; Ventilation 4-81 3 4-12 Releases to Environment - No Suppression Pool; 3 No Ventilation 4-82 4-13 Best Estimate Release Categories 4-83 E 4-14 Conservative Estimate Release Categories 4-84 5 4-15 Comparison of Release Categories 4-85 4-16 Revised C-Matrix for New Source Term Categories 4-86 4-17 Revised C-Matrix for New Source Term Categories 4-87 5-1 Variation of Parameters in Consequence Uncertainty Estimates 5-9 5-2 Consequence Assessment Discrete Probability E Distributions 5-10 3 5-3 Additional Parameters Varied During CRACIT Runs 5-11 I I I Il l l viii 1339P120685 !
LIST OF FIGURES Figure Page S-1 Comparison of Seabrook Station Risk (With No Immediate Protective Action) With Background and Safety Goal I S-2 Individual Risk Levels Acute Fatality Risk as a Function of Protective Action S-3 S-4 S-3 Comparison of Updated Seabrook Station Results With I S-4 NUREG-0396 - 200-REM and 50-REM Whole Body Dose Plots for No Immediate Protective Action Comparison of Updated Seabrook Station Results with S-5 NUREG-0396 REM and 1-REM Whole Body Dose Plots for I Releases Within 24 Hours of Warning and No Immediate Protective Action S-6 1-1 Contents of SSPSA Report Volumes 1-28 I 1-2 1-3 Block Diagram Structure of Seabrook Risk Model Standard Form of Accident Sequences in SSPSA Risk 1-29 Model 1-30 I 1-4 1-5 1-6 Overview of SSPSA Event Sequence Model Structure Generalized Transient Early Response Event Tree (TRAN) Block Diagram Showing Support Systems for Emergency 1-31 1-32 Feedwater System 1-33 1-7 Illustration of Release Category Definition 1-34 2-1 Comparison of Updated Early Fatality Risk Curves for Seabrook Station (No Immediate Protective Action) with I 2-2 SSPSA and WASH-1400 (PWR)--Mean Values Comparison of Updated Early Fatality Risk Curves for 2-23 Seabrook Station (No Immediate Protective Action) with I 2-3 SSPSA and WASH-1400 (PWR)--Median Values Comparison of Updated Latent Cancer Fatality Risk Curves for Seabrook Station (No Immediate Protective 2-24 I 2-4 Action) with SSPSA and WASH-1400 (PWR) Comparison of Seabrook Station Risk (with No Immediate Protective Action) with Background and Safety Goal 2-25 Individual Risk Levels 2-26 2-5 Spatial Distribution of the Expected Frequency of Acute Fatalities for Seabrook Station Based on Updated Results for No Immediate Protective Action 2-27 I 2-6 Impact of Different Energency Planning Options on Risk of Early Fatalities (Results of This Study Compared Against WASH-1400) 2-28 2-7 2-29 I 2-8 Acute Fatality Risk as a Function of Protective Action Early Fatality Risk Reduction for Different Protective Action Strategies 2-30 2-9 Comparison of Updated Seabrook Station Results with I 2-10 NUREG-0396 - 200-REM and 50-REM Whole Body Dose Plots for No Immediate Protective Action Comparison of Updated Seabrook Station Results with 2-31 NUREG-0396 REM and 1-REM Whole Body Dose Plots for No Immediate Protective Action 2-32 I I 1339P120685 iX
I LIST OF FIGURES (continued) l Figure Page 2-11 Comparison of Updated Seabrook Station Results with NUREG-0396 REM and 1-REM Whole Body Dose Plots for g Releases within 24 Hours of Warning and No Immediate g Protective Action 2-33 3-1 Cold Leg Injection Path Arrangement 3-107 3-2 RHR Suction Path Arrangement 3-108 3-3 Frequency of Check Valve Leakage Events 3-109 3-4 Seabrook Emergency Plan Optimization - VI Tree 3-110 3-5 Seabrook Emergency Plan Optimization - VS Tree 3-112 g 3-6 Probability of Pipe Failure 3-113 3 3-7 Leak or Relief Valve Flow Rate Versus Pressure 3-114 3-8 Makeup Paths to the RWST, BAT, and VCT 3-115 3-9 Fault Tree for Environmental Failure of RHR, CBS, or SI Pumps 3-116 3-10 Exceedance Probability for Time to Containment Failure for a Station Blackout Accident at Seabrook Station 3-117 3-11 Portsmouth Area Electrical Transmission System One-Line Diagram 3-118 3-12 Exeter and Hampton Electric Company Transmission System One-Line Diagram 3-119 3-13 Seabrook Station Temporary Power and Light General 3-14 Arrangement Example from a Quick Review of Emergency Power 3-120 l Suppliers 3-121 4-1 Type B (Seabrook) Lower Reactor Cavity Configuration 4-88 l 4-2 Pressure - Hoop Strain Relation for Containment E Cylindrical Wall 4-89 4-3 Conditional Cumulative Probability Distribution for g Fuel Transfer Bellows Failure 4-90 5 4-4 Conditional Cumulative Probability Distributions for Feedwater Penetration Failure (Fluehead or Pipe Crushing) before Hoop Failure as a Function of Failure Pressure 4-91 4-5 Discrete Probability Distribution for All Other Containment Failure Modes Combined 4-92 4-6 Combined Discrete Probability Distributions for All Benign Containment Failure Modes 4-93 4-7 Composite Containment Failure Probability Distributions a for Type B (Leak) Failure, Type C (Gross) Failure, and g Total Failure 4-94 4-8 Engineered Safety Feature Flow Diagram 4-95 4-9 Plan Views of RHR Equipment Vault at Three Elevations 4-96 4-10 Elevation View of RHR Equipment Vault 4-97 4-11 Primary System Pressure PSIA 4-98 4-12 Core Water Temperature F 4-99 E 4-13 Upper Compartment Pressure PSIA 4-100 3 4-14 Upper Compartment Gas Temperature F 4-101 I x 1339P120685 m
I LIST OF FIGURES (continued) Figure Page 4-15 Vessel Water Level (Bottom Core = 7.94) Feet 4-102 4-16 Greatest Temperature in a Core Node F 4-103 I 4-17 4-18 Cavity Melt Mass LB Cavity Axial Concrete Penetration Feet 4-104 4-105 4-19 Seal LOCA Flow Rate into Vault LBM/ Hour 4-106 I 4-20 4-21 4-22 Vault Water Level - Feet Fraction of Noble Gas Release to Auxiliary Building Fraction of CSI Release to RHR Equipment Valut 4-107 4-108 4-109 I 4-23 4-24 4-25 Fraction of TE Release to RHR Equipment Vault Fraction of SR Release to RHR Equipment Vault Fraction of RU Release to RHR Equipment Vault 4-110 4-111 4-112 4-26 Fraction of CS0H Release to RHR Equipment Valut 4-113 4-27 Discrete Characterization of Source Term and Site Model Uncertainties 4-114 5-1 Dose Versus Distance Curve for Release Category S6B-H I 5-2 for No Immediate Protective Action for 24 Hours Risk versus Distance Plot for Release Category S6B-H for No Immediate Protective Action 5-12 5-13 5-3 Health Effects CCDF for Release Category S6B-H, No Evacuation 5-14 5-3d Health Effects CCDF for Release Category S7C-H with 10-Mile Evacuation 5-17 5-3e Health Effects CCDF for Release Category S7C-H with 2-Mile Evacuation and Sheltering to 10 Miles '5-18 I I I I I I I xi 1339P120685
I ACRONYMS Acronym Definition BAT boric acid tank g BIT boron injection tank 3 BWR boiling water reactor CBS containment building spray CCDF conditional cumulative distribution function CVCS chemical and volume control system DMW demineralized water ECCS emergency core cooling system EFW emergency feedwater EPZ emergency planning zone ESFAS engineered safety features actuation system FAI Fauske and Associates, Inc. E FPS fire protection system FSAR final safety analysis report IDCOR Industry Degraded Core Rulemaking LER Licensee Event Report LOCA loss of coolant accident LPI low pressure injection LPR low pressure recirculation LWR light water reactor MCB main control board MOV motor-operated valve i l NERC National Electric Reliability Council NPCC Northeast Power Coordinating Council I NPE Nuclear Power Experience NRC Nuclear Regulatory Commission PAG Protective Action Guideline PCC primary component cooling water system PCS pressurizer control system 3 PDT primary drain tank E PLG Pickard, Lowe and Garrick, Inc. - PORV power-operated relief valve PRA probabilistic risk assessment PRT pressurizer relief tank PWR pressurized water reactor RCS reactor coolant system RDMS radiation detection monitoring systen l RHR residual heat removal 3 ! RSS Reactor Safety Study E RWST refueling water storage tank - 1250P100385
ACRONYMS (continued) Definition I Acronym ! SGTR steam generator tube rupture SMA NTS/ Structural Mechanics Associates, Inc. SSPS solid state protection system SSPSA Seabrook Station Probabilistic Safety Assessment TMI Three Mile Island visual alarm system I VAS VCT volume control tank WEC Westinghouse Electric Corporation YAEC Yankee Atomic Electric Company I I I I I I I I 1 I I I 1250P100385 Xiii
I EXECUTIVE
SUMMARY
The purpose of this report is to present the results of a technical evaluation of emergency planning options and other risk management actions under consideration for Seabrook Station. These results include an update of the Seabrook Station Probabilistic Safety Assessment (SSPSA) (Reference S-1) to account for new insights regarding radioactive release I source terms and the progression of sequences involving loss of coolant events that bypass the containment. I The principal focus of this study was the evaluation of the impact of various protective actions such as evacuation and sheltering to various radial distances from the plant site. A variety of risk measures were I used to evaluate emergency planning options from different perspectives. These include traditional PRA-type frequency of exceedance of consequence curves, spatial distribution of risk, risk as a function of evacuation distance, and the risk factors that were employed as the basis for setting the current EPZ distance at 10 miles (Reference S-2). A key result of this study was the quantification of acute fatality risk, I defined as the expectation of the annual frequency of health effects. The result for the assumption of no immediate protective actions (e.g., evacuation), is presented in Figure S-1. This figure also I includes the corresponding NRC safety goal for average individual risk within 1 mile of the site boundary together with the " background" risk from nonnuclear accidental causes from which the NRC safety goal was derived. These results and others presented in this report show that I even under the assumption of no immediate protective actions, the risks of early and latent health effects are very low in relation to the NRC safety goals and in reference to any implicit or explicit standards of I acceptable levels of risk (e.g., WASH-1400 and NUREG-0396). Therefore, the requirement for any protective actions such as evacuation cannot be established on the basis of achieving an acceptable level of risk. Of the very low levels of risk reduction that can be achieved through evacuation, most of the risk thereby avoided is realized by an evacuation to short distances from the site. This is graphically illustrated in the I plot of total risk of acute health effects versus protective action in Figure S-2. The dotted line at the top of this figure indicates how much the risk would have to increase before the average individual risk to the I population witnin 1 mile of the site boundary reaches the safety goal. In fact, between 70% to 95% of the risk that can be avoided through evacuation would be realized for an evacuation distance between 1 and 2 miles, respectively. The extremely small additional risk reduction I that is achieved with evacuation from 2 to 10 miles is matched by sheltering the same population segment. It is emphasized that the total potential for risk avoidance through protective actions of any kind is extremely small in view of the very low absolute levels of risk involved. l In addition to evaluating emergency planning options from a PRA perspective, the deterministic and historical bases for the current 10-mile EPZ were reviewed in light of new insights about source terms and lI 1323P120585 t
I the role of plant specific factors in the determination of plant safety. It was determined in this study that an EPZ level of 1 mile or less can be justified for Seabrook Station based on criteria similar to those used in NUREG-0396 to justify a 10-mile EPZ. This conclusion is supported by the results in Figures S-3 and S-4. These figures compare the dose versus distance relationships that were calculated in this study for Seabrook with those calculated in NUREG-0396 for whole body doses of 1, 5, 50, and 200 rem. The numerical estimates of risk obtained in this study exhibit significant uncertainties. Sens1tivity analyses were performed to determine how robust the above conclusions are with respect to different assumptions regarding the magnitude and impact of these uncertainties. l It was determined that the conclusions of this study are generally insensitive to a reasonable range of alternative assumptions regarding the magnitude of radioactive release source terms, the subjective weights assigned to different source term estimation procedures, and uncertainties in the definition and frequency quantification of accident sequences. An independent technical peer review group reviewed this report, generally concurred with the qualitative conclusions of the study, and agreed that these conclusions are insensitive to the inherent uncertainties. Risk management actions in addition to those of interest in emergency planning are evaluated in this report. These actions include refinements g and additions to the emergency operating procedures. The emphasis of a this evaluation was with respect to the dominant risk sequences identified in the SSPSA. Operator actions were identified to reduce the risk of loss of coolant events that bypass the containment, and this led to a refinement of the associated procedures. In conjunction with this evaluation, a new event sequence model for this class of events was developed and quantified. The remaining risk management actions identified and evaluated in this study were operator actions to recover containment heat removal during station blackout sequences that progress to core melt. Small reductions in latent health risk were estimated for these recovery actions. REFERENCES S-1. Pickard, Lowe and Garrick, Inc., "Seabrook Station Probabilistic I Safety Assessment," prepared for Public Service Company of New Hampshire ar.d Yankee Atomic Electric Company, PLG-0300, December 1983. S-2. Collins, H. E., et al ., " Planning Basis for the Development of g State and Local Government Radiological Emergency Response Plans in 3 Support of Light Water Nuclear Power Plants," prepared for the U.S. Nuclear Regulatory Commission, NUREG-0396, December 1978. I I I 1323P120585
I . I I I 10-2 I S BACKGROUND ACCIDENTAL FATALITY RISK E 10-3 I / (5 FATALITIES PER 10,000 POPULATION PER Y "o n b 104 - k a I u. k 10-0 - a SAFETY GOAL (.001 TIMES BACKGROUND RISK)
$ 10-0 -
I i ca 10-, - V I I O E THIS STUDY FOR SEABROOK STATION J l (NO IMMEDIATE [ j 10-8 - PROTECTIVE ACTION)
=
\
I
^
,0-0 A .
I FIGURE S-1. COMPARIS0N OF SEABROOK STATION RISK (WITH NO l IMMEDIATE PROTECTIVE ACTION) WITH BACKGROUND AND SAFETY G0AL INDIVIDUAL RISK LEVELS I I I S-3 I
I 10-2 , , I N NRC SAFETY GOAL FOR INDIVIDUAL RISK MULTIPLIED BY POPULATION WITHIN 1 MILE 10-3 - - E s a 5 5 m 10'A -- ~ ! p 3 e I 8 z 10-5 - b 2 I 10'O ~ ' REDUCTION IN 2-MILE = EVACUATION RISK WITH SHELTERING TO 10 MILES g i 1 3 X i f f ! I' 10-7 0 2 4 6 8 10 l EVACUATION DISTANCE (MILES) FIGURE S-2. ACUTE FATALITY RISK AS A FUNCTION OF PROTECTIVE ACTION ! i , I S-4
I I I I - i I ai1 l 4 i i e i a lij i e a a i a e i-I - , = -
$ 0.1 --
~~
E 8 3 I ag - Ea - 5Q , 50 REM wb - N*2 m 0 0.01 -- go
-~
z< - I $b5 200 REM N \ 50 REM 8 \ - I 5 O 0.001 -
\
O MEDIAN NUREG - 0396 7
\ _
l _ \ = = =
~
O MEAN
- THIS STUDY FOR
) SEABROOK STATION
~
I -
~
O I 200 REM 200 REM MEDIAN CURVE OFF O bD
~
l
/ l I 0.0001 i
SCALE C ' ' ' ' ' ' ' '' ' ' ' ' ' 1 10 100 1,000 DISTANCE (MILES) FIGURE S-3. COMPARISON OF UPDATED SEABROOK STATION RESULTS WITH NUREG-0396 - 200-REM AND 50-REM WHOLE BODY DOSE PLOTS FOR NO IMMEDIATE PROTECTIVE ACTION S-5 I
I I I 1 REM g y SREM 3 o 0.1 -- O _ _ - s a E 3 Oz h%
~
Sg g - EU - O - 8E g Eb - O , - h \ I b 0.01 g g O go : O
\ g
- g z< -
g O E
$5
- O \ -
82 mo -
\ \
O
\ 1 REM -
u us \ am - O g - jo \ O co -
\ \
b \ O O \o \ O 0.001 -.- \ \
,p OO -
1 \ NUREG - 0396 : l -
\ MEDIAN -
o 0 0 0 - 5 REM
^" -
THIS STUDY FOR
- SEABROOK STATION ~
l O- -O - MEDIAN 0.0001 ' ' ' l ' ' ' ' ' ' ' ' 1 10 100 1,000 l W DISTANCE (MILES) l FIGURE S-4. COMPARIS0N OF UPDATED SEABROOK STATION RESULTS WITH l NUREG-0396 REM AND 1-REM WHOLE B0DY DOSE PLOTS FOR 3 i RELEASES WITHIN 24 HOURS OF WARNING AND NO IMMEDIATE E j PROTECTIVE ACTION S-6
I
- 1. INTRODUCTION The purpose of this report is to document the results of a risk-based I evaluation of emergency plan options and other risk management actions for Seabrook Station. This evaluation is part of an ongoing risk management program at Seabrook Station and is intended to provide a key I input to the final form of the Seabrook Station emergency plan and to the emergency operating procedures. The information provided in this study, although not required by any regulatory bodies, was requested to help E determine prudent courses of action and the appropriate allocation of E resources in the risk management program.
The principal emphasis of this study is to evaluate the risk impact of I alternative protective actions, such as evacuation and sheltering. This evaluation is timely in view of the need to obtain NRC approval of an emergency plan for Seabrook Station. It is an especially appropriate I time for a reexamination of the risk bases for emergency planning because of advances in risk analysis, new insights about the nature and magnitude of source terms, and the mounting evidence that specific and unique I features of nuclear power plant systems, containment structures, and sites characterize risks. The insights obtained from the full-scope, Level 3 PRA recently completed for Seabrook Station and the risk model it presented have provided a much improved basis for understanding the risk I impact of emergency planning than was available previously, based on extrapolations from the Reactor Safety Study (Reference 1-1) made in Reference 1-2. The roles of the large dry primary containment, the unique secondary containment, the specific systems, and the unique site characteristics can now be fully reflected in all aspects of risk management, including the optimization of the emergency plan and emergency operating procedures. 1.1 OBJECTIVES The objectives of this study are to: o Evaluate the risk sensitivity to alternative emergency plan factors, I such as evacuation distance and sheltering. o Evaluate potential improvement to the emergency operating procedures j for risk dominant accident sequences. e Update the SSPSA containment response model to make use of new insights and tools for analyzing radioactive source terms. t I e Update the SSPSA plant model to incorporate more realistic treatment of sequences involving interfacing system LOCA and to account for I recovery of containment heat removal systems during station blackout-induced core melt sequences.
- g e Provide documentation suitable for review by the NRC and the l 3 technical community famil&r with PRA and nuclear reactor safety.
1 1272P112685
I
1.2 BACKGROUND
In December 1983, a full-scope Level 3 PRA was completed for Seabrook Station (Reference 1-3). The purpose of the PRA was to provide a baseline risk assessment and an integrated plant and site model for use throughout the period of plant operation as a risk management tool. Although the PRA was not required by the U.S. Nuclear Regulatory Commission, it was provided to the NRC and to the public for information in January 1984. While the Seabrook Station Probabilistic Safety Assessment was in progress and since it was completed, two separate research programs nave provided significant new insights about the nature and magnitude of radioactive source terms that could potentially result from hypothetical g severe core damage accidents. One of these was sponsored by the NRC 3 (Reference 1-4) and the other by the nuclear industry (Reference 1-5). Until these new insights became available, the best knowledge about g source terms for severe core damage accidents in LWRs had been the g Reactor Safety Study, which was completed in 1975. Prior to the SSPSA, source terms for PRAs have either been taken directly from the RSS or have been calculated by using the RSS methodology, or a comparable approach. While the same methodology (i.e., the CORRAL code, Reference 1-6) was also used to calculate an initial set of source terms in the SSPSA, adjustments were made to these source terms based on hand E calcalations and an evaluation of key uncertainties. This evaluation 3 established uncertainty distributions on the source term parameters in light of the source term research results that had been published in the 1982 to 1983 time frame when the SSPSA was carried out. Since the completion of the SSPSA, the initial phase of the industry research effort, IDCOR, has been completed and has provided additional information about source terms. Also, new tools to perform the necessary computations have been developed, particularly the MAAP code (Reference 1-7). Another important product of the IDCOR effort was new g insights about the event sequence progression of interfacing system LOCA E scenarios. These scenarios, which were first postulated in the RSS, are of particular interest at Seabrook Station. They involve multiple valve failures at the interfaces between the reactor coolant system and the residual heat removal system and have the potential for RHR system overpressurization. The interest in these scenarios stems from the fact that, as analyzed in the SSPSA and in other PRAs, they have often been found to dominate the risk of early health effects. The IDCOR research raised questions about some key conservative assumptions built into the PRA analyses of these sequences. The conservatisms relate to the g response of the RHR systems to overpressurization. The risk impact of g interfacing systems LOCA scenarios is reexamined in this report. Presently, the focus at Seabrook Station is to complete construction and obtain an operating license. In addition, the risk management program (References 1-8 and 1-9) is moving forward. A recent product of this program was an application of the SSPSA risk model to evaluate system g importance and to optimize the plant technical specifications 3 (Reference 1-10). I 1-2 1272P120585
I I This report is the next major product of the risk management program since completion of the SSPSA. Its purpose is to investigate alternative approaches to controlling and maintaining risk at acceptably low levels, especially those that mitigate the consequence of hypothetical severe I core damage accidents. Of particular interest are those approaches addressed in the emergency plan and the emergency operating procedures. This study provides an update of the source terms for the SSPSA risk I model, a more realistic treatment of interfacing system LOCA sequences, and incorporates small adjustments to the plant and systems models that were made in Reference 1-10. The key application of this enhanced risk I model is the evaluation of the risk impact of setting the evacuation zone at various distances from tne plant and tne effects of sheltering. Improved emergency operating procedures for interfacing systems LOCA sequences and those for maintenance of containment integrity during I severe core damage sequences involving a station blackout are also considered. The reason for focusing on these sequences for refining emergency procedures is that they were identified in the SSPSA as major I risk contributors. 1.2.1 OVERVIEW 0F SSPSA RESULTS This section briefly reviews key aspects of the SSPSA results. The full report and the Technical Summary Report (Reference 1-11) provide more comprehensive discussions of these results. The key results are as I follows: e The mean and median values of the uncertainty distribution for core melt frequency were found to be 2.3 x 10-4 and 1.9 x 10-4 events per reactor-year, respectively. I e Both the societal and individual risk provisions of the NRC safety goals were met by wide margins; hence, the risk to public health and safety was estimated to be extremely small. e Different risk factors were found to have different key contributors. Interfacing systems LOCA events and, to a lesser extent, seismic-induced transient events were the principal contributors to early health risk. The contributors to core melt I frequency and latent health risk were made up of a large group of initiators, including loss of offsite power, transient events, fires, and seismic events. e The dominant contributors to core melt frequency were support system I faults, external events, and internal hazards that affected both the core cooling and containment heat removal systems. As a result, a major fraction of the core melt frequency, 73%, was associated with sequences in which long-term containment overpressurization was l indicated, while only 1% was associated with early containment l failure. l E e In contrast with previous containment analysis, the timing of g containment overpressurization in the above sequences was found to be measured in units of days rather than hours. I I 1272P120585 1-3
~ I The above results are summarized for ease of reference in Table 1-1. In l the SSPSA, a systematic procedure involving matrix operations was used to 5 determine the principal contributors (see Section 13.2 of the SSPSA) according to the different groups of accident sequences that are defined g by the initiating events, plant damage states, and release categories. 3 Then, after the most important groups of accident sequences are identified, the examination of individual sequences was limited to those in the important groups. One important product of the matrix decomposition procedure in the SSPSA was the determination of the plant damage states that make significant g contributions to risk factors and to core :rcit frequency. Plant damage B states are particular types of severe core damage states that have a unique containment event tree analysis. The plant damage states of the g SSPSA risk model are presented in Table 1-2. The risk-significant plant g damage states are presented in Table 1-3. Of the 39 plant damage states defined in Table 1-2 for the SSPSA plant model, only 9 make significant contributions to risk or core melt frequency. A sequence-level perspective of the SSPSA risk contributors is provided in Table 1-4, which includes the top 20 sequences ranked by core melt g frequency contribution plus the dominant sequence contributing to early 5 health risk. The top sequence is the well-known " station blackout" sequence initiated by loss of offsite power. This sequence, and many g others in the table, involves a reactor coolant pump seal LOCA, and the g importance of support system faults and cc mon cause initiating events is evident. All the events that contribute to 'the sequence frequency (i.e., help reduce it by having probabilities less than 1) are indicated in the first two columns. The resulting dependent failures all have a conditional probability of 1, given the postulated failure of the associated support system or systems and result from functional 3 dependencies between systems or are the direct result of the initiating 3 event. Whil<s the dependent f ailures do not affect the numerical value of the sequence frequency, they need to occur to produce an accident sequence. The first update of the SSPSA was reported in Reference 1-10 to account for changes in the plant technical specifications and refinements to the systems and plant models. A second update is provided in this report to account for additional plant model refinements and new source terms, as well as alternative emergency plans. The cumulative effect of both 3 updates on the results of the Seabrook Station risk and core melt frequency assessment is described in Section 2. 3 1.2.2 EMERGENCY PLANNING The size of the emergency planning zone represents a considerable impact on both onsite and offsite emergency planning organizations. In 1978, a E task force of NRC and EPA representatives was formed to help establish a 5 planning basis and guidance for emergen"y planning requirements (NUREG-0396, Reference 1-12). Tne task force considered risk, g probability, cost effectiveness, and a spectrum of accident g consequences. However, the final determination of a generic 10-mile EPZ I 1-4 1272P112185
l l e I distance was based on a spectrum of accident sequences with a consideration of accident frequency. The Reactor Safety Study, issued in 1975, was used to characterize core melt accidents. In this study, a Seabrook Station-specific assessment of risk, frequency, and consequences is provided. The Federal government established the EPZ distance in NUREG-0396 and provides detailed planning guidance in NUREG-0654 (Reference 1-13). NUREG-0654 (page 12) summarizes the bases for the EPZ distance established in NUREG-0396:
- 1. " Projected doses from the traditional design basis accidents would not exceed protective action guide levels outside the zone."
- 2. " Projected doses from most core melt sequences would not exceed protective action guide levels outside the zone."
- 3. "For the worst core melt sequences, immediate life threatening doses would generally not occur outside the zone."
- 4. " Detailed planning within 10 miles would provide a substantial base for expansion of response efforts in the event that this proved necessary."
These bases are further described below with regard to the NUREG-0396 analyses and Seabrook Station-specific insights: I 1. The higher protective action guide (PAG) plume exposures of 25 rem (thyroid) and 5 rem (whole body) would not be exceeded for design-basis accidents beyond a distance of 10 miles. Most of the time, the lower PAG exposures of 5 rem (thyroid) and 1 rem (whole I body) would not be exceeded for design-basis accidents beyond the distance of 10 miles. At Seabrook Station, a secondary containment was constructed to reduce offsite exposures to the public from design-basis accidents.
- 2. The doses from less severe core melt releases, derived from WASH-1400, would not exceed the most restrictive PAG beyond 10 miles. At the time NUREG-0396 was written, it was believed that approximately 30% of all core melt accidents would exceed the PAG exposure outside the 10-mile EPZ. This result occurs because WASH-1400 concluded that most sequences melted through the basemat and did not catastrophically fail the containment.
At Seabrook Station, a full-scope plant-specific PRA provided a number of new insights regarding the timing of releases, the strength of the containment and other factors.
- a. The less severe core melt accidents for the Seabrook Station are l those with the containment intact rather than those with basemat I melt-through.
- I I 1272P112185 1-5
- b. Overpressure failure would occur very late, a matter of days E rather than the RSS assessment of several hours. E
- c. Early gross failure of the containment is very unlikely, about 1%
of the core melt frequency compared with 34% evaluated in the RSS for PWRs.
- 3. A substantial reduction in early severe health effects for the more severe core melts occurs at a distance of about 10 miles.
Figu're 1-11 in NUREG-0396 is a plot of the probability, given core E melt, of exceeding certain doses as a function of distance. The 5 highest dose plotted is 200 rem, characterized in NUREG-0396 as the dose above which "significant early injuries start to occur." Over g the years, it has been interpreted that this 200-rem dose curve drops 3 off significantly at about the EPZ distance (i.e., there is a knee in the curve at about 10 miles).
- 4. The fourth consideration was not analytic, but the following quote from NUREG-0396 (Page III-3) is believed to provide some insight:
I The Task Force had to decide whether to place reliance on general emergency plans for coping with the events of Class 9 accidents for emergency planning purposes, or whether to recommend developing specific plans and organizational capabilities to contend with such accidents. The Task Force believes that it is not appropriate to develop specific plans for the most severe and most improbable Class 9 events. The Task Force, however, does believe that consideration should be given to the characteristics of Class 9 events in judging whether emergency plans based primarily on smaller accidents can be expanded to cope with larger l events. This is a means of providing flexibility of g ~ response capability and, at the same time, giving 5 l reasonable assurance that some capability exists to minimize the impacts of even the most severe accidents. Guideline "4" addresses preparedness indices for evacuation beyond the EPZ should that need arise. This specific question was considered to be l beyond the scope of this analysis. For most source term categories, the l radionuclide release occurs much later and over a longer time period than was predicted in WASH-1400. Thus, the longer time available for emergency actions should certainly be a positive element in assessing guideline "4." Furthermore, it must be easier to expand emergency plan l actions from an EPZ radius of 1 mile to 2 miles than it would be to I implement the same actions for an EPZ radius of 10 miles to 20 miles. l The NRC's selection process for the 10-mile plume pathway EPZ incorporated observations drawn from quantitative probabilistic assessments of reactor accident consequences similar to those performed in this study to assist in defining an appropriate planning zone. l 1-6 l 1272P112185
l I 1 l I A number of critical assumptions were made in NUREG-0396 that had a direct impact on the results and, hence, the decision to set the EPZ at a distance of 10 miles for all plants. These include the assumptions that: 1 I e The plant and systems portions of the RSS risk models that are used to define and quantify accident sequences, which were based on the Surry and Peach Bottom plants, are representative of all U.S. PWR and BWR plants, respectively. e The site, meteorological, and demographical characteristics of the Surry plant are representative of all U.S. sites. e The source terms developed in the RSS adequately represent all risk-significant accident sequences at all U.S. sites. Large, dry containment structures would fail at relatively low pressures and at relatively early times during overpressurization scenarios and containment failure, at least by basemat melt-through, is an inevitable consequence of core melt. The results of more recent studies indicate that none of the above critical assumptions are valid. First, significant advancements have I been made since the completion of the RSS in various aspects of PRA methodology that have a direct impact on the risk assessment calculations that were performed in the RSS and NUREG-0396. These advancements have enhanced the completeness of accident sequences, improved the accuracy of frequency estimates, especially with respect to dependent events, and have taken advantage of an improved PRA data base upon which the accident I frequency estimates are based. In addition, major improvements have been made in the consequence assessment methodology, which was originally developed in the RSS, to enable a fully site-specific analysis to be made. With use of these advancements, a number of plant and site-specific PRAs have been completed whose results make it unnecessary to assume that two plants and one site adequately represent all the plants and sites in the l U.S. (References 1-14 and 1-15). The RSS provided the best information on nuclear reactor accident risk that was available in the mid-1970s. ! With the benefit of 15 years of continued development and application of l PRA and nuclear safety assessment, we are currently in a better position I to estimate risks, to characterize the unique and specific features of a plant, and to evaluate the risk impact of specific emergency plans. There have also been advancements in the ability to analyze containment capability to mitigate accident consequences. These new insights greatly l lengthen the time scales of many of the risk significant releases in the l SSPSA. This enhanced capability, the availability of a full-scope, level 3 risk model from the SSPSA, and the new insights on source terms now available dictate a reexamination of the risk impact of emergency planning options at Seabrook Station. 1.3 PROJECT OVERVIEW This project began in March 1985 and was completed in two phases. Both I phases included extensive applications of the SSPSA site and consequences model, which utilizes the CRACIT computer program--an advanced version of I 1272P112185 1-7
I the CRAC code that was developed and used in the RSS. In Phase 1, source terms calculated for Zion in the IDCOR program by using the MAAP code were input to CRACIT, and the applicability of these source terms to Seabrook Station was examined. It was concluded in Phase 1 that the best 3 set of realistic and upper bound source terms that could currently be 3 defined for Seabrook Station would consist of some of the source terms calculated in IDCOR for Zion, some calculated during the SSPSA, and new a plant-specific source terms using the MAAP methodology. This mix was g found to be necessary to adequately cover the full spectrum of scenarios in the SSPSA, to account for all risk-significant plant and site-specific features of Seabrook Station, and to adequately account for uncertainties. In Phase 2 of the project, the MAAP program was used to develop plant-specific source terms for sequences involving an interfacing g systems LOCA. Then, based on additional source terms from the SSPSA, 3 Zion (IOCOR), and NRC-sponsored research programs, a full set of best estimate and upper bound source terms were defined. The CRACIT code was g then executed to rebaseline the site model results of the SSPSA and to g investigate the impact of alternative emergency plans. In parallel, the plant model was updated to incorporate a new event tree model for interfacing system LOCA sequences. In addition, a new containment recovery model was developed to formulate and evaluate procedures for recovery of containment heat removal systems for station blackout-induced core melt sequences. In both phases, the quantification of uncertainties in the plant, containment, and site models was emphasized. A more detailed description of the project and its technical approach is provided in the following section. 1.3.1 TECHNICAL APPROACH The technical approach followed in this project consists of the performance of a PRA, as supplemented to achieve the particular objectives of this project. PLG's technical approach to the performance of a PRA is thoroughly documented in the SSPSA report. In view of the g large volume of material in the report, a pictorial guide to particular 3 topics is provided in Figure 1-1. A summary of the SSPSA technical approach is in Section 4 of the main report (Reference 1-3). The containment and source term analysis and the site and consequences analysis are described in Sections 11 and 12, respectively, of the SSPSA. A summary of the technical approach, with an emphasis on the departures from the SSPSA approach made in this evaluation, is provided in the following sections for each of the three major elements of the risk model: the plant model, the containment model, and the site model. i The major elements of these models and the roles they play in the El definition of accident sequences are provided in Figures 1-2 and 1-3, g respecti vely. 1.3.1.1 P1 ant Model As illustrated in Figure 1-2, the overall Seabrook risk model is composed of the plant model, the containment model, and the site model . While each of the three models plays an important part in the definition of accident sequences, whose standard form is illustrated in Figure 1-3, the I 1-8 1272P120585
plant model wholly contains the analysis of plant systems. Hence, all the information about the contribution of systems to accident sequence frequencies is contained within the plant model. An important characteristic of the risk model is Step 5 in Figure 1-3, the interface between the plant and containment models. This is known as the plant damage state where the sequences progress either to successful termination or to severe core damage state. Hence, core melt frequencies can be calculated using only the plant model. I The approach use( to construct and quantify the plant-level portion of the event sequenr.2 model is called the modularized event tree approach. The modularized event tree structure employed in the SSPSA is seen in Figure 1-4 to c.:nprise 16 separate event tree modules. One represents I all the plant auxiliary or support systems, 9 cover the early response of the frontline systems to 9 groups of initiating events, and 6 cover the long-term (i.e., after switchover to recirculation cooling) response of I the frontline systems. Within each event tree block, there is a separate event tree module, such as that illustrated in Figure 1-5 for the TRAN module. The TRAN module models the response of frontline systems to the I transient group of initiators indicated in Figure 1-4. The 58 initiating events and the 39 plant damage states, which start and end the plant model sequences, are defined in Tables 1-5 and 1-2, respectively. The remaining event tree modules are defined in Table 1-6. If each sequence of the plant model is defined by each specific initiating event and all paths through the event tree modules, this plant I model includes over four and one-half-billion sequences. However, at each interface between event tree modules, the sequences were pinched or
" grouped" according to the similarity of their downstream impact. This grouping process greatly facilitated the handling of such large numbers of scenarios in an efficient and organized manner.
The modularized event trees and the pinch points that are used to connect the modules together contain all the information related to intersystem dependencies. Initiating event dependencies are accounted for by performing a separate quantification for each of 58 initiating events, each with a set of input data to reflect the impact the initiating event has on the plant. The fine structure of the event trees greatly simplifies the analysis of individual systems or subsystems. Take, for example, the event tree top event in column three of the general I l I transient event tree in Figure 1-5, "EF - Feedwater and Steam Generator Relief." Part of the system analysis procedure for the emergency feedwater system is to construct a reliability block diagram like the one I in Figure 1-6. The block diagram is used to determine the minimal cutsets of the system (normally can be found by inspection) and to define the different sets of boundary conditions dictated by the initiating events and support systems, for which separate quantifications need to be performed. The whole TRAN event tree, for example, is quantified for each combination of initiating event and support system states. In a given event tree quantification, the appropriate quantification for EFW I and all the other systems is used. I 1272P112185 1-9
The SSPSA plant model was recently updated in support of a technical g specification improvement program (Reference 1-10). This update g incorporated two significant changes to modeling systems and accident sequences. One change was a more complete modeling of the contribution of common cause events to the unavailability of the primary component cooling water system. The second change was a more realistic treatment of containment isolation for sequences involving seismically induced loss of offsite power. The more realistic treatment took into account that air-operated valves in the containment purge penetrations would eventually close because the station air compressors, supplied by nonclass 1E electric power supplies, would stop running and the air lines would experience pressure decay. These changes and some minor changes that were made in the system analyses to account for changes in the technical specifications led to an adjustment in the plant model results and the ranking of accident sequences for core melt frequency and risk. Hence, Reference 1-10 was the starting point for the plant analysis update in this study. Additional changes were made to the SSPSA plant model in this evaluation. The first change is the reevaluation of accident sequences 3 involving interfacing system LOCA scenarios. In the SSPSA, the modeling 3 of these scenarios was similar to that of other PRAs that have been performed since, and including, the RSS. This approach to modeling the interfacing system LOCA identifies all the interfaces between the reactor coolant system and interfacing systems that are not designed to RCS pressure. At Seabrook, the key interfaces were the double motor-operated valve and double check valve interfaces between the RHR system and the RCS. A model and component operating experience data are then used to estimate the frequency of multiple valve ruptures (necessary to produce shock waves) in these lines. In the SSPSA, the total frequency of these events were estimated to have a mean value of 1.8 x 10-6 per reactor-year. This frequency estimate exhibited a large uncertainty because of the scarcity of data (zero event statistics). Once postulated, these LOCA scenarios were assumed to result in a rupture of the low pressure RHR piping outside containment. Consequently, a core melt with containment bypass condition was assumed to result. In the updated analysis in Section 3.1, a full event tree analysis of these sequences is performed to take into account realistic pressure transients that could result within the RHR system; the pressure capacity of the RHR piping, pump seals, and heat exchangers; the behavior of the RHR system relief valves inside and outside the containment; additional data on check valve leaks and ruptures; and an improved initiating event model. These changes bring the level of sophistication of the analysis for these sequences more in line with the rest of the sequences modeled in the SSPSA. They also enhance the perspective on risk significance of interfacing systems LOCAs. The second major change to the plant model in this evaluation is the incorporation of a containment recovery model, which is described in Section 3.2. The approach followed in the SSPSA for modeling recovery 1-10 1272P120585
I actions was to first evaluate the plant model accident sequences without recovery. Then, these actions were incorporated on a sequence-by-sequence basis to the extent necessary to achieve results that are insensitive to further recovery actions. With a few exceptions, all the recovery actions considered in the SSPSA were those necessary to prevent severe core damage. Following the onset of core melt, the only recovery actions that were considered were those associated with manually closing containment isolation valves in some of the penetrations with small (< 3-inch) lines for sequences in which the automatic means of containment isolation was postulated to be unavailable. The top ranking accident sequence with respect to core melt frequency and latent health risk in the SSPSA and in the updated results of Reference 1-10 is a station blackout sequence. The steam-driven I emergency feedwater pump is successful, but an unmitigated pump seal LOCA is assumed to occur. Recovery accions to re: tore electric power were considered up to the time of core damage--about 14 hours--and those I actions significantly reduced the frequency of severe core damage for this sequence. In this evaluation, recovery actions after this time and before containment overpressurization failure are incorporated. I Containment failure occurs sometime between 1-1/2 and 5 days after the initiating event. Although these particular actions do not reduce core melt frequency, they reduce the potential for consequences of such sequences. The results of the SSPSA have clearly shown that, with successful operation of the containment building spray system and with conditions for debris bed cooling established, long-term integrity of the containment is very likely and offsite consequences are, as a result, negligible. Hence, the identification of procedures for containment recovery in Section 3.2 provides a potentially effective approach to postaccident risk management as well as a better perspective on risk due to potential accidents at Seabrook Station. 1.3.1.2 Containment Model The core and containment response analysis comprises the development of the containment model. Exparience has shown that the severity of the threat to containment integrity can be categorized according to certain physical conditions occurring during an accident sequence involving severe core damage. Such conditions include the following: s The time at which the core becomes uncovered with water, e The size of the leakage path for water leaving the primary vessel and l the residual pressure inside the vessel. 1 e The presence or absence of water in the containment below the primary vessel. e The availability of containment sprays and heat removal systems. l For example, without the availability of containment heat removal, the threat to the containment structural integrity can be substantial. Also, the longer the time until core uncovery, the lower the heat source which threatens containment integrity. 1-11 1272P112185 l
I As explained previously, event sequences terminate in plant damage states at the end of the plant model. Each plant state generally contains a number of sequences stemming from different initiating events. However, these sequences have connon conditions with respect to the containment impact assessment. Each plant state does not lead to a unique containraent condition or fission product release. Given a particular plant state, there is a vector of conditional frequencies of possible containment damage states which are translated into release categories. The collection of vectors for all plant states is termed the containment damage matrix, or S "C" matrix for short. The major focus of the containment analysis is to E quantify the conditional release category frequencies in the C-matrix, with probability distributions rather than point estimates being desired. Typical containment damage states are: early containment overpressure failure; late containment overpressure failure; basemat melt-through; and steam explosion. Together with several plant state conditions, such as containment spray availability and initial containment isolation, the containment damage states help define the release categories. The way this is done is illustrated in Figure 1-7 for cases in which the E containment is initially isolated. A three dimensional discretization is g made, with the coordinate axes defined by: (1) containment failure mode; (2) fission product release characteristic; and (3) containment spray g availability. Typical containment failure modes are early g overpressurization, late overpressurization, and basemat concrete mel t-th rough . Typical release characteristics are large or small leakage paths through the containment structure and oxidation of radionuclides. Each combination of conditions identifies a unique release category. There are substantial uncertainties in the magnitudes of fission product release, given the release conditions. The elements of the containment analysis consist of a study of degraded core and core melt processes, an evaluation of thermodynamic conditions in the containment, an assessment of the capability of the containment structure to withstand these thermodynamic conditions, and a radioactivity release analysis. These elements are fully described in i Chapter 11 of the SSPSA. l The principal departurcs of this evaluation from the SSPSA in the containment model area are the approach to defining source terms and the g reassessment of source term uncertainties in light of the NRC and IDCOR E findings on this subject. In the SSPSA, the CORRAL code was used in l conjunction with the MARCH /C0C0 CLASS 9 series of codes (see Section H.2 of I Appendix H in the SSPSA) to develop an initial upper bound set of source terms. These results, supported by hand calculations and uncertainty propagation calculations, were used to construct a discrete set of four different scerce terms for each release category found to make a significant contribution to risk. Uncertainty propagation calculations were made to assign probability weight to each of the four source terms. Separate CRACIT runs were made to cover each part of these source term uncertainty distributions. 1-12 i 1272P112185 1
1 In this evaluation, the objective was to incor the IDCOR rasearch in a cost-effective manner.porate the contributions In Phase 1 of this of evaluation, it was determined that all of the Zion source terms were applicable to Seabrook Station except for the interfacing system LOCA source term. On the other hand, there were not enough source terms calculated for Zion in the 1DCOR program to cover the complete set of accident sequences defined in the SSPSA. A detailed comparison of Zion I and Seabrook containment and plant characteristics was made to support this conclusion regarding source term applicability. Because of its risk dominance in the SSPSA and significant plant differences between Zion and I l Seabrook for interfacing system LOCAs, new MAAP calculations and model refinements were made in this study to generate new source term information for this sequence only. Because the Zion IDCOR source terms did not fully represent the spectrum of accident sequences of the SSPSA, I the final set of source terms in the evaluation came partly from the SSPSA, partly from the Zion IDCOR work, and partly from the new MAAP analysis for the interfacing systems LOCA. The uncertainty distributions for source terms were then reassessed. Because of the great reduction in source term uncertainty since the SSPSA, the number of different sets of source terms to represent uncertainty was reduced from four to two. A best estimate and a conservative or " upper bound" source term, for each release category was defined. The containment and source term analysis is more fully described in Section 4 1.3.1.3 Site Model The consequence methodology used in this study is the same as that used in the SSPSA, which is described in detail in Chapter 12 and Appendix J of the SSPSA. Thus, only a brief summary is presented here. The health effects from released radioactive material are dependent on the radiation exposure or dose. The dose, in turn, depends on the time spent in areas where radioactive material is present. This could be in the plume or in areas where radioactive materials have deposited on the ground. The number of random variables associated with the determination of dose and health effects dictates the need for a probabilistic l calculation, in this case a Monte Carlo simulation, of consequences. The probabilistic calculation of consequences requires a computerized model that incorporates site-specific meteorology, population, evacuation, shielding (sheltering), and other site-specific effects, such as sea breeze. The CRACIT code was used in the SSPSA to compute consequences and is used here to evaluate the effects of revised source term information as well as to determine risk sensitivities to variations in the emergency plan. This program was derived from CRAC. The CRAC code was developed by the Nuclear Regulatory Commission for the Reactor Safety Study and was the first code used to perform a comprehensive probablistic assessment of l 3 consequences from a severe reactor accident. The assessments included l l treatment of the effects of plume rise, wet and dry deposition, and
- changes in meteorological conditions (except wind direction) during plume 1-13 1272P112185
t ransport. It also simulated the impact of evacuation and other mitigative measures, and it modeled doses and health effects from both early and chronic phases of exposure. The CRAC program was intended for use in computing the composite risk from a number of nuclear plants in different regions throughout the United States. The CRACIT program was developed for use in plant- and site-specific PRAs, such as the SSPSA. It incorporates such features as variable direction wind trajectories, actual evacuation routes, and topological features that enabled a more realistic simulation of site-specific conditions. The CRACIT code and the PRA theory of accident occurrence assume that an accident release occurs at any random point in time. The consequences of an accident depend on when the radioactive material is assumed to have been released from the plant. This is true because the meteorological 3 conditions that exist during dispersion of the released material greatly 3 influence results. Wind speed, wind direction, atmospheric mixing, and precipitation determine the affected area and the concentrations of airborne and deposited radioactive material at all locations. Doses are calculated by using the distribution of the radioactive material in the environment, the evacuation routes, and the location of the population, all as a function of time. Each postulated release is dispersed by many weather condition scenarios to simulate damage for a range of possible environmental conditions. These results are used to develop the cumulative probability distributions. Weather scenarios are selected randomly from a representative 1-year period of hourly data. Sequential hourly meteorological data are used to calculate plume trajectory and concentration changes as a function of time. Thus, the effect of time-varying weather conditions on plume transport and dispersion is simulated in CRACIT. Accident sequences that have similar release characteristics are grouped into release categories. A separate set of consequence analysis computer runs (with many weather scenarios) is made for each release category. A g set of runs is used for each category to represent a discrete uncertainty g distribution of CRACIT input parameters. This distribution quantifies uncertainties in the source term and in key parameters of the consequence calculation (e.g., the number of health effects per man-rem exposure). The use of release categories and a small number of uncertainty cases enables the use of relatively few computer runs to represent a large number of accident sequences. During the computer run, a frequency distribution for each health effect is plotted for each release category. This distribution is called the conditional risk distribution (conditional on the occurrence of the release category). The conditional risk distribution is combined with the frequency of occurrence for the particular release category to define an absolute risk distribution. The total absolute risk from the plant is the sum of the absolute risk distributions for all relase categories. The only significant departures in the technical approach to the site 3 model for this study in comparison with the SSPSA are the risk f actors 3 that were calculated and the treatment of uncertainties. In addition to I 1-14 1272P112185
' f the standard risk curves normally calculated in a PRA, additional curves that portray the spatial distribution of risk and the amount of risk averted, as a function of evacuation distance, were gene, rated, as shown in Section 2. The number of site model uncertainty cases was reduced ,
from three to two to account for both the source terms and site model I uncertainty. Therefore, a total of four CRACIT runs were made for each significant release category, two runs each of two sets of source term assumptions. The site model is more fully explained in Section 5. 1.3.2 PROJECT TEAM Pickard, Lowe and Garrick, Inc., was responsible for the SSPSA and has the lead responsibility for conducting this evaluation. In addition to PLG, two of the major subcontractors of the SSPSA team were retained as key participants in this project. They are Westinghouse Electric Corporation and Fauske and Associates, Inc. They bring to the team, not only the experience from the SSPSA and the reactor vendor, but also those responsible for the IDCOR research products that were used in this study. FAI was responsible for developing the MAAP methodology and for conducting major elements of the IDCOR source term research. Westinghouse was responsible for applying the MAAP methodology to Zion, making key contributions to the IDCOR research, and, in the SSPSA, for applying the RSS methodology to define an initial set of source terms. Individuals who made up the project team are acknowledged on the inside cover of this report and include the project manager, key principal investigators of the SSPSA, the principal authors of the CRACIT and MAAP computer codes, and recognized experts in PRA, nuclear reactor safety, consequence assessment, and emergency planning. 1.3.3 TECHNICAL REVIEW AND QA I The SSPSA as well as this evaluation were performed in accordance with a set of technical review and quality assurance procedures that PLG had developed for use in PRA projects (Reference 1-16). These procedures included those elements of 10CFR50, Appendix B (Reference 1-17), that are to be applicable to applied risk and reliability evaluations. They call I for internal and client audits to ensure the procedures are being followed. A multistage technical review was required for the SSPSA, as summarized in Table 1-7. This review recognizes a number of different responsibilities for the performance of technical reviews. These included different responsibilities for the project team members, the plant owner, and a technical review board. The SSPSA technical review board consisted of the following people: e Frank R. Hubbard, Chairman, Pickard, Lowe and Garrick, Inc. e George Apostolakis, University of California, Los Angeles e William K. Brunot, Private Consultant e Vijay K. Dhir, University of California, Los Angeles e William T. Hussey, QA Manager, Pickard, Lowe and Garrick, Inc. e Mohammad Modarres, University of Maryland I 1272P112185 1-15
I e Donald A. Norman, University of California, San Diego e Norman C. Rasmussen, Massachusetts Institute of Technology e James E. Shapley, Pickard, Lowe and Garrick, Inc. e Walter B. Sturgeon, Public Service Company of New Hampshire e Juliette Zivic, Public Service Company of New Hampshire The SSPSA technical review board reviewed two preliminary drafts of the report, conducted several meetings at which team members gave 3 presentations and answered questions and provided written comments for E resolution in the final report. For this evaluation, the normal internal technical reviews called for by the QA procedures were performed. This led to the production of a draft report that was submitted to New Hampshire Yankee and their technical peer review group in October 1985. This technical peer review group included the following individuals: e Robert Budnitz, Chairman, Future Resources Associates, Inc. 3 e David Aldrich, Science Applications Incorporated 3 e Joseph Hendrie, Consultant e Norman Rasmussen, Massachusetts Institute of Technology e Robert Ritzman, Electric Power Research Institute e William Stratton, Consultant e Richard Wilson, Harvard University After completion of their review of the report, the peer review group conducted a 2-day review meeting with the study team. At the meeting, presentations were provided and the review group's questions were answered. This final report dated December 1985 incorporates the peer review group comments. The peer review group conclusions are documented in Reference 1-18 and are included in this report as Appendix E. 1.4 REPORT GUIDE The technical results of this study are summarized in Section 2, together with the overall conclusions of the evaluation. These results include an update of the SSPSA results, an assessment of the risk impact of emergency planning options, and a recommendation to enhance emergency operator procedures for interfacing system LOCA and station blackout sequences. Detailed aspects of the results are provided in separate sections for the plant model (Section 3), source terms (Section 4), and site model (Section 5). For the convenience of the reader, acronyms and selected definitions used in this report are provided after the table of contents.
1.5 REFERENCES
1-1. U.S. Nuclear Regulatory Commission, " Reactor Safety Study: An Assessment of Accident Risks in U.S. Commercial Nuclear Power Plants," WASH-1400, NUREG-75/014, October 1975. 1-16 g 1272P121685 g, 1
I 1-2. U.S. Nuclear Regulatory Commission, " Final Environmental Statement Related to Operation af Seabrook Station Units 1 and 2," NUREG-0895, December 1982.
;- t. Pickard, Lowe and Garrick, Inc., "Seabrook Station Probabilistic Safety Assessment," prepared for Public Service Company of New Hampshire and Yankee Atomic Electric Company, PLG-0300, December 1983.
1-4. Gieseke, J. A., et al., "Radionuclide Release under Specific LWR Accident Conditions," Volume VI, "PWR Large, Dry Containment I Design (Zion Plant)," Batelle Columbus Laboratories, BMI-2104, July 1984. 1-5. Technology for Energy Corporation, " Nuclear Power Plant Response to Severe Accidents, IDCOR Technical Suntiary Report, November 1984. 1-6. Burian, R. I., and P. Cybulskis, " CORRAL II Users Manual," Battelle Columbus Laboratory, January 1977. 1-7. "MAAP - Modular Accident Analysis Users Manual," Technical Report I on IDCOR Tasks 16.2 and 16.3, May 1983. 1-8. Fleming, K. N., J. H. Moody, and K. L. Kiper, "The Seabrook PRA Viewed from Three Perspectives," presented at International ANS/ ENS Topical Meeting on Probabilistic Safety Methods and Applications, San Francisco, California, Fettruary 24-28, 1985. 1-9. Thomas, George S., New Hampshire Yankee, leater to G. Knighton, U.S. Nuclear Regulatory Commission, " Supporting Analysis for Technical Specifications Improvement Prograni," August 22, 1985. 1-10. Pickard, Lowe and Garrick, Inc., " Risk-Based Evaluation of Technical Specifications for Seabrook Statf or ," prepared for New I Hampshire Yankee Division, Public Service of New Hampshire, PLG-0431, August 1985. 1-11. Pickard, Lowe and Garrick, Inc., "Seabrook Station Probabilistic Safety Assessment Technical Summary Report," prepared for Public Service Company of New Hampshire and Yankee Atomic Electric Company, PLG-0365, June 1984. 1-12. Collins, H. E., et al., " Planning Basis for the Development of State and Local Government Radiological Emergency Response Plans I in Support of Light Water Nuclear Power Plants," prepared for the U.S. Nuclear Regulatory Commission, NUREG-0396, December 1978. 1-13. Federal Emergency Management Agency, " Criteria for Preparation and Evaluation of Radiological Emergency Response Plans and Preparedness in Support of Nuclear Power Plants," prepared for the Nuclear Regulatory Commission, NUREG-0654, January 1980. I 1272P112185 1-17
1-14. Garrick, B. J., "Recent Case Studies and Advancements in Probabilistic Risk Assessment," Risk Analysis, Vol. 4, No. 4, 1984. 1-15. Garrick, B. J., " Experience and Advancements in Risk Assessment," E Lecture, Nuclear Power Plant Risk and Safety Course, Massachusetts E Institute of Technology, July 16, 1984. 1-16. Pickard, Lowe and Garrick, Inc., " Quality Assurance Program," PLG-0223, Revision 4, March 1985. 1-17. U.S. Nuclear Regulatory Commission, Title 10, Code of Federal Regulations, Part 50, January 1,1983. 1-18. Budnitz, R. J., Future Resources Associates, Peer Revie 4 Group Findings, letter to New Hampshire Yankee Division of Public Service of New Hampshire, November 9,1985. l 11 I I 1-18 1272P112185
M M M M M M M M TABLE 1-1.
SUMMARY
OF PRINCIPAL CONTRIBUTORS TO RISK IN TERMS OF ACCIDENT SEQUENCE GROUPS AND INITIATING EVENTS FROM THE SSPSA Containment Response - Group Group Fraction of Contributing Contribution Frequency Total Release Seq e e roup Percent Frequency Initiating Events (mean values) Group I Early Containment f ailure 2.4 x 10-6 per .01 Early Health - Interfacing LOCA 76 Reactor Year or Effects - Seismic 24 Once in 410,000 T06 Reactor Years Group II Delayed Containment Failure 1.7 x 10-4 per .73 Latent Health - Loss of Offsite Power 40 Reactor Year or Effects, - Transients 19 Once in 6,000 L - Fires 15 Reactor Years u) - Seismic 15
- Others 11 100 Group III Ccntainment Intact No Health - Transients 57 6.0 x 10-5 per .26 Effects - SLOCA 29 Reactor Year or
- Others 14 Once in 17,000 TDU Reactor Years Total 2.3 x 10-4 per 1.00 Reactor Year or Once in 4,300 Reactor Years 1296P092485
l ' M M T e F ' AH " D RAS )A E C R( RC F "' S U T M IA O N
)
P T D U E HF( B GC M e R . NN R E L E D E T L I F N N I E 3TE P<M O IE A 8 L L B E I D S SO OM U I D P K O O YS M N H(
)
F LI R L AA M K T GN NE CR CS S R I P I SS I R . A T S N 3E E P >M IT
@ YS H
PI N A S T L E O A l O M P M S E D 8 :[- S R O E 8 . C R 8 l E N I F O E T L I F E ( 8 s M D E M I T M lg
- l S T U A S S
T C E N ) E I T M E A O D ( N T N I T N A A I T R T T N E S CY E E M N TL NT UL I M G A I NCA F B I A OUV Y OA IDO LN()C M T N S
@ [C/ 9A E P A
D T N A O C S E Y SOM O FIRE PR L M - SA UC AR CE EL BO DO EC M L T A S P A V Y L) U N A 9 E HM O N (B E O [ TF O NI N M 3 F O D TL R
/7 8 /
8 M e [) NNCA O N I AOUV) TS I DOA 8 8
)
T ASOM( EFI RE I N I F l H PR 2 ' M E H ND TI OET
> ) ) >
D RG ST " 4 ( 7 ( - OU WC I A S - S TO e
. SC R NA TH RET J NNI ~
E Y O N - ) 9 ( KE OR OU 2 1 OE I R - IT F LT I I E R U RT BA AE M DOE R
) DT E F N EM A EP SN E O ML E O I TU L CI E R TLS AR YG B T S UCES0P ,
L BI S S S AS0 O E DE A T TS AE V E E E3 R RV >g
, S B I U NT O R ED DN M
PN U LI O I ( N O HT CT TA EA RT I W ER PS E R M O T TCW I N S R TN LE M U EG T L AO YO M M E D L E EA R H E ' RT T R E M E U A 6 A OT E CH CS R N S OI C S E < L M (l O ll ll I l 1ll1l L
I I I I TABLE 1-3. IDENTIFICATION OF IMPORTANT PLANT DAMAGE STATES IN THE SSPSA RISK MODEL Significance Relative To: I Accident Sequence Plant Damage State (see Dominant Containment Risk Occurrence I Groups Table 1-2) Response Early Effects latent Effects Frequency I 1F Large (> 3") Major Major Minor 3F Open Penetration Major Major None or Bypass 7FP Small (< 3") Minor Major Minor 3FP Open Penetration Minor Major Minor I II anj Overpressuri-zation 80 Long-Term None Major Major 7D Overpressurization None Major Major 30 None Major Minor 8A Long-Term None None Major III 4A Leak-Tight None None Minor Integrity I lI I lI I 1-21 l 1296P092485
TABLE 1-4.
SUMMARY
OF ACCIDENT SEQUENCES WITH SIGNIFICANT RISK AND CORE MELT FREQUENCY CONTRIBUTIONS FROM SSPSA Sheet 1 of 2 Sequence Ranking Additional System Failures / equence Initiating Event Human Actions Resulting Dependent Failures Frequency g Latent Early (per reactor year) Heal th Health Mel t Risk Risk Loss of Offsite Onsite AC Power, No Recovery of AC Power Component cooling, high pressure makeup 3.3-5 1 1
- Power Before Core Damage (ECCS), reactor coolant pump seal LOCA, containment filtration and heat removal.
Loss of Offsite Service Water, No Recovery of Offsite Onsite AC power, component cooling, high 9.2-6 2 2
- Power Power and low pressure makeup (ECCS), reactor coolant pump seal LOCA, containment filtration and heat removal.
Small LOCA Residual Heat Removal None. 8.9-6 3 *
- C:ntrol Room None Component cooling, high and low pressure 8.7-6 4 3
- Fire makeup (ECCS), reactor coolant pump seal LOCA, containment filtration and heat y removal.
N Loss of Main Solid State Protection System Reactor trip, emergency feedwater, high 8.3-6 5 4
- Fcedwater and low pressure makeup (ECCS), contain-ment filtration and heat remov'a1.
Steam Line Operator Failure to Establish Long-Term 5.6-6 6 *
- Break Inside Heat Removal Containment Rmactor trip Component Cooling 4.6-6 7 5
- High and reactor low pressure coolant makeup pump seal (ECCS),
LOCA, conta in-ment filtration and heat removal. Loss of Offsite Train A Onsite Power, Train B Service Train B onsite power, component cooling, 4.4-6 8 6
- Power Water, No Recovery of AC Power Before high and low pressure makeup (ECCS),
Core Damage reactor coolant pump seal LOCA, contain-ment flitration and heat removal. Loss of Offsite Train B Onsite Power. Train A Service Train A onsite pcwer, component cooling, 4.4-6 9 7
- Power Water No Recovery of AC Power Before Core 6amage highandlowpressuremakeup(ECCS) reactor coolant pump seal LOCA, con bin-ment filtration and heat removal.
PCC Area Fire None Component cooling, high and low pressure 4.1-6 10 8
- makeup (ECCS), reactor coolant pump seal LOCA, containment flitration, and heat removal.
- Negligible contribution to risk.
NOTE: Exponential notation is indicated in abbreviated form; i.e., 3.3-5 = 3.3 x 10-5,
~1296P082885 M M M M M M M M M M M
M M M M M M M M M M M M M TABLE 1-4 (continued) Sheet 2 of 2 Sequence Ranking
" * "9 ^ " ""
H Resulting Dependent Failures Fr qu n y Latent Early (per reactor year) Health Health Risk Risk Partial Loss of Component Cooling High and low pressure makeup (ECCS), reactor 3.8-6 11 9
- Main Feedwater coolant pump seal LOCA, containment filtra-tion, and heat removal.
Cable Spreading None Component cooling, high and low pressure 3.5-6 Room Fire 12 10
- makeup (ECCS), reactor coolant pump seal LOCA, containment filtration, and heat removal.
Less of One DC Emergency Feedwater, No Recovery of Bleed and feed cooling Train A 3.2-6 *
- 13 Bus Emergency or Startup Feedwater containment filtration and heat removal.
Reactor Trip Operator Failure to Establish Long. Term None. 3.0-6 14 *
- Heat Removal, e-* Turbine Trip Component Cooling 2.8-6 15
- High and low pressure makeup 11 L reactor coolant pump seal LOCA,(ECCS),
conta in-w ment filtration, and heat removal. Less of Service None Component cooling, high and low pressure 2.3-6 16
- Water 12 makeup, reactor coolant pump seal LOCA, containment filtration, and heat removal. .
Partial Loss of Operator Failure to Establish Long-Term None. 2.3-6 17 *
- Feedwa ter Heat Removal Turbine Building Onsite AC Power, No Recovery of AC Power Offsite power, component cooling, high 2.3-6
- Fire Before Core Damage 18 13 and low pressure makeup (ECCS), reactor coolant pump seal LOCA, containment filtra-tion, and heat removal.
Small LOCA Train B Safety Features Actuation. Train A high and low pressure makeup and 2.2-6 19 *
- Train A Residual Heat Removal residual heat removal; train B containment filtration and heat removal.
Small LOCA Train A Safety Features Actuation, Train B high and low pressure makeup and 2.2-6 20
- Train B Residual Heat Removal residual heat removal; train B containment filtration and heat removal.
Interfacing None Low pressure makeup, residual heat Systems LOCA 1.8-6 ~ 27 14 1 removal co filtration.ntainment isolation and l
- Negligible contribution to risk.
) NOTE: Exponential notation is indicated in abbreviated form; f.e., 3.8-6 = 3.8 x 10-6, 1296P082885
I TABLE 1-5. INITIATING EVENTS SELECTED FOR QUANTIFICATION OF THE SEABROOK STATION RISK MODEL Sheet 1 of 2 Group Initiating Event Categories Selected Code for Separate Quantification Designator e Loss of Coolant 1. Excessive LOCA ELOCA Inventory 2. Large LOCA LLOCA
- 3. Medium LOCA MLOCA
- 4. Small LOCA SLOCA
- 5. Interfacing Systems LOCA V
- 6. Steam Generator Tube Rupture SGTR e General 7. Reactor Trip RT Transients 8. Turbine Trip TT E
- 9. Total Main Feedwater Loss TLMFW E
- 10. Partial Main Feedwater Loss PLMFW
- 11. Excessive Feedwater Flow EXFW
- 12. Loss of Condenser Yacuum LCV
- 13. Closure of One MSIV IMSIY
- 14. Closure of All MSIVs AMSIV
- 15. Core Power Excursion CPEXC l 16. Loss of Primary Flow LOPF
- 17. Steam Line Break Inside Containment SLBI i
- 18. Steam Line Break Outside Containment SLB0
- 19. Main Steam Relief Valve Opening MSRV
- 20. Inadvertent Safety Injection SI e Coman Cause 3 Initiating E l Events
- Support 21. Loss of Offsite Power LOSP i
System Faults 22. Loss of One DC Bus L1DC 1
- 23. Total Loss of Service Water LOSW l 24. Total Less of Component Cooling LPCC l
Water
- Seismic Events
- 25. 0.79 Seismic LOCA
- 26. 1.0g Seismic LOCA
- 27. 0.2g Seismic Loss of Offsite Power E.7L El.0L E.2T l
l.
- 28. 0.3g Seismic Loss of Offsite Power E.3T E
- 29. 0.4g Seismic Loss of Offsite Power
~
E.4T E
- 30. 0.5g Seismic Loss of Offsite Power E.5T
- 31. 0.79 Seismic Loss of Offsite Power E.7T
! 32. 1.0g Scismic Loss of Offsite Power E1.0T l I I 1296P082885
I TABLE 1-5 (continued) Sheet 2 of 2 Group Initiating Event Categories Selected Code for Separate Quantification Designator
- Fires 33. Cable Spreading Room - PCC Loss FSRCC
- 34. Cable Spreading Room - AC Power Loss FSRAC
- 35. Control Rocm - PCC Loss FCRCC
- 36. Control Room - Service Water Loss FCRSW
- 37. Control Room - AC Power Lo'ss FCRAC
- 38. Electrical Tunnel 1 FET1 I 39. Electrical Tunnel 3
- 40. PCC Area
- 41. Turbine Building - Loss of Offsite FET2 FPCC Power FTBLP
- Turbine 42. Steam Line Break TMSLB Missile 43. Large LOCA TMLL l 44. Loss of Condenser Vacuum
- 45. Control Room Impact
- 46. Condensate Storage Tank Impact TMLCV TMCR TMCST
- 47. Loss of PCC TMPCC
- Tornado 48. Loss of Offsite Power and One MELF Missile Diesel Generator I 49. Loss of PCC
- 50. Control Room Impact MPCC MCR I - Aircraft Crash
- 51. Containment Impact
- 52. Control Room Impact
- 53. Primary Auxiliary Building Impact APC ACR APAB
- Flooding 54. Loss of Offsite Power FLLP
- 55. Loss of Offsite Power and One Switchgear Room I 56. Loss of Offsite Power and Two Switchgear Rooms FL1SG FL2SG
- 57. Loss of Offsite Power and Service Water Pumps FLSW
- Others 58. Truck Crash into Transmission Lines TCTL I
I I 1296P082885
I I TABLE 1-6.
SUMMARY
OF SSPSA PLANT EVENT TREE MODULES Event Tree Module Code Name AUX Auxiliary Systems Response of all plant suppcrt systems to all initiating events. LLj Large LOCA Early Response Early response of frontline systems to large LOCAS inside containment. I LL2 Large LOCA Long-Term Response Long term response of frontline systems to large LOCAs inside containment. APCi Aircraft Crash into Primary LLi modified for LOCAs caused by aircraft Containment - Early Response crash into containment. APC2 Aircraft Crash into Primary LL2 modified for LOCAs caused by aircraft Containment - Long-Term Response crash into containment. ML Medium LOCA Early response of frontline systems to medium LOCAs inside containment. SL Small LOCA Early response of frontline systems to small LOCAs inside containment. TRAN General Transient - Early Response Early response of frontline systems to transient events with successful scram. SLI Steam Line Break Inside Containment Early response of frontline systems to steam line breaks inside containment. SLO Steam Line Break Outside Containment Early response of frontline systems to steam line breaks outside containment. ATWS Transient without Scram Early response of frontline systems to all nonlarge LOCA events without scram. LTj, General Transient - Long-Term Long-term response of frontline systems to LT2 Response all nonlarge LOCA events. SGTR Steam Generator Tube Rupture - Early response of frontline systems to steam Early Response generator tube rupture. SGTR2 , Steam Generator Tube Rupture - Long-term response of frontline systems to SGTR3 Long-Term Response certain steam generator tube rupture events. I 1 1-26 g 1296P083085 g
I TABLE 1-7.
SUMMARY
OF TECHNICAL REVIEW RESPONSIBILITIES I Several stages of technical review will be carried out and documented for all project deliverables. At each stage of the review process, the indi-vidual members of the team have different responsibilities and review objectives as follows: Stage Review Objective Person Responsible 1 Check all calculations, computer input Analyst / Author and output, proofread documents I prepared by publications department for technical accuracy. I 2 Double check all calculations, review documentation for technical accuracy, ensure consistency of documentation Task Leader within technical area (e.g., systems), ensure that the right tools are used. 3 Spot check calculations, ensure that Technical Review Board / l acceptable PRA methods and procedures are utilized, perform independent review of all deliverables and Independent Technical Reviewer supporting calculations and documents as necessary focusing on reasonableness of results and conclusions and whether project documentation adequately I reflects what was done, recommend corrective action when appropriate. I 4 Review all deliverables, ensure project objectives are met, ensure consistency among technical areas, Project Manager I responsible for resolution of all review comments and assignment of work needed to resolve review issues. 5 Review results and conclusions of key Project Director deliverables for technical credibility and efficacy of methods employed. 6 Review all deliverables for appropri- Client ateness of assumptions regarding interpretation of plant documentation, safety analyses, and modeling of plant and site unique characteristics. 7 Perform QA audits, conduct QA training, PLG QA Manager maintain QA records. 1-27 I 1296P083085
l MAIN REPORT TECHNICAL APPENDICES VOLUMES SECTIONS SECTIONS VOLUMES
,L E. INTRODUCTION I
- 2.
SUMMARY
OF REsutis I ANO CONCLussONs VOLUME 1
- 3. SEA 8ROOE STATION PLANT ANO site PERSPECTIVE
- 4. sE ABROOK $TATION Resa A PROSAglLisfiC ResK ASSESSMENT METMODOLOGY It 46 s. SEA 0 ROOK STATION PLANT MODEL
-4 8 EVENT sEOUENCE MOOEL SUCCESS CRITER A VOLUME 5 VOLUME 2
] s oATA Analysis { 4 C EVENT SEQUENCE MOOEL DETAILS p U o 7. sysnMs ANALvsEs
{ n
--> o OETAiLEosysTEus ANai vsts VOLUME 6 o
8 DEPENDENT f AILURE ANALYSIS il 4 E sP ATIAL INTER ACTIV E ANALvsis DETAits AND DOCUMENT Af SON VOLUME 3 9 E ATERNAL EVENTS Analysts a sEisMsC MAzaRo ANo pMAGiuf v g VOLUME 7 10 MuMAN ACTION $ Analysis G sEASROOK siMut ATOR E NPERIMENT DOCUMENTATIOra ir jl 11 CORE AND CONTAINMENT RESPONSE Analysis 4 CORE ANO CONTAINMENT l M PME NOME NOLOGICAL Analysts l VOLUME 8 jia sE APRoom sf ATiON site MODE L{ VOLUME 4 W ,',o",**,'*,y'** I B3 Rism as5E M9L T AND D E COMPOsifiON it 1 FIGURE 1-1. CONTENTS OF SSPSA REPORT VOLUMES 1 1-28 I
M M M M M M M M M M M M M M M i i ( 3 r 3 r 3 l 1 PL ANT CONTAINMENT W EVENT $EQUENCE W W ECO WEEES EVEN.T sODEL SEQUENCE MODEL aAOctOGiCAt i [ (C) s,sT E Ms MODELS +-- wuMAN INT E RACTION -4 EXTERNAL EVENTS M CONTA,NMENT FAILURE + ACCiOE NT SIMULATON + Tc= OUAg M ECONOM,C IMPACT TO,0GR o y e- DE MOGR APH , + MODE LS MODELS MODEL MODE L MODEL METEOROLOGY RESPONSE TIF OTION MODE LS i OATA PLANT StTE UNIQUE ga$g UNSOUE FEATURESAND FEATURES EVACUATION PL ANS a ( Pt ANT M*E L y ( CO.aAi M Nr Om j ( saE MOOu j i l FIGURE 1-2. BLOCK DIAGPNI STRUCTURE OF SEABROOK RISK MODEL l l I
I I
.11 j. SPECIFICATION OF THE INITIATING EVENT.
I RESPONSE AND STATUS OF SUPPORT SYSTEMS 2 (SUCCESS / FAILURE COMBINATIONS). PLANT - 3 EARLY RESPONSE AND STATUS OF FRONTLINE MODEL SYSTEMS (SUCCESS / FAILURE COMBINATIONS). LONG TERM RESPONSE OF FRONTLINE SYSTEMS 4 (SUCCESS / FAILURE COMBINATIONS). 05! SPECIFICATION OF PLANT DAMAGE STATE. PHENOMENOLOGICAL RESPONSE OF (DEGRADED) A NMENT
"^ "
RESPONSE
7 RESPONSE OF CONTAINMENT STRUCTURE. SPECIFICATION OF RELEASE CATEGORY:
- b. 8 '
i.e., SEVERITY OF RADIOACTIVE RELEASES. 9 METEOROLOGICAL DISPERSION SEQUENCE. SITE _ MODEL 10 EVACUATION / EMERGENCY ACTION RESPONSE. SPECIFICATION OF PUBLIC HEALTH AND PROPERTY II DAMAGE LEVELS. o e N1 SeOUeNce ,,Nce PO,N1S I I FIGURE 1-3. STANDARD FORM 0F ACCIDENT SEQUENCES IN SSPSA RISK MODEL I I 1-30 I I
M M M M M M M M M M M M M M M INITIATING EVENTS AUXILI ARY SYSTEM EARLY SYSTEM RESPONSE LONG TERM SYSTEM RESPONSE TRANSIENT EVENTS OTHER EVENTS EVENT TREE EVENT TREES EVENT TREES LLOCA RT FSRAC ELOCA TT FCRCC E.7 L > > LLI > LL2 > TLMFW FCRSW E 1.0L PLMFW FCRAC TMLL EXFW FET1 LCV FET 3 l APC l > > APC j > APC2 > 1 MSIV FPCC A MSIV FTBLP CPEXC TMLCV l MLOCA l 5 > ML LOPF TMCR TO PLANT St TMCST STATES LOSP TMPCC l SLOCA j > > SL
- L1DC MELF 0 LOSW MPCC AUX LPCC MCR > > TRAN > LT1 E.2T ACR E.3T APAB E.4T FLLP l SLBI l > > SLI LT2 >
E.5T FL1SG E.7T FL2SG SLBO E 1.0T FLSW TMSLB : > SLO : FSRCC TCTL MSRV 1r 1r1r1r
---> SGTR 2 >
l SGTR l 0 SGTR g M SGTR 3 l V l - FIGURE 1-4. OVERVIEW 0F SSPSA EVENT SEQUENCE MODEL STRUCTURE
s o-5 d e 52 - 2 { ![}T $ 5! Ed s I I 5 lE ;. ;is ES E 5. ! 5 !! is; !*=l ip- di a: .: !! c5 :r :r _6. a* l > i. =r 53
= u :
..g "i:
tu.
':a 4
4
-5 bu :s E
U" as "n"5 E. > E. > 1;5
==
1;5
==
u u 04 RW top OP RV om ose RA RG L1 L2 CA CS it TTI EF NL NN.m*_ NN ba.I. NN emen. pen _e NNNee._I I 3 3 4 NN .$m NN __ N90 __ NN .- 5 6 I I GF 8 A GF _ G. _ G. _ .. I I ii N. GP .*= NN __ NN __ psN _i l l l Ia , _ t3-21 GP .*_ NN .=*== sees .e_ NN _ l l _. n- 30 "I" Gy _ op - 31 GF ==*= NN _*_ NP0 .*= psN _ f l l l 32 GP - GF 33 34 35 1 1 34 II Gp l 28
,,,,,, - Gs 3.
'O Nee a.m. NN a.
48 Ge ==_ Go
.m.=-.--.a.m.--===..m.as_._e. 42-50 NN
~ ~ - < - N.,_N_,,,,_L,,,,.______.__.__._.___....
i GP a-o_ peni .me.m i NN e-o esse i . NN GP - Gp- 60 I I I
- es N,,
l >-- i
~~- _.__________.____..,,
.-_.____.___.__n0 l i ee NN NN - NN -
i i en G,- e N _.l N l I l . ___________.___..... G,__NN NN N_
-_ e m. 2 100 a e___ a-tot G F a=.m. NN .me=. NN .e=. NN a.
103 i . __________.____........ a - .e_ m amm_ an. emm. e-.a._ ==. _ em. .m.' II
- I 30
=.Gpa-.mNN I 's*
Go - NN - \C ___-_-_.___._.,n-.0 e--. .ns. ===_ _m _ I 81
- I 8I a _e amo e- . ex . es Nas au-o p,N a .-_ _ .--. . % 153 I
3 G. LEGEND G7
- GuaRANTEGOftaLURE IC * 'NCONSEQve nstsaL NN
- NOT NGCS ESARY N0ft A en,Nfso Larven oggicNafos aveting toes sawg avgNt f RG E STRUCTURG OUT Dippgegner ENO ST ATER FIGURE l-5. GENERALIZED TRANSIENT FARLY RESPONSE EVENT TREE (TRAN) 1-32 I
m M M M M M M M M M M M M M 12SV DC 125V DC CABINET CABINET 112A 1128 I St SIGNAL SI SIGNAL i TRAIN A TRAIN B I EMERGENCY FEEDWATER _ PUMP FLOW TO STEAM TDP-37A i GENERATOR A (EFWAl
, (TDP) (v125) v125 EMERGENCY FEEDWATER _
FLOW TO STEAM GENERATORB Y CST - PIPING - (V126) V126 (EFWB) 2/4 E$ EMERGENCY ! (CST) (EFW/ PIPE) FEEDWATER - FLOW TO STEAM (V127) V127 GENERATOR C (MDP) (E FWCl ,I PUMP l _ EMERGENCY l MDP-378 FEEDWATER -
; a FLOW TO STEAM
- GENERATOR D 4,160V AC ' # '
l SI SIGNAL i BUS E6 TRAIN B LEGEND: O COMPONENT BLOCK O INTERMEDIATE GATE O SUPPORT SYSTEM FIGURE 1-6. BLOCK DIAGRAM SHOWING SUPPORT SYSTEMS FOR EMERGENCY FEEDWATER SYSTEM I l l
I l I I I I Il Ru AND Te l OXIOATION RELEASE l m CHARACTERISTIC 3 l N 3 3 RELEASE )
;;; q e CATEGORY S2 ,
)$ .
