ML20028H600: Difference between revisions
StriderTol (talk | contribs) (StriderTol Bot insert) |
StriderTol (talk | contribs) (StriderTol Bot change) |
||
| Line 3,434: | Line 3,434: | ||
* O e> c gg 1 O O O O O u *'a W ct I' 0 L C C " W C | * O e> c gg 1 O O O O O u *'a W ct I' 0 L C C " W C | ||
< n n , a n '' ,,) s w y a 2 < < & C O, 0, 0, 0, 6, O, | < n n , a n '' ,,) s w y a 2 < < & C O, 0, 0, 0, 6, O, | ||
(( | |||
SR R C E | SR R C E | ||
* C 0 R | * C 0 R | ||
| Line 4,497: | Line 4,497: | ||
. . . - . . - ~ . _ _ _ . _ - _ _ - _ . . _ . _ _ _ . _ , . _ - . _ - , _ . - . - , _ . _ . _ - - . _ , , . _ _ _ . . - _ _ _ , _ . _ _ . . _ _ . . - | . . . - . . - ~ . _ _ _ . _ - _ _ - _ . . _ . _ _ _ . _ , . _ - . _ - , _ . - . - , _ . _ . _ - - . _ , , . _ _ _ . . - _ _ _ , _ . _ _ . . _ _ . . - | ||
i ! It ; |
| Issue date: | 12/31/1990 |
| From: | Bohn M, Susan Daniel, Hashimoto P, Jerrica Johnson, Lambright J, Mraz M, Ravindra M, Tong W EQE, INC., SANDIA NATIONAL LABORATORIES |
| To: | NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| References | |
| CON-FIN-A-1228 NUREG-CR-4550, NUREG-CR-4550-V3R1P3, NUREG-CR-4550P3, SAND86-2084, NUDOCS 9101090373 | |
| Download: ML20028H600 (393) | |
Text
{{#Wiki_filter:. .. _ _ NU REG /CR-4550 SAND 86-2084 Vol. 3, Re',.1, Part 3 Ana:ysis of Core Damage Frecuency: Surg Power Station, Unit 1 External Events Prepared by M. P. Ilohn, J. A. Lambright, S. L Daniel, J. J. Johnson, M. K. Ravindra, P. O. Hashimoto, M, J. Mraz, W. II. Tong Sandia National Laboratories Operated by Sandia Corporation Prepared for U.S. Nuclear Regulatory Commission PD A OC O 000 80 P PDR l 1
- - . - . . . - _ . _ _ _ . . _ x . .. -.
- - - - - . . - ;-. . - .. _ . - - .-. --- _ -- - ~ ~ . . - . . , , ~
9 l AVAILA'DILITY NOTICE
- Avaltabihty of Reference Materialr> Cated in NRC Pubhcatms ,
Most documents cited in NRC publications will be available from one of the following sources:
- 1. The NRC Public Document Room 2120 L Street, NW, Lower Level, Washington, DC 20555
- 2. The Superintendent of Documents, U.S. Government Printing Office P.O. Box 370A2, Washington,
- DC 20013 7082 -
- 3. The National Technloal information Service, Springfield, VA 22161 Although the 6 sting that follows represents the majority of documents cited h NRC pubhcations, it la not intended to be exhaustive, Referenced documents available for hspection and copylng for a fee from the NRC Pubuo Document Room
. Include NRC correspondence and internal NRC memoranda: NRC Office of inspection and Enforcement bu6etins, circulars, information notloes, inspection and investigation notloes: Licensee Event Reports: ven-dor reports and correspondence: Commission papers; and appicant and Scensee documents and corre-sDondence.
The following documents in the NUREG series are avanable for purchase from the GPO Sales Program: formal NRC staff and contractor reports NRC-sponsored conference proceedings, and NRC booidets and brochures Also avalable are Regulatory Guides NRC regulations h the Code of Federal Regulaflons, and Nuclear Regulatory Commission Issuances, Documents avalable from the National Technical information Servloe include NUREG series reports and technical reports prepared by other federal agenotes and reports prepared by the Atomic Energy Commis-sion, forerunner agency to the Nuolear Regulatory Commission, Documents avalable from publo and spoolal technical libraries include.aI open IterMure items, such as books, joumal and periodloal articles, and transactions, Federe! Register nottoet, federal and state legista. tion, and conpessional reports can usually be obtained from these traries. Documents such as theses, dissertations, foreign reports and translations, and non-NRC conference pro. i ceedings are available for purchase from the organization sponsoring the publication cited, Single ooples of NRC draft reports are avalable free, to the extent of eJpply, upon written request to the Office of Information Resources Management, Distribution Section, U.S Nuclear Regulatory Commission,
; Washington, DC 20555,
..Coples of industry codes and standards used in a substantive manner in the NRC regulatory process are maintained at the NRC Library,7920 Norfolk Avenue, Bethesda, Maryland, and are avaRable there for refer-ence use by the pubbo, Codes and standards are usuaBy copyrighted and may be purchased from the originating organization or, if they are American National Standards, from the American National Standards Institute,1430 Broadway, New York, NY .10018, DISCLAIMER NOTICE This report was propared as an account of worl< sponsored by an agency of the United States Govemment.
Neither the United States Govemment nor any agency thereof, or any of their employees, makes any warranty, =
- exprosed or implied, or assumos any legal liability of responsibility for any third party's use, or the results of I such uso, of any information, apparatus, product or process diseinsed in this report, or represents that its uso {
by such third party would not infringo privatoly owned rights, l l H
m NUREG/CR-4550 SANDS 6-2084 Vol. 3, Rev.1, Part 3 Analysis of Core Damage Frequency: Surry Power Station, Unit 1 External Events . Manuscript Completed: November 1990 L) ate Published: December 1990 Prepared by M. P. Bohn, J. A. Iambright, S. L Daniel, J. J. Johnson *, M. K. Ravindra', P. O. liashimoto', M. J. Mraz*, W 11. Tong
- i Sandia National l2boratories
}-. Albuquerque, NM 87185 Prepared for Division of Systems Research Ollice of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission Washington, DC 20555 NRC FIN A1228
-
- EQE, Inc,
ABSTRACT
'The U.S. Nuclear - Regulatory Commission has sponsored probabilistic risk assessments of five operating commercial- nuclear power plants . as part of a major update of the understanding of risk as provided by the original WASH 1400 risk assessments. In contrast to the WASH-1400 studies, the NUREG 1150 risk assessments include a detailed analysis (for two plants) of risks due to earthquakes, fires, floods, etc. (which are collectively i known as " external events") . This report presents the external events probabilistic risk assessment for the Surry Power Station (Unit 1).
In keeping with the philosophy of the internal events analyses for NUREG-1150, which are intended to be " smart" PRAs making full use of all insights gained during the past ten years' developments in risk assessment methodologies, the . corresponding external event analyses performed by newly developed methods which are an improvement over past methodologies in terms of completeness and reproducibility and which, in many cases, provide significant simplifications in calculational effort. These methods have been development at Sandia National Laboratories (SNL) under the sponsorship of the NRC's Division of Systems Research as part
-of their Dependent Failure Methodology Development Program.
As. -a first step, an extensive screening analysis was performed which showed thct all external events had a negligible contribution except w fires and seismic events. Detailed analyses for fire and seismic events were then performed. The final analysis of internal fires resulted in a total (mean) core - damage f requency of 1.13E 5 per year. The final analysis ' of the seismic risk resulted in a total (mean) core damage frequency of li16E 4 per year using hazard curves developed by Lawrence Livermore National Laboratory -(LLNL) . The mean seismic core d,' mage frequency was also- calculated using hazard curves devolepod- by the Electric Power Research Institute (EPRI) and found to be 2.50E-5 per
-year. Uncertainty analyses were performed for both fire and seismic events, and dominant components and sources of uncertainty were identified.
-111/1v-
i TABLE OF CONTENTS
-Seetion . Efqgt
-ABSTRACT 111 FORWARD xiii EXECUTIVE
SUMMARY
EXEC 1
1.0 INTRODUCTION
11 1.1_ The NUREG 1150 Risk Analyses 11 :I 1.2 The External Event Methodology 1-2 1.3 -Steps in the Analysis 14 i 1.3.1. Plant Walkdown and Data Gathering 14 1.3.2 Screening of Other External Events 15 1.3.3 _ Seismic Risk Assessment Methodology-16
- 1. 3 ~. 4 Fire Risk Assessment Methodology 1 10 1.4 References 1 12
- 2.0 PLANT DESCRIPTION 21-2.1 Plant,-Site and Conoral Characteristics- 21 '
2.2 Description of Plant Systems 2-1 2.2.1 Introduction 2-1 i 2.2.2 Containment Spray System 21 2.2.3 High Pressure Injection / Recirculation System 22 ' 2.2.4- -Accumulator System 24 2,2. 5 - Low Pressure Injection / Recirculation System 28 2.2.6 .Inside Spray Recirculation System 2-10 2.2.7 Outside Spray Recirculation System 2-10
~2.2.8 Auxiliary Feedwater System 2-12 2.2,9 Primary: Pressure Relief System
~
2 15 2.2.10 Power Conversion-System 2 15 2.2.11 Charging Pump Cooling System 2 15 2.2.12 Service Water System 2 17_ , 2.2.13 Component Cooling Water System 2 19 i
~2.2.14 Emergency Power System 2 19 2 '. 2 .15 - Safety Injection Actuation System 2-22 2.2.-16 Consequence Limiting Control System 2 26 2.2.17 _ Recirculation Mode Transfor System 2 26 !
2.2.18 Residual Heat Removal System 2 26 3 1 2,3 . Initiating Events and Accident Sequences- 2 30 2.3.1 Introduction' 2 30 2.3.2 Ti (Loss of Offsite Power) Event Tree 2-30 2.3.3 T3 (Turbine Trip with MFW Available) Event Tree 2 37 2.3.4 le.rge LOCA Event Tree 2-40 2.3.5 Medium IDCA Event Tree 2 i 2.3.6 Small LOCA Event Tree 2-44 2.3.7 Very Small LOCA Event Tree 2-46
-v-4
i CONTENTS (Continued) Section Par.e
-3.0 SCOPING QUANTIFICATION STUDY 31 i
3.1 General Description 31 3.1.1 Site 31 3.1.2 Plant 38 3 1.3 Site Visit 38 3.2 Initial Screening of External Events 3-9 3.3 Screening of External Events Based on FSAR and Site Hazard Studies. 3 16 3.3.1 Accident in Industrial and Military Facil tties 3 16
- 3. 3.2- Transportation Accidents 3-18 3.3.3 Release of On-site' Chemicals 3-23 3.4 Bounding Analyses 3 24 3.4.1 Extreme Winds and Tornadoes 3-28 3.4.2 Pipeline Accidents -3 34 3.4.3 Turbine Missiles 3 35 3.4.4 External Flooding 3 39 3.4.5 Aircraft Impact 3-41 3,4.6 Internal Flooding 3 46 3.5 Summary 3-52 3.6 References 3 53 4.0 SEISMIC PRA 41
- 4. l~ Seismicity and llazard Curves 4-1 i 4.1. l' Ceneral Considerations 41 4.1.2 Hazard Curves Used for.Surry 4 4.2 ' Response _ Calculations; 4-6 4.2.1 Introduc. tion 46
~
4.2.2 Site and Seismic Characteristics 48-4.2.3 Probabilistic Response Analysis. 4-11
~
4.2.4 Safety Related Component Responses 4-38 4.3 Seismic Fragilities 4-50 4.3.1 Generic Fragilities 4-50 4.3.2 Surry Site-Specific Component Fragilities 4 54 4.3.3 Site-Specific Building Fragilitles 4 54 4.3.4 Structure Frag 111 ties Derived for Surry 4-61 4.3.5 Liquefaction 4-70
-vi-
CONTENTS (Continued) S_tction IMLt 4,4 Core Damage and Risk Computations 4 70 4.4.1 Initiating Events 4 70 4.4.2 Event Trees 4 75 4.4.3 Failure Modes of Safety Systems 4-75 4.4.4 Accident Sequence Evaluation 4 81 4.4.5 Base Case Surry Results 4 84-4.4.6 Base Case Importance Studies 4 100 4.4.7 Summary and Plant Specific Insights 4 105 4.5 References 4 106 5.0 SURRY FIRE ANALYSIS 51
- 5.1 Introduction 51 5.2 Fire Locations Analyzed 5-4
- 5. 2 .1 - Cable Vault / Tunnel (Fire Area 1) 54 5.2.2 Emergency Switchgear Room (Fire Area 3) _5-4 5.2.3 Control Room (Fire Area 5) 54 5;2.4 Emergency Diesel Generator Rooms (Fire Areas 6,.7, and 8) _
5-7 5.2.5 Primary Containment (Fire Area 15) 5-7
'5.2.6 Auxiliary building (Fire Area 17) 5-7 5,2.7 Safeguards Area (Fire Area 19) 5-7 5.2.8- Turbine Building (Fire Area 31) 57 5.2.9 Mechanical Equipment Room #3 (Fire Area 45) 5-8 5.2.10 Charging Pump Service Water Pump Ronm (Fire Arca 54) 58
.5.3 Initiating Event Frequencies 5-8 5;4 Determination 1of Fire-Induced "Off-Normal" Plant- States- 5 11 5.5_ Detailed Description _of thefScreening Analysis- 5-11 5.6 Fire Propagation Modeling 5 7 5.7 Barrier _ Failure Analysis 5 19 5.8 Recovery Analysis 5 20 5.9 Uncertainty Analysis 5 23 5.10 Description of Unscreened Fire-Induced Core Damage Scenarios-and Their Associated Fire Areas 5-24
-5.10.1 Introduction- .5-24
-5.10.2 Auxiliary Butiding 5 24 5.10.3 Cable vault / Tunnel 5-25
.5.10.4 Control Room 5 27 5.10.5 Emergency Switchgear Room 5-29 5.10.6; Charging Pump Service Water Pump Room 5-30 5.11 Conclusions - 5-32 5.12_ References 5-33
-v11-
CONTENTS _(Concluded) APPENDIX A. Surry Structural Floor Response Spectra APPENDIX B Numerical Values of Building Response at Three Excitation Lovels APPENDIX C. Cross Reference File, Boolean Expressions and i Accident Sequences APPENDIX D Critical Components by Fire Area APPENDIX E -Fire Event Data APPENDIX F Soils Liquefaction Analysis for Surry i l i
'l l
')
l
-viii-
---_~ _
FIGURES - Fi p. ore h 2.1- _ Containment Spray System Schematic 23 2.2_ High Pressure Injection / Recirculation System Schematic 25 2.3 Accumulator System Schematic 27 2.4 Low Pressure Injection / Recirculation System Schematic 29 2.5- -Inside Spray Recirculation System Schematic 2-11 2.6 Outside Spray Recirculation System Schematic 2-13 2.7 Auxiliary Feedwater System Schematic 2-14 2.8 Primary Pressure Relief System Schematic 2 16 2.9 Charging Pump Cooling System Schematic 2-18 2.10 Service Water System Schematic 2 20 2.11 Component Cooling Water System Schematic 2 21 2.12 Emergency Power System Logic Schematic 2 23 2.13 Safety Injection Actuation System Logic Diagram- 2-25 2.14 Consequence Limiting Control System Logic Diagram 2 27 2.15 Recirculation Mode Transfer System Logic Diagram 2-28 2.16 Residual Heat Removal System Schematic 2-29 2.17 Event Tree for Ti -Loss of Offsite Power 2-36 2.18 Event Tree for T3 -Turbine Trip with MW 2 39 2.19 . Event Tree for A -Large LOCA 2 41 2.20 Event Tree for S -Medium t LOCA 2-43 2.21 Event Tree for $ --Small 2 LOCA 2-45
- 2.22 Event Tree for S -3 Very Small LOCA 2-48 3.1 Immediate Environs of Plant Site: Surry Power Station 3-3 3,2 General Topography: Surry Power Station 34 3.3 Local Topography: Surry Power Station 35 3.4 Airports Within 10 Miles of Plant Site 36 3.5 Natural Gas Pipelines Within 10 Miles of Plant Site 3-7 3.6 Computed Surge Levels at the Surry Power Station 3 42 L4.1 LLNL Surry Hazard Curve 4-3 4.2= EPRI Surry Hazard Curve 44 4.3 Surry Power Station General Arrangement 47 4.4- Surry Power Station Substructure Profile 4-9 4.5 Variation of Scil Shear ' Modulus and Damping Ratio
-With Depth 4-14 4.6 .Surry Power-Station Medim Free-Field Input Motion 4 15 4.7 Surry Power Station Reactor Building Model 4 19 4.8 Surry Power Station Auxiliary Building Model 4 20 4 . 9_ Surry Power Station Control Room Structure 4-22 4.10 Surry Power Station Safeguards Area Model 4 4;11 Surry Power Station Containment Spray Pump Enclosure Model 4-25 4.12 Surry Power Station Emergency Generator Enclosure 4 26 4.13 Surry Power Station Intake Structure Model 4-27 ix-
FIGURES (continued) Uturt Llte 4.14 Containment Building Instructure Response'-Level 2 4-28 4.15 Auxiliary Buildlng Instructure Response Level 2 4-30 4.16 Control Room Structure Instructure Response Level-2 4 32 4.17 -Safeguards Building Instructure Response Level 2 4 34 4.18 Containment Spray Building Instructure Response Level 2 4-35 4.19 Emergency Generator Enclosure Instructure Response Level 2 4 36 4.20 Intake Structure Instructure Response Level 2 4 37 4.21_ _ Auxiliary Building Median Responses 4-39 4.22 Control Room Structure Median Responses 4 41 4.23 Emergency Generator Enclosure Median Responses 4 42 4.24 Safeguards Area Medlan Responses 4 43 4.25 Reactor-Building Median Responses 4-44 4.26 RVR_ Initiating Event Frequencies 4 72 4.27 Frequencies of Pipe Breaks Causing LOCAs 4-73 4.28 Initiating Event Hierarchy Event Tree 4 74 4.29- Large IDCA Seismic Event Tree 4 76 4.30~ Hedium LOCA Seismic Event Tree 4 77 4.31 ' Small LOCA Seismic Event Tree 4 78 4.32 Ti (Loss of Offsite Power) Seismic Event Tree 4 79 4.33 T 3 (Turbine Trip) Seismic Event Tree 4 80 x.
TABLES Inble Engg 1.1 List of External Events 13 2.1 Initiating Event Categories Used in the External Events Analysis 2-31 2.2 Event Tree lleadings 2 32 3.1 Preliminary Screening of External Events for Surry Nuclear Power Station 3 11 3.2 Chemical Compounds Used and/or Stored Near Surry 3 17 3.3 Chemical Compounds Shipped on the James River 3 19 3.4 Chemical Compounds Transported by Truck on Virginia Highway 10 3 22 3.5 Surry On-Site Chemical Spill Analysis 3-24 3.6 Surry 1 and 2-Toxic Chemical Sourco Locations 3-25 3.7 Peak Concentration of Chemicals in Control Room 3-27 3.8 Structures and Components Designed for Seismic and Tornado Criteria 3-29 3.9 Maximum Probable Flood Protection Levels for Class I Structures 3-43 3.10 Airports Within 25 Miles of the bite 3-44 ; 3.11 Surry Flooding Vital Area Analysis Summary 3-48_ ! 4.1 LLNL Hazard Curves Values 45 4.2 EPRI Hazard Curves Values 45 4.3 Surry Power Station Low Strain Soil Properties 4 11 4.4 Surry Power Station Strain Compatible Soil Properties 4 12 4.5 Free-Field Acceleration Time Histories 4-13 4.6 Surry Power Station Foundation Models 4 17 4.7 Surry Seismic Response Locations 4 45 4.8 Rules for Assigning Response Correlation 4-48 4.9 _ Correlation Coefficients Between Responses 4 49 4.10- Generic Component Categories 4-52 4.11 Generic Component Fragilities 4 53 4.12 - Summary of Surry Sito-Specific Fragility Functions 4 55 4.13 Surry Structural Fragilities Summary 4 62 4.14 Seismic Accident Sequences 4-85 4.15 - Safety System Nomenclature 4-87 4.16 Accident Sequence and Total Core Damage Frequencies 4 89 4.17 Base Case Accident Sequence Frequency - LLNL Hazard 4 90 4.18 Base Case Accident Sequence Frequency - EPRI Hazard 4 91 4.19 Mean Initiating Event Frequencies LLNL llazard 4-94 4.20; Mean Dominant Accident Sequence Frequencies - LLNL Hazard 4-95
- 4. 2 l- Mean Core Damage Contributions at Intervals LLNL Hazard 4 96
.4.22 Mean Initiating Event Frequencies - EPRI Hazard 4 97 4.23 Mean Accident Sequence Frequencies - EPRI Hazard 4-98 4.24 Mean Core Damage Contributions From Dominant EPRI Hazard 4-99 l
-xi-l
TABLES (continued) IIthlt fafa 4.25 Dominant Component Contributors to P(cm) Ranked by Risk 4 101 4.26 Comparison of Contributions of Modeling Uncertainty in Response, Fragility and Hazard Curves to Core 4-102 4.27 . Comparison of Contributions of Modeling Uncertainty in Response, Fragility and-Hazard curves to Core 4 102 4.28 Comparison of Mean Hazard Curve Probabilities from Ten Discrete Hazard Curves - LLNL Hazard 4-104 4.29 Comparison of Mean Hazard Curvo Probabilities f rom Ten Discrete Hazard Curves EPRI Hazard 4 104 5.1 Surry Fire Area Core Damage Frequency 5-1 5.2 Dominant Accident Sequence Core Damage Frequency Contributors 52 5.3 Surry Fire Areas Containing Safety Related Components . 55 5.4 Statistical Evidence of Fires In LWRs (as of June 1985) 59 5.5 Surry Fire Initiating Event Frequencies (/yr) 5 12 5.6 Surry Fire Induced Initiating Events Analyzed 5-13 5.7 Modified COMPERN III Input Parameters 5 19 5.8 Time to Damage Critical Cables (minutes) Using the Modified Version of COMPBRN III 5 20
- 5.9~ Critical Area Ratios- 5 21 5.10 Approximate Number of Barriers at a Plant 5-22 5'11 Estimates of Single Barrier Failure Rates
. 5 22 5.12 Auxiliary Building Fire Scenario Factors and Distribution 5 26 5.13 Cable Vault / Tunnel Fire Scenario Factors and Distribution 5 28 5.14 Control Room Fire Scenario Factors and Distribution 5-29 5.15. Emergency Switchgear Room Fire Secuario Factors and Distribution 5 31 5.16- CPSWPR Fire Scenario Factors and Distribution 5-31 I
1
*xil-
FOREWORD l This is one of numerous documents . that support the preparation of the NUREG ll5O document by the NRC Office of Nuclear Regulatory Research, Figure 1 Illustrates the front end documentation. There are three interfacing ~ programs performing this work: the Accident Sequence Evaluation Program (ASEP), the Severe Accident Risk Reduction Program (SARRP), and the Phenomenology and Risk Uncertainty Evaluation Program (PRUEP). The Zion PRA was performed at the Idaho National Engineering Laboratory and at Brookhaven National Laboratory. Table 1 is a list of the original primary documentation and the corresponding revised documentation. There are several items that should be noted. First, in the original NUREG/CR-4550 report, Volume 2 was to be a summary of the internal analyses. This report was deleted. In Revision 1, Volume 2 now is the expert judgment clicitation covering all plants. Volumes 3 and 4 include external events analyses for Surry and Peach' Bottom, respectively, The revised NUREG/CR 4551 covers the analysis included in the original NUREG/CR-4551 and NUREC/CR 4700. However, it is different from NUREG/CR-4550 in that the results from the expert judgment clicitation are given in, four parts to Volume 2 with each part covering one category of issues. The accident progression event trees are S iven in the appendices for each of the plant analyses, Originally, NUREG/CR-4550 was published without the designation " Draft for Comment. " Thus, the final revision of NUREG/CR-4550 is designated Revision 1. The label Revision 1 is used consistently on all volumes except Volume 2, which was not part of the original documentation. NUREC/CR-4551 was originally published as a "Draf t for Comment" so, in its final form, no Revision 1 designator is required to distinguish it from the previous documentatation.
.There aro several other reports published in association with NUREG 1150.
'These are:
NUREG/CR-5032, SAND 87 2428, Modeling Time to Recovery and initiating Event Frecuency for Loss of Off site Power Incidents at Nuclear Power Plants, R. L. Iman and S. C. Hora, Sandia National Laboratories, Albuquerque, NM, January 1988. NUREG/CR 4840, SAND 88-3102, Procedures for Externni Event Core Damage Freauency Analyses for NUREG-1150, M. P. Bohn and J. A. Lambright, .
)
Sandia National Laboratories, Albuquerque, NM, November 1990, 1
-xiii-
i
- A u_
[ METHODOLOGY gg
\ TE
[ EXPERT PANEL RESULTS g [ PROJECT STAFF RESULTS W ?. c
\ 11 h [ INTERNAL EVENTS g
g m [lNTERNAL EVENTS APPENDIX gc "E EXTERNAL EVENTS G$ y
$! [\
,o m T $
@ z y flNTERNAL EVENTS g g
$ ]{ [lNTERNAL EVENTS APPENDIX g' g M $ g [ EXTERNAL EVENTS ~{ga !
q e ~
\ w >
/ O 0
8 g () g (INTERNAL EVENTS C E [lNTERNAL EVENTS APPENXIX f@ , S [lNTERNAL EVENTS y
$ [lNTERNAL EVENTS APPENXIX (@;g h 3$
a, --t
=
a y g y o n
<y 2 $ [ INTERNAL EVENTS e5 C a x y= I g N m o ' C
[ RISK MANAGEMEPR
\ a
- Ul I
1 Q[ g O z s E [8ACK END A g me-a g f- - g O [ DOCUMENTATION _
= AA L= o/I --
h$h
$WM
~ ,1 \
m n-
I Table 1:. < NUREG-1150 Analysis Documentation
-i Oririnal Documentation .
NUREG/CR-4550 . NUREG/CR-4551 NUREG/CR-4700 Analysis of Core Damage FrequencyE Evaluation of Severe Accident Containment Event Analysis From Internal Events Risks and the Potential for for Potential Severe Accidents' Risk Reduction
.t Volume 1 Methodology .
-Volume 1 Surry Unit 1 Volume 1 Surry Unit 1 2 Summary (Not Published)- 2 Sequoyah Unit 1 2 Sequoyah Unit-1 3 Surry Unit 1 3 Peach Bottom Unit 2 3 Peach Bottom Unit 2 4 Peach' Bottom Unit 2 4 Grand Gulf Unit 1 4 Grand Gulf Unit 1 5 Sequoyah Unit'1 ,
6 Grand Gulf Unit 1 7 Zion Unit 1
, Revised Documentation ,
x 7 NUREG/CR-4550, Revision 1 NUREG/CR-4551, Evaluation Analysis of Core Damage Frequency of Severe Accident Risks Volume 1 Methodology Volume 1 Methodology -i 2 Part 1 Expert ' Judgment Elicit. --Expert Panel 2 Part 1 Expert Judgment Elicit.--In-vessel ' Part 2 Expert Judgment Elicit.--Project Staff Part 2 Expert Judgment Elicit.--Containment 3 Part 1 Surry Unit 1 Internal Events Part. 3 Expert Judgment Elicit.--Structural Part 2 Surry Unit 1 Internal Events App. Part 4 Expert Judgment Elicit. --Source-Term Part 3 Surry Unit 1 External Events Part S Expert Judgment Elicit. --Supp. Cale. 4 Part 1 Peach Bottom Unit 2 Internal Events Part 6 Expert Judgment Elicit.--Proj . Staff Part 2 Peach Bottom Unit 2 Internal Events App. Part 7 Expert Judgment Elicit. --Supp. Calc. Part 3 Peach'Bott';m Unit 2 External Events Part 8 Expert Judgment Elicit. --MACCS Input 5 Part 1 Sequoyah Unit 1 Internal Events 3 Part 1 Surry Unit 1 Anal. and Results Part.2 Sequoyah Unit 1 Internal Events App. Part 2 Surry Unit 1 Appendices 6 Part 1 Grand Gulf Unit 1 Internal Events 4 Part 1 Peach Bottom Unit 2 Analz and Results Part 2 Grand Gulf Unit 1 Internal Events App. Part 2 Peach Bottom Unit 2 Appendices 7 Zion Unit 1 Internal Events 5 Part 1 Sequoyah Unit 2 Anal. and Results Part 2 Sequoyah Unit 2 Appendices , - 6 Part 1 Grand Gulf Unit 1 Anal. and Results l Part 2 Grand Gulf Unit 1 Appendices 7 Part 1 Zion Unit 1 Anal. and Results 4 Part 2 Zion Unit 1 Appendices
NUREG/CR-4772, SAND 86 1996, Accident Secuence Evaluation Procram Humnn Reliability Analysis Procedure. A, D. Swain III, Sandia National Laboratories, Albuquerque, NM, February 1987. NUREG/CR 5263, SAND 88 3100, The Risk Management Implications of NUREG-
,1150 Methods and Resulth A. C. Camp et al., Sandia National Laboratories, Albuquerque, NM, December 1988.
A Human Reliability Analysis for the ATWS Accident Secuence with MSIV Closure at the Peach hottom Atomic Power Station._ A 3272, W. J. Luckas, Jr. et al. , Brookhaven National Laboratory, Upton, NY,1986. A brief flow chart for the documentation is given in Figure 2. Any related supporting documents to the back end NUREG/CR 4551 analyses are delineated in NUREG/CR 4551. A complete list of the revised NUREG/CR-4550, volumes and parts is given below, General NUREG/CR 4550, Volume 1, Revision 1, SAND 86 2084, Analysis of Core Damage Frecuency: Methodolony Guidelines for Internal Events. NUREG/CR-4550 Volume 2, SAND 86 2084, Analysis of Core Damage Freauency from Internal Events: Expert Judgment Elicitation on Internal Events
' Issues - Part 1: Expert Panel Results. Part 2: Project Staff Results.
Part 1 and 2 of Volume 2, NUREG/CR 4550 are bound together. This volume was not part of the original documentation and was first published in April 1989 and distributed in May 1989 with the title: Analysis of Core Damage Frequency from Internal Events: Expert Judgment-Elicitation. In retrospect, a more descriptive title would be; Analysis of Core Damage Frequency: Expert Judgment E11 citation on Internal Events Issues, SURRY NUREG/CR-4550, Volume 3, Revision 1, Part 1, SAND 86-2084, Analysis of Core Damnge Frecuenev: Surry Unit 1 Internal Events. NUREG/CR-4550, Volume 3 Revision 1, Part 2, SAND 86 2084, Analvnis of Core Damage Freauency: Surry Unit 1 Internal Events Annendices. NUREG/CR-4550, Volume 3, Revision 1, Part 3, SAND 86-2084, Analysis of Core Damage Freauenev: Surry Unit 1 External Events.
-xvi-
FR0!!T-END ANALYSIS BACK-EtID A!!ALYSIS IlUREC/CR-4550 NUREC/CR-4551 REVISIDH'l FLANT DAMACE STATE FREQUENCIES StfRRY ACCIDENT FROGRESSION SURRY : UNIT I UllIT I' 8 RISK REDUCTION AND A!!D RISK-UNCERTAItiTY llEASURES l I i l _ -NURE0/CR-4550 REVISI0ti'l VOL.1 HETil0DOLOGT ] BACK-END SUFFORT
'l DOCUMENTATION
, _. HURE0/CR-4550 REVISIDH 1 ,
x VOL. 2 EIFERT OFINION .SURRY
,FTACM BOTTIM NUREG-
_ HURED/CR-4840 EXTERNAL 1150 ,SEQUOTLL L EVENTS METHODS .gRATID CULF
, ZION ,
_ NURE0/CR-4772 IIRA PROCEDURES _ NURE0/CR-5032 LOSF IE FREQ AND RECOVERY Figure.2. Surry Related Docunentation.
Peach Bottom NUREG/CR 4697, EGC 2464, Containment Venting Analysis for the Peach Bottom Atomic Power Station. D. J. Hansen et al., Idaho Nativaal Engineering Laboratory (EG6G Idaho, Inc.) February 1987. NUREG/CR 4550, Volume 4, Revision 1, Part 1, SAND 86 2084, Analysis of Core Damare Frecuency: Peach Bottom Unit 2 Internal Events. NUREG/CR 4550, Volume 4, Revision 1, Part 2, SAND 86 2084, Analysis of Core Damage Freauency: Peach Bottom Unit 2 Internal Events Annendices. NUREC/CR 4550, Volume 4, Revision 1, Part 3, SAND 86 2084, Analysis of Core Damare Frecuenev Peach Bottom Unit 2 External Events, Seodovah NUREG/CR 4550, Volume 5, Revision 1, Part 1, SAdD86 2084, Analysis of Core Damage Freauenev: Seouovah Unit 1 Internal Events, NUREC/CR-4550, Volume 5, Revision 1, Part 2, SAND 86-2084, Analysis of Core Damare Frecuency: Seouovah Unit 1 Internal Events Anoendices. Grand Gulf NUREC/CR-4550, Volume 6, Revision 1, Part 1, SAND 86-2084, Analysis of Core Damare Frecuency: Grand Gulf Unit 1 Internal Events. NUREG/CR-4550, Volume 6, Revision 1, Part 2, SAND 86-2084, Analysis of Core Damage Frecuency: Grand Gulf Unit 1 Internal Events Annendices. 2d2D NUREG/CR-4550, Volume 7, Revision 1, EGG-2554, Analysis of Core Damage Frecuenev: Zion Unit 1 Internal Events,
-xviii-
EXECUTIVE
SUMMARY
The U.S. Nuc1 car Regulatory Commission (USNRC) is sponsoring probabilistic risk assessment of five operating commercial nuclear power plants as part of a major update of the understanding of risk as provided - by the original WASH-1400 assessments. In contrast to the WASit 140C studies, .two of the NUREG 1150 risk assessments will include a detailed analysis of ricks due to earthquakes, fires, floods, etc., which are collectively known as "exte rnal - events . " The two plants for which external events were analyzed are Surry and peach Bottom, a PWR and a BWR, respectively. This report presents the results obtained for - the Surry (Unit 1) cxternal events core damage frequency assessment. In koopin6 with-the philosophy of the internal events analysen for NUREG-1 11150, which are intended to be " smart" PRAs making full use of all-insights: gained' during the - past ten years' developments in risk assessment methodologies, the corresponding external event analyses have been performed by newly developed methods. The methods have been - developed under NRC sponsorship and represent, in many cases , . both advancements and simplifications over techniques. that have been used in past years. They also include the most up to date data bases on equipment seismic fragilities, fire occurrence frequencies -and fire damageability thresholds. These methods were developed at Sandia National' Labratories under the sponsorship of the USNRC's Division of Systems Research as part of their Dependent Failure Methodology Development Program. The first application of these new methods-was to the seismic analysis of six- power plants as part of the NRC program 'for the resolution 1of Unresolved Safety Issue _ USI A 45 - Adequacy. of Decay
!! cat Removal Systems. _ Extension of these methods to fire, . flood, etc. ,
= bas been continuing'during recent years.
In contrast to most ~ past external - event analyses ,- wherein rudimentary
-systems models --wero developed reflecting each external -event : under consideration, the -NUREG-1150 ' external event analyses are based on the full internal event PRA systems models (event. trees and fault trees) and make _use of extensive computer-aided screening to reduce them to accident sequencercut sets - important. to each external event. This provides two major advantages in -that both consistency and scrutability with respect to the internal event analysis is achieved,'and the full gamut of random and_ test / maintenance unavailabilities are -automatically included, while-only those probabilistically important survive - the screening - process.
- Thus, full bonefit of tho' internal event- analysis is obtained by performing the internal and external event analyses sequentially.
~
EXEC-l 1
The external event analysis began with a review of the FSAR, related design documents and the systems descriptions in the internal events PRA. Important components were located on general arrangement drawings. The utility fire study prepared to meet Appendix R of 10CPR50 requirements formed-the basis for the initial identification of fire and flood area boundaries and barriers. Shortly thetcaf ter, a plant visit of 3 days duration was made, involving an integrated team of specialists in the various external events. Based on the plant walkdown and the screening analysis described in Chapter 3, all external hazards were screened out based on probability considerations except for seismic and fire events. The seismic risk assessment was the critical path item due to the time required to assemble-the structural drawings and.models, A best estimate-structural dynamic response calculation for each building containing equipment important to safety was made using models used in the original
-design. The results were distributions for floor slab accelerations, and estimates of variability .and correlations. Component fragilities were obtainad either from a generic data base or derived on a plant specific basis as needed. Dual probabilistic screening methods were used to determine important cutsets while allowing for explicit incorporation of correlation. The scismic hazard itself was obtained by extrapolation from the results of the - NRC sponsored Eastern Seismic Hazard Characterization Program performed at Lawrence Livermore National Laboratory (LLNL) and the industry-sponsored Electric Power Research Institute (EPRI) Seismic Hazard Methodology for the Central and Eastern United States Program.
The detailed fire analysis tasks were perfcrmed in parallel. Fire initiator frequencies were obtained from an updated historical data set developed at SNL. Partitioning of building fire frequencies (for which data are availabic) down to sub-area frequencies was based on cable loading, electrical eabinet locations and transient combustible estimates based on walkdown observations and a transient combustible data base developed at-Sandia. Component damage temperatures (rather than auto-ignition- temperatures) were based on SNL fire - tests. The COMPBRN III code was - used to predict component . temperatures in fire areas where growth and separation are important considerations. Critical area analyses using the SETS code provided sequence cut sets for quantification, including barrier failure and random failures as appropriate, A fire detection / suppression histogram developed at SNL was used to incorporate firefighting timing into the analysis. Similar approaches were used for internal and external floods, tornadoes, winds, etc. A maj or economy is achieved by analyzing fires and floods together, and s e i s mic ,- wind and tornado events together, due to the commonality of the analysis _ processes. For example, it is a minor task to extend the seismic fragility derivations to be applicable to wind fragilities. SimLlar economics arise in the screening steps for fires and floods. . EXEC-2
Detailed analysts of internal fires resulted in a total -(mean) core damage frequency of 1.13E-S per year. A detailed seismic analysis resulted in a total (mean) core damage frequency of 1.16E 4 per year using hazard curves developed by Lawrence Livermore National Laboratory. The mean seismic core damage frequency was also calculated using hazard curves developed by the Elcetric Power Research Institute and found to be 2.SOE.S per year. Uncertainty analyses were performed for both fire and seismic events; and dominant components and sources of uncertainty were identified. In general, it was fo~und that only a icw accident sequences dominated the seismic and fire analysis results. For the seismic analysis, the most dominant sequence is a loss of offsite power (LOSP) transient sequence in which the auxiliary feedwater system fails (due to loss of a condensate storage tank) and the high pressure inj ec tion (HPI) system (and hence, the feed and bleed function) fails due to either-failure-of the refueling . water storage tank or failures of the onsite AC power systt - The recond most significant seismic sequence is also a loss of o.fsite power transient sequence, except that this transient sequence leads to a seal LOCA. This is caused by failure of both the llPI system and the component cooling water (CCW) system which leads to the seal LOCA. The.llPI system fails as described above while the CCW system fails due to loss of onsite AC power. Together, these two sequences constitute approximately 67% of the computed seismic core damage frequency. The fire core damage frequency was found to be due to hypothesized fire events in four areas: (a) the emergency switchgear r oo m', '(b) the auxiliary building, (c) the control room, and (d) the cable vault / tunnel. In the case of the emergency switchgear room, cable vault / tunnel, and the auxiliary building, a reactor coolant pump seal LOCA leads to core-damage. The fire itself fails cabling for both the llPI and CCW systems resulting in a seal LOCA, For the control room, a general transient with a subsequent stuck-open PORV 1cada to a small LOCA, Failure to control the plant from the auxilia ry shutdown panel . results in core damage. Together, these four areas gave rise to 99% of the-fire risk. EXEC-3
. - .. . ~. . - - . - -
1.0 INTRODUCTION
1.1 ThE NUREG-1150 Risk Analyses This report describen the Level 1 external events probabilistic risk assessment (PRA) performed for the Surry commercial nuclear power plant as part of the NRC sponsored Accident Sequence Evaluation Program (Ref.
- 1) power plant risk reevaluations, of ten referred. to as the NUREG-1150 program (after the principal document summarizing the results of the program). In contrast to the original WASH 1400 risk assessments (Ref.
2), . both internal and external events risk analyses are being performed in this program. A Level 1 PRA consists of an analysis of plant design and operation focusing on accident sequences that could lead to core damage, their basic causes, and frequencies. Two kinds of accident initiators are considered for a Level 1 PRA, initiating events that occur within the power plant systems themselves and accident initiators caused by events external to the power plant systems. Exampics of external initiators include earthquakes, floods and high winds. The results of both analyses provide assessments of plant safety, design and procedural adequacy, and insights into how the plant functions from the perspective of preventing i core damage. This report documents work performed for the Level 1 external events PRA. It describes the methodology used, assumptions, data _ and models that provide the basis for the work, and the final results. The methods utilized in the NUREG-1150 external events PRAs represent both advancements , and, in many cases , simplifications over techniques that have been used in past years. They include the most up to dato data bases on- equipment seismic fragilities, fire occurrence frequencies and fire damageability thresholds. In :addi ti'on , they provide for minimization of execution time and cost reduction through the use of past
.PRA experience, generic data bases and def ensible ' methodological simplification where possibic. A full description of these procedures is given in Bohn and Lambright (Ref. 3), .The methods were developed to meet the following objectives
- a. _To be consistent with the internal event PRA analyses. The same event trecs/ fault- trees and random, common mode failure and test and-maintenance data are used,
- b. To be transparent. A standard report format provides the data to enable the reader to reproduce the any of the point estimate results.
- c. To be realistic. Best estimate data and models are used. All important_ plant specific failure modes are analyzed,
- d. To.be consistent, The external event analyses are intended to be consistent with the internal event analyses due to common generic
-data , . and methodology, and common level of detail .
1-1
a I 1 i 1.2 The External Event Methodolony The simplified PRA procedures described in this section are based on the following general concepts:
- a. -The external event analyses are based on the internal event risk assessment plant system models and fault trees, and - (other than preliminary data gathering) are not started until the internal events systems analysis (event trees and fault trees) has been finalized.
- b. Vigorous and systematic screening of the full range of external events to which the plant could conceivably be exposed - (e . g. ,
aircraf t crash, external flooding, tornado, extreme wind, etc.) is performed to climinate early all unimportant contributing events.
- c. Simultaneous and coordinated evaluation of all non-negligible external events is performed to minimize data gathering efforts and
. prevent duplication of effort. For example, building fragilities for extreme winds can be derived directly from seismic fragilities.
Also, simultaneous evaluation produces insights into interactions (for - example , seismic-fire interactions) no t - otherwise readily perceived.
- d. In the analysis of each types of external event, computer aided screening techniques and generic failure data are used prior to detailed component failure analysis calculations.
The general steps in the analysis of any external event risk analysis are shown below:
- a. Determine the hazard,
- b. Model plant and systems.
- c. Solve fault trees with screening- techniques to determine non-negligible. cut sets,
- d. Determine responses, fragilities, and correlation for basic events in non negligible cut sets,
- c. Evaluate point estimate sequence and core damage frequencies,
- f. Perform uncertainty analysis and sensitivity studies.
These general steps apply to the full range of external events to which a
- power plant may be exposed. Table 1.1 presents a reasonably complete list af such events. Past PRA experience (Ref. 3) shows that only a very few of these are significant contributors to risk at any given site. In f a r. t , the seismic and fire events are commonly the me-t important 1
1-2 j
. ., . -- . -- . . ~ - - . - . - - . - . . . . . . . - - -
1 1 contributors. In addition, external flooding, tornado or aircraft crashes are less frequent (and usually less significant) contributors. Simplifications in Step (a), hazard determination, have been identified for both the seismic and fire analyses. Computer-aided screening techniques are used for Step (c) for fire, flood and seismic analyses to i reduce the required number of plant-specific component failure , calculations. For Step (d), response determination, seismic design fixed-base structural models are utilized in conjunction with an accurate and fully defensible soil-structure interaction model. While not a simplification, this process has been made very officient by standardization, and use of variabilities and correlation factors derived from previous detailed seismic PRA work. Thus, in each step, defensible simplifications are identified which results, overall, in a cost-effective yet defensible analysis. The procedures used here have been applied (in whole or in part) to six pow 3r plants as part of the U.S. NRC-sponsored ' Unresolved Safety Issue A-45 resolution program (Ref. 4), and have been applied at the N Reactor (Ref. 5) and Savannah River (Ref. 6) Department of Energy reactor facilities. Table 1.1 List of External Events Major PRA Consideration Minor PRA Consideration seismic Lightning Fire Low Lake / River Level Internal Flood Ice Cover Avalanche Forest Fire Industrial Facility Accident _ Landslide Meteorite Volcanic Activity Hall Occasional PRA Consideration External flood Transportation accidents Pipe line accidents Aircraft impact Extreme winds l Tornado I-1-3
t 1.3 Steps in t.ht Mysis 1.3.1 Plant- Walkdown and Data Cathering The Surry external events analysis began with a plant visit in April 1987. The initial visit served as the bassis for the in'tial plant tnformation rer;ue. t s'Amittal. Prior to La first plant visit, the 1 external events team was briefed by the int. .a1 events systems analyst as to _ the general cheracter of safety systems, support systems, system success criteria and critical interdependencies identified to date. In addition, applicable Final Safety Analysis Report (FSAR) sections were reviewed, and a basic set of plant general arrangement drawings were obtained for each team membe.. The team consisted of the followiOB personnel: PRA project Manager M p. Bohn Team 1.cader _ J . A.1.ambright Struct. ural- Fragility Analysts J. J. Johnson, p. O. liashimoto Fire and Flood PRA Analyst J. A. lamt,right External Event Screening Analyst R. Ravidra During the initial walkdown, team members visited all areas conta!ning safety or support equipment except the conteinment. Two full days were adequate for this initial visit. At the completion of this initial visit, the following had been obtained,
- n. A list of components suspected -of being vulnerable to seismic damage and requiring site specific fragility analysis.
- b. A list of potential secondary scismic rtructural failures (masonry walls, etc.) and components potentially damaged by these secondary failures, i
c,. A copy of the civil / structural drawing index for the plant from which needed drawings may be identified.
- d. Sketches of typical anchorage details for important tanks, heat exchangers, electrical cabinets, etc.
- c. A visual evaluation of structural connectivity of floor slabs, wall-to ceiling connections, location of diaphragm cut outs etc., which define load carrying paths. These were to be compared with structu-t1 drawings later.
- f. For each room or. compartment containing essential safety equipment, an identification ' of fire sources -(power cables, pump motors, solvents, etc.), locations of fire barriers, ffre/ smoke detectors, separation of cable trains, etc., and a list of equipment in che room.
1-4 h
- g. For each room or compartment, an identification of flooding sources (tanks, high or low pressure piping), floor drains, purop s , flood valls, flood detectors, etc.
- h. A brief list of key plant personnel or utility engineering / licensing personnel to be contacted later if specific questions arose.
Following the initial plant visit, a list of needed drawings and docuroentation was prepared and sent to the designatsd plant contcet. A second visit to the plant was made by the fire analysis personnel to aDow for cable path tracing and verification. This was undertaken after the prelfininary fire screening analysis had been perfors d based on a review of the plant Appendix R submittal. A final plant visit was made in September 1988. During this final visit initial conclusions as to plant vulnerabilities were reviewed with plant personnel, assumptions wete verified, and final required data was obtained. 1.3.2 Screening of other External Events As mentioned in Section 1.1, the full range of possible external events was considered, but based on the FSAR and the initial plant visit, the + vast majority of the external hazards was shown to have negligible irepac t . The set of general screening criteria which was used is given in the PRA Procedures Guide (Ref. 7) and is summarized as followa: An external event can be excluded if:
- a. The event is of equal or lesser damage potential than the events for which the plant has been designed. This requires an evaluation of plant , design bases in order to . estimate the resistance of plant structures and systen s to a particular external event. For example, it is shown by rennedy, Blejwas and Bennett (Ref. 8) that safety-related structures designed for earthquake - and tornado loadings in T t.e 1 can safely withstand a 3.0 psi static pressure from etplosions, llenc e , if the PRA analyst demonstrates that the overpressure resulting from explosions at a source (e.g., railroad, highway or industrial facility) cannot exceed 3 psi, these postulated explosions need not be considered,
- b. The event has a significantly lower mean frequency of occurrence than other events with similar uncertainties and could not result in worse consequences than those events. For example, the PRA analyst may exclude an event whose mean frequency of occurrence is less than some small fraction of those for other avents. In this case, the uncertainty in the frequency estimate for the excluded event is judged by the PRA analyst as not sL 6nificantly influencing the total risk.
1-5 i
- c. The event cannot occur close enough to the plant to affect it. This is also a function of the stagnitude of the event. Exaroples of such events are landslides, volcanic eruptions and earthquake fault ruptures,
- d. The event is inc it.de d in the definition of another event, or example, storm surges and seiches are included in external flooding; the release of toxic gases from sources external to the plant is included in the ef fects of either pipeline accidents, industrial or military facility accidents, or transportation accidents.
Thesa criteria are usually sufficient to exclude all but a few "other" external events, for those remaining, a simple bounding analysis (Ref.
- 9) will often provide sufficient justification for e xc it.s i o n . The screening and bounding analyses for Peach Bottom are given in Chapter 3.
1.3.3 Seismic. Risk Assessment Methodology A nuclear power plant is designed to ensure the survival of all buildings and escrgemy safety systems in a worst case (" safety shutdown") carthquake, The _ assumptions underlying this design process are deterministic and subject to considerable uncertainty. -It is not-possible, for_ example, to accurately predict the worst earthquake that will occur at -a given site. Soil properties < mechanical properties of buildings, and damping in buildings and internal structures also vary significantly, To model and analyze the coupled phenomena thet contribute to, the total' risk of radioactive release requires consideration of all significant sources of uncertainty as well as all significant interactions. Total risk is then obtained by considering the entire spectrum of possible earthquakes and integrating their calculated consequences. This point underscors.: an important requirenent for a seismic pRA; the nuclear power plant must be examined in its entirety, as a system. A second important aspect which must be addressed in a seismic PRA is that during an carthquake, all parts of the plant are excited simultaneously. Thus, during an earthquake, redundant safety system components-experience highly correlated base motion, and there is a high likelihood that multiple redundant components would be damaged if one is. Hence, the planned for redundancy would be comprised. This " common-cause" failure possibility represt,nti a potentially significant risk to nuclear power plants during earthquakis. S v The simplified seismic risk methodology reported here is basod, in part, on the results of two earlier NRC sponsored programs. The iltst was the Seismic Safety Margins Research Program. In the SSMRP, a detailed seismic risk assessment motl dology was developed. This program
,_ culminated in a detailed evaluation of the seismic risk at the Zion nuclear power station, Bohn-(Ref. 10). In this evaluation, an attempt 1-6
~ - _ .
was made to accurately compute the responses of all walls and floor slabs in the Zion structures, moinent s in the all important piping systems, accelerations of all irnportant valves, and the spectral acceleration at each safety system component (pump, electrical buss, mot or control center, etc.). Correlation between the responses of all components was cornputed from the detailed dynamic response calculations. The irnportant safety and auxiliary systerns functions were analyzed, and fault trees were developed which traced failure down to the indiddual cimponent level. Event trees related the system failures to accident sequei:ces and radioactive release modes. Using these detailed roodels and celcu: ations, it was possiblo to evaluate the seismic risk at Zion, and di tertaine quantitatively the risk iroportance of the components, initiating avents, and accident sequences. The second is the NRC sponsored Eastern Seismic Hazard Characterization program (Ref. 11) which performed a detailed earthquake hazard assessment of all sites east of the Rocky mountains. Results of thess two programs formed the basis for a number of simplifications used in the seismic rnethodology reported here. There arc seven steps required for calculating the seismic risk at a nuclear power plant:
- a. Deterinine-the local earthquake hazard (hazard cure and site spectra),
- b. Identify accident scenarios for the plant which lead to radioactive release (initiating events and event trees),
- c. Determine failure modes for the plant safety and support systems F (lault trees).
- d. Determine the responses (accelerations or forces) of all structures and cornponents (for each carthquake Icvel).
- c. Determine fragilities (probabilistic failure criteria) for the
- important structures and components,
- f. Compute the probability of core damage using the information from Steps (a) through (c) .
g, Estimate uncertainty in the core damage frequencles. only _ the level of detail differentiates a simplified seismic analysis from a detailed seismic PRA, The seven steps of the NUREG 1150 seismic risk analysis procedure are summarized below. . Etep a - Seismic Hazard Characterization e The NUREG.1150 seismic analyses make use of hazard curves obtained from two recent programs aimed at developing sets of hazard curves based on 1 17 j L l
consistent data basca and assunptions. The firrt is the Eastein United States Seismic Hazard Characterization Program supported by the USNRC at Lawrence Livermore National Laboratory. The second is the industy-sponsored Seismic Hazard Methodology program performed by the Electric Power Research Institute. In both these programs, hazard curves were developed for all U.S. commercial nucicar power plant sites east of the Rocky Mountains. Sten b - Initiatirm Events and Event 'Ireen
- The scope of NUREG 1150 includes all potential initiatinh eve nes ,
including loss of coolant accidents (vessel rupture, and large, nedium, and small IDCAs) and transient events. Two types of transients are being considered: those in which the power conversion system (PCS) is initially availabic (denoted Type T3 transients) and those in which the PCS is failed as a direct consequence of the initiating event (denoted Type T1 transients). The event trees derived for the internal event analyses are utilized. , The reactor vessel rupture and large LOCA event frequencies were based on a Monte Carlo analysis . of steam generatorr and reactor coolant pump support failures. The medium and small LOCA event frequencies are obtained from detailed piping failure calculations performed in the , SSMRP, The frequency of Type Tl transients is based on the probability of seismically induced loss of offsite power (LOSP). This is the dominant type of transient (for the majority of plants for which IASP counes loss - of main feedwater) . The frequency of the Type T3 initiating event is computed from the cendition that the sum of the initiating event probabilities must be unity. The hypothesis is that, given an earthquake of reasonabl.c size, at least one of the initiating events will occur. Etto e - Fault Trees Fault trees for the_ safety systems at Surry have been developed in the internal events analysis for random failures only. These fault trees are used, with modification to include basic events for seismic failure modes. .The trees are re solved fi pertinent seismic cut sets - to be included in the probabilistic calcuintions. Probabilistic culling -is used in re solving these trees in such a way as to assure that in;)ortant , correlated seismic failure modes are not lost, j Sten d Component and Structure Failure Descriptions Component seismic fragilities are obtained both from a generic fragility i data base and from plant specific fragilities developed for components ) identified during the plant walkdown. ' The generic data base of fragility functions for seismically induced failures was- originally developed as part of the SSMRP (Ref. 10). Fragility functions for -the generic categories were developed based on a combination of experimental data, design analysis reports, and an 18 i
a k l t l l l j extens,1ve expe rt opinion survey, The experimental data utilized in i developing iragility curves were obtained irom the results of component manuf ac t.ure r's. quali f icat ion test s. , independent testing lab fnilure data and data obtained f rom the U.S. Corps of Engineers extensive SAFEGUARD , Subsystem Hardness Assurance Program (Ref. 12). Ther.c data were s,tatistically combined with the expert opinion survey data to produce fragility curves for each of the generic component entegories as reported
- - in Reference 10. This generic data base was then updated by an j evaluation of 19 site specific se l s.mic PRAs to yelid the final generic !
i fragility data base used for the Surry and Peach Bottom NUREG 1150 PRAs. I { Detailed structural f ragility analyses were perforrted for all important , safety related structures at the Surry plant. In addition, an analysis ! of liquefaction f or the underlying soils was performed. These were [ included directly in the risk assessment. l Steo e Seir,mic Refs pse of Structures and Components
- Building and component seismic responses are estimated from peak ground
! accelerations at several probabilit.y intervals on the hazard c urve . I Three basic aspects of seismic response best estimates variability, and correlation are generated. Zion analysis results from SSMRP and ,
- siinpli fied methods studies form the basis for assigning scaling, j variability and correlation of responses.
I { In each case, SHAKE code (Ref. 13) calculations are perforrned to ansess the effect. of the local soil column (if any) on the surface peak ground acceleration and soil structure interactions. This permits an evaluation of t.he effects of non homogenous underlying soll conditions which can stron61 y affect the building responses. ) Fixed base mass spring (eigen system) models, are either obtained from the j plants architect /en6 1 ncer or are developed from the plant drawings as ! needed. Using these models one can compute the floor slab accelerations using the C1.ASSI code (Ref. 14). This code takes a fixed base _elgensystem model of the structure and input specified frequency dependent soil impedances and computes the structural response (as well as variation in structural response if desired). Variability in responses (floor and spectral accelerations) is assigned
' based on the SSMRP results. The recommended uncertainties (expressed as standard deviations of the logarithms of the responses) are shown below:
Quantitv- Random Peak Ground Acceleration 0,25 Floor Zero Period Acceleration 0.35 Floor Spectral Acceteration 4 AS -- Correlation between component failures is being included explicitly. In computing the correlation between component failures (in order to
' quant if y - the cut sets) -it in neccusary to consider correlations both in the responses and . In the fragilities of each component. Innsmuch as 19
there are no data as yet on orralation between fragilities, the fragility correlations between like cortponents are taken as rero, and the possibic effect of such correlation can be quantified in a sensitivity study. The correlation between responses is assigned according to a set of rules that are explained in Chapter 4.0. Step f - Probabilistic Pallure and Core Dartare Calculationg Given the input from the five steps above, the SETS (Ref. 15) code and mean basic event frequencies are used to calculate the required output (mean probabilitics of failure, core damage, etc.). Sten g Estimate Uncertainties Complete uncertainty distributions were computed for all accident sequences and core damage f requencies using a Monte Carlo approach. 1.3.4 Internal Fire Assessment Methodology Based on nuclear poact plant operating experience over the last 20 years, it has been observed that :ypical nuclear power plants will have three to four significant fires over tncir operating lifetime. Previous probabilistic risk assessments (PRAs) have shown that fires are a significant contributor to the overall core damage frequency, cont ributing . anywhere from 7 percent to 50 percent of the total (considering contributions f rom internal, seismic, flood, fire, and other events). Because of the relatively high core damage contribution, fires are always examined in detail. . An overview of the simplified fire PRA methodology is as follows: j A. Initial Plant Visit Based on the internal event and seismic analyses, the general location of cables and components of the systems of interest is known. The plant visit _ provide s the analyst with a means of socing the physical
- ' arrangements in - each of these areas. The analyst will have a fire zone checklist which will aid the screening analysis and in the quantification step.
The second purpose of the initial plant visit is to confirm w'ith plant personnel that the documentation being used is, in fact, the best availabic information and to get clarification about any questions that might have arisen in a review of_ the documentation. Also, a thorough review of firefighting procedures is conducted. B. Screening It is necessary to select important fire locatit.ns within the power plant under investigation having - the greatest potential f>r producing - risk-dominant accident sequences. The objectives of lacation__ seicetion are
.1 10 l
4 J 4 1 I l j soniowhat competing and should be balanced in a meaningful risk assessment j study. The first objective is to maximize the possibility that all important locations are analyzed, and this leads to the consideration of 3 a potentially large number of candidate locations. The second objective is to minimiec the effort spent in the quantification of event trees and fault trees for fire locations that turn out to be unimportant. A proper balance of these objectives is one that results in an ideal allocation of 4 recources and efficiency of assessment. 1 The screening analysis is comprised of:
- 1. Identification of relevant fire zones.
- 2. Screening fire zones on probability of fire induced initiating i
events, 2 l 3. Screening of fire zones on both order and frequency of cut sets. 4
- 4. Numerient evaluation and culling based on probability for each remaining fire zone.
C. Quantification After the screening analysis has climinated all but the probabilistically significent fire zones, quantification of dominant cut sets is completed as follows:
- 1. Determine temperature response in each fire zone.
i
- 2. Compute component fire fragilities.
d
- 3. Assess the probability of barrier failure for all remaining combinations of fire zones.
- 4. Perform a recovery analysis.
Finally, an uncertainty analysis is performed to estimate error bounds on the computed fire induced core damage frequencies. The Surry fire analysis is presented in Chapter 5, d l j v l l 1 l I 1 11
., . . . - - _ - _ . _ . - . . - . - . _ . - . . - - - _ _ ~ . _ _ _ . ~ _ _ _ _ . - _ _ . _
l v 1.4 References
- 1. U.S. Nuclear Regulatory Commission, Severe Accident Risks! An Assessment for Five U. S. Nuclear Power Plants, NUREG 1150, Second Draf t for Peer Review, Vol.1, June 1989, 1
- 2. U.S. Nuclear Regulatory Commission. Reactor Safety Study: An Assessment of Accident Risks in U.S. Nucicar Power Plants, WASil 1400, NUREG-75/014, 1975. l i 3. M. P. Bohn, and J. A. Lambright, Procedures for the External Event Core Damage Frogntney Analyses for NUREG 1150, Sandia National .
- Laboratories, Albuquerque ,- NM, NUREG/CR-4840, SAND 88 3102, November 1990. )
- 4. D. M. Ericson, Jr. et al, Shutdown Decav llent Removal Analysis Plant Case Studies and_Rpecial Issues, NUREG/CR-5230, SAND 88 2375, April 1989.
- 5. J. A. Lambright and M. P. Bohn, Analysis of Core Damage Freauency Due to External Events at the DOE N Reactor, Sandia National-Laboratories,__ SAND 89-1147, August 1990.
- 6. J. A. Lambright,. Analysis of Coro Dam,ay Frecuency Due to Fire at the levannah River R-Reactor, Sandia National Laboratories, SAND 89 1786, August 1990.-
- 7. U.S. . Nuclear Regulatory Commission, PRA Procedures Cuide, NUREG/CR.
2300, January 1983.
'8. R. P. Rennedy, Blejwas T. E., and D..E. Dennett, Canacity of Nuclear Power Plant Struc tures to Resist Blast Loadings, Sandia National Laboratories, NUREG/CR 2462, SAND 83 1250, September 1983.