/
! SMALL l /
$ LEAK k/
n / 3 mm
@g LARGE
/ 5 g$ LEAK / /
bR 8 u >< EARLY LATE BASEMAT OVERPRESSURE OVERPRESSURE ME LT-THROUGH
- CONTAINMENT FAILURE MODE I
FIGURE 1-7. ILLUSTRATION OF RELEASE CATEGORY DEFINITION I I I 1-34
I 2. I RESULTS AND CONCLUSIONS The purpose of this section is to present the overall technical results and conclusions of this study. The remaining sections of the report and the list of references describe the detailed analyses, models, data, and assumptions used to develop these results and conclusions. These results include an update of the SSPSA results (Reference 2-1), which were published in December 1983, to account for what of importance has been learned since then about plant and systems modeling and source terms analysis. More importantly, the results provide a technical evaluation of the benefits of specific risk management actions, such as evacuation, sheltering, and emergency operating procedure enhancement. In Section 2.1, a rebaseline of the SSPSA results is provided to account for enhancements that have been made since the SSPSA was completed in the plant and systems analysis area, as well as in the definition of radionuclide source terms for accident consequence analysis. To help provide a neutral basis to begin the evaluation of emergency plan options in Section 2.2, the rebaseline risk assessment of Section 2.1 is performed for the case of no immediate protective actions; i.e., no evacuation and no sheltering. Then in Section 2.2, the risk impact of several emergency planning options are compared, including cases of no evacuation, evacuation for various distances without sheltering, and an evacuation and sheltering combination. The risk impact is quantified in several different ways to maximize the insight necessary for prudent risk management. These include the type of risk curves traditionally calculated in PRAs; i.e., the frequency of exceedance of consequence curves, curves that portray the spatial distribution of risk, risk as a functior, of evacuation distance, and the type of curves used in NUREG-0396 (Reference 2-2) that provide information related to specific radiation dose levels. The impact of emergency planning options on NRC safety goal calculations is also quantified. The remaining risk management actions addressed in this study are specific enhancements and additions to the Seabrook Station emergency operating procedures. In the SSPSA, the dominant contributors to early and latent health risks were found to be interfacing systems LOCA and I station blackout sequences, respectively. Therefore, specific risk management actions were investigated to control and reduce the risk contributions of these sequences, as appropriate. To provide meaningful recommendations regarding the procedures and to incorporate new insights into the modeling of these sequences, it was decided to develop and evaluate a new event sequence model for interfacing systems LOCA sequences for use in procedure evaluation. Details of the new model for interfacing systems LOCA are presented in Section 3.1, while the results regarding the risk impact of the procedures are discussed in Section 2.4 The second area identified for emergency procedure evaluation was the recovery of cor.tainment heat removal systems during core melt sequences 2-1 1321P120585 (
B involving a station blackout. In the SSPSA, recovery of electric power, g core cooling systems, and many other recovery actions were considered for 3 these and other sequences. However, no consideration was given to operator recovery of containment heat removal systems after core damage . was predicted to occur. Such consideration is warranted in view of new insights provided by the SSPSA about the strength of the containment and the implications of this strength in estimating the time needed to challenge the pressure capacity of the structure. Containment recovery actions that take advantage of the several days available for recovery in these high risk sequences are identified and evaluated in Section 3.2. The results regarding risk impact are provided in Section 2.4. The overall conclusions of this study are provided in Section 2.5. 2.1 REBASELINING SSPSA RESULTS FOR NO IMMEDIATE PROTECTIVE ACTIONS The purpose of this section is to present an update of the SSPSA results for the case of no-emergency-plan protective actions; i.e., no immediate evacuation and no sheltering. In these calculations, a 24-hour period for radiological exposure for nonevacuees is assumed for consistency with normal practice in consequence modeling for emergency planning purposes; e.g., NUREG-0396 used thesc same assumptions. In comparison with the SSPSA results, which assumed a 10-mile evacuation distance, these results reflect the inherent risk characteristics of the plant and site without the risk reduction afforded by emergency protective actions. As such, this provides a neutral starting point for evaluating the risk reduction benefits of various protective action strategies. A comparison of risk curves for acute fatalities is presented in Figures 2-1 and 2-2 based on mean and median values, respectively. Included in these comparisons are the SSPSA (10-mile evacuation distance), WASH-1400 results for PWRs (25-mile evacuation distance) (Reference 2-3), and updated results for Seabrook Station (no evacuation). Each curve in Figure 2-1 represents the mean value of the family of curves that characterize the uncertainties in each study. Rather than plot the whole family of curves for each study, the project team chose to plot the mean in this figure as a single curve to characterize both the central tendency and the spread of the uncertainty distributions. Although the use of a single curve for each study in Figure 2-1 does not fully characterize the range of uncertainty associated with the risk estimates, it is sufficient for making judgments about the relative risk significance of the various alternatives considered because the uncertainties are correlated for the different evacuation strategies. The comparison in Figure 2-2 based on median results, on the other hand, might be regarded as a "best estimate" comparison without explicit consideration of the effects of uncertainty. As shown in Figures 2-1 and 2-2, the updated risk results for Seabrook Station for the no-immediate-protection-action option provided in this study are substantially lower than results in both the WASH-1400 and the 3SPSA. The comparison with WASH-1400 is made for an important reason. The current 10-mile emergency planning zone is based, in part, on a characterization of risk using WASH-1400 results for sequence definition, quantification, source terms, and consequence assessment methodology. It 1 l 2-2 1321P120585 J
has been argued that the WASH-1400 results have represented a de facto statement of risk level acceptability for a long time. The characterization of risk levels provided by its results and methodology in NUREG-0396 appears to have been a major justification for the decision to set the EPZ boundary at 10 miles for all U.S. LWR plants. Two methods were used to calculate mean risk curves for WASH-1400 in Figure 2-1. 011y median curves and ranges of uncertainty about these curves were published in the report. Method 1 is the most faithful representation of the information provided in the report about ranges of uncertainty and makes an assumption about the shape of the underlying distribution; i.e., that the frequency of exceedance is lognormally distributed at each given level of consequence. Method 2 assumes broader distributions to account for a criticism that WASH-1400 had understated I ranges of uncertainty in the report. The broader distribution of Method 2, which tended to elevate the mean risk curve, assumed the same ratio of mean-to-median values of the early fatality risk curve that was calculated in the SSPSA. For subsequent comparisons made in this study, Method 1 is used because, as a limit of acceptability, it sets the most stringent limit for evaluating risk levels that can be derived from WASH-1400. I Figure 2-2 shows even greater differences among the various sets of results when the comparison is made to median values. This is because the mean values are influenced by the range of uncertainty, which was calculated to be much greater in this study in relation to that in WASH-1400. In comparison with the SSPSA results, which assumed a 10-mile evacuation distance, the updated results in Figures 2-1 and 2-2 reflect several important changes that explain the substantial reduction in risk despite the assumption of no immediate protective action. In order of importance, the most significant changes are: (1) the significantly reduced source terms and (2) a more realistic treatment of interfacing I systems LOCA sequences. An in-depth discussion of source terms is provided in Section 4. Of the major differences in the updateo treatment of interfacing systems LOCA sequences, which is discussed in Section 3.1, the most important with respect to risk are: e The role of the RHR system relief valves in reducing the frequency of high pressure challenges to the low pressure RHR piping. e The pressure capacity of the RHR low pressure piping. e The potential for operator actions and plant hardware to prevent core damage. e The effects of RHR pump vault flooding in reducing the source term. The effects of several of these factors can be seen by examination of Table 2-1. While the frequency of valve ruptures increased, the frequency of a leak in the RHR pressure boundary decreased because of the role of the relief valves inside the containment. The ability to isolate 2-3 1321P120585
some of the valve ruptures and to prevent core damage reflected in the updated analysis, resulted in more than an order-of-magnitude reduction in the frequency of the so-called "V-sequence," as it was termed in WASH-1400 and in the SSPSA. In addition to this frequency reduction, the reduction in the source terms for this and other sequences explains the very low risk levels assessed in Figures 2-1 and 2-2 for Seabrook Station, despite the assumption of no immediate protective action. Significant advances have occurred since the accident at Three Mile Island in understanding the behavior of radionuclides under accident conditions that involve core damage. An anticipation of potential reductions in accident source terms existed when the SSPSA study was performed, and the uncertainty analyses for the SSPSA source terms accounted for potentially reduced source terms in the form of uncertainty distributions for the release fractions and release time parameters. Since the SSPSA was published, IDCOR has issued its assessment of accident source terms, culminating in the definition of new accident source terms for four lead plants. Of these, the Zion station most closely compares to the Seabrook Station. The NRC has very recently published the first draft of NUREG-0956, which contains new accident source terms for the Surry station. The accident source terms used in this study were derived either from the g appropriate IDCOR source terms for the Zion station, corrected for the I major differences in the containment pressure capacity, or from the SSPSA source terms, accounting for further advances in the source term state of the art since the work for the SSPSA was performed. The SSPSA source terms, in turn, were based on the RSS methodology (e.g., CORRAL code), and were corrected for then new insights on source terms resulting from NRC and IDCOR research efforts. Table 2-2 compares the source terms utilized in this study with those published by IDCOR, with those in NUREG-0956, or those in WASH-1400. Si x release categories were defined representing the five important containment failure modes and an intact containment. For the first four l release categories involving failure, two source terms were defined. A set of best estimate source terms (designated by a "B") was based on the IDCOR source terms, when available, and a set of conservative source terms (designated by a "C") was derived from the SSPSA. A comparison of the Seabrook Station design to the Zion station design identified one significant design difference with respect to accident source terms. The building in which the RHR and ECCS equipment is located provides for significantly different transport pathways for radionuclides for the so-called interfacing systems LOCA (or V-sequence). Therefore, a Seabrook-specific analysis of the V-sequence accident source terms was performed using the IDCOR analysis methods. These source terms are shown as release categories S7B and S7C in Table 2-2. The accident source terms for an intact containment, SS, were taken directly from the SSPSA. In the analysis of accident risks, both the best estimate and the conservative source terms were used to define a mean risk profile. Because of the shift in importance of release I 2-4 1321P120585
g categories in comparison with the SSPSA results, a greater number of g categories were analyzed using a more realistic multipuff source term in the CRACIT model. This also contributed to a more realistic assessment of consequences. From a comparison of the applicable NUREG-0956 source terms and the conservative source terms defined for this study, two enveloping source I terms were developed for release categories S1 and S6. These are also listed in Table 2-2 and are designated as S1E and S6E, respectively. The updated results for latent cancer fatality risk in comparison with the SSPSA and WASH-1400 results are presented in Figure 2-3. A similar level of latent cancer risk is characterized by each of these curves. In the high-frequency, low-consequence region, the small variations among these curves reflect differences in core melt frequency as well as differences in the method used to calculate the mean WASH-1400 risk curves. The mean frequency of core melt in the WASH-1400 SSPSA,angtheupdatedanalysisare9.9x10-5,2.3x10-4(PWR),the ,and 2.7 x 10 , respectively. In the low frequency-high consequence region, the updated results for Seabrook Station, with no evacuation, are significantly within the remaining curves although to a lesser extent I than in the case of early fatality risk. Significant advancements have been made since WASH-1400 in the I development and application of PRA methodology. Many of these advancements were incorporated into the SSPSA and into the updated results for Seabrook Station. Therefore, it is doubtful whether the above differences in core melt frequency reflect unfavorably on Seabrook Station in relation to the single PWR plant analyzed 13 years previously in WASH-1400--the Surry plant. The use of a more complete analysis of various types of dependent events, such as external events, spatial interactions, and common cause events, and the benefit of a larger data base in the case of Seabrook probably explain the difference. In the opinion of the authors of this report and the SSPSA, it is also doubtful I whether any U.S. plant without operating experience, when subjected to the same methods and data and scope of evaluation, would exhibit core melt frequencies substantially lower than those indicated above for I Seabrook Station without the benefit of post-PRA modifications to reduce them. The above comparisons indicate a very low level of risk at Seabrook l Station in relation to a de facto level of risk acceptability represented in the WASH-1400 results. A second and more recent point of reference as a limit of risk acceptability is the current NRC policy on safety goals. In the SSPSA, it was shown that both the individual and societal risk
'I safety goals were met by large margins, using what now are viewed as conservative source terms and an assumed 10-mile radial evacuation. As shown in Figure 2-4, the updated results for Seabrook Station, assuming no immediate protective actions, yield individual risks that are more than two orders of magnitude less than the safety goal and, by definition of the safety goal, more than five orders of magnitude less than the nonnuclear sources of risk to which the public is generally exposed. As expected, the societal risk goal is also still met with margins
- comparable to those found in the SSPSA.
2-5 1321P120585 l
To provide another perspective on the updated results and a basis for E evaluating the risk impact of evacuation and sheltering in the next 5 section, it is instructive to examine how the no-evacuation risk is spatially distributed about the plant. This spatial distribution of no-evacuation risk (which should not be confused with risk versus evacuation distance) reflects the population distribution around the site as well as the reduction in radiation doses at progressively greater distances from the plant due to the processes of radioactive material transport. To evaluate spatial distribution, risk is defined here in terms of the expected frequency of health effects at different radial distances from the site. The spatial distributions of acute fatality risk.for the updated Seabrook Station results for no immediate protective actions (normalized) are plotted in Figure 2-5. This figure shows that, of the very low risk levels calculated, most of it is located quite close to the site for acute fatalities and significantly farther from the site for latent cancer risk. For example, less than 5% of the no-evacuation acute fatality risk is located outside 2 miles, while about 70% of the latent fatality risk is located outside 10 miles. 2.2 EVALUATION OF EMERGENCY PLANNING OPTIONS To evaluate emergency planning options, five sets of analyses were performed based on the following immediate protective action strategies that are listed in approximate ascending order of potential risk reduction. (Nonevacuees were assumed to receive a 24-hour dose.) e No Evacuation e 1-Mile Evacuation e 2-Mile Evacuation e 2-Mile Evacuation and Sheltering from 2 to 10 Miles e 10-Mile Evacuation Insofar as the site model calculations are concerned, the number of calculations that were performed in this study through the use of the CRACIT computer code (see Section 12 of Rcference 2-1) is approximately equivalent to five level 3 PRAs. The results for acute fatality risk in frequency of exceedance format are shown in Figure 2-6. Although the l curves show that evacuation for a distance of up to 2 miles has a visible l impact in lowering the risk, the amount of additional risk reduction E l associated with protective actions beyond 2 miles is seen to be very small. It is only because these results are plotted on a log-log scale that any perceptible difference can be noted. The results for latent cancer risk for all five cases of protective actions were found to be identical to the updated no-evacuation cases plotted in Figure 2-3. The reason for the insensitivity of latent health risk to protective action strategy is that evacuation and sheltering l reduce the radiation doses associated with the radioactive plume, while l the latent health risk is dominated by long-term exposure to very low doses after the plume has dissipated. Another perspective on the absolute and relative risk levels of alternative emergency plan options at Seabrook Station is provided in I I 2-6 l 1321P120585 l L
I I Figure 2-7. This figure measured by the frequency,is a ploteffects of health of the per risk year, of acute as afatalities functionas evacuation distance. The effects of sheltering for a distance of 2 to 10 of miles is also indicated. An appreciation of the extremely low absolute level of risk exhibited in the curve, even at the point of no evacuation, can be seen by comparison with the NRC safety goal applied to the population within 1 mile of the Seabrook Station site boundary. This goal sets an individual risk level of acceptability of one-tenth of 1% of the risk of accidental fatality due to nonnuclear hazards to which the public is generally exposed. This very low risk goal is met with a margin of nearly two orders of magnitude even with no evacuatio.i. A I summary of the safety goal results calculated in this study is provided in Table 2-3. l Additional perspective on the risk significance of alternative protective action strategies is provided in Figure 2-8. Tnis figure is similar to Figure 2-7 except that it employs a linear risk scale; it measures the I percentage of the nonevacuation risk avoided instead of the absolute risk level. This figure graphically illustrates that, of the small amount of risk reduction that can be achieved by the protective actions considered I in absolute terms, most of this reduction is realized for a close-in evacuation. Between 70% and 95% of the acute risk ~ reduction benefit to be realized by evacuation is realized within 1 mile and 2 miles of the site, respectively. Additionally, of the very small additional risk reduction that can be expected in moving the evacuation distance from 2 miles to, say,10 miles, this incremental reduction is nearly matched by simply sheltering in this area around the plant. A different viewpoint from which to evaluate protective action strategies is afforded by the type of risk estimates performed in NUREG-0396. This type of estimate was used as a basis for establishing a constant 10-mile emergency planning zone and consists of a conditional frequency of exceedance of various dose levels as a function of distance from the site, given a core melt accident. Unlike the previous risk estimates I provided in this section, this type of estimate does not include the accident frequency in an absolute sense, only in a relative sense, and relates only indirectly to health effects, as explained below. A I comparison of the NUREG-0396 and the updated Seabrook Station curves using this risk measure is provided in Figure 2-9 for the 200-rem and 50-rem curves and in Figure 2-10 for the 1-rem and 5-rem curves, respectively. All curves in both of these figures assume no immediate I protective action. Although the NUREG-0396 results were only quoted for median values, both means and medians are provided for Seabrook Station. The distributions for the 1-rem and the 5-rem whole-body doses are indicative of the spatial extent of low-level exposures and represent the EPA PAG levels. The 50-rem dose level is approximately the threshold for injuries. The 200-rem whole-body dose distribution is approximately the threshold for acute fatalities. At a distance of approximately 10 miles, the 200-rem distribution from NUREG-0396 begins to fall off sharply. This is the so-called " knee" in the 200-rem curve. It is widely believed that this knee in the 200-rem currve is part of the justification for establishing a 10-mile emergency planning zone. 2-7 I 1321P121885
1 For the relatively high ooses in Figure 2-9, there are very large differences between the updated Seabrook Station results and the NUREG-0396 results. The latter used the WASH-1400 plant and systems analysis results for an average LWR plant, WASH-1400 source terms, and the Surry site. The 200-rem curve for Seabrook Station is seen to fall off rapidly within 2 miles, while the corresponding NUREG-0396 curve at the same exceedance frequency indicates 200-rem levels for a distance of more than 10 miles. The corresponding plots for 1 rem and 5 rem also show a much lower risk for Seabrook Station than characterized generically for LWRs in NUREG-0396, especially for distances beyond several miles. I A limitation of the type of calculation performed in Figures 2-9 and 2-10 I in the evaluation of emergency planning options is that the timing of the releases is not taken into account in the calculation of radiation doses. In the case of the results for Seabrook Station, for example, the curves for low doses,1 rem and 5 rem, are dominated by releases (S3 release category) that occur over time intervals starting several days after the warning is given to emergency response organizations. Even ad hoc protective actions should be effective in the prevention of such doses without the assumption of an existing evacuation plan. To see the effect of this factor, the curves in Figure 10 were recalculated for the accidents with releases within 24 hours of the warning time. The results for 1 rem and 5 rem are compared with NUREG-0396 curves in Figure 2-11. The Seabrook Station curves are reduced by a factor of 10 in relation to E Figure 2-10. The NUREG-0396 results for PWRs were based on releases and 5 24-hour exposure times, starting a short time after the initiating event. Doses close-in to the site for the updated Seabrook Station results, which do not include doses that start more than 24 hours from the warning time, are much lower than those for the NUREG-0396 doses that are 30 to 40 miles from the site, regardless of whether the median or mean values are used in Seabrook Station. A further perspective on the risk significance of protective actions can be obtained by examining the types of releases used to characterize the E source terms whose release fractions and timing dictate the amount of a risk reduction to be realized by various protective action strategies. A gross, but revealing categorization of these release types is provided in Table 2-4. Compared in this table are the core melt frequencies and the distribution of core melt frequency into three major release types. The first type, gross early containment failure, is the most important for evaluating protective actions because, as shown in this study, protective acticns primarily impact the early health risk. PRA studies have consistently shown that this release type dominates early health risk. The second type, characterized by such failure modes as long-term, or g gradual, overpressurization; basemat melt-through; or containment leakage 3 is of primary interest in the determination of latent health risk. These modes have generally been found not to be significant contributors to early health risk. Radiation dosas from accidents in this category are either very low in the absence of protective actions (evacuation), or occur over such long periods of time (i.e., several days) that even ad hoc protective actions should be successful in the prevention of doses. As shown in this study, latent health risk is not appreciably affected by the kind of protective actions considered in this study. The I 1321P121885
I third release type represents an intact containment and releases that do not produce significant latent health effects and no early health effects whatsoever. In WASH-1400, 34% of the releases were assigned to the first I category and, hence, contributed to early health risk. This fraction was affected by more conservative assumptions than were made in the SSPSA about the strength of large, dry containments, yet it heavily influeaced the characterization of risk that was used in NUREG-0396 to determine the I EPZ distance. Another important aspect of the WASH-1400 results was that no credit was given for potential long-term containment integrity following a core melt accident. By contrast, the SSPSA results show that only about 1% of the core damage sequences are in the first category and of interest in emergency planning, with 26% in the intact containment category. Radiation doses I from the intact category have been shown in Reference 2-4 to be well within protective-action guideline levels at close-in distances to the site. As will be described more fully in Section 2.4 below, operator I actions have been identified in this study to increase the percentage of accident sequences in this benign category to about 40%. The releases in the second category are either small or occur over such long time I intervals that ad hoc protective actions can be effective in mitigating exposures. A very important aspect of the updated results is that only one-tenth of 1% of core melts are now assessed to be in the first category. Hence, 99.9% of the core melts are of little or no concern in I planning for offsite protective actions. As a final comparison of the risks of acute fatalities, Table 2-5 shows I point estimate risk values from a range of sources; namely, WASH-1400, NUREG-0956, the SSPSA, and from this study. The WASH-1400 risk values were obtained by graphical integration of the corresponding curves in Figure 2-1. These point values were obtained as the mean value of a I complementary cumulative distribution function representing either the median or the mean frequency of exceedance of a consequence level. Columns 5 and 6 represent ccmparisons of the mean value of the median I CCDFs, while the last two columns represent comparisons of the mean value of the mean CCDFs. In each category, the second column normalizes the risk to the WASH-1400 value. For riormalization of the mean CCDF values, I method 1 for determining the mean CCDF for WASH-1400 was used as the basis for comparison, consistent with the other comparisons made in this section. However, for comparison of the SSPSA mean value to the WASH-1400 mean value, comparison should be made with the WASH-1400 I method 2 value since it incorporates an assessment of the uncertainties that are consistent with the SSPSA. The no-evacuation risk for Seabrook constitutes only 6% of the WASH-1400 risk, and it is roughly equal to the NUREG-0956 risk for Surry, which used the BMI-2104 suite of codes and a reevaluation of the containment performance. 2.3 SENSITIVITY ANALYSES OF KEY ASSUMPTIONS , The estimates of risk used to evaluate emergency planning options in the previous section exhibit substantial uncertainties. These uncertainties l 2-9 1321P121885
preclude the attainment of high accuracy in the estimates, but should permit the order of magnitude-type comparisons that are made in this study. The technical approach to the treatment of uncertainty in this study, I which is fully documented in Section 4 of the SSPSA (Reference 2-1), is 5 to include a range of assumptions, models, and data and other sources of uncertainty and to quantify the subjective assessments of the study team g probabilistically. When quoting a single number for risk as the expected g frequency of health effects, the mean value of the underlying uncertainty distributions that was formally propagated through the risk model is quoted. These point values of risk are influenced not only by the set of assumptions included, but also on the probabilities or weights that were assigned to these assumption sets, based on the professional judgment of the study team. For this reason, a number of sensitivity analyses were E performed to evaluate the change in individual risk resulting from 5 differences in assumptions and in the numerical weights that were assigned. In these sensitivity calculations, the following factors were g evaluated and the results are presented in Table 2-6: E e Probability weights assigned to source term and site model assumption sets. e Contribution of short-term and long-term releases to risk. o Numerical values of the conservative source term parameters for risk significant accident categories. The first sensitivity evaluation was performed to determine the margin in the safety goal comparison when the conservative upper-bound source term (C) is probabilistically weighted at 1.0 and the conservative high (H) consequence assumptions are probabilistically weighted at 1.0. The second sensitivity evaluation was performed to determine the potential conservatism associated with calculating the' contribution of g risk due to releases that occur with a warning time longer than 24 hours. 3 The third sensitivity evaluation was performed to investigate a potential lower-bound risk associated with the best estimate release categories and the best estimate consequence assumptions. The fourth sensitivity evaluation was performed to determine the margin in the safety goal when an enveloping source term is used. The I-enveloping source terms included the NUREG-0956 release fractions when they were higher than the Seabrook Station conservative source term. In some cases, there was not a corresponding NUREG-0956 source term to match the Seabrook Station accident category. This sensitivity case was performed even though there is sufficient evidence to support differences in the source terms, as described in Section 4. In addition, a number of technical issues associated with source terms are discussed in Section 4. The uncertainties in the Seabrook Station source terms are believed to adequately represent the uncertainties associated with these issues. I 1321P121685
I Key results of the sensitivity evaluation are presented in Table 2-6. I. The conclusion that the updated results for no evacuation yield risk levels below the safety goal is shown by Table 2-6 to be insensitive to variations in the key assumptions, within a reasonable range. Even under I the worst combination of assumptions, a margin factor of 5 is maintained for the safety goal. 2.4 EVALUATION OF OTHER RISK MANAGEMENT ACTIONS The purpose of this section is to summarize tbs results of the evaluation of specific additions and enhancements to the emergency operating procedures for their effectiveness as risk management actions. The focus of this evaluation was on two categories of accident sequences identified in the SSPSA as the ranking contributors to early and latent health I risk. In the SSPSA, the first category, interfacing systems LOCA, was found to represent about 76% of the frequency of releases with early containment failure or bypass and therefore was the principal contributor to early health risk. The second category, loss-of-offsite-power-induced I station blackout sequences, was found to contribute about 29% of the frequency of core melt and about 40% of the releases involving gradual containment overpressurization. These releases were the dominant I contributors to latent health risk. Therefore, these categories of sequences represented logical candidates for the investigation of risk management actions. There are existing emergency operating procedures for loss of coolant events outside the primary containment in addition to those for such I events inside the containment. Under the assumptions of the so-called V-sequence analysis of the SSPSA, these procedures may appear on the surface to be ineffective since the event was assumed to result in a nonisolable LOCA outside the containment, core melt, and containment bypass of the release. To provide a more realistic basis for evaluating and refining these procedures and to take into account new insights about the plant behavior following such sequences, a new plant event sequence I model for this class of accidents was developed and quantified, as described more fully in Section 3.1. Based on this new model of interfacing system LOCA sequences, a high I potential was identified for operator actions to prevent a core melt and to mitigate the consequences of those events. Because the procedures had been written prior to this analysis, however, the potential for l misdiagnosis was found to be high. Fortunately, specific suggestions
= were made and are being accepted by the plant operators to modify these procedures to reduce the potential for misdiagnosis of a V-sequence as a I LOCA inside the containment. The ability to make these suggestions was made possible on the basis of insights obtained from the more realistic assessment incorporated into the accident sequence model of the plant l response to these sequences.
l l The motivation to modify the procedures was an assessment of the risk contribution by these sequences with and without the precedure modifications in place. It is noted that operator actions play a key role in reducing the core melt f requency from 5.9 x 10-7 aer year, the I 1321P121685
frequency of a LOCA outside the containment, to 3.4 x 10-8 per year. Without the procedure enhancements, the potential for misdiagnosis was assessed to be high and the core melt frequency then approached the 5.9 x 10-7 per year value. The impact on risk with and without the procedure enhancements was calculated for each emergency planning option, and the results summarized in Table 2-7. As seen in this table, the acute fatality risk is reduced by factors ranging from about 4 to about 10, depending on the evacuation scenario assumed. This was an g excellent example of cost-effective risk management in the sense that the 3 potential for a large reduction in risk was identified without requiring significant costs to effect the change. The second category of risk management actions evaluated in this study I was the recovery of the containment heat removal function during station blackout-induced core damage sequences. In the SSPSA, actions to recover E electric power and core cooling for these sequences were considered for E the period up to the time estimated for core melt, but subsequent actions to recover containment heat removal were not included. In the containment failure analysis for these sequences, time periods ranging a day to a week were estimated to be available before the containment would fail due to overpressure. The analysis presented in Section 3.2 explores several options to restore containment heat removal, including extending the chances for recovering I power from the normal 345-kV grid, keeping in place and enhancing the 3 capabilities of the 34.5-kV grid now being used for plant construction, 5 and the use of portable electric generators. Assuming the specific recommendations described in Section 3.2 are followed, it was determined that a major fraction of the station blackout sequences identified in the SSPSA that lead to core melt can be expected to have the containment heat removal function restored before containment overpressurization. The containment recovery analysis was applied to 23 such sequences from the SSPSA where combined frequency was 4.7 x 10-5 per reactor-year. The conditional frequency of successful containment recovery for these sequences was found to be about .93. This would bring about a reduction g of about 30% in the expected frequency of health effects in the latent 5 health risk calculated in this study. It should be noted that the above evaluation only covers the specific risk management actions identified and evaluated in this study. Because the risk was found in the SSPSA and in this update to be caused by a relatively large number of different contributors, the number of risk management actions that could be defined is potentially large. Therefore, the amount of risk reduction achieved by any particular action may not be very large on a percentage basis. There are many other potential actions that were not evaluated in this study and whose risk reduction potential is unknown. Several specific actions have already been identified as having significant potential for effecting further risk reduction either through concrete actions to reuuce risk or by way of enhanced analysis of plant behavior in degraded modes. These include additional recovery actions and enhanced procedures to reduce core melt frequency, capability for reflex cooling via steam generators for interfacing system and RCP seal LOCA scenarios, actions to wet the I 1321P121685
I containment and reactor cavities during postulated " dry scenarios," and a I wide spectrum of recovery actions during seismically induced accident scenarios. The evaluation of the risk impact of these actions and analyses shoald be considered in any subsequent updates of the risk I assessment for Seabrook Station. This is one reason why risk management must be a continual process in order to be able to realize its full potential.
2.5 CONCLUSION
S The conclusions of this study are summarized as follows: I e The updated risk assessment provided in this study shows that the acute health risk is very low in absolute terms as well as in I relation tc any known standards of acceptability or safety goals. Even under the assumption of no immediate protective actions, the acute health risk estimated for Seabrook Station is:
- More than an order of magnitude less than that estimated in the SSPSA, which assumed a 10-mile evacuation distance.
I - More than an order of magnitude less than that estimated in WASH-1400, which assumed a 25-mile evacuation distance. I - About 2 orders of magnitude less than the NRC safety goal for individual risk within 1 mile of the site. I Substantially less than the level of risk achieved with an EPZ distance of 10 miles as perceived in NUP.EG-0396.
- Spatially located close to the plant site, with over 954 located within 2 miles of the containment.
e The above conclusions are based, to a large extent, on significant advancements and new insights about the: Nature and magnitude of radioactive release source terms. Strength of the Seabrook Station large, dry containnent and implications regarding timing and magnitude of radioactive releases. Progression of sequences involving loss of coolant events outside
- the containment.
e The latent cancer risk estimated in this study for Seabrook Station is: i Comparable to that estimated in the SSPSA and in WASH-1400. More than a factor of 250 less than the NRC safety goal for societal risk within 50 miles of the site. 1 l 2-13 1321P121685
l
- Not sensitive to assumptions regarding evacuation because of the role of long-term exposures to low dose levels in the models used to estimate latent health effects.
e Because the acute health risk levels are already very low under the assumption of no evacuation, the potential for risk reduction due to evacuation or sheltering to various distances from the site is also very low in absolute terms. e Evacuation has a negligible effect in reducing latent health risk as calculated in this study. e Of the small amount of risk reduction to be achieved by evacuation, a very large portion of this reduction is achieved with close-in I evacuation. More than 70% of the risk benefits from evacuation are E realized with a 1-mile evacuation distance. More than 95% of the 5 risk benefits from evacuation are realized with a 2-mile evacuation distance. e There is no measurable difference in risk reduction between evacuation to 10 miles and the combination of evacuation to 2 miles and sheltering for a distance of to 10 miles, e Using the same rational basis as used in NUREG-0396 to select a 10 -mile EPZ for all U.S. sites, the results of this study support an 3 EPZ of less than 1 mile. E e Refinements to the existing procedures for loss of coolant events outside the containment have been identified and found to make a significant decrease in acute health risk and no appreciable cost of implementation. e Procedures and operator actions to restore containment heat removal during station-blackout-induced core melt sequences have been identified and evaluated in this study. Incoporation of these E procedures and modifications was estimated to make a small reduction 3 in latent cancer risk. e It would have been very difficult to support any or all of the above conclusions to the extent they are currently supported without the foundations that were laid in the full-scope, Level 3 plant-specific and site-specific PRA that was completed for Seabrook Station. e Owing to large margins between calculated risk levels and levels of acceptability, the above conclusions are generally insensitive to key g uncertainties in the risk estimates. 3
2.6 REFERENCES
2-1. Pickard, Lowe and Garrick, Inc., "Seabrook Station Probabilistic I Safety Assessment," prepared for Public Service Company of New Hampshire and Yankee Atomic Electric Company, PLG-0300, December 1983. I 1321P121685
I 2-2. Collins, H. E., et al., " Planning Basis for the Development of I State and Local Government Radiological Emergency Response Plans in Support of Light Water Nuclear Power Plants," prepared for the U.S. Nuclear Regulatory Commission, NUREG-0396, December 1978. 2-3. U.S. Nuclear Regulatory Commission, " Reactor Safety Study" An Assessment of Accident Risks in U.S. Commercial Nuclear Power Plants," WASH-1400, NUREG-75/014, October 1975. 2-4. Yankee Atomic Electric Company, Report YAEC-1502, December 1985. I I I I I I I I I I I I I 2-15 I 1321P121685
I I I I I TABLE 2-1. UPDATE OF INTERFACING SYSTEMS LOCA KEY RESULTS I Frequency (per reactor-year) Event Updated SSPSA Results Valve Ruptures, LOCA 1.8 x 10-6 6.5 x 10-6 Valve Ruptures, LOCA, 1.8 x 10-6 5.9 x 10-7 Containment Bypass Valve Ruptures, LOCA, 1.8 x 10-6 3.4 x 10-8 E Containment Bypass, 5 Melt I I I I I I I 1322P120285 2-16
I TABLE 2-2. COMPARISON OF RELEASE CATEGORIES Release Time (hours) nergy e ease actions Release 6 Source Category 10 Calories Start Duration Warning per Second Xe I Cs Te Sr Ru La EARLY CONTAINMENT FAILURE
- This Study S1 B 2 12 1 < 10 1 .052 .052 .01 3 .006 .005 2.-4 I NUREG-0956 This Study NUREG-0956 NUREG-0956 V-Pool SIC V-No Pool TMLB'O 2.5 1
1 1 14 10 2 2
.8
.5
.8
.5
< 10
< 10
< 10
< 10 1
1 1
.85
.08
.135
.4
.07
.08
.135 4
.058
.025
.032
.12
.055
.0022 ' 1.-4
.01 6
.011
.01 7.-4 4.-4 7.-5
.0056 6.-4
.0013 2.-4 WASH-1400 PWR-2 2.5 .5 1 12 .9 .7 .5 .3 .06 .02 .004 This Study S1E** 1 2 0.5 < 10 1 .4 .4 .12 .01 6 .006 6.-4 EARLY INCREASE CONTAINMENT LEAXAGE This Study S2B 13 76 5 < 10 1 .013 . 01 3 .004 .002 9.-4 1.-4 This Study S2C 5 51 .6 < 10 1 .025 .025 .008 .003 .0018 3.-4 LATE OVERPRESSURE CONTAINMENT FAILURE This Study S3B 89 0 74 < 10 1 .001 .001 .002 1. 5 1.-5 1.-5 This Study S3C 54 0 42 < 10 1 .002 .002 .01 2.-4 2.-4 3.-5 10COR-Zion ID-SB0 32 0 30 < 10 1 .002 .002 2.-5 1.-5 1.-5 1.-5 CONTAINMENT PURGE ISOLATION FAILURE I This Study This Study IDCOR-Zion S6B S6C ID-IMPAIR 4 4
2 16 12 3 1 3.5
< 10
< 10
< 10 1
1 1
.01
.052
.01
.01
.052
.01 3.-4
.033 3.-4 6.-4
.0062 6.-5
.005 6.-4 6.-5 6.-5 6.-5 2.-4 NUREG-0956 TMLB'B 2 10 0 < 10 1 .022 .013 .11 .058 .0053 2.-4 L' ASH-1400 PWR-4 2 3 2 < 10 .6 .09 .04 .03 .005 .003 4.-4 This Study S6E** 2 10 0 < 10 1 .05 .05 .11 .06 .006 2.-4 I This Study IOCOR-Zion S78 10-BYPASS 8.5 24 CONTAINMENT BYPASS (V-SEQUENCE AT RHR PllMP SEAL) 7 5.5 4
< 10
< 10 1
1 3.-4 8.-5 3.-4 8.-5 2.-4 8.-5 1.-6 5.-5 3.-6 1.-5 3.-C 1.-G This Study S7C 8.5 7 2 < 10 1 .094 .094 .033 2.-4 4.-4 4.-4 I INTACT CONTAINMENT This Study 4.3 I 24 .6 < 10 .009 4.-8 4.-8 6.-9 4.-9 1.-9 1.-10 SSB This Study SSC 2 24 .4 < 10 .014 5.-7 5.-7 1.-7 6.-8 2.-8 2.-9
- Includes V-sequences involving pipe rupture outside containment.
** Enveloping source terms used for sensitivity analys 's.
NOTE: Exponential notation is indicated in abbreviated form; i.e., 2.-4 = 2.0 x 10-4 l l 1322P121685
I I I I TABLE 2-3. EVALUATION OF EMERGENCY PLANNING OPTIONS AGAINST NRC SAFETY G0ALS Early Fatalities Latent Cancer within 1 Mile Fatalities within of Site Boundary 50 Miles 3 Source E Risk to Risk to Risk
- Safety Goal Risk
- Safety Goal Ratio Ratio Safety Goal 5.0-7 -- 2.0-6 --
SSPSA 8.6-8** .17 6.3-9 .0032 This Studyt No Evacuation 2.4-9 .0048 7.3-9 .0037
- 1-Mile Evacuation 4.0-10 .0008 7.2-9 .0036 2-Mile Evacuation 1.6-11 .000032 7.0-9 .0035 I
- 2-Mile Evacuation, 1.6-11 .000032 6.5-9 .0033 10-Mile Sheltering 10-Mile Evacuation 1.6-11 .000032 6.7-9 .0034
*Mean frequency of health effects per person per year.
** Based on population within 1 mile from the containment.
t Evacuation distances are taken from center of containment; site boundary is located aproximately one-half mile from the containment. NOTE: Exponential notation is indicated in abbreviated form; i.e., 5.0-7 = 5.0 x 10-7, I I 2-18 1322P120285 _ _ _ _ _ _ _ _ _ _ __. _ _ _ _ _ ____ _. _ _ _ _ _ _ _ _ _ _ _ _ _ _ ~
I I I I I TABLE 2-4. COMPARIS0N OF CORE MELT FREQUENCIES AND DISTRIBUTIONS OF RELEASE TYPES I Risk Parameter WASH-1400 SSPSA Updated PWR Results e Mean Core Melt Frequency (events 9.9-5* 2.3-4 2.7-4 per reactor-year) e Percent Contribution of Release Types Gross, Early Containment 34 1 0.1 Failure Gradual Containment 66 73 60 Overpressurization or Melt-Through Containment Intact 0 26 40 I
- Based on WASH-1400 uncertainty ranges.
I NOTE: Exponential notation is indicated in abbreviated form; i.e., 9.9-5 = 9.9 x 10-5, I I I 1 I I 1322P120585 2-19 l ._-- --
TABLE 2-5. COMPARIS0N OF ACUTE FATALITY RISKS FROM DIFFERENT SOURCES Evacuation Mean CCDF Median CCDF Study Plant Case Distance (miles) Percent Risk
- Percent Risk
- WASH-1400 PWR Method 1 Mean** 25 7.5-5 100 1.9-4 100 Method 2 Mean** 25 7.5-5 100 9.7-4 51 0 NUREG-0956 Surry Smoothed, WASH-1400 10 6.4-5 85 - -
Unsmoothed, WASH-1400 10 4.0-5 53 - - BMI-2104, WASH-1400 Containment 10 1.1-5 15 - - BMI-2104, Containment Reevaluated 10 3.1-6 4 - - 10 3.7-5 50 4.8-4 250 7 SSPSA Seabrook Seabrook PRA E$ This Study Seabrook 10-Mile Evacuation 10 9.2-11 1 1.5-7 0.1 lb fe S e5 e $ng 25F 9.7-11 <<1 2.3-7 0.1 2-Mile Evacuation 2 2.1-10 <<1 6.7-7 0.4 1-Mile Evacuation 1 6.4-8 0.1 6.5-6 3 0-Mile Evacuation 0 3.1 -7 0.4 1.1-5 6
- Expected annual frequency of acute fatalties in population surrounding plant.
**See text for definition of Methods 1 and 2. ,
rEvacuation to 2 miles with sheltering to 10 miles. NOTE: Exponential notation is indicated abbreviated form; i.e., 7.5-5 = 7.5 x 10-5, l l l 1322P121885 M M M M M M M M M M M m M M M M M M
I I TABLE 2-6. SENSITIVITY ANALYSIS OF EARLY FATALITY RISK / SAFETY GOAL RATIO FOR NO IMMEDIATE PROTECTIVE ACTIONS I Treatment of Source Term and Site Release > 24 Hours All Releases Included after WarnTng Excluded Model Uncertainties BEST ESTIMATE AND CONSERVATIVE SOURCE TERMS ) I Probabilistically Weighted (mean)*
.0043 .0048 Probability [B,M] = 1 .0002 .0002 All Weight Placed on Best Estimate Source I Term and Site Model Assumptions Probability [C,H] = 1 .062 .092 All Weight Placed on Conservative Source I Term and Site Model Assumptions ENVELOPING SOURCE TERMS SUBSTITUTED I Probabilistically Weighted (mean)*
.0074 .0079 Probability [C,H] = 1 .15 .18 All Weight Placed on Conservative Source I Term and Site Model Assumptions
- Weights of .9 and .1 placed on the best estimate (B) and conservative (C) source terms, respectively; weights of .8 and .2 placed on the best estimate (M) and conservative (H) site model assumptions, respectively.
I I I 1322P121685 2-21
I I I I I TABLE 2-7. RATIO 0F MEAN HEALTH EFFECTS RISK WITH AND g WITHOUT A V-SEQUENCE EMERGENCY PROCEDURE E Ratio = Risk without V-Sequence Procedure a Risk with V-Sequence Procedure I Acute Latent Evacuation Scenarios Fatality Cancer Ratio Ratio No Evacuation 4 1 1-Mile Evacuation 9 1 2-Mile Evacuation 8 1 10-Mile Evacuation 9 1 2-Mile Evacuation /10-Mile Shelter 10 1 I I I I I I 1322P112085 2-22
I I I I "' ' ' ' ' I 10-4 I I N o WASH - 1400 MEAN (METHOD 2) b 10-5 - I 5
=
E l
- yso-e -
/
SSPSA (MEAN) g [ WASH - 1400 MEAN (METHOD 1) o THIS STUDY, NO I > l,o-7 N#0NTEAN) IMMEDIATE 0 I E 2 h I 2 10-8 _ I 3 ,,.0 0 i , , , 10 10 1 10 2 10 3 4 5 10 10 l EARLY FATALITIES I FIGURE 2-1. COMPARISON OF UPDATED EARLY FATALITY RISK CURVES FOR SEABROOK STATION (N0 IMMEDIATE PROTECTIVE ACTION) WITH SSPSA AND WASH-1400 (PWR)--MEAN VALUES I 2-23
I I 10-3 g i I I I I 10 4 - _ k e 10-5 __ _ e 6 a: CI: O 10-8 - - I 8 g WASH - 1400 MEDIAN I !10-7 8 SSPSA MEDIAN l E ! l mD,AN ,OR Te,S S1uDv.lTe NO IMMEDIATE PROTECTIVE ACTION I IS OFF SCALE < 10-9 I ' I ' 10-0 0 1 2 10 3 4 5 l 10 10 10 10 10 EARLY FATALITIES I FIGURE 2-2. COMPARIS0N 0F UPDATED EARLY FATALITY RISK CURVES , FOR SEABROOK STATION (N0 IMMEDIATE PROTECTIVE ACTION) WITH SSPSA AND WASH-1400 (PWR)--MEDIAN VALUES 2-24 i
l I I 10 4 g g g g l j I l 10 4 - - x WASH - 1400 MEAN (METHOD 2) I cc 10-5 d _ ~~~~\ N I O b WASH - 1400 MEAN (METHOD 1)
< \
\ SSN MEAN _
10-6 I : 0 I e
$ 10-7 g
THIS STUDY, NO IMMEDIATE* PROTECTIVE ACTION
\
\ _
2 I \
\
\
\
10~8 I \\ 10-9 0 I 1 I 2 I 3
\\ I 4 5 10 10 10 10 10 10 LATENT CANCER FATALITY RATE (YEAR-l) 'THIS CURVE NOT SENSITIVE TO EVACUATION ASSUMPTIONS I FIGURE 2-3. COMPARIS0N OF UPDATED LATENT CANCER FATALITY I RISK CURVES FOR SEABROOK STATION (N0 IMMEDIATE PROTECTIVE ACTION)
WITH SSPSA AND WASH-1400 (PWR) I 2-25
l I I I 10-2 I a S BACKGROUND ACCIDENTAL FATALITY RISK I E 10-3 / (5 FATALITIES PER 10,000 POPULATION PER YEAR) I
~
o H h 104 - N Q 10-0 - 2 SAFETY GOAL (.001 TIMES y 10-6 BACKGROUND RISK) I c2 10-7 - I o THIS STUDY FOR 5 E SEABROOK STATION J (NO IMMEDIATE y 10-8 - PROTECTIVE ACTION)
\
10-9 I FIGURE 2-4. COMPARISON OF SEABROOK STATION RISK (WITH NO IMMEDIATE PROTECTIVE ACTION) WITH BACKGROUND AND SAFETY G0AL INDIVIDUAL RISK LEVELS I I I 2-25
I I I I 1.00 g i g i l i i I 0.9 - - 0.8 - I 5 d 0.7 - -
$- 0.6 - -
E_
$ 0.5 - -
8 I u y 0.4 - - 8 z 0.3 - - 9 D I g 0.2 - - m 0.1 - -
' ' I I I I 0.00 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 DISTANCE (MILES)
I FIGURE 2-5. SPATIAL DISTRIBUTION OF THE EXPECTED FREQUENCY OF ACUTE FATALITIES FOR SEABROOK STATION BASED ON UPDATED RESULTS FOR NO IMMEDIATE PROTECTIVE ACTION 2-27
I I I I 10-3 l i i i I 10 4 - - I 2 e I i 10-5 .- I h a. I 10-6 - I 0 WASH - 1400 MEAN (METHOD 1) e o
/ I 10-7 8
p1 - MILE EV LEGEND < EV = EVACUATION $ SH = SHELTERING 10-8 _ j 2- MILE EV to - MILE EV 2 - MILE EV
~
10-9 NI I g 0 2 3 4 5 10 5 1 10 10 10 10 10 EARLY FATALITIES FIGURE 2-6. IMPACT OF DIFFERENT EMERGENCY PLANNING OPTIONS ON RISK 0F EARLY FATALITIES (RESULTS OF THIS STUDY COMPARED AGAINST WASH-1400) 2-28
I 10-2 , i 6 i I _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ N NRC SAFETY GOAL FOR INDIVIDUAL RISK MULTIPLIED BY POPULATION WITHIN 1 MILE OF SITE BOUNDARY 30-3 _ _ E I c { 10 4 - - I 5 P 3 5 I 1 8 I < z 6 10-5 - 2 I 10-0 -- - REDUCTION IN 2-MILE EVACUATION RISK WITH SHELTERING TO 10 MILES I i X I 10-7 0 2 4 6 8 to EVACUATION DISTANCE (MILES) FIGURE 2-7. ACUTE FATALITY RISK AS A FUNCTION OF PROTECTIVE ACTION 2-29 l ?
100 I I l M - 80 - I z 70 - I e
$ 60 -
X 2-MILE EVACUATION l g WITH SHELTERING TO 10 MILES 5 E N s s* - c s
=
I 40 I u
$m - _
l 20 - _ l I 10 I l - l l 0 I I l l ) 0 2 4 6 8 10 l EVACUATION DISTANCE (MILES)
- NORMALIZED AGAINST RISK REDUCTION OF 10-MILE EVACUATION "
FIGURE 2-8. EARLY FATALITY RISK REDUCTION FOR DIFFERENT I PROTECTIVE ACTION STRATEGIES 2-30
I I I i i
- a i iiil 6 i e i i all i i e a i a i.
I g 0.1 -- w _- _ os - h3 - e8 - yy , 50 REM mb - W= m2
- u. W o$ 0.01 --
go - z< - I wz - Sy - EG - 200 R EM -
<m G /
\ 50 REM g8 - \/ -
l z O 0.001 -
\
O NUREG - 0396 7 _ MEDIAN
\
\
O MEAN THIS STUDY FOR l -
) SEABROOK STATION I O O
200 REM MEDIAN EDE 200 REM CURVE OFF
~ ""
/ I SCALE 1
0.0001 t ' I ' ' ' ' ' ! ' ' ' ' '' I 10 100 1,000 1 DISTANCE (MILES) ! FIGURE 2-9. COMPARISON OF UPDATED SEABROOK STATION RESULTS WITH NUREG-0396 - l 200-REM AND 50-REM WHOLE BODY DOSE PLOTS FOR NO IMMEDIATE PROTECTIVE ACTION l j 2-31
I I I
- 1 t i t i I I I a i l-I I I i iiIl I iI l ll [
t - i 4, - O"% , O "'
\ 1 REM g 0.1 --
5 REM
}
8 _- - g g - os ~\ ~ gz \ - , e9 b
\ \
g
\
wb n O s O 4 / o $ 0.01 - o O t8 : \
} gi E
Ss -
\ _
is :
' vi -
s
\
- I
@o!
z O go - - h 5 REM o,001 O "" 0 NUREG - 0396
\ MEDIAN -
l g
\
m MEAN TH'S STUDY FOR _ SEABROOK STATION
- ~
O--O MEDIAN g f l l 0.0001 1 10 100 1,000 - DISTANCE (MILES) FIGURE 2-10. COMPARIS0N OF UPDATED SEABROOK STATION RESULTS WITH NUREG-0396 - I 5-REM AND 1-REM WHOLE BODY COSE PLOTS FOR N0 E IMMEDIATE PROTECTIVE ACTION 3 2-32
I I I 1 i e i iiiil i e i i s ii l e i i e ist I I o 01 7 1 REM SREM I s ;% 3g g - h \ e< r wb - O , - Nw m2
\ s.
g O % O 0.01 \ g go
- O g -
5 : s o : Dw - O \ S> \ \ - I O
\ 1 REM -
aw - O g - 58 go
\ O
\
\ \
t- O s O
\o
\
\
\
0.001 -- I
<p -.-
OO -
\ NOREG - 0396
\ MEDIAN -
! O - 0 0 9 - 5 REM
^ -
THIS STUDY FOR I SEABROOK STATION o- -o - MEDIAN ' I
' ' ' ' ' ' ' I ' '
0.0001 ' ' ' ' ' ' 1 10 100 1,000 DISTANCE (MILES) FIGURE 2-11. COMPARIS0N OF UPDATED SEABROOK STATION RESULTS WITH I NUREG-0396 REM AND 1-REM WHOLE BODY DOSE PLOTS FOR RELEASES WITHIN 24 HOURS OF WARNING AND NO IMMEDIATE PROTECTIVE ACTION 2-33
- 3. SSPSA PLANT MODEL UPDATE I The purpose of this section is to present an update of the SSPSA plant model results. This update is responsible for specific differences in the definition and quantification of accident sequences between those published in the SSPSA and those used to evaluate risk management options in Section 2. Those differences in part reflect what has been learned since the SSPSA was completed, modifications to the plant technical specifications, and the incorporation of new and revised emergency I operating procedures into the Seabrook Station plant model. Because the SSPSA was originally intended and is still considered as part of an ongoing risk management program, the need for providing updates is a potentially continual need. The objective in providing these particular changes to the SSPSA plant model is to ensure that the evaluation of emergency plan options and other risk management actions in this study is based on the most current and best available information regarding the I definition and quantification of potential accident sequences for Seabrook Station.
I Presented in this section is the cumulative effect of two discrete parts of the update, one that was recently published as part of an evaluation of changes to the technical specifications (Reference 3-1) and a second that reflects additional changes for the evaluation of risk management and emergency planning options. The specific changes covered in both parts of the update are itemized below.
- 1. Modifications to plant technical specifications.
- 2. Detailed reevaluation of the contribution of common cause failures to I five important systems with respect to risk contribution: the electric power, primary component cooling water, service water, emergency feedwater, and containment isolation systems.
- 3. Removal of a conservative assumption in the SSPSA about the response of air-operated containment isolation valves during seismically initiated station blackout scenarios.
- 4. Incorporation of more detailed and realistic models for the j initiation and progression of sequences involving interfacing system loss of coolant accidents.
I
- 5. Evaluation of changes to existing emergency operating procedures for i E new sequences defined in item 4.
! g l 6. Identification and evaluation of procedures for the recovery of < containment heat removal for core melt sequences involving station blackout. ! A partial update covering the cumulative effect of items 1 through 3 l above was published in Reference 3-1. The cumulative effect of all six l items is described below. l I 1319P112285 3-1 L
I The plant model portion of the Seabrook Station risk model includes the definition of accident sequences from the initiating events to plant states. As described more fully in Section 1, the plant states include a successful termination state and a total of 39 pl' ant damage states, all of which involve severe core damage. The total frequency of core damage (i.e., the frequency of all sequences assigned to all plant damage states) is simply the sum of the frequencies of the plant damage states. Therefore, a full explanation of the effect of the above six changes on the SSPSA plant model can be made in terms of changes to the plant damage state frequencies. These changes are summarized in Table 3-1. Table 3-1 includes the results of the two-part update process for a total of 15 plant da.nage states. This set of plant damage states includes the nine states identified in the SSPSA as making significant contributions rg to risk and core melt frequency (see Section 1, Table 1-3). 5 Three additional plant damage states from the SSPSA that exhibited potential for significant changes in the technical specification study in Reference 3-1 (1FP, 2A, 7F), and three new plant damage states introduced ( in this study to characterize new source terms for interfacing systems ' LOCA sequences (1FV, IFPV, 7FPV). As seen in Table 3-1, the changes in plant damage state frequencies Il, included in the first part of the update (Reference 3-1) were significant l only for states IFP, 3F, 3FP, 7F, 4A and 8D. The first four of these g changes reflected the reassessment of the behavior of air-operated 5 containment purge isolation valves during seismically initiated station blackout sequences. In the SSPSA, it was assumed that if the valves are , initially open (an assumed .10 fraction of time) and the actuation signal I for valve closure failed, these valves would remain open for the duration i of the accident; i .e., until completion of the " source term." In l Reference 3-1, it was assessed that prior to core damage or uncovery, I these valves would most likely close because of the loss of instrument air that would occur after loss of offsite power /and the failed closed feature of these valves. This change resulted in the reassignment of g some sequences from "F" type states (i.e., large, unfiltered containment E : bypass) to the corresponding "FP" states; i.e., small containment leakage. This explains the redistribution of frequency from states IF, l l 3F, and 7F to IFP, 3FP, and 7FP, respectively. l The remaining plant damage state that changed significantly in l Reference 3-1 was 8D, whose increase almost fully explains the small l increase in core melt frequency from 2.3 x 10-4 to 2.8 x 10- per I reactor-year. This change is the result of a reassessment of the common l cause contribution to the unavailability of the primary component cooling i water system. It stems from an application of an enhanced systematic procedure for common cause analysis to the five systems identified in Reference 3-1 as having the greatest risk significance. Application of this more systematic variation of the same common cause analysis procedure that had been applied in the SSPSA revealed only minor changes in the remaining systems analyses. The slight decrease in core melt frequency to 2.7 x 10- is due to a reassessment of the interfacing systems LOCA sequence in this update. i i 1319P112285
I The second stage of the plant model update, described in Table 3-1, are the results of items 4 through 6 listed above. These changes reflect a more realistic treatment of the plant response and operator actions in response to interfacing systems LOCA (i.e., "V-sequence") scenarios and the incorporation of operator actions to restore containment heat removal during core melt sequences initiated by station blackout conditions. The I enhanced treatment of the V-sequence has resulted in the removal of the previously analyzed V-sequence, which dominated plant state IF in the SSPSA, and the addition of new sequences to the plant model, primarily in j the new plant damage states IFV, 1FPV, and 7FPV. The final change, j incorporation of containment recovery, has resulted in a shift in the l assignment of accident sequences from states 3D, 7D, and 8D to state 8A. I The "D" states denoted an initially isolated containment with no containment heat removal and a dominant containment failure mode of long-term overpressurization. By contrast, the "A" states have I successful containment heat removal, conditions established for debris bed cooling, and a very high likelihood of indefinite containment recovery. Therefore, this shift in accident frequency from D to A plant damage states, while reflecting no change in core melt frequency, I represents a reduction in the likelihood of radioactivity release and a corresponding decrease in latent health risk in relation to the situation without such a shift. The plant model changes described above also had an effect on the ranking of the dominant accident sequences originally identified in the SSPSA and partially updated in Reference 3-1. In the updated results of this
.I study, there was a total of 43 sequences whose mean point estimates had a value of 1 x 10-6 per reactor-year or greater. A ranking of these sequences with respect to core melt frequency contribution is presented I in Table 3-2, which includes individual sequence frequencies and a definition of the sequence according to specific sequences within specific plant damage states. These sequences are defined further in I terms of initiating events and syste:n top events and boundary conditions in Table 3-3. These events and boundary conditions are, in turn, defined in Tables 3-4 and 3-5. Any reviewer of this document who wishes to attempt to recreate some or all of these results should be able to do so I with two additional sources of information. The additional sources of information are the system models and the data base published in Section 6 and Appendix D of the SSPSA and the update to the systems models provided in Reference 3-1.
The balance of this section is devoted to the reassessment of interfacing I systems LOCA scenarios and the evaluation of containment recovery actions in Sections 3.1 and 3.2, respectively. 3.1 INTERFACING SYSTEM LOCA SEQUENCES The interfacing systems LOCA (or event V), described in the SSPSA, results from failure of the two series check valves in one of the four I separate low pressure RHR injection lines, or from failure of the two series motor-operated valves in one of the two RHR hot leg suction 1319P120285
I lines. Failures of these valves were also assumed in the SSPSA to fail the low pressure piping in the RHR system, creating a LOCA and a path for primary coolant to bypass the containment. A key factor in making this assumption was a concern identified in the Indian Point Probabilistic Safety Assessment (Reference 3-2) that the RHR system could experience transient overpressures significantly greater than the RCS pressure as a result of shock waves if the multiple valve failures were postulated to E1 occur in a very sudden, gross manner. In the SSPSA, it was assumed that 34 such failures led directly to core melt because of difficulties in i maintaining reactor coolant inventory control under such conditions. I;; Although the frequency of interfacing systems LOCAs (and, hence, the contribution to core melt) was found to be relatively small (a mean value of 1.8 x 10-6 per year for all sequences), the V-sequence was found to 4 be a dominant contributor to the risk of early health effects. Because i of this, any efforts to manage the risk of early health effects, such as ! the definition of protective actions in the emergency plan or the l optimization of emergency operating procedures, need to address this 1 class of accident sequences. Since the V-sequence was first identified and analyzed in the Reactor Safety Study (Reference 3-3), the models and assumptions used to characterize its risk contribution have not changed very much until after , the SSPSA. The most significant variations in the analysis of the V-sequence among PRAs has been the modeling of plant-specific piping and valve configurations and the use of different valve rupture failura rates based on different data. Based on what has been learned since the original V-sequence models were developed in the RSS, there are several ; reasons to believe the SSPSA analysis was conservative; therefore, to ensure that a realistic estimate can be made of the effectiveness of risk i management actions, a reexamination of the V-sequence analysis is ' l warranted.
- Given that the series check valves or the series motor-operated valves l between the RCS and RHR systems have failed, it was assumeo in the SSPSA that the low pressure RHR piping would fail. Recent calculations 1 l performed in support of the IDCOR program for the revised Zion V-sequence ,
i source term (Reference 3-4) indicate that the Zion RHR low pressure l l piping has sufficient capacity to survive both the static and dynamic ! loads of the sudden pressurization. l l Given that the RHR system piping and components survive the accidental pressurization, the IDCOR evaluation concluded that the RHR pump seals represent the most probable point of failure in the system. The IDCOR j source term calculation for the Zion V-sequqnce assumed an upper bound seal leak area of 14.4 square inches (93 cmd ) (total for both RHR pumps); however, it was noted that the expected seal leak flow area for i such an event would be considerably less than this conservative value. Another factor that pointed to the conservatisms of the SSPSA analysis was experience with a number of events at BWRs (Reference 3-5) in which the low pressure RHR piping was pressurized to RCS pressure. While the experience and the design pressures were both lower for the BWR events, I 1319P112285 . 1
I the fact that only minor RHR pump seal leakage was observed in these events is supportive of the assumption made in IDCOR for Zion. I In the process of applying these new insights about RHR piping capacity to Seabrook Station, the possibility of flooding the vaults that house the RHR pumps was identified. Such flooding would cover the pump seals with water, creating a mechanism for fission product " scrubbing" prior to I any release to the enclosure building atmosphere. Although the IDCOR evaluation addressed the possibility of flooding, no such credit was taken in their analysis since extraordinary efforts must be performed by the operator to produce RHR pump flooding in the IDCOR reference design. The Seabrook vault configuration, however, has an inherent capacity to provide this flooding. Even though no credit was taken'for flooding, the I IDCOR-calculated source terms for the V-sequence were considerably lower than those used in the SSPSA. The lower source terms resulted from a detailed MAAP calculation of the Zion V-sequence in which plateout in the primary coolant system and RHR system were accounted for as well as I natural depletion mechanisms of the aerosols in the auxiliary building. There are implications of the IDCOR results for RHR piping integrity that point to possibly ncnconservative elements of the SSPSA V-sequence model. The IDCOR results point to the need to redefine the initiating event as any valve failures that lead to RHR system pressurization rather I than those that occur in the catastrophic manner needed to produce shock waves. Such a redefinition would lead to an increase in the initiating event frequency because less severe valve ruptures would cause this, depending on the response of the relicf valves in the RHR system inside I and outside the containment. On the other hand, without the assumption of gross RHR piping failure, these higher frequency events would not necessarily lead directly to core melt and a significant contribution to I early health risk, as assumed in the SSPSA. In summary, there is a good technical basis for performing a new, more realistic assessment of the V-sequence for Seabrook Station to avoid misleading conclusions about the effectiveness of any risk management actions. This more realistic I assessment is presented in the following section. 3.1.1 SEABROOK CONFIGURATION An extensive review of the Seabrook design features related to the V-sequence has been conducted during this study. This review concluded I that the design features of Seabrook relevant to initiating the V-sequence are quite similar to those for Zion. Of special note are the configuration of the Seabrook RHR/CBS vaults and the arrangement of the RHR and CBS system components within these vaults. Failure of the RHR I system within the vaults has an inherent potential for submerging the failure location and release path. l The Seabrook RHR system consists of two heat exchangers, two pumps, and l the associated piping, valves, and instrumentation necessary for j operation and control. A P&I diagram of this system is provided in Figure 4-8. The inlet lines to the system (i.e., the pump suction) are connected to the hot legs of two of the four reactor coolant loops, while the return lines are normally aligned to the cold legs of each reactor coolant loop. These retJrn lines also serve as the ECCS low pressure injection lines. I 1319P112285 3-5
Each RHR pump is an Ingersoll Rand unit that has a design pressure of 600 psig. The RHR pump seal is a mechanical seal unit that is designed for 600 psig (the seal undergoes a cold hydro at 1,200 psig in the shop). No additional tests at 1,200 psig are planned (e.g., after seal maintenance or replacement). The series check valves are located in the injection lines. The series motor-operated valves and the relief valves on the RHR pump suction lines located inside the containment. The remainder of major components of the system are located either in the pipe penetration area, in the pipe tunnel, or in the RHR/CBS equipment vaults. The RHR suction line joins the reactor coolant system at I Elevation -19'0", passes through the primary containment wall at Elevation -18'5", enters the RHR equipment vault at Elevation -29'5", and terminates at the RHR pump at Elevation -57'4". The suction line is approximately 100 feet in length inside the containment and 125 feet in length outside of the containment. The line has nine 90-degree elbows inside the containment. The RHR discharge line joins the RHR heat exchanger at Elevation -28'10", passes out of the equipment vault at Elevation -21'8", passes into the contair. ment at Elevation -18'5", and ends at the RCS at Elevation -10'3". The length of the RHR discharge line is approximately 180 feet inside the containment and 110 feet outside the containment. The line contains thirteen 90-degree elbows inside the containment. Each RHR pump suction line contains a relief valve, which is rated at the combined flow of all the charging pumps at a setpoint pressure of 450 psig. These valves are 3-inch by 4-inch Crosby-type L valves, which are fully open at a 10% accumulation (i.e., N 500 psig). Under these conditions, each valve has a rated flow of approximately 990 gpm.* The suction-line relief valves relieve to the pressurizer relief tank inside the containment, which, in turn, relieves to the containment atmosphere at the setpoint of the tank rupture disk (106 psid). Each RHR pump discharge / cold leg injection line to the RCS contains a small capacity B relief valve located outside the containment. These valves are rated at 3 20 gpm each at a setpoint pressure of 600 psig and relieve to the primary drain tank which is also outside the containment. The transition from low pressure to high pressure designed piping occurs inside the containment for the RHR pump suction lines and inside the tunnel region for the pump discharge / cold leg injection lines. Suction line piping is as large as 16 inches in diameter, but the maximum diameter of injection line piping is 8 inches. Each of the RHR injection paths has a normally open motor-operated isolation valve (RH-V14 for 3 train A and RH-V26 for train B) located in the pipe penetration area. 3
*The corresponding flow rate at an RCS pressure of 2,250 psi would be approximately 2,100 gpm per valve. The flow through a total RHR pump seal leak area of 2.6 square inches at 2,250 psi would be between 2,900 and 4,800 gpm, depending on the flow coefficient, which is assumed.
Because of the large length-to-diameter ratio of the seal flow path, the lower value is believed to be representative of the seal leak. 3-6 1319P112285 ,
I These valves are the transition point from high to low design pressure and can be used to isolate the RHR system from the RCS if necessary. I In the normal standby mode of ECCS operation, the discharge of either RHR pump can supply all four injection lines. This is accomplished by a crosstie line that connects the downstream side of both RHR heat exchangers. Two normally open motor-operated valves (RH-V22 and RH-V21) I facilitate this flow communication. Each of the RHR pumps is provided with a 3-inch diameter minimum flow I line to prevent overheating when the RHR injection paths are isolated from the RCS. Each of the minimum flow lines contains a motor-operated valve, which is open in the standby mode, but which is closed I automatically when flow is sensed in the injection lines. Because these valves and the crosstie valves are normally open, the entire RHR system will tend to pressurize uniformly after valve failure (neglecting the time it takes for pressure waves to traverse the system). As noted, the Seabrook RHR pumps are located in watertight equipment vaults that have three connected cubicles. These vaults also house the I RHR heat exchangers, the containment building spray pumps and heat exchangers, and the high pressure safety injection pumps. Separate vaults are provided for the A and B trains of these systems. The third cubicle is the stairway and accessway to the two equipment cubicles. Each of the vaults extends from the roof, Elevation 25.5', to Elevation -61' where the sumps are located. The coor connecting to both vaults is at Elevation -30', and there are no openings or ducts below this elevation. The RHR and containment spray pumps are installed at the lowest elevation of the vault (i.e., -61'). The RHR pump seals are I located approximately 6 feet off the floor. Each intermediate floor in the vault contains gratings, so water from a line break or component failure will fall through the grating and accumulate at Elevation -61' I where it would be directed to the sump located in the CBS pump cubicle. Each sump has two 25-gpm sump pumps and a 920-gallon capacity (to overflow) . Sump high and low-level alarms are provided at the waste management system control panel (CP-38A) and remotely alarmed on the MCB in the main control room. The Seabrook equipment vault ventilation system is part of the emergency I enclosure building ventilation system and has a flow rate to and from the vaults of about 24,000 scfm per vault. The system provides makeup air to several lower levels of the equipment vaults (19,020 scfm to Elevation -61'0" and 4,560 scfm to Elevation -50'0") and exhausts from I the uppermost level at Elevation 25'6". system is processed through a filtration system consisting of moisture The exhaust from the ventilation I separator, absolute filter, carbon filter, backdraft damper, and fan and exhausted to the plant vent. The system also contains fire dampers in the exhaust line that close when a fusible link reaches a temperature of 165 F. Due to the passive heat sinks between the break location and I the fusible links and the creation of a suppression pool for condensation of break flow, it is uncertain whether this event would trigger the isolation of the ventilation system with high temperature. I 1319P112285 3-7
l Il 1 l All the pumps located in the vault are assumed to fail once they become , submerged. Splash shields would protect the pump motors from damage prior to, or in the absence of, submergence. It is estimated that a l water level of 2 feet will fail the CBS and sump pumps and a level of l 6 feet will fail the RHR pump. The high pressure safety injection pumps are installed at Elevation -50'0" and will fail when the water level reaches a depth of 13 feet. I As shown in Table 3-6, the RWST serves as the initial source of suction for the two RHR pumps, the two centrifugal charging pumps, the two high l pressure safety injection pumps, and the two containment building spray I pumps. As a requirement of the Technical Specifications for the Seabrook Station, the RWST contains a minimum of 450,000 gallons of borated water at a maximum temperature of 85 F. After a low-level signal from the RWST (after%350,000 gallons have been depleted), valving associated with the RHR and CBS pumps automatically realign to take suction from the containment sump. Because of net positive suction-head requirements, the high pressure ' safety injection and centrifugal charging pumps are not automatically I aligned to the containment sump. Instead, the operator must manually align the suction lines of these pumps to the d;scharge lines of the operating RHR pumps. The charging pumps are aligned to take suction from the discharge of RHR pump P-8A, and the high pressure safety injection E pumps are aligned to take suction from the discharge of RHR pump P-88. 5 Thus, for a normal LOCA event, a recirculating source of water is available to provide core cooling. The positive displacement and centrifugal charging pumps are aligned to , take suction from the 4,725-gallon volume-control tank during normal l operation. Upon receipt of an S-signal, the centrifugal charging pumps are automatically aligned to take suction from the RWST. ll ' l l In analyzing the configuration described above for its impact on the E 'l l various V-sequence scenarios, a number of important considerations E arise. First, because the suction-side relief valves relieve to the pressurizer relief tank, the tank is very likely to overpressurize, g relieve to the containment, and cause containment pressure to increase as gi long as the RHR relief valves remain, or cycle, open. Thus, in addition to initial indications that a PORV may be open (i.e., PRT level), overpressure of the PRT will present the operator with indications that a LOCA inside the containment has occurred. Second, if containment pressure reaches the safety actuation P-signal setpoint of 18.0 psig, the containment spray pumps would be signalled to start. The containment spray pumps are each rated at 3,000 gpm and will deplete the RWST inventory quite rapidly if both operate for an appreciable period (i.e., approximately 50 minutes) of time. Third, if the high pressure challenge to the RHR system is greater than can be mitigated by the capacity of its relief valves, it is likely that a failure will occur somewhere in the system, most likely at the RHR pump seals. Other l candidates are RHR piping, valve bodies, and the RHR to PCC system heat exchangers. 3-8 1319P112285
[ Any failure of the RHR system located in the vault will subject all of p the high pressure safety injection, RHR, and CBS pumps to an abnormal L steam, humidity, and thermal environment as well as to the damage from submerged pump motors. If the CBS pumps fail early, RWST water will be available to the RCS for core cooling for a longer period of time and an [ RHR pump seal failure would most likely become submerged. However, i f the CBS pumps start and continue to run for an extended period of time, much of the RWST inventory will be directed to the containment, reducing F the flow of water to the RHR vault and the extent the pump is submerged. If the RHR pumps survive the seal failure, the water in the containment could be used for recirculation. It is assumed, however, there is a high probability that the RHR pump would fail to operate, given any sizable seal leak. If the RHR pumps fail, normal RHR recirculation would be negated and only the centrifugal charging pumps (taking suction from the RWST) would be available for extended core cooling and maintenance of inventory control.* Thus, it is possible that extended operation of the CBS pumps in environments for which they were not designed could produce detrimental effects in mitigating the V-sequence. However, such operation is unlikely since these pumps are mounted in a horizontal configuration in the bottom of the vault and would become submerged very quickly, given any significant leakage to the vault. The splash shields would not protect the motors from shorting since they are not watertight. A considerable reduction in the V-sequence source terms can be justified even if the vaults do not flood, as is demonstrated in Section 4. Should the quantity of RWST and RCS water released to the containment via the RHR relief valves and/or the operation of the containment spray pumps be sufficient to flood the reactor cavity, any core debris released to the p cavity would be covered by water. Since the debris would be covered, L core-concrete interaction and the hot gases produced by such interaction would be minimized. The absence of these hot gases will decrease the potential for any revaporization of fission products deposited on RCS or [ RHR surfaces or on RHR and CBS vault surfaces. Small pump seal leaks of less than 50 gpm are within the capability of the RHR vault sump pumps. Normally, one of the sump pumps is designated { as the preferred pump and the other as the backup pump by the operators at the waste management system control panel (located in the waste management building). The preferred pump starts automatically on a high-level signal . If the water level in the sump container continues to rise, the backup pump starts automatically when the water level reaches the high-high level setpoint. A high-level alarm for each sump is E annunciated at CP-38A in the waste management building and on the MCB in the main control room as well . Leaks greater than 50 gpm will quickly flood the sump pump motors. Since these motors are not submersible, they are expected to fail when they become submerged. { [ *At longer times, the capacity of the positive displacement charging pump (taking suction from the volume control tank) could provide a suitable { source of water for core cooling. 3-9 - 1319P112285
I Each RHR/CBS vault contains a low-range and a high-range area radiation monitor located at an elevation in the vicinity of the RHR pumps. Although this arrangement should provide an indication that RCS radioactivity is entering the vaults when a leak is postulated to occur, it is assessed that these monitors would fail or become ineffective if they become submerged. Note that~in the first few months of core power operation, the primary coolant circulating activity levels approach 3 equilibrium values. Early in the time period, detectors may not respond 5 to RCS leakage into the vault. In addition to the vault radiation monitors, there is also a containment enclosure air monitor that initially can indicate a release from either the charging pump or the RHR/CBS vaults. However, any radioactivity in the enclosure building is quickly distributed, and the location of the source soon becomes difficult to pinpoint. The plant stack radiation monitor may also provide an indication of any releases outside containment. All of these radiation monitors are alarmed in the control room on the RDMS panel, the visual alarm system, and the computer alarm system. 3.1.2 INITIATING EVENT ANALYSIS 3.1.2.1 Plant Response ^The initial plant response to a V-sequence will be somewhat dependent on the size of the breach in the RHR system. In general, it will be similar g to that of a small or medium LOCA, inadvertent opening of a PORV, or g steam generator tube rupture event although the timing might be different. The first indication of trouble, depending on the size of the LOCA, will be the actuation of pressurizer low pressure or low-level alarms. Normal charging pump flow from the volume control tank (VCT) will increase in an effort to maintain pressurizer level. Low pressurizer level isolates reactor coolant letdown. After receiving a E low-low VCT level signal, the charging pumps are isolated from the VCT E and pump suction is taken from the RWST. If valve leakage exceeds normal makeup capacity of the c arging pumps, the continued loss of RCS g inventory will lead to a reactor trip signal and a safety injection g signal (the latter is called the S-signal) generated by low pressurizer pressure. The S-signal automatically performs the following functions:
- 1. Initiates Phase A containment isolation as well as normal feedwater isolation.
- 2. Actuates the emergency feedwater system.
- 3. Starts emergency diesels.
- 4. Initiates safety injection,
- a. Starts all ECCS pumps.
3-10 1319P112285
I
- b. Opens RWST pump suction valves (high pressure safety injection, charging, and RHR).
- c. Opens CVCS injection valves (ECCS mode).
The containment building spray system does not automatically actuate on the S-signal. Automatic actuation of the CBS system requires a P-signal, I which is generated by high containment pressure (18 psig). V-sequences in which the RHR system fails inside the containment and sequences that produce sufficient relief valve flow to the pressurizer relief tank to fail its rupture discs would have a high potential for activating containment spray. I As plant pressure decreases, the RWST continues to supply borated water through the CVCS pumps to the RCS at an increasing flow rate. When pressure falls below the shutoff head of the high pressure safety injection pumps (N1,540 psi), these pumps begin to inject borated water I from the RWST into the RCS cold legs. If RCS pressure falls below that of the accumulators (615 psi), their contents are also discharged into the RCS. The RHR pumps start upon receipt of the S-signal, but will not I provide low pressure injection until RCS pressure falls below their shutoff head (185 psi). Prior to reaching this pressure, flow is circulated through the miniflow lines of the RHR pumps to prevent pump I overheating. As soon as the pumps begin to produce flow to the RCS, valves in the miniflow lines close and all RHR pump flow is injected into the reactor vessel via the RHR cold leg injection lines. The low pressure injection function performed by the RHR pumps is expected to be seriously degraded or negated entirely for a large number of postulated V-sequence failures. While flooding of the vaults will I produce a lower source term because of " scrubbing," it will cause failure of the RHR pumps as well; hence, the likelihood of progressing to core melt is greater. Furthermore, many potential failure locations cause diversion of low pressure injection flow away from the reactor vessel. I As noted earlier, RHR system failures within the vault can cause eventual failure of the high pressure safety injection and CBS pumps as well. l Failures that preclude low pressure injection will most likely preclude both high and low pressure recirculation as well. The anticipated plant response to the maximum expected RHR pump seal leak I area (i.e., approximately 1.3 square inches per pump) has been calculated by the MAAP program (Reference 3-6) for the V-sequence, as summarized in Table 3-7. For this scenario, it is predicted that ECCS and a reactor trip will be initiated within 5 seconds, reactor coolant pumps will be I tripped within approximately 21 seconds, the PRT rupture discs will fail within about 26 seconds, and the RCS is expected to be solid within 30 seconds. Within approximately 10 minutes, the water level in the RHR I and CBS vaults is expected to be sufficiently high to submerge the CBS pumps. This submergence occurs before the containment pressure P-signal is generated; therefore, containment spray is preempted unless the I operators manually initiate the pumps prior to their failure in 10 minutes. Within approximately 12 minutes, RCS pressure has decreased to the setpoint pressure of the RHR system relief valves, and they begin I I 1319P112285 3-11 i n
i I to modulate'. This modulation continues for approximately 2.8 hours at which time the high pressure safety injection pumps become submerged and fail. Because of the relatively large seal leak, the RHR pumps were assumed to fail at the inception of this scenario as a result of seal leak spray injection into the motors. Without any makeup, the RWST supply of 350,000 gallons of water would be 3 exhausted at approximately 6.4 hours after the initiating event, based on 3 the MAAP analysis assumption of one charging and one safety injection pump operating. At this time, only the charging pumps remain operational, but they cannot be lined up to the containment sump. Hence, neither injection nor recirculation is available for cooling, and core damage results. For this scenario, the core uncovery is predicted to begin at approximately 8.1 hours, the core begins to melt at approximately 9.9 hours, and reactor vessel melt-through occurs at approximately 11.5 hours after the initiating event. Since RCS pressure at the time of vessel failure is less than 300 psia, the event is 3 categorized as nondispersive. In addition, only about 47,000 gallons of E water is predicted to reside on the containment floor. This quantity is well below that required to spill over the curb into the reactor cavity. Hence, so-called " dry" containment conditions exist at the time of vessel failure. After initiation of reactor trip or safety injection signals, the operators are trained to implement Emergency Instruction E-0. After verifying that ESFs are functioning as required, that the main steam lines need not be isolated, and that the steam generator pressure boundary is intact, the operator is instructed to check if the RCS is intact (Step 24 of E-0). If containment radiation, pressure, or building water level is greater than normal, the operator is instructed to go to Emergency Procedure E-1, Loss of Reactor or Secondary Coolant, Step 1. Step 30 of E-0 requires the operator to check for radiation in the ' auxiliary building and to transfer to Emergency Procedure ECA-1.2, LOCA Outside Containment, Step 1 if above normal radiation is detected. Because the PRT is predicted to fail as a result of the opening of the RHR relief valves, conditions in the containment will appear to be those of a LOCA, and the operator will transfer to E-1 before he reaches Step 30 of E-0. Although Step 12 of E-1 requires verification of RHR pump operation and a check for containment leakage, E-1 has no transfer points to ECA-1.2. Step 12 of E-1 does require transfer to ECA-1.1, Loss of Emergency Coolant Recirculation, Step 1, but ECA-1.2 is bypassed. Emergency procedure ECA-1.1 provides actions to restore emergency coolant recirculation capability, to delay depletion of the RWST by adding makeup and reducing outflow, and to depressurize the RCS to minimize break fl ow. Thus, the path through existing emergency procedures will initiate makeup, but will not necessarily result in positive actions to diagnose the event and to terminate the escape of coolant from the RCS if possible. 3.1.2.2 Initiating Event Analysis In general, the frequency of failure for two valves, Vi and V2 , in series (V1 is assumed to be nearest to the RCS) can be expressed as As = A(V1 )*P(V2 lV1 ) + A(V2 )*P(VilV2 ) (3.1) 1319P112285
I where As = the frequency of failure of both series valves. l A(VI ) = the frequency of random, independent failure of valve VI . P(VlV)=theconditionallikelihoodthatV2 2 1 is failed, given that Vi fails. A(V2 ) = the frequency of random, independent failure of V2 (events per hour). I P(V1lV2)=theconditionalprobabilitythatV1 is failed, given that V2 fails. P(V2lV1) and P(V 1lV 2) are composed of both random, independent, I and demand type faiTures of the second valve. In some cases, the random, independent failure frequencies and I conditional probabilities for the two valves will be approximately equal, but in other cases, they will not. For example, if V1 leaks slightly but V2 does not, V would be exposed to the differential pressure I loading to which V is normally exposed. In this situation, V1 would have RCS pressure on both sides of the disc and would be expected to have a lower failure rate than'V ,2 which is exposed to a greater differential pressure. Thus, Equation (3.1) could be written as As =A(V)*P(V!Y)*(1-P)+A'(V)*P'(V!v)*P 1 2 1 y 1 2 1 y
+A(V)*P(VlV)*(1-P)+A'(V)*P'(VlV)*P (3.2) 2 1 2 I 2 1 2 I where PI = the probability that the space between valves is pressurized to RCS pressure.
l A'(VI ) = the frequency of a random, independent failure of VI, given that the space between valves is I pressurized (events per hour). j P'(V2lV1 ) = the conditional probability that V2 fails, given
- that Vi has failed and the space between valves is l pressurized.
= the frequency of a random, independent failure j
I A'(V2) of V2, given that the space between valves is pressurized. P'(VlV)=theconditionalprobabilitythatV1 1 2 fails, iven l that V2 has failed and the space between va ves is pressurized. I 1319P112285 3-13
I On the basis of the loadings across the valve discs, the following assumptions appear to be reasonable for the lines that contain the check valves.*
- 1. A'(V2 ) = A(V1 )-
- 2. A'(V; is small compared to A(V ).
- 3. A(V2 ?)is small compared to A'(V ).
- 4. P'(VilV2) = P(V2lV1 )-
SubstitutingforA'(V)andP'(VlV) 2 i 2 As =A(V)*P(VlV)*(1-P)+A'(V)*P'(V!v)*P 1 2 1 y 1 2 1 g (3.3)
+ A(V2 )*P(Vg lV2 )*(1-Py ) + A(Vg )*P(V2lV1)*P g or l
A s =A(V)*P(V!v)+A'(v)*P'(V!v)*P; g 2 1 2 1 (3.4) I 1
+A(V)*P(VlV)*(1-P) 2 i 2 y The third term in Equation (3.4) is small compared to the first; therefore As=A(V)*P(VlV)+A'(V)*P'(VlV)*PI 2 1 2 1 (3.5)
I 1 1 As a conservative upper bound, it can be argued that As*A(V)*P(VlV)*(1+P) 1 2 1 I (3.6) Because only a minute amount of leakage is required to pressurize the space between valves, it is assumed that PI approaches 1.0. Therefore As = 2*A(V1 )*P(V2lV1 ) ( 3.7 ) Given that Vi has failed independently, V2 could fail upon demand (due to the sudden pressure challenge), or it may fail randomly in time, sometime after failure of V .1 The latter failure mode is represented by the standby redundant system model used in the SSPSA. Equation (3.7) g conservatively reflects the potential for discovery of the outboard valve 5 rupture before the next testing opportunity of the inboard valve because of the ability to alarm and indicate this condition to the operator via a the accumulator pressure sensors. g
- An additional failure mode is considered in the line containing the i
motor-operated valves; i.e., disc open and indicating closed. I 1319P112285
I The Seabrook RHR cold leg injection and hot leg suction path arrangements are shown in Figures 3-1 and 3-2, respectively. In the SSPSA, the following three V-sequence events were identified and quantified:
- 1. Disc rupture of the check valves in the cold leg injection lines of the RHR.
- 2. Disc rupture of two series motor-operated valves in the RHR hot leg suction.
- 3. Disc rupture of the M0V equipped with a stem-mounted limit switch and
" disc failing open while indicated closed" in the other motor-operated valve in the normal RHR hot leg suction.
The hot leg injection lines are isolated from the RCS by two normally closed check valves and one normally closed MOV in each line. All three valves must fail simultaneously in the open position in order to expose I the remainder of the system to the RCS pressure. Simultaneous failure of three valves is an even more unlikely scenario and was judged to have a frequency that is insignificant in relation to the analyzed two-valve cases. The following valve failure modes were excluded from the analysis for the following reasons: e "The disc failing open" failure modes for the check valves of the injection lines are excluded because these check valves are leak I tested after RCS depressurization to ensure disc seating. The testing occurs after system use in the RHR mode prior to reactor startup on an average of three times per year, e "The disc failing open while indicating closed" failure mode due to disengagement of the gear drive from the valve stem is excluded for I valves RC-V22 and RC-V87 because they are equipped with stem-mounted limit switches, thus ensuring accurate indication of disc position. The disc rupture failure mode has not been reported in the nuclear I industry data base. Accordingly, the initiating event frequency considered in this study I addresses the frequency of exceeding certain leakage through the valves, based on available data. A review of pertinent data has been conducted and is discussed in Section 3.1.2.2.1. The probability distribution i resulting from the review is shown in Figure 3-3. Nondestructive testing I and inspection of valve components by using magnetic particle, ultrasonic, and dye penetrant techniques should disclose flaws of critical size. It is extremely unlikely that flaws smaller than the critical size can propagate catastrophic failures. In addition, the check valves will be inspected each time the plant goes to cold shutdown (assumed once a year), and the MOVs will be inspected at each refueling shutdown (every 1.5 years). Even though the disc rupture mode of failure is extremely unlikely, as I shown in the following discussion, it is implicitly included in the check I 1319P112285 3-15
l I valve leak data shown in Figure 3-3. The injection line check valves are in a 6-inch diameter line. The flow of subcooled water through the valve would yield an initial flow rate of approximately 64,000 gpm (assuming an upstream pressure of 2,250 psia, a discharge coefficient of 1.0, and a density of 40 lbm/ft 3). The best estimate frequency of exceeding such a leak pa t one valve (as determined from Figure 3-3) is less than 5.3 x 10- per year compared to the mean disc rupture failure rate of g 1.4 x 10- per year that was used in the SSPSA study. 3 In this study, it is assumed that failure of the two valves in series may cause a high pressure challenge to the RHR system piping and components, but not necessarily failure of these components. Of particular importance is the frequency of check valve and M0V leakage (past both series valves) that exceeds the capacity of a charging pump and also exceeds the total capacity of the RHR relief valve system (which is at least 1,800 gpm at the 450-psig setpoint pressure for the two suction side relief valves). Assuming that both RilR relief valves lift,* only those valve failures that result in leaks larger than the capacity of the relief valves challenge the integrity of the RHR system piping and components. Leaks that are larger than the capacity of the charging system will depressurize the RCS to the containment and will take on the characteristics of a small or medium LOCA when the pressurizer relief tank fails. Leaks smaller than the capacity of the charging system will not depressurize the RCS over the short term, but the operator will be alerted to the event by the charging pump operation, pressurizer relief tank conditions, and potential high pressure and temperature in the containment. This is because the normally running charging pumps would maintain coolant inventory control. For the purposes of this study, scenarios in which the leakage past the I series valves is less than 150 gpm are essentially " nonevents." Thus, the initiating event frequencies, VI (injection) and VS (suction), refer to the frequency of exceeding a leakage of 150 gpm. This treatment is consistent with the way in which other LOCA-type events were handled in the SSPSA. 3.1.2.2.1 Valve Failure Data Analysis The failure modes of interest are (1) disc rupture or gross leakage of a seated check valve on the pressure boundary and (2) failure of the check valve downstream of the first check valve to hold, given the failure of the first check valve. The parameter associated with the first failure mode is the rate of failure per hour, but for the secord mode, a frequency of failure on g demand is needed, g
*The mean frequency of single relief valve failing to open on demand is I
approximately 2.4 x 10-1319P112285 .