- 9. ' M. K. Ravindra and 11. Eanon, Methods for External Event Screening Ouantificatiou, Sandia National Laboratories, NUREG/CR 4839, S AND8 7'.
7156, August 1990.
- 10. M. P. Bohn,- et al., Apolication of the SSMRP Methodolory to thq
'St ismic Risk at the Zion Nuclear Power Plant, Lawrence Livermore National Laboratory, Livermore, CA, NUREG/CR 3428, 1983, i
' 11. - D . L. Bernreuter, et al., Seismic Hazard Characterization of thn-Ecstern United States: Methodology and Interim Results for Ten Si tes ,. Lawrence Livermore National Laboratory, NUREC/CR.3756, April l 1984.
!? 12. U.S. Army Corps of Engineers, Subsystem Hardness Assurance Report, l llNDDSP 72-156 ED.R, Vols I and-II, 1975. ; i
- 13. P. B. Schnabel, J . Lysmer, and ll. B. Seed, SHAKE A Computer Program for . ithogake Response Analysis of Horizonrally Lavered Sites,.
Earthquake Engineering Research Center, University of California, _ Berkeley, CA. EERC 72-12, 1972. l 1 12
1 4 . 11 . L. Wong ar.d J . E. Luco, Soil - S t ruc_ture Internetlon: A_l,i ntar Continuum Mechanics Approach (CI ASSI) , Dept. of Civil Engineering, University of Southern California, Los Angeles, CA, CE79 03, 1980.
- 15. R. li , Vorrell, SETS Reference finnual, Sandia National laboratories, Albuquerque, NM, SAND 83 2675, f;UREC/CR-4213, May 1985.
1 1 13-
- - . - . .-. - -. - . - - - - - - - - - - - - - .- . ~ - - _ -
) 2.0 PIANT DESCRITTION 2.1 j'l a n t . Site and Cencral Characteristics The twin PVR units (Surry 1 and Surry 2) of Virginia Electric and Power Company are each rated at 781 MV. The reactor and generator for both the units were supplied by Westinghouse Electric Corporation. The plant began commercial operation in 1972 1973. Stone and Vebster Engineering Corporation was the Architect / Engineer / Constructor for these plants. A type 3D containment design was used. Other Class I structures include the auxiliary building; control room area, including switchgear and relay rooms; fuel building; auxiliary generator cubicles; auxiliary containment buildings that contain main steam and feedwater isolation valves; recirculation spray and low head safety inj e c tion pump cubicles; safeguards ventilat. ion room and circulating water intake structures, including the high-level canal. All these structures have been designed to meet both earthquake and tornado criteria. 2,2 Description of Plant Systems 2.2.1 Introduction This section discusses the system descriptions and system models of the maj or frontline and support systems identified as important to safety. In addition to the event trees discussed in Section 2.3, component fault trees also developed by the internal events analysts were utilized. Use of the same event trees, fault trees, and accident sequences developed during the internal events analysis ensured consistency between these major studies. The discussion of the systems that follow includes:
- a. A brief functional description of the system with reference to the one line diagrams that were developed to indicate which components were included in the model;
- b. Safety related success criteria that were applied to the system; c, Interfaces and safety actuation provisions between the frontline Jystems and the support systems.
2.2.2 Containment Spray System The containment spray system (CSS) provides the initial containment pressure reduction following an accident by spraying cool water from the reactor water storage tank -(RVST) to condense steam in the containment. The curry CSS is composed of cwo 100 percent capacity _ spray injection trains. The CSS has no recirculation or pump cooling capability. Each spray train draws water from the RVST through independent suction lines. 2-1
Each CSS pump takes suction through a normally open MOV and an in line filter assembly. Each CSS pump dischargta through a pair of normally closed MOVs arranged in parallel and through a check valve to its associated containment spray header. Both CSS pumps also feed a common third spray header (located on the outside of the crane wall) through separate check valves. A sittplified schematic of the CSS is shown in Tigure 2.1. The CSS automatically starts on receipt of a }{1111 (25 psia) containment pressure signal from the consequence limiting control system (CLCS). The CLCS signale open the pump inlet and outlet valves and start the CSS pumps. An agastat timer in the pump start circuit delays pump start for 30 seconds af ter receipt of the signal. The success criterion for the CSS ir one of the two CSS trains that provides flow to any one containment spray header. 2.2.3 liigh Pressure Injection / Recirculation System The Surry charging system provides normal coolant makeup to the reactor coolant system (RCS) and cooling flow to the reactor coolant pump (RCP) seals under normal operating conditions. The high pressure injection / recirculation (HPI/llPR) system uses the same charging pumps to provide primary coolant injection and recirculation following an accident, as well as maintaining flow to the RCP seals. The itPI system also functions to deliver boric acid to the RCS from the boric acid transfer system if emergency bor . ion is required. Under norr.. . operating conditions, one of the three charging pumps p;o-vides normal RCS makeup and cooling to the RCP seals by taking suction from the volume control tank (VCT) through two motor operated valves (MOVs) in series. Upon indication of a loss of RCS coolant or steam line break (i.e., low pressurizer level, high containment pressure, high pressure differential between main steam header and any steam line, or high steam flow with low t average temperature (Tavo) or low steam line pressure), the safety injection actuation system (SIAS) initiates emergency coolant irgection. l The SIAS signals the normal charging line isolation valves to close, the standby charging pumps to start, the valves from the VCT to close, the normally open pump inlet and outlet MOVs to open, and a parallel set of normally closed MOVs to open to provide suction from the RVST. Also on receipt of an SIAS signal, a parallel set of normally closed F open to ! provide flow from the pump discharge header to the three rcd cold legs. An additional path to the RCS cold legs through a manually operated normally closec' MOV is also available. Flow through this line to the RCS l is treated as a recovery action. The line to the RCP seals remains open throughout the event. The HPI system may also be used in the " feed and biced" cooling mode. The only difference in this mode of operation from that discussed above is that a SIAS signal is not necessarily generated so the llPI system is manually placed in service. In the recirculation mode of operation, the charging pumps draw suction from the discharge of the low pressure safety injection pumps in the low 22 i
#AOVCS1010 g PS53 rm
/ r, 9;
Ne L 4 D InVST MOVCS101A CV13 CV 105 i MOVCS1010 D Q PSSS Ym XV8 CV7 n.
'r$ W XV15 CV14 r ,2 N.
7 1 A D MOVCS101C pSr7 YY ~ _ ~_ (1 ) C P B) TO gR PS50 MOVCS100A 1-CS-FL-1 A ( J MOVCS1000 ( PSS1 1-CS.FL-1B Figure 2.1. Containment SPray System Schematic i
pressure recirculation (LPR) system. Upon receipt of a low RWST 1cvel signal, the recirculation mode transfer (RMT) system signals the charging pump suction valves from the RVST to close and the suction valves from the LPR pump discharges to open. In the emergency boration mode, the HPI functions as described in the llPI description above with the exception that the boric acid transfer (BAT) pumps deliver boric acid from the BAT tanks to the charging pump suction header. To perform this operatica, the operator must switch the normally operating BAT pump to fast spe N operation and open the MOV allowing flow into the charging pump suction header. To enhance boric acid addition to the RCS, the emergency procedure calls for the PORVs be opened (to provide pressure reduction). A simplified schematic of the HPI/HPR system, including the relevant portionc of the BAT system is presented in Figure 2.2. The success criteria for the HPI modes of operation require flow from any one of three charging pumps to the RCS cold legs in response to a LOCA (automat.ic actuation), flow from any one of three charging pumps to the RCS cold legs in the " feed and bleed" mode (manual actuation), flow from any one of the three charging pumps to the RCP seals, or flow from any one of three charging pumps to the RCS with flow from one of two BAT pumps operating at fast speed (emergency boration mode). The success criterion for the HPR mode of operation is continued flov from any one of the three chargirg pumps taking suction from the dis-charge of the low pressure recirculation system, given successful low pressure system operation. 2.2.4 Accumulator System The accumulators provide an initial influx of borated water to reflood the reactor core following a large LOCA or a medium LOCA on the upper end j of the IDCA size definition. The accumulator system consists of three tanks filled with borated water and are pressurized with nitrogen. Each of the accumulators is connected so one of the RCS cold legs by a line containing a normally open MOV and two check valves in series. The check valves serve as isolation valves during normal reactor operation and open to empty the contents of the accumulator when the RCS pressu- falls below 650 psig. A simplified schematic of the accumulators is snown in Ggure 2.3. The success criterion for the accumulators following a large LOCA; which assumed a cold leg break, is injection of the contents of the two accumu-lators associated with the intact cold legs into the RCS. The success criterion for the accumulators following a medium 14CA is injection of the contents of two or more accumulators into the RCS. 2-4
. .. _ , , . --u. .._- -.x _1.a.a -- .n.-w-a -a .u .- -- +a_ . _ = . -r_ _ _ ~ ------.-,-a =_ aa..n .. . .a- - - .a 1
1 s
*m
-r.,-w * .-
i (2 *$2I I *,*o,: t
". t!-
o 3,* '
- E 4 b I.
b E
* ^
r N k a g sa
, 0 R -
* :- g
'E "gg Em! l U
!! 8 "** f -
I 3 = 2 = 5 r."
.E.8o e
a 4,'t 1 e e
- 2- 5 . ; 3 .c t,I ,
fAf e AwI 1n > ' o 6 Y a Q U N u t E g m 2 2 :
- h
> g . >
E a 2
' I E m f a 8 2 T.
v 2 : U h{ $
- g A- . 'A
=
! N 5
. E '
r I.
- f I i -
S
> x
> N ~ u
? --4 n) & f l .a t
- C
%m G k - 5 6 i i E l e EI 0
(! n 5 E E r .: e 3 : g 5 a j u : A ,
- --A Ia e ee
- i e I.
=
r g
- a. [ -
.l^ I
- 0 A s
i a m d 3 _R ,:U.
. Q
' e:"r i 5 fo i
- i 5= I C
/ &. . --+:
i
- u-
; -e: ;< gI
" ; ~ : 12
' o E a2 6 tl lC 2 2Cl g! -
gr -
. ja
;a
~
5 E!!o 2-5
BAT BAT TANK TANK 1A 1B NI N NI X X MDPCH2A D [3 MDPCH2B (1. CH . P 2A) \ (1.CH P 28) PS 132 Z Z
>A ><
X X (l!!!!!ililiillt) TO BORIC ]f AC:D BLENDER
^ "
X L2 FROM VCT rm 3y nC TO CHARGING O MOV 1350 PUMP SUCTIONS k FAI 4 Figure 2',2, High Pressure Injection / Recirculation System Schematic (2 of 2) 1 1 26 l
t i I r t I FDOPJ 1- S1 - TK - 1 A i WNST m ., l l 92 77 i LEG 1865 A CV107 CV109 ; i i 1-S1 TK-1B i
" i h SA I EG 1865 B CY120 CV130 A
1-St-TK-1C l i A Vm
\
CV145 CV147 EG 1865 C [ I Figure 2.3. Accumulator System Schenutic l i i i t 4
- _ - . - - . - ~ - - - - - _ - - - . - - . - - - - - _ ~ . - . - - - . -
k i 2.2.5 Low Pressure Injection / Recirculation System j The Surry low pressure injection / recirculation (LP1/LPR) system provides
- emergency coolant injection and recirculation following a loss of coolant accident when the RCS depressurizes below 300 psig. In addition to the direct recirculation of coolant during the recirculation phase once the RCS is depressurized, the LPR discharge provides the suction soir
- < tor the HPR system following drainage of the RVST.
The Surry LPI/LPR system is composed of two 100 percent capacity purnp trains. The LPI/LPR has no heat removal capability. In the inj ec tion tr; ode , the purep trains share a common suetion header from the RVST. Each pump draws suction from the header through a normally open MOV, check valve, and locked open manual valve in series. Each pump discharges through a check valve and normally open MOV in series to a common in-jection header. The injection header contains a locked open MOV and branches to three separate lines, one tv each cold leg. Each of the lines to the cold legs contain two check valves in series to provide isolation from the high pressure RCS. . In-the recirculation modo, the- purep trains draw suction from the contain-roent sump through a parallel arrangement of suction lines to a common header. Flow from the suction header is drawn through a normally closed MOV and check valve in series. Discharge of the pups is directed to either the cold legs through the same lince used for injection or to a parallel set of headers which feed the charging pumps, depending on the RCS pressure. In the hot leg injection mode, system operation is identical to normal recirculation with the exception that the normally open cold leg injection valve must be manually closed from a remote location and one or more normally closed hot Icg recirculation valves must also be manually opened from a remote location. Upon indication of a loss of RCS coolant or a main steam line break (i.e., low pressurizer level, high containment pressure, high pressure differential between main steam header and any steam line or high steam flow with low TAVO or low steam line pressure), the safety inj ec tion actuation system (SIAS) initiates LPI operation. The SIAS signals the low pressure pumps to start All valves are normally aligned to their injection position. If primary system pressure remains above the LPI purep shutof f head, the pumps will discharge to the RWST through two normally open rainimum flow recirculation lines until the RCS pressure is sufficiently reduced to allow inilow. Upon receipt of a low RWST level signal, the recirculation mode transfer system (RMTS) signals the low pressure pump suction valves from the RVST and the valves in the minimum flow recirculation lines to the RWST to close and the suction valves from the containment sump to open. A simplified schematic of the LPI/LPR system is shown in Figure 2.4. e 2-8
i
- ~ a I2 II $$
2 2
. e.
EI E 5 .i 5 o d he " : e i 3 I ..e /. :c
- g. .[.c i
I i i g_!_jlE4!iz!ifil $
$g a 4 A
'E I "
[E ~ t 26U c zu te
*a* 20g g 6 5 g .$
W U o Ir 'E 1: j h,3{d nt e g
.s W ff!fl !n S$( ! 8 4
ze z. g* -'g* - t u ej es
- E- 2: RL n r 8
- e. -
u t e Y .Y
- A
!E :E EE $! fg fh.! ;
f[ir i vai 8l N Z$ [#' e.. l ..
.h ~
45 5 0 b>6E *g o r
- ) :
4 W < l(i- l(i . W ii Oy ig .v. , c Ig *v ,, ce, A @ c v. H n H W t o o
!(6,2 :
W*4,2 : E'4.: 6:z E*q 3 E -<< = s g
.E g d
29
. - . . . -. . . . . , - . _ . - . . . . _ . , . _ . , _ - _ .. _ . _ - . . _ . . . . . . . . , . _ - _ , . ~ . _ _ - - , , . _ . - _ - ~ - . _ . . . . . . . . . _ _ - . _ . . . . - - - . - ,
- - . - - - - . - - -_ -- _ ~ - - - . - - - . . - - - - - . ~ ...._- -.. ~ _ . . .
l d l l The success criterion for the LPI mode of operation is flow from one or ! more low pressure pumps to the RCS cold legs in response to a loss of l primary coolant inve nto ry . The success criteria for the LPR modes of i operation are continued flow from either of the two low pressure pumps to the cold legs and switchover to hot leg recirculation at 16 hours or sufficient flow from either of the two low pressure pumps to the charging pump suction header. , 2.2.6 Inside Spray Recirculation System The inside spray recirculation (ISR) system provides long term contain-l ment pressure reduction and containment heat removal following an acci-I dent by drawing water from the containment sump and spraying the water into the containment atmosphere. The Surry ISR system is composed of two independent 100 percent capacity recirculation spray trains. Each spray train draws water from the con. tainment sump through independent suction strainers and lines. The ISR and outside spray recirculation system (OSR) draw from the same sump, although the sump is compartmentalized and each ISR train has a separate i sump compartment. Each ISR system pump discharges to a service water heat exchanger. The cooled water is then directed to an independent spray header, In order to ensure adequate net positive suction head (NPSil) for the ISR pumps during the initial phases of a LOCA, a recirculation line diverts a small amount of the cooled ISR flow back to the sump, close to the pump inlet. A simplified schematic of the ISR system is shown in Figure 2.5, The ISR system automatically starts on receipt of a lli ili (25 psia) con-tainment pressure signal from the consequence limiting control system (CLCS). The CLCS signals start the ISR pumps. An agastat timer in the pump start circuit delays pump start for two minutes to ensure adequate sump inventory and the correct diesel generator loading sequence in the event of loss of offsite power. The success criterion for the Surry ISR system is that at 1 cast one of the two ISR trains provides flow to its containment spray header with service water bdng supplied to the heat e x change r .- 2.2.7 Outside Spray Recirculation System The outside spray recirculation (OSR) system provides long term contain-ment pressure reduction and containment heat removal following an acci-dent by drawing water from the containment sump and spraying the water into the containment atmosphere, , l The Surry OSR system is composed of two independent, 100 percent capacity recirculation spray - trains. The spray trains draw water from the con-tainment sump through two parallel suction strainers and lines which are headered together. The OSR and ISR draw from the same sump, although the j sump is compartmentalized. Each OSR train has its own separate com-partment. Each OSR system pump has an individual suction line from the l l 2-10 l
l )
- r
( ) ( ) PS60 . ( ) - PS61 HXRS1A SWS HXRS1B SWS r .. r JL JL MDPRS1A MOPRS1B Tigure 2.5. Inside Spray Recirculation System Schematic 2-11
header with a normally open MOV. Each pump discharges through a normally open MOV, check valve end a service water heat exchanger. The cooled water is then directed to an independent spray header. In order to ensure adequate NPSil for the OSR system pumps during the early phase of a LOCA, a line is provided which diverts a small amount of the cool CSS flow to the sump, close to the puep suction strainers. A simplified schematic of the OSR system is shown in Figure 2.6. The OSR system automatically starts on receipt of a Hi Hi (25 psia) con-tainment pressure signal from the consequeme limiting control system (CLCS). The CLCS signals start the OSR system pumps and ensure that the pump inlet and discharge valves are open. An agastat timer in the pump start circuit delays pump start for five minutes to ensure adequate sump inventory and the correct diesel generator loading sequence in the event of loss of offsite power. The success criterion for the OSR system is that at least one of the two OSR system trains provides flow to its containment spray header, with service water provided to the heat exchanger. 2.2.8 Auxiliary Feedwater Syst e The auxiliary feedwater (AW) system provides feedwater to the steam generators to provide heat removal from the p-imary system after reactor trip. The Surry AW is a three train system, two electric motor driven pumps and one steam turbine driven pump. Each pump draws suction through an independent line f rom the 110,000 gallon condensate storage tank (CST). In addition, a 300,000 gallon CST, a 100,000 gallon emergency makeup tank and the fire main can be used as water supplies for the AW pumps. Each AW pump discharges to two parallel headert. Each of these headers can
- provide auxiliary feedwater flow to any or all of the three steam gener-ators (SGs). Flow from each header to any one SG is through a normally open MOV and a 'ocked open valve in series, paralleled with a line from the other header. These lines feed one line containing a check valve which joins the main feedwater line to a steam generator. A simplified schematic of the AW is shown in Figure 2.7.
The motor driven AW pumps automatically start on receipt of an SIAS signal, loss of main feedwater, low steam generator level in any steam generator, or loss of offsite power. The turbine driven A W pump auto-matica11y starts on receipt of indication of low steam genera. tor 'evel in two of the three steam generators or undervoltage of any of the three main RCS pumps. These signals also ensure that the system MOVs are in the correct position. The success criterion for the AW following all events is flow from any one AFV pump to any of the three steam generators. 2-12
t f a-( -) , ( ) j- - ( k I l- 4- 4- j 4 4 HXRSID SWS HXRS1C SWS .{ l ! MOVRS1568 cv11 I U i N > ~~~> ' 4 -j MOVRS156A c CV17 7 C PS71 E N i-i PS72 rnoucss l i j _L _ L i i MDPRS2B MDPRS2A -- -- i k k ,
,7 7
- MO 155A mw l m ,
4 1 i f i MOVRS1550 , 1 E ! i r i ! I Figure 2.6. Outside Spray Recirculation System Schematic l I i 4 i b ,- 1
. m. . . _ _ _ _ _ _ _ _ _ -
.. .-. . . . - = - - . . ~
. . . - - - - - . - - - . . . . . . ~ . . -- . - - - . - . ~ -
N g3 S EO = E 85 - : :., a< g
- e ;e = g og y '
[P '4- 8E gE 3 2 -V . V Wh
!E T j [; jY E !
- - " . n*:
l! 4 4 e o@ c@
. :e o ::
sess Og sq[y[
.N M > $
,, m. , , , M .
ggg, 3
$$ 10Sd 'e N ,!
M
- o' ' '.T',.
N ,OttbMd O o
$2u "
40
** $ e
"" E t t ,09 MA d g
- 068d o u 7
. ,Q0 \ I IM s.. fu g h
.e, O
E3 h
- T h
E 0 2 [*[.8 t as a e b
! c~ JE.iEE o
a - e
=a 3 -3 +
- ;o;* g -
;": :>c- ,5> 5 8! !!!h 5 53 8 s a
} }8E e E ,
v-= d
'h w w
f Rn g$u y e a v n y W L.5.- W
. n &g f g ;:
n
<-* ~ . --
Sp C -! h" i% _
.-_ , i _
gJ D r5Cg - _a_*tn l aos <- 9 R neuan hahnh I 2 14
d 2.2.9 Primary Pressure Relief System 2 The prirr.ary pressure relief systett (PPRS) provides protection from over-pressurization of the primary systern to ensure that primary integrity is maintained. The PPRS also provides tht- means to reduce the RCS pressure if necessary. The Surry PPRS is coreposed of three code safety relief valves (SRV) and two power operated relief valves (PORVs). The code safety valves were important only for the AT'.'S analysis. The PORVs provide RCS pre s t.ure rolief at a set point below the SRVs. The PORVs discharge to the pressurizer relief tank. Each PORV is provided with a motor operated i-block valve. A simplified schematic of the PPRS is shown in Figure 2.8. ' The PORVs automatically open on high RCS pressure or are reanually opened at the dheretion of the operator. The block valves are normally open unless a PORV is Icaking, The success criterion for the PPRS following a transient event demanding PORV opening is that the PORVs successfully reclose, The success cri-terion for the PPRS following a transient and failure of the AWS is that both PORVs successfully open on demand, The success criterion for the PPRS following a small LOCA with failure of the AWS and for the support system function provided to HPI in the ettergency boration mode is that one e more PORVs succe efully open on demand, 2.2.10 Power Conversion System The power conversion system (PCS) can be used to provide feedwater to the steam generators following a transient. The PCS, as modeled in this study, consists of the main feedwater pumps, the condensate pumps, -the condensate booster pumps , - and the hotwell inventory. Because Sarry has electrically driven MFW pumps , it is possible to supply feedwater using the MW system, without having the turbine bypass and steam condensing systems available. The inventory of the hotwell (with the CST as a backup supply) was calculated to be sufficient for all mission times of interest, The feedwater regulating valves will close after a reactor scram, due to plant control logic, The feedwater pumps remain on, and the miniflow valves will open. Feedwater can then be provided- to the SCs, through the feedwater regulating valve bypass valve. The success criterion for the PCS are restoration of flow from one or more main feedwater pumps to one or more steam generators. 2,2,11 Charging Pump Cooling System The charging pump cooling (CPC) system is a support system which provides lube oil cooling - and seal cooling to the three charging pureps in the HPI/HPR system. r 2-15
TO PRESSURIZER REUEFTAtM 1-RC-TK-2 6- 6- 6* A SV 1551 A SV m SV 1551C 15518 r 6* F Y LV3V FC PCV 5 4* 1535 1456 MOV FC 1536 PCV PRESSURIZER 1455C Figure 2.8. Primary Pressure Relief System Schematic
The Surry CPC system provides two spreific cooling functions for the char &i ng pumps, lube oil cooling and seal cooling. The CPC system is composed of two subsystems, the charging pump service water system and the charging pump cooling water system. The charging pump service water syi. tom is an open cooling system which provides cooling to the lube oil coolers and to the intermediate seal coolers in the charging pump cooling water system. The charging pump cooling water system is a closed cycle system which provides cooling to the charging pump seal coolers. The charging pump service water system is composed of two 100 percent capacity pump trains, each providing flow to one intermediate seal cooler and all three char 61 ng purep lube oil coolers. Flow is drawn from the condenser inlet lines through independent lines by the charging pump service water pumps. Upstream of each pump are two separate, independent strainer assemblies. Each pump discharges through two check valves. Downstream of the check valves the flow is split with a portion of the flow directed to an intermediate seal cooler and the other portion directed to a common header feeding the lube oil coolers. From this header, flow is directed through the lube cil solers for the operating charging purcps. Temperature control valves control the flow through the lube oil coolers to prevent overcooling of the lube oil. The service water flow is discharged to the discharge canal. 2 The charging pump cooling water system is a closed cycle system composed of two 100 percent capacity pump trains, each containing a charging pump cooling water pump and intermediate seal cooler which provide cooling water to the charging pump seal coolers. Each pump dcaws suction from the outlet of either of the two interacdiate seal coolers and discharge to a common header. The common header provides flow to the seal coolers for each charging pump. Two seal coolers in parallel are provided for each charging pump. The discharge of the seal coolers is returned to the intermediate seal coolers where it is cooled by the charging pump service water system. Makeup to the charging pump cooling water system to ac-count for seal leakage is provided by a surge tank which is supplied by the component cooling water system. A simplified schematic of the CPC system is shown in Figure 2,9. One of the charging pump service water pumps and one of the charging pump cooling water pumps are normally in operation. Upon indication of low discharge pressure of one of the pumps, the parallel pump receives a signal to start. With the exception of the pumps and the lube oil cooler temperature control valves, all other components in the system are manu-ally actuated. 2.2.12 Service Water System The service water system (SWS), as defined for this analysis, is a sup-port system which provides cooling to the heat exchangers in the ISR system and OSR system. The SWS provides heat removal from the contain-ment following an accident. 2-17
1TNYS 90WTM0$10 01 4-l E
* - 3 8 "I,a 8, 2 s.( CX 2l,,
M M M iP U 11
- " g :"g ^y
<; q a ,
o o y* E y,8
- 5,( e*E y O a
- a e
a g a 1 o
"(O
!l(
EX
!l(
- v "i R. $.
o, fa w i Eg g lw e l s. i i. E L !* -
- s t
t W: *t W:*t .: /.: U .e r- - .-- e.a - c ty'g
- o'
. , _ = w
$ c' -
- *5
(.- O o
;; A E u
I ;; ^; 2, ( hc .gagh.
;, y'.:
- e o a * . p e . .o. =
tAgE 2"E" te
. U g [ ,,.
g, e
't.' z t -
- g c gw g g. .
n < _lE . 3-: -: : . !
"U [ 3 l( E f :l: ,c.
3 6 li3 E jx
. l. .
~
E;' ; I 5h 5 d ! c c s. .tg .=1Ji.J.,n -
~
g IslL8 Igo
- et Es 0
seu u a "6 C=E @@3*. 8 j,( $1( g4
. g, I:h ! 3 !3 lf! ! I E i
, g, --
m
.,,p O i
$4 g g 5g t t w, C ,2 a a e e2 - -
i< I 5. 5 5 a v =I 55. < =! t' : E E1 g i "I-l I
- ko
;I-t . E: :
c C c :r e} - t g*s g=s . :t_: A - - x l"-
.g - lE
.i o o 2 18
__ -_-__m._.m. _ . . . _ _ _ _ _ _ _ _ _ - _ a_. .m___ _ . _ . - _ . _ _
The Surry SWS is a gravity flow system. The service water supply to the containment spray heat exchangers consists of two parallel inlet !!nes which provide SW from the condensor cooling pipes each through two nota-ally closed MOVs in parallel to individual headers. The headers each provide flow to one ISR and OSR heat exchanger. The two headers are cross connected by two normally open MOVs in series erh that flow from , either inlet lino can be used to cool all four ISR and OSR heat ex-changers. Service water flows through each heat exchanger and discharges through a normally open MOV to two headers which flow to the discharge tunncl. A simplified schematic of the SWS is shown in Cgure 2.10. The SWS automatically starts on receipt of a Hi Hi (25 psia) containment pressure signal from the consequence limiting control system (CLCS). The CLCS signals open the header inlet valves. No other actions are required to place the SWS in service. 2.2.13 Component Cooling Water System The component cooling water (CCW) system, as defined for this analysis, includes only-that portion of the CCW system required to provide cooling water to the RCP thermal barriers. The CCW system is composed of two CCW pumps in parallel and two CCW heat exchangers. The CCW system is a closed cycle system. Be CCW pumps take suction from the return line from the RCS pump thermal barriers and are headered together at their discharges. The header feeds the two CCW heat exchangers arrangec' in parallel, The discharge of the heat exchangers is delivered to the thermal barriers. After cooling of the thermal barri-ers, the flow is returned to the CCW pump suction. Makeup to the CCW system is provided from a surge ':ank in the system. A simplified sche-matic of _ the portions of the CCW system required for thermal barriet cooling is shown in Figure 2.11, one CCW pump and heat exchanger are normally in operation. In thc event of failure of either component, the parallel component is manually placed in service. Following a loss of offsite power, the stub buses powering the CCW pumps are shed from the emergency 'sures and must be manually reconnected to restore power to the CCW pumps. The throttle valve on the thermal barrier cooling water outlet closes on loss of instrument air or receipt of a consequence limiting contd system (CLCS) Hi-Hi signal, 4 resulting in loss of flow to the thermal barriers. The success criterion for the Surry CCW system is that continued CCW flow is provided to the RCS pamp thermal barriers following reactor shutdown. 1 2.2.14 Emergency Power System The emergency power system (EPS) provides AC and DC power to safety-related components following reactor scram. The EPS consists of two 4160 V AC buses, four 480 buses, four 120 V AC vital instrumentation buses, two 125 V DC buses, one dedicated and one l 2-19 l
i 1 l
< c b '$ aW I E $$
$ c#
m o JL o JL g 2 r is m n - Md 37e 37e V D- 3 D- D w JLs o N JL9 o
- 2 2 E 3 m C N $ m a
E so g LA m A C 2 > o n o g e v g 2 b ' j
~
Y
~
,, g g m
3 R- r-R - E O r D,J D,JL>oL>a m E E 8 w x m m 2 o J I > - w 9 N $ $ $ g a A - a 3 h Y E h S u m Ak Ak m lf O M O ) a, ,b o 8s=N . u
. t.. m 8 {cy I
8 g E A e 2 g C O Y $ $ 44 *
, Y? m o >8 G 2 SE E m R
C 5 E 8 3 A e 2 OS m 29m O V 2-20 l _ _ _ _ _ _ _l
{ ill l Ii, , 4
,son
_o
.a T
e m s n 7
.eb e q
A . R . E C c g..
,, M
, or
, e
,, m el E
T , l A . c
]A"
[~ r l' Am 4, N P a i t a m
.f f e.
,- f n
u P R
- h e
- c H
M S fe . w m. D m Esa ~ w e Oc SA t e srTp s y O S C r e T- t N EE Osa S s. ' W TA UT g OW n r 4 i C l o o s, e C Ts es oLJ . t r,to . n c N e Nh
?p .. .
n o p - A. l
. m S
o i t t W
- c
- c. a, S C p
7 jF, , 3 , 1 r N j gT . s, 1 2 e r e u si g
- ts as i "a F so d ta c t i
N.t fN s 7 c x m, N J, . g f 5
^ m,n.
~\
P v 1 f llllllll,lI (Il
shared diesel generator, and their associated moter control centers, breakers, transformers, chargers, inverters, and batteries. Each 4160 V AC . bus is normally powered from offsite power sources. On loss of offsite power the breakers open, the diesel generators start and their associated breakers close to load the diesels on the emergency buses. Surry has three diesel generators, one dedicated to each unit and a third swing diesel generator shared by the units. The dedicated diesel at Unit 1 is attached to the 111 4160 V AC bus while the swing diesel can be connected to the IJ 4160 V AC bus. In the event that the swing diesel is demanded by both units, the diesel will be aligned to the unit at which a safety injection actuation system SIAS or CLCS Hi Hi exists. If signals exist at both units, the diesel will be aligned to the unit whose breaker closes first. Each diesel is a self contained, self cooled unit with its own battery for starting power. The 4160 V AC buses provide power to the large pumps such as the high pressure inj ectico pumps, the stub buses which each power one CCW and residual heat removal pump and is shed on undervoltage on the main bus , and the 480 V AC buses through transformers. The following description applies to the 111 related buses. Since tre 1H and lJ related buses are symmetrical, the deceription is equal 1 7 a7pli-cable to the 1J related buses with the appropriate changes to the cLesig. nators. The 111 4160 V AC bus feeds two 480 V AC buses (1H and 1H 1) through transformers. The 111480 V AC bus is primarily used to power pumps such as the A train low pressure injection pump. The 1H 1480 V AC bus feeds two motor control centers (MCCs), MCC 1H1 1 and 1H1 2, which provide power to-a multitude of MOVs and small pumps such as the charging pump cooling water pumps. MCC 1H1 1 also provides power to two battery char-gers used to charge DC battery A, and to the 11 120 V AC vital instrumentation by DC bus 1A through an inverter. The 1A 125 V DC bus provides control power to the switchgear for the pumps powered from the 1H buses. The 1A 125 V DC bus is powered from a 480 V AC bus, as noted above, and in the event of loss of the AC power source is powered from DC battery A. A simplified electrical diagram of the EPS is included in Figure 2.12. 2.2.15 Safety Injection Actuation System The safety injection actuation system (SIAS) automatically initiates the high and low pressure injection systems following an indication of the need for primary coolant makeup. The Surry SIAS is composed of two independent trains used to auto- l matically actuate the low and high pressure inj ection systems and the ' motor driven AFV pumps. The signals which actuate SIAS are shown in Figure 2.13. 2-22 1
l SwfTCWARO TO $W'tCWARD FWR et gg, , H R$4 C RSS A
, t
- t. t 1 :
. (. F - ,
y TO UNff 8 5 Q 5 Q 5 [] II sw C,] 1J II lA 10 tC g'g gg g I i W . II oc Sivo $fwa NORW AL 4150 V Butt $ (NOT MALY ED FURNER) 41s0 V 1H H1 J 1 i
// //
r WCC tW1-2 WCC1J11 WCC 1J12 WCC tut.1
- Ft9A FESA Ft98 Fito UPS 102 j l UPS tai l VP81 A2 l l UPS 181 l l 1
I I l gM8LS 1 o A yl[ g[- M B_US B g1n
~
BATTERY 1 A 6 ATTERY 18 l 35 1H 1N SS iM AC . IM C IM AC 1M C i.i m i.a iu Figure 2.12. Emergency Power System Schematic (1 of 2) 2-23
=
i
MCC1J11 ,MCC 1J12 FE98 FE9EU UPS 181 TRANSFORMER TRANSFORMER 43o y . 181 1 181 2 RECTIFIER E ASSEMBLY.
'24 00 lNVERTER INV UPSB1
. O
(' NOTE: UPS 181 SHOWN. UPS 1 A1,1 A2,182 ARE SIMILAR 3.; l: VITAL BUS 111 L-- Figure 2.12. Emergency Power System Schematic (2 of 2) l-1 2-24 n_ - , -- , .
S 11 M OV.1865 A A Chg Pp C 125 VDC A MOV.1289 A Si2 MOV.186S A 120 V AC A LCV.W1 111 SB u LCV.W1 1115C RELAY LOGIC - h A
$13 A
MOV.1867 C DGs1 NETWORK SI4 Chg Pp A A AFWP 3A SIS LHSIPs1 A TRAIN A ( S 11 MOV.136SB B Chg Pp C 125 VDC B LCV.W1 1115E 120 VAC - Sl2 LCV.W1 111SD B MOV.186SC y o MOV 1289B RELAY LOGIC h B sl3 B MOV 18670 DG #3 NETWORK Sl4 Chg Pp B B AFWP.3A
-SIS LHSI P s2 B
TR AIN B Figure 2.13. Safety Injection Actuation System Logic Diagram 2-25
2.2.16 Consequence Limiting Control System The consequence limiting control system (CLCS) automatically actuates the containment safeguards systems following receipt of an inoication of Hi Hi (25 psia) containment pressure. The Surry CLCS is composed of four containment pressure senaors, each feeding a signal comparator. The output of each signal comparator is input into two separate three out of four logic trains. These logic trains automatically actuate the containment safeguards system com-ponents. A simplified CLCS logic diagram is shown in Figure 2,14, 2.0.17 Recirculation Mode Transfer System The recirculation mode transfer (RMT) system automatically initiates the switchover of the suction of the low pressure injoction pumps from the RWST to the containment sump and the suction of the high pressure inj ec tion pumps from the RWST to the low pressure injection pump dis + charges on low RWST level. The Surry RMT system is composed of four independent RVST Icvel sensors, each feeding two separate two out of four relay matrices, laese two re-lay matrices automatically actuate the components required to perform the svitehover to the recirculation mode of the low and high pressure sys-tems, A simplified RMT system logic diagram is shown in Figure 2.15, 2.2.18 Residual Heat Removal System The residual heat. removal (RHR) system provides shutdown cooling when the reactor coolant system (RCS) depressurizes below 450 psig and is less than 3500F, The RilR is a front line system '(although nonsafety grade) designed to provide long term decay heat removal. The following sections _ provide a physical description of the RllR system, and identify the interfaces and dependencies of the RilR system with other front line and support systems. A simplified RHR system schematic is shown in Figure 2,16. The Surry RilR system is composed of two pumps and two RHR heat exchangers in parallel. The RllR pumps take suction from the RCS loop 1 hot-leg through two normally shut motor-operated valves (MOVs) and a manual iso- .lation valvo. The discharge of the pumps is headored together and feeds two heat exchangers arranged in parallel. The RHR pumps and heat ex-changers are cooled by component cooling water (CCW). An air operated valve - ( A0V) controls bypass flow around the heat exchangers, another controls flow through the heat exchangers. The two A0Vs work together to control the cooldovn rate of the RCS. The discharge _ of the flow control valves feeds into the SI/ accumulator piping and is delivered to the RCS loop 2 and loop 3 cold legs. Each path has a normally shut MOV isolating the LIIR from the high pressure RCS during normal plant operations. Make-up to the RilR system is provided by the RCS, The RilR is ma, ually initiated. An interlock prevents opening the RHR isolation MOVs until RCS pressure is below 450 psig. Following a loss 2 26 1
. . , . _ _ . _ _ _ _ . . . _ _ _ . . ~ . . . _ _ . . _ . ._._._m . ~ . _ . _ ._ _ . . _ . . . ._.
CR CLS 2A1 CR CLS 2A2 l PT CR j CLS LM. PS LM100A1 2A3 100A SIGNAL COMPARATOR CR VB11 ] - CLS l_ 3/4 2A4 3 l' : RELAY MATRIX CR CLS l 2A PT 2A5 d 'A CR LM. PS.LM100B1 1000 SIGNAL COMPAR ATOR CLS V B 1 11 CR CLS 2A7 TRAIN A TRAIN B CR CLS PT 281 LM-SIGI,
,M100C1 I"'
'O M P AR ATO h p ._'
CR CLS 100C ,,
-. 282 VI 1 111 . CR 3/4
" CLS RELAY 283 MATRIX
~
PT CLS 125 Vde B 2B4 LM- PS.LPi10001 1000 SIGN AL COMPARATOR CR CLS VB 1 IV = s 285 CR CLS 286 CR CLS 287 Figure 2.14. Consequence Limiting Control System Logic Diagram r-2 27
I i 120 VAC h' OPEN 1963A ---
'LT
" CLOSE 1885A TRAIN A 2/4 = - = -
R-CS100 A1- RELAY RELAY CLOSE 1885C CS I VB 1-I 100A J MATRIX R-CS100A2 OPEN 1860A LT : CLOSE 1862A
- ( 2 M:N TD }
" CS R CS10051 CLOSE LCV-1115B VB 1-11 IM8 R-CS10082 -
LT OPEN 18638 2/4 CS R-CS100C1 '
- RELAY RELAY : CLOSE 1885B TRAIN 8 VB 1-111 100C R-CS100C2
- MATRIK CLOSE 1885D
+
f OPEN 18606 l 120 VAC CS R-CS10001 : CLOSE 18628 VB 1-IV 1MD - ( 2 MIN TD )
- R-CS10002 CLOSE LCV-1115D Figure 2.15. Recirculation Mode Transfer System Logic Diagram 4 " ' ' '
v-,-.-% ,,_ G arM +m.-: .,
COMPOf1ENT.
^
COOUNG WATER TO dk iTNST aL r1L m R11 RH-19 MOV100 D Ril-11 M IS p RELIEF Ril-8 1F TANK Ril-12 v Ril-29 r 4 10- 10" Ril EI A p3 7 J( 14' PS4 . I II e
> CCW 7
) p e 1
FCC RIIR PUMPS ! S1 A dL llCV 1700 1701 0 1758 FE w ( SEALS / , 1605 14" to- r 4 h pgy E Vg Hil 5 Ril-6 4 M M Hil-2 (llX) 1-Ril-P-1 B Ril-20 All-24 MOV gl I 3 - 17200 A sL g. LOOP 1 ilOT LEG MM5 R 23 _ mg 1T RC-24 P89 17 29" ' f LOOP 2 1605 29" 'f 6 LOOP 3 COLDLEG Figure 2.16. Residual 11 eat Removal System Schematic - - _ . - . - - - - - - _ _ _ . - - _ _ _ _ _ - - _ - . - . _ - - - - - _ _ . _ -_ --- - - - - - - - - -- -___t- - - ----
l of offsite power, the stub buses powering the RilR pumps are shed from the emergency. buses and must be manually reconnected to restore power to the RHR pumps. 2.3- Jnitiatine rvents and Accident Seauences 2.3.1 Introduction This task involved the identification of potentially significant external event induced initiators at nuclear plants, identifying the applicability-of them _ to the Surry plant, and grouping the initiators into categories based on similar plant response and similar success criterla for successful initiator mitigation. It is not the intent of a focussed PRA to - explicitly evaluate (i.e., perform event sequence quantification) every possible initiating event. The intent is rather to evaluate those initiators which have previously-been shown to be important and to ensure that all other potential initiators can be adequately represented by those initiators chosen for explicit evaluation. The final list of initiating events which formed the basis for accident sequence quantification. are shown in Table 2,1. These aither seismically or fire-induced event sequences are described in the following sections. Table 2.2 details a description of the event headings for the event trees. From this list of potential initiating events the non recoverable loss of a DC bus was eliminated because the frequency of fire-induced failures was an order of magnitude below that of the internal event frequency and it is judged to be-highly unlikely that the portulated fire would spread _beyond these buses and cause other damage. Also, interfacing LOCAs were screened because a valid fire related mechanism that had not been addressed by the Appendix R submittal could not be identi;ied. It should also be noted that small LOCA (S2 ) fire and very small 1DCA (S )3 fire and seismic sequences had to be transient induced. 2.3.2 Ti (Loss of Of fsite Power) Event Tree - This st tion presents and discusses the event trees for the offsite power t initiacing event, This event is identified by the symbol Ti in the event tree. Loss of, offsite power will.- deenergize the normal and emergency 4160V buses, which will de energize all lower level buses. The DC buses and the vital buses would-be available, unless random failures of these buses were postulated. The reactor protection system will de energize, thus signaling the control rods to, insert. The main feedwater and condensate system will bc i unevallable for the duration of.the event. l I
-l u
l l
'2 30
Table 2,1 Initiating Event Categories Used in the External Events Analysis 1 1 External Event Description Catenory AbbreviatinD T3 Loss of Offsite Power Seismic / Fire T3 Transients with MFW Initially Available Seismic / Fire T 54 Mon Recoverable Loss of DC Bus A Fire Ts3 Non Recoverable Loss of DC Bus B Fire A Large LOCA, 6 in, to 29 in. Seismic Si Medium I'JCA. 2 in, to 6 in. Seismic S2 Small LOCA, 1/2 in, to 2 in. Seismic / Fire S3 Very Small LOCA, less than 1/2 in, Seismic / Fire V Interfacing LOCA Fire The T 2 ovent will affect both Unit 1 and Unit 2, Shruld DG 2 (dedicated to l' nit 2) fail to start or run, DG 3 would be aligned to Unit 2, thereby making it unavailable for Unit 1. In the event that both DG 1 and DC 2 fail to start, DG 3 was always assumed to align to Unit 2. The -four primary functions required in response to T2 are reactor scram, primary system integrity, auxiliary feedwater, and RCP sen1 cooling. If all these functions are provided, the transient is mitigated at a very early stage. Failure to provide reactor scram transfers to.the ATWS tree. Failure of PORVs to reclose transfers to the S2 LOCA tree. Failure to provide RCP seal cooling results in a seal vulnerabic condition which is evaluated separately. Failure to provide AFW 1eads to a demand for " feed and bleed" cooling. For feed and bleed, failure to-provide charging flow and open two PORVs leads to core damage, Successful feed and bleed cooling leads to a demand for the containment systems and coolant recirculation systems, These sequences are developed on the tree. The event. tree for T3 is shown in Figure 2,17. One event tree was uset .o evaluate the loss of offsite power initiating event which assumes at It one diesel initially available at Unit 1. 2-31
_ _ _ _ _ - _ - _ . _ _ = _ _ _ _ _ - - __ - ._ __ Table 2.2 Event Tree lleadings-Part 1: Description of Events ,, Abb r., llendinn Deter'iption of Event A LARGE IE - large IDCA (6 in, to 29 in,) LOCA CS CONT SYS Toi, level event for containment heat removal, includes CSS, ISR, and OSR system functions CV CORE VUINR Probability of core damage for core TO CD vulnerable statc6 (the core is being cooled but containment cooling has failed) D1 IIPI Failure of charging pump system in high pressure injection mode t D2 ilPI Failure of charging pump system in feed and bleed mode D3 SEAL COOL Failure of chargin6 pump systen in seal injec-tion flow mode D$. .ACC Failure of accumulators in injection mode D6 LPI ' Failure of low head safety injection system in injection mode ill- ~ LPR Failure of low head safety injection system in recirculation modo 112- HPR Failure of charging pump system in high pressure recirculation mode K RPS Failure of reactor protection system L AIN Failure of auxiliary feedwater system for transients with reactor trip L3 AFV Auxiliary feedwater: failure of 1/3 AFWPs to 1/2 SCs M MEN Failure of. main feedwater 2-32 Q__----_------_-__-__-______-___-________----.___--____-__-____
I l Table 2.2 (Continued) Event Tree Headings Part 1: Description of Events Abbr. Heading Description of Event OD~ OPER DEPRES Operator fails to depressurize RCS during small break initiators P PRV Failure of both PORVs to open for feed and bleed P1 PORV Failure of one PORV to open for Sat sequences PL PWR LEVEL Power level less than 25% of rated power Q RCI Failure of pressurizer SRV/PORV to close after transient QC RCI Failure of PORV to reclose after very small LOCA (SI causes relief valve to open) R MAN SCRAM Failure to effect manual reactor trip S1 MEDIUM IE - medium LOCA (2 in. to 6 ,) LOCA S2 SHALL LOCA IE small LOCA (1/2 in. to 2 in.) S3 VERY .!bCLL IE very small LOCA (less than LOCA 1/2 in.) SL RC? SEAL RCP seal leakage, limited to less JLDCA than 2 lb/sec/ pump Tl LOSP IE - loss of offsite power T3 TURB TRIP IE - turbine trip with MFW available W/MFW W CCW Failure oof component cooling water to thermal barriers of all reactor cooling system pumps W3- RilR Residual heat removal in shutdown cooling mode i. 2-33
Table 2.2 (Continued) Event Tree Headings Part 2: Definition of Events C - Less than 1/2 CSS trains taking suction from RWST and injecting into associated containment spray sparger. D3 - Less than 1/3 high pressure injection pumps taking suction from RWST and injecting through MOV 1867 C/D into 1 of 3 RCS cold legs. Initiated by SI signal. D2
- Same as D , except must be initiated by operator.
3 D3 - Less than 1/3 charging pumps injecting through MOV 1370. D4 - Less than 1/3 charging pumps injecting through the normal charging lines with the BAT pumps on fast speed, MOV 1350 open, and one PORV open within 10 min from initiator. SI alignment not re-quired. D3 - For A, less than 2/2 accumulators injecting into their associated cold legs, For S, 3 less than 2/3 accumulators injecting into their associated cold legs. De
- Less than 1/2 LHSI trains taking suction ' from the RWST and injecting through MOV 1890C to 1/3 RCS cold legs.
Fi - Less than 1/z ISR trains taking suction from the sump and injecting through associated spray sparger, with service water being provided to the secondary side of the heat exchanger. F2 - Less than 1/2 OSR trains taking suction from the sump and injecting through associated spray sparger, with service water being provided to the secondary side of the beac exchanger, j H- - Less than 1/2 LHSI pumps-taking suction from the sump and ~ i injecting to MOV 1890C, or injecting to -the charging pump suction.
.Plus switch to hot leg recirculation et 16 hr for A and St LOCAs.
H2 - Less than 1/3 charging pumps taking suction from the LHSI discharge and injecting through MOV 1867 C/D. > 'K - Failure of automatic insertion of sufficient control rods to pro-duce suberiticality at hot snutdown. L - Less than 1/3 AW pumps delivering water to 1/3 steam generators 2 34
l Table 2.2 (Concluded) l Event Tree Headings Part 2: Definition of Events l I 19 - Less than 2 motor driven feed water pumps (MDWP) or 1 turbine-driven auxiliary feed water pump (TDAWP) delivering flow to 2 to 3 stream generators. i La
- Less than 1/,' AW pumps delivering water to 1/2 steam generators.
M - Failure of at least 1 main feedi,,ater pump delivering flow to at least one steam generator, and a source of water from the hotwell or CST which is sufficient for 24 hr. P - Failure of at least 2 PORVs and associated block valves to open. Initiated _by manual action. P:
- -Less than 1/2 PORVs and associated ulock valves open. Initiated by operator.
Q - Failure of pressurizer PORVs to reclose or be manually isolated after a transient. W - Failure of component cooling water-supplied to the lower bearing b at exchanger of all reactor coolant pumps. The T 3 event tree represents sequences where at 1 cast one diesel is available at ~ Unit 1. Sequence 1 of- the T 3 event tree represents successful mitigation of the initiator;- diesel generators start, auxiliary feedwater is available, and the charging system provides seal injection flow to. the RCP seals. The plant is in a stable condition and attention can be directed to restoration of the offsite= power. Sequence 2 L is similar to 1, except that seal injection flow from the charging system is unavailable. RCP seal cooling is provided by CCW to the thermal barrier heat exchangers. Sequence 3 represents a condition with no seal cooling- available. 'Both CCW to the thermal barriers and seal injection flow have failed. Auxiliary feedwater is = available , however, and all essential safety functions are being provided at the time seal cooling is lost. This represents a seal vulnerable condition and is handled with the seal LOCA m del. Sequence 4 represents failure of all . steam generator heat removal with successful core cooling via feed and bleed, using one charging . pump. and opening of both PORVs. ECCS recirculation from the sump and successful operation of the containment spray- recirculation heat exchangers provide long term cooling. Sequences 5 and 6 lead'to core damage through failure to provide long term feed and bleed cooling in the recirculation mode. Sequence 5 is due, to failure of ; the high pressure recirculation- systern. Sequence 6 is due to failure of J l I 2-35 i l l
LOSP RPS RCI AFU SEAL CCW HPI PRV CONT CORE LPR HPR COOL SYS WLNR TO CD T1 -K -0 -L -D3 -u -02 -P -CS -CV -H1 -H2 Segjence l CORE l CormENTS
- 1. I1 OK l ! 2. T1-03 OK I 3. T1-D3-V --
SEAL VtfLN
- 4. T1-L OK I 5. T1-L-H2 CM
- 6. T1-L-H1 CM
- 7. T1-L-CS OK 7 I 8. T1-L-CS-H2 CM
- 9. T1-L-CS-H1 CM
- 10. T1-L-CS-CV CH
- 11. T1-L-P CM
- 12. T1-L-D2 CM
- 13. T1 --
GO TO S2
- 14. T1-K --
60 TO A1US Figure 2.17. Event Tree for T1 --Ioss of Offsite Power
. . ~
the low <5 essure recirculation system. Sequences 7 through 10 represent the occ) - -" of a core vulnerabic state and its possible outcomes. A core vu , state occurs when containment heat removal fails af ter feed an. tile e d is initiated. Coolant makeup _ to the core is being p rovi c.ed and heat is being removed from the RCS through the PORVs. However, containment heat rernoval (CHR) has failed, thereby leading to gradual containment pressure increase. Should the containment pressure increase continue, unmitigated by containment venting or restoration of CHR system:, containment overpressure failure will ocesr. Events occurring during containment failure could cause failure to ECCS systems, which in turn would lead to core damage. This is represented by Sequence
- 10. Sequence 7 represents containment failure, but survival of the ECCS and continued core cooling. Sequences 8 and 9 represent containment failure, followed by ECCS failure due to causes other than containment failure.
Sequer.cc 11 represents f ailure of steam generator heat removal followed by f ailure to establish feco and biced cooling, due to failure to open both PORVs. Sequence 12 is similar to 11, except feed and bleed core cooling fails due to failure to establish safety injection flow with the charging system. Sequence 13 represents transient induced LOCAs caused by a transient related PORV demand, followed by failure to reclose PORV. This condition transfers to the S 2 event tree for further evaluntion, Sequence 14 is an ATWS condition. 2.3.3 T3 (Turbine Trip with MW Available) Event Tree This section presents and discusses the event tree for the turbine trip initiating event group in which the main feedwater system remains availa-ble. Transients in which one or both MFW pumps remain available are considered. This event is identified by the symbol T3 in the event tree. This initiating event group represents a fire or seismic induced manual scram or turbine trip, PORV demand for this class of initiators is con-sidt. red to be a random occurrence, due to degraded control system perfor-
.mance or degraded balance-of plant (BOP) components performance. The probability of PORV demand was assigned a value of .014, for high power initiators only, based on historical Westinghouse experience. The MFV control system at - Surry is - such that if the reactor trip breakers are closed and Tm is less than 543cF, the main feedwater regulating valves will close, the miniflow lines will open, and the MW pumps will stay on.
This was assumed to be the course of all T 3 initiating events. Although the MFW pumps are isolated from the steam generators, they remain a viabic source of SG inventory makeup, should AFW be unavailable. AFW is the preferred source of SG makeup, but MFW pumps can easily be used by opening the feedwater regulating valve bypass valve. Because AFW is the preferred source of SG makeup, it appears on the tree before main feedwater. Four primary functions were required to successfully mitigate the T3 events. These functions are reactor scram, RCS integrity, SC inventory makeup, and RCP seal cooling. If all those functions are provided, the transient will be mitigated at a very early stage. Failure to provide 2-37
~ . _ . _ _ .
reactor scram transfers to the ATWS tree. Failure of PORVs to reclose transfers to the S2 LOCA tree. Failure to provide RCP seal cooling leads to a seal vulnerable condition. Failure to provide feedwater leads to a demand for " feed and bleed" cooling. For feed and bleed, failure to provide charging flow and open two PORVs leads to core damage. Successful feed and biced and cooling leads to a demand for containment systems and coolant recirculation systems. The event tree for T3 is shown in Figure 2.18. The first sequence re-presents successful acabilization of the reactor at hot shutdown. Reactor scram is successful. AFn starts and provides water to at least one of three steam generators. Heat removal is via the steam dumps to the condenser. Seal cooling is provided by seal injection flow. At this juncture in the tree, the reactor is stable in hot shutdown. This is considered successful termination and no further syst .a availability questions are asked. Particularly, the availability of RHR which is necessary to reach cold shutdown - is not asked. Sequence 2 is also a success state, with seal cooling being provided by CCW to the thermal barrier, Sequence 3 is a seal vulnerable condition. All critical safety functions are being provided, but RCP seal cooling is not available. The potential for this sequence to lead to care damage depends on the sus-ceptibility of seals to failure after loss of all cooling and the potential recovery options to restore seal cooling prior to seal failure. The seal vulnerable evaluation will be cone on an individual sequence basis, should the quantification show this state to be important. Sequence 4 represents stabic hot rhutdown with SG inventory being pro-vided by main feedwater, after failure of auxiliary feedwater. This is a , success state similar to Sequence 1, except of a much lower probability. Questions of seal cooling were not asked on this branch, because the additional sequences would be subsets of Sequences 2 and 3. Sequence 5 represents loss of auxiliary feedwater and all main feedwater, but suc-cessful feed and bleed cooling, using contaitunent heat removal systems and reactor coolant recirculation systems. Long term feed and bleed cooling requires-high pressure coolant recirculation. Sequence 6 repre-sents core damage due to failure to provide high pressure recirculation for long term cooling. Sequence 7 is similar to 6, except that the low pressure recirculation systems are unavailable. Sequences 8 through 11 represent successful feed and biced cooling, but failuro of containment heat removal. In Sequence 8, containment failure does not lead to structural or phenomenological failure of the ECCS, therefore, core cooling is successful. Sequences 9 and 10 represent ECCS survival of the containment failure, but failure due to random other causes. Sequence 11 represents ECCS failure due to containment failure, Thus, Sequence 11 represents containment failure prior to core damage. Sequences 12 and 13 represent failure to initiate feed and bleed cooling af ter loss of auxiliary feedwater. In Sequence 12 feed and bleed fails due to failure of 2 of 2 PORVs to open, while in Sequence 13 feed and bleed fails due to failure to establish safety injection flow. 2-38
TURB RPS RCI AFU MFU SEAL CCU HPI- PRV CONT CORE LPR HPR TRIP COOL SYS VUL NO.
'J MFU TO CD T3 -K -0 -L -H -D3 -u -D2: -P -CS -CV -H1- -H2 Sequence l CORE ] COMMENTS l
- 1. 13 OK I 2. 13-D3 OK I 3. T3-D3-U --
SEAL VULN
- 4. T3-L OK I 5. T3-L-M OK
! 6. T3-L-M-H2 CM qa 7. T3-L-M-H1 CM g 8. 13-L-M-CS OK
! 9. T3-t-M-CS-H2 CM
- 10. T3-L-M-CS-H1 CM
- 11. T3-L-M-CS-CV CH
- 12. T3-L-M-P CM
- 13. T3-L-H-D2 CH
- 14. 13-0 --
Go 10 S2
- 15. T3-K --
Go 70 Alus Figure 2.18. Event Tree for T3 - Turbine Trip With MFU
l Sequence 14 is a transient induced LOCA, which transfers to the Sg tree for further evaluation. Sequence 15 is an ATWS sequence. 2.3.4 Large L?CA Event Tree This section presents and discusses the event tree for the large LOCA initiating event. This event is identified by the symbol A in the event tree and covers break sizes ranging from 6 to 29 in. The event tree for large lhCAs is shown in Figure 2.19. to the initiator Sequence 1 represents a completely successful response in which all systems function as intended. The accumulators inject water immediately to accommodate the initial high volume surge of water from the reactor cooling system. Iow pressure injection subsequently provides the high volume, low pressure flow required for continued core cooling, g The containment heat rencval systems successfully maintain containment pressures and temperatures at acceptable levels, and recirculation p cooling is established from the containment sump to provide long term cooling. Sequence 2 leads to core damage because of a failure to provide low pres-sure recirculation cooling. No other system can provide the volume of flow needed under lar3e IDCA conditions. Sequences 3, 4, and 5 representA the occurrence of a core vulnerable state and its possible outcomes. > core vulnerable state occurs when containment heat removalUnder fails afte: such core cooling has been established by low pressure inj ec tion. circumstances, heat is being transferred from the core to the containment via the water flowing through the opening in the RCS pressure boundary. As a result, the pressure and temperature in the containment rise due to the lost containment heat removal (CHR) capability. If the containment pressure continues to increase without being mitigated by containment venting or restoration of CHR systems, containment overpressure failure will occur. Events occurring during containment failure could cause ECCS systems to fall, which would lead to core damage. Such a scenario is represented by Sequence 5. Sequence 5 represents containment failure, but the ECCS survives and continues to cool the core. Sequence 4 repre-sents containment failure together with independent f ailure -of the ECCS (i.e., due to causes other than the containment failure). Sequence 6 represents failure to the ECCS to respond early in the sce-nario to-provide the high volume , low pressure injection flow needed to
-cool the core, thereby leading to core damage. In Sequence 7 the occumulators fail to inj ec t water immediately as the pressure in the reactor coolant system drops suddenly as a result of-the large break in the cooling system pressure boundary. This sudden loss of coolant inven-tory causes core damage.
2-40
l LARGE ACC LPI CONT CORE LPR LOCA SYS VULNP , . TO CD A -D5 -D6 -CS -CV -H1 Sequence l CORE
- 1.'A. OK
! 2. A-H1 CM
- 3. A-CS OK
! 4. A-CS-H1 CM w 5. A-CS-CV CM .
+
. t
- 6. A-D6 CM
$ 7.'A-D5 CM' ,
i i
?
i i Figure 2.19. Event Tree for A - Large LOCA s
?
~ s
_2.3.5 Medium 1DCA Event Tree This section presents and discusses the event tree for the medium LOCA initiating event. This event is identified by the symbol Si in the event tree and covers leak sizes ranging from 2 to 6 in. Success criteria for Si are distinctively different A and S. 2 These differences were derived from requirements for AW, accumulators, llPI/R and LPI/R. The Si events will maintain the reactor moderately pressurized during the early time frame, thus requiring early inventory makeup from HPI. As the pressure declines the accumulators and LPI are required. A requirement for high pressure recirculation is not necessary, because pressure will be below shutoff head for low head safety injection (LHSI) pumps at tne time of recirculation. The event tree for medium LOCAs is shown in Figure 2.20. Sequence 1 represents a completely successful response to the initiator in which all systems function as intended. High pressure injection im-mediately provides the high pressure initial flow required for core cooling. The accumulators inject water to accommodate the initial high-volume surge of water from the reactor cooling system. The containment heat removal systems successfully maintain containment pressures and temperatures at acceptable levels, and low pressure injection and recir-culation cooling are established to provide long term coolin6-Sequence 2 leads to core damage because of the failure to provide low pressure recirculation cooling. No other system can provide the volume of flow- needed under the low pressure conditions that follow a medium LOCA. Sequence 3 denotes failure to establish low pressure inj ec tion, which is required before enough water accumulates in the containment sump to allow recirculation cooling. Sequences 4, 5, 6, and 7 represent the occurrence of a core vulnerable state and its possible outcomes. A core vulnerable state occurs when containment heat removal (CHR)' fails after core cooling has been es-tablished by high pressure injection. Under such circumstances, heat is behg transferred from the core to the containment via the wcter flowing through the opening in the RCS pressure boundary. As a result, the pres-sure and temperature in the containment rise due to the failed contain-ment heat removal capability. If the containtnent pressure continues to increase without being mitigated by containment venting or restoration _of CHR systems, containment overpressure failure will occur. Events occur-ring during containment failure could cause ECCS systems to fail, which would lead.co core damage. Such a scenario is represented by Sequence 7. Sequence 4 represents containment failure, but the ECCS survives and i continues to cool the core. Sequences 5 and 6 represent containment failure together with independent failure of the ECCS (i.e., due to causes other than the containment failure). In Sequence 8 the accumulators fall to inj ec t water immediately as the pressure in the reactor coolant system drops suddenly as a result of the ; i 2-42 l l_
b 1%.; m, L 4 4 INTER HPI 'ACC CONI. CORE LPI LPR i t4EDI A SYS WLNR LOCA 10 CD , S1 -D1 -D5 -CS -CV -D6 -H1 Sequence l_iOREl
- 1. S1 -OK
! 2. SI-M1 CM
- 3. St-D6 CM
- 4. 51-CS OK y
' I 5. SI-CS-H1 CM
; ') " 6. 51-CS-06 CM
' 7. SI-CS-CV CM
- 8. 51 CM
- 9. 51-D1 CM Figure 2.20. Event Tree for S2 - Medium LOCA i
l j
._z._ ___
. _ _ , ~
1 I o medium break in the cooling systere pressure boundary. This sudden loss of coolant inventory ccuses core damage. Sequence 9 represents failure of the ECCS to respond early in tne scenario to provide the high pressure injection flow needed to cool the core, thereby leading to core damage. 2.3.6 Small LOCA Event Tree This sect /n presents and discusses the event tree for the small LOCA initiating event. This event is identified by the symbol S2 in the event tret and covers leak sizes ranging from 1/2 to 2 in. Sa success criteria are a combination of t rar.s i e n t and LOCA type criteria. The break is not sufficient to depressurize the reactor, so that large volume ECCS systems are not effective. Thus the need for
- control rod insertion, because the ECCS boration function vill not be perform 6d.
AFW is required for successful Sa mitigation, because the break size itself is not sufficient to carry away decay heat and pump heat. If AW is unavailable, afeed and bleed" cooling is viable if the operator opens -. one PORV. The event tree for S 2 is shown in Figure 2.21. N ' Sequence 1 represents a completely successful response to the initiator in which all systems function as intended. The reactor protection system successfully scrams the reactor. High pressure injection provides the initial high pressure flow required to replace the lost inventory. The . auxiliary feedwater system provides core heat removal via the steam generators. The containment heat removal systems successfully maintain containment pressures and temperatures at acceptable levels. The oper-ator successfully depressurites the RCS, and recirculation cooling is established to provide long term cooling, using the low pressure recir- ' culation systems. Low pressure recirculation from the sump was required for successful mitigation, because shutdown cooling on RRR may not be possible due to break location, t Sequence 2 leads to core damage because of a failure to provide low pressure recirculation cooling. Sequence 3 represents successful miti-gation af ter the failure of the operator to depressurize the RCS. Failure to depressurize the RCS leads to the requirement for high pressure recirculation. If either low or high pressure recirculation fails, core damage results as indicated by Sequences 4 and 5. { Sequences 6 through 11 cover the case in which the containment heat re-moval systems fail af ter core inventory is being maintained via high pressure inj ection and core cooling has been established by the AW system. Whether or not this can lead to a core vulnerable state depends on whether or not the operator depressurized the RCS. If operator de-pressurization occurs, SG heat removal is not effective and a core vulnerable state can occur. Under such circumstances, heat is gradually being transferred from the core to the containment via the water flowing unrough the opening in the RCS pressure boundary. As a result, the pressure end temperature in the containment rise gradually due to the 2-44
HPI AFU PRV cot!T - OPER CORE LPR HPR SMAtt RPS SYS DPRES VULNR LOCA TO CD
-CV -H1 -H2 Sequence 0011MENTS S2 -K -D1 l -L -P1 -CS ,-00' l CORE l
- 1. 52 OK
! 2. 52-H1 CM
- 3. 52-00 OK I 4. 52-00-H? CM
- 5. 52-00-H1 CM
- 6. S2-CS OK I 7. 52-CS-H1 CM
- 8. 52-CS-CV CM
- 9. 57_-05-00 OK I 10. 52-CS-00-H2 CM e
- 11. S2-CS-00-H1 CM
- 12. 52-L OK h ! 13. 52-L-H2 CM
- 14. S2-t-F1 CM
- 15. S2-L-CS OK
! 16. 52-L-CS-H2 CM
- 17. 52-L-CS-H1 ' 25
- 18. 52-L-CS-CV b;
- 19. 52-t-P1 CM
?O. 52-01 EM
- 21. 52-K CM CO TO ATits Figure 2.21. Event Tree for 52 --Srnall WCA
l lost containment heat removal (CHR) capability. If the containment pressure continues to increase without being mitigated by containment venting or restoration of CllR systems, containment overpressure failure will occur. Continued heat removal through the steam generators has been shown to be sufficient to prevent containment overpressure failure in these cases. Events occurring during containment failure could cause ECCS failure which would lead to core damage. Such a scenario is repre-sented by Sequence 8. Sequence 6 represents containment failure, but the ECCS survives and continues to cool the core. Sequence 7 represents containment failure together with the independent failure of the ECCS (i.e., due to causes other than the containment failure). If the oper-ator keeps the RCS pressurized and thus supports steam generator heat removal (as represented by Sequences 9, 10, and 11), then the containment - overpressure failure is averted, even though containment heat removal systems have failed. Under such circumstances the containment is net expected to fail, and the "CV" question is not asked. Sequence 9 repre-sents successful functioning of the ECCS in the recirculation mode. Sequences 10 and 11 represent ECCS failure, which results in core damage. Sequences 12 through 19 address the sequences with auxiliaty feedwater failure. If AW is lost, core cooling can be accomplished by opening a PORV to increase the breakflow. Now sufficient water is lost from the RCS to carry away all decay heat. The charging pump is known to be suc-cessful at this point in the event tree. Sequence 19 represents failure of either PORV to open. Sequences 12 through 18 address the potential for a core vulnerable state due to failure of ChR. A core vulnerable state occurs when containment heat removal fails after feed and bleed core cooling has been es-tablished. Under such circumstances, heat is being transferred from the core to the containment. The pressure and temperature in the containment rise due to the lost containment heat removal capability, If the con-tainment pressure continues to increase without being mitigated by containnent venting or restoration of CHR systems, containment over-pressure failure will occur. Events occurring during containment failure could cause ECCS systems to fail, which would lead to core damage. Such a scenario is represented by Sequence 18. Sequence 12 is A W success and no core damage. Sequences 13 and 14 are AW success but long term recirculation failure leads to core damage. Sequence 15 represents containment failure, but the ECCS survives and continues to cool the core. Sequences 16 and 17 represent containment failure together with independent failure of the ECCS (i.e. due to causes other than the y containment fa12ure). In Sequence 20 the ECCS fails to respond to the small IACA initiator and to provide the initial high pressure ' nj ec tion flow needed to cool the core. In Sequence 21 the RPS fails to scram the reactor. 2.3.7 Very Small LOCA Event Tree This section presents and discusses the event tree for the very small LOCA initiatint; event, This event is identified by the symbol S3 in the event tree. This group of LOCAs includes spontaneous seal LOCAs and very small breaks, with leak sizes equivalent to less than approximately 1/2 in. break. , 2-46
The system success criteria are very similar to the S 2 criteria. Iloweve r , timing considerations due to the impact of the very small Icak rate have a significant impact on the recirculation requi'ements, lle at removal from the RCS by the AW combined with the containment fan coolers ad natural cooling / condensation processes are expected to main-tain containment pressure well below the spray actu stion point. With only the HPI flow draining the RWST, S breaks could remain in the in-3 jection phase for a long time. If the operator takes action to depressurize the RCS, thus reducing the leak rate from the RCS, the reactor can be depressurized and in cold shutdown long before depletion of RVST inventory forces a switch to recirculation. The event tree for S3 is shown in Figure 2.22. Sequence 1 represents a completely successful response to the initiator in which all systems function as intended. The reactor protection system successfully scrams the reactor, liigh pressure injection provides the high pressure initial flow required for continued core cooling. The RCS relief valves reclose if opened, auxiliary feedwater cooling is initi-ated, the operator depressurizes the RCS, and the residual heat removal system is available to provide shutdown cooling. Sequence 2 addresses the case whero residual heat removal system is un-available and low pressure recirculation cooling is required to provide long-term core cooling, If LPR fails (as in Sequence 3), then core damage vill result. Sequences 4, 5, and 6 address the cases where the operator does not depressurize the RCS, Continued blowdown leads to RWST depletion which forces recirculation. Sequence 4 represents wuccessful switch to high pressure recirculation. Sequences 5 and 6 represent core damage due to failure of high and low pressure recirculation. Sequences 7 through 21 represent all cases in which the primary mode of steam generator feedwater supply is lost. In Sequences 7 through 13, main feedwater supplies steam Senerator feed flow. These sequences have much the same characteristics as Sequences 1 through 6. Sequences 14 through 21 address the case that both AW and MFW have been lost. In this instance, it is necessary to establish feed and bleed cooling. Both PORVs must open to allow water to flow f rom the RCS, to l l remove decay heat. A single charging pump is required to supply makeup to replenish the PORV discharge. If feed and bleed cooling is lost l (Sequence 21), then core damage results. Sequence 14 represents suc-cessful feed and bleed cooling followed by long term cooling in the recirculation mode. If either high pressure or low pressure recir-culation cooling is lost (as in Sequences 15 and 16), then core damage results. Sequences 17 through 20 represent the occurrence of a core vulnerable state during successful feed and bleed cooling. A core vulnerable state 1 I-2-47 1 1
..l
VERY RPS HPI RCI AFU MFW PRV CONT CORE OPER RHR LPR HPR SMALL SYS' VULHR DPRES LOCA TO CD S3 -K -D1 -0C -L -H -P -CS -CV -0D -U3 -H1 -H2 Sequence l CORE l CoeU4ENTSl
- 1. S3 OK I 2. S3-u3 OK I 3. S3-ul-H1 CM
- 4. S3-00 OK
! 5. 53-00-H2 CM
- 6. S3-00-H1 CM 7..S3-L OK
! 8. 53-L-H1 CM
- 9. S3-L-U3 OK I 10. 53-L-U3-H1 CM s,
- . 11. S3-L-00 OK
"* I 12. S3-L-00-H2 CM
- 13. S3-L-00-H1 CM
- 14. S3-L-M OK I 15. 53-L-M-H2 CM
- 16. S3-L-M-H1 CM
- 17. S3-t-M-CS OK
! 18. 53-L-N-CS-H2 CM
- 19. S3-L-M-CS-H1 CM 90.'S3-L-M-CS-CV CH
- 21. S3-L-M-P CM
- 22. 53-0C --
GO TO S2
?3. S3-D1 CM
- 24. 53-K --
CD TO ATuS Figure 2.22. Event Tree for S3 --Very Small LOCA e_ > _ _ _ _ _ _ . _ . _ _ _ . _ _ ___._m.
- . .. -- _ _ .- -- . . ~
l occurs when containment heat removal fails after core cooling has been established in the feed and bleed mode. Under such circumstances, heat is being transferred from-- the core to the containment. (A core vulner-able state cannot occur in Sequences 2 through 13 in the event tree because an insufficient amount of hot water is transferred into the con-tainment to cause overpressure.) As a result, the pressure and temper-ature in the containment rise due to the lost containment heat removal capability. If the containment pressure continues to increase without being mitigated by containment venting or restoration of CHR systems, containment overpressure failure will occur. F onts occurring during containment failure could cause ECCS systems to ' which would lead to core damage. Such a scenario is represented by aquence 20. Sequence 17 represents containment failure, but the ECCS survives and continues to cool the core. Sequences 18 and 19 represent containment failure togeth-er with independent failure of the ECCS (i.e., due to causes other than the containment failure). Sequence 22 represents the case in which SI flow causes the RCS relief valves to open, and one of the valves fails to rescat. This leads to a larger LOCA size, which requires analysis via the small LOCA event tree. In Sequence 23 the ECCS fails to respond to the LOCA initiating event and to provide the initial high pressure injec-rion flow needed to cool the core. In Sequence 24 the RPS fails to scram the reactor. 1 2-49
3.0 SCOPING QUANTIFICATION STUDY A scoping quantification study was performed for Surry Power Station site to determine which external events should be included in the detailed PRA study, This scopin6 study considered all potential external hazards at the site except-for seismic and fire events, since these N 'nts were already scheduled for a detailed risk analysis. The PRA Proc < a es Guide (Ref. 1) was used as a guidaline for systematic identification of the external events at the site. Next, an initial screening process was carried out to climinate as many events as possibic from the list. For this purpose, a set of screening criteria was developed and then each external event was examined for possible climination based on these criteria. After the initial screening process was completed, it was found that the following events could not be screened out based on the general screening criteria:
- a. Aircraft Impact
- b. External Flooding
- c. Extreme Winds and Tornados
- d. Industrial or Military Facility Accidents
- e. Pipeline Accidents
- f. Release of Chemicals from om Site Storage
- g. Transportation Accidents
- h. Turbine Cencrated Missiles 1, Internal Flooding A bounding analysis was done for each of these events The degree of sophistication in the bounding analysis for each event dupended on whether the event could be eliminated based on only a hazard analysis or whether a complete analysis including hazard analysis, fragility evaluation and plant response analysis was required.
This chapter covers the screening and boundin6 analyses for the external events as part of the scoping quantification study of the Surry . Power Station. Section 3.1 is a general description of the plant and its location. Section 3.2 deals with the idencification and screening of external events for this site. A number of the events could be screened based on the Surry Updated Final Safety Analysis Report (FSAR) (Ref. 2) and its supporting documents as discussed in Section 3.3. Finally, the remainin6 cxternal hazards were screened out using a bounding analysis as described in Section 3.4. Section 3.5_ summarizes the results of the screening study. 3 .L 1 General Description 3.1.l Site The Surry Power Station is located in Gravel Neck, Virginia at approxi-mately-37' 10 ft N, 76* 42 ft W. The peninsular site is bordered by the James River and the llog Island Waterfowl Refuge. This wildlife area is marshy and covered by many streams and creeks. The. site is 8 miles from the town of Surry and is at the end of Route 650 (a state secondary route). This road provides the only land access to the area. Also, a-1 I 3-1
.o n n = -a--- - -..+~>-~~u..mraswa,- - s. a a - - - .am uu n. s ~x-- . s , .. -~ ,
'l l
public access road- to the waterfowl refuge runs through the power plant site. The topography in macro and micro scales is shown in Figures 3.1 through 3.3. The site occupies 840 acres and the area within 10 miles of the site' is predominantly rural, with a few small urbanized segments. The neighbor-ing area is characterized by farmlands, marshy wetlands, swamps, and small streams. The water table is near the surface throughout the area and drainage is toward Hampton Roads, on the Atlantic Ocean and near the mouth of the Chesapeake Bay. The ground surface at the site is generally flat, with steep banks sloping towards the river and to the low level waterfowl refuse, pre construction elevation whhin the site boundaries varied from river level to 39 ft, with a mean olevation of 34 ft. Station ground grade for the site was established at 26.5 ft above the 1 mean sea 1cvel. The resident population in 1980 was estimated to be 1,759 within 5 miles of the site and 61,711 within 10 miles. The nearest city is Newport News, with a population of 114,903 which is , however, only 41/2 miles across the James River. In addition, there is a transient population of 25,000 per year at the public recreational facilities (beaches, boat landings, fishing areas, etc.), 2.16 million at the Busch Gardens / Anheuser Busch brewery (6 miles north of the site), and 1.5 million to i 2.5 million per year at the historical attractions in the Williamsburg-Jamestown area (4 to 7 miles north of the site). Further details regarding population projections are available in FSAR. The roads, railways, and airports in the vicinity of the site are shown in Figures 3.2 and 3.4 The location of the natural gas pipelines is shown in Figures 3.3 and 3.5. As seen from these, two pipelines cross the southeast corner of the site. The closest industrial facilities to , the site are a brewery plant (6 miles), a synthetic fibers factory (5 miles), and some food processing units. The U.S. Army Transportation Center at Fort Eustis is within 5 miles of the site. There are no known mines or stone quarries within 5 miles of the site. The Surry site experiences a high' variability in temperature extremes. For example, extreme . temperatures recorded at nearby Richmond range from 12'F to 105'F. Temperature data from Norfolk indicates a range of 5'F to 104*F. The maximum recorded precipitation for a 24-hou_r period was 8.79 in, at Richmond and 11.4 in, at Norfolk. The maximum 24 hour snowfall observed at the two stations was 21.6 in, and 12.4 in., respectively. The local climatological data indicates an average of 29
-days per year of heavy . fog (i.e. , visibility of 1/4 mile or less) for Richmond and 21 days for Norfolk. The site experiences a wide spectrum of extreme winds and tornadoes. The one hundred year wind speed is estimated to be 105 mph 'and
- sing a gust factor of 1.3, the highest instantaneous gust expected is 137 mph. During the period 1951 through 1982, a total of 30 tornadoes were reported within 50 miles of the site.
In addition, an average of two storms / hurricanes per year bring torrential; rainfall to the tidewater areas, and high tidos result in flood conditions for low-lying areas along the coast. 3-2
i.Aq.k.ad AmmAh4.im.htz JiaaA-e-ase 5 a _a mmL4 waEn-&chJ ha A J AJ".--A R-+444,J.h as d 4 -4p.-S.a=as c--A-4-44.-a+A-mL.il es aan2<WA--M.E .A--dah-e-a4a.NJ amm-A .L,1.M.5 4aa _.hc ..+,6 I E F v % [.,,%y j y W g x % vgf V d 9rfid!Widds ... Q) . f, c , y, ,.
~ p /~
y% ,,. 4g .....\,:sg.
.~s
- ny>
g
+.
$ ;+ W j ,,A ,
k N j Nyt mjf,
!?
f '/
~' .
+ /
y
,h o
- ~
v ; b p g ., J .g y ?/'(n '
' ' ;i,. n p fe 3 y l
s s : 2. ,.
+
i
. ./ .a i ;
ger- C ., . - , , s m i u Fi l T
.r , u e g s .. . , ,
[ . . ~ m 9 N, ..ie N .* m
sW _s V
, hg k,e 1, . . .i. ;- , ,,
g.'k'.,A E
,1 >,.r.$,e, Jo - - v ,
,n \ 7-j ~. y gei z. ' ,,
g '
- 'C-
.s .,,,, t= ,,,, y o
,4 ., .. , t g . , , , .
lL
, . f,'t'n ~
N ' Y
~
"' -Q _ f y r ...
3 y ;y N:
- s .
- 3. .
3, f 3 y c
,D '
W f ,. i e "" . ,' -
, ! , [-
/ = =P '4 \'p .
( - d 4 3
< xg .[n.. , ,
! 3 l 9.-s.a%
= .
l .
- ~
q , s
- s. . .
.\
c . - e s, ; s
% f
,ej f ' ;. y '. n f. g h.,,h, -!
p 3 N, g.. %7 , 3 i l .s ':i Vs ., .a aV1, l i t i..% . " u l. 5
- s ,
e T. - ' n
. s , , _ .
'e ',
I
'a y4 .
\' M- -
~~
g
.3
'.g,Dg N,
.e _ s , . -
e
-s M
- -Cg '
, r E' , . . g -- . if w
i s ; '
/j - -
< ,< . , . . - t ,.
'., j ,; ,
N ,
-j f).f - _.-
.1 . ~
.e l
'.&' e~
h
~
$ Ts / / 4. v, / *
,- , - t
) .. . e / .
.. t s '
i
- [. /.,-
-{ f-i
/ Q,' , . ...'N e .
. .n. .
4 f 3-3
ss@*'fmeevw[! tko a'\1 p h pv 4/ %'u
't .
t : x :-~Uc.
. s4 -
.kW]' g f ', , s
% l Os + .
' '," /h ,,,, _
*Qi / >
[w #} 4 N q ( - p
/
Psr ' - Yd j ' S ot kC J
%.y j , , ,
t ;,j . 7- 'Y g
' g& %4, *
-/
p 6 [-' *
#Ab , /
, ,, 5 4N *['~&!~.fig 7/j. T ,. . , 4, u m[, '
- l 4(
w' y[ s Xa-- g ,Y -
'._ . }
= ;
*M3 .
N / \c 4 - D, , .
-N s ., go
.h ._ ,
N ,( M .-
^^
6,
/ .
,h I y ,e p' .
SP7'" W,p'* E kJdI , 30 ' f f
.- /.'
[ y7'p yitg s A
,r,N %L p ' {= f g' L fA.'a er _.
sW 1
~- . . d s.iL.,.:
A . _ i.,;.' q - v - w 3m . Z- . . s g , S * 's 7 ) [T*' W gh4 ,. \ ' Qf
'h ,,, f6
'@ U
# ' '^I i ,y N '
k
-% ' .i , '
- 'I' ! , i #
+ h l \ y r= -
t >
< L.a '>,'
i
.1. E ' s I .
. ~~
1 % .
# p,at LE C vg
. N. W N . f.. ' . =rt'n. - =
? ? ^? , V . '? . $? GENERAL TOPOGRAPHY Figure 3.2. General Topography: Surry Power Station 1
! 3 - l.
H G ISLb. D ST TE
)h.,...
n .
/s .
1 .
% s ., g s .
f ,
,G A ENT AREA
, j W ILDLIF E*.)1 p
w ~c
/ '*. [.ar
<' CP -
i
,,s \
\
,r t . .
',,.e 1
.j zr ,
**,\..,,,.
' .e,*. ,
2, e p
,!s
,'I' M i &l /g 1 m ee > ~
)! sfits W k, ( ,E N
q.
~
p
$ VISITOR.-::: . '
' CENTER.Q ,: :,lt 4 S.,
N: ; f - D R. h{t I ; [$
.a .8,RIDG E, UN ;.,
F ,:# , t D , C [ c$ &
% .. *; POWER ;
,'s, f, ,e :P NT f, . g ,- , y C'
- . g s 2 . .- ,
- y )
..% w
^
' ' *+ WELL
/ f"*.,/
o /k
,I l
,,j '.J-[
> z0 I
~ . 4 ;
J'#, e
, t
# /
IdWi 'N6Mg /' . *I*,', flPEUNES g . . m n
%m
- _ 4 t- ,,,, :-
t,. t q. .... a G . h se*,c,RO4e - i",4R
' , . l
. ASR8IN g '
gSWITCHYAg.y.,y, T: ,. '..g ES . . j'f.-Q. g ., iis&ddi@
/ ? , / 5"C'"."! -;
.- . \s-"%i..
1 '?<,,'l*,. . av . ' " , , . . . . . . . . . wu f.- -
..%s_ .
a.t , Ms
- l, l .n
,,,,,..Myqug.p% w 4 ...
uu"
'" ,* t.
e
- ,h. , - ACCESS
- 4 s,
..,.*\.* ,.g W
- e. '. . q s.- .p $ : '. ' CHANNEL
,, I:. -- . , ,, , . ,
M - M r e,.. e ..; 4..,.1,g>'(> u..... r **** 0 .:... %, 4i;@1, o /i,...{ :piQ ,..gj 'C.,?,-! f TRANSMISSION r 1 ,'
?, , CONCRETE
..S, , 6, '
BATCH f RIGHTS OF WAY.'- D *' M\%h PLANT
>- k , 'd- M'- ',
k >.qvk N . l\\Y&,6, %, ig /g&.y &fR,.. .;i # W,I g* ~ li.L. )-
""U. Y . N h I' .
s@ N- '
/ -llN5 , '
wit.0Lirt . MANgf,EMENT k'REA N!. ' a.
.:f)le)i N
( .
.:q-
- 1+'& N
- A , ' p' /
,,-% 'I!P
,:%g/. '. .
yK [. . / .'Q
.. l .; - ~.
t - }s }ll .;, . . . -
'$ :tn 7.h 'Y'
,.3
- i nseu m m, Figure 3.3. I.ocal topography: Surry Power Station 3-5
t E
.^
!i 3 l., <. ..
j/'~'Nn5$+.w ,'. . . *
.k, l ,, ~,s
. \ '.,, ,
/+
j s .- _ 'g ----
&/$ %
,? l/
<' : ; #e.sa.,A; .
c'e
'y 3fij },_ r -
,~
' ?j
, /*
j f
- 7. u.
. m y
'x ,..
.x y.
j .',
. qdnr e
/ ' '
'-b' ,
5(+ /"v,., ~ ' ,' g yc[m\,'.' 1
, =
[ k
.h.a- x,b,
't v, s
A > p1
,, ,J
'\ f e
a w.. /, n E g
*- \ .-
j i 1: /. ,,; s,
\.
4
\ .
- k. -
,J ,_
~
,' q , e i
\
., .9 h . lI ' '
l
,, '\ -
1 s i I e/ e i
..g4 c1a5 \ \]i j.
,&-$< L,i ,
,\j . , _ .
..\ l
(. 3-6
t p 4 g s t q q4m gyp 4 . [JQg;.fC 3
.~$f
./ .
#ge.<*
! \
i 1 a ,4 ,... cx
~- - [! -
g
>A, t
m
~
.,.'a .,.
r
/ +,
,/
s y3 e a 3,
'g.
, a h *
.! b k <
-fu .a / f/ s.
s jiz 8 y n , s 2:(). . .\ , ,, ...- -
,e y jij '
- k. , M c:
}
b (f
/
) T'
/ g# g\y l
\ l M($gg' frA.
v -
, j si sti s , 1 2
it) ,
\
* ,Az x , e s%$
f 3 i
\ a 6 \ E
), ,y i
.' q
.~..
e s% ,g ' e a 8 e ;/. I
- .... ..f9 s - s'b O, i j.
E
,\ l ' $1 h'
1 x3J c If ll : ,,h.\ '
\
*h I
L [ _ . . . , . 37
l l 3.1.2 plant
-The twin PWR units (Surry 1 and Surry 2) belonging to the Virginia Power Company are each rated at 781 MW. The reactor and generator for both the units were supplied by Westinghouse Electric Corporation. The plant began commercial operation in 1972-73. Stone and Webster Engineering Corporation was the Architect / Engineer / Constructor for these plants.
The reactor containment structure is a steel-lined, reinforced concrete unit with vertical cylindrical walls and a hemispherical dome. The sup-porting flat base of the foundation mats is approximately 66 ft below finished ground grade. The contairunent structure below grade is con-structed inside a cofferdam. Dimensions for each of these units are as follows:
- a. Inside diameter 126 ft-0 in,
- b. Springline of dome above the top of 122 ft 1 in, foundation mat
- c. Thickness of mat 10 ft 0 in,
- d. Thickness of dome 2 ft-6 in,
- c. Thickness of cylindrical walls 4 ft-6 in, f, Thickness of steel liner:
(i) base mat 0.25 in. .75 in. l (ii) hemisphere 0.5 in. (iii) cylindrical wall 0.375 in. Access to the containment stracture for personnel and equipment is pro-vided by two hatch penetrations with internal diameters of 7 f t-0 in and 14 f t 0 in, respectively. Besides these, there are several smaller penetrations for pipes and conduits. Other Class - I structures (i.e., except the reactor containment) are the auxiliary building; control room area, including switchgear and relay rooms;-fuel building;-auxiliary generator cubicles; auxiliary containment buildings that contain main steam and feedwater isolation valves; recir-culation spray and low _ head safety injection pump cubicles; safeguards ventilation room and circulating water intake structures, including . the high level canal. All-these structures were designed to meet both carth-quake and tornado design criteria. 3.1.3 Site Visit The screening analysis began with a site visit conducted in April 1987, ' The purpose of the site visit was twofold: first, to confirm the infor-mation in the FSAR which was used in the Surry scoping quantification l l 3-8 t l
. - . _ . _ . _ - - . . . ~ . ~ , . - . ,- . - , - - - - . . , - . - _ . .-,-,,.c . . ,~ . . , - - -
study , _ and _- second to _ collect new information and look for possible changes in the plant and site conditions which could affect the risk from external hazards to the site. The site visit included a tour of the plant structures as well as a survey of the plant boundary . and surrounding areas. Following is a highlight of the issues which were resolved by the site visit:
- a. No major changes or deviations from the information in the Surry FASR (which could affect the external event screening) were observed in the plant or its surroundings,
- b. A survey of the structures in Surry revealed that all the doors which open to the outside of the plant are above the plant grade which is considerably higher than the probabic maximum hurricane-induced flood level. The circulating water intake structure and emergency service water pumphouse have doors and air intake louver openings at levels below the probable maximum surge level. However, the doors are Icaktight and the air intake is not used in the event of a probable maximum surge,
- c. During the site visit, a survey of the objects in the plant boundary which could potentially become tornado generated miasiles was carried out. The' site visit confirmed that the potential number of missiles at the Surry site is less than the number used in the tornado missile simulation study (Ref. 3) utilized in the bounding analysis study discussed in Section 3.4.2.
- d. The site viait confirmed that there are no new industries, major air-ports,-pipelines, or major highways in the vicinity of the site that are not described in the Surry FSAR.
3.2 Initial Screenine of External Events An extensive review of. information on the site region and plant design was made to identify all external events to be considered, The data in the Surry FSAR as well as other data obtained from the utility, and the information gathered in the site visit were reviewed for this purpose. A set of screening criteria was utilized to identify those external hazards which could be screened from further consideration based on very general considerations, as described in Section 1.3.2. These criteria, based on those in the PRA Procedures Guide (Ref. 1), are listed again below: An external event can be excluded from further consideration if: Criterion 1 The event is of equal or lesser damage potential than the events for.which the plant has been designed. This requires an evaluacion of plant design bases in order to estimate the resistance of plant structures and systems to a particular external event. 3-9
Criterion 7 The event has a significantly lower mean frequency of occurrence than other events with similar uncertainties and could not result in worse consequences than those events. Cr_iterion 3 The event cannot occur close enough to the plant to affect it. This is also a function of the magnitude of the event. Criterion 4 The event is included in the definition of another ' event. Criterion 5 The event is slow in developing and there is sufficient time to climinate the source of the treat or to provide an adequate response, The use of these criteria minimizes the possibility of omitting any significant risk contributors while at the same time reducing the amount . of detailed bounding analysis required. Table 3.1 is a listing of external hazards for the Surry Station based on the augmentation of Table 10 1 of the PRA Procedures Guide (Ref. 1). For each external event, the applicable screening criteria and a brief description of the basis for the screening (if any) is included in the table. In summary, the findings of the preliminary screening are that, aside from seismic -and fire events which have already been included in the detailed. external hazards analyses, the following events were identified as requiring further bounding study,
- a. Aircraft Impact s
- b. External Flooding
- c. Extreme Winds and Tornadoes
- d. Military and Industrial Facilities Accidents
- e. Pipeline Accidents
- f. Release of Chemicals in On-site Storage
- g. Transportation Accidents
- h. Turbine Missiles
- 1. Internal Flooding The bounding analyses performed for these events are discussed in Sections 3.3 and 3.4.
3-10 --
l l Table 3.1 Preliminary Screening of External Events for Surry Nuclear Power Station Applicable Screening Events Criteria _ Remarks Aircraft Impact -- A bounding analysis is performed for this event. Avalanche 3 Topography is such that no avalanche is possible. Biological Events 1 The only biological event which may affect the safety of the plant is fish in the river, i.e., fish may block flow of water in the intake structure. This event is not further considered because there would be adequate warning, and the re fo re , remedial action can be taken before supply of the intake canal is exhausted. Coastal Erosion 3 The site is located on the banks of the James River on three sides. The area is covered by marshy wetlands and swamps. Therefore, erosion is
- not a significant possibility, i
i Drought 1 The stretch of the river between l Richmond and the mouth of the river is essentially a tidal estuary. There are no known or planned river control structures and the possibility of water shortage is unlikely. -However, under certain circumstances, vir', from the northeast could cause .ionormally low river levels at the oite for up to 24 hours. Iloweve r , the design of l the plant can accommodate this event. The high level intake canal contains a minimum of 45 million gallons of water for use in recirculation spray heat exchangers during a LOCA incident in one unit combined with loss of power in both units. This storage volume can be used up to 100 hours to maintain the station in a safe shutdown condition. 3-11
Table 3.1 (Cont'd) Preliminary Screening of External Events for Surry Nuclear Power Station Applicable Screening Events Criteria Retrarks External Flooding - A bounding analysis is performed for this event. Extreme Winds and -- A bounding analysis is performed for this event. Tornadoes Fog 4 Fog can affect the frequency of occurrence of other hazards such as highway accidents or aircraft landing and take-off accidents. The effects of fog on highway, railway, or barge accidents are implicitly taken into account by assuming a worst possible transportation accident near the site. Transportation accidents are considered in detail for the present study. Forest Fire 3 Site itself is cleared, while scrub pine exists beyond site boundary. Fires cannot directly affect the plant. Fire suppression systems at Surry not automatically activated, so no chance of incidental actuations. Prost 1 Loads induced on structures due to frost are much lower than snow and ice loads, i.e., frost loads can be safely neglected in the plant hazard analysis. Hall 1 liail is less damaging than other missiles which are generated outside of the plant such as tornado missiles and turbine missiles. Therefore, hail is not considered further in the scoping study. Iligh Tide or liigh 4 Included under external flooding. River Stage 3-12
Table 3.1 (Cont'd) Preliminary Screening of External Events for Surry Nuclear Power Station Applicable Screening Events Criteria
- Remarks liigh Summer 1 As mentioned under drought, it is Temperature possible to safely shut down the plant due to unavailability of water. Therefore, high temperatures.
on record were indirectly included under drought conditions.
.llurricano 1 The effects are included under flooding and tornado effects.
Ice Cover 1 Ice or snow loading is considered in the-plant design. Ice blockage-of the. river is included in flood. Industrial or - A bounding analysis is performed for Military Facility this event. l Accident-Internal Floodin6 - A boundin5 analysis is performed for this event. Landslide 3 The Surry plant .is built on flat land where landslides are not possible. Low' Lake or River 4 This event is considered under Water Level drought.
~ Thermal' stresses and embrittlements.
Low Winter 1,4 . Temperature are insignificant and are covered by' design codes and standards for plant de s ign . .. Generally, there is ade-quate warnin6 of icing on the ulti-mate heat sink (i.e., river)-so that i remedial action could be taken. 3 13 l
~ .- . _ - . .-. - - - ..-. - - - . . . - . . - _ _ .. . - .. _ . .._ .. - - - - _
1 1 I i Tablo 3.1-(Cont'd) Preliminary Screcoing of External Events for Surry Nuclear Power Station Applicable Screening _ Events Criteria Remarks Meteorite 2 This- event has a very low probability of occurrence. A study by Solomon et al. (Ref. 4) showed that the probability of a meteorite impacting any nuclear power plant in the U.S. is negligible, and therefore, meteorites need not be considered in this study. Pipeline Accident - A bounding analysis is performed for this event. Intense Precipitation 4 Included under internal and external flooding. Release of Chemicals - A bounding analysis is done for this ! in On site' Storage event. River Diversion ~ 3 This event is not credible for the site under consideration. Sandstorm 3 This is not relevant for this re-Bi on. Seibhe 4 Included under external flooding. Snow 1 Plant is designed for snow load, ponding effects, and combinations of snow with other loads. l Soil Shrink Swell 1 Plant structures are all designed Consolidation for the effects of consolidation. Such - offects occur over a long period and they do not pose a. hazard during plant operation, i.e.e, the-plant can be safely shut down if needed. Storm Surge 4 Included under external flooding. Transportation - A bounding analysis is done for this Accidents event, 3 14 L
l Table 3.1 (Concludtsd) Preliminary Screening of External Events for Surry Nuclear Power Station Applicable Screening Events Criteria _ Remarks Tsunami 3 Tsunamis are rare on the East Coast. Plant location is inland from sea coast, Toxic Gas 4 Included in transportation accident, on site chemical release, and indus-try and military facilities acci-dents. ' Turbine-Generated -- A bounding analysis is performed for Missiles this event.
^
Volcanic Activity 3 The site is not close to any active volcanos. Waves 4 This . event included under external 1 flooding. 3-15
l 1 1 l 3,3 Screening of External Events Based on FSAR and Site Hazard Studies , l This section describes the external events which could be screened out 1 based on the updated FSAR information supplemented with new data. Section 3.3.1 discusses the military and industrial facilities accidents, Section 3.3.2 deals with the transportation accidents and Section 3.3.3 covers on site chemical release. It is concluded that these events can be screened out, 3.3.1 Accidents in Industrial and Military Facilities According to the Surry FSAR, the areas to the north and south of the site, except for the Williamsburg area, are principally rural and agri-cultural; The nearest industrial facility is located 4-1/2 miles from the site, and this is the only industrial facility within a five mile radius. Table 3,2, which is duplicated from an NUS Corporation study on toxic chemicals at the Surry site (Ref. 5), lists all the chemical compounds used by, and/or stored, at this facility There ' are three possibic effects from an industrial accident near the site: (1) incident over pressure on plant structures due to an explosion, (2) seepage of toxic- chemicals into the control room, which could incapacitate the operators, and (3) flammable vapor clouds leading to a heat-hazard-at the site. Industrial accidents at distances farther than 5 miles to the site are not expected to cause significant over pressure loads on the plant structures, For example, of all the chemicals stored at the industrial facility (Table 3.2), only acrylonitrile and methyl
. acrylate are explosive, Assuming an explosion of the entire quantity of these chemicals, the peak over pressure experienced on wall panels at the site _ would be less than 1 psi, As the Surry plant Category 1 structures are ' designed for tornado wind loads, with a minimum capacity of 3 psi against blast loads.,an over-pressure hazard due to industrial accidents can be screened out.
-Release o f _--toxic chemicals near nuclear power plants -can potentially resuit in the control room being uninhabitable. This condition can happen if (1) large quantities of toxic chemicals are released, (2) there are favorabic wind conditions and insufficient dilution of chemicals such that these chemicals reach the control room air incakes, and (3) there l= are no detection systems and air isolation systems in the control room.
According to Regulatory Guide 1,78 (Ref, 6), chemicals stored or situated at- distances- greater that 5 miles need not be-considered as an external hazard. This is due to the fact that if a release occurs at such a distance, atmospheric dispersion will _ dilute and disperse the incoming f_ plume to such a degree that there should be sufficient time for the
-control room operators to take appropriate action. As the amount of stored chemicals is small and at n' distance of nearly 5 miles from the site, the accidents in -the only industrial facility near the plant do not pose an unacceptable risk. This same conclusion was reached in the NUS Corporation study (Ref, 5)- ,
3-16
___;_ _ m _ _ - _ . _ -_ . . .. __ Table 3.2 Chemical Compounds Used and/or Stored Near Surry Type Dis';nce Container Quantity per Unit Container Miles Berm Cbemical Size 1 Metal Tank 4.9 50'x30'x4.5' Acrylonitrile 50,000 gal (5,000 gal) 4 ea (30'x15'x4.5') u 1 Metal Tank 4.9 30*x20'x5.5' i Methyl Acrylate 25,000 gal
" (5,000 gal) 1 (30*x15'x4.5')
Me 1 Tank 4.9 40'x20'x2' Sulfuric Acid 5,000 gal 3 ea 1etal Tank 4.9 40'x20'x2' Hydrochloric 5,000 gal 3 ea Acid l
. ~.- -- . -.- - . . . _~. ~ . - - =... - - - - . - ~ - ~
3.3.2 Transportation Accidents The plant is located on the - banks of the J mes River, which is a navigable river used for transportation of bult goods. The type of chemicals and their quantities are shown in Tabic 3.3. Virginia flighway 10 is the only major surface route near the plant besides the state secondary access Route 650 to the site. The access road ends at the llog Island Waterfowl Refuge, north of the site. Small amounts of chemicals required in plant operations are transported along the access road and these hazards are considered under on site chemicals in Section 3.3.3. The chemicals transported on Virginia 10 are given in Table 3.4 (from Reference 5). There is no rail traffic within a five mile radius of the station and the risk from the air transport mode is considered separately in Section 3.4.5. A transport accident near the site can pose risk in one of the following ways: (1) a chemical explosion due to a transportation accident may cause damage to Category I structures and safety related equipment, and (2) toxic chemicals which are spilled in a transportation accident may drift into the control room and cause incapacitation of the operators, A chemical explosion near the plant structures may cause over pressure, dy namic pressures, blast induced ground motion, or blast generated missiles, lloweve r , from previous research in this area, it has been determined that over pressures would be the controlling consideration for explosions resulting from transportation accidents (Regulatory Guide 1.91, Ref 7). An accident over-pressure at the site can also occur due to vapor cloud explosions drifting towards the structures. This type of explosion involves complex phenomena which depend on the material involved, combustion process, and topographical and meteorological conditions. According to a study by Eichler and Napadensky (Ref. 8), present theoretical -and empirical knowledge is too limited to quantitatively evaluate realistic accidental vapor cloud explosion scenarios. However, vapor cloud explosions are implicitly included in the-TNT equivalents whicP are used to represent transportation accidents. According to the Regriatory Guide 1.91 (Ref. 7), chemical explosions which would result in free field over pressures of less than 1 psi at the site do not need to be considered in the plant design. Based on experimental data on hemispherical charges- of TNT, a 1 psi pressure would be translated into a safe-distance R (ft) which is defined as: l .R > kwu3 t
;where k and w is an equivalent weight of TNT charge.
According to Table 3.4, the maximum-possible explosive charge is due to 8,500 gallons of gasoline, which is an (approximate) equivalent of 50,000 l lbs. of TNT charge. Using the relation given above, the distance for a l' pressure pulse less than 1 psi is ' calculated to be 1,658 f t. Based on this result, it is concluded that explosions on Virginia 10 highway will not pose an over pressure hazard to the plant -structures. 3-18
r Table 3.3 Chemical Compounds Shipped on the James River Container Quantity Type Distance- -l Chemical Size per Unit Container Miles Diaminocyclo Nexane 55 gal / barrels 4,400 to Closed Van 1 1/2 Corrasive Liquid .80'to 140 7,700 gal Ocean Vessel Ethanol / Inflammable 55 gal / barrels 4,400 to Closed Van 1 1/2 Liquid 80 to 140 7,700. gal Ocean vessel .} Tiazinetrione Dry 50 lb bags 40,000 to closed van 1 1/2 Oxidizer Pelletized 60,000 lb Ocean Vessel u 1 Napthyl Methyl 50 lb bags 40,000 to Closed Van 1 1/2 o Carbonate - Poison Pelletized 60,000 lb Ocean vessel Ethyl Alcohol 55 gal / barrels 4,400 to Closed Van _ 1 1/2 Flammable-Liquid 80 to 140 7,700 gal _ Ocean Vessel Sodium Meta 50 lb bags 40,000 to Closed van 1 1/2 Periodate - Oxidizer Pelletized 60,000 lb Ocean vessel Nitro Imidayol 50 lb bags 40,000 to Closed van 1 1/2 Poison - Solid Pelletized 60,000 lb' Ocean Vessel Ethyacloxysilane 55 gal / barrels 4,400 to closed van 1 1/2 Corrosive Liquid 80 to 140 7,000 gal Ocean vessel Dinitrochloro 50 lb bags 40,000 to Closed Van 1 1/2 Benzene - Poison Pelletized 60,000 lb Ocean Vessel
.j l
Table 3.3 Chemical Compounds Shipped on the James River (Continued) Container Quantity Type Distance Chemical Size per Unit Container Miles Monochloracetic Acid 50 lb bags _40,000 to closed Van 1 1/2-Corrosive Pelletized 60,000 lb Ocean Vessel 2-Methox 4-2-3 Dyhydro' 55 gal / barrels _4,400 to - Closed Van 1 1/2 . 4-H Inflammable Liquid '80 to 140 7,700 gal Ocean Vessel [ Ortho-Phenylenediamine 50 lb bags 40,000 to closed van 1 1/2 i Poison . Pelletized 60.000 lb Ocean vessel u Chloro Benzo Tri Fluoride 55 gal / barrels 4,400 to Closea va.' 1 1/2 l, Inflammable Liquid 80 to 140 7,700 gal Ocean Vessel - o Caustic" Alkali 55 gal / barrels 4,400 to Closed Van 1 1/2 [ Liquid Corrosive 80 to 140 7.700 gal Ocean vessel l Thionyl Chloride 55 gal / barrels 4,400 to closed van 1 1/2 ; Corrosive 80 to 140 7,700 gal Ocean vessel Gasoline,-*:6 Oil, Steel Tanks 168,000 gal ea Barge 1 1/2 Diesel Oil, p2 Oil 8 Compartments 1,300,000 total , Phenol Steel Tanks 1,325 tons en Barge 1 1/2 2 Compartments 2,650 total i Oleum Steel Tanks 1,500 tons ea Barge 1 1/2 2 Compartments 3,000 total 4
i. i .i ,
.j 4
i ! I i i Table 3.3 i = Chemical' Compounds. Shipped on the James River'(Concluded) i Container Quantity- Type Distance Chemical' Size per Unit Container Miles Sulfur (Liquid ' Steel Tanks' 10,000 tons es Barge 1 1/2 ; at 260*F to 275*F) 2 Compartments 20,000 total Liquid Fertilizer . Steel Tanks 5,000 tons ea Barge 1 1/2 y (Uran) 2 Compartments 10,000 total Ammonium Sulfate- 50 lb bags ,1,500 to Barge 1.1/2 i Pelletized 12,000 tons } Ammonium Sulfate "50 lb bags 8,000 to closed van 1 1/2 4 Pelletized 25,000 tons Ocean Vessel
.p j
Y 15 5 4 0
. . . ~ . - -- -
a Table 3.4' ; Chemical Compounds Transported by Truck on Virginia' Highway 10 Container ' Quantity. Type Distance-Chem'. cal Size per Unit Container Miles , Sulfurit Acid 25 ton truck tank. 3,300 gal Metal Tank 4 1/2 u Nitric Acid 25 ton truck tank 4,000 gal Metal Tank 4 1/2 , i N w Muratic Acid 25 ton truck tank 5,000 gal Metal Tank 4 1/2 Petroleum 25 ton truck tank 8,500 gal Metal Tank 4 1/2 Gasoline, Oil t i i L t
i 3 I
.. \
f Assuming a typical maximtun probable equivalent TNT charge of 1 x 107 lbs. j l for any of the chemicals t ransported on a river barge and the distance of 1 . the barge from the nearest plant structure to be 1.5 miles , an over- ' ' pressure of around 1 psi vill be experienced. This is well within the ! design limit of 3 psi, postulated for tornado designed st ructures. i Flam:nable vapor clouds also do not present any explosive hazard. Ac-
- cording to a study by richler, Napadensky and Mavec (Ref. 9), the accidents in an empty barge due to vaporization of liquid left in 'he t ank woui,d lead to a maximata TNT equivalent explosive load of 1000 lbs Since this type of accide,1t do u not produce a more severe condirlon, it
- is not considered further.
1 h A toxic chemical spill neat the site would pose a danger to the plant if toxic chemicals penetrate into the control rocm through air intakes. This can happen if (1) large quantities of toxic chemicals are released, (2) there are favornble wind condit. ions which would cause a drift of chemicals towards the control room air intakes av excessive concentra-tions, and (3) *here are no detection systems and air isolation systems in the control room. l
- j. Among the various transportation modes near the site, a barge accident in "
the James River would result in the largest. armount of cho,nical spill. The NUS Corporation study (Ref. 5) also estiinated the danger from toxic chemicals spilled in an of C site transportation accident-. According to this report, from the quantities, distances and properties of the chemicals, the toxicity limit and the estimated cloud center i concentration at the control room air intake of most chemicals were not cause for concern, only concentrations of gasoline exceeded the toxicity ! limit. It was estimated that the control room personnel would have 2,390 i seconds (40 min.) of warning if notified iminediately of the accident, i This *.ime includes the tieto required for the vapor cloud to drift to the air intake and then to build up to the toxicity limit in the control L.. room. The amount of warning time available without knowledge of the l accident is 192 seconds, if detectors are placed at the air intake.
- In respunse to NRC review of this study, VEPCO agrood to modif1 cations to
- assure control room habitability. With these modifications, the risk to control room personnel due _t o a transportation accident will be negligible.
3.3.3 Release of On site Chemleals
- The chemicals stored on site at tha Surry plant are listed in Tau,e 3,5 L and their storage locations are shown in Table 3,6. The NUS Corg uration study (Ref. 5) analyzed the consequence of release of a single _ eonb iner of these chemicals, its distersion and subs ,uant build up in the control room air. The amounts of each chemical . alyzed for spill and their toxicity limits are listed in Table 3.5. The results in terms of peak
- concentration of chemicals in the control room are given in fable 3.7.
This table shews that most of these- chemicals (morpholine, acetone, 3-23 l
. . . . . _ . . - - ~ - - . . - . _ . _ . _ . . .
l l Table 3.5 , l Surry On Site Chemical Spill Analysis i 1 4 I Toxicity Quantity Limit Chemical Assumed Sullied (m g/m3)_ }. Morpholine 55 gal 105 Acetonc $5 gal 4,800 Cyclohexylamine 55 gal 40 lI Sulfuric Acid 8,000 gal 2 Aminoniurn Hydroxide 3,000 gal 70 , I carbon Dioxide l' tons 18,000 Diesel Fuel 210,000 gal 1,355 Chlorine 64 lb 45 4 Ilydrazine . 55 gal 0.3 Dimethylamine '135 lb 28-
)
cyclohexylatnine , sulfuric acid, aminonium hydroxide, and - diesel fuel) present no hazard to controi room personnel. _ The _ peak concentration in the control roorn exceeds the - toxicity limits due to release of dimethylamine, carbon dioxide, chlorine and hydrazine. Tht- time required to reach the limits are also indicated. Time t2 gives the warning time if detectors are present at the chemical storage location whereas ta represents the warning time available for detectors at the air intake. Pacific Northwest Laboratories reviewed the - NUS Corporation report on control room habitability (Ref. 10) for the NRC. - VEPCO agreed to certain modifications listed in USNRC letter of June 28, 1982 (Ref. 11). These inodifications will provide safe, habitable conditions 'within control room under both normal and accidental toxic gas conditions and the risk from these hazards can be expected to be negligible. I 3.4 Bounding Analyses The bounding analyses for the external events which could not be screened out by the Seneral criteria as described above are given in this section. I~ i 3 24 l i G-.-...-. -. . . . w_._._--.,...____-_-__._..__ . . . _ . . , _ . , _ . . . _ . , _ . , _ . . . . . . . , _ . , , _ _ , , . . , _,, . , . , , , , _ ,
, _ _ _ . . . . _ . . _ . _ . . _ _ . . _ _ _ _ . . _ . _ . _ _ . . . . . _ _ ___._.m_._.m..._.___ 1 2 , t . Table 3.6 1 j Surry 1 and 2 Toxic Chemical Source locations j Distance From Air Intake I _ _: ChtmlS.a1 ._ {fti Loem ig.p j
- Dimethylacine, Argon, Helium 125 Outside NW of Intake East
- l. Ilydro6en, Nitrogen, Oxygen, of Security building 3- Carbon Dioxide, Acetyltne, J 3
Breathing Air, Specialty Cas ! Mixes ) 4 1
!!orpholine Anhydrous 190 Outside NNW of Intake East liydrazine Acetone, Sodium of Security Building i liypochloride, Cyclohexylamine ;
llydrogen Bank 276 _Outside V of Intake, SW of ' Condensato Storage Tanks
- Sulfuric Acid 410 Room Within Condensate Pol. '
ishing Building, Berm With. , in Room, 2 (Self Closing) . Doors - Between Emergency Intake. 567 ft. From Con-densate Polishing Building liVAC Exhaust Stack 'to Nor-mal Intake Arnmoniwn llydroxide 426 Room Within Condensate Pol. ishing Building, 2 (Self. Closing) Doors Between Emergency Intake. 620 ft. From Ammonium Room Exhaust Stack to Normal Intake flydrazine 374 Condensate Polishing Build-ing. 1 (Self Closing) Door Between Emergency Intake i ~ Carbon Dioxide 157 Outside Adjacent to Double Doors South Side of Turbine Building. Sulfuric Acid 131 Inside Turbine Building ' Across From Emergency Intake I l l 3 25
Tabic 3.6 Surry 1 and 2 Toxic Chemical Source locations- (concluded) o 1 Distance From i Air Intake Chemical (ft) location Diesel Fuel 400 Outside Separate Tank 60'x 60'x 9' Dike Chlorine 472 Inside Sowerage Treatment Building Off Plot f Ilydrazine 1,476 Inside Warehouse Building - Ammonium Hydroxide Off Plot ! -) The probabilistic models used in these bounding analyses integrate the / . randomness and uncertair,ty associated with loads, response analysis, and capacities to predict the annual frequency of the plant damage from conservative models. If the mean frequency computed with a conservative '1 model is predicted t.o be suf ficiently low (e.g. , less than 10-5/ year), the external event nay be eliminated from further consideration, The bounding , analysca thus provides a second screening of t.be external hazards, allowing additional hazards t_o be deleted from further consideration,- and identifying those remaining external events which need to be~ analyzed in
' detail as part of the PRA,
- In addition to calculating and scrocuing on a best estimate frequency.. of ;
4 core damage, the uncertainties in hazard and component fragilities may be ! l used to find the high confidence (95 percent) _ bounds on; the frequency of , core damage, llowever, such an uncertainty analysis is required only if the < - best estimate of the core damage frequency of the external event leads to a value which is close to the (usual) mean rejection frequency of 10'8/ year, Often, simplifications in the above analyses are introduced. As an example, in case of_ aircraft impact, back face (inside) scabbing of the exterior barrier walls of safety related structures can be assumed to ' resul t- in core damage even though, actually, . a suitable combination of component failures is necessary to Icad to this- damage state, llowever , if-the resulting frequency of core damage computed with the conservative model is suf ficiently small,- no furt.ber consideration in required, L In_ addition, for some external events, it is possible to perform a bounding analysis without performing a structural response analysis. In effect, one shows that the frequency of exceeding the _ design loads is very small, and _ thus, infers that the har.ard can be neglected due to the conservatism in 3-26 i
,m.,nn ,
_._..___.m___._____ i Table 3.7 Peak Concentration of Chemicals in Control Room Chemien1 TL C, .- Lt ta Morpholine 105 9. 2x10-1 *
- Acetone 4,800 2.7x101 *
- Cyclohexylamine 40 1.1 *
- Sulfuric Acid 2 4.3x103 *
- llydrazine 0.3- 2.1x101 946 36 ,
Diesel Fuel 1,355 5.2x101 *
- Ammonium flydroxide 70 3.8 *
- Carbon Dioxide 1.8x104 3.9x10' 159 01 Carbon Dioxide 1.8x104 2. 2x104 ( E) 180(E) 82(E)
Chlorino 45 8.9x102 280 17 Dimethylamine 28 6.5x10) 68 7 TL - Toxicity Limit (mg/m3) . Og - Peak concentration in centrol room (eng/m3), t2- Time from spil1~ until TL is reached in control room air (seconds),
- indicates TL not reached, t2- Time from reaching TL at intake to reaching TL.in control room.
E - Emergency air intake. 3 27 T,, e___ - r . _ n n __m_,_ ._, , , _ _ _ , , _ _ _ __ , _ _ _ _ , _, _ _ , _ _ _ _ - . _,e__ _- t
___ w a swta w s w w. T-E d These, and other simplifications are utilized as the design process. appropriate in the following bounding analyses. 3.4.1 Extreme Vinds and Tornadoes Extreme winds from tornadoes, hurricanes or wind storms present a likely threat to the nuclear power plants due to (a) direct damages from the dynamic wind loadings, (b) rnissiles generated and, (c) pressure differentials. The winds associated with hurricanes and storma are usually less intense and lower in magnitude than t.hos e associated with tornadoes. Hence, it is sufficient to to consider risk to the structures due to tornadoes. This section describes the analysis of Surry $ structures for the effects of tornadoes. Regulatory Guide 1.117 (Ref.14) specifies the plant systerns, structures, components, areas, etc., to be protected against tornadoes. Both seismic category I structures and non category I structures were considered for this task. Seismic category I structures have been designed for extrerne winds, scismic, and tornado loadings. Non category I structures were L generally designed against wind loads. L 3.4.1.1 Plant Design Criteria for Category I Structures The category I structures of Surry were designed to withstand a Design Basis Tornado (DBT) which is defined as follows: Rotational velocity 300 rnph Transintion velocity 60 mph Pressure drop 3 psi in 3 sec Overall diameter 1200 ft Radius of maximum winds 200 ft As per the FSAR, the structures can resist a rnaximum wind velocity associated with a tornado of 360 mph; and were also checked for tornado pressure loacing, pressure drop and combinations of the two. For the purpose of structural analysis, dynamic wit.d pressures on the structures were converted into equivalent static forces which vary along t.he height of each structure. Since the natural periods of buildings at Surry are short compared with the rise in time of applied design pressures, the above assumption is well justified. The safety related rtructures were also designed for the effects of postulated tornade missiles. The postulated tornado missiles used in the design of category I structures were as follows:
- a. Vooden pole 40 ft long, 12 in, diameter, weighing 50 lbs/ft3 and traveling in a vertical or horizontal direction at 150 reph,
- b. 1 ton autornobile traveling at 150 mph.
The FSAR gives details regarding different structures and sys t ern s designed for tornado loadings. (Table 3.8) i I i 3-28
1 i i i Table 3.8 I
, Structures and Components Designed for Seismic and Tornado Criteria (i i
Earthquake Tornado Typical Thickness
' Item Criterion Criterion Note of Concrete I I
i 1 Structures ; i ! ]' Reactor Containment. NA t t Reinforced-concrete substructure- I P 54" (Cylinder '.lalls) i Reinforced-concrete superstructure I T 30"
- . Reinforced-concrete interior
- shields and walls I NA NA t
- Steel plate liner I P P for containment NA j integrity, l y Piping.. duct, and electrical. I T T for shield wall 14" .
w- penetrations and shield wall and critical i j' system penetra-tions only
- Personnel access hatch I P NA }
, Equipment access hatch I P NA j i [ f Cable Vault and Cable Tunnel I T 24- ! I j Pipe Tunnel to Containment from
. Auxiliary Building- I T 24*
1
; . t Auxiliary Steam-Cenerator Feed Purep i Cubicle I- T 36* 5 i
l 1 I i i< ;
< )
I f I i
- b. - .
i f .. i-i i Table 3.8 i Structures and Components Designed for Seismic and Tornado Criteria (Continued) 1 Earthquake Tornado Typical Thickness ! Item Criterion Criterion Note of Concrete f' Structures 36" - Cubicle for Main Stea:n and I T i Feedwater Isolation Valves ! Recirculation Spray and Low-Head Safety Injection Pump Cubicle i and Pipe Tunnel i u Safeguards Ventilation Room I NA NA l 5 o Auxiliary Building
- Reinforced-concrete Structure I T 18" to 24*
SteeI superstructure I NA NA
- Vacuum equipment. area I NA NA i
Fuel Building Reinforced-concrete structure I. T T for horizontal Drawings Steel superstructure I T missile only. Not Spent-fuel storage rack: I P T for tornado Available P for horizontal missile only 7 a 1 - i o
1 i Table 3.8 \=
-Structures and Components Designed for Seismic'and Tornado Criteria (Continued) j Earthquake.' Tornado Typical Thickness l Item Criterion Criterion Note of Concrete i
) Structures j Fuel Building (continued) ! Fuel-handling trolley support I P" T for tornado NA structure winds only Control Room I T 18" 4 T 18" to 24" y Emergency Switchgear and Relay Room I
- - w a
*~
Battery Rooms I T 12" i Air-Conditioning Equipment Rooms I T For control room 18" to 24" i and relay room only i
- Reactor Trip Breaker Cubicle I T l
1 Auxiliary Diesel-Cenerator Cubicles I Reinforced-concrete floor I T 24* ! Valls, excluding louvers' I T 24" Structural steel-supported roof I T Protected by ] i and roof slab missile rack i i 4 i k 4 .s
l l Table 3.8 Structures and Components Designed for Seis=ic and Tornado Criteria (Continued) Earthquake Tornado Typical Thickness Item Criterion Criterion Note of Concrete l Structures Turbine Building NA NA By design, building NA collapse vill not damage any Class I structures and com-ponents during earth-quake, or tornado-resistant structures w l and corponents d, 1
" during tornado.
T T for emergency 12" to 36" Circulating ater Pump Intake I ser-rice water pump Structure cubicle only T T, no missile 30" to 36" High-Level Intake Structures I protection required I i 1 no missile 18" Seal Pita protection required NA NA High-Level Intake Canal I t _ _ _ _ _
I t i Table 3.8 1
- Structures and Components Designed for Seismic and Tornado Criteria (Concluded) i t
, t ! Earthquake. Tornado Typical Thickness i ' Item Criterion Criterion Note of Concretc Structures ! 1 !' Fire-Pump House I T Engine-driven pump 24' ; only j i i Fuel-Oil Transfer Pump Vault 1 T 24* t 4 Boron Recovery Tank Dikes I T 24* V i
! '[
I - Refers to Seismic Class I criteria. All Class I components and structures are designed to resist ! 4 the operating-basis earthquake within allowab'Le working stresses. A check has been made to { determine that failure to function vill not occur with a design-basis earthquake. l i T - Refers to structures, systems, and components that will not fail during the design tornado. $ i' P - Refers ' to systems and components that vill not fail during the design tornado since they are , designed to be protected by tornado resistance structures. NA - Not applicable. , 4 I 4
; L i
t t i i
,. i 3 I a .
1 i l d According to Ravindra and Banon (Ref. 15), if the plant has been designed against tornado effects, there are no me tal sided walls or roofs in seismic category I buildings, if the reinforced concrete walls of seismic category I buildings are at least 18 in, thick, and if there are no non-redundant outdoor unprotected safety related equipment, the contribution of tornado and extreme wind induced accidents to the plant risk is jud Sed to be very low. A review ot the engineering drawings revealed that there are no metal sided walls or roofs in Seistnic Category I buildings and the j walls of these buildings are either 18 in, or more in thickness. It was also confirmed that the outdoor equipment such as the condensate storage tank and refueling water storage tank are either protected against ,
- tornado missiles or have redundant items that are protected frorn tornado effects. It is therefore concluded that the risk of damage from tornado and tornado missile impacts is negligibly small.