E To estimate the rate of the first failure mode, all check valve failure r events in the U.S. LWRs, as reported in Nuclear Power Experience L (Reference 3-7), were reviewed. NPE is an LER-based compilation of failure events; the data base of this study covered the period 1972 through 1984 Among the several hundred check valve failures identified as the result of the data search, only those associated with PWR, ECCS, [ and RCS were considered the most relevant for the valves considered here, which are initially seated and testable. No disc rupture event was E, identified, and the maximum leak rate observed was 200 gpm. Due to L ambiguity and lack of details in the event descriptions, the actual leak rate was not available for a large number of events. In those cases, the leak rates were estimated by considering other indirect indications, such as pressure reduction or the rate of change in boron concentration and similarity to other occurrences for which the leak rates were known. In general, the leak rates were estimated conservatively and grouped into several categories by leak size. { A recent review of eight BWR events (Reference 3-5) that.could be E considered as precursors to an interfacing LOCA indicated that a testable L check valve was involved in each of the events. Five of these eight events were associated with " interference by the attached air operator," and it was recommended that the nonsafety-related air operator be disabled to prevent future occurrences. All eight events were considered in the analysis leading to the development of Table 3-8 and Figure 3-3. However, those events that were directly associated with the air operator or that would have been detected during leak testing after refueling were L judged to be inappropriate for inclusion in this study. 7 The leakage events used in developing the failure frequency are L summarized in Table 3-8. Table 3-9 lists the number of events in various leak-rate categories. Table 3-9 also provides the estimated frequency per hour of check valve leakage events for each of the leak-rate [ categories. All together, 21 events were identified with leak rates ranging from less than 5 to 200 gpm. It can be argued that many of these 21 events, for example the several involving accumulator check valves, do not fully represent conditions of the check valve in the SI and RHR lines. The inclusion of these events is therefore viewed as a conservative assumption. A total exposure time of about 1 x 108 check valve hours was estimated by counting the number of check valves in all power plants in the data _ base of the RCS and ECCS systems (Reference 3-8). Finally, the data points in Table 3-9 (frequency of exceedance) were plotted against the leak rate, and a best line fit was obtained by using Bayesian regression techniques (Reference 3-9). This line and the calculated bounds at 90% confidence are shown in Figure 3-3. The bounds represent only the statistical uncertainty associated with the data presented in Table 3-9. To account for the uncertainty stemming from estimation of the leak rates, the uncertainty range was subjectively increased to a factor of 10. These bounds were.then used as the 95th and 5th percentiles of a lognormal distribution representing the overall uncertainty. I 1319P112285 3-17
I No data applicable to the second failure mode were found, and the frequency of fail to operate on demand for check valves was used in the l analysis. The distribution of this frequency has the following characteristics (Reference 3-10, Table 6.2-1). Parameter Frequency (events per demand) , Mean 2.7 x 10-4 I Sth Percentile 5.3 x 10-5 1 95th Percentile 6.3 x 10-4 Median 1.4 x 10-4 3.1.2.2.2 Injection Line Frequency Based on Figure 3-3, the median frequency of a single check valve failure resulting in leakage that exceeds the capagity of one charging pump (i.e.,150 gpm) is approximately 1.7 x 10-o per hour. Assuming a lognormal distribution for this frequency and a range factor of 10 yields: I Parameter Frequency (events per reactor-year) 95th Percentile 1.5 x 10-3 Mean 4.0 x 10-4 E Median 1.5 x 10-4 E Sth Percentile 1.5 x 10-5 I Similarly, the median frequency of exceeding 1,800 gpm is 2.3 x 10-9 per hour. Assuming a lognormal distribution with a range factor of 14 yields: I Parameter Frequency (events per reactor-year) 95th Percentile 2.8 x 10-4 I Mean 7.3 x 10-5 Median 2.0 x 10-5 5th Percentile 1.4 x 10-6 I 3-18 1319P112285
-. __ )
I ThetermP(VlV)inEcuation(3.7)containstwocomponents: 2 1 one representing random failures of the second valve, given that the first valve has failed, and the second representing a demand failure at the time the first valve failed. As shown in Section 6.6 of the SSPSA (Reference 3-10), the determination of the frequency of occurrence of random failures is facilitated by I assuming that the two series check valves in each path represent a standby redundant system, and failure of the downstream check valve cannot occur until failure of the check valve nearest to the reactor I coolant system loop has occurred. The probability of random failure (unreliability) for a single injection path is given by Qpath a 1 . e-At (1 + At) (3.8) , where A is the appropriate failure rate of a single check valve. In this study, A is the frequency of exceeding leakages of 150 gpm. This I expression was then used to derive a failure (or hazard) rate for the path. That is, I A path (t) = (y ,1 9 h[1-Qpath] (3.9) I or A path (t) = (3.10) 3 As noted earlier, the plant is expected to go to cold shutdown once a year at which time these valves will be inspected. If it is determined that the system is not functioning, it is repaired at that time. I Therefore, the time-dependent failure rate is bounded at 1 year. The average failure rate over a time period, T, is given by T 1 Adt
#A path per reactor year *
- T 1_,
0 (1, At) (3.11) 1
= T [AT - s.n (1 + AT)]
When AT << 1, this result can be expanded to obtain
<Apath*
- A (* }
The demand component of the path failure frequency is merely the product of A and the demand failure rate, Ad . Thus, <A path> I I 1319P112285 3-19
I calculated for the SSPSA can be expanded as follows to account for the demand failure.
<A pa h> " A E +A] d (3.13)
Finally, the above expression for <1 oath > is multiplied by a factor of 2 to account for the logic used ih developing Equation (3.7). This logic is that the two valves can fail in either sequence because of an assumed high likelihood of inboard valve leakage and pressurization of the space between valves. Thus, the final expression for the series valves in the injection lines is
<A path >=2A[f+A] d (3.14)
As an upper bound, the check valve fail to operate on demand data given E in Table 6.2-1 of Reference 3-10 and summarized in Section 3.1.2.2.1 will 5 be used to estimate <Anath> for the injection lines. The estimated frequency of failure of two series injection check valves that produces 3 leakage to the RHR system in excess of 150 gpm is 3 Parameter Frequency (events per reactor-year) I 95th Percentile 3.3 x 10-6 Mean 1.1 x 10-6 Median 9.0 x 10-8 g 5th Percentile 5.2 x 10-9 l Since there are four injection paths, the distribution for VI is I Parameter Frequency (events per reactor-year) 95th Percentile 1.3 x 10-5 I Mean 4.4 x 10-6 Median 3.6 x 10-7 Sth Percentile 2.1 x 10-8 Top Event LR in the injection path event tree represents the fraction of the initiating event frequency, VI, in which the leakage not only exceeds 150 gpm, but also exceeds 1,800 gpm. The product of LR and VI thus represents the frequency of pressure challenges to the RHR system due to I 1319P120285
I I failures of both check valves in the four injection paths. Based on the above distributions, LR has a mean value of .093. 3.1.2.2.3 Suction Line Frequency For a V-sequence to occur in the RHR hot leg suction paths, failure of two series MOVs must occur. Given that such failures occur, the low I pressure piping and the RHR system components downstream of M0V RC-V23 or MOV RC-V88 would be exposed to RCS pressure if the failure results in leakage in excess of the relief valve capacity (1,800 gpm). Three cutsets of valve failures that lead to a suction line V-sequence are possible. The first cutset involves independent failures of both MOV I valves, causing excessive leakage. The second cutset involves independent failure of one of the valves and a demand failure of the second valve. These two cutsets are similar to those for the series check valves in the injection lines. The third cutset involves gear I drive disengagement for the first M0V, which is not equipped with a stem-mounted limit switch, causing it to fail open while indicating closed, and a random failure of the second MOV, which results in I excessive leakage past the disc. Thus, the equation for <A Dath> for the suction path has an additional term to account for the third cutset:
#A path > " 2A [ + Ad3 + A* A g (3.15) where Ag = the frequency of failure of an MOV to c';se on demand and indicate closed.
T = the interval between refueling shutdowns (13,140 hours). The third term (or cutset) in Equation (3.15) is not multiplied by 2 since one M0V has the stem-mounted limit switch, which precludes the mode of failure postulated for the other M0V. I The frequency of MOV valve disc leakage and failure upon demand (due to a i sudden pressure loading) were conservatively assumed to be identical to , E that for the check valves. The frequency of failure of an MOV to close ! E on demand and indicate closed [i.e., Ag in Equation (3.15)] was obtained from data reported in Nuclear Power Experience. The distribution for A g is (see Table 6.2-1 of Reference 3-10): Parameter Frequency of Failure on Demand I 95th Percentile Mean Median 3.1 x 10-4 1.1 x 10-4 7.5 x 10-5 5th Percentile 2.1 x 10-5 I I 1319P112285 3-21
I This distribution, as well as those for valve disc leakage and failure on demand, was propagated through Equation (3.15) to obtain the following distributicn for the frequency of a single suction line V-sequence: I Parameter Frequency (events per reactor-year) 95th Percentile 4.7 x 10-6 Mean 1.6 x 10-6 Median 1.2 x 10-7 5th Percentile 7.0 x 10-9 I Since there are two such paths, the total suction side V-sequence g frequency, VS, is given by the following distribution: g Parameter Frequency (events per reactor-year) 95th Percentile 9.4 x 10-6 Mean 3.2 x 10-6 Median 2.4 x 10-7 Sth Percentile 1.4 x 10-8 The split fraction LR for the fraction of V5 in which the leakage past the series MOVs is greater than the capacity of the relief valves is 0.09. 3.1.3 EVENT TREE MODEL Except for the isolation capability inherent in the cold leg injection lines, the characteristics of the V-sequence are nearly identical for the suction lines. Thus, one event tree was initially developed for both types of lines. Differences between the injection and suction paths are l accountec' for by small differences in the event tree structure and end ! states and by assuming different event tree split fractions appropriate j for each line type. The event trees for the injection path and suction path sequences are shown in Figures 3-4 and 3-5, respectively. A leakage rate up to 150 gpm (equivalent to approximately 0.1 square inches at normal RCS pressure) is within the capacity of the charging, letdown, and seal water system. Normal charging is provided by one of three charging pumps taking suction from the 4,700-gallon volume control tank. If the level in the volume control tank decreases to an emergency low setting, a signal is provided to open the RWST suction valves to the 3 charging pumps and to close the stop valves in the volume control tank 5 outlet line. This action makes 350,000 gallons of water available from the RWST. Even if the leakage remained at 150 gpm, it would take approximately 40 hours to exhaust the available water. Thus, any 1319P112285 3-22 l
I I postulated failures involving leak rates of 150 gpm or less are assumed to be " nonevents." l I The combined capacity of the RHR system relief valves is more than 1,800 gpm. Thus, V-sequences that lift these valves, but do not cause any other failures in the low pressure RHR piping, will release primary coolant to the containment via the pressurizer relief tank. A small I amount of coolant could be released to the primary drain tank via the two injection line relief valves. Although emergency core cooling will be required, normal means should be available to provide this function, if a I leak less than 1,800 gpm occurs. Hence, additional system or human failures would be required for these sequences to result in core melt. I Therefore, the sequences of interest are those (valve leaks
> 1,800 gpm) that pressurize the RHR system and challenge low pressure piping in that system.
Some V-sequences have an inherent potential for mitigating the release of fission products due to flooding of the vault if the leak occurs in the vault at a level below that of the door to the RCA tunnel I (Elevation O'). The two RHR vault sump pumps have a combined capacity of only 50 gpm. Thus, their impact on depleting the water inventory in the RHR vaults for leaks significantly larger than .09 square inches will be I negligible. Furthermore, it appears that the sump pump motors would be flooded shortly before the RHR pump seals would become covered with water. The sump pumps will fail when their motors are submerged. If the event is diagnosed by the operator, there are a number of possible actions that might be taken to mitigate the interfacing LOCA. I 1. Motor-operated isolation valves RH-V14 (train A) and RH-V26 (train B) can be closed to isolate the 8-inch diameter injection lines. The RWST stop valves (CBS-V2 for train A or CBS-V5 for train B) can I 2. be closed to prevent the loss of RWST inventory.
- 3. Motor-operated crosstie valves RH-V22 or RH-V21 could be closed to I allow an intact RHR train to operate successfully.
None of the above actions can terminate a leak caused by failures of the RHR suction line valves. The potential exists for failure of the low pressure RHR system inside the containment, in the pipe tunnel, or in the RHR vaults. Failures of I the suction line inside the containment would behave like a LOCA with at least one train of low pressure injection and RHR unavailable. Failures of the injection lines within the containment are less likely because I these lines are already designed for RCS pressure. It is not clear whether failures of any of the RHR piping in the pipe tunnel would be mitigated by flooding, but such failures are predicted to be of very low I frequency. Failures of piping and components located at lower elevations in the RHR vaults will produce sufficient flooding to cover the leak, producing a suppression pool-like fission product scrubbing effect. I 3-23 1319P112285
I In addition to those considerations discussed earlier, two other important considerations emerge for the V-sequence. The first of these relates to the potential loss of a source of water once the RWST is depleted. The obvious solution to this problem area is to provide makeup to the RWST to either extend the initial injection period or to provide a suction source for the charging pumps following unsuccessful recirculation. Alternatively, an external recirculation path could be established using the water in the vault as a source. This approach uses portable pumps and although possible, was not addressed in detail in this study. The second consideration is contingent on the interfacing LOCA being located in the RHR vault at an elevation higher than that required to get significant scrubbing due to flooding. If such a leak has occurred, the configuration of the vault is such that the leaking primary coolant itself will flood the vault. External sources for flooding the vault could also be employed. 3.1.3.1 Top Event Descriptions A total of 14 top events is used in the V-sequence event tree. These top events define the relative size of the valve leak, whether or not RHR relief valves open, whether or not the RHR piping survives the pressure challenge, the size of the RHR seal leak if such a leak occurs, operator actions to diagnose and mitigate the event, and survivability of the pumps located in the vault. A detailed description of each of the top events, including the initiating events for each tree, is given below, e Initiating Event VI. Models all important modes of failure of the g two series check valves in the four RHR cold leg injection paths that 'g could result in leakage to the low pressure portion of the RHR system that exceeds the capacity for a single charging pump (150 gpm). e Initiating Event VS. Models all important modes of failure of the two normally closed M0V valves in each of the two RHR suction lines that could result in leakage that exceeds the capacity of a single charging pump (150 gpm). e Top Event LR. Asks whether the leak created by either an injection g or a suction-line valve failure is within the capacity of the g appropriate RHR relief valves (assumed to be 1,800 gpm at 450 psig). Sequences in which the arswer to the question posed by this top event is yes are not interfacing system LOCAs. e Top Event V0. Asks whether the two large, suction-side RHR system relief valves lift on demand. Failure of these valves to lift is assumed to overpressurize the RHR system in a manner equivalent to the SSPSA V-sequence. e Top Event Pl. Asks whether the high pressure challenge to the RHR system causes either the piping or the heat exchanger to fail. Failure of the piping or heat exchanger is assumed to result in a plant damage state similar to 1F, which had been used in the SSPSA E for characterizing the V-sequence. To account for new source terms 5 uniquely appropriate to the V-sequence, these sequences are assigned to 1FV. 3-24 1319P112285
g e Top Event SI. Given that neither the piping nor the heat exchangers E fail, this top event asks whether the RHR pump seals remain leak free at RCS pressure. e Top Event L1. Given that the RHR pump seals have failed, this top event asks whether the total leak area of both pump seals is less than or equal to .09 square inches. This leak area produces leakage of approximately 50 gpm per pump, which equals the combined capacity of the two sump pumps installed in each sump. Since the sump pumps prevent flooding of the vault, success of LI negates failure of the pumps in the vault because of submersion. e Top Event L2. Asks whether the total RHR pump seal leak area lies in the range of 0.09 to 1.05 square inches. The lower bound of this I range corresponds to the upper bound of L1, The upper bound of L2 represents a leak area that produces a leak flow of 150 gpm at an RCS pressure slightly below the suction-side relief valve pressure of I 450 psig. Makeup to the RWST appears to be limited to approximately 150 gpm. Top Event L3. Asks whether the total RHR pump seal leak area lies in I e the range of 1.05 to 2.6 square inches. The lower bound of this range corresponds to the upper bound of L2, while 2.6 square inches corresponds to the maximum total flow area (two RHR pumps) expected for pump seal failure. Failure of L3 guarantees failure of all pumps located in the vault. I e Top Event 01. Models the ability of the control room operators to diagnose the V-sequence event. Failure of 01 is assumed to guarantee failure of 02. Failure probabilities used to quantify the event trees are based on emergency procedures that are updated to reflect I this consideration in Section 3.1.4.3. e Top Event 02. Models the ability of the operators to terminate the I leak past the failed series valves and isolate the RHR train that is involved in a timely manner. Termination of the leak will not be possible if the suction-side MOVs have failed; therefore, 02 is I always failed for VS. In addition, failure of this event is assumed to be guaranteed if the operator fails to diagnose the interfacing systems LOCA. e Top Event CS. Models the ability of the containment spray pumps to survive the hostile vault environment that is created by the RHR pump seal failure. Containment spray pump failure is assumed if 02 fails and the seal leak is greater than 0.09 square inches or if the seal leak is greater than 1.05 square inches. e Top Event RS. Models the ability of the RHR pumps to survive the hostile vault environment created by failure of their seals. RHR pump failure is assumed if 02 fails and the seal leak is greater than 0.09 square inches or if the seal leak is greater than 1.05 square I inches. 3-25 I 1319P112285
I e Top Event SS. Models the ability of the high pressure safety injection pumps to survive the hostile vault environment created by failure of the RHR pump seals. Safety injection pumps are assumed to fail if 02 fails and the seal leak is greater than 0.09 square inches E or if the seal leak is greater than 1.05 square inches. E e Top Event VC. Asks whether or not the RHR system relief valves seat properly once RCS pressure falls below their setpoint pressure. These valves are expected to cycle open and closed numerous times during the course of the accident. Failure of these valves to close will depressurize the RCS more quickly. e Top Event 03. Models the ability of the operator to follow Emergency Procedure ECA-1.1 to restore emergency coolant recirculation E capability, to delay depletion of the RWST by adding makeup and W reducing drawdown, and to depressurize the RCS to minimize flow through the failed pump seals and through the RHR relief valves to the containment via the PRT. Failure of the CBS, RHR, and high pressure safety injection pumps is assumed to be guaranteed if their motors become submerged. It was also assumed that seal failure on both RHR pumps is much more likely than having the seal on only one pump fail; therefore, only the former was considered in this analysis. Existing emergency procedures do not guarantee that the operator will initiate the steps taken in ECA-1.2 since the hierarchy of operator g actions and accident conditions may circumvent the latter procedure g altogether. The quantification of this tree assumes that revised proce/ures are in place. The valt as used for the top event split fractions in quantifying the event tri.'s are summarized in Table 3-5. The basis for these split fractions 's provided in Section 3.1.4 3.1.3.2 Evert Tree Structure and End States Application of ti.' split fractions summarized in Table 3-5 results in the event trees for VI md VS shown in Figures 3-4 and 3-5, respectively. Except that the operCor cannot terminate the interfacing LOCA (i.e., 02 is always 1.0) when the. suction MOVs are postulated to fail, the trees E are very similar in struc. are. Hence, only the VI tree will be discussed 5 in detail. As indicated in both Figures 3-4 and 3-5, event tree sequences were mapped to one of the following: e An SSPSA plant damage state (70, 8C). l e New plant damage states (1FV, IFPV, or 7FPV), whose release characteristics are similar to their SSPSA counterparts (1F, IFP, E and 7FP), have been given the suffix, V, to indicate that their 3 respective source terms were specifically calculated for this study. I 3-26 ' 1319P112285
e The small or medium LOCA event trees (simply referred to as LOCA). e Success states DLOC or DILOC for which core makeup is being provided by the charging pumps. DLOC refers to the cases where the interfacing LOCA has been terminated, while DILOC refers to cases where it has not been terminated. In the SSPSA, plant damage state 70 represented sequences with delayed c're melting (longer than 6.0 hours), a high RCS pressure at vessel me:t-through, and a cry reactor cavity; i.e., no spray or ECCS I i nj e.r.ti on . Plant damage state 8C represents accident sequences where the RCS prassure remains relatively high until vessel failure, core melt occurs iJte (longer than 6.0 hours), the RWST is injected, and containment heat removal is unavailable. SSPSA plant damage state IF (or I 1FV in this 3tudy) represents accident sequences in which the containment is failed or bypassed from the inception of the accident. Core melt occurs early with a low RCS pressure and a dry cavity at vessel I mel t-th rough . SSPSA plant damage state IF was dominated by the V-sequence. Plant damage state IFP was similar to IF except for the size of the bypass. Bypasses with larger diameters than an equivalent I 3.0 inches were mapped to 1F, while those with smaller diameters than this were mapped to 1FP. SSPSA plant damage state 7FP is characterized by high RCS pressure and a dry cavity at vessel melt-through in addition to an unisolated containment with an equivalent diameter smaller thar. I 3.0 inches. Core melt occurs late (more than 6.0 hours). Sequence 1 in Figure 3-4 represents the sequence in which leakage past the series check valves in the injection lines exceeds the capacity of one charging pump, but does not exceed the RHR system relief valve capacity and therefore does not threaten the integrity of the RHR system unless the relief valves fail to open. Accordingly, this sequence is mapped to the LOCA end states. Sequence 2 represents the case in which the relief valves fail (V0). Since it was known a priori that this sequence would have a very low frequency, due to the relatively high reliability of the relief valves, sequence 2 was conservatively mapped to end state 1FV. For sequences 3 through 71, the leakage past the series check valves in the injection lines is greater than the capacity of the RHR relief valves, and a challenge to the integrity of the RHR system is presented. Sequentes 70 and 71 were also mapped to plant damage state 1FV. These sequences represent cases in which either the RHR system relief valves fail to open (similar to sequence 2) or failure of the RHR piping or heat exchanger (PI) occurs. Sequence 3 assumes the RHR pump seals survive the pressure challenge to the system if no other piping failures have occurred. Since the interfacing LOCA is precluded in this sequence, it is mapped to the LOCA end state. Sequences 4 through 42 represent RHR pump seal failures in which the total leakage area lies within the range of 0.0 to 0.09 square inches. For this range of failures, failure of the CBS, RHR, and SI pumps due to submergence is precluded becaust. the capacity of the vault sump pumps is greater than the leak flow rate. Failure of these pumps due to the humid environment in the vault, or as a direct result of the RHR pump seal leak, is possible. Sequences 4, 5, and 6 were mapped to DLOC because, in each 3-27 1319P120285
s case, the operator was able to terminate the interfacing systems LOCA (W) and either the RHR pumps were available or the operator was able to supply makeup to the RWST (U) for those sequences in which the RHR pumps fail. Because the frequency for DLOC was expected to be small, subclasses for various combinations of available pumps were not identified. Sequence 7 represents the case in which both the RHR pumps have failed (RS) and the operator fails to provide makeup'to the RWST (03). This sequence is mapped to PDS 8C since the CBS pumps did not fail in this sequence. Sequences 8 through 11 are similar to 4 through 7 except the CBS pumps have failed; therefore, sequence 11 was mapped to PDS 7D rather than 8C. Sequences 12 through 30 represent those in which the operator has diagnosed the interfacing systems LOCA event ('0T) bLt.is-unsuccessful in terminating the bypass (02). Thus, although capability to provide makeup to the core may be available (e.g., sequences 12,14, ; 16, 18, 20, 22, 24, 26, and 28), the sequence is mapped to end state DILOC to indicate that the interfacing systems LOCA +as not terminated. Since the interfacing systems LOCA is not terminated, makeup to the RWST (M) must be provided to prevent core melt. If makeup is not provided (03), the sequences in this grouping are mapped to either end state 7FPV or 1FPV, depending on whether or not RCS pressure was expected to be high or low at the time of reactor vessel failure. In general, it was assumed that RCS pressure was low at the time of failure if RUST makeup was not provided (03) and the RHR system relief valves failed to close (VC) but that it was high if the relief valves closed (W). Sequences 31 to 42 represent scenarios in which the operator fails to diagnose the event (01). For such scenarios, failure of the operator to terminate the event is guaranteed (i.e., 02 = 1). Mapping to end states for sequences 31 to 42 follows the same prescription as that discussed above. Sequences 43 to 55 represent scenarios where' the RHR pump seal leak lies in the range of 0.09 to 1.05 square inches, and sequences 56 through 64 represent leaks of 1.05 to 2.6 square inches. Sequences 65 through 69 represent seal leaks greater than 2.6 square inches. The same prescription for mapping sequences to end states, as discussed above for sequences 4 to 42, is followed for these sequences as well. 3.1.4 EVENT TREE QUANTIFICATION 3 3.1.4.1 RHR Piping and Heat Exchanger Strength The entire RHR system is classified as Nuclear Safety Class. RHR system g piping and heat exchangers are designed to ASME Section III standards. gs Design pressure and temperature of the rystem are based on normal operation some 4 hours after reactor shutdown when temperature and a pressure in the RCS are 350*F and 425 psig, respectively. g The RHR system has a design pressure of 600 psig. The system piping is composed of Schedule 405, Type 304 stainless steel piping. Low pressure piping in the suction lines is as large as 16 inches in diameter, but the maximum pipe diameter in the injection path is 8 inches. The 3/4-inch tubing in the RHR heat exchanger is 18-gauge, SA-241, Type 304 stainless steel. The tubing is designed for 600 psig and for a temperature of 400*F. The shell of the heat exchanger la made from j 3-28 1319P120285
I I carbon steel and is designed for a pressure and temperature of 150 psia and 200 F, respectively. The IDCOR analysis concluded that dynamic effects of accidental pressurization of the RHR system were not an important consideration and, therefore, only quasi-static pressurization to RCS pressure needs to be addressed. This conclusion was based on dynamic evaluations performed for both the case in which the low pressure segment of the piping was full of water at a static pressure equal to the head of water in the RWST and for the case in which it is assumed that a gas void exists in the low pressure piping, thus providing a " pocket" for acceleration of the liquid during the pressurization of the systen. These cases were discussed in Appendices A I and B, respectively, of Reference 3-4. For the case in which the low pressure portion of the system is filled with water, the assumption of a catastrophic failure of the valves separating the high and low pressure segments of the system results in the propagation of a compression wave into the low pressure region and a rarefaction wave into the high pressure region. The velocity increment (AU) for the waves can be expressed by AU = 0 pC g where AP = the pressure change across the wave, p g = the density of the water. C = the acoustic velocity in the water. Conservation of mass requires that the compression and rarefaction wave velocities be equal. Therefore PRCS - Pg,Pg-PLP PgG pG g where PRCS = the initial RCS pressure. PI = an intermediate pressure for the wave propagation. PLP = the initial pressure in the low pressure segment of the piping. [ g For a PWR with an initial RCS pressure of 2,250 psia, the intermediate l g pressure would be approximately one-half this value, or 1,125 psia. If a l I 1319P120285 3-29
I wave of this magnitude were to encounter a solid wall (such as a closed g valve), it would be " reflected in a like sense" with essentially twice N the value of the incident wave. Thus, the pressure behind the reflected wave would be equal to the initial RCS pressure. In reality, friction and form losses diminish the stren'gth of these traveling waves. When the compression wave reaches the first elbow in the low pressure RHR system piping, it undergoes a complex process in which some of the wave energy is transmitted through the elbow and continues downstream and some is reflected back upstream into the piping already at the intermediate pressure. The solid wall case serves as an upper bound for the reflection process; thus, the pressure cannot exceed the initial RCS pressure. The IDCOR analysis indicated that if the volume of dissolved gases in the a system was insufficient to provide a gas volume that could occupy the g total cross-sectional flow area at the expected location of the void, a pressure wave traveling into the expected two-phase (i.e., stratified) zone would experience a mixture compressibility with an acoustic velocity equal to that of the gas. A compression wave encountering this highly compressible media would be substantially attenuated. Thus, a gas void in the low pressure segment of the piping would significantly reduce the 3 import of the hydrodynamic transient. E To our knowledge, the IDCOR evaluations discussed above have not been refuted by the NRC. It should also be noted that the preliminary case study report on overpressurization of ECCS in BWRs (Reference 3-5) did not address the dynamic effects of low pressure system overpressurization. At the normal RCS pressure of 2,250 psia, the h00p stress in the larger RHR piping approaches the yield strength of the piping material of about 35,000 psi. The probability of pipe failure at 2,250 psia was estimated from the distribution shown in Figure 3-6. The failure probability of the piping was assumed to be the combination of a flat distribution that accounts for undetected design errors, material defects, and inspection oversights and a losnormal distribution for which it was assumed that the probability of failure at the yield strength of the material is .01 and the probability of failure at the ultimate strength of the material is .99. A conservative value of 10-* was assumed for the pressure independent or " flat" distribution component of the fragility curve. Based on these estimates, the probability of p failure at a pressure of 2,250 psia is estimated to be about 6 x 10 gpe . g Valve bodies and other components in the RHR system are expected to have g similar design margins. The above approach is compatible with the approach followed in the SSPSA to assess the containment pressure capacity. Calculations similar to those performed for the RHR system piping indicate that the RHR heat exchanger tubing will have even greater margins against failure. 3.1.4.2 RHR Pump Seal Failure Area A review of the RHR pump seal design indicates three representative scenarios following an overpressurization event: (1) the seals remain I 3-30 1319P112285
intact and excessive leakage develops; (2) the seals are blown out, but the mechanical seal assembly remains in place; and (3) the entire mechanical seal assembly is blown out. To completely blow out the mechanical seal assembly, the four 3/4-inch studs that hold the seal cover plate to the top of the pump casing must fail. Using a pressure of 2,250 psia, the load on the seal assembly can be calculated as that acting on the motor shaft, shaft sleeve, and 0-ring. This results in a diameter of approximately 4.4 inches, a load area of approximately 15.5 square inches, and a corresponding load of approximately 35,000 lbf. The average stress in each of the four 3/4-inch bolts is approximately 19,800 psi. The bolts are made of ASME SA453 Grade 660 material, which has a tensile strength in excess of 50,000 psi . Thus, failure of the bolts under these conditions is extremely unlikely. The equivalent break area for this f ailure, taken to be the annular ring between the pump casing and the motor shaft, would be approximately 6.5 square inches per pump. The RHR pump seal failure area, characterized by the destruction of the seal assembly bushings, seals, 0-rings, etc., would be characterized by the clearance between the mechanical seal assembly and the motor shaft. This distance is estimated to be 0.12 inches. This yields an area of approximately 1.3 square inches per pump. The corresponding hydraulic diameter (i.e., 4 times the cross-sectional flow area, divided by the wetted perimeter) is 0.235 inches, the characteristic length is 1 inch, and the length to diameter ratio is approximately 4. With the pump seals in place, the clearance between the shaft and the seals is estimated to be less than 1/32 inch. If the clearance were exactly 1/32 of an inch, the corresponding flow area would be about
.17 square inches per pump. The hydraulic diameter is approximately 0.063 inches, and the characteristic length is 7.0 inches. For this case, the length to diameter ratio is approximately 112.
Given that no other failures of the RHR system have occurred, it appears that seal failure due to exposure to high temperature coolant is very a unlikely. Such exposure could result if RCS water passed through the g pumps as it travels to the relief valves. However, for failures of the suction line M0Vs, the relief valves are upstream of the pumps. Since each pump discharge contains a check valve to prevent backflow through the pumps, primary coolant will bypass the RHR pump via the miniflow line for injection line valve failures. To account for uncertainty in predicting the size of the RHR punp seal leak, the event trees were quantified for each of the following total leak areas (i.e., the sum of the leak areas for both pumps): , RHR Pump Seal Estimated Probability l Leak Area of Occurrence (square inches) per Challenge l 0.0 .01 l 0.0 to 0.09 .08 l .09+ to 1.05 .4 1.05+ to 2.6 .5
>2.6 .01 3-31 1319P121685
Figure 3-7 depicts the flow rate as a function of pressure difference for representative areas. The intermediate break point of 0.09 square inches represents a leak area that produces a leak flow of 50 gpm per RHR pump at RCS pressure. This flow rate represents the capacity of the sump pumps in each RHR/CBS vault. The selection of a break point of 1.05 square inches is somewhat arbitrary; however, it is roughly one-half of the total maximum expected leak area of 2.6 square inches (or equal to that which would be obtained if the seals on one of the RHR pumps survived and the other failed). As shown in Figure 3-7, an area of 1.05 square inches produces a flow of 550 gpm at 425 psig, which is slightly below the setpoint pressure of the RHR relief valves and is the run-out flow for a single centrifugal charging pump. Makeup to the RWST is limited to 150 gpm which is the positive displacement charging pump capacity. 3.1.4.3 Operator Actions and Emergency Procedures Operator actions can mitigate V-sequences that result in leakage that exceeds the capacity of the RHR relief valves and causes subsequent failure of the RHR pump seals. Initially, conditions in the pressurizer relief tank will be diagnosed as an open PORV condition. At the same time, the high radiation level in the auxiliary building and the sump level and the pump operation alarms in the RHR vault may indicate reactor coolant is being transferred outside the containment. Finally, the g sustained transfer of reactor coolant to the pressurizer relief tank will 3 cause failure of its rupture discs. This could be diagnosed as a LOCA inside the containment, as containment pressure increases, and as sump level increases. The simultaneous occurrence of the's e indications may cause some degree of confusion among the operating staff. This section describes the analysis of three sets of operator actions. E The term operators refers to a control room team. The plant is in a E normal full-power condition prior to the initiating event, and all support systems (e.g., electric power) are assumed to be available if required. e Operator Action 01. Operators diagnose the RHR system LOCA. e Operator Action 02. Operators isolate the RHR system LOCA. j e Operator Action 03. Operators provide makeup to the RWST. The operators are required to follow special procedures for a LOCA l outside of containment and reduce emergency cooling system flow to maintain an adequate supply of borated water in the refueling water storage tank. Because the ability to recirculate the reactor coolant system via the RHR pumps is not available, it is especially important for this type of LOCA that the operators match ECCS flow to that required for ! adequate decay heat removal early since the amount of borated water in ! the RWST is limited. Maintenance of maximum ECCS flow, while l conservative for most situations, almost assures long-term failure for l this type of LOCA. Maximum makeup of borated water to the RWST is about l 150 gpm. The minimum RCS flow to remove decay heat is presented in Procedure ECA-1.1 (Figure ECA-1.1-2). This flow rate is about 600 gpm immediately after reactor shutdown and 150 gpm at 13.5 hours after 1 3-32 E 1319P112285 N
reactor shutdown. One category of RHR system rupture is modeled in the Seabrook-specific MAAP analysis described in Section 4. The upper bound leak area of 2.6 square inches is modeled in MAAP since this results in faster draining of the RWST and allows less time for operator actions. The results of the operator action analysis are presented in Table 3-10. These operator actions include the hardware contribution, where applicable, and are based on enhanced procedures and instrumentation in order to aid the operators in their diagnosis of the event. The sequence time intervals are based on the ability to ensure adequate RWST level for long-term cooling when the maximum makeup rate to the RWST is 150 gpm. The sequence time interval may be longer for many ciasses of RHR system LOCAs (sizes of leaks). The judgmentally assessed human error rates are not expected to change significantly if the time for operator action increases after diagnosis of the event and acceptable operator action is considered to be very dependent on adequate procedures and instrumentation within the time ranges of interest. The following procedures are applicable to the RHR system LOCA event. e Procedure E-0 (Reactor Trip or Safety Injection). e Procedure ECA-1.2 (LOCA Outside Containment). This procedure provides guidance on isolating the rupture, e Procedure ECA-1.1 (Loss of Emergency Coolant Recirculation-ECR). This procedure provides guidance for supplying adequate ECCS flow and plant stabilization. e Procedure E-1 (Loss of Reactor or Secondary Coolant). This procedure provides guidance for long-term cooling and stabilization. Within a few moments after the event occurs, the Seabrook Station control room team will be following Procedure E-U for Reactor Trip or Safety Injection. The RCS pressure is a function of leak size and is expected to be between 450 and 250 psia during en RHR system LOCA. The maximum pressure is expected to be limited by the operation of the RHR relief valves. The lower pressure is expected to be controlled by RHR pump seal leakage or a failed open RHR relief valve when at least one SI or charging pump is operating. While the LOCA remains unisolated, the RCS cannot be repressurized; RCS temperature is expected to remain near the saturation temperature; and the pressurizer level may remain below the indicating level. There are a number of alarms and indicators in the control room that may be used to aid in the diagnosis of an RHR system rupture. However, these alarms are either of low priority during a LOCA or, if treated alone, do f not provide for an unambiguous determination of RHR system rupture. These indications and/or alarms are: e RHR system flow. t RHR pump current. 3-33 1319P112285
l 1 l e RHR pump discharge pressure. I l e RHR heat exchanger inlet temperature. e RWST level, e Containment sump level . gi E1 e RHR sump (vault) level. ; e Pressurizer relief tank level . e PORV valve position indication. ) e PORV block valve position indication. ) e Radiation monitoring system (containment enclosure ventilation and 3, RHR vault area monitors). E Operator response for Procedure E-0, as with other reactor trips, proceeds through Step 15 in which the operators check the status of the RHR system. If the pumps are not running, the operators may notice one or more indicators listed above, but would probably just restart the RHR l. pumps since the chance of a RHR system LOCA is perceived to be remote. l Step 24 instructs the operator to check if the RCS is intact. It is uncertain whether the containment radiation, pressure, or level alarms will activate. If such alarms do occur, the operator would probably g think that a PORY is open. Step 30 instructs the operators to check the g auxiliary building radiation monitor. The radiation monitoring equipment currently installed in each RHR vault consists of one high-range and one low-range (alarm at 0.1 mR/ hour) area monitors. This instrumentation is located at the floor level in each RHR l vault. These monitors are more than likely under water during an RHR l system LOCA and are not expected to provide adequate aid for the ! operators. Even considering a case where the monitors do not flood, it i is unlikely that they would be useful because of the low levels of radiation in the coolant compared to the alarm level. The G-M (Geiger-Muller) type radiation monitor in the common containment enclosure ventilation system duct may not alarm or provide adequate l assistance to pinpoint the source of radiation. A negative report of radiation, in fact, may make it harder to diagnose the RHR system LOCA. It is recommended that Steps 24 and 30 be combined in Procedure E-0. 3 Additional instrumentation beyond the radiation monitoring should be E l provided to help diagnose the RHR system LOCA. A condition in which the l pressurizer relief tank level is high, RHR system pressure is high, or t RHR vault sump (one or both) level indication and sump pump are operating should provide positive indication of an RHR system LOCA. The combination of these alarms written into the procedures (at Step 24) with a reference to Procedure ECA-1.2 (LCCA Outside Containment) should assist the control room team's ability to diagnose the RHR system LOCA. 1 I 1 l 3-34 1319P112285 l
I I The RHR discharge pressure instrument should have an extended range because a pegged-high RHR pressure instrument may be interpreted as a mal function . RHR system pressure may be greater than 450 psig for a short time. The RHR discharge pressure under normal conditions is expected to be below 195 psig, the shutoff pressure for the RHR pumps. The Seabrook Station operator aid, called the visual alarm system or the VAS, may be able to check the signals from these instruments and present I an alarm of high priority to alert the operators should conditions exist for an RHR system LOCA. This alarm would have a higher priority than the other alarms, taken one at a time, and serve as a backup to the written procedures. The quantification of operator action 01 is based on the availabil'ity of the pressurizer relief tank level instrumentation and RHR system pressure indicators. The human error rate used is that recommended in Table 20-6 of NUREG/CR-1278 (Reference 3-11) for following a procedure under abnormal conditions. This human error rate is interpreted to have a mean I value of 0.005 and to be represented by a lognormal distribution with a range factor of 10. The equation used to quantify operator action 01 is 01 = OP + (ZITRLR)(24.0) + (ZITRPR)(24.0)(2.0) where 0P is the basic human error rate, and ZITRLR and ZITRPR are the data designators from Section 6 of the SSPSA for level and pressure instruments. I Diagnosis of the RHR system LOCA in Steps 24 and 30 is necessary in order to guide the operators into Procedure ECA-1.2 for isolating the LOCA if possible. If an RHR system LOCA is definitely identified as the source of leakage, it may be advisable to shut an RHR system crosstie valve in order to isolate or reduce leakage if the seal in only one RHR pump had been severely damaged. I The quantification of the operator action sequence 02 in which operators isolate the LOCA is based on a mean human error rate of 0.005 and the f ailure of one motor-operated valve to close on demand. The equation used to quantify operator action 02 is 02 = (0P) + (ZIVM00) where 0P is the basic human error rate and ZIVMOD is the data designator developed in Section 6 of the SSPSA for the failure to close on demand I for a motor-operated valve. The last step in Procedure ECA-1.2 guides the operators to Procedure E-1 (Loss of Reactor and Secondary Coolant) if the LOCA is isolated successfully and Procedure ECA-1.1 (Loss of Emergency Recirculation) if the RHR system LOCA is not isolated. If both RHR suction valves fail, the LOCA cannot be isolated. In this case, the RHR vault level will rise until the height of water in the RHR I 3-35 1319P121685
vault is approximately equal to the highest filled portion of the reactor coolant system. If the operators are able to diagnose the RHR LOCA early and reduce ECCS flow to the minimum required to remove decay heat (Figure ECA-1.1-2 in Procedure ECA-1.1), there may be sufficient water and time, in reality, to cool and recycle a portion of the borated water in the RHR vault back to the RWST or to the boric acid tanks and into the RCS. If the operators do not diagnose the event in a timely manner or try to repressurize the reactor coolant system, the time to empty the RWST will decrease and core damage is mere certain to occur. Repressurization, in fact, is not possible until the RHR system LOCA has been isolated. The ECCS termination criteria cannot be met in this condition. Figure 3-8 depicts the normal makeup paths to the RWST, BAT, and VCT. Operator action 03, makeup to the RWST, is developed from the following equation: 03 = OP + [3] [(ZP2S + (ZP2R)(24.0)]2 where OP is the basic human error rate and ZP2S and ZP2R are the data designators from Section 6 of the SSPSA for a standby pump to start (ZP2S) and a standby pump to run (ZP2R). The addition of borated water to the RWST at 150 gpm is assumed to require the operation of one of two boric acid transfer pumps, one of two reactor makeup water pumps, and 3 one of two demineralized water transfer pumps. The makeup water inlet E valves to the RWST (CSV 446 and CSV 444) must be opened manually at the valve. Additional valves may have to be repositioned in order to complete the valve lineup. Assuming that these valves are easily accessible and have positive identification, the error rate in not completing the valve lineup correctly is included in the basic human error rate for completing the procedure. The procedure for emergency RWST make-up is described in Attachment A to Procedure ECA-1,1. 3.1.4.4 Pump Operation in Adverse Environments According to Reference 3-12, all safety-related motors used in the Seabrook plant are either drip proof or totally enclosed. In addition, the reference document states the following:
- 1. " Drip-proof motors are protected by housings from falling water or falling objects, and cooling air intakes are protected by baffles to E prevent ingestion of solid objects or water spray. Heavy spray, E directed into the air intakes, could be drawn into the motor, and if permitted to remain for long periods of time, could eventually cause a deterioration of the motor insulation and lubrication, thereby E
shortening motor life. Short-term or immediate failure of the motor as a result of water spray is not a credible event."
- 2. " Complete immersion due to flooding is considered to render motors inoperable because of degradation of insulation and bearing lubrication, and because of possible short circuiting of internal 3 ci rcui t ry. " E I
3-36 1319P112285 l
g_ _ _ .
. . . = . . . . . .
I I Based on this information, two failure modes (due to environmental effects) were considered for the CBS and safety injection pumps, and three failure modes were considered for the RHR pumps. The first two failure modes, common to all three sets of pumps, are failure resulting from a humid or steam environment in the vault and failure resulting from submersion of the pump motor. For the RHR pump, direct failure of the pump / motor caused by the seal failure per se is also considered. This I latter failure mode is assumed to include failure of the motor due to any jet of water emanating from the clearance between the shaft and seal i I assembly created when the seal fails. Figure 3-9 depicts the fault tree I l for environmental failures of the three sets of pumps. Point estimates ! for each of the failure modes and top events are shown in Tables 3-11 l through 3-13. These estimates are extrapolated from the MAAP calculation of the transient that pressurizes the RHR system to RCS pressure after a series failure of either the suction side MOVs or pump discharge / injection line check valves with a subsequent failure of the pump seals, resulting in a leak area of 2.6 square inches. Based on the 1 values assumed in Tables 3-11 through 3-13, it is predicted that the RHR l and CBS pumps are very likely to fail, and there is a high probability l that the safety injection pumps will fail as well. 3.1.5 SSPSA PLANT MODEL INTEGRATION I The assignment of sequence end states was discussed in Section 3.1.3.2. The results of this initial assignment of end states is summarized in Table 3-14. End state LOCA contains those sequences in which the leakage past the series check valves in the injection lines or the series MOVs in the suction lines exceeds 150 gpm, but does not exceed the RHR system relief valve capacity. For these sequences, the RHR system remains intact but primary coolant is released to the containment via the pressurizer relief tank while the RHR system relief valves are open. The sequences , I represented by this state are essentially medium LOCAs. When combined with the SSPSA medium LOCA initiating event frequency of 4.6 x 10-4 per year the contribution from the V-Sequence becomes insignificant and the SSPSA results for the medium LOCA event need not be adjusted. 1 1 End state DLOC contains sequences in which the interfacing LOCA has been terminated, and the ECCS has been degraded (D) (RHR or SI pumps have I I failed). If such sequences were to proceed to core melt due to random failures of the ECCS equipment that survived the terminated interfacing LOCA, they would fall into SSPSA plant damage states 3D, 70, or 8D. The 3 point estimate frequency of DLOC is 4.0 x 10-7 per year. The i 3 additional failures required to achieve core melt would lower this i frequency by at least one order of magnitude. This contribution is negligible compared to the rebaselined values for any of these states. Plant damage state DILOC represents sequences in which coolant makeup is being supplied to the core but the interfacing systems LOCA has not been terminated. Random failure of either of the charging pumps assumed to be operating would result in a late core melt with a containment bypass via the RHR pump seals. The frequency contribution from DILOC is multiplied I 3-37 1319P112285
I by the unavailability of one of two centrifugal charging pumps (.013) and reassigned to plant damage state 7FPV. The predicted frequency of plant damage state 8C is 7.1 x 10-10 per year. If it is conservatively assumed that this frequency is assigned to 80, its contribution is negligible (six orders of magnitude difference) in comparison with other contributions to 8D. A similar argument can be made for neglecting the 7D contribution given in Table 3-14. After reassignment, the three plant damage states of interest have the following frequencies: 7FPV: 1.2 x 10- year 1FPV: 2.7 x 10 / year 1FV: 4.6 x 10 / year The evaluation presented above implicitly assumes the availability of all support systems. The postulation of additional support and frontline system failures was considered and found not to contribute signficantly, based on a review of the SSPSA results. 3.2 CONTAINMENT RECOVERY ANALYSIS FOLLOWING AN EXTENDED LOSS OF ALL AC I POWER This section describes the analysis of the recovery of containment heat removal during core melt sequences initiated by loss of offsite power and involving station blackout. .In these scenarios, molten core material has a penetrated the reactor vessel, causing an increase in the sealed g containment atmosphere pressure and temperature. If the containment heat removal systems are not recovered, decay heat will cause the containment atmosphere temperature and pressure to continue to rise until the containment fails. In the SSPSA, it was determined that the dominant containment failure mode for these scenarios is long term containment overpressurization failure. The time of containment failure along these sequences cannot be determined precisely due to uncertainties in the strength of the containment, behavior of penetrations, etc. However, the uncertainty distribution quantified in the SSPSA for the time of containment failure indicates that there is a high degree of confidence that the containment wili remain intact for at least one and possibly several days after the initiating event. The containment failure probability as a function of time for an adiabatic heatup is developed in Section 11.6 of the SSPSA (Reference 3-10). It is presented here in Figure 3-10. Figure 3-10 E shows that the median time of containment failure, for example, is almost E 3 days (the mean is somewhat greater than 3 days). By contrast, the time of the beginning of core damage along these sequences was estimated to be about 14 to 16 hours. The high probability of a long period of time between core melt and containment failure is a principal motivation for the evaluation of containment recovery actions for these sequences. The analysis of the electric power system hardware and the operator actions to restore AC power prior to core melt, following a loss of all I 3-38 1319P112285
I offsite power initiating event, is described in Section 10.4 of Reference 3-10. In that analysis, the time available for electric power recovery (from 2 to 24 hours) depended upon the competing factors of DC I power availability, emergency feedwater availability, and reactor coolant pump seal leakage. Recovery of electric power was considered from either normally installed emergency diesels or the offsite 345-kV power grid. Containment recovery analysis is applied to those scenarios where core I melt has occurred and electric power has not yet been recovered. Containment recovery is assumed successful once the containment spray and recirculation functions have been accomplished. Containment integrity is preserved by providing another source of 4,160V power to operate at least one containment spray pump, one component cooling water pump, and one service water pump, or by providing an alternate means for spray and recirculation. Functionally, it is necessary to establish steady state heat removal from the containment atmosphere at a rate that exceeds decay heat generation rate. 3.2.1 RECOVERY MODEL I There is currently a set of accident sequences in the SSPSA that involve a loss of offsite power, a loss of onsite power, core melt, and an assumed failure of the containment due to overpressurization. The I containment recovery model consists of a new top event, CR, that is used to generate two new sequences, for each SSPSA sequence to which this model is applied, as follows. For each applicable SSPSA sequence, S, two new sequences are produced: S*CR - the sequence with successful containment recovery. S*N - the sequence with failure to recover the containment. What we seek to estimate is the conditional frequency of containment a failure given that core melt occurs for each sequence, F(NjS). This can be expressed as X(S*N) F(NlS)= g3) where I A(S*N) = frequency of core melt and containment failure along sequence S (events per reactor-year). I A(S) = frequency of core melt along sequence S (events per reactor-year) . A(S*N ) = (AL0SP) (EPR-5) (Qu) (3.16) l 3-39 1319P120285
where AL OSP = total frequency for the loss of all offsite power per site-year. EPR-5 = conditional frequency of onsite power failure and failure to recover onsite power prior to core melt, given a loss of offsite power initiating event. The number 5 refers to a specific electric power recovery model from Section 10 of Reference 3-10.* , Qu
= conditional frequency that offsite power and containment heat removal is not restored from t=0 to the time of containment failure, given loss of offsite power, loss of 3 onsite power, and nonrecovery of onsite power. 5 The reason for organizing the model in this way is to be able to make use of the part of this problem that was already solved in the SSPSA (LOSP and EPR-5), such that all time dependent aspects are incorporated into the Qu term.
Further refinement is afforded to the model by noting that
- ALOSP " ARLOSP + ANRLOSP where ARLOSP = frequency of recoverable loss of offsite power (events per site-year).
ANRLOSP = frequency of nonrecoverable loss of offsite power (events per site-year). The above refinement was made to avoid overestimating the benefits of E recovery and to be able to explicitly model common cause events that g could fail all transmission lines coming into the site. Having incorporated this requirement, the term A(S) can now be estimated as A(S) = (ARLOSP)(EPR-1) + (ANRLOSP)(EPR-5) (3.17) I where (EPR-1) is the conditional frequency of core melt due to loss of g onsite power and f ailure to recover both onsite and offsite power. m I
*The electric power recovery models in Reference 3-10 incorporate the contributions of diesel generator failures. By contrast, the recovery factors in Table 3-4, denoted by ER - X, X = 1, 2, ..., 10, do not have the diesel failures included, but are otherwise consistent with these factors.
I 3-40 1319P112285
I The frequency for ARLOSP, the' recoverable loss of offsite power, was developed in Section 6 of the SSPSA. The parameters for ARLOSP are I 95th Percentile: Mean: 50th Percentile: 3.4 1.3 7.2 x x x 10 10 10
/
/
/
plant-year plant-year plant-year 5th Percentile: 1.6 x 10 / plant-year The frequency for ANRLOSP, the loss of offsite power that is not recoverable, from the offsite power grid was developed by NTS/SMA (Reference 3-13). Nonrecoverable loss of offsite power is modeled to avoid overestimating the likelihood of recovery after an extended length of time; i.e., the several days after the initiating event are those of I interest in containment recovery analysis. The most likely cause of a fully nonrecoverable loss of offsite power at this site was assessed to be a seismic event. Other causes, such as weather related causes, were assumed to be recoverable. In,the SSPSA, seismically induced loss of I offsite power because of damage done in the switchyard and on the site were analyzed, considered to be nonrecoverable, and the risk contributions thereby accounted for. What is being accounted for here is the contribution to seismically induced loss of offsite power resulting from damage outside the plant; i.e., within the external grid. The analysis of this event used the SSPSA seismic risk analysis methodology I (see Section 9.2 of Reference 3-10), the Seabrook Station seismicity curves from the SSPSA, and a fragility for the offsite grid developed by the SSPSA seismic consultant, SMA, for use in the northeastern United States. The median capacity, Sr, and Su values of this fragility I curve are 0.3, 0.15, and 0.5, respectively (Reference 3-13). The results for ANRLOSP are Mean: 2.4 x 10-4 5th Percentile: 1.8 x 10-6 I 50th Percentile: 95th Percentile: 9.4 x 10-5 1.6 x 10-3 The factor EPR-5 is defined in Table 10.4-7 of the SSPSA (Reference 3-10). The product (ANRLOSP) (EPR-5) may be viewed as an estimate of the frequency of core melt when a loss of all offsite power is the initiating event and when offsite power is not restored. This I condition is important to avoid double counting the nonrecovery of offsite power prior to core melt, which is part of the parameter Qu. I In general, the likelihood that containment cooling is not initiated and maintained within a 168-hour period after the loss of all offsite AC power initiating event, Qu(t), can be calculated from the expression 168 Qu (t) = &c (t) 47 (t) d t +Q H (3.18) I I 1319P112285 3-41
I where
&c(t) = probability density function for containment failure at t hours after the initiating event at t=0.
&F(t) = cumulative probability that power is not recovered within t hours after its failure at t=0.
QH
= unavailability heat removal (ofe.g.,
components required containment for containment spray pumps, component cooling water pumps, etc.) for 24 hours after electric power is restored. The hardware unavailability contribution to the operation of the service water system, component cooling water system, and containment spray recirculation was estimated from the following split fractions developed in Appendix D of Reference 3-10. QH = (WA2)(WBB) + (PA1)(PBA) + (XA1)(XBA) (3.19) where (WA2)(WBB) = service water system unavailability. (PA1)(PBA) = component cooling water system unavailability. 3 (XA1)(XBA) = containment spray recirculation system unavailability. 5 The parameters for QH pre Mean: 1.0 x 10-4 5th Percentile: 1.4 x 10-5 50th Percentile: 5.1 x 10-5 E 95th Percentile: 4.5 x 10-4 E The frequency of power nonrecovery from any source within t hours after a loss of all AC power at t=0 is calculated from a Boolean combination of, in this case, three recovery models. This formulation is necessary to account for the fact that power recovery from any source is sufficient.
&p(t) = [1-F][(1-4345(t)) (1-434.5(t))(1-$other(t))]
+ [F] [(1-$other(t))]
where
$345 (t) = probability that power is recovered in t hours from the existing 345-kV offsite power grid.
$34.5 (t) = probability that power is recovered in t hours from the 34.5-kV offsite power transmission lines and grid.
&other (t) = probability that power is recovered in t hours from emergency transportable diesels or gas turbines.
I 3-42 1319P112285
I F = (ANRLOSP)/ (ARLOSP + ANRLOSP); the fraction of loss of offsite power events that are nonrecoverable. I The recovery integral for the containment does not include the possibility of recovery from the normally installed emergency diesels. These diesels are not included in this integral because the entry state I for this model specifies that core melt has occurred and electric power has not yet been restored. The emergency diesels have experienced a long-term type of failure. The probability that a single diesel can be recovered after having failed for longer than 8 hours is less than 20%. This assumption is broadly based on generic recovery data (e.g., References 10.4-1 and 10.4-2 in the SSPSA). The lower rate of recovery (as compared to the bulk offsite power grid) for diesels that failed initially, plus the probability of diesel failure to run for 24 hours, offers a significantly lower probability of recovery than that from the offsite power grid or from providing additional transportable emergency diesel power. In addi+1on, the possibility of recovery from the I installed diesels is i..cluded in the analysis of the recovery of electric power prior to core melt for those scenarios when the diesels are available. That is, the diesel recovery factor up to the time of core melt is included in the EPR-5 value. The calculations for the containment recovery model are performed by I Monte Carlo simulation methods of the STADIC3 computer program (Reference 3-14). The cumulative probability of containment failure as a function of time during a station blackout and heatup (Figure 3-10) is I input in tabular form as Table 3-15. This table is taken from the SSPSA containment analysis (see Section 11 of Reference 3-10). 3.2.2 345-kV 0FFSITE POWER REC 0VERY The probability that power is recovered within t hours from the existing 345-kV offsite power grid (4345 (t)) was developed in Section 10.4.3.1 I of Reference 3-10 from a review of records of forced outage reports on PSNH's 345-kV lines. The records included outage information applicable to losses of power for less than 24 hours. PSNH has had several years of operating experience from the 345-kV transmission lines. There have been I no instances of total 345-kV transmission grid unavailability up to 1985 (Reference 3-15, Section 8.2.2.2). This period of time includes the February 6 to 8,1978, snowstorm that struck New England. This was one I of the most intense, persistent, severe winter storms on record (Reference 3-15, Section 2.3.1.2). Table 3-16 presents the tabular values for offsite power recovery from I the 345-kV grid (all three lines). The cumulative frequency of power recovery from this source increases up until 24 hours. It is conservatively assumed that additional recovery beyond 24 hours is not l available from this source beyond that already obtained in the first period. I I There is a degree of uncertainty in estimating the cumulative probability of restoring electrical power within 24 hours because of the limited number of data points. The most up to date analysis of this data I l I I 1319P120285 3-43
I' provided in an NSAC Report (Reference 3-16) reports that in 588 site-years at nuclear power plants (all years through 1984), only 13 outages occurred that lasted longer than 1 hour. The longest outage , lasted about 9 hours. In an EPRI report (Reference 3-17), the cumulative i probability of recovering electric power was assessed. The probability ' reported for the recovery of electric power within 24 hours (averaged over all NERC regions) is 0.966, with a range of from 0.930 to 0.998. E The result reported for the NPCC is 0.950. These results are not 3 directly applicable to the Seabrook Station, but may be used for a comparison with the results from the review of PSNH 345-kV line outage ! data. The 5th and 95th percentiles for the 345-kV power recovery at ! t = 24 hours are 0.940 and 0.999. . 1 l l 3.2.3 34.5-kV 0FFSITE POWER REC 0VERY l The source of power for the Seabrook Station during construction has been two 34.5-kV power lines. The source of power for these lines is the g l Schiller Station located in Portsmouth, New Hampshire (about 15 miles E from Seabrook Station). Schiller Station is presently connected to the Seabrook Station temporary radial feeder that provides power to the I station via two parallel lines (one from Guinea and one from Hampton Switching Station). The 34.5-kV transmission system originating at I Schiller Station is presented in Figures 3-11, 3-12, and 3-13. l Figure 3-13 shows the 34.5-kV grid coming in from the west (U.S. Highway E I Route 1) and from the east (across the wetlands adjacent to Seabrook 5 l Station). Presently, Schiller Station has five generating units. Three units (50-MW each, coal-fired) are base-loaded. One 25-MW oil unit and one l 24-MW combustion gas turbine are used for peaking loads. It is estimated l that one of these units can restart in about 30 minutes if they are taken l offline by a widespread loss of all AC power. PSNH estimates that it I might take from about 4 hours (hot plant) to about 8 hours (cold plant) l to restart one of the base-loaded units if it were not operating at the g i time when all AC power was lost. Schiller Station, with 5 generating 3 units, should be a reliable additional source of offsite power to the i Seabrook Station. A detailed analysis of the 34.5-kV transmission system, or the Schiller Station, is beyond the scope of this analysis. Existing information has been taken into account to estimate recovery from this source. The present 34.5-kV line is installed into the temporary construction power ' supply. An estimate of the cumulative power recovery frequency for the 34.5-kV offsite power grid to Seabrook Station is presented in Table 3-17. The mean recovery fraction (e.g., .95 for t = 48 hours) is derived from the review of the PSNH 345-kV transmission line outage data. The 5th percentile for the 345-kV transmission system represents an estimate for a single 345-kV transmission line. This estimate is used for the expected or mean recovery fraction for the 34.5-kV transmission system. This analysis, therefere, assumes that the installation and performance of the 34.5-kV lines are similar to the 345-kV lines that are the basis for the recovery fraction. In the analysis, the recovery frequency for the 34.5-kV system was assumed to be equivalent to a 3-44 1319P112285
1 l l single 345-kV line for the first 24 hours after recovery starts. Beyond 24 hours, to account for possible common cause effects, additional recovery of offsite power was only considered for one system. For i I convenience, the additional electric power recovery after 24 hours was modeled in the 34.5-kV system. The uncertainty in the recovery factor at each time is assumed to be represented by a lognormal distribution with a l i range factor of 5. The development or extension of transmission line recovery factors beyond 24 hours is uncertain because of the lack of data. The factors that I might cause loss of power for extended periods (from 18 to 168 hours) at Schiller Station, or in the approximately 15 miles of transmission line, seem remote. It might be reasoned that, in an emergency, large segments I of the line could be entirely replaced. The upper bound for recovery reported in Reference 3-17 was 0.998 at 24 hours. This analysis has judgmentally extrapolated the mean electric power recovery fraction as 0.999, as presented in Table 3-17. This extrapolation is believed to be I conservative and within the bounds of our present state of knowledge about the 34.5-kV transmission service for Seabrook if this grid is similar in performance to the 345-kV system. Because it was assumed that these temporary 34.5-kV lines would be removed after construction, they were not included in the SSPSA I analysis. Their incorporation into the analysis assumes that these lines will remain in place and that steps will be taken to ensure the 34.5-kV system is of comparable reliability, recoverability, and resistance to severe weather as a single 345-kV line coming into the site. To account I for the competition between various recovery efforts in the early phases of the accident, it is assumed that efforts to restore electric power via the 34.5-kV line do not begin until after core melt occurs (18 hours). 3.2.4 RECOVERY OF POWER FROM OTHER TRANSPORTABLE EMERGENCY POWER SOURCES I Containment spray and recirculation may be provided by supplying temporary power directly to the containment spray pumps, component cooling pumps, and service water pumps. The approximate power requirements for one of each type of pump (see Appendix D of Reference 3-10) are: Pump KVA KW Volts Service Water 3,709 590* 4,160 Component Cooling Water 3,647 577 4,160 Containment Spray 2,941 494 4,160
- Estimated from the KVA requirements.
I l I 1319P120285 3-45
- . . ~
Z b A rough estimate of the power requirements to sequentially start and run - all three pumps might be on the order of 2 MW, supplied by one or more diesel or gas turbine mobile power units. These mobile power units may be purchased in case of emergency by a group of utilities in a regional area, or may be available from a list of sources developed as part of a station's emergency plans. A quick check of a readily available guide (Reference 3-18) produced a number of sources. Figure 3-14 presents one L possible source of 2-MW (2,000 kW Allison 501 KA gas turbine generator 5 sets) mobile power units that may be available in a reasonable length of time in a high priority situation. This example is not meant to be conclusive of the availability of the units. It is just an example of a quick review of possible sources. _ The selected mobile power source may be attached to either Seabrook 3 Station IE Emergency 4,160V bus, or attached directly to the motor and 5 run from a control console at the generator. The availability of all C three system pumps is not required to prevent core damage because their g .- function may be provided by temporary diesel-driven pumps or, in some g cases, by the installed diesel-driven fire pump. It is estimated that a pump of 1,000 gpm at 100 psig would be sufficient to provide enough heat removal capacity for the secondary side of the containment spray recirculation system. The diesel-driven fire pump is rated at a maximum of 1,500 gpm at 125 psig. N Tne diesel-driven fire pump system, or a fire truck, could supply the 5 7 service water system function on a once-through basis. In addition, a - temporary diesel-driven pump could be brought in to function for the g E component cooling water pumps. While the possibility exists that a g & temporary pump could be installed for the containment spray pump, the use of the diesel-driven fire pump is more likely to be hooked up in time. The containment spray recirculation system would be circulating highly contaminated water. This system must be capable of operating unattended for long periods of time. This requirement means that a temporary system - would require more formal installation requirements (e.g., preoperational B T testing, hydrostatic test) to ensure proper operation prior to pumping 5 contaminated fluid. The requirements for temporary service water or component cooling water are less stringent in this case. b The recovery of the containment spray and recirculation function includes a large number of recovery scenarios, weather conditions, and plant
=- hardware availability conditions. Table 3-18 presents a subjective cumulative probability recovery distribution for the containment spray and recirculation function. The source of recovery is the installation n-and operation of mobile power supplies from commercial sources to start g and operate a normally installed service water pump, component cooling 5 i water pump, and containment spray pump. The distribution presented in
= It is believed that the cumulative recovery Table 3-18 is subjective. probability can be significantly increased with formal procedures and emergency plans. Additional sources of emergency power may be available t through the military and flown in to Pease Air Force Base. Temporary m pumps, fittings, etc., may be available from the Portsmouth Naval l Shipyard. Both installations are located near Portsmouth, New Hampshire, 5 5 approximately 12 miles northeast of Seabrook Station. Additional power p-E'- L K'" [ 1319P112285 3-46 7
I generation may be delivered by rail. The closest approach of a branch rail line to the site is about 2,000 feet (Reference 3-15, Section 2.2.2.1.b.1). 3.2.5 RESULTS I This section presents the recovery results and lists assumptions important in the estimation of containment survival during an extended loss of all AC when core melt has occurred. The results from the model are presented in Table 3-19. The analysis of the operator actions to restore electric power after a loss of power can be found in Section 10.4 of Reference 3-10. Sources of recovery for I electric power in the SSPSA were the 345-kV grid and the station diesels. The station blackout core melt fr equency is estimated in Equation (3.17) from the sum of the nonrecoverable and recoverable (with I the EPR corrective factor for recovery) loss of all offsite power initiating event frequencies. Electric power recovery factors that included the possibility of recovery from the 345-kV grid are not included because recovery from the 345-kV grid is included within the I containment recovery model. The integral for the containment [ Equation (3.18)] includes recovery from t = 0 to 168 hour from either the 34.5-kV, 345-kV, or mobile power sources. Recovery fri i the 34.5-kV line is considered only after core melt has occurrea. The unavailability of the regularly installed hardware required to I provide containment spray and heat removal recirculation contributes about 3% to containment failure. The availability of this equipment sets a lower limit or target for electric power nonrecovery of approximately 1 x 10-4 for Qu. Recovery below this value would require provisions for additional and diverse heat removal and spray equipment. ! Operator error and recovery is not explicitly included in this recovery l E analysis. Tne inclusion of operator error and recovery factors would be i 5 masked by the analysis results. It is therefore implicitly assumed that the operator error rates within the considerable length of time available for recovery are low and are less than the unavailability of electric lI power and hardware to provide the containment cooling function. The failure of the containment is expected to occur less often than the core melt frequency because of the added time available. This analysis l estimates that the factor I F(NjS), containment failure frequency / site year core melt frequency / site year the conditional frequency of nonrecovery of the containment, given I station bla;kout-induced melt, is about .07 This nonrecovery factor would have approached .5 without the assumptions regarding the 34.5 kV , system. This factor is conservative because the possibility of recovery l of the onsite diesel generators was not considered in the time interval between core melt and containment failure. It is believed that comparably favorable recovery factors (such as .07) could be achieved if the northeast utilities pooled their resources and purchased a dedicated I i 3-47 l 1319P112285
I mobile electric generator that could be easily transported to specific plants in an emergency. Recovery of electric power from the 345-kV switching station will require operation of the 345-kV breakers. The control and indication of the 345-kV breakers are on an independent DC system. The gas system used for closing and tripping these breakers is designed to store enough high pressure gas for a minimum of three close/open operations at full rating without the need to operate air compressors. This analysis assumes that these breakers are operable (compressed air is available) and that the ,' operators have been trained on the manual operation of these and the operation of all 13.8-kV and 4.16-kV circuit breakers. The 13.8-kV and 4.16-kV breakers may be operated via a manual charging handle (Section 10.4.4.3 of Reference 3-10). The containment analysis assumes that containment recovery is successful if power restoration has commenced. In actuality, the initiation of g containment recirculation may be required prior to the times of B containment f ailure provided in Figure 3-10. This assumption of no time delay is believed to be insignificant to the overall results. The service water system and component cooling water system have a mission time of 24 hours for success, while containment recirculation is 72 hours. This analysis assumes that all required service water valves, component cooling water valves, and containment spray system valves can be operated manually. As part of emergency planning, these systems can be walked down in order to locate areas where it may be feasible to attach 3 E temporary pumps and fittings. The containment spray system (Reference 3-19) shows a removable spool piece at the suction of each containment spray pump. Although it may t,e relatively easy to open the g system at this point, better locations with far easier access may exist. g As was recommended for core melt scenarios, a list of instrumentation not dependent on AC or DC power to monitor the containment level, pressure,
~
and temperature should be developed. A list of possible sources of temporary 4,160V motive power, as well as standby lower voltage AC and/or DC power for instrumentation should be developed. Diesel-driven pumps, or pumps of lower voltage and power requirements, may fulfill the containment cooling requirement. A quick review of possible sources by PSNH personnel during the course of this analysis located a 1,000-kV, 480V diesel only 20 miles from Seabrook Station. The pumping requirements (pump head, capacity, and cooling water supply required) should be formally established. The containment spray pump head must be sufficient to overcome containment pressure and elevation losses and still provide sufficient pressure difference across the spray nozzles to assure atomization. A review of the losses of offsite power at U.S. nuclear power plants presented in Reference 3-16 shows that about half of the offsite power outages that lasted longer than 1 hour were caused by weather (e.g., tornadoes, lightning, hurricanes, snow storms, ice storms, and . salt use). Although Seabrook Station has not suffered a loss of its 345-kV transmission grid through a 100-year snow storm, Seabrook Station should continue to review all losses of offsite power for the applicability of the grid prior to and during station operation. 1319P112285
i If the permanent installation of the 34.5-kV transmission grid should be considered, it should be tested at reasonable intervals. A review of previous outages (e.g., losses caused by excessive icing and salt buildup on insulators and substation equipment) should be included in a design review of the 34.5-kV system. This analysis assumes that the 34.5-kV grid is installed and that its performance is similar to present I 345-kV transmission lines. The risk-reduction benefits of this action are described in Section 2.
3.3 REFERENCES
3-1. Pickard, Lowe and Garrick, Inc., " Risk-Based Evaluation of Technical Specifications for Seabrook Station," prepared for New Hampshire Yankee Division, Public Service of New Hampshire, PLG-0431, August 1985. I 3-2. Pickard, Lowe and Garrick, Inc., Westinghouse Electric Corporation, and Fauske & Associates, Inc., " Indian Point Probabilistic Safety Study," prepared for the Power Authority of the State of New York and Consolidated Edison Company of I New York, Inc., March 1982. 3-3. U.S. Nuclear Regulatory Commission, " Reactor Safety Study: An I Assessment of Accident Risks in U.S. Commercial Nuclear Power Plants," WASH-1400, NUREG-75/014, October 1975. I 3-4. Fauske & Associates, Inc., " Evaluation of Containment Bypass and Failure to Isolate Sequences for the IDCOR Reference Plants," Draft Report FAI/84-9, July 1984. 3-5. U.S. Nuclear Regulatory Commission, " Preliminary Case Study Report, Overpressurization of Emergency Cooling System in Boiling Water Reactors," February 1985. 3-6. "MAAP - Modular Accident Analysis Program Users Manual," Technical Report on IDCOR Tasks 16.2 and 16.3, May 1983 3-7. S. M. Stoller Corporation, Nuclear Power Experience, updated monthly. 3-8. " Data Summaries of Licensee Event Reports of Valves at U.S. Commercial Nuclear Power Plants," prepared by EG&G Idaho, Inc. for U.S. Nuclear Regulatory Commission, NUREG/CR-1363, Vol.1, June 1980. 3-9. Box, G. E. P., and G. C. Tiao, Bayesian Inference in Statistics, Addison-Wesley, 1973. 3-10. Pickard, Lowe and Garrick, Inc., "Seabrook Station Probabilistic Safety Assessment," prepared for Public Service Company of New Hampshire and Yankee Atomic Electric Company, PLG-0300, December 1983. I 1319P112285
e E 3-11. Swain, A. D., and H. E. Guttmann, " Handbook of Human Relialility Analysis with Emphasis on Nuclear Power Plant Applications," Sandia National Laboratories, prepared for U.S. Nuclear Regulatory Commission, NUREG/CR-1278, SAND 80-0200, August 1183. 3-12. United Engineers & Constructors, Inc., "Seabrook Station Moderate Energy Line Break Study," Revision 2, November 1983. 3-13. NTS/ Structural Mechanics Associates, Inc., private communication regarding seismic-induced loss of offsite power events assessed originally for a northeast U.S. site. 3-14 Cairns, J. J., and K. N. Fleming, "STADIC--A Computer Code for Combining Probability Distributions," General Atomic Report for U.S. Department of Energy, GA-A14055,1977. 3-15. Seabrook Station Final Safety Analysis Report, Amendment 48, January 1983. 3-16. Nuclear Safety Analysis Center, " Losses of Offsite Power at U.S. Nuclear Power Plants - All Years Through 1984," prepared for Electric Power Research Institute, NSAC-85, June 1985. 3-17. Science Applications, Inc., " Loss of Offsite Power at Nuclear Power Plants: Data and Analysis," prepared for Electric Power Research Institute, EPRI-2301, March 1982. 3-18. McGraw-Hill, Power, Vol. 128, No. 11, November 1984. 3-19. Seabrook Station P&I Diagram 9763-F-805023, Rev. 6. I I l I I I 3-50 l 1319P112285
)
M M M M M M M M M M M M 'l TABLE 3-1. UPDATE OF SSPSA PLANT MODEL RESULTS FOR CORE MELT AND PLANT DAMAGE STATE FREQUENCIES Event Frequency Basis for Significant Plant (events per reactor-year) Changes to SSPSA Damage - State
- SSPSA Reference 3-1 Current In Reference 3-1 In Current Results Results Results Results Results IF 1.89-6 1.86-6 2.00-8 --
Enhanced V-Sequenr.e 1FV -- -- 4.60-9 -- -- IFP 8.52-7 1.38-6 1.38-6 Reassessment of Seismic -- IFPV -- -- 2.70-8 -- -- 2A 1.85-6 1.86-6 1.86-6 -- -- 3D 1.94-5 1.94-5 1.55-5 -- Containment Recovery 3F 5.00-7 2.81-7 2.81-7 Reassessment of Seismic -- 3FP 6.21-6 8.87-6 8.87-6 -- -- w 4A 1.28-5 1.37-5 1.37-5 Reassessment of Systems --
& 70 7.06-5 6.86-5 2.22-5 --
Containment Recovery 7F 3.55-8 2.20-8 2.20-8 Reassessment of Seismic -- 7FP 1.09-5 9.98-6 9.98-6 -- -- 7FPV -- -- 1.20-8 -- Enhanced V-Sequence 8A 4.50-5 4.64-5 9.67-5 -- Containment Recovery 8D 5.51-5 1.03-4 1.03-4 Reassessment of Systems -- Others ~5-6 ~5-6 ~5-6 -- -- Core Melt 2.30-4 2.76-4 2.74-4 Reassessment of Systems Enhanced Treatment of V-Sequence
*See Table 1-2 for definitions of "lant damage states those ending in V are variations on those corresponding states without V and'uenote that new and separate source terms are defined in this study in Section 4.
NOTE: Exponential notation is indicated in abbreviated form; i.e., 1.89-6 = 1.89 x 10-6, 1251P121685
I TABLE 3-2. UPDATE OF SSPSA PLANT MODEL ACCIDENT SEQUENCES RANKED BY CORE MELT FREQUENCY CONTRIBUTION Ranking Plant Damage State : Sequence Number
- Frequency (events /yr)
CN1 : PDS-8A : Sequence-31 3.163E-05 l C52 : PDS-80 : Sequence-2 1.988E-05 W CM-3 : PDS-80 : Sequence-4 1.611E-05 l C54 : PDS-80 : Sequence-6 1.239E-05 C55 : PDS-8A : Sequence-1 8.900E-06 CF-6 : PDS-8D : Sequence-9 9.779E-06 C57 : PDS-80 : Sequence-1 8.576E-06 CN8 : PDS-30 : Sequence-1 8.300E-06 CM-9 : PDS-8A : Sequence-2 5.600E-06 CN10: PDS-8A : Sequence-32 4.876E-06 CM-ll: PDS-8A : Sequence-38 4.14?E-06 CM-12: PDS-SA : Sequence-39 4.142E-06 C N 13: PDS-4A : Sequence-1 4.073E-06 CN14: PDS-80 : Sequence-3 CM-15: PDS-8D : Sequence-5 3.997E-06 E CM-16: PDS-80 : Sequence-13 3.429E-06 g 3.559E-06 CM-17: PDS-8A : Sequence-3 3.000E-06 CN 18: PDS-80 : Sequence-8 2.591E-06 CM-19: PDS-80 : Sequence-7 2.400E-06 C520: PDS-8A : Sequence-4 2.400E-06 C521: PDS-70 : Sequence-1 2.381E-06 CM-22: PDS-70 : Sequence-5 2.352E-06 CM-23: PDS-8A : Sequence-5 CM-24: PDS-8A : Sequence-6 2.200E-06 E l CN 25: PDS-BA : Sequence-27 2.200E-06 g i 2.051E-06 CS26: PDS-80 : Sequence-10 2.003E-06 CH-27: PDS-70 : Sequence-6 1.990E-06 CM-28: PDS-8A : Sequence-7 1.900E-06 CW29: PDS-4A : Sequence-2 1.900E-06 C&30: rDS-8A : Sequence-9 1.749E-06 C&31: PDS-8A : Sequence-10 1.749E-06 CM-32: PDS-80 : Sequence-11 l l C533: PDS-7FP: Sequence-1 1.519E-06 1.337E-06 l g l C534: PDS-80 : Sequence-12 1.320E-06 l CM-35: PDS-8A : Sequence-8 1.300E-06 C536: PDS-70 : Sequence-7 1.255E-06 [ C537: PDS-4A : Sequence-3 1.200E-06 i CM-38: PDS-7FP: Sequence-5 1.122E-06 CM-39: PDS-3FP: Sequence-1 1.106E-06 CN 40: PDS-30 : Sequence-3 1.102E-06 CM-41: PDS-70 : Sequence-8 1.098E-06 l CM-42: PDS-80 : Sequence-14 1.085E-06 g CM-43: PDS-80 : Sequence-14A 1.085E-06 Total Identified Core Melt Sequences - This Table 1.938E-04 Unidentified Core Melt Sequences - Total 7.998E-05 SEABROOK - Total Core Melt - 2.738E-04
$ Sequences defined in terms of initiating events and system top events and boundary conditions in Tables 3-3, 3-4, and 3-5.
I I I 1320Pil2185
E TABLE 3-3a. UPDATE OF SSPSA PLANT MODEL - PLANT DAMAGE STATE 1F SEQUENCES PDS : Failure Expression
- Frequency Rank (events /yr)
I 1F -1 : E.7L
- OG7
- NGA6
- NGB6
- SAQ
- SCC
- NPAQ
- NPCB
- NRWF
- C24 1F -2 : El.0L
- OG8
- NGA7
- NGB7
- SAS
- SCF
- PAS
- PCF
- RWG
- C24 Ir -3 : E.7L
- OG7
- NGA6
- NG86
- SAQ
- SCC
- NPAQ
- NPCB
- RWF
- C24 1.091E-14 4.805E-15 5.620E-15 1F -4 : El.0L
- OG8
- NGA7
- NGB7
- SAS
- SCF
- NPAS
- NPCE
- RWG
- C24 4.563E-15 IF -5 : E.7L
- OG7
- NGA6
- NG86
- SAQ
- SCC
- PAQ
- PCC
- NRWF
- C24 4.277E-15 1F -6 : El.0L
- OG8
- NGA7
- NGB7
- SAS
- SCF
- PAS
- PCF
- RWG
- C24 4.80EE-15 IF -7 : El.0L
- OG8
- NGA7
- NGB7
- SAS
- SCF
- NPAS
- NPCE
- RWG
- C24 4.563E-15 1F -8 : E.7L
- OG7
- NGA6
- NGB6
- SAQ
- SCC
- PAQ
- PCC
- RWF
- C24 2.203E-15 Plant Damage State - IF - This Table I Other Sequences in Plant Damage State - IF Flant Damage State - IF - Total 3.908E-14
?.000E-08 2.000E-08
- Definition of initiating events and boundary conditions provided in Table 3-4.
NOTE: An "N" preceding any three-letter top event split fraction identifier is defined as follows: NXYZ = 1 - XYZ where XYZ is the three-letter top event split fraction identifier. I I . I I I I 3-53 1320Pil2185
l _ _ TABLE 3-3b. UPDATE OF SSPSA PLANT MODEL - PLANT DAMAGE STATE 1FP SEQUENCES PDS : Failure Express *on* Frequency Rank (events /yr) IFP-1 : El.0L
- OG8
- GA7
- GBJ
- SAS
- SCF
- RWG 2.068E-07 l IFP-2 : E1.0L
- OG8
- GA7
- GBJ
- SAS
- SCF
- NRWG 1.267E-07 5 1FP-3 : E.7L
- OG7
- GA6
- GBH
- NSAQ
- NSCB
- NRWF 1.087E-07 1FP-4 : E.7L
- OG7
- GA6
- GBH
- SAQ
- SCC
- NRWF 8.945E-08 1FP-5 : E1.0L
- OG8
- GA7
- GBJ
- NSAS
- NSCE
- RWG 8.772E-08 1FP-6 : E.7L
- OG7
- GA6
- GBH
- NSAQ
- NSCB
- RWF 5.598E-08 IFP-7 : El.0L
- OG8
- GA7
- GBJ r NSAS
- NSCE
- NRWG 5.376E-08 1FP-8 : E.7L
- OG7
- GA6
- GBH
- SAQ
- SCC
- RWF 4.608E-08 1FP-9 : E.7L
- OG7
- NGA6
- NGB6
- SAQ
- SCC
- NRWF 1.658E-07 1FP-10: E1.0L
- OG8
- NGA7
- NGB7
- SAS
- SCF
- PAS
- PCF
- RWG 5.217E-08 1FP-11: E.7L
- OG7
- NGA6
- NGB6
- SAQ
- SCC
- NPAQ
- NPCD
- RWF 6.101E-08 1FP-12: El.0L
- OG8
- NGA7
- NGB7
- SAS
- SCF
- NPAS
- NPCE
- RWG 4.954E-08 1FP-13: E.7L
- OG7
- NGA6
- NGB6
- SAQ
- SCC
- PAQ
- PCC
- NRWF 4.644E-08 1FP-14: E1.0L
- OG8
- NGA7
- NGB7
- SAS
- SCF
- PAS
- PCF
- RWG 5.217E-08 1FP-15: E1.0L
- OG8
- NGA7
- NGB7
- SAS
- SCF
- NPAS
- NPCE
- RWG 4.954E-08 IFP-16: E.7L
- OG7
- NGA6
- NGB6
- SAQ
- SCC ^ PAQ
- PCC
- RWF 2.392F-08
. Plant Damage State - IFP - This Table 1.276E-06 Other Sequences in Plant Damage State - 1FP 1.050E-07 Plant Damage State - IFP - Total 1.381E-06 (Definition of initiating events and boundary conditions provided in Table 3-4. NOTE: An "N" preceding any three-letter top event sp>it fraction identifier is defined as follows:
. NXYZ = 1 - XYZ where XYZ is the three-letter top event split fraction identifier.
I I I I I I I I I 1320P112185
l i TABLE 3-3c. UPDATE OF SSPSA PLANT MODEL - l PLANT DAMAGE STATE 1FV SEQUENCES PDS : Failure Expression
- Frequency Rank (events /yr) l IFV-1 : VI
- LRI
- PI 2.506E-09 !
IFV-2 : VS
- LRS
- PI 1.757E-09 Plant Damage State - IFV - This Table 4.263E-09 Other Sequences in Plant Damage State - IFV 3.770E-10 Plant Damage State - IFV - Total 4.540E-09
- Definition of initiating events and boundary conditions provided in Table 3-4.
I NOTE: An "N" preceding any three-letter top event split fraction identifier is defined as follows: NXYZ = 1 - XYZ where XYZ is the three-letter top event split fraction identifier. I I I I I I 3-55 1320Pil2185
1 l s. TABLE 3-3d. UPDATE OF SSPSA PLANT MODEL - l PLANT DAMAGE STATE 1FPV SEQUENCES PDS : Failure Expression
- Frequency Rank (events /yr)
IFPV-1 : VS
- LRS
- L3
- VC 1.482E-08 1FV-2 : VS
- LRS
- L2
- VC 1.186E-08 Plant Damage State - IFPV - This Table 2.668E-08 Other Sequences in Plant Damage State - IFPV 6.200E-10 Plant Damage State - IFPV - Total 2.730E-08 ODefinition of initiating events and boundary conditions provided in Table 3-4.
NOTE: An "N" preceding any three-letter top event split fraction identifier is defined as follows: NXYZ = 1 - XYZ where XYZ is the three-letter top event split fraction identifier. I I I I I 3-56 1320P112185
TABLE 3-3e. UPDATE OF SSPSA PLANT MODEL - PLANT DAMAGE STATE 2A SEQUENCES PDS : Failure Expression
- Frequency Rank (events /yr)
I 2A -1 : LLOCA
- LAI
- LBA 2A -2 : MLOCA
- Lil
- L2A 2A -3 : ELOCA 8.200E-07 3.200E-07 2.500E-07 2A -4 : MLOCA
- EBB
- L22 6.900E-08 2A -5 : MLOCA
- EAA
- L12 6.900E-08 2A -6 : LLOCA
- LC1
- LDA 4.100E-08 2A -7 : MLOCA
- WBO
- L12 3.509E-08 2A -8 : MLOCA
- WAA
- L22 3.509E-08 2A -9 : LLOCA
- EAA
- LB2 2.700E-08 2A -10: LLOCA
- EBB 5 LA2 2.700E-08 Plant Damage State - 2A - This Table 1.693E-06 Other Sequences in Plant Damage State - 2A 1.630E-07 Plant Damage State - 2A - Total 1.856E-06 I
- Definition of initiating events and boundary conditions provided in Table 3-4.