3.4.2 Pipeline Accidents There are two natural gas pipelines passing through the southeast end of the site. These pipelines are operated by Commonwealth Natural Gas Corporation and Colonial Pipeline Company and come from across the James River and join another pipeline with a northwest-southeast orientation (Figure 3.5). The pipelines cross the canal near the intake structure ! (Figure 3.4) and one branch of the pipeline supplies natural gas to the i combustion turbine building located south of the cooling canal. There are no automatic check valves in the vicinity of the power plant. The Sorry FSAR shows that the probability of damage - to plant structures due to a pipeline t.ccident is negligibly small. However, according to Ravindra and Banon (Ref. 15), if there are pipelines transporting natural gas, propane and other flammable explosive or toxic gases near the nuclear power plant, a scoping analysis of the hazard posed by the pipelines should be pet formed. The safety hazards posed by pipelines include thermal radiation, blast overpressure, missile generation, and plant contamination by gas at an unacceptable concentration. Among these, hazards due to thermal radiation, missile generation and plant - contamination by gas at an unacceptabic concentration are negligible. The annual frequency of failure of a large pipeline near the plant, p, is calculated as: P - N D fs fw ft fd/L where N - number of gas transmission line failures per year in the United States e L - miles of transmission pipeline in the United States D - length of pipe near site (miles) fs - fraction of failures that are large 3 34
fw - fraction of time wind will blow toward the plant from pipeline ft- fraction of failures due to construction related failures and corrosion fd - fraction of leaks going undetected The distance from the gas pipeline at the closest approach to the nearest plant structures is approximately 0.02 miles. The length D of pipe con-sidered is based on the quantity of natural gas that would produce an explosive force equivalent to 25,000 pounds of TNT, and as pc. FSAR, it is 2.6 miles. Other values for use in equation are estimated to be f t - 0.25 N/L - n of pipeline ruptures / year / mile - 1.2 x 10-' L - 200,000 miles fs - 0.329 fd - 0.10 fw - 0.5 (estimated from wind direction roses for the site.) Hence, it is found that p - 1. 2 x 10-* x 2. 6 x 0. 3 2 9 x 0. 5 x 0. 2 5 x 0.1
- 1. 2 x 10-6 The annual frequency of failure of the pipeline near the plant is, there-fore, 1.2 x 10-5 It is judged that the probability of this event leading to core damage is extremely small.
3.4.3 Turbine Missiles Failures of large steam turbines in both nuclear and fossil fueled power-plants, although rare, have occurred occasionally in the past. These failures have occurred because of one or more of the following broad classes of reasons: (1) metallurgien1 and/or design inadequacies, (2) environmental e f fects , (3) r.ut of phase or generator field failures and (4) failures of overspeed ,,rotection systems. The failures have resulted in loss of blades, d i r.x cracking, rotor and disk rupture and even missiles. Interior mirsiles are highly energetic and have the potential y to damage safety related structures housing critical components. In a total of 2,500 years of interior operation in nuclear power plants in the free world, only four failures have occurred: Calder Hall (1958), Hinkley point (1969), Shippingport (1974), and Yankee Rowe (1980). Missiles were produced in the Hinkley point and Calder Hall failures. Although the causative mechanisms of these failures have been identified and are generally corrected in the modern plants, there is no assurance 3-35
that turbine failures will not secur in the future. Recent discovery of widespread stress corrosion cracking in the disks and rotors of operating nuclear turbines has revived the industry's interest in the issue of such failures. Turbines rotate at 1800 rpm with the low-pressure (LP) and high pressure (llP) sections on a contiguous shaft. The LP sections have blade hubs (called " wheels" or " disks *) shrunk onto the rotor. Depending on the manufacturer and rated capacity of the turbine, there could be 10 to 16 disks on each LP section. The disks are massive components each weighing between 4 and 8 tons. These disks, because of their relatively large radius, are the most highly stressed spinning components in the interior. With the interior unit running at less than 120 percent of the rated speed, the disks are stressed well below the yield strength of material so that failures can be caused only by undetected material flaws that may be aggravated by stress corrosion and fatigue. At 180 percent of the rated speed, the disks are stressed at or above their ultimate strength so that they burst-into fragments. At intermediate speeds (i.e., 120 to 180 percent), rupture of disks may be caused by a combination of flaws and weaker material in the disks. Turbine missiles are spinnbg, irragular fragments with weights in the range of 100 to 8,000 pounds, and velosities in the range of 30 ft/see to ' 800 ft/sec. .It is conventional to discuss two types of turbine missile trajectories: low trajectory inissiles (LTM) and high trajectory missiles (llTM) . The low trajectory missiles are those which are ejected from the turbine casin6 at a low angle toward a barrier protecting an essential system. liigh trajectory missiles are ejected vertically (almost) upward through the interior casing and may strike critical targets by falling on theta . The customary ballistic distinction between LTM and itTM is the initial elevation angle (p) of the missile (LTM is for 4 < 45' and itTM is for 4 > 45') . Turbine manufacturers have specified that the maxituum deflection angic for the missiles produced in the burst of the last disk on the rotor is 25'. Based on this, the NRC has defined a low trajectory missile strike zone in the Regulatory Guide 1.115 (Ref. 16) and recommended that the essential systems be located outside this LTH strike zone. If a turbine missile impacts a barrier enclosing a safety related component, interest lies in knowing if the missile perforates or scabs the barrier to cause sufficient damage to the component. Using empirical formulas for scabbing derived on the basis of full scale and model tests, it is estimated that concrete barriers should be at least 4 ft thick to prevent scabbing. The need for providing such barriers depends on the probability of turbine failure and the arrangement of safety related components with respect to interior missile trajectories. In the design of a nucinar power plant, the designers have many alternative approaches for treating the potential effects of turbine failures (Sliter, Chu and Ravindra, Ref. 17). These approaches can be grouped as: (1) prevention of turbire failure, (2) prevention of reissiles, (3) prevention of strike on critical cotrponents, and (4) performance of probabilistic analysis to demonstrate that the probability of turbine missile damage is acceptably low. 3 36
3.4.3.1 Probabilistic Methodology The probability of serious damage f roin turbine inissiles to a specific system in the plant is calculated as (Bush, Ref.18): P4 -P P2 Pa where P3 - probability of turbine failure leading to inissile generation P - probability of missiles striking a barrier which encloses the safety systern given that the missile (s) have been Senerated P3 - probability of unacceptable damage to the system given that one
-or more missiles strike the barrier In practice, the evaluation of P. should include' consideration of differ- l ent speed conditions, distribution of missiles and all the safety related l components and systems in the plant.
Turbine missile damage in the older-plants was usually considered on the basis of a deterministic safety review according to RG 1.115 and SRP2.2.3 (NUREG 0800, Ref. 19), i.e., the probability of unacceptable damage froin - turbine missiles (P ) was ireplicitly_ shown to be less than 10'7 per year. The ' new guidelines concerning safety of nuclcar power plants against turbine inissile strikes are best summarized in NUREG 1068 which is a review of the Limerick PRA-(Ref. 20). The following paragraphs have been reproduced from NUREG 1068 describing the NRG position on calculating the probability of turbine inissile damage: In the past, analyses for construction permit and operating license review assumed the frequency of missile generation (P ) to be approximately 10-4 per turbine year, based on 1 the historical failure rate. The strike probability (P 2) was estimated (SRP 3.5.1.3) based on postulated inissile
- sizes, shapes, and energies , and on available plant specific information such - as turbine placement and orientation, number - and type of intervening barriers ,
target geometry, and potential missile traj ec to ri es . The damage probability (Pa) was generally assumed to be 1.0. The overall f requency of unacceptable --damage to safety-related systems (P 4 ), which is the surn over all targets of the. product of these frequencies, was then evaluated for cornpliance ' with the NRG safety objective.
- This logic places the regulatory emphasis on t% strike probability.
That is, having established an individual plant safety objective of about 10*7 per yaar, or less, for the probability of unacceptable dat. 3ge to safety related
, systema as a result of turbine mi.siles, this procedure requiras that P P3 be less than or e tual to 10 3 Although the calculation of strike probability (P 2 ) is not difficult in principle, for the most part reducing it to a straightforward ballistics analysis presents a problem in 3-37
_ ____. __ ____.m . _ . . _ _ _ _ _ . _ . . _ _ _ _ _ _ . _ _ _ _ _ _ . _ _ _ _ I 1 practice, The problem st(as frotn the fact that numerous I i modeling app ro x irna t. i ons and simplifying assumptions are required to make tractabic the incorporation into acceptable models of available data on the (1) propertles of missiles, (2) interactions of missiles with barriers and obstaclec, (3) trajectorien of missiles as they interact with or perforate (or are deficcted by) barriers, and (4) identification and location of safety related targets. The particular approximations and assumptions inade tend to have a large effect on the resulting value of Pa. Sitailarly, a reasonably accurate specification of the damage probability (P3 ) is no simple matter because of difficulty of defining the missile impact energy required to make given safety-related systems unavailable to perform their safety function, and the difficulty of postulating sequences of events that would follow a rnissile producing turbine failure. Because of the uncertainties involved in calculating P , the NRC staff concludes that P analyses are " ball park" or
" order of inagnitude" type calculations only. Based on simple estirnates for a variety of plant layouts, the NRC staff further concludna that the strike and darna ge probability product can be reasonably taken to fall in a characteristic narrow range that is dependent on the gross features of turbine generator orientation because (1) for favorably oriented turbine generators, P P 3tend to lie on the range 10 ' to 10 8, and (2) for unfavorably oriented turbine generators, P: P3 tend to lie in the range 10 3 to 10*2 For these reasons (and becau'se of weak data, controversial assumptions, and modeling dif ficulties), in the evaluation of P4 , the NRC staff gives credit for the product of the strike and darnage probabilities of 10 3 for i an unfavorably oriented turbine, and does not encou age calculations of them. In the opinion of the NRC staff, these values represent where P: P 3 lie, based on calculations done by the NRC staff and others.
l It is the view of the NRC staff that the NRC safety j objective with regard to turbine missiles is best expressed " in terms of criterion applied to the missile generation f reauency which requires the demonstrated value of turbino - missile generation frequency (P ) be less than 104 for 3 initisi startup and that corrective action be taken to return Pi to this value if -it should become greater than 109 during operation. It is the staff's view that the frequency of unacceptable I damage to safety related structures, systems and components as a result of turbine inissiles is acceptably low (i.e.,
'ess than 10 7 per year) provided that the above criterion on turbine missile generation is inet. This criterion is to be met by the maintenance of an appropriate in service inspection and testing program on the turbine throughout the plant's life as discussed in detail in the Lirnerick PRA.
3-38
i l 4 Ft t the proceding paragraphs, it is seen that the emphasis is on turbine 1 mamtenance and in service inspection to assure a value of the f requency l of turbine missile generation (P ) less than 10 5 per year. 3 i Also, if a plant has an in service inspection program which assures ! missile generation frequency of less than 104 per year, then based on a 3 miniraum P2 P3 value of 102 per year, turbine missiles can be excluded
! fren external events analysis. For plants which do not have an inspec.
1 tion program, but have a favorable turbine orientation, the argument for f excludin6 turbine missiles from further consideration is as follows.- Based on historical failure data (Ref. 18), the probability of turbine , missile generation has been calculated to be approximately 10*' per year. , Also, Patton, et al (Ref. 21) conducted a comprehensive study which , estimated- the probabilities of turbine missile generation at operating _I speed and overspeed a s . 1. 2 x10*' per year and 0.44x10** per year, , respectively. Since damage due to turbine missiles in a favorably i oriented turbine is almost entirely due to the high trajectory missiles, , i the P: P3 probability estimate of 10 8 pet year which was accepted by the NRC staff is judged to be conservative. Therefore, the frequency of i turbine missile damage in plants which have favorably oriented turbines 1 is conservatively estimated to be on the order of 107 per year, i 3.4.3.2 FSAR Analysis Westinghouse turbine generators which have never experienced any disk failure have been used at the Surry plant. It has been estimated that for failure- at the normal rated speed or at 120 percent of rated speed, only 2 shrunk - on disks out of 16, in the low pressure turbine could generate external missiles. All other fragments would be incapable of penetrating the turbine casing and would remain within the stationary
- turbine parts. It was judged that the external missiles produced by the two disks vill range from 3,711 lb at 287 fps to 2,665 lb at 416. fps at 120 percent of rated speed.. As all class I structures are designed for
; tornado, the penetration of these structural barriers by missiles is not 4
expected.- In addition, most important areas of the contaitunent and other structures are also shielded by moisture separators /rcheaters or other parts of the turbine building structure. The probability of turbine missiles entering the spent fuel pool is estimated as approximately 10 5, According to NRC, if turbines are maintained and in service inspection is carried out periodically, the frequency of turbine missile generation , less than 10 5 per year can. be assured and _the frequency of turbine missile damage can - be expected to be less than 10 7 per year and a . bounding - _ analysis - is not required, Site - data on the frequer.cy of L inspe ction -- a t Surry -- was not known. However, as per Surr.y ? R , - in . addition to design provisions associated with turbine control and protection system, valves are exercised on a regular basis during unit
- operation -to minimize the possibility of valve stem sticking. Analyses of oil- samples are performed regularly. The turbine is periodically oversped to check the tripping speed. _The remaining tripping devices are regularly - checked. In addition, design, manuf acturing and inspection l
3 39
technique for turbine rotors and disk forgings make the possibility of an undetected flaw very remote. Thus, likelihood of a turbine risk hazard is considered negligibic. 3.4.4 External Flooding l The Surry Nuclear Power Station is located on the banks of the James River on a peninsular site. The ground surface at the site is ilar with a station grade of 26.5 ft above the mean sea level and steep banks sloping towards the river and to the low level waterfowl refuge. Much of the region is characterized by marshes, swamps and streams. The water table is approximately at an elevation of 4 f t and drainage is towards llampton Roads, on the Atlantic Ocean and near the mouth of the Chesapeake Bay. The effects of flooding on the plant components may include (1) inundation, (2) hydrostatic or dynamic forces, (3) Erosion, (4) sedimen-tation, and (5) corrosion. All these consequences, except inundation, 4 are insignificant, The water level in the James River at any time is determined by three components: (1) freshwater discharge from the James River watershed, (2) flow due to the oscillatory ebb and flood of the tide, and (3) flow due to circulation patterns caused by intrusion of saline water within the estuary. Therefore, the water 1cvel rise due to river discharge , high tide, hurricane, intense local precipitation, storm sur60, ice blockage and the effects of waves is to be considered for the Surry site. The drainage area of the river above the station site is 9517 square
- miles. The river between Richmond and the mouth of the river is a tidal estuary and is subjected to tidal motion. The sem1 diurnal tide has two high waters and two low waters in each lunar day. The oscillatory tidos constitute the dominant motion near the site, much larger than downstream flow required to discharge the freshwater to sea. In addition, there is a net nontidal circulacion due to movement of less saline water towards the sea and deeper saline layers up the estuary. The volume rate of this flow is smaller than the oscillatory tidal flow, but it is several times larger than the river discharge.
Due to the wide flood plain at the site, even severe meteorological events produce only a small rise in water level. For example, it is estimated in FSAR that for a 50 year river flood, the 1cvel at the site will not rise more than 1 ft. Even during flurricane Agnes in 1972, peak flood discharge due to excessive rainfall led to flood icycls of 4 ft to 5 ft in Richmond, but negligible levels at the site. Based on 11 years of observations at the site, there has been no significant high water level due to storm surge during the hurricanes. The highest water icvel over reached at Norfolk in 100 years cf records is 8.6 ft. A study of meteorological means and extremes in the Surry site region leads one to l conclude that ice formation c.n the river is unlikely to obstruct the flow l and cause flooding due to salinity of river below the site. The analysis in FSAR identifies the flooding resulting due to storm surge from the probable maximum hurricane given below to be the most severe source of flooding at the site. [ 3-40
Central Pressure Index 26.97 in, of mercury P.adius of Maximum Winds 35 nautical miles Forward Speed of Translation 22 knots Maximum Wind Speed 135.4 mph Based on theoretical models, the surge at the power station was computed and is shown in Figure 3.6. This ",cludes the contributtor' of the highest astronomical tide, an initia. (;se to account for short period anomalies, and the rise due to atmoupheric pressure reduction. For this hurricane, the size, period and length of the waves impinging on the east and west ends of the site, and the resulting run up on the slopes, was found to be small. Calculations indicate that the probable maximum hurricane would not pro-duce a high enough level of water at the site to be considered as a source of risk. For example, the maximum water elevation at the site was calculated to be approximately 22 ft, which is considerably less than the plant grade elevation of 26.5 ft. As only eight hurricanos have passed within a 100 mile radius of the site in the last 100 years , the likelihood of water icvel reaching the peak for the probable maximum hurricane is considered to be negligibly small. In any case, further ! protection is offered by engineered structures such as berms, scavalls, levees, etc. Moreover, for a flood to pose any danger to the plant, the water level has to reach the openings of safety related structures, most of which are either at or above the station ground grade (Table 3.9). Only the circulating water intake structure and emergency service water pumphouse located above it, are the exception. As the sill of the pump t' room door - entrance and air intake louver openings are at 21 f t 2 in.,
- assuming the maximum probable hurricane plus maximum wave run up on the east side, inundation of emergency service water pump diesels is possible, but leak tight construction for doors will prevent this.
Moreover, external flooding events likely to damage the plant generally take time to develop. It can be safely assumed that ample warning time is available for emergency procedures. As per FSAR, air intakn louvers can be sealed with warning of a design basis flood and air for the operation of diesel-driven emergency service water pumps can be provided by the motor operated dampers located in the top of_the pump house structure with a roof elevation of 33 f t 6 in., and beyond the ! reach of waves. Hence, the risk of. external flooding is considered negligible. 3.4.5 Aircraft Impact An assessment of the risk from aircraf t crashes into the Surry structures is presented in this section. For this purpose, information in the PSAR was used. Section 3.4.5.1 describes the information in FSAR, and Section 3.4.5.2 describes the bounding analysis. 3-41
9 S H P AN RO I GT E OAT tt S* h 8 R S
% D F
Y A B E T n
*t s 3s s
e u Y HW R E S - N E 7 O n O E C 2. ST F EP o I K A T Oa A c L E GY i t T RR R a A E P S R UO 7 H S U U t T ECe s S A G C E SS S
/ ' R N S E
H RO r St c U L A T r e E C r t E t N v _ W OwHT E e _ O 1 N r P ST E An I _ USE 6 T y Y N r
/ f U R
R L arr t Ca T O C S r u S E S A T E os S e e S h O t 5 R t C
/ a L s E l V e O v S
I l e T1 4 H S P e A g O W r C 0 u 9 S N E R d P E e O T t 3 F u A m S o R C U O H .
- 6 2 E 3
- w t
T e r u g i F g e o
. 0 6 r e 4 o g 2 e o 3 = 2 d$m a = $s E -e
(______ .- _ _ -__ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ - _ _ - _ _ i i i l-j Table 3.9 Maximum Probable Flood Protection levels for Class I Structures
~
l Flood Protectic.n Level, Class 1 Structure Pt s 'SL
.i i
j Containment Structure 26.5 , Cable Vault and Cable Tunnel 26.5 1. Pipe Tunnel Between Containment and 26.5 1 Auxiliary building i 4 Main Steam and Feedwater Isolation 27.5 Valve Cubicit,- Recirculation Spray and Low.llead 26.5 [ Safety Injection Pump Cubicle ) Safeguards Ventilation Room 26.5 } Auxiliary Building 26.5 Fuel.BuildinS 26.5 { Control Room 27.0 i Emergency Switchgear and Relay Room 26.5 Relay Room 26.5 i Battery Room 26.5 Air. Conditioning Equipment Room 26.5
. Reactor Trip Breaker Cubicle 45.25 i
Auxiliary Diesel.Cenerator Cubicle 26.5 4 Circulating' Water Intake Structure 24.0 (Emergency Service Water Pump-llouse) liigh Level. Intake Structure 36.0 t l-Seal Pit Not Applicable
- i il .
{- i-
.3 43 h .-._,_.,..-.__.;.._.,._...__....~~..__.m.~.-~. .__-,_._~,___.._:..-...._.,_._..-..___...-._,,...,_,,.-,.J,_-._,- - ,
3.4.5.1 FSAR Information The Surry TSAR includes a description of airports and aircraft activity near the site. There are two main airports near the site. Williamsburg-Jamestown Airport, 5 miles north northwest of the site, has a 3,200 f t long paved runvoy. Melville, 6 miles vest southwest of the site, is a private field with a 2,900 ft long unpaved runway. This airfield is used_ by a few small aircraft. These and other airports within 25 miles of the site are given in Table 3.10. There are no federal airways within 5 miles of the plant. TSAR estimated the probability of an aircraf t accident due to flights from the two airports within 5 miles of the site to be 7 x10-7 per year and from Patrick Henry Airport to be 2.7x10 e per year. According to the Standard Review Plan, the possibility of aircraft acci-dents resulting in unacceptable radiological consequences is less than about 10*7 per year if the following requirements are met:
- a. The plant to airport distance D Lis _ between 5 and 10 miles statute miles, and the projected annual numbers of operations _is less than
-500 D2, or the plant to airport distance D is greater than 10 statute tniles, and the projected number of operations is less than 1,000 D2, b, The _ plant is at least 5 _ statute miles from _the edge of military training routes, including low level training routes, except for those associated with a usage greater than 1,000 flights per year, or where activities (such as practice bombin6) may create an unusual stress situation.
- c. The plant is at least 2 statute miles beyond the nearest edge of a federal airway, holding pattern, or approach pattern.
The Standard Review Plan requires that a detailed review of aircraf t impact risk be performed if the above requirements are not met or if
'.. sufficiently hazardous military activities are identified.
~, In the present case, there are two airports at 5 miles from the plant. The project annual number of operations at these airports is greater than 500(5)2(-12,500) operations. Therefore, a bounding analysis is required. 3.4.5.2 Aircraft Impact Bounding Analysis The evaluation of probability of an aircraft crash at Surry Power Station is considered from Felker AAF (5 miles SE, 81,500 movements) and Williamsburg Jamestown (5 miles NN"J, 45,000 movements) . Only accidents within a few miles of the airports are relevant here since there is no air corridor passing directly above the Surry station. 3-44
a i l l I Tabic 3,10 1 I I I Airports Within 25 Miles of the Site ! J J Distance Number of Type of Airport - (mi) Sector Movements /vr Airport Felhor AAF $ SE 81 500 P. M (30) 2 Melville 6 SW - E, R (29) 2 Williamsburg Jamestown 5 NtN 45,000 E, P (32) 1 2 Patrick llenry 11 ESE 172,000 F, C (80) i Langley AFB 19 ESE - P, M (100) NAS Norfolk 24 SE - F, M (37) F Aerodromes with facilities (land)
- - E Acrodromes with emergency or no facilition (land) ,
P 'Public use i C Civil M Military R Restricted () Longth of longest runway in hundreds of ft l l i a 3-45 l
.. _ . , , . _ . , _ _ ._ _ _ , . _ . _ _ _ ,,_.i.__.,...______.i
There is no exact way to model a problem of this type and some of the factors cannot be easily quantified, llowever, an approximate value of the strike probability / year can be estimated from p-pfA where P - Probability of aircraft strike / year p - Aircraf t strike probability per square mile / flight of an aircraft along a given flight pattern f - Number of movements or flights per year of aircraf t along a given flight pattern A - Effective target area of critical portions of the plant The effective target area A includes the base area of the structure plus additional areas accounting for the possibility of skidding of an air-craft after hitting the ground as well as consideration of shadow areas of structures. The numerical values assumed here allow for aircraft hitting up to 100 f t short of a structure and sliding into it. The structures considered as targets include containment building, aux-iliary building, control building, fuel storage building, service water punphouse and tank farm. The exposed area for these is calculated by assuming a 30' slope for the approaching aircraft. This 30' abovo hori-zontal shadow of the height of the structures is considered to be an average trajectory of a ground aviation aircraf t in a landing or takeoff ground collision. A review of the site plan shows that the containment building is the dominant one and shields a large number of adjacent buildings. The shielded structures are thus covered under ariy aircraf t hitting the reac tor - dome. The area is calculated- for four -different directions of aircraft travel and the inaximum value is chosen. Due to the complexity of the site plan, such area computations necessarily involve some approximations. Based on these computations and approximations, the target area is estimated to be less than 3 x 10 3 square miles, leading to a strike probability of 6.6 x 10*7/ year from the Felker AAF and 3.6 x 10-7/ye a r from the Williamsburg-Jamestown airport, i.e., a total probability of 1 x 10-6/ year. (This is different f rom the FSAR due to conse rvative bias in area computation.) llence, the risk of aircraft crash and resulting plant damage is considered negligible. 3.4.6 Internal Floodinn 3.4.6-1
. Introduction A nuclear power plant contains many potential sources of flooding and flood locations. In order to make the analysis of these floods tractable, a process was defined to identify candidate sources and critical flooding areas and to estimate their contribution to core damage frequency if required. The process consisted of the following steps:
3-46
- a. Identification of htportant flood sources and critical flooding areas during the initial plant walkdown. Critical areas can be thought of as those plant areas where flooding could not only result in a plant trip but also damage safety related equipment needed to mitigate the ef fects on cry potentially induced plant transient.
- b. Definition of all initiating events which have the potential to be flood induced for each flood source in each critical area. This step of the analysis results in the spectrum of potential flood rates but is also used in quantification of initiating event frequencies,
- c. Perforre a screening analysis. The screening analysis.is comprised of the following steps:
- 1. Eliminate all plant areas not identified either by the initial plant walkdown or by computer mapping of critical equipment.
- 2. Perform a computer aided critical area analysis which allows for the incorporation of random failures (i.e., failures not related to the ficod itself) as well as all flood related damage. This is a similar process to what occurs in the fire analysis so refer to Chapter 5 for more details on this procedure. This step resulted in flood zone singles, singles with randoms, and double combinations that are listed in Tabic 3.11,
- 3. Screen on frequency for each remaining flood scenario. For Surry this step resulted in elimination of all remaining flood areas and scenarion under consideration. Details of why each of the Table 3.11 areas were screened from further consideration are given in Section 3.4.6.2.
- d. Quantify core damage sequences for each remaining flood scenario,
- c. Perform an uncertainty analysis utilizing the TEMAC computer code for all remaining scenarios.
3.4.6.2 Screening based on critical area analysis. As - de scribed above , a complete critical area analysis was performed for all the areas within -the plant and for all the potential flood induced accident- snquences identified as part of a review -of all internal events accident initiators. This analysis identified those singles, singles in conjunction with random failures, or multiple areas (with or without random _ failures) which, if all equipment- in the zone is assumed to be f ailed by the flood, results in the occurrence of an accident scenario. The results' arc shown in Table 3.11. The zones themselves are defined in Table 5.3 of Chapter 5. The fire zones of that Tabic 5.3 correspond directly with the flood zones of Table 3.11. In addition, the equipment located - in each fire zone is described in Appendix D. Tabic 3,11 3-47
I:. . - . , . . _ _ i f 1. i Tabic 3.11 [. 1 i Surry Flooding Critica1' Area Analysis Summary 1 1 i l b j Sinr.le Zone Single Zone Plus Randoms Double Zones
- . Zone 1 Zone 15 Nono
- l
- Zone 3 Zone 19 l l:
Zone-5 Zone 31 i. i Zone 17 Zone 45 i i Zone 54 Zone 54 L i i \ I l
~
Note: Zones are' identical to the fire r.ones defined in Sect. 5.2
- 8 o
i i l. F < 1 i. 1 a 4 9 l= 3-48 i
.~ - _ ~ ._ - . . - - - - - - - - - - - _=._-
) : presents all the zones that survived the screening analysis and these are ! the zones which were analyzed for the possible occurrence of floods in i this section. Note that the same zone (for example, Zone 2) can occur i cither as a single or as a singic plus random in dif f erent accident ! sequences. (Of course, the same zone cannot occur as a single and as a !- single plus random in .he same accident sequence or it would be non-l minimal). As can be 4een, a total of only ten zones survivad the screening process. Four zones were identified as singles, while eleven
; zones in conjunction with random f ailures were identified. Note that j each of these zones in general was associated with a number of different
, random failures, so each zone itself could actually occur in a number .of j different singic plus random cut sets. Finally, eight combinations of two zones (again, some in combination with random failures) were j -- identified. In the following, each one of these zones or zone plus j random failure combinations are analyzed to determine any potential non-i negligible flooding scenarios. 1
.Qable Vault / Tunnel (Zone 1) l This area adjacent to the emergency switchgent room on 9'6" elevation.
- Most of the safety cabling for Unit 1 passes through the cabic vault. and j tunnel. The only water source within this zone is a deluge fire i suppression system. Two doors enter this area; onc from the emergency j switchgear room and the second via a spiral staircase leaves the area at a higher elevation going to the outside. The only water source in
,. adj acent rooms is a 3 inch pipe running through the emergency switchgear d room in a channel in the floor. Any break from this 3 luch line would be detected by one or more of the three existing flood alarms. The critical equipment in this area are power and control cables for the itPI and CCW systcas. .The lowest point that this cabling is relative to the floor is approximately three feet above floor level. As a consequence, water i level in the tunnel would have to be approximately three feet high before
- postulated damage could occur. Given that the only adjacent water source 4 is in the emergency room this scenario can be bounded and neglected in comparison to flooding within the emergency switchgear room itself which ,
i contains both safety trains of the emergency 4KV switchgear. Since the switchgear are lower. relative to the floor than the cabling in the cable
- vault / tunnel it is clear that flooding in the emer6cncy switchgear room
- . would effect the 4gV switchgear which would result -in station blackout) long before -any failures of the UPI and CCW system occurred in the cable p vault / tunnel area, llence, this scenario may be screened out.
l. EmergtDi;y Switehntar Poom (Zone 3) I This room-is at elevation 9'6". As mentioned above, the only water
' ~ source in this room is a three inch pipe laid in the channel in the floor
- and protected by three flood alarms, There are two doors into this area for Unit 1. It is connected to the Turbine Building (through the Unit 2 cmcrgency switchgear room) through a fire door and a 2 foot flood
- barrier. This door Icads out to the bottom floor of he Turbine Building also at elevation 9' 6" . Secondly, there is the ir into the cable l.
3-49 i __..m....,,_ . ~ . _ , . . l ~... _ _ , __ - . . . . , . _ , _ . _ . . , _ . , . , . , _ _ _ _ _ , . . _ , _ , _ . . , . _ _ . , _ _ _ _ , . _ , ,,,_.,.w, , ,-..-m_.w,..-
i
)
l vault / tunnel described above. This zone is a single inasmuch as failure of both 4KV switchgear due to flooding would result in station blackout and also a seal LOCA. It is estiinated that flooding at least one foot high in the entire area would be required to fail the 4KV switchgear, i Two scenarios need be considered for the emergency switchgear roorn. The first is the case of the break of the 3 inch pipe within the room. In this case, for a problem to occur it would be necessary for all threc
; flood alarms to simultaneously fail and for the sump pumps also to fail.
- Given the small volume of water available through the 3 inch pipe, and the low probability that all three flood alarms would fail this scenario can be screened from further consideration.
4 The second scenario would involve an unisolatable flood in the adj acent Turbine Building, raising the water level of the Turbine Building above the two foot flood barrier allowing water to flood the entire emergency switchgear room (of both Units 1 and Units 2). An unisolatable flood is possible because intake canal level is (normally) approximately 8 feet above Turbine Building basement level. If the inlet piping failed (low pressure lines) two random value failures would also have to occur to make this scenario valid. Therefore, failure of the inlet piping can be eliminated frorn further consideration. Surnp pump capacity is such that failure of the shell side of the condenser would not provide a sufficient water source to exceed the 2 foot barrier at the entrance - to the emergency switchgear roorn . Therefore, any postulated mechanism for an unisolatable flood can be screened. Mechanient Eaulpment Room (Zone 45) This room contains a service vaeer system which provides cooling to the lube oil- supply for the llPI system. If this equipment were to fail, it would fail the llPI system. There is only one door into this room (from the emergency switchgear roorn) and the only water sources in the room are from the small pony pumps t'< . alves and the three inch supply line in a channel in the floor. Th!. .a is in a cut set n conjunction with two random failures. The two p a t' e scenarios are flood induced failure of the llPI in conjunction with n e stuck open relief valve (random) and random failure of the remaining service water pump located in a different flood zone. The second scenario would involve flood induced failure of the llPI system in conjunction with random failures of the CCW system and again, random failure of the other service water pump. These two scenarios can be eliminated based on the random failure probabilities and a conservative pipe-break frequency estirnate of 1E 3/yr.
- Charrine. Pumn Service Water Pumn Room (Zone 54)
This zone is a single room on the wall of the Turbine Building (on the
- l. opposite side of the wall from the mechanical equipinent room #3) and has a single door connecting this area into the Turbine Bay at elevation 9'6". This area contains one of the two charging pump service water system pumps and in addition, contains a cable for the other charging i
3-50 i l
. _ . . . _ -~ _ _ __ _ - _ _ _ . _ _ _ _ .__ _ . - _ _ _ _
pump service water system pump. *rhc flood scenarina in this room could damage both trains of the charging pump service water system and hence, fail the lipl . Again, tuo scenarios can occur which are a small LOCA involving flood induced failure of the HPI in conjunction with a stuck open safety relief valve or a seal lhCA due to flood induced failures of the HPI in conjunction with random failure of the component cooling water sys um . The only water sources in the room are two small capacity pony pumps, lloweve r , 'llnods in the Tttrbine Building could enter this area under the ow. These two scenarios can be subrmed with the flooding scenarins associated with the emergency switchgear room inasmuch as both these scenarius require additional random failures whereas scenarios associated with the emergency switchgear room Icad direct:y to a station blackout scenario. The cabic associaced with the charging pump service water system ptunp in the adj acer c mechanical equipment room enters through the common Turbine Building wall at an elevation of approximately 4 foot above the floor level and then exits through the ceiling. Hence, a flood in this room would have to flood the entire area over 4 feet in order to fai; both of the service water pumps. Wth a 1E-3/yr pipe break frequency, which is clearly conservative, and the stuck opeu relief valve probability of approximately 1E-4/ demand this scenario can be screened out, Similarly, random failure of the component cooling water system is associated with a failure probability of approximately 1E 3/ demand and thus, would also screen out in conjunction with failures of the pony motors or pipes within_the room alone. Also, for the seel LOCA case, a readily available recovery action is to cross connect to - the unit #2 component cooling water system. Turbine Buildine (Zone 31) The Turbine Building elevation 9'6" was found to -be a single zone in conjunction with random failures in the vital area analysis. This arose due to the fact that cables from both charging pump service water system pumps enter the wall of the Turbine Building at approximately 7 foot elevation above the floor and hence, any flood which shorted those cables - out would fail the llPI system in exactly the same scenarios as dircussed for zones 45 and 54. However, a flood 'in the Turbine Building up to elevation 7 foot above t!'e floor level would by then have exceeded the barrier into the emergen y switchgear room and hence, gave rise to the scenarios associated with - that - zone which are more-severe (station blackout) than the - scenarios which would result in this case. In addition, those scenarios for the Turbine Building require random failures of the component cooling water system or stuck open relief valve as discussed for zone 45. Hence, this flood zone can be screened since it is subsumed by the scenariot associated with the emergency switchgear room. Control Room (Zone 5) This is at elevation 27 foot adj acent to the Turbine Building. The control room itself has no water sources other than those associated with air conditioning and normal domestic water supply. Rooms surrounding the l
)
3-51
_ - -- . - . = . _ _ . _ .- . - _- -- . - - - - 1 l control room consists . of lunch room and - of fice space, Again, these adj acent areas have no significant sources of water. Above the control room is the normal (non emergency) cabic spreading room which has no water sources, lle nc e , the only flooding that could occur in this room , J l would be due to fiooding in the Turbine Building. This would require flooding the Turbine Building to elevation 27 feet which, as discussed above, would have already resulted in flooding of the energency switchgear room with its associated station blackout scenarios, llence , floods in the control room. (which is a single vital area analysis) are subsumed by floods in the emergency switchgear room, Auxiliary Buildinn (Zone 17) The Auxiliary Building is a single vital area analysis zone because it contains both the caponent cooling water pumps and the high pressure injection pumps r.nd failure of those systems together leads to a seal LOCA. All theso pumps are located at the bottom (two foot elevation) IcVel. The cubicles for the high pressure inj ec tion pumps flood are isolated at the 2 foot floor elevation from the main floor area which contains tne CCW pumps, These walls extend to the 13' elevation, Since there are no significant water sources either above or adjacent to this area and flooding would have to reach the 13' elevation, this zone was climinated from further consideration.
- 'afeguards Area (Zone 19)
This area is comprised of the rooms surrounding the containment and contains the auxiliary feedwater system pumps and the low pressure injection pumps. There are several elevations in the safeguards area. The auxiliary feedvater and LPI _ pumps are on the ground floor elevation level. This zone occurs in cut sets associated with additional random failurea and two types of scenarios are possible. The first is associated with a random failure of the feed and bleed function in conjunction with. flood induced failure of the auxiliary feedwater system. Random f e " <res of feed and bleed are due, for exampic, to random failures or ao PORV or random failures of the llPI system, The second type of scentrio is associated with a stuck open safety relief valve and involves flood induced failure of both the auxiliary feedwater system and the LPI system which thus results in failure of the long term recirculation function. Since the random failure probabilities for the l'ORVs , safety relief valves, and the llPI system are approximately 1E-4/ demand and random pipe break frequency which might 1 cad to a flood is smaller than lE-3 per year it can be seem that these sequences (conse rvatively) are less than lE-6/yr and hence, can be screened out from further consideration. Containment (Zone 15) The containment occurs as a single zone in conj unction with random failure of the auxiliary feedwater system. The containment flood must fall a PORV. The PORVs are located on the top of the pressurizers. 3-52
Cabling for the PORVs runs down the pressurizer, then is routed along the
-containment wall out through the upper elevations of the safeguards area and then directly into the cable vault / tunnel. This would require flooding of the containment structure to approximately the 18 foot elevation, This scenario can be screened by virtue of a frequency of pipe break being bounded by 1E-3 per year, the probability of a spurious actuation induced by the flood in the PORV (approximately lE 1 per demand) and the random failure probability of the auxiliary feedwater system which is approximately 1E 3/ demand. Taken together these factors demonstrate that the scenario can be screened from further consideration.
3,5 Summary The scoping quantification study considered all possible external events at the site except for seismic and fire events, since these two events were included in a detailed external events analysis. The PRA Procedures Guide (Ref. 1), suitably augmented with other available information, was used as a guidelines for identification of all possible external events at the Surry site. Next, an initial screening process was carried out to eliminate events not applicabic to Surry from the list. For this purpose, a set of screening criteria was developed and then each external event was examined for possible elimination based on these criteria.
' Af ter the initial screening process was completed, the following events were found to be potential contributors to the plant risk,
- a. Aircraf t Impact
- b. External Flooding
- c. Extreme Winds and Tornadoes
- d. Industrial or Military Facility Accident
- e. Pipeline Accidents
- f. kolease of Chemicals from On-Site Storage g, Transportation Accidents
- h. Turbino Cienerated Missiles
- 1. Internal Flooding The degree of sophistication in the bounding analysis for each event depended on whether the event could be eliminated based on only a hazard analysis or a complete analysis including hazard analysis, fragility evaluation and plant response analysis. The detailed plant response analysis was conservatively neglected in evaluating the impact of these external events.
The risk due to an aircraf t striking the plant structures and causing unacceptable radiological consequences was screened out on the basis of the probability of strike and the design of different structures. Evaluation of the potential for floodin6 as a result of the most conservative combination of Probable Maximum Flood (computed from conserm tve estimates of probable maximum precipitation) and wind-gene re.: waves showed that the essential structures in the plant are locate: .ove the probable maximum surge level and the risk of flooding is neg;.glbly small. ! 3-53 L L
_ . - _ - - . . . - ~ - . . . _ . . . - - . . . _ ..
-Tornadoes and tornado missile impacts were eliminated on the basis of a detailed computation of tornado strike probability and other.fcatures of plant structures and components designed to withstand the effects of a Design Basis Tornado. -
The information availabl6 from the Virginia Power Company was used as the basis to assumed the safety of essential plant structures from damage due to turbine missiles. Finally, explosions due to pipeline accidents, transportation accidents and both on site and off-stte chemical release were determined have a low probability of affecting the site. Thus, all external hazards except fire and seismic events were found to be negligible' contributors to the risk of core damage at the Surry plant. Detailed evaluations of fire and seismic events are contained in the remainder of this report, l 3-54
- . , . . - - .~ .- . . - . .. - . . . - - . - - . . - ..- - -
3.6 ' References 1, USNRC, PRA Procedpres Guide, NUREG/CR-2300, January 1983.
-2. Virginia Electric & Power Company, ]Jpda t ed Final 'Sa fety Analysis B.eport, Surry Nuclear Power Station, Richmond, Virginia, 1983.
- 3. L. A. Twisdale and W. L. Dunn, Tornado Missile Simulation and Desing Mc.thodology . Vol. 1, Research Triangle Institute, NP 2005, August 1981.
- 4. K. A. Soloman et al., D timate of the llazards -to a Nuclear Reactor From the Random Impact of Meteorites, UCLA - ENG - 7426, March 1974, 5, NUS Corporation, Surry Onsite Toxic Chemical Relense Analysis (Vol.
E) and Surry Offsite Toxic Chemical Release Analysis (Vol. II),
#NUS 3735, 1981.
- 6. USNRC, " Assumptions for Evaluating liabitability of nuclear power Plant Control Room During a Postulated llazardous Chemical Release,"
Regulatory Culde 1.78, U.S. AEC Directorate of Regulatory Standards, June 1974.
- 7. USNRC, " Evaluation of Explosions Postulated to _ Occur on Transportation Routes Near Nuclear Power Plants," USNRC Office of Standards Development, Revision 1, Regulatory - Guide 1.91, February 1978.
- 8. T. V. Eichler and 11 . S. Napadensky, Accidental Vanor Phase l- Exclosions on Transportation Routes near Nuclear Power Plants, USNRC, NUREG/CR-0075, May 1978.
- 9. T . _V . Eichler et al, Evaluation of Risks to the Marble 11111 Nuclear Generating Station from Traffic on the Ohio River. Engineering Division, IIT Research Institute, Chicago, Illinois, October 1978.
10, D , _W . Murphy, Battelle Pacific Northwest Laboratories Letter to L. A. Ilulman, USNRC, March 23, 1982, regarding Control Room liabitability Evaluation.
- 11. S. A. Varga, USNRC letter to R. 11. Leasburg, VEPCO, June 28, 1982, Docket No. 50-280 and 50-281.
l2, NUS Corporation, Severe Accident Ri sk Assessment , Limerick Generating Station, Philadelphia, Pennsylvania, prepared for NUS Corporation, Philadelphia Electric Company, 1983,
- 13. A~. D, Swain and 11 . E. Cuttman, llandbook of Iluman Reliability Analysis, NUREC/CR-1278, SAND 80-0200, Sandia National Laboratories, Albuquerque, NM, August 1983.
3-55 1
- 14. USNRC - " Protection of Nuclear Power Plants Against Ex t renm Winds and Tornadoes." ErgRLa t o ry Gui de 1. l LZ , 19 7 7.
- 15. M. R Ravindra and 11. Banon, litt.lto1 s for IMnntn.l.lv.ent ScirenLug 1
Qunntifita11nu, n NUREG/CR 4839, to be published.
- 16. USNRC - " Protection Against Low Trajectory Turbine Missiles," Rev.
1, Rego.lnlory Gu i d_e._LJlh , J uly 19 7 7 .
- 17. G. E, Sliter et al, "EPRI Research on Turbine Missile Effects 1n Nu c l e a r Powe r Pl a n t s , " Ir_ajl. sac _tlons_o f 7 t b I n t e nt[LLj o2Daj Go p f e r e nc e gn...ELnic turAL.1[e_thn al e s in Renctor Technology, Paper J815, Chicago, Illincis, 1983, pp. 403 409,
- 18. S. 11 . Bush, " Probability of Damage to Nuclear Components due to Turbine Pallure, in thicJent Safety, Vol. 14, No. 3, May - June 1973,
- 19. USNRC, Staniard t Review Plan for the Review of S n.fe t y Attn l ys i s Reports for Nuclent Power PlanLE, USNRC, Office of Nuclear Reactor Regulations, Washington, D. C. , NUREG-0800 LWR Edit lon, July 1981.
- 20. M. A. Acarm et al., A Review of the Li me_tttk Generat ing Etn11cn Severe Accident Ansessment, NUREC/CR 3493, July 1984
- 21. E, M. Patton et al, ProbnhLlistic Analysis of 1.ow Pressur Steam Turbine Missile Generation Events, Battelle Pacific Northwest Laboratories, EPRI NP 2749, August 1973.
3-56
v n --_ 4,0 SEISMIC .'RA A detailed seismic risk assessment was performed for the Surry Plant. This analysis utilized dynamic response calculations for all important structures, a generic seismic fragility data base for components, and detailed component fragility derivations for a number of components identified during the plant visit as f alling outside the generic data base, Hazard curves developed by the USNRC sponsored Seismic llazard Characterization Program at Lawrence Livermore National Laboratory for the USNRC and by the Electric Power Research Institute were used, Mean values of accident sequence and core damage frequencies were obtained using a Monte Carlo approach. Each of those aspects of seismic risk are described in the following subsections. 4.1 S e i sn'i c i t y and Hazp_rd Curves The earthquake hazard at a given power plant site is characterized by a hazard curve and a site ground motion spectra, The hazard curve is a f requency plot which gives the probability of exceedance (per year) of different peak ground accelerations, The site response spectra describes the relative frequency content of the carthquakes expected at the site, and also the influence of the local soil column and layering in modifying the earthquake frequencies transmitted to the site. 4.1.1 Ceneral Considerations For a given site, the hazard curve is derived from a combination of re-corded ear '.. quake data, estimated earthquake magnitudes of known events for which no data are available, review of local geological investigations, and use of expert j udgme nt from seismologists and geologists f amiliar with the region in question, The region around the site (say within 100 km) is divided into zones, each zone having an (assumed) uniform mean rate of earthquake occurrence. This mean occurrence rate is determined from the historical record, as is the distribution of earthquake magnitudes, Then, for the region under consideration, an attenuation law is determined which relates the ground acceleration at the site to the ground acceleration at the carthquake source, as a function of the earthquake magnitude. The uncertainty in the attenuation law is specified by the standard deviation of the data (from which the law was derived) about the mean attenuation curve. These four pieces of information (zonation, mean occurrence rate and magnitude distribution for each zone, and attenuation law) are then combined statistically to compute the hazard curve, The low level of seismic activity and the lack of instrumental records make it difficult to carry out seismic hazard analyses for the central and eastern United States using historic data alone, To augmer,t the data base, current methodologies make use of the judgment of experts f amiliar with the area under consideration. Approa. used to generate the subjective input, to assure reliability by feceu loops and cross-checking, and to account for biases and modes o f j udgme n t are described in detail in Bernreuter (Ref. 1). l l 4-1
-- .- - - . . _ ~ - -
4.1.2 Hazard Curves Used For Surry The- hazard curves used in the NUREC 1150 PRAs were taken from two sources. The first set of curves was obtained from the USNRC sponsored Eastern US Seismic Hazard Characterization Program (Ref. 1) being performed by Lawrence Livermore Laboratories ,LLNL). From this program one can obtain a median hazard curve and an estimate of the distribution about the median curve. This is shown in Figure 4.1 where the mean, median, the 15th percentile and 85th percentiles are shown. According to the principal investigator of this prograa, the distribution about the median is nearly log normal so for use in the NUREG 1150 analyses a log normal distribution was fit using the median and mean curves. From this fit any particular percentile curve of the hazard curve family can be F computed. Table 4.1 lists the numerical values used in fitting the LLNL hazard curves. A second set of hazard curves was obtained from the industry sponsored Electric Power Research Institute's Seismic Hazard Methodology Development program (Ref. 2). The corresponding curves are shown in Figure 4.2. These were also fit with a log normal model. The numerical-values used in fitting the EPRI curves are listed in Table 4.2. Note that the mean hazard curves of Figure 4.1 and 4.2 are near or above the 85th percentile hazard curve shown, This mean hazard curve will be found to drive the calculation of mean core damage frequency estimates as explained in Section 4.4. The two sets of hazard curves shown in Figures 4.1 and 4.2 are significantly different, bath in regard to location of the mean hazard curve as - well as to the range of uncertainty about the median curve. This is not too surprising inasmuch as the emphasis of the t. programs was somewhat different. The EPRI Program focused on very detailed geological studies of the sites in question, and resulted in a somewhat finer zonation of each site. However, only three attenuation (ground motion) models were used. Further, while a number of' teams of seismological and geological experts were essembled, each team was proscribed to reach a consensus on the final hazard curve families developed by that team. By contrast, in the LLNL program, considerable emphasis was placed on the full range of attenuation models, and rather than a number of teams, a total of 11 seismicity experts and five ground motion experts were individually polled, and a full set of 2750 hazard curves were developed for each site by considering caen expert's input equally likely. The curves developed in this process encompass somewhat more uncertainty than those produced by the EPRI process, and the increased uncertainty leads to higher probabilities of nonexceedance for the LLNL mean curve peak ground acceleration val.ues than are obtained from the EPRI distributions. At this time, both sets of hazard curves are viewed by the US NRC staf f as being equally credibic. As such, calculations of the scismic core damage and plant damage state frequencies at Surry are presented for both sets of hazard curves in this report. l l 4-2
Probability of Exceedance (per year) 1.0 E - 01 3 j
- Mean 1.0 E - 0 2 s-1.0 E - 0 3 s 1.0 E - 0 4 s
_ . N, 1,0 E - 0 5 s . 85 .- I - 50 - 1.0 E - 06 s is N
~
1.0 E - 07 O 0.2 0.4 0.6 0.8 1 L Peak Ground Acceleration (g) l. Figure 4.1, LLNL Surry Hazard curve: Mean, Median, 15th and 85th Percentilo curves i 43
Probability of Exceedance (per year) 1.0E- 01 = E
~
Mean 1.0E - 0 2 s 1.0 E - 0 3 s
~
1.0E - 04 s ,
~
1.0 E - 0 5 s N, , 85 s _ 50 ' s. 1.0E - 06 - 15 , 1.0 E - 0 7 O 0.2. 0.4 0.6 0.8 1 Peak Ground Acceleration (g) Figure 4.2. EPRI Surry Hazard Curve: Mean, Median, 15th and 85th Percentile Curves 4-4
. - . - . . . _ . . . . . - . -. -. . ~.
Tabic 4.1 LLNL Mean and Median llazard Curve Values Mean llazard Median llazard, PGAff l f.xce. dane. Der year f e r e..d an: . per venr 0.05 4.10E-3 1.67E 3 0.15 4.24E 4 9.45E 5 0.25 1.25E-4 2.03E 5 0 35 5.40E-5 6.85E-6 0.45 2.78E-5 2.92E-6 0.55 1.61E 5 1.43E 6 0.65 1.01E-5 7.75E-7 0.75 6.74E 6 4.77E 7 Table 4.2 EPRI-Mean and Median llazard Curve Values
- Mean llazard Median llazard' PGM n) . fenceedance. per voar Eexce dane, per year 0,05 1.92E-3 1.11E 3 0;15 1.35E 4 4.68E 5
'O.25 3.28E 5 8.52E 6 0'35- 1.21E 5 2.56E 6 0.45- 5.54E-6 1.01E-6 0.55 2.92E 6 4.77E 7.
0.65 1.67E-6 2.38E 7 0.75 1.07E 6 1.19E 7
- Note that numerical values for the EPRI curve shown here differ slightly from those published in -the final version of Reference 2. The final core damage frequency results reported here would be decreased by 12% using the latest EPRI hazard curves, with the relative importance of components and sequences being unchanged.
45
l l l l
~
4.2 Response Calculations 4.2.1 Introduction i As previously described, seismic probabilistic risk assessments (PRAs) can be considered in a series of steps: seismic hazard characterization, scistnic response of structures and components, structure and component failure descriptions, plant logic models, and probabilistic failure cal-culations. Section 4.2 deals with the frequency characteristics of the free field ground rnotion (an olement of the seismic hazard characteriza-tion) and the seitmic response of structures and components. In a seismic PRA of a nuclear power generating plant's safety systems, only tho' components affecting the operation of the systems and those structures housing or supporting these components need to be analyzed. Plant-logic models .' identify the components. Plant general arrangement and mechanical drawings are _ then used to locate the components and identify the relevant supporting structures. For the Surry Power Station the specific safety related components are housed in the Reactor Building, Auxiliary Building, Safeguards Area, Emergency Generator Enclosure, Containment Spray - Pump Enclosure, Control Room, and Intake
. Structure. Figure 4.3 illustrates the general plant layout showing relative location of these structures.
Seismic PRAs require as input best-estimate structural response, varia-tions of -response and correlation of response. A seismic PPA considers '
-earthquakes over the entire - range of the seismic hazard curve; hence, seismic - responses ' must be determined over this range. Often, seismic response -determined as part of the plant design process is available.
However, this data reflects the conservatism associated with the seismic
-design analysis methodology and considers only low seismic excitation levels.
To determine structural response at the higher excitation levels required by a seismic PRA, either the design analyses must be extrapolated or reanalyses of the structures must be made. For this study, analytical models of each structure identified above as housing safety-related components.were developed and used in a probabilistic response analysis to determine the best-estimate seismic response of these structures.
.The balance of this section will describe and summarize:
- a. site and seismic characteristics
- b. probabilistic response analysis of each structure i c. in-structure responses which define the responst. ci i safety related components l
l l' 4-6
If i P i
) FUEL BUILDING f. -
REACTOR 1 REACTOR 2 , SAFEGUARDS' BLDG AUXILIARY . BUILDING-
-f .
NTAINMENT-SPRAY p EMERGENCY. GENERATOR-
- t CONTROL SERVICE' BUILDING 3 BUILDING . t i
I TURBINE. BUILDING' l s i t i l 7
)
.. Figure 4.3 Surry Power Station General Arrangement 4
I i 3
.mm.. ..mm. _ %. . _ - . .
p_ _ _ _ . . , . . , _ _ -, ,-.. _7.,-
.--..,,:-,,-u_-- .
3 ___ ______. __ ______ __
4.2.2 Site and Scismic Characteristics
~4.2.2.1 Site Description.
The Surry Power Station site is characterized as a deep soil site of alternating strata of clay and sands of the Pletstocene age. The Pleistocene age strata lie unconformably on Mioenne clays beginning at elevation 38. Original ground elevation through the area of the site was 434 ft. Finished grade exists at an elevation of 426.5 f t. The Miocene clay is heavily over-consolidated extending to -280 ft in olovation. -Formations of the Eocene, Paleocene, Cretaceous and Crystalline age exists beyor.d the Miocene clay strata, Figures 4.4a.and 4.4b show the Pleistocene and Miocene age strata and foundation elovations for the Surry Power Station structures. 4.2.2,2 Soil Properties and Earthquake Definition. Two interrelated objectives for the initial portion of this _ investigation were to:
- a. define strain compatible soil properties over the range of seismic excitation levels defined by the seismic hazard curves,
- b. define the input motion for the probabilistic response analyses of the structures The safe shutdown earthquake (SSE) for the Surry site is defined to have a peak horizontal ground acceleration of 0.15g.' Three seismic excitation levels were considered and defined by their peak ground acceleration in the horizontal direction'-- 0.15g (1 SSE), 0,30g (2 SSE), 0.45g (3 SSE).
They are denoted acceleration ranges 1, 2, and 3 in subsequent discussions. These excitation levels were treated explicitly -- input motions and probabilistic. response for other levels defined by the hazard curve can then be interpolated from the results. In general, soil properties such as_ shear modulus and damping are a func-tion of soil strain and consequently a function of excitation level, i.e., acceleration ranges 1, 2, and 3 defined above, With higher excitation levels, soil shear modulus tends to decrease while soil damping tends -to increase. Equivalent linear visco-olastic soil properties as a function of excitation -level were developed using the program SHAKE (Ref. 3). The soil deposit is idealized as a series . of horizontal layers. Low strain soil properties were derived by relationships between blow counts and shear wave - velocity for the sand strata (Ref, 4), The blow counts for the sand layers are given in Reference 5. Shear modulus for the clay ' s tra ta , both Pleistocene and Miocene deposits, are reported in Reference 5 as derived from quick shear ! test results on undisturbed samples. There are three principle layers of strata with varying low strain soil properties as given in Table 4.3. Estimation of equivalent linear strain compatible properties is preceded by defining the relationship between soil shear modulus and strain (shear modulus degradation curve), and soll material damping and strain. No 4-8
CONTAINMENT FUEL CONTAINMENT 40' _ l STRUCTURE NO. 1 DUILDING STRUCTURE NO. 2 I 0 8a. 10'._ .,.. .. .... ..-. :.- ~.--
~- .-
s . - . . . .' -: < ' . . . --. y .
- .~.-^
. ~
e - . . , . . . .. .
~.-r. :; . . '- .-
~- *
"' .f.-.: e:. . . .~., : . . -.*
0,- . .' :<
- \
-h..*':.T 6..N.
;t., ': Q:% V z
o -1g* - ' .i:.s..
. . - '. w < ~-
H ...s .. p 20, - ......1
.u 9.. ......y.
- .~ .
.< :,. . . a' . ... : - - -
.* '^;~~ n -
T* * ~ ' * ' ' '
' ^ *
~
'~A***
- O '*^*L~'*
D -30' -
'l..':~. -ll.5'.~$.u. ~.M*Y. - -
a -40' r.: - l
-50' _
i -60' - PILES i ' - -70' _ e I t I CLAY l l COMPACTED BACKFILL SAND Figure 4.4a. Surry Power Station Subsurface Profile E-U Section I - ---- _ _ _ _ _ - _ _ _ _ - _
I 40'- 30' __ TURBINE ROOM CONTROL AUXILIARY FUE 20, - ROOM DUILDING DL 10* **'.M..,.~~,.'..<.-*.:*; E-* . ga _. M Oe - g "
t._10 ~ c o - n y n ;. .s,.
vp ~l=: >: ::.. ; - '*.;.
. -: _ . .-: ~. .
,~
x . #. 9:.t.;q;;rfg.1'.'.~;.
' :.r , .*:.*.
. -~ -- ... . -
z ~20, - -- O 30, ..
. . . e 1... . .:. : s.
r5.* .,. .. .. ... c_.. .. ,, .. -- .-, ,. ..*
...,, t w -
-~ . '. .:. y.~:-
* . .. . g ; - . ., . .* ...- , . . . m . 4.. . A. .* .*: .~ . s
_ . ;.; . ,- .-- .n.. 9<:-40' _ y r- '..' . ... .- .: ..
= . . c: . .. y i s4 -50' -
a o ga
~ -60 ' -
o -70'~ PILES PILES
-80' _
i ! CLAY COMPACTED BACKFILL
- t
= i
"- SAND
- h. . t i
i i t Figure 4.4b. Surry Power Station Subsurface Profile N-S Section t r
l 1 i 1 i Surry Power Station specific dat- were availabic, thus curves developed by Seed and Idriss -(Ref. Q were used i_n the present investigation, i Vertically propagating shear waves are assumed to_be the wave pregation mechanism by which horizontal motion propagates to the soil free surface. 1 Nominal strain compatibic soil properties for the three acceleration - ranges computed using the SHARE- code are shown in Table 4.4 and Figure 4.5. Tabic 4.3 Surry Power Station Low Strain Soil Properties Shear Wave Unit Thickness Velocity Weight Laver No. (ft) (ft/s) (nef) 1 20 740 110 2 40 810 110 3 280 820 110 The input motion used to develop these values and also to perform the probabilistic analysis was - developed from recorded carthquakes at soll sites. A suite of ten earthquake acceleration time histories was defined and scaled to each of the three excitation levels. .F - " of the acceleration time histories consists of recorded motiont af ' actual l carthquakes from similar soil sites. A total of five recorded earthquake l acceleration time histories were selected and listed in Tabic 4.5. For the _ purpose of the analyses a total of ten input acceleration time histories in each orthogonal horizontal direction was created by rotation of the two horizontal components. The-median acceleration response spectrum.of-the ten horizontal components is shown in Figure 4.6, along with median response spectra for two types of similar soil sites: soft to 1 medium clay and deep cohesionless soil as reported in iteference 7 The comparison shows that frequency content and amplification for the median response of the ten horizontal components adequately the represent the expected motion at the Surry Power Station.
-4.2.3 _Probabilistic Rcsponse Analysis In recognition of the importance of the effects of embedmont and soil-structure interaction (SSI), probabJlistic soil structure - interaction -
building response analyses were_ used to generate median responses for the Surry Poter Station structures housing safety related components. The L methodology used is that of SMACS (Ref. 8) as implemented in the computer L progra.n CLASSI (Ref. 9) utilizing the substructure approach. The substructure approach to' SSI is composed of the following elements: specification of the free field ground motion; determination of the 4-11
._ _ _ _ _ _~_ _ _ __
l 1 l Table-4.4a Surry Power Station Strain Compatible Soil Properties - Acceleration Range 1 Elevation Shear Modulus Shear Wave
- (ft) (KSP) Velocity (fos) Damoing 27.5 1805. 727. .016 21.5 1172. 586. .026 16.5 883. 508. .033
'11.5 699. 452. .038 65 825, 491. .039 1.5 755, 470. .041
-3.5 684. 447. .044 13.5 620. 426. .047
-23.5 580. 412. .049
-33.5 588. 415. .049
-43.5 585. 414. ,049 53.5 596. 418. .049
-63.5 612, .423. .047 73.5 624, 427. ,047 83.5 652. 437. .046 93.5 628. 429. .047 Table 4.4b Surry Power Station Strain Compatible Soil Properties -- Acceleration Range 2 Elevation Shear Modulus Shear Wave (ft) (KSF) yp'acity (fos) Damoint 27.5 1388. 637. .022 21.5 722. 460. .037 16.5 534. 395. .046 11.5 410. 346. .055 6,5 490. 379. .055 1.5 435, 357. .061
-3.5 374. -331, .068
-13.5 325, 308. .074
-23.5 319, 306. .074
-33.5 372. -330. .069
-43.5 388. 337. .068
-53.5 398, 341. .066
-63.5 344, 317. .072 73.5 291, 292. .078
-83.5 278. 285. .080
-93.5 241. 266. .088 4-12
s 1 Tabic 4.4c Surry Power Station Strain Compatible Soil - Properties -- Acceleration Range 3 Elevation Shear Modulus Shear Wave (ft) (KSF) Velocity (fos) Damping 27.5 1190. 590. .026 21.5 623, 427. .041 16.5 444. 361. .050 11.5- 346, 318. .063 6.5 409, 346. .064 1.5 358, 324. .070
-3.5 302, 298. .076
-13.5- 266. 279. .082
-23.5 252, 272. .084
-33.5 264. 278. .083
-43.5 266, 279. .083
-53.5 273. 283. .082
-63.5- 260. 276. .084 73.5 251, 271. .086
-83.5 247, 269. .087
-93.5 221, 255. .092 Table 4.5 Free-Field Acceleration Time Histories for Probabilistic Responso Analysis Site Date El Centro May 18, 1940 Hollywood July 21, 1952 Storage Ferndale December 21, 1954 City Hall Hollister April 8, 1961 8244 Orion February 9, 1971 Los Angeles 4-13
; ;Il ili ,!l i .)l , 1 il l ; l 0
1
- 4, r 1
- F.
f g ! 0
) t x,.
( - '
. e ,,.]
9 4 g _ [ t 13 I 3* ( - #. q n. i ryf .el t p { 5 m ~I' ~ s ? h c O
}' t p
O' 1(-
- a. D e
h 8 ,~
"% 2 t
i 1 - -
- w I - -
o
- i
- - 0 - -
t
- a c- 7 4 - R
# 2 a. &- R'. o. ?. .
- - -
- s 1 1 g o - - - - n t
- i
- p x
m g CO .sg3o"t e t D a d n _ a s u l y u d o 50 1
- M 2 e X r l 1 23 a
~
b e i eec h S 0 t gga 2 annn l l oaaa i f nmn9R o s m. I l o S
- k. aC f -
;e I .
5 o d t n s. 1 n Si n u I e a o - l g *. r i t u t . e ot a
- d. 0 LLS i o.
M, 1 r a V e.
- c. .
- ":['T.
T -- 5 5 h.' ! S
'[3f.b l
]I l
l e et , [: a y
3 t
a a(J I,
. 4 r
e u g
- 0 i 0
4 F I. - e - s. 2 a. &. &. 2 l
- -
- t 1 e
t
'N 3 b C0 seyeno a et
- 3. n
' . l'
4 m to ' 3.O e
- 2. 5--
Ts ' 's t
's f ee
- 2. e-- f . eb I
"'(' s 1
i e ff 3 f h' r t
,1g t.
1.5-- g e u g
\ s,\
o si \ 4 ,s -d ' -
/ If s . e--
9 s
> /
- i i /
/
w T' /
/-
/
/
/ /
,/
3 : : - - : :: -
- - ue: -
le 80 to IO SO' Frequency fitz t Legend: Notes: COMestentest Sotts AtI spectra calculated at 5% semping Sort to peentum Ctes Acceleretton in units of g E. Send sist, _ _ _ ___ . Heen of 10 Comp. - - _ _ _ - _ _ . Figure 4.6. Surry Power Station Median Free-Field Input Motion Compared to Soft Cohesionless Soils and Clay Spectra (Ref. 7) 9 .. .. _.w
foundation input mocion and 11mpedances ; calculation of the _ dynamic characteristics of the structure; and analysis of the coupled soilstructure system. Each element as pertaining to the Surry Power Station structural analyses is discussed below. 4.2.3.1 Free Field Ground Motion Specification of the free field ground motion includes specifying its f requency characteristics , spatial variation, and control point. The frequency characteristics and spatial variation of the free field motion were discussed above. The elevation at which the free field is specified for each - structure is the control point. Generally, this would be the existing free _ surface at elevation 27.5 f t for the Surry power Station Site. 'However, some of the structures analyzed and founded at elevation-2 ft_are surrounded by other structures on all sides founded at the same or deeper elevation. Thus the control point was conservatively defined as -2 f t for the Auxiliary Building, and Control Room- Structure, while for other structures the control point was specified at elevation 27,5 ft. 4.2.3.2 Foundation Input Motion The foundation input motion varies from the free field motion for all cases except surface founded foundations, i.e., control point and foundation at the same elevation, subjected to vertically propagating shear ' and dilatational waves. A scattering function relates the three translational free-field components to the six degrees of freedom on the foundation. The scattering function is frequency dependent and complex valued. For this investigation all waves are assumed to be vertically propagating, The variation between free field motion and foundation motion- is due to the variation of free field motion with depth and wave -< scattering at the soll foundation interface for embedded foundations. This follows since the foundation is modeled as rigid and massless, and points -- on the foundation are constrained to move according to its geometry in plan and depth of embodment. 4.2.3.3 Foundation Impedances, Foundation impedances are the force-displacement characteristics of the soil. Foundation impedances depend on the soil layering and soil material behavior, frequency of excitation, and geometry and embedment of the ' foundation. For a rigid foundation,- the force displacement
-characteristics _are uniquely defined by a 6 x 6 matrix, complexed valued and' frequency dependent, relating a resultant set of forces and moments to - the six rigid body _ degrees of freedom of the foundation. The foundations . of all' structures- analyzed here are approximated - as an equivalent surface-founded or embedded cylinder. The soil _ column is idealized as . a half-space with properties taken at an elevation of half the characteristic length below its foundation. Depth of embedmont for each structure's foundation model is given in Table 4.6 below.
4-16
Table 4.6 Surry Power Station Foundation Models Depth of Embedment Embedment (ft) Reactor Building 67 Auxiliary Building 0 Control Room Structure 0 Safeguards Area 19-Containment Spray Pump Enclosure 19.5 Emergency _ Generator Enclosure 13 Intake Structure 36.75 4.2.3.4 Structural Lynamic Characteristics. Structural dynamic characteristics are described by their _ fixed base
-eigensystem and modal damping factors. Eigensystema, fundamental modes of vibration .and eigenvectors are determined from fixed base lumped mast- ' beam element models. The beam elements represent stiffness between floor . levels located at the' shear centroid of the reinforced concrete walls or diagonal steel bracing, including shear deformation. The contribution to lumped mass at each floor level is from the half height of the wall above and below, floor slab, and equipment at that floor. Nominal values of structure damping were taken to- be 0.07, 0.085, and 0.10 (fractions of critical damping) for the three seismic acceleration ranges considered here. These were based on published damping values and assumed stress 1cvels achieved.
4.2.3.5 Soil Structure Interaction Analysis.
- The probabilistic SSI analysis- procedure is to perform a series of deterministic analyses,- each simulating an earthquake occurrence, including variability -in seismic - input, soil-structure interaction, and-structure representation. The seismic input' variability is normally introduced by -considering _an- ensembic of carthquake motions. For this l study, the five earthquake motions described earlier were used. A series l, of ten earthquake simulations for each acceleration range were performed each using the identical free-field input motion as a starting point.
Soil structure interaction and structure response -- va riability are introduced through a limited number of parameters -- soil shear modulus, l 4-17 i
soil damping, structure frequency and s t ruc tu ra l modal damping. Variability in SSI was incor, orated through modelling soll shear modulus and soil material dampin,. as random variables with lognormal dist ributions w; th medians co rresponding to the nominal values of Table 4.4 and coef f .clents of variation of 0.4 and 0.5, respectively. Variability in otructure dynamic behavior was also modelled by t.reating structure f requ(ncy and modal damping as random variables, parameter variations in tach step of t.h e response analysis were selected to represent random variability, and not to include modelling uncertainty . The assumed paramater variability corresponds to that developed in the SSMRp (Ref. 10). The parameter values for each of the ten simulations were celected fr(m the probability distributions by dividing the distributions into equally probable segments, sampling from each segment and combining the samples using a Latin hypercube experimental design. The responses calculated from the simulations are combined to estimate median responses conditional on the occurrence of an earthquake described by a particular ha::ard curve parameter, e.g. , peak ground acceleration. Instructure spectra were calculated at 5 percent damping at the mass centroid of each floor elevation translational component for the ten input motions. The ten spectra were then combined to form median centered spectra assuming a lognormal diotribution, The structures for which best-estimate dynamic responses were computed based on the 10 selected time histories were shown in Table-4.6. Each structure considered is described below. Reactor Buildinn Internal Structure. The reactor building is a reinforced concrete structure, circular in plan (68 ft radius) supported on a 10-f t-thick reinforced concrete mat at clovation 39.5 f t and which extends in height to elevation 95 ft. The foundation supports two independent structures, the containment shell and the internal structure, coupled only at the base. Figure 4.7 shows the 3-D fixed base model used to calculate the nominal eigenvalues and eigenvectors of the reactor building. Both the containment shell and internal structure are represented in the model. The internal structure's fundamental mode of vibration has a frequency of
~
6.48 llz in the E W direction and 6.89 lir. in the N-S direction accounting for 86 percent and 89 percent,- respectively, of the mass participating in the horizontal directions. Auxiliary Building. The auxiliary building is a reinforced concrete structure up to elevation 27.5 f t and steel frame to elevation 66 ft, It is rectangular in plan (150 ft x 111 - f t) supported on a 4 ft thick reinforced concrete mat.at elevation 2 ft. The auxillary building la surrounded on all sides by other structures founded at the same elevation or deeper. Figure 4.8 shows the 3-D fixed base model used to calculate the nominal eigen values and vectors of the auxiliary building. The auxiliary building fundamental mode of vibration has a frequency of 20.8 Itz in the 4-18
O Mall LOCATION s innra RIC1b ELtntWT SLM ELtMINT EL.186*9'-..() E.L. 140'1' .. () n.1:4 's ... ( > EL.10B'?'...() EL. sa's ... (p II -- EL '5'n' EL.61'11'...()
... EL. 46'3' EL. 31'5'-..(>
2
$*..EL.16'3' L EL. C ' 3 0' -.. ()
[ -. EL. 4'S' d i y EL. 29'7' ... . .. . , y EATE MM. INTE M A:. O Figure 4.7. Surry Power Station Reactor Building Structural Model 4-19
g MAS 5 LOCATION mr rs R202 D E.LLKIN: BEAM Ell'trWT
-. EL. 66'C'
-- EL. 45'1C'
-. EL. 2'1'6*
o' I i b EL. 13'C'- % _ Y
/// fill x
Figure 4,8. Surry Power Station Auxiliary Building Structural Model l l 4-20 l l l l l
E-W direction and 22.1 ilz in the N-S direction accounting for 80 percent j and 65 percent, respectively, of the mass participating in the horizontal l directions. The structural steel frame on the upper elevations has fundamental modes at approximately 6 ilz in both directions. Control Room Structure. The control room resides inside the service building as a separate structure isolated by expansion joints resting on an independent foundation mat. It is constructed of reinforced concrete to elevation 45 ft and structural steel frame to 77 ft. The service building foundation in the control room area is tied into the adj acent turbine building strip footing running in the E-V direction. In addition the service and turbine building share the lateral force resisting system of a structural steel frame above elevation 45 ft. Stiffness and mass contributions from the turbine building are incorporated in the structural model to the extent structural details and load paths dictate. The control room area is rectangular in plan (185 ft x 75.5 ft) founded on a 4-f t thick reinforced concrete mat at elevation -2 f t, surrounded on all sides by other structures founded at the same elevation, Pigure 4.9 shows the 3 D fixed base model used to calculate the nominal eigenvalues and eigenvectors of the control room structure, j Contributions to the stiffness betwoon floor 1cvels and mass from the turbine buildin6 were incorporated in the model to the extent necessary as dictated by lateral force resisting systems and load paths by the single bent shown in Figure 4.9. The control room structure fundamental ' mode of vibration has a frequency of 17.0 llz in the E W direction and 22.6 11z in the N-S direction accounting for 53 percent and 47 percent, respectively, of the mass participating in the horizontal directions. The s truc tural steel frames of the- upper elevations have fundamental modes between 1.8 and 6 Itz,
' Safeguard Area. The safeguards area is of irregular shape, a segment of a circular are conforming to the circular plan of the reactor building containment shell. It was idealized as a rectangular structure. The safeguard building is 68 f t x 14 f t, founded on a 2.5-f t thick reinforced concrete mat at elevation 10.0'ft extending in height to elevation 42.5 ft- Reinforced concrete shear walls and diaphragms transmit lateral-loads to its base. The roof is a steel frame metal deck and concrete slab.
Figure 4.10 shows the 3 D fixed base model used to calculate the nominal eigenvalues and eigenvectors of the safeguards area. The safeguards building significant modes of vibration - are 34 Itz accounting for 56 percent of tl e mass in a direction tangential to the containment shell. In the orthogonal direction (towards the center of the reactor building) the first significant mode of vibration is 21.5 llz with 57 percent of the mass participating. Containment Sorav Pumn Enclosure. The containment spray pump- enclosure is-of irregular shape. It was idealized as a rectangular structure. The containment spray pump enclosure is 38 ft x 30 ft founded on a 2.5-f t-thick reinforced concrete mat at elevation 9.0 ft, extending in 4-21
1
-j
)
$ MS5 LOCATION
_ Racio Ez.E m 7 - n. :: 'e-SEAM ELEMENT
-* EL. 17'6'
-- EL. 5,'6*
I-*EL.45'3' (I TVMINE SVILOING EL. 37'C' =* CONTROL ROOM STRUOTUAI (I q .. u , 5 ... n . s's - p l ()-- EL. 45'3'
/g L
* ,/\
g1 j 1 1 Figure 4.9. Surry Power Station Control Room Structure Structural Model 4 22
O m is to:x: es
..... P3C!t' EttxtN; BtM ELtKIN;
.. EL. 42'4' EL. ,,... .. - ~
EL. 19'C' ... Lt N EL. 10'9' ... ,u ss;g; X Y l l l [ Figure 4.10. Surry Power Station Safeguards Area Structural Hodel 4 23
~ - . - _ _ . - _
)
i- l 1 I i l l 4 l height to elevation 42,5 ft. Reinforced concrete shear walls and 4 diaphragms transmit lateral loads to its base. The roof is a steel frame
- metal deck and concrete slab.
i
- Figure 4.11 shows the 3 D fixed base model used to calculate the nominal
! eigen values and vectors. The containment spray pumps enclosure i significant modes of vibration are at 10.82 and 21.0 llz accounting for 20 i and 52 percent of the mass in a NS direction In the orthogonal 1 direction, E V the first sign
- ficant mode of vibration is 21.6 ilz with 43 percent of the mass participating.
J- fErgtDIV Generator Enclosure. The emergency generator enclosure is a i rectangular structure 64-ft x 110 ft founded on a perimeter strip footing
- ranging in elevation from 3.5 ft to 20.5 ft. Reinforced concrete shear walls and diaphragms transmit lateral loads to its base.
The emergency generator enclosure foundation was modeled as a rigid, massless and embedded equivalent circular plate at an average footing r j clevation 13 ft in an idealized half space. The soll strain compatible properties represent those at a depth of half the characteristic length (c1 - 47.ft) below the foundation clovation. I Recognized sa a very stiff structure relative to the soil and since the only required response is at the free field elevation (27.5 ft), the entire mass properties of the structure (all six degreen of freedom) were
- calculated about the foundation reference point. A single rigid massless
! element translates the response to the desired nievation of 27.5 ft. Figure 4.12 shows the 3 D SSI model of the emergency ;;enerator enclosure. Intake Structure. The intake structure is a rectangular structure 74 ft x 180 ft founded on a reinforced concrete mat 3 ft thick at elevation 9.25 ft. Reinforced concrete shear walls and diaphragms transmit lateral loads to its base. F Recognized as a very stiff structure relative to the-soil and since the l only required response is at the free field elevation (27.5 ft) the structural model is simplified. The entire mass properties of the structure, all six degrees of freedom, were calculated about the foundation reference point. A single rigid, massless element translates the response to: the desired elevation of 27,5 f t. Figure 4.13 shows the 3+D SSI model of-the intake structure. Response Results. For each of these structures, the dynamic structural response - for each of the' ten suites of time histories was computed at - l each of the three earthquake excitation levels. From the computed time history responses at the different floor levels, response spectra were generated. As exampics of the output, the computed response spectra for the 2 SSE acceleration range for each structure are shown in Figures 4.14
- through .4. 20. In each figure, spectra in the E W, N-S and Vert
!: ' directions are showr, Each spectra plot has s.everal building elevations l l~ 4 24 !.., _ . . . _ . - _ _ _ . _ _ _ . . , _ . . - _ . . , . . _ . _ . . . . . _ _ . ~ . . . _ . _ . . _ _ _ , . . _ _ , - , . . . . __ -
O m,$$ LCCATION l tru.~r pggge gLt e s-3LA.P. ILD.tv7 l
.. EL. 52'C'
! N,
.. Et. 2('(*
\
N X Y ,ss,3,s .. EL. 10'3' Figure 4.11. Surry Power Station Contaircent Spray Pun;r Enclosure Stuctural Model 4 25
l l enemme con:utt roc Attes ucio n. ten I. '.,I a u i i t 4 b
<6$ % p/
s,,*, l Tigure 4.12. Surry Power Station Emergency Generator Enclosure OSI Model 4 26
d 1 i J i-e e n 4 4 i mean con:Ar;g rouwoA;20N -I ---- k2stD ELsxtx; i i i-monumammune h i Figure 4.13.- Surry Power Station Intake Structure SSI Model 4 27 __..-....__..-_a___. ,_.c.~., .. . . . _ - . _ . . _ _ . . _ _ .- . . _ . . . ._.,.-_;__.,.,_,-~-__-_,_,..._,-...-....
L A R 99 0 g,y. , g, O N-S 3e.0 30.0 E5.9- - 25.0- - ! C C 29.9-- :meec.0--
= =
l C L i, .e.15.7- -
- 15.O--
i e = g 19.9--
/k
%gg l 4 10.0--
S.9-- - -,- , _ S . O-- q%( __ l
- .;
- : :-x : : : : - - --- - --
.c : - ::::-- -
0 0 :--- - - - - ::e I si' n' .' i.* i6' i.' i.' ..' Freguener Mel 8'reguency 941 ! R 99 * - tic.i s4.0 t.egend-r- L 'F *' o> free-field , i
, , , ,, fnd, ref. pt. __.__ _ _ _. _ _ _
3
,, t.c.*. et -3*6' . _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ . i e
78.9- - t.o.s. el 18*4' . __ ; t
= s.9-- %tes: ,
- 4. 0--
AtI spectre tattuteteo at St demolog Atteteratten tritarttts or fl#s/s : t . 9- -
.o - -
- :- = - : - - - - - - - - --
ee-
-t t to 19 se to 2
Freguentf P+tl l Figure 4.14a. Surry Nuclear Power Plant Reactor Contairsent Internal Building Instructure Response Acceleration Range 2
. i N l
- e e Ili _r l l- 1l P, j '
I t O N m [ r . l 8 rs 1 , O e.e
. I I m
~
/ F e :: ll W A : . z ,a
, e f .
s .
, -u -
# . y . -
~
e
'/ >{',t u
Is = 5 e E
.s es
-c& e
%/ . . .T . u ok M !.!8L.
E4F .. _5 cu 3 E e . m en W *** 3 0 u u W W W L lh g, '
- b. =I 0b ta . * .. D 0 ft 4./
C e. r. r g r= g u ft
*E
.* p p
g . w Ep4 6 6 6 6 6 e = .e o. o. s .o -,a c #.O << v v-s t0
= o. o. - *R R R C
- r t'
= uoussele23v EE cc c uL cv ft 0 E
L w ' i 7 u N N N r :: 2 . E ;; u - ww g m 4 p . ., O C e=e M
. lli .-
l e,
. ge
. ., .c a
. I w-1; e ; O
.. = w s'.'
# > ]
, =w m sc
[
.ji : . & :: E
.- gp h s
.,, .. ., .d
' *
- 0 b 5
e s !.!' u U m y d' . l w . 3 ..
= . . . . -
e o. 6 6 6 4 6 o. 's e o. 6 6 6 6 6 6 e. 'E
% W OW W w uottesete 3, x uettesetessy l
i 4 29 l l . , - - - . - . . . . _ .
e 0 8
-_ g
_ n
- i
-_ w e
_. d _ _. T s _ S/ s _
- 0 l _ t / _
a t 1 H e
-_ F
. e f o _
t y e t i n _ e ti a . r e l s
- u ut q ci e ' ' l n _
r 06 au 0 F c E 37 7 g 1 D1 2 a o n g rt i
- g - t l t t d2 S - , ' ee ca l i e
- g *
" ..: p e e~ ug
- s B n r_
- s. s. se n sl a N ?
e g - e yR g g t I c r
,. d" o c. o t c an t
- e. r ' t c N A A i o d l i 8 C.
0 S 0 S 0 0 o- o. t i t xa o0 5 0 5 ur t 3 E P 1 1 A e x C3u***~ l t e nc a c l A P e rs e n wo op P s 0 2 e e 1 4 0 rR . 4 e 4 2 a e _ 4 e ee l r t
?
cu ut e _ : N c yr u s - rt r s un
- '9 1 : 'o l SI g
- 1 t
- : t r
- i -
i .
- l
- G a y 5
- 0 f 1 ts .
- =
e T 4 w e g - u e q r e e u r r g r S I l 0
'es F i F
a 7
- c W : i t
- r ;
e E : V :
' 0 p 9
9 9 9
'e 59 : 0 o.
s.
- o. 0
'6
. . . . e. 1 s
e0 3 S 5 9 5 9e 2 os s , e s3 3 P 1 1 1 t 1 x c3 ,,Wt*e g N e* *Le .,o* e.uO
i g gg 8 gf d le N-S E 30 = so.: ts.e .' & ES.0-- ) - g - - s ro.s- - _. ee .e..
- e
- 13.8--
\
o 35.9-- -
=
u . ._
=
Y 10.0--
\ %- . , _ .
% 19.0-- N_ g _ __
3.0-- s.*-
.o -
e e -r - - - : r- - -
- - ee
- : : ---- - - - em e 16 8
to' to' to 16 to' to' 10 rey Mel Freesency f542) x to ' fertical t.egend p u._. 12.#-- tree-ftold rnd. rer. pt. _ _ _ _ _ _ _ _ e ,, , ,_ t.o.s. el 45*10' _ _ _ _ _ _ _ _ _ _ . 3,
= e . 0--
t.o.s. el 66*0' . . .. L s . s-- Dm ::: :-_- Note 3: u All spectre calculated et 52 de* ping 4 . 0-- Acceleration units in ft/s/s t .0--
.o : - :-- r : : : : : :--
e::e
-t e e to te le: 10 Yms Mel Figure 4.15b. Surry Nuc1 car Power Plant Auxil'ary Building Instructure Response Acceleration Range 2
h' ; , ei ; l ,i 2 _ ._ _
- 0 __ _
g 1 __ n
- i _
__ . p
- . m e
- _ d
- _ _ % s _
. 5/ _
_ _ . s _ 1 _ t / _ n 9 1 et 9 4 5 __ f
- 1 d -
y e n
- t i c e
- n l s e ut g
g ct e O r r '6'03
' l n a o c
e r u
=o t t 97 5 n t
+
p 24 e o c
- d rt u 0 t .t i t t t r2 S t e 'e eee ce 0
S e
- l e r g 0 r- r . . . : p e e n r
1 s sl oa e . s. s. s. N - e e e oR g R edno. r o. o. t o l c l c n t o e_ F r t t t N A A l o ot o0 0 o. 5 e. o 0 0 0 c. t r t t a 5 0 3 nr t 3 2 r 1 1 cI = E= = o2 oe x Cl e t c nc l aA P s e rs e n wo op 2 o r P s
- o e
- : t e f rR
- a ee
_ l r _ : - - cu ut N c
- yr u
rt rs un Q : 8
- 91 l f
t i
= o..t SI
- : a t , ,
- : 6 T 1 e.
'v-r m
e r s : e 4 e r r9 9 F 3 l
=
'. re t i F
u g
~
e : t
- c I
t : r : _ : e : e E : f :
- - - t -
- - - o -
a p 8 0 E. o. 0 9 o. t ' : 8 e.
- e. 0 S o os s
. o_ t o0 t 3 E r 5
9 1 5 t t 2 1 , s a e x e: a.L= = u Y x ,S=L* .=u*4 a e.w"
I i i . h I
. e
$e i
l . l i,i l
- C l l1 4 i l e
'l ." I I p p t 1 0%
/l l
I l e
- us
,# Y '$%
$ lt @U s' / 2 C e
( -, o e p ., >
- ai
- p 3 w
., y p
- e " f 4
- 6. MC R 3 w
,,O L **
- W D
> *3 W CD 5 C U V
$ D ON 9 0
.- u u N k4 O W h &W V 4 0 L Mg i
.. - 0 OO g c e
O h L * * ..
- C l eV e O
- oe a- k & * *
- O e> c gg 1 O O O O O u *'a W ct I' 0 L C C " W C
< n n , a n ,,) s w y a 2 < < & C O, 0, 0, 0, 6, O,
(( SR R C E
- C 0 R
x venesiteens 6 .tO u U C U ft *C N 4 we
& C 3 O C Ci fv N C. m O 0 i
i l I l
.O
- D l
1
- W gg g
- a o e 1, ? l'
~ w
- O 3 i
y I lll I I a (' ~ 4 EE x3 WW
.J l I w m
- em D C l
, ,j .
, a ., a .
F ',,s : * : , 5 W- r ., t . y ~ f.' 4 (' . , N
% C *-
t
.oD w t e e
(
.O g- -
4,, m : u ., 3 .,
"a" ,
j a> L ., s W LU a* N D. . . . O. O U. . . , O,
- 2% 2 S * *
- N x vettesetenst x votieseteen, 4 33
x te ' g-w < to ' 9 -s so.O so.c ms.e- - 25 . *-- c L c n_ Sro.o s to.e-- ~. m -> f~.
. .=
- h. *
- hY. '\
= es.o.- = 15.0--
3 3 to.*- 4 ____ Y to.o-- k\
%g __ l
- 3. 0-- 5 . o-- ,
,e . . . . . . . - . . . . . . . . . . .;,,, ,g._ . . . . . . . . . .
-t e t 2 -t 9 1 2 I to 19 te to to lo to to Freguoncy M41 frequew y (Hel a to ' vertical l
14.0 Legend: l 7
$ 12.e-~ Y free-ftold
" ** * ~ ~ ~ ~ ~ ~ ~ ~ ~
c 19.e-- 3 t.o.s. el 19'O* --
=
L s .e-- t.o.s. el 27*5* _
- I
. t.o.s. el 42*5* __ __ ___
= s.e--
u
= Notes:
4 . 0-- All spectre calculeted at 5% demoing
- r. 0-- Acceleration in units of ft/s/s 1 1
4444-2 16' to' te' to l Fwp Mel , I i i i rigure 4.17. Sorry truclear Power Plant Safeguards Building Instructure Responses Acceleration Range 2
I 1 h l 1 8 I F l1 i li ., I i .e-
! p l
g 'i T%
\l ., yE l 1 8 eaw fl
- I 4 a
.[ !.!.
- g ll e-o
/. '/ l - Ee a ,. / .. s a a-f s .. .e.,
J l' .- 3
~ c [
e, . . . L 60 0 # * =c O . u n
}l$g to sD N P =
0
. nNr 4 0*
+ 2 L E
., w e .= .a w y th ., aw ec y e A
.. - 0 D L 4N e
a 9 i g .. ,, e np r-t {g e ,, e 7
& W . e. e. 0 g V U % D D G ~ U an c gg b s n * '
T 4, L C d * ** N N
+*
g ** *u e a n, 6, e, 4, o, o, 6, O
~~
[ E 9 0 SR R R v ucite.stestv y" cw u-4 U v cv n< 5Vw k V l 8 o
- I, (o " hk sm
*{ ..
I ,, we 81 ( n
, g1 a' .- g, g a .. ., - w
;I f ..
E3 2 y
,I r - -
sE wa [ !!E% !!!T $E
,b :: E :: a 5, -
e ., ., p .> .6 s s (* .- .> f at ., .- 4 e a w e 6 w l!I , ! 2 E
- e .. w
.. u .,
* ., =
e au .-
>g .,
o in 6 6 6 6 6 an 'E e en 4 6 6 6 6 6* an '* EN' N N N N * 'N N N N
- w uttleselssty x Wettesetessy L. 35
l'i lf f; l t i 2 ot g n i ws n/ _ os/
% t 5f g t ol ._
t F a o
- t r _
k - W f d s
^ et
- y t i e
t . a t f r
- n o q ep t
S l u e i s
- t g 6 c n l i o
e a l f - c eo'fr e c s n
- r7 t r E t
2 so r
- d n et o
- l ol t t S - ot e cn t a7
; t t ce r d f e . : oe ce n -d s. s sl ng N -
e en e e e n g t t C a e u o.
- e. fr f ot t
o t c y R t o N a a
~ - -
1 cn
. &. O.
0 O. &. . t nn ej
- e0 3 0 3 9 3 gt s3 2 2 1 1 r a a C3==*l* ="o< cr El ec r
e t c nc aA l P s _ e rs en 2 o 2 o wo op e t t
- P s
- ~ -
- rR e
- a
- l cc r
et e
- - l f
ut e r _- N yr t i rt g M ot r - et r s
- t - t r un u
k : f u e - f Si y - g c 9 0 - n e 1 w
- e s 4 e e r o e e M 9 o re r t
s t l : 3 n
- e :
i F
- t :
s g t g M : E
- - - t * , - - - - t
- - - - o * -
9
; g 0 9 9 ;t
- 0 &. 1
- o. - O. &.
e 0 3 9 S 94 E 8 & 2 90 h 93 2 3 1 11 S K c3 ,.mOe ="w. 1 3 ,. = g h I e u*
i l P
,t. '
t_. ...' u-s j 30.0 30.0 l l I es.o.- 25.o-- 1 i e e 1 2 es.0-.- 3 Po.o-- g l *
= W h s.o.-
i 6 T i t s .e-v [A%
. 3 3 to.e-- # to.o- -
N s l - -
- 5. 0-- 5. o--
7 ,3 2 2 16 8 to' to' to 16' to' to' 10 Frequency fetal rreewe,cr f%rl
,to ' Weeticet n l t.cgenc:
h tr.01 Fr ce -F teId
\
1 foundat ion cef. pt. _ _ _ _ _ _ _ . c so.e--
\ t.o.s. o f 27 - 6 _ _ _ _ _ _ _ _ _ _ _ _ _ .
- T
= e
- e--
- t
* \\ Notes:
. N
= e.e-- ( e t t spectre calculated at 51 anoo t,q
~
$ acceterations in untts of f t/ s /s
- 4. 0--
- t. 0--
1
,, .....3 16 89 99 10 Fregue,cr fHst l l
Figure 4.20. Surry Nuclear Power Plant Intake Structure ; Instructure Responses Acceleration Range 2 , t
L i i i corresponding to maj or floor slabs). Similar spectra are given in i Appendix A for the other acceleration ranges. Taken together, the
- spectra at the three different acceleration ranges provide all the response input needed, i
l 4.2.4 Safety Related Component Responscs i l The in structure spectra presented and discussed in the previous section are used to determine safety-related component response. Assuming that the dynamic characteristics of a given component can be represented by a [ single dominant mode of vibration, the component response can be j approximated by the spectral acceleration of the approprinto in structure spectra at the frequency of the dominant mode. Thus, at each structural location, numerical response values at different i
- frequencies or frequency ranges are computed directly from these spectra, l These ranges span the probable natural frequencies of the components housed at that location. The median zero period acceleration response is
' calculated from the ten values given by the probabilistic response analysis assuming a lognormal distribution. The median response over a frequency range is over the range from the median spectra given by the ten earthquake simulations. Given the natural frequency of the component of interest, the appropriate frequency interval and component response is then defined. Numerical values of the median component responses for the 1
!- _three IcVels of ground motion (1 SSE, 2 SSE, and 3 SSE) taken from these spectra are presented in Appendix B.
~
4.2.4.1. Responses in Terms of peak Ground Acceleration The - responses in Appendix B are given at three peak ground acceleration values (0.15g, 0.30g, and 0.45g). One could directly interpolate between these ' three _ values to obtain any specified response at any arbitrary value of peak ground acceleration. However, a more direct approach which greatly simplifies computation of the component failure probabilities is to compute the average ratio between the median PGA and the median response spectral acceleration at each specified component location. Figures 4.21 through 4.25 are plots of the response location accelerations in each building (at various building elevations) versus:PGA. It can be seen that a linear relation exists up to - free field accelerations of 0.4g or greater. Furthermore, for. those curves which show significant non linearity at higher acceler. ation. levels, the linear relation provides a conservative estimate of the i local response. From these figures, ratios between the various responses and PGA were determined, as listed in Table 4.7. (No_to that not all responses listed in Appendix B; are -included on this table, as not all floor _ slabs supported critical components -identified on the seismic fault trees.) Using these response amplification ratios, the local spectral , acceleration' response at any floor level of any of the buildings can be computed at any pga level. 4-38
- -. - - - _ .-. . , _ , . . _ - - - . - - . . . , _ - _ _ , - - - , . - - - . - , ~ _ . - . -
I l l i f Spectral Acceleration (g) i 1 ,
-- q ;
~- E l. 2'-O' 2 PA 9 El. -2'-O' 7 Hz
/
/
-(+- E l . - 2 '- 0
- 2 - 6 H z /
0.8 -
/. '
,/
/
/
0.6 - 7': i
/
/
/,4 ,/
0.4 - I /,/ /'/s
/ / \
h' v'
/'
,/
,/
/ ,/
/l/
/
/
/
0.2 -
/ /
g//
/
. ep 05 '
O 0.1 0.2 0.3 0,4 0.5 Peak Ground Acceleration (c) t l l Figure 4.01a. Auxiliary Building Median Responses 4 39
j i l l i 1 Spectral Acceleration (g) - 1 l El 13'-O' Z PA
~3- El 13'-O' 6-10 H z a
0.8 - i 1 I I 0.6 - 1
/
/i !
/ ~i
,./ /
i 0.4 -
/ ,A
/
R
/
/
/ / /,/
/
0.2 - o/7,/
- ./
/
V 0'8,/-
/
O 0.1 0.2 0.3 0.4 0.5 Peak Ground Acceleration (g) Figure 4.21b. Auxiliary Building Median Responses 4 40
., - - . . _ . . . . , - . - . . - . - - _ - . . _ . - . _ _ . . _ . . _ _ , . - . . _ _ . . , . - . , . _ , . - - - _ - . _ _ _ , _ , . _ _ ~ , , - . _ _ . , . . - - . _ , , . - . - - . . ,
Spectral Acceleration (g) 1
~- E l. 9'-6' Z PA
-E- E l. 9'-6' 10 H z El. 9'-6' 6-10 Hz 0.8 -
O El. 9'-6' 7 Hz 0.6 --
/
/
,K/,
//'/Y 0.4 -
f/If'/ l
// -C l
h
/
0.2 -
,-/- j OCI O 0.1 0.2 0.3 0.4 0.5 Peak Ground Acceleration (g)
Figure 4.22 Control Room Structure Median Respotises ( 4 41
. . - . . ~. . ._.-_ . _ . _ ~ - - - . ..- - . _ _ . - . - - . . _ .
Spectral Acceleration (g) 1 ,
~*-
El. 27'-6' Z PA , El 27'-6* 5-10 Hz l i 0.8 - l 0,6 - l 1
,/ i
/ ,/.~.. j h 0.4 - / ,' /s i g.
/
5', l 0.2 -
/ // '/
/ /,
t/ ' ' ' h 01 2 ' O 0.1 0.2 0.3 0.4 0.5 Peak Ground Acceleration (g) Figure 4. 23. Emergency Generator Enclosure W.dian Responses 4-42 i.
1 1 1 Spectral Acceleration (g) 1 ,
-+- El. 2 8'-6* Z PA ,
i i El. 28'-6' 7 Hz 4 0.8 r i 1 I 0.6 - : i
/?
i 7'
- j. C
/
,/
/ , /l 0.4 -
p/-
/ /
l
/ [ ,/r ^
0.2 07
/
/
77, 03 ' ' ' ' - J 0 0,1 0.2 0.3 0.4 0.5 Peak Ground Acceleration (g) Figure 4,24. Safeguards Area Median Responses 4 43
-l i . , . , .= ,.. , . , . - - - - . - . , , . . , . - , , , , . . _ . . . . . - , . . . - . . . - , - - . , -
l Spectral Acceleration (g) 1
~
El.18' 4' ZPA !
-9 El.18' 4' 7 Hz I
0.8 - l 0.6 - l-1 0.4 - t p/ i-
/
/
/ /
O.2 -
/
/ /
OL
' ' d 0 0.1 0.2 0.3 0.4 0.5 Peak Ground Acceleration (g)
Figure 4.25. Reactor Building Median Responses .. l 4 44 -l 1
~
i Table 4.7 Surry Seismic Response locations 1 Renponse Multiple Hynber location Elevation Frecuency of PGA 1 _ Yard (PGA) ZPA 1.0 2 Service Building (53) 9'6" ZPA 1.0 3 5 10 1.2 4 " - " 7 1.3 5 10 1.1
- 6. " "
27' 0" ZPA 0.9 , 5 10 1.2 7 8 Aux, Building (AB) *2' 0" ZPA 1.0 9 * " " 5 1.5* 10 " " " 7 1.3 ! 11 Safeguards Area (S0) 27' 6" ZPA 0.9 l 12 " " " 7 1.2 13 Turbine Building (TB) 15' 0" ZPA 1.0+ 14 Cable Vault / Tunnel (CVT) 15' 0" 5 10 1.2+ 15 Reactor Building (RB) 18' 4" ZPA 0.5 16 a " " 7 0.6 17 " " 27'-7"++ 7 0.8
*Estiinated from 2 5 and 7 li values. Not used. +Used Auxiliary Building at elevation 13' 0". -i ++Used RB 47' 4". ;
i 4-45
)
a
4.2.4.2 Variabilit.y in Response Variability in responses (floor and spectral accelerations) was assigned based on SSMRp results (Ref. 10). Confidence bounds were computed for the final core damage probabilit.ics using both random (irreducible) and systematic (modeling) uncertainties. The uncertainties (expressed as standard deviations of the logarithms of the responses) are shown below: Ounntity Random Systematic Peak Ground Acceleration 0.25 See cross-Floor Zero Period Acceleration 0.35 reference ) Floor Spectral Acceleration 0.45 table in Appendix C 4.2.4.3 Correlation-1 In computing the probability of cut sets involving correlated component failures, it is necessary to consider correlations both in the responses and in the frag 111 tier. of each pair of components in the cut set. Once this is dono, the correlation coef ficient be tween any two component failures is computed from the expression OR10R2 0F10F2
#~ #
#R1R2
- F1F2 2 2 2 2 2 2 2 2 O 0 F1 OR2
- OF2 O 0 F1 OR2
- OF2
, RI , , R1 ,
in which p - correlation coefficient between the failure of components 1 and 2
~ pas, OR2 - standard deviation of the logarithms of t.he responses of components 1 and 2 pri, Ar2 - standard deviations of the logarithms of the
< fragilities of components 1 and 2 FR1R2 - correlation coefficient between responses of components 1 and 2 [ l prira - correlation coefficient between the fragilities of components 1 and 2-
~
This relation shows that the correlation between the failure of any two components depends not only on the correlations between the respective responses and the respective fragilities, but also on the variances in the responses ans' fragilities. L 4 46 ..--_ _._-_ _._-__..__ _ ___ _ .__-..._.-_ __ __ _,_ ..,....._- . ~ . . . . , . _ . _ . - , _ . - , , _ - . . , , . .
_ _ . . _ - __ ~ I i l i. j With the correlation between the failure events in the cut set known, the evaluation of the cut set probability is performed by evaluating the t- multivariate probability distribution for the cut set. Methods for j evaluating such correlated cut s e t.s are described in Reference 3 of Chapter 1 of this report. j The pairwise correlations between the responses are assigned according to 4 the rules on Table 4.8. Using the rules given and the definitions of the i responses given on Table 4,7, the response correlation matrix shown in
! Table 4,9 results.
l l Inasmuch as there are no data as yet which prove or disprove correlation i between fragilities, the fragility correlations between both like and
- -unlike components were taken as zero.
In general, there exists some degree of correintion between any two components excited by the same earthquake by virtue of the common ground rno tio n , lloweve r , it is not necessary to compute correlated failure probab t.11 ties when the degree of correlation between the failure events j is small (e.g...less than 0.25) as the result will be very close to the u uncorrelated value. By examining the response and (in genert1) the t fragility correlations, it is possible to identify those pairs of l components for which correlation effects may be neg1ceted, and those for which correlation must be considered, In general, it is found that correlation between like components (identical components which are ! sensitive to the same spectral acceleratiot0 in the same location should always considered as they are usually the most si6nificant. Ilowever,
- while - correlations between two unlike components can (in principle) exist, these are usually of lesser significance, and can usually be neglected, especially when dealing with components located on different floors of a building or in seperate buildings.
For Surry, a review of the response correlation tabic in conjunction with the fact that fragility correlations are taken as zero allowed screening i of the components for those dif fering components which might be assi6ned correlation. For unlike components, it was found that only correlation between the - RWST and the CST had any potential significance, -By contrast, a number of identical components in the same location were found to be significantly correlated. These components are listed below: 4
- a. 4 kV busses
- b. 125 volt busses
- c. diesel generators
- d. PCS motor driven pumps ;
- e. Pilot operated relief valves i
For these - components, the correlation coefficient was computed and a
- proper evaluation of the correlated pairs of failures occurring in the various cut sets was made during quantification of the accident sequences.
l ( 4 47 ! l i
i 2. i i 4 1 1 ! I
- Tabic 4.8 j Rules for Assigning Response Correlation ruu
] 1. Components on the same floor slab, and sensitive to the En.c spectral i frequency range (i.e., ZPA,- 5 to 10 liz, or 10 to 15 liz) will be ! assigned response correlation - 1.0, 2 1 2. Components on the same floor slab, sensitive to different
- - ranges of spectral acceleration will be assigned response correlation - 0;5.
] i ! 3. Components on different floor slabs (but in the same building) and sensitive to the Eng spectral frequency l range (ZPA,- 5 to 10 ilz or 10 to 15 ilz) will be assigned response l correlation - 0.75. L
! 4. Components on the ground surface (outside tanks, etc.) _
L shall be treated as if they were on the grado floor of an adjacent building.
- 5. ' Ganged" valve configurations (either parallel or series) will have response correlation - 1.0.
- 6. All other configurations will have response correlation equal to zero.
1. e i i .; i L t 1
,4 48 ti' n" '"'i= ==gl'Q. e---es+ d'rWre5w4e.,.,9-g>.yL. ..eny - p,., u-,--4y:, y grey.., ,.;_m p.,,,,j
_ py..qq.g, ,,,u. , ,,4, _j j , f pa ng y qf %,.g y y..y ,,- ,, ,g,, p _ , , , _ ,
7 5 1 1 7 6 5 1 1 5 1 1 4 1 1 3 1 _ 1 _ 2 5 1 1 y r r u 1 1 S 1 s e 0 5 1 - s 1 n o . p s e 9 5 1 R n e e 9 w 8 1 t 4 e B e l s b t 7 5 5 1 a n 7 T i e c i f 6 5 1 . f 7 e o . C - n 5 5 5 5 1 o i o t r - _ a e . , l . c z nw e 4 5 5 1 i , r r s o r t n h , o e o s C m i m t e 3 5 1 y a s S o l e h s r t i l r _ o t 2 1 e c p _ l e b l c a x T l A e ~ 1 1 0 1 2 3 4 5 6 7 1 2 3 4 5 6 7 8 9 1 1 1 1 1 1 1 1 oI$
;l !# ;[!! !il!! -
1 !iji j li!i! i!. ;Iai. ll .j. 1I 1 a j1 al, ;
l q. i l i I 4.3 Seismic Fragiljti n Component failure is taken as either loss of pressure boundary integrity or loss of operability. Failure (fragility) is characterized by a cumu. lative distribution function which describes the probability that failure j has occurred given a value of loading. Loading may be described by loen1 3 spectral acceleration or moment, depending on the component and failure mode. The fragilities are related to the appropriate local response to 7 permit an accurate assessment of the effects of common cause seismic j failures in the evaluation of the accident sequences. I 4.3 1 Generic Fragilities F A generic data base of fragility functions for seismically induced failures was developed in the SSMRP (Ref. 10). At a first step, all components were grouped into generic categories. for example, all motor operated valves located on piping with diameters between 2 1/2 and 8 inches were placed 1 sto a single generic category, and similarly, all motor control centers were placed into another generic category. Fragility functions for the generic categories were developed based on n [ combination of experimental data, design analysis reports, and an j extensive expert opinion survey. The experimental data utilized in de-i veloping fragility curves were obtained from the results of component manufacturer's qualification tests, independent testing lab failure data and data obtained f rom the extensive U.S. Corps of Engineers SAFEGUARD Subsystem liardness Assurance Program. These data were critically . examined for applicability and then statistically combined with the expert opinion survey data to produce the fragility curves for the SSMRP . generic component categories reported in Reference 10. Finally, a review of more recent site specific component fragilities i contained in the Lawrence Livermore data base (Ref. 11) was made. Based on these reviews, several of the SSMRP generic fragilities in Reference 10 were updated. The final generic categories and the corresponding. fragility medians and uncertainties are shown in Tables 4,10 and 4.1" These fragilities are used as the starting point in the simplified ic PRA. As in the use. of any generic data base, one must be cognizani c the source of the data and the .cquipment to which it applies. An important aspect of using this data is to examine the equipment in the plant being analyzed and compare - it with the data base for which the generic fragilities were developed,
- Any deviation is noted and examined carefully, and new site specific fragilities developed as necessary.
4 Fragilities for electrical components represent a special problem in that there is a wide variety of electrical gear found within a plant. Typi-cally, all this gear is enclosed in switchgear cabinets or motor control . - conters. The two lowest failure modes that were identified in the SSMRp fragility data base were relay chatter and inadvertent trip of circuit l 4-50 l l
- _ . ., m . . __ ._ .__ .._ _ _
t Table ~4.10 Generic Component Categories i Fragility Category Comporent Clast: Typical Components Frecuency (Hz) t I 1 IDSP Ceraric Insulators ZPA 2 Relays 5-10 ! 3 circuit Breakers 5-10 4 Batteries .ZPA 5 Battery Rachs ZPA 6 Inverters 5-10 7 Transformers 4KV to 480V and 480 to 120V 10 8 Motor Control t ariters Control for ESF Pumps and valves 5-10 9 Aux.~ Relay Cabinets . 5-10 7 10 Switchgear (Inc. Transformers.
.416V and 480V 5-10 'y Buses and Breakers)
ZPA 11 Cable Trays -[ 5-10 12 Control Panels and Racks RPS Process Control 13 Local Instruments Misc. Pressure and Temperature 5-35 Sensors l 4160 AC Emergency Power Uaits 22 14 Diesel Generators 15 Horizontal Motors Motor-Generator Sets ZPA 16 Motor-Driven Pumps and AFUS, RHR, SIS, Charging Pumps, 7 Con: pressors Lube Oil Pumps, Diesel Starting , Compressors 17 Large Vertical, Centrifugal Servict Mater Pumps 5 Pumps (Motor-Drive) 18 Large Motor-Operated Valves (>10") ZPA 19 Small Motor-Operated Valves (<10") ZPA 20 Large Pneumatic / Hydraulic Valves Includes MSIf, ADP, and PORV ZPA 21 Large Check and Relief Valves ZPA , 22 Miscellaneous Small Valves (<8") ZPA
b i Table 4.10 Generic. Component Categories (Concluded) i Fragility . Catecorv Component Class Tvolcal Components Frecuency (Hz) 23 Large Horizontal Vessels and Pressurizer Relief Tank, f W 'ZPA Heat Exchangers Heat Exchangers 24 Small to Medium Heat Exchangers Boron Injection Tank 20 and Vessels 25 Large Vertical S.torage Vessels RHR Heat Exchanger, Accumulator ZPA with Formed Heads Tank , 26 Large Vertical Flat-Bottomed CST, RUST , Storage Tanks 27 Air Handling Units Containment Fan Coolers S w I w L _ . . , _ ._ m. . _
Table 4.11 Generic Component' Fragilities, in units of gravity (g) Catecorv Generic Component Median
- 1 Ceramic Insulators 0.25 0.25 0.25 2 Relays 4.00 0.48 0 75 3 Circuit Breakers 7.63 0.48 0.74 4 Batteries 0.80 0.40 0.39 5 Battery Racks 2.29 0.31 0.39 6 Inverters 2.00 0.26 0.35 !
7 Dry Transformers 8.80 0.28 0.30 8 Motor Control Centers 7.63 0.48 0.74 9 Auxiliary Relay Cabinets 7.63 0.48 0.74 10 Switchgear 6.43 0.29 0.66 , d., 11 Cable Trays 2.2.3 0.34 0.19 u 12 Control Panels and Racks 11.50 0.48 0.74 13 Local Instruments 7.68 0.20 0.35 14 Diesel Generators 1.00 0.25 0.31 15 Horizontal Motors I?.10 0.27 0.31 16 Motor-driven Pumps and Compressors 2.80 0.25 0.27 17 Large Vertical Centrifugal Pumps 2.21 0.22 0.32 18 Large Motor-Operated Valves (>10 in.) 6.50 0.26 0.60 19 Small Motor-Operated Valves (<10 in.) 4.83 0.26 0.35 20 Large Pneumatic / Hydraulic Valves 6.50 0.26 0.35 21 Large Relief, Manual, and Check Valves 8.90 0.20 0.35 22 Miscellaneous Small Valves 12.50 0.33 0.43 23 Large Horizontal Vessels and Heat Exchangers 3.0 0.30 0.53 24 Small to Medium Vessels and Heat Exchangers 1.84 0.25 0.45 25 Large Vertical Vessels with Formed Heads 1.46 0.20 0.35 26 Large Vertical Tanks with Flat Bottoms 0.45 0.25 0.29 27 Air Handling Units 6.90 0.27 0.61
*All medians in . terms of spectral acceleration at 5% damping and for frequency (or frequency range) shown.on Table 4.10.
n - random uncertainty B.e - systematic uncertainty
- - .~. - . - - .-- - . - - . - - - -.. .- -.- .~-.- - -.- - breakers. Virtually all ' the electrical switchgear and motor control conters in a nuclear: power plant include these two types of. components, so these two fragilities were used as - the generic failure modes for electrical gear in the SSMRP analysis. Relay chatter is the lowest failure mode and, if included blindly in a risk analysis, would be the dominant failure. Because, in many cases, circuits are protected by time delay circuits and because, in most cases, chatter of relays would not cause a change in the state of a system being controlled, the NUREC 1150 analyses chose not to include relay chatter as a failure modo for electrical gear but rt.ther to include circuit breaker trip as the lowest functional failure mode. 4.3.2 Surry Site Specific Component Fragilities During the; initial plant visit, the following components were identified as requiring plant specific fragility derivations: 1, RWST and CST Tanks
- 2. CCW llent Exchangers
- 3. 4Kv Busses 4 Diesel Generator Load D' -bution Cabineta
- 5. 480 MCC Cabinets in Cab: Vault Tunnel The RWST and CST tanks were identified both because of their height to-diameter ratios as well as due to the time period during which they were designed and installed. The CCW heat exchangers were mounted on concrece pedestals with relatively few anchor bolts. The 4kV busses and the diesel generator cabinet were anchored with relatively small welds. The 480 MCC cabinets in the cable vault / tunnel area were evidently bolted to a concrete mounting pad, however, due to the very high aspect ratio, it was suspected that the bolting might be marginal.
Based on'these observations, site specific fragilities were developed for
-the above mentioned items. The resulting component site specific fragilities are summarized on Table 4.12.
It should be pointed out that all the components above were found to have ' median failure 'acceleratior. levels well pbove the - SSE. However,- they did
-have less margin of safety above the _SSE than the other components examined during the plant visit, and, hence, were anticipated to be sig-nificant contributors to the accident sequence probabilities.
4.3.3 Site Specific Building Fragilities 4.3.3.1 Method of Fragility Evaluation The-fragilities of Surry Unit 1 structures were generated using the basic
- methodology described in Reference 12, with certain modifications. The fragility of a structure can be expressed in terms of its peak ground acceleration capacity, A, as follows:
4-54
. ,. . . - , - - , - . . . . _ . - . _ _ . . _ . . _, _-._ _ _- _. _ . _ - - . . . . - . _ .~ _ - _
-l l4 Table 4.12 Summary of Surry Site Specific Fragility Functions i
Median Base f Acceleration a l Component Failure Mod.g at Failure ! RWST Buckling with 0.46g 0.34 Anchor Bolt Yiciding CST: Buckling with 0.45g* 0.35* ,
' Anchor Bolt Yielding 'CCW llTX Support Failure 0.29g 0.30 Diese'l Load Center Weld 0,76g 0.25 Generators Anchorage Failure 480 MCOs Anchorage Failure 0.70g 0.25 '(BAC 1111 2, BAC 1J1 2)
!
- No CST drawings were located. Used value from generic data base, which was consistent'vith that computed for the RWST.
i 4-55 l l t
. - ._- -_- - . - _ ~ . - .. - = . . . - . - . - . - - - -. .
A=A c R U In this formulation, A, is the median peak ground acceleration (PGA) capacity, and eg and cu are random variables with unit median, repre-senting the-inherent randomness about the median and the uncertainty in the median value. The variables eg and eu are assumed to be lognormally distributed with logarithmic standard deviations a and fu, respectively. The properties of the lognormal distribution are presented in Reference 10. For convenience, the median peak ground acceleration capacity, A., was formulated as the p ro duc '. of the SSE peak ground acceleration, Asst " 0.15g for Surry site, and a median factor of safety against this ground motion level, F . Thus, the median peak ground acceleration capacity can be expressed as:
^m~ mASSE The median . factor of safety, F., was in turn expressed as the product of the following two median factors of safety:
- a. The median strength factor, F,, which is defined as the ratio of the median structure strength to the median structure loads for the SSE ground motion input
- b. The median inelastic energy absorption factor, Fu, which accounts for the ability of the structure to withstand seismic loads in excess of those corresponding to yield through ductile, nonlinear response.
The strength - and inelastic energy absorption factors have associated lo6arithmic standard deviations, , and u. . From the properties of the lognormal distribution, the logarithmic standard deviation associated with the total factor of safety is calculated as follows: p 2,
'2g S
,g U2 ' 1/2 6 s
-These variabilities are composed of randomness and uncertainty, which are defined'as follows:
- a. Randomness consists of variabilities that cannot be reduced by more detailed evaluation or data collection.
- b. Uncertainty consists of variabilities resulting from lack of L knowledge.
The only source of random variability reported in this section results from the effect of certain earthquake characteristics on the structure i i 4-56 I
inelastic energy absorption capability. Uncertainties result from varia-bles such as material _ strength, member capacity, member ductility, etc. Structure seismic response contributes additional variabili'> to the structural fragilities. Logarithmic standard deviations fc_ seismic response variability are not included in the values reported in t.h i s section as they are included in the responses directly, 4.3.3.2 Development of Structural Capacities The Surry structural fragilities were expressed in terms of factors which account for structure ultimate strength and inelastic energy absorption capability. The basic techniques used to determine the median values and-associated variabilities of the terms were essentially those described in Reference 11, with certain modifications, Structure Element Ultimate Strengths Two major exsiderations are involved in the determination of the ulti-mate strengths of individual structural elements, one is the definition of the strengths of the materials composing the members, The other is the determination of the ultimate strength capacities of the structural members given the type - of loading, - materini strength, member configu-ration, etc. The Surry plant specific material strength data were not available. The following values, which were used in the fragility evaluation, were esti-mated based upon data from other nuclear power plants (Ref. 12): Concrete Compressive Strength Minimum Sneci fied (psi) tiedlan (psi) o 3000 4900 0.17 4000 6000 0,15 Steel Reinforcement Yield Strength Grade . Median (ksI) M 40 48 0,10 50 55 0.10 60 69 0.07 The Grade 50 steel was used' for #L4S and #18S reinforcement in the Surry fragility analyses. 4-57
l 1 1 The median ultimate strength capacities of the structural elements were found using the median material strengths and member configurations (i.e., geometry, reinforcement, etc.) in conj unc t ion with available predictive formulation or approaches. The approaches and formulations used were those appropriate for the type of element (i.e., shear wall, reinforced concrete - cylinder, etc.) and loading (i.e., shear, flexure, etc.). They were typically found to provide essentially median-centered capacities when compared to the results of available experimental testing. For example, the predictive equations used to determine the median ultimate strengths of the Surry shear walls subjected to in plano shear and flexure are presented in Reference 13. Median strength factors, F, were calculated for individual structural elements as follows: V um,1 p , sm V SSE,1 V a ,g - Median ultimate strer.gth for element i Vsst,t - Median load due to SSE ground motion input for element i The median strength factor for a structure was generally taken to be the lowest value of the individual elements composing its primary seismic load resisting system. This is slightly conservative if the structural elements are ductile and redundant. In certain cases, load redis-tribution among such structural elements was censidered when -determining the structure strength factor. Variability of the structural element ultimate strengths was considered to be composed of uncertainty since it is associated with a lack of knowledge. Uncertainty attributed to caterial strength was based upon the estimated variabilities listed above. Comparisons of the predicted strength capacities to the available test results' provided estimates of the uncertainty in the predictive strength formulations. . Additional uncertainty attributable to variabilities associated with other sources, such as member geometry, reinforcement spacing, openings, workmanship, differences between field and laboratory conditions, accuracy of the predicted load distributions, etc., were also included. Structure Inelastic Enercy Absorption The ability of a structure to withstand seismic levels in excess of those corresponding to yield through_ ductile, nonlinear response was accounted for by the inelastic energy absorption factor, Fu . This factor was based upon the Riddell Newmark response deamplification factor, du (Ref. 14). 4-58
. .- . -. . .= .~ .-
The median inelastic energy absorption factor, F., corresponding to some ductility ratio, p, is given by the following equation: S,, F - au S.. - Median clastic spectral acceleration for median structure damping at the dominant structure frequency S.o - Deamplified spectral acceleration at the dominant structure frequency For frequencies in the amplified acceleration range (between about'1.8 Hz and 3 liz) of the Surry median ground response spectrum: S au - fu S.>'Sadr a
- 4. - (pp.' q)"
p-q+1 q - 3,0 p-0 30 r - 0.48_ p-0 08
- System damping For frequencies in the amplified velocity region (less than 1.8 itz), the q and r terms-in the above equations are defined as:
l q . 2,7 p-o 40 r - 0.66 p-0 04 For frequencies greater than the frequency at which the median spectral acceleration returns to the peak ground acceleration (about 15 liz); 0 S.u.== S wr - f D PGA
'PGA Peak-ground acceleration.
The Riddell-Newmark response deamplification factor was based upon a series .of nonlinear analyses utilizing single degree-of freedom (SDOF) ( fixed base models subj ec t to time histories of large magnitude, long l duration earthquakes. Nonlinear response of the Surry structures would I be expected'to differ from the response calculated using these deamplifi. cation factors . for the following reasons: l u l l 4-59
.. - ~ - .- .- . _ _ . - - ._
l
- a. The Surry structures are founded on reintively sof t soil. As a result, significant soil-structure interaction (SSI) is expected,
- b. The Surry structures are typically multi-degree-of-freedom (MDOF) systems,
- c. Small magnitude eartnquakes are expected for the Surry site.
To account for these differences, an effective ductility, p., was used in the. equations above. The system ductility, p,y , for use with the Riddell Newmark deamplifi-cation factor is a measure of the nonlinearity threughout the structure, r For fixed-base SDOF structures, the system ductility is equal to the story drif t ductility, p t. However, for MDOF structures, the system ductility may be less than the story ductility if the ratio of the story demand to story capacity is not uniform through the structure, Also, i nonlinear behavior has less effect on structures with significant SSI effects as compared to fixed base structures for the following reasons:
- a. Structure nonlinearity causes only slight f requency shif t in system : nodes dominated by soil flexibility,
- b. Increased damping due to hysteretic behavior is small compared to soil. radiation damping.
In the fragility evaluation, the system ductility, ,y , was reduced from the story ductility. From the study of Reference 15, the system ductil-ity, py. can be related to the story ductility by a factor M.