NOTE: An "N" preceding any th.'ee-letter top event split fraction identifier is defined as follows: NXYZ = 1 - XYZ where XYZ is the three-letter top event split fraction identifier. I l l [ 3-57 1320P112185
1 TABLE 3-3f. UPDATE OF SSPSA PLANT MODEL - l PLANT DAMAGE STATE 3D SEQUENCES m PDS : Failure Expression
- Frequency Rank (events /yr) 3D -1 : ALOMF
- SA6
- SBK 8,300E-06
! 3D -2 : LOSP
- gal
- GBA
- EF2
- ER2
- FR1
- CR 1.544E-07 l 3D -3 : LOSP
- NGAl
- NGB1
- WA3
- WBC
- EF2
- ERA
- FR1 1.102E-06 30 -4 : LOSP
- gal
- GBA
- NEF2
- ON2
- ER2
- CR 5.539E-08 3D -5 : E.7T
- OG7
- NGA6
- NGB6
- NSAM
- NSBV
- NRTF
- NPAQ
- WPCB
- EKB
- RWD 7.24?E-07 3D -6 : E.3A
- NOG 4
- SAG
- SBN
- NPAA
- NPBD
- PL1 6.400E-07 3D -7 : E.3A
- NOG 4
- SAG
- SBN
- NPAA
- NPBD
- FLIN 6.400E-07 3D -8 : LOSP
- NGAl
- NGB)
- WA3
- WBC
- NEF2
- ON2
- ERA 3.954E-07 3D -9 : L.5T
- OG6
- NGA5
- NGB5
- NSAK
- NSBS
- NPA0
- NPBY
- EJB
- RWC 4.447E-07 3D -10: LOSP
- GB2
- WA4
- EF2
- ER4
- FR1
- CR 2.887E-08 3D -11: LOSP
- GA2
- WB4
- EF2
- ER4
- FR1
- CR 2.887E-08 3D -12: E.7T
- OG7
- NGA6
- NGB6
- NSAM
- NSBV
- NRTF
- FAQ
- PCC
- EKB
- RWD 2.839E-07 30 -13: E.4T
- OG5
- NGA4
- NGB4
- NSAI
- NSBP
- NPAM
- NPBV
- EIB
- RWB 3.341E-07 3D -14: RT
- SA6
- SBK
- OR5 2.100E-07 30 -15: LOSP
- DA2
- GB2
- EF2
- FR1
- ERA
- CR 2.771E-09 l 3D -16: LOSP
- DB2
- GA2
- Er2
- FR1
- ERA
- CR 2.771E-09 3D -17: LOSP
- GA2
- WB4
- NEF2
- ON2
- ER4
- CR 1.036E-08 I 3D -18: LOSP
- GB2
- WA4
- NEF2
- ON2
- ER4
- CR 1.036E-08 l 3D -19: ATT
- SA6
- SBK
- PR6 1.600E-07 l Plant Damage State 3D - This Table 1.347E-05 Other Sequences in Plant Damage State - 30 2.000E-06 Plant Damage State Total 1.547E-05 ODefinition of initiating events and boundary conditions provided in fable 3-4.
NOTE: An "N" preceding any three-letter top event split fraction identifier is defined as follows: , NXYZ = 1 - XYZ where XYZ is the three-letter top event split fraction identifier. I l
~
l i l l 3-58 1320Pil2185
TABLE 3-39 UPDATE OF SSPSA PLANT MODEL - PLANT DAMAGE STATE 3F SEQUENCES PDS : Failure Expression
- Frequency Rank (events /yr) 3F -1 : E.3A
- NOG 4
- SAG
- SBN
- NPAA
- NPBD
- PL1
- H35
- C25 6.400E-08 3F -2 : E.3A
- NOG 4
- SAG
- SBN
- NPAA
- NPBD
- PLIN
- H35
- C25 6.400E-08 3F -3 : E.7T
- OG7
- NGA6
- NGB6
- SAM
- SBW
- NRTF
- NPAQ
- NPCB
- EK2
- RWD
- C24 5.710E-14 3F -4 : E.7T
- OG7
- NGA6
- NGB6
- SAM
- SBW
- NRTF
- PAQ
- PCC
- EK2
- NRWD
- C24 4.345E-14 3F -5 : E.7T
- OG7
- NGA6
- NGB6
- SAM
- SBW
- NRTF
- NPAQ
- NPCB
- EK2
- NRWD 3.103E-14
- H28
- C24 3F -6 : E.7T
- OG7
- NGA6
- NGB6
- SAM
- SBW
- NRTF
- PAQ
- PCC
- EK2
- RWD
- C24 2.239E-14 3F -7 : E.5T
- OG6
- NGA5
- NGB5
- SAK
- SBT
- NRTC
- NPA0
- NPBY
- EJ2
- RWC
- C24 1.239E-14 3F -8 : E.5T
- OG6
- NGAS
- NGB5
- SAK
- SBT
- NRTC
- NPA0
- NPBY
- EJ2
- NRWC
- H27 1.066E-14
- C24 3F -9 : El.0T
- OG8
- NGA7
- NGB7
- SAO
- SBZ
- NRTI
- PAS
- PCF
- EL2
- RWE
- C24 8.270E-15 3F -10: El.0T
- OG8
- NGA7
- NGB7
- SAO
- SBZ
- NRTI
- NPAS
- NPCE
- EL2
- RWE
- C24 7.853E-15 3F -11: E.5T
- OG6
- NGAS
- NGB5
- SAK
- SBT
- NRTC
- PA0
- PBZ
- EJ2
- NRWC
- C24 1.045E-14 3F -12: E.4A
- NOG 5
- SAI
- SBQ
- NPAC
- NPBG
- PL1
- EIL
- NRWB
- C25 1.014E-08 3F -13: E.4A
- NOG 5
- SAI
- SBQ
- NPAC
- NPBG
- PLIN
- EIL
- NRWB
- C25 1.014E-08 3F -14: E.3A
- NOG 4
- SAG
- SBN
- NPAA
- NPBD
- PL1
- EHL
- NRWA
- C25 1.124E-08 3F -15: E.3A
- NOG 4
- SAG
- SBN
- NPAA
- NPBD
- PLIN
- EHL
- NRWA
- C25 1.124E-08 Plant Damage State - 3F - This Table 1.707E-07 Other Sequences in Plant Damage State - 3F 1.100E-07 Plant Damage State - 3F - Total 2.807E-07
- Definition of initiating events and boundary conditions provided in Table 3-4.
NOTE: #n N" preceding any three-letter top event split fraction identifier is defined as follows: NXYZ = 1 - XYZ where XYZ is the three-letter top event split fraction identifier. 4 I I 1320Pil2185
TABLE 3-3h. UPDATE OF SSPSA PLANT MODEL - PLANT DAMAGE STATE 3FP SEQUENCES PDS : Failure Expression
- Frequency Rank (events /yr) 3FP-1 : E.7T
- OG7
- GA6
- GBH
- NSAM
- NSBV
- NRTF
- EK2
- NRWD 1.106E-06 3FP-2 : E.7T
- OG7
- GA6
- GBH
- SAM
- SBW
- NRTF
- EK2
- NRWD 9.088E-07 '
3FP-3 : E.7T
- OG7
- GA6
- GBH
- NSAM
- NSBV
- NRTF
- EK2
- RWD 5.698E-07 3FP-4 : E.5T
- OG6
- GAS
- GBF
- NSAK
- NSBS
- NRTC
- EJ2
- NRWC 5.843E-07 3FP-5 : E.7T
- OG7
- GA6
- GBH
- SAM
- SBW
- NRTF
- EK2
- Rbo 4.682E-07 3FP-6 : E.4T
- OG5
- GA4
- GBD
- NSAI
- NSBP
- EI2
- NRWB 4.863E-07 3FP-7 : El.0T
- OG8
- GA7
- GBJ
- SAO
- SBZ
- NRTI
- EL2
- RWE 3.559E-07 3FP-8 : El.0T
- OG8
- GA7
- GBJ
- SAO
- SBZ
- NRTI
- EL2
- NRWE 2.181E-07 3FP-9 : El.0T
- OG8
- GA7
- GBJ
- NSAO
- NSBY
- NRTI
- EL2
- RWE 1.514E-07 3FP-10: E.5T
- OG6
- GAS
- GBF
- SAK
- SBT
- NRTC
- EJ2
- NRWC 1.653E-07 3FP-II: E.3T
- OG4
- GA3
- GBB
- NSAG
- NSBN
- EH2
- NRWA 2.018E-07 3FP-12: El.0T
- OG8
- GA7
- GBJ
- NSA0
- NSBY
- NRTI
- EL2
- NRWE 9.276E-08 3FP-13: E.5T
- OG6
- GA5
- GBF
- NSAK
- NSBS
- NRTC
- EJ2
- RWC 9.512E-08 3FP-14: E.4T
- CG5
- GA4
- GB0
- SAI
- SBQ
- EI2
- NRWB 6.648E-08 3FP-15: El.0A
- OG8
- GA7
- GBJ
- SAO
- SBZ
- RTI
- PL1
- ELL
- RWE 5.620E-08 3FP-16: El.0A
- OG8
- GA7
- GBJ
- SAO
- SBZ
- RTI
- PLIN
- ELL
- RWE 5.620E-08 3FP-17: E.7A
- OG7
- GA6
- GBH
- NSAM
- NSBV
- RTF
- PL1
- EKL
- NRWD 4.811E-08 3FP-18: E.7A
- OG7
- GA6
- GBH
- NSAM
- NSBV
- RTF
- PL1
- EKL
- NRWD 4.811E-08 3FP-19: E.2T
- OG3
- gal
- GBA
- EG2 7.506E-08 3FP-20: E.7T
- OG7
- NGA6
- NGB6
- SAM
- SBW
- NRTF
- NPAQ
- NPCB
- EK2
- RWD 6.199E-07 3FP-21: E.7T
- OG7
- NGA6
- NGB6
- SAM
- SBW
- NRTF
- PAQ
- PCC
- EK2
- NRWD 4.718E-07 3FP-22: E.7T
- OG7
- NGA6
- NGB6
- SAM
- SBW
- NRTF
- NPAQ
- NPCB
- EK2
- NRWD
- H28 3.369E-07 3FP-23: E.7T
- OG7
- NGA6
- NGB6
- SAM
- SBW
- NRTF
- PAQ
- PCC
- EK2
- RWD 2.430E-07 3FP-24: E.5T
- OG6
- NGA5
- NGB5
- SAK
- SBT
- NRTC
- NPA0
- NPBY
- EJ2
- RWC 1.346E-07 3FP-25: E.5T
- OG6
- NGA5
- NGB5
- SAK
- SBT
- NRTC
- NPA0
- NPBY
- EJ2
- NRWC
- H27 1.157E-07 3FP-26: El.0T
- OG8
- NGA7
- NGB7
- SAO
- SBZ
- NRTI
- PAS
- PCF
- EL2
- RWE 8.979E-08 3FP-27: El.0T
- OG8
- NGA7
- NGB7
- SAO
- SBZ
- NRTI
- NPAS
- NPCE
- EL2
- RWE 8.526E-08 3FP-28: E.5T
- OG6
- NGAS
- NGB5
- SAK
- SBT
- NRTC
- PA0
- PBZ
- EJ2
- NRWC 1.135E-07 Plant Damage State - 3FP - This Table 7.965E-06 Other Sequences in Plant Damage State - 3FP 9.000E-07 Plant Damage State - 3FP - Total 8.865E-06
- Definition of initiating events and boundary conditions provided in Table 3-4.
NOTE: An "N" preceding any three-letter top event split fraction identifier is defined as follows: NXYZ = 1 - XYZ . where XYZ is the three-letter top event split fraction identifier. l 3-60 1320Pil2185
l TABLE 3-31. UPDATE OF SSPSA PLANT MODEL - PLANT DAMAGE STATE 4A SEQUENCES PDS : Failure Expression
- Frequency Rank (events /yr)
I 4A -1 : LlDC
- EF0
- FR1
- FR2 4A -2 : ATT
- RTS
- OH1 4A -3 : ALOMF
- RTS
- OH1 4.073E-06 1.900E-06 1.200E-06 4A -4 : RT
- EFA
- ORS
- FR2
- FR1 3.310E-07 4A -5 : LOSP
- NGAl
- NGB1
- EFB
- ORS
- ERA
- FR1 3.289E-07 4A -6 : LOSP
- DA2
- NGB2
- EFD
- ERA
- FR1 2.924E-07 4A -7 : LOSP
- DB2
- NGA2
- EFD
- ERA
- FR1 2.924E-07 4A -8 : E.7T
- OG7
- NGA6
- NGB6
- NSAM
- NSBV
- NRTF
- NPAQ
- NPCB
- EKB
- NRWD
- H28 3.936E-07 I 4A -9 : E.5T
- OG6
- NGA5
- NGB5
- NSAK
- NSBS
- NRTC
- NPA0
- NPBY
- EJB
- NRWC
- H27 4A -10: E.4T
- OG5
- NGA4
- NGB4
- NSAI
- NSBP
- NPAM
- NPBV
- EIB
- NRWB
- H26 4A -11: SLBI
- L13
- L2C 4A -12: LOSP
- GA2
- EFD
- ORS
- ER4
- FR1 3.748E-07 3.664E-07 2.300E-07 3.153E-07 4A -13: LOSP
- GB2
- EFD
- OR5
- ER4
- FR1 3.153E-07 1
4A -14: LOSP
- NGA1
- NGB1
- WB4
- EFD
- ORS
- ERA
- FR1 2.228E-07 1 4A -15: LOSP
- NGAl
- NGB1
- WA4
- EFD
- ORS
- ERA
- FR1 2.228E-07 I 4A -16: LlDC
- PA4
- EF0
- FRI'
- FR2 2.553E-09 4A -17: LIDC
- PB4
- EFD
- FR1
- FR2 2.553E-09 I Plant Damage State - 4A - 1his Table Other Sequences in Plant Damage State - 4A Plant Damage State - 4A - Total 1.086E-05 2.800E-06 1.366E-05
- Definition of initiating events and boundary conditions provided in Table 3-4.
NOTE: An "N" preceding any three-letter top event split fraction identifie: is defined as follows: NXYZ = 1 - XYZ where XYZ is the three-letter top event split fraction identifier. h I i I . 1320P112185 3-61
TABLE 3-3j. UPDATE OF SSPSA PLANT MODEL - PLANT DAMAGE STATE 7D SEQUENCES PDS : Failure Expression
- Frequency Rank (events /yr) 70 -1 : LOSP
- gal
- GBA
- NEF2
- ERI
- CR 2.381E-06 70 -2 : LOSP
- NGAl
- NGB1
- WA3
- WBC
- NEF2
- ER9
- CR 3.670E-07 70 .3 : LOSP
- GA2
- bE4
- NEF2
- ER3
- CR 3.118E-07 70 -4 : LOSP
- GB2
- WA4
- NEF2
- ER3
- CR 3.llSE-07 70 -5 : FTBLP
- gal
- GBA
- NEF2
- ERS 2.352E-00 70 -6 : FCRAC
- NEF2 1.990E-06 70 -7 : FLLP
- GA1
- GBA
- NEF2
- ER5 1.255E-06 7D -8 : TCTL
- gal
- GBA
- NEF2
- ERS 1.098E-06 70 -9 : LOSP
- GA1
- GBA
- EF2
- NFR1
- ERI
- CR 4.OllE-08 7D -10: FSRAC
- NEF2 4.666E-07 70 -11: LOSP
- NGAl
- NGB1
- WA3
- WBC
- EF2
- NFR1
- ER9 8.831E-08 70 -12: RT
- OG1
- gal
- GBA
- NEF2
- ERI
- CR 1.463E-08 7D -13: PLMFW
- OG1
- gal
- GBA
- NEF2
- ERI
- CR 1.185E-08 70 -14: LOSP
- DA2
- GB2
- NEF2
- ER3
- CR 9.527E-09 70 -15: LOSP
- DB2
- GA2
- NEF2
- ER3
- CR 9.527E-09 70 -16: TT
- OG1
- gal
- GBA
- NEF2
- ERI
- CR 9.121E-09 70 -17: LOSP
- GB2
- WA4
- EF2
- NFR1
- ER3
- CR 5.252E-09 7D -18: LOSP
- GA2
- WB4
- EF2
- NFR1
- ER3
- CR 5.252E-09 70 -19: EXFW
- OG1
- gal
- GBA
- NEF2
- ERI
- CR 6.458E-09 70 -20: LOSP
- DA2
- NGB2
- WB4
- NEF2
- ER9
- CR 3.221E-09 70 -21: LOSP
- DB2
- NGA2
- WA4
- NEF2
- ER9
- CR 3.221E-09 Plant Damage State This Table 1.071E-05 Other Sequences in Plant Damage State - 70 1.150E-05 Plant Damage State - 7D - Total 2.221E-05 CDefinition of initiating events and boundary conditions provided in Table 3-4.
NOTE: An "N" preceding any three-letter top event split fraction identifier is defined as follows: NXYZ = 1 - XYZ where XYZ is the three-letter top event split fraction identifier. I I: 1 I l 1320Pil2185
TABLE 3-3k. UPDATE OF SSPSA PLANT MODEL - PLANT DAMAGE STATE 7F SEQUENCES PDS : Failure Expression
- Frequency Rank (events /yr)
I 7F -1 : E.7T
- OG7
- NGA6
- NGB6
- SAM
- SBW
- NRTF
- PAQ
- PCC
- NEKB
- NRWD
- C24 7F -2 : E.7T
- OG7
- NGA6
- NGB6
- SAM
- SBW
- NRTF
- PAQ
- PCC
- NEKB
- RWD
- C24 7F -3 : E.7A
- OG7
- NGA6
- NGB6
- SAM
- SBW
- RTF
- PAQ
- PCC
- PL1
- NEKB
- RWD 3.695E-14 1.903E-14 8.276E-16 4
l l
- C24 I
7F -4 : E.7A
- OG7
- NGA6
- NGB6
- SAM
- SBW
- RTF
- PAQ
- PCC
- PLIN
- NEKB
- RWD 8.276E-16 '
- C24 l 7F -5 : E.5T
- OG6
- NGA5
- NGB5
- SAK
- SBT
- NRTC
- PA0
- PBZ
- NEJB
- NRWC
- C24 1.856E-14 7F -6 : E.4A
- NOG 5
- SAI
- SBQ
- NPAC
- NPBG
- PL1
- RWB
- C25 2.423E-09 I 7F -7 : E.4A
- NOG 5
- SAI
- SBQ
- NPAC
- NPBG
- PLIN
- RWB
- C25 7F -8 : E.5A
- NOG 6
- SAK
- SBT
- NPAE
- NPBJ
- PL1
- RWC
- C25 7F -9 : E.5A
- N0G6
- SAK
- SBT
- NPAE
- NPBJ
- PLIN
- RWC
- C25 7F -10: E.3A
- NOG 4
- SAG
- SBN
- NPAA
- NPBD
- PL1
- RWA
- C25 2.423E-09 2.084E-09 2.084E-09 1.501E-09 7F -11: E.3A
- NOG 4
- SAG
- SBN
- NPAA
- NPBD
- PLIN
- RWA
- C25 1.501E-09 7F -12: E.4T
- OG5
- NGA4
- NGB4
- SAI
- SBQ
- PAM
- PBW
- NEIB
- NRWB
- C24 1.167E-14 Plant Damage State - 7F - This Table 1.201E-08 Other Sequences in Plant Damage State - 7F 1.000E-08 Plant Damage State - 7F - Total 2.201E-08
- Definition of initiating events and boundary conditions provided in Table 3-4.
NOTE: An "N" preceding any three-letter top event split fraction identifier is defined as follows: I NXYZ = 1 - XYZ where XYZ is the three-letter top event split fraction identifier. I I lI lI 3-63 1320Pil2185
TABLE 3-?m. UPDATE OF SSPSA PLANT MODEL - PLANT DAMAGE STATE 7FPV SEQUENCES PDS : Failure Expression
- Frequency Rank (events /yr)
I 7PV-1 : VS
- LRS
- P(L4)
- NYC 7PV-2 : VS
- LRS
- P(L3)
- CP
- NYC 7PV-3 : VS
- LRS
- P(L2)
- CP
- NVC 2.633E-09 1.773E-09 1.418E-09 7PV-4 : VI
- LR1
- P(L3)
- 01
- NVC 1.220E-09 7PV-S : VI
- LR1
- P(L2)
- 01
- NVC 9.759E-10 7PV-6 : VS
- LRS
- P(L1)
- CS
- NVC
- NRS
- NSS 9.189E-13 7PV-7 : VS
- LRS
- P(L3)
- 01
- NYC 8.566E-10 7PV-8 : VS
- LRS
- P(L2)
- 01
- NVC 6.84SE-10 7PV-9 : VS
- LRS
- P(L3)
- 03
- NVC 6.450E-10 7PV-10: VS
- LRS
- P(L2)
- 03
- NYC 5.160E-10 Plant Damage State - 7FPV - This Table 1.072E-08 Other Sequences in Plant Damage State - 7FPV 1.280E-09 Plant Damage State - 7FPV - Total 1.200E-08 I
- Definition of initiating events and boundary conditions provided in Table 3-4.
preceding any three-letter top event split fraction identifier is defined as follows: I NOTE: An "N" NXYZ . 1 - XYZ where XYZ is the three-letter top event split fraction identifier. I I I I I I I I 1320Pil2185 3-65 .
TABLE 3-3n. UPDATE OF SSPSA PLANT MODEL - PLANT DAMAGE STATE 8A SEQUENCES PDS : Failure Expression
- Frequen Rank (events / )
8A -1 : SLOCA
- L13
- L2C 8.900E-06 8A -2 : SLBI
- ON2 5.600E-06 8A -3 : RT
- ONI 3.000E-06 BA -4 : PLMFW
- ONI 2.400E-06 8A -5 : SLOCA
- EB0
- L14 2.200E-06 E 8A -6 : SLOCA
- EAB
- L24 BA -7 : TT
- ON1 2.200E-06 1.900E-C6 5
8A -8 : EXFW
- ON1 1.300E-06 8A -9 : SLOCA
- WAA
- L24 1.749E-06 BA -10: SLOCA
- WBD
- L14 1.749E-06 8A -11: RT
- EFA
- 031 9.960E-07 8A -12: RT
- EFA
- 031
- EFA
- L13
- L2C 7.152E-07 BA -13: TT
- EFA
- 031 6.210E-07 8A -14: LCPF
- ONI 5.400E-07 4.398E-07 l
g 8A -15: EXFW
- EFA
- 031 8A -16: TT
- EFA
- L13
- L2C 4.459E-07 8A -17: LCV
- ON1 4.000E-07 8A -18: IMSIV
- ONI 3.400E-07 8A -19: SLBI
- 031 3.300E-07 8A -20: TLMFW
- ON1 3.200E-07 8A -21: EXFW
- EFA
- L13
- L2C 3.158E-07 BA -22: SGTR
- EFB
- 005 2.649E-07 BA -23: SLOCA
- L13
- L2C
- XA1 2.300E-07 7.152E-07 l
g BA -24: RT
- EFA
- L13
- L2C BA -25: LCV
- EFB
- 031 1.915E-07 BA -26: LOPF
- ONI 1.800E-07 8A -27: LOSP
- gal
- GBA
- EF2
- ER2
- FR1
- NCR 2.051E-06 8A -28: LOSP
- GB2
- mA4
- EF2
- ER4
- FR1
- NCR 3.835E-07 BA -29: LOSP
- gal
- GBA
- NEF2
- ON2
- ER2
- NCR 7.359E-07 BA -30: LOSP
- GA2
- WB4
- EF2
- ER4
- FR1
- NCR 3.835E-07 BA -31: LOSP
- GA1
- GBA
- NEF2
- ERI
- NCR 3.163E-05 BA -32: LOSP
- NGAl
- NGBl
- WA3
- WBC
- NEF2
- ER9
- NCR 4.876E-06 8A -33: LOSP
- gal
- GBA
- EF2
- NFR1
- ERI
- NCR 5.328E-07 BA -34: LOSP
- DA2
- GB2
- EF2
- FR1
- ERA
- NCR 3.682E-08 BA -35: LOSP
- DB2
- GA2
- EF2
- FR1
- ERA
- NCR 3.682E-08 8A -36: LOSP
- GA2
- WB4
- NEF2
- ON2
- ER4
- NCR 1.376E-07 8A -37: LOSP
- GB2
- WA4
- NEF2
- ON2
- ER4
- NCR 1.376E-07 BA -38: LOSP
- GA2
- WB4
- NEF2
- ER3
- NCR 4.142E-06 8A -39: LOSP
- GB2
- WA4
- NEF2
- ER3
- NCR 4.142E-06 BA -40: RT
- OG1
- gal
- GBA
- NEF2
- ERl
- NCR 1.944E-07 BA -41: PLMFW
- OG1
- GA1
- GBA
- NEF2
- ER)
- NCR 1.574E-07 8A -42: TT
- OG1
- gal
- GBA
- NEF2
- ERl
- NCR 1.212E-07 BA -43: EXFW
- OG1
- gal
- GBA
- NEF2
- ERI
- NCR 8.580E-08 BA -44: LOSP
- DA2
- GB2
- NEF2
- ER3
- NCR 1.266E-07 8A -45: LOSP
- DB2
- GB2
- NEF2
- ER3
- NCR 1.266E-07 8A -46: LOSP
- GB2
- WA4
- EF2
- NFR1
- ER3
- NCR 6.978E-08 8A -47: LOSP
- GA2
- WB4
- EF2
- NFR1
- ER3
- NCR 6.978E-08 8A -48: LOSP
- DA2
- NG82
- WB4
- NEF2
- ER9
- NCR 4.280E-08 BA -49: LOSP
- DB2
- NGA2
- WA4
- NEF2
- ER9
- NCR 4.280E-08 Plant Damage State - 8A - This Table 8.831E-05 E g
Other Sequences in Plant Damage State - 8A 8.370E-06 Plant Danage State - 8A - Total 9.668E-05
- Definition of initiating events and boundary conditions provided in Table 3-4.
NOTE: An "N" preceding any three-letter top event split fraction identifier is defined as follows: NXYZ = 1 - XYZ where XYZ is the three-letter top event split fraction identifier. 3-66 1320P112185
I TABLE 3-30. UPDATE OF SSPSA PLANT MODEL - PLANT DAMAGE STATE 8D SEQUENCES I PDS : Failure Expression
- Rank Frequency (events /yr)
I 80 -1 80 -2 80 -3 80 -4
- FCRCC
- NEF2
- RT
- pal
- PBA
- NEF2
- FPCC
- NEF2 8.576E-06 1.988E-05 3.997E-06
- PLMFW
- PA1
- PBA
- NEF2 1.611E-05 I 80 -5 80 -6 80 -7 80 -8
- FSRCC
- NEF2
- TT
- pal = PBA. *
- LOSW
- NEF2 NEF2
- LOSP
- NGAl
- NGB1
- PA2
- PBB
- NEF2 3.429E-06 1.239E-05 2.400E-06 2.591E-06 80 -9 : EXFW
- PA1
- PBA
- NEF2 8.779E-06 I 80 -10 : FCRSW
- NEF2 80 -11 : FLSW
- NEF2 80 -12 : LPCC
- NEF2 2.003E-06 1.519E-06 1.320E-06 80 -13 : LOPF
- PA1
- PBA
- NEF2 3.559E-06 I 80 -14 : SLOCA
- WBD
- EAB
- NEF2 80 -14A: SLOCA
- WAA
- EBD
- NEF2 80 -15 : LOSP
- GA2
- PBS
- NEF2
- ER3 8D -16 : LOSP
- GB2
- PA5
- NEF2
- ER3 1.085E-06 1.085E-06 7.597E-07 7.597E-07 80 -17 : LOSP
- NGAl
- NGBl
- WB4
- PA5
- NEF2
- ER9 I
2.364E-07 80 -18 : LOSP
- NGAl
- NGB1
- WA4
- PBS
- NEF2
- ER9 2.364E-07 8D -19 : RT
- WA2
- WBB
- NEF2
- SR1 2.839E-07 80 -20 : PLMFW
- WA2
- WBB
- NEF2
- SRI 2.300E-07 80 -21 : TT
- WA2
- WBB
- NEF2
- SR1 1.770E-07 I 80 -22 : FCRCC
- NEF2
- ON2 8D -23 : SLCCA
- WAl
- WBA
- NEF2
- SR2 80 -24 : SGTR
- WA1
- WBA
- NEF2
- SR2 80 -25 : EXFW
- WA2
- WBB
- NEF2
- SRI 1.076E-07 3.479E-il 2.085E-11 1.254E-07 Plant Damage State This Table I Other Sequences in Plant Damage State - CD Plant Damage State Total 9.165E-05 1.160E-05 1.032E-04
- Definition of initiating events and boundary conditions provided in Table 3-4.
NOTE: An "N" preceding any three-letter top event split fraction identifier is defined as follows: NXYZ = 1 - XYZ where XYZ is the three-letter top event split fraction identifier. I I I I I 1 I I I 1320P112185 3-67
I TABLE 3-4. DEFINITION OF INITIATING EVENTS, TOP EVENTS, AND B0UNDARY CONDITIONS DEFINED IN TECHNICAL SPECIFICATION UPDATE
- AND THE SSPSA l
Sheet 1 of 21 ELECTRIC POWER SYSTEM Top Event Value Definition Split Fraction Offsite Grid OG1 2.6600E-04 Loss of Offsite Power after other Initiating Events OG2 1.0000E+00 Loss of Offsite Power (LOSP) Initiating Event = OG3 3.5310E-01 Loss of Offsite Power Af ter Seismic Event - 0.2g OG4 6.5510E.01 Loss of Offsite Power After Seismic Event - 0.3g OG5 8.3500E-01 Loss of Offsite Power After Seismic Event - 0.4g l OG6 9.2290E-01 Loss of Offsite Power After Seismic Event - 0.5g 3 OG7 9.8510E-01 Loss of Offsite Power After Seismic Event - 0.7g OG8 9.9950E-01 Loss of Offsite Power After Seismic Event .1.0g AC Power With Offsite Power Available EPl 2.5410E-07 Loss of Power at E5 and E6 - Offsite Power Available DC Power System EP4 3.0400E-10 Loss of DC Power, Train A and Train B No LOSP EP5 4.7610E.07 Loss of DC Power After LOSP DC Power - Train A DAl 5.1310E-04 Loss of DC Train' A after LOSP DA2 5.1260E.04 Loss of DC Train A . Loss of AC Power at Train A only DA3 1.0000E+00 Loss of DC Train A - Guaranteed Failure - IE DC Power . Train B CBI 5.1310E.04 Loss of DC Train B after LOSP DBA 5.1210E-04 Loss of DC Train B after LOSP - DB Failed DB2 5.1260E.04 Loss of DC Train B - Loss of AC Power at Train B only DB3 1.0000E+00 Loss of DC Train B - Guaranteed Failure - IE Electric Power Recovery ERl 3.4090E.02 Electric Power Recovery .1 ER2 6.6360E-02 Electric Power Recovery - 2 ER3 3.2070E.02 Electric Power Recovery - 3 ER4 8.9130E.02 Electric Power Recovery - 4 ER5 5.3220E.01 Electric Power Recovery - 5 ER6 6.0500E.01 Electric Power Recovery - 6 l ER7 6.1730E.01 Electric Power Recovery - 7 6.4620E-01 Electric Power Recovery - 8 ER8 ER9 4.4360E.02 Electric Power Recovery - 9 ERA 2.8000E-01 Electric Power Recovery - 10 Diesel Generators (As Presented in SSPSA) EP2 8.8730E-03 Loss of Power at E5 and E6 After LOSP EP3 6.4220E.02 Loss of Power at E5 or E6 After LOSP Diesel Generator - Train A gal 7.2820E-02 Loss of Diesel Generator - Train A GA2 6.4220E-02 Loss of DG Train A . Single Train GA3 8.2730E-02 Loss of DG Train A . Seismic Event - 0.3g GAA 7.3570E.02 Loss of DG Train A - 0.39 - Single Train GA4 1.2240E.01 Loss of DG Train A - Seismic Event - 0.4g GAB 1.1100E-01 Loss of DG Train A - 0.4g - Single Train GA5 1.9070E.01 Loss of DG Train A . Seismic Event . 0.5g GAC 1.7550E-01 Loss of DG Train A - 0.5g - Single Train GA6 3.6720E-01 Loss of DG Train A - Seismic Event - 0.7g
- Reference 3-1.
1251P111985 3-68
I TABLE 3-4 (continued) I ELECTRIC POWER SYSTEM Sheet 2 of 21 Top Event Val ue Definition I Split Fraction GAD GA7
.....4210E-01 3 Loss of DG Train A - 0.79 - Single Train 6.4370E-01 Loss of DG Train A - Seismic Event - 1.0g GAE 6.0320E-01 Loss of DG Train A - 1.0g - Single Train GA8 1.0000E+00 Loss of DG Train A - Guaranteed Failure Diesel Generator . Train B GBl 7.0740E.02 Loss of DG Train B after LOSP GBA 1.0620E-01 Loss of DG Train B after LOSP - GA Failed GB2 6.4220E-02 Loss of DG Train B - Single Train GB3 7.1520E-02 Loss of DG Train B - Seismic Event - 0.3g I GBB GBC GB4 GBD 2.3650E.01 Loss of DG Train B - 0.3g - GA Fail 7.3570E.02 Loss of DG Train B - 0.3g - Single Train 7.4830E.02 Loss of DG Train B - Seismic Event - 0.4g 5.0280E-01 Loss of DG Train B - 0.49 - GA Fail GBE 1.1100E-01 Loss of DG Train B . 0.4g - Single Train I GB5 GBF GBG 8.1320E-02 Loss of DG Train B - Seismic Event - 0.5g 6.8170E-01 Loss of DG Train B - 0.5g - GA Fail 1.7550E-01 Loss of DG Train B . 0.59 - Single Train GB6 1.0490E.01 Loss of DG 1 rain B - Seismic Event - 0.7g I GBH GBI GB7 GBJ 8.3210E.01 Loss of DG Train B - 0.79 - GA Fail 3.4210E-01 Loss of DG Train B - 0.7g - Single Train 1.9300E-01 Loss of DG Train B - Seismic Event - 1.0g 9.0290E-01 Loss of DG Train B - 1.0g . GA Fail GBK 6.0320E.01 Loss of DG Train B - 1.0g - Single Train GB8 1.0000E+00 Loss of DG Train B - Guaranteed aflure iI I
i I I I I I I ,,,,,,,,,,, 3.e,
I TABLE 3-4 (continued) Sheet 3 of 21 SOLID STATE PROTECTION SYSTEM Top Event Value Definition Split Fraction E
---------------------------------------------------- -------------------------- g SSPS - Train A sal 1.8490E-03 SSPS Train A - LLOCA/MLOCA SAA 1.4540E-03 SSPS Train A - LLOCA/MLOCA - Single Train SA2 1.1570E-03 SSPS Train A - SLOCA SAB 1.1560E-03 SSPS Train A - SLOCA - Single Train SA3 1.1610E-03 SSPS Train A - SGTR SAC 1.1580E-03 SSPS Train A - SGTR - Single Train l SA4 1.8290E-03 SSPS Train A - SLBOC 3 SAD 1.4340E-03 SSPS Train A - SLBOC - Single Train SA5 1.1570E-03 SSP 5 Train A - SLBIC SAE 1.1560E-03 SSPS Train A - SLBIC - Single Train SA6 1.1610E-03 SSPS Train A - GT SAF 1.1580E-03 SSPS Train A - GT - Single Train SAG 4.1160E-02 SSPS Train A - Seismic Event - 0.3g SAH 4.ll10E-02 SSPS Train A - 0.3g - Single Train SAI 1.2120E-01 SSPS Train A - Seismic Event - 0.4g l SAJ 1.2100E-01 SSPS Train A - 0.4g - Single Train 3 SAK 2.2110E-01 SSPS Train A - Seismic Event - 0.5g SAL 2.2090E-01 SSPS Train A - 0.5g - Single Train SAM 4.5110E-01 SSPS Train A - Seismic Event - 0.7g SAN 4.5060E-01 SSPS Train A - 0.7g - Single Train SAO 7.0100E-01 SSPS Train A - Seismic Event - 1.0g SAP 7.0030E-01 SSPS Train A - 1.0g - Single Train SA0 4.5170E-01 SSPS Train A - Seismic Event - 0.7g - LLOCA SAR 4.5080E-01 SSPS Train A - 0.7g - Single Train - LLOCA 7.0150E-01 SSPS Train A - Seismic Event - 1.0g - LLOCA l
SAS 3 SAT 7.0040E-01 SSPS Train A - 1.0g - Single Train - LLOCA SA7 1.0000E+00 SSPS Train A - Guaranteed failure SSPS - Train 8 SBl 1.4560E-03 SSPS Train B - LLOCA/MLOCA SBA 2.1360E-01 SSPS Train B - LLOCA/MLOCA - SA Failed SBB 1.4540E-03 SSPS Train B - LLOCA/MLOCA - Single Train l SB2 1.1580E-03 SSPS Train B - SLOCA 3 SBC 7.3780E-04 SSPS Train B - SLOCA - SA Failed SBD 1.1560E-03 SSPS Train B - SLOCA - Single Train SB3 1.1600E-03 SSPS Train B - SGTR 3 SBE 2.5170E-03 SSPS Train B - SGTR - SA Failed 1.1580E-03 SSPS Train B - SGTR - Single Train E SBF SB4 1.4360E-03 SSPS Train B - SLBIC SBG 2.1600E-01 SSPS Train B - SLBIC - SA Failed SBH 1.4340E-03 SSPS Train B - SLBIC - Single Train l SBS 1.1580E-03 SSPS Train B - SLBOC 3 SBI 7.3780E-04 SSPS Train B - SLBOC - SA Failed SBJ 1.1560E-03 SSPS Train B - SLBOC - Single Train SB6 1.1600E-03 SSPS Train B - GT 3 SBK 2.5170E-03 SSPS Train B - GT - SA Failed 1.1580E-03 SSPS Train B - GT - Single Train g SBL SBN 1.2080E-03 SSPS Train B - Seismic Event - 0.3g SBN 9.7170E-01 SSPS Train B - Seismic Event - 0.3g - SA failed SB0 4.ll10E-02 SSPS Train B - 0.3g - Single Train SBP 1.3180E-03 SSPS Train B - Seismic Event - 0.4g SB0 9.9030E-01 SSPS Train B - Seismic Event - 0.4g - SA failed SBR 1.2100E-01 SSPS Train B - 0.4g - Single Train SBS 1.4870E-03 SSPS Train B - Seismic Event - 0.5g SBT 9.9470E-01 SSPS Train B - Seismic Event - 0.5g - SA failed SBU 2.2090E-01 SSPS Train B - 0.5g - Single Train SBV 2.1110E-03 SSPS Train B - Seismic Event - 0.7g SBW 9.9730E-01 SSPS Train B - Seismic Event - 0.7g - SA failed SBX 4.5060E-01 SSPS Train B - 0.7g - Single Train 1 1 1251Pil1985
I i I TABLE 3-4 (continued) ' Sheet 4 of 21 SOLIO STATE PROTECTION SYSTEM Top Evert Value Definition I Split Fraction SBY SBZ 3.8760E-03 SSPS Train B - Seismic Event - 1.0g - SA failed 9.9830E-01 SSPS Train B - Seismic Event - 1.0g I SCA SCB SCC SCO 7.0030E-01 SSPS Train B - 1.0g - Single Train 2.6 5'0E-03 SSPS Train B - Seismic Event - 0.79 - LLOCA 9.9670E-01 SSPS Train B - Seismic Event - 0.7g - SA failed 4.5080E-01 SSPS Train B - 0.79 - Single Train - LLOCA SCE 4.8710E-03 SSPS Train B - Seismic Event - 1.0g - LLOCA SCF 9.9790E-01 SSPS Train B - Seismic Event - 1.0g - SA failed SCG 7.0040E-01 SSPS Train B - 1.0g - Single Train - LLOCA SB7 1.0000E+00 SSPS Train B - Guaranteed Failure I I I lI I I I I I I I 1251Pil1985 3-71 ,
I TABLE 3-4 (continued) Sheet 5 of 21 ESFAS FUNCTION Top Event Value Definition Split Fraction ESFAS - Train A EAl 1.1750E-02 ESFAS Train A - LLOCA/MLOCA EAA 1.1640E.02 ESFAS Train A - Single Train - LLOCA/MLOCA EA2 9.4490E-03 ESFAS Train A . SLOCA/SGTR EAB 9.3410E.03 ESFAS Train A . Single Train . SLOCA/SGTR EA3 1.2190E.02 ESFAS Train A - SLbOC EAC 1.2080E-02 ESFAS Train A - Single Train - SLBOC l EA4 1.0060E.02 ESFAS Train A - SLBIC 5 EAD 9.9490E.03 ESFAS Train A - Single Train - SLBIC EA5 1.3320E-03 ESFAS Train A - GT EAE 1.2690E.03 ESFAS Train A . Single Train . GT E EA6 1.0000E+00 ESFAS Train A - Guaranteed Failure g ESFAS - Train B EB1 1.1780E-02 ESFAS Train B - LLOCA/MLOCA l EBA 9.6670E.03 ESFAS Train B . LLOCA/MLOCA . EA Failed 5 EBB 1.1640E.02 ESFAS Train B - Single Train - LLOCA/MLOCA EB2 9.4300E-03 ESFAS Train B - SLOCA/SGTR EBC 1.1380E-02 ESFAS Train B - SLOCA - EA Failed EBD 9.3410E-03 ESFAS Train B - Single Train - SLOCA/SGTR EB3 1.2220E-02 ESFAS Train B - SLBIC EBE 9.3790E.03 ESFAS Train B . SLBIC - EA Failed EBF 1.2080E-02 ESFAS Train B . Single Train - SLBIC EB4 1.0050E-02 ESFAS Train B . SLB0C EBG 1.0740E-02 ESFAS Train B - SLBOC - EA Failed EBH 9.9510E-03 ESFAS Train B - Single Train - SLBOC EB5 1.1740E-02 ESFAS Train B - GT EBI 9.7310E-03 ESFAS Train B . GT - EA Failed EBJ 1.1600E-02 ESFAS Train B - Single Train - GT EB6 1.0000E+00 ESFAS Train B - Guaranteed Failure I I I I I I,' 1251P111985 3-M
TABLE 3-4 (continued) I REACTOR TRIP SYSTEM Sheet 6 of 21 Top Event Value Definition I Split Fraction Reactor Trip System
== -- .................-...--...
5.5530E-04 Reactor Trip - Both SSPS (No Operator Action) I RT1 RT2 4.6670E-03 Reactor Trip - Single SSPS (No Operator Action) RT3 5.4280E.06 Reactor Trip - SSPS Not Required (LOSP etc.) RT4 0.0000E-01 Reactor Trip - Not Asked (LLOCA, RT Initiating Events) RTS 8.0770E.05 Reactor Trip - Both SSPS Signals Present I RT6 RTA RTB RTC 6.4410E.04 Reactor Trip - Single SSPS Signal Present 2.0540E-02 Reactor Trip - RTl - Seismic Event - 0.5g 2.4580E.02 Reactor Trip - RT2 - Seismic Event - 0.5g 2.0000E-02 Reactor Trip - RT3 - Seismic Event - 0.5g I RTD RTE RTF RTG 8.0510E-02 Reactor Trip - RTl - Seismic Event - 0.79 8.4300E-02 Reactor Trip - RT2 . Seismic Event - 0.7g 8.0000E-02 Reactor Trip - RT3 - Seismic Event - 0.7g 2.4040E.01 Reactor Trip - RTl - Seismic Event - 1.0g RTH 2.4350E-01 Reactor Trip - RT2 - Seismic Event - 1.0g RT! 2.4000E-01 Reactor Trip - RT3 - Seismic Event - 1.0g I I I I I I I I I 1251Pil1985 3-73
TABLE 3-4 (continued) Sheet 7 of 21 SERVICE WATER SYSTEM Top Event Val ue Definition Split Fraction Service Water - Train A WA1 5.3090E-03 Service Water A - after SI - Offsite Power Avail. WAA 5.0590E-03 SW Train A - Single Train - After SI WA2 3.5440E-04 SW Train A - no SI - Offsite Power Available WAB 3.3630E-04 SW Train A - Single Train - No SI WA3 1.8030E-02 SW Train A - LOSP WA4 1.6770E-02 SW Train A - Single Train - LOSP WA5 1.0000E+00 SW Train A - Guaranteed Failure Service Water - T.ein B WB1 5.1110E-03 SW Train B - after SI - Offsite Power Avail. WBA 4.3320E-02 SW Train B - after SI - WA Failed WBD 5.0590E-03 SW Train B - Single Train - After SI WB2 3.3650E-04 SW Train B - no SI - Offsite Power Available WBB 5.7450E-02 SW Train B - no SI - WA Failed l WBE 3.3630E-04 SW Train B - Single Train - No SI E WB3 1.7310E-02 SW Train B - LOSP WBC 5.9010E-02 SW Train B - LOSP - WA Failed WB4 1.6770E-02 SW Train B - Single Train - LOSP WB5 1.0000E+00 SW Train B - Guaranteed Failure Service Water System Results WC1 2.5020E-04 SW System - After SI WC2 1.8160E-05 SW System - No SI WC3 1.2580E-03 SW System - LOSP WC4 5.0590E-03 SW System - After SI - Single Train WC5 3.3630E-04 SW System - No SI - Single Train WC6 1.6770E-02 SW System - LOSP - Single Train Service Water System Recovery SRI 4.4570E-03 Service Water Recovery - 1 SR2 6.5920E-06 Service Water Recovery - 2 I I I I I I
~
1251P111985
F L TABLE 3-4 (continued) Sheet 8 of 21 g PRIMARY COMPONENT COOLING WATER Top Event Value Definition Split Fraction f Primary Component Cooling System Results Pl A 5.4480E-06 PCC - Boundary Condition lA P2A 6.2680E-04 PCC - Boundary Condition 2A r PIB 2.2740E-05 PCC - Boundary Condition IB P2B 7.4460E-04 PCC - Boundary Condition 2B PIC 5.4200E-06 PCC - Boundary Condition 1C P2C 5.9700E-04 PCC - Boundary Condition 2C
~
Reactor Coolant Pumps Thermal Barrier Cooling
~
RlA 1.4370E-04 RCP - Boundary Condition I A q R1B 2.8050E-04 RCP - Boundary Condition IB R2A 7.6070E-04 RCP - Boundary Condition 2A R2B 6.7900E-03 RCP - Boundary Condition 2B PCC Area Ventilation System
~
Fl B 7.7200E-06 FH - Boundary Condition IB F2B 2.1160E-03 FH - Boundary Condition 2B , Primary Component Cooling - Train A PA1 6.3210E-04 PCC - Train A - No P Signal - Offsite Power Available PA2 2.8910E-03 PCC - Train A - LOSP I l PA3 PA4 PAS 6.0230E-04 PCC - Train A - P Signal - Offsite Power Available 6.2680E-04 PCC - Train A - Single Train - No P Signal 2.8600E-03 PCC - Train A - Single Train - LOSP 1 PA6 5.9700E-04 PCC - Train A - Single Train - P Signal I PA7 PAA PAB PAC 1.0000E+00 PCC - Train A - Guaranteed Failure 1.0630E-02 PCC - Train A - Seismic Event - 0.3g 1.0620E-02 PCC - Train A - Seismic Event - 0.3g - Single Train 5.0630E-02 PCC - Train A - Seismic Event - 0.4g l PAD 5.0590E-02 PCC - Train A - Seismic Event - 0.4g - Single Train I PAE PAF PAG 1.2060E-01 PCC - Train A - Seismic Event - 0.5g 1.2060E-01 PCC - Train A - Seismic Event - 0.59 - Single Train 2.8060E-01 PCC - Train A - Seismic Event - 0.7g l PAH 2.8050E-01 PCC - Train A - Seismic Event - 0.79 - Single Train PAI 5.1060E-01 PCC - Train A - Seismic Event - 1.0g 'I PAJ PAK 5.1030E-01 PCC - Train A - Seismic Event - 1.0g - Single Train 1.2890E-02 PCC - Train A - Seismic Event - 0.3g - Af ter LOSP PAL 1.2830E-02 PCC - Train A - Seismic Event - 0.3g - Single Train PAM 5.2880E-02 PCC - Train A - Seismic Event - 0.4g - Af ter LOSP I PAN PA0 PAP PAQ 5.2710E-02 PCC - Train A - Seismic Event - 0.4g - Single Train 1.2290E-01 PCC - Train A - Seismic Event - 0.5g - After LOSP 1.2250E-01 PCC - Train A - Seismic Event - 0.5g - Single Train 2.8290E-01 PCC - Train A - Seismic Event - 0.7g - Af ter LOSP 3 PAR 2.8200E-01 PCC - Train A - Seismic Event - 0.7g - Single Train g , PAS PAT 5.1280E-01 PCC - Train A - Seismic Event - 1.0g - After LOSP 5.1140E-01 PCC - Train A - Seismic Event - 1.0g - Single Train Primary Component Cooling Train - B PB1 6.2840E-04 PCC - Train B - No P Si nal - Offsite Power Available PBA 1.0060E-02 PCC - Train B - No P Signal - Off. Power - PA fail PB2 2.8710E-03 PCC - Train B - LOSP PBB 8.0700E-03 PCC - Train B - LOSP - PA failed PB3 5.9860E-04 PCC - Train B - P Si nal - Offsite Power Available PBC 1.1170E-02 PCC - Train B - P Si nal - Off. Power - PA failed PB4 6.2680E-04 PCC - Train B - Sin e Train - No P Signal PBS 2.8600E-03 PCC - Train B - Sin le Train - LOSP I PB6 5.9700E-04 PCC - Train B - Sin le Train - P Signal PB7 1.0000E+00 PCC - Train B - Guaranteed Failure 1251Pil1985
I TABLE 3-4 (continued) Sheet 9 of 21 PRIMARY COMPONENT COOLING WATER Top Event Value Definition Split Fraction PBD 6.3470E-04 PCC - Train B - Seismic Event - 0.3g PBE 9.4700E-01 PCC - Train B - Seismic Event - 0.3g - PA failed PBF 1.0620E-02 PCC - Train B - Seismic Event . 0.3g - Single Train PBG 6.6150E-04 PCC - Train B - Seismic Event - 0.4g PBH 9.8790E-01 PCC - Train B - Seismic Event - 0.4g - PA failed PBI 5.0590E-02 PCC - Train B - Seismic Event - 0.4g - Single Train PBJ 7.1430E-04 PCC - Train B - Seismic Event - 0.5g PBK 9.9480E.01 PCC - Train B - Seismic Event - 0.5g - PA failed PBL 1.2050E-01 PCC - Train B - Seismic Event - 0.5g - Single Train PBM 8.7360E.04 PCC - Train B - Seismic Event - 0.7g PBN 9.9770E-01 PCC - Train B - Seismic Event - 0.7g - PA failed PB0 2.8040E.01 PCC - Train B - Seismic Event - 0.7g - Single Train PBP 1.2860E-03 PCC - Train B - Seismic Event - 1.0g PBQ 9.9870E-01 PCC - Train B - Seismic Event - 1.0g . PA failed PBR 5.1030E-01 PCC - Train B - Seismic Event - 1.0g - Sir.31e Train PBS 2.9010E-03 PCC - Train B - Seismic Event - 0.3g - Af te- LOSP PBT 7.8800E-01 PCC - Train B - Seismic Event - 0.3g - PA f. iled PBU l .2830E-02 PCC - Train B - Seismic Event - 0.39 - Sing' e Train PBV 3.0230E-03 PCC - Train B - Seismic Event - 0.4g - Afte LOSP PBW 9.4670E-01 PCC - Train B - Seismic Event - 0.4g - PA failed PBX 5.2710E-02 PCC - Train B - Seismic Event - 0.4g - Single Train PBY 3.2650E.03 PCC - Train B - Seismic Event - 0.5g - Af ter LOSP PBZ 9.7680E-01 PCC - Train B - Seismic Event - 0.5g - PA failed PCA 1.2250E-01 PCC - Train B - Seismic Event - 0.5g - Single Train PCB 3.9940E-03 PCC - Train B - Seismic Event - 0.7g - Af ter LOSP PCC 9.8980E-01 PCC - Train B - Seismic Event - 0.79 - PA failed PCD 2.8200E-01 PCC - Train B - Seismic Event - 0.7g - Single Train PCE 5.8840E.03 PCC - Train B - Seismic Event - 1.0g - After LOSP PCF 9.9440E-01 PCC - Train B - Seismic Event - 1.0g - PA failed PCG 5.1140E-01 PCC - Train B - Seismic Event - 1.0g - Single Train I I I I I 1251Pll1985 3-76
E TABLE 3-4 (continued) Sheet 10 of 21 ENCLOSURE BUILDING YENTILATION SYSTEM Top Event Value Definition Split Fraction EAH System EH1 1.8900E.05 Enclosure Building Ventilation (EAH) . T si nal or GT EH2 7.5800E.03 EAH - LOSP and One AC Bus (Single PCC Train EH3 8.2000E-04 EAH - Single T signal EH4 8.3800E-03 EAH - Single T signal, Single PCC Train EHS 1.4700E.04 EAH - LOSP EH6 1.0000E+00 EAH - Guaranteed Failure (Both T NOAC) EAH in Long Term Trees CVI 0.0000E.01 EAH - Operatin (Yes) - LT Trees CV2 1.0000E+00 EAH - Operatin (No) - LT Trees E E E E E E 1251Pil1985 3-77
I TABLE 3-4 (continued) Sheet 11 of 21 EMERGENCY FEEDWATER SYSTEM Top Event Val ue Definition Split Fraction E Emergency Feedwater Recovery
.-.......................-- ....... g FR1 6.6410E-01 EFW Recovery - 1 (EFR1 - TDFP) 3 FR2 1.5130E-02 EFW Recovery - 2 (EFR2 - SFP) g Emergency Feedwater System EFl 4.1620E.04 EFW Normal Configuration - No Startup Feed Pump EF2 4.7750E-02 Turbine Driven Pump Only EF3 7.0990E-06 EFI
- Startup Feed Pump EF4 7.0550E-04 EF2
- Startup Feed Pump EF5 6.0070E-03 Motor Driven Feed Pump Only EF6 5.8660E-03 EFW Feeding Both Steam Generators - ATWS EF7 1.4390E-02 Startup Feed Pump Only EF8 1.0000E+00 Guaranteed Failure EFA 4.1620E-04 EF1 + SC1 : (SCI = Atmos. Relief
- Steam Dump)
EFB 5.9930E-04 EFl + SC2 : (SC2 = Atmos. Relief Only) EFC 2.4020E-02 0.5
- EFl + 0.5
- EF2 + SCl EFD 2.4200E-02 0.5
- EF1 + 0.5
- EF2 + SC2 EFE 6.1890E-03 EFS + SC2 : TT Fails, No TD Pump EFF 5.0310E-01 0.5
- EF5 + 0.5 + SC2 : TT Falls, Loss of One signal EFG 7.1550E.06 EF3 + SCI : LOMF Events EFH 3.5370E-04 0.5 * (EF3 + EF4) + SCl :
EFI 1.4390E-02 EF7 + SCl : Startup Pump Only, No EFW Actuation
~
EFJ 5.8660E-03 EF6 + SCl : EFW ATWS EFK 2.6830E-02 0.5*(EF6+EF6A)+SCl : EFW ATWS, Loss Of One Signal EFL 4.8090E.02 EF6A + SC2 : LOAC - ATWS, No MD Pump Emergency Feedwater System - Seismic Events EGl 3.0400E-02 EFW : Seismic Event 0.2g EG2 7.6310E-02 TDP Only : Seismic Event 0.2g EGS 3.5830E-02 Motor Driven Feed Pump Only : Seismic Event 0.2g EG6 3.5690E-02 EFW Feeding Both Steam Generators - ATWS : Seis. 0.2g EGA 3.0400E-02 EFl + SCl : Seismic Event 0.2g EGB 3.0580E-02 EF1 + SC2 : Seismic Event 0.2 EGC 5.3300E-02 0.5
- EF1 + 0.5
- EF2 + SCI :gSeismic Event 0.2g EGD 5.3480E-02 0.5
- EFl + 0.5
- EF2 + SC2 : Seismic Event 0.2g EGE 3.6000E-02 EF5 + SC2 : TT Fails. No TD Pump : Seismic Event 0.2g E EGF EGJ 5.1800E-01 0.5
- EF5 + 0.5 + SC2 : Seismic Event 0.2g 3.5690E-02 EF6 + SCl : Seismic Event 0.2g g
EGK 5.6030E-02 0.5 * (EF6 + EF6A) + SCI : Seismic Event 0.2g EGL 7.6640E-02 EF6A + SC2 : Seismic Event 0.2g EH1 1.1040E-01 EFW : Seismic Event 0.3g EH2 1.5250E-01 TDP Only : Seismic Event 0.3g EHS 1.1540E.01 Motor Driven Feed Pump Only : Seismic Event 0.3g EH6 1.1520E-01 EFW Feeding Both Steam Generators - ATWS : Seis. 0.3g EHA 1.1040E.01 EFl + SCl : Seismic Event 0.3g E 1.1050E-01 EFl + SC2 : Seismic Event 0.3g EHB EHC 1.3140E.01 0.5
- EFl + 0.5
- EF2 + SC1 : Seismic Event 0.3g E EHD 1.3150E-01 0.5
- EF1 + 0.5
- EF2 + SC2 : Seismic Event 0.3g EHE 1.1550E-01 EF5 + SC2 : TT Fails. No TD Pump : Seismic Event 0.3g EHF 5.5780E-01 0.5
- EF5 + 0.5 + SC2 : Seismic Event 0.3g EHJ 1.1520E.01 EF6 + SCI : Seismic Event 0.3g EHK 1.3390E.01 0.5 * (EF6 + EF6A) + SCI : Seismic Event 0.3g EHL 1.5280E-01 EF6A + SC2 : Seismic Event 0.3g Emergency Feedwater System - Seismic Events E!1 2.3030E-01 EFW : Seismic Event 0.4g EI2 2.6670E.01 TDP Only : Seismic Event 0.4g EIS 2.3460E-01 Motor Driven Feed Ptap Only : Seismic Event 0.4g EI6 2.3450E-01 EFW Feeding Both Steam Generators - ATWS : Sets. 0.4g
~
1251Pil1985
l TABLE 3-4 (continued) Sheet 12 of 21 EMERGENCY FEEDWATER SYSTEM Top Event Value Definition I Split Fraction EIA EIB 2.3030E-01 EFl + SCI : Seismic Event 0.4g 2.3040E.01 EF1 + SC2 : Seismic Event 0.4 EIC 2.4850E.01 0.5
- EF1 + 0.5
- EF2 + SCI :gSeismic Event 0.4g EID 2.4860E-01 0.5
- EFl + 0.5
- EF2 + SC2 : Seismic Event 0.4g EIE 2.3470E.01 EFS + SC2 : TT Fails No TD Pump : Seismic Event 0.4g EIF 6.1740E 01 0.5
- EF5 + 0.5 + SC2 : Seismic Event 0.4g EIJ 2.3450E-01 EF6 + SCl : Seismic Event 0.4g I EIK EIL EJ1 EJ2 2.5060E-01 0.5 * (EF6 + EF6A) + SCl : Seismic Event 0.4g 2.6700E-01 EF6A + SC2 : Seismic Event 0.4g 3.4020E-01 EFW : Seismic Event 0.5g 3.7150E-01 TDP Only : Seismic Event 0.5g I EJ5 EJ6 EJA EJB 3.4400E.01 Motor Driven Feed Pump Only : Seismic Event 0.5g 3.4390E.01 EFW Feeding Both Steam Generators . ATWS : Seis. 0.5g 3.4020E.01 EFl + SCI : Seismic Event 0.5g 3.4040E-01 EF1 + SC2 : Seismic Event 0.5g EJC 3.5580E-01 0.5
- EFl + 0.5
- EF2 + SCI : Seismic Event 0.5g I EJD EJE EJF EJJ 3.5590E.01 0.5
- EF1 + 0.5
- EF2 + SC2 : Seismic Event 0.5g 3.4410E.01 EF5 + SC2 : TT Fails N 6.7200E.01 0.5
- EFS + 0.5 + SCE :o Seismic 3.4380E-01 EF6 + SCI : Seismic Event 0.5g TD Pump : Seismic Event 0.5g Event 0.5g I EJK EJL EKI EK2 3.5770E.01 0.5 * (EF6 + EF6A) + SCl : Seismic Event 0.5g 3.7170E 01 EF5A + SC2 : Seismic Event 0.5g 5.3020E.01 EFW : Seismic Event 0.7g 5.5240E-01 TDP Only : Seismic Event 0.7g EKS 5.3280E.01 Motor Driven Feed Pump Only : Seismic Event 0.79 I EKti EKA EKB EKC 5.3280E-01 EFW Feeding Both Steam Generators - ATWS : Seis. 0.7g 5.3020E.01 EFl + SC1 : Seismic Event 0.7g 5.3020E.01 EF1 + SC2 : Seismic Event 0.7g 5.4120E-01 0.7
- EFI + 0.7
- EF2 + SCl : Seismic Event 0.7g I EKD EKE EKF EKJ 5.4130E-01 0.7
- EFl + 0.7
- EF2 + SC2 : Seismic Event 0.79 5.3290E.01 EFS + SC2 : TT Fails N 7.6640E-01 0.7
- EFS + 0.7 + SCE :o TD 5.3270E.01 EF6 + SC1 : Seismic Event 0.7g
- Seismic PianpEvent Seismic 0.7g Event 0.79 EKK 5.4260E-01 0.7*(EF6 + EF6A) + SCI : Seismic Event 0.7g EKL 5.5260E.01 EF6A + SC2 : Seismic Event 0.7g EL1 7.3010E.01 EFW : Seismic Event 1.0g EL2 7.4280E-01 TDP Only : Seismic Event 1.0g ELS 7.3160E-01 Motor Driven Feed Pump Only : Seismic Event 1.0g I EL6 ELA ELB ELC 7.3163E.01 EFW Feeding Both Steam Generators - ATWS : Seis. 1.0g 7.3010E-01 EFl + SCI : Seismic Event 1.0g 7.3010E.01 EFl + SC2 : Seismic Event 1.0 7.3640E.01 0.5
- EFl + 0.5
- EF2 + SCl :gSeismic Event 1.0g l ELD 7.3650E-01 0.5
- EFl + 0.5
- EF2 + SC2 : Seismic Event 1.0g '
I ELE ELF ELJ ELK 7.3160E.01 EF5 + SC2 : TT Fails, No TD Pump : Seismic Event 1.0g 8.6580E.01 0.5
- EF5 + 0.5 + SC2 : Seismic Event 1.0g 7.3150E-01 EF6 + SCI : Seismic Event 1.0 7.3720E.01 0.5 * (EF6 + EF6A) + SCl : Se smic Event 1.0g I ELL 7.4290E.01 EF6A + SC2 : Seismic Event 1.0g I
I I 1251Pil198 ; 3-79
I TABLE 3-4 (continued) Sheet 13 of 21 MAIN STEAM SYSTEM FUNCTIONS Top Event Value Definition Split Fraction MSIV Isolation MSI 8.9800E-05 MSIV Isolation - SLBOC or Turbine Trip Failure MS2 8.9800E-05 MSIY Isolation - SLBIC MS3 1.0000E-03 Main Steam Line Intact - SL Tree - SGTR Turbine Trip TTl 4.4900E-06 Turbine Trip (non - TT events) TT2 0.0000E-01 Turbine Trip (TT events) - Guaranteed Success TT3 1.0000E+00 Turbine Trip - Guaranteed Failure TT4 4.0320E-10 Turbine Trip
- MS3 Safety Valve Action =
SV1 1.0000E+00 Safety Valve Action for ATWS Steam Dump System 501 0.0000E-01 Steam dump Available - SL Tree - SGTR SD2 1.0000E+00 Steam dump Available - SL Tree - SGTR - GF MSIV and Bypass Valves SGTR - SL Tree IVI 1.5200E-03 MSIY and Bypass Isolated - SL Tree - SGTR IV2 1.0000E+00 MSIY and Bypass Isolated - SL Tree - SGTR - GF Steam Generator Isolation SGTR - SL Tree SGI 0.0000E-01 Steaming SG Isolated - SL Tree - SGTR Steam Generator Atmospheric Valves SGTR - SL Tree A01 4.2700E-03 Atmos. Relief Valves Open - SL Tree - SGTR A02 1.0000E+00 A*mos. Relief Valves Open - SL Tree - SGTR - GF ACI 2.5000E-02 Atmos. Relief Valves Close - SL Tree - SGTR AC2 1.0000E+00 Atmos. Relief Valves Close - SL Tree - SGTR - GF All 1.0000E-02 Atmos. Relief Valve Isolated - SL Tree - SGTR AI2 1.0000E+00 Atmos. Relief Valve Isolated - SL Tree - SGTR - GF Steam Generator Safety Valves SGTR - SL Tree SN1 1.0000E-01 Safety Valves Not Demanded - SL Tree - SGTR SN2 9.0000E-01 Safety Valves Not Demanded - SL Tree - SGTR (OR Fail) 5 01 9.2800E-03 Safety Valves Open and Close - SL Tree - SGTR S02 2.0100E-01 Safety Valves Open and Close - SL Tree -SGTR (OR Fall) Steam Generator Secondary Leak SGTR - SL Tree SL1 1.0800E-04 No Secondary Side Leak to Atmosphere - SGTR SL2 5.4600E-03 No Secondary Side Leak to Atmosphere - SGTR SL3 4.2000E-04 No Secondary Side Leak to Atmosphere - SGTR SL4 5.7700E-03 No Secondary Side Leak to Atmosphere - SGTR (OR Fail) SL5 1.1100E-02 No Secondary Side Leak to Atmosphere - SGTR (OR Fail) SL6 2.0100E-01 No Secondary Side Leak to Atmosphere - SGTR (OR Fail) SL7 9.3900E-03 No Secondary Side Leak to Atmosphere - SGTR SS1 5.3600E-03 No Secondary Side Leak to Atmosphere - SGTR Secondary Cooling Function SCI 5.6500E-08 Atmos Relief Valves & Cond Steam Dump SC2 1.8200E-04 Atmos Relief Valves Only 1251Pil1985
I TABLE 3-4 (continued) i l Sheet 14 of 21 EMERGENCY CORE COOLING SYSTEMS Top Event Val ue Def f rition Split Fraction 1 Refueling Water Storage Tank ' RW1 2.6600E-08 Refueling Water Storage Tank (RWST) . LLOCA RW2 5.3300E-08 RWST - MLOCA RW3 1.6000E-07 RWST - Other Events RWA 2.0000E.02 RWST - Seismic Event 0.3g . General Transient RWB 6.0000E-02 RWST - Seismic Event 0.4g - General Transient RWC 1.4000E-01 RWST - Seismic Event 0.5g . General Transient RWD 3.4000E-01 RWST - Seismic Event 0.7g . General Transient RWE 6.2000E-01 RWST - Seismic Event 1.0g - General Transient RWF 3.4000E.01 RWST - Seismic Event 0.79 - LLOCA RWG 6.2000E.01 RWST - Seismic Event 1.0g - LLOCA RWST Outlet Valves RAI 3.3500E.05 RWST Outlet Valve - Train A - LLOCA I RA2 RA3 RBI 3.3600E-05 RWST Outlet Yalve - Train A - MLOCA 3.3900E-05 RWST Outlet Valve - Train A - SLOCA, etc. 3.3500E-05 RWST Outlet Valve - Train B - LLOCA RB2 3.3600E-05 RWST Outlet Valve - Train B - MLOCA RB3 3.3900E-05 RWST Outlet Valve - Train B - SLOCA, etc. High Pressure Injection For MLOCA H11 2.4330E-05 High Pressure Injection (HPI) - MLOCA l I H12 H13 H14 3.2130E-02 HPI - MLOCA - Loss of One AC Power Bus 1.3470E-02 HPI - MLOCA . Loss of One PCC Train 1.0000E+00 HPI . MLOCA - Guaranteed Failure I H21 H22 High Pressure Injection For SLOCA's etc. 1.0310E-06 HPI . SLOCA etc. 1.9520E-04 HPI - SLOCA - Loss of One AC Power Bus H23 6.4090E.05 HPI - SLOCA - Loss of One PCC Train H24 1.0000E+00 HP1 - SLOCA - Guaranteed Failure H25 2.0000E.02 HPI - SLOCA etc. - Seismic Event 0.3g H2A 2.0190E-02 SLOCA - Loss of One AC Power Bus - Sets 0.3g H28 2.0060E-02 SLOCA . Loss of One PCC Train - Seis 0.3g H26 6.9990E.02 HPI - SLOCA etc. - Seismic Event 0.4g H2C 7.0180E.02 SLOCA - Loss of One AC Power Bus - Seis 0.4g H2D 7.0060E.02 SLOCA - Loss of One PCC Train . Seis 0.4g H27 1.4000E-01 HPI . SLOCA etc. - Seismic Event 0.5g H2E 1.4020E-01 SLOCA - Loss of One AC Power Bus - Seis 0.5g H2F 1.4010E-01 SLOCA - Loss of One PCC Train - Seis 0.5g H28 2.8000E-01 HPI - SLOCA etc. - Seismic Event 0.79 H2G 2.8010E-01 SLOCA - Loss of One AC Power Bus - Seis 0.7g H2H 2.8000E-01 SLOCA . Loss of One PCC Train - Seis 0.7g H29 4.6000E-01 HPI - SLOCA etc. - Seismic Event 1.0g H2! 4.6010E.01 SLOCA - Loss of One AC Power Bus - Seis 1.0g H2J 4.6000E.01 SLOCA - Loss of Die PCC Train - Seis 1.0g High Pressure Injection For ATWS Events H31 1.0580E-03 HPI - Anticipated Transients Without Scram (ATWS) H32 2.5180E-02 HPI - ATWS - Loss of One AC Power Train H33 6.5180E.03 HPI - ATWS - Loss of One PCC Train H34 1.0000E+00 HPI - ATWS - Guaranteed Failure H35 2.1040E-02 HPI - ATWS . Seismic Event 0.3g H3A 4.4680E-02 HPI - ATWS - Loss of One AC Power Train - Seis 0.3g H3B 2.6500E-02 HPI - ATWS - Loss of 01e PCC Train . Seis 0.3g H36 7.0980E.02 HPI - ATWS - Seismic Event 0.4g H3C 9.3420E.02 HPI - ATWS . Loss of One AC Power Train - Seis 0.4g H30 7.6440E-02 HPI - ATWS - Loss of One PCC Train . Seis 0.4g I 1251Pil1985 3-81
TABLE 3-4 (continued) Sheet 15 of 21 EMERGENCY CORE COOLING SYSTEMS Top Event Value Definition Split Fraction H37 1. C OE-01 HPI - ATWS - Seismic Event 0.5g H3E 1.6170E-01 HPI - ATWS - Loss of One AC Power Train - Seis 0.5g H3F 1.4640E-01 HPI - ATWS - Loss of One PCC Train - Seis 0.5g H38 2.8080E-01 HPI - ATWS - Seismic Event 0.7g H3G 2.5200E-02 HPI - ATWS - Loss of One AC Power Train - Seis 0.7g H3H 2.8620E-01 HPI - ATWS - Loss of One PCC Train - Seis 0.7g H39 4.6060E-01 HPI - ATWS - Seismic Event 1.0g H3! 4.7360E-01 HPI - ATWS - Loss of One AC Power Train - Seis 1.0g H3J 4.6600E.01 HPI - ATWS - Loss of One PCC Train - Seis 1.0g Low Pressure Injection - Train A LA1 1.2300E 02 LPI - LLOCA - Train A LA2 1.2250E 02 LPI - LLOCA - Train A - Loss of One AC Power Bus LA3 1.0000E+00 LPI - LLOCA - Train A - Guaranteed Failure , LA4 2.9110E-01 LPI - LLOCA - Train A - Seismic Event 0.7g LAS 2.8880E-01 LLOCA - Train A - Loss of 01e AC Power Bus - Seis 0.7g LA6 4.7030E-01 LPI - LLOCA - Train A - Seismic Event 1.0g LA7 4.6660E-01 LLOCA - Train A - Loss of One AC Power Bus - Seis 1.0g Low Pressure Injection - Train B LB1 8.0540E-03 LPI - LLOCA - Train B LBA 3.5330E-01 LPI - LLOCA - Train B - after LA fails LB2 1.2250E-02 LPI - LLOCA - Train B - Loss of One AC Power Bus LB3 1.0000E+00 LPI - LLOCA - Train B - Guaranteed Failure LB4 1.1220E-02 LPI - LLOCA - Train B - Seismic Event 0.7g
- LB0 9.7270E-01 LPI - LLOCA . Train B . after LA fails - Seis 0.7g l
LB5 2.8880E-01 LLOCA - Train B - Loss of One AC Power Bus - Sets 0.7g ! LB6 1.5020E-02 LPI - LLOCA - Train B - Seismic Event 1.0g l LBF 9.8310E.01 LPI - LLOCA - Train B after LA fails - Seis 1.0g l LB7 4.6660E-01 LLOCA - Train B - Loss of One AC Power Bus - Seis 1.0g l Low Pressure Miniflow - MLOCA's - Train A Lil 1.5390E-02 Low Pressure Miniflow (LPM) - MLOCA - Train A ! L12 1.4880E.02 LPH - MLOCA etc - Train A - Loss of One AC Power Train L13 1.5390E-02 LPM - SLOCA etc - Train A Ll4 1.4880E-02 LPM - SLOCA etc - Train A - Loss of One AC Power Train L15 1.0000E+00 LPM - Train A - Guaranteed Failure Low Pressure Miniflow - MLOCA's - Train B L21 1.5110E-02 LPM - MLOCA etc - Train B L2A 3.2960E-02 LPM - MLOCA etc - Train B - Af ter Train A failed L22 1.4880E-02 LPH - MLOCA etc - Train B - Loss of One AC Power Train L23 1.5270E-02 LPH - SLOCA etc . Train B L2C 3.5260E-02 LPM . SLOCA etc - Train B - Af ter Train A failed L24 1.4880E-02 LPM - SLOCA etc - Train B - Loss of One AC Power Train L25 1.0000E+00 LPH - Train B - Guaranteed Failure RHR Shutdown Cooling LR1 1.2420E.03 RHR Shutdown Cooling - Both Trains LR2 '.1830E-02 RHR Shutdown Cooling - Single Train LR3 1.0000E+00 RHR Shutdown Cooling - Guaranteed Failure ( Containment Sump Isolation Valves - Train A sal 5.0850E-03 Cont. Sump (CS) Isolation Valve - Train A - LLOCA SA2 4.8600E-03 CS ! sol. Valve - Train A - Single Train . LLOCA e i SA3 4.5160E.03 CS Isol. Valve - Train A . SLOCA etc. g i 1251P111985
I TABLE 3-4 (continued) I EMERGENCY CORE COOLING SYSTEMS Sheet 16 of 21 Top Event Value Definition Split Fraction SA4 4.2970E-03 CS ! sol. Valve - Train A - Single Train - SLOCA etc. SAS 1.0000E+00 CS Isol. Valve - Train A - Guaranteed Failure Containment Sump Isolation Valves - Train B SBl 4.8650E-03 CS Isol. Valve - Train B - LLOCA SBA 4.4170E-02 CS Isol. Valve - Train B - LLOCA - SA failed SB2 4.8600E-03 CS Isol. Valve - Train B - Single Train - LLOCA SB3 4.3160E-03 CS Isol . Valve - Train B - SLOCA SBC 4.8490E-02 CS Isol. Valve - Train B - SLOCA - SA failed SB4 4.2970E-03 CS Isol . Valve - Train B - Single Train - SLOCA etc. SB5 1.0000E+00 CS Isol. Valve - Train B - Guaranteed Failure Long Term LP Recirculation - Train A 1.3650E-03 Long Term Low Pressure Recirculation (LPR) - Train A I LC1 LC2 1.1440E-03 LPR - Train A - Loss of One Train of AC Power or PCC LC3 1.0000E+00 LPR - Train A - Guaranteed Failure Long Term LP Recirculatio'n - Train B LD1 1.1450E-03 LPR - Train B LDA 1.6220E-01 LPR - Train B - After LC failed LD2 1.1440E-03 LPR - Train B - Loss of One Train of AC Power or PCC I 1.0000E+00 LPR - Train B - Guaranteed Failure LD3 Low Pressure Recirculation Heat Exchanger Cooling - Train A HA1 4.5320E-03 LPR - Heat Exchanger Train A - LLOCA HA2 4.3130E-03 LPR - HExch. Train A - One Train PCC, AC Power failed HA3 1.0000E+00 LPR - Heat Exch. Train A - Guaranteed Failure Low Pressure Recirculation Heat Exchanger Cooling - Train B HB1 4.3330E-03 LPR - Heat Exchanger Train B - LLOCA HBA 4.8340E-02 LPR - Heat Exch. Train B - After HA failed HB2 4.3130E-03 LPR - HExch. Train B - One Train PCC, AC Power failed HB3 1.0000E+00 LPR - Heat Exch. Train B - Guaranteed Failure Low Pressure Recirculation - Train A l LSI 1.1310E-03 LPR Train A (Includes Heat Exchanger) L52 9.6010E-04 L53 1.2400E-03 LPR HPR Train Train A - Loss of One A (Includes Heat Train PCC,)etc. Exchanger L54 1.0690E-03 HPR Train A - Loss of One Train PCC, etc. L55 1.0000E+00 LPR/HPR Train A - Guaranteed Failure Low Pressure Recirculation - Train B L61 9.6120E-04 LPR Train B (Includes Heat Exchanger) l L6A 1.5150E-01 LPR Train B - After Train B fails l L62 9.6010E-04 LPR Train B - Loss of One Train PCC, etc. i L63 1.0700E-03 HPR Train B (Includes Heat Exchanger) L6C 1.3830E-01 HPR Train B - After Train B fails LE4 1.0690E-03 HPR Train B - Loss of One Train PCC, etc. L65 1.0000E+00 LPR/HPR Train B - Guaranteed Failure High Pressure Recirculation RCl 2.7400E-08 High Pressure Recirculation (HPR) HPI Pumps RC2 8.6100E-07 HPR - Loss of Train A LPR RC3 2.8500E-08 HPR - Loss of Train B LPR 1251P111985 3-83
I TABLE 3-4 (continued) Sheet 17 of 21 EMERGENCY CORE COOLING SYSTEMS Top Event Value Definition Split Fraction RC4 1.1800E-06 HPR - Loss of One Train of AC Power RC5 1.1100E-06 HPR - Loss of One Train of PCC RC6 0.0000E-01 HPR - Guaranteed Failure Reactor Pressure for Recirculation RPl 0.0000E-01 180 psig RP2 1.0000E+00 Reactor Pressure l 180 psig Reactor Pressure Water in Containment WS1 0.0000E-01 Water in Containment (Yes) WS2 1.0000E+00 Water in Containment (No) I 1251Pil1985 3-84
I TABLE 3-4 (continued) I REACTOR COOLANT SYSTEM FUNCTIONS Sheet 18 of 21 Top Event Value Definition I Split Fraction
....................-----.-......... - .........==
No RPV Rupture I RV1 RV2 0.0000E-01 No Reactor Pressure Vessel Failure 1.0000E-02 Reactor Pressure Vessel Failure No RCP Seal LOCA NL1 0.0000E-01 No Reactor Coolant Pump Seal Failure NL2 1.0000E+00 Guaranteed Reactor Coolant Pump Seal Failure Primary Pressure Relief PSl 1.5600E-03 Primary Pressure Relief . ATWS - 2/2 PORY PS2 9.8500E-04 Primary Pressure Relief - ATWS - 1/2 PORV PS3 5.2600E-03 Primary Pressure Relief - ATWS - 1/1 PORY PS4 1.0000E+00 Primary Pressure Relief ATWS . Guaranteed Failure PS5 0.0000E.01 Primary Pressure Relief ATWS . Not Required PORY and Safety Valves Reseat P21 5.8600E-02 Safety and Relief Yalves Rescat - ATWS PORV's PRI 1.0500E-02 PORV in Feed and Bleed PR2 5.7200E.04 PORY Lif t . ATWS - Chemical Shutdown 1/2 PORY PR3 4.2700E.03 PORY Lift - ATWS . Chemical Shutdown 1/1 PORY PR4 1.0000E+00 Feed and Bleed Guaranteed Failure PR5 2.3500E.02 PRl + OR : OR - Oper. Initiates Feed and Bleed PR6 1.3570E-02 PR2 + OR : OR - Oper. Initiates Feed and Bleed PR7 1.7270E-02 PR3 + OR : OR - Oper. Initiates Feed and Bleed Plant Power Level PL1 5.0000E-01 P1 ant Power Level - ATWS I I I 1251Pll1985 3-85
l l l TABLE 3-4 (continued) Sheet 19 of 21 CONTAINMENT BUILDING SPRAY Top Event Value Definition Split Fraction j CBS Injection - Train A CA1 1.0920E-02 CBS Injection - Train A CA2 1.0200E-02 CBS Inj. - Train A - Loss of One AC Power Bus, etc. CA3 1.0000E+00 CBS Inj. - Train A - Guaranteed Failure CBS Injection - Train B CBI 1.0310E-02 CBS Injection - Train B CBA 6.6390E-02 CBS Inj. - Train B - After CA fails CB2 1.0200E-02 CBS Inj. - Train B - Loss of One AC Power Bus, etc. CB3 1.0000E+00 CBS Inj. - Train B - Guaranteed Failure 1 l CBS Recirculation - Train A ( XAl 6.6170E-03 CBS Recirculation - Train A i XA2 6.3900E-03 CBS Recirc. - Train A - Loss of One AC Power Bus, etc. XA3 1.0000E+00 CBS Recirc. - Train A - Guaranteed Failure XA4 1.0980E-02 CBS Recirc. - Train A - No cooling Required XA5 1.0720E-02 CBS Recirc. - Train A - No cooling Required - Sgl Trn CBS Recirculation - Train B - XB1 6.4330E-03 CBS Recirculation - Train B XBA 3.4310E-02 CBS Recirc. - Train B - After XA fails XB2 6.3900E-03 CBS Recirc. - Train B - Loss of One AC Power Bus, etc. XB3 1.0000E+00 CBS Recirc. - Train B - Guaranteed Failure XB4 1.0840E-02 CBS Recirc. - Train B - No cooling Required XBD 2.4430E-02 CBS Recirc. - Train B - No cooling Required - XA fail XB5 1.0720E-02 CBS Recirc. - Train B - No cooling kequired - Sgl Trn CBS Recirculation Cooling - Train A val 4.3950E-03 CBS Recirc. Cooling - Train A f VA2 4.3600E-03 CBS Rectrc. Cooling - Train A - AC Lost to One Bus i VA3 1.0000E+00 CBS Recirc. Cooling - Train A - Guaranteed Failure VA4 0.0000E-01 CBS Recirc. - No Cooling Required - Train A CBS Recirculation Cooling - Train B VB1 4.3790E-03 CBS Recirc. Cooling - Train B l VBA 7.9520E-03 CBS Recirc. Cooling - Train B - After VA fails VB2 4.3600E-03 CBS Recirc. Cooling - Train B - AC Lost to One Bus VB3 1.0000E+00 CBS Recirc. Cooling - Train B - Guaranteed Failure VB4 0.0000E-01 CBS Recirc. - No Cooling Required - Train B CBS Recirculation - Train A X31 1.2270E-02 CBS Recirculation - Train A - LT1, LT2 etc. X32 1.1520E-02 CBS Recirc. - Train A - Loss of One AC Power Bus, etc. X33 1.0000E+00 CBS Recirc. - Train A - Guaranteed Failure CBS Recirculation - Train B [ X41 1.1670E-02 CBS Recirculation - Train B - LT1, LT2 etc. X4A 6.0580E-02 CBS Recirc. - Train B - After XA fails X42 1.1520E-02 CBS Recirc. - Train B - Loss of One AC Power Bus, etc. X43 1.0000E+00 CBS Recirc. - Train B - Guaranteed Failure l 1251P111985
I TABLE 3-4 (continued) Sheet 20 of 21 CONTAINMENT ISOLATION Top Event Val ue Definition I Split Fraction
..===== ___..... ..__.___..__.........___.__...........____..__....___
3" Penetrations CII 2.6710E-04 Containment Isolation (CI) . All Support Available I CI2 CI3 CI4 9.7140E-03 CI - Single S signal Train A 9.9860E-03 CI - Single S signal Train B 8.8450E-03 CI - LOSP, Loss of One 5 Signal CIS 4.3330E-03 CI . LOSP, Loss of One AC Bus CI6 9.8500E.03 CI . Loss of Either S Signal CI7 1.0000E+00 CI - Guaranteed Failure CIA 1.3350E.08 CI - LOSP, Both S. Loss of One AC Bus (Oper. Action) CIB 4.9250E.07 CI - Loss of Either S Signal (Oper. Action) CIC 4.4230E-07 CI - LOSP, Loss of One S Signal (Oper. Action) CID 2.1660E-07 CI - LOSP, Loss of One AC Bus (Oper. Action) CIE 5.0000E-05 CI . No AC Power (Oper. Action) l 3" Penetrations C21 4.7220E.04 C22 3.4950E-05 C2 Containment
- LOSP, BothBuildinfoss S, of One AC BusPurge Lines (C2) - All Support C23 6.4030E.03 C2 - Single S signal, Offsite Power Available C24 9.2090E-08 C2 - No AC Power C25 9.9990E-02 C2 - No S Signal, Offsite Power Available P01 1.0000E.01 Containment Purge Valves Open I
I I I 1251Pil1985 3-87
TABLE 3-4 (continued) Sheet 21 of 21 OPERATOR ACTIONS Top Event Value Definition Split Fraction OD1 2.6400E-02 Operator Action to Depressurize - MLOCA - 30 min 0D2 1.2900E-02 Operator Action to Depressurize - MLOCA - I hour 0D3 1.0000E+00 Operator Action to Depressurize - Guaranteed Failure 004 5.0000E-02 0041 in SGTR Tree - Pzr Spray and SG Depress. ODD 7.0000E-02 0D42 in SGTR Tree - Feed and Bleed Depress. 0D5 5.0000E-02 OD51 in SGTR Tree - No EFW Depress. ODE 9.0000E-02 OD52 in SGTR Tree - No HPI Depress. OM1 6.2400E-02 Operator Action to Control EFW Flow - Overcooling OM2 0.0000E-01 Operator Action to Control EFW Flow - Not Asked OfG 1.0000E+00 Operator Action to Control EFW Flow - GF OP1 2.2600E-02 Operator Action to Control HPI Flow - Overcooling OP2 0.0000E-01 Operator Action to Control HPI Flow - Not Asked OP3 1.0000E+00 Operator Action to Control HPI Flow - GF ORI 1.6900E-02 Operator Action - Feed and Bleed , SGTR Break Flow OR2 0.0000E-01 Operator Action - Feed and Bleed , Not Asked OR3 1.0000E+00 Operator Action - Feed and Bleed , GF OR4 5.0000E-02 Operator Action - Feed and Bleed , SGTR Break Flow ORS 2.5300E-02 Operator Action - Feed and Bleed , ORI + PRI ONI 1.0000E-06 Operator Action - Plant Stabilization ON2 1.3000E-02 Operator Action - Plant Stabilization - Dep SG's ON3 1.0000E+00 Operator Action - Plant Stabilization - GF 031 8.0000E-04 Operator Action - LPR or HPR in LTl 032 1.0000E+00 Operator Action - LPR or HPR in LTl - GF HEl 8.0080E-04 Operator Action - LPR - LLOCA HE2 8.0110E-04 Operator Action - LPR - LLOCA - Single Train HE3 1.0000E+00 Operator Action - LPR - LLOCA - GF HS1 8.0080E-04 Operator Action - Hot Leg Recirc. - LLOCA HS2 8.0110E-04 Operator Action - Hot Leg Recire. - Single Train HS3 1.0000E+00 Operator Action - Hot Leg Recirc. - GF OE1 6.7000E-04 Operator Action - Early SGTR Diagnosis OG1 2.2000E-02 Operator Action - SGTR - Isolate Failed Steam Gen. 0 51 1.0000E-04 Operator Action - Isolate SG - SGTR OS2 1.0000E+00 Operator Action - Isolate SG - SGTR - GF OTl 1.0000E+00 Operator Action - Manual Trip Turbine - ATWS - GF OT2 , 0.0000E-01 Operator Action - Manual Trip Turbine - ATWS - NA OH1 5.3100E-03 Operator Action - Manual Reactor Shutdown - ATWS OH2 1.0000E+00 Operator Action - Manu11 Reactor Shutdown - ATWS - GF OX1 1.3700E-01 Operator Action - Manual Reactor Scram - Support Tree OX2 1.0000E+00 Operator Action - Manu.11 Scram CF - Support Tree All 1.0000E-02 Operator Action - Isolate Leaking Relief Valve - SGTR I I 1251P111985 3-88
I TABLE 3-5. DEFINITION OF INITIATING EVENTS, TOP EVENTS, AND B0UNDARY CONDITIONS DEFINED IN THIS STUDY Sheet 1 of 3 Sp i Value Definition Fraction I VS 3.26-6 VS SEQUENCE Suction Line MOV Leakage > 150 GPM LR 0.09 Leakage < Relief Valve Capacity VO 4.8-5 Relief Valves Open PI 6.0-3 RHR Piping / Heat Exchanger Remains Intact SI 0.99 RHR Pumps Seals Remain Intact I L1 0.919 0.0 < Pump Seal Leak < 0.09 square inches L2 0.56 0.09 < Pump Seal Leak < 1.05 square inches L3 0.02 1.05 < Pump Seal Leak 12.6 square inches 01 6.5-3 Operator Diagnoses Event 02 1.0 Operator Terminates Valve Leakage I CSA 0.11 CBS Pumps Survive Vault Environment (seal leak 5 0.09 square inches) CSB 1.0 CB"S Pumps Survive Vault Environment (seal leak
> 0.09 square inches)
RSA 0.56 RHR Pumps Survive Vault Environment (seal leak 10.09 square inches) I RSB 1.0 RHR Pumps Survive Vault Environment (seal leak
> 0.09 square inches) l I SSA 0.11 SI Pumps Survive Vault Environment (seal leak 1 0.09 square inches)
SSB 1.0 SI Pumps Survive Vault Environment (seal leak l
> 0.09 square inches)
VC 0.1 Relief Valves Close NOTE: Exponentialnotationisigdicatedinabbreviatedform; i.e., 3.26-6 = 3.26 x 10 . 1251P112285 3-89 ,
I TABLE 3-5 (continued) Sheet 2 of 3 Value Definition Sp r tion 03A 1.0 Operator Establishes RWST Makeup (given that operator fails to diagnose the event) 03B 1.0 Operator Establishes RWST Makeup (given that seal leak > 2.6 square inches) 03C 4.9-3 Operator Establishes RWST Makeup (given that operator diagnoses event and seal leak < 2.6 square inches) VI SEQUENCE
- VI 4.5-6 Injection Line Check Valve Leakage > 150 GPM LR 0.093 Leakage < Relief Valve Capacity 02A 9.1-3 Operator Terminates Valve Leakage (given that operator diagnoses event) 02B 1.0 Operator Terminates Valve Leakage (given that operator failed to diagnose event)
CSA 0.1 CBS Pumps Survive Vault Environment (seal leak
< 0.09 square inches)
CSB 0.44 CBS Pumps' Survive Vault Environment (operator terminates interfacing LOCA and 0.09 square inches < seal leak < 1.05 square inches) CSC 1.0 CBS Pumps Survive Vault Environment (operator fails to terminate interfacing LOCA and seal leak > 0.09 square inches)
*0nly those top events that have different values from the VS sequence are listed.