#st -1 ,1 878 , M ,
The median story ductility for typical nuclear plant shear walls is esti-mated to be about five, Values for the-t' actor M were estimated on a case by case basis depending on the. extent of the soil structure interaction-effects and the distribucion of structure nonlinearities. For the con-tainment spray pump enclosure and the safeguards area, an M value of 4,5 was estimated. For the containment.and the concrete internal structures where soil structure interaction is more significant and localized non-linearity is expected, M values of 6 and 7 were estimated, respectively. The service building, auxiliary building, emergency generator enclosure, and intake structure all essentially behave as rigid structures on flexi-ble soil. This conclusion is based on the observation that there is littic or no amplification of the foundation level input motion through-out the height of these structures. For these buildings, S o was.calcu-lated using the equation for the rigid frequency range (>15 Hz) along with the median story ductility. j L i 4 60 j
t The Riddell Newmark response deamplification factors were based only on large magnitude earthquakes. It is well known that lower magnitude carthquakes are not as damaging to structures and equipment as higher magnitude carthquakes with the same peak ground acceleration (Ref. 15). The lower magnitude earthquakes have lower energy content and shorter durations which develop fewer strong response cycles. Structures are able to withstand larger deformations (i.e., higher ductility) for a few cycles compared to the larger number of cycles resulting from longer duration events. Earthquake magnitude effects were accounted for by using an effective ductility, p., in the Riddell-Newmark response deamplification factor approach. The effective ductility was calculated as follows: p, - 1. 0 + Co (p,y, - 1.0) where the duration coefficient, Co, is a function of the earthquake magnitude and p,y, is the previously defined system ductility. The results of the analyses performed in Reference 15 were used to pro-vide estimates of the duration coef ficient, Co, as.a function of earth- 7 quake magnitude. For earthquakes having magnitudes ranging from 4.5 to ) 6.0, a duration coef ficient of 1.4 was determined to be appropriate by correlating the inelastic energy absorption factor from the Riddell-Newmark formulation to the results of Reference 15. Similarly, a duration coefficient of 0.7 was estimated for earthquake magnitudes in the 6.5 to 7.5 range. A duration coefficient of 1.3 was estimated for the Surry structures. This is a representative value for castern United States nuclear plants. It should be noted that, for purposes of this study, structures are con-sidered to fail functionally when inelastic deformations of the structure under seismic load are estimated to be sufficient to potentially inter-fore with the operability of safety-related equipment attached to the structure. The element and system ductility limits chosen for structures are estimated to correspond to the onset of significant structural damage. For many potential modes of failure, this is believed to repre-sent a conservative bound on the level of inelastic structural defor-mation which might interfere with the operability of components housed within the structure. It is important to note that considerably greater margins of safety against structural collapse are believed to exist --for these structures than many cases reported within this study. Thus, the structural element capacities reported herein should not be inferred as corresponding to structure collapse. 4.3.4 Structure Fragilities Derived for Surry Fragilities for the Surry structures are listed in Table 4.13. In general, several potential failure modes were investigated for each structure. Fragilities for the governing failure modes are reported. These failure modes are typically associated with structural failure which would result in damage to the safety-related equipment located in the building. 4-61
Table 4:13 Surry Structural Fragilities Summary , Critical Failure Consequence of Structure Mode & & & Failure Incorooration in PRA Containment Shear failure 7.7g .09 .24 Loss of liner Not included since Building near the base integrity and negligible loss of reactor coolant pressure boundary Concrete Shear failure 1 8g
. .14 .27 Ioss of lateral Results in RVR internal at the base support of steam initiating event > generators and a coolant pumps ands w loss of primary coolant pressure boundary Safeguards Shear wall 1.5g- .06 .23 Damage to equip- No initiator results.
Building failure ment throughout LPI, HPR, LPR the structure Impact damage 1.6g .26 .31 Loss' of' anchorage Fails LPI, HPR to slab due of low head to sliding safety injection pumps mounted on the slab Spray Pump Shear wall 2.lg .06 .23 Damage to equip- No initiator. Fails Enclosure failure ment throughout CSS and OSR. the structure Sliding 1.8g .26 .30 Potential damage Induced damage to components to the slab housed in the enclosure
l Table 4.13 Surry Structural Fragilities Summary (Concluded) Critical Failure Consequence of Mode fu Failure Incorporation in PFA Structure b da Service Shear wall 1.7g .05 .24 Damage to equip- Causes T1 (IESP) Building failure ment throughout initiator and fails all the structure electrical (AC & DC) systems. R.. Its in sequence T RQLD 1 2-EGE Shear wall 4.2g .05 .21 Damage to equip- Negligible, so not failure ment mounted on included. the wall or roof. Probably no I e damage to the 4 diesel generators j u Intake Failure of 2.0g .05 .24 Damage to equip- No initiator results. Structure guide wall ment throughout Fails pumps which fill the struct fe. canal. Does not fail , canal. Would affect only after 12 hrs. Sliding 1.7g .33 .35 iamage to the service water pipes and other lifelines pene-trating outer walls. Auxiliary Shear wall 1.8g .05 .23 Damage to equip- Results in seal IDCA. 3 Building failure ment throughout Does not cause LOSP. - the structure. Fatis HPI, CCU (D1, D2, D3, W) systems. Notes: ,
- 1. Median capacities are calculated by multiplying the factor of safety A of the critical failure mode by 0.15g free field peak ground acceleration.
- 2. a and Bu reported are variabilities associated with capacity only except those reported for sliding which include variabilities of both capacity and response.
I p In developing the capacity factors, structural wall and beam re sul tant forces were deterrained from the dynamic response inodels, The building's structural dynamic characteristics are described by their fixed base eigensystem and modal damping factors, Eigensystems, fundarnental modes of vibration and eigenvectors, are determined from fixed base lumped mass beam element models. Beam elements represent stiffness between floor 1cvels located at the shear centroid of the reinforced concrete walls or dia,r,onal steel bracing, including shear deformation. The contribution to lur< ped mass at each floor level is f rom the half height of the wall above ar.d below, floor slab, and equipment at the floor. National values of rtructure damping were taken to be 0.07, 0,085, and 0.10 (fractions of critical damping) for the three seismic acceleration ranges considered here, These were based on published damping values and assumed stress levels achieved. Failure modes for each teructure are described below.
.Qg.ntainment and Internal Structures The containment structure is a reinforced concrete structure consisting of a circular cylindrical wall capped by a hemispherical dome. The con-taitunent wall is supported by a basemat founded on soil. The bottom of the basemat is at elevation (-)39 ft-7 in. A continuously operating drainage system is provided to keep the groundwater below the top of the basernat such that the hydrostatic pressure is not significant. Principal dimensions of the containment structure are:
Mat Radius 71 ft-4 5/8 in. Thickness 10 ft-0 in. Liner plate thickness 3/8 in, Cylinder Inside radius 63 ft 0 in, Wall thickness 4 ft-6 in. Liner plate thickness 3/8 in, lleight to springline 122 ft-1 in. Dome Inside radius 63 f t-0 in. Wall thickness 2 ft 6 in. Liner plate thickness 3/81n. Concrete with a design compressive strength of 3000 psi at 28 days - was used to construct the wall. Crade 50 n18S reinforcing bars with e mini.- mum specified yield strength of '50 ksi were provided in the meridional and hoop directions. Additional two layers of #185 diagonal reinforce-ment were provided in the cylindrical wall .to resist horizontal seismic shear force. Both the flexural and shear strengths were evaluated for the containment structure, The controlling fai?ure mode was found to be shear failure of
-the cylindrical vall near the base. tiocizontal shear forces due to scis-mic response of the conta inme r.t structure introduce tangential- shear a
stress in the wall. The median shear strength was determined using an 4-64
empirical equation derived from testing of scale model prestressed and reinforced concrete containment structures. Resistance to horizontal seismic shear force is provided by the concrete, meridional and hoop reinforcement and diagonal reinforcement. This mode of failure was found to have_a median PGA capacity of 7.7g. Loss of liner integrity and loss of reactor coolant pressure boundary will result. The concrete internal structure of the Surry containment structure consists of the primary shield wall, cylindrical crane wall, concrete floor slabs and refueling pool. The internal structure provides biologi-cal shielding and missile protection and also supports major components such as RPV, coolant pumps, etc. The main lateral load carrying elements of the internal structure are the crane wall and the primary shield wall. , These structures are founded on the basemat common with the containment I structure. Cadwelds were used to provide continuity of vertical wall reinforcing steel of these structures across the basemat liner plate. I Dimensions of the crane wall and the primary shield wall are: Crane wall Outer radius 53 ft-0 in. Thickness 2 ft 9 in, lie ight 124 ft-5 in. Primary shield wall Inner radius 11 ft-0 in. Thickness 4 ft.6 in. lleight 47 ft-11 in. A review of the internal structure indicates that failure due to seismic response will probably occur toward the base of the structure. Near the base, the crane wall is perforated by several large openings that result in a series of wall segments, typically 2 f t 0 in, thick by 8 f t-0 in.
- wide spanning from the top of the basement at elevation (-)29 ft 7 in, to the slab at elevation (-)3 ft 9 in.
Failure of the concrete internal structure was found to be governed by shear at the base. Shear yiciding is expected to occur first at those wall segments near the base of the crane wall. The primary s1* eld wall was found to have higher capacity than the crane wall. Since the primary shield wall and the crane wall are structurally tied together by - the floor slabs at elevation (-)3 ft-6 in., elevation 18 ft-4 in, and elevation 47 f t 4 in, and by radial walls , some load redistribution is expected to occur after ductile yielding of the crane wall. Additional load can be resisted by the primary shield wall. The median PGA capacity of the concrete internal structure- accounting for this load redistribution was found to be 1.8g. Failure of the concrete internal structure will result in loss of lateral support for the steam generators, coolant pumps and RPV, and loss of primary coolant pressure boundary. Safeguards Area The safeguards area is a reinforced concrete enclosure located outside of the contr!nment structure with planar dimensions of about 17 ft in the 4-65 __ . . ~ . _ . __ .____. _
radial direction of the containment structure by about 70 ft long in the circumferential airection. The structure is founded on a 2 ft 6 in.- thick basemat at elevation 12 ft 0 in, with a total height of about 30 ft. This area is enclosed on three sides by reinforced concrete walls and by the - containment structure shell on the fourth side. The safeguards area is separated from the containment structure with a 3-in. gap throughout its height. Safety-related equipment in this enclosure include the containment recirculation spray pumps and low head safety injection pumps. The controlling failure modes of the safeguards area were found to be concrete shear wall failure and structure sliding-induced failure. Both failure modes would occur in the short direction of this enclosure, i.e., radial direction of the containment structure. There are fewer concrete shear walls in this direction to resist the lateral force. Also, due to the backfill outside of the long wall, both static and dynamic lateral carth pressures are present in this direction. The governing shear wall was found to be the 1-f t wall between the safeguards area and the spray pump enclosure. The failure mode of this wall is governed by flexure. The median PGA capacity of this mode of failure was determined to be 1,5g. The potential consequence of this failure mode is damage to equipment throughout the safeguards area. The second controlling failure mode was found to be sliding towards the containment structure. Resistance to sliding is primarily provided by friction at the base of the safeguards structure. No buoyancy force was considered as the ground water table is about 10 ft below the basemat. The median capacity for sliding was based on Newmark's approach (Ref. 16). 1 ,ause structural backfill is present only at one side of the i safegu .ds area, and causes relatively significant earth pressure, Newer.rk's sliding equation for the unsymmetric resistance case was used as shown below: 2 u - 1 V A,y m ,2.5, 2gN ,N .. l' N - Coefficient of friction A - Peak ground acceleration V _ Peak ground velocity u, - Structure displacement The 2.5 median factor of safety associated with this equation was de-termined based on the data given in Reference 16. Should the structure slide 3 in, and impact the containment structure, there is a possibility of concrete spalling with subsequent damage to anchorage of the low head safety injection pumps anchored close to the edge of the - slab. The . sliding failure mode was found to have a median PCA capacity of about I 1.6g. l 4-66 1
Containment Sprav Pump Enclosure The containment spray pump enclosure is located outside of the con-tainment structure and houses safety related e qui preen t such as the containment spray pumps and emergency auxiliary feedwater pumps. The building is enclosed by an "L" shaped reinforced concrete wall on two sides, by the containment structure shell on the third side and the main steam valve enclosure wall on the fourth side. The spray pump enclosure is founded on a 2-ft 6 in, concrete basemat at elevation 11 ft 6 in. Similar to the safeguards area, this enclosure is separated from all adjacent structures by a 3-in, gap throughout its height. Two controlling failure modes were identified for the spray pump on-closure: Concrete shear wall flexural failure and structure sliding induced failure, The median PGA capacity of shear _ wall failure was found to be 2.lg. Torsional response of this enclosure was found to be sig-nificant due to the unsymmetric "L" shaped layout of the major shear walls and was considered in the evaluation. Similar to the safeguards area, the unsymmetric sliding equation was used to evaluate the median sliding capacity. Upon closing of the 3 in, gap after initiation of sliding, impact between the pump enclosure and the containment structure would occur with subsequent potential damage to the containment spray pumps and the auxiliary feodwater pumps. The median PGA capacity of the slidin6 failure mode was determined to be 1.8g. Service Building The service building consists of a reinforced concrete substructure from 2 ft-0 in, up to elevation 45 ft-3 in, and a structural steel mperstructure abova elevation 45 f t 3 in. The areas in the service building which house safety related systems or equipment are the control room at elevation 27 f t 0 in. and the switchgear and battery rooms at elevation 9 f t 6 in. All safety-related equipment are enclosed in the reinforced concrete substructure. It is judged that the -- failure of- the steel superstructure will not damage the safety related equipment. Thus, the fragility evaluation of the service building is focused on the reinforced concrete substructure. The seismic induced . lateral forces are resisted by the typical 2 ft-thick reinforced concrete shear walls and concrete floor diaphragms. The L governing failure mode was found to be the shear wall failure in flexure in the transverse (N S) direction. The median PGA capacity was deter-mined to be-1.7g. Damage to the safety related equipment throughout the service building is expected as a result . of this failure mode. In the longitudinal direction (E-W), the service building was found to have much more higher capacity. Emergency Generator Enclosure The emergency generator enclosure (EGE) is a single story reinforced concrete structure and houses the four emergency diesel generators and related equipment. The ECE structure consists of concrete roof slab and 4-67 l
._. _. __ ~-- -
load bearing concrete shear walls. The exterior walls are founded on strip footings at different elevations. The interior partitioning walls are founded on strip footings near grade. Each diesel generator is sup-ported on its own mat near grade which is separated from the EGE structure. The controlling shear wall failure in the transverse (N-S) direction of the ECE was found to have a median pCA capacity of 4g. Damage to equip-ment mounted on the walls is expected as a result of this failure mode. 1 For exterior walls where significant backfill are present, the effect of both lateral earth pressure and hydrostatic pressure were considered in the wall capacity evatuation. The exterior walls were found to have higher PGA capacities. The roof diaphragm of the EGE was evaluated and was found to have higher capacity than the controlling shear wall. Auxiliary Buildinn .) 1 The auxiliary building is composed of a reinforced concrete substructure below elevation 27 f t 0 in, and a structural steel superstructure above elevation 27 f t 0 in. The top of the reinforced concrete foundation basemat is at elevation 2 ft. Numerous concrete walls and columns are present throughout the substructure of the auxiliary building. Most of the safety-related equipment are located in the concrete substructure. The superstructure consists of a metal roof deck at elevation 66 f t, an 8 in. concrete slab on metal deck at elevation 45 ft 10 in, and vertical
- braced frames. The seismic capacity of the steel superstructure was not evaluated. All safety related equipment located above elevation 27 ft are enclosed by three separate reinforced concrete enclosures and should not be damaged by failure of the superstructure.
A number - of shear walls and diaphragms were evaluated. The controlling
- failure mode -was found to be failure of the cast west oriented reinforced concrete shear walls at the center core of the auxiliary building bounded by Column Lines H, K, 8, and 10. The median PGA acceleration capacity of this' failure mode was found to be 1.8g with inelastic load redistribution among these center core walls considered. Failure of these walls is - expected to lead to equipment damage throughout the auxiliary building.
The floor diaphragms were found to have higher capacities. Intake Structure l The intake structure is a reinforced concrete structure founded on a basemat bearing on the soil at approximately elevation (-)26 ft. Plan dimensions of the structure are approximately 177 f t in the north south by 64 ft in the east west._ The-reinforced concrete oil and pump storage room, which houses the safety related service water pumps, is supported on the operating floor of the intake structure at elevation 12 ft 0 in. ! The major lateral force resisting system consists of concrete shear walls and slabs. Both _ structural failure modo and sliding induced failure mode were evaluated. Sliding was considered as a potential failure mode due to l 4-68 l r 4 . . -
lack of keyways at the basemat and foundation soil interf ace. The intake structure is backfilled on the north, south and west sides with the east side open to the water. Thus, sliding in the eastward direction was evaluated. Resistance to sliding is provided by the static friction between the basemat and the foundation soil. The normal water level was assumed at elevation 0 in. Reduction of the static friction resistance due to buoyancy force at the bottom face of the basemat was considered. The median capacity for sliding was determined using Newmark's equation for symmetric resistance (Ref.16) as given below:
~2 I'
"m ~ , ,
gN , , The median factor of safety associated with this equation was estimated to be 2.0. The equation for symmetric resistance was used in consideration of the massiveness of the intake structure. The sliding displacement at which damage to the service water lines is expected was estimated at three in. This criteria is based on the line configuration, the backfill depth above the lines outside of the intake structure and the line anchorage at the intake structure outer wall. The median PGA capacity for sliding induced failure was found to be 1.78 A number of the shear walls and the diaphragms of the intake structure were evaluated. In addition to the seismic inertial loads, forces due to both static and dynamic effects of the backfill and the water inside and outside of the s t ruc ture were considered. The controlling structure failure mode of the intake structure was found to be the ficxural-failure of interior guide walls with a medial PGA capacity of 2.0g. Failure of these walls is expected to lead to damage of service water pumps. Masonry Block Walls
-The reevaluation effort on I&E Bulletin No. 80-11 activities at Surry Power Station (Ref. 17) identified all safety-related (Class I and II) masonry block walls in the Category .I structures of the Surry Power Station. Class I masonry walls are defined as those walls located in areas with high probability of impacting a significant amount of safety-related equipment if wall failure resulted. -Class II masonry walls are those with limited safety related equipment in its proximity. Some of these walls such as the ones in the pump and oil storage room of the intake structure were modified as a result of the reevaluation effort (Ref. 17), These modified walls were j udged to have high capacities.
_0ther walls were found acceptable without any modification necessary. The fragility evaluation os Surry masonry block wall is limited to these unmodified walls. The 8-in, block wall in the control room of the service building was judged to have the lowest seismic capacity. This wall separates the control room and the computer room at elevation 27 f t-0 in. The wall was constructed with the lightweight C90 masonry units using Type N mortar. The wall spans vertically with a span of about 16 ft high between the l 1 4-69 l
floors at olevation 27 ft and elevation 45 ft 3 in. It was assumed that the top joint of the wall is mortared into the overhead slab based on review of the available drawings of similar walls in the plant. The , seismic capacity of the wall was determined assuming the wall can develop arching action. The median capacity of this wall governed by - the compressive stress of the masonary unit was found to be about 3.5g. Failure of this wall is expected to damage equipment -in the control and computer rooms. 4.3.5 Liquefaction An analysis for the potential of soils liquefaction at the Surry site was made by CooMatrix, Inc. as contained in Appendix D. Their analysis showed that some liquefaction would be expected at peak ground acceleration values of 0.3g to 0.4g. The effect of this liquefaction would be relative displacements between the containment and other important safety buildings (Auxiliary Building, Safeguards Area, Service Building and Turbine Building) of approximately 2 to 4 inches, llowever, this displacement is limited by the depth of the. liquifying layers. A site examination of the piping systems and cable penetrations going from these buildings into the containment indicated that such displacements were not likely to cause failure. Hence, liquefaction, while it is to be expected at earthquake . levels above the SSE, is not expected to affect the plant. Thus, liquefaction was not included explicitly in the Surry base case seismic PRA results 4.4 Core Damage and Risk Corroutations In the event of an earthquake or any other abnormal condition in a nuclear power plant, the plant safety systems act to bring the plant to a safe shutdown condition. In this . step of the risk analysis process, we identify the possible paths that a nuclear plant would follow, given that an earthquake related event has occurred which causes shutdown. These paths involve an initiating event and a success or failure designation for systems affecting the course of events, and are referred to as acci-dent sequences. 4.4.1 Initiating Events The seismic analysis . performed for Surry is based on the same set of event trees developed for the internal event analyses of the plant. The initiating events considered are:
- a. Reactor Vessel Rupture (ECCS ineffective)
- b. Large LOCA
- c. Medium LOCA I
- d. Small LOCA
- e. Transient Type 1 (PCS failed by initiator)
- f. Transient Type 3 (PCS initially available) l l
4-70 i
The reactor vessel rupture RVR and large LOCA (ALOCA) were computed baued on the failure of the supports of the steam [,enerators and reactor cool-ant pumps. Specific values for support fragility were taken from the SSMRP analysis of the Zion plant, however, a review of fragilities of other plants as contained in Reference 11 showed that the values used were typical. Surry is a 3 loop plant, and hence, the definition of the RVR event is the simultaneous failure of at least one steam generator or reactor coolant pump in at least two of the loops. Similarly the defi-nition of the large LOCA is a failure of at least one steam r,enerator or one reactor coolant pump in any one of the three loops. Since these failures are due to the same floor response and their fragilities are expected to be highly correlated, it was necessary to do an exact evalu-ation of these failure events explicitly including all correlation. In particular, it was necessary to include correlation between cut sets (combinations of component failures) as well as correlation between the failure events in each cut set. This wss accomplished by performing a Monte Carlo evaluation of the Boolean equations describing the RVR and ALOCA events. This resulted in the failure probability distributions shown in Figure 4,26. The independent variable in these figures is the concrete internals response at 7 Hertz computed for Surry. This failure distribution was satisfactorily fit in log normal form and input as a component for the analysis , The small and medium LOCA initiating events were computed based on the failure of appropriately sized piping in the reactor coolant loop. These distributions were generated from the calculations of piping failures for all the pipes considered in the SSMRP Zion analysis. These distributions are shown in Figure 4.27, The independent variable for this figure is peak ground acceleration, with a random variability of 0.25g. These distributions were also input in log normal component form for the analy-sis. The Type 1 transient initiating event was based on the probability of loss of offsite power (LOSP). This has been found to be the dominant source of such transients in all seismic PRA's to date (wherein LOSp results in loss of the main feedwater system). , In computing the frequency of the initiating events, a hierarchy between them must be established. The order of this hierarchy is such that, if one initiating event occurs, the occurrence of other initiating events further down the hierarchy are of no consequence. Thus, for example, if a large LOCA occurs , we are not concerned if a small LOCA or transient occurs. Thus, the most serious initiating event is assumed to be the RVR event. The probability of the lar;;c LOCA event is then computed as the probability of the anchorage failures causing the large LOCA initiating event times the complement of the RVR event, and similarly, for the MLOCA, SLOCA and T1 events. Figure 4.28 illustrates the hierarchy in an event tree- format, and shows the expressions used to calculate the initiating event frequencies. Implicit in the hierarchy definition is the requirement that events in the hierarchy above a given initiating event cannot occur in the accident sequence for that event. For example, LOSP can occur as a basic event in any of the LOCA sequences, but cannot occur as a basic event in the T3 accident sequence. ( 4-71
Conditional Probability of Occurrence 1.0E+ 00 n > E F
- y 1.O E - 01 g Y '
L ' l 1.0 E - 0 2 's 1.0 E - 03 s
- ALOCA RVR 1.0 E - 0 4 : >
t
? !
1,0 E - 05 s ,
= ,
i 1.0 E - 06 g 1.0 E - 07 O.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Peak Ground Acceleration (g's) Figure 4.26 RVR Initiating Event Frequencies Due to Steam Generators & Reactor Coolout Pump Support Failures at-Surry 4-72
1 I 1 1 Conditional Probability of Occurrence 1.0E+00 g 1.0E- 01 = 5 1.0 E - 0 2 ,' i l 1.0E - 03 = b
- SLOCA MLOCA
'1.0 E - 0 4 e
=
1.0 E - 0 5 g - E : i 1.0E-06 g 1.0E - 07 O.0 0.2 0.4 0.6 0.8 1.0 1.2 1,4 Peak Ground Acceleration (g's) l Figure 4.27 Frequencies of Pipe Breaks Causing LOCAS Derived
- from SEMRP Piping Calculations 4-73 1
1 RVR ALOCA MLOCA SLOCA T1(LOSP) IE-T3 EO IE-T1(LOSP) IE-SLOCA lE-MLOCA IE- ALOCA lE-RVR P[IE(RVR)] - RVR P[IE(AMCA)) - E5
- A14CA
~
P[IE(MLOCA)) - RVR
- ALOC4
- MLOCA P[IE(SIDCA)) - 5VR
- ALOC'A
- MLOCA
- SLOCA P[IE(T 2)] - RVR.* ALOCA
- MLOCA
- SLOCA
- LOSP P[IE(T3 )) P[IE(RVR)) - P[IE(ALOCA)) P[IE(MLOCA))
- P[IE(SIDCA)) - P[IE(Tt ))
Figure 4.28 Initiating Event Hierarchy Event Tree 4-74 l l
, . -- r
With the hierarchy established, the Type 3 initiating event >robability is computed ' from the condition that the sum of the initioting event probabilities considered.must be unity. The hypothesis is that, given an earthquake -of reasonabic size, at least one the initiating events will occur. At the least, we expect the operator to manually SCPJJi the plant given an earthquake above the OBE level. Numerical values for the initiating events at various earthquake levels are given in Section 4.4.5. Numerical values for the parameters of the fitted distributions are listed in Appendix C. 4.4.2 Event Trees The event trees developed for the internal event analyses were used, so c as to be able to compare the final core dama6e frequencies due to seismic and internal events on a common basis. The complete internal event trees were shown in Section 2.3. For the scismic compration of core damage frequency, wherein failure - of the containment safety systems does not play a role, the_ internal event trees were simplified by deleting the containment systems. These trees,-used for the seismic calculations, are shown in Figures 4.29 thtough 4.33. Note that no event that is shown for the RVR initiating event as the initiator itself leads directly to core damage since the ECCS mitigating systems are assumed ineffective. Assignment of the accident sequences and their cut sets to the different damage states was performed by examination of the cut sets in both the accident sequences and the containment system sequences. 4.4.3' Failure Modes of Safety Systems To determine failure modes for the plant safety systems, fault tree methodology is used. This methodology systematically identifies all groups - of components in a system which, if they f ailed simultaneously, would result in failure of that system. Construction.of a fault tree begins by identifying the immediate causes
-of system failure. Each of these causes is then examined for more funda-mental _ causes , until one has constructed a downward branching tree, at the bottom of which are failures not further reducible, i.e., failures of mechanical or electrical components due to all causes such as structural failure, human- error, maintenance outage, e tc .
These lowest order
- failures- on the -- fault tree are called basic events. Failures of basic events-due to seismic ground motions, random failures, human error,.and test and maintenance outages are included in the seismic analyses.
The main difference between an internal event fault tree for a safe ty system and an external event fault tree is that consideration must be given - to - the physical location of the components, because the physical location determines to what extent secondary f ailures' become important. Examples of this would be secondary failures due to local masonry wall collapse or.due to a high temperature / steam environment from a broken steam line. Hence, in performing the seismic analyses, the locations of all important pieces of equipment must be determined from the general l 4-75
ACC LPI LPR D5 Dg Hg ALO +1 i ALOCR-2 ~~ \ t x ,
' "~
RLOCR i RLOCR-4 L L Figure 4.29 Large IDC Seismic Event Tree
.HPI RCC LPI LPR UI DS D.
6 Hg ' ox MLoca-i MLOCR-2 i'
~
i .
~
* '"~3
, MLOCR 4 MLOCA-4
'[
-i Figure 4.30 Medium IDCA Seismic Event Tree a: -- '
- ,_ s<-% ,___ _ _ _ _ ___ _ _ ______ _ _ __ ,__
w OPER. RPS HPI PJH PORV DEPRESS. LPR HPR K Di L P; OD Hj H 2 OK SLOCR-1 OK SLOCR-2 SLOCR-3 e OK i U , SLOCR-4 SLOCR-5 l SLOCR-S SLOCR St0CR-7 SLOCR-K RTWS Figure 4.31 Small IJOCA Seirmic Event Tree.
t i FNB ! i RPS- SRV RFH HPI CCH HPI PORV LPR' 'HPR -l K 0 L- D3 H P Hg H2
- D2 i OK I t -
OK f i t f Ti -1 SERL LOCH ! Tg -2 CORE h E RABLE l i i Tg -3 - f i i i 1 [ 1 Ti -4 l - t i i I ' l Tg -5
. [
t I l -b l i Il Tg -0 5' 'CR l Ty -K Rir3 [ i t i Figure 4.32 .T1 (Loss of Offsite . Power) Seismic Event Tree j I r I
- ?
t 4 , I ' i
- r ,
. _ . _ . - _ . _ - _ _ _ _ _ _. . _ . _ . _ _ . . _ _ . . . = -
l d i FNB j r i RPS SRV RFH PCS- -HPI CCW HPl .PORV .LPR HPR ; i K 0 L M D3 H D2 P H; H2 ! OK OK j' T 3 -l' SEAL LOCR f OK t OK or CV
; T3 -2 !
f 5 T3 -3 i3 -4 i T3 -5 I
- 3 Tr6 T3-0 SLOCR T3-K RTHS l
l Figure 4.33 T 3(Turbine Trip) Seismic Event Tree i I 4 t
1 l j- arrangement drawings for the plant, and then a systematic examination for j secondary iallure possibilities is stade during the plant walkdown. As stated earlier, the internal event pRA fault trees form the basis for i the fault trees used in the seismic analy s,i s . This allows for a con- ! sistent level of detail between internal and external event analyses, and ! assures the consistent inclusion of random and test /isnintenance outage ! unavailabilities in the seismic analysis. ! Since the internal event fault trees are assumed to exist and scismic ! failure modes are to be added, one must toodify the internal event fault j trees to include: J l ,
- n. Local structural failures (block walls, erancs, etc.)
A
- b. Failure of crittent passive components (tanks, cable
- tray failures, and pipes.) often missing in internal events analysis.
- This is accomplished in several ways. First, the secondary or passive
! failure event can be added directly to the fault tree structure and the-
" gate" definition data file tuodi f i ed . Alternatively, the fragility
~j definition of a relatively strong component on the tree may be redefined in terms of the (relatively weaker) associated secondary failure. Finally, events globally affecting a safety system or an accident j sequence (such as building failure or liquefactien) can be added directly. L to the Boolean expression for the system failure or accident sequence. l
.4.4.4 Accident Sequence Evaluation Accident sequence probabilities are used in determining the frequencies
! of core damage and of - radioactive release for a given release category. Core damage frequency is defined as the sum of the frequencies of all accident sequences leading to core damage. 4 A. General Considerations ! Each accident sequence consists of groups of events (successes or failures of safety systems) which must occur together. The failure of each safety system can be represented in terms of minimal cut sets, which are groups of component failure which will cause the safety system' to L fail. These cut sets and the accident sequences are combined together so ) that every accident sequence can be expressed in a Boolean expression of the form T ACCj - IE y [C CgCy or C 4C3 or . . . o r C C)Cg ) - 3 3 l in which IE3 is the initiating event and the Ci are basic events (i.e., failure of individual components) identified on the system f ault trees. If at least one of -the component f ailure groups C i3 C Cy occurs, then the-j ._ 4 81
._._._.2_._,_._--.____. _ . _ ._ __ _ ___ _._.. _ _..
t.c c ide nt sequence occurs. Computation of each accident sequence proba-bility consists of determining the probability of each cut set, t.nd then combining them to get the accident sequence probability. Each basic event seismic failure probability is computed assuming that the response and fragility distributions are in log normal form. Calcu-lations in the SSMRP showed that responses were reasonably fit by log norms'. distributions. The liraited data on f rat.111 ties can be fit with log normal distributions as well as any other type. Hence, for con-venience the log normal distribution is used for both. The equation used to calculate seismic failure frequencies is given as in 'm R !*r) Pf = 4-
, FR
- RR where 4 is the standard normal cumulative distribution function, eg, or are the medians of the response and fragility, ,
l'ra, fu are the corresponding random variabilities Note that the use of log normal distributions is not essential to the-calculation of process used in these calculations, and, in fact, any arbitrary pair of distributions could be used for the responses and fra-gilities provided they are physically meaningful. When the individual basic failure events in a cut set Ct C)C, are not independent, correlation between the basic events must be explicitly included. When only two of the basic events are correlated the joint probabilities may be computed directly by the use of tables, When more than two basic events in a cut set are correlated, numerical multiple integration may be used (such at contained in the SEISIM code developed in the SSMRP). Finally, the accident sequences defined above are a function .of peak ground acceloration, and as such, are conditional on the hazard curve. They are subsequently un conditioned by integrating these sequences over the hazard curve as described subsequently. B. Accident Sequence Quantification Quantification of the accident sequences is a multi step procedure involving several levels of screening. In the first step, the SETS code is used to evaluate all potential accident sequences using point estimato input screening values for al. the seismic failure events (and using the internal events point estimat e failure values for all random events). The same fault trees used by the internal events analysis are solved with 4-82
additions as noted in Section 4.4.3. The f.cismic screening values are ! taken as some conservative estimate, usually the component failure probabilities evaluated at three times the SSE. A dual probabilistic o cullint, criterion is used in this culling process. This dual criterion l Is used in recognition of the fact _ that potentially large correlations i can exist between basic events in the sarne cut set due to the pervasive ( nature of the seistric input motion. The result of this screening step is a reduced set of Boolean equations describing each of the safety and j support systems. j In the second step, again utilizing the SETS code, these Boolean
- equations are merged together to form the accident sequences, again as i defined for the internal events analysis. At this stage, truncation is l performed based both on the order of the cut sets as well as the proba.
! bility of the cut. sets. The result of this step is Boolean equations } describing each accident sequence and containing all the important scis-
- mic and random failure events.
l The f i n a*. step involves the actual quantification of the accident J j sequences. These accident sequence expressions are utilized both to L compute point estimates of the accident sequence frequencies and to per. l form the uncertainty analysis calculattens. A cross reference tabic is set up which relates each component to a component-1D number, its randot. .
- point estimate and error factor value, and to its associated seismic fragility cateBory and seismic responso category. This cross reference
- table thus provides all the itaformation required to compute the proba-I bility of~ failure of any basic event (random or seismic or combined) at
- any peak ground acceleration level. The cross reference table for Surry is presented in Appendix C.
{ Finally, a complete uncertainty analysis is performed on the dominant
- accident sequences (and on the dominant cut sets in each accident sequence) as determined in the point estimate evaluations. A true Monte J Carlo analysis was used for the NUREG 1150 studies - Thus, the expression for the unconditional accident sequence frequencies (and for core damage frequency), shown as below
ACC 3 - [ P(ACC ,PGA)f,q(PGA)d(PGA) 3 where-I- P(ACC),PGA) is the conditional accident sequence frequency as a function of PGA, and f,q(PGA) is the probability distribution function for the hazard curve, i is randomly sampled varying the hazard curve parameters, the random
-failure frequencies, and the seismic response and fragility parameters.
From the accumulated values of accident. sequence f requency - and core ! damage frequency, exact statistics on their distributions are directly obtainabic. 4-83 i l
j' 1 1 Note that in pe r forming the uncertainty analyses, full correintion between random samples taken from each response category and from each j f ragility category was enforced. This is correct, and consistent with the philosophy utilized in the internal everit NUREG 1190 uncertainty 5 calculations, j in addition to the full uncertainty analysis (which produces exact mean i ) values and exact percentiles of the distributions of the accident i 3 sequences and total core damage frequency) a "mean point estimate" is , computed. The mean point estimate is useful for illustrating various intermediate results (conditional accident sequences frequencies, initiating event frequencies, e tc . ) which explain the flow of the calculations, for demonstrating convergence of the numerical integration, j and for performin6 sensitivity studies in a cost effective manner. Specifically, the mean point estimate is used to understand the i l contributions of the various basic events to the total frequencies and to understand the contributions to the total uncertainty bands. The mean point estimate is computed by using the mean random failure frequencies, the mean seismic hazard curve, and the mean values for the seismic failure event frequencies in evaluating the accicent sequences. Only one reevaluation of the accident sequences is required. This mean point estimate will be seen to be nearly equal to the exact mean values
- of the accident sequence and core damage frequencies as obtained from the uncertainty analysis. This is to be expected because mean values probabilistically add to yield the mean value of each accident sequence (conditional on the hazard), and the only difference between the true mean and the - mean point estimate has to do with integration of the conditional accident frequencies over the hazard curve. Experience has
- shown, however, that the difference between these is very small.
. 4.4.5 Lase Case Surry Results This section presents the results of the base case seismic risk analysin for the Surry Nuclear Power Plant. The base case is our best estimate of ' the current configuration of the plant and its emergency procedures. In particular, the seismic component failure probabilities were taks. from the generic fragility data base (Table 4.11) with the exception of the site specific component and building fragilities given in Tables 4 '1 and 4.13. As described in Section 4.4.2, a total of six initiating sients l are included for the seismic analysis. A total of 28 accident sequences are identified on these trees which were solved for the Surry seismic analysis. These 28 sequences are presented in Table 4.14 along with identification of the Boolean sequences that were solved for cach accident sequence. -(The number of Booleans solved using the SETS code is less than the number of accident sequences because several accident sequences may utilize the same Boolean expression even though the initiating event may be different.) Also identified on this table are the complement expressions which must be included in the numer-ical sequence quantification at high pCA levels at which success proba-4-84
. . . - . . - ~ . _ _ _ . _ - _ _ - _ . . _ . _ _ _ . _ , . _ - . _ - , _ . - . - , _ . _ . _ - - . _ , , . _ _ _ . . - _ _ _ , _ . _ _ . . _ _ . . -
i ! It ;![ ii[!I!; {lr !l I "};:L l i' [ t((+ Ii!II {tj > ; If, Ih1 . e _ l - b a s r - e e t n l o u N V e r o . C t n , er D3 D r Hg g mo s s S i J 3 et D5 D5 Dg Dg w T L3 Pg P l c 3
- 3 3 g r 3 pa 1 D D _. D 1 D D D 1 D D _D D D 5 1 mF o
c o ) ) ) ) ) } ) ) ) s SN ) ) ) ) ) 0 I 2 9 8 3 2 1 I e T 1 2 3 8 9 1 I 3 2 2 2 2 2 I c ED ( ( ( ( ( ( ( ( ( ( ( ( ( ( n SI e u q e 4 S 1 t n 6 n o e ni 2 2 e d as 1 3 Hg g Hg l i es H3 H 2 b c l e 1 3 D5 D 3 3 D H _D HD H H g g a c or 1 3 s 5 O P P P T A op D D 5 Dg D1 D t L1 L 1L L 3 B x g i c E 1 1 D _D 3 D 3 D D D D 1 w_Dm_DD D 3 D U D g 1 m s i e S rn eo ii l s ps - i e t r 1 l' 1 1 1 1 1 1 1 _K _K _K _ K K K E l p ux ME e r u t A 2 2 Hg t e R p u A C O Ng A C O i H HD H g H1 i g nc L J Hs s L i D _D g l g g en d e l' e s e r s Hs 1 s m u D5 D 3 D1 Dg D 1 3 1 wOL wLP PLP L g 1 i i 3 3 2 _ i u D5 D D i 1 1 D D D D D D D cq s R r 3 3 5 d D1 Dg D D a K K K2 K 2K K2 K 2
' ce e V a D D D D e g 1 m 2 z 2
' AS V R L A A _A A M S S S S S S S S S S 5 S
_ 1 2 3 4 1 2 3 4 1- 2 3 4 - S 6- 7 1 A C A'- C A C R O O V W L L R A M S
,Im* -
, . :Ili; lt!!.fti ;13! {!jIiI ]l4j .j i'Iij .j1!* jl . 1
1 e l s b _ e a _ t r A o A e C _ N C nl D _ O I L u v l _ l a a e e e r S o S C t 2 t n mP I a r I t : m ot I I 1 3 2 2 D3 2 e F D3 D3 l p a c L D 2 D2 U: P P 6 I 2 I, D D D D 1 3 m F o C
)
d-s u n . .
) ) )
i o ) ) ) ) ) ) ) ) ) t S N 3 4 3 2 4 5 6 0 9 8 7 6 n T 3 3 1 1 1 1 2 2 1 1 1 1 o E D ( ( ( ( ( ( ( ( ( ( ( ( . C S I ( yo _ l t s n e og . 4 c t in n 1 _
. e 2 i d 4 u a q
e n o 2 I I I I ee rl
'e
- I I I S n i I I I e
l b a s y I i t
- p F F P z hA T
a t n e s e Dg WP z l P D1 D3 2 2 D3 D 3D 2 3 3 wC
,iD l 2 2 2 7D
.e d
i o P, o p r x I L GL D G D D I D D L L D DI EI D L e el c E r a c t e A A s c C a i r n D m s e o i i I S e s i l s S u e p s Q Q QK ~Q T2 _ S i e t r T) E IQ 5 KQ Q K II Q ~Q i K M K M M M M 5 I h e ca l p t o u x o t M E t r e I 2 sd r r I 2 : t n t H H 3 feo I I l l e y P P P sn t e ) F D H F t WP i H r P i s D3 D 2 3 2 D D D 3 2 3 2 3 3 ni a n c P i 2 2 2 2 2 n D D LD L6 LD rl . e n S D EL D Dl DL DL a D L I ti d e O L r L M M M M M a L T ef i u c n c e ( I EL IQ @ E 3 3 i QQg E IQg 3 3 K K K K %ff4 Q Q Q Q 3 3 3 3 3 3 cnet'og e A S T T T T T T T T T T T T T T a u q 1 m - eda s d - s sde 1 2 3 4 5 6 1 2 3 4 3 6 i er
- - - - - - - - - - - - h eo
. 1 3 T nc T T *
,$e m
i
-j j lfi j' l,1Ijj! , 4
~ . _ . . ..._ _ .... _- _ __ _ _ _ _ _ _ _
}__....__._.__________. i i i
. Table 4.15 t
Safety Systems Nomenclature O Containment spray system (CSS) D liigh pressure injection (llPI) i D2 Same as llPI i D3 High pressure injection for seal cooling i D3 Accumulators (ACC) l De Low pressure injection (LPI) F3 Inside spray recirculation (ISR) F2 Outside spray recirculation (OSR) I li t Low pressure recirculation (LPR Lil) , j !! 2 Low pressure recirculation (LPR illt) I 1 l L Auxiliary feedwater system (AFWS) j M Main feedwater (PCS) OD Operator depressurization (OD) l P3 Block valves and PORV system (one valve required) (PPS2) P Block valves and PORV system (both valves required) (PPS1) l W Component cooling water system (CCW) 1 i i I 1 i I l i- 4 87 __ _ _ ._.. . . _ . __ . _ _ _ _ _ _ . . . ~ . . , ..- _ . _._._. -... _ . .. _ _ . . . - . . - _ , _ _ . , -
bilities may be significantly less than unity. The multiplier expression column lists those events specified by algebraic equations rather than by Boolean logical expressions. The analytical equations used for calculat-ing the multipliers, the Boolean sequences, and the complement factors are presented in Appendix C. Table 4.15 describes the abbreviations used for the accident sequences in Table 4.14 A total of 10 accident sequences survived the seismic screening process. These 10 non negligible accident sequences were fully requantified using best estimate random failure frequencies and best estimate seismic fragilities and responses plus associated variabilities. The total mean core damage frequency for the Surry base case was computed to be 1.16E 4 per year using the LLNL hazard curves and 2.50E 5 per year using the EPRI hazard curves. The mean contributions of the accident sequences are shown on Table 4.16 for both hazard curves. Percentiles of the frequency distributions from the Monte Carlo analyses are shown on Tables 4.17 and 4.18. (Relative importance of the basic events to these results is given in the point estimate results presented later.) Based on this final - quantification, seven dominant sequences were identified. These dominant sequences are (in order of importance): LLNL Hazard EPRI Hazard e T1 6 44% 40% T1 1 23% 27% T3 1 6% 8% SLOCA 7 6% 5% T1-5 6% 5% T3-6 4% 3% ALOCA 3 4% 3% The percentage contributions were taken from the Monte Carlo uncertainty results on Tabic 4.16. Note that the same dominant accident sequences were obtained from the two dif ferent hazard curves , and it will be seen _later that the order of importance of the major contributors is the same. A description of the dominant accident sequences follows. Descrintion of Accident Secuences The dominant sequences computed for the Surry seismic risk are related to loss of AC power and failures of the auxiliary feedwater system and the high pressure injection systems. Civen an event which does not cause a LOCA, there are two ways to remove heat. First, there is the auxiliary feedwater system and, second, there is the feed and bleed operation. This latter operation requires both high pressure injection and the pilot operated relief valves (PORVs). In addition, two seal LOCA sequences were identified. At Surry, there are two sources of cooling water for the reactor coolant pump seals, namely, the high pressure injection (HPI) 4 88
Table 4.16 Accident Sequence and Total Core Damage Mean Frequencier.(1/yr) Ace Sea LINL Hazard EPRI Hazard T1 6 5.1 e 5 1.0 e 5 T1-1 2.7 e 5 6.8 c 6 T3-1 7.2 e 6 2.1 e 6 SLOCA 7 6.8 e 6 1.3 c 6 T1-5 6.4 c 6 1.3 e-6 T3 6 4.9 e 6 8.7 e 7 ALOCA-3 4.3 e-6 7.4 e*7 ALOCA 2 J.4 e 6 5.9 e-7 RVR 3.3 o 6 5.5 e 7 MLOCA 4 1.5 e-6 1.7 e-7 Total 1.16 c 4 2.50 e 5 4-89
i Table 4'.17 l i. I Base Case Accident Sequence Frequency Distribution Percentiles (LLNL Hazard) No. Secuence Mean var 51 }0_1 221 l'- T1-6 5.1E-5 2.21E-08 1.29E-07 5.54E-06 1.80E-04 4.51E-07 1.69E-05 2 T1-1 2.7E-5 5.05E-10 6.91E-09 3 T3-1 7.2E-6 3.21E-08 2.53E-08 2.22E-06 9.22E-05 4 SLOCA-7 6.8E-6 2.84E-09 2.87E-10 9.85E-08 1.47E-05 l- ! e. 5 T1-5 6.4E-6' 1.87E-09 5.10E-09 3.79E-07 2.09E-05 6 T3-6 4.9E-6. 3.42E-10 6.72E-10 1.43E-07 1.39E-05 7 AIDCA-3 4.3E-6 S.52E-10 1.72E-10 5.53E-08 7.40E-06 f i 8 AIDCA-2 3.4E-6 1.05E-11 1.02E-10 3.50E-08 3.37E-06 9 RVR 3.3E-6 2.10E-10 5.31E-10 1.20E-07 1.43E-05 10 MLOCA-4 1.5E-6 2.40E-09 0.00E-01 4.34E-13 5.02E-06 i TOTAL 1.16-04 1.40E-07 3.92E-07 1.48E-05 4.38E-04 i 1,
I t l i Table 4.18 l Base Case Accident Sequence Frequency Distribution Percentiles (EPRI Hazard) No. Secuence Mean var H 50% 95% i I 1 T1-6 1.0E-05 7.94E-10 9.50E-08 2.21E-06 4.51E-05 2 T1-1 6.8E-06 -2.38E-11 4.43E-09 2.00E-07 4.70E-06 i 3 T3-1 2.1E-06 1.07E-09 1.65E-08 9.53E-07 2.41E-05 < I i 4 SIDCA-7 1.3E 06 2.71E-10 1.86E-10 5.55E-08 5.30E-06
.~
I ' 5 T1-5 1.3E-06 3.60E-11 2.90E-09 1.43E-07 4.96E-06 6 T3-6 8.7E-07 9.34E-12 2.89E-10 5.03E-08 2.82E-06 7 ALOCA-3 7.4E-07 1.79E-11 6.48E-11 1.86E-08 1.93E-06 t f i ! ! 8 ALOCA-2 5.9E-07 3.85E-13 5.39E-11 1.28E-08 8.15E-07 5.53E-12 2.44E-10 3.96E-08 2.6?E-06 9 .RVR 5.5E-07 t 10 MIDCA-4 1.7E-07 1.37E-10 0.00E-01 1.27E-13 2.02E-06 :
! TOTAL 2.50E-05 5.04E-09 3.00E-07 6.12E-06 1.03E-04 i i
I j' ! t i
' t i i i
I
_ _ _ . . . m.__...__.-.-_. _ . _ . - - _ _ _ _ _ . _ _ _ . . _ _ . _ . _ _ . _ _ _ . _ . _ _ _ _ i i 4 system and secondly, the component cooling water (CCW) system. Both these systems must f ail in order to fail cooling to the reactor coolant , purops . 4 l Tim most important sequence is sequence T1 6. This is a loss of offsite l power (LOSP) sequence in which both the auxiliary feedwater system and 4 high pressure injection fail. The auxiliary feedwater system fails i j primarily due to failure of the condensate storage tank while the high (
- pressure injection system falls either due to failure of the refueling
] water storage tank (RWST) or loss of onsite AC power. The loss of
!. onsight AC power is due primarily to failure of the 4KV emergency switchgear anchorages, and secondarily to failure of the diesel generators to start given the seismic event.
I - The second most dominant sequen~ is T1 1. This is a loss of offsite
- _ power sequence leading to a seal LOCA. Note that this is a loss of '
l offsite power sequence, however, the auxiliary feedwater system does . succeed. Ilowever, failures of high pressure injection and cornponent
- cooling water lead to a seal lhCA, Failures of these two systems are due
- -either to the RWST or are onsite power related. There is a small contribution from the PJIR heat exchanger support failures to the failure l j - of the component cooling water system. The third most dominant sequence l T3 1 is identical to sequence T1 1 except that now the transient is caused by some other failure (or manual scram) leading to shutdown, and offsite power is available. The predominant contribution to this sequence is due to failures of the 4KV emergency switchgear which effectively cause loss of all emergency AC power.
The fourth most important sequence is SLOCA 7 which is a small LOCA and
- which involves failure of the high pressure injection system. Again, the high pressure injection system fails either due to the RWST or onsite AC power failures as in the sequences aircady discussed.
The fifth most important sequence. T1 5, is also a loss of offsite power sequence in which both the APWS and - the feed and bleed function have failed. In this case, feed and bleed fails due to_ failure of the PORVs
- and their associated block valves. This is caused by failure of one train of AC power in conjunction with one set of block valves being closed. (At Surry, both sets of block valves and PORVs must be available for feed and bleed.)
Sequence T3 6 is the same as T1 6 except that offsite power is initially available. In this case, failures of the AFWS and the llPI systems are caused by failures of the water sources (the CST and the RWST, respectively). The two sequences ALOCA-3 and ALOCA 2 are next in importance. In the i former, the accumulators function properly but the low pressure injection system fails due to electric power failures. In the latter, both the accumulators and LPI succeed, but long term low pressure injection falls. 4 92.
(..-......_-. . - .- . - - _.- - - - _ - - -...- - - ) J 1 O e i The RVR sequence is next in importance. It's occurrence is totally due to failure of the reactor coolant pump and steam generator supports. j Lastly, the medium LOCA sequence MLOCA-4 occurs due to failure of the HPI system. This occurs primarily because loss of power and also due to loss of the-RWST. 1 Mean Point Estimate Using the LLNL Ha: ard Curve,,g a As described earlier, this point estimate is based on using the mean values for all variables. The mean initiating event frequencies at , different p0A values are given in Tabic 4.19 As can be seen, at the I lower earthquake levels the transient sequence initiating events 1 dominate, and as the earthquakt acceleration level increases, the LOCA initiators increase until, fi nt.11y , at the highest earthquake icvels - there is a contribution from the reactor vessel rupture (RVR) event. Also note th a t ., at each earthquake level, the initiatin6 events sum to 1.0, Values of the dominant accident sequence conditional frequencies at various earthquake icvels are presented in Table 4.20 These are the values that- are integrated over the hazard curve to obtain the unconditional accident sequence frequencies. Table 4.21 presents the mean core damage contributions at seven intervals over the hazard cucve for each accident sequence. (Integration over the hazard curve was performed from 0.05g to 0.75g and in the uncertainty analysis computations, integration increments of 0.025g were utilized. However, for explanatory purposes the results presented here are based on an integration increment of 0 lg.) The right hand column presents the total contribution of each accident sequence to the total core damage , frequency of 1.12c 4. As can be seen, the incremental contributions from the LOCA events do not become significant until the hi Cher acceleration levels. The reactor vessel rupture sequence does not make a significant contribution until the highest PGA increment. An important thing to note f rom Table 4.21- is the sum of the accident sequence contributions at each earthquake level, as shown at the bottom of each column on the table. The contributions are seen to be small at the first increment, increasing to a maximum at the fourth earthquake increment, and then decreasing at higher earthquake levels. -This indicates that the bulk of the risk is occurring in the range of 0.25g to 0.65g which roughly corresponds to the range of 2 4 SSE. Further, this-shows that the bulk of the risk has been captured by integrating over the range 0.05g to 0.75 6. Monn Point Estimate Using the EPRI Hazard Curva Tables 4.22 through 4.24 presents similar results for the mean point i estimate using the EPRI hazard curves. In this case, a total core damage-frequency of 2.21E-$ was computed. This was very close to the Monte l Carlo estimate of mean core damage frequency of 2.50E 5 computed using the same equations in the uncertainty analysis. Similar comments with 4-93 _ - . - . _ . _ . - _ . , _ . _ _ . - _ . , , _ _ _ _ _ , ._ .._ ___-,--a;.._,,._,_,_ -
Table 4.19 Mean Initiating Event Frequencies - LLNL Hazard 0.5r 0.6e 0.7c 0.2r 0.3r 0.4r _ 0.1r 1.66E-02 4.22E-02 7.80E-02 1.19E-01 1.79E-06 3.59E-04 4.03E-03 1.87E-01 RVR 1.35E-02 4.37E-02 9.18E-02 1.43E-01 ALOCA 1.63E-05 1.71E-03 5.01E-02 6.57E-02 4.82E-04 4.12E-03 1.43E-02 3.14E-02 MLOCA 4.59E-06 1.2SE-01 1.63E-01 1.78E-01 1.84E-04 6. 73 E-03 3.28E-02 7.79E-02 SLOCA 7.06E-01 6.74E-01 5.65E-01 4.51E-01 3.36E-02 3.25E-01 6.08E-01 0.00E-01
,, TI(LOSP) 1.41E-01 3.27E-02 0.00E-01 8 T3 9.66E-01 6.66E-01 3.38E-01 1
,I f!!t ' e' j : ,h1 ltI llt[ij m -
=
~
142111121 000000000 e - - - - - - - - - 7 EEEEEEEEE + 156089977 0 519077158 - 424011161 132111221 000000000 r - - - - - - - - - 6 EEEEEEEEE 608029073 d 0 541062894 r 528011741 a z a H L. 121212222
'2 000000000 1 c - - - - - - - - -
Y 1 EEEEEEEEE 5 317 823298 0 850021291 s' 511217429 e i c) nd . er y ua qz 122222222 ea 000000000 7 rh r - - - - - - - - - 0 F 4 EEEEEEEEE - 2 n 237619627 - 690214613 eo 0 4 c nl 339562114 *6 " e ea 9g l b un qo a ei T St 222223332 i td 000000000 r - - - - - - - - - nn eo 3 EEEEEEEEE 836323385 dC 0 187441073 i( c 823414411 c A' % t n a 333244453 n 000000000 i - - - - - - n r - - r o 2 EErEEEEEE F' D 050561941 0 962245537 n~ 224171351 . a e M 765478685 000000000 r 1 EEEEEEEEE 610828973 0 507179716 . . 7 4 6 3 8' C121 73 42 AA AA 2311CC CC 1I13II VLL DDROO TTTTSARMA TJ' _ I > , !i \l;h: ' , : , ' .I 3!f : 1 : , J 4 3 ! , I ,
Table 4.21 Mean Core Dawage Contributions (Median) at Intervals of FCA - I.IRL Hazard 0.45-0.55r 0.55-0.65r 0.65-0.75r Total 0.25-0.35r 0.35-0.45e 0.05-0.15e 0.15-0.25r 9.46E-06 4.28E-06 1.81E-06 5.12E-05 1.87E-07 '4.63E-06 1.47E-05 1.61E-05 7.19E-06 T1-6 1.50E-06 2.10E-07 1.40E-08 5.61E-10
.5.38E-08 1.78E-06 3.63E-06 2.48E-07 2.50E-05 T1-5 9.61E-06 6.38E-06 2.62E 8 55E-07 Y1-1 4.01E-07 4.94E-06 1.18E-08 1.96E-09 4.55E-06 2.19E-06 1.14E-06 '3.15E-07 6.56E-08 T3-1 8.28E-07 1.94E-06 1.34E-06 8.31E-07 8.08E-06 4.'79E-09 3.07E-07 1.44E-06 2.22E-06 4.67E-06 SIDCA-7 9.45E-07 1.19E-06' 1.13E-06 9.18E-07 AIDCA-3 3.35E-10 6.16E-08 4. 30E-07 6.46E-07 3.17E-06 3.93E-07 5.94E-07 6.87E-07 6.95E-07 RVR 8.52E-09 1.42E-07 4.16E-07 3.19E-07 1.82E-06 2.20E-08. 1.82E-07 4.08E-07 4.76E-07:
MIDCA-4 1.19E-10 1.18E-06 1.15E-06 9.32E-07 4.35E-06 ) 1.19E-10 1.93E-08 2.58E-07 8.08E-07 2.05E-06 AIDCA-2 2.25E-07 6.99E-08 1.97E-08 5.27E-09 l T3-6 2.73E-07 8.78E-07 5.81E-07 1.79E-5 9.91E-6 5.7E-6 1.12E-04 1.74E-6 1.49E-5 3.24E-5 2.95E-5 Total e i c 7
- i 4' t
!- Table 4.22- , i Mean Initiating Event Frequencies - EPRI Hazard ,
?
I i 0.1c 0.2e 0.3r 0.4e 0.5r O.6c 0.7c
! RVR 1.79E-06 3.59E-04 4.03E-03 1.66E-02 4.22E-02 7.80E-02 1.19E-01 :
l I ALOCA 1.63E-05 1.71E-03 1.35E-02 4.37E-02 9.18E-02 1.43E-01 1.87E-01 , MLOCA 4.59E-06 4.82E-04 4.12E-03 1.43E-02 3.14E-02 5.01E-02 6.57E-02 ' SLOCA 1.84E-04 6.73E-03 3.28E-02 7.79E-02 1.28E-01 1.63E-01 1.78E-01 TI(LOSP) 3.36E-02 3.25E-01 6.0BE-01 7.06E-01 6.74E-01 5.65E-01 4.51E-01 l T3 9.66E-01 6.66E-01 3.3SE-01 1.41E-01 3.27E-02 0.00E-01 3.00E-01 ! i
> l 1
1 C 4 i 4 i i i i l 4 __ r' -- __ _ ' _ _ _ . _ _ - _ _ . _ . . _ _ . . _ . _
a Table 4.23 Mean ^ccident Sequence Frequencies (per year) Conditional on Hazard - EPRI Hazard 0.4r 0.5r 0.6r 0.7c 0.le 0.2r 0.3r
'8.18E-02 3.62E-01 5.83E-01 5.56E-01 4.51E-01 TI-2 7.56E-07 2.90E-03 2.15E-04 2.65E-03 2.83E-02 3.93E-02 1.51E-02 2.40E-03 T1-3 4.01E-06 8.18E-02 4.96E-02 6.70E-05 4.20E-03 3.76E-02 9.07E-02 1.07E-01 T1-1 2.08E-02 0.00E-01 0.00E-01 T3-1 3.18E-04 1.25E-02 4.43E-02 5.26E-02 1.42E-02 6.11E-02 1.22E-01 1.62E-01 1.78E-01 SIACA-7 8.72E-07 7.46E-04 1.79E-01 1.51E-04 4.13E-03 2.49E-02 7.13E-02 1.29E-01 ALOCA-3 6.98E-08 7.90E-02 1.19E-01 1.79E-06 3.59E-04 4.03E-03 1.66E-02 4.22E-02 7 RVR 5.34E-05 1.78E-03 1.12E-02 2.99E-02 4.97E-02 6.57E-02 e MLOCA-4 2 17E-08 1.43E-01 1.87E-01 1.63E-05 1.71E-03 1.35E-02 4.37E-02 9.18E-02 ALOCA-2 l
l l l I
Table 4.24 Mean Core Damage Contributions From Dominant Accident Sequences - EFRI Hazard 0.55-0.65r 0.65-0.75r Total 0.15-0.23r 0.25-0.35r 0.35-0.45r 0.45-0.55r O_05-0.15e 1.57E-06 6.56E-07 2.36E-07 9.85E-06 7.03E-08 1.20E-06 3.14E-05 2.98E-06 1.57E-06 T1-6 2.79E-07 3.48E-08 2.14E-09 7.28E-11 2.02E-08 4.60F-07 7.74E-07 3.22E-08 5.26E-G6 T1-5 2.05E-06 1.19E-06 4.33E-07 1.31E-07 T1-1 1.51E-07 1.28E-06 1.82E-09 2.55E-10 1.19E-06 5.69E-07 2.42E-07 5.85E-08 1.08E-08 T3-1 3.11E-07 3.21E-D7 2.05E-07 1.08E-07 1.43E-06 7.95E-08 3.08E-07 4.12E-07 7.73E-07 l 51DCA-7 1.80E-09 1.97E-07 1.73E-07 1.19E-07 1.26E-10 1.60E-08 9.16 E-08 1.76E-07 5.38E-07 AIDCA-3 1.1CE-07 1.14E-07 1.06E-07 8.39E-08 3.20E-09 3.68E-08 8.38E-08 4.14E-08 3.04E-07 RVR 3.87E-08 7.58E-08 7.87E-08 6.38E-08 MIDCA-4 4.4SE-11 5.69E-09 1.77E-07 1.21E-07 7.02E-07 ( 5.01E-09 5.50E-08 1.50E-07 1.95E-07 AIDCA-2 4.48E-11 1_16E-08 3.01E-09 6.84E-10 5.11E-07 1.02E-07 2.28E-07 1.24E-07 4.19E-08 T3-6 2.96E-6 1.52E-6 7.42E-7 2.21E-5 6.55E-7 3.86E-6 6.90E-6 5.47E-6 Total l l
l t b s 2 to the variation of initiat.ing event frequencies and accident sequence l f requencies with earthquake level as described for the LLNL mean point
- . estin. ate case apply. P 4.4.6 Base Case Importance Studies l A. Basic Event Importance to Mean Values The irtportance of the basic seismic failure events was evaluated by
< setting the seismic failure probability to zero, which gives a measure of the not reduction in risk that would occur if that component could never
- fail due to seismic shaking. The results of these calculations for both sets of hazard curves are shown in Table 4.25 and the results are both i qualitatively and quantitatively similar. (Note that the sum of the risk i reduction percentages do -not -and should not equal unity, since many of
{ the important n.mponents occur together in the same cut sets, and hence, , a zero failure probability of one component causes the entire cut set to l-vanish.) it can be seen that the largest risk reduction occurs for ceramic h insulators. This occurs, of course, because the ceramic insulators are the basis for the T3 transient sequences. The two vertical water storage tanks (CST and RWST) have risk reductions of 26 percent and 10 percent i respectively. The 4Ky busses together represent a risk reduction of 36 percent, which is due to the fact that all 4Kv power, including emergency power from the diesel generators, go through these busses. The
- two diesel generators represent a risk reduction of 22 percent when taken together. The remainder of the components have significantly less risk
- reduction potential.