NOTE: Exponential notation is indicated in abbreviated form; i.e., 4.9-3 = 4.9 x 10-3, I 1251P112285 3-90
TABLE 3-5 (continued) Sheet 3 of 3 Value Defi ni t. ion Sp r tion ~ VI SEQUENCE * (continued) CSD 0.75 CBS Pumps Survive Vault Environment (operator a terminates interfacing LOCA and 1.05 square inches < seal leak < 2.6 square inches) CSE 1.0 CBS Pumps Sursive Vault Environment (seal leak
> 2.6 square inches)
RSA 0.55 RHR Pumps Survive Vault Environment (seal leak
< 0.09 square inches)
RSB 0.85 RHR Pumps Survive Vault Environment (operator - terminates interfacing LOCA and 0.09 square inches < seal leak < 1.05 square inches) RSC 1.0 RHR Pumps Survive Vault Environment (seal leak
> 1.05 square inches)
SSA 0.1 SI Pumps Survive Vault Environment (seal leak ~
< 0.09 square inches)
SSB 0.33 SI Pumps Survive Vault Environment (cperator - terminates interfacing LOCA and 0.09 square inches < seal leak < 1.05 square inches) SSC 1.0 SI Pumps Survive Vault Environment (operator fails to terminate interfacing LOCA and seal leak > 0.09 square inches) SSD 0.64 SI Pumps Survive Vault Environment (operator terminates interfacing LOCA and 1.05 square inches < seal leak < 2.6 square inches) SSE 1.0 SI Pumps Survive Vault Environment (seal leak
> 2.6 square inches)
I
*0nly those top events that have different values from the VS sequences are listed here.
l 1251P112285 L
TABLE 3-6. PUMP ALIGNMENT Suction Discharge
"*E
- Normal Injection Mode Recirculation Mode Normal Injection Mode Recirculation Mode Operation of ECCS of ECCS Operation of ECCS of ECCS Positive Displacement i Volume Control Not part of Not part of RCS Cold Leg Not part of Not part of Charging Tank ECCS ECCS via Regenera- ECCS ECCS tive Heat Exchanger and RCP Seal Injection Centrifugal Charging 2 Volume Control RWST RHR Pump RCS Cold Leg Four RCS Cold legs Four RCS Cold Legs Tank Discharge via Regenera- via BIT and RCP via BIT and RCP tive Heat Seal Injection Seal Injection Exchanger and RCP Seal Injection High Pressure 2 N/A RWST RHR Pump N/A Four RCS Cold Legs Four RCS Cold Legs Y
e Safety Injection Discharge or Four RCS Hot Legs or Four RCS Hot Legs RHR (Low 2 Two RCS Hot RWST Containment Four RCS Cold Four RCS Cold Four RCS Cold Legs Pressure Injection) Le gs* Sump Legs
- Legs or Two RCS Hot Legs Containment 2 N/A RWST Containment N/A Containment Containment Spray Sump Spray Headers Spray Headers
- Normal alignment in the RHR mode.
NOTE: N/A = not applicable. 1251P111985 W M M M M M M M M M M
TABLE 3-7.
SUMMARY
OF V-SEQUENCE ANALYZED WITH MAAP l l 1. INITIAL CONDITIONS l , e Reactor coolant system at 100% power. e Containment conditions normal. e Auxiliary building conditions normal. l e RHR train A/B cross-connect line open. l
- 2. INITIATING EVENT e Simultaneous failure of both MOVs in the RHR suction path of RHR
, train A (or B) with a leak rate exceeding the capacity of the I low pressure RHR relief valves on both trains.
- 3. ACCIDENT PROGRESSION l
The 990-gpm capacity (at 495 psig) relief valves on the suction I e side of each RHR pump open on both trains (due to the open cross-connect). These relief valves relieve to the pressurizer relief tank inside the containment. e Pump seals fail on both RHR pumps (due to the open I cross-connect). The pump seal leak area is determined by assuming that all j nonmetallic parts of the pump seal assembly " disappear". However, the pump seal hold-down ring plate, which is held in place by four 3/4-inch diameter bolts, remains in place and l limits the leak area to the clearance between this ring plate and the pump shaft (approximately 1.3 square inches per pump). e All ECCS and the containment building spray system are available l and start if appropriate setpoints for automatic start signals are reached. Charging flow continues as normal. l e All auxiliary feedwater is available. e No operator action is taken to depressurize the steam generator secondary side since the sequence will look like an intermediate-sized LOCA inside containment. e No low pressure injection or recirculation cooling occurs due to failure of the RHR pumps. It is assumed that the water jet along the pump shaft will fail the pump motor even if the pump is not submerged. e Containment spray system operates as long as the spray pumps are not submerged.
" ~
I TABLE 3-8. CHECK VALVE LEAKAGE EVENT DATA BASE Sheet 1 of 2 NPE Plant " * " "9' Reference (date) (gpm) V11. A.126 Zion 2 A leak rate of ~0.25 gpm was detected from y -~0.25 (October 1975) the "A" accumulator check valve - wrong size gasket installed. V11.A.32 Turkey Point 4 One of the three check valves in the high-head y ~0.33 (May 1973) SI lines to the RCS cold legs developed 1/3 gpm leakage with 180 psi of water pressure applied. Two other check valves showed only slight leakage - failure of sof t seats. V11. A.175 San Onofre 1 A tilting disc check valve located in the LPI y<5 (May 1978) system as the first valve inside containment, failed to close with gravity - valve installed in a vertical rather than a horizontal pipeline. V11. A.114 Surry 1 Check Valves 1-51-128, 130 leaked causing boron y < 10 m (July 1976) dilution in the "B" accumulator. y < 10 I V11. A.182 Calvert Cliffs 2 The outlet check valves associated with the y < 10 (September 1978) safety injection tanks 218 and 22B 1eaked y < 10 reducing the boron concentration from 1,724 and 1,731 ppm to 1,652 and 1,594 ppm in 1-month period, respectively. V11. A.291 St.rry 2 Check valve associated with the SI accumulator y < 10 (January 1981) "C" leaked, resulting in accumulator baron dilution - cause unknown. V11.A.306 McGuire 1 Discharge check valves associated with the cold y < 10 ( April 1981) leg injaction accumulator A leaked - cause y < 10 unspecif t ed. V11.A.343 Point Beach Check valve 1-853C, serving as the first-off y < 10 (October 1981) check valve from the RCS for the low head SI, V11.A.63 Ginna Accumulator "A" check valve leaked leading to y < 20 (September 1974) boron dilution (from about 2,550 down to 1,617 ppm) - cause unknown. V11. A.85 Surry 1 Check valve associated with the 1C accumulator y < 20 ( August 1975) failed to seat, resulting in increase in accumulator level - cause unspecified. developed 6 gpm leakage. g V11. A.105 Robinson 2 "B" SI accumulator check valve developed y < 20 5 January 1976 leakage - cause unspecified. V. A.122 Zion 1 Discharge check valve on the accumulator 1D y < 20 E (June 1976) developed back leakage - cause unspecified. 3 V.A.407 McGuire 1 Cold leg injection accumulator check valve 20 < y < 50 (May 1983) leaked, resulting in low accumulator boron concentration - cause unspecified. I I 3-94 1251P112285
[ TABLE 3-8 (continued) { [ Sheet 2 of 2 l ~ NPE Plant Leak Rate Eveat DescriP tion Range L nererence (date) (gpm) V.A.452 St. Lucie 2 The SIT outlet check valve developed excessive 20 < y < 50 (December 1984) leakage - foreign material caused ball galling leading to joint binding. V.A.456 Calvert Cliffs 2 SIT check valve developed excessive leakage - 20 < y < 50 (January 1985) ethylene propylene 0-ring material degradation. Y.A.437 Farley 2 Loop 3 cold leg S! check valve developed 50 < y < 100 (September 1983) excessive leakage - incomplete contact between disc and seat. V.A.273 Davis Besse 1 Gross back leakage through core flood check 20 < y < 50 (October 1980) valve - cause unspecified. V11. A.384 Calvert Cliffs 1 JIT outlet check valve leaked at the rate of y -~200 (July 1982) 200 gpm ring deteriorated. E I l I L I l I 1 I l I t I I i g 3-95 ) 1251P112285 l
l I I I I I I TABLE 3-9. STATISTICAL DATA ON CHECK VALVE LEAKAGE EVENTS IN PWR, ECCS, AND RCS SYSTEMS Leak Rate Number of gf F en pr ce Frequency of I (gpm) Events Exceedance (per hour)
.5 3 2.94-8 2.06-7 10 7 6.86-8 1.77-7 20 5 4.90-8 1.08-7 50 4 3.92-8 5.90-8 100 1 9.80-9 1.96-8 200 1 9.80-9 9.80-9 NOTE: Exponential notation is indicated in abbreviated form; i.e., 2.94-8 = 2.94 x 10-8, I
I I I 3-96 1251P120385
M M M M M M M M M M M M M TABLE 3-10. OPERATOR ACTION SEQUENCES USED IN THE RHR OR V-SEQUENCE LOCA ANALYSIS Operator Action Sequence Sequence Time 5th 50th 95th Mean Interval
- Percentile Percentile Percentile.
01 - Operators fail to diagnose 1/2 to I hour. 6.4-3 1.1-3 3.7-3 1.7-2 the RHR system LOCA. 02 - Operators fail to isolate 1/2 to 1-1/2 hours. 9.2-3 2.0-3 6.0-3 2.4-2 the RHR System LOLA and throttle flow into the primary system. ?> 03 - Operators fail to provide 1/2 to 2 hours. 5.1-3 2.3-4 2.0-3 1.6-2 $3 makeup to the RWST. 04 - Operators fail to isolate -- 1.0 -- -- -- the RHR system LOCA and throttle changing flow into the primary system, given that the operators have failed to diagnose the event (operator action 01). OMeasured from time of initiating event. NOTE: Exponential notation is indicated in abbreviated form; i.e., 6.4-3 = 6.4 x 10-3, 1251P111985
I TABLE 3-11. POINT ESTIMATES FOR ENVIRONMENTAL FAILURES OF THE RHR PUMPS I Fault Tree Events
- Valve Failures SP LS OP EN SF E t**
Injection Valve 0.0 < A 10.09 5.8-3 0.0 1.5-2 .1 .5 .55 0.09 < A < 1.05 5.8-3 1.0 .2 .25 .75 .85
- 1. 05 < A 1 2. 6 5.8-3 1.0 4 .5 1.0 1.0 A > 2.6 5.8-3 1.0 1.0 .5 1.0 1.0 Suction Valve 0.0 < A 10.09 5.8-3 0.0 1.0 .1 .5 .56 0.09 < A < 1.05 5.8-3 1.0 1.0 .25 .75 1.0
- 1. 05 < A 1 2. 6 5.8-3 1.0 1.0 .5 1.0 1.0 A > 2.6 5.8-3 1.0 1.0 .5 1.0 1.0 I
*See Figure 3-9. ** Top Event RS in V-sequence event trees.
NOTES:
- 1. The total area of RHR pump seal leaks is in square inches.
- 2. An area of .09 square inches corresponds to 50-gpm RHR pump seal leak at each pump at RCS pressure.
- 3. An area of 1.05 square inches corresponds to 550 gpm (total for both pumps) at 425 psi.
- 4. An area of 2.6 square inches is the maximum seal leak area (both RHR pumps) determined by Westinghouse.
- 5. Exponential notation is indicated in abbreviated form; i.e., 5.8-3 = 5.8 x 10-3,
~
1251P111885
I l l TABLE 3-12. P0 INT ESTIMATES FOR ENVIRONMENTAL FAILURES OF THE CBS PUMPS
- Fault Tree Events
- Valve Failures SP LS OP EN SF E t**
Injection Valve 0.0 < A < 0.09 5.8-3 0.0 1.5-2 .1 0.0 .1 0.09 < A < 1.05 5.8-3 1.0 .25 .25 0.0 .44
- 1. 05 < A < 2. 6 5.8-3 1.0 .5 .5 0.0 .75 A > 2.6 5.8-3 1.0 1.0 .5 0.0 1.0 Suction Valve 0.0 < A < 0.09 5.8-3 0.0 1.0 .1 0.0 .11 0.09 < A < 1.05 5.8-3 1.0 1.0 .25 0.0 1.0
- 1. 0 5 < A < 2. 6 5.8-3 1.0 1.0 .5 0.0 1. 0 A > 2.6 5.8-3 1.0 1.0 .5 0.0 1.0 I *See Figure 3-9.
** Top Event CS in V-sequence event trees.
NOTES: I 1. 2. The total area of RHR pump seal leaks is in square inches. An area of .09 square inches corresponds to 50-gpm RHR pump seal leak at each pump at RCS pressure.
- 3. An area of 1.05 square inches corresponds to 550 gpm (total for both pumps) at 425 psi.
4 An area of 2.6 square inches is the maximum seal leak area (both RHR pumps) determined by Westinghouse. I 5. Exponential notation is indicated in abbreviated foria; 1.e., 5.8-3 = 5.8 x 10-3, I 3-99 1251P111885
I TABLE 3-13. POINT ESTIMATES FOR ENVIRONMENTAL FAILURES OF THE SAFETY INJECTION PUMPS I Fault Tree Events
- Valve Failures SP LS OP EN SF E t**
Injection Valve 0.0 < A < 0.09 0.09 < A < 1.05 5.8-3 5.8-3 0.0 1.0 1.5-2
.1
.1
.25 0.0 0.0
.1
.33 l
- 1. 0 5 < A < 2. 6 5.8-3 1.0 .25 .5 0.0 .64 A > 2.6 5.8-3 1.0 1.0 .5 0.0 1.0 Suction Valve 0.0 < A < 0.09 5.8-3 0.0 1.0 .1 0.0 .11 0.09 < A < 1.05 5.8-3 1.0 1.0 .25 0.0 1.0
- 1. 05 < A < 2. 6 5.8-3 1.0 1.0 .5 0.0 1.0 A > 2.6 5.8-3 1.0 1.0 .5 0.0 1.0
*See Figure 3-9. ** Top Event SS in V-sequence event trees.
NOTES:
- 1. The total area of RHR pump seal leaks is in square inches.
- 2. An area of .09 square inches corresponds to 50-gpm RHR pump seal leak at each pump at RCS pressure.
- 3. An area of 1.05 square inches corresponds to 550 gpm (total for both pumps) at 425 psi.
- 4. An area of 2.6 square inches is the maximum seal leak area (both RHR pumps) determined by Westinghouse.
- 5. Exponential notation is indicated in abbreviated form; i.e., 5.8-3 = 5.8 x 10-3, I
I 3-100 1251P111885
[ [ l E E , TABLE 3-14 V-SEQUENCE RESULTS - INITIALLY ASSIGNED PLANT DAMAGE STATES l [ F requency Plant Contribution From Total Damage State F requency { VI VS LOCA 4.1-6 3.0-6 7.1-6 E DLOC 4.0-7 0 4.0-7 { DILOC 3.3-9 2.6-7 2.6-7 8C 7.1-10 0 7.1-10 7D 5.2-9 0 5.2-9 7FPV 2.5-9 5.6-9 8.1-9 E 6.1-10 1FPV 2.7-8 2.7-8 { 1FV 2.7-9 1.9-9 4.6-9 Totals 4.5-6 3.3-6 7.8-6 NOTE: Exponential notation is indicated in j abbreviated form; i .e. , 4.1-6 = 4.1 x 10-6, 1 1251P111985 3-101
I I I I I TABLE 3-15. CUMULATIVE PROBABILITY OF CONTAINMENT FAILURE WITHIN t HOURS AFTER A LOSS OF ALL AC POWER l
-(N0 CONTAINMENT SPRAY AND RECIRCULATION)
[$c(t)] t (hours) 4c 18 2.00,x 10-4 24 3.00 x 10-3 48 1.80 x 10-1 72 5.00 x 10-1 96 7.60 x 10-1 120 9.30 x 10-1 144 9.86 x 10-1 l 168 9.98 x 10-I I I I I I I 1251P112285
l 1 l l I TABLE 3-16 CUMULATIVE POWER REC 0VERY FROM 0FFSITE POWER I FOR THE 345-KV SOURCE WITHIN t HOURS AFTER THE LOSS OF ALL AC POWER [4345(t)] I 5th 50th 95th I t(hours) Mean Percentile Perceqtile Percentile 18 .953 .905 .953 .999 24 .970 .940 .970 .999 l 48 72
.970
.970
.940
.940
.970
.970
.999
.999 96 .970 .940 .970 .999 120 .970 .940 .970 .999 144 .970 .940 .970 .999 I 168 .970 .940 .970 .999 I
I I I I I I I 1251P112285 3-103
j TABLE 3-17. CUMULATIVE REC 0VERY FREQUENCY FOR THE 34.5 KV SOURCE i l l 1 34.5(t) Time After Time After Initiating Event Recovery Starts $34.5(t) 5th 50th 95th t t-18 Mean Percentile Percentile Percentile 18 0 0.00 1.00 1.00 1.00 1.00 24 6 0.73 2.7-1 3.4-2 1.7-1 8.4-1 i' 48 30 0.95 5.0-2 6.2-3 3.1-2 1.6-1 5 72 54 0.983 1.7-2 2.2-3 1.1-2 5.4-2 96 78 0.994 6.0-3 7.5-4 3.7-3 1.9-2 120 102 0.998 2.1-3 2.6-4 1.3-3 6.5-3 144 126 0.999 1.0-3 1.2-4 6.2-4 3.1-3 168 150 0.999 1.0-3 1.2-4 6.2-4 3.1-3 NOTE: Exponential notatian is indicated in abbreviated form; i.e., 2.7-1 = 2.7 x 10-1, 1251P112285 W M M ~ _____
I J I I TABLE 3-18. CUMULATIVE PROBABILITY OF REC 0VERY OF CONTAINMENT SPRAY AND RECIRCULATION OF ADDITIONAL INDEPENDENT SOURCES WITHIN t HOURS AFTER A LOSS OF ALL AC POWER [&0ther(t)] 5th 50th 95th t(hours) 40ther Mean Percentile Percentile Percentile 18 0.0 -- -- -- 24 0.0 -- -- -- 48 0.0 -- -- -- 72 0.3 0.1 0.30 0.50 96 0.4 0.2 0.40 0.60 120 0.5 0.3 0.50 0.70 144 0.6 0.4 0.60 0.80 168 0.7 0.5 0.70 0.90 I I 1251P112285 3-105
,-r -.w- * +
I I I I TABLE 3-19. CONTAINMENT REC 0VERY ANALYSIS RESULTS 5th 50th 95th Parameter Mean Percentile Percentile Percentile Qu 4.6-3 1.6-4 1.5-3 1.9-2 A(S*CE) 1.2-6 5.4-8 5.4-7 4.7-6 A(S) 3.6-5 2.7-6 1.8-5 1.3-4 F(CElS) 7.3-2 2.2-3 2.7-2 3.1-1 NOTE: Exponential notation is indicated in abbreviated form; i.e., 4.6-3 = 4.6 x 10-3, I I I I 1251P112285
O I I . I ACC SI 21 l HIGH LOW PRESSURE PRESSURE l R A
< M SIV20 x
R H-V61 M RH V31 C m LPI SYSTEM T ACC TRAIN A a< z x x I RH V14 SIV5 RH 59 RH V15 ACC l v < E M I X M SIV50 RH V65 RH V30 S ACC Pi SYMEM l TRAIN B E L < Ml x M RH V26 l Si.V35 RH V63 RH 29 i INSIDE CONTAINMENT PIPE TUNNEL RHR VAULT FIGURE 3-1. COLD LEG INJECTION PATH ARRANGEMENT I I 3-107
I I I i n TO PRT i I RC V89 g R HOT LEG TO RHR PUMP A RH V87 RH V88 C T 1 o l
" I l
l v S i j M ""'I RC.V24 HOT LEG
> TO RHR PUMP RH V22 RH V23 I
l l HIGH > - LOW PRESSURE ' l
" PRESSURE l lNSIDE l CONTAINMENT FIGURE 3-2. RHR SUCTION PATH ARRANGEMENT I
3-108
E
-6 1 x 10 l 3 I i8iill 3 6 6 16IIil l 6 i i i iill l 6 6 66 I I I.
s s - N 3 - g g N - 2 - \
- N
\ N
\ \
[ "'i -
\N N
s.s \ - 6
~
g \ g Ng %
\ \ -
3 -
\ \ \ -
[ e 2 -
\
\
\
N N s
\
N
\-
S 1 x 10j -- g e s - k 8 - N ~ W 7 6
\ \ -
N -
\ \ -
g 5 -
\ N -
w 4 - k 3 - u 2 - N
\g -
$ \
9
- 1x10*3
\ _
7 LEGENO: \ g - 6 - 5 -
-- = = STATISTICAL BOUNDS g .
4 . AT 90% CONFIDENCE \ [ eEST riT s NN 3 - N\ \ \ 2
- - ASSUMED 95TH AND STH PERCENTILES -
~
' ' 'O iO N g {
\ 7 7
6 - 5 - 4 - 3 - ' 2 - 1 10I ' Il! ' ' ' I ' ' '! ' ' l 1 10 100 1,000 10,000 CHECK VALVE LEAK RATE (GPM) FIGURE 3-3, FREQUENCY OF CHECK VALVE LEAKAGE EVENTS b 3-109
SEA 9 ROOK EMERGENCY PLAN OPTIMIZATION - VI TREE trcEND: g GF CUARANTEED FAILURE D ~ $ b$ b$ y & $o3 3 lg 25 -8 88 88 0 g IE !E E l $ rW kl SG !s 35 59 E: ! AV 85 35 gi g E-is D
- in n5 ss sg sg g g5 ji jl d !! *
!! !! ! !$ $! Ib !! I! $I $5 8! Ek !! ! kE M Ut v0 Pi 9 Li L2 L1 01 02 CS RS S3 VC 03 SEO END STATE FREQ 1 LCCA 4.0813E-06 1 2 1FV 1.9591E-10 . 3 LOCA 4.1597E-09 I 4 DLOC 1.1969E-08 I 5 DLOC 1.3299E-09 F 6 DLOC 1.6175E-08 I 7 7,9649E-11 BC 0 DLOC 1.3299E-09 I 9 DLOC 1.4777E-10 F 10 DLOC 1.7972E-09 I 11 7D 8.8499E-12 12 DILOC 9.8446E-11 I 13 7FPV 4.8476E-13 14 DILOC 1.0938E-11 I 15 1FPV 5.3862E-14 16 DILOC 1.0938E-11 I 17 7FPV 5.3862E-14 18 DILOC 1.2154E-12 I 19 1FPV 5.9847E-15 F 20 DILOC 1.3369E-10 I 21 7FPV 6.5831E-13 22 DILCC 1.4855E-11 I 23 1FPV 7.3146E-14 24 D1 LOC 1.0992E-11 I 25 1FPV 1.2214E-12 26 DILOC 1.2214E-12 I 27 1FPV 1.3571E-13 p 28 DILOC 1.4855E-11 I 29 7FPV 7.3146E-14 30 IFPV 1.6586E-12 F , F- 31 7FPV 7.1127E-11 F- 32 IFPV 7.9030E-12 g F- 33 7FPV 7.9030E-12 F- 34 1FPV B.7811E-13 F l F- 35 7FPV 9.6592E-11 F- 36 1FPV 1.0732E-11 j F- 37 DILOC 7.9030E-12 F- 38 1FPV B.7811E-13 F- 39 D1 LOC 8.7011E-13 g 7- 40 1FPV 9.7568E-14 7 g F- 41 7FPV 1.0732E-11 F- 47 1FPV 1.1925E-12 FIGURE 3-4. SEABROOK EMERGENCY FLAN OPTIMIZATION - VI TREE (Sheet 1 of 2) 3-110
s SEABROOK EMERGENCY PLAN OPTIMIZATION - VI TREE I D E
*- g h
CI ki E h cI g E LEGEND: CF CUARANTEED FAILURE 5on yg g s, gg gg gg y y- g- ,. W n I
$^ -a g Ia o 2 g5 55 y5 g o E,rU 3
Q , bb e y
$ $ 5 !x n3 n ! hg y *8 55 83 s H 3 l.n ,. sg Ng NaI !< 55 %w 2,E S F5 6i !!
I 5: E> be
<g UyW I B b E5 a
- e. <
zw E
==
!. Io Lo
$8 to $s8 E5 o
5y# 5' s-ow b> "8 >IE5 m> F8 d E a EE om w ut vo m s u L2 ts ci or es as ss vc 03 I SEO END STATE FREO I 43 DLOC ?.2260E-09 I 44 DLOC 4.5442E-09 F 45 DLOC 7.7649E-08 I 46 BC 3.8235E-10 I I 47 48 DLOC DLOC 7.2490E-09 3.5704E-09 F 49 DLOC 6.1010E-08 I I 50 7D 3.0042E-10 F-F-F g 51 DILOC 1.3483E-09 52 7FPV 6.6391E-12 53 1FPV 1.5055E-10 W-F-F-F F- 54 7FPV 9.7414E-10 I W-F F-F g j F- 55 56 57 1FPV DLOC BC 1.0824E-10 5.0866E-C8 2.5047E-10 g 58 DLOC 1.5260E-07 I "-F-5 g 59 60 61 62 7D DILOC 7FPV 1FPV 7.5142E-10 1.6817E-09 8.2800E-12 1.8777E-10
"-#~# #- 7FPV 1.2150E-09 I l 63
'- 64 1FPV 1.3500E-10
#-F-# "- 65 7D 4.1728E-09
{ 66 7FPV 3.4489E-11 l 67 1FPV 3.8321E-12 I l 68 69 70 71 7FPV 1FPV 1FV 1FV 2.4796E-11 2.7551E-12 2.5109E-09 2.0088E-11 l I l FIGURE 3-4 (Sheet 2 of 2) 1 lI 3-111
SEABROOK EMERGENCY PLAN OPTlMlZATION - VS TREE LEGEND: CF OUARANTEED FAILURI g g D -lii $$ f i - g w
$g 59 iE %
5 8 @k e)
== f$ og N E
< W- W-56 56 W6
- a h
"5
$0 12h y 0b $y ! $ 3 h5 Ib bb ff y $m
!! !n !! d i N S!- !! !! !! ! !!
!h a
3d $n m a o sw Iz' n
==
Ni i, Lo ie sii n g so ma Dw
-v b> 8> kss e"s!
m> a bee I VS LR VO PI 9 L1 L2 L3 01 02 CS RS SS VC 03 SEO END STATE FREO 1 LOCA 2.9665E-06 I 2 1FV 1.4240E-10 . 3 LOCA 2.9163E-09 I 4 F DILOC 7.2520E-09 I 5 7FPV 3.5710E-11 6 DILOC 8.0577E-10 I 7 1FPV 3.9677E-12 8 DILOC 8.9631E-10 I 9 7FPV 4.4135E-12 10 DILOC 9.9590E-11 g I 11 IFPV 4.9039E-13 F 12 DILOC 1.0371E-08 I 13 7FPV 5.1066E-11 14 DILOC 1.1523E-09 I 15 1FPV 5.6740E-12 16 DILOC 9.OO72E-10 I 17 1FPV 1. OOC 8E-10 I 1e 19 DILOC 1FPV 1.1133E-10 1.2369E-11 l g F 20 DILOC 1.2817E-09 I 21 7FPV. 6.3115E-12 22 1FPV 1.4312E-10 F F- 23 7FPV 4.7680E-11 I F- 24 1FPV 5.2978E-12 F- 25 7FPV 5.8930E-12 I F- 26 1FPV 6.5478E-13 7 F- 27 7FPV 6.8184E-11 I F- 28 1FPV 7.5759E-12 F- 29 DILOC 5.8930E-12 I F- 30 SFPV 6.5478E-13 F- 31 DILOC 7.2835E-13 I F- 32 1FPV 8.0928E-14
----F F- 33 7FPV 8.4272E-12 I
F- 34 1FPV 9.3635E-13 F-F-F-F ; 35 DILOC 1.0387E-07 36 7FPV 5.1149E-10 37 1FPV 1.1598E-08 l F--F--F---F g F- 38 7FPV 6.8294E-10 l F- 39 1FPV 7.5883E-11
- F-F-F-F ;
40 DILOC 1.2956E-07 41 7FPV 6.3796E-10 42 1FPV 1.4466E-08 F-F-F-F g F- 43 7FPV 8.5182E-10 F- 44 1FPV 9.4646E-11 F-F-F-F F- 45 7FPV 2.6571E-09 g F- 46 1FPV 2.9523E-10 F-F-F-F W- 47 7FPV 1.7384E-11 l F- 48 1FPV 1.9316E-12 49 1FV 1.7603E-09 50 1FV 1.4083E-11 l FIGURE 3-5. SEABROOK EMERGENCY PLAN OPTIMIZATION - VS TREE 3-112
I .
'j - '
l l l I I l I 6 7 5
~
~
~
l j 4 - - 3 - - 2 - " l - ~ I 10'9 8 7 6 P' F -
~
~
l l 5 - " 4 - - 3 - - 2 - - I 10-2 g 8 Pp =PO + Il - 0) Pp ' __, 7 - WHERE Pp' CORRESPONDS TO A - I 6 5 4 LOGNORMAL DISTRIBUTION Pp' = .01 AT YlELD y = 8.1673 o =.1776
~
~
PF' = .99 AT ULTIMATE I 3 2 Pp (2,250) = 10-3 + 5 x 10-3 = 6 x 10-3 I 10-3 _P O- - - -- 9 - I 8 7 6 5 l 4 - - l 3 -
~
j MATERIAL YlELD MATERIAL ULTIMATE E (o = 35,000) (o = 80,000)
-4 1 l t l V l i l t l V 1 ! t 1,000 2,000 3,000 4,000 5,000 6,000 7,000 PR ESSUR E (PSI A)
I l FIGURE 3-6. PROBABILITY OF PIPE FAILURE iI 3-113
I I I I I 5,000 I g g , g , , , , , , g
- SCALED FROM 1,890 GPM AT 450 PSI i ** DETERMINED FROM GPM = 53.03 AOC geAP V P E' S W 4,000 -
WH E R E C = 0.6, @ - p = 60 lbm/ft3
- qgf.Y I
4*s#ps' 2 3,000 - - I b b .. 04" ggCS l E S0 B ko ZM - - I 1,000 - A"' 0 A 0= 0.09 SQUARE INCHES **
! l l i i i
! i e 200 400 600 800 1,000 1,200 1.400 1,600 1,800 2,000 2,200 2,400 PRESSURE DIFFERENCE (PSI)
I FIGURE 3-7. LEAK OR RELIEF VALVE FLOW RATE VERSUS PRESSURE I 3-114
m M M m M m W W W m m m m M M M DMW BAT A BATU "^ 24,000 G A L 24,000 GAL ANK 200,000 G AL DMW TRANSFER BORIC ACID PUMPS Mk TRANSFER PUMPS REACTOR l l l - M AK E UP I STORAGE < 75gom 75 gpm TANK DMV12 DMV11 170gpm 112,000 GAL , REACTOR M AKEUP l l 3 FCV 110A WATER PUMPS J FO ._ 170 gpm WATER l f TREATMENT [ g SYSTEM g BORIC ACID , y, BLENDER 150 gpm c1
-[ FCV.111 A FC & 480,000 GAL / DAY j
(333 opm) I !
*1P j CS V446 150gpm FIRE FIRE PROTECTION PROTECTION CS-V444 FCV.1118 FCV 110B RMW-V30 0 GAL GAL l v v v v RWST VCT VCT RCP INLET OUTLET SEAL COOtiNG pp, PUMPS DE -
}
DE - {} l uuST BE OPENED LOCALLY FIGURE 3-8. MAKEUP PATHS TO THE RWST, BAT, AND VCT
PUMPS Fall BECAUSE OF ENVIRONMENTAL CONDITIONS r% HUMID SPRAY FROM ! ENVIRONMENT RHR PUMP FLOODING F AILS PUMPS All PUMP F AILS PUMPS l EN sp LEAK OPERATOR SUFFICIENTLY F AILS TO LARGE TO TERMINATE FLOOO MOTORS B LOWDOWN ' OP e INSUFFICIENT SUMP PUMP LEAK > CAPACITY SOGPM LS SUMP PUMP SUMPPUMP NO.1 FAILS NO,2 FAILS
' PRIOR TO SUBMERGENCE SP SP FIGURE 3-9. FAULT TREE FOR ENVIRONMENTAL FAILURE OF RHR, CBS, OR SI PUMPS m m m m m M M M M M M M m m m m m
f L . E I I I I I I I I I l i I I I I I I I i i 180 - - 170 - - u 160 - 150 - - 140 - - E 130 - - i s g 120 - - E _ E 110 - r E
- a g 100 -
l
- y * -
l - E ~ 80 - ! - 70 - - 60 - - 50 - 40 - 30 -- - L 7 20 - -
~
3g I i ! I I I I I I I I I I I I I I I I 0.01 0.050,10.20.5 1 2 5 10 20 30 40 50 60 70 80 90 95 98 99 99.99 PROBA8ILITY OF EXCEEDANCE (%) 1 FIGURE 3-10. EXCEEDANCE PROBABILITY FOR TIME TO CONTAINMENT FAILURE FOR A STATION BLACK 0UT ACCIDENT AT SEABROOK STATION 3-117 o - - - - - - - - - --- -- y
8 EIE N' $
-e-
^g r.za a me J h e e
[g p $$; y tia er 4 .1 $ t- R !!' j lb
, -1 i 5 i l ! di$p If l!!
== _g - _ z_._;. 7 l
.W . ' wa m
3
.e i g
[ *h - a n 3< !.i
.a e E IIy 3, E
..-..__.l a
. i z ,, , wp 3, _ - --
8,! @
= e 3 g p
.y '
m hN __;h
*,g^[-
l r > e_ vg 5 M- !([d d _ , .; .. y g
,>i t - g-
- m
.~ t ; .;
gi i -- i p g_ (_ , y ', ,
~~
' (. ,
o
\ .
.o :
is- ; g,f,: g, ' ' .'s.g l. $syr g r e 4 .-- . .. M, _
,:- [_..!,:. .'.
.-m
- i-f y, z
- P, -
,i f J !c 1- , i
- g 1.<g a
as
.!. -. 3,s.;,..U
.c.? <.: . -
g,g .
'-c.. -
@ c
~
f
'e78 ! Ij I 9.. :'s . Is @2 < e P ', E!! d- ,,c , p sibfro 2- . 1e' !. b
^ i(+is .@'i " ->-6p I
~@ 4e .., o "i w
E 1 -
- i-+ ..
- p E
- .l t-tr--
~ ,<
.e- <
!!I t*
+- a . t r- ;- -- ,
. g
~#-*
} 'f-+ y :. 15 1! 75 d :
< E I
sg.a.5g.f s;4.':tA; r hu ,_' .I ' . . z E I a !< .- - i[
'-W!!is b
! ! ! i @ 9
-: ( n, di' -
Yy E m E s : - - ( s, b.
*- @ 5 l *-
t j !.x lf.-e_9 LL* 4;s,- , ' E v$1I 3 'a.4 , ,. 2
' u i -. e . , -.
!! E P -
g M. @r-+ IU I
"r
.g.- *- -e
!5@r'~ a-- '1 t-+ROis y :,,
~
5 i"; g%- e ji: 5 L .F@s [ . ]1 [E s . 1
-" U s :
'r r ' !
1:$ i-
'e (f_->-{4h-{
~
! ,'3 _.,I 10
; y m
E l
- b. !: !lH .. 4 - -
2,_g! E lill : re - B * }Sr-
- E a"Jh; 8
,s gi]!'
ir r o . ggja g) e [<
. I
!? *[j'E3 3 )*g s
-l* !!!N !!!! . !h .'l' -44 l' .
gi!! - ki @i:
- .Pg
-@ 4
;5 ,,5 @ @.@
. .g ;4ht-
'jn4
= 5~ @@ '
M,I, E: @- ,,- -
, g s
@,,- - g 4
'" 5 3-118
X
. , . . m n. + + .
r. 9 14 2 .' lW E, lg -
. . .h. i' - > I-'.
,$k Ih ,4 a
I3 @ E s , _,, vg 4, , e
.s., i, E'..
- n. .- .-
- M' 3.q 9 h .k:li@f. m,g .n ;
m se
% "E 0
si
; .SId .
-> e
.e
,i i/I - 9?;8 --
j.- ~i2
! "j& y.
i
~
P-d. ,, ,e -g X T j i If1
'd- k" 3: j fl! }I g p% -l.t k I e
. . i a
- : 4 -
e
. a s - . . . , . , . . <
i . . . , - { j .
\ tApp 9e ,
""jN s *" Q
. . ._ l --
w ma4%. wf Gy ersa . I arroT ' ens -
- f.
- O%
,c ;L ..
J, S. -
**5M *
/
+1 j i
w I l ~*ser-arr. 7* * "* ' son.-c~e--\ . .$ .= z [ 4- g Q o-
\
i , ;i * ; 3 a ,, 4 O I -j r , hM ; 3 ' ga,}/. . . . reu r v8 i r
- 1. W
. < t ,sa w s.g~ ~a
- P 4
.= :g 3
w m t .. s Q .4 I _ . .a
. 4
. -s . ; . ,
Q ,es - m.e e {- m i ' ', ,%,,, I E E ' i \ x b 'i ca s***.m ..s eD-*
/ -
,' N
,g Z
\
lQy_ 1-#= = s *= s*
- 4
*+c 1
s% o
- * * .L.. scuss '
a-k k 6 Mr. ..c.,- s . u . , 2 t # l a: g .$
}h )4) A^ ft f ,'
M I
$- ;; q y {!
. , ye .-s 7 i
\EE
-L =ser n v., .
Eg= f .I 4 ' .g (*, e = WW*h *#**-"'" M i =
- m. = rts 2.'s ?o . W .-
4-*j } ll? * - - - ars.uwswo '
- Q
!! --\y - -
{-- g .a.- . g .,3* u.. ._4 Jg % . > ,Q{. < }
- g i
' . h..{ Q.-
s - - z r,io e=
*tc tra,. -
i ,,L,,_
. u
_m_ . _ - .43 r[ ... g < s+ y* e
-e -== *** -a ,gy b, i
, J ) L' tM b.t
- H I ':; 5.' IC l 2g,
.D N@ 2E 5!E5!
- A.
> a:i z r ;ih,f 'h . W, ~ . -
- q. nQ5 p9 v&,,$...%. ,,,.5-~=>-M bh+N &
?r i t . ,u s i r .p ,,2, e-r n y pt
,,, .. o N M I 3 T ::i it g , *,**9"'
,&(
.I 2 7 yi 1
9 g
.I^*'*f
*/ ,
ss as W$ J' s
'"'.~
U y 6
;, *. 5 ga
& ~
/E ,,
g
& L' ) { :* /,f s/a-<>.M I 'W' g., .
i 5* F--
*: U R
# . w
, p* I I Asr(6tr f h, A
=
3 g,
- i. s&sg%
g- w
- z I
- c. , , , - ..
. . , r_ e . _,
o [ ~ ,c... . ,=c. .r. x r ? g . [ x
""' 0 :I a E,i, :- h
- g s
I o{ 4 > c e
- w. otsu h"'
n = 4 { t z
- g* , e*-egv" E @_.
h
, T_1"I ll *s' .$
y u..r. ew ( - w g
= se } ,,
} re,tc g3,gy %,
I m h y [q n e.': ? 2 + 2C es =a..e '.s-{ 1 y i 3
- - =, 3
.. t , g p. .
y lore / l M9iO s win / 5 s * .g {,; eee me. .-l "a K + . . . c r- -
- ' 2 1 '
f eC S l I : i
. ... _3 ~
i
~r
-t,h u,33y -Q
'If
?4*
t A W c' T.'.
'%'t O I
- e O b
I I 3-119
o
-~
l o I i . > I [5 s ,_ ^ ) !!$g j (jlNih N! j
,j}! l!
,/k!hr j
i j l j .:,.. , i[L#' ['j "
+ f ip
_fg j I; a j' i,' j
s \ [ h' k I
e p;;p,,- z/ g ph,1 ..,. v's.w- n. y - .. p1,,
-w.
W; I il
.- .fi,@ o-
=
' \. l m;h..i ~~' l1 l x - t - w E
.bm._.s%V d / '> j. ' '.
,.*I 5 5, 1lsiji odi ji /j/?,; (L ,3s - )-,) ' jjl 3
- i. ii e
e
.rI*,. p7,v ' '"
~$
c ,
.. o -j3 Q .s.a; ,sk .
r--
!.I U m$yj' 5 h 2Y ' y,\,
ll..] ,a q$
,g
.i!!
e l
\<
v ,' -.N y NtI h 4 >1 + K4 W,. t
- @j n i (A\ sI; n5 I il>I hil xj
, } {g@'
~ c ~. .
i
\
is ,
\ t 2 I ,
f r? ml _ t
'Q,~ .n( $,\g,io.y1._r..s t c g_ s. 4. i. - ,
m l_ i. . t' , i - . (\/ i - o L 4 j p = j j'$ g . '{.D.h y i gj ; e
$ il bu \ $L...'i lA TDi ? t +s @l T lhrj E I
.. . .'i;, .
t > '. ' s us
' \ %y-
.: e l
$lh msL_1 s -
i-
- k. A u.y /
"g t I.
lij i it I ! r
;. t , / 'l . . _ 5
[,6 Qi! p.I r l gj !:l; li !\ N >g 5 7 ' /. A'g3 .,,fy / gj ye
~ g m
, a lI 7 . G, D i 1\ g
.: .l " . ,
'p s p w -> p, .
x
!.Il / ,,/! . 8
_._ , . _s,' -
.v I
i in,, . i .!!: p!!i !. ej .;:..
; i!! ;;.
.jl..h. f7 gf y,I ho .
,h
, g 5
l ey~$.h W;K ~ ~. f ii;#
,.1.
- i t
'.: @ /.* y1.
y 1., [P
~l I m
gl s l p> ! l Q. 'lN, .
'j; 'Ie i l"h s2 /,' d -
.O N . // A s
1q: li
/ K ss ',!,j j
I w E f- ,,,,,;- 9 --- -. L f gg II i!g;i E 4 'I-,
{ ,l 7 ,
E o L.-3 i 1!! . . . , . ;, j j. .
~
l ' .5 !. l ] $: ?'Q ?
'l ljN i' j ( I Tsis 4=-:5 NA n! !
9! i
, B'Kb 'e M'(/
ih i i li '!!!!l;,ii!!i n ti li i !! . dr~./ ;! O t- ; ' s\: 4 no.j<s+....ltH.:
'- ttt '
E E
',.. a' '
f .i.4.' E . ..
~~*
.il !!r .) U .
x -
. :( ..-
- . :,; ' t' \,
s ,ss - p .', tfl.. f s. . 6 iI;g ,N ( s' t is u {1l2i 2 4 (: i ;M' r o' s ,' , L-ll'f
..S ' - ) E.
3-120
I YOUR a O TOR I SETIS N STOCK. I YOU'LL FINDIT LISTED BELOW. For cogeneration, standard continuous duty or standby application, you'E do better With a used generator trorn o'Brien. We offer reliable, top quality equiprnent from aR the major manufacturers I at a fraction of new cost. And we offer it from our huge stock inventory. Here's a sample-Steam Turthators 150tMW GL 550 PSL 7501,160 PSL Condensmg. Cordsermng Extrachon Sets extuust. 480V. Dou Act 300 250 PSI 40 exh 160V I , 6CD to 7501. with 140 PSI, and 30 PSI. extrache. 3/60/480V 4000KW GE.,250 PSI SM,20 PSI. exh 2400/4160V. 5000KW GE, ,250 PSI guestmghome. 850 PSI /90&F, I 3500KW CL 6(D /650T,55 P5t 3500KW Westmghouse. 400 PSI /75&F 150 & 50 P5L extract. 2400V. 5000KW GE,400 P51/650T,150 & g 400 PSL 75&F/ Others 10MW to 50 MW.400 PSL to 1450 PSL,some with extract Diesel Generator Sets I 500k 10 nat. m e 400 PSL/750T, Desel 50 AW Wesh 400 PSI /75&F, k jg .16V71T. GE,400 PSI SOT,175 & 7 7 M Y00V besel.16V149T. CL 507,50 PSL hDe277/m I I 10.000K 625 P51/780T, 4h . 1000KW CM Em 567. 720 W Dual Fuel, C.E,300 P51/650T,132KV 15.000KW GE.,650 P51/825T,12.500V. 25400KW GL 850 P51/825T,60 P5L Gas Turbine Generator Sets ex1ract.13.8KY (1975L 1250KW Ruston IA 1500,3/60/277/ I 30.000KW Mesunghome. 850 P51/900T, 13.800V. 480V. Dual fuel Recently Overhautes 2000KW Aikson 501KA. 3/60/4160V. Non-Condensing Sets 2880KW soiar centaur with Didd I 300KW Terry. 200 PSL D & 115 PSL exhaust,2400V 20 ext 400/ 700KW Worttungton,250 PSI D & i Fuel Unused 3065KW Aikson 501KB.3/60/ i 10. 3/60/144KV. 4 16.000KW CE Frame 5 with wtB. 3/60/ 10 PSL exhaust. 240CV4160V. 13 SKY 2 I 1000KW 40 PSL D & i 5-25 PSL 1500KW Elhott. 250 PSL 60N,40 PSI exhaust,2400V. 27 W 301. Dual. 31.000KW Mshnghouse W-251 with wtE. 3/60/138KV [HE QBRIEN Maculnew Ca I 214 Power Drive. Downingtown. Pennsylvania 19335 Phone (215) 269-6600 TELEX 835319 I I FIGURE 3-14. EXAMPLE FROM A QUICK REVIEW 0F EMERGENCY POWER SUPPLIERS Source: McGraw-Hill, Power, Vol .128, No.11, November 1984. 3-121
I
- 4. SOURCE TERMS AND CONTAINMENT ANALYSIS i
4.1 SOURCE TERM STATE-0F-THE-ART ASSESSMENT 4.
1.1 INTRODUCTION
This section discusses the current state of the art in source term research for the nuclear industry, both from the industry viewpoint and I from the regulatory viewpoint. The nuclear industry is performing its source term research and assessment via two programs; namely, the Industry Degraded Core Rulemaking program and the relevant research performed at the Electric Power Research Institute. The government I program is coordinated by the Accident Source Term Program Office (ASTP0) in the NRC Division of Research. The program consists of a multiple tier research program addressing many facets of the severe accident I phenomenology. Other research programs that contribute to enhancing the state of knowledge about source term issues are in progress abroad, particularly in Germany and, to a lesser extent, in other countries such I as England, France, Sweden, and Japan. However, except for a few specialized areas of source term research, the overall leadership in advancing the state of knowledge about accident source terms is clearly in the United States. Each of the major programs is briefly discussed in I tre following sections. 4.1.2 SSPSA SOURCE TERMS The SSPSA (Reference 4-1) source terms were analyzed and quantified in the 1983 time frame by using the MARCH /C0C0 CLASS 9/ CORRAL series of computer codes (References 4-1 (Appendix H), 4-2, and 4-3). This I analysis was completed before any of the new source term analysis methods became available. However, because shortcomings in the WASH-1400 (Reference 4-4) source term methodology had been identified, the uncertainties in the calculated source terms were explicitly quantified, using the information available by mid-1983. I In the SSPSA (Section 11.6 of Reference 4-1), a total of 13 point estimate source terms were quantified with the MARCH /C0C0 CLASS 9/ CORRAL codes, representing different containment failure modes, different reactor cavity conditions, and the availability of active containment I heat removal and fission product scrubbing systems. The containment failure models represented in the C0C0 CLASS 9 code, and particularly in the CORRAL code, were based on a comprehensive structural failure I analysis of the Seabrook Station containment design, including analyses for the membrane elements, for each penetration type and for the discontinuities in the containment structure. This analysis established I the Seabrook containment as the strongest containment for overpressure failure capacity of any nuclear power plant analyzed to date. The high pressure capacity and the dominant failure modes and locations were then explicitly modeled for the source term analysis. The risk contribution from each of the 13 point estimate release categories was examined, and four release categories were found to 4-1 1281P120685
I dominate the health effects risk. These represented the following containment failure modes:
- 1. Containment bypass conditions when the reactor coolant system leak 3 occurs outside the containment. E
- 2. Containment isolation failure in which the containment purge valve fails to close.
- 3. Early increases in the containment leakage rate, which is followed by late overpressurization failure.
- 4. Late overpressurization failure of the containment.
For each of these four release categories, uncertainty distributions were developed by first identifying known model weaknesses or omissions in the MARCH /C0C0 CLASS 9/ CORRAL codes, which were being questioned as being g unrealistic or overly conservative. The uncertainties introduced into g the calculated source terms due to each major modeling assumption about primary system retention, containment retention, and auxiliary building retention were estimated as release ratios for each major radionuclide category. The release ratio expresses the ratio of the expected release to the release calculated by CORRAL. It was estimated by comparing the effect of assumptions built into the analysis codes with observations 3 from experiments, from first principal analyses, or from data obtained in 5 the TMI-2 accident. Release ratios are informed judgments that are expressed as confidence distributions, stating the level of confidence that the release ratio is less than a certain value between zero and one. Release ratios were estimated separately for each accident sequence that contributed significantly to a given release category and for important variations within a given accident sequence. For example, differences in the primary system release between a hot leg LOCA and a cold leg LOCA were accounted for in this manner. Accident j sequence-specific release ratios were obtained for the primary system g i release, the containment release, and the auxiliary building release by 5 combining the uncertainty distributions for the individual contributing effects. l Using a similar procedure, uncertainty distributions were also developed for the time of release, for the release duration, and for the release warning time. The containment release ratios were correlated to the 3 l difference in time between the time of vessel melt-through and the time 5 of containment failure. Finally, overall uncertainty distributions for the release fractions were obtained by combining the release ratio distributions for the primary system with those for the containment or auxiliary building, as appropriate, then combining the distributions for I the different accident sequences that contribute to a release category. Dependencies between release timing and the release fractions were l recognized by correlating short release times, short release durations, short warning times, and high release fractions. The release category l uncertainties were finally modeled by defining four discrete release g i subcategories for each release category, each of which was associated W I with a level of confidence that the source terms for the represented accident sequences would be no higher than those defined by the release 4-2 l 1281P120685 3 l
b subcategory. The four subcategories were designated a, b, c, and d, with [ subcategory a representing the highest release derived directly from the CORRAL calculatien and subcategory d representing the lowest release.
.Probabilistically, these four subcategories are defined as follows:
e ce ( Subcategory Probability [ a .02 0.99 b .08 0.95 ' c .30 0.75 d .60 0.5 (median ) Therefore, subcategory d represents the most likely release, and there is a 99% confidence that the source term will not be more severe than the release defined by subcategory a. Thus, the results of'the source term [- uncertainty analysis indicated a confidence level of 99% that the source terms calculated for limiting accident sequences by the g MARCH /C0C0 CLASS 9/ CORRAL codes would not be exceeded. L The SSPSA source terms account for the specific Seabrook design features, and they quantitatively account for uncertainties in the WASH-1400 methodology for determining source terms. Therefore, the SSPSA source { terms contain a large amount of useful information for this project. 4.1.3 THE INDUSTRY DEGRADED CORE PROGRAM The IDCOR program was started in 1981 with the objective to develop a unified nuclear industry position with respect to accident source terms ( and consequences. The findings from this research program would be used in a rulemaking hearing in which the Nuclear Regulatory Commission considered to establish what changes, if any, were required to accommodate core damage accidents. Four reference plants were selected; { namely, (1) the Zion PWR with a large, dry, posttensioned concrete containment, (2) the Sequoyah plant with an ice condenser containment. (3) the Peach Bottom boiling water reactor with a Mark I containment, and [ (4) the Grand Gulf boiling water reactor with a Mark III containment. A program of 24 tasks was defined to cover the selection of risk dominant accident sequences, an assessment of all important accident and source ( term phenomena, operator actions during a degraded core accident, equipment survivability under accident conditions, and an integrated accident response and source term quantification. The technical objective of the program has essentially been completed and has resulted [- in the publication of a technical summary (Reference 4-5) report and a 4-3 [ 1281P120685
I series of task reports that separately address each task and subtask issue. Of the four power plants analyzed under the IDCOR program, the Zion plant most closely resembles the Seabrook Station. Both contain large Westinghouse PWR nuclear steam supply systems housed in large, dry concrete containments. For the Zion plant, three new source terms were calculated, representing four significant accident sequences and their corresponding containment failure modes (Table 4-1). These are:
- 1. A station blackout accident sequence with a concurrent loss of the turbine-driven auxiliary feedwater pump, leading to a late overpressurization containment failure mode. This initiating event and containment failure mode was analyzed with two separate variations, as shown by the first two cases in Table 4-1. The first sequence involves intact RCP seals; in the second sequence, a $
concurrent failure of the RCP seals is postulated. These two 5 sequences yielded identical results. However, the timing for these two accident sequences could be different if failure of the g turbine-driven auxiliary feedwater pump were not postulated as part g of the initiating event.
- 2. The third accident sequence was a station blackout sequence with simultaneous failure of the turbine-driven auxiliary feedwater pump, RCP seal LOCA, and a concurrent failure of the 10-inch diameter containment purge isolation valves to close on one open purge line.
- 3. The fourth accident sequence represented an interfacing systems loss of coolant accident that postulates the rapid failure of the E isolation valves between the high pressure section and the low g pressure section of the residual heat removal system.
The research and review findings from all the 10COR tasks were incorporated into a new transient accident analysis program called MAAP (Reference 4-6). The MAAP code in its current version models both the thermal hydraulic aspects and the radionuclide transport and release g aspects of severe accidents in light water reactors. Therefore, it is 3 possible to determine accident sequen::e-specific releases of I radionuclides from a single, integrated accident analysis calculation a with the-MAAP computer code. In all other previous and current analyses, I separate calculations are performed for the thermal hydraulic transient response and for the radionuclide transport and release portion of an i accident analysis. The MAAP code, therefore, not only incorporates the E ! new radionuclide deposition and transport phenomena, particularly those E i related to primary coolant system deposition and revolatilization, but it also allows the integrated treatment of feedback effects between the g radionuclide transport phenomena and the thermal hydraulic phenomena. 3 l The results from the integrated core accident analyses for the Zion Station are shown in Table 4-1, which is taken from the IDCOR Summary l Report (Reference 4-5). For the station blackout sequence and for the impaired containment sequence, the accident sequence timing (time of core l uncovery, time of vessel melt-through, time of containment failure) is E i not significantly different from results published in the SSPSA. 5
- However, the fraction of radionuclides released, as determined by the i
l I l 4-4 1281P120685 l L
l l l IDCOR program, are significantly lower than those previously calculated with such analysis codes as MARCH and CORRAL. The reduction in release fractions is principally due to the changes in the iodine chemistry and due to the retention of radionuclides in the primary system. An improved I aerosol behavior model in the containment has also contributed to reduced I source terms. The IDCOR results for the interfacing systems LOCA are substantially different from any previously published results, both with j respect to timing and with respect to radionuclide releases. previous risk analyses, the interfacing systems LOCA was assumed to cause In all a' double-ended rupture of the low pressure piping system as a result of a I shock wave traveling from the high pressure side to the low pressure side. The IDCOR program concluded that the low pressure piping system would not structurally fail, even assuming a shock wave develops from an instantaneous rupture of the RHR system isolation valves. Furthermore, the static pressure capacity of the low pressure RHR piping was found to l have a large margin beyond the design pressure and would not be expected to fail even if pressurized to the full primary system pressure. l As a result, the IDCOR program concluded that the postulated failure of all isolation valves either on the RHR suction side or on the RHR I discharge side would cause a failure of the primary system boundary at the RHR pump seal. The corresponding leak area was bounded at a maximum of 0.1 square feet. With this failure mode, the accident sequence timing is stretched out considerably and results in a substantial amount of time for potential operator recovery actions. The duration over which l radionuclides are released also increases significantly as a result of the initial deposition of most radionuclides on the reactor coolant surfaces. Following vessel melt-through, the radionuclides retained in l the primary coolant system will eventually heat the surfaces on which i they have been deposited to sufficiently high temperatures that revolatilization of the radicnuclides occurs. The containment is isolated and the uncooled debris in the reactor cavity penetrates into concrete and thereby releases noncondensable gases. The sweeping action from these gases through the primary coolant system and through the RHR pump seal leak area acts to eventually sweep the material first deposited l on the reactor coolant system surfaces out into the auxiliary building. The release fractions to the environment are nevertheless very small. The radionuclides released from the primary coolant systems diffuse very I slowly through the large volume auxiliary building at Zion. The large I surface areas and the steel heat sinks in the auxiliary building act as very effective deposition areas. l Overall, the IDCOR program concluded that the releases of radionuclides from very severe reactor accidents are much smaller than previously anticipated. The releases also occur much more slowly than previously anticipated. These source terms represent a substantial advance in our understanding of accident behavior and about the threat to the environment from reactor accidents. Uncertainties in these release fractions and in the timing of releases have not been quantified by the IDCOR program. Such a quantification of uncertainties would appear to be an essential element in substantiating a generic conclusion that radionuclide releases from reactor accidents are small and occur slowly. All previous calculations of radionuclide releases from reactor accidents have shown much larger releases; however, these releases have always been suspected to be upper bound releases and not realistic best estimates. l 1281P120685
I 4.1.4 THE NRC SOURCE TERM PROGRAM The behavior of radionuclides released from the core during the Three l Mile Island accident indicated that the radionuclide release and E transport model used in the past to estimate reactor accident releases 5 and consequences was overpredicting radionuclide releases. Initially, this was believed to be due to the significant differences in the g chemistry of iodine during reactor accidents. Analytical models treated 3 a large fraction of the iodine released from the core as a gas. However, evidence from the TMI accident supported a theory that f odine would be forming cesium iodide; therefore, the transport of iodine would follow that of cesium; namely, as a particulate. This led to the initiation of a NRC-sponsored source term research program that developed, as its major products, several reports among which are NUREG-0772 (Reference 4-7) and, E BMI-2104 (Reference 4-8). A further milestone report, NUREG-0956 5 (Reference 4-9), has just been released in draft form for comments. Until comments are resolved, this report cannot yet be considered to g represent the current NRC position on accident source terms. None of the g other NRC-sponsored reports published to date has actually published new source terms. The NRC research to date has led to improvements in a series of first principal computer codes for the various phases of a reactor accident. These computer codes, developed at Battelle Columbus Laboratory, have become known as the BMI-2104 series of codes. These computer codes are used in a chained calculation that uses the results g from one computer code to develop the input for the next computer code to 5 determine the radionuclide behavior in each phase of selected accident sequences. The BMI-2104 series of reports for four reference plants documents these results. Due to the chained nature of these calculations, no complete release categories have been published to date from the NRC-sponsored research program. Although it might be possible to develop a new set of release categories from tne existing documentation, this would require interpretation of results beyond what is supported by the documentation. The current plan for publishing the NRC-sponsored results of the source term research program is to issue g NUREG-0956, after review comments are resolved, as an interim document 3 that will contain updated accident source terms and a risk perspective for the Surry plant only. These new source terms only address either g early containment failure modes or basemat melt-through failure modes. 5 Therefore, they do not address the two most likely accident source terms identified by the SSPSA; namely, those for our intact containment and those for a late overpressure failure mode. The longer term goal for the NRC research program is to publish a complete update of accident source terms and risk profiles for all five g reference plants (Surry, Zion, Sequoyah, Peach Bottom, and Grand Gulf) in 3 NUREG-1150. The schedule for this report is mid-1986, and no other source term information is scheduled to be published in the interim. Because NUREG-0956 was not available for this study, no direct use of results from the NRC source term research program have been used in this study. A comparison of the source terms used in this study with the 3 corresponding source terms published for Surry in NUREG-0956 and with 5 those calculated by the IDCOR program will be discussed in Section 4.5. I 4-6 1281P120685
L [ 4.1.5 OTHER SOURCE TERM RESEARCH PROGRAMS Some limited source term research is sponsored by virtually every organization funding reactor safety research. These programs are not at a point at which they are projecting new complete source terms for consequence analyses. However, important contributions to the state of knowledge of radionuclide behavior is coming from the EPRI-sponsored ~ research program and from the German-sponsored research program. The German source term research program is generally regarded as a valuable contribution in key areas, such as the SASHA experimental program and the NAVA computer code (Reference 4-10). However, the German source term F research program is focused specifically on the German containment design L conditions, which are somewhat different from the containment configuration of U.S. nuclear power plants. More recently, the Beta facility has, and continues to provide, valuable German research data for the core concrete interaction problem. -These experiments are of particular interest because they are of a similar magnitude and approach as the Sandia core concrete interaction experiments. The results from the German experiments indicate significantly smaller releases of the lanthanide radionuclide group r compared to the releases indicated by the SANDIA experiments. The L conditions that govern the release of lanthanide fission products during core concrete interaction are not yet fully understood. Several possible reasons for the differences in the experiments have been identified, I including the quantity of unreacted zirconium in the debris pool, the pool temperature, and the concrete composition. However, the root cause I l for the differences in the results have not been isolated and all of the identified potential causes may be interrelated. The exothermic energy from the oxidation of unreacted zirconium will increase the pool temperature. The concrete aggregate principally used in German reactor designs is a basaltic material, while, in U.S. nuclear power plants, the l most commonly found aggregate is a limestone material. Limestone is mcstly made up of calcium carbonate, which decomposes into calcium oxide and carbon dioxide at temperatures below the melting temperature of the l concrete. Basaltic aggregate, on the other hand, does not contain any appreciable quantities of materials that decompose and yield noncondensable gas products. The total gas flow rate from the concrete to the debris pool includes both the steam released from the cement and the carbon dioxide released from the calcium carbonate. Therefore, the gas flow rate for basaltic concrete can be substantially lower than for the limestone concrete. For this reason, the driving force for stripping l fission products out of the molten debris pool during concrete attack is I smaller for basaltic concrete compared to limestone concrete. Furthermore, differences in the melting behavior of basaltic and limestone concrete can alto yield differences in the debris pool temperature during concrete penetration. Isolation of the root causes that are responsible for the potentially enhanced release of lanthanide radionuclides from the debris pool are important. The American Physical Society peer review of the source term state of knowledge identified this particular phenomena as a key uncertainty. It was the single predominant reason for the American 4-7 1281P120685
I Physical Society's conclusion that wholesale reduction of source terms may be premature. For the Seabrook Station, resolution of this phenomena is of specific significance because the Seabrook Station is one of the few U.S. nuclear power plants that employs a basaltic aggregate in the E concrete mix; therefore, the German research results may be more 5 appropriate for interpretation of concrete penetration phenomena at Seabrook Station than the U.S. experiment data. The American Physical 3 Society peer review report (Reference 4-11) explicitly identified the g difference in concrete composition as one of the suspect causes for the differences in research results. 4.1.6 EVIDENCE AND CONCLUSIONS A large source term research effort has been under way for several E years. Most countries with nuclear power programs are either sponsoring 5 source term research or they are participating in multinational research programs. To date, only results from the IDCOR program have been g published in the form of complete new' source terms (Reference 4-5). The g IDCOR source terms are substantially lower than the source terms from the WASH-1400-based methodology. The only other fully integrated source term analysis program is sponsored by the NRC. Intermediate results published by the NRC and by NRC contractors, such as the Reference 4-9 series of documents, also indicates a trend toward a substantial reduction in source terms. This review of the current state of knowledge about radionuclide source terms leads to the conclusion that there are three major sources of source term information available now; namely:
- 1. The IDCOR documentation.
- 2. The NRC interim documentation.
- 3. The SSPSA documentation.
These were used in developing source terms for this study. 4.2 ACCIDENT PHENOMENA AND SOURCE TERM CONSIDERATIONS This section will discuss accident phenomena and source term issues that are of importance to the assessment of accident consequences. This will E be accomplished by summarizing the state of knowledge that formed the E basis for the SSPSA assessment, by addressing major advances in the assessment of accident phenomena since the SSPSA study, and by 3 summarizing open issues in the understanding of accident phenomena, which g were identified in the NRC-IDCOR review meetings. 4.2.1 MODELING 0F ACCIDENT PHENOMENA IF THE SSPSA The SSPSA core and containment analysis vas performed in 1983 by the same organizations that performed the Zion and Indian Point PRAs. The 3 modeling of accident progression and sour:e term phenomena adopted the 3 methodology developed for the Zion and Indian Point studies. The Il 4-8 1281P120685
l l modeling of debris penetration into concrete used the CORCON code with a 1 water cover and debris quench model . An analysis by the Pittsburgh l testing laboratory of the concrete aggregate used in the containment at Seabrook indicated that a basaltic aggregate was used. It contains no i i calcium carbonate or other components that would decompose at elevated temperatures and liberate noncondensable gases as decomposition l l products. The analysis of debris concrete interactions accounted for the ! Seabrook-specific concrete composition. l A Seabrook-specific conteinment pressure capacity and failure mode analysis was performed. This analysis was used to define I l Seabrook-specific containment leakage and failure models for the radionuclide release analysis. The containment capacity and failure models used in this study will be described in more detail below. 4.2.2 ADVANCES IN MODELING ACCIDENT PHENOMENA AND SOURCE TERMS The chemical form of radionuclides liberated from the reactor core during I the TMI-2 accident was at substantial variance with the accident models used up to that time to estimate radionuclide source terms. The ensuing l research in radionuclide transport phenomena has led to major I j improvements in predicting accident source terms. These affect all phases of analysis starting with the core heatup to the release of radionuclides to the environment. The most outstanding advances have I been achieved in the retention of radionuclides in the primary system and in the recognition that containments are very strong structures. They l are not as likely to fail early in an accident sequence as indicated by I l the WASH-1400 analyses. Maintaining containment integrity for a substantial time period after the core debris is released to the containment has been identified as the single most important aspect in the progression of an accident. This has the potential for significantly I reducing accident source terms. In a series of technical exchange meetings between the industry-sponsored IDCOR program and the NRC, many areas of agreement have been established. As a result of these technical exchange meetings, a total l of 18 issues have been identified that require further efforts for resolution. These 18 isues will be addressed in the next section. l 4.2.3 SEVERE ACCIDENT TECHNICAL ISSUES Following completion of the initial IDCOR research program, a series of technical review meetings was held with the NRC staff. These resulted in identifying the 18 technical issues shown in Table 4-2 that required further resolution. The IDCOR/85 program is focused on fulfilling the IDCOR commitments to resolve these issues. In subsequent meetings with l the NRC staff, a path for resolving each issue was agreed upon. The 18 issues principally include key physical processes that govern the release, transport, and deposition of the fission products, as well as those physical processes that could potentially threaten containment integrity. The IDCOR/85 Program developed an integrated response to these issues. Agreed-upon paths to resolution, IDCOR/85 actions taken, and the results of these actions are documented in Table 4-2a taken 4-9 1281P121685
I from the the Issue Resolution report (Reference 4-12). Issues of particular interest for Seabrook are discussed below. Although the 10COR responses have not been reviewed by the NRC and their technical consultants, it is very encouraging that no major changes in the original IDCOR conclusions have resulted from these 18 issues. Of the 18 issues listed in Table 4-2, all except Issue 13 apply in some way to large, dry PWRs such as the Seabrook design. Issues 1 to 6 g address questions about modeling physical phenomena in the primary system 3 up to the time of vessel failure. Issues 7 to 9 aadress physical phenomena that can affect early containment failure. Issues 10, 11, 12, and 17 address modeling the physical phenomena after vessel failure. Issue 14 addresses a modeling assumption about the completeness of emergency response evacuation, while Issues 15, 16, and 18 address the performance of structures and essential equipment. For the PWR system with a very strong, large, dry containment, which is the case for Seabrook, many of these issues have a negligible influence 3 on fission product releases to the environment. In particular, these g include the rate of fission product release from the fuel prior to vessel failure, natural circulation in the reactor vessel, modeling of in-vessel hydrogen generation, core slump, core collapse, reactor vessel failure, secondary containment performance, and essential equipment performance. Elements that should be addressed for the Seabrook study relate to direct heating of containment by ejected core material, the aerosol deposition 3 mechanisms within the primary system and containment, ex-vessel heat 5 transfer models during core-concrete attack, ex-vessel fission product release models, revaporization of fission products in the reactor vessel g upper plenum, hydrogen ignition and burning, and containment g performance. In addition, plant-specific issues should be addressed for Seabrook Station. Issues relating to alpha mode containment failure by in-vessel steam explosions and modeling of emergency response have been resolved, as has Issue 3, which is related to the release models for , control rod materials. Four of the 18 issues are plant-specific. These are as follows: e Issue 8. Direct heating of the containment atmosphere by ejected core debris. e Issue 15. Containment performance. e Issue 16. Secondary containment performance. e Issue 18. Essential equipment performance. These four plant-specific issues will be discussed in the next section in the context of the Seabrook design. The remaining 13 issues are generic. The resolution of these issues is documented in Reference 4-12. The important generic issues are discussed below. Aerosol deposition within the primary system and the containment was E initially treated within the integrated G P system response through an 5 I 4-10 1281P120685 E
I experimentally based correlation. As part of the IDCOR/85 effort, extensive numerical calculations were performed with a sectionalized aerosol code to determine if (1) fundamental correlations exist for I aerosol agglomeration and deposition and (2) the nature of such correlations. The numerical calculations demonstrated that basic correlations exist and can be characterized through two asymptotes. The I first represents a steady-state aerosol in which a fine particulate source is continually available and the airborne aerosol concentration is constant. A second correlation characterizes the decaying aerosol in which the airborne concentration decreases exponentially in time. These I correlations have been compared extensively with experimental data and demonstrated to be in excellent agreement with a broad spectrum of experimental studies. In addition, this approach allows the particle size distribution to be calculated at each time interval so that the integrated decontamination factor of overlying water pools can be evaluated mechanistically. However, for the very stong Seabrook I containment, all aerosol models predict effective deposition within the primary system and the containment. As a result, agreement on the details of aerosol modelin.g is not fundamental to the conclusions of the Seabrook Emergency Planning Study. Considerations of the nonvolatile fission products released during core-concrete attack requires an assessment of both the concrete attack I models and the modeling for stripping of fission products by gases bubbled through molten core debris. The IDCOR model for core-concrete attack has been benchmarked against available experiments, as well as I# against sample calculations from the German code WECHSL. IDCOR agreed to continue this benchmarking process as additional results became available and encouraged the NRC staff and consultants to also benchmark their own models. The MAAP model (DECOMP) assumes that the molten material is horlogeneously mixed and, as a result, predicts that the debris teriperatures decrease rapidly once core-concrete attack is initiated. Large-scale experiments recently performed in Germany show that the materials are homogeneously dispersed'and that the core debris temperatures decrease very rapidly after the molten debris is discharged into the concrete crucible and thermal attack is initiated. In addition, y the integrated analyses performed in the IDCOR Program demonstrated that f about half of the core material would be involved in direct core-concrete attack immediately following vessel failure (as opposed to 100% assumed for the NCR analyses) and that substantial water could be available within the containment and, also, the primary syste,m that could quench the debris following reactor vessel failure (such processes are not considered in the core-concrete thermal attack modeling in the NCR I analyses). Due to a smaller amount of core debris and quenching once the debris is discharged into the containment, the overall process leading to thermal attack of the concrete is one of heating the core debris to I temperatures sufficient to initiate the attack. Consequently, there is no extended period of high debris temperatures (> 2,000 K) during which significant fractions of nonvolatile fission products could be scrubbed from the core material. As a part of the issue related to nonvolatile fission product releases, IDCOR committed itself to developing a basic approach for modeling the I 1281P120685
I chemistry of fission products in a molten debris pool and the stripping potential for these products, as noncondensable gases are bubbled through a molten configuration. In this assessment, IDCOR has found that the chemical forms of the pure metals, the oxides, hydroxides, and double E hydroxides are important in determining this chemical balance. This is 5 in agreement with the NRC models. In addition, IDCOR has also determined that chemical forms, such as silicates and zirconates, are very 3 influential in determining the dominant chemical form of strontium, g barium, and lanthanum in the core debris. Currently, these forms are not included in the NRC models. As a result of including such additional compounds, the releases of strontium, barium, and lanthanum are small fractions of the core inventory (less than 1%) and would only be liberated after extended time intervals. Most of the fission products released from the debris would be deposited in the containment. E Consequently, the release of nonvolatile fission products from the core 5 debris during core-concrete thermal attack does not appear to have a significant influence on the Seabrook Emergency Planning Study. Long-term revaporization of fission products in the reactor vessel upper plenum could potentially lead to an increase in the environmental releases. As part of the IDC0R/85 Program, information was obtained from several operating plants about the primary system heat losses during normal operation. The information obtained demonstrated that such heat losses are dominated by not-through-insulation losses that result at 3 joints in the insulation, penetration of piping through the insulation, 5 and seismic supports. In general, this heat loss can be 1 to 2 MW or greater. However, at the level of 1 MW for normal operation, the primary g system heat losses are sufficiently great that the potential for g long-term revaporization within the primary system for a PWR with a large, dry containment is negligible. Therefore, this issue does not influence the Seabrook Emergency Planning Study. The issue of hydrogen ignition and burning within the containment relates to the potential for hydrogen recombination during core-concrete attack. 3 Specifically, this issue addresses hydrogen produced as a result of the 3 concrete thermal attack and its recombination within the reactor cavity with oxygen that has been circulated into the cavity by the natural circulation currents in the containment. For the Seabrook reactor cavity configuration, natural circulation would not be in question and the IDCOR/85 analyses show that the temperatures within the reactor cavity during core-concrete attack are more than sufficient to ensure that 3 recombination would occur. As a result, the process of core-concrete 5 attack does not lead to hydrogen accumulation within the containment, but rather causes oxygen consumption, so the potential for global burning a within the containment decreases following reactor vessel failure. This g issue should not influence the Seabrook Emergency Planning Study. The last issue relevant to Seabrook relates to the containment performance, which includes the overall containment thermal-hydraulic response and the mode of containment failure. Since Seabrook has a very strong containment, the long-term response is similar to that obtained g for the IDCOR PWR large, dry containment reference plant (Zion) except 5 that the time to containment failure is considerably longer (due to aoth I 4-12 1281P120685
the stonger containment and the basaltic concrete used in Seabrook). Consequently, more time would be available for fission product deposition. Also, in the Seabrook PRA, the failure modes for the containment were analyzed extensively, with the conclusion that local penetration failures would lead to a leak-before-break condition as opposed to catastrophic failure of the containment. The combination of I these two elements leads to environmental releases that are dominated by noble gases and are released days after core damage. As a result of the above considerations, the IDCOR/85 Issue Resolution I efforts demonstrate that fission products in vapor and aerosol form would be effectively deposited within the containment except for the noble gases. Other elements related to hydrogen accumulation and large-scale I burns sufficient to threaten containment integrity, as well as direct containment heating, are not applicable for Seabrook because of both generic issues (hydrogen recombination) and Seabrook's specific geometry (direct containment heating). Consequently, no physical process or I combination of processes have been identified that could lead to early containment failure and direct release of fission products. If the accident is assumed to progress for days without corrective action, the containment would eventually fail, but the releases would be limited to noble gases. The APS review was limited to the review of NCR-sponsored work, and it was published after the 18 issues were identified. Therefore, these 18 issues are not explicitly aimed at the APS review comments. 4.2.4 ISSUE 8 - DIRECT HEATING OF THE CONTAINMENT ATMOSPHERE BY DEBRIS For postulated severe accidents that result in reactor vessel failure, core debris would be discharged from the reactor vessel into the reactor cavity. Such failures would likely involve a few tens of percent of the core inventory, and the debris may be liquefied or molten at the time of I discharge. In addition, for sequences in which the primary system pressure is elevated at vessel failure, the blowdown of primary system gases into the containment could provide a significant driving force for displacing core debris from the reactor cavity region. Thus, an issue to be addressed in the containment response is the potential for directly heating the containment atmosphere by the core debris as it is displaced from the reactor cavity. Direct heating of the containment atmosphere is of interest since the direct exchange of energy from the core debris to the containment I atmosphere could result in an increase of the atmosphere temperature. A substantial impact would result if 30% or more of the core inventory is i nvol ved. Therefore, the major focus for addressing this phenomena is one of discussing the potential for directly transferring tens of percent of the core into the containment atmosphere and distributing this material as a fine particulate, so it could exchange heat within a short interval of time. For this to occur, the debris must be: (1) displaced from the reactor cavity region, (2) particulated into fine aerosol, and (3) distributed through the containment volume. Each of these is a necessary element for direct heating and will be discussed individually I 4-13 1281P120685
below. The detailed methodology, analysis, and documentation developed to address this issue as part of the resolution phase for the IDCOR/85 program (Reference 4-12) is used as the basis for this evaluation. Before discussing how the individual phenomena relate to the Seabrook E containment configuration, a brief background of pertinent experimental 5 information will be presented. 4.2.4.1 Experiments Conducted to Date Debris dispersal experiments in reactor-like configurations have been carried out at Sandia National Laboratories (Reference 4-13) by using a 1/10 linear scale mockup of the Zion reactor cavity, which does not represent the remainder of the containment geometry or containment internal structure. In these experiments, the core debris is simulated 3 through the use of an iron thermite mixture that reacts in a short time 5 to produce temperatures approaching 3,0000K. This is sufficient to both melt the constituents and achieve a substantial energy level above a the melting point (superheat). In these experiments, debris dispersal g has been observed in addition to a fine aerosol mist. The iron thermite temperature is sufficiently high that the iron vapor pressure is close to 1 atmosphere and the aluminum oxide vapor pressure is about U.1 atmosphere. Therefore, one would anticipate a significant amount of aerosol formation when these materials are discharged into a gaseous atmosphere, which has a negligible vapor pressure from both g constituents. As stated in Reference 4-12, the debris temperature 3 anticipated for a severe core damage event would not yield a significant vapor pressure for the uranium dioxide, zirconium, or zirconium dioxide, g which make up the vast majority of the core inventory. Consequently, g while some materials, such as small amounts of stainless steel or control rod constituents, may provide for some aerosol formatioq, there would be no potential for a significant fraction of the core to attain an aerosol E state by vaporization and subsequent condensation. Therefore, the core E behavior following release of the materials from the reactor vessel would be considerably different than that observed in the Sandia tests. Other experiments have been carried out at Argonne National Laboratory (Reference 4-14). These are smaller in scale, but include a more complete-representation of the containment compartments and also use a thermite that is more prototypic of the reactor materials. In these tests, little aerosol formation was observed, which is in agreement with the above discussion concerning vapor pressures at the time of debris discharge from the primary system. In addition, these tests demonstrated that the structures outside of the reactor cavity / instrument tunnel region substantially influence (inhibit) the progression of debris into the containment atmosphere. Specifically, these experiments demonstrated that the overhanging structure at the exit of the instrument tunnel would catch debris and distribute it on the containment floor in the immediate vicinity of the tunnel. This was also verified by using a small transparent facility (Reference 4-15) and simulant fluids of nitrogen, water, woods, and metal to represent the various material densities of interest for this phenomenon. The simulant fluid experiments also 3 demonstrated accumulation of molten material on overhanging structures in 5 the near vicinity of the instrument tunnel . Such structures were not represented in the Sandia experiments. I 4-14 1281P120685
4 As a result of the Sandia and Argonne tests, it can be concluded that debris could disperse from the reactor cavity / instrument tunnel region l into the lower compartment of the containment. As the debris enters the I lower compartment, overhanging structures in the near vicinity of the instrument tunnel have a substantial effect on removing the dense debris from the gas-flow stream and distributing it on the containment floor. Analyses of the containment should include overhanging structures in the near vicinity and the potential for establishing a complete flow path from the lower containment compartment to the remainder of the containment atmosphere. 4.2.4.2 Debris Dispersal Characteristics for the Seabrook Configuration I The discussions above already addressed the potential for fine aerosol formation. Given the temperatures at which the zircaloy cladding can liquefy the fuel, no significant vapor pressures would be anticipated. Hence, the only mechanism for fine particulation would be hydrodynamic I fragmentation. As reviewed in the IDCOR report on Technical Support for Issue Resolution, the Seabrook reactor cavity is much like that of the Indian Point reactors. As shown in Figure 4-1, the geometry provides for an entry of the in-core instrument tubes into the reactor cavity and for a I separate manway for personnel access. During the blowdown of the primary system following reactor vessel failure at an elevated pressure, both of these would act as gaseous flow paths from the reactor cavity into the lower containment compartment. However, given the large density I difference between core debris and the flowing gases, the debris would be anticipated to follow along the outermost path and be dispersed through the entry port for the in-core instrument tubes. As a result, the key I consideration is the magnitude of overhanging and adjacent structures in this region. A review of the Seabrook drawings demonstrates that the support structure for the seal table and the seal table itself are in g close proximity to this flow path where substantial core material could 3 be stripped out of the gaseous flow stream and prevented from being dispersed directly into the containment atmosphere. Thus, the tunnel exit is essentially enclosed in a room, and there is no viable mechanism for providing sustained entrainment of debris as fine particulate, so tens of tons could be swept into the containment volume. In fact, this configuration provides for very effective separation of the core debris from the high velocity gases. 4.2.4.3 Material Available for Direct Containment Atmosphere Heating In the IDCOR report on Technical Support for Issue Resolution (Reference 4-12), criteria were developed to estimate the amount of material that could be directly dispersed into the containment atmosphere. Since little aerosol would be provided under reactor accident conditions, hydrodynamic fragmentation would be the only method by which substantial particulation could occur. The simulant fluid tests I carried out at Argonne National Laboratory demonstrated that the high pressure blowdown forces of gases and molten material from the primary system into the reactor cavity region would quickly disperse the major I 1281P120685 4-15
fraction of the molten material as a fluid wave. This is in contrast to a process in which fine-scale entrainment from the surface of the molten debris would occur. Consequently, the major fraction of material is quickly removed from the reactor cavity and is not available for l long-term particulation by hydrodynamic forces. The material that could a be available for long-term entrainment and subsequent dispersal into the containment atmosphere is that which could be held up on a g downward-facing surface in the high velocity gas strean. Carrying out g representative calculations for the Seabrook geometry and gaseous blowdown velocities for severe accident analyses demonstrates that only a few percent of the core inventory could be particulated and transferred directly into the containment atmosphere. This amount is well below the material mass that would be required to threaten the Seabrook containment as a result of direct heating. Therefore, it is concluded that, for the Seabrook design, debris dispersal could occur from the reactor cavity into the lower containment compartment, but only a small fraction of this material would be available for directly heating the containment g atmosphere. At Seabrook, direct heating creates no significant challenge g to containment integrity. 4.2.5 ISSUE 15 - CONTAINMENT PERFORMANCE Establishing the containment pressure capacity, including the shell structure and the penetrations and discontinuities, was identified by 3 both the NRC and by IDCOR as the most important single issue with respect 5 to the magnitude and release timing of accident source terms. A comprehensive probabilistic containment failure analysis was performed as g part of the Seabrook Station Probabilistic Safety Assessment. This g section provides a summary of the pressure capacity analysis. More details are found in Reference 4-1. The Seabrook Station containment is a large, dry, reinforced concrete structure, which houses an 1,200-MW(e), four-loop Westinghouse pressurized water reactor. It is a dual containment, designed and constructed by United Engineers and Constructors, with a design pressure of 52 psig. Recent source term analyses by the NRC and the industry have established the fact that accident source terms are significantly reduced if the containment remains intact for several hours after vessel melt-through. In order to determine the time and rate of radionuclide releases, it is necessary to know at what internal pressure the containment would realistically fail, where it fails, and what the leak area associated E with the failure is. Consequences of reactor accidents are also E influenced significantly by the time of containment failure and by the rate of radionuclide leakage, which, in turn, is determined by the leak g area. For each containment failure mode, size of the leak area is g required to determine whether the failure would result in a rapid release or in an extended slow release of the airborne radionuclide inventory. 4.2.5.1 Pressure Capacity, Leak Area, and Uncertainties for the Failure Pressure of Individual Failure Modes Failure of the containment is defined as significant leakage in excess of the design limit. Many different failure modes could lead to such 4-16 1281P120685
leakage. In most cases, the failure pressure for a given failure mode, P, is expressed as P = Pm.M.S in which P is the failure pressure; Pm is the median pressure capacity; M is a lognormally distributed random variable with a unit median value and a logarithmic standard deviation b3 , representing the analytical model uncertainty; and S is a lognormally distributed random variable with a unit median value and a logarithmic- standard deviation, bS. representing the material strength uncertainty. The distinction between modeling uncertainties and strength uncertainties is important when considering the correlation of failure modes. For a given pair of I failure modes, one of the uncertainty factors may be correlated, while the other may be independent. For example, shear failure in both the basemat and the cylindrical wall involves shear failure of a reinforced I concrete section. The strength uncertainty factor, S, incorporates the variability in the strength of the constituent materials, steel, and concrete. These uncertainties are common to both failure modes. Therefore, the random factors for strength are correlated. However, the I random factors for modeling are independent since modeling considerations for the shear force in different sections are, for the most part, not
.related. Correlation matrices are used to define uncertainty factor I dependencies for strength and modeling, which are assumed to be either perfectly correlated or perfectly uncorrelated.
I Failure modes that involve shell elements will be referred to as structural failure modes. Table 4-3 shows that the critical failure modes involve membrane failures. The cyliMrical wall and dome carry the pressure loads mostly by membrane tension. Since the concrete cracks at relatively low pressures, the tensile forces must be carried entirely by the reinforcing steel and the steel liner. The cylinder wall hoop is the most critical membrane tension. The median pressure that causes yielding of both the liner steel and the reinforcing bars is 157 psig. At larger pressures, large strains occur (see Figure 4-2), but this is not accompanied by leakage until the liner tears. The liner tears if the system is strained to its ultimate capacity, which occurs at a median estimated pressure of 216 psig. Membrane capacities are slightly higher for critical sections in the dome and considerably higher for other locations. Both the modeling and strength uncertainty factors for the I membrane failure modes are judged to be correlated. Hence, only the most critical membrane failure mode need be considered. The consequences of hoop failure depend on whether the liner or the reinforcing bars fracture first. Fracture of reinforcing bars propagates in an unstable fashion and is followed immediately by liner tearing and an almost instantaneous blowdown of the ' containment pressure. The I proportion of strength contributed by the reinforcing steel is significantly greater than the liner capacity. However, if the liner fails first and if the force carried by the liner can be transferred to I the reinforcing bars, leakage would occur through cracks in the concrete. Such cracks would continue to open until an equilibrium is reached where the leak rate matches the rate at which internal pressure is generated. I 1281P120685 4-17
The hoop strain in the reinforcing bars may differ locally from that in the liner, but the average hoop strain around the circumference is the same. Therefore, the failure sequence is determined by the average hoop strain for the liner and the reinforcing bars at failure . In constant uniaxial tension, a typical Seabook reinforcing bar (ASTM A625 Grade 60) fractures at a strain of about 10%. However, when embedded in concrete, the bar stress varies between a maximum at crack locations and a minimum at intercrack locations where. part of the membrane tension is carried by the concrete. At the critical section, the difference between the maximum and minimum stress in the steel is estimated to be 17.7 psi when the tensile strength of concrete (520 psi) is developed at intercrack locations. The strains in the reinforcing bars just prior to fracture are then about 10% at crack locations and 2% at intercrack locations. The average hoop strain at fracture of the reinforcing bars is about 4.7%. The liner consists of mild ductile steel (ASME SA526 Grade 60), for which the elongation at fracture in uniaxial tension is about 30%. The maximum hoop strain is reduced by biaxial effects and gauge-length effects. A 3 biaxial strain reduction factor of 1.73 is obtained by a method described 3 in Reference 4-16. The liner gauge length (anchor separation) is 20 inches, compared to 2 inches for the test specimen, resulting in a fracture strain reduction factor of 1.5 to 2.5. Even considering all reductions of the liner strain at fracture, it is unlikely that the liner will fracture at a hoop strain of less than 4.7%. Therefore, liner fracture before reinforcing bars fracture is considered unlikely. Even if the liner should fracture first, the reinforcing system could not carry the load shed by the liner, and reinforcing bars would fracture as a consequence. Hoop failure would therefore lead to a large blowout failure with an almost instantaneous loss of containment pressure. In contrast to the structural failure modes, the failure modes of penetrations do not occur as a direct consequence of applied pressure loads. They are caused by the displacements expected for the containment wall before hoop failure can occur (see Figure 4-2). Where restraints do not allow the pipe to move with the containment wall, large stresses are g induced. A containment structure includes a large number of penetrations 5 and local discontinuities that are normally overdesigned to assure that they are not controlling the design. It is therefore not a foregone conclus-ion that local failures will occur before gross structural failure. In order to take credit for such local failure modes, they must be proven to exist at pressures lower than the lowest pressure for gross structural failure. The penetrations and discontinuities were screened by reviewing the drawings, by a site inspection, and by performing screening calculations to identify the most vulnerable areas. The penetrations studied in more detail include the feedwater lines, the fuel transfer tube, and the mechanical penetrations. The feedwater line penetrations are forged, flued heads with pipe restraints inside and outside the containment. The external restraint is not designed for loading, due to radial displacements of the containment wall, and would fail without inducing a large load on the penetration. Inside the containment, however, the restraint prevents the line from moving with the containment wall. The induced tensile forces increase 4-18 1281P120685
E with wall displacement until, at a median estimated pressure of 179 psig, E either tne flued head penetration or the pipe fails. Based on an assessment of the relative strengths of the penetration and the pipe, and I the uncertainties therein, the probability that the penetration would fail is 17%. In the more likely case (83% probability) that the pipe fails instead of the penetration, the containment pressure boundary is maintained by the external feedwater pipe and valve. However, at even l larger displacements, the external feedwater line fails by interference with external structures. The probability of this occurring before the radial displacement for hoop failure is reached, given that the penetration does not fail, is estimated to be 25%, based on consideration of a number of contributing factors. The leak area for both feedwater penetration failure modes increases with the increasing containment wall deflection and is thus a direct function of the containment failure pressure. Such failures exhibit self-regulating leaks. An equilibrium is established, so the leak area is just large enough to relieve the pressurization source inside the containment, which is usually water boiloff driven by the decay heat. - The critical element of the fuel transfer tube pressure boundary is a set of stainless steel convoluted bellows to acconnodate the radial wall I di spl acement. Only a 6-inch radial displacement is possible before the containment wall bears on the fuel storage building, which occurs at a median pressure of 172 psig. The rigidity of the fuel storage building prevents further displacement. The stresses due to the restraining force acting on the containment are not critical because of the additional reinforcement of the containment wall at lower elevations. The resulting I discontinuity in the pressure-deflection relationship introduces a discontinuity in the probability distribution for the fuel transfer-tube penetration failure pressure. The mechanical penetration pipes pass through a welded ring plate, which forms the pressure boundary between the pipe and the sleeve. For three penetrations (X-25, X-26, and X-27) the pipe support conditions and the fillet weld strength to pipe strength ratio caused failure of the fillet welds before hoop failure, with an estimated median leak area of 0.5 square inches per pipe. This leak area would not increase with increasing pressures. After large containment wall deformations begin to occur, a number of failure modes are conceivable for which probability distributions for the failure pressure have not been explicitly calculated. These failure modes include the following:
- 1. Pipe penetrations not considered in detail.
- 2. Failure at electrical penetrations.
- 3. Personnel or equipment hatch penetrations.
- 4. Seals at purge line valves, personnel air lock and equipment hatch.
- 5. Liner tearing due to friction and adhesion of the liner to the concrete.
- 6. Microcracking due to imperfections in the liner or in welds.
I 1281P120685 4-19
For each of these potential failure modes, scoping analyses or other technical rearoning showed that either the expected failure pressure was well above those calculated for other failure modes, or that the failure mode was not likely to exist, or that it would yield only a small leak area. Nevertheless, each of the six failure modes was estimated to have a 5% probability of occurring before the wall hoop failure mode, which yielded a composite (mutually exclusive) failure probability of 26% for the group. 4.2.5.2 Containment Failure Categories Three distinct containment overpressure failure categories ^were defined to address differences in potential release consequences. These are: e Containment failure type A includes small leak-area failures for which the leak rate is too small to terminate the continued pressure rise. Thus, a type A containment failure will subsequently lead to 3 either a type B or a type C failure. Type A leak paths are limited g to an area of about 6 square inches. e Containment failure type B includes those failures where the leak area increases with increasing pressure until the leak rate balances the pressurization source. Type B failures thus are self-regulating, long-term releases. Type B leak paths include leak areas in the E range from about 6 square inches to about 0.5 square foot. E e Containment failure type C includes failures with large leak areas g and rapid containment blowdown. Type C failures typically include g the membrane failure modes. The leak area for type C failures is greater than 0.5 square foot. Each failure mode was associated with the appropriate failure category. As seen from Table 4-3, type A containment failures are dominated by mechanical penetration failures, type B. failures are dominated by high E energy piping penetration failures, and type C failures are dominated by 5 the cylinder wall hoop failure. 4.2.5.3- Composite Probability Distribution for the Containment Failure Pressure Containment f ailure types B and C can occur either by themselves or in E combination with a type A failure. The probability that a type A failure E occurs before a type B or C failure is 90%. In order to distinguish probabilistically between failure types B and C, the failure modes in g each category were probabilistically combined, while accounting for any E dependencies. For any given failure mode the strength uncertainties are independent of model uncertainties. However, between failure modes, the uncertainties can be correlated. This was accounted for by treating the standard normal variates as either correlated (equal) or independent (random) where appropriate. The combined median failure pressure for type C alone is 212 psig. 4-20 1281P120685
Local f ailure modes are independent of each other. However, all local failure modes are directly dependent on the containment wall displacement and therefore on the pressure at which a certain displacement occurs. All local failure modes were analyzed as being fully correlated to the pressure at which the ultimate hoop capacity is reached. Six discrete hoop capacities associated with their discrete probabilities, which are based on the uncertainties in the ultimate hoop capacity, were defined. Six cumulative distributions were then developed for the failure pressure of each type B failure mode (one for each discrete hoop capacity), expressing the conditional probability that the type 8 failure mode occurs, given that a type C failure mode has not occurred. Figure 4-3 shows the results for the fuel transfer tube penetration. Closing of the 6-inch gap between the containment and the fuel storage building causes the discontinuity. Figure 4-4 applies to the feedwater penetration where the combination of the two failure modes is responsible for the shape of the curves. Figure 4-5 applies to the combination of all other failure modes. Since these failure modes are correlated to hoop capacity, the three failure modes were combined separately for each discrete hoop I capacity, resulting in the family of combined discrete probability distributions shown in Figure 4-6. Now the composite probability distribution for type B containment failure modes is obtained by combining the six curves in Figure 4-6 according to their probabilities. The results are shown in Figure 4-7, separately showing the total cumulative probability distribution for the containment failure pressure for dry containment co'n ditions (no injection of the RWST) and for wet conditions (with RWST injection). For wet conditions, the contributions from type B failures and type C failures are also shown. Figure 4-7 indicates that the Seabrook Station containment is unusually strong. The median total f ailure pressure is 195 psig for wet conditions and 172 psig for dry conditions, respectively. These failure pressures correspond to 3.3 and 2.9 times the integral test pressure, respectively. Furthermore, it is noted that type B containment failures (leaks) clearly dominate over the type C (gross) failure modes. However, this conclusion is containment design-dependent. The bottom curve in Figure 4-7 shows the fraction of failures that are type C, as a function of pressure. If failure of the Seabrook containment occurs at 125 psig or lower, the conditional probability that the failure is of type C is 0.04 or less. At 185 psig, this conditional probability increases to 0.18, and, at 215 psig, it is 0.34. At even higher pressures, the type-C conditional failure probability reaches an asymptotic value of 0.4. Therefore, at Seabrook Station, leak-type containment failures are indeed more likely than gross containment failures. The containment enclosure building was also analyzed, and it was concluded that radial deflection of the containment wall in the area of the equipment hatch caused the enclosure building to fail before any of the containment failure modes occurred because the clearance in that area is only 3 inches. 4.2.5.4 Conclusions A methodology has been developed and implemented to realistically determine the failure pressure, failure mode, and leak area of failure for containments. Applied to Seabrook, the containment was found to be 4-21 I 1281P120685
unusually strong, but the uncertainties in the failure pressure are also significant. Therefore, quantification of uncertainties are also essential to a full understanding of the containment pressure capacity. Local failure modes can be quantified, and their assessment is design-specific. For Seabrook, local failure modes with extended slow release characteristics clearly dominate containment failure. These conclusions are plant specific but the methodology is readily applicable to other containment designs. 4.2.6 ISSUE 16 - SECONDARY CONTAINMENT PERFORMANCE The issue of secondary containment performance at Seabrook station has two distinct aspects. The first aspect addresses the performance of the secondary containment enclosure building, and the second aspect addresses the performance of the auxiliary building as a fission product retention structure. The secondary enclosure building at Seabrook Station is designed to mitigate the release of radionuclides for design-basis accidents when the release from the primary containment occurs because of leakage at the design-basis leak rate. The performance of the enclosure building at Seabrook Station was analyzed in the SSPSA. It was concluded that, for accident sequences where the containment reaches a high pressure before failure, the enclosure building fails because of structural interference with the primary containment before failure of the primary containment occurs due to either type B or type C failure modes. This is basically due to the narrow 3-inch clearance between the equipment hatch bulkhead and the enclosure building. The bending loads imposed on the enclosure building when. the equipment hatch bulkhead expands against it cause the enclosure building to fail locally before failure of the primary containment occurs. For accident sequences when the primary containment is bypassed, the enclosure building is also bypassed. Therefore, no consideration was given to fission-product retention in the enclosure building for any accident sequence that failed or bypassed the containment. However, for accident sequences where the containment remained intact, the performance of the enclosure building for holdup and filtration of fission products, as a result of design-basis leakage from the primary containment, was accounted for. The auxiliary building comes into consideration as a fission-product retention barrier applies to interfacing systems LOCA sequences when the radionuclides are released from the reactor coolant system into the ; auxiliary building. Such accident sequences, referred to as V-sequences, ' have been specifically analyzed for this project and the Seabrook-specific design features have been fully accounted for. The analysis is documented in Section 4.4. l 4.2.7 ISSUE 18 - ESSENTIAL EQUIPMENT PERFORMANCE Essential equipment to mitigate the release of radionuclides following a j core melt accident at Seabrook Station includes the equipment used to l provide containment heat removal and containment fission-product scrubbing. At Seabrook Station, both of these functions are provided by 1 i I 1 1 4-22
- 1281P120685 I i -- _ - - _ - _ - - _ - - - - - - - - - - - - - - - - - - - - - - - - - - - -
the containment spray system. The portions of the containment spray system that are located inside the containment include the containment sump and the containment spray headers, both of which are passive structural components that are not sensitive to the containment atmosphere conditions created by the accident. The only identifiable impact that the in-containment accident conditions could have on the operation of the containment spray system are related to the potential for plugging the sump suction. The sump is physically protected by a missile shield. The containment floor is sloped away from the sump to prevent any particulate matter with a specific gravity greater than 1.0 from being swept into the sump. A trash rack and a fine-mesh miner's screen prevents lightweight particles larger than 0.1 inch-diameter from entering the sump. Suspended particles of this size would not harm the RHR pumps or prevent the spray pumps from performing their function. Furthermore, the recirculation suction line penetrates into the sump 2.5 feet above the sump floor, so approximately 900 cubic feet of debris would have to accumulate in the two sumps combined before both containment spray suction paths could become blocked. This volume I corresponds to approximately twice the volume of the core debris that could conceivably be generated from melting the entire core region. On this basis, it was concluded that blockage of the containment spray recirculation suction path was not considered credible. The containment spray system was assumed to perform its intended function if the active components in the system were available and an actuation signal had been generated. 4.2.8 RELEASE CATEGORIES The radionuclide release categories need to be defined, so the risk from all potential accident sequences at the Seabrook Station can be properly represented. Having the SSPSA report available is a significant factor I in establishing the source terp categories because the SSPSA quantified a large number of accident sequences that were ranked according to their risk significance. In the SSPSA, six release category sets were defined to distinguish significant containment failure modes. With two changes, the six release category sets were adopted for this study. The first change relates to the release categories for containment bypass accident sequences. In the SSPSA, the containment bypass failure mode, which was designated release category set S6, represented two types of accident sequences. The first type included I accident sequences where the containment failed to isolate properly and a direct path from the containment atmosphere to the environment existed. Accident sequences that included failure to isolate a containment purge line penetration while in the purge mode were the most visible I contributors in this group of accident sequences. The second type of accident sequences represented by source term category S6 included the so-called V-sequences that are characterized by ruptures in the low pressure portion of the RHR system outside the containment. For this type of accident sequence, the release path leads directly from the reactor coolant system into the auxiliary building where the release path proceeds to the environment. As a result of significant advances in the interpretation of likely failure modes for this accident sequence, both I 1281P120685
by NRC-sponsored research and by IDCOR-sponsored research, it was decided to represent the sequences as separate release categories. This choice was further supported by the identification of significant differences in the s elease path characteristics with respect to retention of fission products, which is particular to the Seabrook design. Therefore, release category S7 was established to represent the V-sequences. The second modification in the source term categories for this study related to the basemat melt-through containment failure mode. In all recent aaalyses, the radionuclide source terms associated with basemat melt-through failure modes were shown to have very low release fractions. This conclusion is valid if the basemat melt-through containment failure mode g is not accompanied by a simultaneous atmospheric release. These could be g caused by the containment atmosphere leaking from the site of basemat melt-through to the environment. Since the Seabrook Station is built on rock soil, it was not possible to exclude the existence of such a direct atmospheric release following basemat melt-through. Therefore, in the SSPSA, two possible release categories were acknowledged for the basemat melt-through containment failure mode. The first possible release g category resembles that associated with an intact containment, which g would apply to a rock site if there is no direct leakage to the containment atmosphere. The second type of release was acknowledged as being potentially similar to the source term associated with late overpressurization of the containment. This source term would apply if a significant unfiltered leakage of containment atmosphere does exist. In the SSPSA, each of the two potential release types associated with the E basemat melt-through failure mode was assigned an equal probability. 5 Release category S3 represents the containment failure mode associated with late overpressurization, and release category SS represents source g terms for intact containment conditions. In the SSPSA, release category E S4 was initially designated to represent basemat melt-through failure modes. For the above reasons, it is no longer used to designate a specific release category. The release categories used in the SSPSA and in this study are shown in Table 4-4. Release category S1 includes all early containment failures ? that result in the rapid release of the airborne radionuclide at approximately the time of vessel melt-through. Accident sequences such as an aircraft crash into the containment are represented by this source E term. Release category S2 includes all source terms that are g characterized by a significant increase in the containment leak rate at or near the time of vessel melt-through, resulting in a continuous long-term release that may or may not be followed by a late overpressurization failure. The important characteristics of this source term is that most of the leakage occurs over an extended period of time, beginning at the time of vessel melt-through. This type of source term category has not been identified by other studies. It has been defined for Seabrook as a direct result of the containment failure analysis summarized earlier. It accounts for two facts: (1) the Seabrook containment is an unusually strong containment and (2) the Seabrook containment includes failure modes at identifiably lower pressure levels with a fixed leak area that is insufficient to prevent further pressurization of the containment. Release category S3 represents all accident sequences leading to late overpressurization of the I 4-24 1281P120685
r containment. Accident sequences in this category are characterized by successful isolation of the containment, but all containment heat removal [ systems fail. Release category SS represents all accident sequences where the containment remains intact and the radionucli'de release is determined by leakage from the containment into the enclosure building 7 interspace at the design-basis leak rate. Release category S6 represents L all accident sequences where a containment penetration greater than 3 inches in diameter has failed to isolate and therefore a direct leak , path from the containment atmosphere to the environment exists. Release category S7 represents all accident sequences in which the containment is bypassed by a leak path directly from the reac' tor coolant system into the RHR equipment vault. This source term category therefore' includes all the V-sequences. Two source terms will be defined for each release category representing 7 both a best estimate definition and a conservative definition of the L source term. The applicability of source terms quantified by the IDCOR program will be evaluated by carefully comparing the Seabrook de' sign and , the Zion design with regard to important design features for determining accident source terms. The Zion design was selected for this comparison because the Zion Station out of the four reference designs analyzed by the IDCOR program most closely resembles the Seabrook Station. This design comparison is documented in the next section. 4.3 ZION /SEABROOK DESIGN COMPARISON 4.3.1 PURPOSE One objective of this study was to make maximum use of available information about new accident source terms to the extent that these new source terms are applicable to the Seabrook Station design. The two major sources of Seabrook-relevant source terms available were from the IDCOR program and the SSPSA. Since the 10COR program studied the Zion _ Station as one of the four reference plants, a comparison between its design and the Seabrook Station design was made. This will develop a basis for determining which accident source terms determined by IDCOR for the Zion Station are applicable to the Seabrook design. Those applicable source terms will be adopted for this study. The remaining source terms ~ will be derived from the source terms developed for the SSPSA to the extent feasible. Any remaining source terms that cannot be meaningfully derived from existing sources of information will be separately analyzed by using the IDCOR source term methodology (the MAAP computer code, Reference 4-6). It is the objective of this design comparison section to determine which source terms require a Seabrook-specific study. 4.3.2 DESIGN COMPARISON TABLES 1 A tabulated design comparison between the Zion and the Seabrook Station 'I was developed for all systems, structures and design features that are considered to be important in the definition of accident source terms. This design comparison is documented in Table 4-5. Section number 1 in the comparison table indicates that the containment spray system for the two designs compares very closely and no significant differences in I I 1281P120685 4-25 l .
I source terms could arise from the small differences. A significant design difference exists for the containment fan coolers. The Zion Station includes five safety-related fan coolers. The Seabrook Station
~
does not include safety-related fan coolers nor does the design include nonsafety-related fan coolers of sufficient capacity for containment heat removal under core damage accident conditions. In the Seabrook design, the containment heat removal function is integrated into the containment spray system design by including a separate spray heat exchanger that is not present in the Zion design. The nonexistence of fan coolers in the Seabrook design is reflected in the definition of plant damage states and is fully accounted for in the analysis of accident sequences in the SSPSA. Section number 4 of the comparison table compares the emergency core cooling systems and indicates that the design of the two systems is so similar that no significant differences in accident source terms would arise. Section number 5 compares the containment isolation system and also indicates only insignificant differences. The auxiliary feedwater system is compared in Section number 6 of the table and also indicates 3 insignificant differences. Section number 8 compares the free 3 containment volume and dimensions. The containment free volume is very similar for the two designs. This is very important because the free containment volume is one of the key parameters for determining the pressure level and the rate of pressure buildup inside the containment under accident conditions. Item 89 (and 11m) in Table 4-5 does point to a significant difference in the two designs. The Seabrook Station design includes a 30-inch high curb on the containment floor surrounding the reactor cavity, while in the Zion design, this curb is only 6 inches high. The volume of water that has to collect on the containment floor before spillover into the reactor cavity can occur is therefore much larger for the Seabrook design. This means that in the Seabrook design, essentially all of the RWST contents must be injected into the containment before significant spillover of water into the reactor cavity occurs. In the Zion design, only a small fraction of the RWST must be injected before flooding of the E reactor cavity occurs. However, in both cases; the reactor cavity is dry E without RWST injection and wet with RWST injection. Another-important design difference is indicated by item 81. The Seabrook Station containment is designed and built as a reinforced concrete containment. The Zion design is a prestressed concrete design. This design difference has important implications for the potential failure modes and on the pressure capacity of the containment. This , design feature has been fully reflected in the Seabrook containment pressure capacity analysis discussed in Section 4.2.5. It is one of the g reasons for the very high pressure capacity of the Seabrook containment. 3 Section number 9 of the design comparison table addresses containment operating conditions and does not indicate any significant design differences. Structural heat sinks in the containment are addressed in Section number 10, indicating several substantial differences that tend to compensate for each other. The Seabrook design includes less exposed I 1281P120685
L
~
' steel, but compensates for that by a signficantly larger amount of exposed concrete. For the early containment failure modes or for the 7 containment bypass f ailure modes, the structural heat sinks cannot play L an important role. Therefore, no substantial difference in containment response is expected to result from these differences in the containment , structural heat sinks. For the late containment overpressurization [ failure mode, the concrete structural heat sinks absorb a signficant fraction of the total heat. Section number 11 compares the reactor cavity area, indicating a { - substantially larger reactor cavity for the Seabrook containment. However, in both designs, the RWST inventory is sufficiently large to 7 fill both the containment and the reactor cavity to the curb level. u Thus, there is no particular significance in this difference in volume. Associated with the volume difference is a larger reactor cavity floor area for the Seabrook design. This would translate into a smaller debris I depth, which would be more easily cooled for the Seabrook design. Both L designs contain sufficient reactor cavity floor area to quench and cool the anticipated core debris mass if the reactor cavity is filled with r water. Therefore, no particular significance is attached to the L difference in reactor cavity floor area. J Section number 12 addresses the reactor and reactor cooling system design
~
parameters. There is no significant difference between the two designs. The containment concrete composition is compared in Section number 13. The Seabrook Station uses a concrete aggregate of a basaltic composition, while the Zion concrete aggregate is a limestone composition. This
, difference is significant because the Seabrook basaltic aggregate does not contain decomposable components. The concrete penetration by core debris and the resulting core concrete interaction would not generate
_ noncondensable carbon dioxide and carbon monoxide in the Seabrook case. The limestone concrete includes a large element of decomposable calcium carbonite that releases carbon dioxide upon decomposition. The aggregate used at Seabrook Station is typical of the aggregate composition used in European reactor designs, while the aggregate used in the Zion design is typical of the concrete composition in U.S. power plants. This difference in concrete composition has been identified by the American Physical Society peer review of the NRC source term program as one of the reasons suspected for the differences observed between the European core-concrete interaction experiments and the U.S. core-concrete interaction experiments. In the U.S. experiments, much larger aerosol release rates are observed compared to the German experiments. [ Section number 14 compares containment leakage data, and item 14b identifies the existence of a secondary containment for the Seabrook I Station. The secondary containment at Seabrook Station has not provided a substantial mitigation of severe accident source terms and therefore is not expected to be a significant design difference. Section number 15 addresses containment penetrations and uses the appropriate FSAR tables as references for the comparison basis. The containment penetrations of most interest are the containment atmospheric purge lines. They can be open during normal operation and they provide a direct patn to the environment. The purge lines are 8 inches in diameter in the Seabrook I 4-27 1281P120685
design and 10 inches in diameter for the Zion design. The flow area in the Seabrook case is, thus, only two-thirds of the Zion flow rate. If such a penetration f ailed to isolate the initial radionuclide, release rates at Seabrook would be reduced until the release rates are determined by the rate of gas generation. The individual penetration design and its structural capacities has not been evaluated for Zion; therefore, no comparison is made. Section number 16 compares auxiliary building data. Comparison of the auxiliary building design features is important for the radionuclide release path characteristics for the V-sequence. This release is expected to occur in that portion of the auxiliary building that is occupied by the RHR system. Significant differences between the two designs in both the release path and the potential for mitigating the radionuclide source term associated with the V-sequence are identified. In the Seabrook design, the entire RHR system is contained in the lowest portion of the RHR cubicles. These cubicles are designed as thick-walled vaults that are separate from the primary auxiliary building. The vaults 3 are deep and have a relatively small cross-section. They represent a E small volume with relatively few surface areas for deposition of radionuclides before release to the environment. However, the RHR cubicles have no openings in the lower 30 feet. Approximately 24 feet of water can accumulate in both vaults. Therefore, a deep pool of water would cover the most likely release site for radionuclides. In the Zion design, the RHR system is located in the main auxiliary building, and the release of radionuclides would occur through the large volume auxiliary building that contains a large amount of surface area for deposition. However, in the Zion design, there is no potential for flooding the g location where radionuclides are expected to be released. g Finally, Section number 17 compares the containment failure characteristics. It identifies the very high pressure capacity of the Seabrook containment. This pressure capacity is particularly impressive when considering that the Zion containment pressure capacity is the third highest identified to date for any nuclear power plant containment. Only the Midland containment has a pressure capacity intermediate between the Zion Station containment and the Seabrook Station containment. Furthermore, for the Seabrook containment a specific failure location and leak area was identifed as the dominant failure location. No specific failure location was identified for the Zion containment. 4.
3.3 CONCLUSION
S Several design differences were identified between the Zion and Seabrook containment that affect the frequency of accident sequences or the plant g damage state to which a given accident sequence is assigned. These are 3 not further evaluated because they all have been acccounted for in the SSPSA. Other differences affect the timing of degraded core phenomena during core damage accidents. These affect the available time to recover failed equipment, either to prevent core damage or to prevent containment failure if core damage has occurred. The latter aspect of recoverability is a significant aspect of this study, as discussed in Section 3.2. Additionally, a few differences have been identified that affect the I 1281P120685
radionuclide source term itself. Two differences are potentially important. The first relates to the composition of the concrete aggregate. In the Seabrook design this may preclude an enhanced release of the lanthanide radionuclide group during penetration of core debris I into the concrete. The second significant design difference for source terms pertains to the layout of the room in which the RHR systems are located. Without flooding of the release location, the smaller cubicle I layout at Seabrook would be expected to result in an enhanced release of radionuclides compared with that analyzed for the Zion configuration. Overall, the design comparison leads to the following conclusions: e For accident sequences with an isolated containment, but no containment heat removal, the Seabrook containment will exhibit a I longer time until containment overpressure failure occurs. This is expected to result in a source term reduction for Seabrook, particularly with respect to the release of tellurium, which occurs, in part, at the time when the core debris reheats to high I temperatures. For accident sequences with failure to isolate a direct atmospheric I e pathway penetration, the source term for the Seabrook Station would also be expected to be somewhat smaller and with a longer release duration. This is basically due to the smaller size of the containment purge line penetration. e For containment bypass sequences (V-sequences), the release paths and the radionuclide mitigation mechanisms are significantly different. I Without analysis, it would not be possible to conclude whether the source term associated with a V-sequence at Seabrook would be larger or smaller than the IDCOR calculated source term for the Zion Station. I After reviewing the IDCOR source term analysis for the Zion Station and comparing the Seabrook containment design with the Zion containment I design, it is concluded that the conditions and specific configurations for which the Zion source terms were determined are equally applicable to the Seabrook Station for all accident sequences except the V-sequence. I Therefore, information from the IDCOR-calculated Zion source terms for the station blackout sequence and for the containment isolation failure sequence can be used and are relevant to the Seabrook Station design. Furthermore, neither the SSPSA source terms nor the IDCOR source terms . for the containment bypass sequences (V-sequences) are applicable to the conditions encountered at Seabrook Station. For this reason, a Seabrook-specific analysis of accident source terms associated with the I V-sequence was performed by using the computer codes that were developed under the IDCOR program. The same codes were used to determine the accident source terms for the Zion Station. The analysis of the V-sequence accident source terms for Seabrook Station is discussed in Section 4.4. 4.4 V-SEQUENCE ANALYSIS A containment bypass event, or V-sequence, was modeled to determine the environmental fission-product release from the Seabrook containment and I 4-29 1281P120685
auxiliary buildinq. The following section discusses the Seabrook g physical plant, the core and containment analysis, fission product 5 transport and behavior, and the resultant environmental release of fission products for the Seabroook V-sequence. 4.
4.1 DESCRIPTION
OF PHYSICAL PLANT AND SYSTEMS The residual heat removal system for the Seabrook plant consists of a E dual train system, each of which includes a pump and a heat exchanger. 5 The system is used to remove decay heat from the reactor core during shutdown conditions and is also part of the low-head safety injection system for both the injection and recirculation phases. Each train is normally isolated from the reactor coolant system by two motor-operated isolation valves in series on the suction side (hot leg) and two isolation check valves on the discharge side (cold leg). The two trains are cross-connected by two normally open motor-operated valves. Each train has overpressure protection in the form of spring-loaded relief valves on both the suction and discharge portions of the system. The g suction side of the RHR system has a relief valve with a 20-gpm capacity 3 and a 2,485-psig setpoint located between the two isolation valves and a 900-gpm capacity relief valve with a 450-psig setpoint located downstream a of the second isolation valve. The discharge from both of these valves g is routed to the pressurizer relief tank. The discharge side of the RHR system has a relief valve located between the RHR heat exchanger and the outboard isolation valve with a 20-gpm capacity and a 600-psig setpoint. The discharge from this valve is routed to the primary drain tank located in the auxiliary building. The discharge side of the RHR system also contains a normally open motor-operated valve between the RHR relief g valve and the containment wall. A schematic of the RHR system is 3 presented in Figure 4-8. A description of the RHR suction and discharge-side piping is also given in Table 4-6. The RHR system is routed from the reactor coolant system through the containment, the emergency enclosure building, and into the equipment I vaults, which are part of the auxiliary building. The RHR pumps and the RHR heat exchanger are located in the equipment vaults along with the containment spray pumps, containment spray pump heat exchangers, and the high-head safety injection pumps. A simplified plan and elevation view of the equipment vaults is given in Figures 4-9 and 4-10, respectively. The RHR suction line joins the reactor coolant system at Elevation -19'0", passes through the containment at Elevation -18'5", enters the equipment vault at Elevation -29'5", and terminates at the RHR pump at Elevation -57'4". The suction line is approximately 100 feet in length inside the containment and 125 feet in length outside the E containment. The line has nine 90-degree elbows inside the containment. 5 The RHR discharge line joins the RHR heat exchanger at Elevation -28'10", passes out of the equipment vault at Elevation -21'8", passes into the contait snt at Elevation -18'5", and ends at the RCS at Elevat a -10'3". The length of the RHR discharge line is approximately 180 feet inside the containment and 110 feet outside the containment. The line contains thirteen 90-degree elbows inside the containment. I 4-30 1281P120635
The RHR system is isolated from the ECCS system by: e A check valve and a locked-open valve in the line from the RWST. e A normally closed motor-operated valve in the line from the containment recirculation sump. A check or a locked closed valve in the line to the containment spray I e pump. e One normally closed motor-operated valve in the line to the safety injection pump and charging pump. Overpre:surization of the RHR system is unlikely to result in the propagation of the overpressure incident to other systems. Additionally, the action of the SI actuation system and the subsequent valve realignments have been evaluated. The valve realignments will not result in degradation of the capability of the high-head or charging pumps to I operate in the injection mode and draw water from the RWST because the valve realignments maintain isolation between the low-head and high-head portions of the system. The RHR system has a design pressure of 600 psig. The system piping is composed of schedule 40S, 304 stainless steel piping throughout. Pipe I sizes are given in Table 4-6. The RHR pump is an Ingersol Rand unit, which also has a design pressure of 600 psig. The RHR pump seal is a mechanical seal unit that is designed for 600 psig, but which undergoes a shop cold hydro at 1,200 psig. The Seabrook equipment vault ventilation system is part of the emergency enclosure building ventilation system and has a flow rate to and from the I vaults of 23,580 scfm per vault. The system provides makeup air to several lower levels of the equipment vaults (19,020 scfm to Elevation -61'0" and 4,560 scfm to Elevation -50'0") and exhausts f rom the uppermost level at Elevation 25'6". The exhaust from the ventilation I system is processed through a filtration system consisting of moisture separator, absolute filter, carbon filter, backdraft damper, and fan and is exhausted to the plant vent. The system also contains fire dampers in the exhaust line that close when a fusible link reaches a temperature of 1650F. Due to the passive heat sinks between the break location and the fusible links and the creation of a suppression pool for condensation I of break flow, it is uncertain whether this event would trigger the isolation of the ventilation system on temperature. 4.4.2 ANALYSIS OF THE OVERPRESSURE EVENT The failure of the isolation valves in either the suction or discharge side of one RHR train would be expected to overpressurize both trains of I the RHR system. Analysis of the capability of the piping to withstand such an overpressure (Reference 4-17) shows that, for a fully liquid system (such as the Seabrook RHR system) the piping stresses from the I dynamic and static loading of the overpressure event would remain well below the ultimate strength of the material. Analysis of the dynamic I 1281P120685
loading for a system that is not water filled also indicates that the piping stresses will not exceed the ultimate pipe strength for the construction materials. The location of highest stress is one of the 90-degree elbows in the pipe. Based on the layout of the RHR system as described in the previous section, any pipe failure that is a result of the dynamic impact of the isolation valve failure overpressure would be expected to occur inside the containment, and the event sequence would resemble a small LOCA. However, this is not the most probable case as most of the RHR piping inside the containment is designed to RCS pressure. All of the piping inside the containment on the discharge side of the RHR is designed to primary system pressure, and the piping from a the RHR relief valves to the RCS is designed to primary system pressure. E The weakest point in the RHR system was determined to be the RHR pump seal mechanisms. Analysis of the mechanical pump seal assembly indicates that the mechanical seal assembly would remain in place for this event although some degradation of the sealing components may occur. A conservative evaluation of the impact of the overpressure on the pump 3 seal would be to assume the loss of all of the internal components of the 3 mechanical seal. This would result in a leak area of approximately 1.3 square inches per pump, or a total leak area of 2.6 square inches. Destruction of only the 0-ring seals in the mechanical assembly would result in a leak area of 0.166 square inches per pump. The discharge of reactor coolant to the equipment vault would flood the lower levels of the equipment vaults because the break flow is significantly greater than the equipment vault sump pump capability for the large leak case. For flows from a break somewhat smaller than that E corresponding to the small break size, the operation of the sump pumps 3 may prevent flooding of the equipment vaults. However, such a leak is within the makeup capacity of the charging pumps and would thus be considered a nonevent. A flood in the equipment vaults would fail the containment spray pumps and the RHR (low-head safety injection) pumps early in the event. The high-head safety injection pumps would fail at some later time in the event. The break flow to the equipment vault would also result in a pressurization of the vault area. The pressure ! relief would be provided by the ventilation system, the entrance / exit i doors at Elevation O'11", and the concrete plugs on the roof of the g l equipment vaults at Elevation 25'6". E 4.4.3 DESCRTPfl0N OF EVENT ANALYSIS AND MODELS The V-Sequence for the Seabrook plant was analyzed by using the MAAP computer code, Version 2B. This code gives the fission-product release l from this event sequence to the equipment vault. The V-Code, g l Reference 4-17, which is a follow-on code to the MAAP code for analysis 5 of the fission-product retention in the equipment vault area was used to l analyze the fission-product releases to the environment. The MAAP code l allows for only one break location (except for the RCP seal LOCA). In order to model the RHR pump seal LOCA and RHR relief valves for the Seabrook V-Sequence, a scoping run was used, with the break modeled as the total LOCA of the RHR pump seal and relief valves, to obtain the proper timing results. The accident sequence was then modeled by using I I 1281P120685 l l
I the break flow to model the RHR pump seal leak to the auxiliary building I and by using the pressurizer PORVs in the code to model the RHR relief-valve flow to the pressurizer relief tank. I In the MAAP code, a primary system break can be modeled as either a hot leg break or a cold leg break. A model for the flow from the primary system, through the RHR piping, and into the auxiliary building is not provided in the code. The model relieves water directly from the RCS to I the auxiliary building. Thus, the model neglects the break flow reduction due to the pressure drop in the long RHR piping. This results in an overprediction of break flow. Also, MAAP does not include I deposition of fission products in the long RHR piping run to the pumps, which may result in an overprediction of fission-product release. However, calculations show that the amount of fission-product retention may not be significant. The V-Code is run sequentially after the MAAP code to analyze the fission-product retention in the auxiliary building equipment vaults for Seabrook. The code process adjusts flow through the stacked nodes in order to maintain the pressure at the initial value in all nodes. This can result in modeling slightly shorter residence times for fission I products in the equirment vaults, which would tend to slightly overpredict the totai fission-product releases. Finally, the V-code balances the ventilation system mass flow into and I out of each nodal compartment. Since the volumetric ventilation flow into and out of the equipment vaults is specified as input, the code predicts a bypass flow around the filtered ventilation system in order to maintain an unpressurized system. This bypass flow is caused by the mass and ener gy added by the break flow and allows fission products to escape filtration. Thus, the overall fission-product release will be overpredicted. 4.4.3.1 Core and Containment Behavior Initially the reactor is assumed to operate in an equilibrium, full-power condition with letdown flow and makeup flow via the charging system. The two motor-operated valves on the RHR suction line (hot leg side) fail I catastrophically and pressurize the RHR system to reactor coolant system pressure. Since the cross-connect between the two RHR trains is open, both RHR trains are pressurized. As a result of the overpressurization I of the RHR system, the RHR pumps seals are assumed to fail and create a loss of coolant accident into the equipment vault area of the auxiliary building. The LOCA size is equivalent to a 2.6-square-inch break. Also, as a result of the RHR overpressurization, the relief valve on each RHR I suction line will open (setpoint of 450 psig) and discharge to the pressurizer relief tank. I The event chronology for this analysis is given in Table 4-7. The time histories for the reactor coolant system pressure and temperature are given in Figures 4-11 and 4-12, respectively. The containment pressure I and temperature transient for this event, as represented by the upper containment compartment above the operating deck, are given in I 1281P120685
I Figures 4-13 and 4-14, respectively. The analysis shows that the reactor coolant system pressure quickly dreps to the ECCS actuation setpoint that initiates the HPI pumps. The LPI pumps are assumed to be inoperable due to the pump seal LOCA. The system pressure continues to drop to the saturation pressure of the steam generators. As the cold SI water is delivered to the reactor coolant system, the system pressure and . temperature continue to fall and the accumulators begin to inject. The system has become water solid during this time interval. The pressure g decreases to the RHR relief valve setpoint and modulation of the relief g valves begins at approximately 12 minutes. The RCS pressure remains constant at this point, and the accumulator water is depleted in a little over 1 hour. At this time, the containment spray pumps in the equipment vault are already inoperable due to the flooding caused by the RHR pump seal LOCA. The RCS pressure falls below the relief valve setpoint at approximately 2.7 hours, and the relief valves close. Approximately E 2.8 hours after the event begins, the water level in the equipment vault E reaches the safety injection pumps, thereby rendering these pumps inoperable. RWST injection continues at this point through the charging g pumps. The flow from the charging pumps is sufficient to maintain RCS g inventory, and the break flow through the RHR pump seals is sufficient to remove the core decay heat. This condition continues until the RWST water is depleted in approximately 6.4 hours. At the time the RWST inventory is depleted, switchover to the l recirculation mode is unsuccessful due to the unavailability of the RHR E pumps. From this point, inventory continues to be lost from the RHR pump 5 seal LOCA to the equipment vaults and core uncovery, heatup and melting occur. The core water level, as a function of time, is given in E Figure 4-15. The core temperature representation from the hottest core node is presented in Figure 4-16. The reactor vessel fails after I approximately 11.5 hours. The reactor coolant system pressure and temperature at the time of failure, as shown in Figures 4-11 and 4-12, are 140 psia and 290 F, respectively. The containment conditions at this time, as shown in Figures 4-13 and 4-14, are 30 psia and 211 F. At this time, approximately 750 pounds of hydrogen have been produced, which is 3 equivalent to a 36.1% invessel zircalloy-water reaction. The containment 5 contains approximately 1,635,000 pounds of water, which is less than that required for spillover to the reactor cavity. Based upon the conditions g at the time of reactor vessel failure, the vessel failure phenomena can g be characterized as a nondispersive event into a dry reactor cavity. A plot showing the mass of core material in the reactor cavity, as a function of time, is given in Figure 4-17. Since the accumulators have previously emptied into the RCS, no aporeciable water is introduced into the reactor cavity at vessel failure, and the containment only slightly pressurizes by an additional 6.7 psia. After vessel failure, the g containment pressure increases very slowly since a direct pathway is open E to the environment through the failed reactor vessel and the RHR pump seals. Core-concrete interaction occurs throughout the remainder of the event, with a maximum vertical penetration of approximately 2.3 feet at 24 hours, as shown in Figure 4-18, which shows the vertical concrete penetration time history. l Figure 4-19 shows the RHR pump seal break flow rate into the RHR va Jlt, and Figure 4-20 shows the water level in the vault as a function of t 4-34 1281P120685
l 1 l time. It is seen that the spray pumps flood at about 30 minutes, the RHR pumps flood at less than 2 hours, and the SI pumps flood at less than 3 hours after the initiating event. The vault water level continually I reaches a depth of 28 feet (Elevation -33'0") in each vault. At this depth, each RHR pump seal is approximately 20 feet below the water I l surface. l 4.4.3.2 Fission-Product Behavior I The fission-product release begins after approximately 8.5 hours, and the l principal portion of the release is completed by 24 hours. The fission products are classified into six groups, based on similarities in their chemical behavior, as shown in Table 4-8. The time history for the l fission-product release to the equipment vaults for each group are presented in Figures 4-21 through 4-26. l As shown in the figures, the fission-product release is composed of three distinct stages. The first phase, from 8.5 to 9 hours, represents fission-product release f rom the fuel due to temperature-enhanced diffusion and occurs before the beginning of core melt. During this time, the major portion of the fission product release from the fuel rods is retained within the reactor coolant system. The second phase of the fission product release occurs from 9.75 hours until 10.5 hours. This l phase of the release is associated with the core melt and, as in the previous release, a major portion of these fission products are retained I l in the reactor coolant system. The third phase of the fission-product release begins after approximately 11.25 hours and is associated with the I revaporization of the fission products retained in the reactor coolant system, as the hot gases from the core concrete flow through the reactor coolant system and out the RHR pump seal break. I l The fission-product release into the equipment vault for the large RHR pump seal break case is from the pump seals, which are submerged under approximately 20 to 30 feet of water that has accumulated in the equipment vaults from the LOCA. The water release to the equipment vaults for ;he event is approximately 2,676,000 pounds, of which approximately 48% is released prior to the flooding of the SI pumps at 2.8 hours and the remainder is released prior to the time the RCS water level drops below the level of the RHR line at 7.3 hours. This pool of water is subcooled throughout the event sequence and will therefore remain in place and act as a large suppression pool for scrubbing the l I fission products released to the equipment vaults. It is further expected that some additional depletion of fission products will occur in the equipment vaults. 4.4.4 FISSION-PRODUCT RELEASE The release of fission-product material to the environment from the equipment vault is a function of the release rate of material to the equiprant vaults, the scrubbing efficiency of the suppression pool, the deposition of fission products in the eauipment vaults, the availability I of the equipment vault ventilation system, and the efficiency of the equipment-vault ventilation systen. filtration. The release of fission I 1281P120685
I product material to the equipment vault, as a function of time, is given in Table 4-9. For the case of the large pun.c-seal failure break size, the equipment vault would contain at least 30 feet of subcooled water at the beginning of the fission-product release to the vault. This water level will be maintained throughout the event sequence. The pool becomes saturated after approximately 24 hours. Thus the fission-product release to the 3 equipment vault would be subject to efficient scrubbing by the E suppression pool, and a decontamination factor of 1,000 is justified, based on the results presented in the IDCOR Tll.2 report (Reference 4-18) for the Grand Gulf fission-product release quantification. The releases to the environment for this case are presented in Table 4-10. A second case was analyzed in which no suppression pool was present. In this case, the releases to the environment would be a function of the operability of the equipment vault ventilation system. As described in Section 4.4.1, it is not clear that the fire dampers would isolate the g equipment vault ventilation system. In addition, no procedures exist to 3 instruct the operator to isolate the ventilation system for this type of event. Analysis of the fission-product release to the environment with the ventilation system in operation are given in Table 4-11. The analysis of these releases includes fission-product deposition in the equipment vaults and a decontamination factor of 100 for the filtration system. This analysis also reflects bypass leakage from the building through the roof hatches due to the pressurization effect of the break flow. Although most of the fission products were transported through the ventilation system, a filtered route, some fission products were released into the bypass flow and escaped filtration. Thus, the analysis results indicate that over 99% of the fission-product release to the environ aent occurs via the bypass route. The case with no suppression pool and with the ventilation system completely isolated was also analyzed, and the results are presented in Table 4-12. In this case, only fission-product reduction due to E deposition in the equipment vaults as the fission products traverse the E upward path from the RHR pump seal to the roof hatches was included. 4.4.5 CONSIDERATION FOR EMERGENCY OPERATING PROCEDURES The V-Sequence, as analyzed in this study, is included in the plant Emergency Operating Procedures. However, the operator, based on information available, must diagnose the event as a V-sequence and respond accordingly. The initial indications of the event in the control room would include loss of pressurizer pressure and inventory, a safety injection signal, and pressurizer relief tank pressure rise. Since the PRT rupture discs quickly blow out, a containment pressure rise will also be observed shortly after the event begins. The radiation monitors in the equipment vault may alarm, depending on the amount of activity in the reactor coolant system. A sump-level increase in both the containment and the equipment vault will also be recorded. Since the event will resemble a I 1281P120685 t
L c L medium LOCA with a loss of the PRT, the operator would not be expected to manually blow down the steam generators to maintain a secondary side heat p sink as he would be expected to do for a small LOCA for which the t procedures direct him to blow down the steam generators. Additionally, the operator would not be expected to manually initiate containment spray because the pressure never reaches the spray setpoint although the I pressure increases to approximately 30 psia and remains at that point. Within 10 minutes, the RHR failure due to the pump seal LOCA will give 7 the operator an indication of a V-sequence. If there is no pump failure, L then, in approximately 1 hour, the operator will receive an indication that the RHR pumps are no longer operational due to flooding in the ~ equipment vaults. After approximately 2.8 hours, the high-head safety L injection pumps will shut down due to flooding of the pump motors. At this point, the operator has only the charging pumps to maintain reactor coolant inventory. It will be apparent to the operator that the ECCS recirculation is not operational due to the loss of the RHR pumps and the safety injection pumps. Several operator action states could be assumed in the analysis of this - event, as described above, based on information available to the operator and the current Emergency Operating Procedures for the Seabrook plant.
- Operator recognition of the exact event sequence and prompt actions could
~ mitigate the event consequences substantially. These operator actions include: ~ e Manual actuation of the steam generator steam relief valves to blow down the steam generators and provide a long-term heat sink at low ] e RCS pressures. Throttling the HPI pumps (safety injection and charging) to minimize break flow and conserve RWST inventory. e Initiation of RWST makeup. o Closure of the RHR cross-connect line to prevent break flow from both RHR pump seals and prevent flooding of one safety train. e Closure of all isolation valves in the RHR system to attempt to terminate the break flow. I e No manual initiation of the containment sprays and termination of the sprays should automatic initiation occur. Depressurization of the reactor coolant system to minimize break I e flows and prevent flooding of the high-head SI pumps. o Continued operation of the equipment vault ventilation system through the filtration system to minimize environmental releases. e Establishing an alternate recirculation path from the containment I sump or equipment vault to the high-head safety injection pumps and/or charging pumps. I 1281P120685
I Based on the establishment of proper procedures and training, the probability of a degraded core event resulting from the V-Sequence could be substantially reduced. Additionally, these operator actions would reduce the consequences from a degraded core event resulting from a V-Sequence. 4.4.6
SUMMARY
The V-Sequence for the Seabrook plant exhibits a different behavior from that previously analyzed for the IDCOR reference plants, both in terms of the core and the RCS behavior and the potential fission-product releases. The analysis of the V-Sequence for the Seabrook plant was based on the specific characteristics of the plant, which include the RHR relief valves that relieve to the pressurizer relief tank and the assessment of the RHR pump seal LOCA-equivalent break area. This results E in a transient that the operator might first diagnose as a medium-break E LOCA inside containment, as opposed to diagnosing a small-break LOCA outside containment for the transient analyzed for the IDCOR reference g plants. In addition, the design of the area in which the RHR pumps are E located is a small building that would contain the break fluid in the immediate area of the pumps. Thus, while fission-product retention is limited in the equipment vault -(due to short fission-product residence times), the break fluid creates a pool similar to a BWR suppression pool, which provides excellent fission-product scrubbing. In the analysis of the V-Sequence, the fission-product releases for Seabrook were found to be higher than those reported for the IDCOR reference plants. However, the fission-product releases for Seabrook in this analysis are still lower than those reported in the Seabrook, PSA and other PRA studies. The Seabrook V-Sequence analysis shows that the larger RHP. pump seal LOCA E is not necessarily the most conservative. Although a smaller RHR pump 5 seal LOCA would give the operator more time for intervention before core degradation, in the event of a release of fission products from the fuel, g a small RHR pump seal LOCA would not result in significant break flow to E the equipment vaults, which would minimize the effect of the suppression t pool fission-product scrubbing. A small pump seal LOCA would not result j in a significant break flow to the equipment vaults, which would minimize l the effect of the suppression pool fission-product scrubbing. The analysis shows that without fission-product scrubbing of the suppression pool, the release fractions would be one to two orders of magnitude 3 higher than those with the scrubbing effect. Thus, it is important to 5 maintain the pool in the equipment vault for the Seabrook V-sequence case. The Seabrook analysis did not include pressurizing the containment to the containment spray initiation setpoint. Pressurizing to the spray setpoint would have resulted in the spray system drawing from the RWST l leaving less water for emergency core cooling and, barring operator action to terminate the spray, would have resulted in a much earlier core degradation time. l l Finally, the Seabrook analysis shows that the ventilation system might not be isolated, as was the case for the Zion IDCOR reference plant. E 4-38 1281P120685
L Failure to isolate the ventilation system, however, does not significantly impact the fission-product release fractions. [ 4.5 RELEASE CATEGORIES 4.
5.1 INTRODUCTION
AND OVERVIEW In this section, the release categories are defined quantitatively for use as accident source terms in the consequence calculations that are p described in Section 5. No consensus of experts yet exists about the L timing and magnitude of accident source term releases for specific accident sequences. The industry has presented its assessment of accident source terms in the form of the IDCOR technical summary report ( and in the IDCOR technical task reports (References 4-19 and 4-20). There has not been the same definitive statement from the NRC. The NRC technical report on this issue, NUREG-0956, has just been released for comments. A cursory review of this report indicates that quantitative { source term information for large, dry PWRs is only provided for the Surry plant, except for a basemat melt-through source term portion; p furthermore, all accident sequences analyzed involve either early L containment failure or basemat melt-through. No accident sequences are analyzed in which the containment either remains intact or in which the containment f ails due to late overpressurization. In the Seabrook SSPSA, [ these two types of accident sequences were found to contribute almost 90% of the total frequency of core damage. A first comparison of accident source terms calculated by the NRC contractors for the Surry Station with { those calculated by IDCOR for the Zion Station indicates that there are still differences in the timing and in the release fractions for the individual radionuclide groups, even for very similar accident r sequences. At this point, it is difficult to determine the extent to L which these differences are due to differences in the plant designs or whether they are largely due to differences in the analysis methods and assumptions used by the different groups. In light of this evidence, it [ appears that a fair degree of uncertainty still exists about accident source terms. However, even the NRC-published accident source terms in NUREG-0956 show a reduction in the source term magnitude compared to the WASH-1400 source terms. In order to reflect these remaining uncertanties in the source term magnitudes, two source terms were defined for each release category. The first is a best estimate source term and is intended to represent the radionuclide release that would most likely occur for the most likely accident sequence contributing to a particular I l release category. The second source term is intended to represent a conservatively defined source term category. This conservative source term is not intended to represent an absolute upper bound source term, but rather it is intended to represent a source term that is not very I likely to be exceeded for the spectrum of accident sequences that contribute to a particular release category. All accident release categories analyzed for Seabrook Station exhibit an extended release time history spanning several hours. Any release category that exhibited a time-phased release or an extended-duration release was defined as a multipuff release in which the individual releases were timed to correspond to the major release phases. A 4-39 1281P120685 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ \
I single-puff release model was only used for the source terms representing an intact containment and for the source terms representing a late overpressure failure of the containment. The radionuclide release for an intact containment (SS) was so small that the contribution to health effect risks is negligible, and a multipuff release analysis is not E warranted. The radionuclide release for a late overpressure failure (S3) 3 occurs in the time frame of 2 to 4 days, and there is a 10% to 20% probability that the release duration is less than 2 hours. For such g late releases, the difference in health risks between a puff release and 3 an extended release is small. Again, a multipuff release analysis is not warranted. With the exception of the intact containment source term, all other source terms were analyzed and quantified under two assumptions that tend to overpredict the accident source terms. First, it was assumed that the containment spray system did not operate to mitigate the radionuclide source term in the containment atmosphere. Second, it was E assumed that the RWST is not injected into the containment during the 5 accident sequence, which results in a dry reactor cavity with an uncooled debris bed penetrating into the reactor cavity basemat. These were g defined as dry accident sequences in the SSPSA. Dry sequences g contributed the following percentages to the total frequency of all sequences in the respective release category sets: I Release Category Percent of Frequency Set from Dry Sequences S2 98 S3 58 S4 100 S6 100 Representing all sequences as dry sequences only represents a slightly conservative assumption. The accident source terms used as the basis for consequence analyses include consideration of radionuclide retention in the reactor coolant system and in the containment or in the auxiliary building. No credit was taken for radionuclide depletion in the leakage path from the containment or from the auxiliary building to the environment. The type B and type C containment failure modes have large E enough leak areas that no substantial deposition of radionuclides would 5 be expected in the passage through the containment wall. For type A containment failure modes, the leak path geometry would not be very g f avorable for deposition. However, for all type A and type B containment 5 failure modes, the leak path is known to go into another building; namely, the pipe chase, the penetration area, the spent fuel storage building, or the enclosure building. These buildings may have experienced some structural damage due to interference with the primary containment structure before the leak occurs. Nevertheless, they would offer some radionuclide retention, which was neglected because the E location and type of damage in the buildings surrounding the containment 5 are not known. I 4-40 1281P120685
4.5.2 BEST ESTIMATE RELEASE CATEGORIES I The best estimate release categories are shown in Table 4-13. In order to distinguish a best estimate release category from a conservative I category, the letter B or C is used after the source term category designator. Therefore, a release category designated as SlB would designate the best estimate release category for source term category Sl. Multipuff releases are modeled as three consecutive puff I releases. They are designated -1, -2, and -3. Thus, release category S1B-2 indicates the second puff for the best estimate release category of source term category Sl. Release category S1B represents the best estimate release category for accident sequences leading to early containment failure due to either overpressurization or the generation of missiles that penetrate the containment wall. Due to the exceedingly high failure pressure of the containment, with a median capacity in excess of 190 psia, no accident sequences have been found in the SSPSA that are expected to cause containment failure as a direct result of rapid pressurization events. Missiles with sufficient energy to penetrate the 3-1/2-foot heavily reinforced concrete dome are difficult to imagine. In the past, I in-vessel steam explosions have been considered in this context, but there is now agreement that such a missile is exceedingly unlikely to fail the containment. In the SSPSA, aircraf t crashes into the containment have been found to dominate the low frequency of this source term category. Due to the proximity of the Pease Air Force Base from where F-lll military aircraft operate, aircraft crashes into the containment with sufficient energy to penetrate the containment are I remotely possible. A new accident sequence contributing to release category S1B is due to the reanalysis of the V-sequence reported in Section 3.1 The small likelihood of an RHR piping failure was retained I in order to account for the possibility that a piping flaw may exist that does not exhibit itself at the hydro test pressure, but which could cause piping failure at pressures significantly higher than the test pressure. Since the accident source terms developed for the V-sequence specifically address the pump seal failure mode, the small frequency of accident sequences associated with V-sequences in which the piping fails have also been assigned to source term category Sl. V-sequences involving a pipe l failure now dominate source term category S1. Thus, the dominant accident sequence for S1 now involves a containment bypass failure that preexists prior to core overheating. Therefore, the release of I radionuclides will be a slowly evolving process that follows the core overheating and migration of fission products through the primary system and into the containment or auxiliary building. The IDCOR program did not analyze an accident sequence of this type. In the SSPSA, this type I of accident sequence was represented by release category 55V. The SSPSA release categories are summarized in Section 4.1.2. Release category T6V-d was used as the basis to represent the best estimate I source term for release category 51 with one modification. The time for the release start was reduced from 4 hours to 2 hours to include the range of release timing of V-sequences. Current uncertainties in the release characteristics of Te and Ru for this release category are addressed in Section 4.5.5 on enveloping source terms. 4-41 1281P120685
Source term category S2 is dominated by contributions from plant damage g
' states 3FP and 7FP. These two plant damage states contain accident 5 sequences in which a containment penetration of less than 3 inches in diameter has failed to isolate and provides a direct release path to the envi ronment . Over 95% of the total frequency of source term category S2 comes from these plant damage states. The same plant damage states dominated this release category in the SSPSA; therefore, release category T2V-d from the SSPSA was used to represent release category S2B in this study. Release parameters for the type of containment failure represented by this source term category were not determined by either the IDC0R or NRC source term research programs because this failure mode is specific to the Seabrook design.
Source term category S3 is dominated by contributions from plant damage state 80 and from the plant damage state combination 30 and 70. All these plant states represent accident sequences in which the containment has successfully isolated, but all active containment heat removal systems and fission-product scrubbing systems fail. The same accident sequences and plant damage states dominated this source term category in the SSPSA in which the specific accident sequence used to model this release category was a station blackout sequence. The IDCOR program a explicitly analyzed station blackout sequences for the Zion station, and, g based on the design comparisons discussed in the previous section, the release fractions from the IDCOR analysis for Zion were adopted as the best estimate release fractions for release category S38. However, due to the differences in containment strength, the timing of the release was adopted from the SSPSA analysis for this accident sequence. The longer time to release for the Seabrook Station in this accident sequence would even further reduce the release fractions calculated by the MAAP code for Zion. Two changes were made to the IDCOR release fractions as follows: The release of Cs and I was reduced by a factor of 2 to account for the much longer time of release (89 hours for Seabrook versus 32 hours for Zion).to 2x10-}he based on therelease of Terecognition more recent was increased fromofthe that a portion theIDCOR value of tellurium release to the containment is now expected to occur after vessel melt-through if a core-concrete interaction takes place. If containment failure occurs during or shortly after the time of Te release 3 from the debris, the Te release fractions to the environment have been 3 estimated to range from 1% to 10% of the Te inventory. For the best estimate source term, a release fraction of 2% was adopted and this value was reduced by a factor of 10 for Seabrook to account for deposition during the much longer time difference between concrete attack and containment failure. According to Reference 4-21 these changes in the release fractions for Cs, I, and Te will not result in a noticeable 3 impact on accident consequences. 5 Source term category S5 represents accident sequences in which the g containment remains intact. This means the containment has been g successfully isolated and the containment heat removal system, (containment spray) operates successfully. Therefore, fission product scrubbing in the containment by the spray system is guaranteed for this l release category. The release fractions for this release category are so 1 4-42 1281P120685
low that the SSPSA release fractions and release timing from release category SS have been directly adopted to represent release category SSB for this study. Source term category S6 represents accident sequences that involve failure to isolate a containment penetration with a diameter greater than 3-inches, resulting in a direct release path to the environment. The containment atmospheric purge line is the only penetration that can be opened during normal operation and provide a direct release path. Therefore, a transient-initiated accident sequence with failure to isolate the containment atmospheric purge line was selected both in the SSPSA and by the IDCOR program to represent this source term. The dominant contributions to this release category come from plant damage states 3F and 7F. According to the SSPSA analysis, the radionuclide release for this source term occurs over a time period of approximately 16 hours. Therefore, in the SSPSA, this source term was analyzed as a multipuff release. The same release category was analyzed as a single puff release under the IDCOR program. For the best estimate relsase category S6B, the total radionuclide release fractions from the IDCOR analysis were adopted and combined with the Seabrook-specific release timing, determined for this accident sequence in the SSPSA. Furthermore, I the release was divided into three sequential puff releases in proportion to the releases determined in the original SSPSA analysis. Current uncertainties in the release characteristics of Te and Ru for this release category are addressed in Section 4.5.5 on enveloping source terms. Release category S7 represents V-sequence releases that are assigned to I the new plant d3 mage states IFPV and 7FPV. The best estimate release Category S78 was derived from a Seabrook-specific analysis of the RHR pump seal failure V-sequence, as discussed in Section 4.4 This release category represents the radionuclide releases associated with a release through a pool of subcooled water in the RHR vault. A scrubbed pool release was adopted for the best estimate representation of this release I category. An RHR pump seal V-sequence that would not flood the pump seals requires a specific and narrowly defined pump seal leak history, which is considered unlikely. This specific set of circumstances will be addressed in the next section where the conservative release category I estimates are discussed. The release fractions and the release timing for this release category are represented as a multipuff release, which is derived directly from the results of the MAAP code analysis discussed in Section 4.4 , 4.5.3 CONSERVATIVE ESTIMATE RELEASE CATEGORIES For each of the release categories, a conservative estimate source term was defined to explicitly account for the disparity of opinion about the numerical values of release fractions and accident sequence timing. The conservative estimate release categories are shown in Table 4-14 The IDCOR release categories adopted for this study were all treated as best estimate release categories. The conservative estimate release categories were all derived from the SSPSA results with the exception of release category S7C the V-sequence release category, for which a Seabrook-specific analysis was performed. In the SSPSA, an explicit 4-43 1281P120685
uncertainty analysis was performed for each significant release category. This uncertainty analysis resulted in the definition of four subcategories for each release category. Each subcategory was associated with a confidence level for nonexceedance of the release subcategory. Release subcategories a and b represented the 99% confidence level and I the 95% confidence level for nonexceedance. Release subcategory a was E derived directly from the MARCH and CORRAL calculations without any account of radionuclide deposition and removal processes not modeled in these codes. Based on the current assessment and understanding of accident progression and radionuclide behavior, accident source terms calculated directly by MARCH and CORRAL are no longer believed credible. As a result, the a and b subcategories of the SSPSA release categories are no longer considered to be meaningful statements for accident source terms. Release subcategories c and d represented the 75% confidence level and the 50% confidence level, respectively. These subcategories accounted in an approximate manner for the radionuclide transport and deposition processes not explicitly modeled in the MARCH and CORRAL codes. This was accomplished through a probabilistic analysis of release reduction factors based on an interpretation of the published evidence available at the time and supported by limited analyses. These analyses are documented in the SSPSA. Therefore, the conservative release category estimates for this study were derived, in general, from the c subcategory of the corresponding SSPSA release category. The source for each conservative estimate source term is indicated in the right-hand column of Table 4-14 For release category S3C, the Te released was increased by the same procedure that was used for S3B except that a base release fraction of 10% was used instead of 2%. Release category S6C is represented by the SSPSA release category S6V-d. TE the SSPSA, only a release subcategories a and d applied directly to open purge sequences, g where a is the release directly calculated by MARCH / CORRAL. Subcategories b and c applied to V-sequences that are not represented
. separately by release category S7. Furthermore, the IDCOR source term for open purge sequences is substantially lower than release subcategory d. The release fractions for Te and Ru for release category S6C were increased by a factor of 2.5 to account for current uncertainties in these releases. This increase is judgmental, and the issue is addressed again in Section 4.5.5 on enveloping source terms.
Release category S7C was derived directly from the Seabrook-specific MAAP code analysis. A conservative estimate of this release category was determined by modeling the radionuclide release path from the RHR pump seals to the environment without a flooded vault condition. Due to the E small volume of the RHR vaults, substantially higher release fractions E are determined by the combined analysis with the MAAP code and the V-code, compared to the best estimate analysis. The pnysical conditions a under which a V-sequence could occur and not develop a flooded RHR pump 5 seal condition, requires the RHR pump seal leak rate be below the combined capacity of the RHR vault sump pumps. Therefore, the leak rate at each RHR pump seal would be required to be less than 50 gpm or a long-term primary system loss rate of less than 100 gpm total. Following reclosure of the RHR system relief valves inside the containment, an RCS makeup rate of 100 gpm is well within the capacity of the charging pumps. Continuous operation of the charging pump and continuous makeup 4-44 1281P120685
for the charging pump supply could maintain the core in a long-term stable condition. At an RCS loss rate of 100 gpm, the remaining RWST inventory would last for approximately 1 day. Therefore, an unflooded RHR pump seal condition can only exist if the small pump seal leak rate I condition is combined with additional failures; namely, failure of the charging pumps and their normal makeup supply or failure of the charging pumps combined with failure to establish RWST makeup at a rate of 100 gpm in a time period of 1 day. An additional conservatism in the analysis of the S7C accident source term arises from the fact that the analysis was performed assuming an RHR pump seal leak area of 2.6 square inches. This corresponds to the large pump seal leak area that would guarantee I flooding of the RHR pump seals. The leak area corresponding to a 100-gpm leak rate would be less than 10% of the assumed leak area. Since a substantial portion of the release occurs following vessel melt-through in the radionuclide revaporization phase, this large leak rate assumption is expected to overestimate the radionuclide release to the environment. 4.5.4 RELEASE CATEGORY UNCERTAINTIES AND COMPARIS0N OF RELEASE -FRACTIONS The IDCOR published source terms and the NUREG-0956 source terms are the only sources available for comparison that take into account the current I level of knowledge. The IDCOR source terms were available at the time when the release categories for this study were defined; however, the NUREG-0956 source terms were published substantially later. Table 4-15 I compares the release categories defined in this study with the IDCOR and NUREG-0956 release categories where a meaningful cnmparison exists. Included also, for reference, are the corresponcing WASH-1400 release categories where appropriate. When IDCOR source terms were available, they closely correspond to the best estimate release categories defined in this study as they are derived from the IDCOR source terms. Since a careful review of NUREG-0956 was not performed in this study, no definitive statements are made to explain the differences between the source terms used in this study and the NUREG-0956 source terms. However, the following factors are expected to contribute to the differences:
- 1. The accident analyses in NUREG-0956 were performed with the intent to exercise the BMI-2104 series of computer coaes for the purpose of demonstrating that they constitute a usable methodology for developing new source terms. NUREG-0956 states that the calculated source terms are only a demonstration of that methodology and that new source terms will be calculated in other studies leading to publication of a future report to be identified as NUREG-1150.
- 2. Recent experiments on the behavior of tellurium indi: ate that it may be chemically bound to unreacted zirconium in the core and can be released ex-vessel when the remaining zirconium oxidizes during concrete penetration if the debris is not cooled. This can increase the tellurium source tern in the containment atmosphere at a time in the accident sequence that is closer to the time of containment failure. The tellurium release is then sensitive to the time between vessel failure and containment failure, resulting in an increased tellurium release if the time interval is short.
4-45 1281P120685 .
- 3. The significantly higher pressure capacity of the Seabrook containment compared to the Surry containment would result in a significantly longer time interval between vessel breach and containment failure; therefore, lower releases for tellurium and other materials would be anticipated for Seabrook.
- 4. Release categories for accident sequences with either late overpressure failure of the containment or with the containment failure were not reported in NUREG-0956. Analyses reported in BMI-2104 for these cases showed that the release fractions for all radionuclides except noble gases would be very low and consistent with the results from this analysis.
- 5. Differences in the decontamination factor for water pool scrubbing for the V-sequence in suspected to be due to two major reasons.
First, in the Seabrook case the pool is 30 feet deep and subcooled and secondly, the SPARC code used by the NRC tends to yield lower decontamination f6ctors than the SUPRA code used by EPRI.
- 6. For the V-sequence without pool scrubbing, it is suspected that the analyses reported in NUREG-0956 did not account for deposition in the auxiliary building, whereas this was modeled in the Seabrook analysis.
Reference 4-21 has examined the effect on accident consequences of releasing varying quantities of the other radionuclides. It is shown that the early health effect consequences are not sensitive to the
- release fractions of Cs, I, and Te below about 10% and to release fractions of the remaining radionuclides below about 1%. Below these release levels, the noble gases contribute a significant fraction of the early exposure.
4.5.5 ENVELOPING SC'JRCE TERMS In order to assess the potential impact of the differences between the source terms used in this study and those published in NUREG-0956, a set of enveloping source terms was defined to be used in a risk sensitivity analysis. Enveloping source terms could be defined for release category S1 and 56 since corresponding NUREG-0956 source terms for Surry or Zion ~only exist for these two source terms, as indicated in Table 4-15. The enveloping source terms will be identified as S1E and S6E. They are derived by selecting the worst value of each listed release category parameter from Table 4-15 from all sources other than WASH-1400. Table 4-16 lists the enveloping source terms in the form of multipuff releases as they were used in the sensitivity analyses, which is discussed in Section 2.3. As can be seen by comparison to Table 4-15, approximately half of the enveloping source term parameters are derived from NUREG-0956, with the other half from the conservative source terms used in this study. It is also noted that for those release fractions for which the NUREG-0956 values are higher, the difference is typically a factor of 3, which is not expected te have a very large effect on consequences since only a portion of the release groups are affected. I 4-46 1281P120685
I 4.6 ACCIDENT SE0VENCE MAPPING INTO RELEASE CATEGORIES 4.6.1 GUIDELINES FOR ACCIDENT SEQUENCE MAPPING In the SSPSA, the mapping of accident sequences into release categories , was accomplished by the C-matrix. An entry in the C-matrix gives the i conditional probability that a given plant damage state leads to the corresponding release category. The conditional split fractions in the C-matrix are derived from a containment event tree analysis that traces the outcome of physical phenomena during the accident progression. The assessment of these physical phenomena may differ for different plant I damage states. Furthermore, for a given plant damage state, uncertainty in predicting the outcome of a given physical phemonema may require that alternate outcomes be acknowledged. However, the intent pursued in defining the plant damage states and the release categories is to have a one-to-one correlation between plant damage states and release categories if the accident progression could be predicted without any uncertainties. It is therefore not surprising there is a release category with a high conditional probability of occurring for each plant damage state. Contributions to other release categories reflect the possibility that the accident sequence progression may differ from that predicted by currently available knowledge and analysis tools. A new containment response analysis was not performed for this study. Rather, the containment response analysis from the SSPSA was adopted for the new release categories defined in Section 4.5. Table 3-1 in Section 3 lists the mean frequencies for all important plant damage states according to the current analysis. All other plant damage states need not be considered because their frequencies are much lower than the frequency of other plant damage states with similar consequences. A new C-matrix was developed to correspond to the new release term categories. 4.6.2 C-MATRIX FOR CURRENT SOURCE TERM CATEGORIES The new C-matrix is shown in Table 4-17. This new C-matrix was derived from the SSPSA C-matrix by reassigning C-matrix split fractions in the SSPSA to the current plant damage states. Documentation of the split fraction numerical values is provided in the SSPSA. The mean frequency of each plant damage state is indicated in parentheses. By multiplying the mean frequency of each plant damage state by the conditional C-matrix split fractions, it is possible to determine which plant damage states dominate each of the seven release categories. The result of this multiplication is shown in parantheses under each of the split-fraction entries in Table 4-17. The total mean frequency of all plant damage states and of each release category is shown at the bottom of the table. It is evident that release category S1, which represents early
- containment failure accident sequences, is dominated by plant damage state 1FV. This represents the residual possibility of an RHR piping failure mode for a V-sequence. Release category S2 represents early increased containment leakage sequences. It is dominated by plant damage state 3FP and 7FP. Release category S3 represents accident sequences with late overpressure containment failure. This release category is dor,inated by plant damage states 8D, 3D, and 70. As expected, all of these plant damage states represent "D" conditions where no containment 4-47 I 1281P120685
heat removal is available. Release category SS represents accident E sequences where the containment remains intact. This source term is 3 dominated by plant damage states 4A and 8A. The "A" plant damage states are those where containment heat removal is available. Release category S6 represents releases associated with the failure to isolate a large containment penetration. Plant damage states 3F and 7F dominate this release category. The new release category, S7, represents releases associated with the V-sequences. It is dominated by the new plant damage states IFPV and 7FPV. These were explicitly defined for the RHR pump seal V-sequences. 4.7 TREATMENT OF SOURCE TERM AND SITE MODEL UNCERTAINTIES The approach to the quantification of source term and site model assumption uncertainties in this study is to define two discrete sets of assumptions for each source, as illustrated in Figure 4-27. For each of these two sources of uncertainty, a "best estimate" and " conservative" set of assumptions were defined. With the exception of the new source term category S7 (defined for the RHR pumps seal LOCA type "V" sequence) for which the weights of .80/.20 were used for the best estimate / conservative assumption set, the weighting factors for source terms were generally set at .90/.10. The weights are based on an examination of the specific sequences assigned to each source term category. In the case of site model assumption uncertainties, weights of .80/.20 were generally assigned to the best estimate / conservative assumption set. Details of the site model assumption sets are discussed in Section 5. A full implementation of the four discrete combinations of source term and site model would require four site model evaluations (i.e., four executions of the CRACIT computer program) for each release category and each emergency plan protective action case. All four CRACIT runs were evaluated for selected release categories for the no-evacuation case. These results showed that there was very little difference between the BH and CM interior cases. For the remaining release categories for the no-evacuation case, the four discrete cases were represented by three CRACIT runs, one for each of the BM, CM for BH, and CH cases. For the evacuation and sheltering cases, some release categories were represented by only the two extreme assumption sets, BM and CH. In all cases where fewer than four CRACIT runs were made, the accident sequences and probability weights were conserved and combined conservatively. These practical considerations were used to keep the number of CRACIT runs to a reasonable level without disturbing the numerical results.
4.8 REFERENCES
4-1. Pickard, Lowe and Garrick, Inc., "Seabrook Station Probabilistic Safety Assessment," prepared for Public Service Company of New Hampshire and Yankee Atomic Electric Company, PLG-0300, December 1983. 4-2. Wooten, R. 0., and H. I. Avci, " MARCH Code Description and Users Manual," Battelle Columbus Laboratory, NUREG/CR-1711, (BMI-2064), October 1980. I 4-48 1281P120685
E 4-3. Burian, R. I., and P. Cybulskis, " CORRAL-II Users Manual," Battelle Columbus Laboratory, January 1977. 4-4 U.S. Nuclear Regulatory Commission, " Reactor Safety Study: An Assessment of Accident Risks in U.S. Commercial Nuclear Power Plants," WASH-1400, NUREG-75/014, October 1975. { 4-5. Technology for Energy Corporation, " Nuclear Power Plant Response r to Serve Accidents," IDCOR Technical Summary Report, L November 1984. 4-6. "MAAP-Modular Accident Analysis Program Users Manual," Technical [ Report on IDCOR Tasks 16.2 and 16.3, May 1983. 4-7. U.S. Nuclear Regulatory Commission, " Technical Bases for E Estimating Fission Product Behavior during LWR Accidents," L NUREG-0772, June 1981. . 4-8. Gieseke, J. A., et al., "Radionuclide Release under Specific LWR Accident Conditions," Vol. VI, Batelle Columbus Laboratories, BMI-2104, July 1984. { 4-9. U.S. Nuclear Regulatory Commission, " Reassessment of the Technical Bases for Estimating Source Terms," NUREG-0956, draft report, July 1985. 4-10. Bunz, H. , M. Kayro, and W. Schoch, "NAUA-Mod 4: A Code for Calculating Aerosol Behavior in LWR Core Me',t Accidents," KfK-3554, August 1983. 4-11. " Report to the American Physical Society of the Study Group on Radioactive Release from Severe Accidents at Nuclear Power Plants," draft report, February 1985. { 4-12 Fauske and Associates, Inc., " Technical Support for Issue r Resolution," FAI/85-27, submitted to IDCOR for publication as an L IDCOR Technical Report. 4-13 Tarbell, W., J. Brockman, and M. Pilch, "High Pressure Melt Streaming (HIPS) Program Plan," NUREG/CR-3025, August 1984. 4-14 Spencer, B. W., et al., " Overview and Recent Results of ANL/EPRI Corium-Water Thermal Interaction Investigations," Paper TS-15.2, proceedings of the International Meeting on LWR Severe Accident Evaluation, Vol. 2, Cambridge, Massachusetts, August 28 to September 1, 1983. 4-15 Spencer, B. W., D. Kolsdonk, and J. J. Sienicki, "Corium/ Water Dispersal Phenomena in Ex-Vessel Cavity Interactions," Paper [ TS-15.5, proceedings of the International Meeting on LWR Severe Accident Evaluation, Vol.2, Cambridge, Massachusetts, August 28 to September 1, 1983. ~ 4-49 1281P120685
I 4-16 Manjoine, M. H., " Ductility Indices at Elevated Temperature, J. Engineering Materials and Technology, ASME, Vol. 97, Series H, No. 2, pp.156-161,1975. 4-17. " Evaluations of Containment Bypass and Failure to Isolate Sequences for the IDCOR Reference Plants," IDCOR Task 23.5 Report, July 1984. 4-18. " Identification of Fission Product Release Pathways," Technical Report on IDCOR Task 11.2. 4-19 " Zion Nuclear Generating Station Integrated Containment Analysis," IDCOR Technical Report 23.1Z, draft, March 1985. 4-20 Fauske and Associates, Inc., " Evaluations of Containment Bypass and Failure to Isolate Sequences for the IOCOR Reference Plants," FAI/84-9, July 1984. 4-21 Kaiser, G.D., "The Implications of Reduced Source Terms for Ex-Plant Consequence Modeling," paper presented at the American Nuclear Society Executive Conference on The Ramifications of the Source Term, Charleston, South Carolina, March 12, 1985. I I I I 4-50 1281P120685
l TABLE 4-1. ACCIDENT SOURCE TERMS AND CONSEQUENCES CALCULATED BY THE IDCOR PROGRAM FOR THE ZION STATION Sequences Leading to Environmental Releases Station Blackout with Blackout with Seal Interfacing I Results B!ackout (14 TE) Seal LOCA (2 SE) LOCA andimpaired Ccatainment System LOCA (16-V) I Prcbability per reactor year
- 2(- 7) 6(- 6) 3(-8) -
1(- 7) Top of core uncovered, hr 2.3 2.2 2.2 20.0 Start of fuel melting, hr 3.1 3.0 3.0 ~ 23.0 Vessel breach, hr 4.0 3.8 3.8 26.0 I Containment overpressure failure, hr Time of fission product release, hr 32.0 32.0 32.0 32.0 0.0 4.0 24.0 Fission Product Release Fractions
- Xe-Kr 1(0) 1(0) 1(0) 1(0) 1 Br 2(-3) 2(-3) 1(- 2) B(- 5)
Cs-Rb 2(-3) 2(-3) 1(- 2) 8(- 5) Te-Sb 2(- 5) 2(-5) 3(-4) SrBa 8(-5)
< 1(- 5) < 1(- 5) 6(-4) 5(- 5)
Ru-Mo < 1(- 5) < 1(- 5) 6(- 5) < 1(- 5) I Offsite Consequences Early fatalities Earlyinjuries 0 0 0 0 0 0 0 0 Laieni cancer fatiiity inoex b I Offsite costs, $106 Whole body man rem 1(-4) 7(1) 8(5) 1(-4) 7(1) 8(5) 7(-4) 1(2) 3(6) 2(- 5) 6(1) 9(4)
- Numbers in parentheses are exponents of 10.