B. Basic Event Importance to Overall Uncertainty The relative contribution of the hazard curve, the seismic response and
.the seismic fragility uncertainties (Su's) to the overall core damage frequency was ascertained. The results of these comparisons (for both sets of hazard curves) are shown on Tables 4.26 and 4.27 The base case mean, 95 percent and 50 percent core damage frequencies are shown in the
.first column, The second column shows the corresponding values with the
. hazard curve fixed at its median value (i.e., with no modeling uncertainty), for the LLNL hazard - curve case, it can be seen that the .
error factor (EP) associated with these results is 3.6, whereas the corresponding error factor for the base case--was 29.6. Similarly, for the EPRI hazard curve case, the base caso error factor was 16.8 while with no . uncertainty in _ the hazard curve, the error factor is reduced to > 4.2. Clearly, the hazard curve is contributing the vast majority of the uncertainty in the base case results. The third column shows the calculation wherein all the fragility and ! response modeling uncertainties are simultaneously r,et to zero. . For the LLNL hazard curves, the error factor is 23.6. For the EPRI hazard curves, the corresponding error factor is 12.6. These results show that the reduction in the response or fragility uncertainties has only a I 4 100 i
, - _ _ . - . -,_,_.m. , - _ , - - . , - _ . _ . - - . . , . . _ _ - . _ _ _ _ . . , _ , _ . . _ . . , . _ . _ .. _. ..__.-. _._ m , . _ . - . - .
Table 4.25 Dominant Compe nent Contributors to P(cm) Ranked By Risk Reduction Potential LISL EPRI Component flezard llazard Ceramic Insulators 50% 68% 4KVill 36g 27 g. 4KV1J CST 26% 21% DG1 FS 22% 13% DG3 FS RWST 21s 22% BAC 11R-2 9% 8% BAC 1J12 AFW XCONN 3% 24 OEP DG 3U2 3% 2%
- , CRB FT 15113 <1% <1%
CRB FT 15J3 <1% <1% DG1-MA <1% <1% DG3 MA <1% <1% OEP DG CCF 13 <1% <1s BATTlA <1% <1% BATT1B <1% <1% 4 101
. . . - - ~ . . - - . , . - . - - . - ._-. - = _ _ _ _ - - - . - _ . - . . . . - .. . . . . - _ - -
Tabic 4,26 Comparison of Contributions of Modeling Uncertainty in Response, Fragility and llazard Curves to Core Damage Frequen:y LLNL llazard Base llazard ru.0 Ca s e .. _&e0 &t-0 Mean 1.16E 4 1.76E 5 6.31E-5 95% 4.38E 4 4.66E 5 2.30E 4 50% 1.48E 5 1.28E-5 9.73E 6 E ,(95%) 29.6 3.6 23.6 Pc .(50%) Table 4.27 Comparison of Contributions of Modeling Uncertainty in Response, Fragility and llazard l Curves to Core Damage Frequency EPRI llazard Base llazard Bru-0 Case B,mo ht,-0 Mean 2.50E 5 8.09E 6 1.29E-5 95% 1.03E 4 2.29E 5 4.86E 5 I 50% 6.12E 6 5.47E 6 3.86E 6 fe,(95%) 16.8 4.2 12.6 Pe.( 50 % ) - l 4 102
u d l secondary offeet on the overall core damage uncertainty (no mat ter which J set of hazard curves is used). These results show quite clearly that the uncertainty in the hazard curve is _ the dominant factor in both the mean value of core damage frequency i and in the uncertainty of the core damage frequency. Further, na was seen in the discussion of the tocan point estimate case, it is the tacan hazard curve which drives the mean estimate of core damage frequency. Again, this shows the dorninant influence of the hazard curve uncertainty (which deterrnines the tocan hazard curve) in determining the rocan core damage frequency. C. Effect of Hazard Curve Discretization All the results discussed so far have been based on a inodel of the hazard curve uncertainty in which the variation is assumed to be log normal (at each value of PGA). The principal investigator or the Eastern US Seismic Hazard Characterization Program has indicated that this uncertainty dis. tribution is approximately log normal, and this was substantiated by the calculated mean hazard curve shown earlier, lloweve r , the log normal distribution does have an extended tail. To assess the potential effect of contributions from the tail of the assumed distribution an alternate approach was taken. In this sensitivity study, a family of ten hazard curves was generated from the assumed log normal distribution corresponding to confidence levels of 5 percent, 15 percent, . . . 95 percent. Each of these curves is assumed to be equally weighted. Tabic 4.28 compares the. LLNL mean hazard curve ordinates derived from the family of discrete hazard curves- used above with the mean hazard curve obtained from the full log normal distribution model. As can be seen from this - table , the mean hazard curve is significantly less for the discrete family. A point estimate calculation was made using the mean hazard curve for the family and mean seismic accident sequence frequen-cies which _ resulted in a mean point estimate value of core damage frequency given by 6.40E 5. This compares to the base case incan value of 1.12E 4 This reduction in core damage frequency from the base case is due to both cliniinating the tails of the distribution and due to a chift in the mean hazard curve. i Table-4.29 compares the EPRI rocan hazard curve ordinants derived from the family of _ discrete hazard curves with a full log normal distribution model. Again, repeating the analysis _ resulted in a mean core damage frequency of 1.67E 5 as contrasted to the b_ase case result of 2,21E 5 per year. From these results, one would infer that the use of a limited number of discrete hazard curves re sul ts : in a reduction in computed core damage frequencies from 24 percent to 43 percent, and that the reduction.is due to the reduction in the mean hazard curve which results from cutting off the tails of the full hazard curve distribution. From a PRA perspective, I 4-103
..= .-... -.. - - .
9 Table 4.28 Comparison of Mean Hazard Curve Probabilities From Ten Discrete llazard Curves and From llazard Curve with Assumed Log Normal Distribution LLNL llazard 10 Discrete Curves Full Distribution Mean Hazard Probability Mean Hazard Probability _EGA_ D.15g 3,63E 4 5.65E 4 0.25g 9.58E-5 1.70E 4 0.35g 3.74E 5 7.30E 5 0.45g 1.79E 5 3.77E 5
-0.55g 9.78E 6 2.19E 5 0.65g 5.78E 6 1.37E 5 i 0.75g 3.70E 6 9.12E 6 Table 4.29 Comparison of Mean llazard Curve Probabilities From Ten Discrete Hazard Curves and From Hazard Curve with Assumed Log Normal Distribution EPRI Hazard 10 Discrete Curves Full Distribution PGA . - Mean 11arard Probability Mean Hazard Probability 0.15g 1.10E 4 1.35E 4 0.25g 2.42E 5 3.28E 5 0.35g 8.34E 6 1.21E 5 0.45g 3.64E 6 5.54E 6 0.55g 1.85E-6 2.92E-6 0.65g 1.01r-6 1.67E 6 0.75g 5.96E-7 1.07E-6 4 104
l.- reductions, while not insignificant, would not affect the conclusions resulting from a seismic PRA. Thus, one would conclude that. knowledge of the exact form of the tails of the hazard curve distribution (as determined by the LLNL nazard curve development process) is not essential to a robust understanding of the plant's s e i s m . .- risk and vulnerabilities. 4.4.7 Summary and Plant Speelfic insichts This chapter has presented the seismic risk results for the Surry Plant using both industry sponsored (EPRI) and NRC sponsored (LLNL) hazard curve estimates. The differences between these sets of hazard curves resulted in a significant difference in computed total core damage frequency (1,16E 4 per year for the LLNL hazar.1 curves and 2.50E_5 per - year for the EPRI hazard curves). This rather significant difference is expected to bound the seismic risk at Surry, llowever, the seismic risk was found to be dominated by relatively few accident sequences and the same dominant accident sequences were found using both sets of hazard curves. Furthermore, it was found that the relative contribution of individual component failures was the same (both qualitatively and quantitatively) for both sets of hazard curves. Thus, insights as to important contributors to risk at Surry and to the identification of important accident scenarios are relatively robust and did not depend on the particular hazard curves chosen. In general, it was found that only a few accident sequences dominated the results. The most dominant sequence was a loss of offsite power (LOSP) transient sequence in which the auxiliary feedwater system fails (due to loss of the condensate storage tank) and the high pressure inj ec tion (llPI) system (and hence, the feed and bleed operation) fails due to either failure of the refueling water storage tank or failures of the onsite AC power system. The second most significant sequence is also a loss of offaite power transient sequence, except that this transient sequence leads to a seal LOCA. This is caused by failure of both the HP1 system and the component cooling water (CCW) system which leads to the seal LOCA. The llPI system fails as described above while the CCV system fails due to loss of onsite AC power. Together, these two sequences constitute approximately 67% of the computed core damage frequency. Finally, a sensitivity study in which the continuous lognormal uncertainty model for the hazard curves was replaced by a discrete family of hazard curves (and, hence, the extreme t a il. s of the lognormal distribution were truncated) was made. This study showed that the tails of the hazard curve distribution did not dominate the core damage frequency results obtained, l i 1 l l 4-105
4 i i l 4,5 lie fe renc e s
; 1. D. L. Bernreuter et al., Seismic . linrard Characterization of 69 4 Huelcar Plant Sites Enst of the Rocky Mountains, Lawrence Livermore National Laboratory, NUREC/CR 5250, October 1988, i i- 2. Electric Power Research Institute, kismic llazard Methodolony for the c^
EtnLrni and Eastern United States, EPRI NP-4726, Vol. 1-10, July 1986, d
- 3. P. B. Schnabel, J. Lysmer, and 11. B. Seed, SilARE - A Comput er Pro-tram for Entthaunke Resronse Annlysis of Ilorizontally Layered Sites.
Earthquake Enginenting 1cicarch Center, University of California, Berkeley, CA, EERC 72 12, 1972.
- 4. Ohta and Coto, " Empirical Shear Wave Volocity Equations In Terms of l
Characteristic Soil Indexes," Earthaunke Fug Necring 6 gLtuctural s Dynarnien, Vol, 6, pp 169 187,1978. i
- 5. Surry Power S ta t ion -- Upda t ed Final Safety Annivsis Report, Section 2.0, June 1983.
- 6. 11. B. Seed and I. H. Idriss, Soil Moduli and Damoinn Factors for i Dynnmie Resnonse Annivres, Earthquako Engineering Research Center,
! University of California, Berkeley, CA, EERC 7010, 1970.
- 7. 11. B. Seed, C. Ugas, and J. Lysmer, Site Dependent Spectra for Earth-gunke Resistant Desigu, Earthquake Engineering Research Center, Uni-versity of California, Berkeley, CA, EERC 74 12m, 1974 l 8. J. J. Johnson, C. L. Coudreau, S. W. Bumpus, and 0, R. Maslenikov, hismic Safety Margins Research Pronrnm Phase I Finni Renort -
SMACS - Seismic Methodolony Analysis Chnin with Statistics (Proicet 21111 Lawrence Livermore National Laboratory, Livermore, CA, UCRL-53021, Vol. 9 NUREC/CR 2015, Vol. 9, 1981.
- 9. 11 . L. Wong - and J . E, Luco, Soil Structure Interaction; A Linear l
Continuum Mechanics Approach (GASSI) Dept. of Civil Engineering, University of Southern California, Los Angeles, CA, CE79 03, 1980,
- 10. M. P. Bohn et al., APJ11 cation of_the SSMRP Methodology to the Scis-mic Risk at the Zion Nuclear Power Plant, UCRL 53483, NUREC/CR 3428,
. Lawrence Livermore National Laboratory, Livermore, CA, 1983, l
- 11. L. E. Cover, et al., llandbook of Nucient Power Plant Seismig Fragilities, NUREC/CR 3558, December 1983.
- 12. R. D. Campbell, et al., Compilation of Fragility Information From Available Probabilistic Risk Assessments,-Lawrence Livermore National Laboratory Report UCID 20571 Rev. 1, September 1988.
- 13. D. A. Wesley and P. S. liashimoto, seismic Structural Frattlity I nve s t i gitt ion for the Zlop Nuclear Power Plant, Lawrence Livermore National Laboratory, NUREC/CR 2320, October 1981.
4-106
,. . - . ~ _ . _ _ _ . _ , . - , . , . _ . . _ _ . _ . . _ . . _ _ . _ _ _ _ - _ . _ _ _
- - . - . . - .-. . .. . - -. _- - - - .- - _ . . = . - - . -. .. __ ..
14 R. Riddell' and N. M. Newmark, Statistical Analysis of the Responseg Ngnlinear Systems Subiected to Earthounkes, UILU ENG 79 2016, Univer-sity-of Illinois, August 1979.
- 15. R. P. Kennedy, et al, Engineering Characteristics of Ground Motion -
Task 1: Effects of Characteristics of Free - Flei d Motion on Structural Response, Lawrence Livermore National Laboratory, NUREC/CR 3805, May 1984.
- 16. N. M. Newtua rk , " Effects of Earthquakes on Dams and Embankments,"
Geotechnioue, Volume XV, No. 2, 1965.
- 17. Letter from Virginia Electric and Power Company to NRC,
Subject:
IE 15u11ctin 80-11, Masonry Wall Design Final Report, Surry Power Station Units 1 and 2, dated October 29, 1981 (Scrial No. 604). 4-107 i
- i. . . _ . _
5.0 SURRY FIRE ANALYSIS 5.1 Jntroduction The objective of the analysis reported here was to estimate the contribution of fire induced core damage and plant damage state frequencles. The overall ftre induced core damage frequency for Surry Unit 1 was found to be 1.13E 5 per year. The various fire area contributions are given in Table 5.1. The accident sequences these scenarios mapped into are listed in Table 5.2. Tabic 5.1 Surry Fire Area Core Damage frequency Core Damage Frequency (/yr) 5th 95th Fire Area Mean fercentile Median Percentile Emergency Switchgear Room 6.09E-6 3,93E 9 3.15E-6 1.98E 5 Control Room 1.58E-6 1.20E 10 4.68E-7 6.95E-6 Cable Vault / Tunnel 1.49E 6 6.51E-10 6.99E 7 5.79E-6 Auxiliary Building 2.18E 6 5.32E-7 1.59E 6 5.64E-6 Charging Pump Service Water Pump Room 3.92E-8 1.43E 10 5.66E 9 1.58E 7 Total 1.13E 5 5,37E-7 8.32E-6 3.83E 5 Based on plant operating experience over the last 20 years, it has been observed that typical nuclear power plants will have three to four significant fires over their operating lifetime. Previous probabilistic risk assessments (PRAs) haue shown that fires are a significant contributor to the overall core damage frequency, contributing anywhere from 7 percent to 50 percent of the total (considering contributions from internal, seismic, flood, fire, and other events). Because of the relatively high core damage contribution, fires need to be examined in more detail. An overview of the simplificd fire PRA methodology is as follows: ? 51
Table 5.2 Dominant Accident Sequence Core Damage Frequency Cor S.tGUCnte F1re Area D.nntagt ,,c ,, el _(fy,r,), T 3D3VD3 Emergency Switchgear Room 6.09E 6-Auxiliary Building 2.18E-6 Cable Vault / Tunnel 1.49E-6 T 3QD3 Control Room 1.58E 6 Charging Pump Service 3,92E 8 Vater Pump Room I~ A. Initial Plant Visit Based on the internal event and seismic analyses, the general location of cables and components of the systems of interest is known. The plant visit provides the analyst with a means of socing the physical arrangements in each of these areas. The analyst will -have a fire zone checklist which will aid the screening analysii, and in the quantification step. The second purpose ' of the initial plant visit is to confirin with plant personnel that the documentation being used is, in fact, the best available information and to get clarification about any questions that might have arisen in a review of the documentation. Also, a thorough j review of firefighting procedures is conducted. B. . Screening It is necessary to specify the important fire locations within the power plant under investigation that have the greatest potential for producing risk-dominant accident - sequences . The objectives of this location selection are somewhat- competing and should be balanced in a- meaningful risk assessment study. The firs t - obj ec tive is -to ma>1mize the possibility that all important locations are analyzed, this leads to the consideration of a potentially large number .of candidate locations. The second objective is to minimize the effort spent in the quantification of t event ~ trees and fault trees for fire locations that turn out to be-L unimportant. A proper balance of those objectives is one that results in an ideal- allocation of resources and efficiency -of assessment. The screening analysis is comprised of:
- l. 1. Identification of relevant fire zones. Those Appendix R identified fire ' zones which had either safety related equipment or power and control cables for that equipment were identified as requiring further analysis. This group of fire zones (areas) is briefly described in Section 5 . 2 -- All critical safety components within the.c fire areas are given in Appendix D.
5-2 l
- 2. Screen fire zones on probable fire-induced initiating events.
Determination of the fire frequency for all plant locations and determination of the resulting fire induced initiating events and "off-normal" plant states is delineated in Sections 5.3 and 5.4 respectively.
- 3. Screen fire zones on both order and frequency of cut sets.
- 4. Each fire zone remaining is numerically evaluated and culled on frequency.
The screening methodology (Section 5.5) describes how reduction of the initial group of locations from Section 5.2 to the five remaining with contributions to core damage frequer v of greater than 10-6 per year was accomplished. C. Quantification After the screening analysis has eliminated all but the probabilistically-significant fire zones, quantification of dominant cut sets is completed as follows:
- 1. Determine temperature response in each fire zone,
- 2. Compute component fire fragilities. The latest version of the fire growth code COMPBRN with some modifications was used to calculate fire propagation and equipment damage. A description of these results for steps 1 and 2 is given in Section 5.6. These fire calculations were only performed for the fire areas that survived the screening analysis.
- 3. Assess the probability of barrier failure for all remaining combinations of fire zones. A barrier failure analysis was conducted for those combinations of two adj acent fire zones which, with or without additional random failures, remained after the screening analysis. The methodology to assign barrier failure probability to the fire zone combinations is described in Section 5.7.
- 4. Perform a recovecy analysis. In similar manner to that used for the internal event analysis recovery of non-fire related random failures was addressed. Appropriate modifications to recovery probabilities were made as described in Section 5.8.
- 5. An uncertainty analysis is performed to estimate error bounds on the computed fire-induced core damage frequencies. As in the internal events analysis, the TEMAC code was utilized in the uncertainty analysis as described in Section 5.9.
5-3
.l 1
In Section-5.10 a detailed description of all fire scenarios with contributions to core damage frequency of greater than 10'8 per year and their associated fire areas is given. Distributions and a description of all factors used in the final quantification of all fire areas are delineated, 5.2 Fire Locations Analyzed In this _ section, the plant areas (fire zones) analyzed are listed in Tabic 5.3. A list of components contained in each of these fire zones is given in Appendix D. Table 5.3 also provides a brief physical description of each fire zone. This study was conducted with cable routing information on a limited set of components considered to be those most vital to mitigating the effects of any potential fire-induced "off-normal" plant state. Some of this cable routing information was obtained from the Appendix R submittal while other routings were obtained from utility routing information and confirmed during a plant walkdown. These lists of components as well as cable traced vital components formed the basis of the computer aided screening analysis. All other fire areas not contained in Table 5,3 were screened, as they did not contain either vital equipment or cabling for that equipment. The following subsections provide a discussion of the fire detection and manual or automatic fire extinguishment capabilities that presently exist in each fire zone. 5.2.1 Cable Vault / Tunnel (Fire Area 1) Ionization smoke detectors are provided in Fire Area 1. These detectors alarm in the control room. In addition, heat detectors which actuate an automatic CO 2 system are located in the CV/T. A manually activated deluge system, located at the top of the' high ceiling vault, and a manually activated closed head sprinkler system, located within the tunnel, covers Fire - Area 1. Portable extinguishers and hose stations are available in each area for firefighting purposes. 5.2.2 Emergency Switchgear Room (Fire Area 3) Fire detection consists of ionization detectors in conjunction with a l . manually actuated total flooding Halon system. There are also portable I' extinguishers located within the area and a hose station located in the turbine building at the door to the emergency switchgear room. p 5.?.3 Control Room (Fire Area 5) l The control room has ionization smoke detectors mounted at the ceiling. There is no automatic suppression system. 5-4
_ Table 5.3 Surry Fire Areas Containing Safety Related Components Fire Area Physical Description 1 Outside containment-penetration vault; Cable tunnel; Service building cable vault. 3 Emergency switchgear room (Elev. 9 ft 6 in. - Service Building) contains switchgear area, 2 battery rooms, and a relay room, as well as the auxiliary shutdown' panel. 5- Main control room (Elev. 27 ft) in the Service Building for operation of primary and secondary systems of each unit. 6 Emergency Diesel Generator Room #1 for Unit 1 (Elev. 27 ft) in the Service. Building. 7 Emergency. Diesel Generator Room #2 for Unit 2 (Elev. 27 ft) in the Service Building. i 8 Emergency Diesel Generator Room #3'as backup for Unit 1 or 2 (Elev. 27 ft).in the Service Building. 15 Primary Containment for Unit 1, multilevel structure with floor elevations of 46 ft 4 in., 27 ft 7 in., 18 ft 4 in., 13 ft (partial elevation only), and 3 ft 6 in., with personnel airlock access hatch at the 45 ft 10 in. elevation of the. auxiliary building. l i
Table 5.3
-Surry Fire Areas Containing Safety Related Components (Concluded)-
1 Fire Area Physical Description 17 . Auxiliary Building, Fuel Building, and Decentamination Building. The buildings are located side by side in a north-south orientation, with
'the auxiliary building to the south, the decontamination building to t the north, and the fuel building in the center. The auxiliary building is a four-story structure consisting of .the 2 ft, 13 ft, 27 ft 6 in.,
and 45 ft 10 in. elevations. 19 This fire area, collectively referred to as the sa feguards area, consists of the main steam valve house, containment.' spray pump house, and the safeguards area. 31 .The Turbine Building. consists of three primary elevations: the 9 ft y 6.in. basement, the 35 ft mezzanine, and the 58 ft 6 in. turbine deck. 2 45 Mechanical Eqvipment Room #3 is located in the service building basement at elevation 9 ft 6 in.
'54 The Charging Pump Service Water Pump Room is on the 9 ft 6 in. level adjacent to the main turbine building and mechanical equipment room #3. !
r l l
.. . . - -_ -~ _ --, _ . .
Manual-fire suppression in provided for by fire extinguishers interior to the control room and a hose station located in the turbine building.
-5.2.4 Emergency Diesel Generator Rooms (Fire Areas 6, 7, and 8)
Each emergency diesel generator room is equipped with a total flooding . Iow pressure carbon dioxide (CO2 ) fire suppression system. The system can be manually actuated either locally at the CO2 control panel directly outside the door or remotely in the control room. Doors and dampers are equipped with blow of f caps to close upon CO 2 initiation. Rate compensated heat detectors (190* F) are located in each room and provide remote annunciation to the control room. All the EDO rooms have at 1 cast two fire extinguishers. IIose stations and a portabic firefighting foam cart are located nearby in the turbine building corridor. 5.2.5 Primary Containment (Fire Area 15) The boundary fire barrier for Fire Area 15 is of a heavy reinforced concrete construction with an inherent fire rating in excess of three houra. Fire detection consists of heat, smoke,.and duct detectors, which are alarmed in the control room. There are portable fire extinguishers located just outside the containment at the personnel hatch. Dry hose standpipes are available inside containment. Adequate hose lengths to reach all portions of the containment can be brought in during emergencies. 5.2.6 Auxiliary Building (Fire Area 17) An automatic detection system that alarms in the control room is provided in the auxiliary building portion of Fire Area 17. Smoke detectors are located .-on each elevation of the auxiliary building, consisting of both ceiling mounted smoke detectors and duct detectors. One ceiling mounted detector and one _ duct detector is provided in each charging pump cubicle. Two ceiling mounted detectors are installed above each unit's charging pump-component cooling water pumps. Portable extinguishers and manual hose stations are provided on all levels of the auxiliary building for fire fighting purposes. 5.2.7 Safeguards Area (Fire Area 19) The safeguards area is equipped with ionization smoke detectors. All of the smoke detectorn alarm in the control room. In-addition, each area contains portable extinguishers. An exterior hose station, located in the yard, is availabic for manual firefighting purposes. 5.2 8 Turbine Building (Fire Area 31) A full area automatic sprinkler system is installed on the 35 f t and the 9 ft 6 in. elevations. Upon sprinkler system water flow, an alarm is transmitted to the control rcom. 5-7
l The major lube oil components have individual deluge systems actuated by heat detectors. These also provide annunciation to the control room upon system actuation. There are a number of portable fire cxtinguishers and hose stations located _in the turbine building as well as a portable firefighting foam cart. 5.2.9 Mechanical Equipment Room tt3 (Fire Area 45)- Smoke detectors are provided in Fire Area 45. These detectors alarm in the control room. Some of these smoke detectors are designed to operate MOVs in the event of a fire to allow the redundant charging pump service water pumps to operate. There are fire extinguishers in the area and hose stations are located in the turbine building at the door to the emergency switchgear room. 5.'2.10 Charging Pump Service Water Pump Room (Pire Area 54) Fire detection consists of two ionization detectors which alarm in the control room. This area could be entered from the turbine building for l firefighting purposes, i 5.3 Initiatinn Event Frecuencies Data' on fires in Light Water Reactors have been analyzed in several studies (Refs. 1,2,3). Although they have been done independently, they have some common aspects. For example, almost all studies have used License Event Report (LER) data from the Nuclear Regulatory Commission (NRC). All have reported the overall frequency of fires of approximately 0.16 per reactor year on a plant wide basis. To determine fire initiating event frequencies, there are two kinds of information needed: (1) the number of fire incidents that have occurred in specific compartments during commercial operation, and (2) the number of compartment years that the nuclear industry has accumulated. Most of-the data for the first part comes from reports of insurance inspectors to American Nuclear Insurers (ANI), although other sources are also used, e.g., the U.S. Nuclear Regulatory Commission. While the NRC requires the reporting of fires-that, in some way, affect the safety of the plant, the ANI has more stringent requirements in the sense that all fire events must be reported. Compartment years are computed' by adding the age of all compartments (within a certain category of compartments) of units that were in commercial operation by the end of June 1985. The age is defined - as the time between first commercial operation and the end of
- June 1985 (or date of decommissioning). The combination of specific fire locations and compartment age is given in Table 5.4. Even though fire events that occurred when the plant was shutdown were used, an event was only included if it could be postulated that it could also occur when the plant was at power. Eight areas are typically found in nuclear power plants. These are (1) the control room, (2) cable spreading room, l
1 58 l
(3) diesel generator room, (4) reactor building, (5) turbine building. (6) auxiliary. building, (7) electrical switchgear room, and (8) battery room. In most plants, the first three areas and the electrical switchgear room and battery room are single compartments while the other three are typically large buildings. A listing of all generic data used for each of the four types of fire areas that survived screening is given in Appendix E. Table 5.4 Statistical Evidence of Fires in IRRs (As of June 1985) Number Number of of Fires Compartment Years Area r T Control Room 3 681.0 Cable Spreading Room 2 747.3 Diesel Generator Room 37 1600.0 Reactor Building 15 847.5 Turbine Building 21 654.2 Auxiliary Building 43 673.2 Electrical Switchgear Room 4 1346.4 Battery Room 4 1346.4 To obtain fire zone specific initiating frequencies, a partitioning method is required. Partitioning allows the analyst to subdivide the frequency of fire occurrence from a large building (e.g., auxiliary
- building) to a specific room or area within that building. Also, further partitioning can occur within a specific room or area. One method of partitioning is comprised . of ratioing the areas of fire zones within a building- (e . g. , auxiliary building) . The assumption here is that the probability of fire occurrence is dependent only upon the amount of area a fire zone contains. Another method of partitioning would look at each fire zone and analyze factors important- to probability of fire initiation. These factors are the amount of electrical components and
-cabling, the fire loading, whether the fire zone is controlled, and how often the fire zone is occupied.
5-9
The fire events and operating years for the eight plant areas were obtained using the fire data base developed by Wheelis (Ref. 4). To determine operating years for electrical switchgear roomc and battery rooms, auxiliary building operating years were doubled. A survey of all U.S. light water reactors indicated that there is an average of 2.25 trains of emergency switchgear and their associated batteries per plant. However, it is known that some plants such as Surry locate both trains of their emergency switchgear in one fire zone. So it was assumed that an average number would be close to two per plant. To aid partitioning within a large building or within a specific fire zone in that building a checklist was used on the initial plant visit to ' determine the most probable fire initiating sources. Also, data on past fire occurrences was thoroughly reviewed. For instance, control room data indicate that fires have only occurred in cicctrical cabinets. Therefore, area ratios were developed based on cabinet area within this respective area. Since transient combustible initiated fires have never occurred, they were eliminated from further consideration. The generic fire occurrence data was updated using a method developed by Iman (Ref. 5) to determine plant specific fire occurrence frequencies. This Bayesian approach models the incidence rate for each plant relative to the incidence rates of all other plants, and the posterior distribution is found for the incidence rate for each plant. For this analysis the gamma distribution is used as a model, although many other distributions could be used. In this way plant specific fire initiating event frequencies and distributions were developed. Table 5.5 lists the Surry Unit #1 specific fire initiating event frequencies for the five types of fire areas with contributions to core damage frequency of greater than 104 per year. It should be noted that fire frequency for the CPSWPR was based strictly on generic data. There was no ready means of determining how many pump rooms there are on average per plant. Therefore, two were assumed and auxiliary building operating years were doubled. Since a breakdown of the number of pump rooms per plant could not be obtained, the distribution for the CPSWPR was assumed to be lognormal with an error factor of three. Surry Unit #1 had no recorded fire occurrences in any of .u five areas (cable spreading room, control room, electrical switchgear room,
. auxiliary building, pump room) that - survived the screening process.
g Surry, however, did have four fire occurrences between 1972 and 1980 that occurred in other plant areas. These fires were located in the safeguards area, transformer yard, diesel generator room, and in a local control tunnel for a control room chiller. Since none of these areas survived the screening analysis described in Section 5.5 no attempt was made to update their fire initiating event frequency. ) 5-10
. . . - - - . . . - . -- . - - . __~ - ~ - --- - - .. . .
5.4 Determination of Fire-Induced "Of f-Normal" Plant States One of the m at critical steps in a fire analysis is to determine on a plant specific basis which events in a wide range of possible initiating events have the potential to be induced due to a fire occurrence. As in the NUREG 1150 internal events analysis, a comprehensive list of initiators was identified for further study. It is known from a review of previous fire PRAs that only a limited set of initiating events have the potential to be significant contributors to fire induced core damage frequency. Typically, initiating events such as large or rnedium LOCAs caused directly - by the fire have not been analyzed because the vulnerabilities of a piping system or tanks to fire events are considered to be insignificant. Tabic 5.6 lists the initiating events that were analyzed during the screening process and provides a brief explanation as to why a particular initiating event was included or excluded from further study. The same fault trees and event trees which were used in the internal events analysis were utilized in the fire analysis. Thus, the level of analytical detail was consistent with the 1cvel in the internal event analysis. 5.5 Detailed Description of the Screeninn Analysis A comprehensive screening analysis is required to reduce the number of potential fire-induced scenarios to only those which have the potential to be probabilistically significant to core damage frequency. The screening analysis is composed of the following four steps; Step 1. Identification of Relevant Fire Zones Fire -zones containing equipment or cables associated with safety related systems which mitigate the ef fects - of the unscreened fire-induced "off-normal" plant states were identified. All other fire zones wera then eliminated from further analysis. This resulted in the fire zones which are described in Section 5.2. Step 2. _ Screen Fire Zones Based on Fire Area Analysis The remaining fire zones underwent a fire area analysis (location mapping) of components as well as control-and power cables for a limited set of " vital" components that were located within these areas. This information resulted in a transformation block used in conjunction with the SETS computer code (Refs. 6, 7) to solve all f ront line systems and then solve all of the ider.tified sequences (Table 5.6) of Section 5.4 l 1 5-11 l l
_ = _ Table 5.5
)
Surry Fire Initiating Event Frequencies (/yr) l 5th 50th 95th Mean Percentile Percentile Percentile Fire Area 1.2E-6 9.6E-4 7.4E-3 Control Room 1.8E-3 3.OE-6 1.8E-3 1.6E-2 Cable Vault / Tunnel 7.5E-3 2.0E-5 2.4E-3 1.7E i Electrical Switchgear 8. 0 E- 3 Room 2.7E-2 5.9E-2 1.6E-1 Auxiliary Building 6.6E-2 3.7E-3 (Iognormal EF - 3) Pump Room ? N
Table 5.6 Surry Fire Induced Initiating Events Analyzed Initiating Event Comments Loss of Offsite Power Offaite power was excluded because redundant trains were found to be widely separated when routed through common areas which were of sufficient size to preclude buildup of a hot gas layer. Transient with PCS Similar to the seismic methodology if no initially available other initiator could occur it was assumed rhat the operator would either manually scram the plant or an automatic trip would occur due to the fire. Transient induced seal The probability of one unisolatable stuck-or stuck open open relief valve was sufficiently high PORV LOCA (>10-5 demand) to require further analysis. V-Sequence LOCA Screened from further analysis because no probabilistically significant mechanism could be identified which had not been addressed by the Appendix R submittal. The fire occurrence frequency for each zone was set to 1.0 and, given a fire, all components within that zone were assumed to fail. The output of this process was accident cut sets which has fire zone combinations as well as random failures (i.e., not fire related) included. Truncation -of cut sets at a random failure probability - of 10 was accomplished. This is equivalent to truncation of internal event cut - sets at approximately 10 8 since the fire frequency is arbitrarily set for screening purposes to 1.0. Cut sets which required three or more fire zones were eliminated. This was deemed appropriate since these cut sets imply the failure of two or more three hour rated fire barriers. Cu, e"s which contained two fire zones were screened on the following three criteria: (1) no adj acency between zones, (2) no penetrations in the adjacency between zones, -and (3)_if there were penetrations by numerical culling with barrier penetration failure set to a screening value of 0.1. It is known from the analysis of many fire barriers-that typical failure rates are on the order of 10-2 to 10-3 Therefore, this screening value has been set high enough to insure potentially important fire zone combinations are not truncated-in this screening step. L 5-13 l'
l I l I One additional important piece of information gained from these cut sets was identification of the remaining plant locations where zone to zone barriers needed to be analyzed. Dominant cut sets which contained adjacent fire zones were analyzed for barrier failure in the quantification process. Step 3. Cull Fire Zones on Frequency Cut sets not climinated in the first two screening steps were resolved with fire zone spe:ific initiating event frequencies that were calculated as described in Section 5.3. Also, operator recovery of non fire related random failures was included. For screening purposes only all short term (less than 24 hours) recovery actions (of non fire failures) were increased from their respective internal events probabilities by a factor of five to allow for the additional confusion of the fire situation occurring in conjunction with other random failures. If recovery actior,s were long term (greater than - 24 hours) no modification to internal event probabilities were deemed appropriate. It is felt that by this time the fire will be extinguished and any spurious signals will have termilated in open circuits. It must be noted that Steps 2 and 3 of the screening process reduced the number of cut sets under consideration by at least two orders of magnitude. Also, there were only a few remaining sequences which had not been screened, Step 4 Confirmatory Plant Visit For those remaining fire zones all fire related failure scenarios were identified.- A scenario can be thought of aa a combination of one or more fire related equipment failures - within a fire zone with or without additional' non-fire related (random) failures outside of the- fire area. These failure combinations must minimally lead to core damage. Each fire zone can have one or more scenarios depending on the equipment combinations-which must fail due to the fire in that particular area. A second plant visit was then conducted to determine which of these scenarios were valid based upon cable or equipment locations within ' a particular fire zone. For instance, if a given scenario required the fire-related failure of cabling. for components A and B and- it could be j_ shown that these cables were always separated by greater ' than 40 f t
-within a room of sufficient size to preclude buildup of a hot gas layer, or one of the component's cabling was in-a 3 hr rated fire wrap, then these types of scenarios were eliminated from further consideration.
-Past experience with fire code calculations, which is discussed in the following section, and fire testing, provided much of the basis for assessing the validity of the scenarios. About one-quarter of the remaining cut sets (scenarios) were eliminated as a result of this confirmatory plant visit.
5-14
Those scenarios remaining after screening on physical location of components or their associated cabling within a fire zone was determined had fire propagation calculat. ions run to determine equipment damage. It must be noted for some fire areas that the exact location of a particular components cabling could not be determined. In such cases a best estimate of cable routing was used, 5.6 Fire Propagation Modelinn The COMPBRN fire growth code (Ref. 8) was used to calculate fire propagation and equipment damage. COMPBRN was developed specifically for use in nuclear power plant fire PRAs. The code calculates the time to damage critical equipment given that a fire has started. This failure time is then used in conjunction with experiential information on fire suppression in nuclear power plants to obtain the p robab il ' ty or frequency that a given fire will cause damage which leads to cort. damage before the iire can be suppressed. The latest version of the code, COMPBRN III (Ref. 9), with some additional modifications was used for the calculations. COMPBRN follows a quasi static approach to. simulate the process of fire during the pre flashovar period in an enclosure. COMPBRN uses a zone model, breaking the fire environment into three zones: flame / plume, hot gas layer, and ambient (see Figure 5.1). Simple fire and heat transfer models and correlations are employed to predict the thermal environment as a function of time. The thermal response of various targets in the fire scenario' is modeled to predict the amount of time for a fire to damage or ignite critical equipment. The critical equipment is generally taken to be a cable tray carrying cables necessary for safe shutdown of the plant, although other critical components such as pump motors may be modeled. The original version of COMPBRN, now referred to as COMPBRN 1, has been used to calculate damage times in the majority of fire PRAs to date. However, the code -calculations are thought to be highly conservative-due to the neglect of-heat losses from the targets. A-critical assessment of-the code detailing this and other problems has been performed (Ref. 10). In= response to these problems with COMPBRN. I, two later versions of the code were developed: COMPBRN II and COMPBRN . III (Ref. 9). Neither of the later versions of the code has been extensively validated or compared to data, but presumably represent various degrees of improvement.
-As a part of a recent s tudy (Re f . 3) on nuclear power plant fire risk assessment, the latest version of the code (COMPBRN III) was selected-to _
requantify fire damage times from several fire PRAs. ' Initial attempts to use COMPBRN _ III in -. the requantification resulted in the observation of problems- with and nonphysical behavior of the code. Many of the code calculations could not be explained on a physical basis. As a result of the observed nonphysical behavior of the code, an effort was undertaken to identify problem areas and to suggest and implement modifications to the code which would make the code predictions more reasonable on_ a physical basis. It was this modified version of the COMPBRN code which 5-15
/
1-4 COMPBRN Modeling ; Celling l C..l. t//,t/ ,
**Y '
I
...................~eas I I I /
') Layer
/ -
ffff// i w \ / k \. I
./ "
i Plume / Convection i and f Radiation -! k .[ Medeied
\ /
l / ,
\ /
\ l Fire l
;t
- Fire Modeled as a Cylinder r
- Cable Trays Discretized into Fuel Elements
- Hot Gas Layer Effects !
- Figure 5.1. COMPBRN Zone Model :
I i
was used to provide the fire propagation analysis for this report. References 3 and 11 provide detailed discussions of the goblems noted and recommended modifications for the COMPBRN III code. The following is a brief listing of the major problems which were identified and addressed .in the modified version of the code:
- a. An' error, and nonconservative assumption, exists in the forced ventilation hot gas layer model, predicting low hot gas layer temperatures.
- b. Radiative heat transfer directly above the flame is not modeled, yielding cooler temperatures directly above the flame than off to the side of the flame,
- c. Two errors in the calculation of view factors overpredict the heat radiated to targets to the side as compared to objects directly above the flame.
- d. Only convective heat transfer, not the dominant radiative heat ,
transfer for objects directly engulfed in the flame, is modeled. Time to ignition is highly nonphysical.
- c. The conduction algorithm is unstable, often resulting in premature termination of the code, especially for cases involving i objects in the flame or thermal response of barriers.
- f. The mass burning rate of burning objects is underpredicted due to lack of thermal feedback modeling, g, Cable insulation ignition and damage failure threshold criteria are not currently well understood and the results are quite sensitive to the input parameters chosen.
Both small and large fires were postulated in the calculations. A small
. fire was assumed to be 2 ft. (.61 m) in diameter and consist of 1 gallon (3.8 1) of oil. A large fire was assumed to be 3 ft (.91 m) in diameter and consist of 10 gallons (38 1) of oil. Analysis of a data base on transient combustibio ' fuel sources found at nuclear power plants
- indicates that oil sources less than or equal to 1 gallon (3.81) were found approximately 70 percent of the time. Oil sources larger than this were found roughly 30 percent of the time. A similar partitioning between small and large quantities- in terms of heat content (BTU or KJ) can . .be made for other credible transient combustible - sources such as solvents or trash paper. Again, analysis indicates that a 70/30 partitioning between small and large fuel sources is appropriate (within i 10 percent). It can also be shown that 10 gallons (38 1) of oil bounds any large solvent or - trash paper combustible source in terms of heat.
content and is, therefore, an appropriate upper bound on transient combustible fuel source size. A walkdown of the Surry Power Plant was performed to obtain vital information for the COMPBRN calculations. This information included the
- Transient Combustible Fuel Sources Found at Nuclear Power Plants (Data),
Letter Report by W. Wheelis, Sandia National lebs, July 1984 5-17
location of critical equipment and cable trays, separation between redundant trains, types of cable present, and any shielding or fire barriors which may be present. Several " pinch points" were identified where critical cables from redundant trains passed from one room to another, Thin sheets of corrugated aluminum were observed on top of many cable trays. Ilowever, because of i r r. low melting point this aluminum was neglected in the COMPBRN calculatiens to be conservative. Similarly, in several cases the power cables to critical pumps were routed in metal sleeves. In the COMPBRN calculations, these cables were assumed to be incapable of igniting. Iloweve r , damage was assumed to occur when the surface temperature reached the temperature corresponding to cabic failure. Cable insulation ignition and damage thresholds are currently not well known (Ref. 12). For this study, a cabic insulation ignition temperature of 773*K (932'F) was assumed along with a damage temperature of 623'K (662*F). For the large fire simulations these thresholds are not as critical to the fire damage time calculations because of the intensity of the flames. A list of input parameters for the COMPBRN calculations is shown in Table 5.7. These parameters were selected to represent typical qualified cabic insulation. It was assumed that the cabling in the areas of interest included typical brands of nuclear qualified cable insulation materials, such as Rockbestos Firewall III, Brand Rex, or Okonite. Because of the good flame resistance properties of these cables, no self-ignited (electrically initiated) cable tray fires were postulated. The COMPBRN results are shown in Table 5.8 for the critical areas noted in Section 5.2. One general comment is in order: The modified version of COMPBRN III used in these calculations predicts that it is very difficult to ignite qualified cable insulation unless the cables are actually in the flames. For cases where the cables are not within the flames (or very close to them), the modified version of COMPBRN - 111 predicts that they will not be damaged (infinite damage time). One exception to this is the charging pump service water pump room which is so -small that the hot gas layer from a fire anyvhere in the room would quickly damage critical cables. For cases where the cables are immersed in the flames from a transient combustible source, the modified version of COMPBRN III predicts that these cables ignite very quickly (1 to 4 min), The modified version of COMPBRN III also calculated that a small fire would have to occur within 2 ft (.61 n) of a cable tray (horizontal distance) to damage it. Large fires were capable of damaging cable trays if they were located within 3 ft (.91 m) horizontally of the cable tray. Using these results the area in which a fire would have to occur to damage critical cables can be estimated. An area ratio can then be calculated by dividing this area by the total floor area of the room. This reduction factor can then be multiplied by the initiating frequency to estimate the frequency of fires which occur in a critical portion of a given room. 5-18 i
Table 5.7 Modified COMPBRN 111 Input Parameters fable Insulation Parameters Density 1715 kg/m 3 Specific Heat 1045 J/kg-K Thermal Conductivity 0.092 W/m-K Heat of Combustion 1.85 2.31E7 J/kg Combustion Efficiency 0.6 0.8 Critical Temperature Piloted Igr
- tion 773*K Spontaneous Ignition 773'K Damage 623*K Surface Controlled Burning Rare 0.0001-0,007 5 kg/m2 .s Burning Rate Radiation Augmentation 1.86E-7 kg/J m 2 Radiative Fraction 0.3-0.5 Smoke Attenuation Factor 1.4 Reflectivity 0.1-0.3 011 Parameters Density 900 kg/m 3 Specific Heat 2100 J/kg-K Heat of Combustion 4.67E7 J/kg Combustion Efficiency 0.9 Surface Controlled Burning Rate 0.06 Radiative Fraction 0.3-0,5 Mass of 011 3.4-34.0 kg The area ratios for the rooms of interest are presented in Table 5.9.
Note that for the charging pump service water pump room, an area ratio of 1.0 was assumed because the small size of the room enables the hot gas layer from a fire anywhere within the room to damage critical cables. 5.7 Barrier Failure Analysis In the unscreened cut sets where a potential for barrier failure had been identified, barrier failure probability was estimated using barrier failure rates developed as described below. Barriers were grouped into three types: (1) fire doors, security doors, water tight doors, and fire curtains; (2) fire dampers and ventilation dampers; and (3) penetration seals and fire walls. The data base contains 628 records f rom when construction began on any given plant to the end of June 1985. The number of barriers of each type at a plant is required to estimate the rate at which a specific component fails. The number is not known precisely for each plant, but a nominal figure that has oeen estimated for each barrier type is given in Table 5.10, 5-19
Tabic 5.8 Time to Damage Critical Cables (minutes) Using the Modified Version of 00MPBPJ1 111 Small Large Area Scenario Fi rt Fire Tray at 12 ft
- 2 Auxiliary Building Cabic Vault Tray at 7 ft 3 3 Tray at 4 ft 3 3 Emergency Tray at 7 ft 3 3 10 ft and 12 ft Trays
- 3 Switchgear Room Relay Room
- 4 3 Safeguards -Auxiliary Feedwater
- 3 Pinch Point
- 3 Mechanical Equipment Cable to Pump 1 1 Room 3 Charging Pump Junction Box 1 1 Service Water Pump Room Anywhere Else 3 2
*No damage predicted (infinite time)
The statistical uncertainty of each estimate, reflecting sampling variation and plant-to plant variation, is represented by 90 percent confidence bounds. These estimates and confidence bounds are given in Table 5.11 where units of both estimates and bounds are failures / year. During the confirmatory plant visit scenarios which required barrier failure had those barriers inspected. No plant specific vulnerabilities vero noted as a result of this inspection which would require modification of generic barrier failure rates. After multiplying barrier failure rates by the number of penetrations at each appropriate fire zone adjacency and utilizing the probabilities developed in screening Step 4, all remaining barrier failure scenarios did not survive the 10-S per year frequency screening criteria. S.8 Recoverv Annivsis For those remaining cut sets which survived the screening process and where the COMPBRN code predicted fire damage would occur, recovery of random failures and credit for extinguishment of the fire before the COMPBRN predicted time to ftre damage was applied. 5-20
l Table 5.9 Critical Area Ratios Critical Area Ratio Area Scenario Small Fire larne Fire Auxiliary Building Tray at 12 ft NA 6.34E-4 Cabic Vault Tray at 7 ft 0.022 0.027 Tray at 4 ft 0.022 0.027 Emergency Tray at 7 ft 0.027 0.033 Switchgear Room Tray at 10 ft NA 0.027 or 12 ft Relay Room NA 0,074 Safeguards Auxiliary Feedwater NA 8.93E-3 Pinch Point NA 5.36E 3 Mechanical Equipment Cabic to Pump 0.1 0.1 Room 3 Charging Pump Service Junction Box 1.0 1.0 Water Room NA - Not applicable because a small fire will not result in damage for this scenario. Recovery of random failures (non fire related) was treated in a similar fashion as in the internal events analysis. All operator recovery actions that were used in the internal events analysis were inspected for use whnre appropriate in the remaining cut sets. If a sequence was long term (greater than 24 hours), two recovery actions were allowed. In short term (less than 24 hours) sequences only one recovery action was allowed. A recovery action was chosen if the possibility of multiple recovery actions was present and on a hierarchy based on recovery probabilities established by the internal events analysts. For short term sequences recovery action probabilities were modified when deemed appropriate. In the areas where firefighting activity takes place, no credit was given for local recovery actions until after the fire was extinguished. In non affected areas, local recovery was allowed for valve manipulation or pump operation when damage to power cabling of an applicable component had not occurred. 5-21
Table 5.10 ' Approximate Number of Barriers at a Plant Type Nominal 1 150 2 200 3 3000 Table 5.11 Estimates of Singic Barrier Failure Rates 5% 95% Barrier Barrier / Confidence Confidence Type Unit Estimate Bound Bound 1 150 7.4E 3 0.0 2.4E-1 2 200 2.7E 3 0.0 2.2E-1 3 3000 1.2E-3 0.0 3.7E-2 In conjunction with human factors analysts and the " Handbook of Iluman Reliability Analysis With Emphasis on Nuclear Power Plant Applications" (Ref.13), any additional recovery actions not developed by the internal events recovery procedure were quantified. Only one additional recovery action was added for the Surry analysts. This recovery action was necessitated by failure of control cabling in the control room requiring control of the plant from the remote shutdown panel. Even though explicit procedures - were in place for this situation, a high stress recovery probability was applied. This was deemed appropriate due to timing of the sequence (less than one hour) and the fact that some amount of time would be required to make the decision to abandon the control room and man the remote shutdown stations. The probability of manual non suppression of a fire before the COMPBRN predicted time to damage was quantified using the Wheelis' data base (Ref. 4) Valch contained information on 69 fire events which had time to suppression ast.ociated with them. As part of the Fire Risk Scoping Study (Ref. 3) a distribution was fit to this data. A probability of non-suppression was then associated with any COMPBRN predicted time to fire damage. 5-22
- . . . - . . - - - . ._.. - - - - . - - . - - - - - . . . . . ~ .. ,
Credit was also given for automatic suppression systems in areas where they were located. In the case of Surry the only unscreened area which contained such a system was the cable vault / tunnel. Generic reliability data indicates approximately a 96 percent success rate for such systems (Ref. 14). Ilowever, a modification to this- reliability value was deemed appropriate due to the predicted short time to damage (~3 minutes), the half minute system actuation time delay, and the fact that five fixed temperature (190'F) heat detectors actuate the system and none was in close proximity to the postulated fire. S.9 Uncertaintv Analysis Distributions on fire frequency, fire suppression probability, fire code calculations, random failure probability, barrier failure probability, and operator recovery actions, generated uncertainties on fire induced core damage frequencies. The uncertainty of these values was propagated through the accident sequence models using two computer codes. A Latin liypercube Sampling (Ills) algorithm was used to generate the sampics for all of the parameter values (Ref.15), The Top Event Matrix Analysis Code (TEMAC) was used to quantify the uncertainty of the accident sequence equation using the parameter value sampics generated by the Ills code (Re f. 16) . IJIS is a constrained Monte Carlo technique which forces all parts of the distribution to be sampled. ~ The Ills code is also flexible in that it can sample a variety of random variable distributions. Furthermore, parameter distributions for similar events were correlated. For example, if two similar components (e.g., MOV XX-FTO and MOV YY-FTO) are modeled from the same probability distribution, then the sampling of these two distributions is perfectly correlated, meaning the same value is used for both events in a given sample member. For basic events which are modeled ' with' very similar_ but slightly different distributions - (e.g., MOV XX fails to remain closed for 100 hours and MOV YY fails to remoin closed l for 200. hours), the Ills code permits an - induced correlation between the l samples. iloweve r , LilS does not allow the correlation- coef ficient for this case to be equal to 1.0. IJIS does ' permit sampling with - a coefficient of 0.99 in these cases. TEMAC uses the Ills parameter samples and the accident sequence equations (cut sets) as input to quantify the core damage estimates. TEHAC generates a sample of the accident sequence frequency, a point estimate of the frequency, and various- importance measures and ranking for the base events. l Uncertainty on fire initiating event frequency was developed - when the generic fire frequencies were updated using Surry specific data. This process which was briefly discussed in Section 5.3 is covered in more detail in Reference 5. 5-23 l
Uncertainty on fire non suppression probabilities (Q(r ))o was addressed by modification of COMPBRN predicted time to damage. The COMpBRN Code predicted time to damage and its associated non-suppression curve probabili ty were taken to be a best estimate of a maximum entropy disturbed variabic. Fifteen minutes were added and subtracted from the COMPERN predicted time to allow for uncertainty in its result and the uncertainty in the probability of non suppression distribution. These probabilities were taken as a minimum and maximum of the maximum entropy distribution respectively. Uncertainty associated with the fire size estimate factor ( f.) . was developed utilizing information associated with an I&E inspector report (see footnote on pg. 517) on a survey of different types of combustibles and their amounts found in nucicar power plants. Two fire sizes, a large and small fire were modelled as described in Section 5.6. These fire sizes (BTU content) were compared to the distributions on possible fire sizes developed for the different combustibles from the I6E data. The best estimate for percentage of fires that were either large or small was taken from an average of the different types of combustibles for an equivalent BTU 1cvel fire modelled by -COMPBRN, This probability was
-assumed to be the best estimate value of a maximum entropy distribution.
Maximum _and minimum probabilities for this distribution were assumed to be based on one individual type of combustible with either the maximum and minimum percentage corresponding to applicabic fire size (BTU rating). Random failure events and operator recovery actions were treated identically as in the internal events analysis. Uncertainties and types of distributions.were not modified for the fire analysis. All other factors and their associated uncertainties are not common to all fire sequences and will be addressed individually in the approp: Tate '
- subsections of Section 5.10.
l 5.10_ Description of Unscreened Fire-Induced Core Damare Scenarios and i Their Associated Fire Areas 5.10.1 Introduction This section will describe the fire scenarios- and their associated fire l' zones _ which are listed in Table 5.1. All other fire zones and all l adj ac ent fire zone combinations were either screened as described in Section 5.5 or had scenarios that dropped below 10-8/yr af ter either. operator recove ry - o f non-fire related failures, -COMPBRN code calculations, or barrier failure probabilities were applied. 5.10.2 Auxiliary Building One fire scenario in the auxiliary building remained af ter screening. This scenario was a large fire on the 13 it elevation which irrecoverably 5 24 ' l l
1-- - . .. i I damaged power or control cables for both the HPI and CCW systems. These fire-related failures with no additional random failures required led to a reactor coolant pump seal LOCA occurrence. The recovery for this particular scenario required the operation of two manual HPI system cross connect valves which were located in the immediate vicinity of the large fire, No recovery was allowed until 15 min after the fire was extinguished. The core damage equation is as follows:
-A aur a fsQ(rC) R op 4 f cm where b - fire-induced core damage frequency for the auxiliary building A,o - frequency of auxiliary building fires j
- f. - area ratio within the auxiliary building where critical damage occurred f, - severity ratio (based on generic combustible fuel loading) for a large fire Q(rg ) - that percentage of fires within the suppression data base where the fire was not manually extinguished before the COMPBRN predicted time to critical damage occurred R, - f ailure to cross connect of Unit 2 high pressure injection system to either prevent seal LOCA occurrence or mitigate its effect Table 5.12 gives the values of each of these factors as well as their associated distribution and upper and lower bounds. In the case of log-normally distributed variables the upper and lower bounds represent the 95th and 5th percentiles of the distribution, respectively, while the best estimate represents the mean value.
5.10.3 Cable Vault / Tunnel The one remaining scenario which survived screening and is similar to the one described for the auxiliary building in that the postulated fire irrecoverably damages power or control cables for both the HP1 and CCW systems leading to a seal LOCA. Credit was takcn for the automatic CO2 system suppressing the fire before critical damage occurred. COMPBRN predicted 3 min time to damage for thin particular ccenario. The automatic CO 2 system is actuated by fixed ter perature heat detectors at 190'F. There is one heat detector located Two at the end of the critical area of influence for this scenario. 5-25
Table 5.12 Auxiliary Building Fire Scenario Factors and Distribution Factor Distribution Lower Bound Best Estimate Upper Bound A,o gamma 0.027 0.066 0.16 f, maximum 2.4E-4 6.3E 4 1.lE 3 entropy f, maximum 0.19 0.30 0.67 entropy Q(ro) maximum 0.69 0.80 1.0 entropy R op maximum 0.19 0.26 1,0 entropy others are located such that v. nilation flow would force the hot gas layur in their direction, The system actuation delay time to allow for evacuation is 30 s. Therefore, the heat detectors must respond to fire ignition and the CO 2 system must suppress the fire within 2.5 min, to prevent critical damage. For these reasons, system reliability data for automatic CO 2 suppression systems was modified to account for this relatively short time to prevent critical damage. Operator recovery for this scenario is similar to that for the auxiliary building scenario except that the fire is not in the immediate vicinity or even same fire area as where the local recovery actions must take place. Also, since no control room operators respond to the fire itself the same recovery value for operator action was applied as was used in the internal events analysis. The core damage equation is as follows: CM ~ ^CSR a s S(# G AUTO op where 4m - fire-induced core damage frequency for the cable vault / tunnel A CSR = frequency of cable vault / tunnel fires i 5-26
1 l f, - area ratio within the cabic vault / tunnel where critical damage occurred j f, - seve rity ratio (based on generic combustibic fuel loading) Q(ro ) - that percentage of fires within the data base where the fire was not manually extinguished before the COMPBRN predicted time to critical damage occurred R op - failure to cross connect of Unit 2 high pressure injection system to either prevent seal LOCA occurrence. or mitigate its effect QAuto - probability of the automatic CO2 not suppressing the fire before COMPBRN predicted time to critical damage occurred Tabic 5.13 gives the values of each -# these factors as well as their associated distribution and upper and lower bounds, 5,10.4 Control Room One scenario survived the screening process for the control room, As was
';he case for the auxiliary building and cable vault / tunnel, no additional random failures were required to lead directly to core damage. This scenario was a fire interior to benchboard 1 1 leading to the spurious actuation of -one PORV located on this benchboard. Because of' the cabinet configuration within the control. room and based on Sandia cabinet fire tests (Ref, 17), the fire was assumed not to spread and damage any components outside of benchboard 1-1, However, due to Sandia large scale enclosure tests (Ref, 18) where smoke engulfed _a control room within 5 to 10 min, of time from ignition within a cabinet even with ventilation rates of up to 10 room changes per hour, this scenario postulates forced abandonment of the control room and subsequent plant control from the auxiliary' shutdown panel located in the emergency switchgear room.
Credit was given for quick extinguishment of the fire within benchboard 1-1 since the control room is continually staffed. None of.the four control room fires in the data base lead to abandonnent of the control room. It was assumed that one in ten control room fires would result in abandonment of the control room and a factor of ten reduction in control room fire frequency was the modification made to allow credit for continuous occupation.
- The area ratio for = fire involvement was developed ratioing the area of benchboard 1 1 to the total cabinet area in the control room. This is warranted based on fire event data that all control room fires - have
. occurred within electrical cabinets. Therefore, this is postulated to be the most likely fire ignition source within the control room.
5-27 w- am ' w',
Table 5.13 D Cable Vault / Tunnel Fire Scenaria factors and Distribu'lan Facter Matributi.nD LQwer liound Esst Estimate Upper itound Aca gamma 3.0E 6 7.5E 3 0.d16 i f, maxiinum 0.011 0.025 0.047 entropy j f, maximum 0.50 0.99 1.0 entropy Q(to) maximum- 0.69 0,80 1.0 entropy QAuto maximum 0.50 0.70 0.90 entropy R yp maxiinum 4.42 3 0.044 0.44 entropy Once abandonment of the control room takes place, operators would control the plant from the auxiliary shutdown panel, However, p0RV indication is not provided at this panel and in conversations with the utility it was learned that the PORV disabic function on the auxiliary shutdown panel is not electrically independent of the control roorn. Therefore, it was i assumed that the p0RV disabic function would fall and, consequently, the operators would be in high stress recovery mode, The core damage equation is as follows:
~ I A
CM CR a op t where (m - fire induced core damage frequency for the control room Aca - frequency of control room fires f, - probability that opet : tors will not successfully extinguish the fire before smoke forces abandonment of the control room
- f. - area ratio of benchboard 1 1 to total cabinet area within
, the control room Rep - probability that operaten will unsuccessfully recover the plant from the auxiliary shutdown panel 5 28
_ _ . ___ . _ ..._ _ m _ _.m. _ ._ .. . _ _ __._ _ _ _ _ _ _ _ _ _ _ _ . . _ I i l i i j Table 5,14 gives the values of each of these factors as well as their l as.sociated distribution and upper and lower boundc. Tabic 5,14 Control Room Fire Scenario Factors and Distribution l 1 Factor fistribul10D Lower Bound Best Estimate Unner Bound Aca gamma 1.2E 6 1,8E 3 7,4E 3
- f. maximum 0,028 0 184 0,12 entropy R op maximum 7.4E 3 0,074 0,74 entropy f, maximurn 0.01 0.1 0,25 entropy i
5.10.5 Emergency Svitchgear Room i One fire scenario remained in the emergency switchgear room after screening. This scenario was a fire that damaged either power or control cables for HPI and CCW pumps thus leading to a reactor coolant pump seal LOCA. No additional random failures were required for this scenario tc lead directly to core damage. As vas. che case for the cable vault / tunnel and auxiliary building, recovery from _ this scenario was cross connecting ilPI from Unit 2. The 4 fire itself would not affect local auxiliary building recovery actions. l Therefore, similar to the cable vault / tunnel the same probability for recovery was used as-in the internal events analysis, The core damage eque' ton is as follows: CM SWGR W)G op !f al f s1 a2 fs2I where pm - fire induced core damage for the emergency switchgear room i Asma - frequency of emergency switchgear room fires l 5-29 ...--.....-.a.--_.=..-.....:- . _ .-. _ .... - . - - . _ . - - ---... ,. . . . . - . -
f.3 - area ratio within the emergency switchgear room for a small fire where critical darage occurred f,1 - severity ratio (based on generic combus tible fuel loading) of small fires f.2 - aren ratio within the emergency switchgear room for a large fire ist - severity ratio (based on generic combustibic fuel loading) of large fires Q(f o) - that ' percentage of fires in the data base where the fire was not manually extinguished before the COMPBRN predicted time to critical damage occurred R., - failure to cross connect of Unit 2 high pressure injection system to either prevent seal 1DCA occurrence or miti6 ate its effect Tabic 5.15 gives the values of each of these f actors as well as their ascociated distribution and upper and lower bounds. c 5.10.6 Charging Pump Service Water Pump Room One fire scenario remained in the chargin6 pump service water pump room after screening. Fire related component damage included Units 1 and 2 charging pump service water pumps (CPSWP) 10A and. control power for Unit 1 CPSWP 10B. As in the internal events analysis it was assumed that orm service water pump provides insufficient cooling flow for both unf o-charging pumps given a small iDCA. Either a large or small fire will fail all cabling and components within this relatively small fire area due to a rapid buildup of a hot gas layeri This scenario requires a PORV demand and subsequent failure to reclose and isolate the leak. The internal events failure rate for tha non isolatabic stuck open PORV was used. The core damage equation is as follows: (eg - Arn Q(fo) Qroav where 3 des - fire. induced core damage frequency for the CPSWPR A3 - frequency of pump room fires (small pumps only) Q(f o) - that percentage of fires in the data base where the fire was manually extinguished before the COMPBRN-predicted time to critical damage occurred Qroxy- stuck open PORV.with failure to isolate the leak path i 5 30
Table 5.15 Emergency Switchgear Room Fire Scenario Factors and Distribution Factor Distribution L2wer Bound Best Estimat.g Unner Bound Anag gamma 2.0E 5 8.0E 3 0.017 f,3 maximum 0.02 0.039 0.099 entropy f,i maximum 0.33 0.70 0.81 entropy f.: maximum 0.051 0.10 0.24 entropy f s2 maximum 0.19 0.30 0.67 entropy Q(ro) maximum 0.67 0.80 1.0 entropy Rop maximum 4.4E 3 0.044 0.44 entropy Note that neither an area or severity ratio factor appear in the core damage equation. This is because a fire of any. size no matter where it was located in the room led to the rapid development of a hot gas layer which failed all components and cabling. Therefore, both these f actors are taken to be unity. Tabic 5.16 given the values of each of these factors . as well as their associated distribution and upper and lower bounds. Table 5.16 CPSWPR Fire Scenario Factors and Distribution Factor Distribution Lower-Bound Best Estimate Upper Bound A cgum lognormal (E.F -3) 3.7E 3 Q(ro) maximum 0.67 0.80 1.0 entropy Qmay determined by TEKAC computation 5 31
]- l 1,. 5.11 Conclusion 1 The overall fire induced core damage frequency for Surry Unit I was found j to be 1,13E-5 per year, The dominant contributing plant areas are the (a) emergency switchgear room, (b) auxiliary building, (c) control room,
- and (d), cabic vault / tunnel. These four areas comprise 99% of the total i fire risk. 1 In the case of the emergency switch6ear room, cable vault / tunnel, and the auxiliary building, a reactor coolant pump seal 14CA leads to core damage. The fire itself fails cabling for both the llP1 and CCW systems resulting in a seal IhCA,
[i For the control room, a general transient with a subsequent stuck open PORV Icada to a small LOCA, Failure to control the plant from the auxiliary shutdown panel results in core damage, , j h f 5 32
. , , . . - , , _ .. _. . . . _ . _ - _ . . , _ - _ . _ . - . . - . . _ . . _ - _ _. ~._ . _ . . . _ . . . . _ . . _ _
5.12 References
- 1. Seabrook Station PC2b ab111stic Safety Assessment, Section 9.4, 1 Public Service Ccmpany of New Ha:tpshire and Yankee Atomic Elcetric Company, December 1983,
- 2. Severe Acefdent Risk Assessment Limerick Generatinn Station, Chapter 4, Main Report, Philadelphia Electric Company, Report #4161, April 1983.
- 3. J. A. Lambright, S. P. Nowlen, V. F. Nicolette, and M. P. Bohn, f.lre Risk Scopinn Study' Current Perception of Unaddressed Fire Risk Issues, Sandia National Laboratories, Albuquerque, NM, SAND 88 0177, NUREG/CR-5088, December 1988.
l 4. W. T. Wheelis, Users Guide for a Personal Computer-Based Nuclear Power Plant Fire Data Base, Sandia National Laboratories, Albuquerque, NM, SAND 86 0300, NUREG/CR 4586, August 1986.
- 5. R. L. Iman, S. C. Hora, Modelinn Time to Recovery and Initiatinn Event Freauency for Loss of Of f-Site Power Incidents at Nuclear Power Plants, Sandia National Laboratories, Albuquerque, NM, SAND 87 2428, NUREG/CR 5032, January 1988.
- 6. R. B. Worrell, SETS Re ference Manual . Sandia National Laboratories, Albuquerque, NM, SAND 83 2675, NUREG/CR 4213, May 1985.
- 7. D. W. Stack, A SETS ' User's Manual for Accident Scauence Analysis, Sandia National Laboratories, Albuquerque, NM, SAND 83-2238, NUREC/CR 3547, January 1984.
- 8. N. O. Siu, COMPBRN - A Computer Code f.qr Modelinn Compartment Fires, University of California, UCLA ENG.8257, NUREC/CR 3239, May 1983.
- 9. V. Ho, N. O. Siu, G. Apostolakis, COMPBRN III - A Computer Code for l19.delinn Compartrnent Fires, University of California, UCLA ENG 8524, November 1985.
- 10. C. Ruger, J, L. Boccio, and M. A. Azarm, Evaluation of Current Methodolony Employed in Probabilistic Risk Assessment (PRA) of Fire Events at Nuclear Power Plants, Brookhaven National Laboratory, NUREC/CR 4229, May 1985.
- 11. V. F. Nicolette, S. P. Nowlen, J. A. Lamb right , Observations krncrdre the COMPBRN III Fire Growth Code, Sandia National Laboratories, Albuquoque, NM , _ S AND8 8 21600, Presented at International Topical Meeting, Probability, Reliability, and Safety Assessment, April 1989 in Pittsburgh, PA.
- 12. V. F. Nico1cete, and S. P. Nowlen, A Critical Look at Nuclear Qualified Electrical Cable Insulation Innition and Damage Thresholds, SAND 88-21610 presented at Operability of Nuclear Systems in Normal and Adverse Environments, Lyon, France, September 18-22, 1989.
l 5 33 1
)
l i
- 13. A. D. Swain, !!. E. Guttmann, ligtgtbook of Human Reliability .6naly.!ila with Emphasis on NucLeer Power Plent App _lient ions, Sandia National Laboratories, Albuquerque, NM, SAND 80 0200 Nt' REC /CR 1278, August 1983.
- 14. M. P. Bohn, and J. A. Lambright, Proeninres for the Externni Event fore Damnge Frecuency Annivses for NtfREG 1150, Sandia National Laboratories, Albuquerque, NM, NUREG/CR 4840, Novembet 1990.
- 15. R. L. Iman, M. J. Shortenearrier, A FORTRAN 77 Prorrne and tiser's Guide for the Generation of Latin Hypercube and Rndem $ 3 .mpics for
]Jse with Computer Model.a. Sandia National Laboratories, Albuquerque, NM, SAND 83 2365, NUREG/CR 3024, March 1984.
- 16. R. L. Iman, M, J. Shortencarrier, A User's Guide for the Top Event Matrix Annivsen Code (TEMAC), Sandia National Laboratories.
Albuquerque, NM, SAND 86 0960, NUREG/CR 4598, August 1986.
- 17. J. M, Chavez, An Experimental Investigation of Internally Irnited Fires in Nuclear Power Plant Control Cabinets. Part I - Cabinet Effects Testa, SAND 86 0336, NUREG/CR 4527, Albuquerque: Sandia National Laboratories, April 1987,
- 18. J. M._Chavez, An _ Experirnental Investigation of Internally Tr.nfted
-Fires-in Nuclear Power Plant Cabinets. Part II Roorn E f f e; ty.. l e s t g .
SAND 86 0336, NUREC/CR 4527, Albuquerque: Sandia National Laboratories, October 1988, i f l 5-34
,__._..-_u_.___ . . _ . _ . _ _ _ _ _ _ _ . ___ - __ _ . . . _ _ - . . _ . _ _ _ . _ . . _
L i-I 4 APPE!1 DIX A , 4 Surry Structural floor Response Spectra e i I n I I i i 1 l" l'
- i. _ _ _ . __ _
+
t mN.v.-...--++...-,e..%m-' *.w-,e,--.__ J______' wm-.-__ __,.__www,.,wwww,m,.,e.w.g,w.,,,q.,my+ww.,,+.m.,ym-9 , .
+Ii t .i , i' o !.LL .1 ij , j;rI ! .ijt ii e
rgr - -e e ,. g i - s &E = ^ ow9ouee,
<s.oo t-e
,2 o __ g e t _ n i
p m _ - e s d/ s _ %/ 5t
- 1 t
f o) af t I H ( _ d o e s y t t c ai N
/
- n l n l
a e uu n c u c r e *
- l n e1 g\ r 64 ai c
t ne O F t 38 n I g
- 0
" - 1 p 1 e o n ri t a
- d t t nR l .l l c S
- ef ee ca -
n n
- i e e p e r no
- c f r . . :
s sl i i n - at N
- e e . s. s. e t l c e t a nr g
e e rno o. o. o l c oe Cl
- - t L f f t C N A A 0 e 0
0 0 0
- c. 1 rc C. o. o. oc o4 2 1
0 1 e s 4 2 t A c t 1 c3 ele 73 ae x es _ R n o t p ns ae l R P e rr eu 2 2, wt oc 0 - P u _ 1 - -
- - r
-- : - rt
- as en
= : l I
- cug
~
=_
- N n i
yd 1 - 1, ) rl ri
- 0 ) , z
- 1 I H uu
( H - ( SB
- y y c )
c - n a A : n e ( e u 1
. u :
q
\x s q e A e r
/g g' :
00 F 1 r l a o, F e r u g
- i
- c -
- F i
w t - r
- e -
- V E
1 1 - d 6
- 0 - .
0 0 0 0, ; 1 0 C. 0 0 c. 1 C. o. o. o. , 08 s 4 2 04 2 0 1 e s 4 2 1 11 1 Co ole; X 3Le7o% X k;j! 2 ,:a , * ,
o y" x to N-S r-X to E-w 14.C 14.C se 12.o-- o 12.o.. c 10.o--- h - c to.o-- o
- b. .,.
.a. % e.o-- 3
~
(
% e a-. r g tM -
1
- /s ,'ts. -
= s .o-- /
.e s .
s "..
=
a s.o-- ("\y t
-Q- $
4.o-- P % $s,A(v Q ( - - 4.o.. s_.3~~-_ -- s..________ - P' N z.o-- , e.o-. :
^^'^^^; ^ ^^ "
^ " " " " " " " " : " :""" : " " """; C'
^ ^ " ^^'-too a ^^"z to
=
C o i z - to lo I to 10 g6 10 Faequency Nzi Fcequency 0121 o 4 8 h* X 10 Vertical g 7 8.C _egeno: to free-field Fod. ref. pt. --- - s .0-- t.o.s. e1 47*4* _ _ __ ___ . c t.o.s. el 95*6* . . .. L
= 4.0-- Notes:
0a All spectre calculated at 5% demping
% s 2.&-
~
Acceleration in units of ft/s/s
" " ^ ^^^
": """ : - "^^""":
.C 5 so' so 8
so" l s5 ree - y ma, [ Internal Figure A.l(b) Surry Nuclear Power Plant Reactor Containment Building Instructure Response Acceleration Range 1
J i- ., 5 l 2r .4 x to 0 N-3 y x'10 E-w 30.C t-
-4 f'
30.C 25.0-- P i 25.0--- o o e C C 3 eo.o-- 3 20.o-- s,,
- n
. - e U e t " t- e 15.0-- .
- D
.e 15.0-- e e
4 10.O-- . 10.0-- , 3 5 .0-- .---.--a 5.o-- b-- ? s 9
" " " """" ^ ^ " ^ "
" " " " ^ ^ - h
.C
" " """ " .O 2 m 2 -1 0 1 10
-1 0 10 1
10 to 10 to 10 10 Frequentg Mz1 Frequency 0411 o M a i U o x to Vertical ? s4.c m Legend: a 12.0-- free-f1 eld fnd. ref. pt. _ _ _ _ . _ _ _ C 10.0-- 2 t.o.s. el -3*6' _ _ . _ _ _ _ _ _ _ _ _ . o . t .o.s. e l 18*4' . . ._ a s . 0-- t NO ces: s o
- 5. O--
a At t spectra ca tet;steo at 5% damoing 4 . 0-- Acceteralkon in units of f E /s /s
- 2. 0--
.0- " " " " """".
0 10 10 ' .10 10 Frequency-StII Figure A.2(a) Surry Nuclear Power Plant Reactor Contairwent Internal Building Instructure Response Acceleration Range 2
Lil .v I[t ! ' jm 2yr u ".oo ev l U"'.o**" .u
. to .W oMeos.5m t
2 0 g . =.. 1 _ n
- _ i
= -
_ o m
- _ _. a s
= _ . d/
_ . . s
%/
5t f e 1 t
\ oi __ ar z __
t i o D o -
- ,+ -
f c n es t t ai l n l e u uu a c c n
\ e ' ' l n r r 4 6 ai e2 S -00 F .*
- c t ne -
1 t 7 5 n I g p4 9 ao n Ak, J - a g .t t a i c t t a nR S - ef ee ca er d n t e r_ c . . : s o e sl eo i i n N - e , . s. s. e e at g c t a e edn o. o. r o t c t c nr oe - 1 0 L g f t t N A A Cl 0 C. 0 o. 0 0 o. c. 1 e o0 5 0 5 0 5 rc oc
~
t 3 2 2 1 1 - tA X C3 .L=*'- oY c ae - es R n o t p ns ae l R P e rr
._ 2 2 0
eu 0 + wt 1 - 1 oc
=- -
P u rt r
==. _
as en l I
=..
- cug Nn i
w 1
'0
)
1 ol r yd rl
- 1 z - t ri
- H - 4
( uu ( -
- SB
- f -
y c c ) n - n
. ~ e e b
( u u q q 2 _. \ e e N r r A
/ 00 F
- 00 F ;
- e W4 1
1 l r a - u g' I t c - i i W : t F r e _ E V - 1 - _ - - - - 1
- 0 - _ - - - - 0 0 0 0 , 0 0 0 0 1 0 C.
0 0 o. 0 c. 1 C. , o. 00 5 0 5 0 5 o4 t 1 2 1 g e s 4 2 2 . 1 1 1 3 2 X C 3 s . (=** $ E x e3o=L.& eau <
>sc a '
1' 1 : l1 ,' a e I , 2: ,a ; :; ;4' <r *, t
j i 4 . ] RSP6V V3.09 3"cbl . pit 10:01'21 97-90 86 I e t i N l' .
- r a, e I I ]
! l . I i .C- ,
j
- I I
le P 1 l 4 . I-i I cN l [ t e '
" es llb l l D u ,
fk j'
, I l b i 3 2% lIi em o
4 1 r(i[/ : E o v M a a
) o g a
/%E'
- l w c
/
N q .
- f. , ,
fU
- e w
't
\s D L L.
.T g
E W [E M M E O I = @O C' b pc0ge y ao L -
<> e.e e mm a u c M . W 6 W W W 4 u et
..
- W v L CN
. i e m L . . ,, og -
C C y I g s e. M. M y e* y O 2 ' e u -u
, W ve L e o. o. o -o 5,O
' ; ; ,o J
- w u u 7 4 < cw e c w < r. A. rv. . o. - t.h, ,
o 4
- uu x untnessteDay cv a<
v nv 0 m nC u 8. ' ) ee it C M N 4 N o , %- u8v [i l E"
!I L &
L wu n m-oC
. . . mn.
l a v ts
- - 2 C
) E%
S%
. ! pE
( : d- E wO
-l>.
vv
.- > .. v3 sc
-) s> ..
u c . u 3
/. t- - t.
a 5 q r\ m
% E E
- .o
- L : .O. L 4 e ! ,
8' L-
~
x : - : a ** l' L. v 54.
-tu -
I:
- a. o. 'E o o. 6 6 6 6 6 o. S w -- < r. A. ' s.
_x uctnessteeny x votnesstency (= l A-5 l-
- . - - .. .- - . . . - - - . . , - . ~ . . . . - . . . - , - - - , , . ..
i .i i 4 2 2 g" X 10 E -.W E to N -S 4 4 e
-e w
4 o 4
}- .}- o c C 3 3 w a ==
2 E. . e-- .
. . e-- .~
O i 7u N w e o \ - u n u/ f. / ,sf , % A a o , 4 y
- ' # ^9
.1--
.1-- j r [g s ,
,ft j \ s- ( ~~--% _---- _ _ .. ;
N ~_u
- -_ _ _ ___ _ _ __ ,e 1
.C
.C m
i l 16 10 10 10 16 to 10 10 . > Frequency (Hz) Frequency DIzl
- ' o 4 O' M
' 0 j x to Vertical g' 30.C Legend: h j 25.0-- free-Fleid i fno. rer. pt. _._ _ _. _ _ _
, c j
3> 20.0-- t.o.s. et'47*4' _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
= t.o.s. el 95*S' . . ._
L
= 15.0--
! O No tes- !! u 4
% 10.0-- Att spectra calculated at 5% demping Acceleration in units of ft/S/S
- 5. 0-- _ _ .
1
.c " """:" .. "
16 10 10 to . Frequency 181 2 ) Figure A.3(b) Surry Nuclear Power Plant Reactor Containment Internal Building Instructure Response Acceleration Range 3 4 _e- _ _ _ _ . _ . _ _ _ _ _ _ _ _ . _ _ _ . _ . _ _ . .
j/' ![ll 2 g r ho ~.= E ~=*",* *M ms )m
- c , em7e 2
0 1 __ g __ n i p _ m _ _ a d
- _ . % s
_ ._ 5/ 1 __ _ t / s o) __ at t z . f H (
- d y
e n t i c n a e l s
" u ut q ci e *
- l n r O6 a u
".001 F . *
- c t 3 7 n p1 2 ao o ri g
" l .l l t t n S er ee c a i
- i e e r dI o r c . . : p e l n - s i e e s . , s. s.
sl ug N " e e g B n 1 e reOn o. o. t o l c l c yR a
- o L f r t t N A A r O C. 0 0 0 0 0 0 C.
t an 04 2 0 8 6 4 2 i o 1 1 1 1 l i i t x c*.a= Leo _e,o4 xa ur A e l t e nc ac l A P e rs en 2 0 2 wo op
" . 0 1 1 P s e
rR a ee l r cu ut N c u
- yr 1 I rt 0 ) o) rs 1
z t z un
\ " (
H ( SI y . y ) c c a
" n .
n ( e e 4
" w u o .
q A e e r r e
,0 F 00 F r 0 1
r l . 1 u
" a . g i
" C .
F W " 1 t
- ^
r e E V
- - - - - 1 1
- - - - - - -0 - - - 6 O 0 0 0 0 0 0 1 O 0 0 g 1 C. O. C. o. _
o4 t1 2 1 0 1 8 6 4 2 08 1 s 4 2 x C3a=Oee* e u< X c3 ae Le ' Y
>f "
rlllll ll!l .Il i{ l ;lll1! lI llf!lll,lllf'1ilIlll 1l '
. - - - _ - - ~ . . . -- - . .
r i a f i y
- x 10 18.0 E-W 'x'to ' N-S E f 18.C [4
= 16.0--- - 16.0-- # t <
- f. 14.0--
',4
)
e 3, 12.0-- -1 4' e 14.0-- 3 12.0-- e [t f' o' I
- n. . t E 10.0--
. 8.0-- -
g T 1 710.0-- g p E ir i 4 3 e 6.0-- k ~._ g
- 8. 0--
s Y . _ __ r [ k _m - 6.0-- N t
)
^4 . 0-- 4 . 0--
, ~ ' 2. 0-- M [ 2.0-- m m -li
.0
^
;'". ^
.0 " - - -
a l i 16 10 i to to tE to 10' 10
- .CreQuerncp M11 Frequenc; fHr1
}
}' ' > o l
' 0 4
x to Vertical 7 ro 8.0 l Legeno: i g ; free-fteld
- c s.0-- fna. ref. pt. _ _ _ _ _ _ _
t j 3 g t.o.s. el 45*10' _ _ r , % 4 t o
- t.o.s. el 66*0' . ._
4.0-- No tes: u % - u-; [ % AII spectra calculated at 5% damping ! ! 2 . 0-- Acceleration units in ft/s/s
,o . .
3 I 16 10 ' 10 10 k v reouency fHz) 1 2
- Figure A.4(b) Surry Nuclear Power Plant Auxiliary Building i Instructure Response Acceleration Range I t-
- t. ,
i: 2 _
!i.i* ,tt s t k!f , ! [;.t .i(ytt* hIk I t t tt:irt[ttlI p *t(!.l,L[ f iI ['i> - Devr o. <Y ,o ~ ""** E '
o* wM o* oM.oi=e 2 .
. 0 g
" 1
._ n
- _ . i o _
v a d
._ _. . % s
. _ 5/
s 1 t / ot at t M i __ o f
- cn
- s c i c a
. - n t s e ut m ci e ' ' l n r 06 au
- c0 F .*
- c
" 4 t 3 7 7
- O 1 2 a0
- g r1
- g .l t t t g S ef ee ca n
- g e ca i o r_ r . . : oe d2 n s. s. s st l N
e c e e i e ug g ed t c r n o. o. t e o t c B n a t L g f t t N A A yR
- - - - o r 0 0 0 0 0 t O C. 0. - an i o o0 S 0- 5 0 5 t 3 E 2 1 1 li i t
. x c2 u e L * .,
- o E xa ur A c l ~
t - _ ne at l A P
.e rs 2 2 en 0
1 4 4 o t wo op
- + P s e
- _ rR
- - a ee l r cu ut Nc u
1 1 yr o1 0 i rt t 1 1 z rs . - 4 0 i l f un _
- SI y f
- c. )
- x -
a _ e ( w e we 5 _ r r A 0 0o F e
- o t
_c 4 t r
'- l a
u 3 g
- c -
i W - i - F t
- r
- e E
V 1 . - 1
- ~o
- - 0 0 3 0 0 0 1 0 c. t 0 O 0 0 0 c. &. C.
C. 0. . oe 2 0 s s a 2 00 5 0 5 0 5 13 2 e 1 1 t t 1 1 X e ,, e o = ., . u E x c* =t* aa<
>ee L .
li ll
, n s s E".* <@ O ,. Ua o: - .o2 t m onE e*
- 0 __ g u1 __
n i p __ m a _ _ d
~
. % s
_ _ 5/ t / s
% - o) __ at t z __
T x 3 4 0 d en f y t i c a p1 n l s e ut u q ' ci e 0' l n r 1 C a u 00 F -.-
- c 1 t 56 n
' P4 6 ao
- d ri g
" i t l t t S " ef ee c a n
- t e e r i d2 d f r . . : p e l n - s sl N
- e g
e - s. s. e e i e ug 1 e ed r n o. o. t o l c l c B n a
- 6 L f f t t N A A yR r
"00c. 0 s
o. 0 o. s 0 0 0 s C. 1 an i o 1 3 e 2 i 1 li c i t
* .L3 3< xa ur A e l
t e nc ac l A P e rs en o . C wo op t 1
- m. .
P s e rR g_
. a ee
. . l r cu b ut Nc yr u
\ .
I 0 l 1 H f z 0 1 zi i p rt rs un SI y y Q . c n c n
)
b Y, e e (
. u u 5
/, c .
q e e A e r
,00 r F e 1 .
. 0 r
. 1 l .
u
. a .
g
. c i W -
i t F r . e E . V .
- - - - - . 6 - - - - - 6 8
- c. o. o. o. o. 0 g1
, C.
0 0 0 0 0 , 1 00 s 0 '5 0 s 04 2 0 8 6 , 2 3 3 e 2 1 1 1 1 1 1 x- c3 .L. . < , X c3
- o<
y.o ,II1llllilll!l'\ ,(;,!llli [lll f lil lit \(llIIf ll\
a i i a ! i i 2 m 2 X to N-S w x 10 E-W c i
.' 4 r- i
- r
" I u~ r i
o
.3- o c .> ,
c ' ' O O w r 4 _. - a
, u
- e. e U" L -
. L. .s- ) ~
. .s- ). *
~
. O r
. u - !
u o " u - - --
. t-- .1-- * ,
I ut '
.C ^
.0
- " " " " " - " " " " " ^ ^ " " " ^
u 0 2 "
-t O 1 2 10
-t to 10 t
to to 10 to to i f Frequency (Hzl Frequency IHz) ; f o .
~ eu !
x to 0 9,,,t ic a l E i 30.c . Legend: g 4 ' ! 25.0-- fr ee-f te td ! fnd. r'ef. pt. _ _ _ _ _ _ . _ _ _ c j 3 20.0-- t.o.s. e1 13*0' - - - - - - _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ r u l = t.o.s. el 27*6' . . ._ i <_
- 15.o--
. No tes:
U :
< 10. 0-- ( _ All spectra calculated at 5% camping Accelerat ton units in fc/s/s l
- 5. 0-- [
b
.c " ""
-1 0 1 2 f to 10 to to .
reeaueicy - fuz) i i Figure A.6(n) Surry Nuclear Power Plant Auxiliary Building i Instructure Response Acceleration Rany,e 3 d .- _ ___ - _ _ . _ . . _ _ _ -_.4
2 27 a X 10 N-S e x to E-W r U 4 i
*1 Y ,
c i
.}- )E o ,
q c
.1-- % c g ,
5
- a. gg a. ;;
L 'J'g L e J *'
. 2--
e .2-- l . 5 y Y% .
.- 5 y h i .1--- .1-- _
O 9 un (
!v
.C
.C - - -----: - - - - - - - - --
te i 0 2 m
-1 0 1 2 -t 10 t
10 10 to 10 10 to to , y reequency !%29 Frecuenct (%ZI e o [
~ ru ;
x 10 Vertical b j 30.C
- e 1 egeno:
- r 25.0-- Free-f te td [
fnd. r'er. pt. _ - - _ _ _ _ _ _ 9 20.0-- t.o.s. el 45*10- _ _ _. _ ___ _ _____ _. __ {
% t.0.s. e1 66*0' . . .. i 1 L .
= 15. 0-- i O Noces: i h 10.0-- All spectra calculated et 5% damping Acceleration units in ft/s/s
- 5. 0-- !
l
.C - --:" - - - - - - .
-1 0 1 2 .
10 10 10 10 [ Frequency (Hz)- l l i
- 7) ' Figure A.6(b) Surry Nuclear Power Plant Auxiliary Building ,
j
- Instructure Response Acceleration Range 3 l a
~ g .,
-______m. . . _ _ _ _
i a. R$ PLT v3.00 istl. pit 11:18: 22 OPatt-60 1 0 t e I 1 I N I i o
- U l ) r
. I
. I i -
i 6 0
,t .' g i C-e i
I O Y y e f, , ( t ' ON l t e MN < / .
"O T 9 .W
]: e>
." E
~l l O b
WC
.> > a w
'! e
)>
- M D&
a u - e . .* s a c L 60 .o m e D y
, 0* 6 .
0 W@NO
; C o . El N 4 e o -
M-
. t **
-> .e .m -m a ,a V N W *- W W W O e e
=>
*D
. *' W
- L * * * ..
D OW L. u be r i M M M 0-7
- g g . M. M. M. g "g . g
& O D 0 0 0 u ** v C 0 L C *
- O ~ u 85 w
4 . .
';'a a s u a a a' 2 < < "
o o. . 6. - 6 6 o. 6 6 o. - C C o*
-- N. e. c 4 g N -'ok w U w DDi%IJD)SDDY p g CW C 0 V-4 U U C U tt <
N Ch. @ 0 Wm ' tu N 4 C
$ l l l$
. nm ff 0
i ' W gr. I a 3
.' g P 0 W e.4 w U U
' ' 3W
- 2. U l
o == . O sE MM
'"*d
= *d WM f l' h.
- h YC Mw a
h. C .
)c g
* . b w - $ ~
< a . o .
b t
%. ..*. w . .co i .
.. - m s-m
'- -2 : 3 : [ a I L . O
- i. _
Li.,i *
'A .
I i n~ i~ i i
- s a , M 1 o o. 6. - 6 6 6 6 6 o. 'S o o. - 6 6 6 o, 'E !
l 09 N O E @ T N OC @ 9 N e ee e t x boitesotonov x -Uenesotoony l l I I A-13 l l h t L ,r .y -4. . , e .,ny,..,_J.,,,..,,,,,,,-,,,,,-__.~r..
, o.,.-J...m-.. .-- _ _. ..-._,. ,.._ - - ._-..- - -~ m ,. . - . - - - -
l l l l li' l ,1 !! iIili _. 2iner e_ y . E eemb ~ ;[e .ts o 7 e.ee 2 _ 31 o __ g
=,,
^
t _ n
^
^
i p
^
__ m a _ _ c
^
% 5
(\ _ 5/ . t __ t / 5
^
^
0 I _ at m
^
^
1 l l _ ._ f
%. \'
l . ^ I s \ d m_
^
y en t i c a
- n e l s
^ u ut 0 ci
/ e ' ' l n Oo F r 36 a u
. *
- c n
t t 8 7 e d p57 a o r _ " ri u , S l .t l t t t c ef ee c a
- " : t e e r u r ,
d f- r . .
- p e n s t 4* sl e . s. s.
e e e S g
- - - - - - - 1 e edn o. o.
r t o l c l c m o' 0 0 0 0 0
- - - 0 L F f t t N A A o C. &. 0 0 1 R
C. 1 , 06 4 2 0 6 6 4 2 1 1 1 1 1 l m 4 eoO.t*,e h oe rg t n na _ oR C . t c nc l aA P s e 2 rs o 2 en t . 0 1 wo op Ps e rR a _
. ee l r cu ut Nc
- _. - - t 1 yr u
' ol .
t t . 0 l rt s g' . D 4 1 e H rs un
.. 1 "
. (
( . SI c . y 1, - n c ) t t e . n b e
- b. \
" u q .
q u ( 7 1, e r e 00 F A f "
" 1 0 . F 0
r e - l .
" 1 r
" a u.
W " c . g.
" i i -
t . F r . E e V 2 4 3 2 1
- - o 1
t O c. 0 0
/ 1 0
C. 0 g 1 0 '. 1 3s s 4' 2 X eoO ,te e u X e .%te;
>a '*
11I: ;iU ' , j!.. j, 1j;li,1l': ,s1!ii!L. 4 4* , $;
- i. - I i
1 1 a 4 t l I a 0 m i tn 0 x 10 N-S 1, - x to E .W 30.0 ? _a i 30.C i < l i 25.0-- 25.0-- ." l o > o > r s e c '
} 3 20.0-
# w
. f' 3,2020--
, ..- ., e
- . ** f
- L -
1 i L'
- 15. 0-- U i' e 15. 0-- .% - -
3
.- o
" t u ..
j Y 10.0-- E 10.0-- h _. _ _ _ __ _
- 5. 0-- [
- 5. 0--
w ! i o. . l, 4- m,
.C ,
.0 " " " " " " . -
0 0 1 2 -1 0 1 2 i -t 10 10 10 10 10 10 to - ,- 10 L Freo,entf f%zi Fvecuency (Hz) [
> o )
4 t 92
~ e 1
- ' w 0 ea t
X 10 VeetlCal 7 14.C cqeno: e : e L t 12.0-- s free-r le td [. s rnd. re ' . p t. _ _ _ _ _ _ _ _ . _ _ _ _ _ i
, t c 10. 0--
3 j 2 ; t.o.s. el 9*6' _ _ _ _ _ _ . . _ _ _ _ _ _ _ . . l. i j . e . 0-- t t.o.s. el 27*0' .. . L i t.o.s. el 45*3' l i j = s . 0-- u - u N0tes: n ! , n . 0-~ All spectra esiculeted at 5% camping l
- 2. 0-- Accelerat1on un1ts in ft/s/s !
l L
. . ... . j
,e . . . ,. j
-1 0 1 2 :
to 10 10 10 r r ! Frequentf f4z1 - i Figure A.8(a) Surry Nuclear Power Plant Control Room Structure Instructure Responses Acceleration Range 2 l I ! i ;
n x so g ,, E-W X to O N-S y
- 30.C r-
.- -4 25.0--
y 3- 4 .o 4-- _g\ '
- c I \ c
!. u 3 1 3 20.0-- j e 3-- ' y
\t f= 15.0--
v h. s k
- '\ s ~
= "
. 2--
n u %\ u 3 3 10.0-- N' - j 'g
.1-- %---- 5. 0-- N m
^ O t+ : : " ^: 00+++ .C
^^:::^0 " : :^^ . ^
..-^++4 w m
.C- - : " :^::":
-t O 1 2
-1 O 1 2 to to 10 10 3- to to 10 10 s
frequency (Hz) frequency (Hz) o
- N e
m t X 10 Vef't ica 1 7 14.C Legeno: e i
- i 12.0-- rg.ee-f ie 10 i
fno. rer. ot. _ _ _ _ _ _ _ i c 10.0-- i 3 t.o.s. el 58'3' _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _
= e . 0--
t.o.s. el 77*6* . . ._ t
.e_ .
- e. s . o-- Notes: !
u m__- All soectr a calculated at 5% damping E a . 0-- Accelecation untcs to ft/s/s
- 2. 0--
.0
- . ":. - ---^^^: ^ ; : ^ : : :*-
0 10 10 to to 1 Frequency fitzt , I Figure A.8(b) Surry Nuclear Power Plant Control Room Structure l Instructure Responses Acc. Range 2 : 1 i m_ _x _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ .
x 10 2 g,g g x to ' -E-W
. ?e w
O o
- , .>- /g' ,
3--- 3 w 3 a.
'q .
o L-L
- . e-- ,-- 2-- . ;, .
- 3. . .
u " o y
. s-- .- -
ns
.C es
.C 2 w
-t O t 2 to
-t 10 O
t0 t 10 10 10 10 to > F'equency fHzl Frecuency (Hzl b Y " 0 es X 10 Vert 1CaI so.c 7 Legeno: e m 25.0-- 'r*ee-f te id 9nd. c*. pc'. _ _ . - . 3 20.0--- t.o.s. e1 9*6' _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . a a- c t.o.s. el 27*0' . . .. L 3 15 O'-- t.o.s. et 45*3' _ . . _ _. . . ._.
. e U
s Notes; y ,o,S w AtI spectra calculated at 5% damptng 5 . o-- Acceteration units to ft/s's i l l
.C -
* * +++: "
"^*+! "
"""4+++ -
0 16 I 10 10 ' to recouency fuzi Figure A.9(a) Surry Nuclear Power Plant Control Room Structure Instructure Responses Acceleration Range.3 m..
i D t,rv E 10 N -S e a K to E-W 4-w
/ o U
(
.T- y<f ~
o "
. fr-o r :
c , il \'%N v e' t s-\ *.&- K\
- 4-- ) .
bb <. L-
#\t g A D N
,(- - .1-- ..
n
,7. q-_______________
_ ?
,e .
10
. . . . . . . . . . . . , to
. .. . 16 to
,g .
to 10 10 reequency Nzt o 16 rr-equency Nrl m. N 7 e O m 0 Vertital m x to t_cgend: 30.0-f ce-rtotd __ 25.o-- fnd. ref. pt. t.o.s. et 58*3' ._ I t.o.s. el 77*6' 3 20.0-- Notes; L N
- 15 . 0--
= All spectre calculated at 5% comping 10.0,- .
*- Acceleration units in ft/s/s 5 . 0--
,g_ 10 0 to 16' 10 Frecuency (S4r)
Figure A.9(b) Surry Nuclear Power Plant Control I:oom Streveture Instructura Responses Acc. R arer,e 3
.- Nl
RSDLT V3.00 togi. pit id: 52: 48 02-21-80 j l i N
- o a
,31; r
I .- {f h 9 0 l i o% i O l . Y% n
, D 'A g
i w (
.ro
/ : ,
u e
- e. c.
-r e'
/
[3 ) D a u e a .* . - C= L m
, ,0 u . .O 10 .th p 8 u CD k N C I O ** N *t e o a L *
= .Ca , .* - ** u u M = W s @ W W u9 h.
..
- U U L w D 6 L * * .. D b =0 m C I @@@ @ @ ad w O p .a g W W e . .
+
& W D C 0 0 u aa u D t.D 8C c W L + O ** u n 7 a s .C .u u' u' 2 .s <1 E"
i i
- U. b. 6 6 * *. 6
- EE
* *
- N 23 $ 2
, cetteselency L'O 4 e WW ns W&
4 u v cv
~
e< cm G N N % e iS gC o
~
- c. L m
)
f 4 uh n l , l 4 4 L
-u uD
,o oa w w I- c. - .
o-
~ suas.
~y .
e u o oc sv
- M" J .
W W s [t f i [o o
~
3 o
- v v <
- I
- o I , .g o
e : te M E 2 ss i L u \ tu
* , , , e i i i i i i 6 6 6 6 6 o. E o o. 6 6 6 o 'S o o. O E 9 o.
N OC @ at N O4 N - m = x vottesolenny x uottesote3DY A-19 l l _ .
. _ _ _ _ _ m _ _ . _ _ .
_ 1 . 4 4 I 'i
' n o e
< to 9 -S ?
E o E-W -e X 10 30.C i .i 30.0 - < t P I 25.0-- I 25.0-- s 8 [ C' I L% ! C A
~ 3 20.o - .
E I 3 20.o-- .
$ \ *
' g. ~ N s b
' h ,.
- r. Ig -l ' 3 15.0-- :
'
- 15.0--
0 \ : k\ c 1 4 Y 10.0-- k(6- ' Y 10.O-- 5.0-- W_Z : __--_e__
- en o
j r 5.0-- M I . . . . . . . . . . . ...... . .... , .;
^ *
. . . . . . . . . . .. , ,, 2
! . . . . ..... O 1 t
,e 2 t 10 10 to
' -t O t to 10 10 to j to fYeque7cir Nzl [ Frequency mil o y fu i Es .I ' N X 10 Vertical 7 ! e
- s4.c t_egend: tz: -
free-f1 eld . 1 12.F l fnd. ref. pt. - _ _ _ _ _
! I i c to.o-- t.o.s. el 19*0* ... -
r i 3 a t.o.s. el 27*5* . . ._ l , I
- 8.0-- _ _ _ -__ ___
, t.o.s. el 42*5* _ l _ _ - o s.o.. Notes: ,i i # All spectra calculated at 5% demning )
- 4. o-- t
' Acceleration tre units of ft/s/s r l
2.0-- l i " " " " " " "
.C f
0 to ttr f 16' 10 F*~ecuenc ; M11 , e t 1
' Figure A.11 Surry Nuclear Power Plant Safeguards ButIding !
Instiveture Responses Acceleration Range 2 t I }- i i i
I h I X 10 N-S h X 10 E-W 4- l 5 4 <
'f o
rQ - o 3-- C .T %
} na l
*3 s o s *
-e l c
n m N
!s
\
U e D "* ,, I C . g b e . 2-- e . 2-- '. T 0
~.
( .. . . _ o Q. ~ w __c _ =- y r. I-~
.?
.1--
.4
*16 .0 .
I 10 to 0- . 16 10 16 I 10 to to Frequency (Hz) ec.,e,c f f-z) o eu
> J l
to 0 e X 10 Ver* t ica 1 g 30.C Legeno: free-flelo 25.0-- fnc. ref. Ot. _ _ _ C t.o.s. el 19'O' _ _ _ _ _ _ _ _ _ _ _ _._ 3 20.0-- t .o .s. e l 27 *5' - . .-
% t.o.3. el 42*5' __ __ ___
E 15.0-- e Nott 0 l l # 10.0-- %11 socctra calculatec et 5% damoing l scceteratto7 in units of f t /s /s 5 . 0--
.o s. ..
t I to to 16 10 c reauency (Hzl Figure A.12 Surry Nuclear Power Plant Safeguards Building Instructure Responses Acceleration Range 3 i
O x to # ' N -E' E x to E-W ? 14.C 14.C , d l + 12.0-- 12.0-- 1 N 'o ' C 10.0-- b c 10.0-- ft -' O '
'i 3 3 1 ..
n i I a 3 -
.g.' a .w -
e 8 . O-- Y g
- 8 . 0--
C F L
% 1 g U !,
i
* . 6 . 0-- \\
- 6 . 0-- gg %
m u u 9 " ~ " " " ' " ' ~ ^ 4 . 0-- 4 . 0-- E 2 . 0-- 2 .0-- un E
.c *
.o * - ' ' ' - - - ---
1 10 10 10 15 10 10 10 i Frequency fitzt Fr'equency 34I1 b 8
" 4, x to Vertical ;
e.c Legen0: m m free-field 6.0--- fnd. ref. pt. _ _ _ . _ _ . _ _ _ c t.o.s. el 26*6* ___________._________
= t.o.s. el S2*O* . . ._
L e 4 . 0-- e tJo tes: u All SDectra calculated at 5% camping , 2.&- Acceleration in units of ft/s/s ,
.o ;w. -
16 to 10 - 10 Frequency fitti Figure A.13 Surry Nuclear Power Plant. Containment Spray Building. Instructure Responses Acceleration Range 1.
-. - - ~
- ~< .. ._.
D o to O ( 10 *J -S X 10 E - vf 30.C ? w 30.C es .o-- o I es .o-- o
~
C b C 3 20.O-- b hN 3 20.0-- fg- , ru
. N's L
M = L = 15. 0-- {hs b e,15. F - {I . o (1 s u \%
# to.o- -
4k_m.=:..-=-_.
# to.o-- Q% - -___]. -
- 5. 0--
9 5 . o-- - 9
- ^
^ :+;;
" ^^
. rt+++ -
" . ;+ .0
.C 10
.0 I 10 16' 10 10
> 16 10 10 1 e Frequency (Hz) to Feequency (Hz) o
" eu e
ru O X 10 Vertical 7 a 14.C Legend: m 12.0-- f ree-f te td f nd. re f . p t. _ _ ___ __ . c ,g ,, t.o.s. e1 26*G' _ _ _ _ - _ _ _ - . 3 . , u t.o.s. c1 52*O' .. ! = e . 0-- t L e::::: , Notes:
- 6. 0-.
At t s:)ectra calculateo at 5% damping 4.04 Accc ierat to, tn un t ts of f t 's /s
- 2. 0--
0
. - :+
I ..0 10 I to 16 10 crequency (Hz) t Figure A.14 Surry Nuclear Power Plant Contal.nment Spray Building Instructure Responses Acceleration Range 2
)IIf Il)I
,* ro t. omaN 'g "Er_a <W.oo wE "o - m 2
0 _ _. g _ ^ 1 _ ,
^
_ _ k o m __ a s _ c/ _ _ s _ %/ 5 : 1 0 l _. _. 1 z h
. a 'o
( o g i s N -
^
y e s c t
\\
sY \ c n ah l u , q
- e. c .
\hf
-00 o
e r F .-
- 6 '0 l
at c, o g n t 6 2 i 1 D2 5 aot d d - l l .l l c c i ef e e ca u S -
; i e c - B
- d f r . . : oe y n -
- s. s.
5 st a3 N e e . 'e e r t c pe g e edn o. r o. t o t c S g n
- L f f t t N A A
- ~
- - t_o t a 2 4 } 2 1
. C.
t nR e 0 mn 1 no
- 3% SE ii at t a nr oe Cl e
t c nc aA l , P s 2 0 e
. 0 '
1 rs 1 - ' en
-w-wo op P s e
rR _ - a _ ee l r cu g ut Nc
\%_' 0 1
^
o) yr u
^
t 1 :
\- .
z, ' 4 [ ( rt
" rs y y un s \' ' c c SI i
n n s\ i\ r'( Y . e. q e u g 5 1 f e e
$' ,o0 -
e r r A
. 0 1 e 1
l " r a u g c i i W t F
- r e
E V 30 - - - - - 6
# .. y 2 g g 1 0 C.
0 0 0 0 0 C. 1 o 00 5 0 5 0 5 s 1 3 2 2 1 1 x oU. = e I X c3 a*c.= euo<
>eNo lllllll!il l,lI .l? \ l. ,i flIi'lllil!I! ll,lf;i,!lII'l>I,ijl1l-
1 \lI) l'
<i sk o y ? m, o" om n Dgrw
' *o* O i F
- o t
_ _. g n
... i p
m s a/
.. _ d s
. /
._ % t 5f
\ \
( ; 0
+
1
)
H ( z t f a o o s et
% - / t i c . a n n t l u e
a u p u q 0 c n e l i e r f - a r
. 0 F e c s u 1 r7 n s 2 a o l o
0 n ri 1 ol t t c S eie c a n
- 1 t e r E
- a . p e d f "
r n s e- d n s. sl o N ' e e e t g e u 0 t l c a e_ f f ot r l c 0 r1 6 t N a e e 0 0 0 0 0 0 0 0 O 1 ne e g o4 2 0 8 6 4 2 Gn t t 1 1 a yR
< r3 >
- b oS< c nn eo gi rt ea mr E e l
t e nc ac o o l A t P
- t e
- - rs en wo
- - op Ps e
- - rR a
ee
~ \
s l r cu
\ 0 1
)
z o) z t ut Nc
\ "
H ( H f yr u y " y y rt rs c c
, n n un e e SI n u c u c
a e e 6 r 1
. 10 8
- ot F" .
l
+
A a
- e c r i u W t g i
r F
- e V
E
- - - I i
- - - - - - 6 - - - g 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 s c.
o4 2 0 8 6 4 2 08 6 a 2 t 1 1 1 1 O x c3 aeI e3< 4 3amI " o* y [ n s. 1 l1lllll 1Il!
t 4 2 0 y, x to 0 E'-W X to N-S 30.C r-30.C -a 25.D-- ." -6 25.0-- o C Y(. C @( e l 3 20.0-- 1 3 20.0-- o a *
.a m
- n L \('- L
- 15.0-- E
*, 15.0--
. \.' . \ -
u \ g---- U
< 10.0--
\g
< 10. 0-- ___
- 5. 0-- R 5 . 0-- a.
.C
- <- - -^ c)
.C .
0 2 *'
..0 1 2 10
-1 +
to to t LO 10 - 10 10 to
<-equency (Hz) Frentsoncy (Mz) o e n ;
w 0 c') -
- X ' to y,,, t t c a l i
14.C Legend: m ..t m 12.&- free-field .. foundation ref. Ot. j .; e 10.0-- t.o.s. el 27 - 6 j 3 . o e s . 0-- g (A - Notes: all soectr a calculated at 5% camoing
*, 64- -- -
u accelerations in units of fc/s/s 4 . 0-- 2.0--
.C , . . .
16 to 10 to
'Feequency (Hz) L I
Figure A.17 Surry Nuclear Power Plant Emerr,ency Generator Enclosure Instructure Response Acceleration Range 2 i A n c y -- - -- -,, ,-
+
l l l i. l 2 . g" X 10 2' E '- W t to N - S 4 tj
.A es
. 3--
*g y .}- -
' g
\ 5 1 \ 3 5 1 h .2-- {l h .2-- { "'
i u [ u [ i 4
/ ~_-- J ___ ,
. 1-- f
.1--
7 _
- / /
/ Z >
/ - * ' ' '" -
N .. su 3
.c -
- ' :- .o -
I 80 10 16 10 to 10 16 10 Frequency fHz) Fecquency fHz) > o' I -J to *
" 0 -
o i j X 10 Vertical 20.C T
- P Legend: e
~
p rl free-field ' .founo m on ref. pt. _ _ _ _ _ _ _ _ _ t 15.o-- \A . 3 k t.o.s. e1 27 - 6 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
- .)
h go,o__ [ \ No tes:
- N N -- a t t spectra calculated at 5% damping 3 '\
# ! .I~ acce'terattons in units of f t /s /s S . o-- .
,g . . ,, ..
16 to 10 ' to Frequenct fitz) Figure A.18 Surry ' Nuclear Power Plant Ernergency Generator Enclosure Instructure Response Acceleration Range 3
.- - , t.-,-p. + . . . - a. v - c=~... r.wm -
- p 0 * *O N-S Y.
r x 10 E-W 14.0 _4 14.C < 12.&- F
- 12. 0.- - o o
to.&- n c 10.0-- o. 3 s O
& X e.0--
\ S
.-{ "" *~ / E r/ N -
Ns u
. s o.._ %
e 6 . 0-- N. u u v
- g. 4 o %
d 0-- \ --_ - _ _ , 4 . 0-- \_____,_, _
- 2. 0-- y 2.0 .
r te C 10 10 0 I 16 to
~ gi 10 10 to Frequency DizI
> FaeQuenc( (HZ) o M N2 E)
- O m X 10 Ver't 1ca 1 e.c Legend: h f ree-f te ld foundatLon ref. pc. - - - - - - - .
- 6. 0-- _________________
t t.o.s. el 27 - 6 3, 4 T S No tes: g ,, g 4
\-
all spectra calculated at 5% demping 3 \ - - --- _ accelerations in units of ft/s/s
- 2. 0--
.c "-- ".
I 10 10 10 16 Frequencr tHz) Figure A.19 Surry Nuclear Power Plant Intake Structure Instructure Response Acceleration Range 1
~,e-I 4
2 ~ 0 4 - x 10.0 'E-W X 10 N-S g t 30.C r-30.C -4
.g 25.D-- 25.0-- ? !
O ,7. c I c 3 20.0-- 3 20.0-- - 3 . a " ! a e
- .e I
- i. ts.0-- /Ah i v fVsK
- 1. is.0-- v v-9 u-e 10.o-- d*k # to.o-- %
\~_-~
\- _ _ _ _ - . _
n
- 5. 0-- 5. 0-- I o
N
.C
*~~ . . . m-1
.C .
-1 2 ^ '
-1 ,0 1 2 to 0
10 1 10 to 10 to 10 to , y Frequency : 01z1 8~recuencf (51 r1 ,
' o 9
N
'# o port,c,g j, '
x to 14.c m. Legenc: e e i 12. F ( rr.cc-rtotd c ,g ,,, -}. Foundatton cef pt. _ _ _ _ _ _ _ _ _ _ _ _ _ .
- 8 . 0--
(
.\
t.o.s. c1 27 -6 ______________ , i t\ Notes" , b
- T t
- 8. 0-- \ a t t spectr a calculated at 5% damp hg
~ >
u accelerattons an untts of f t /s /s
- 4. 0--
i 2.0-- i .C 0-3 i6 10 to 10 i l Frequency fHz) ; t Fir,ure A.20 Surry Nuclear Power Plant Intake Structure Instructure ' Response Acceleration Ran;;e 2 '1 "C-' **' A
""W*" " " * ' * - - # " " ' ' ' ' '"
__ _m. _ _ _ __ _ . _ _ . . _
i 2: 2 x to 2 E: W < to 'J -S . 7, r-4 " w o
. 3-- o
- 3-- C c O ._ ,
l'.. O 3 b, " l .O e e t h-_
.L 3 o- l%%
v% -
= . 2-- *j % vsf% Ng "
e D3 e-j % 3 '\g
. t-- %_ __ . i-- L _._ ___ ___ _
9 i o . m-n,
.C
""++ .C .
m 2 -1 0 1
-1 .0 1 10 to 10 t o' 2 1
, 10 10 to 10 '
> Frequency I4z) Feequency (Hz) s o i w M o
0 i X to Vertical 30.C 1.e ge no' g P i 25.0-- free-fteld i founoation ref. pt. _ _ _ _ _ _ _ _ e - 3 20.0-- c o.s. el 27 - 6 _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _
- a. /g Otes: ;
k 15.0-- \ t
- all socctra ca tculated at 5% damp tog
'% accelerations an tsp Es o f i t 's 's .
E 10.0-- VX %--~
- 5. 0--
i -I " --"
.C
-1 O t 2
. to 10 10 to Frequency (Hz) i f
Figure A.21 Surry Nuclear Power Plant Intake Structure - Instructure Response Acceleration Range 3 l
" ' ' " " '^ #
m h'__ _ _ - . . - . _ - -
l l l i APPENDIX B Numerical Values of Building Response at Three Excitation Levels >
-1 I
Table B la Reactor Building Median Response Acceleration Range 1 (g) Freauency Elevatip_D DJ.I 2-5(Hz) 5 10(Hz) 7 (Hz) 10 (Hz) gna 39' 7" x .15 .10 .11 .08 .07 39' 7" y .15 .10 .11 .08 .08 39'-7" z .13 .08 .08 .07 .07 3'-6" x .13 .10 .10 .08 .08-
-3'-6" y .13 .10 .10 .08 .08
-3'-6" : .14 .09 .08 .08 ,07 18' 4" x .12 .10 .10 .09 .08 188 4" y .12 .10 .10 .09 .08 18' 4" : .14 .09 .09 .08 .07 47'-4" x .14- .11 .12 .10 .10 47' 4" .y .14 .11 .11 .09 .09 47' 4" z .14 .09 09 ,08 .07 95' 6" x .20 .15 .16 .14 .13 95'-6" y .20 .15 .17 .13 .12 95' 6" - z .14 .09 .09 .08 .07 Table B lb Reactor Building Median Response Acceleration Range 2 (g)
Frecuency . Elevation. Dir 2-5(Hz) 5-10(Hz) 7 (Hz) 10 (Hz) 123 l -39' 7"' x .30 . 21 - .22- . 16 - .15 i -39'-7" 'y .30 .21 ,22 .16 .14
-39'-7" z .23 .15 .16 .14 .12
-3'-6" x .24 .18 .19 .15 .13
-3'-6"' y .24 .18 .19 .15 .13
-3'-6" z .23 .16 .17 .14 .13 L 18'-4" x .21 ,18 .19
.14 .14 18'-4" y .22 .18 .19 .15 .14 18' 4" z .23 .16 .17- .14 .13 47'-4" x .22- .19 .21 .17 .16 47'-4" y .23 .19 .20 .16 .15 47'-4" z .23 .16 .17 .14 .13 95'-6" x .32 .28 .30 .25 .23 95' 6" y .32 .28 .30 .24 .22 95'-e' z .23 .16 .17 .14 .13 B-1 i
. - . . .. -. .. .. . - _ - .~ --. . -. _ . - .
Table B 1c Reactor Building Median Response l Acceleration Range 3 (g) Frecuenev Elevation Dir 2-5(Hz) 5 10(Hz) 7 (Hz_1 10 (Hz) gna 39' 7" x .46 .38 .42 .28 .23 39'-7" y .46 .38 .42 .30 .23 39'-7" z .24 .18 .19 .16 .14 3' 6" x .36 .31 .34 .22 .18 3'-6" y .36 .31 .34 .24 .19 . 3'-6" z .25 .20 .21 .17 ,14 18' 4" x .31 .28 .32 .20 .17 18'-4" y .31 .28 .31 .21 .18 18' 4" z .25 .20 .20 .17 .14 47' 4"- 'x ,25 .29 .34 .21 .17 47' 4" y .27 .29 .33 .21 .18-47' 4" z .25 .20 .21 .17 .14 - 95' 6" x .30 .41 .46 .34 .25 95' 6" y .32 .43 .50 .33 .25 95'-6" z .25 .20 .20 .17 .15 Tnble B 2a Auxiliary Building Median Responso Acceleration Range 1 (g) Frecuency Elevation Elr 2-5(Hz) 5-10(Hz) 7 (Hz) 10 (Hzl gna 0" x .32 .19 .20 .17 .15
-2 0" y .32 .19 .19 .16 .15 2 0" z .18 .15 .15 .13 .09 13'- 0" x .32 .19 .20 .17 .16 13'- 0" y .32 .19 .20 .16 .15 13'- 0" z .18 .15 .15 .13 .09 L 27'- 6" x .32 .19 .20 .17 .16 27'- 6" y .32 .19 .20 .16 .15 27'- 6" z .18 .15 .15 .13 .10 45'-10" x .31 .19 .20 .17 .16 45'-10" y .32 .19 .20 .17 .15 h' 45'--10" z .18 .15 .15 .13 .10
-66'- 0" x .48 .34 .37 .24 .21 66'- 0" y .46 .36 .40 .25 .21 66' 0" z .18 .16 .16 .14 .10 l
B-2 l
I 1 l I J Table B-2b Auxiliary Building Median Response Acceleration Range 2 (g) Freauency Elevation Qir 2 5(Hz) 5-10(Hz) 7 (Hz) 10 (Hz) gna 0" x .56 .35 .35 .31 .29 2 y .55 ,34 .35 .30 .28 2 - 0" 2 - 0" z .34 .27 .27 .24 .18 x .55 .35. .35 .31 .29 13'- O' y .55 .34 .35 .30 .28 13'- 0" 0" z .34 .27 .27 .24 .18 13' ,29 27'- 6" x .55 .34 .35 .31 27'- 6" y .55 .34 .35 .30 .28 27'- 6" z .34 .27 .27 .24 .18 x .54 .34- .35 .30 .29 45'-10" 45'-10" y .55 .34 .35 .30 .29
.34 .27 .28 .25 .18 45'-10" z 66' 0" x .79 .54 58 40 .36
.76 .55 .60 .40 .36 66'- 0" - y 66' 0" z .35 .29 .29 .26 .19 Table B-2c Auxiliary Building Median Response Acceleration Range 3 (g)
Freauency Elevation Dir 2 5(Hz) 5-10(Hz) 7 (Hz) 10 (Hz) gna
-2 0" x .76 .48 49 .43 41 y .76 47 48 42 .40 0"
z .49 .38 .38 .35 .25 0" x .76 48 .49 .43 .41 13'- 0" y .75 .47 .48 .42 .40 13'- 0" 13'- 0" z 49 .38 .39 .35 .25 x .75 .47 .48 .42 .40 ! 27'- 6" 47 .48 .42 40 27'- 6" y .75
.35 .25
.38 27'- 6" z 49 .38
.74 .47 .48 ,42 .40 45'-10" x 40 y .75 47 .48 .42 45'-10" .26
.50 .39 .39 .35 45'-10" z
.49 66' 0" x 1.03 .69 .74 .53 y 1.00 .69 .75 .52 .48 66'- 0" p
z .50 40 .41 .36 .27 66'- 0" B-3
Table B 3a Control Room Structure Median Response Acceleration Range 1 (g) freauency Elevation Elr 2 5(Hz) 5-10(Hz) 7 (Hz) 10 (Hz) Ena 00" x .30 .19 .20 .16 .15 0" y .32 .19 .19 .16 .15 0 - 0" z .17 .13 .13 .12 .09 9' 6" x .31 .19 .20 .17 .16 9'- 6" y .31 .19 .19 .16 .15 9'- 6" z .18 .14 .14 .12 .09 27' 0" x .29 .19 .19 .16 .15 27'- 0" y .31 .19 .20 .17 .15 27' 0" z .17 .13 .14 .12 .09 45' 3" x .28 .19 .19 .16 .15 45'- 3" y .31 .19 .20 .17 .15 45'- 3" z .17 .13 .13 .12 .09 58'- 6" x .50 .28 .28 .25 .24 58'- 6" y .38 .26 .28 .20 .17 58'- 6" z .17 .13 .13 .12 .09 77'- 6" x .74 .47 .51 .36 .32 77'- 6" y .39 .32 .35 .24 .18 77'- 6" z .17 .13 .13 .12 .09 B-4
. - _ - . . . . . . . . .. . . =_- ._. .- . . .
Tabic B 3b Control Room Structure Acceleration Rango 2 (g) Frecuency Elevation Dir 2 5(Hz) 5 10(Hz) 7 (Hz) 10 (Hz) ggg 00" x .53 .34 .35 .30 .28 0 0' y .56 .34 .35 .30 .28 0" z .32 .24 .24 .23 .17 9'- 6" x ,55 .35 .36 .31 .29 9'- 6" y .55 .33 .34 .30 .28 9' 6" z .33 .25 .25 .23 .17 27'- 0" x .50 .33 .34 .29 27 27'--0" y .55 .34 .35 .30 .8 27' 0" z .32 .25 .25 .23 .37 45'- 3" x ,48 .32 .33 .29 .28 45'- 3" y .53 .34 .34 .30 .28 45'- 3" z .32 .24 .25 .23 .17 58'- 6" x .78 44 .43 .40' .39 58'- 6" y .64 .43 .46 .35 .31 58' 6" z ,32 .23 .23 .22 ,16 77'- 6" x 1.11 ,71 ,75 .56 ,51-77' 6" y .66 .51 .55 .41 .32 77'- 6" z .32 .23 .23 .22 .16 i B-5
i Table D 3c Control Room Structure Acceleration Range 3 (g) l~ Frequency Elevation gir 2-5(Hz) 5-10(Hz) 7 (Hz) 10 (Hel tan 0 O* x .74 ,47 ,48 ,42 .39 0 - 0" y .78 48 49 .43 ,41 0 - 0" z .45 .35 .35 .33 ,24 9' 6" x .76 .49 .50 .43 .41 9'- 6" y .76 47 .48 .42 .40 9'- 6" z 48 .36 ,37 .34 .24 27' 0" x .69 .46 .47 .41 .39 27'- 0" y .75 .48 49 .43 .41 27'- 0" z ,46 .36 .36 .33 ,24 45' 3" x .65 .45 .46 .40 .38 45' '3" y .73 .47 .48 .43 .40 45' 3" z 45 .35 .35 .33 .24 58' 6" x .99 .57 .57 .52 .51 58' 6" y .87 .58 .61 .48 .43 58'- 6" a 45 .34 .34 .31 .23 77'- 6" x 1,38 .87 ,91 ,71 .65 77'- 6" y .88 .68 .72 .55 44 77' 6" z 45 .34 .34 .31 .23 Table B 4a Safeguards Area-Median Response Acceleration Range 1 (g) Frecuenev F_levation Dir 5(Hz) 5 10(Hz) 7 (Hz) 10 (Hz) Ipa
.9'- 6" x .27 .16 ,16 .14 .13 9'- 6" y ,27 .16 ,16 .14 .13 9'- 6"- z , 18 .16 .16 .14 ' .10 19'- 6" x .29 .17 .17 .14 .13 19' 6" y .29 ,17 .18 .14 .13 19'- 6" z .18 .16 ,16 .14 .10 28'- 6" x .30 ,19 .20 .15 .14 28'- 6" y .30 .19 .20- .15 ,14 -
28'- 6" z .18 .16 .16 .14 .10 42'- 6" x .33 ,24 .25 .19 ,15 42'- 6" y .33 .22 .24 .17 .15 42'- 6" z .18 ,16 .17 .14 .10 B6
. . . _ - . . . -}}