- Latent cancer fatahty index is fraction increase over normalincidence within a 50-mile radius of the plant over 30 years, in the event that the accident occurs.
Source: Reference 4-5 (Table 10-3, Summary of Zion Results). 4-51
I TABLE 4-2. NRC/IDCOR TECHNICAL ISSUES FOR SEVERE ACCIDENTS r Description I 1 Fission Product Release prior to Vessel Failure 2 Recirculation of Coolant in the Reactor Vessel 3 Release Model for Control Rod Materials 4 Model for Fission Product and Aerosol Deposition in the Primary System 5 Modeling of In-Vessel H2 Generation - 6 Core Slump, Core Collapse, and Reactor Vessel Failure Models 7 Alpha Mode Containment Failure by In-Yessel Steam Explosions 8 Direct Heating of Containment by Ejected Core Material 9 Ex-Vessel Heat Transfer Models from Molten Core to Concrete / Containment 10 Ex-Vessel Fission-Product Release Modeling 11 Revaporization of Fission Products in the Upper Plenum 12 Deposition Model for Fission Products in Containment 13a Amount and Timing of Suppression Pool Bypass 13b Fission Product Removal in Ice Condensers 14 Modeling of Emergency Response 15 Containment Performance 16 Secondary Containment Per,formance 17 Hydrogen Ignition and Burning 18 Essential Equipment Performance l I 1300P083085 4-52
TABLE 4-2a. SUttt1ARY--TECHNICAL SUPPORT FOR ISSUE RESOLUTION
- Sheet 1 of 5 Agreed Upon Path to Resolution IDCOR/85 Actions Taken Result of IDCOR/85 Studies Issue Incorporate release models e NUREG-0772 models incorpo- No iubstantive change from l 1 - Fission Product Release rated into MAAP along with previous IDCOR analyses.
Prior to Vessel failure from NUREG-0772 and consider Te released in-vessel or EPRI steam oxidation model, ex-vessel. e User option Tc released in-vessel or ex-vessel - recom-mended uncertainty analysis. Incorporate in-vessel natural e Incorporated in-vessel natu- No significant change in 2 - In-Vessel Natural the hydrogen generation Circulation circulation model into MAAP ral circulation model into HAAP-PWR. or the upper plenum and benchmark with TMI-2. temperature. e Benchmarked against THI-2. e Benchmarked against EPRI-W, experiments. Assume that Ag-In-Cd control Consistent with modeling as- No change from previous i 3 - Release Model for Control sumptions in MAAP. IDCOR analyses,
$ Rod Materials rod material melts and relo-cates away from high tempera-ture regions.
Extensive numerical experi- Correlations are developed 4 - Fission Product and Aero- Perform extensive numerical from fundamental princi-sol Deposition in the experiments with a sectional- ments have provided more ccm-ized aerosol code to validate prehensive aerosol deposition pies and more deposition Primary System mechanisms are included, and/or update aerosol deposi- correlations for the MAAP codes. Preliminary NAAP results tion correlations in MAAP. show no significant dif-ferences from previous IDCOR analyses. e Core debris levitation model Hydrogen generation rates 5 - In-Vessel Hydrogen IDCOR would benchmark their and magnitude are essen-Production calculations against THI-2 added to MAAP. tially the same as previ-behavior as well as the SFD ous IDCOR studies. tests - NRC would investigate o MAAP models benchmarked the possibility of carrying against THI-2 observations. out the same benchmarking calculations, e NAAP heatup models banch-marked against SFD tests.
*Taken from Reference 4-12 (Table 19.1) l
TABLE 4-2a (continued) Sheet 2 of 5 Issue Agreed Upon Path to Resolution IDCOR/85 Actions Taken Result of IDCOR/85 Studies 6 - Core Slump, Collapse and A core melt progression model Core melt progression models Vessel Failure should be incorporated into Core debris temperatures incorporated into the MAAP-0WR are somewhat lower and the MAAP. and MAAP-PWR codes. molten material mass at vessel failure is less variant than previous IDCOR studies. 7 - In-Vessel Steam Explosions NRC sponsored Steam Explosion IDCOR agrees with the conclu-and the Alpha Failure Mode Group concludes that the alpha sfon of the Steam Explosion failure mode is highly Review Group. unlikely. 8 - Direct Containment Heating IDCOR would survey the various e Reactor cavity and instru- Only a few percent of core cavity configurations for PWR ment tunnel configurations a plants and establish a crt- material would be avail-have been surveyed for the able to directly heat the A terion for estimating the various PWR designs. containment atmosphere. fraction material that could No significant change in participate in directly, heat- e A criterion has been devel- containment loading. in the atmosphere. oped to estimate the mass of material that could directly heat the contain-ment atmosphere. e The criterion has been ap-plied to the spectrum of PWR reactor cavity designs. 9 - Ex-Vessel Heat Transfer IDCOR would benchmark test re- Models for ex-vessel heat Models No substantive change in suits for core-concrete inter- transfer have been benchmarked. IDCOR nodels, action for the Sandia test against detailed calculations program and the Beta test as and experiments performed to results become available, da te. As results become available this comparison base will be expanded. g g g g m m M M M M M M M M
mR n FN F7 FR F7 FR F7_ ra ro rm r r r TABLE 4-2a (continued) Sheet 3 of 5 Issue Agreed Upon Path to Resolution IDCOR/85 Actions Taken Result of IDCOR/85 Studies I 10 - Ex-Vessel Fission Product IDCOR will increase the number e IDCOR has developed a chemi- More non-volatile fission Release of chemical species tracked cal thermodynamic equillb- products tracked. The re-during ex-vessel core-concrete rium model for tracking more leases from the debris are attack, fission product species in dependent upon debris flow the core debris and their from the vessel and the ul-release due to stripping, timate debris disposition. e A stand-alone version is be-ing exercised. e The model will be incorpo-rated into MAAP by 7/31/85. l i 11 - Revaporization of Fission e Revaporization models should e MAAF models have been bench- IDCOR models in general l ? Products in the Upper be benchmarked against marked against the ANL agreement with the data. g Plenum available experiments, experiments. Survey of primary system heat loss data shows a e Uncertainty calculation e Uncertainty calculations greater potential for re-should be carried out with have been carried out with taining volatile fission l lower vapor pressures for lower vapor pressures for products. l the deposited fission the deposited fission prod- ! products, ucts and these demonstrated lower releases to the , environment. ' e Survey of available plant data show the primary heat losses to be more extensive than previously considered thereby substantially re-ducing the environmental releases. , 12 - Fission Product and Aero- See Issue No. 4 See Issue No. 4. See Issue No. 4. sol Deposition in the Containment
~
TABLE 4-2a (continued) Sheet 4 of 5 Issue Agreed Upon Path to Resolution IDCOR/85 Actions Taken Result of IDCOR/85 Studies 13a - Amount and Timing of e Incorporate an aerosol plug- e An aerosol plugging model Essentially eliminates any l Suppression Pool Bypass ging model into the HAAP has been incorporated into sensitivity to normal dry-l codes. MAAP. well/wetwell leakage for Mark III designs. Reduces l e Carry out sequence evalua- e Grand Gulf sequences run all releases except noble tions using the plugging with the plugging model show gases to negligible levels model and for a sequence as- environmental releases to be for Mark I sequences with suming a stuck-open vacuum dominated by noble gases. wetwell venting. Substan-breaker with the plugging tial reduction in Mark I l model overridden, o Release fractions for a se- source terms. quence with an assumed stuck-open vacuum breaker and the plugging model over-ridden result in 1% of the volatile species released.
?
m 13b - Fission Product Removal Future calculations will spe-cify how much the aerosol is Future analyses will make this No substantive change in
- in Ice Condensers distinction. IDCOR models.
deposited as a result of steam condensation and how much is due to other deposition processes. 14 - Modeling Emergency Issue resolved. IDCOR will continue to assume No change. Response that 5% of the population would not participate in an evacuation nor an early relocation. 15 - Containment Perfonnance e IDCOR will include a con- e Strain induced model for IDCOR will continue to tainment strain methodology containment failure modes consider uncertainties in for evaluating likely fail- will be incorporated into containment failure sizes ure modes, the MAAP codes through EPRI as performed in previous sponsored work. IDCOR analyses. e IDCOR will consider uncer- e Uncertainties will be evalu-tainties in the failure mode ated for various failure for source term evaluations. modes. s An IDCOR/NRR Sub-Group will e A small IDCOR/NRR Sub-Group be established to evaluate will be established including proposed containment failure EPRI and Owners' Group methodology, pa rticipation. M M M M M M M M M M M M M M M M
M M M M M M M M M M M M M M M M M M M TABLE 4-2a (continued) Sheet 5 of 5 Issue Agreed Upon Path to Resolution IDCOR/85 Actions Taken Result of IDCOR/85 Studies 16 - Secondary Containment Careful consideration should IDCOR analyses will continue to No change in the IDCOR Performance be given to the detailed geo- carefully consider the building analyses, metry of the reactor building, geometry, thermodynamic condi-ventilation systems and tions, ventilation system per-thermal-hydraulic conditions, fonnance, etc. In future analy-ses for individual plant analyses. 17 - Hydrogen Ignition and e IDCOR and NRC Fould define a e The thermodynamics of ny- H recombination in con-Burning standard problem to compare drogen recombination were t$1nments will occur and models for hydrogen assessed and fcund to be is very influential in the
, combustion. very rapid for typical con- contairsent response. Ad-i ditions in severe accidents. ditional benchmarking with $ e NRC will assess the effects of IDCOR compartmentaliza-This supports the modeling large scale H, combustion approach used in the MAAP tests with ighiters show tion. codes, the IDCOR uncertainty ranges to be conserva-e IDCOR hydrogen combustion tive. The effective drag models have been benchmarked coefficient in the HAAP against a wider data base model should be increased including the VGES and to 1-100 with a default Nevada test data, value of 100. This means that igniters are more ef-e IDCOR and NRC will define a ficient than credited in standard problem for compar- the previous IDCOR Con-ing hydrogen combustion tainment Analyses, models.
_ m m_ _ _ _ _ _ _ _ _ _ _
I TABLE 4-3 CONTAINMENT FAILURE MODES AND TYPE Median . Lognormal Median Failure Failure Failure Standard Mode Pressure k*p Type (a) Deviation E (psia) S E Structural Failure Modes Cylinder Wall Hoop 231 Large(b) C .12 Dome Hoop or Meridional 238 Large(b) C
.12 l
Wall Meridional 296 Large(b) C .12 l Base Slab Shear 338 Large(b) C .23 Base Slab Flexure 415 Large(b) C .25 Wall Shear at Base 423 Large(b) C .30 Local Failure Modes l Feedwater Penetration 194 Self-Regulating (c) B 0.5 Flue Head Feedwater Pipe Crushing 231 Sel f-Regulating (c) B .12 Fuel Transfer Tube > 260(d) Sel f-Regulating (c) B (e) Bellows Penetrations X-25, X-26, 181 0.5 Square Inch A 0.16 X-27 Each . l All Others(f) -(d ) Sel f-Regul ating(c) B (e) l l
- a. Containment failure types A, B, and C are defined in Section 4.2.5. gI
- b. Much larger than 0.5 square foot. El
- c. Leak area is self-adjusting to stop pressure rise. l
- d. Probability of failure is less than 50% at ultimate wall hoop capacity.
- e. Failure pressure model not lognormal.
- f. Composite estimate of liner adhesion, microcracks, weld faults, equipment hatch, l other mechanical penetrations, and electrical penetrations.
I! l l l
~ ~
1300P100285
E L [ TABLE 4-4. SOURCE TERM CATEGORIES E L Source Identified Analyzed Term Containment Failure Mode in the in This [ Category SSPSA _ Study S1 Early Containment Failure Yes Yes S2 Early Increased Containment Leakage Yes Yes [ S3 S4 Late Overpressure Failure Basemat Melt-through Yes Yes Yes No* p S5 Containment Intact Yes Yes S6 Containment Not Isolated Yes Yes r S7 Containment Bypassed (V-sequence) No Yes L
- Based on the SSPSA results, basemat melt-through sequences were assigned
[ to category S3 in this study. W [ [ W
TABLE 4-5. CONTAINMENT DESIGN COMPARISON TABLE FOR SEABROOK STATION AND Tile ZION STATION Sheet 1 of 14 Seabrook Station Zion Station Model Source Value ) 1. Containment Spray System I
- a. Spray pump flow (indicate data for: injection /
ricirculation). (1) Total number of pumps available. FT 6.2-2 2 2 Motor-Driven; 1 Diesel-Driven (2) Mimimum flow per spray pump (gpm). FT 6.2-2 3 (3) Maximum flow per spray pump (gpm). FT 6.2-2 3,,000 300 2,614 (4) Number of pumps available with station FS 6.2.2.2 None None blackout (LOSP, no diesels),
- b. Containment pressure setpoint for spray pump FS 6.2.2.1 20.4 23.0 actuation (psig).
4 c. Maximum post-accident delay time for effective FS 6.2.2.1 62 30 a o spray flow to enter containment after containment pressure setpoint in item lb is reached (seconds),
- d. As item Ic but with station blackout FS 6.2.2.2 N/A N/A (no offsite power, nc diesels) (seconds).
- e. Spray pump cooling requirements (flow (gpm) SD No. 23 26/PCC Self Cooled per pumpA.oaH n system).
- f. Limits of oper aility under adverse FT 6.2-75 300*F/300 psig 400'F/200 psia environment conditions; i.e., pressure, Design Design temperatures, etc.
- 2. Fan Cooler Availability
- a. Do the fan coolers perform a safety related FS 9.4.5.3 No Yes function? (yes/no)
- b. Are they shut down and isolated by the YAEC Yes No containment isolation or safeguards actuation system? (yes/no)
NOTE: FT = FSAR table; FS = FSAR section; SD = system description FF = FSAR figure; SBU = intercompany memorandum from UE&C to Seabrook; N/C = not calculated; N/A = not available. 1303P092785 m m M M M M M M M M M M M M m m
M MMU M R R R R R TABLE 4-5 (continued) Sheet 2 of 14 Seabrook Station 3 Source Value
- c. How many fan coolers can be powered from the YAEC 2--Diesel A 2--Diesel A diesels? 2--Diesel B 5 2--Diesel B l--Diesel C
- d. How may fan coolers can be restarted, YAEC None All including cooling water with the containment isolated?
- e. Is restart under item d manual or automatic, YAEC No Procedure Manual l and are procedures available? for Restart Yes
- 3. Fan Cooler Perfonnance (answer only if they YAEC N/A l are operable witn containment isolated) I
- a. Minimum / maximum number available per diesel N/A 1/2--Diesel A train (train A, train B). 1/2--Diesel B l O/1--Diesel C '
i b. Containment pressure setpoint or other N/A SI Signal $ starting signal for fan cooler initiation.
- c. Maximum post-accident delay time for effective N/A 30 heat removal by fan coolers after starting signal (item 3-b) is reached (seconds).
- 4. Emergency Core Cooling 4.1 Injection Mode
- a. (1) Number of charging pumps. FT 6.2-2 2 2 (2) Charging pump design flow rate FT 6.2-2 150 at 2,518 150 (gpm at psi).
(3) Charging pump cooling requirements SD-23 64/PCC 300/CCS (gpm/ system),
- b. (1) Number of HPSI pumps. FT 6.2-2 2 2 (2) HPSI pump design flow rate (gpm at psi). FT 6.2-2 440 at 1,160 400 (3) HPSI pump cooling requirements SD No. 23 10/PCC 30/CCS (gpm/ system).
NOTE: FT = FSAR table; FS = FSAR section; SD = system description; FF = FSAR figure; SBU = intercompany memorandum from UE&C to Seabrook; N/C = not calculated; N/A = not available. I i I 1303P092785
TABLE 4-5 (continued) Sheet 3 of 14 Seabrook Station Zion Station Source Value
- c. (1) Number of LPSI pumps. FT 6.2-2 2 2 (2) LPSI pump design flow rate (gpm at psi). FT 6.2-2 3,000 at 163 3,000 (3) LPSI pump cooling requirements SD No. 23 5/PCC 60/CCS (gpm/ system).
- d. (1 ) RWST volume available for injection (gallons). FT 6 3-6 376,070 350,000 (2) Fraction of f(l) remaining at switch to RAI 440.28* 0.069 0.23 spray recirculation.
(3) Fraction of f(l) remaining at switch to RAI 440.28* 0.069 0.23 ECC recirculation.
- e. (1) Number of accumulators. FT 6.3-1 4 4 (2) Water volume per accumulator (f 3t ). FT 6.3-1 850** 850***
(3) Total volume per accumulator (ft ). FT 6.3-1 1,350 1,350 (4) Setpoint for accumulator injection (psia). FT 6.3-1 615 Minimum 615 Minimum
- f. (1 ) Safety injection actuation signals S Signal S Signal 3, (signals,setpoints).
Q (2) Recirculation switchover signals RAI 440.28* RWST Low-Low 1 RWST Low Level (signals, setpoints). at 117.5 Inches
- g. (1 ) Cold leg injection (normal. FS 6.3.2.1 Normal Normal alternate,no).
(2) Hot leg injection (normal, Al ternate Al ternate alternate,no). 4.2 Recirculation Mode List any differences from data in 4.1 for injection None None mode, use same designator [(i.e., 4.2c(2)] for difference to item 4.lc(2).
- a. Minimum sump volume to avoid pump Calculation 19,156 9,358 cavitation (gallons). 4.3.22F6
*NRC request for additfogal information, FSAR, Vol.15. **Does not include 40 f t3fn discharge piping. ***Does not include 53 ft in discharge piping.
NOTE: FT = FSAR table; FS = FSAR section; SD = system description; FF = FSAR figure; SBU = intercompany remorandum from UE8C to Seabrook; N/C = not calculated; N/A = not available. 1303P092785 m M M M M M M M M M M M M M M
7 __n n n T- L__ f 1 T-~1 n n n _n Ff n n n _n n_ _n v TABLE 4-5 (continued) Sheet 4 of 14 Seabrook Station Zion Station Source Value
- b. (1) Number of RHR heat exchangers. FT 5.4-8 2 2 (2) Primary side flow rate (gpm). FT 5.4-8 3,000 3,750 (3 Secondary side flow rate (gpe). FT 5.4-8 5 000 5 000 (4)) Primary side inlet temperature (*F). FT 5.4-8 IE5 157.5 (5 FT 5.4-8 85 107.1 (6) Secondary
) Primary sideside inlet outlet temperature temperature (T).(T). FT 5.4-8 102.7 122.3 (7) Rated capacity (Btu / hour). FT 5.4-8 35.1x10 6 28.0 x 10 6
- c. (1) Number of spray heat exchangers. FT 6.2-2 2 (2) Primary side flow rate (gpe). FT 6.2-2 3,010 (3) Secondary side flow rate ( FT 6.2-2 4 800 Uses RHR Heat (4) Primary side inlet temperabpm).
re ( Y). FT 6.2-76 245 Exchangers for (5) Secondary side inlet temperature (*F). FT 6.2-76 120 Spray (6) Primary side outlet temperature (T). FT 6.2-76 181 6 Rechulation (7) Rated capacity (8tu/ hour). FT 6.2-76 96.7x10 )
, d. List and describe any other pumps that can
serve as ECCS pumps af ter containment isolation. - None None 4
- e. List and describe any other heat exchangers that could serve as ECCS heat exchangers. None None
- f. List iloits of operation under adverse environment conditions for active ECCS equipment; i.e., pressure, temperature, etc.
FF 6.3-5 3 600 3 427 (1) SISIpump (2) pump shutoff head (feet)(.T). design temperature FT 6.3-1 300 350 (3) Recirculation pump shutoff head (feet). N/A N/A (4) Recirculation pump design temperature ( T). N/A N/A FF 6.3-3 430 392 (5) RHRpump (6) RHR pumpdesign shutoff head (feet)lT). temperature FT 5.4-8 400 400
- 5. Containment Isolation System
- a. Actuation signals for isolation A. FS 6.2.1,4 5 psig SI Signal or
. High Containment Pressure (4.5 psig)
- b. Actuation signals for isolation B. FS 6.2.1,4 18 psig 23 psig NOTE: FT = FSAR table; FS = FSAR section; SD = system description; FF = FSAR figure; SBU = intercompany memorandum from UE8C to Seabrook; N/C = not calculated; N/A = not available.
1303P092785
TABLE 4-5 (continued) Sheet 5 of 14 Seabrook Station Zion Station Source Value
- c. Lines isolated for isolation A. FS 3.6-1 See Source All Except ESF FS 3.6.2-83 and Phase B
- d. Lines isolated for isolation B. FS 3.6-1, See Source RCP Cooling Lines FS 3.6.2-83
- 6. Secondary Coolant Auxiliary feedwater system design flow rate FT 6.8-1 One Electric Two Electric for each pump (gpm). Motor- Motor-Driven:710 Driven: 495 Each One Steam One Steam Turbine- Turbine-Driven:710 Driven:900
- 7. Additional Containment Heat Removal Features List and discuss any other systems available hone None
$ for containment heat removal, other than those discussed under items 1, 2, and 10, which would be available (automatic or manual start) following containment isolation.
- 8. Containment Configuration and Dimensions
- a. Nominal containgnt net free volumes (f t ).
(1) Upper Compartment (2) Lower Compartment (1 ) (1) 2.1 x 10 (1) 2.076xig 6 (2) (2) 2.5 x 10 (2) 4.27 x 10 5 (3) Annular Compartment (3) (3) 3.0 x 10j (3) 3.43 x 10 3 (4) Reactor Cavity (4) (4) 1.7 x 10 6 (4) 7.94 x 10 6 (5) Total (5) FT 6.2-1 (5) 2.704 x 10 (5) 2.86 x 10
- b. Height from ground level to spring SBU-24859 94 141 line (feet).
NOTE: FT = UE&C from FSAR table; FS = FSAR to Seabrook; N/C =section;lculated; not ca N/A = not available.SD = system description FF = FSAR figure; SGU = intercompany memorandum 1303P092785 m M M M M M M M M M M M M M M M M M
M M M M M M M M M M~ M M M M M M TABLE 4-5 (continued) l Sheet 6 of 14 Seabrook Station Zion Station l Source Value
- c. Height from spring line to top of inner SBU-24859 70 48 containment (inside) (feet).
- d. Volumes and locatfogs of compartments within Open Open the containment (f t ).
! e. Provisions for mixing containment atmosphere :--- None None (including compartments) for total loss of AC power.
- f. Containment Sump: Calculation 4.3.22F6 (1) Location. FF 6.2 Periphery of Outside Wall Containment 3
(2) Vol ume ( f t ) . Calc. 4.3.22F6 2,561 706 (3) Curb height (feet). FF 6.2 None* None a (4) Screen mesh size. FF 6.2 Vertical: Not Available
& 1" x 3-11/16" ui Open Horizontal:
0.097 Inch Open
- g. Water volume discharged onto containment f1 r Estimate 25,130 9,622 3
before spillover into reactor cavity (ft ). { l
- h. Containment type (i.e., large, dry). PSS Section 11.2 Large Dry Large Dry l i
- 1. Containment construction (i.e., steel lined. FF p. 6.2-5 Steel Lined Steel Lined, reinforced, or prestressed). Reinforced Prestressed, Concrete Post-Tensioned J. Intermediate floor openings (grated, concrete PSS Section 11.2 Grated Grated Concrete hatch). Operating Deck
- k. Fan cooler ducting within 10 feet of containment PSS Section 11.2 N/A No floor (yes/no). -
- Floor slopes away from sump.
NOTE: FT = FSAR table; FS = FSAR section; SD = system description; FF = FSAR figure; SBU = intercompany memorandum from UE&C to Seabrook; N/C = not calculated; N/A = not available. 1303P092785
TABLE 4-5 (continued) Sheet 7 of 14 Seabrook Station Source Value
- 9. Containment Operating Conditions
- a. Range of nonnal containment pressures (maximum / UEAC Drawing +0.5/-1.5 +0.3/-0.1 minimum) (psig). 9763-F-300219
- b. Range of nonnal containment temperatures 120/50 120/65 (maximum / minimum) (*F).
- c. Range of refueling water storage tank FS 6.2.2.3 86 Maximum 100/40 temperatures (maximum / minimum) (*F).
- d. Range of temperature outside containment SBU-24859 104/50 95/-10 (in shield building annulus, if any) (maximum /
minimum) (*F). , 10. Structural Heat Sinks in Containment b' a. osed steel inside containment (including 3,548,930 6,110,905 Mass liner,ofexc exfuding RCS, Ibm),
- b. Surfaceofexpogedsteel(excludingRCS)inside 227,592 327,387 containment (ft ).
- c. Mass of concrete inside containment (Ibm). Calculated from 33,835,392 13,522,508 FT 6.2-3
- d. Mass of concrete in containment shell (excluding 56,837,644 39,264,725 basemat, Ibm),
- e. Mass of steel in containment shell concrete 6,121,416 Not Available (excluding basemat, excluding liner Ibm).
- f. Surface of exposed concrete inside containment:
2 I 152 43,091 (1 )Greater (2) Less thanthan22feet feet thickness (ft ).2). thickness (ft 54,262 25,736
- 11. Reactor Cavity Area
- a. Reactor cavity volume to top of curb (ft 3). FS 6.2.2.3.a 16,935 7,940
- b. Instrument tunnel volume. Included in Included in Item 11.a Item 11a NOTE: FT = FSAR table; FS = FSAR section; SD = system description; FF = FSAR figure; SBU = intercompany memorandum from UE&C to Seabrook; N/C = not calculated; N/A = not available.
1303P100285 m M M M M M M M M M M M M M M M M
M N-~M W- - 7 M T~~l M7 n R U U TABLE 4-5 (continued) Sheet 8 of 14 Seabrook Station 3 Model Source Value j
~$
- c. Fraction of tunnel area occupied by instrument Small Small and other structures at limiting location,
- d. Elevation from cavity floor to top of curb (feet). UE8C Drawing 29.3 23.5 9763-F-805056
- e. Position of reactor vessel relative to cavity 17.5* 14.5*
(feet).
- f. Reactor vessel support. Biological Biological Shield Shiel d Supports Nozzles Supports Nozzle s
- g. Flow paths between reactor cavity and main containment volume. '
PLG Letter
? (1 Instrument tunnel (ft 2), 11/1/82 76.56 63.75 m (2)) Manway access (ft2 ). Torri to Tsai 7.11 N/A " 2 2.8 53.0 (3) Around regetor vessel (ft ). 5.38 (4 ) Othar (ft ).** 7.74 (5) Total 91.4 122.1
- h. Reactor vessel insulation. TRANSCO Dwg. 3.5 Inches SS 3 Inches SS JM-4421-02 Reflective Reflective i Reactor cavity floor area (ft2). Horizontal-- Horizontal--344.6 476.4 Total--399
- j. Debris discharge path from reactor cavity Smooth Sloped Smooth Sloped (smooth, restrained). PLG Letter 11/1/82
- k. Room configuration above debris discharge location Torri to Tsai Open to Lower Open to Lower (closed volume, open to main containment volume). Containment Containment
- 1. Reactor cavity sump dimensions 4' x 5'4" x 2' Approximately (length x width x depth). 6' x 4' x 3'
- m. Reactor cavity curb height above containment 30 6 floor (inches).
- Clearance from bottom of reactor vessel to cavity floor.
** Tunnel bypass.
NOTE: FT = FSAR table; FS = FSAR section; SD = system description; FF = FSAR figure; SBU = intercompany memorandum from UE&C to Seabrook; N/C = not calculated; N/A = not available. MM . _ _ _ _ _ _ -
TABLE 4-5 (continued) Sheet 9 of 14 Seabrook Station Zion Station Source Yalue
- 12. Reactor and RCS Parameters
- a. (1) Fuel rod and assembly geometry and FT 4.1-1 17 x 17 Array 15 x 15 Array dimensions.
(2) Geometry and dimensions of grid plates. FT 4.1-1 8 7 (3) Thickness / diameter of reactor vessel YAEC 0.448/14.7 0.443/14.7 bottom head (feet).
- b. (1) Mass of UO2 in core region (1ba). YAEC 222 739 216 600 (2) Mass of Zr in core region (lbm). YAEC 45,E34 44,500 i
(3) Mass of silver in core region (1bm). YAEC 6 150 6,150 1 l (4) Mass of other materials in core region (Ibm). YAEC IE850 N/C
- c. Core power (MWth). FT 6.2-1 3,411 3,236
- d. Flow area of core (ft2 ). FT 4.1-1 51.1 51.4
- e. Makeup, let down flow (gpm). YAEC 75-120 75-120 l f. Mass of water which can be stored in bottom Estimate 50,000 50,552 at 585'F head (Ibm). and 2,200 psi
- g. Clad thickness (feet). FT 4.4.-1 001875 00206
- h. (1) Operating temperature (hot leg, cold leg, 'F). FT 5.1-1 618.2/558.8 594.3/530.2 (2) Operating pressure (psia). FT 5.1-1 2,250 2,265
- 1. (1) Safety relief valve set point (psia). FT 5.4-6 2,500 2,500 (2) Rated flow of safety relief valves 420,000 Each 420,000 Each (pound / hour).
J. Masses, materials, location of materials in j upper and lower plenums. (1) Upper Estimate 132 000 See Item 12r (2) Lower Estimate 79,$00 See Item 12r
- k. Initial primary steam volume (ft3 ). FT 5.1-1 741 720
- 1. Initial primary water volume (ft3 ). FT 5.1-1 11,524 12,281 1
NOTE: FT = FSAR table; FS = FSAR section; SD = system description FF = FSAR figure; SBU = intercompany memorandum from UE8C to Seabrook; N/C = not calculated; N/A = not avaliable. 1303P092785 W M M M M M M M M M M M M
TABLE 4-5 (continued) Sheet 10 of 14 Seabrook Station gg Source Value
- m. Mass of water in steam generator NAH*-U-1961 448,000 357,400 secondary side (lbm) (total of four),
- n. Layout of primary system hot and cold legs. FF 5.1-1 29 Inches 29.2 Inches Hot leg Hot Leg 27.5 Inches 27.7 Inches Cold Cold Leg Leg
- o. Steam generator secondary pressure relief 1,135 1,050 Relief Yalve set point (psia), 1,065 Safety Valve
- p. Exposed gurface area of steel internals above 10,764 10,764 core (ft ).
- q. (1) Reactor vessel inside surface above core (ft2 ). 484 484 (2) Reactor vessel mass above core (Ibm). 88.105 88,105 a (3) Reactor vesgel inside surface area adjacent 328 328 to core (ft ).
@' (4) Reactor vessel mass adjacent to core (Ibm). 74,890 74,890 (5) Reactor gessel inside surface area below Included in Included in core (ft ). Item 12s(3) Item 12s(3) (6) Reactor vessel mass below core (1bm). Included in Included in Item 12s(4) Item 12s(4) 2
- r. (1) RCS piping inside surface - hot le 21,528 21 528 (2) RCS piping mass - hot legs (poundsfs (ft ).
. 136,564 136,564 2
3 RCS fping inside surface - cold legs (ft ). 2 691 2 691 (4) ( ) RCS f ping mass - cold legs (pounds). 658,767 638,767
- s. (1) Pregsurizer and surge line inside surface 1,245 1,245 (ft ).
(2) Pressurizer and surge line mass (pounds). 170,000 170,000
- t. (1) Steam generator tubing inside surface (ft2 ), 21,528 21,528 (2) Steam generator tubing mass (pounds). 145,374 145,374
- u. (1) PORY pressure setpoint (psia). 2,350 2,350 (2) PORY rated flow (pounds / hour). 210,000 210,000 NOTE: FT = FSAR table; FS = FSAR section; SD = system description; FF = FSAR figure; SBU = intercompany memorandum from UE8C to Seabrook; N/C = not calculated; N/A = not available.
1303P092785
TABLE 4-5 (continued) Sheet 11 of 14 Seabrook Station Zion Station Source Value
- v. (1) Volume of water in quench tank (f 3t ). 900 900 (2) Building elevation of quench tank rupture 4.92 4.92 disks (feet).
- w. (1) Reactor coolant pump model. FS 5 93A-1 93A (2) Reactor coolant pump seal design. Weir Conventional Three Seal
- 13. Basemat Concrete Below Reactor Cavity
- a. Composition by weight of elements of concrete. Pittsburgh SiO2 :0.622 Free H20:
Testing 2.7 w/o Laboratory CO :0,015 2 Bound H2 0: Report 2.0 w/o Ca0:0.025 CO2 : 2I 2 W/0 CACO 3 :0.0343 i b. Density of concrete (1bm/f t 3). UE&C 144 142.9 $ 603-SEABR00K
-DOC-60
- c. Thickness of basemat:
(1) Inner surface to ifner (feet). UEAC Drawing '! .0 (2) Liner thickness (feet). 9763-F .J21 .0313 (3) Liner to lower surface (feet). 101,402 6 Feet, 3.5 11 Inches
- d. Weight percent of water (bound and free) in NUREG/CR-2142 Free: 2.7 Free: 2.7 reactor cavity concrete. Bound: 2.0 Bound: 2.0
- e. Weight percent of steel in reactor cavity concrete. PLG Letter 7.54 Not Available 11/1/82 Torri to Tsai
- 14. Containment Leakage Data
- a. Primary containment design leak rate 0.1 0.1 (percent / day),
- b. ndary containment design leak rate 3,704 Not Applicable Secg/
(ft day). NOTE: FT = FSAR table; FS = FSAR section; SD = system description; FF = FSAR figure; SBU = intercompany memorandum from UE&C to Seabrook; N/C = not calculated; N/A = not available. 1303P092785 W W W W W W W W M M M
1 m
. M M l
TABLE 4-5 (continued) Sheet 12 of 14 Seabrook Station Zion Station Item Model l Source Value
- c. Containment interspace (annulus) width (ft). FT 6.2-82 4.5 to 5.5 Not Applicable
- d. Containment interspace volume (ft3 ). FS 6.2.3.1 524,344 Not Applicable
- e. Containment interspace pressure FS 6.2.3.1 -0.25 Not Applicable (psid or inches of water),
- f. Containment enclosure emergency exhaust filtration system ,
1 (1) Status during normal operation SD No. 53 Standby Not Applicable (2) Maximum exhaust flow rate (cfm}. SD No. 53 2x2000 Not Applicable (3) Exhaust filtration. SD No. 53 HEPA Moisture Not Applicable
- 15. Containment Penetrations See FSAR See FSAR Table 6.2-83 Table 6.6.5-1 i Containment atmospheric purge line diameter 8 10 g (inches).
- 16. Auxiliary Building Data
- a. RHR cubicle volume (ft3 ). 133,208 1,465,400
- b. Elevation of lowest opening (feet). (-) 31 feet 342 feet 10 inches
- c. Water fill volume to elevation in (b) (ft3 ). 49,860 0
- d. Water level after RCS injection (feet). 6.7 -O
- e. Water level after RCS and RWST injection (feet). 31 feet -0 10 inches
- f. Elevation of RHR pumps (feet). (-) 56 feet 342 feet 4 inches
- g. Elevation of pressure relief valve (s) (feet). (-) 18 feet Not Available 5 inches NOTE: FT = FSAR table; FS = FSAR section; SD = system description; FF = FSAR figure; SBU = intercompany memorandum from UEAC to Seabrook; N/C = not calculated; N/A = not available.
1303P092785
TABLE 4-5 (continued) Sheet 13 of 14 Seabrook Station Zion Station Mcdel Source Value
- h. Elevation of RHR piping penetrations (feet). (-) 18 feet Not Available 5 inches
- 1. Elevation of first RHR piping elbow in auxiliary (-) 18 feet Not Available building (feet). 5 inches J. Elevation of RHR piping high point in auxiliary (-) 18 feet Not Available building (feet). 5 inches
- k. RHR valving. FS 5.4.7 Suction: Suction:
One MOV Inside Two MOV Inside (1) Location. Missile Barrier Containment (nor.aally (normally closed); One closed); One MOV MOV Inside Pump Room Containment (normally open). (normally Discharge (hot A closed). leg): Two CV 4 (2) Actuation. Discharge Inside Con-ru (cold leg): tainment; One One CV Inside MOV Inside Missile Containment Barrier; One (normally CV Inside closed); One (3) Nomal post-trip position. Containment; MOV Outside One MOV Outside Containment Containment (normally (normally o closed). Discharge (pen), hot Discharge (cold leg): One CV leg): Three Inside Missile CV Inside Con-Barrier; One tainment; One CV Inside Con- MOV Gutside tainment; One Containment MOV Outside (normally open). Containment (nomally open). NOTE: FT = FSAR table; FS = FSAR section; SD = system description; FF = FSAR figure; SBU = intercompany memorandum from UE&C to Seabrook; N/C = not calculated; N/A = not available. 1303P092785 E E E E E
i TABLE 4-5 (continued) Sheet 14 of 14 I Seabrook Station Zion Station Item Model Source Value j l
- 1. RHR piping dats (outside containment).
(1 ) Design pressure (psia). FS 5.7 600 psig 600 l (2) Design temperature (*F). FS 5.7 400 400 (3) Pipe schedule. FF 5.4-10 Suction Hot Leg Suction 12 inches 14 inches Discharge Hot Leg Discharge 8 inches 12 inches No Schedule Cold Leg Discharge l Given 10 inches ' SS 310 l
- 17. Containment Failure Characteristics
- a. Ultimate pressure capacity (psia). 211 (wet 149 sequence) 190 (dry sequence) i b. Ultimate containnent temperature (*F). 450 w
" (1) Atmosphere Not Available 450 (2) Wall PSS Section 11.3 700 ' Not Available
- c. Dominant component failure modes. Feedwater 1% Strain of Penetration Shell
- d. Characteristic leak area (ft2 ). Self-Regulating Self-Pegulating NOTE: FT = FSAR table; FS = rSAR section; SD = system description; FF = FSAR figure; SBU = intercompany memorandum from UE8C to Seabrook; N/C = not calculated; N/A = not available.
1303P092785
l TABLE 4-6. DESCRIPTION OF SEABROOK RHR SYSTEM I Sheet 1 of 3 RHR System Description Pump Suction-Side From Reactor Coolant System (12-inch line) Flow Path (B Train) Through Normally Closed MOV (RC-V87) Pass PRV (RC-V361 with 2,485-psig setpoint) I Through Normally Closed MOV (RC-V88) Pass PRV (RC-V89 with 450-psig setpoint) Penetrate Containment (penetration X-10) Pass "T" to Reactor Coolant Filter [3-inch line with check valve (CS-V497) and locked-closed MOV (CS-V829)] Penetrate Equipment Vault Pass "T" to RWST [12-inch line with check valve (CBS-V56) and locked-open MOV (CBS-V24)] Pass "T" to RHR Heat Exchanger Recirculation [3-inch line with normally open MOV (FC-V611)] Through Reducer to 16-Inch Line Pass "T" to Containment Recirculation Sump Through Reducer to 14-Inch Line Pass "T" to WLD DT (3/4-inch line with locked closed H0V) Into RHR Pump I Pump Discharge-Side From Reactor Coolant System (10-inch line) i Flow Path (B Train) Through Check Valve (SI-V35) ' Pass "T" to Accumulators [10-inch line with normally open l MOV (SI-V32) and check valve (SI-V36)] Through Reducer to 6-Inch Line Through Locked-Open MOV (RH-V63) 1279P083135 4-74
TABLE 4-6 (continued) Sheet 2 of 3 RHR System Description Pump Discharge-Side Pass "T" to 3/4-Inch Drain Line [normally closed MOV Flow Path (B Train) (RH-V113)] (continued) Pass "T" to SI Pumps (2-inch line with check valve SI-V130) Through Check Valve (RH-V29) Through Reducer to 8-Inch Line - Pass "T" to Test Line [3/4-inch line with normally closed MOV(RH-V27)] Pass "T" to Drain Line [3/4-inch line with normally closed MOV (RH-V110)] Penetrate Containment (penetration X-12) Through Normally Open MOV (RH-V26) Pass "T" to Drain Line [3/4-inch line with normally closed MOV (RH-V106)] Pass "T" to PRV (RH-V25 with 600 psig setpoint) Pass "T" to 8-Inch Line [normally closed / fail closed MOV (FC-V619)] Pass "T" to 8-Inch Line [normally open MOVs (RH-V22, RH-V21)] Through Normally Open/ Fail Open MOV (HC-V607) Pass "T" to RHR Recirculation Line [3-inch line with normally open M0V (FC-V611)] Pass "T" to SI System [8-inch line with normally closed MOV (RH-V36)] Into RHR Heat Exchanger Out of RHR Heat Exchanger (12-inch line) I 4-75 1279P083185
I TABLE 4-6 (continued) Sheet 3 of 3 RHR System Description Pump Discharge-Side Through Reducer to 8-Inch Line Flow Path (B Train) (continued) Through Locked-0 pen HOV (RH-V45) Pass "T" to RHR Recirculation [8-inch line with normally closed / fail closed M0V (FC-V619)] Pass "T" to 3/4-Inch Line [normally closed MOV (RH-V43)] Through Check Valve (RH-V40) Pass "T" to Flush Connection [3/4-inch line with locked closed MOV (RH-V39)] Into RHR Pump I I I 1279P083185 4-76
I TABLE 4-7. V-SEQUENCE CHRON0 LOGY I I Approximate Time Event 0 Two Series Valves Leak Excessively 0 RHR Pump Seal Failure I 0 5 seconds RHR Relief Valves Lift HPI On 27 seconds Pressurizer Relief Tank Rupture Disk Fails I 29 seconds Reactor Coolant System Solid 7.7 minutes Accumulator Discharge Begins 12.2 minutes RHR Relief Valves Begin To Modulate 30 minutes Spray Pumps Flood 1.8 hours RHR Pumps Flood 1.0 hour Accumulator Water Depleted 2.8 hours Safety Injection Flooded in Equipment Vault 6.4 hours RWST Water Depleted 6.4 hours ECCS Recirculation Fails I 7.4 hours RCS Water Level Falls below RHR Piping Level 8.1 hours Core Uncovery Begins 8.5 hours Zircalloy-Water Reaction Begins 10.0 hours Core Melting Begins l 11.5 hours 11.5 hours Reactor Core Support Plate Fails Reactor Vessel Fails 11.5 hours Reactor Cavity Dry; Core-Concrete Interaction Begins 24.0 hours End of Analysis
- I I
I 1300P120585 4-77
I I' TABLE 4-8. DEFINITION 0F FISSION PRODUCT GROUPS I Group Fission Product 1 Noble Gases 2 Cesium Iodide 3 Tellurium 4 Strontium 5 Ruthenium and Lanthanum 6 Cesium Hydroxide I 4-78 1279P083185
t E I l I l I l I l I l TABLE 4-9. RELEASES TO EQUIPMENT VAULT l Fraction of Core Inventory Released Time l (hours) Group 1 Group 2 Group 3 Group 4 Group 5 Group 6 11.49 9.04-1 2.99-1 1.76-1 3.02-4 3.84-4 2.29-1 16.0 9.16-1 3.23-1 1.99-1 1.44-3 2.83-3 2.63-1 l 24.0 9.27-1 3.25-1 2.04-1 1.47-3 2.87-3 2.84-1 l l NOTE: Exponential notation is indicated in abbreviated form; I i.e., 9.04-1 = 9.04 x 10-1 I l l i 1279P083185 4-79 1 j
I' I TABLE 4-10. RELEASES TO ENVIRONMENT - SUPPRESSION POOL SCRUBBING l Fraction of Core Inventory Released Time (hours) Group 1 Group 2 Group 3 Group 4 Group 5 Group 6 11.49 9.04-1 2.99-4 1.76-4 3.02-7 3.84-7 2.29-4 16.0 9.16-1 3.23-4 1.99-4 1.44-6 2.83-6 2.63-4 24.0 9.27-1 3.25-4 2.04-4 1.47-6 2.87-6 2.84-4
- NOTE
- Exponential notation is indicated in abbreviated form; l
1.e., 9.04-1 = 9.04 x 10-1 I I l
)
I 1279P083185 I [ 4-80 l
I I I 1 I TABLE 4-11. RELEASES TO ENVIRONMENT - N0 SUPPRESSION POOL; VENTILATION Time Fraction of Core Inventory Released (hours) Group 1 Group 2 Group 3 Group 4 Group 5 Group 6 11.49 9.04-1 5.35-2 4.17-2 4.26-5 4.05-5 4.69-2 16.0 9.16-1 5.70-2 4.38-2 1.48-4 2.08-4 5.02-2 24.0 9.27-1 5.71-2 4.41-2 1.51-4 2.72-4 5.20-2 NOTE: Exp'onential notation is indicated in abbreviated form; i.e., 9.04-1 = 9.04 x 10-1 I I I 1279P083185 4-81
l I I I l l l I l TABLE 4-12. RELEASES TO ENVIRONMENT - < N0 SUPPRESSION P00L; N0 VENTILATION l Fraction of Core Inventory Released Time (hours) Group 1 Group 2 Group 3 Group 4 Group 5 Group 6 11.49 9.04-1 8.83-2 7.89-2 6.58-5 3.45-6 8.29-2 16.0 9.16-1 9.17-2 8.20-2 1.88-4 3.28-4 8.54-2 24.0 9.27-1 9.39-2 8.31-2 2.27-4 3.84-4 9.19-2 NOTE: Exponential notation is indicated in abbreviated form; i.e., 9.04-1 = 9.04 x 10-2 l 1279P083185 l 4-82 ' l
I TABLE 4-13. BEST ESTIMATE RELEASE CATEGORIES Rd m Release Time (hours) Energy elease Racdons Source
- Category Start Duration Warning 106 cal /sec XE CS,1 TE SR RU LA EARLY C0NTAINMENT FAILURE SIB-1 2 2 1 < 10 .2 .022 .004 .003 8.-4 8.-5 SSPSA: T5V-I d S1B-2 4 4 3 < 10 .3 .028 .005 .003 .001 1.-4 M -2d S1B-3 8 6 7 < 10 .5 .002 .004 2.-4 2.-4 4.-5 T5V-3d
$1B-Total 2 12 1 < 10 1 .052 .013 .006 .0002 2.-4 EARLY INCREASE CONTAINMENT LEAXAGE S28-1 13 12 5 < 10 .15 .004 7.-4 5.-4 2.-4 2.-5 SSPSA: 32V-1d S28-2 25 8 17 < 10 .2 .007 8.-4 8.-4 6.-4 1.-4 Tfv-2d 528-3 33 56 25 < 10 .65 .002 .002 2.-4 1.-4 2.-5 32V-3d S28-Total 13 76 5 < 10 1 .013 .004 .002 9.-4 1.-4 LATE OVERPRESSURE CONTAINMENT FAILURE S3B 89 0 74 < 10 1 .001 .002 1.-5 1.-5 1.-5 IDCOR CONTAINMENT INTACT SSB 2 24 0.4 < 10 .009 4.-8 6.-9 4.-9 1.-9 1.-10 SSPSA: SS CONTAINMENT PURGE ISOLATION FAILURE I S68-1 568-2 568-3 4
6 10 2 4 10 3 5 9
< 10
< 10
< 10
.2
.3
.5
.004
.005
.001 9.-5 1.-4 9.-5 3.-4 3.-4 2.-5 2.-5 3.-5 1.-5 2.-5 3.-5 1.-5 IDCOR S6B-Total 4 16 3 < 10 1 .01 3.-4 6.-4 6.-5 6.-5 CONTAINMENT BYPASS (V-SEQUENCE)
S78-1 8.5 1 5.5 < 10 .2 6.-5 6.-5 0 0 0 MAAP-SB:** Pool 578-2 9.5 1 6.5 < 10 .6 1.-4 1.-4 4.-8 0 0 MAAP-SB: Pool S7B-3 10.5 5 7.5 < 10 .2 2.-4 4.-5 1.-6 3.-6 3.-6 MAAP-SB: Pool S78-Total 8.5 7 5.5 < 10 1 3.-4 2.-4 1.-6 3.-6 3.-6 I ***S:e text for sourceMAAP Seabrook-specific term modification. analysis (see Section 4.4). NOTE: Exponential notation is indicated in abbreviated form; i.e., 8.-4 = 8.0 x 10-4 1300P120685
I TABLE 4-14. CONSERVATIVE ESTIMATE RELEASE CATEGORIES I Release Time (hours) Energy Release Fractions gg 6 Soume* alories Category per Second Xe Cs,I Te Sr Ru La Start Ouration Warning EARLY CONTAINMENT FAILURE SIC-1 1 1 .5 < 10 .2 .06 .01 .007 .002 2.-4 SSPSA: M-1 c 3 S1C-2 2 3 1.5 < 10 .3 .07 .013 .009 .003 3.-4 TSV-2c SIC-3 5 5 4.5 < 10 .5 <.005 .009 4.-4 6.-4 1.-4 M -3c SIC-Total 1 9 .5 < 10 1 .135 .032 .016 .006 6.-4 I EARLY INCREASE CONTAINMENT LEAKAGE S2C-1 5 7 .6 < 10 .15 .007 .001 .001 3.-4 3.-5 SSPSA: M-1 c S2C-2 12 6 7.6 < 10 .2 .015 .002 .002 .001 2.-4 T2V-2c S2C-3 18 38 13.6 < 10 .65 .003 .005 4.-4 3.-4 5.-5 M -3c S2C-Total 5 51 .6 < 10 1 .025 .008 .003 .002 3.-4 l LATE OVERPRESSURE CONTAINMENT FAILURE S3C 54 0 42 < 10 1 .002 .01 2.-4 2.-4 3.-5 SSPSA: M -c CONTAINMENT INTACT SSC 2 24 0.4 < 10 .014 5.-7 1.-7 6.-8 2.-8 2.-9 SSPSA: SS l CONTAINMENT PURGE ISOLATION FAILURE l 56C-1 2 2 1 < 10
< 10
.2 .022 .01 003
.003
.002
.0026 8.-5 1.-4 SSPSA: T6V-1d T5V-2d l
l 56C-2 4 4 3 .3 .028 01 3 3 S6C-3 8 6 7 < 10 .5 .002 .009 2.-4 6.-4 4. 5 T6V-3d S6C-Total 2 12 1 < 10 1 .052 032 .006 .005 2.-4 l CONTAINMENT BYPASS (V-SEQUENCE) S7C-1 8.5 1 2 < 10 .2 .038 .04 7.-6 0 0 MAAP-SB:**No Pool S7C-2 9.5 1 3 < 10 .6 .038 .026 3.-5 0 0 MAAP-SB: No Pool S7C-3 10.5 5 4 < 10 .2 .01 8 .017 2.-4 4.-4 4.-4 MAAP-SB: No Pool i 57C-Total 8.5 7 2 < 10 1 .094 .083 2.-4 4.-4 4.-4
*S:e text for source term modification. l **S:abrook-specific MAAP analysis (see Section 4.4). 5 l NOTE: Exponential notation is indicated in abbreviated form; f.e., 2.-4 = 2.0 x 10-4
! I I 4-84 , 1300P120685
TABLE 4-15. COMPARIS0N OF RELEASE CATEGORIES Release Time (hours)
- 9# * ** * * "*
Rd m 6 Source * ' *
- Cate9ery Start Duration Warning per Second Xe I Cs Te Sr Ru La EARLY CONTAINMENT FAILURE
- This Study S1B 2 12 1 < 10 1 .052 .052 .013 .006 .005 2.-4 I NUREG-0956 This Study NUREG-0956 NUREG-0956 V-Pool SIC V-No Pool TMLB'D 2.5 1
1 1 14 10 2 2
.8
.5
.8
.5
< 10
< 10
< 10
< 10 1
1 1
.85
.08
.135
.4
.07
.08
.135 4
.058
.025
.032
.12
.055
.0022
.01 6
.011
.01 1.-4 7.-5
.0056 6.-4 7.-4 4.-4
.0013 2.-4 W ASH-1400 PWR-2 2.5 .5 1 12 .9 .7 .5 .3 .06 .02 .004 This Study SIE** 1 2 0.5 < 10 1 .4 .4 .12 .01 6 .006 6.-4 EARLY INCREASE CONTAINMENT LEAKAGE I This Study This Study 528 S2C 13 5
76 51 .6 5 < 10
< 10 1
1
.013
.025
.013
.025
.004'
.008
.002
.003 9.-4
.0018 1.-4 3.-4 LATE OVERPRESSURE CONTAINMENT FAILURE This Study S3B 89 0 74 < 10 1 .001 .001 .002 1.-5 1.-5 1.-5 This Study S3C 54 0 42 < 10 1 .002 .002 .01 2.-4 2.-4 3.-5 10COR-Zion ID-SB0 32 G 30 < 10 1 .002 .002 2.-5 1.-5 1.-5 1.-5 CONTAINMENT PURGE ISOLATION FAILURE This Stutr 56B 4 16 3 < 10 1 .01 .01 3.-4 6.-4 6.-5 6.-5 This Study 56C 2 12 1 < 10 1 .052 .052 .033 .0062 .005 2.-4 IDCOR-Zion ID-IMPAIR 4 --
3.5 < 10 1 .01 .01 3.-4 6.-4 6.-5 6.-5 NUREG-0956 TMLB'8 2 10 0 < 10 1 .022 .013 .11 .058 .0053 2.-4 WASH-1400 PWR-4 2 3 2 < 10 .6 .09 .04 .03 .005 .003 4. 4 This Study S6E** 2 10 0 < 10 1 .05 .05 .11 .06 .006 2.-4 CONTAINMENT BYPASS f V-SE0VENCE AT RHR PUMP SEAL) This Study S78 8.5 7 5.5 < 10 1 3.-4 3.-4 2.-4 1.-6 3.-6 3.-6 IDCOR-Zion ID-BYPASS 24 -- 4 < 10 1 8.-5 8.-5 8.-5 5.-5 1.-5 1.-5 This Study 57C 8.5 7 2 < 10 1 .094 .094 .083 2.-4 4.-4 4.-4 INTACT CONTAINMENT This Stutr SSB 4.3 24 .6 < 10 .009 4.-8 4.-8 6.-9 4.-9 1.-9 1.-10 This Stutr SSC 2 24 .4 < 10 .014 5.-7 5.-7 1.-7 6.-8 2.-8 2.-9 OIncludes V-sequences involving pipe rupture outside containment. I NEnveloping source terms used for sensitivity analysis. NOTE: Exponential notation is indicated in abbreviated form; i.e., 2.-4 = 2.0 x 10-4 I l i 4-85 1300P121685
TABLE 4-16. ENVELOPING SOURCE TERMS FOR SENSITIVITY ANALYSES Release Time (hours) Energy Release Fractions Rd m 6 Start Duration Warning pr nd Xe Cs, I Te Sr Rr La . SIE-1 1 1 0.5 < 10 1 3 .06 .008 .003 3-4 This Study - SIC NUREG-0956 - TMLB'D SIE-2 2 1 1.5 < 10 -
.1 .06 .008 003 3-4 NUREG-0956 - V -
No Pool SIE - Total 1 2 0.5 < 10 1 .4 .12 .01 6 006 6-4 .so cn 56E-1 2 1 0 < 10 .5 02 .02 .02 002 8-5 This Study - S6C NUREG-0956 -TMLB'B 56E-2 3 3 1 < 10 .5 .02 .04 .03 .003 1-4 56E-3 6 6 4 < 10 - 01 05 .01 001 4-5 56E - Total 2 10 0 < 10 1 .05 .11 .06 .006 2-4 NOTE: Exponential notation is indicated in abbreviated form; i.e., 3.4 = 3 x 10-4, 1300P120685 m M M M M M
I TABLE 4-17. REVISED C-MATRIX FOR NEW SOURCE TERM CATEGORIES e S urce Term Category S1 S2 S3 SS S6 S7 (frequ ncy) 1F 1.0 (2.0-8) (2.0-8) 1FV 1.0 (4.3-9) (4.3-9) 1FP 1.0 (1.4-6) (1.4-6) IFPV 1.0 (2.7-8) (2.7-8) 2A 3.4-5 1.4-4 1.0-2 0.99 (1.6-6) (5.5-11) (2.3-10) (1.6-8) (1.6-6) 3D/ 7D 2.0-6 8.0-5 0.95 0.05 (4.8-5) (9.6-11) (3.9-9) (4.6-5) (2.4-6) 3F/7F 1.0 (3.0-7) (3.0-7) 3FP/7FP 1.0 (1.9-5) (1.9-5) 4A/8A 3.1-6 1.3-4 5.2-3 0.995 I (1.1-4) (3.3-10) (1.4-8) (5.5-7) (1.1-4) 7FPV 1.0 (1.1-8) (1.1-8) 80 1.1-6 3.1-5 0.9999 (1.0-4) (1.1-10) (3.2-9) (1.0-4) Total 4.9-9 2.0-5 1.5-4 1.1-4 3.2-7 3.7-8 Frequency NOTE: Exponential notation is indicated in abbreviated form; i.e., 2.0-8 = 2.0 x 10-8 1300P120585 4-87
l I I N
\ ,
\ gl
> ' 5 m
' p O.
i
/
e l
/ $ g
_ _ _ _ _ _ _. / - g E l VA l l l l EA
" S z
l 2 m a y W l
,7 ,
4-88
o . m r _. - ~ m r X m r I
; 4 m L L
A W L m ) t A C n I R e D c N r m ( e p I L Y C e T s I N n , 3 aB E M N e I h t A T m N m o r O C F R y O a F w D - A N O I l l T a A W L m I l 2 t n E R e N I m A i n R m t a T S n P o . O C O e H m h t - n E i R U S m I l 1 i n a S E
. r R t P S 9 m p o
o 2 H 4 E R m U G I F m , .
. ~ _
O 0 0 0 0 0 0 5 0 5 2 1 1 m 9'mo.~ E E YE5. 380 m 1 E
- i! ! ; I ! ll ll 1t
1.0 _ 2 6 3 10'I - 1 2 3 4 5 6 CURVE DISCRETE ULTIMATE CURVE PROBABILITY HOOP (PSIA) 1 0.05 184 g
>- 10 2 -
2 0.15 202 g b 3 0.20 217 d m 4 0.20 231 E 5 0.20 245 6 0.20 268 g e - 9 b
- l S
8 10 3 _ l I
\ HOOP CAPACITY l I 10 4
7
< 6" - = > 6" l
_~ RADIAL DEFLECIlON l
' j 140 160 180 200 220 240 260 PRESSURE (PSIA)
FIGURE 4-3. CONDITIONAL CUMULATIVE PROBABILITY DISTRIBUTION FOR FUEL TRANSFER BELLOWS FAILURE 4-90
I 1.0 _ Z I _ _ CURVE 1 2 3 4 5 6 I 10'I -. _ ~ I E D 10'2 T 3 - V V y 'P FH PC DISCRETE g - CURVE (PSIA) (PSIA) PROBABILITY E 1 155' 184 0.05 j 2 170 202 0.15 z 3 182 217 0.20 9 - 4 194 231 0.20 t-I @ 0 10
-3 7
5 6 206 224 245 268 0.20 0.20 I :
~
l .17 T I l 10 T l l : - - FLUE PIPE HEAD CRUSHING I FAILURE FAILURE
~
5 I I I l I I I 10 140 160 180 200 220 240 260 FAILUR E PRESSURE (PSfA) I FIGURE 4-4. CONDITIONAL CUMULATIVE PROBABILITY DISTRIBUTIONS FOR FEEDWATER PENETRATION FAILURE (FLUEHEAD OR PIPE CRUSHING) BEFORE HOOP FAILURE AS A FUNCTION OF FAILURE PRESSURE 4-91
i 1.0 _ _ g g CURVE 1 2 3 4 5 6 10'l -- 1 - 1 i - i i y 10'2 -- t ~ d _- DISCRETE ULTIMATE t @
~
CURVE PROBABILITY HOOP (PSI A) l o. __ j 1 0.05 184 2 _ 2 0.15 202 9 3 0.20 217 I b 4 0.20 231 l C 5 0.20 245 l 10 3
-- 6 0.20 2G8 3 1 -
I
~
10'4
=
I I I I I I I l 05 140 160 180 200 220 240 260 l PRESSURE (PSIA) FIGURE 4-5. DISCRETE PROBABILITY DISTRIBUTION FOR l ALL OTHER CONTAINMENT FAILURE MODES COMBINED g l E l 4-92 4
I 1.0 _ Z I ~ CURVE 1 2 3 4 5 6 I 10'l -- I E
~
I - DISCRETE ULTIMATE CURVE PROBABILITY HOOP (PSI A) 2 1 0.05 184 g 10 T 2 0.15 202 I -3 co g o 3 4 5 0.20 0.20 0.20 217 231 245 _ 6 0.20 268 E E 3 a - 2 9 - I b O Z
-3 8 10 -
I E I 4 -- 10 I E I 6 5 i 10 I 140 160 180 200 PRESSURE (PSI A) 220 240 260 I FIGURE 4-6. COMBINED DISCRETE PROBABILITY DISTRIBUTIONS FOR ALL BENIGN CONTAINMENT FAILURE MODES 4-93
1.0 _ ,,,........ eee- = - _ TOTAL FAILURE e'
.**.** # ~~ ~
_ g
- PR ESSURE:
- g
- og -
e / y - Q/ TYPE B (LEAK) E
. / / FAILURE 3
/ /
/ WET SEQUENCES
* /
* /
10'I -- e
. / TYPE C (GROSS)
FAILURE E
. / WET SEQUENCES 3
.' /
. / g M ~
/ 3
/
- /
l y 10- = : l 3 : 5 B
/
/
/
l G
- l y _ : / -
0.4 l
% : I e ' 10- =
- I 1 .
e l
- / 3 i
/ -
0.3 3
. l
- I 7
- / =
C g m l f 0.2 2 m l 9 l
/
10
~
j E
/ E
/ -
0.1 _ /
/
- l
/
.s i 140
/
160 i i 180 i 200 i 220 240 i i 260 o PR ESSU RE (PSIA) FIGURE 4-7. COMPOSITE CONTAINMENT FAILURE PROBABILITY DISTRIBUTIONS FOR TYPE B (LEAK) FAILURE, TYPE C (GROSS) FAILURE, AND TOTAL FAILURE 4-94
-W ml R R R R R R R M n G n_ n n. P
~~9 ~ N..
m
. . . . s ,,,,,,_.,,,
C
%y, 2
..N. .
.s
.;. .. y r
@@ m m I
f g--.,. g.(Lt ) J -. -
-~
g [ i~l
. @d L:.,_N.f .~
A .. m s7. ..
-.e., .a,e -.=
. ~ . -
- ;~. .,
.N
.- ~ s_. _9 a J
~
~~ . su,- A
~ q.a.*a):-~ - +;:.. n m -
.. e . .. -
- 7. . . . . .- V ;:- y..
. -.. k. 'g..
e<. . .. -- '",~ !":: . . -
. - - . . -<... .. __. w ~~
O ,
~~
7 N
=
f"" #
'T , 'T' *
-::.:r *~ ~
f.
'{W. . ..S.,,Y . , . . .
*;(4- - ;.i'"
3 l e= ":::: r.im 9 .M.~e: ;8 ' S i
,b:
- - ~_ >.. 4[:- ;..~ ~.~,.;:. d... .!!*i. (
@~+y-. f -*c.!: . . . .
.. . .l*;. h. d
.-e).
- s. a.
- c. ,
i
--.-6-1 :.> N 'a-* :: _- ,
'"a ,,Q, l t . ,b,>-- 2 ---e6 . j
),
c .a.r / ' g' Y -
- 4. -.G ,,,,,,, n. - - -
s _ Y. __ .. nrr so '7-
' sma se ..
trl .i- *:
- L N ?- -
7 -
'" ~~
. ., ; V ..;;- .;... V
....~ .r N ~.
O 'c. '[
-~ j' l - J;j, 4.
. r.g. .. 9 . -.= e ' 3e >--N
. . ~ e__ .
- - Q. . . . .
i 4.. g A) t .. . ., ., 9** ..t
-_ y: =2 9~ ..
". Q ~
{ -. 9: - .l l **_ .
~
$- @ Q l L*~ d+
.g. m G - ...
4 3 I. ~- ._e +:, .
- h. .- . .
. .~ . = . . .I.. .N
~* V " " :"l::::*'4~:
3 ggj ..
.. .. 1 m
l
.N . _ .. E.d.,
. l, FIGURE 4-8. ENGINEERED SAFETY FEATURE FLOW DIAGRAM
, N l ) @ n - - . - s.w 0 5Y s,Cf " *' y N.1-~f 'i
. i ...
J p-r ...;: ,c.. %- g.;g.-2 - N ,w-- -
.s. ..-
p e o,_mme :~
}..:. lp. l M... $@'W
.i ?, * -
a
.n -
m l" s wa r - ., , ;,- i i en 'fut *-
. , _t . .:.
,1 E i i
Lg- a 13..,... A vautt +1 $ (
- m = u. / .
. /
%} (,i _ ';
..- .. 3
.: 5;.
ww'ps \m
., mrr ,
N f ./2 g
. T ?'t'
) 3
. g- .,- ..
;%p..
, e N 4_ h., f
.i
,,, . m , i.e T ' . .
9- -T . .. 7-
- y. gt-.%c
. , l]c*~ 7
@t -
.- -F-i- u-Nr, : ,1 ", .
1 Q1 e' " ';a. ; E g 2
@ & ?* O' l ia.c PL AN AT El *. Al' O' m C
. .. , ' n .
- , .~:,
^-* ...*
- 9. 5 ya*.i
.d. . . .
*h g
.,s k \. 4, '.'. D, k)
>> .u euw i- e g- - . s- .x a 3 ..
L.u Y i D
.b.
e .- 's ,
, .:i " /- m, .= .y ..
( .. .N. ..... N L h e.
.. .a'S___
,67 t '\
, / ~
s Gu n.w.8 . , . p,.L;~ 4 ..:,, e' . w- ._.- .y3,. N .' l v uu,* r ,g] b DM'
, _ .. .m an.. 7, ,' f.N,
& d .%_ Kg. / d, 'i
'8
/
/ .
\
.L.4 '
_1'f Lt.ti. 7 I - b c l . W **** i
. . q. . '
, f .... . is #
.. .ni 'y'.:
.., p ,qua,J u,
;r..ae~ - p,./)
1 ;\
'. 21 9 , . . . . . - . _ ...
$m d 6
Dib [ ;."., J._. h . .e- g . -dI - h ,
*,*9 A
, .-. . g h w w' ,
***e..
\ 5 3-l
.c.l g
C, -
. ,e 6 . .,
~# .- .,
81 4-y 6- er-Iw g i
- e. (
t 3
----@d+ y +-
.*.- a yt'6]._ :,=
1
. pt a y or
"+.'.'i..'..' a m........,
et. 30.o g v.7 n'
- i
/ .'}. ,fif- Pr -
I
'[ /i ' .,' t J== 5
.? .->. ; .i. t9 y*{
^
s "m+- ;M'2y r &*s \ i e i 't b ;
- y M..,p._
g i-
- p. s .
- 4R u., i ,, t h.3 ..s g d, ._./_,\
*\, EL .tyh.. ,
l v~c--l',t?., . . . - ,. ). / " ' p'. >y L , ..... , ,s, nj a - > , a
, it_ ,
I .
- i 'r
.i./ l, G!,- 6/--
nt w c n ._. v.w FIGURE 4-9. PLAN VIEWS OF RHR EQUIPMENT VAULT AT THREE ELEVATIONS 4-96
I I @ l ._ _ _. - - - - - _ _ _... .-...w<.
,c_-_________. ;-
t l l e_,r.. . ....
\
l .
; 3._a ,\,
si
,s.i_.. ._,,. _,r. .-
7-.
.m . . .-
1 _ _ __ _ _ _ .w l b >-(. a, _+, ._\. . . .J.7et. J .
. n w.. .
e - ve
.. .s . . n o.;yn- e s l'
I ~' ! g g
..._J.U___....[\i.__________.....
. uu I
4.. . . . . , . .. _ _ _ _ _ ...... . . . _ _ _ _ _ _ _ _ . . . . . ,_ =.
.5 r1. w.--p g ,
c l l*2=!
.;. _=. . - . .
I
. i .*i!'*En% .l \?E :
44 i ei i
. i
['A [=F
.r va ,...w
. +g.. b1,. . j , -. . . . .s I
l 5: i lA
, :: ., g
-.c.-. =-
u,,e z m...
. en.. .
I
.- , . _=1 . -
i:1 ......._.~.;
, . a.u t1
@ =
i . - y I 1
. pe r:r. - 6-
*=, l ,.
v s a l = i a .s iu,mpi 1 m Ll 1 .e, gv_ .=. ,oi C . I ;
~m A=1 wuc 1 ,
9 gw 1 II
,.,T .
g3
.c
/ s __ i s/
.i q
=_
a w , , . . . l . ~ ., 1 m ~'?.. tl l
- r- i=[ .
.. o
+ \m -
- f .. .c r V
=
. A r' s.... r- ; = '.,%
j ' 7 p i"'i ~-+ f-*N '%. y E
$;: l. M l
l :. .
~m. .
!.\ m' i , , , , , , .
;e 6 i f"' .r; :-
ra m .
.a.
i - t 1 I Idj -- c 2 I : .m.3 w , i x.. _= - _ M_ . . , (J q .[t:d_ .f, C==*' w 3. ..:.
. .a .p. a a -
d, 5 ?j '
*:.'?*'$r
,, .W ,
, T. .J* ..31
. . e-l y e a. .
- a. .. . a I E t t* V AT ION C*a 9M 3 7 6C*/40
*A . A
- FIGURE 4-10. ELEVATION VIEW 0F RHR EQUIPMENT VAULT I -
4-97
24tB 2220 2228 19C8 1688 5 E des I N W
$1205 5
? E g g SEED SCO Se5 agg '
I fY-l V [# J ) 200 , 8 2 4 6 9 le 12 14 16 IS 25 22 24 TIMC IHOURSI FIGURE 4-11. PRIMARY SYSTEM PRESSURE PSIA (by Westinghouse) M M M M M M W W M M M M M M M M
O [- l O R R R R R R R EW M R R- R R. V 600 l 575 558 525 520 475 5 458 D ^V ^M g i e e a 425
- u dC3 575 -
558 y 525 - 528 L 275 0 2 4 6 9 10 I2 14 16 IS 20 22 24 TINC 189JR53 FIGURE 4-12. CORE WATER TEMPERATURE F (by Westinghouse) D
M M M 4 2 2 M 2 2 5 M
" - S I
A I M S _ P _ 6 1 E R U S M 4 S 1 8 E S R 0 P 0 21 1 0 6 T) Ne M C Es
' 1 t Mu i To T Rh 0
1 Ag P n Mi M Ot Cs e 9 RW E Py M Pb U( 6
\ .
\ 3 1 M 4 4 E
R e U G M I F j
"j 4
5 2 5 8 5 s 2 6 2 2 2 0 2 S I 6 1 4 I 0 M
;s5 5a5 M
M
?Eo m
P R R R G 4 2 2 R 2
/ 0 2
R /
- F E
/
S R I U T A R 6 R E
/ 1 P
M E T R 4 3 I S S A G) 6 e 21 Ts 1 Nu R E M Eo Mh N 0 I T Tg Rn Ai 1 Pt
' Ms R Oe CW s ,
S Ry Eb P( R_ P U l T N G 4 4 1 4 l
% 2 E
R U G I l F 8 8 0 , O l 0 s O 3 9 6 4 2 . e 8 O - 2 2 2 S 4 2 C - ' 2 l I 1 1 I f E:I8 =MS R n M R
?~S R
26 26 ( 24 22 C 2e Tr 7 is W 8
,16 8
p g l4 i E
= 12 e
iI g 15 y W'
\ .
s a 2 0 0 2 4 6 5 le 12 le 16 15 20 22 24 TINC IWUR31 FIGURE 4-15. VESSEL WATER LEVEL (BOTTOM CORE = 7.94) FEET (by Westinghouse) M M M M M M M M M M M M M M M M M M
M M M M M M M M M M M M M M M M M M M ssee sees # asee t- dees I S
?
W ssee 8 a Sete U
$ 25e6 A
2eee Isee leen see O g 2 4 6 9 je I? 14 16 IS 2e 22 28 TIME lHCJRSI FIGURE 4-16. GREATEST TEMPERATURE IN A CORE N0DE F (by Westinghouse) e
.SSE 6
.5BE 6 45E 6
/
/
/
.eDE 6 55C+6 sn E
r 5DE*6 / 5 Y
& E .2SE*6 8
s EE 8 2DC 6 15E*6
.lOE*6 50E*5 0
e 2 4 6 S le B2 34 16 , 19 20 22 24 TIME IMOURSI i i FIGURE 4-17. CAVITY t1ELT MASS LB (by Westinghouse) l l M M M M M M M M M m m m m m
l 1 n m__ n m_ n
/ 4 2
/ 2
/ 2 T .
E _ n 8 2 E F N O
/ I n / 5 1
T A R T E F 6 N E n / 1 P E
< T)
Ee
/ . ss 1
e Rs Cu No M , x M Oh Cg
) 2I 1
E n M Li I At n e l T I s Xe AW Yy Tb I( n S V A C S 8 n 1 4
. E R
U R G I 2 F n 4 2 2
. 9 6 E
- l
- e. .
2 8 2 2 1 1 I I [ 5ca5e Muu
!.$5 .
G n n n
.3O
- 7 1Il?I l
.65E 6
.60E 6 (y, E .55E 6 2
3 .50E 6
>- l
$ .45E 6 E
o .40E 6 3
.55E 6 ;
\
E .50E+6 h
? > L 5 3 .25E 6 E f g .20E 6 y .15E 6
.10E 6
.50E 5 O.
O. 2. 4. 6. 8. 10. 12. 14. 16. 18. 20. 22. 24. TIME (HOURS) FIGURE 4-19. SEAL LOCA FLOW RATE INTO VAULT LBM/ HOUR (by Westinghouse) M M M M M M M M M M M M M M M M M M
, i ;
i n S m
- S P
M S P U P M P M U U P Y n P I R H A R P S R S O + : * . 4 2 n 2 2 0 n 2 8 T 1 T E F ,
. E 6 F 1
)
n S L E 4 R V 1 U E O L)
. H e 2I Rs n 1 E
E u To Ah
. M Wg 0I n 1 T Ti Lt n .
Us Ae VW 8
/ y
.b n / 6 0(
2 4
/ i E
. R n / j/
4 U G I I
/ F
/ 2 n /
[ . 8 6 4 2 c
- . . . ' . . . . . .O 8 6 4 E 0 8 6 4 2 0 2 2 2 2 2 n
1 1 1 I 1 c dU3- eg3 3a$ n n n n
,l. o -
i r ,
M M i M M 4 2 T M L 2 U 2 A V T N 8 E 2 M P I 5 U Q M 1 E R H R 1 6 O M T E 4 S) Ae M 1 8 5
- R Es
- U Lu O Eo H Rh 2I g 1
C Sn f N I T Ai Gt s M e Ee l LW B Y 0y 5 Nb ( M F O N O 6 I T M C A R 4 F M 1 2 2 - 4 E R m q 8 U 8 7 6 6 4 5 2 I G I 8 o EwN2 F m m m aI am m N
M M M M M M M M M M M M M M M M 525 5 . 2,5 l 25 l 225 i s at
- 115 o
b
'5 .
g ! E m 125 I 575 95 925 5 8 2 4 g ' 30 28 22 24 TlHg s q qg, FIGURE 4-22. FRACTION F CS EL SE O RHR EQUIPMENT VAULT
M M M M 4 2 M 2 2 T L U A 8 V - 2 T N E M S I P I M U Q E 6 _ 1 R H R M 7 _ 4 O 1 1 T) 5 e R U 0 Es Su M 21 6 Ao 1 Eh Lg tr En Ri j f 8 i f Es t M 1 Te _ W r F Oy b 6 N( O M I T C 6 A R F M 4 3 2 4 M 2 E R U _ G 2 2 2 e l 6 1 4 1 2 1
' s e
6 5 4 5 2 e 8 I F M_ _ e b 5=WE M_ M, M M
.L. 5 M
~
M M M M M M M M M M M M M M M M eens l eena 1 i eel2 een d
= , ysees C 5 W
g sees seea 7 ese2 e J e 2 4 s e le i2 14 is is 2e 22 24 IINE IHOURSI FIGURE 4-24. FRACTION OF SR RELEASE TO RHR EQUIPMENT VAULT (by Westinghouse) _
ee5 l es27 7 ee25 8022 ee2 s sel7 . k' l B esis l C 5 N eel 2 e es /
., I i
Bees ese2 0 8 2 4 6 e je 12 14 16 IS 2e 22 24 TINC IHOURSI FIGURE 4-25. FRACTION OF RU RELEASE TO RHR EQUIPMENT VAULT (by Westinghouse) M M M M M M M M m m a e a e g g
M M M M M M M M M. M M M M M M M i 5 l 27s /
/
2s / 22s 2 / y 17s 5 I
- U is g b j w 5 gits e l I
ers e es e2s e 2 4 6 8 le 12 14 is 33 28 22 24 flHC IHOURSI FIGURE 4-26. FRACTION OF CSOH RELEASE TO RHR EQUIPMENT VAULT (by Westinghouse)
N RELEASE CHARACTERIZE CHARACTERIZE SOURCE PROBABILITY CATEGORY. SOURCE TERM SITE MODEL TERM (SUBJECTIVE 9,J = 1,2,...,6 UNCERTAINTY UNCERTAINTY IDENTIFIER WEIGHT)
^ 9 m .8
-BM .72 BEST BEST ESTIMATE ESTIMATE (B) (y)
.2 "IO CONSERVATIVE i (H)
[
.p.
1 n .8 SJ-CM .08 CONSERVATIVE BEST (C) ESTIMATE (M)
.2
.02 CONSERVATIVE (H)
- TREATMENT OF RELEASE CATEGORY S7 IS THE SAME EXCEPT THAT THE PROBABILITY WEIGHTS FOR BEST ESTIMATE AND CONSERVATIVE SOURCE TERMS ARE .8 AND .2, RESPECTIVELY, FIGURE 4-27. DISCRETE CHARACTERIZATION OF SOURCE TERM AND SITE MODEL UNCERTAINTIES
s
- 5. SITE ANALYSIS The site (or consequence) analysis calculations for this study were made using the CRACIT (calculation of reactor accident consequences including trajectories) computer program. The CRACIT program is described in detail in the SSPSA (Reference 5-1); therefore, only a brief summary of the model is presented in this section. Run setups used in this study were essentially the same as those for the SSPSA analysis with the exception of certain release category and site-specific assumptions that are discussed in this section.
The primary intent of this section is to present the results of individual CRACIT runs. Thus, this section is intended to characterize results that are conditional only on the occurrence of the individual release category being run. The combination of these results with the results obtained for the plant and the containment analysis to form frequency-weighted total risk curves is presented in Section 2. This section also describes the modifications made to CRACIT and associated output (postprocessor) routines to enable spatial evaluation I of dose and health consequences for different protective action strategies. This section describes the plotted CRACIT output obtained for individual release categories and presents examples. Results for all runs are summarized in Appendices A through D. 5.1 REVIEW 0F SSPSA SITE MODEL AND MODIFICATIONS FOR THIS STUDY The site model used in the SSPSA incorporated Seabrook site features including population distributions, meteorological data (from the site tower and other regional sources), estimates of evacuation trajectories I based on highway locations, and evacuation times. The CRACIT program uses these data to evaluate accident consequences for 96 randomly selected weather scenarios for each release category. In each scenario, doses are computed by taking into account the time-dependent plume and evacuee locations. Results of these scenarios are combined to form plots of frequency versus number of health effects. CRACIT features the ability .to treat wind direction and evacuation track changes; thus, it is referred to as a variable trajectory model. Options in this CRACIT calculation are used to account for many other effects of interest in this study, such as sheltering, delay time before evacuation, and evacuation distances, which are discussed in detail in this section. The site (consequence) analysis methodology for this study is the same as I that described in detail in the SSPSA (Reference 5-1, Section 4.5, " Site Model Analytical Procedure"). Section 12 of the SSPSA, " Site Consequence Analysis," provides further details concerning the consequence analysis methodology. Analytical details regarding the underlying models can be found in Appendix I of the SSPSA entitled " Site Model ." One objective of the analytical technique used in this study is to quantify the site matrix part of the master assembly equation that was 1 I 5-1 I 1325P120285
I described in Section 4.2 of the SSPSA using revised source terms. The elements of the site matrix computed in this study are the conditional frequencies of exceeding damage levels (see Section 5.2.3) for health effect components of accident consequences; i.e., acute fatalities and latent cancer fatalities. All results presented in this section are conditioned on the occurrence of an accident characterized by the " release category" being considered. The major differences between the SSPSA consequence analysis and this study are the revised source term characteristics resulting from plant and containment evaluations presented in Sections 3 and 4 of this report. There are also differences in the treatment of uncertainties described below. A feature of this study was the modification of CRACIT and its associated output (postprocessor) routines to compute and present the spatial distribution of risk for various protective action strategies. Thi s included saving additional information on disk for use by the postprocessor routines. This necessitated changes, in addition to source term changes, associated with the time and duration of release and the fraction of each isotope group released. The 96 meteorological scenarios and the evaluation trajectories are all identical to those used in the SSPSA. Changes to the CRACIT input compared with that used in the SSPSA are described below. e In the SSPSA, uncertainty in the consequence analysis was expressed g by calculating three sets of consequence analysis results (H, M, and g L for high, medium, and low, respectively) and assigning probabilities to each (see SSPSA, Section 12.4). For this study, it was determined that uncertainties could be adequately characterized by using two categories (M and H). The following parameters (summarized in Table 5-1) were varied between the two uncertainty categories: Delay time between warning and the start of evacuation. Latent cancer fatality health-risk conversion factors. Air and ground concentrations. (These quantities were varied by
. scaling the power level.)
The values selected for these parameters are discussed in Reference 5-1; however, a summary is presented below due to their importance to this study. The " low" category was deleted from the consideration of uncertainties in this study because it was judged that it would have little impact on the evaluation of protective action strategies. The assumptions of emergency response for the " medium" case allows for an extended delay of evacuation in up to 10% of the weather scenarios. The extended delay can be used to represent the impact on risk of severe weather effects or ineffective evacuation for other reasons, such as earthquakes. Delays of this type would be expected only rarely. 5-2 1325P120285
Emergency response assumptions for the "high" case reflect more extensive evacuation delays for every weather scenario and arbitrarily increase the doses by a factor of 2. For all cases, the assumption of no emergency response for 24 hours beyond the evacuation zone is pessimistic. Even in the absence of emergency response planning, relocation of the affected population beyond the evacuation zone could be expected before 24 hours. The shelter fraction of 86% (Reference 5-2) is only used for the runs designated as " shelter" cases. For these cases, sheltering was assumed between the 2-mile evacuation zone and 10 miles. Normal activities with minimal sheltering were assumed to exist beyond the evacuation zone (or shelter zone for shelter cases). I The variations in emergency response assumptions primarily influence early effects. Factors used to convert population doses to numbers of thyroid cancer cases were varied to represent uncertainty in latent effects. The conversion from population doses to numbers of fatalities from cancer other than thyroid cancer were similarly treated. Emergency response strategies do not have a strong influence on latent effects. However, the latent fatality risk factors were modified for the medium and high cases in this study to the same extent as they were for the SSPSA. A description of these uncertainty factors is found on page 12.4-5 of the SSPSA, o In addition to the uncertainties considered in the consequence analysis (medium and high cases), uncertainties in the source term were also studied for two categories, "best estimate" and I " conservative" (referred to as "B" and "C" cases). CRACIT input conditions for these cases are provided in Section 4. e Application of the uncertainty categories defined above and in Table 5-1 requires that probabilities be assigned to each consequence uncertainty group. For this study, the probability that was assigned I in the SSPSA to the " low" category was deleted and probabilities assigned, as shown in Table 5-2. e As a result of the plant and source term analyses described in I Sections 3 and 4, most release categories result in long duration releases that required multiphase (or multipuff) treatment in the , CRACIT calculation. In the SSPSA, it was necessary to treat less I than one-half of the release categories with the multiphase processing capability. In this study, more than 80% of the CRACIT runs required use of the multiphase release capability to obtain realistic results, o Population distributions near the plant were reviewed to ensure that there were no significant anomalies (e.g., people located in sectors that were over water). This review confirmed that the population distributions used in the SSPSA were appropriate (Reference 5-3). The only population change from the SSPSA was the deletion of the I 2,000 Unit 2 workers assumed to be within the first 1/2-mile in the direction of Sector 25. The focus of this study is directed toward i l 5-3 1325P112685 l
-= :.. .
i assessment of risk aversion for different evacuation and shelter ; strategies for the' general public. Therefore, it was appropriate to delete the temporary population from the analysis. Note that by deleting this population, the spatial distribution of risk from the site is more realistically determined. This deletion has no ' i ' significant impact on the conclusions presented in Section 2. The effect of this change on evacuation trajectories and timing in CRACIT was not taken into account because it was judged to be minimal. e In the SSPSA, the evacuation distance was always assumed to be 10 miles with shelter to 50 miles and normal activities beyond , 50 miles. In thi.s study, evacuation and shelter zones are varied. e In the SSPSA, the assumption was made that evacuees traveled to the b edge of the evacuation zone and then received an additional 4-hour E = dose. For this study, it was considered to be realistic to assume B e( evacuees would c'ontinue to travel beyond the evacuation zone. This :j is particularly true for the smaller evacuation distances studied g 7 (e.g.,1 or 2 miles). To maintain consistency in comparisons, the 3 _ same assumption was made for the 10-mile evacuation cases. f A summary of important CRACIT input parameters used in this study is a ,. provided in Table 5-3. . 5.2 CRACIT POSTPROCESSOR FUNCTION The following discussion describes the CRACIT postprocessor calculations ;
=
and resulting evaluations. Emphasis has been placed on computing and ; characterizing health risk as a function of distance (and evacuation 4 distance) from the containment; however, the conventional cumulative distribution of probability versus number of health effects (CCDFs) are also provided. Risk point estimates are summarized on spreadsheets, as - discussed in Appendix D. j g
't 5.2.1 ASSESSMENT OF DOSE AS A FUNCTION OF DISTANCE The frequency of exceeding whole-body dose levels as a function of _
distance, assuming no immediate protective acticn, was calculated for g j each re16ase category. Exposures were allowed to continue for 24 hours g 'm after the time of release. Doses due to long-term exposure after 5 reoccupation of land areas are not included in this study. These . calculations were made primarily for comparisons with the dose versus " distance calculations presented in NUREG-0396 (Reference 5-4) used - primarily for developing emergency planning strategies. The 1, 5, 50, r and 200-rem whole-body doses were included in the computation. The g - calculation proceeded as follows: 3 , e For each scenario, the whole body dose exceeds the selected dose i level at any one of the population grid distances in CRACIT (see Appendix I of the SSPSA for distances), a counter is incremented for that location. These occurrences are then accumulated over all scenarios for each e evacuation distance and divided by the total number of scenarios to - determine frequency of exceeding each dose level at each distance. ; _= 5-4 - 1325P112685 J
e Since a dose in excess of the given level in any direction will ! increment the counter for the given distance, the results are j independent of direction. ' I e Doses are only evaluated at locations on the grid where at least one person is located in the population table. The results are not population-weighted; i.e., results are not related to the number of I l people who receive a dose at or above the given dose levels. Plots of dose versus distance are provided in Appendix A. Doses in these plots assume that residents take no protective action for period of 24 hours after the release starts. An example of a dose versus distance plot used for screening purposes is shown as Figure 5-1 for a typical release category, which was found to have a signi.ficant contribution to the risk of early health effects. Dose versus distance curves cannot be used individually for risk assessments; rather, they should be weighted by the frequencies attributable to each release category and summed as described in Section 2. Although thyroid dose and its effects were computed in this study, thyroid dose results are not presented. In making risk assessments, it is customary to represent the effects of thyroid dose in terms of thyroid cancers. Five percent of the thyroid cancers were assumed to be fatal and were added into the total cancer health effect. Also, the thyroid doses have reduced risk importance in this study due to the considerable reduction of iodine as compared with the SSPSA in the source terms described in Section 4. Thus, th6 assessment of protective action strategies in this report are based primarily on calculations of whole-body dose and early fatality risk. 5.2.2 EVALUATION OF RISK AS A FUNCTION OF DISTANCE Computations and presentations of dose as a function of distance as discussed in Section 5.2.1 should not be used alone to characterize the risk of health effects. To compute risk, the number of individuals affected, as well as the accident occurrence frequency, must be accounted for. Additionally, the spatial distribution of risk must be computed for use in evaluating the need for, and spatial extent of, protective action strategies. It is very important to note that the " distance," as presented in all CRACIT results, is the initial location of the resident. Some dose may be received at a location different from the initial location during evacuation. Thus, the doses (and health effects) reported are for the residents at the calculation grid distances assumed in CRACIT. Of importance to this study is the spatial characterization or quantification of risk that could potentially be averted. This is done by computing a set of curves that depict the risks outside each distance in the distance grid for the various dose and health effect categories ( assuming no immediate protective actions are taken. Values for each dose and health effect category in these tables are computed by CRACIT postprocessor routines that use data from the CRACIT output files and the f 5-5 1325P121685
l l population file to produce computer plots. Calculations of risk versus distance are made as follows: o CRACIT runs a wide variety of results that are written to a disk file for each meteorological scenario as a function of position on the CRACIT population grid (32 directional sectors by 34 distance segments). These data include the fraction of the population in each grid element with each health effect index (e.g., acute fatality). The maximum dose incurred by a person in that element is also included. All results are tabulated by the starting position element for a given person since, in the event of evacuation, a person could receive dose from each of several elements along the evacuation route. e For each weather scenario, the risk versus distance calculation starts by multiplying the population by the fractional risk for each health effect or dose index for each position element. e These results are summed for all directions at each distance, weighted (multiplied) by the meteorological scenario frequency,* and summed over all weather scenarios. e The total risk for a given health effect index is found by summing over all distances. The risk outside a given distance is found by summing the number of effects for all sectors outside that distance. e The fraction of total risk as a function of distance is obtained by dividing the risk at each distance by the total risk for that index. The risk versus distance curves have particular significance for the cases that assume n'o evacuation or sheltering. In this case, the curve g for a particular index shows the residual fraction of risk that would g remain if an instantaneous and perfect evacuation were performed out to each distance. In particular, if a given curve drops off sharply beyond a given distance, it means that relatively little is to be gained in mitigating that effect by evacuation beyond that distance. The example plot shown in Figure 5-2 is for a typical release category. In this example, a significant reduction " knee" occurs for acute fatalities well within 2.0 miles. Such observations are useful in the screening of results for individual release categories; however, the complete risk " story" is not available until all release category results have been frequency-weighted and combined, as is done in Section 2. Results of the risk versus distance calculations, assuming no evacuation for 24 hours, are presented in Appendix B for each release category. It is important to note that these plots represent the risk at various distances, and not the risk averted, using different evacuation distances. This point is clarified in Section 2.
*In the CRACIT Monte-Carlo scheme, more severe scenarios are chosen more often, but lower frequencies (weights) are assigned in order to increase the accuracy of the results.
5-6 1325P120285
I 5.2.3 CONDITIONAL CUMULATIVE DISTRIBUTION FUNCTIONS The most common way to express risk in consequence analyses is through the use of the CCDFs, which are tables or curves representing the probability (based on the number of weather scenarios run) versus the number of effects (e.g., acute fatalities) conditional on the release. A typical CCDF is illustrated for a single release category in Figure 5-3. The CCDFs are generated in CRACIT as described in Section 12 of the SSPSA. Separate distributions are computed for each release category, uncertainty level, and mitigation strategy. CCDFs are provided for all CRACIT runs in Appendix C. Total risk curves that use individual CCDFs to account for the frequencies of occurrence for each release category are provided in Section 2. The methodology uses the CCDFs in the same way as described in Section 13 of the SSPSA. 5.3 SENSITIVITY ANALYSES AND OBSERVATIONS An evaluation of the results has identified several observations, which are discussed below. e Risk is considerably lower for early health effects compared with equivalent cases in the SSPSA due to the considerable reduction of particulate and iodine releases in the revised source term. e In several cases, additional reductions in the particulate source term cases would not significantly reduce doses or the distance at which " knees" occur because doses are primarily the result of plume shine doses from the noble gases in the plume. Noble gas releases remain at levels essentially the same as for the original SSPSA release categories. Noble gas doses can be reduced by longer delay times and by using the multipuff model. Further reductions in the particulate source term can, on the other hand, reduce latent effects; mean early effects; and high consequence, low frequency tails of CCDFs. e Comparisons of runs with the following mitigation assumptions showed little difference: Evacuation to 2 miles, normal activities beyond. Evacuation to 10 miles, normal activities beyond. Evacuation to 2 miles, shelter between 2 to 10 miles, normal activities beyond. This is explained when it is considered that the shielding factor for plume shine decreases only a small amount (from .75 to .5) for the shelter cases. This is in sharp contrast to the change from a shielding factor of 0.33 to .08 (normal activities to shelter) for ground shine (due to deposited particulates and iodines). The shift in the importance 5-7 1325P112685
of sheltering from classic WASH 1400-based analyses (Reference 5-5) is due to the reduction of particulates and iodines that contributed the ground shine dose. The noble gases (which now contribute a large fraction of the plume shine dose) dominate the early effects.
5.4 REFERENCES
5-1. Pitkard, Lowe and Garrick, Inc., "Seabrook Station Probabilistic Safety Assessment," prepared for Public Service Company of New Hampshire and Yankee Atomic Electric Company, PLG-0300, December 1983. 5-2. Aldrich, D. C., D. M. Ericson, Jr., and J. D. Johnson, "Public Protection Strategies For Potential Nuclear Reactor Accidents: Sheltering Concepts with Existing Public and Private Structures," SAND 77-1725, February 1978. 5-3. Lee, Dr. S., Yankee Atomic Electric Company, letter to K. Woodard, Pickard, Lowe and Garrick, Inc., July 29, 1985. 5-4. Collins, H. E., et al., " Planning Basis for the Development of State and Local Government Radiological Emergency Response Plans in Support of Light Water Nuclear Power Plants," prepared for the U.S. Nuclear Regulatory Commission, NUREG-0396, December 1978. 5-5. U.S. Nuclear Regulatory Comission, " Reactor Safety Study: An Assessment of Accident Risks in U.S. Commercial Nuclear Power P1 ants," WASH-1400, NUREG-75/014, October 1975. I I 5-8 l 1325P112685 l
m M M TABLE 5-1. VARIATION OF PARAMETERS IN CONSEQUENCE UNCERTAINTY ESTIMATES Parameter Medium Case High Case Emergency Response Incremental Delay One in 90% of Weather Scenarios Two in 90% of Weather Scenarios in Evacuation (hour)(a) , Four in 7% of Weather Scenarios Six in 7% of Weather Scenarios Six in 3% of Weather Scenarios Eight in 3% of Weather Scenarios Fraction of Population 0 (except for " shelter" cases. 0 (except for " shelter" cases, Sheltered between Evacy which are 0.86). which are 0.86) Distance and 10 Miles tgion Ground Dose Period for 24 24 Population beyond Evacuation Zone (hour)ICI Dose VariationIdI Used as Calculated Increased by a Factor of 2 Conversion Factors for Probability of Latent Effect versus Dose
- ui
& Cancers Other than Thyroid Cancer 2.00-4 5.00-4 (cancer fata i per man-rem)gies Thyroid Cancer 1.34-4 4.02-4 (cancer case per man-rem)gg
- a. Delay of entire evacuee population after warning is given to government authorities by plant personnel. 1001 evacuation is assumed.
- b. Fraction not sheltered is assumed to pursue normal activities. Doses are assumed to be reduced to a limited extent by structures. The fraction sheltered is assumed to have doses reduced to the extent attainable by taking shelter in the basement of a single-family house, allowing for some dose accumulation due to infiltration of airborne material and due to possible exposure during relocation.
- c. Period between the beginning of exposure and relocation to an unaffected area. During this period, dose accumulates due to exposure to radiation from material deposited on the ground and other surfaces.
- d. Doses were modified uniformly for all locations.
- e. High case is effectively equivalent to BEIR III, linear / relative risk model. Medium case is effectively equivalent to BEIR III, linear-quadratic / relative risk model.
- f. High case effectively treats I-131 as equal to X-rays in thyroid cancer induction. Medium cases effectively treat I-131 as one-tenth as effective as X-rays, as was assumed in the RSS.
NOTE: Exponential notation is indicated in abbreviated form; f.e., 2.00-4 = 2.00 x 10-4 1283Pil2085
l I I I TABLE 5-2. CONSEQUENCE ASSESSMENT DISCRETE PROBABILITY DISTRIBUTIONS I Consequence Uncertainty B Damage Group Probability l Index High Medium Early and Latent 0.20 0.80 I Effects
\
l l I ;
)
l l l l
)
l l I l l j 1283P112685 5-10 l
M M M M M M M M M M Table 5-3. ADDITIONAL FARAMETERS VARIED DURING CRACIT RUNS Parameter Assumed Values Fraction of Evacuation Zone Population Evacuating 0 for Nonevacuation Cases; 1.0 for All Evacuation and Shelter Cases Fraction of Sheltering Zone Population Sheltered 0.86 for Shelter Cases; O for All Others Maximum Distance of Sheltering Zone (miles) 10 (for shelter cases) Last Evacuation Element Stay Time (hours) 0 Starting Distance Segment Number 1 Ending Distance Segment Number 34 Maximum Evacuation Distance Segment Number 0 = No Evacuation, 2 = 1 Mile, 4 = 2 Miles, 15 = 10 Miles Cloud Shielding for Evacuees 1.000+00 7 C Cloud Shielding for Normal Activities 7.500-01 Ground Shielding during Evacuation 5.000-01 Ground Shielding for Normal Activities 3.300-01 Cloud Shielding for Sheltered Nonevacuees 5.000-01 Ground Shielding for Sheltered Nonevacuees 8.000-02 Power Level Fraction of 3,300 MWth 0.9997 for "M" Runs,1.999 for "Hi" Runs Ground Dose Exposure Time for Nonevacuees (hours) 2.400+01 NOTE: Exponential notation is indicated in abbreviated form; i.e., 1.000+00 = 1.0 x 100 ; 7.500-01 = 7.5 x 10-1 1283P112085
18 : 1 E u Ua s .
-1 R 18 .. .18
-t M
E z 5 E -2 -2 m 18 --
, 18 r e a
iS L= Ei -3 -3
>.18 __
, ,18 C
a
% x288.88REN a:
+58.88 ret 1
@ 45.88 REN a: 1.88 REN n.
4 4 18 ' i 18 18-1 ' 'i S 8 ' 't S t iS2 DISTAMCE (NILES) FIGURE 5-1. DOSE VERSUS DISTANCE CURVE FOR RELEASE CATEGORY 56B-H FOR NO IMMEDIATE PROTECTIVE ACTION FOR 24 HOURS M M M M M M M M M M
I I l
\
l I W (13 O WW I CD w .M cn c w C W Q CD > > tn tn Q Q Q Q M E5
- O H w: bl w wl m! QQ lC3o l' cm a I J
MI Ga Q alwlm g G Q Z JlW,Wa.a. Q ' H 4 7lHl3 z
. Gs ZuQ* O 2 *E O
W l i 6 M *.lH6 l* ElE3 3 3 M w w m
- wl E >-
H H,
- w l m l w Ew "' W 8 E
=> = : H . . l Ex - O U U ; J I tJGlW EEPl -
G C.D G e LD I LD e qmp td 9 W
* <C to o
{
.i J
w I
- 5 WZ cc O l e wm I f
.h #
.T 8
JW wU ww
.f
- O>
0 La. - 1
& H 8
. HU
, OW
{' ' JH I
./
=
ag wm j '
-> Ew I
( l' e to W
.J CC W H CC mm
. 9 HO I l I
/
i I j
.i-J m l
-5 u
E Ow mE 3~ M MO I I i
/ l l
WH G
.tE.P (13 (D o wZ MC m
.! / Mk I' .. Ill . x l /),/
./ l 4
1
. ./ /
s
, ./ .. / ,s
##s -
4'1
- w x
/* -
e Sa massaa W "W n e #. f /a %....
.s * *. ** a* ~" ;*.
..... . y,7. .
155 Lt. w e u u u 5 I I Et3 CD b. m tg y 43p (E3 e CD. N. LO. Y. N.
, EEP e
M E O O E @ g E i l g 33NUISIG H30ID 301S100 XSIH 30 NDI130H3 5-13
I 4 4 i 0 188 ,j g3 18 ..181 . . . . . .182 ,,j g4
'- - , , .,.j a 5
.............;a.............,'+*4,s.N. -
g - i' a r m - ' .-.., \.\ : ACUTE FATALITIES
$g ,
., g A..C.U.T.E...I.M.J.U.R.I.E..S..
E,d 18 , _18 , o m
'N -
T- . TOTAL LATENT EFF du u o i-
\-
X w - w u ,
\ -
m=m i* k
= -2 .
- 2 u
z o:' 18 -- -
., ) . _. .
18 - g8 i \ i T w 5 c 5a '\ : a - _a M - E ad 18 -3 -3 gw .-. .18 sJ -: . o - 5
-4 -4 1g ... .. .. ... ... tg 188' "'181 ' "182' 'tg3' "'t g 4 ' "'t g 5 ' 'tg6 NUMBER OF HEALTH EFFECTS FIGURE 5-3. HEALTH EFFECTS CCDF FOR RELEASE CATEGORY S6B-H, NO EVACUATION M M M M M M M M M M
I l I I lI APPENDIX A DOSE VERSils DISTANCE CURVES I l l I I - I I I I I
I APPENDIX A FIGURES Figure Page A-1 Dose Versus Distance Curve for Release Category S1-CM for No Immediate Protective Action (Run Number 137) A-2 A-2 Dose Versus Distance Curve for Release Category SI-CH for No Immediate Protective Action (Run Number 177) A-3 A-3 Dose Versus Distance Curve for Release Category S2-BM for No Immediate Protective Action (Run Number 138) A-4 A-4 Dose Versus Distance Curve for Release Category S2-CM for No Immediate Protective Action (Run Number 139) A-5 A-5 Dose Versus Distance Curve for Release Category S2-CH for No Immediate Protective Action (Run Number 163) A-6 A-6 Dose Versus Distance Curve for Release Category S3-BM for No Immediate Protective Action (Run Number 140) A-7 A-7 Dose Versus Distance Curve for Release Category S3-CM for No Immediate Protective Action (Run Number 141) A-8 A-8 Dose Versus Distance Curve for Release Category S3-CH I A-9 for No Immediate Protective Action (Run Number 156) Dose Versus Distance Curve for Release Category S6-BM for No Immediate Protective Action (Run Number 131) A-9 A-10 A-10 Dose Versus Distance Curve for Release Category S6-CM for No Immediate Protective Action (Run Number 178) A-11 A-11 Dose Versus Distance Curve for Release Category S6-CH for No Immediate Protective Action (Run Number 176) A-12 A-12 Dose Versus Distance Curve for Release Category S7-BM for No Immediate Protective Action (Run Nymber 142) A-13 A-13 Dose Versus Distance Curve for Release Category S7-CM for No Immediate Protective Action (Run Number 143) A-14 A-14 Dose Versus Distance Curve fer Release Category S7-CH for No Immediate Protective Action (Run Number 158) A-15 I I - I I 1339P120685 iii
APPENDIX A DOSE VERSUS_ DISTANCE CURVES I This appendix contains plots generated from the output of the CRACIT computer code for dose versus distance. Each plot corresponds with a particular release category source term case (B for best estimate, C for conservative) and consequence model case (M for medium, H for high). All plots are conditional frequency of exceedence curves, given the release. I I I I A-1 1324P100385
I i la - If
=
'i
! -1 -1 5 10 .. _10 W
Z 5 2 e 2 -2 e la _-._ ..10 S
> 0 E>
Ei -3 -3 3 10 19 a E x200.00 ret 1 cc +50.00 ret 1
@ 45.00 ret 1 m 1.00 REN n-
-4 4 10 e i 10 10-1 '100 't a l 102 DISTAtlCE (NILES)
FIGURE A-1. DOSE VERSUS DISTANCE CURVE FOR RELEASE CATEGORY SI-CM FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 137) i ' uma uma uma sus ama aus amm uma sus amm
M M M M M M 10 ; 1 3 8 m U s
-1 -1 R 10 - _ 10 a
8 5 2 o -2 -2 m 18 --- -- 18 5 S U 30 d b -3 < >
-3 3 la -- 10 s
3 d 2 x200.00REN cr
+50.00 REN a::
a5.00 REN 1.00 REN n.
-4 -4 la i i 10 10-1 '100 't a l 102 DISTANCE (NILES) i 1
FIGURE A-2. DOSE VERSUS DISTANCE CURVE FOR RELEASE CATEGORY SI-CH FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 177)
I 18 : 1 E
-1 -1 E 10 __ __18 M
a
\ k to -2 -2
, 10 _._18 5
5 "i
?
4 6 5 -3 -3
>. l a - - ---10 a
2 cr x200.00 ret 1
+50.00 ret 1
@ 5.00 RE cx 1.00 RE
- a. _4 _4 10 e e la 10-1 '100 '181 Ig2 DISTANCE (NILES)
FIGURE A-3. DOSE VERSUS DISTANCE CURVE FOR RELEASE CATEGORY S2-BM FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 138) e e e
M 1 3 18 : If S m U s
-1 -1 R 10 __ __ta W
8 5 E o -2 -2 e 18 -r -7 18 5 0
- u d
?
m m
*3 -3 j 4 9g
> 9g 97 -r U;
j 2 a x 200.00 REM
+50.00 REM L.
@ 4 5.00 REN
= 1.00 REN i -4 -4 la e i 10 10-1 '198 't a l 102 DISTANCE (NILES)
FIGURE A-4. DOSE VERSUS DISTANCE CURVE FOR RELEASE CATEGORY S2-CM ! FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 139) l ) i
/
3 18 : _ ff S
=
o s
@ la' -- - 10' '
w _ E N c2 z: 5 5 -a .a m 18 -- 18 S ? m U j Ei -3 -3
> 18 -:- -- 10 b
2 x200.00 ret 1 cr +58.08 ret 1 L
@ 5.00 ret 1 cx 1.00 ret 1 t'-
-4 -4 18 i . 18 10-1 '100 '181 182 DISTAtlCE (t1ILES)
FIGURE A-5. DOSE VERSUS DISTANCE CURVE FOR RELEASE CATEGORY S2-CH FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 163) sum um amm
M M M M M ^ 3 10 . If c, 1 m U
-1 -1 R 10 -- - 10 m
8 5 , E e -a -a
! e la -- - 10 l 5 i 8 F
u ! Ei -3 -3
> 10 -- - 10 a
2 cc x200.00REN 50.00 REN
@ +5.00 Rett 4
1.00 REN
=.
n -4 -4 la e ' la 10-1 '100 't a l 102 l DISTANCE 01ILES) FIGURE A-6. DOSE VERSUS DISTANCE CURVE FOR RELEASE CATEGORY S3-BM FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 140) 1
1 18 g if E
-1 -1 R 10 __ __16 N
E 5 : E w -2 -2 m 18 --
-- 18 5
S w
> M .
En " Ei -3 -3 w ie __ __is U, 2 cc x288.00REN
+50.00 REN
@ 4 5.00 REN
= 1.00 REN n- 4 4 la i ' 18 10-1 '100 ' 't a l 102 DISTANCE (NILES)
FIGURE A-7. DOSE VERSUS DISTANCE CURVE FOR RELEASE CATEGORY S3-CM FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 141) ums mumi aus uma ums - aus sum nas sus
1 J l n n n R R F1 R R R. R R R R R n o n F ta' _ 18 E
'iO 5
-t -1 E 18 .. ..18 W
8 m 5 5 -2 -2 m 18 -- - 18 5
> 0 a 0 ti Ei -3 -3 3 18 __ __td C
_a 2 a x288.08REN
+58.98 REN
@ 4 5.08 REN m
- o. et.88 REM 4 4 16 i ' 10 is-1 'i es ' 'st i tea DISTAMCE CNILES)
FIGURE A-8. DOSE VERSUS DISTANCE CURVE FOR RELEASE CATEGORY S3-CH FOR NO IEEDIATE PROTECTIVE ACTION (RUN NUMBER 156)
18 1 3 , S m U o 3
-1 -1 E ta ._ _ 18 W
8 z 5 5 -2 -2 a 10 _ 18 5 0
> u a
o 5 Ei -3 -3 s ta __ __18 m" x200.00REN i e +50.00 REN
@ 5.00 REN
=.
o 1.00 REN 4 4 10 ' e 18 10-1 '180 '181 1g2 DISTANCE (NILES) FIGURE A-9. DOSE VERSUS DISTANCE CURVE FOR RELEASE CATEGORY S6-BM FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 131) , uma mas num uma sus uma uma ame uma um ums
R R R R R R R R R R R 7 M R. F l 3 18 - If S
= .-
w 5
-1 -1 s la __ __18 m
8 z: 5 5 -a -a e ta __ __i8 5
? =
M w
' {;
Ei -3 -3
>.ia __ __t8 U
a 2 x288.88REN a: # 58.88 REN
@ 5.88 REN =.
a .t.88 REN 4 4 18 i
~= 18 18-1 ~ 's S S 'St i iS2 DISTANCE (r1ILES)
FIGURE A-10. DOSE VERSUS DISTANCE CURVE FOR RELEASE CATEGORY S6-CM FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 178) l
10 : 1 3 - M
! -1 -1 5 18 __ _18 N
E z: 5 E -2 -2 m ie __ __18 5 > 0 a 0 X " N Ei -3 -3
> 16 . _ 18 b
a 2 a x200.00 Rett
+58.00 REN 4
@ 5.00 ret 1 m 1.00 REN
- o. _4 4 10 i e 18 10-1 188 ~ '181 182 DISTANCE (t1ILES)
FIGURE A-11. DOSE VERSUS DISTANCE CURVE FOR RELEASE CATEGORY S6-CH FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 176) mas sum uma ums uma umm uma mas amm aus sums uma amm uma uma sus ums
7 n n n n n n n n n n n n n n_ n n n m r" 18 _ 1 E 5
-1 -1 E tt -- - 18 M
8 m 5 18 g 18-- -:
?
t; M
~
1
-a t, -a 18 __ - 18 3
s U d
,288.88REN cc +58.08 REN 8 4 5.08 REN ix
- a. .t.88 Rett 4
4 18 .i i 18 ta-t GE i ' 't G 1 182 DISTAMCE (NILES) FIGURE A-12. DOSE VERSUS DISTANCE CURVE FOR RELEASE CATEGORY S7-BM FOR NO IPMEDIATE PROTECTIVE ACTION (RUN NIMBER 142)
1 18 - 1 E \ 5 5
-1 -1 5 18 __ __10 N
E z 1 5 I E 2 -2 cs l a --- - 10 5
> S
, a 0 4 6 " 1
$5 -3 -3 s ta __ __ts C
_a E x200.08 Rett a +58.80 ret 1 8 cx 5.00 REN 1.00 Rett a 4 4 10 * ' 18 te-1 't aa 't e t tea DISTANCE (t1ILES) FIGURE A-13. DOSE VERSUS DISTANCE CURVE FOR RELEASE CATEGORY S7-CM FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 143) num ame uma amm uma muu muu aus aus uma sus mas amm
M M M M M M M M M M M M 18 5 tk
$ ~
Z 1
-1 E 18 ,-- ~~
M Z
@ ~
. E -2 a 18 - -,te 5 ' S
? u
\ G l 6
% -3 -3
! ,_ t a -- _ 18 D ,! d
' * *200.88 REli 56.06 REM i
- a. 4 11:31 li!I 4
10 . 18-1 '188
'g g t ,;!
DISTattCE (MILES) FIGURE A-14. DOSE VERSUS DISTANCE CURVE FOR RELEASE CATEGORY S7-CH FOR N01.'ffEDIATE PROTECTIVE ACTION (RUN NUMBER 158) I
I ) I I I l l APPENDIX B. HEALTH RISK VERSUS DISTANCE CURVES I I I lI I I 'I
~
APPENDIX B FIGURES Figure Page [ B-1 Health Risk Versus Distance Curves for Release Category S1-CM for No Immediate Protective Action (Run Number 137) B-2 B-2 Health Risk Versus Distance Curves for Release Category _ S1-CH for No Immediate Protective Action (Run Number 177) B-3 - B-3 Health Risk Versus Distance Curves for Release Category l S2-BM for No Immediate Protective Action (Run Number 138) B-4 B-4 Health Risk Versus Distance Curves for Release Category p S2-CM for No Immediate Protective Action (Run Number 139) B-5 B-5 Health Risk Versus Distance Curves for Release Category S2-CH for No Immediate Protective Action I (Run Number 163) B-6 l I B-6 Health Risk Versus Distance Curves for Release Category S3-BM for No Immediate Protective Action (Run Number 140) B-7 Health Risk Versus Distance Curves for Release Category B-7 S3-CM for No Immediate Protective Action l (Run Number 141) B-8 i B-8 Health Risk Versus Distance Curves for Release Category l I B-9 S3-CH for No Immediate Protective Action (Run Number 156) Health Risk Versus Distance Curves for Release Category B-9 S6-BM for No Immediate Protective Action (Run Number 131) B-10 l B-10 Health Risk Versus Distance Curves for Release Category S6-CM for No Immediate Protective Action l I B-11 (Run Number 178) Health Risk Versus Distance Curves for Release Category B-11 S6-CH for No Immediate Protective Action I B-12
. (Run Number 176)
Health Risk Versus Distance Curves for Release Category S7-BM for No Immediate Protective Action B-12 (Run Number 142) B-13 B-13 Health Risk Versus Distance Curves for Release Category S7-CM for No Immediate Protective Action l B-14 (Run Number 143) B-14 Health Risk Versus Distance Curves for Release Category S7-CH for No Immediate Protective Action l (Run Number 158) 8-15 I I iii 1339P120685 t
l APPENDIX B HEALTH RISK VERSUS DISTANCE CURVES l I This appendix contains plots generated from the output of the CRACIT computer code for health risk versus distance. Each plot corresponds l I with a particular release category source term case (B for best estimate, C for conservative) and consequence model case (M for medium, H for high). All plots are conditional frequency of exceedence curves, given the release. l l I l I I 1 I l l I l I l . I I I lI I I 1324P100385 B-1
m z m o a WW o o 4 m
>= Q Q tip a w w o - >- >- >-
w .w c e o i- w:
- w. -l wi
-: mgo.cl c .o e w
w ~
.J l a
_J
" aial5,swl.o
.e.o
_J I d' d' w w im ys vm~
- l 5 !.. ' w*l r ' 5cl5 5 w me w im c. c
- W
=> m:;6.e
- m=l i W-l a: w wi m
5wJE m U U lC1W LaJ D G G . , J l OJ LD l LD ** s g dZ ZZ w a0
= wM j- #
. v l D s i e wZ l
e
/ f
- e. >O gw
,f *
. 9
- DH OU I <
)
f J *
/ f W
M w uW
. Z>
s
- s gm e' ( .N '
p-- 6.-
) DO l ww I. !
e CH o
- ma J.
I
/ / l l
.g, .
am m f
- I 477 W
% L&J L&J l=
.J
/
/ l
[ ll w* w~ E J ><~ MC mW i[ l# . l I V~y
;l i .
m I
,f ! -
MW
==
WO aZ 1/
/ /
7 j i
-a- m-
=a w
/ / ) .
// / / ll ."'
'T w
..' , .i
/ ,s' ****. ' W p
s #'...=*_
/j w W a
,,..=**#
",# G L, - ~ ~
v,Q....-c' <7,_;;m;.-:.c......:
- -~~
63 (D b m og 9 g W M Po LA T W
*
- e o e N. , e O O E E g W 33HUISIG N3 AID 30ISinD MS!H JD NDI130H3 l
l ( l B-2
I I ' I w
=
=
u O Ww I C tn en &
> O O m c o en a w >.
w m o - > >. e
- w ..-w l .g oc ol ca oct o
>- m :. ~ . w i . e
- w a - a l o i, m
.:..o a wl .
,w' r_
es
"! al 5: d ut 2 5 !D-" rj w nl5g 5l e w ME I
w W: = <w wm l . s m.w:l c.wi. r
= = : .w..i U
a a.JI.!.I. U: 4 cu i m i m -
-.i x Wm ez Mz
~
om ( <
.! J m
j - f: .. wz
>o e e-e
.I
. 1 : . ,- . as j / UU w
[' . yw i Jl z> qw l
) .cu >->-
. / -
ma ww I' : f I :.: .. og ma ll
/ l ::
I#l . ( ) J
.J*n tn mA m
M LAJ W LAJ f-
/ :l:
l l'/ i
/ / :
=eo r
w we mw l
/ l ~~u -
ak -
/ / .
=
-e I
/ ! .2 d=
>I
/ /
l == ugw
. /
lj
- )
/ . .
I / .- .*,.
*I . ,/ ..- /
I 9 e
#/ 12...., #
/ -
7* _............ -, :: g e p[g,, _gg - y, . l M, i e i i g ED f.O == t% t's (7) y . tD CD Lf3 Y C4 ,. . e N. ,
* .e e .s .e . .e 33NB1SIO N3010 301S100 MSIN 30 NOI130HJ I
I I B-3
I w vs ca o w w e O 173 tri N
> Q Q
= 58W>> a 5 l al aua' g t: W,q"m j p! "!j m = = w I e
i- --2 a i d ".h o" , em um a g si s_ l' 5: 5 l o 5 Fs jEwj E 5 5 g3 t E w '-* me at w
= = Wz *! .ac t wi" el'E' "j!E o
C u n a i .j ID a . .J l tu
=r I LAl *=
el g wm gg W
- NZ up om
* - uM
./ i e M
wZ
/ l W >Q s
- ge
,/ .v
- or vo
[ t
- W
" O uJ
/ a a2 E
./ ,
t;t;
-w a
! W O j l
.=, ,
m8 5" l m
=
ww
. wr l
d ><
, 3-2 u8
/
mw
- :r E
/ #
l -E U z *
/ -
-m Ee
/ Q aZ g
i / : =e l l -
/ ..
]5 iy l l / /* /
/ /
~ ~~.
E g E
-a y/_s '
h d [ p .* # ',, _ ,.. -- -...
**a* ~ " " ** *'",.....=*
y
~
E
,.y _
- E i
W W
=
b M m y e
. g 43 M Pm U3 T M W
& Y 33uuksta u3nt0 301c1no xsin 30 not13eni
( I i B-4
I I I w
=
r u l o w w 8 o (73 fn N o o M '
> c o tn a w m o . >. > >-
w .- cs e O H w: w.- sl mI wl ol en o o c: o as e w
. -a ag w gm a -:. o, a w E?! al 5i 3 w,d'
" <O
=
e um~ 5i" i ri w
# w:HIw'rl5
=
**wl=iw.m.c' 5 w
ME
<w
= =: .e 3 ml w: - w&
u a u:ale
- a. u>l m - el. -
wo tu e mZ e EZ
= Os
. t ** LM i N so
/ e e
WZ
>O I *
./
I
.I I
I t 9 W
- E *-*
DH uu W uW aC W
/ 4 Z>
/
/ -
;$ ~
mu
-w
- O
*
- mm
./ om
> 4. n
. w (n & I.LJ I
d w wH
> c:!"
./ / ./ w m.I -
E MC 1
* .* MW
/ w ;;;E
.I > .I M a x
/ ep HO l
I l
./
/
/
l i .to. w We m aC wm IO aZ W [
/
/
] e W T
/
/ ..- -4 a j
- s s ** s =
y
./ , / . .
' 's ' . .
- w w
,// ""... a' " "
tu
* ', . . . . * ,-- . ---r. rv r Qs,_, .g. r- ,4* _.,".----- ------ ------ -~
I tD ED e CD e h s h W3 EM g-e tF) (4 o y e e e
" N O @ m e g a 33NU1SIO N30!D 30I5100 MS!H JD NOI130HJ m
t l 5 B-5
i 1.88 g a.ss- i, '. \ %s, 5 D\ . ' ~., h j \i ' -~.
~
o s.7s_ gg , .s,
'%- ACUTE FATALITIES y ( ; ,
G 1
- [ \ s ACUTE IMJURIES 0.57- l '
\ \
g i J *.} N - LATENT FATALITIES E, i.j k,
\, 2_8_8_R_E_N_U_H_O_L E_ _B O DY D O S E o 8.43- *
.) -
, s 58 REN UHOLE DOSE i
?
cn x 5 I :t
- l i
\
\ s 5 REN UHOLE BODY DOSE I I} - \ t REN UHOLE BODY DOSE g 8.29.
*l '\ -- - -
5 ' il '.
- N 8.14 _ .
= ;\ .
i., y \. - y g,gg l , N. - - - - _~_ 7 2'__ -- ~ .,._ a.es 2.se 4.es s.se s.es te.se 12.se 14.se ts.es DISTANCE (MILES) FIGURE B-5. HEALTH RISK VERSUS DISTANCE CURVES FOR RELEASE CATEGORY S2-CH FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 163) uma imus e uma sum umn mum amm uma em
[.-- . [ r L [ r m x E EE h
.sm a =
5 e p "!b m,"'R:t,
.W i 8
5 !ddim i l: m N 3 l id' 58 [ lf E'! 5 ! ' j al
=c'l 5l 5 5 5l c:
5 e m
~
m g it c G5 WiSI:l:=,iW..'=l='e'l we W 1. -
==
OD
, u. x
[ l w i
- . 05 3 ms
./ , oa I w
- uw
. z>
.I ,
'm <-
,I t;t; m
[ #
. OH l
5 M J
. = ,
= a' ew r J.
, ; w w-L ./ _d
.c -s ue
- ww
.I , ..m " p
/
= -
l , ; -n
-=
Wo az j -da
/ '
5x [ / .' .
*2
)
l ,-
/ 5 i
/..-
l' ,-
. . W
./ /
... =
.~
8
,/,< ~' ~.
u [ & _,,. e " ...
=
= T .
E J J 4 4 4 4 g 4 33Nulsta N3nto 301sino xsrs 30 NOI13083 M B-7
I w (13 E U I o w w 8 Q t#3 US M b Q Q
= a w y W W D (#3 ). ). m d
s w'pl"lClCo o wi we 08 o e
- ~.a d $ agwlj W H
- 5,i al 5:1d 5 o:
l gs ve-
= = w.
w w:.lr i 3- $- w mg W WIE'wlr
- W ' sag l w g 'g y' $W u $ ! N e.l .* , y I i A e u: a.s=lcui mimi.l 8 wa
%z g
6-EE we -
,/
/
./ =
=
Oz wo
- e-
., or
./ ,
i Om i >
- a. QC
(' sr
/ ,
Mu
~W c r.
[ , . M l S' S l
- %w ww g
, WH l "$ ><
l = u8 l l 5 ww a l
/
/ ."o 5
~E x
no
~
E
=n *z
./ , **
/ ,/ .
/ .wa 3a zo
.- LA.
/
./
/ .=
'?
-4 m
/
/
./ -
./
/ .. / .. $
/ ,... f..* .a: s s
.r./, - -
' E k.8y=-=',,
-c===....,.------ ------ ------ ------ -. m W
Y =
- - e e a s : =
=
0 N3hID 30!SinD )lSIN JD ND!IDUHJ I B-8
l I i I I I w z 55 c o w
" gW>> a
= . pl ai O t; W!5l bl mgi -g" "g' S
.s ,!.W'5'\d 1
- b,
%G
$ 8 5 5 w
~
5 l r W! 'i.'>= mg W = ll wj 5i 5=e l5 ts"
= = : W . .i af W WiSl"!: I m-l- -l gg I
- j <
J
=
O Oz f / W >O I g
.! DH
./
- UU 4
l E e W l '
/ =
Ms l
- I j
j'
/
i i 2
#C ne os M
l ,f (
/
= m~
mw
./ w w >-
W I ./ /
/ / .""m bis
-x I /
l y "5 l '
/ l -3 50 l ~"
// $*
I
'~.
/- l I
=
j '. , O, _../../ /:
,/. . . '
s D
.37,. ; .. /- . a
.=.=.==.....
. . . '___a ,,;
fy......... I W h W T a a a a a a a 33NU1SIO N30!D 3016100 MS!H JD NDI130H3 I I B-9 -
-,--,,,.---,-e, - --
l I w x I E
=
w w
= =
T e E
> 88 E
= 5aW>> g a
W t: .!plaigigug' W 5 j wl w,
= = Fu a =:a.o w.,w E d d' $m ~
w Mi$15l Ei" w:H j r' E
= w! w
_5l c 55 w Mg
<w
$ l W $iWj WiSI:el:Iml E'
*lE r; l .
Wj @ gg w
- mz oo l WM mm I = >o
./ J x-
.v
- m i--
./ 8 vo
/ [ , w ow J ,
l r . z> n, <-
- wn e
I MU
-W CH
/ , e
* = w8
,l
- f I
-*n w
W s' ew ww j : d ><
. 1 35 u8 ww o I *
.I .
w
" -E
/ I M a x l s .i en
=w wo az
/
-do de
/
/ zo w
./ /^ - ...* ,
j/
/ -
/ -4
=
e
- 9 f /. y.- ;
w g
--4 -
g g .._ ~ _...- g.,,.e..--:..~~~..~~~~""" , _ . , g a d d i A 5 l m
- 9 9 i 9 m
7 e
- e a e e 33NU1 BIO N30!D 301S190 ,NSIN JD NOI130HJ B-10
[ [ [ [ W 5 WW h
's
>- 88 r . 58W>> E L
- C'"' 18' - -
a' 8 t m *ini":i:m !" d
- N,Ti al 5ii s o'i o
's , ;$s s
*i"jm'gl5
*l-E lE 5 W-W
= =: Wi 5. j ."l =l E! 5 55 um F
L M M i 5 lit! : 1. -l . EE z 9 e oz i ( g mv
/ i W
I' I , G5 w .-- p j " o= UW
. . w E .i / / = =Nm
,I
.a- 15 m t,_
My I . ( Ow L .i i
". ma me
,f /
-rg ga-
/ m ww l
l' I
*dE #Q M
s i ?" mQ
-w g -
/ i co m z-E
. . i u j ,/
- =
f i
./ ..-
/ b=S
,;-a:
L m= [ t' .-
/ /i .
i *8
. u-
/ / ) /,\
i :- 9
=
// /. - /
// / ,s'- ,? g y l g(
'4
- 7.r. ' -: : .:=== = c a[,,,,,a A .. s- ,
-a 2 u.
l j g.:-:=3 , l . . . . . . 4 E " F : E E O 2 I a a a a i a a 33NUISID N3 AID 301S100 MSIM 3D HDI130HJ B s-u -
I l w
= 5i l
a E E o o w N a * >.
= o w =
w w o . >. >. o. o
!.Io,o I
I .- w:H Up
- m.J wi o3. o. o. U3 w
-: a oalwlw a w t-3 "i el 5! gl d: di "R i
5
- D- r l 5 'i~
W 5!'j wl' El 5 55 Wi 5i e." el.
= =:..
u el 515 - am WE a u a!m mIm a.s -
=g a
g oz
- u. m
/ l m-w x
>z l . =, .
I
/
/ J .
mo 0-v ,- o
/ -
I we u
.i , zw I
/
J.:
* . l;*m >- .
mp ( , l- -O [ . a w-
.I : : . wo
.I,/ .
1: i:.- .w m ex me
=
. (: WW I - > +-
.I,/ I. =aE uS ma
;I :
/ ..m "w -w xy l
11 / = z-l l ) =3 n= be
<=
/ -wo ya a.
/ r m
- o u-
/./
/ i *
=,
)
.I / / : ,
=
1 ..
- ...:' i)
/
.. / s .. .. .
w e g s
-.........==c==u........,
-- n g
g, g*g ,.fru'. _... .,,.,_ - ,
- E E3 CD b
C4 D)
.. g y
CD En =. t= IJ3 T N W g 3DNUISIO N3 AID 30ISinD MSIM JD NDI130H3 I B-12
l l
)
l E W e. I w
~
m wIpl W Q w O o w "i o O c ce le WW
= =
O O
). y, o
M 8 w w' w I
* = ~
~
~ i a l w= augw H a w w =c ^
H
. ,m . = -lel o a. a u~
a a::
~. w . 5 3 W3 W
g s= sl5
=:sl :=le!c. MIS l5 5Iz'5 * =.
w w
<m ww am WE g u;el
**a mil mlml
., af a sl =
mm e 2 25
-E !
0 5 I / .
.* >=
MO
/ ,
am h f l .Y
- uH e
f g4
! zu
/
/ =$ Hm m e-.
E l ' r 58 H
/. / ,. mo
==
E ,t I'-GWa N' LaJ Lag I / >H
,f
}
/
- g2~ us W8 l
E ! l dw xy
/ , **
H ae
/ .
[% .l l
,l
** -m
*~
XM O
~
, ,/
l -4 7 'E , d ./ / m
,w.*/Mj'.** ,,,/*** (D g
D V M e.r: . = .- ..v....= = =..= = r..T.7
- -- v
=
=
=
6 s m q m P
~ = .
- * , N. e M g . g (D I
4D 33NUISIO N30ID 30IsinD )fSIs 3D NDI13eg3 .E B-13
I I r u u i w w 8 w w O O N W
> Q c w c w w
w .M C w > >O x e e
$w m
w
- 8
-, N : *'* l w-w, l w &
a m._ w o,=lw,l5:5 o =' a, b5
=: a.
b 7 a I5 gh b] 5 ms m F
- c. , *t" M e
- e 3!! s= j =iI" m f W* *l Lw8 ww W
W N
- a $a el w *mlwlw _l U:
ty =a
. x J
- $5C
~
l l M* i ! w i sx go l
' .T E
- / un wy E
/ / !
o Zw I ' l
- I J f -5 ;5 m R,-
/
a f - se~
/ l .- me
! / .
BE E, i <
/ '*2- ~~
l I >s
! : .~ M
,t
$V mQ <
ce ~ ~ a ~. j
- / ,/ u x~
g5 a
.i / .-
/ . we.n ae mz xx g
i >,
/ <
l w If /t R m' 3 i / l s '/ v C3 g
/
/'
s- -) + u . w M l
- *...- R g
/ g :.:-J' . _ .
,C_....
. .* *- *; y y :.: at u u no u o u u s.t u asu:,s s:u t -
k fg:, s W
. l = .cA w 3' N
n *
- 4 7 9 i
" U IO N3f51D 30iging yggy ,9 99, i
B-24 -
i I I I 1 1 a z d o w w V I s aa a a m w
. tw w a m > >
E l w .- ! . I oo !oo ai o
- w = 0. CD w: o. W
._,l wiai wl .
a .:..o w w i
" Mi el 5, a <-
05
" Ei'i r'w'glrld: 5 5 d' M~ :
I w:e = t-w <E i F He wl
=, : .- i wl w: r. cwl WW
=
a a a : . .mimIm .i ma.al.i.i= = - aE W ED I M oz
. . u. O f
f l E mv
/ W 5
J'
, / .
>z j *v .
I xo t # Dm l l l O&
, 1 O i
, ,f , W< o l i l< . ZW I J Ia J e t I ( i d
, tu pw ms
-U OW p
W mo
" j
- l
- pg m c.
I ff l..*n 'm &
- f l W WW l '
> l"=
JI J j: gr- < l u~ i, / .- ma e f ww I cm, d ( -
= "E z~
)
I .
./ /
/
=
.== .$.
bo
<z Mm I,? .
/ / i.
?
I I.I -
/ . 'a 2, o
,/
/ / 's
/
v D
/l,f o' s
/ ,,,.="*..*-,1 w Ij ,/ ,s $ b J
- .-- .C . . -- : c c , , ,-- . . =- ' -' -d S V;,,.s _-;;c --
u-yx; - I CD (D Uk 03
=
b IJ3 t'3 Y m N y
=
k 33NUISIO N3nID 30ISinD MSIB 3D NDI1305d I I B-15
I I l I I l APPENDIX C CONDITIONAL RISK CURVES I I I I I I ~ I I I I I I
I APPENDIX C FIGURES Figure Page
- C-1 Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category SI-CM for No I C-2 Immediate Protective Action (Run Number 137)
Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S1-CH for No C-2 Immediate Protective Action (Run Number 177) C-3 C-3 Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S2-BM for No I C-4 Immediate Protective Action (Run Number 138) Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S1-CM for No C-4 Immediate Protective Action (Run Number 139) C-5 C-5 Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S2-CH for No Imediate Protective Action (Run Number 163) C-6
,I C-6 Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S3-BM for No Immediate Protective Action (Run Number 140) C-7 ,g C-7 Complementary Cumulative Distribution Functions for .g Health Effects Risk for Release Category S3-CM for No Immediate Protective Action (Run Number 141) C-8 C-8 Complementary Cumulative Distribution Functions for ;I Health Effects Risk for Release Category S3-CH for No Immediate Protective Action (Run Number 156) C-9 C-9 Complementary Cumulative Distribution Functions for I C-10 Health Effects Risk for Release Category S6-BM for No Immediate Protective Action (Run Number 131)
Complementary Cumulative Distribution Functions for C-10 Health Effects Risk for Release Category S6-CM for No I C-11 Immediate Protective Action (Run Number 178) Complementary Cumulative Distribution Functions for C-11 Health Effects Risk for Release Category S6-CH for No I C-12
- Immediate Protective Action (Run Number 176)
Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S7-BM for No C-12 Immediate Protective Action (Run NJmber 142) C-13 C-13 Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S7-CM for No Immediate Protective Action (Run Number 143) C-14 C-14 Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S7-CH for No Immediate Protective Action (Run Number 158) C-15 I C-15 Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category SI-CM for 1-Mile Evacuation (Run Number 172) C-16 Complementary Cumulative Distribution Functions for I C-16 Health Effects Risk for Release Category SI-CH for 1-Mile Evacuation (Run Number 179) C-17 I iii
I' APPENDIX C FIGURES (continued) l Figure Page C-17 Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S2-BM for a 1-Mile Evacuation (Run Number 144) C-18 g C-18 Complementary Cumulative Distribution Functions for ) Health Effects Risk for Release Category S2-CH for 1-Mile Evacuation (Run Number 164) C-19 C-19 Complementary Cumulative Distribution Functions for ' Health Effects Risk for Release Category S3-CH for 1-Mile Evacuation (Run Number 148) C-20 g' C-20 Complementary Cumulative Distribution Functions for l Health Effects Risk for Release Category S6-BM for 1-Mile Evacuation (Run Number 133) C-21 3l g i C-21 Complementary Cumulative Distribution Functions for E Health Effects Risk for Release Category S6-CM for 1-Mile Evacuation (Run Number 183) C-22 C-22 Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S6-CH for 1-Mile Evacuation (Run Number 152) C-23 C-23 Complementary Cumulative Distribution Functions for B Health Effects Risk for Release Category S7-CH for 3 1-Mile Evacuation (Run Number 159) C-24 j C-24 Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S1-CM for 2-Mile Evacuation (Run Number 173) C-25 C-25 Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category SI-CH for l 2-Mile Evacuation (Run Number 180) C-26 m C-26 Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S2-BM for 3 2-Mile Evacuation (Run Number 145) C-27 3 C-27 Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S2-CH for Mile Evacuation (Run Number 165) C-28 C-28 Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S3-CH for 2-Mile Evacuation (Run Number 149) C-29 E j C-29 Complementary Cumulative Distribution Functions for 5 l Health Effects Risk for Release Category S6-BM for 2-Mile Evacuation (Run Number 134) C-30 g C-30 Complementary Cumulative Distribution Functions for g Health Effects Risk for Release Category S6-CM for 2-Mile Evacuation (Run Number 184) C-31 C-31 Complementary Cumulative Distribution Functions for l Health Effects Risk for Release Category S6-CH for 2-Mile Evacuation (Run Number 153) C-32 ! C-32 Complementary Cumulative Distribution Functions for E Health Effects Risk for Release Catege y S7-CH for E i l 2-Mile Evacuation (Run Number 160) - C-33
- i
I APPENDIX C FIGURES (continued) Figure Page C-33 Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S1-CM for I C-34 10-Mile Evacuation (Run Number 174) Comolementarv Cumulative Distribution Functions for Hea'lth Effects Risk for Release Category SI-CH for C-34 I C-35 10-Mile Evacuation (Run Number 181) Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S2-BM for C-35 C-36 I 10-Mile Evacuation (Run Number 146) , C-36 Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S2-CH for 10-Mile Evacuation (Run Number 166) C-37 I C-37 Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S3-BM for 10-Mile Evacuation (Run Number 170) C-38 I C-38 Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S3-CH for 10-Mile Evacuation (Run Number 150) C-39 C-39 Complemer.tary Cumulative Distribution Functions for I Health Effects Risk for Release Category S6-BM for 10-Mile Evacuation (Run Number 135) C-40 C-40 Complementary Cumulative Distribution Functions for I C-41 Health Effects Risk for Release Category S6-CM for 10-Mile Evacuation (Run Number 185) Complementary Cumulative Distribution Functions for C-41 Health Effects Risk for Release Category S6-CH for I C-42 10-Mile Evacuation (Run Number 154) Complementary Cumulative Distribution Functions for C-42 Health Effects Risk for Release Category S7-CH for I . C-43 10-Mile Evacuation (Run Number 161) Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S1-CM for C-43 I C-44 Mile Evacuation and Sheltering to 10 Miles (Run Number 175) Complementary Cumulative Distribution Functions for C-44 Health Effects Risk for Release Category SI-CH for I 2-Mile Evacuation and Sheltering to 10 Miles (Run Number 182) C-45 C-45 Complementary Cumulative Distribution Functions for > Health Effects Risk for Release Category S2-BM for 2-Mile Evacuation and Sheltering to 10 Miles (Run Number 147) C-46 C-46 Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S2-CH for ! 2-Mile Evacuation and Sheltering to 10 Miles (Run Number 167) C-47 1 I l 1 lI "
I APPENDIX C FIGURES (continued) Figure Page C-47 Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S3-CH for g l 2-Mile Evacuation and Sheltering to 10 Miles g (Run Number 151) C 48 C-48 Complementary Cumulative Distribution Functions for Health Effects Risk for Release Category S6-BM for l 2-Mile Evacuation and Sheltering to 10 Milas a (Run Number 136) C-49 C-49 Complementary Cumulative Distribution Functions for g Health Effects Risk for Release Category S6-CM for 5 2-Mile Evacuation and Sheltering to 10 Miles i (Run Number 186) C-50 1 C-50 Complementary Cumulative Distribution Functions for t Health Effects Risk for Release Category S6-CH for 1 2-Mile Evacuation and Sheltering to 10 Miles (Run Number 155) C-51 C-51 Complemene.ary Cumulative Distribution Functions for Health Effects Risk for Release Category S7-CH for 2-Mile Evacuation and Sheltering to 10 Miles (Run Number 162) C-52 I I I
- I l
I I I I vi
APPENDIX C CONDITIONAL RISK CURVES This appendix contains plots generated from the output of the CRACIT h computer code for conditional frequency of exceedance of health effects. Each plot corresponds with a particular release category source term case (B for best estimate, C for conservative) and consequence model case (M for medium, H tor nign). All plots are conditional frequency of i exceedence curves, given the release. i lI I I lI I I I .I }I g C-1 1324P100385
se e 188 ,1 s 1 . . . . . ,t e2 ,j es .ja4
.m uy, , .
. . ,,t e s
._s.
'. T w
E E w ' , 'N
~.*
a
- ACUTE FATALI~*IES 5E '. . \
~
A.Cy,TE,,,IM,y,U R I,[:S,
*wg 1g-t__ .
, t- N __ta
-t TOTAL LATEMT EFF o= .
' N* ::
d u o
' \
50 '. i : o$m -2 I. \-
-2 h5 I8 -- I.
- \, --:18 W8 \ E
? 55 : N B, - _a o . E8 -3 -3 5 18 -"- 18 t "i _E E : 5 :
-4 -4 1g ... .i .. ... ... Ig 188' "'181 ' t e2 ' '183 ' "'t 8 4 ' "'185 ' 'tg6 NUMBER OF HEALTH EFFECTS FIGURE C-1. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY SI-CM FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 137) aus uma uma aus aus sus ese ens sum
- sus amm one aus
M M M M M M M ' M M M M M M M te a ias -
. . ,t e l
.,t e2 . . . . . .j e 3 . ,1 s 4 ..s s5 s "i ..
- - - -- - - -r . . . ;...,; ,~, w g s , -
m \.,\ N. : E . \, ACUTE FATALITIES E \. - g ., s, A.C. .U.T.
. . E. . .I.M. .J.U. .R.I. E.S. .
E ws tg ,
, '. N -
__t g , TOTAL LATENT EFF o m
, N :
d u. ' u o
- X w '
w a . z :. b g: i 3m l -2 : -2 u5a-5u 18 o o - o 55 . a "Eo i - s : .
@d 3 ! -3 o 5 18 j _18 tg : :
e : 5 ~:
-4 -4 ig ... .i .. ... ... 1g 1g8' "'t g 1 ' "'t g2 ' 'tg3' "'t g 4 ' "'i g 5 ' 'ggG MUMBER OF HEALTH EFFECTS FIGURE C-2. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY SI-CH FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 177)
i 18 8 188 I . . . . . j g2 ,j g3 ,,j g4 ,,j g5
..,J 8 , ,
-- n..:, .., .
g .. g . e r \ . ACUTE FATALITIES
$w $ \, -
A.C.U.T.E...I.M..J.U..R.I.E.S..
,zd,18 ,..
__tg , TOTAL LATEMT EFF Em \ -
"b ....., 'g :
da i. i :
\
b$ -
~
tzg " 18... \, , is ' so \ : n 5z - i *E : a= . E8 -3 3 5 18 -.- ' 18 bU _5 e : 5 R
-4 -4 10 ... .i .i ... ... 1g 1g8' "'t s t ' "'t g2 ' 'tg3' "'t g 4 ' "'t g 5 ' '136 MUMBER OF HEALTH EFFECTS FIGURE C-3. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S2-BM FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 138) num uma aus uma een ame uma men aus ses sus sue uma sus ame
N -- m - 7 R R n n M n F7 R n n n n P 1 l 18 8 188 ..,j s t - .j a2
..j a3 ja4 ..,j a 5 w \
5 r w
- ACUTE FATALITIES i
\* l gg .. - -- ..-.....,.., \, ,
A.C..U.T.E...I.M..J.U.R.I.
. . .. E.S.. l E g, t g ,._
g _t g , TOTAL LATEMT EFF S,w , \ : w - - . y '. \ - wu ',-- - u5 : '$. . l o= -2 :
- 1 -2 i
3m 18 l g5u -- 1 l- -:18 l to o i. ) : z .
? Es w-i l :
cn a o : . E" -3 ! -3 o sw ta _~_ i. __ts E : -
~
8 i' 4 :'-" 4 1e - a .i i ... i 1e tes ' i
"'st' "'182 ' 183 ' "'t a 4 ' "'185 ' '186 Mut1BER OF HEALTH EFFECTS I
FIGURE C-4. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S2-CM FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 139) i i
l 18 180 ..J 8 I ..,J 8 2
.......J83 ..j a4 ja5
.m g_s.
i
. \ ACUTE FATALITIES l
b5
,m .
\, -
A.C.u.T.E...I.M..J.U..R.I.E.S. l z d t g ,._ ' __tg , TOTAL LATEMT EFF E= '.
\ :
5e x
'.'.. '\.
wu
- \ -
l b5 ! i - 3 E
-2 : *
-2 i
g g te ,_ i 1 ._ig w8 l :
= i. . , ? 5 "5a i i m i -
d" r -3
\
-3 5l od w is _- -_ :. ._sg J _:
M l . 8
-4 -4 18 ... .. .i ... ... tg ige' "'181 '_ 1 g2 ' t g 3 ' "' ig 4 ' "'t g 5 ' 'tgG Mut18ER OF HEALTH EFFECTS FIGURE C-5. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S2-CH FOR N0 IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 163) ums une aus uma uma
M M ~ M s 18 e l -as_ .m ,,,. ,.. ..,3 8 1 3 82,
, . ,3 8 3 .,j e4
,.,j s w ,\. -
E \. : ACUTE FATALITIES Cu T - -
$ =g ,
g , A. .C.U. .T.E. . .. .I.M. J. .U. R.I. E. .S. . E ' __1 g , TOTAL LATENT EFF
$ Wg i s ,__
dw m o l. X I* w w u
~
wg o g
\*
> = 18 .
-? I 18
-2 u5 r u 1-Eo 1 o = :
L "
=E \ -
a= 3 1* Ed -3 . I -3 o > 1e ' * -
\ _'.
_1s [" o g : u .
-4 -4 10 ..i .i .i ..i ..i 18 188 ' "'181 ' 192 ' 1 g 3 ' "'134 ' "'1 g5 ' '1g6 Mui1BER OF HEALTH EFFECTS FIGURE C-6. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S3-BM FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 140)
1 i 1 8 188 ..,1 a t ,,1 e2 18 . .,t a 3 . .,1 e 4 ..)e5
- - - - - , . . c.,.
- ,' s. -
! w N. - E \. ACUTE FATALITIES r w E v'
" N.7 -
ACUTE IMJURIES l _ .........., _ i E g t g ,__ ,
\, _t g , TOTAL LATEhr EFF
- o=
w
\. :
m no 1
\ -
i dd - I - i - ' Ei D= $,18 _. 2 *--
\
\ 18
-2 z:8 -- .
a= . ! : o 5 "5 i l : a E l g o '
"d -a
\. .. l -s
.i o n te _--_
,_ w,
- t. _:
_is g : a i : . -4 -4 1g ... ... .. ... ... ja jg8' "'1 g 1 ' t g2 ' 183 ' "'18 4 ' "'18 5 ' 'gg6 Mur1BER OF HEALTH EFFECTS FIGURE C-7. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S3-CM FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 141) aus e e m
m m m se e 189 je3 ja5 a _. ,. ..,j e t
. . _ . . . . .,j a
. ..,j e 4
~.s.% ::
! m ......., ' -s . y ,
'N : ACUTE FATALITIES E5 -
9.cu,TE,,J,y,Ug,IE,S,
~y1g'I4 E
x
', . __t g'I TOTAL LATEMT EFF m .
.' \ .
S - - Oo '. I. : dw
'O U
Y m
)
)e
-2 . -2 U8 = I0 -r 1 - 18 5u a o l
o 0* k :
=y 4
g . s o - { Ed -3 .
-3 o
gws> 18 -- - 1 -_18 l - g . 5
-4 -4 1g ... .. .i ... ... ta jg8' 'tg3'
"'t g i ' t e2 ' "'t g 4 ' "'t g 5 ' '1g6 nut 1BER OF HEALTH EFFECTS FIGURE t-3. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S3-CH FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 156)
te s 188 ..j e t . ,j ea . je3 ..,t e d ..j e5 "s m. a s s "a - w N'g - E - l ACUTE FATALITIES r w \*
$ g= ,
\, ,
A.C.U.T..E...I.N..J.t.I
.. .. R.I.E.S..
E g t e ,__ T __ta TOTAL LATENT EFF E= g : b ------ ....
}
wdz i.- 'g - m og - g
' -a tzg5 is'8, -----
'g __is o Wo o i.
b b
,o .
E a -a 5 18 -r - 18 w .
.s J .
E : 5 l -1 -4 l ta ... .i .. ... ..' ta tes' "'t e t ' i ea ' '183 ' "'t e 4 ' "'18 5 ' 'teG Mut1BER OF HEALTH EFFECTS FIGURE C-9 COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S6-BM FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 131) M M ] M M
R R R R R R R O V .R F-~1 ...!
)
tg 8 188 ..j 81 . . . . .j g2 ,j g3 ,,j g4 ,,j g5
's , , ;
- . . . . . . . . . . . . . . a . . . . . . . . . . . . .w. , 4 w . -
u . g s'
\.
E . . ACUTE FATA'.ITIES r - . g
$mg .
.,g A.C.U.T.E...I.N..J.U.
... . .E.I.E.S. .
tg , l E ,s t g ,. _ . N TOTAL LATENT EFF l Em - T- : l d' u o i
- a. \ :
x . - l WW *
., }
b5. .2 i-
-a o
m g 18 - '.. \. ..18 W8 . \ :
? 5= '\ : ; *B
_,= E8 -3 -3 S 5 10 -- - 18 sW sa ..
~
i 5 : I
-4 -4 1g ..' .i .' ..' ..'
18 180' "'181 ' "'t g2 ' t g 3 ' "'t g 4 ' "'t g5 ' 'gg6 HUMBER OF HEALTH EFFECTS FIGURE C-10. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S6-CM FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 178) i
l jg IS O ,,j g5
. . . ..1. g. .3. . , ; ;,,,1
. . . . .. . . . . . . j a I .. .. y.-..,7,. . . .. ..._. g4
.... 1 0 ... .
w
- s. g :
l E .
- ACUTE FATHLITIES r , \
$ w=
g ',
'\,
A.C.U.T.E...I.M..J.il
... .. R.I..E.S..
E d,18, ._ . N, _gg
. TOTAL LATE.NT EFF
'\
$b i :
3O \ \ : y5 ", k - d z u 18' _ k- .. _ t 8' w u g : m o . . m o z W w I - C E3o - a . Ed -3 -3 5 gw318 -_ __.18
~-a .
E 5
-4q ..
-4 10 .. .. ... ... tg ig8' j
'gt' t g2 ' g g 3 ' "'g g 4 ' "'185 ' 'ggG HUMBER OF HEALTH EFFECTS FI611RE C-11. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S6-CH FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 176)
M M M M M M M M M M M
t s' 8 88 _. . ._. _. . . J 8 8 jaa js3 ..,j a4 s
..,j s
'N, .:
ut N i { 'N *
- ACUTE FATALITIES Eh ,
\,
A.C.U.T.E...I.M. l.u..R.I.E.S.. { ,18 d ,._ \ _t g , TOTAL LATEMT EFF g* mg
................ (.
N w w N. I. u g 5h -2
\*
'\.
-2 3m .
g 818 -- ! \ .- 18 n US '\ !
!5o \
b ! _s . E' -3 -3 5 518 . _tg CU _i e : 5
-4 -4 , t0 ... .. ... ... ... tg 188' "'t S t ' 182 ' g g 3 ' "'g g 4 ' "'g g 5 ' ' jg6 Mul1BER OF HEALTH EFFECTS 1
FIGURE C-12. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S7-BM FOR NO D1 MEDIATE PROTECTIVE ACTION (RUN NUMBER 142) 4
l 18 18 .08I
. . . . . . .s . . . . . . . 4 . . . , , , , , , ; 7, . J 8
. . . . . J83 ..J 8 ..j g5 i i -, ..;
w
,,.4,%.N
\., h, h ACllTE FATALITIES
{$
$g ,
\.y -
A. .C. U. .T. E. ... . .I.M. .J.U. R.I. E.S. .
" g is ,
., 'g __t g , TOTAL LA TENT EFF 5"
e w m . 's. 5 y :. \ . uln : -
\.
m5 O m 35 -2 ; \. -2 a g 18 .- - g - 18 5u a o m o '.
\.
L u5 \ a ma - a o . E8 -3 -3 o 5 18 -
. t8 t "i :
E 5 :
-4 -4 18 ... .. .i ... ... tg 1g0' "'181 ' t g2 ' t g 3 ' "'t g 4 ' "'18 5 ' 'tg6 MUMBER OF HEALTH EFFECTS FIGURE C-13. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S7-CM FOR NO IMMEDIATE PROTECTIVE ACTION (RUN NUMBER 143) sus sum um sme muu uma muu um um aus um sem muu em ums uma uma num
7 r7 W m I i n n n n n n_ n. . v n .o rm m te e tea_ . . . . ,_,,,,j et a ,,s es
, . . . . . . . .,,. .... ,j. a , , , , . . . ,,s e3 ,s e4
. . , - - - ~ ,% ., ., , _
j
%.,' N. -
w ',-
"\-
g
., . ACUTE FATALITIES e -
N aM o a '. g
.A.C. .U.T.
. E.. .I.M. .J.U..R.I..E.S. .
-8 g w'!,g-1-- .,
s- s _ _is TOTAL LATENT EFF 3 wg u l. W * - o . . k * *
- o. .
y $$ . ;7
=o
,o .
1 : O O g , u, - m e . Ea 3 -3 o 5 18 .'. s hj
- 1G 5
U
-4 -4 1g ... ... .. ... ... 1g
,g- i y- ig . - it ,3 - -i,4-t 135' 'gg6 MUNeER OF HEALTH EFFECTS FIGURE C-14. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S7-CH FOR N0 IMEDIATE PROTECTIVE ACTION (RUN NUMBER 158) !
i l I i
a, aaaa m....b .a..a.a.aa.a...
.. . ....a .., 7, ; , , . .a, aa,
.. am,
- L N. :
w '\ - E E
$yg ' . '\
- ACUTE FATALITIES
\.
A. .C.
. U.T. E. . . .. . .I.M. .l.u. R.I. E. .S. .
" ,d 18 ,.. . '-\ __t a , TOTAL LATEMT EFF o= .
N : d' u
\ :
6u '.
~
\ -
b$m -2 \.*
\ -
-2 h y18 _.. ,,, 5, _;.18 o Wo o
\ :
I w E 1 . _, o . E' -3 -3 2518
,_ w __ --
__18 wJ . 5 :
~4 -A 18 ... a. .i a.. ... tg 188' "'181 ' t g2 ' t g 3 ' "'t g 4 ' "'i g5 ' 'tgG MUMBER OF HEALTH EFFECTS FIGURE C-15. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY SI-CM FOR 1-MILE EVACUATION (RUN NUMBER 172) aus ums nun sus muu num smus tua sur sus sus sum uma an unmi aus uma ame
i n_ v . n n_ i i n o o n o o r, e r, _r m m i ie s tes _ . ..j e t- - . . . . , j s2
..je3 ,t ed
.., ..,j e5
.g*g,
-- ., .. . y .... ,;,7
', N, :
m ., E ',
- \, : ACUTE FATALITIES
$g$
s, ,
.A.C..U.T.
. E...I.M.. .J.U..R.I.E.S. .
E d 1e , . N, __t e , TOTAL LATEMT EFF E" '.
\ : l bg '.
- : I y .
- i wg , .
u= og l* ~ l 3m -2 : -2 I u g 1e __ g .; ;.1 e w a a o . n o z - w - a "w . l , y ' ~a a o i - E' -3 i -3 o 51s __
,_ w --
- - __te J .
E : l 5 :
-4 -4 ie ... .. .. ... ... 1e 1se' "'t e t ' t e2 ' t e 3 ' "'t e4 ' "'t eS ' 'ta6 MutieER OF HEALTH EFFECTS FIGURE C-16. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY SI-CH FOR 1-MILE EVACUATION (RUN NUMBER 179)
l 18 8 tas . .,,i s t
, , , , , j aa ..j e3
..,j e4 ..,j a5
, .a, -, - -
m \ ':
~
E n \ ACUTE FATALITIES E5 . A.C,u,TE,,,I,y,U,f,IE S, E$ta _ k ,_j g'I TOTAL LATEMT EFF 5" \ 'E Os \
, Yd \ -
55 \ - 2;aE ta' __ '\. __t s'" m ,. WE \ i n, o= - I E $~5 : a e _ l E" -3 -3 5 te -- __ts t;" : E : 5 :
-4 -4 ta 4 .. ..i ... ... 1a
, tas' 181' "182 ' t a 3 ' "'t e 4 ' "' is5 ' 'teG I NUttBER OF HEALTH EFFECTS FIGURE C-17. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S2-BM FOR 1-MILE EVACUATION (RUN NUMBER 144) 1
t n rm m n. p. v rm e v w r, f- , m m te8 tes ,,j e t ,,j ea , , , , , ,,,j aa
, ,j e 4
,,j es
, _. _g ..
m
----~ ..... ,~~' ..,
N,'\ - a r \ : ACUTE FATALITIES m
.,'. \,
$ g" ,
A.C. .U.T. E. . . .I.M. .J.U. .R.I. .E.S. .
" g t e ,._ ,, \, __te , TOTAL LATENT EFF o= . N- :
am o o
, \
5g. '.'. ') : m= : . oW -2 i. 1-
-2 3= :
u =>te _-_ ( __te zg : . go i l :
? 5= i
- i. :
g 55 ' a o : - l e . z- -3 3 o s' s -_e _- gmu i. __te m : . o : 5 u d
-4 -4 1e ... ... ... ... ..i 1e see' "'t e t ' t e2 ' "'t e 3 ' "'t e 4 ' "'t e5 ' 'teG Mut1BER OF HEALTH EFFECTS FIGURE C-18. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S2-CH FOR 1-MILE EVACUATION (RUN NUMBER 164)
s se s see .- , , -, ., .,t .a, l_ ,1 o2 ,t a3 ..j a5
. ..,j e 4
'w.%,s. -
w \ ~
"--', ACUTE FATALITIES
'N . ,
E{Eg "
-, ' T. -
A.C.U..T.E...I.M.J.U.R.I.E.S..
, g ,
E d te , '- . _tg , TOTAL LATENT EFF g= ' ..
\ :
ag -
' i. :
M , - wd ',
}
==
E i.' 1 - j -2 Z; g is 2__ i . _ta 5u : I : i
? g 1 j :
s ' 05 w- : : e
-- 'I. ,
Es -3 .
-3 o s se __ l __ts ha t :
g - o u (.
-4 -4 la ::' -- . . : ' :' ::' 18 168 '181 '182 '183 ::184 '1 g 5 . igg MUMBER OF HEALTH EFFECTS FIGURE C-19. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S3-CH FOR 1-MILE EVACUATION (RUN NUMBER 148) aus aus sua sus sus uma sum uma sus as aus sus use ens uma mas amm sum
I Fl U Tl. Fl__Fl F L_f l F U L_.J7 Fl .F7 .F7_ _r 7 cm_ rl . q l i 1s 8 ies jst. .
..,.s. . ,s e2 . .,s s 3
. .,s ed s
..,s e .
g w 'g - E r l ACUTE FATALITIES
\ * -
E =w: a \ . A.C. U.T. E. . . .I.M.
.. . .J.U. .R.I. E.S. .
~I E]1g-I,_ N g , _tg TOTAL LATEMT 'EFF E= g :
$5 ' " " '
..... i i dO i.
w r ; i. oE
-a i
- l. -a 3
g 18 _
. ~ ~ -
g _18 so k 5
? 5= i : i O "Ew -
a Ea -3 -3 o $ 18 -- __ ._18 9w -: nn . E ~ 5 .
-4 -4 1g ... . .. ... ... ta
- g. ...y. .j . . t ,3 i,4 t "'t g 5 ' 'tg6 Mut1BER OF HEALTH EFFECTS FIGURE C-20. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S6-BM FOR l-MILE EVACUATION (RUN NUMBER 133)
18 0 100 ..101 . . . . .j g2 ,j g 3 ,,j a 4 ,,,1 g 5
.............ca....... ..... _w c , % . ,
~~-
- . ACUTE FATALITIES sw '. -
N* oEw _y
',, A.C.U.T.E...I.M..J.U..R.I.
... . E.S..
\'N z d 10 -
__1g TOTAL LATENT EFF E=
- W 6 *
. ' \. -
u o . X w w , \. - a .
. (
a Z
- oe .
\.
>=g la-2.
a
-2
) _18 a
$E o z
'\ -
_l . A e s \ -
~ 6- -
J U . E' -3 -3 5 10 -- _.18 MU _i e : 5 -
-4 -4 1g ... .. ., ... ...
1e 188' "'t e l ' i g2 ' g g 3 ' "'g g 4 ' "'1 g 5 ' 'tg6 HUMBER OF HEALTH EFFECTS FIGURE C-21. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S6-CM FOR 1-MILE EVACUATION (RUN NUMBER 183) l l l
1 cm u v rm u . rm . n v. .r m w n r- _. m is 8 1e8 . . ,. j e t .j e2 . . . . .. .j e 3
. ,j e4 ..,j a5 a......
.g,s.
. . - . ....... ....;7, ..
,,, ~ .. -
w ' '. .* N. E *
. \. ACUTE FATALITIES
$r$ ".- \. -
A.C. U.T. E. . . .I.M.
.. . .J.U. .R.I. E.S. .
.Ewg818 ,.. '.. N
__1 g , TOTAL LATEMT EFF o* . -3 : l guO ; u . k ~ X w
- u k.
w o =z
-2 l
2 U =:318 _~_ l __18 mg . -: so \ : 7 Sr : C$ "$ _, w E' -3 -3 5 18 --.
-:18 t;"! :
E : 8 ."
-4 -A 18 i i
i ,,s
,i
,,,. . .. ' , , , . .. ', g . i,3
. . ', , 4 ., ,5 Mut1BER OF HEALTH EFFECTS FIGURE C-22. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S6-CH FOR 1-MILE EVACUATION (RUN NUMBER 152)
- - , _ _ . . . . . , . . ....b b b j$ .. . . . . . ,,,,;~,.,
N
., Nw 5
,N m ,"-
E ,
- l ACUTE FATALITIES r m - \
E=g
'x ,
tg , A. .C. U.T. E. . . .I.M. .J.U. R. .I. E. .S. . TOTAL LATENT EFF E g 1 g ,__ ',
's'N E= '
b5 . . ' 66 \. \. w6 o m
\ \.
-2 .
g -2 3m u5 18 - i . - 18 5g
=>
( i. l. . p 5= g E3 o= . E' -3 -3 18 5510 t "i : E 5 :
-4 -4 la i i e ..i i 18 1g8' "'t s t ' 'tg2' 'tg3' "'t g4 ' "'t g5 ' 'tgG l MUMBER OF HEALTH EFFECTS FIGURE C-23. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S7-CH FOR 1-MILE EVACUATION (RUN NUMBER 159) mas as man aus aus sua sus num aus an uma sus man sua sus mas sus aus
t rm v v rm v v w . rm . rm- rm. rm. v a a .j e5 te s te ,. _ . .. . . . . . . . . . .j s
, ,t
. . .., . . ;, .7,. r. . fe.Je3
, ,. --, -i. J. s4
,s. :
w , N* - E .- s ACUTE FATALITIES r m '. ' s ~ EE
. \ A,CU,TE,,,I,M,J,URJES,
-1
_E $ t a._ 't. 'N __te TOTAL LATEMT EFF S m *m '.-
. 'N. E k *
=0 '
m= 't. . M
-a : \. -a
>8 g I8 --
- 1' - 18 wg a
e..., 1 - O E
- O, w m N N 3 '
On 38- o-e J G . E' -3 -3 5 51e -- -:.t e t' : E : 8 :
-4 -A l 1g ... . ...
..,. 1e
..,. i
,,,. , ,,, . ., g . . . ., , 3 ,,4 ,,5 . ies 1 MUMBER OF HEALTH EFFECTS FIGURE C-24. C0:iPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY SI-CM FOR 2-MILE EVACUATION (RUN NUMBER 173) l
'8' . . A 88 .08 18 . . ,A 8 ' . m.. . . . 0 8 ' .. A 8
. . . . . . w,"-
m '* ..,. . .g ' N , 5 y "., \, - N E z-
, \, l ACUTE FATALITIES
$h _
\,
s, A.C. U.T.E...I.N. .J.U..R.I.E.S.. Egs t g ,__ , N _1 g , TOTAL LATENT EFF
. w ,
0 ',
- \, 25 Oo ',
dd - w5 o= -2 3m , 2 o 5 18 .-- -
-- 18 EM .
n 8 : : I _a o . . E8 -3 i -3 5 18 - i - 18 tb : : E
-A -A 18 ... .i .. ... ... ig 180' "'181 ' 182 ' 18 3 ' "'18 4 ' "'18 5 ' '186 NUMBER OF HEALTH EFFECTS FIGURE C-25. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY SI-CH FOR 2-MILE EVACUATION (RUN NUMBER 180) aus sus aus sus aus uma ses aus e sus een aus mas see ese ses sus sus num
\ p u v m U U M w rm- rm. r se e tes ..,1 e t . . . . .j e2
. .,j e 3
.j e4
..,j e5 c N. :
w \ - E "\ ACUTE FATALITIES r w aE
'\. ACu,JE,,J,MJ,UR,IE,S, E U te-I _ \ _ _t e'I TOTAL LATEMT EFF l EE :
w m \. - Oo \ : 60 w m
\ -
o- -2
'\-
-2 U 5= 18 -- \ -
- .18 ho \ 5
'? 5= :
C "Eto a . E" -3 -3 5 le -- -
- le t5 :
E 5 :
-4 -A
. ,t ,,e ie ...
-.i
,,,. ...,,,. ., g . . . ., ,3 ,, ,4 ,,,,.
MUMeER OF HEALTH EFFECTS FIGURE C-26. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY'S2-BM FOR 2-MILE EVACUATION (RUN NUMBER 145)
l 18 IE8 ..10I
.., . ..,8-. 0 . _. . .w.
....80 . .,8 0
.. ..,j g 5 w , :
m \* - E ' ACUTE FATALITIES r \*
$m g (
......., A.C..U.T.E...I.M.J.U..R.I.E.S..
E bis ._ '
'\- _ta
-I TOTAL LATENT EFF o= .
"'0
- yw ......, g .
m , u* *
- o. -2 1 5 3m -2
{ g 810 u
- - i j - 10 m : .
, o az i 1- : i m$m 1 a . . E-8 -3 : -3 o 5 18 '~ __ i _tg t "i i,
~-
E i 0 \: '.
-4 .
-4 1g ... .i .. ... ... tg tes' "'t 81 ' '182 ' t g 3 ' "'t g 4 ' "'t g 5 ' 'tg6 Hut 1BER OF HEALTH EFFECTS FIGURE C-27. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S2-CH FOR 2-MILE EVACUATION (RUN NUMBER 165) 1 M M M M M M M M M M M M M M M M M
jg IO ,
,......f0 0 . .. ..
- - - - -- mv s . c,- -
g g "3 "3 . w N-s.s. N . o - ACUTE FATALITIES c \ . gE. , A. .C. U.T. E. . ... .I.M. .J.U. R.I. E.S. . g . o z U gg-1.- . t g'I TOTAL LATEMT EFF o g g l $m a o l. nwu .
. \. . ,z ............,,, g ,
- o. -2
-2 3m ....-.., ) yg u
r u818 .-
)*
n U ' o . l.
>g ....
I E' -3 -3 o 5 18 ~- 'I, - 18 b5
~
l i e i :
-4 -4 18 ..' ..i ... .. ..i 18 188' "'181' '182 ' t g 3 ' "'t g4 ' "'t g5 ' 'tg6 MUMBER OF HEALTH EFFECTS i
l FIGURE C-28. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S3-CH FOR 2-MILE EVACUATION (RUN NUMBER 149)
f ta' 188 ..J 81 j e3 je3 ..,j s4 ,,j e5
.r.g , , ..
"N. f
~
M '\- cr ACUTE FATALITIES gw \ o - w -1 '\ -t f.Cu,TE,,J,y,U3,QS,
" d 1a __ 'S __tg TOTAL LATEMT EFF o=
w ,i w m ) - o . - x ) - l w w . . m= og 1 - D zgE 18_- -:--- )
._t e
n ao o j : z , a o . E" -3 -3 s te _~ _ _ts C' _? i E : 5
~
-4 -4 1e ... .i .. ... ... tg 188' "'t s t ' t g2 ' t g 3 ' "'164 ' "'t g 5 ' 'tg6 Mu?1BER OF HEALTH EFFECTS FIGURE C-29. COMPLEMENTARY CUMULATIVE DISTRIBUTION FUNCTIONS FOR HEALTH EFFECTS RISK FOR RELEASE CATEGORY S6-BM FOR 2-MILE EVACUATION (RUN NUMBER 134)
)
i l M M M M M M M M M M M M M M M M M M
T- l M_ M R R. O M M. U M _. .O M_ M. M_ M
- t. M. T~~T
] \
Ig 8 188 ..j 8I -- . . . . . j 8 2 .j 83 4- .j g5
.. ..,j 8 ..
w , E \. \ ACUTE FATALITIES
, , A,Cy,JE,,,I,M,J,y,R,IE,S, EwU i g'I__ * .. . ._1g
-I TOTAL LATENT EFF o m w m s*g :
Uo X N.*
\. ~
~
wU \ \
~
$6 E -2 i j .
-2 l D z
58 -p 1
)
- __ta i a -
l Mo \ : l l
?w 55 "3o \. :
a . E8 -3 -3 o U 18 . __10 t-M : E 5 :
-4 l -4 10
,,,,. 'g.
., .,.' ,3
, , ,4 ,>, ,5 1
. , ,,0}}